Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 2015 – Main Text

Transcript

1

2 Front cover photo credit for cow and digester: Vanguard Renewables.

3 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 2015 , inclusive, will be made available for the final report published on Ap ril 15, 2017 at the internet site mentioned above. FOR FURTHER INFORMATION Ms. Mausami Desai Contact , Environmental Protection Agency, (202) 343 - 9381 , [email protected] . o – 9897, weitz.meli [email protected] r Ms. Melissa Weitz, Environmental Protection Agency, (202) 343 For more information regarding climate change and greenhouse gas emissions, see the EPA web site at < https://www.epa.gov/climatechange >.

4 Acknowledgments The Environmental Protection Agency would like to acknowledge the many indi vidual and organizational contributors to this document, without whose efforts this report would not be complete. Although the complete list of researchers, government employees, and consultants who have provided technical and editorial support is too long to list here, EPA’s Office of Atmospheric Programs would like to thank some key contributors and reviewers whose work has significantly improved this year’s report. Work on emissions from fuel combustion was led by reco . Amy Bunker, Sarah Fro man, Susan Vincent Cam ob , and Sarah Roberts directed the work on mobile combustion and transportation. Work on industrial processes Burke and John Steller . Work on fugitive methane emissions from and product use emissions was led by Mausami Desai irected by Melissa Weitz and Cate Hight. Calculations for the waste sector were led by the energy sector was d Rachel Schmeltz. Tom Wirth directed work on the Agriculture and the Land Use, Land - Use Change, and Forestry , with support from John Steller . Work on emissions o f HFCs, PFCs, SF chapters , and NF was directed by Deborah 3 6 Ottinger and Dave Godwin. Within the EPA, other Offices also contributed data, analysis, and technical review for this report. The Office of Transportation and Air Quality and the Office of Air Quality Pl anning and Standards provided analysis and review for several of the source categories addressed in this report. The Office of Solid Waste and the Office of Research and Development also contributed analysis and research. The Energy Information Administrat ion and the Department of Energy contributed invaluable data and analysis on related topics. - The U.S. Forest Service prepared the forest carbon inventory, and the Department numerous energy of Agriculture’s Agricultural Research Service and the Natural Res ource Ecology Laboratory at Colorado State University contributed leading research on nitrous oxide and carbon fluxes from soils. The National Oceanic and Atmospheric Administration prepared the estimates of emissions from Coastal Wetlands. Other government agencies have contributed data as well, including the U.S. Geological Survey, the Federal Highway Administration, the Department of Transportation, the Bureau of Transportation Statistics, the Department of Commerce, the National Agricultural St atistics Service, the Federal Aviation Administration, and the Department of Defense. We would also like to thank Marian Martin Van Pelt , Leslie Chinery, Alexander Lataille and the full Inventory team at ICF including Diana Pape, Robert Lanza, Lauren Marti , Mollie Averyt, Mark Flugge, Larry O’Rourke, Deborah Harris, Jonathan Cohen, Sabrina Andrews, Bikash Acharya, Claire Boland, Rebecca Ferenchiak, Kasey Knoell, Kevin Kurkul, Cory Jemison, Matt Lichtash, Jessica Kuna, Emily Kent, Emily Golla, Rani Murali, D rew Stilson, Cara Blumenthal, Tim Storer , Louise Huttinger, and Jessica Klion for synthesizing this report and preparing many of the individual analyses. Finally, we thank the following teams for their significant analytical support: Eastern Research Grou p team (Casey Pickering, Brandon Long, Clint Burklin, Gopi Manne, Deb orah Bartram, Kara Edquist, Ami e Aguiar and Brian Guzzone); RTI International (Kate Bronstein, Meaghan McGrath); Raven Ridge Resources, and Ruby Canyon ntha Phillips, and Phillip Cunningham). Engineering Inc. (Michael Cote, Sama

5 Preface The United States Environmental Protection Agency (EPA) prepares the official U.S. Inventory of Greenhouse Gas Emissions and Sinks to comply with existing commitments under the United Natio ns Framework Convention on Climate Change (UNFCCC). Under decision 3/CP.5 of the UNFCCC Conference of the Parties, national inventories for UNFCCC Annex I parties should be provided to the UNFCCC Secretariat each year by April 15. In an effort to engage the public and researchers across the country, the EPA has instituted an annual public review and comment process for this document. The availability of the draft document is announced via Federal Register Notice and is posted on the EPA web site. Copies a re also mailed upon request. The public comment period is generally limited to 30 days; however, comments received after the closure of the public comment period are accepted and considered for the nex t edition of this annual report . Public review of this re port occurred from comments received are posted to the EPA web site. February 15 to March 17, 2017 and iii

6 Table of Contents ... ... ... VI TABLE OF CONTENTS ... ... ... ... RES, IX LIST OF TABLES, FIGU AND BOXES ES ... ... ... - 1 EXECUTIVE SUMMARY ... ... ... ... ... . ES.1 Background Information - 2 ES ES.2 Recent Trends in U.S. Greenhouse Gas Emissions and Sinks ... ... ... ES - 4 ES ... ... ... ES.3 Overview of Sector Emissions and Trends - 18 ... ES.4 Other Information ... ... ... ... ... ES - 23 1. INTRODUCTION ... ... ... ... 1 - 1 1.1 ... ... ... ... 1 - 3 Background Information ... 1.2 ... ... ... National Inventory Arrangements 1 - 10 1.3 Inventory Process ... ... ... ... ... 1 - 13 1.4 ... ... ... ... Methodology and Data Sources 1 - 15 1.5 Key Categories ... ... ... ... ... 1 - 16 ... 1.6 Quality Assurance and Quality Control (QA/QC) ... ... 1 - 19 Uncertainty 1.7 Analysis of Emission Estimates ... ... ... 1 - 21 1.8 ... ... ... ... ... 1 - 23 Completeness ... ... ... ... 1 - 23 1.9 Organization of Report TRENDS IN GREENHOUSE GAS EMISSIONS ... ... ... 2 - 1 2. 2.1 ... ... ... 2 - 1 Recent Trends in U.S. Greenhouse Gas Emissions and Sinks ... 2.2 ... ... ... Emissions by Economic Sector 2 - 23 2.3 Indirect Greenhouse Gas Emissions (CO, NO , NMVOCs, and SO 35 ) ... ... 2 - 2 x ... ... ... ... ... 3 - 1 3. ENERGY ... ... ... Fossil Fuel Combustion (IPCC Source Category 1A) 3 - 5 3.1 3.2 Carbon Emitted from Non - Energy Uses of Fossil Fuels (IPCC Source Category 1A) ... 3 - 45 3.3 ... ... ... Incineration of Waste (IPCC Source Category 1A1a) 3 - 52 3.4 Coal Minin g (IPCC Source Category 1B1a) ... ... ... 3 - 56 ... 3.5 Abandoned Underground Coal Mines (IPCC Source Category 1B1a) ... 3 - 61 - ... 3 ... 65 3.6 Petroleum Systems (IPCC Source Category 1B2a) ... 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 vi

7 3.7 ... ... ... 3 - 77 Natural Gas Systems (IPCC Source Category 1B2b) Energy Sources of Indirect Greenhous 3.8 ... ... ... 3 - 93 e Gas Emissions 3 ... International Bunker Fuels (IPCC Source Category 1: Memo Items) - 94 3.9 ... - Wood Biomass and Biofuels Consumption (IPCC Source Category 1A) ... 3 ... 98 3.10 INDUSTRIAL PROCESSES AND PRODUCT USE ... ... ... 4. 4 - 1 ... ... ... 4 - 8 4.1 Cement Production (IPCC Source Category 2A1) Lime Production (IPCC Source Category 2A2) ... ... ... 4 - 12 4.2 4 ... ... ... - 17 4.3 Glass Production (IPCC Source Category 2A3) Other Process Uses of Carbonates (IPCC Source Category 2A4) ... ... 4 - 20 4.4 Ammonia Production (IPCC Source Category 2B1) ... ... ... 4 - 24 4.5 Urea Consumption for Non Agricultural Purposes ... ... ... 4 - 28 4.6 - 4 Nitric Acid Production (IPCC Source Category 2B2) ... ... ... - 30 4.7 Adipic Acid Production (IPCC Source Catego ry 2B3) ... ... ... 4 - 34 4.8 ... ... 4.9 4 - 38 Silicon Carbide Production and Consumption (IPCC Source Category 2B5) 4 ... ... ... - 41 4.10 Titanium Dioxide Production (IPCC Source Category 2B6) Soda Ash Production and Consumption (IPCC Source Category 2B7) ... ... 4.11 - 43 4 4.12 Petrochemical Pro duction (IPCC Source Category 2B8) ... ... 4 - 47 ... 4.13 HCFC - 22 Production (IPCC Source Category 2B9a) ... ... ... 4 - 53 Carbon Dioxide Consumption (IPCC Source Category 2B10) 4.14 ... ... . 4 - 56 4.15 ... ... ... 4 - 59 Phosphoric Acid Production (IPCC Source Category 2B10) Iron and Steel Production (IPCC Source Category 2C1) and Metallurgical Coke Production ... 4 - 63 4.16 Ferroalloy Production (IPCC Source Category 2C2) 4 ... ... ... - 72 4.17 4 ... ... ... - 75 4.18 Aluminum Production (IPCC Source Category 2C3) Magnesium Production and Processing (IPCC Source Category 2C4) ... ... 4.19 4 - 80 4.20 Lead Production (IPCC Source Category 2C5) ... ... ... 4 - 85 ... ... ... 4.21 4 - 88 Zinc Production (IPCC Source Category 2C6) 4.22 ... ... ... Semiconductor Manufacture (IPCC Source Category 2E1) 4 - 93 4.23 Substitution of Ozone Depleting Substances (IPCC Source Category 2F) ... ... 4 - 104 4.24 ... ... 4 - 112 Electrical Transmission and Distribution (IPCC Source Category 2G1) 4 ... ... Uses (IPCC Source Category 2G3) - 119 4.25 Nitrous Oxide from Product Industrial Processes and Product Use Sources of Indirect Greenhouse Gases ... ... 4.26 - 121 4 5. AGRICULTURE ... ... ... 5 - 1 ... 5.1 Enteric Fermentation (IPCC Source Category 3A) ... ... ... 5 - 3 5.2 ... ... ... Manure Management (IPCC Source Category 3B) 5 - 8 5.3 Rice Cultivation (IPCC Source Category 3C) ... ... ... 5 - 16 ... 5.4 Agricultural Soil Management (IPCC Source Category 3D) ... ... 5 - 21 39 5.5 Liming (IPCC Source Category 3G) ... ... ... ... 5 - vii

8 5.6 ... ... ... 5 - 42 Urea Fertilization (IPCC Source Category 3H) 5 Field Burning of Agricultural Residues (IPCC Source Category 3F) ... ... - 44 5.7 LAND USE, LAND - USE C HANGE, AND FORESTRY ... ... ... 6 - 1 6. ... ... ... ... 6 - 7 Representation of the U.S. Land Base 6.1 Forest Land Remaining Forest Land ... ... ... ... 6 - 21 6.2 Land (IPCC Source Category 4A2) Land Converted to Forest 6 ... ... . 6.3 - 42 6.4 ... ... .. 6 - 48 Cropland Remaining Cropland (IPCC Source Category 4B1) ... ... ... Land Converted to Cropland (IPCC Source Category 4B2) 6 - 57 6.5 6.6 Grassland Remaining Grassland (IPCC Source Category 4C1) ... ... 6 - 63 6.7 Land Converted to Grassland (IPCC Source Category 4C2) ... ... ... 6 - 71 . 6.8 Wetlands Remaining Wetlands (IPCC Source Category 4D1) ... ... 6 - 78 6.9 Land Converted to Wetlands (IPCC Source Category 4D2) ... ... ... 6 - 96 99 - ... 6.10 Settlements Remaining Settlements 6 ... ... ... 6.11 3B5b) ... ... 6 - 115 Land Converted to Settlements (IPCC Source Category ... ... Other Land Remaining Other Land (IPCC Source Category 4F1) 6 - 120 6.12 6.13 Land Converted to Other Land (IPCC Source Category 4F2) ... ... . 6 - 120 ... 7. ... ... ... ... WASTE 7 - 1 7.1 Landfills (IPCC Source Category 5A1) ... ... ... ... 7 - 3 ... 7.2 Wastewater Treatment (IPCC Source Category 5D) ... - ... 7 19 7.3 ... ... ... 7 - 34 Composting (IPCC Source Category 5B1) 7.4 ... ... ... Waste Incineration (IPCC Source Category 5C1) 7 - 37 7.5 Waste Sources of Indirect Greenhouse Gases ... ... ... 7 - 37 8. OTHER ... ... ... ... ... 8 - 1 ... 9. RECALCULATIONS AND I MPROVEMENTS ... ... 9 - 1 - 1 ... 10. REFERENCES ... ... ... 10 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 viii

9 List of Tables, Figures, and Boxes Tables Potentials (100 - Year Time Horizon) Used in this Report ... ... ES - 3 - Table ES 1: Global Warming Sinks (MMT CO ES Eq.) ... . 2: Recent Trends in U.S. Greenhouse Gas Emissions and - 5 Table ES - 2 3: CO - Use Sector (MMT CO Table ES Eq.) - ... ES - 11 Emissions from Fossil Fuel Combustion by End 2 2 4: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO Table ES Eq.) - 2 ... ... ... ... ... ES - 18 ... 5: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land Use Change, and Table ES - - Eq.) ... ... Forestry (MMT CO ... ... ES - ... 22 2 6: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO Table ES Eq.) - ES - 24 ... 2 - 7: U.S. Greenhouse Gas Emissions by Economic Sector with Electricity - Related Emissions Distributed Table ES ... (MMT CO ... ... ... ... Eq.) ES - 25 2 - 8: Recent Trends in Various U.S. Data (Index 1990 = 100) ... ... . ES - 26 Table ES 1: Global Atmospheric Concentration, Rate of Concentration Change, and Atmospheric Lifetime of - Table 1 ... ... ... ... Selected Greenhouse Gases 1 - 4 ... - 2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report ... 1 - 9 Table 1 - Year GWP values - Table 1 ... ... ... .. 1 - 10 3: Comparison of 100 - ... y Categories for the United States (1990 - 2015) Table 1 ... ... 1 - 16 4: Ke ... - Eq. and Percent) 5: Estimated Overall Inventory Quantitative Uncertainty (MMT CO 1 - 22 Table 1 2 Table 1 - 6: IPCC Sector Descriptions ... ... ... ... 1 - 23 ... Table 1 - 7: List of Annexes ... ... ... ... 1 - 24 Table 2 - Greenhouse Gas Emissions and Sinks (MMT CO 1: Recent Trends in U.S. Eq.) ... ... 2 - 3 2 Table 2 ... ... 2 - 5 - 2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (kt) 3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO Eq.) - Table 2 2 .. ... ... ... ... ... 2 - 8 ... - 4: Emissions from Energy (MMT CO Eq.) ... ... Table 2 2 - 11 ... 2 - 5: CO Emissions from Fossil Fuel Combustion by End - Use Sector (MMT CO Table 2 Eq.) ... 2 - 12 2 2 ... - Eq.) Table 2 ... 6: Emissions from Industrial Processes and Product Use (MMT CO 2 - 16 2 Table 2 - 7: Emissions from Agriculture (MMT CO Eq.) ... ... ... 2 - 19 2 Table 2 8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land - Use Change, and - ... ... ... ... ... Eq.) 2 - 20 Forestry (MMT CO 2 - 9: Emissions from Waste (MMT CO Table 2 Eq.) ... ... ... 2 - 22 2 Table 2 - 10: U.S. Greenhouse Gas Emissions All ocated to Economic Sectors (MMT CO Eq. and Percent of Total in 2 2015) ... ... ... ... ... ... 2 - 24 Table 2 - Related Greenhouse Gas Emissions (MMT CO Eq.) 11: Electricity Generation ... ... 2 - 26 - 2 Table 2 - 12: U.S. Greenhouse Gas Emissions by Economic Sector and Gas with Electricity - Related E missions ... Distributed (MMT CO Eq.) and Percent of Total in 2015 ... ... 28 2 - 2 30 Table 2 - 13: Transportation - Related Greenhouse Gas Emissions (MMT CO - Eq.) ... ... 2 2 ix

10 Table 2 14: Recent Trends in Various U.S. Data (Index 1990 = 100) ... ... ... 2 - 34 - 15: Emissions of NO - Table 2 (kt) ... ... ... 2 - 35 , CO, NMVOCs, and SO x 2 Table 3 - , and N O Emissions from Energy (MMT CO 2 Eq.) 1: CO ... ... 3 - , CH 2 2 4 2 2: CO 3 , CH Table 3 , and N O Emissions from Energy (kt) ... ... ... 3 - - 2 2 4 3: CO , CH - Table 3 , and N O Emissions from Fossil Fuel Combustion (MMT CO 5 Eq.) ... ... 3 - 2 2 2 4 4: CO , CH Table 3 , and N O Emissions from Fossil Fuel Combustion (kt) - ... ... 3 - 5 2 4 2 5: CO - Emissions from Fossil Fuel Combustion by Fuel Type and Sector (MMT CO Eq.) ... Table 3 3 - 6 2 2 Table 3 - 6: Annual Change in CO Emissions and Total 2015 Emissions from Fossil Fuel Combustion for Selected 2 ... Eq. and Percent) ... ... ... Fuels and Sectors (MMT CO 3 - 7 2 Table 3 7: CO - , CH , and N 12 O Emissions from Fossil Fuel Combustion by Sector (MMT CO - Eq.) ... 3 2 2 2 4 8: CO 13 , CH Table 3 , and N O Emissions from Fossil Fuel Combustion by End - Use Sector (MMT CO - Eq.) ... 3 - 2 2 4 2 9: CO Table 3 Emissions from Stationary Fossil Fuel Combustion (MMT CO - ... ... 3 - 14 Eq.) 2 2 - 10: CH Emissions from Stationary Combustion (MMT CO Table 3 Eq.) ... ... 3 - 15 2 4 ... - O Emissions from Stationary Combustion (MMT CO Table 3 Eq.) ... 3 - 16 11: N 2 2 - 12: CO Emissions from Fossil Fuel Combustion in Transportation End - Use S ector (MMT CO Table 3 Eq.) ... 3 - 24 2 2 - Emissions from Mobile Combustion (MMT CO 26 Table 3 13: CH ... ... 3 - Eq.) 2 4 Table 3 - 14: N O Emissions from Mobile Combustion (MMT CO Eq.) ... ... 3 - 27 2 2 Table 3 15: Carbon Intensity from Direct Fossil Fuel Combustion by Sector (MMT CO Eq./QBtu) ... 3 - 33 - 2 - Emissions from Energy - Related Fossil Fuel Table 3 16: Approach 2 Quantitative Uncertainty Estimates for CO 2 3 ... ... Eq. and Percent) - 35 Combustion by Fuel Type and Sector (MMT CO ... 2 17: Approach 2 Quantitative Uncertainty Estimates for CH Table 3 and N - O Emissions from Energy - Related 2 4 Eq. and Percent) ... ... 3 - 40 Stationary Combustion, Including Biomass (MMT CO 2 Table 3 - 18: Approach 2 Quantitative Uncertainty Estimates for CH and N O Emissions from Mobile Sources 4 2 ... Eq. and Percent) ... ... ... (MMT CO ... 3 - 43 2 Table 3 - Emissions from Non - Energy Use Fossil Fuel Consumption (MMT CO 19: CO Eq. and Percent) ... 3 - 45 2 2 20: Adjusted Consumption of Fossil Fuels for Non Energy Uses (TBtu) ... Table 3 - 3 - 46 - ... - 21: 2015 Adjusted Non - Energy Use Fossil Fuel Consumption, Storage, and Emissions ... 3 - 47 Table 3 - Emissions from Non - Energy Uses of Fossil Table 3 22: Approach 2 Quantitative Uncertainty Estimates for CO 2 Fuels (MMT CO ... ... ... ... 3 - 48 Eq. and Percent) 2 Table 3 - - Energy Uses of Fossil Fuels 23: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non ... ... ... .. ... 3 - 49 (Percent) ... - 24: CO , CH 52 , and N - O Emissions from the Incineration of Waste (MMT CO 3 Eq.) Table 3 ... 2 4 2 2 - , CH 53 , and N Table 3 O Emissions from the Incineration of Waste (kt) 25: CO ... ... 3 - 4 2 2 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 CO and N O from the Incineration of Waste (MMT 2 2 Eq. and Percent) ... ... CO ... ... 3 - 55 ... 2 Table 3 - 28: Coal Production (kt) ... ... ... ... 3 - 57 Table 3 29: CH 57 Emissions from Coal Mining (MMT CO - Eq.) - ... ... ... 3 2 4 - 3 ... ... 57 Table 3 - 30: CH ... Emissions from Coal Mining (kt) 4 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 x

11 Table 3 - 31: Approach 2 Quantitative Uncertainty Estimates for CH Emissions from Coal Mining (MMT CO Eq. 2 4 ... ... ... ... ... 3 - 60 and Percent) 32: CH ... Emissions from Abandoned Coal Mines (MMT CO Table 3 Eq.) ... - 3 - 62 2 4 ... Table 3 Emissions from Abandoned Coal Mines (kt) ... 33: CH ... 3 - 62 - 4 - 34: Number of Gassy Abandoned Mines Present in U.S. Basins in 2015, grouped by Class according to Table 3 - Abandonment State ... ... ... ... ... 3 - 63 Post - Emissions from Abandoned Underground Coal 35: Approach 2 Quantitative Uncertainty Estimates for CH Table 3 4 ... Eq. and Percent) ... ... Mines (MMT CO ... 3 - 64 2 Table 3 36: CH Emissions from Petroleum Systems (MMT CO 66 Eq.) - ... ... . 3 - 2 4 Table 3 - 37: CH Emissions from Petroleum Systems (kt) ... ... ... 3 - 66 4 Table 3 38: CO - Emissions from Petroleum Systems (MMT CO ) ... ... ... 3 - 66 2 2 - 39: CO Table 3 Emissions from Petroleum Systems (kt) ... ... ... 3 - 66 2 Table 3 40: Approach 2 Quantitative Uncertainty Estimates for CH - Emissions from Petroleum Systems (MMT 4 ... Eq. and Percent) CO ... ... ... ... 3 - 68 2 Table 3 - 41: Oil Well Count Data ... ... ... ... 3 - 71 Table 3 ) ... 42: National Tank Activity Data (MMbbl) by Category and National Emissions (Metric Tons CH 3 - 71 - 4 Table 3 - 43: National Equipment Counts for Fugitive Sources and National Emissions (Metric Tons CH ) ... 3 - 72 4 - 44: Pneumatic Controller and Chemical Injection Pump National Equipment Counts and National Table 3 Emissions (Metric Tons CH ) ... ... ... ... .. 3 - 72 4 - Table 3 3 45: Associated Gas Well Venting and Flaring National Emissions (Metric Tons CH ) ... - 73 4 Table 3 46: Production Segment Gas STAR Reductions Update (Metric Tons CH - ) ... ... 3 - 74 4 Table 3 - Captured and Extracted for EOR Operations (MMT CO 47: Quantity of CO ) ... ... 3 - 76 2 2 48: Quantity of CO Table 3 Captured and Extracted for EOR Operations (kt) ... - 3 - 76 ... 2 - 49: CH Emissions from Natural Gas Systems (MMT CO 78 Eq.) ... ... Table 3 - 3 2 4 - Emissions from Natural Gas Systems (kt) ... ... ... Table 3 3 - 78 50: CH 4 - 51: Calculated Potential CH Eq.) and Captured/Combusted CH Table 3 from Natural Gas Systems (MMT CO 2 4 4 ... ... ... ... ... ... 3 - 79 ... Table 3 - 52: Non - combustion CO Emissions from Natural Gas Systems (MMT) ... ... 3 - 79 2 Table 3 53: Non - combustion CO - Emissions from Natural Gas Systems (kt) ... ... 3 - 79 2 energy CO Table 3 and Non - 54: Approach 2 Quantitative Uncertainty Estimates for CH Emissions from Natural - 2 4 Eq. and Percent) ... ... ... Gas Systems (MMT CO 3 - 81 ... 2 - ) Table 3 55: National Tank Activity Data (MMbbl) by Category and National Emissions (Metric Tons CH 3 - 84 ... 4 Table 3 - 56: Gas Well Count Data ... ... ... ... 3 - 84 Table 3 Counts for Fugitive Sources and National Emissions (Metric Tons CH ) ... 57: National Equipment 3 - 85 - 4 Table 3 - 58: Pneumatic Controller and Chemical Injection Pump National Equipment Counts and National Emissions (Metric Tons CH ) ... ... ... ... .. 3 - 85 4 3 - 59: National Liquids Unloading Activity Data by Category and National Emissions (Metric Tons CH ) Table 3 - 86 4 Table 3 - 60: National Gathering and Boosting Episodic Emission Activity Data (Number of Stations) and National 87 Emissions (Metric Tons CH - ) ... ... ... ... .. 3 4 xi

12 Table 3 - 61: Production Segment Gas STAR “Other Reductions” Data (Metric Tons CH ) and Scaling Factors 4 (fraction) ... ... ... ... .. 3 - 88 ... 62: CH Emissions from Processing Plants (Metric Tons CH Table 3 ) ... ... - 3 - 89 4 4 Table 3 63: Previous (last year’s) 1990 to 2014 Inventory Estimates for Processing Segment Emissions (Metric - ) ... ... Tons CH ... ... 3 - 90 ... 4 - 64: NO - , CO, and NMVOC Emissions from Energy - Related Activities (kt) ... ... 3 Table 3 93 x Eq.) - , CH Table 3 , and N O Emissions from International Bunker Fuels (MMT CO 95 - ... 3 65: CO 2 2 2 4 - 66: CO , CH 95 , and N - O Emissions from International Bunker Fuels (kt) ... Table 3 3 ... 2 2 4 - ... ... 3 - 96 Table 3 67: Aviation Jet Fuel Consumption for International Transport (Million Gallons) Table 3 Fuel Consumption for International Transport (Million Gallons) ... ... 68: Marine 3 - 97 - Table 3 - 69: CO Emissions from Wood Consumption by End - Use Se ctor (MMT CO 99 Eq.) ... 3 - 2 2 ... - Emissions from Wood Consumption by End - Use Sector (kt) ... Table 3 3 - 99 70: CO 2 - 71: CO Emissions from Ethanol Consumption (MMT CO 99 Eq.) Table 3 ... ... 3 - 2 2 ... - Emissions from Ethanol Consumption (kt) Table 3 72: CO ... 3 - 99 ... 2 Table 3 - 73: CO Emissions from Biodiesel Consumption (MMT CO 100 Eq.) ... ... 3 - 2 2 ... - Emissions from Biodiesel Consumption (kt) Table 3 74: CO ... 3 - 100 ... 2 Table 3 - 75: Woody Bioma ss Consumption by Sector (Trillion Btu) ... ... ... 3 - 100 ... Table 3 76: Ethanol Consumption by Sector (Trillion Btu) ... ... - 3 - 100 ... Table 3 - 77: Biodiesel Consumption by Sector (Trillion Btu) ... - ... 3 101 Table 4 - 1: Emissions from Industrial Processes and Product Use (MMT CO Eq.) ... ... 4 - 3 2 Table 2: Emissions from Industrial Processes and Product Use (kt) ... ... ... 4 - 4 4 - ... 3: CO Emissions from Cement Production (MMT CO Table 4 Eq. and kt) - ... 4 - 9 2 2 - 4: Clinker Production (kt) ... ... ... ... 4 - 10 Table 4 - 5: Approach 2 Quantitative Uncertainty Estimates for CO Table 4 Emissions from Cement Production (MMT CO 2 2 ... ... ... ... ... 4 - 11 Eq. and Percent) - 6: CO Emissions from Lime Production (MMT CO Eq. and kt) Table 4 ... ... 4 - 13 2 2 ... - Emissions from Lime Production (kt) Table 4 7: Potential, Recovered, and Net CO 4 - 13 ... 2 Table 4 - 8: High - Calcium - and Dolomitic - Quicklime, High - Calcium - and Dolomitic - Hydrated, and Dead - Burned - Dolomite Lime Production (kt) ... ... ... 4 - 14 ... - ... ... ... ... 4 - 14 9: Adjusted Lime Production (kt) Table 4 - 10: Approach 2 Quantitative Uncertainty Estimates for CO Emissions from Lime Production (MMT CO Table 4 2 2 ... ... ... ... ... 4 - 16 Eq. and Percent) - Emissions from Glass Production (MMT CO 18 Eq. and kt) Table 4 ... ... 4 - 11: CO 2 2 Table 4 - 12: Limestone, Dolomite, and Soda Ash Consumption Used in Glass Production (kt) ... 4 - 19 Table 4 - 13: Approach 2 Quantitative Uncertainty Estimates for CO Emissions from Glass Production (MMT CO 2 2 ... ... ... ... Eq. and Percent) 4 - 19 ... Table 4 - 14: CO Emissions from Other Process Uses of Carbonates (MMT CO 21 Eq.) ... ... 4 - 2 2 Table 4 - 15: CO 21 Emissions from Other Process Uses of Carbonates (kt) ... ... 4 - 2 22 - 4 ... ... Table 4 - 16: Limestone and Dolomite Consumption (kt) ... 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 xii

13 Table 4 - Emissions from Other Process Uses of 17: Approach 2 Quantitative Uncertainty Estimates for CO 2 onates (MMT CO Eq. and Percent) ... ... ... ... 4 - 23 Carb 2 18: CO Emissions from Ammonia Production (MMT CO ... Eq.) ... - 4 - 25 Table 4 2 2 - Emissions from Ammonia Production (kt) ... ... 19: CO 4 - 25 Table 4 ... 2 20: Ammonia Production and Urea Production (kt) ... ... ... 4 - 26 Table 4 - - Table 4 Emissions from Ammonia Production (MMT 21: Approach 2 Quantitative Uncertainty Estimates for CO 2 ... ... ... ... ... Eq. and Percent) 4 - 27 CO 2 22: CO Emissions from Urea Consumption for Non - Agricultural Purposes (MMT CO Table 4 Eq.) - ... 4 - 28 2 2 23: CO Table 4 Emissions from Urea Consumption for Non - Agricultural Purposes (kt) - ... 4 - 28 ... 2 - 24: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt) ... 4 - 29 Table 4 Table 4 25: Approach 2 Quantitative Uncertainty Estimates for CO - Emissions from Urea Consumption for Non - 2 ... ... ... Eq. and Percent) 4 - 30 Agricultural Purposes (MMT CO 2 26: N O Emissions from Nitric Acid Production (MMT CO Table 4 Eq. and kt N O) ... ... - 4 - 31 2 2 2 27: N itric Acid Production (kt) ... ... ... ... Table 4 4 - 33 - - 28: Approach 2 Quantitative Uncertainty Estimates for N Table 4 O Emissions from Nitric Acid Production (MMT 2 ... CO ... ... ... ... Eq. and Percent) 4 - 34 2 - ... 29: N O Emissions from Adipic Acid Production (MMT CO Eq. and kt N 35 O) ... Table 4 4 - 2 2 2 - 30: Adipic Acid Production (kt) ... ... ... ... Table 4 - 37 4 Table 4 - 31: Approach 2 Quantitative Uncertainty Estimates for N O Emissions from Adipic Acid Production 2 (MMT CO 37 Eq. and Percent) ... ... ... ... ... 4 - 2 and CH Table 4 - 32: CO 38 Emissions from Silicon Carbide Production and Consumption (MMT CO - Eq.) ... 4 2 4 2 - 33: CO and CH Table 4 Emissions from Silicon Carbide Production and Consumption (kt) ... 4 - 38 2 4 Table 4 ... ... 4 - 39 34: Production and Consumption of Silicon Carbide (Metric Tons) - 35: Approach 2 Quantitative Uncertainty Estimates for CH and CO - Emissions from Silicon Carbide Table 4 2 4 Production and Consumption (MMT CO ... ... ... Eq. and Percent) 4 - 40 2 - 36: CO Emissions from Titanium Dioxide (MMT CO Table 4 Eq. and kt) ... ... 4 - 41 2 2 ... - Titanium Dioxide Production (kt) ... ... ... 4 - 42 37: Table 4 - 38: Approach 2 Quantitati ve Uncertainty Estimates for CO Emissions from Titanium Dioxide Production Table 4 2 (MMT CO Eq. and Percent) ... ... ... ... ... - 43 4 2 Table 4 - 39: CO Emissions from Soda Ash Production and Consumption Not Associated with Glass Manufacturing 2 (MMT CO Eq.) ... ... ... ... ... 4 - 45 2 Table 4 40: CO - E missions from Soda Ash Production and Consumption Not Associated with Glass Manufacturing 2 ... ... ... ... ... ... 4 - 45 (kt) - 41: Soda Ash Producti on and Consumption Not Associated with Glass Manufacturing (kt) ... 4 - 46 Table 4 Table 4 42: Approach 2 Quantitative Uncertainty Estimates f or CO - Emissions from Soda Ash Production and 2 ... Eq. and Percent) Consumption (MMT CO ... ... ... 4 - 47 2 Table 4 - 43: CO and CH 49 Emissions from Petrochemical Production (MMT CO - Eq.) ... ... 4 2 4 2 - and CH 49 Table 4 44: CO ... ... 4 - Emissions from Petrochemical Production (kt) 4 2 Ta ble 4 - 45: Production of Selected Petrochemicals (kt) ... ... ... 4 - 51 Table 4 - 46: Approach 2 Quantitative Uncertainty Estimates for CH Emi ssions from Petrochemical Production and 4 52 CO - Emissions from Carbon Black Production (MMT CO 4 Eq. and Percent) ... ... 2 2 xiii

14 Table 4 - - 23 Emissions from HCFC - 22 Production (MMT CO 47: HFC Eq. and kt HFC - 23) ... 4 - 53 2 Table 4 - 22 Production (kt) ... ... ... ... 4 - 54 - 48: HCFC 49: Approach 2 Quantita 23 Emissions from HCFC - - - 22 Production Table 4 tive Uncertainty Estimates for HFC ... (MMT CO ... ... ... Eq. and Percent) 4 - 55 ... 2 - 50: CO Emissions from CO 56 Consumption (MMT CO - Table 4 ... ... 4 Eq. and kt) 2 2 2 ... - Production (kt CO Table 4 ) and the Percent Used for Non - EOR Applications ... 4 - 58 51: CO 2 2 - 52: Approach 2 Quantitative Uncertainty Estimates for CO Table 4 Emissions from CO Consumption (MMT CO 2 2 2 ... ... Eq. and Percent) ... ... 4 - 58 ... Table 4 - 53: CO Emissions from Phosphoric Acid Production (MMT CO 60 Eq. and kt) ... ... 4 - 2 2 4 - ... ... - 61 Table 4 54: Phosphate Rock Domestic Consumption, Exports, and Imports (kt) - 55: Chemi cal Composition of Phosphate Rock (Percent by Weight) ... ... 4 - 61 Table 4 - Table 4 56: Approach 2 Quantitative Uncertainty Estimates for CO Emi ssions from Phosphoric Acid Production 2 ... Eq. and Percent) (MMT CO ... ... ... 4 - 62 ... 2 Table 4 - 57: CO Emissions from Metallurgical Coke Productio n (MMT CO 64 Eq.) ... ... 4 - 2 2 ... - Emissions from Metallurgical Coke Production (kt) Table 4 58: CO ... 4 - 64 ... 2 Table 4 - 59: CO Emissions from Iron and Steel Production (MMT CO 64 Eq.) ... ... 4 - 2 2 Table 4 - 60: CO Emissions from Iron and Steel Production (kt) ... ... ... 4 - 64 2 ... Table 4 - 61: CH Emissions from Iron and Steel Production (MMT CO - Eq.) ... 65 4 2 4 - 62: CH Emissions from Iron and Steel Production (kt) ... ... ... 4 - 65 Table 4 4 - ... ... 4 66 - Table 4 63: Material Carbon Contents for Metallurgical Coke Production 64: Production and Consumption Data for the Calculation of CO Table 4 and CH - Emissions from Metallurgical 4 2 ... ... ... ... 4 - 67 Coke Production (Thousand Metric Tons) - 65: Production and Consumption Data for the Calculation of CO Emissions from Metallurgical Coke Table 4 2 3 ) ... ... ... ... ... Production (Million ft 4 67 - Table 4 66: CO - Emission Factors for Sinter Production, Direct Reduced Iron Production and Pellet Production 4 - 67 2 ... ... ... 67: Material Carbon Contents for Iron and Steel Production 4 - 68 Table 4 - - 68: CH Emission Factors for Sinter and Pig Iron Production ... ... .. 4 Table 4 68 - 4 - roduction and Consumption Data for the Calculation of CO and CH Emissions from Iron and Steel Table 4 69: P 2 4 Production (Thousand Metric Tons) ... ... ... ... 4 - 69 Table 4 - 70: Production and Consumption Data for the Calculation of CO Emissions from Iron and Steel 2 3 unless otherwise specified) ... ... ... . 4 - 70 Production (Million ft 71: Approach 2 Quantitative Uncertainty Estimates for CO and CH - Table 4 Emissions from Iron and Steel 2 4 Production and Metallurgical Coke Production (MMT CO ... ... 4 - 71 Eq. and Percent) 2 - 72: CO Table 4 and CH Emissions from Ferroalloy Production (MMT CO 72 Eq.) ... ... 4 - 2 2 4 - 73: CO and CH Table 4 ... ... ... 4 - 72 Emissions from Ferroalloy Production (kt) 4 2 74 Table 4 ... ... ... 4 - 74: Production of Ferroalloys (Metric Tons) - Table 4 - 75: Approach 2 Quantitative Uncertainty Estimates for CO Emissions from Ferroalloy Production (MMT 2 CO Eq. and Percent) ... ... ... ... ... 4 - 75 2 - 76: CO 76 Emissions from Aluminum Production (MMT CO - Table 4 ... ... 4 Eq. and kt) 2 2 76 - 4 ... ... Table 4 - 77: PFC Emissions from Aluminum Production (MMT CO Eq.) 2 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 xiv

15 Table - ... ... ... 4 - 77 4 78: PFC Emissions from Aluminum Production (kt) 79: Production of Primary Aluminum (kt) - ... ... 4 - 79 Table 4 ... and PFC Emissions from Aluminum 80: Approach 2 Quantitative Uncertainty Estimates for CO - Table 4 2 Eq. and Percent) ... ... Production (MMT CO ... 4 - 80 ... 2 81: SF Emissions from Magnesium Production and Processing (MMT , HFC - 134a, FK 5 - 1 - 12 and CO Table 4 - 2 6 CO Eq.) ... ... ... ... ... .. 4 - 81 2 Table 4 , HFC - 134a, FK 5 - 1 - 12 and CO 82: SF - Emissions from Magnesium Production and Processing (kt) ... 4 - 81 2 6 ... 83: SF Emission Factors (kg SF - per metric ton of magnesium) Table 4 ... 4 - 83 6 6 - 84: Approach 2 Quantitative Uncertainty Estimates for SF 134a and CO , HFC - Table 4 Emissions from 2 6 Magnesium Production and Processing (MMT CO Eq. and Percent) ... ... ... 4 - 84 2 Table 4 - Emissions from Lead Production (MMT CO 85: CO Eq. and kt) ... ... 4 - 86 2 2 87 Table 4 ... ... ... ... 4 - - 86: Lead Production (Metric Tons) - 87: Approach 2 Quantitative Uncertainty Estimates for CO Emissions from Lead Production (MMT CO Table 4 2 2 ... ... ... ... ... 4 - 87 Eq. and Percent) Table 4 - ... ... ... ... 4 - 89 88: Zinc Production (Metric Tons) - 89: CO Emissions from Zinc Production (MMT CO Eq. and k t) ... ... Table 4 4 - 90 2 2 - 90: Approach 2 Quantitative Uncertainty Estimates for CO Table 4 Emissions from Zinc Production (MMT CO 2 2 ... ... ... ... Eq. and Percent) 4 - 92 ... Table 4 - 91: PFC, HFC, SF 94 , NF - , and N 4 O Emissions from Semiconductor Manufacture (MMT CO Eq.) ... 2 2 3 6 - Table 4 92: PFC, HFC, SF , NF 94 , and N - O Emissions from Semiconductor Manufacture (kt) ... 4 2 3 6 - 93: F HTF Compounds with Largest Emissions Based on GHGRP Reporting (tons) ... 4 - 95 Table 4 - 94: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SF , NF and N O Emissions from Table 4 - 3 2 6 Eq. and Percent) Semiconductor Manufacture (MMT CO ... ... ... 4 103 - 2 95: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO Table 4 Eq.) ... - 4 - 104 ... 2 - 96: Emissions of HFCs and PFCs from ODS Substitution (Metric Tons) ... Table 4 4 - 105 ... Table 4 97: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO - Eq.) by Sector ... 4 - 105 2 Table 4 98: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions from ODS Substitutes - ... Eq. and Percent) ... ... (MMT CO ... . 4 - 108 2 ... - Eq.) Table 4 99: U.S. HFC Consumption (MMT CO ... 4 - 109 ... 2 Table 4 - 100: Averaged U.S. HFC Demand (MMT CO Eq.) ... ... ... 4 - 110 2 Table 4 101: SF - Emissions from Electric Power Systems and Electrical Equipment Manufacturers (MMT CO Eq.) 2 6 ... ... ... ... ... ... 4 - 113 - 102: SF Electric Power Systems and Electrical Equipment Manufacturers (kt) Emissions from Table 4 ... 4 - 113 6 Table 4 103: Transmission Mile Coverage and Regression Coefficients (Percent) ... - 4 - 116 ... Table 4 - 104: Approach 2 Quantitative Uncertainty Estimates for SF ission and Emissions from Electrical Transm 6 Distribution (MMT CO Eq. and Percent) ... ... ... 4 - 117 ... 2 Table 4 - 105: N O Production (kt) ... ... ... ... 4 - 119 2 ... - 106: N O Emissions from N 119 O Product Usage (MMT CO - Eq. and kt) ... Table 4 4 2 2 2 Table 4 - 107: Approach 2 Quantitative Uncertainty Estimates for N O Emissions from N O Product Usage (MMT 2 2 121 CO - Eq. and Percent) ... ... ... ... ... 4 2 xv

16 Table 4 108: NO , CO, and NMVOC Emissions from Industrial Processes and Product Use (kt) ... 4 - 122 - x - Eq.) ... ... ... 5 - 2 1: Emissions from Agriculture (MMT CO Table 5 2 5 ... ... ... ... - 2 2: Emissions from Agriculture (kt) Table 5 - 3: CH Emissions from Enteric Fermentation (MMT CO Table 5 Eq.) ... ... .. 5 - 3 - 4 2 Table 5 4: CH Emissions from Enteric Fermentation (kt) ... ... ... 5 - 3 - 4 - Emissions from Enteric Fermentation (MMT Table 5 5: Approach 2 Quantitative Uncertainty Estimates for CH 4 Eq. and Percent) ... ... CO ... ... 5 - 7 ... 2 - and N O Emissions from Manure Management (MMT CO 10 Eq.) 6: CH ... ... 5 - Table 5 2 2 4 - 7: CH ... and N O Emissions from Manure Management (kt) ... Table 5 ... 5 - 10 2 4 - 8: Approa ch 2 Quantitative Uncertainty Estimates for CH Table 5 and N O (Direct and Indirect) Emissions from 2 4 Eq. and Percent) ... ... ... 5 - 14 Manure Management (MMT CO 2 - from Table 5 9: IPCC (2006) Implied Emission Factor Default Values Compared with Calculated Values for CH 4 ... ... ... ... 5 - 14 Manure Management (kg/head/year) ... - Emissions from Rice Cultivation (MMT CO Table 5 Eq.) ... ... 5 - 17 10: CH 4 2 - 11: CH Table 5 ... ... ... 5 - 17 Emissions from Rice Cultivation (kt) 4 - ... ... ... Area Harvested (1,000 Hectares) 5 - 19 Table 5 12: Rice - 13: Average Ratooned Area as Percent of Primary Growth Area (Percent) ... ... 5 - 19 Table 5 Table 5 14: Approach 2 Quantitative Uncertainty Estimates for CH - Emissions from Rice Cultivation (MMT CO 2 4 ... ... ... ... Eq. and Percent) 5 - 20 ... Table 5 - 15: N O Emissions from Agricultural Soils (MMT CO Eq.) ... ... ... 5 - 24 2 2 Table 5 - 16: N O Emissions from Agricultural Soils (kt) ... ... ... 5 - 24 2 Table 5 - O Emissions from Agricultural Soils by Land Use Type and N Input Type (MMT CO 17: Direct N Eq.) 5 - 2 2 24 18: Indirect N ... O Emissions from Agricultural Soils (MMT CO Table 5 Eq.) ... - 5 - 25 2 2 Table 5 - 19: Quantitative Uncertainty Estimates of N O Emissions from Agricultural Soil Management in 2015 2 Eq. and Percent) ... ... ... ... ... (MMT CO 5 - 37 2 Table 5 20: Emissions from Liming (MMT CO - Eq.) ... ... ... 5 - 39 2 Table 5 21: Emissions from Liming (MMT C) ... ... ... ... 5 - 39 - - ... Table 5 ... ... 5 - 41 22: Applied Minerals (MMT) ... Table 5 23: Approach 2 Quantitative Uncertainty Estimates for CO Eq. and - ions from Liming (MMT CO Emiss 2 2 Percent) ... ... ... ... ... ... 5 - 41 Table 5 - 24: CO Emissions from Urea Fertilization (MMT CO Eq.) ... ... ... 5 - 42 2 2 - 25: CO Table 5 Emissions from Urea Fertilization (MMT C) ... ... ... 5 - 42 2 43 26: Applied Urea (MMT) Table 5 ... ... ... ... 5 - - Table 5 - 27: Quantitative Uncertainty Estimates for CO Emissions from Urea Fertilization (MMT CO Eq. and 2 2 Percent) ... ... ... ... ... ... 5 - 43 44 Ta 28: CH ble 5 and N O Emissions from Field Burning of Agricultural Residues (MMT CO - Eq.) ... 5 - 2 4 2 - 29: CH , N O, CO, and NO Table 5 Emissions from Field Burning of Agricultural Residues (kt) ... 5 - 45 4 2 x ... Table 5 30: Agricultural Crop Production (kt of Product) ... ... - 5 - 47 ... 5 ... 47 - Table 5 - 31: U.S. Average Percent Crop Area Burned by Crop (Percent) 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 xvi

17 Table 5 - ... 5 - 48 32: Key Assumptions for Estimating Emissions from Field Burning of Agricultural Residues 33: Greenhouse Gas Emission Ratios and Conversion Factors ... . 5 - 48 Table 5 - ... and N O Emissions from Field Burning of Table 5 34: Approach 2 Quantitative Uncertainty Estimates for CH - 4 2 Eq. and Percent) ... ... Agricultural Residues (MMT CO 5 - 48 ... 2 Use Change, and Forestry (MMT CO 1: C Stock Change from Land Use, Land Eq.) ... - 6 - 2 - Table 6 2 2: Emissions from Land Use, Land - Use Change, and Forestry by Gas (MMT CO Table 6 - Eq.) ... 6 - 3 2 3: Emissions and Removals (Net Flux) from Land Use, Land - Use Change, and Forestry (MMT CO Table 6 Eq.) 6 - - 2 4 4: Emissions and Removals from Land Use, Land - Use Change, and Forestry (MMT CO Eq.) - ... 6 - 5 Table 6 2 5: Emissions and Removals from Land Use, Land - Use Change, and Forestry (kt) ... - 6 - 6 Table 6 ... - 6: Managed and Unmanaged Land Area by Land - Use Categories for All 50 States (Thousands of Hectares) Table 6 ... ... ... ... ... 6 - 9 ... - 7: Land Use and Land Use Change for the U.S. Managed Land Base for All 50 States (Thousands of Table 6 - 6 ... ... ... ... ... - 9 ... Hectares) 8: Data Sources Used to Determine Land Use and Land Area for the Conterminous United States, Hawaii, Table 6 - 6 ... ... ... ... ... - 15 and Alaska - 9: Total Land Area (Hectares) by Land - Use Category for U.S. Territories ... ... 6 - 21 Table 6 - Table 6 Flux fr om Forest Pools in Forest Land Remaining Forest Land and Harvested Wood Pools 10: Net CO 2 ... ... ... ... ... Eq.) 6 - 25 (MMT CO 2 - 11: Net C Flux from Forest Pools in Forest Land Remaining Forest Land and Harvested Wood Pools Table 6 25 ... ... ... ... . 6 - ... (MMT C) Table 6 - 12: Forest Area (1,000 ha) and C Stocks in Forest Land Remaining Forest Land and Harvested Wood Pools (MMT C) ... ... ... ... ... 6 - 26 Table 6 - 13: Estimates of CO (MMT per Year) Emissions from Forest Fires in the Conterminous 48 States and 2 a Alaska ... ... ... ... ... ... 6 - 28 - 14: Quantitative Uncertaint Table 6 Flux from Forest Land Remaining Forest Land : y Estimates for Net CO 2 ... ... ... Eq. and Percent) 6 - 31 Changes in Forest C Stocks (MMT CO 2 ... 15: Mean C Stocks, CO and CH Table 6 Fluxes in Alaska between 2000 and 2009 - ... 6 - 34 4 2 - 16: Non - CO Table 6 es (MMT CO 35 Eq.) ... ... ... 6 - Emissions from Forest Fir 2 2 a 35 - CO Emissions from Forest Fires (kt) - ... ... ... Table 6 - 17: Non 6 2 Emissions from Forest Fires (MMT CO - - CO Eq. and Table 6 18: Quantitative Uncertainty Estimates of Non 2 2 a 36 ... ... ... ... ... ... 6 - Percent) a, b Table 6 19: N O Emissions from N Additions to Soils - (MMT CO Eq. and kt N 37 O) ... ... 6 - 2 2 2 - 20: Quantitative Uncertainty Estimates of N Table 6 O Emissions from Soils in Forest Land Remaining Forest 2 ... Land Converted to Forest Land (MMT CO Land Eq. and Percent) and ... 6 - 39 2 a - 21: Estimated CO and Non - CO 40 Emissions on Drained Organic Forest Soils (MMT CO Table 6 - Eq.) ... 6 2 2 2 a (kt) Emissions on Drained Organic Forest Soils (MMT C) and Non - CO 40 Table 6 22: Estimated CO - ... 6 - 2 2 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 CO Emissions on Drained Organic Forest and Non - CO 2 2 a Soils (MMT CO - Eq. and Percent) 42 ... ... ... ... 6 2 xvii

18 Table 6 25: Net CO Flux from Forest C Pools in Land Converted to Forest Land by Land Use Change Category - 2 ... ... ... ... ... 6 - 43 (MMT CO Eq.) 2 Land Converted to Forest Land by Land Use Change Category 26: Net C Flux from Forest C Pools in Table 6 - ... ... ... ... . (MMT C) - 44 ... 6 27: Quantitative Uncertainty Estimates for Forest C Pool Stock Changes (MMT CO Eq. per Year) in 2015 - Table 6 2 by Land Use Change ... ... ... 6 - 46 from Land Converted to Forest Land - 28: Net CO Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT CO Table 6 Eq.) ... 6 - 49 2 2 Table 6 ... Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT C) 29: Net CO 6 - 49 - 2 - Table 6 Cropland 30: Approach 2 Quantitative Uncertainty Estimates for Soil C Stock Changes occurring within . (MMT CO Eq. and Percent) ... ... Remaining Cropland 6 - 55 ... 2 - 31: Net CO Flux from Soil C Stock Changes in Land Table 6 Converted to Cropland by Land Use Change 2 Category (MMT CO Eq.) ... ... ... ... ... 6 - 57 2 Table 6 - Flux from Soil C Stock Changes in Land Converted to Cropland (MMT C) ... 6 - 58 32: Net CO 2 - Table 6 Land 33: Approach 2 Quantitative Uncertainty Estimates for Soil C Stock Changes occurring within ... Eq. and Percent) (MMT CO ... ... 6 - 61 Converted to Cropland 2 - 34: Net CO ... Flux from Soil C Stock Changes in Grassland Remaining Grassland (MMT CO Eq.) Table 6 6 - 63 2 2 - 35: Net CO Table 6 Flux from Soil C Stock Changes in Grassland Remaining Grassland (MMT C) ... 6 - 64 2 Table 6 36: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes Occurring Within Grassland - (MMT CO Eq. and Percent) Remaining Grassland ... ... ... 6 - 67 2 Eq.) - and N 69 O Emissions from Biomass Burning in Grassland (MMT CO - Table 6 37: CH ... 6 2 2 4 Table 6 - 38: CH , N O, CO, and NO Emissions from Biomass Burning in Grassland (kt) ... ... 6 - 69 4 2 x Table 6 - Thousands of Grassland Hectares Burned Annually ... ... ... 6 - 70 39: 6 - - CO Table Greenhouse Gas Emissions from Biomass Burning in Grassland 40: Uncertainty Estimates for Non 2 Eq. and Percent) ... ... ... ... ... (MMT CO 6 - 70 2 - Table 6 Flux from Soil and Biomass C Stock Changes for Land Converted to Grassland 41: Net CO (MMT CO 2 2 ... ... ... ... ... ... 6 - 72 Eq.) Converted to Grassland - Net CO Table 6 Flux from Soil and Biomass C Stock Changes for Land (MMT C) .. 6 - 42: 2 72 - 43: Approach 2 Quantita tive Uncertainty Estimates for Soil C Stock Changes occurring within Land Table 6 (MMT CO Eq. and Percent) Converted to Grassland ... ... ... 6 - 75 2 - Peatlands Remaining Peatlands (MMT CO Table 6 44: Emissions from ... ... 6 - 79 Eq.) 2 Table 6 - 45: Emissions from Peatlands Remaining Peatlands (kt) ... ... ... 6 - 80 Table 6 - t Production of Lower 48 States (kt) ... ... ... 6 - 81 46: Pea 6 - ... ... ... - 81 Table 6 47: Peat Production of Alaska (Thousand Cubic Meters) - 48: Approach 2 Quantitative Uncertainty Estimates for CO Peatlands , CH O Emissions from , and N Table 6 4 2 2 ... (MMT CO Eq. and Percent) Remaining Peatlands ... 6 - 83 ... 2 Table 6 - 49: Net CO Vegetated Coastal Wetlands Remaining Vegetated Coastal Flux from Soil C Stock Changes in 2 Wetlands (MMT CO Eq .) ... ... ... ... 6 - 85 ... 2 Table 6 - 50: Net CO Vegetated Coastal Wetlands Remaining Vegetated Coastal Flux from Soil C Stock Changes in 2 86 (MMT C) ... ... ... ... ... 6 - Wetlands Flux from Table 6 51: Net CH (MMT CO - Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands 4 2 ... 86 ... - 6 Eq.) ... ... ... ... 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 xviii

19 Table 6 52: Net CH Flux from Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands (kt CH - ) . 6 - 86 4 4 Table 6 Vegetated - 53: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes occurring within (MMT CO 6 ... ... Eq. and Percent) - 87 Coastal Wetlands Remaining Vegetated Coastal W etlands 2 Emissions occurring within Vegetated Coastal 54: Approach 1 Quantitative Uncertainty Estimates for CH - Table 6 4 Eq. and Percent) ... ... 6 - 88 Wetlands Remaining Vegetated Coastal Wetlands (MMT CO 2 6 55: Net CO Vegetated Coastal Wetlands Converted to Unvegetated Flux from Soil C Stock Changes in Table - 2 Eq.) ... ... ... ... 6 (MMT CO 89 Open Water Coastal Wetlands - 2 - 56: Net CO Table 6 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands Converted to Unvegetated 2 (MMT C) ... ... ... ... 6 - 89 Open Water Coastal Wetlands - 57: Approach 1 Quantitative Uncertainty Estimates for Net CO Flux occurring within Vegetated Coastal Table 6 2 (MMT CO - Eq. and Percent) ... 6 Wetlands Converted to Unvegetated Open Water Coastal Wetlands 90 2 Table 6 58: Net CO - Flux from Soil C Stock Changes from Unvegetated Open Water Coastal Wetlands Converted 2 to Vegetated Coas (MMT CO tal Wetlands Eq.) ... ... ... ... 6 - 92 2 - Unvegetated Open Water Coastal Wetlands C Flux from Soil C Stock Changes from Table 6 onverted 59: Net CO 2 - ... ... to Vegetated Coastal Wetlands ... 6 (MMT C) 92 ... - 60: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes occurring within Unvegetated Table 6 Coastal Wetlands (MMT CO Eq. and Percent) ... 6 Open Water Coastal Wetlands Converted to Vegetated 93 - 2 ... - O Emissions from Aquaculture in Coastal Wetlands (MMT CO Table 6 Eq. ) ... 6 - 94 61: Net N 2 2 - 62: Net N O Emissions from Aquaculture in Coastal Wetlands (kt N 94 O) ... ... Table 6 6 - 2 2 - O Emissions for Aquaculture Production in Table 6 63: Approach 1 Quantitative Uncertainty Estimates for N 2 Eq. and Percent) Coastal Wetlands (MMT CO ... ... ... 6 - 95 ... 2 Table 6 - 64: Net CO Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT 2 Eq.) CO ... ... ... ... ... .. 6 - 96 2 Table 6 65: Net CO - Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT C) 2 ... ... ... ... 6 - 96 ... ... Eq.) 66: Net CH - Land Converted to Vegetated Coastal Wetlands (MMT CO Table 6 ... 6 - 96 Flux in 4 2 - 67: Net CH Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal Wetlands (kt CH - ) Table 6 6 4 4 97 - 68: Approach 1 Quantitative Uncertainty Estimates for Net CO Flux Changes occurring within Land Table 6 2 (MMT CO ... ... Converted to Vegetated Coastal Wetlands 6 - 98 Eq. and Percent) 2 - Table 6 Land Converted to Emissions occurring within 69: Approach 1 Quantitative Uncertainty Estimates for CH 4 ... tal Wetlands Eq. and Percent) Vegetated Coas (MMT CO ... 6 - 98 ... 2 Table 6 - 70: Net CO Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT CO 100 Eq.) . 6 - 2 2 - Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT C) Table 6 ... 6 - 100 71: Net CO 2 Table 6 72: Thousands of Hectares of Drained Organic Soils in Settlements Remaining Settlements ... 6 - 100 - Table 6 - 73: Uncertainty Estimates for CO Emissions from Drained Organic Soils in Settlements Remaining 2 Settlements (MMT CO Eq. and Percent) ... ... ... ... 6 - 101 2 Table 6 74: Net C Flux from Urban Trees (MMT CO Eq. and MMT C) ... ... - 6 - 102 2 Table 6 - 75: Annual C Sequestration (Metric Tons C/Year), Tree Cover (Percent), and Annual C Sequestration per 2 ... - yr) for 50 states plus th e District of Columbia (2015) ... Area of Tree Cover (kg C/m 6 - 104 Table 6 76: Approach 2 Quantitative Uncertainty Estimates for Net C Flux from Changes in C St ocks in Urban - Trees (MMT CO 106 Eq. and Percent) ... ... ... ... 6 - 2 107 Table 6 - 77: N - O Emissions from Soils in Settlements Remaining Settlements (MMT CO 6 Eq. and kt N O) ... 2 2 2 xix

20 Table 6 - 78: Quantitative Uncertainty Estimates of N O Emissions from Soils in Settlements Remaini ng Settlements 2 . ... ... ... ... Eq. and Percent) 6 - 109 (MMT CO 2 79: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfil CO Table 6 Eq.) - ... 6 - 110 ls (MMT 2 - C) ... 6 80: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills (MMT 110 Table 6 - - 81: Moisture Contents, C Storage Factors (Proportions of Initial C Sequestered), Initial C Contents, and Table 6 s for Yard Trimmings and Food Scraps in Landfills ... Decay Rate ... 6 - 113 ... Table 6 82: C Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C) ... ... 6 - 113 - - Flux from Yard Trimmings and Food Scraps in Table 6 83: Approach 2 Quantitative Uncertainty Estimates for CO 2 (MMT CO ... Eq. and Percent) ... ... ... Landfills 6 - 114 2 Table 6 - 84: Net CO Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to 2 Settlements (MMT CO Eq.) ... ... ... ... .. 6 - 115 2 Table 6 85: Net CO - Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to 2 Settlements ... ... ... ... ... 6 - 116 (MMT C) - Land Converted ccurring within Table 6 86: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes o ... (MMT CO Eq. and Percent) to Settlements ... ... 6 - 119 ... 2 Table 7 - 1: Emissions from Waste (MMT CO Eq.) ... ... ... . 7 - 1 2 Table 7 2: Emissions from Waste (kt) ... ... ... - 7 - 2 ... Table 7 - 3: CH Emissions from Landfills (MMT CO 4 Eq.) ... ... ... 7 - 2 4 Table 7 - 4: CH Emissions fr om Landfills (kt) ... ... ... ... 7 - 5 4 Table 7 - 5: Approach 2 Quantitative Uncertainty Estimates for CH Eq. and Emissions from Landfills (MMT CO 2 4 Percent) ... ... ... ... 7 - 11 ... ... a 6: Materials Discarded in the Municipal Waste Stream by Waste Type from 1990 to 2014 (Percent) ... 7 - - Table 7 18 7: CH - and N Table 7 O Emissions from Domestic and Industrial Wastewater Treatment (MMT CO Eq.) 20 ... 7 - 4 2 2 - 8: CH 21 and N Table 7 O Emissions from Domestic and Industrial Wastewater Treatment (kt) ... 7 - 2 4 - Produced (kt) ... ... 7 - 23 Table 7 9: U.S. Population (Millions) and Domestic Wastewater BOD 5 10: Domestic Wastewater CH Table 7 - Eq. and Emissions from Septic and Centralized Systems (2015, MMT CO 2 4 ... ... ... Percent) ... ... 7 - 24 ... Table 7 11: Industrial Wastewater CH Emissions by Sector (2015, MMT CO 24 Eq. and Percent) - 7 - ... 2 4 Table 7 - 12: U.S. Pulp and Paper, Meat, Poultry, Vegetables, Fruits and Juices, Ethanol, and Petroleum Refining Production (MMT) ... ... ... ... ... 7 - 24 Table 7 - ... 7 - 26 13: Variables Used to Calculate Percent Wastewater Treated Anaerobically by Industry (percent) 3 - /ton) and BOD Production (g/L) for U.S. Vegetables, Fruits, and Juices Table 7 14: Wastewater Flow (m ... ... ... ... ... 7 - 27 Production - Table 7 15: U.S. Population (Millions), Population Served by Biological Denitrification (Millions), Fraction of ailable Protein (kg/person - year), Protein Consumed Population Served by Wastewater Treatment (percent), Av ... - - (kg/person ... ... year), and Nitrogen Removed with Sludge (kt 7 - 31 N/year) Table 7 - 16 : Approach 2 Quantitative Uncertainty Estimates for CH Emissions from Wastewater Treatment (MMT 4 CO Eq. and Percent) ... ... ... ... 7 - 32 ... 2 Table 7 - 17: CH and N 35 O Emissions from Composting (MMT CO - Eq.) ... ... 7 2 4 2 Table 7 - 18: CH 35 and N - ... ... ... 7 O Emissions from Composting (kt) 2 4 - 7 ... ... 36 Table 7 - 19: U.S. Waste Composted (kt) ... ... 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 xx

21 Table 7 - es for Emissions from Composting (MMT CO 20: Approach 1 Quantitative Uncertainty Estimat Eq. and 2 Percent) ... ... ... ... ... 7 - 36 ... 21: Emissions of NO - , CO, and NMVOC from Waste (kt) ... ... ... 7 - 37 Table 7 x - Table 9 - Eq.) ... ... 9 1: Revisions to U.S. Greenhouse Gas Emissions (MMT CO 4 2 Table 9 - 2: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land - Use Eq.) ... Change, and Forestry (MMT CO ... ... 9 - 6 ... 2 Figures 1: Gross U.S. Greenhouse Gas Emissions by Gas (MMT CO Figure ES Eq.) ... - ES - 4 ... 2 - 2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the Previous Year .. ES Figure ES 5 - Figure ES 3: Cumulative Change in Annual Gross U.S. Greenhouse Gas Emissions Relative to 1990 (1990=0, - ... ... Eq.) ... MMT CO ... ... ES - 5 2 Figure ES - 4: 2015 U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO Eq.) ... ES - 8 2 Figure ES 5: 2015 Sources of CO - Emissions (MMT CO 9 Eq.) ... ... ... ES - 2 2 - Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT CO 10 Figure ES Eq.) 6: 2015 CO ... ES - 2 2 Figure ES - 7: 2015 End - Use Sector Emissions of CO from Fossil Fuel Combustion (MMT CO 10 Eq.) ... ES - 2 2 ... - Eq.) Figure ES 8: Electricity Generation (Billion kWh) and Electricity Generation Emissions (MMT CO ES - 13 2 Figure ES - 9: 2015 Sources of CH Emissions (MMT CO 14 Eq.) ... ... ... ES - 2 4 Figure ES - 10: 2015 Sources of N O Emissions (MMT CO 16 Eq.) ... ... ... ES - 2 2 Eq.) Figure ES - 11: 2015 Sources of HFCs, PFCs, SF , and NF - Emissions (MMT CO ES 17 ... ... 2 3 6 - 12: U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Figure ES Eq.) ... ES - 18 Sector (MMT CO 2 - ... ... ES - Figure ES 20 13: 2015 U.S. Energy Consumption by Energy Source (Percent) 24 14: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO - ... ES - Figure ES Eq.) 2 Figure ES - 15: U.S. Greenhouse Gas Emissions with Electricity - Related Emissions Distributed to Economic Sectors Eq.) ... ... ... ... ... ES (MMT CO 26 - 2 - 16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product (GDP) .. ES - 27 Figure ES ... 17: 2015 Key Categories (MMT CO - Eq.) ... ... Figure ES ES - 28 2 Figure 1 1: National Inventory Arrangements Diagram ... ... ... 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 CO Eq.) ... ... 2 - 1 2 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 Eq.) ... ... ... ... ... ... 2 - CO 2 2 ... - Eq.) Figure 2 4: U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO 2 - 8 2 Figure 2 - 5: 2015 Energy Chapter Greenhouse Gas Sources (MMT CO Eq.) ... ... 2 - 10 2 Fi - 6: 2015 U.S. Fossil Carbon Flows (MMT CO Eq.) gure 2 ... ... 2 - 11 ... 2 Figure 2 - 7: 2015 CO Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT CO 13 Eq.) ... 2 - 2 2 Figure 2 - 8: 2015 End - Use Sector Emissions of CO 13 from Fossil Fuel Combustion (MMT CO Eq.) ... 2 - 2 2 14 Figure 2 - 9: Electricity Generation (Billion kWh) and Emissions (MMT CO - Eq.) ... ... 2 2 xxi

22 Figure 2 - 10: 2015 Industrial Processes and Product Use Chapter Greenhouse Gas Sources (MMT CO Eq.) ... 2 - 16 2 Figure 2 - Eq.) ... ... 2 - 18 11: 2015 Agriculture Chapter Greenhouse Gas Sources (MMT CO 2 - ... Eq.) ... Figure 2 2 - 22 12: 2015 Waste Chapter Greenhouse Gas Sources (MMT CO 2 Figure 2 Eq.) ... 13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO 2 - 24 - 2 - 14: U.S. Greenhouse Gas Emissions with Electricity - Related Emissions Distributed to Economic Sectors Figure 2 ... ... (MMT CO ... ... ... Eq.) 2 - 27 2 - ... 2 - 34 Figure 2 15: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product - 1: 2015 Energy Chapter Greenhouse Gas Sources (MMT CO - Eq.) ... ... 3 Figure 3 1 2 Figure 3 2: 2015 U.S. Fossil Carbon Flows (MMT CO Eq .) - ... ... 3 - 2 ... 2 Figure 3 - 3: 2015 U.S. Energy Consumption by Energy Source (Percent) ... ... 3 - 8 Figure 3 - ... ... ... 3 - 8 4: U.S. Energy Consumption (Quadrillion Btu) - - 5: 2015 CO Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT CO Eq.) ... 3 Figure 3 9 2 2 - 6: U.S. Gasoline Consumption (Tbtu) and CO ... Emissions by Sector (MMT CO Eq.) Figure 3 3 - 9 2 2 Figure 3 - 7: Annual Deviations from Normal Heating Degree Days for the United States (1950 – 2015, Index Normal = 100) ... ... ... ... ... ... 3 - 10 Figure 3 - – 2015, Index Normal 8: Annual Deviations from Normal Cooling Degree Days for the United States (1950 - ... ... ... ... = 100) 3 ... 11 ... Figure 3 - 9: Nuclear, Hydroelectric, and Wind Power Plant Capacity Factors in the United States (1990 – 2015, Percent) ... ... ... ... ... ... 3 - 12 ... Figure 3 10: Electricity Generation (Billion kWh) and Emissions (MMT CO Eq.) - ... 3 - 17 2 Figure 3 - 11: Electricity Generation Retail Sales by End - Use Sector (Billion kWh) ... ... 3 - 17 Figure 3 - ... ... ... 3 - 19 12: Industrial Production Indices (Index 2007=100) - 13: Sales Weighted Fuel Economy of New Passenger Cars and Light - Duty Trucks, 1990 – 2015 Figure 3 - ... ... ... ... 3 - 23 (miles/gallon) ... ... 14: Sales of New Passenger Cars and Light - – 20 15 (Percent) ... - 3 - 23 Figure 3 Duty Trucks, 1990 - 15: Mobile Source CH and N 26 O Emissions (MMT CO - Eq.) ... Figure 3 . 3 ... 2 4 2 ... - Related CO Figure 3 Emissions Per Capita and Per Dollar GDP - 3 - 33 16: U.S. Energy Consumption and Energy 2 - 1: 2015 Industrial Processes and Product Use Chapter Greenhouse Gas Sources (MMT CO Eq.) Figure 4 4 - 2 ... 2 ... - Eq.) Figure 4 2: U.S. HFC Consumption (MMT CO ... 4 - 110 ... 2 Figure 5 - 1: 2015 Agriculture Chapter Greenhouse Gas Emission Sources (MMT CO Eq.) ... ... 5 - 1 2 Figure 5 2: Annual CH - Emissions from Rice Cultivation, 2015 (MMT CO Eq./Year) ... ... 5 - 18 2 4 - 3: Sources and Pathways of N that Result in N O Emissions from Agricultural Soil Management ... 5 Figure 5 23 - 2 Figure 5 - 4: Crops, 2015 Annual Direct N O Emissions Estimated Using the Tier 3 DAYCENT Model (MMT CO 2 2 Eq./year) ... ... ... ... ... .. 5 - 25 Figure 5 - 5: Grasslands, 2015 Annual Direct N O Emissions Estimated Using the Tier 3 DAYCENT Model (MMT 2 ... Eq./year) CO ... ... ... 5 - 26 ... 2 Figure 5 - 6: Crops, 2015 Annual Indirect N T Model O Emissions from Volatilization Using the Tier 3 DAYCEN 2 ... Eq./year) ... ... ... ... (MMT CO 5 - 27 2 Figure 5 - 7: Grasslands, 2015 Annual Indirect N O Emissions from Volatilization Using the Tier 3 DAYCENT 2 ... 28 ... - 5 Model (MMT CO ... Eq./year) ... ... 2 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 xxii

23 Figure 5 - O Emissions from Leaching and Runoff Using the Tier 3 DAYCENT 8: Crops, 2015 Annual Indirect N 2 ... ... ... ... ... 5 - 29 Model (MMT CO Eq./year) 2 9: Grasslands, 2015 Annual Indirect N O Emissions from Leaching and Runoff Using the Tier 3 Figure 5 - 2 Eq./year) ... ... ... ... DAYCENT Model (MMT CO 5 - 30 2 10: Comparison of Measured Emissions at Field Sites and Modeled Emissions Using the DAYCENT - Figure 5 ... O per ha per year) Simulation Model and IPCC Tier 1 A ... pproach (kg N 5 - 38 2 1: Percent of Total Land Area for Each State in the General Land - Use Categories for 2015 ... 6 - 11 Figure 6 - 2: Changes in Forest Area by Region for in the conterminous United - Forest Land Remaining Forest Land Figure 6 2015, Million Hectares) ... ... ... States and coastal Alaska (1990 - 24 - 6 - 3: Estimated Net Annual Changes in C Stocks for All C Pools in Forest Land Remaining Forest Land in Figure 6 l Alaska (MMT C per Year) 6 ... ... the Conterminous U.S. and Coasta - 27 ... Figure 6 4: Total Net Annual CO - Flux for Mineral Soils under Agricultural Management within Stat es, 2015, 2 Cropland Remaining Cropland ... ... ... 6 - 50 ... 5: Total Net Annual CO Flux for Organic Soils under Agricultural Management within States, 2015, - Figure 6 2 ... ... ... ... Cropland Remaining Cropland - 51 6 - 6: Total Net Annual CO Flux for Mineral Soils under Agricultural Management within States, 2015, Figure 6 2 ... ... ... Grassland Remaining Grassland 6 - 64 ... Figure 6 7: Total Net Annual CO - Flux for Organic Soils under Agricultural Management within States, 2015, 2 Grassland Remaining Gras ... ... ... ... 6 - 65 sland ... Figure 7 Eq.) ... 1: 2015 Waste Chapter Greenhouse Gas Sources (MMT CO 7 - 1 - 2 Figure 7 - 2: Comparison of the Revised Inventory Methodology to EPA’s GHGRP Subpart HH Emissions ... 7 - 14 2014 Inventory Methodology to the Revised Inventory Methodology Figure 7 - 3: Comparison of the 1990 - ... 7 - 15 Figure 7 - ... ... 7 - 17 4: Management of Municipal Solid Waste in the United States, 2014 gure 7 - ... ... ... 7 - 17 Fi 5: MSW Management Trends from 1990 to 2014 - - ... ... 7 (Percent) 18 Figure 7 6: Percent of Recovered Degradable Materials from 1990 to 2014 Boxes - 1: Methodological Approach for Estimating and Reporting U.S. Emissions an d Sinks ... ES - 1 Box ES Box ES 2: EPA’s Greenhouse Gas Reporting Program ... ... ... ES - 2 - - ES ... Box ES - 13 3: Use of Ambient Measurements Systems for Validation of Emission Inventories ... - - Related Data ... Box ES ES - 26 4: Recent Trends in Various U.S. Greenhouse Gas Emissions Box ES - 5: Recalculations of Inventory Estimates ... ... ... ES - 29 Box 1 - for Estimating and Reporting U.S. Emissions and Sinks ... 1 - 2 1: Methodological Approach - IPCC Fifth Assessment Report and Global Warming Potentials ... ... Box 1 1 - 9 2: The ... - ... ... ... 3: IPCC Reference Approach 1 - 15 Box 1 Box 2 - 1: Methodology for Aggregating Emissions by Economic Sector ... ... 2 - 32 ... Box 2 Trends in Various U.S. Greenhouse Gas Emissions - Related Data 2: Recent ... 2 - 33 - Box 2 - 3: Sources and Effects of Sulfur Dioxide ... ... ... ... 2 - 36 Box 3 - 1: Methodological Approach for Estimating and Reporting U.S. Emissions and Sinks ... 3 - 3 4 Box 3 - 2: Energy Data from EPA’s Greenhouse Gas Reporting Program ... ... 3 - xxiii

24 Box 3 - Fossil Energy Effects on CO 3: Weather and Non from Fossil Fuel Combustion Trends ... 3 - 10 - 2 - of Greenhouse Gas Reporting Program Data and Improvements in Reporting Emissions from Box 3 4: Uses ... ... ... ... 3 - 31 Industrial Sector Fossil Fuel Combustion 5: Carbon Intensity of U.S. Energy Consumption ... ... - ... 3 - 32 Box 3 3 - Product Use in Energy Sector ... - 51 Box 3 6: Reporting of Lubricants, Waxes, and Asphalt and Road Oil - 7: Carbon Dioxide Transport, Injection, and Geological Storage ... ... 3 - 75 Box 3 Box 4 1: Methodological Approach for Estimating and Reporting U.S. Emissions and Sinks ... 4 - 6 - 4 - ... ... - 7 Box 4 2: Industrial Processes Data from EPA’s Greenhouse Gas Reporting Program - 1: Methodological Approach for Estimating and Reporting U.S. Emissions and Sinks ... 5 - 2 Box 5 - Box 5 O Emissions ... ... ... 5 - 31 2: Tier 1 vs. Tier 3 Approach for Estimating N 2 - 3: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach ... Box 5 5 - 40 Box 5 - 4: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach ... 5 - 46 Box 6 - 1: Methodological Approach for Estimating and Reporting U.S. Emissions and Sinks ... 6 - 7 Box 6 - ... ... ... 6 - 20 2: Preliminary Estimates of Land Use in U.S. Territories ... - Emissions from Forest Fires ... ... ... Box 6 6 - 27 3: CO 2 - 4: Preliminary Estimates of Historical Carbon Stock Change and Methane Emis sions from Managed Land in Box 6 ... ... ... . 6 - 34 Alaska (Represents Mean for Years 2000 to 2009) Box 6 Compared to Tier 1 or 2 Approaches ... ... 5: Tier 3 Approach for Soil C Stocks 6 - 52 - Box 6 - 6: Grassland Woody Biomass Analysis ... ... ... ... 6 - 68 Box 7 - ... 7 - 2 1: Methodological Approach for Estimating and Reporting U.S. Emissions and Sinks - ... ... .. Box 7 7 - 2 2: Waste Data from EPA’s Greenhouse Gas Reporting Program Box 7 - 3: Nati onwide Municipal Solid Waste Data Sources ... ... ... 7 - 15 ... Box 7 4: Overview of the Waste Sector ... ... - ... 7 - 16 ... ... 18 - 7 Box 7 - 5: Description of a Modern, Managed Landfill ... 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 xxiv

25 Executive Summary 1 sources and sinks of An emissions inventory that identifies and quantifies a country's primary anthropogenic 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 mechanism that enables Parties to the United Nations Framework Conventi on 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 objecti ve 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 th at 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 2 proceed in a sustainable manner.” 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 a ll greenhouse gases not controlled by 3 The United States views this report as an opportunity the Montreal Protocol, using comparable methodologies...” to fulfill these commitments. This chapter summarizes the latest information on U.S. anthropogenic greenhou se gas emission trends from 1990 through 201 5 . 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 overnmental Panel on Climate Change (IPCC) Guidelines for National Greenhouse Gas Inventories Interg (IPCC 4 The structure of this report is consistent with the UNFCCC guidelines for inventory reporting. 2006). Approach for Estimating and Reporting U.S. Emissions and Sinks - 1 : Methodological Box ES 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 sinks . The net emissions total presented in this report for the from land use, land - use change, and forestry ( LULUCF ) United States includes emissions and sinks from LULUCF. ethods All emissions and sinks are calculated using m 1 The term “anthropogenic,” in this context, refers to greenhouse gas emissions and removals that are a direct result of human activities or are the result of natural processes that have been affected by human activities (IPCC 2006). 2 Article 2 of the Fram ework Convention on Climate Change published by the UNEP/WMO 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 y the Conference of the Parties elaborated the role of Annex I Parties in preparing national inventories. decisions b See . 4 >. e/docs/2013/cop19/eng/10a03.pdf See

26 5 Additionally, the calculated accepted guidelines provided by the IPCC. that are consistent with - internationally emissions and sinks in a given year for the United States are presented in a common manner in line with the 6 UNFCCC reporting guidelines for the reporting of inventories under this international agreement. The use of consistent methods to calculate emissions and sinks by all nations providing their inventories to the UNFCCC ensures that these reports are comparable. The report itself follows t his standardized format, and provides an explanation of the methods used to calculate emissions and sinks, and the manner in which those calculations are conducted. Box ES - 2 : EPA’s Greenhouse Gas Reporting Program On October 30, 2009, the U.S. Environmental 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 Greenhouse Gas R eporting Program (GHGRP). 40 CFR part 98 applies to direct greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers, and facilities that inject carbon dioxide (CO ) 2 7 Reporting is at the facili ty level, except for certain suppliers of underground for sequestration or other reasons. fossil fuels and industrial greenhouse gases. The GHGRP dataset and the data presented in this Inventory report are complementary . The GHGRP data set continue s to be an important resource for the Inventory, provid ing 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 categories of emissions ory estimates in new ways that can highlight differences across regions and sub - invent , along with enhancing application of QA/QC procedures and assessment of uncertainties. uses annual GHGRP data in a number of category estimates and EPA continues to analyze the data on an annual es presented in this Inventory consistent with basis, as applicable, for further use to improve the national estimat 8 IPCC guidance . ES.1 Background Information absorb infrared radiation, thereby heat and mak ping Greenhouse gases ing the planet warmer. The most trap important greenhouse gases directly emitted by humans include carbon dioxide ( CO , methane (CH ) ), nitrous oxide 2 4 O occur naturally (N , and N O), and several other fluorine - containing halogenate d substances. Although CO , CH 2 4 2 2 From the pre - in the atmosphere, human activities have changed their atmospheric concentrations. industrial era (i.e., ending about 1750) to 201 , concentrations of these greenhouse gases have increas ed globally by 44 , 162 , and 21 5 percent, respectively (IPCC 2013 and NOAA/ESRL 201 7 ). 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 gas es, when a gas influences the atmospheric lifetimes of other gases, and/or when a gas 5 See < http://www.ipcc - nggip.iges.or.jp/public/index.html>. 6 http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf >. See < 7 See and . 8 See < > nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf - http://www.ipcc - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 2 ES

27 9 affects atmospheric processes that alter the radiative balance of the earth (e.g., affect cloud formation or albedo). P) concept to compare the ability of each greenhouse gas to The IPCC developed the Global Warming Potential (GW trap heat in the atmosphere relative to another gas. the accumulated radiative forcing within a specific time The GWP of a greenhouse gas is defined as the ratio of horizon caused by emitting 1 ki (IPCC 20 14) . The logram of the gas, relative to that of the reference gas CO 2 reference gas used is CO weighted emissions are measured in million metric tons of CO , and therefore GWP - 2 2 , 10 11 esented in units of MMT CO All gases in this Executive Summary are pr Eq.). equivalent (MMT CO Eq. 2 2 Emissions by gas in unweighted mass tons are provided in the Trends chapter of this report. IPCC Fourth UNFCCC reporting guidelines for national inventories require the use of GWP values from the 12 All estimates are provided throughout the report in both CO Assessment Report (AR4) (IPCC 2007). equivalents 2 A comparison of emission values using the AR4 GWP values versus the SAR (IPCC 1996), and unweighted units. (AR5) (IPCC 2013) GWP val ues can be found in Chapter 1 and, in more IPCC Fifth Assessment Report and the The GWP values used in this report are listed below in T able ES - 1 . detail, in Annex 6.1 of this report. able ES - - Year Time Horizon) Used in this Report T 1 : Global Warming Potentials (100 Gas GWP 1 CO 2 a CH 25 4 O 298 N 2 23 14,800 HFC - HFC 675 - 32 - 125 3,500 HFC HFC - 134a 1,430 HFC 143a - 4,470 124 - 152a HFC - 3,220 HFC 227ea 236fa - HFC 9,810 1,640 HFC - 4310mee 7,390 CF 4 C 12,200 F 2 6 F C 8,860 10 4 C F 9,300 6 14 22,800 SF 6 NF 17,200 3 a The CH GWP includes the direct 4 effects and those indirect effects due to the production of tropospheric ozone and stratospheric water vapor. The indirect effect due to production of CO is not included. 2 Source: IPCC (2007) 9 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. 10 Carbon comprises 12/44 of carbon dioxide by weight. 11 12 One million metric ton is equal to 10 grams or one teragram . 12 > http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf See < . 3 - ES Executive Summary

28 ES.2 Recent Trends in U.S. Greenhouse Gas Emissions and Sinks 5 , total gross U.S. greenhouse gas emissions were 6,586.7 million metric tons (MMT) of CO In 201 Eq. Total U.S. 2 from 201 3.5 percent from 1990 to 201 5 , and emissions decreased to 201 4 emissions have increased by 5 by 2.3 percent ( MMT CO Eq.). The decrease in total greenhouse gas emissions between 2014 and 2015 was driven 153.0 2 in large part by a decrease in CO emissions from fossil emissions from fossil fuel combustion. The decrease in CO 2 2 substitution from coal to natural gas consumption in fuel combustion was a result of multiple factors, including: ( 1 ) 5 in 201 ) creased demand for heating fuel in de resulting in a winter conditions the electric power sector ; ( 2 warmer the residential and commercial sectors and ( 3 ) a slight decrease in electricity demand . Relative to 1990, the baseline ; for this Inventory, gross emissions in 2015 are higher by 3.5 percent, down from a high of 15.5 percent above 1990 Figure ES levels in 2007. - 1 through Figure ES - 3 l illustrate the overall trends in total U.S. emissions by gas, annua changes, and absolute change since 1990. Overall, net emissions in 201 5 were 11.5 percent below 2005 levels as shown in - 2 . Table ES Table ES - 2 provides a detailed summary of gross U.S. greenhouse gas emissions and sinks for 1990 through 201 5 . Eq.) 1 U.S. Greenhouse Gas Emissions by Gas (MMT CO - : Figure ES Gross 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 4 ES

29 Figure ES - Gross U.S. Greenhouse Gas Emissions Relative to the 2 : Annual Percent Change in Previous Year 3 Figure ES l Gross U.S. Greenhouse Gas Emissions Relative to - : Cumulative Change in Annua 1990 (1990=0, MMT CO Eq.) 2 Table ES - 2 : Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO Eq.) 2 2012 2005 2011 Gas/Source 1990 2013 2014 2015 CO 5,123.0 6,131.8 5,411.4 5,362.1 5,514.0 5,565.5 5,569.5 ₂ 4,740.3 5,202.3 Fossil Fuel Combustion 5,049.8 5,746.9 5,227.1 5,024.6 5,156.5 1,820.8 2,400.9 2,157.7 Electricity Generation 2,038.1 2,038.0 1,900.7 2,022.2 a 1,707.6 Transportation 1,493.8 1,887.0 1,696.8 1,713.0 1,742.8 1,736.4 a 782.9 828.0 Industrial 842.5 812.2 806.1 805.5 775.0 319.6 345.4 Residential 338.3 357.8 325.5 282.5 329.7 a 220.4 Commercial 217.4 223.5 196.7 221.0 228.7 246.2 - ES 5 Executive Summary

30 27.6 49.7 40.9 43.5 42.5 41.4 41.4 U.S. Territories 117.6 138.9 109.8 106.7 Energy Use of Fuels 123.6 119.0 125.5 Non - Iron and Steel Production & 101.5 61.1 68.0 55.4 53.3 58.6 48.9 Metallurgical Coke Production 35.2 30.1 35.7 Natural Gas Systems 38.5 42.4 42.4 37.7 Cement Production 46.2 32.2 35.3 36.4 39.4 39.9 33.5 Petrochemical Production 27.0 26.3 26.5 26.4 26.5 28.1 21.3 11.7 14.6 14.0 14.0 14.2 13.3 Lime Production 13.8 Other Process Uses of Carbonates 4.9 6.3 9.3 8.0 10.4 11.8 11.2 13.0 9.2 9.3 9.4 10.0 9.6 10.8 Ammonia Production 8.0 12.5 10.6 10.4 10.4 10.6 10.7 Incineration of Waste Urea Fertilization 2.4 3.5 4.1 4.3 4.5 4.8 5.0 1.4 4.1 4.0 4.2 4.5 4.3 Carbon Dioxide Consumption 1.5 3.9 4.3 3.9 6.0 Liming 3.6 3.8 4.7 3.6 3.9 4.2 3.9 3.7 3.6 3.6 Petroleum Systems Soda Ash Production and 3.0 2.7 2.8 2.8 2.8 2.8 2.8 Consumption 6.8 4.1 3.3 3.4 3.3 2.8 2.8 Aluminum Production 2.2 1.4 1.7 1.9 1.8 1.9 2.0 Ferroalloy Production 1.8 1.7 1.5 1.2 1.7 1.6 Titanium Dioxide Production 1.7 1.5 1.9 1.3 1.2 1.3 1.3 1.3 Glass Production Urea Consumption for Non - 3.7 4.0 4.4 4.0 1.4 1.1 Agricultural Purposes 3.8 Phosphoric Acid Production 1.5 1.3 1.2 1.1 1.1 1.0 1.0 0.9 Zinc Production 0.6 1.0 1.3 1.5 1.4 1.0 0.5 0.5 Lead Production 0.5 0.6 0.5 0.5 0.5 Silicon Carbide Production and 0.2 0.2 0.2 0.2 0.2 Consumption 0.2 0.4 Magnesium Production and + + + + + + Processing + Wood Biomass, Ethanol, and b 219.4 230.7 276.4 276.2 299.8 307.1 291.7 Biodiesel Consumption c 113.1 International Bunker Fuels 111.7 105.8 103.5 99.8 103.2 110.8 780.8 680.9 CH 672.1 666.1 658.8 659.1 655.7 4 164.2 168.9 166.7 165.5 168.9 166.5 164.2 Enteric Fermentation Natural Gas Systems 159.7 154.5 156.2 159.2 162.5 162.4 194.1 179.6 134.3 119.0 120.8 116.7 116.6 115.7 Landfills 56.3 63.0 65.6 63.3 62.9 66.3 Manure Management 37.2 96.5 64.1 71.2 66.5 64.6 64.8 60.9 Coal Mining Petroleum Systems 55.5 46.0 48.0 46.4 44.5 43.0 39.9 16.0 15.3 15.1 14.9 14.8 14.8 Wastewater Treatment 15.7 11.3 16.7 14.1 11.3 Rice Cultivation 11.4 11.2 16.0 8.5 7.4 7.1 6.6 8.0 8.1 7.0 Stationary Combustion 6.6 6.4 6.2 6.2 6.3 6.4 Abandoned Underground Coal Mines 7.2 0.4 1.9 1.9 1.9 2.0 2.1 2.1 Composting a 5.6 2.8 2.3 2.2 2.1 2.1 2.0 Mobile Combustion Field Burning of Agricultural 0.2 0.2 0.3 0.3 0.3 0.3 0.3 Residues 0.2 0.1 + 0.1 0.1 0.2 Petrochemical Production 0.1 Ferroalloy Production + + + + + + + Silicon Carbide Production and + + Consumption + + + + + Production & Iron and Steel + Metallurgical Coke Production + + + + + + – - Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 6 5 201 ES

31 + + Incineration of Waste + + + + + c 0.1 0.1 0.1 0.1 0.1 0.1 International Bunker Fuels 0.2 N 361.6 364.0 340.7 335.5 335.5 334.8 359.5 O 2 256.6 259.8 270.1 254.1 250.5 250.0 251.3 Agricultural Soil Management 21.3 20.2 21.4 22.9 23.4 23.1 Stationary Combustion 11.9 16.5 17.4 17.5 17.5 17.5 17.7 Manure Management 14.0 a 18.5 35.7 22.8 20.4 Mobile Combustion 16.6 15.1 41.2 12.1 11.3 10.9 10.5 10.7 10.9 11.6 Nitric Acid Production 4.8 4.4 4.8 4.9 4.9 5.0 Wastewater Treatment 3.4 5.5 7.1 10.2 15.2 3.9 5.4 4.3 Adipic Acid Production N O from Product Uses 4.2 4.2 4.2 4.2 4.2 4.2 4.2 ₂ 0.3 1.7 1.7 1.7 1.8 1.9 1.9 Composting 0.5 0.4 0.3 Incineration of Waste 0.3 0.3 0.3 0.3 + 0.1 0.2 0.2 0.2 0.2 0.2 Semiconductor Manufacture Field Burning of Agricultural Residues 0.1 0.1 0.1 0.1 0.1 0.1 0.1 c 0.9 1.0 1.0 International Bunker Fuels 0.9 0.9 0.9 0.9 46.6 154.3 120.0 HFCs 155.9 159.0 166.7 173.2 Substitution of Ozone Depleting d 145.3 0.3 99.7 168.5 Substances 150.2 154.6 161.3 HCFC - 22 Production 46.1 20.0 8.8 5.5 4.1 5.0 4.3 0.2 Semiconductor Manufacture 0.2 0.2 0.2 0.2 0.3 0.3 Magnesium Production and Processing 0.0 0.0 + + 0.1 0.1 0.1 24.3 6.9 6.0 5.8 5.2 PFCs 6.7 5.8 2.8 3.2 3.4 3.0 2.8 3.2 3.2 Semiconductor Manufacture 3.4 3.5 2.9 3.0 21.5 2.5 2.0 Aluminum Production of Ozone Depleting Substitution 0.0 + + + + + + Substances 11.7 9.2 6.8 6.4 6.6 5.8 28.8 SF 6 Electrical Transmission and 23.1 8.3 6.0 Distribution 4.6 4.8 4.2 4.8 Magnesium Production and 5.2 2.7 2.8 1.6 1.5 1.0 0.9 Processing 0.5 0.7 Semiconductor Manufacture 0.7 0.4 0.4 0.4 0.7 0.5 + 0.7 NF 0.6 0.5 0.6 0.6 3 + 0.5 0.7 0.6 0.6 0.5 0.6 Semiconductor Manufacture 7,313.3 6,776.7 6,538.3 6,680.1 6,739.7 6,586.7 Total Emissions 6,363.1 e 10.6 23.0 19.9 LULUCF Emissions 19.2 19.7 19.7 26.1 f (830.2 ) (754.0 ) (769.1 ) Carbon Stock Change ) (782.2 ) (781.1 ) (778.7 ) (779.8 LULUCF g Net LULUCF Sector (819.6 ) (731.0 ) (749.2 ) Total (753.8 ) (763.0 ) (761.4 ) (758.9 ) Net Emissions (Sources and Sinks) 5,543.5 6,582.3 6,027.6 5,784.5 5,917.1 5,978.3 5,827.7 Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF. Eq. + Does not exceed 0.05 MMT CO 2 a There was a method update in this Inventory for estimating the share of gasoline used in on - road and non - road applications. The change does not im pact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the transportation sector to industrial and commercial sectors for 2015, creating a break in the time series. The change is discussed further in the Planned Imp . Chapter 3.1 rovements section of b Emissions from Wood Biomass an d Biofuel Consumption are not included specifically in summing Energy sector totals. Net carbon - Use Change, fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for Land Use, Land and Forestry. c Emissions from International Bunker Fuels are not included in totals. d Small amounts of PFC emissions also result from this source. 7 - ES Executive Summary

32 e and N LULUCF emissions include the CH O emissions reported for Peatlands Remaining Peatlands , Forest Fires, Drained 4 2 Organic Soils, Grassland Fires, and Coastal Wetlands Remaining Coastal Wetlands ; CH emissions from Land Converted to 4 ; and N O emissions from Forest Soils and Settlement Soils. tlands Coastal We 2 f LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest erted to Cropland, Grassland Remaining Land, Land Converted to Forest Land, Cropland Remaining Cropland, Land Conv Settlements Wetlands Remaining Wetlands, Land Converted to Wetlands, Grassland, Land Converted to Grassland, Land Converted to Settlements . Refer to Remaining Settlements - 5 for a breakout of emissions and removals for , and Table ES - Use Change, and Forestry by gas and source category. Land Use, Land g The LULUCF Sector Net Total is the net sum of all CH and N O emissions to the atmosphere plus net carbon stock changes. 4 2 Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or sequestration . Figure ES - illustrates the relative contribution of the direct greenhouse gases to total U.S. emissions in 201 5, 4 . Note, unless otherwise stated, all tables and figur weighted by global warming potential es provide total emissions without LULUCF. The primary greenhouse gas emitted by human activities in the United States was CO , 2 representing approximately percent of total greenhouse gas emissions. The largest source of CO , and of overall 82.2 2 greenhouse gas emissions, was fossil fuel combustion. Methane emissions, which have decreased by 16.0 percent since 1990, resulted primarily enteric fermentation associated with domestic livestock, natural gas systems , and from decomposition of wastes in landfills. A gricultural soil management, manure management, mobile source fuel combustion and stationary fuel combustion were the major sources of N Ozone depleting substance O emissions. 2 substitute emissions and emissions of HFC - 23 during the production of HCFC - 22 we re the primary contributors to Perfluorocarbon (PFC) emissions resulted as a byproduct of primary aggregate hydrofluorocarbon (HFC) emissions. aluminum production and from semiconductor manufacturing, electrical transmission and distribution systems ) emissions, and semiconductor manufacturing is the only source of ted for most sulfur hexafluoride (SF accoun 6 nitrogen trifluoride (NF ) emissions. 3 - : 201 5 U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO Figure ES 4 2 Eq.) 5 , total emissions of CO Eq. ( increased by 288.4 MMT CO percent), while total Overall, from 1990 to 201 5.6 2 2 MMT decreased by 125.1 MMT CO Eq. ( emissions of CH percent), and N 24.7 O emissions decreased by 16.0 2 2 4 CO Eq. ( 6.9 percent). During the same period, aggregate weighted emissions of HFCs, PFCs, SF and NF rose by 2 6 3 , HFCs increased by 85.0 Eq. ( 85.3 percent). From 1990 to 201 5 MMT CO 126.6 MMT CO percent), Eq. ( 271.8 2 2 PFCs decreased by 19.1 MMT CO Eq. ( 78.6 percent), SF 79.8 decreased by 23.0 M MT CO Eq. ( percent), and 2 6 2 Despite being emitted in smaller quantities relative to the NF percent). increased by 0.5 MMT CO Eq. ( 1,057.0 2 3 and NF are significant because many of these other principal greenhouse gases, emissions of HFCs, PFCs, SF 6 3 , long atmospheric lifetimes. gases have extremely high global warming potentials and, in the cases of PFCs and SF 6 Conversely, U.S. greenhouse gas emissions were partly offset by carbon (C) sequestration in forests, t rees in urban which, in aggregate, offset and coastal wetlands, areas, agricultural soils, landfilled yard trimmings and food scraps, - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 8 ES

33 11.8 percent of total emissions in 201 . The following sections describe each gas’s contribution to total U.S. 5 greenhouse gas emissions in more detail. Carbon Dioxide Emissions Billions of tons of carbon in the form of The global carbon cycle is made up of large carbon flows and reservoirs. CO are absorbed by oceans and living biomass (i.e., sinks) and are emitted to the atm osphere annually through 2 When in equilibrium, carbon fluxes among these various reservoirs are roughly natural processes (i.e., sources). 13 balanced. have risen Since the Industrial Revolution (i.e., about 1750), global atmospheric concentrations of CO 2 roximately app percent (IPCC 2013 and NOAA/ESRL 201 7 ), principally due to the combustion of fossil fuels. 44 93.3 percent of CO emissions in 201 Within the United States, fossil fuel combustion accounted for 5 . Globally, 2 approximately MMT of CO 33,733 were ad ded to the atmosphere through the combustion of fossil fuels in 201 4 , of 2 14 which the United States accounted for approximately percent. Changes in land use and forestry practices can 16 CO to a net emissions (e.g., through conversion of forest land to a gricultural or urban use) or lead to net sink for 2 CO (e.g., through net additions to forest biomass). F . emissions ossil fuel combustion is the greatest source of CO 2 2 T here are 2 4 additional sources included in the Inventory ( Figure ES - 5 ). - Figure ES 5 : 201 5 Sources of CO Emissions (MMT CO Eq.) 2 2 Note: than 0.05 MMT CO Eq. Fossil Fuel Combustion includes electricity generation, which also includes emissions of less 2 - based generation. from geothermal As the largest source of U.S. greenhouse gas emissions, CO from fossil fuel combustion has accounted for 2 77 percent of GWP - approximately emissions weighted emissions since 1990. The fundamental factors influencing 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 Between 1990 and 201 5 , CO emissions from fossil fuel the economy. 2 13 The term “flux” is used to describe the net emissions of greenhouse gases accounting for both the emissions of CO to and the 2 removals of CO from the atmosphere. Removal of CO from the atmosphe re is also referred to as “carbon sequestration.” 2 2 14 Global CO Emissions from Fossil emissions from fossil fuel combustion were taken from International Energy Agency CO 2 2 Fuels Combustion – Highlights IEA ( - ) . See < https://www.iea.org/publications/freep ublications/publication/co2 - emissions 2016 - . > 2016.html - highlights from - fuel - combustion 9 - ES Executive Summary

34 combustion increased from 4,740.3 Eq. to 5,049.8 MMT CO MMT CO Eq., a 6.5 percent total increase over the 2 2 - six - year period. In addition, CO Eq. emissions from fossil fuel combustion de creased by 697.2 MMT CO twenty 2 2 from 2005 levels, a decrease of approximately 12.1 percent between 2005 and 2015. From 201 4 to 201 5 , these emissions by 152.5 MMT CO decreased Eq. ( 2.9 percent). 2 Historically, changes in emissions from fossil fuel combustion have been the dominant factor affecting U.S. term and emission trends. Changes in CO - emissions from fossil fuel combustion are influenced by many long 2 - population and economic - term factors include fting shi , technological changes, trends short . Long term factors various policies at the national, state, and local level. In the short term, the overall energy fuel choices, and consumption and mix of fossil fuels in the United States fluctuates primarily in response to changes in general energy prices, economic conditions, ove rall the relative price of different fuels , weather, and the availability of non - fossil alternatives. CO Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT : 201 5 Figure ES - 6 2 CO Eq.) 2 from Fossil Fuel Combustion (MMT CO ES - Figure End - 7 : 201 5 Use Sector Emissions of CO 2 2 Eq.) sectors contributing to CO economic emissions from fossil fuel combustion are The five major fuel consuming 2 Carbon dioxide sportation, industrial, residential, and commercial. electricity generation, tran emissions are produced by the electricity generation sector as they consume fossil fuel to provide electricity to one of the other four sectors, - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 10 ES

35 - or “end For the discussion be low, electricity generation emissions have been distributed to each end - use” sectors. use sector on the basis of each sector’s share of aggregate electricity consumption. This method of distributing use sector consumes electricity that is - emissions assumes that each end generated from the national average mix of Emissions from electricity generation are also addressed separately after fuels according to their carbon intensity. the end - use sectors have been discussed. Note that emissions from U.S. Territories are calculat ed separately due to a lack of specific consumption data for the use sectors. Figure ES - 6 , individual end ES - 7 , and Table ES - 3 summarize CO - emissions from fossil fuel Figure 2 combustion by end - use sector. - 3 : CO Emissions from Fossil Fuel Combustion by End - Use Sector (MMT CO Table ES Eq.) 2 2 Sector 1990 2005 2011 20 1 End 2013 2014 2015 - Use 2 a 1,891.8 1,711.9 Transportation 1,700.6 1,717.0 1,746.9 1,740.1 1,496.8 1,493.8 1,887.0 1,707.6 1,696.8 1,713.0 1,742.8 1,736.4 Combustion 3.0 4.7 4.3 3.9 4.0 4.1 3.7 Electricity a 1,529.2 1,564.6 1,399.6 Industrial 1,375.7 1,407.0 1,399.3 1,355.0 828.0 775.0 782.9 812.2 806.1 805.5 Combustion 842.5 686.7 736.6 624.7 592.8 594.7 593.2 549.6 Electricity 931.4 1,214.1 1,116.2 1,007.8 1,064.6 1,080.1 1,003.9 Residential 282.5 357.8 325.5 Combustion 338.3 329.7 345.4 319.6 734.9 684.3 Electricity 593.0 856.3 790.7 725.3 734.7 a 934.7 958.4 897.0 925.5 1,026.8 909.4 755.4 Commercial 223.5 220.4 196.7 Combustion 228.7 246.2 217.4 221.0 803.3 738.0 700.3 704.5 706.0 663.1 Electricity 538.0 b 27.6 U.S. Territories 49.7 40.9 43.5 42.5 41.4 41.4 4,740.3 5,746.9 5,227.1 5,024.6 5,156.5 5,202.3 5,049.8 Total 1,820.8 2,400.9 2,157.7 2,022.2 2,038.1 2,038.0 1,900.7 Electricity Generation a road There was a method update in this Inventory for estimating the share of gasoline used in on - road and non - applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the transportation sector to industrial and commercial sectors for 2015, creating a break in the time series. The change is discussed further in the Planned Improvements section of 3.1 . Chapter b 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. related emissions from electricity generation are - Notes: Combustion allocated based on aggregate national electricity consumption by each end - use sector. Totals may not sum due to independent rounding. Use Sector. When electricity - related emissions are distributed to economic end - use sectors, Transportation End - 34 .5 percent of U.S. CO .The transportation activities accounted for emissions from fossil fuel combustion in 201 5 2 - emissions in 201 5 were passenger cars (4 2.3 percent), medium largest sources of transportation CO and heavy - 2 duty trucks ( 23.6 percent), light - duty trucks, which include sport utility vehicles, pickup trucks, and minivans (17. 1 2.5 percent), commercial aircraft (6. 8 percent), rail ( percent), and percent), other aircraft (2.3 per cent), pipelines ( 2.2 ships and boats (1. percent ). Annex 3.2 presents the total emissions from all transportation and mobile sources, 9 , CH including CO , N O, and HFCs. 2 2 4 5 , total transportation CO In terms of the overall trend, from 1990 to 201 emissions increased due, in large part, to 2 . The number of VMT by light increased demand for travel - duty motor vehicles (i.e., passenger cars and light - duty 15 trucks) increased percent from 1990 to 201 5 , as a result of a confluence of factors includin g population growth, 40 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 15 In 2011, FHWA VMT estimates are based on data from FHWA Highway Statistics Table VM - 1 (FHWA 1996 through 2016). road vehicle changed its methods for estimating VMT by vehicle class, wh ich led to a shift in VMT and emissions among on - - classes in the 2007 to 2015 time period. In absence of these method changes, light duty VMT growth between 1990 and 2015 would likely have been even higher. 11 - ES Executive Summary

36 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 Industrial CO emissions, resulting both directly from the Use Sector. combustion of fossil fuels and - 2 27 percent of CO from indirectly from the generation of electricity that is consumed by industry, accounted for 2 5 . Approximately 59 percent of these emissions resulted from direct fossil fuel fossil fuel combustion in 201 bustion to produce steam and/or heat for industrial processes. The remaining emissions resulted from consuming com In contrast to the other end - use electricity for motors, electric furnaces, ovens, lighting, and other applications. stry have declined since 1990. sectors, emissions from indu 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. - Use Sectors. The residential and commercial end - use sectors accounted for 20 and Residential and Commercial End percent, respectively, of CO 18 emissions from fossil fuel combustion in 201 5 . Both sectors relied heavily on 2 percent, respectively, of electricity for meeting energy demands, with and 73 68 their emissions attributable to electricity consumption for lighting, heating, cooling, and operating appliances. The remaining emissions were due Emissions from the residential and to the consumption of natural gas and petroleum for heating and cooking. percent and percent since 1990, respectively. commercial end - use sectors have increased by 8 20 Electricity Generation. The United States relies on electricity to meet a significant portion of its energy demands. 34 Electricity generators consumed t of total U.S. energy uses from fossil fuels and emitted 38 percent of the percen CO is the main factor from fossil fuel combustion in 201 5 . The type of energy source used to generate electricity 2 - emissions. For example, some electricity is generated through non influencing fossil fuel options such as nuclear, hydroelectric, or geothermal energy. Including all electricity generation modes, electricity generators relied on coal for approximately 34 percent of their total energy requirements in 201 5 . In addition, the coal used by electricity 16 generators accounted for percent of all coal consumed for energy in the United States in 201 5 . 93 Recently, a decrease in the carbon intensity of fuels consumed to generate electricity has occurred due to a decrea se in coal consumption, and increased natural gas consumption and other generation sources. Including all electricity generation modes, electricity generators used natural gas for approximately 32 percent of their total energy requirements in 201 5 . Across the time series, changes in electricity demand and the carbon intensity of fuels used for electricity generation have a significant impact on CO emissions. While emissions from the electric power 2 sector have increased by approximately 4 percent since 1990 , the carbon intensity of the electric power sector, in terms of CO E q. per QBtu has significantly decreased by 16 percent during that same timeframe. This trend away 2 s shown below in 8 . - Figure ES from a direct relationship between electricity generation and the resulting emissions i 16 See Table 6.2 Coal Consumption by Sector of EIA 2016. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 12 ES

37 Figure ES 8 Electricity Generation (Billion kWh) and Electricity Generation Emissions (MMT - : Eq.) CO 2 Other significant CO trends included the following: 2 Carbon dioxide emissions from non - energy use of fossil fuels increased by 7.9 MMT CO Eq. ( • 6.8 percent) 2 from 1990 through 201 Emissions from non - energy uses of fossil fuels were 125.5 . MMT CO 5 Eq. in 2 5 , which constituted 2.3 percent of total national CO 201 emissions, approximately the same proportion as 2 in 1990. Carbon dioxide emissions from iron and steel production and metallurgical coke production have decreased • percent) from 1990 through 201 52.6 MMT CO , due to restructuring of the industry, Eq. ( 51.8 5 by 2 technological improvements, and increased scrap steel utilization. 10.8 MMT CO Carbon dioxide emissions from ammonia production ( Eq. in 201 5 ) decreased by 2.2 • MMT 2 CO Eq. ( 17.2 Ammonia production relies on natural gas as both a feedstock and a t) since 1990. percen 2 fuel, and as such, market fluctuations and volatility in natural gas prices affect the production of ammonia. • Total C stock change (i.e., net CO removals) in the LULUCF sect or decreased by approximately 6.2 2 percent between 1990 and 2015. This decrease was primarily due to a decrease in the rate of net C accumulation in forest C stocks and an increase in emissions from Land Converted to Settlements . Box ES - 3 : Use of Ambient Measurements Systems for Validation of Emission Inventories In following the UNFCCC to develop and submit national greenhouse gas emission requirement under Article 4.1 inventories, the emissions and sinks presented in this report are organized by source and sink categories and 17 calculated using internationally - accepted methods provided by the IPCC. Several recent studies have measured 17 http://www.ipcc See < nggip.iges.or.jp/public/index.html>. - 13 - ES Executive Summary

38 emissions at the national or regional level with results that sometimes of emissions. EPA differ from EPA’s estimate 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 commun ity on ambient measurement and remote sensing techniques to improve national greenhouse gas follows guidance from the IPCC on the use of measurements and modeling to validate emission inventories, EPA 18 inventories. treach efforts is how ambient measurement data can be used An area of particular interest in EPA’s ou 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 ant hropogenic 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 Harvar d 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 nventory is de signed to be consistent with the 1990 to 2014 I U.S. EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks estimates for the year 2012, which presents 19 national totals for different source types. Methane Emissions Methane (CH Over the last two ) is 25 times as effective as CO at trapping heat in the atmosphere (IPCC 2007). 2 4 162 percent (IPCC 2013 and in the atmosphere increased by hundred and fifty years, the concentration of CH 4 ). CDIAC 201 6 include natural gas and petroleum systems, agricultural activities, Anthropogenic sources of CH 4 LULUCF, landfills, coal mining, wastewater treatment, stationary and mobile combustion, and certain industrial Figure ES - 9 ). processes (see Figure ES - : 201 5 Sources of CH 9 Emissions (MMT CO Eq.) 2 4 LULUCF emissions are reported separately from gross emissions totals and are not included in Figure ES - 9 . Note: Refer to Table ES - 5 for a breakout of LULUCF emissions by gas . 18 See . 19 ed emissions - methane - 2012 - >. See < https://www.epa.gov/ghgemissions/gridd - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 14 ES

39 Some si include the following: gnificant trends in U.S. emissions of CH 4 Enteric fermentation is the largest anthropogenic source of CH emissions in the United States. In 201 5 , • 4 enteric fermentation CH emissions were 166.5 MMT CO emissions), Eq. ( 25.4 percent of total CH 4 2 4 2.4 MMT CO percent) since 1990. Eq. ( 1.5 which represents an increase of This increase in emissions 2 from 1990 to 201 generally follows the increasing trends in cattle populations. From 1990 to 1995 , 5 reased from 1996 to 2004, mainly due to fluctuations in beef emissions increased and then generally dec cattle populations and increased digestibility of feed for feedlot cattle. Emissions increased from 2005 to d. Research indicates that the feed di gestibility of dairy 2007, as both dairy and beef populations increase d during this period . Emissions decreased again from 2008 to 201 5 as beef cattle cow diets decrease populations again decreased. • Natural gas systems were the second largest anthropogenic source category of CH emissions in the United 4 St ates in 201 with 162.4 MMT CO 5 Eq. of CH emitted into the atmosphere. Those emissions have 4 2 31.6 MMT CO emissions is largely due Eq. ( 16.3 percent) since 1990. The decrease in CH decreased by 4 2 distribution. The decrease in transmission and to the decrease in emissions from transmission, storage, 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 lastic piping, which has lower emissions than other pipe materials, and station upgrades at metering and p regulating (M&R) stations. • Landfills are the third largest anthropogenic source of CH MMT emissions in the United States ( 115.7 4 5 emissions Eq.), accounting for 17.6 percent of total CH emissions in 201 5 . From 1990 to 201 CO , CH 4 4 2 63.8 MMT CO from landfills decreased by Eq. ( 35.6 percent), with small increases occurring in some 2 increased interim years. This downward trend in emissions coincided with landfill gas collection and control systems, and a reduction of decomposable materials (i.e., paper and paperboard, food scraps, and 20 , yard trimmings) discarded in MSW landfills over the time series which has more than offset the emissions from an increase in the amount of municipal solid waste dditional CH a that would have resulted 4 landfilled. Methane emissions from manure management , the fourth largest anthropogenic source of CH • emissions in 4 increased by 78.3 percent since 1990, from 37.2 the United States, Eq. in 1990 to 66.3 MMT MMT CO 2 Eq. in 201 5 . CO The majority of this increase was from swine and dairy cow manure, since the general 2 trend in manure management is one of increasing use of liquid systems, which tends to produc e greater The increase in liquid systems is the combined result of a shift to larger facilities, and to CH emissions. 4 Also, new regulations limiting facilities in the West and Southwest, all of which tend to use liquid systems. utrients have shifted manure management practices at smaller dairies from daily the application of manure n spread to manure managed and stored on site. Methane emissions from p etroleum systems in the United States ( 39.9 MMT CO • Eq.) account ed for 6.1 2 emissions in 201 5 . percent of total CH From 1990 to 201 5 , CH emissions from petroleum systems 4 4 decreased 15.6 MMT CO emissions have decreased by Eq. (or 28.1 percent). Production segment CH by 2 4 associated gas venting and around 8 percent from 2014 levels, primarily due to decreases in emissions from flaring. Nitrous Oxide Emissions Nitrous oxide (N O) is produced by biological processes that occur in soil and water and by a variety of 2 anthropogenic activities in the agricultural, energy - related, industrial, and waste manage ment fields. While total N O 2 emissions are much lower than CO emissions, N at trapping heat in O is 300 times more powerful than CO nearly 2 2 2 the atmosphere (IPCC 2007). Since 1750, the global atmospheric concentration of N O has risen by approximately 2 The main anthropogenic activities producing N p ercent (IPCC 2013 and CDIAC 201 6 ). 21 O in the United States 2 20 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 . LULUCF or are accounted for in the estimates f 15 - ES Executive Summary

40 are agricultural soil management, stationary fuel combustion, fuel combustion in motor vehicles, manure management, and nitric acid production (see - 10 ) . Figure ES O Emissions (MMT CO 10 5 Sources of N Figure ES - Eq.) : 201 2 2 Figure ES - 10 . LULUCF emissions are reported separately from gross emissions totals and are not included in Note: Table ES - 5 for a breakout of LULUCF emissions by gas . Refer to O include the following: Some significant trends in U.S. emissions of N 2 Agricultural soils accounted for approximately 75.1 • percent of N percent of total O emissions and 3.8 2 emissions in the United States in 201 . Estimated emissions from this source in 201 5 5 were 251.3 MMT CO , although overall Eq. Annual N 5 O emissions from agricultural soils fluctuated between 1990 and 201 2 2 e missions were 2.0 percent lower year fluctuations are largely a reflection of 5 than in 1990. Year - to - in 201 annual variation in weather patterns, synthetic fertilizer use, and crop production. 94.0 percent) from • Nitrous oxide emissions from stationary combustion increased 1 1.2 MMT CO Eq. ( 2 1990 through 201 . Nitrous oxide emissions from this source increased primarily as a result of an increase 5 in the number of coal fluidized bed boilers in the electric power sector. In 201 5 , total N Eq.; O emissions from manu re management were estimated to be 17.7 MMT CO • 2 2 14.0 MMT CO O emissions Eq. in 1990. These values include both direct and indirect N emissions were 2 2 Nitrous oxide emissions have remained fairly steady since 1990. from manure management. Small changes O emissions from individual animal groups exhibit the same trends as the animal group populations, in N 2 1.1 O emissions showed a 26.6 percent increase from 1990 to 201 5 and a with the overall net effect that N 2 percent from 201 4 through 201 5 . increase • by 26.1 MMT CO percent) from Nitrous oxide emissions from mobile combustion decreased Eq. ( 63.3 2 1990 through 201 5 , primarily as a result of N O national emission control standards and emission control 2 - road vehicles. technologies for on • Nitrous oxide emiss ions from adipic acid production were 4.3 MMT CO , and have decreased Eq. in 201 5 2 significantly since 1990 due to both the widespread installation of pollution control measures in the late d production have decreased by 72.0 Emissions from adipic aci 1990s and plant idling in the late 2000s. percent since a peak in 1995. 74.8 percent since 1990 and by - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 16 ES

41 HFC, PFC, SF Emissions , and NF 6 3 Hydrofluorocarbons (HFCs) and perfluorocarbons (PFCs) are families of synthetic chemicals that are used as ne depleting substances (ODS), which are being phased out under the Montreal Protocol and alternatives to ozo Hydrofluorocarbons and PFCs do not deplete the stratospheric ozone layer, Clean Air Act Amendments of 1990. al Protocol on Substances that Deplete the Ozone Layer . and are therefore acceptable alternatives under the Montre and NF These compounds, however, along with SF , are potent greenhouse gases. In addition to having high global 3 6 and PFCs have extremely long atmospheric lifetimes, resulting i n their essentially warming potentials, SF 6 irreversible accumulation in the atmosphere once emitted. Sulfur hexafluoride is the most potent greenhouse gas the IPCC has evaluated (IPCC 2013). - 22 production, electrical transmissio n and distribution systems, Other emissive sources of these gases include HCFC semiconductor manufacturing, aluminum production, and magnesium production and processing (see Figure ES - 11 ). - 11 : 201 5 Sources of HFCs, PFCs, SF Figure ES , and NF Eq.) Emissions (MMT CO 2 6 3 Some significant trends in U.S. HFC, PFC, SF , and NF emissions include the following: 3 6 • Hydrofluorocarbon and perfluorocarbon emissions resulting from the substitution of ODS (e.g., chlorofluorocarbons [CFCs]) have been consistently increasing, from small amounts in 1990 to 168.5 Eq. in 201 5 . This increase was in large part the result of efforts to phase out CFCs and other MMT CO 2 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, Copenhagen A are themselves phased out under the provisions of the . mendments to the Montreal Protocol have increased by GWP - weighted PFC, HFC, SF , and NF • emissions from semiconductor manufactur ing 6 3 34.3 percent from 1990 to 201 5 , due to industrial growth and the adoption of emission reduction technologies. Within that time span, emissions peaked in 1999, the initial year of EPA’s PFC Reduction/Climate Partnership for the Semiconductor Industry, but have since declined to 4.8 MMT CO 2 percent decrease relative to 1999). 47.1 (a Eq. in 201 5 Sulfur hexafluoride emissions from electric power transmission and distribution systems decreased by 82.0 • 19.0 . MMT CO percent ( Eq.) from 1990 to 201 5 There are two potential causes for this decrease: (1) a 2 during the 19 sharp increase in the price of SF 90s and (2) a growing awareness of the environmental 6 impact of SF emissions through programs such as EPA’s SF Emission Reduction Partnership for Electric 6 6 Power Systems. Eq.) MM T CO • Perfluorocarbon emissions from aluminum production decreased by 90.7 percent ( 19.5 2 . from 1990 to 201 5 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. 17 - ES Executive Summary

42 ES.3 Overview of Sector Emissions and Trends accordance with the UNFCCC decision to set the 2006 IPCC Guidelines for National Greenhouse Gas In (IPCC 2006) as the standard for Annex I countries at the Nineteenth Conference of the Parties Inventories - 12 and Table ES - 4 aggregate emissions and sinks by the sectors defined by t (UNFCCC 2014), Figure ES hose Over the twenty - six - year period of 1990 to 201 5 , total emissions in the Energy, Industrial Processes and guidelines. 221.0 MMT CO percent), and Eq. ( 4.1 Product Use, and Agriculture grew by percent), 35.5 MMT CO Eq. ( 10.4 2 2 MMT CO Eq. ( 5.5 percent), respectively. 27.0 Over the same period, total emissions in the Waste sector decreased 2 by MMT CO percent) and estimates of net C sequestration in the LULUCF sector (magnitude of 59.9 Eq. ( 30.1 2 emissions plus CO removals from all LULUCF source categ ories) decreased by 60.7 MMT CO percent). Eq. ( 7.4 2 2 - 12 Figure ES : U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO 2 Eq.) 4 : Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Table ES - Sector (MMT CO Eq.) 2 Chapter/IPCC Sector 1990 2005 2011 2012 2013 2014 2015 Energy 5,328.1 6,275.3 5,721.2 5,507.0 5,659.1 5,704.9 5,549.1 5,746.9 Fossil Fuel Combustion 4,740.3 5,024.6 5,227.1 5,156.5 5,202.3 5,049.8 Natural Gas Systems 189.8 190.2 191.4 197.7 204.9 204.8 231.8 - 117.6 138.9 109.8 Non 106.7 123.6 119.0 125.5 Energy Use of Fuels 66.5 64.1 71.2 96.5 64.6 64.8 60.9 Coal Mining Petroleum Systems 59.0 49.9 52.2 50.3 48.2 46.6 43.4 Stationary Combustion 27.6 28.4 28.0 20.4 30.9 31.5 30.1 a Mobile Combustion 46.9 38.6 25.1 22.6 20.6 18.6 17.1 10.7 Incineration of Waste 12.9 10.9 8.4 10.7 10.9 11.0 7.2 6.6 6.4 6.2 6.2 6.3 6.4 Abandoned Underground Coal Mines 360.9 375.9 379.8 363.7 Industrial Processes and Product Use 340.4 353.4 371.0 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 18 ES

43 Substitution of Ozone Depleting Substances 99.8 145.4 150.2 154.7 161.3 168.5 0.3 Iron and Steel Production & 68.1 61.1 55.5 53.4 58.6 48.9 Metallurgical Coke Production 101.5 33.5 46.2 32.2 35.3 36.4 39.4 39.9 Cement Production 21.5 27.0 26.4 26.6 26.5 26.6 28.2 Petrochemical Production 14.6 14.0 13.8 14.0 14.2 13.3 Lime Production 11.7 12.1 11.3 10.9 10.5 10.7 10.9 11.6 Nitric Acid Production Other Process Uses of Carbonates 6.3 9.3 8.0 10.4 11.8 11.2 4.9 9.2 9.3 9.4 10.0 9.6 10.8 Ammonia Production 13.0 3.6 4.7 4.9 4.5 4.1 5.0 5.0 Semiconductor Manufacture 7.6 6.8 6.4 6.2 5.4 4.8 28.3 Aluminum Production 1.5 1.4 4.1 4.0 4.2 Carbon Dioxide Consumption 4.5 4.3 HCFC - 22 Production 46.1 20.0 8.8 5.5 4.1 5.0 4.3 Adipic Acid Production 7.1 10.2 5.5 3.9 5.4 4.3 15.2 4.2 O from Product Uses 4.2 4.2 4.2 4.2 4.2 4.2 N ₂ Electrical Transmission and 23.1 8.3 6.0 Distribution 4.8 4.6 4.8 4.2 Soda Ash Production and 2.8 3.0 2.7 Consumption 2.8 2.8 2.8 2.8 Ferroalloy Production 2.2 1.4 1.7 1.9 1.8 1.9 2.0 Titanium Dioxide Production 1.8 1.7 1.5 1.7 1.7 1.6 1.2 1.9 1.3 1.2 Glass Production 1.3 1.3 1.3 1.5 Urea Consumption for Non - Agricultural Purposes 3.8 3.7 4.0 4.4 4.0 1.4 1.1 Magnesium Production and 5.2 2.7 Processing 1.7 1.5 1.1 1.0 2.8 Phosphoric Acid Production 1.5 1.3 1.2 1.1 1.1 1.0 1.0 1.0 0.9 Zinc Production 0.6 1.0 1.3 1.5 1.4 Lead Production 0.5 0.6 0.5 0.5 0.5 0.5 0.5 Silicon Carbide Production and 0.2 Consumption 0.2 0.2 0.4 0.2 0.2 0.2 495.3 526.4 541.9 525.9 516.9 514.7 522.3 Agriculture Agricultural Soil Management 259.8 270.1 254.1 250.5 250.0 251.3 256.6 166.7 Enteric Fermentation 168.9 168.9 164.2 165.5 164.2 166.5 Manure Management 51.1 72.9 80.4 83.2 80.8 80.4 84.0 Rice Cultivation 16.7 14.1 11.3 11.3 11.4 11.2 16.0 4.3 Urea Fertilization 3.5 4.1 2.4 4.5 4.8 5.0 Liming 4.7 4.3 3.9 6.0 3.9 3.6 3.8 Field Burning of Agricultural 0.4 0.3 0.3 0.4 Residues 0.4 0.4 0.4 144.4 199.3 139.4 158.2 Waste 142.6 140.4 140.2 Landfills 134.3 119.0 120.8 116.7 116.6 115.7 179.6 20.4 20.1 19.9 19.8 19.7 19.7 Wastewater Treatment 19.1 0.7 3.5 3.5 3.7 Composting 4.0 4.0 3.9 b 6,363.1 7,313.3 6,776.7 Total Emissions 6,680.1 6,739.7 6,586.7 6,538.3 Land Use, Land - Use Change, and Forestry (819.6) (731.0) (749.2) (753.8) (763.0) (761.4) (758.9) Forest Land (729.8) (733.8) (723.6) (733.5) (731.8) (728.7) (784.3) 1.3 Cropland (0.7) 4.0 2.4 3.1 4.0 4.7 Grassland 13.8 25.3 9.9 0.8 0.4 0.9 0.4 (4.0) Wetlands (5.2) (3.9) (3.9) (4.0) (4.0) (4.1) Settlements (47.6) (20.5) (25.4) (28.3) (28.9) (30.4) (31.3) c 5,978.3 Net Emissions (Sources and Sinks) 5,917.1 5,543.5 6,582.3 6,027.6 5,784.5 5,827.7 Net emissions presented with LULUCF. Notes: Total emissions presented without LULUCF. a There was a method update in this Inventory for estimating the share of gasoline used in on - road and non - road applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the - ES 19 Executive Summary

44 transportation sector to industrial and commercial se ctors for 2015, creating a break in the time series. The change is Chapter . the Planned Improvements section of 3.1 discussed further in b Total emissions without LULUCF. c Total emissions with LULUCF. Totals may not sum due to independent rounding. Parentheses indicate negative values or sequestration. Notes: Energy The Energy chapter contains emissions of all greenhouse gases resulting from stationary and mobile energy , and the use of fossil fuels for non - energy purposes . activities including fuel combustion and fugitive fuel emissions Energy - related activities, primarily fossil fuel combustion, accounted for the vast majority of U.S. C O emissions for 2 the period of 1990 through 201 . In 201 5 , approximately 82 percent of the energy consumed in the United States (on 5 a Btu basis) was produced through the combustion of fossil fuels. The remaining 18 percent came from other energy sources such as hydropower, biomass, nuclear, wind, and solar energy (see F igure ES - 13 ). Energy - related activities 12 are also responsible for CH and N percent and O emissions ( 42 percent of total U.S. emissions of each gas, 2 4 84.2 Overall, emission sources in the Energy chapter account for a combined respectively). percent of total U.S. greenhouse gas emissions in 201 5 . igure ES 13 F 5 - U.S. Energy Consumption by Energy Source (Percent) : 201 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. Greenhouse gas emissions are produced as the by - products of many non For energy - rel ated industrial activities. - , example, industrial processes can chemically transform raw materials, which often release waste gases such as CO 2 CH These processes include iron and steel production and metallurgical coke production, cement O. , and N 4 2 pro duction, ammonia production, urea consumption, lime production, other process uses of carbonates (e.g., flux stone, flue gas desulfurization, and glass manufacturing), soda ash production and consumption, titanium dioxide consumption, silicon carbide production and production, phosphoric acid produc tion, ferroalloy production, CO 2 consumption, aluminum production, petrochemical production, nitric acid production, adipic acid production, lead production, zinc production, and N and NF Industrial processes also release HFCs, PFCs, SF O from product uses. 2 6, 3 as ODS substitutes, HFCs, PFCs, of HFCs and some PFCs and other fluorinated compounds . In addition to the use , NF SF l sources in the , and other fluorinated compounds are employed and emitted by a number of other industria 6 3 22 production, semiconductor manufacture, - These industries include aluminum production, HCFC United States. electric power transmission and distribution, and magnesium metal production and processing. Overall, emission - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 20 ES

45 sources in the Indus trial Process and Product Use chapter account for percent of U.S. greenhouse gas emissions 5.7 . in 201 5 Agriculture The Agriculture chapter contains anthropogenic emissions from agricultural activities (except fuel combustion, which is addressed in the Ene rgy chapter, and some fluxes, which are addressed in the Land Use, agricultural CO 2 Land Agricultural activities contribute directly to emissions of greenhouse gases - Use Change, and Forestry chapter). 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 nsible for emissions of 522.3 MMT CO2 burning of agricultural residues. In 2015, agricultural activities were respo Eq., or 7.9 percent of total U.S. greenhouse gas emissions. Methane, nitrous oxide and carbon dioxide were the primary greenhouse gases emitted by agricultural activities. Methane emissions from enteric fermentation an d manure management represented approximately 25.4 percent and 10.1 percent of total CH4 emissions from anthropogenic activities, respectively, in 2015. Agricultural soil management activities, such as application of sition of livestock manure, and growing N fixing plants, were the largest - synthetic and organic fertilizers, depo source of U.S. N2O emissions in 2015, accounting for 75.1 percent. Carbon dioxide emissions from the application of crushed limestone and dolomite (i.e., soil liming) and urea fertil ization represented 0.2 percent of total CO2 Figure 2 - 11 and Table 2 - 7 illustrate agricultural greenhouse gas emissions emissions from anthropogenic activities. by source. Land Use, Land - Use Change, and Forestry - O, and emissions and Use Change, and Forestry chapter contains emissions of CH The Land Use, Land and N 2 4 (C from managed lands in the United States . Overall, managed land is a net sink for CO removals of CO 2 2 s . The primary drivers of fluxes on managed lands include, for example, forest sequestration) in the United State 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 s tocks 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 oc curs predomina n tly from and landfilled yard trimmings and food scraps, is a result of net tree growth and increased urban urban forests forest size, as well as long - term accumulation of yard trimmings and food scraps carbon in landfills. The LULUCF sector in 201 5 resulted in a net increase in C stocks (i.e., net CO Eq. removals) of 778.7 MMT CO 2 2 21 ( Table ES - 5 ). This represents an offset of 11.8 percent of total (i.e., gross) greenhouse gas emissions in 201 5 . 5 Emissions of CH percent of and N O from LULUCF 0.3 activities in 201 represent are 19.7 MMT CO Eq. and 2 2 4 22 Between 1990 and 201 5 , total C sequestration in the LULUCF sector decreased by total greenhouse gas emissions. 6.2 a decrease in the rate of net C accumulation in forest s and an increase in CO emissions percent, primarily due to 2 from Land Converted to Settlements . O emissions for Carbon dioxide removals are p resented in Table ES - 5 along with CH from C stock changes and N 2 4 LULUCF source categories. Forest fires were the largest source of CH totaling F in 201 5, emissions from LULUC 4 emissions of 3.6 resulted in CH 7.3 MMT CO Eq. ( 292 kt of CH Coastal Wetlands Remaining Coastal Wetlands ) . 4 2 4 MMT CO ). Eq. (143 kt of CH Eq. (16 kt of CH ). Grassland fires resulted in CH emissions of 0.4 MMT CO 2 4 4 2 4 Peatlands Remaining Peatlands , emissions and Drained Organic Soils resulted in CH Land Converted to Wetlands, 4 . of less than 0.05 MMT CO each Eq 2 21 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 22 and N LULUCF emissions include the CH , O emissions reported for Peatlands Remaining Peatlands Forest Fires, Drained 2 4 emissions from Land Converted to Organic Soils, Grassland Fires, and Coastal Wetlands Remaining Coastal Wetlands ; CH 4 Coastal Wetlands O emissions from Forest Soils and Settlement Soils. ; and N 2 21 - ES Executive Summary

46 Forest fires were also the largest source of N MMT CO Eq. ( 16 kt O emissions from LULUCF in 2015, totaling 4.8 2 2 O oxide emissions from fertilizer application to settlement soils in 201 5 totaled to Nitrous 2.5 MMT CO Eq. of N ). 2 2 8 O ). This represents an increase of 76.6 of N percent since 1990. Additionally, t he application of synthetic ( kt 2 resulted in N 0.5 O emissions of fertilizers to forest soils in 201 MMT CO Nitrous oxide Eq. ( 2 kt of N 5 O). 2 2 2 455 percent since 1990, but still account for a emissions from fertilizer application to forest soils have increased by s resulted in N Eq. (1 kt O emissions of 0.4 MMT CO relatively small portion of overall emissions. Grassland fire 2 2 O). Coastal Wetlands Remaining Coastal Wetlands and of N resulted in N O emissions of Drained Organic Soils 2 2 O emissions of Eq. each (less than 0.5 kt of N N O), and 0.1 MMT CO resulted in Peatlands Remaining Peatlands 2 2 2 0.05 less than Eq. MMT CO 2 - : U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land - Table ES 5 Use Change, and Forestry (MMT CO Eq.) 2 - Use Category 1990 Gas/Land 2005 2011 2012 2013 2014 2015 a Carbon Stock Change (754.0) (769.1) (830.2) (779.8) (782.2) (781.1) (778.7) Forest Land Remaining Forest Land (697.7) (664.6) (670.0) (666.9) (670.8) (669.3) (666.2) Land Converted to Forest Land (81.4) (75.8) (75.2) (75.2) (75.2) (75.2) (92.0) (26.5) (19.1) (21.4) (19.6) (18.7) (18.0) Cropland Remaining Cropland (40.9) 43.3 25.9 23.2 Land Converted to Cropland 22.7 22.7 22.7 22.7 Grassland Remaining Grassland (4.2) 5.5 (12.5) (20.5) (20.4) (20.9) (20.8) Converted to Grassland 20.7 17.9 19.2 Land 20.4 20.5 20.5 20.5 Wetlands Remaining Wetlands (8.9) (7.6) (7.7) (7.8) (7.8) (7.8) (7.6) + Land Converted to Wetlands + + + + + + Settlements Remaining Settlements (86.2) (91.4) (98.7) (99.2) (99.8) (101.2) (102.1) Land Converted to Settlements 37.2 68.4 70.7 68.3 68.3 68.3 68.3 6.7 11.0 13.3 CH 11.2 14.9 11.3 11.3 4 Forest Land Remaining Forest Land: 7.2 9.4 6.8 10.8 7.3 7.3 Forest Fires 3.2 Wetlands: Coastal Wetlands Remaining 3.4 3.5 3.5 3.5 3.6 3.6 3.6 Wetlands Remaining Coastal Wetlands Grassland Remaining Grassland: 0.3 0.8 0.6 0.2 0.4 0.4 Grass Fires 0.1 Forest Land Remaining Forest Land: + + + + + + + Drained Organic Soils Land Converted to Wetlands: Land Converted to Coastal Wetlands + + + + + + + Wetlands Remaining Wetlands: Peatlands + + + + + + + Remaining Peatlands O 3.9 N 9.7 8.7 11.1 8.2 8.4 8.4 2 Forest Land Remaining Forest Land: Fires 4.5 2.1 6.2 Forest 7.1 4.7 4.8 4.8 Settlements Remaining Settlements: b 2.6 1.4 2.5 2.6 2.7 Settlement Soils 2.5 2.5 Forest Land Remaining Forest Land: c 0.1 0.5 0.5 0.5 0.5 0.5 0.5 Forest Soils Grassland Remaining Grassland: 0.1 0.3 0.9 0.2 0.4 0.4 Grass Fires 0.6 Wetlands Remaining Wetlands: Coastal 0.2 0.1 Wetlands Remaining Coastal Wetlands 0.1 0.1 0.1 0.1 0.1 Forest Land Remaining Forest Land: 0.1 0.1 0.1 0.1 Drained Organic Soils 0.1 0.1 0.1 Wetlands Remaining Wetlands: Peatlands Remaining Peatlands + + + + + + + d LULUCF Emissions 10.6 23.0 19.9 19.2 19.7 19.7 26.1 a Carbon Stock Change (830.2) (754.0) (769.1) (779.8) (782.2) (781.1) LULUCF (778.7) e (819.6) Total (731.0) LULUCF Sector (749.2) (753.8) (763.0) Net (761.4) (758.9) + Absolute value does not exceed 0.05 MMT CO Eq. 2 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 Settlements Wetlands Remaining Wetlands, Land Converted to Wetlands, Grassland, Land Converted to Grassland, ining Settlements, Rema . Land Converted to Settlements and - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 22 ES

47 b Settlements Remaining Settlements and Land Converted to Estimates include emissions from N fertilizer additions on both Settlements. c F Land Converted to and orest Land Remaining Forest Land Estimates include emissions from N fertilizer additions on both . Forest Land d LULUCF emissions include the CH , Forest Fires, Drained Peatlands Remaining Peatlands O emissions reported for and N 2 4 Coastal Wetlands Remaining Coastal Wetlands Organic Soils, Grassland Fires, and emissions from Land Converted to ; CH 4 Coastal Wetlands ; and N O emissions from Forest Soils and Settlement Soils. 2 e The LULUCF Sector Net Total is the net sum of all CH and N O emissions to the atmosphere plus net carbon stock 2 4 changes. tes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration. No Waste The Waste chapter contains emissions from waste management activities (except incineration of waste, which is Landfills w ere the largest source of anthropogenic greenhouse gas emissions in the addressed in the Energy chapter). Waste chapter, accounting for 83.0 percent of this chapter’s emissions, and 17.6 percent of total U.S. CH 4 23 emissions. Additionally, wastewater treatment accounts for 14.2 percent of W aste emissions, 2.3 percent of U.S. CH 1.5 percent of U.S. N O emissions. emissions, and Emissions of CH and N O from composting are also 4 2 4 2 2.1 Eq. and MMT CO accounted for in this chapter, generating emissions of 1.9 MMT CO Eq., respectively. 2 2 2.1 percent of total U.S. greenhouse gas Overall, emission sources accounted for in the Waste chapter generated 5 emissions in 201 . ES.4 Other Information Emissions by Economic Sector Throughout the Inventory of U.S. Greenhouse Gas Emissions and Sinks repor t, emission estimates are grouped into five sectors (i.e., chapters) defined by the IPCC: Energy; Industrial Processes and Product Use; Agriculture; LULUCF; and Waste. While it is important to use this characterization for consistency with UNFCCC reporting characterize , it is also useful to guidelines emissions according to and to promote comparability across countries economic sector categories: residential, commercial, industry, transportation, electricity generation, commonly used agriculture, and U.S. T erritories. shows the trend in Table ES 6 summarizes emissions from each of these economic sectors, and Figure ES - 14 - emissions by sector from 1990 to 201 5 . 23 incomplete degradation of organic materials such as harvest wood products, yard Landfills also store carbon, due to - trimmings, and food scraps, as described in the Land Use Change, and Forestry chapter of the Inventory report. - Use, Land 23 - ES Executive Summary

48 Figure ES - Eq.) 14 : U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO 2 Eq.) Table ES - : U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO 6 2 1990 2005 2011 2012 2013 2014 2015 Economic Sectors 2,078.2 1,941.4 2,079.7 Electric Power Industry 1,862.5 2,441.6 2,197.3 2,059.9 1,864.8 a Transportation 1,551.2 2,001.0 1,806.6 1,800.0 1,780.7 1,790.2 1,815.8 1,551.3 a 1,411.6 Industry 1,626.3 1,467.1 1,378.6 1,365.9 1,413.4 1,418.0 1,620.9 577.6 574.3 Agriculture 526.7 567.5 566.1 570.3 592.0 a Commercial 418.1 400.7 406.5 387.3 410.1 419.5 437.4 563.4 318.4 393.9 372.6 356.3 372.7 Residential 344.9 370.4 418.1 U.S. Territories 33.3 58.1 46.6 46.0 48.5 48.1 46.6 344.9 6,776.7 Total Emissions 6,363.1 7,313.3 6,586.7 6,538.3 6,680.1 6,739.7 33.7 b Net Total (819.6 ) (731.0 ) LULUCF Sector (749.2) (753.8) (763.0) (761.4) (758.9) 5,978.3 5,827.7 Net Emissions (Sources and Sinks) 5,543.5 6,582.3 6,027.6 5,784.5 5,917.1 s : Total emissions presented without LULUCF. Total net emissions presented with LULUCF. Note a There was a method update in this Inventory for estimating the share of gasoline used in on - road and non - road applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasol ine consumption from the transportation sector to industrial and commercial sectors for 2015, creating a break in the time series. The change is discussed further in the Planned Improvements section of Chapter 3.1 . b The LULUCF Sector Net Total is the net sum of all CH and N O emissions to the atmosphere plus net carbon stock 4 2 changes. negative values or sequestration. Notes: Totals may not sum due to independent rounding. Parentheses indicate Using this categorization, emissions from electricity generation accounted for the largest portion ( percent) of 29 total U.S. greenhouse gas emissions in 201 5 . Transportation activities, in aggregate, accounted for the second largest portion ( 27 percent), while emissions from industry accounted for the third largest portion ( 21 total percent) of U.S. due to greenhouse gas emissions in 201 . E missions from industry have in general declined over the past decade , 5 a number o f factors, including structural changes in the U.S. economy (i.e., shifts from a manufacturing - based to a percent of U.S. 22 service - based economy), fuel switching, and energy efficiency improvements. The remaining by, in order of magnitude, the agriculture, commercial, and residential greenhouse gas emissions were contributed 9 percent of U.S. sectors, plus emissions from U.S. Territories. Activities related to agriculture accounted for s were dominated by N emissions; unlike other economic sectors, agricultural sector emission O emissions from 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 24 ES

49 agricultural soil management and CH emissions from enteric fermentation. The commercial and residential sectors 4 7 6 percent of emissions, respectively, and U.S. Territories accounted for 1 p ercent of accounted for percent and was emissions; emissions from these sectors primarily consisted of CO emissions from fossil fuel combustion. CO 2 2 also emitted and sequestered by a variety of activities related to forest management practices, tree planting in urban , and changes in C stocks in coastal areas, the ma nagement of agricultural soils, landfilling of yard trimmings . wetlands Table ES - 7 Electricity is ultimately consumed in the economic sectors described above. presents greenhouse gas - use categories emissions from economic sectors with emissions related to electricity generation distributed into end (i.e., emissions from electricity generation are a llocated to the economic sectors in which the electricity is To distribute electricity emissions among end consumed). use sectors, emissions from the source categories assigned - to electricity generation were allocated to the residential, commercial, indust ry, transportation, and agriculture 24 economic sectors according to retail sales of electricity. from fossil fuel These source categories include CO 2 combustion and the use of limestone and dolomite for flue gas desulfurization, CO and N O from incineratio n of 2 2 and N from electrical transmission and distribution systems. waste, CH O from stationary sources, and SF 6 4 2 When emissions from electricity are distributed among these sectors, industrial activities and transportation account 29 for the largest shares of 27 U.S. greenhouse gas emissions ( percent, respectively) in 201 5 . The percent and residential and commercial sectors contributed the next largest shares of total U.S. greenhouse gas emissions in 201 5 . Emissions from these sectors increase substantially when emissions from electricity are included, due to their relatively large share of electricity consumption (e.g., lighting, appliances). In all sectors except agriculture, CO 2 the combustion of fossil fuels. accounts for more than 80 percent of greenhouse gas emissions, primarily from 15 Figure ES shows the trend in these emissions by sector from 1990 to 201 5 . - - 7 : U.S . Greenhouse Gas Emissions by Economic Sector with Electricity - Related Table ES Eq.) Emissions Distributed (MMT CO 2 1990 2005 2011 2012 2013 201 4 2 015 Implied Sectors a Industry 2,293.9 2,178.1 1,973.6 1,926.7 1,977.4 1,978.7 1,931.1 a 1,810.4 Transportation 1,820.0 1,554.4 2,005.9 1,804.3 1,784.7 1,794.3 a Commercial 968.4 1,217.6 1,158.1 1,100.6 1,128.5 1,139.9 1,114.8 1,057.2 1,241.3 1,161.5 Residential 951.5 1,122.0 1,143.7 1,071.6 620.6 Agriculture 612.4 633.1 561.5 609.9 610.8 612.0 U.S. Territories 33.3 58.1 46.0 48.5 48.1 46.6 46.6 6,586.7 Total Emissions 6,363.1 7,313.3 6,776.7 6,538.3 6,680.1 6,739.7 b LULUCF Sector Net Total ) (819.6 ) (731.0 (749.2) (753.8) (763.0) (761.4) (758.9) 5,978.3 Net Emissions (Sources and Sinks) 5,543.5 6,582.3 6,027.6 5,784.5 5,917.1 5,827.7 a - for estimating the share of gasoline used in on There was a method update in this Inventory road applications. The - road and non change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the transportation sector to industrial and commercial se ctors for 2015, creating a break in the time series. The change is discussed further in the Planned Improvements section of Chapter 3.1 . b O emissions to the atmosphere plus net carbon stock changes. The LULUCF Sector Net To tal is the net sum of all CH and N 2 4 Notes: Emissions from electricity generation are allocated based on aggregate electricity consumption in each end - use sector. Totals may not sum due to indepen dent rounding. Parentheses indicate negative values or sequestration. 24 Emissions were not distributed to U.S. Territories, s ince the electricity generation sector only includes emissions related to the generation of electricity in the 50 states and the District of Columbia. 25 - ES Executive Summary

50 Figure ES - : U.S. Greenhouse Gas Emissions with Electricity - Related Emissions Distributed 15 to Economic Sectors (MMT CO Eq.) 2 - : Recent Trends in Various U.S. Greenhouse Gas Emissions - Related Data Box ES 4 These Total emissions can be compared to other economic and social indices to highlight changes over time. (1) emissions per unit of aggr - related activities are comparisons include: egate energy consumption, because energy related the largest sources of emissions; (2) emissions per unit of fossil fuel consumption, because almost all energy - emissions involve the combustion of fossil fuels; (3) emissions per unit of electricity consumption, because the — utilities and non electric power industry utilities combined — was the largest source of U.S. greenhouse gas - emissions in 201 ; (4) emissions per unit of total gross domestic product as a measure of national economic a ctivity; 5 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. Greenhouse gas emissions in These values represent the relative change in each statistic since 1990. the United States have grown at an average annual rate of 0. 2 percent since 1990. Since 1990, this rate is slightly slower than that for t otal energy and for fossil fuel consumption, and much slower than that for electricity consumption, overall gross domestic product (GDP) and national population (see Figure ES - 16 ). These trends vary relative to 2005, when greenhouse gas emissions, total energy and fossil fuel consumption began to peak. average annual rate of 1 . 0 percent since 2005 . decreased at an Greenhouse gas emissions in the United States have Total energy and fossil fuel consumption have also decreased at slower rates than emissions since 2005, while electricity consumption, GDP, and national population continued to increase. Table ES - 8 : Recent Trends in Various U.S. Data (Index 1990 = 100) Avg. Annual Avg. Annual Change Change 201 201 1 1990 201 2 201 3 4 201 5 since 1990 Variable 2005 since 2005 a Greenhouse Gas Emissions 1.0% - 0.2% 100 104 115 107 103 105 106 b 100 118 115 112 Energy Consumption 115 117 115 0.6% - 0.2% b Fossil Fuel Consumption 0.7% 100 119 110 107 110 111 110 0.4% - b Electricity Consumption 100 134 137 135 136 138 137 1.3% 0.3% c GDP 171 100 159 168 174 178 183 2.5% 1.4% d 0.8% 128 127 Population 100 118 125 126 126 1.0% a GWP - weighted values Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 - 5 201 – 26 ES

51 b Energy content weighted values (EIA 2016) - c Gross Domestic Product in chained 2009 dollars (BEA 7 ) 201 d U.S. Census Bureau (201 6 ) Figure ES - 16 : U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product (GDP) Source: BEA (201 7 ), U.S. Census Bureau (201 6 ), and emission estimates in this report. Key Categories The 2006 IPCC Guidelines (IPCC 2006) defines a key category as a “[category] that is prioritized within the national inventory system because its estimate has a significant influence on a country’s total inventory of 25 greenhouse gases in terms of the absolute By level, the trend, or the uncertainty in emissions and removals.” 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 s hould be performed, in order to capture any key categories that were not identified in either of the quantitative analyses. 5 emission estimates for the key categories as defined by a level analysis (i.e., the Figure ES - 17 presents 201 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 rep orted 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 Key Categories and Annex 1. more information regarding key categories, see Section 1.5 – 25 www.ipcc See Chapter 4 “Methodological Choice and Identification of Key Categories” in IPCC (2006). See 27 - ES Executive Summary

52 Figure ES : 201 5 Key Categories (MMT CO - Eq.) 17 2 Note: For a complete discussion of the key category analysis, see Annex 1. Blue bars indicate either an Approach 1, or Approa ch 1 and Approach 2 level assessment key category. Gray bars indicate solely an Approach 2 level assessment key category. Quality A ssurance 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 systemat ic approach to QA/QC. T he 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 expert and public reviews for both the Inventory , and the UNFCCC reporting guidelines. T he QA process includes I . I nventory report nventory estimates and the Uncertainty Analysis of Emission Estimates emovals Uncertainty estimates are an essential element of a complete inventory of greenhouse gas emissions and r , because they help to prioritize future work and improve overall quality. Some of the current estimates, such as those the for CO emissions from energy - related activities, are considered to have low uncertainties . This is because 2 related activities is directly related to the amount of fuel consumed, the fraction amount of CO emitted from energy - 2 he carbon content of the fue l and, f or the United States, the uncertainties associated of the fuel that is oxidized, and t 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 error associated with the estimates presented . Recogn izing the benefit of conducting an uncertainty analysis, the (IPCC 2006) , Volume 1, 2006 IPCC Guidelines UNFCCC reporting guidelines follow the recommendations of the Chapter 3 nd sink categories. and require that countries provide single estimates of uncertainty for source a - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 28 ES

53 , a In addition to quantitative uncertainty assessments provided in accordance with UNFCCC reporting guidelines 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 - 5 : Recalculations of Inventory Estimates Each year, emission and sink estimates are recalculated and revised for all years in th e Inventory of U.S. Greenhouse Gas Emissions and Sinks , as attempts are made to improve both the analyses themselves, through the use of better 2006 IP CC In this effort, the United States follows the methods or data, and the overall usefulness of the report. (IPCC 2006), which states, “Both methodological changes and refinements over time are an essential Guidelines part of improving inventory quality. It is good practice to change or refine methods when: available data have changed; the previously used met hod is not consistent with the IPCC guidelines for that category; a category has become key; the previously used method is insufficient to reflect mitigation activities in a transparent manner; the capacity for inventory preparation has increased; new inve ntory 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; 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 201 4 ) has been anges in historical data are Ch (IPCC 2006). 2006 IPCC Guidelines recalculated to reflect the change, per the generally the result of changes in statistical data supplied by other agencies. References for the data are provided for additional information. 29 - ES Executive Summary

54 1. Introduction This report presents estimates by the United States government of U.S. anthropogenic greenhouse gas emissions and 5. A summary of these estimates is p rovided in Table 2 - 1 and Table 2 - 2 by gas sinks for the years 1990 through 201 The emission estimates in these tables are and source category in the Trends in Greenhouse Gas Emissions chapter. 1 presented on both a full molecular mass basis and on a Global Warming Potential (GWP) weighted basis 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 ra tified 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 acc ordance 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 ti me - frame sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not threatened 2 , 3 and to enable economic development to proceed in a sustainable manner.” Parties to the Convention, by ratifying, “shall develo p, periodically update, publish and make available...national inventories of anthropogenic emissions by sources and removals by sinks of all greenhouse gases not controlled by 4 the Montreal Protocol, using comparable methodologies...” The United States views t his 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 Intergovernme ntal Panel on Climate Change The role of the IPCC is to assess on a comprehensive, objective, open and transparent basis the scientific, (IPCC). technical and socio economic information relevant to understanding the scientific basis of risk of human - induce d - climate change, its potential impacts and options for adaptation and mitigation (IPCC 20 14 ). Under Working Group 1 of the IPCC, nearly 140 scientists and national experts from more than thirty countries collaborated in the creation Revised 1996 IP of the CC 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 Nat ional Greenhouse Gas Inventories and the further expanded upon the IPCC Good Practice Guidance for Land Use, Land - Use Change, and Forestry methodologies in the Revised 1996 IPCC Guidelines . In 2006, the IPCC accepted the 2006 Guidelines for National Fifth Session (Mauritius, April 2006) ouse Gas Inventories at its Twenty - Greenh . The 2006 IPCC Guidelines built 1 IPCC Fourth Assessment More information provided in “Global Warming Potentials” section of this chapter on the use of Report (AR4) GWP values. 2 nhouse gas emissions and removals that are a direct result of human The term “anthropogenic,” in this context, refers to gree activities or are the result of natural processes that have been affected by human activities (IPCC 2006). 3 MO Information Unit on Climate Article 2 of the Framework Convention on Climate Change published by the UNEP/W Change. See . (UNEP/WMO 2000) 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 elaborate d the role of Annex I Parties in preparing national inventories. See . 1 - 1 Introduction

55 upon the previous bodies of work and include new sources and gases “...as well as updates to the previously technical knowledge have improved since the previous guidelines were published methods whenever scientific and The UNFCCC adopted the as the standard methodological approach for Annex I issued. ” 2006 IPCC Guidelines 23, 201 3). This report presents countries at the Nineteenth Conference of the Parties (Warsaw, November 11 - 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 5 s consistent with The structure of this report i sinks for the United States, including all states and U.S. Territories. the current UNFCCC Guidelines on Annual Inventories (UNFCCC 2014) for Parties included in Annex I of the Convention. 1 - 1 : Methodological Approach for Estimating and Reporting U.S. Emissio ns and Sinks Box national greenhouse gas emissions In following the UNFCCC requirement under Article 4.1 to develop and submit inventories, the gross emissions total presented in this report for the United States excludes emissions and sinks from LULUCF. The n et emissions total presented in this report for the United States includes emissions and sinks from LULUCF. All emissions and sinks are calculated using internationally - accepted methods consistent with the 6 IPCC . Additionally, the calculated emi ssions and sinks in a given year for the United States are Guidelines presented in a common manner in line with the UNFCCC reporting guidelines for the reporting of inventories under 7 The use of consistent methods to calculate emissions and sinks by all nations this international agreement. providing their inventories to the UNFCCC ensures that these reports are comparable. In this regard, U.S. emissions and sinks reported in this Inventory are comparable to emissions and sinks reported by other countries. The report itself follows this standardized format, and provides an explanation of the IPCC methods used to calculate emissions and sinks, and the manner in which those calculations are conducted. On October 30, 2009, the U.S. Environmental Protection Agency (EPA) p ublished 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 g reenhouse gas emitters, fossil 8 CO underground for sequestration or other reasons. fuel suppliers, industrial gas suppliers, and facilities that inject 2 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 . s Inventory, providing not only an important resource for the emissions annual The GHGRP dataset continue to be information, but also other annual information, such as improve and activity data and emissions factors that can and trends over time ate national GHGRP data also allow EPA to disaggreg national emission estimates refine . - categories of emissions. The I nventory estimates in new ways that can highlight differences across regions and sub 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 9 a for a number of categories and uses that dat consistent with IPCC guidance . EPA has integrated GHGRP 10 11 several categories other categorie this year and also identifie s information for where EPA plans to integrate s 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 < http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf >. 8 See < https://www.epa.gov/ghgreporting >. 9 See < http://www.ipcc - nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf > 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 22 Production, Lime Production, Magnesium Production and Processing, ODS Transmission and - Distribution, HCFC Substitutes, Nitric Acid Production, Petrochemical Production, Semiconductor Manufacture); and Waste (Landfills). 11 Industrial Process and Product Use (Ammonia Prod uction, Cement Production, and Other Fluorinated Gas Production) - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 2 1

56 additional GHGRP data (see those categories Planned Improvement sections for in the next edition of this report details). Background Information 1.1 Science For over the past 200 years, the burning of fossil fuels such as coal and oil, deforestation, land - use changes, and other sources have caused the concentrations of heat - trapping "greenhouse gases" to increase significantly in our atmosphere (NOAA 2017). These gases in the atmosphere absorb some of the energy being radiated from the surface of the Earth that would otherwise be lost to space, essentially acting like a blanket that makes the Earth's surface warmer than it would be otherwise. Greenhouse gases are necessary to life as we know it. Without greenhouse gases to create the natural heat - trapping properties of the atmosphere, the planet's surface would be about 60 degrees Fahrenheit cooler than present ( USGCRP 2014 ). Carbon dioxide is also necessary for plant growth. With emissions from biological and geologic al sources, ther Human emissions of e is a natural level of greenhouse gases that is maintained in the atmosphere. greenhouse gases and subsequent changes in s the balance of energy transfers atmospheric concentrations alter (IPCC 2013) . A gauge of these changes is called radiative forcing, which is a between space and the earth system substance’s total net effect on the global energy balance for which a positive number represents a measure of a warming effect and a negative number represents a cooling effect (IPCC 2013). IP CC concluded in its most recent scientific assessment report that it is extremely likely that human influences have been the dominant cause of th warming since the mid - 20 century (IPCC 2013). made sources, the Earth's temperature is As concentrations of greenhouse gases continue to increase in from man - ge land and ocean surface temperature has increased by about 1.2 to climbing above past levels. The Earth's avera each been the warmest decade successively at the 1.9 degrees Fahrenheit since 1880. The last three decades have Earth’s surface since 1850 (IPCC 2013). Other aspects of the climate are also changing such as rainfall patterns, snow and ice cover, and sea level. concentrations continue to increase, cl imate models predict that If greenhouse gas the average temperature at the Earth's surface is likely to increase from 0.5 to 8.6 degrees Fahrenheit above 1986 and the responsiveness of the climate through 2005 levels by the end of this century, depending on future emissions (IPCC 2013). system For further information on greenhouse gases, radiative forcing, and implications for climate change, see the recent 12 13 scientific assessment reports from the IPCC, the U.S. Global Change Research Program (USGCRP), and the 14 National Acade (NAS). mies of Sciences, Engineering, and Medicine Greenhouse Gases Although the Earth’s atmosphere consists mainly of oxygen and nitrogen, neither plays a significant role in enhancing the greenhouse effect because both are essentially transparent to terrestrial radiation. The greenhouse effect is primarily a function of the concentration of water vapor, carbon dioxide (CO ), methane (CH ), nitrous 2 4 oxide (N rface of the O), and other trace gases in the atmosphere that absorb the terrestrial radiation leaving the su 2 Earth (IPCC 2013). CO N , CH Several classes of , Naturally occurring greenhouse gases include water vapor, O , and ozone (O ). 3 2 4 2 halogenated substances that contain fluorine, chlorine, or bromine are also greenhouse gases, but they are, for the most part, solely a product of industrial activities. Chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons 12 See < http://www.ipcc.ch/report/ar5 > 13 See < http://nca2014.globalchange.gov > 14 http://nas - sites.org/americasclimatechoices/ See < > 3 - 1 Introduction

57 (HCFCs) are halocarbons that contain chlorine, while halocarbons that contain bromine are referred to as bromofluorocarbons (i.e., halons). ratospheric ozone depleting substances, CFCs, HCFCs, and halons are As st The UNFCCC defers to this covered under the Montreal Protocol on Substances that Deplete the Ozone Layer. earlier international treaty. Consequently, Parties to the UNFCCC are not required to include these gases in national 15 Some other fluorine - greenhouse gas inventories. — hydrofluorocarbons (HFCs), containing halogenated substances c ) — do not deplete stratospheri ), and nitrogen trifluoride (NF perfluorocarbons (PFCs), sulfur hexafluoride (SF 3 6 These latter substances are addressed by the UNFCCC and accounted for in ozone but are potent greenhouse gases. national greenhouse gas inventories. There are also several other substances that influence the global radiation budget but are short - lived a nd therefore not well mixed , leading to spatially variable radiative forcing effects . These substances include carbon monoxide - (CO), nitrogen dioxide (NO Tropospheric ozone ), sulfur dioxide (SO ), and tropospheric (ground level) ozone ( O . ) 3 2 2 is formed from volatile organic chemical reactions in the atmosphere of precursor pollutants, which include ) compounds (VOCs , including CH ) and nitrogen oxides (NO , in the presence of ultraviolet light (sunlight) . x 4 Aerosols are extremely small particles or liquid drop lets 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. f the atmosphere (e.g., scattering They can affect the absorptive characteristics o incoming sunlight away from the Earth’s surface, or, in the case of black carbon, absorb sunlight) and can play a , as well as role in affecting cloud formation and lifetime the radiative forcing of clouds and precipitatio n 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, usly emitted to and removed from the atmosphere by natural processes , and N are continuo O CH 2 4 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 atmosp heric 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 CO CH N CF O SF 6 2 4 4 2 0.270 ppm - industrial atmospheric concentration 280 ppm 0.700 ppm 0 ppt 40 ppt Pre c a b b b Atmospheric concentration 1.834 ppm 8.6 ppt 0.32 79 ppt 404 ppm 8 ppm e e e d,e 0.7 ppt/yr 2.4 ppm/yr 5 ppb/yr 0.8 ppb/yr Rate of concentration change 0.27 ppt/yr g f g Atmospheric lifetime (years) 12.4 50,000 121 3,200 See footnote a The atmospheric CO concentration is the 2016 annual average at the Mauna Loa, HI station (NOAA/ESRL 2017). 2 b The values presented are global 2015 annual average mole fractions (CDIAC 2016). c global mean atmospheric concentration is from the Advanced Global Atmospheric Gases Experiment (IPCC The 2011 CF 4 2013). d The growth rate for atmospheric CH decreased fr om over 10 ppb/yr in the 1980s to nearly zero in the early 2000s; recently, the 4 growth rate has been about 5 ppb/year . e The rate of concentration change is the average rate of change between 2005 and 2011 (IPCC 2013). f For a given amount of carbon dioxi de 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 y ears, and a small portion of the incr ease will remain for many centuries or more. g This lifetime has been defined as an “adjustment time” that takes into account the indirect effect of the gas on its own resi dence time. 15 - e Emissions estimates of CFCs, HCFCs, halons and other ozon depleting substances are included in this document for informational purposes. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 4 1

58 Source: Pre - nges for CH industrial atmospheric concentrations, atmospheric , N lifetime, and rate of concentration cha O, SF , and 6 4 2 are from IPCC (2013). The rate of concentration change for CO CF is an average of the rates from 2011 through 2016 has 2 4 to 3.0 ppm per year over this period (NOAA/ESRL 2017). 9 fluctuated between 1. The following A brief description of each greenhouse gas, its sources, and its role in the atmosphere is given below. gases as a measure of their relative section then explains the concept of GWPs, which are assigned to individual average global radiative forcing effect. Water vapor is O) Water vapor is the largest contributor to the natural greenhouse effect. Water Vapor (H . 2 fundamentally different from other greenhouse gases in that it can con dense 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 ase temperature. While some human activities such as evaporation from irrigated crops or power plant cooling rele 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 bo th 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 emissions of water vapor can create Aircraft contrails of water vapor. , which may also develop into contrail - induced cirrus clouds , with complex regional and temporal net forcing effects that currently have a low level of radiative 2013 ). scientific certainty (IPCC In nature, carbon is cycled between various atmospheric, oceanic, land biotic, marine biotic, ). Carbon Dioxide (CO 2 and mineral reservoirs. The largest fluxes occur betwee n 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 CO complex function of . Atmospheric CO is part of this global carbon cycle, and therefore its fate is a 2 2 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 40 4 percent ppm v in 201 6 , a 4 4 , 17 16 The IPCC definitively states that “the increase of ... is increase (IPCC 2013 and NOAA/ESRL 201 7 ). CO 2 caused by anthropogenic emissions from the use of fossil fuel as a source of energy and from land use and land use The predomi CO emissions is the nant source of anthropogenic changes, in particular agriculture” (IPCC 2013). 2 Forest clearing, other biomass burning, and some non - energy production processes (e.g., combustion of fossil fuels. . In its Fifth Assessment Report , the IPCC stated “it is CO cement production) also emit notable quantities of 2 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 is the most important (IPCC 2013). ,” of which CO 2 Methane is primarily produced through anaerobic decomposition of organic matter in biological Methane (CH ). 4 systems. Agricultural processes such as wetland rice cultivation, enteric fermentation in anima ls, and the decomposition of animal wastes emit Methane is also , as does the decomposition of municipal solid wastes. CH 4 emitted during the production and distribution of natural gas and petroleum, and is released as a by - product of coal mining and incomp lete fossil fuel combustion. Atmospheric concentrations of CH have increased by about 16 2 4 18 - industrial value of about 700 ppb to 1,8 34 ppb in 201 5 percent since 1750, from a pre although the rate of increase decreased to near zero in the early 2000s, and h as recently increased again to about 5 ppb/year. The IPCC has CH flux to the atmosphere is anthropogenic, from human estimated that slightly more than half of the current 4 activities such as agriculture, fossil fuel use, and waste disposal (IPCC 2007). Met hane is primarily removed from the atmosphere through a reaction with the hydroxyl radical (OH) and is ultimately converted to CO . Minor removal processes also include reaction with chlorine in the marine boundary 2 layer, a soil sink, and stratospheric rea ctions. Increasing emissions of CH reduce the concentration of OH, a 4 CH (IPCC 2013). Methane’s reactions in the atmosphere also feedback that increases the atmospheric lifetime of 4 16 The pre - industrial period is considered as the time preceding the year 1750 (IPCC 2013). 17 riod (i.e., 750 to 1750), a time of relative Carbon dioxide concentrations during the last 1,000 years of the pre - industrial pe climate stability, fluctuated by about  10 ppmv around 280 ppmv (IPCC 2013). 18 6 ). This value is the global 2015 annual average mole fraction (CDIAC 201 5 - 1 Introduction

59 lead to production of tropospheric ozone and stratospheric water vapor, bo th of which also contribute to climate change. emissions include agricultural soils, especially production of O). Anthropogenic sources of N Nitrous Oxide (N O 2 2 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) an d nitric acid production; wastewater The atmospheric concentration of N O treatment and waste incineration; and biomass burning. has increased by 21 2 19 8 ppb in 201 5, industrial value of about 270 ppb to 32 a concentration that has not - percent since 1750, from a pre 800 thousand years. Nitrous oxide is primarily removed from the atmosphere by the been exceeded during the last 13 ). photolytic action of sunlight in the stratosphere (IPCC 20 20 ields the Earth from harmful levels of ). Ozone is present in both the upper stratosphere, Ozone (O where it sh 3 21 where it is the main component of ultraviolet radiation, and at lower concentrations in the troposphere, anthropogenic photochemical “smog.” During the last two decades, emissions of anthropogenic chlorine and e - containing halocarbons, such as CFCs, have depleted stratospheric ozone concentrations. This loss of bromin ozone in the stratosphere has resulted in negative radiative forcing, representing an indirect effect of anthropogenic compounds (IPCC 2013). The depletion of stratospheric ozone and its radiative emissions of chlorine and bromine 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 provid e the fourth largest increase in direct radiative forcing since the pre - industrial era, behind CO Tropospheric , black carbon, and CH . 4 2 ozone is produced from complex chemical reactions of volatile organic compounds (including CH ) mixing with 4 NO lived in the p resence of sunlight. The tropospheric concentrations of ozone and these other pollutants are short - x and, therefore, spatially variable (IPCC 2013). Halocarbons, Sulfur Hexafluoride, and Nitrogen . de ma - Halocarbons are, for the most part, man Trifluoride chemicals that have direct radiative forcing effects . Halocarbons that contain and could also have an indirect effect 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 Although most CFCs and HCFCs are potent global warming Protocol on Substances that Deplete the Ozone Layer. is reduced because they cause stratospheric ozone gases, their net radiative forcing effect on the atmosphere depletion, which itself is a greenhouse gas but which also shield s the Earth from harmful levels of ultraviolet radiation. Under the Montreal Protocol, the United States phased out the production and import ation of halons by 1994 and of CFCs by 1996. Under the Copenhagen Amendments to the Protocol, a cap was placed on the production 22 and importation of HCFCs by non - Article 5 countries , including the U.S., beginning in 1996, and then followed by While ozone depleting gases covered under intermediate requirements and a complete phase - out by the year 2030. the Montreal Protocol and its Amendments are not covered by the UNFCCC, they are reported in this nventory I under Annex 6.2 for informational purposes. Hydrof luorocarbons, PFCs, SF are not ozone depleting substances. The most common HFCs are, , and NF 3 6 however, powerful greenhouse gases. Hydrofluorocarbons are primarily used as replacements for ozone depleting substances but also emitted as a by - product of the HCFC - 22 (chlorodifluoromethane) manufacturing process. Currently, they have a small aggregate radiative forcing impact, but it is anticipated that without further controls 19 (CDIAC 201 6 ). value is the global 2015 annual average This 20 The stratosphere is the layer from the troposphere up to roughly 50 kilometers. In the lower regions the temperature is nearl y constant but in the upper layer the temperature increases rapidly because of sunlight absorption by the ozone layer. The ozo ne - layer is the part of the stratosphere from 19 kilometers up to 48 kilometers where the concentration of ozone reaches up to 1 0 parts per million. 21 The troposphere is the layer from the ground up to 11 kilometers near the poles and up to 16 kilometers i n 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 5 of the Montreal Protocol covers several groups of countries, especially developing countries, with low consumption Article rates of ozone depleting substances. Developing countries with per capita consumption of less than 0.3 kg of certain ozone ances (weighted by their ozone depleting potential) receive financial assistance and a grace period of ten depleting subst - additional years in the phase out of ozone depleting substances. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 6 1

60 their contribution to overall radiative forcing will increase (IPCC 2013). An amendm ent to the Montreal Protocol was adopted in 2016 which includes obligations for Parties to phase down the production and consumption of HFCs. Perfluorocarbons, SF are predominantly emitted from various industrial processes including aluminum , and NF 3 6 lting, semiconductor manufacturing, electric power transmission and distribution, and magnesium casting. sme , and NF Currently, the radiative forcing impact of PFCs, SF is also small, but they have a significant growth rate, 3 6 extremely long atmospheric lifeti mes, and are strong absorbers of infrared radiation, and therefore have the potential to influence climate far into the future (IPCC 2013). Carbon Monoxide Carbon monoxide has an indirect radiative forcing effect by elevating concentrations of (CO) . an d tropospheric CH ozone through chemical reactions with other atmospheric constituents (e.g., the hydroxyl 4 CH and tropospheric ozone. radical, OH) that would otherwise assist in destroying Carbon monoxide is created 4 when carbon - containing fuels are burned in completely. Through natural processes in the atmosphere, it is eventually oxidized to CO . Carbon monoxide concentrations are both short - lived in the atmosphere and spatially variable. 2 Nitrogen Oxides (NO ). The primary climate change effects of nitrogen oxides (i.e., NO and NO ) are indirect . 2 x 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 NO as a precursor to nitrate particles (i.e., aerosols) and due to de struction of x 23 Additionally, NO stratospheric ozone when emitted from very high altitude aircraft. emissions are also likely to x decrease CH concentrations, thus having a negative radiative forcing effect (IPCC 2013). Nitrogen oxides are 4 created from light ning, soil microbial activity, biomass burning (both natural and anthropogenic fires) fuel combustion, and, in the stratosphere, from the photo - degradation of N O. Concentrations of NO are both relatively 2 x - le. lived in the atmosphere and spatially variab short Non - Non - methane volatile organic compounds include methane Volatile Organic Compounds (NMVOCs). substances such as propane, butane, and ethane. These compounds participate, along with NO , in the formation of x oxidants. NMVOCs are emitted primarily from transportation and tropospheric ozone and other photochemical - 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 variabl e. 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 b y 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 24 and carbonaceous aerosols, sulfates, nitrates, volcanic aerosols, industrial dust , aerosols (e.g., black carbon, . Aerosols can be removed from the atmosphere relatively rapi dly by precipitation or through more organic carbon) 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 so lar 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 - aerosol interactions, the contribution of aerosols to radiative forcing are difficult to quantify understanding of cloud because aerosols generally have short atmospheric lifetimes, and have number concentrations, size distribut ions, 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 nges on aerosol forcing, there is high confidence that aerosols have climate). In fact, “despite the large uncertainty ra 25 offset a substantial portion of GHG forcing” (IPCC 2013). Although because they remain in the atmosphere for 23 emissions injected higher in the stratosphere, primarily from fuel combustion emissions from high altitude supersonic NO x 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 h confidence as an indication of strong scientific evidence and agreement in this statement. The IPCC (2013) defines hig 7 - 1 Introduction

61 26 only days to weeks, their concentrations respond rapidly to changes in Not all aerosols have a cooling emissions. effect. Current research suggests that another constituent of aerosols, black carbon, has a positive radiative forcing (IPCC 2013). by heating the Earth’s atmosphere and causing surface warming when deposited on ice and snow Black carbon also influences cloud development, but the direction and magnitude of this forcing is an area of active research. Global Warming Potentials A global warming potential is a quantified measure of the globally averaged relative radiative forcing impacts of a particular greenhouse gas (see Table 1 - 2 ). It is defined as the accumulated radiative forcing within a specific time horizon caused by emitting 1 kilogram (kg) of the gas, relative to that of the reference gas CO (IPCC 20 14) . Direct 2 radiative effects occur when the gas itself absorbs radi ation. Indirect radiative forcing occurs when chemical transformations involving the original gas produce a gas or gases that are greenhouse gases, or when a gas ses. influences other radiatively important processes such as the atmospheric lifetimes of other ga The reference gas CO used is , and therefore GWP - weighted emissions are measured in million metric tons of equivalent (MMT CO 2 2 27 Eq.). The relationship between kilotons (kt) of a gas and MMT CO Eq. can be expressed as follows: CO 2 2 푀푀푇 ) ( ( ) 푀푀푇 퐶푂 푘푡 표푓 푔푎푠 퐸푞.= 퐺푊푃 ×( × ) 2 1, 000 푘푡 where, CO equivalent MMT CO Eq. = Million metric tons of 2 2 ilotons (equivalent to a thousand metric tons) kt = k GWP = Global warming potential MMT = Million metric tons GWP values allow for a comparison of the impac ts 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 able. - year time horizon, although other time horizon values are avail ...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 li sted 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 green house 28 gases over a 100 - year time horizon... ) tend to , , NF Greenhouse gases with relatively long atmospheric lifetimes (e.g., CO , CH N , HFCs, PFCs, SF O 4 6 2 2 3 n be be evenly distributed throughout the atmosphere, and consequently global average concentrations ca lived gases such as water vapor, carbon monoxide, tropospheric ozone, ozone precursors determined. The short - products and carbonaceous particles), however, (e.g., NO , and NMVOCs), and tropospheric aerosols (e.g., SO 2 x vary regionally, and consequently Parties to the it is difficult to quantify their global radiative forcing impacts. 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 onger negative forcing effect (i.e., a few years) (IPCC 2013). stratosphere, which can result in a l 27 ths of carbon dioxide by weight. Carbon comprises 12/44 28 Framework Convention on Climate Change; Available online at: ; 31 January 20 14; Report of the Conference of the Parties at its nineteenth session; held in Warsaw from 11 to 23 November n 2013; Addendum; Part two: Action taken by the Conference of the Parties at its nineteenth session; Decision 24/CP.19; Revisio of the UNFCCC report ing guidelines on annual inventories for Parties included in Annex I to the Convention; p. 2. (UNFCCC 2014). - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 8 1

62 Table 1 2 : Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report - a Gas GWP Atmospheric Lifetime b 1 CO 2 See footnote c CH 25 4 12 N O 298 2 114 HFC 23 14,800 - 270 HFC - 32 675 4.9 HFC - 125 3,500 29 HFC 134a 1,430 - 14 HFC 4,470 143a - 52 - 152a HFC 124 1.4 HFC - 227ea 3,220 34.2 HFC 236fa 9,810 - 240 - 4310mee 1,640 HFC 15.9 7,390 CF 4 50,000 12,200 F C 6 2 10,000 C 8,860 F 4 10 2,600 9,300 F C 14 6 3,200 SF 22,800 6 3,200 17,200 NF 3 740 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 d those indirect effects includes the direct effects an The GWP of CH 4 due to the production of tropospheric ozone and stratospheric water is not included. vapor. The indirect effect due to the production of CO 2 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 elative to previous assessment of climate change. Within the AR5 report, the GWP values of gases were revised r IPCC Second Assessment Report (SAR) (IPCC 1996), the IPCC Third Assessment Report IPCC reports, namely the (TAR) (IPCC 2001), and the IPCC Fourth Assessment Report (AR4) (IPCC 2007). Although the AR4 GWP values straight - forward to review the are used throughout this report, consistent with UNFCCC reporting requirements, it is changes to the GWP values and the ir impact on estimates of the total GWP - weighted emissions of the United State s. In the AR5, the IPCC applied an improved calculation of CO radiative forcing and an improved CO response 2 2 Additionally, the atmospheric lifetimes of some gases have been function in presenting updated GWP values. recalculated, and updated background co ncentrations 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 presents the new GWP values, relative to those presented in the adjusted to match more recent science. Table 1 - 3 AR4 and using the 100 - year time horizon common to UNFCCC reporting. For consistency with internationa l 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 29 s report are also presented in unweighted units. For All estimates provided throughout thi for national inventories. 29 See . 9 - 1 Introduction

63 informational purposes, emission estimates that use GWPs from other IPCC Assessment Reports are presented in detail in Annex 6.1 of this report. - Table Comparison of 100 - Year GWP values 3 1 : AR5 with a b AR5 AR4 Gas feedbacks Comparison to AR4 SAR SAR AR5 AR5 with b feedbacks CO 2 1 1 1 NC NC NC 1 c CH 4 (4) 28 21 25 3 9 34 N O 2 298 265 298 12 310 (33) 0 HFC - 23 11,700 14,800 12,400 13,856 (3,100) (2,400) (944) HFC 32 - 677 817 (25) 2 142 675 650 HFC - 125 (330) 3,500 3,170 3,691 (700) 2,800 191 HFC - 134a 1,300 1,430 1,300 1,549 (130) (130) 119 143a - HFC 4,470 4,800 5,508 (670) 330 1,038 3,800 HFC - 152a 140 124 138 167 16 14 43 227ea HFC - 3,350 130 (320) 640 2,900 3,220 3,860 - HFC 236fa 9,810 8,060 (3,510) (1,750) (812) 6,300 8,998 HFC - 4310mee 10 1,640 1,650 1,952 (340) 312 1,300 CF 4 7,349 7,390 6,630 (890) 6,500 (760) (41) F C 6 2 12,200 11,100 12,340 (3,000) (1,100) 140 9,200 C F 4 10 7,000 8,860 9,200 10,213 (1,860) 340 1,353 F C 6 14 8,780 (1,900) (520) 7,400 9,300 7,910 (1,390) SF 6 700 23,500 26,087 23,900 22,800 3,287 1,100 NF 3 NA 17,200 16,100 17,885 NA (1,100) 685 Not Applicable ) NA ( NC ( No Change ) a The GWPs presented here are the ones most consistent with the methodology used in the AR4 report. b - carbon feedbacks for the non - The GWP values presented here from the AR5 report include climate gases in order to be consistent with the appro ach used in calculating the CO CO lifetime. 2 2 Additionally, the AR5 reported separate values for fossil versus biogenic methane in order to account for the CO oxidation product. 2 c The GWP of CH includes the direct effects and those indirect effects due to the production of 4 The indirect effect due to the production of CO is tropospheric ozone and stratospheric water vapor. 2 only included in the value from AR5 that includes climate - carbon feedback s . Note: Parentheses indicate negative values. Source: (IPCC 2013, IPCC 2007, IPCC 2001, IPCC 1996). 1.2 National Inventory Arrangements The U.S. Environmental Protection Agency (EPA), in cooperation with other U.S. government agencies, prepares Inventory of U.S. Greenhouse Gas Emissions and Sinks . A wide range of agencies and individuals are involved the in supplying data to, planning methodological approaches and improvements, reviewing, or preparing portions of the and state government authorities, research and academic institutions, industry U.S. Inventory including federal — associations, and private consultants. Within EPA, the Office of Atmospheric Programs (OAP) is the lead office responsible for the emission calculations y, as well as the completion of the National Inventory Report and the Common Reporting provided in the Inventor Format tables. EPA’s Office of Transportation and Air Quality (OTAQ) is also involved in calculating emissions for cially submits the annual Inventory to the UNFCCC, EPA’s the Inventory. While the U.S. Department of State offi - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 10 1

64 OAP serves as the Inventory focal point for technical questions and comments on the U.S. Inventory. The staff of n calculations at the EPA coordinate the annual methodological choice, activity data collection, and emissio 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 issu es 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 ficial data for use in the Inventory. The U.S. Department of Energy’s Energy Information Administration provide of provides national fuel consumption data and the U.S. Department of Defense provides military fuel consumption and bunker fuels. Informal relationships a lso 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 ration, the Department of Administration, the U.S. Geological Survey, the Federal Highway Administ Transportation, the Bureau of Transportation Statistics, the Department of Commerce, the National Agricultural Statistics Service, and the Federal Aviation Administration. Academic and research centers also provide activity data a nd calculations to EPA, as well as individual companies participating in voluntary outreach efforts with EPA. Finally, the U.S. Department of State officially submits the Inventory to the UNFCCC each April. 1 Figure 1 - diagrams the National Inventory Arrangements. 11 - 1 Introduction

65 Figure 1 - 1 : National Inventory Arrangements Diagram 201 - 1 12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 – 5

66 Inventory Process 1.3 ’s This section describes EPA approach to preparing the annual U.S. Inventory, which consists of a National Inventor y Report (NIR) and Common Reporting Format (CRF) tables. The inventory coordinator at EPA 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 ource stage process for collecting information from the individual s contractors familiar with the sources. A multi - 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 th e estimation methodology for the individual source categories. Because EPA has been preparing the Inventory for many years, f or most source categories, the methodology for the previous year is applied to the new “current” year of the Inventory, and invento ry 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 tes is implementing a higher Tiered approach for that source category), then the source category (e.g., the United Sta 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 ent ire time series, and conduct a special source - specific review process involving relevant experts from industry, government, and universities. Once the methodology is in place and the data are collected, the individual source leads calculate emissions and s ink 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 y control (QA/QC) checks, and uncertainty analyses. quality assurance and qualit nventory is based on EPA internal guidelines, as I The treatment of confidential business information (CBI) in the 1 safeguard CBI during the applicable to the data used. EPA has specific procedures in place to well as regulations 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 whil e still providing useful information for analysis. For example, within the Energy and Industrial Process and - level data from the Greenhous Gas Reporting Product Use (IPPU) sectors, EPA has used aggregated facility 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 2 ng only data values that meet the GHGRP In the Inventory, EPA is publishi underlying CBI from public disclosure. 3 aggregation criteria. 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 1 40 CFR part 2, Subpart B titled “Confidentiality of Business Information” which is the regulation establishing rules governing handling of data entitled to confidentiality treatment. See < https://www.ecfr.gov/cgi bin/text - - > idx?SID=a764235c9eadf9afe05fe04c07a28939&mc=true&node=sp40.1.2.b&rgn=div6 . 2 Federal Register Notice on “Greenho use Gas Reporting Program: Publication of Aggregated Greenhouse Gas Data.” See pp, 79 and 110 of notice at < https://www.gpo.gov/fdsys/pkg/FR - 2014 - 06 - 09/pdf/2014 - 13425.pdf >. 3 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggrega ted Greenhouse Gas Data, November - reporting> . 25, 2014. See

67 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. Data Compilatio n and Storage Summary The inventory coordinator at EPA collects the source categories’ descriptive text and Annexes, and also aggregates the emission estimates into a summary spreadsheet that links the individual source category spreadsheets together. This summary sheet contains all of the essential data in one central location, in formats commonly used in the Inventory document. In addition to the data from each source category, national trend and related data are also gathered in the summary sheet for use in the Executive Summary, Introduction, and Recent Trends sections of the Inventory report. Electronic copies of each year’s summary spreadsheet, which contains all the emission and sink estimates for the United States, are kept on a central server at EPA under t he jurisdiction of the inventory coordinator. National Inventory Report Preparation The NIR is compiled from the sections developed by each individual source lead. In addition, the inventory coordinator prepares a brief overview of each chapter that summa rizes the emissions from all sources discussed in the chapters. The inventory coordinator then carries out a key category analysis for the Inventory, consistent with the he reporting requirements , and in accordance with t 2006 IPCC Guidelines for National Greenhouse Gas Inventories Also at this time, the Introduction, Executive Summary, and Recent Trends sections are drafted, to of the UNFCCC. reflect the trends for the most recent year of the current Inventory. The analysis of trends necessitates gathering supplemental data, including weather and temperature conditions, economic activity and gross domestic product, population, atmospheric conditions, and the annual consumption of electricity, energy, and fossil fuels. Changes in these data are used to explai n the trends observed in greenhouse gas emissions in the United States. Furthermore, specific factors that affect individual sectors are researched and discussed. Many of the factors that affect emissions Text boxes are included in the Inventory document as separate analyses or side discussions in boxes within the text. are also created to examine the data aggregated in different ways than in the remainder of the document, such as a focus on transportation activities or emissions from electricity generation . The document is prepared to match the specification of the UNFCCC reporting guidelines for National Inventory Reports. Common Reporting Format Table Compilation The CRF tables are compiled from individual tables completed by each individual source lead, which contain source emissions and activity data. The inventory coordinator integrates the source data into the UNFCCC’s “CRF Reporter” for the United States, assuring consistency across all sectoral tables. The summary reports for emissions, emission factors used, the overview tables for completeness and quality of estimates, the recalculation methods, and tables, the notation key completion tables, and the emission trends tables are then completed by the inventory coordinator. hecks on the CRF Reporter, as well as reviews by the source leads, are Internal automated quality c completed for the entire time series of CRF tables before submission. QA/QC and Uncertainty have general QA/QC and uncertainty analyses are supervised by the QA/QC and uncertainty coordinators, who oversight over the implementation of the QA/QC plan and the overall uncertainty analysis for the Inventory (see sections on QA/QC and Uncertainty, below). These coordinators work closely with the source leads to ensure that a consistent QA/QC plan and uncertainty analysis is implemented across all inventory sources. The inventory QA/QC plan, detailed in a following section, is consistent with the quality assurance procedures outlined by EPA and IPCC. The QA/QC and uncertainty findings also info rm overall improvement planning, and specific improvements are noted in the Planned Improvements sections of respective categories. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 14 1

68 Expert , , and UNFCCC Review Periods Public day Expert Review period, a first draft of the document is sent to a select list of technical experts During the 30 - outside of EPA who are not directly involved in preparing estimates. The purpose of the Expert Review is to encourage feedback on the methodological and data sources used in the current Inventory, especially for sources ich have experienced any changes since the previous Inventory. wh Once comments are received and addressed, a second draft of the document is released for public review by b site. The Public Review publishing a notice in the U.S. Federal Register and posting the document on the EPA We - day comment period and is open to the entire U.S. public. Comments may require further period allows for a 30 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 . EPA publishes comments received with publication of the report on its website. details , the report also undergoes review by an Following completion and submission of the report to the UNFCCC 4 independent international team of experts for adherence to UNFCCC reporting guidelines and IPCC Guidance. Feedback from these review processes all contribute to improving inventory quality over time. Final Submit tal 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. The U.S. Department of State sends the official submission of the U.S. Inventory to the UNFCCC. The document is 5 then formatted and posted online, available for the public. Methodology and Data Sources 1.4 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 f or 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 - - emission For more information on data sources see Section 1.2 above, Box 1 1 on use of GHGRP s to an activity. 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. C hoices 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 s ources utilized in the calculation of each source category. 1 - 3 : IPCC Reference Approach Box The UNFCCC reporting guidelines require countries to complete a "top - down" reference approach for estimating CO emis sions from fossil fuel combustion in addition to their “bottom - up” sectoral methodology. This estimation 2 method uses alternative methodologies and different data sources than those contained in that section of the Energy 4 http://unfccc.int/national_reports/annex_i_ghg_inventories/review_process/items/2762.php >. See < 5 See . 15 - 1 Introduction

69 chapter. The reference approach est imates fossil fuel consumption by adjusting national aggregate fuel production user consumption surveys (see Annex 4 of data for imports, exports, and stock changes rather than relying on end - - ba sed fuels are brought into a national economy, they this report). The reference approach assumes that once carbon are either saved in some way (e.g., stored in products, kept in fuel stocks, or left unoxidized in ash) or combusted, Account ing for actual consumption of and therefore the carbon in them is oxidized and released into the atmosphere. national level is not required. - fuels at the sectoral or sub Key Categories 1.5 2006 IPCC Guidelines The (IPCC 2006) defines a key category as a “[category] that is prioritized within the national inventory system because its estimate has a significant influence on a country’s total inventory of 6 the uncertainty in emissions and removals.” By greenhouse gases in terms of the absolute level, the trend, or definition, key categories include those categories that have the greatest contribution to the absolute level of national In addition, when an entire time series of emission and removal estimates i s prepared, a thorough emissions. investigation of key categories must also account for the influence of trends and uncertainties of individual source This analysis can identify source and sink categories that diverge from the overall trend in and sink categories. Finally, a qualitative evaluation of key categories is performed to capture any categories that na tional emissions. were not identified in any of the quantitative analyses. 2006 (IPCC 2006), was implemented to id entify the key categories Approach 1, as defined in the IPCC Guidelines This analysis was performed twice; one analysis included sources and sinks from the Land for the United States. Use Change, and Forestry (LULUCF) sector, the other analysis did not include the LULUCF categories. - Use, Land 2006 IPCC Guidelines g Approach 1, Approach 2, as defined in the (IPCC 2006), was then implemented to Followin identify any additional key categories not already identified in Approach 1 assessment. This analysis, which includes each source category’s uncertainty a ssessments (or proxies) in its calculations, was also performed twice to include or exclude LULUCF categories. In addition to conducting Approach 1 and 2 level and trend assessments, a qualitative assessment of the source categories, as described in the 06 IPCC Guidelines (IPCC 2006), was conducted to capture any key categories that 20 were not identified by either quantitative method. For this inventory, no additional categories were identified using criteria recommend by IPCC, but EPA continues to update i ts qualitative assessment on an annual basis. 1 - Table : Key Categories for the United States (1990 - 201 5 ) 4 2015 Emissions a Qual (MMT Approach 1 Approach 2 CO IPCC Source Categories Eq.) Gas 2 Trend Level Level Trend Trend Level Trend Level With Without Without Without With With With Without LULUCF LULUCF LULUCF LULUCF LULUCF LULUCF LULUCF LULUCF Energy CO Emissions from 2 CO • Mobile Combustion: • 1,463.4 • • • • • • 2 Road Emissions from CO 2 - Stationary Combustion CO • • 1,350.5 • • • • • • 2 Electricity - Coal Generation Emissions from CO 2 CO • • • • 526.1 • • • • 2 Stationary Combustion - 6 “Methodological Choice and Identification of Key Categories” in IPCC (2006). See Volume 1, See Chapter 4 . - - 1 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 16 5

70 Gas - Electricity Generation Emissions from C O 2 CO Stationary Combustion 467.5 - • • • • • • • • 2 Industrial - Gas C O Emissions from 2 CO 272.2 Stationary Combustion - • • • • • • • 2 Industrial Oil - C Emissions from O 2 CO 252.8 Stationary Combustion - • • • • • • 2 Residential - Gas C O Emissions from 2 CO 175.4 - Stationary Combustion • • • • • • • 2 - Commercial Gas C O Emissions from 2 CO Mobile Combustion: 159.2 • • • • • • • 2 Aviation CO Emissions from 2 CO 125.5 Non - • • Energy Use of • • 2 Fuels C O Emissions from 2 CO 81.5 Mobile Combustion: • • • 2 Other O Emissions from C 2 CO • 67.9 - Stationary Combustion • 2 - Commercial Oil C O Emissions from 2 CO 66.8 • - • • • Stationary Combustion • 2 Residential - Oil C O Emissions from 2 CO 65.9 • • Stationary Combustion • • • - • • • 2 Industrial - Coal CO Emissions from 2 CO • • • 42.4 • 2 Natural Gas Systems C O Emissions from 2 CO • 34.3 - • • • Stationary Combustion 2 Oil U.S. Territories - O Emissions from C 2 CO Mobile Combustion: • • • • 32.3 2 Marine C O Emissions from 2 Stationary Combustion - CO • • • • • • • 23.7 2 Electricity Oil - Generation C O Emissions from 2 CO 3.0 Stationary Combustion - • 2 - Gas U.S. Territories CO Emissions from 2 CO 2.9 Stationary Combustion - • • 2 Coal Commercial - O C Emissions from 2 CO • 0.0 • - Stationary Combustion 2 Residential - Coal CH Emissions from 4 CH • • • 162.4 • • • • • 4 Natural Gas Systems ugitive Emissions from F CH 60.9 • • • • • • • • 4 Coal Mining C H Emissions from 4 CH 39.9 • • • • • • • • 4 Petroleum Systems N on - CO Emissions 2 CH • 3.9 from Stationary • • • 4 Residential - Combustion Introduction 17 - 1

71 N on CO - Emissions 2 from Stationary N O 19.5 • • • • 2 - Electricity Combustion Generation N O Emissions from 2 N O 11.2 Mobile Combustion: • • • • • • 2 Road International Bunker Several .8 • 291 b Fuels ndustrial Processes I CO Emissions from Iron 2 and Steel Production & 48.9 CO • • • • • • • • 2 Metallurgical Coke Production CO Emissions from 2 39.9 CO • • 2 Cement Production O Emissions from C 2 • Petrochemical • CO 28.1 2 Production O C Emissions from 2 Other Process Uses of CO 11.2 • • 2 Carbonates N O Emissions from 2 4.3 N O • • 2 Adipic Acid Production E missions from Substitutes for Ozone 168.5 HiGWP • • • • • • • • Depleting Substances HFC - 23 Emissions from 4.3 HiGWP • • • 22 Production - HCFC P FC Emissions from • 4.2 HiGWP • • • • • Aluminum Production Emissions from SF 6 HiGWP • • Electrical Transmission 2.0 and Distribution Agriculture CH Emissions from 4 CH • • • • 166.5 4 Enteric Fermentation Emissions from CH 4 CH 66.3 • • • • • • • • 4 Manure Management Direct N O Emissions 2 N 213.3 O • • • from Agricultural Soil • • 2 Management Indirect N O Emissions 2 38.0 N O • • • • • • • • 2 from Applied Nitrogen aste W CH Emissions from 4 115.7 CH • • • • • • • • 4 Landfills Land Use, Land Use Change, and Forestry CO Emissions from 2 Land Converted to 68.3 CO • • • • 2 Settlements Emissions from CO 2 22.7 CO • Land Converted to • • • 2 Cropland CO Emissions from 2 20.5 CO • Land Converted to • 2 Grassland C O Emissions from 2 • • (18.0) Cropland Remaining • CO • 2 Cropland - 5 1 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 18 201

72 C O Emissions from 2 Grassland Remaining CO • • • (20.9) 2 Grassland C Emissions from O 2 CO • (75.2) • Land Converted to 2 Forest Land Emissions from O C 2 (102.1) CO • • Settlements Remaining • • 2 Settlements CO Emissions from 2 (666. • • • • CO ) 2 Forest Land Remaining 2 Forest Land C H Emissions from 4 CH 7.3 • • 4 Forest Fires N O Emissions from 2 N O • 4.8 2 Forest Fires Subtotal Without LULUCF 6,422.0 Total Emissions Without LULUCF 6,586.7 Percent of Total Without LULUCF 98% Subtotal With LULUCF 5,660.2 Total Emissions With LULUCF 5,827.7 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). 1.6 Quality Assurance and Quality Control (QA/QC) As part of efforts to achieve its stated goals for inventory quality, transparency, and credibility, the United States has developed a quality assurance and quality control plan designed to check, document and improve the quality of its inventory over time. QA/QC activities on the Inventory are undertaken within the framework of the U.S. Quality Assurance/Qua lity Control and Uncertainty Management Plan (QA/QC plan) for the U.S. Greenhouse Gas Inventory: Procedures Manual for QA/QC and Uncertainty Analysis . Key attributes of the QA/QC plan are summarized in Figure 1 2 . These attributes include: - Procedures and Forms: • detailed and specific systems that serve to standardize the process of documenting tion of QA/QC and the analysis of and archiving information, as well as to guide the implementa uncertainty • Implementation of Procedures: application of QA/QC procedures throughout the whole inventory development process from initial data collection, through preparation of the emission estimates, to publication of t he Inventory expert and public reviews for both the • Quality Assurance: nventory report I nventory estimates and the I (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 (IPCC 2006) 2006 IPCC Guidelines - • Quality Control : consideration of secondary data and category specific checks (Tier 2 QC) in parallel and 19 - 1 Introduction

73 coordination with the uncertainty assessment; the dev elopment of protocols and templates, which provides for more structured communication and integration with the suppliers of secondary information General ( Tier 1) Category • - specific ( Tier 2 ) Checks: quality controls and checks, as recommended by and IPCC Go od 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 ed research efforts investigations which provide for continual data quality improvement and guid • Multi - Year Implementation : a schedule for coordinating the application of QA/QC procedures across multiple years , especially for category - specific QC promoting communication within the EPA, across Federal a gencies and • Interaction and Coordination: departments, state government programs, and research institutions and consulting firms involved in supplying data or preparing estimates for the Inventory. itself is intended to The QA/QC Management Plan be revised and reflect new information t hat 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 sour ces. 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 a 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. Simila rly, 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. n the U.S. QA/QC plan occur throughout the inventory process; QA/QC is The quality control activities described i not separate from, but is an integral part of, preparing the Inventory. in the form of both good — Quality control practices and procedures are being practices (such as documentation procedures) and checks on whether good In addition, quality followed — is applied at every stage of inventory development and document preparation. during the assurance occurs expert review and the public review , in addition to the UNFCCC expert technical inventory quality, the public review phase is also phases significantly contribute to While all review . improving essential for promoting the openness of the inventory development process and the transparency of the inventory data and methods. des the process of ensuring inventory quality by describing data and methodology checks, The QA/QC plan gui developing processes governing peer review and public comments, and developing guidance on conducting an The QA/QC procedures also include feedback loops . analysis of the uncertainty surrounding the emission estimates and provide for corrective actions that are designed to improve the inventory estimates over time. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 20 1

74 Figure 1 - 2 : U.S. QA/QC Plan Summary Uncertainty Analys 1.7 is 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 emission estimates for any inventory. Some of the current estimates, such as those for with the development of ) emissions from energy related activities, are considered to have minimal uncertainty - carbon dioxide (CO 2 associated with them. 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 error associated with the estimates presented. The UNFCCC reporting guidelines follow the recommendation in the 2006 IPCC Guidelin es (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 discu ssed. 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 In particular, emissions from some land greenhouse gas emissions are not available at this use time. - and industrial processes are not included in (e.g., emissions and removals from interior Alaska) activities the inventory either because data are incomplete or because methodologies do not exist for estimating 21 - 1 Introduction

75 emissions from these source categories. See Annex 5 of this report for a discussion of the sources of ouse gas emissions and sinks excluded from this report. greenh Improving the accuracy of emission factors Further research is needed in some cases to improve the . • accuracy of emission factors used to calculate emissions from a variety of sources. For example, the CH N and accuracy of current emission factors applied to emissions from stationary and mobile O 2 4 combustion is highly uncertain. Collecting detailed activity data . Although methodologies exist for estimating emissions for some sources, • btaining activity data at a level of detail where more technology or process - specific problems arise in o emission factors can be applied. The overall uncertainty estimate for total U.S. greenhouse gas emissions was developed using the IPCC Approach 2 uncertainty estimation methodology. Estimates of quantitative uncertainty for the total U.S. greenhouse gas Table emissions are shown below, in - 5 . 1 The IPCC provides good p — Approach 1 and Approach 2 — to estimating ractice guidance on two approaches Approach 2 uncertainty analysis, employing the Monte Carlo Stochastic uncertainty for individual source categories. Simulation technique, was applied wherever data and resources per mitted; further explanation is provided within the respective source category text and in Annex 7. 2006 IPCC Guidelines (IPCC 2006), over a Consistent with the - year timeframe, the United States expects to continue to improve the uncertainty estimates presented in this multi report. 1 - 5 : Estimated Overall Inventory Quantitative Uncertainty (MMT CO Table ) Eq. and Percent 2 2015 Emission Standard Uncertainty Range Relative to Emission c b c a Deviation Mean Estimate Estimate Gas Eq.) (MMT CO (MMT CO Eq.) (%) (MMT CO Eq.) 2 2 2 Lower Upper Lower Upper d d Bound Bound Bound Bound 5,411.0 5,305.4 5,652.4 90.2 - 2% 4% 5,474.3 CO 2 e 655.7 599.9 45.3 - 9% 19% 681.8 779.2 CH 4 e 302.5 30.7 357.0 10% 334.8 424.6 - 27% N O 2 e 5.5 184.7 183.1 204.4 193.4 - 1% 11% , and NF PFC, HFC, SF 6 3 106.0 6,586.2 6,505.0 6,919.9 - 1% 5% 6,706.6 Total f - 14.6 38.2 19.7 94% 23.3 6.3 26% LULUCF Emissions g (778.7) (993.1) (620.7) - 94.7 20% 28% (808.4) LULUCF Total Net Flux h 94.8 (785.1) (758.9) (969.7) (597.9) - 21% 28% LULUCF Sector Total Net Emissions (Sources and 5,827.3 5,643.8 6,207.4 - 3% 7% 5,921.5 142.8 Sinks) 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 this table exclude approximately MMT CO Eq. of uncertainty was performed this year. Thus the totals reported in 0.4 2 emissions for which quantitative uncertainty was not assessed. Hence, these emission estimates do not match the final 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 th th percentile and the upper bound corresponding to 97.5 percentile. corresponding to 2.5 c Mean value indicates the arithmetic average of the sim ulated 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 tak e into account the uncertainty in the GWP values for CH O and high GWP , N 2 4 gases used in the nventory emission calculations for 2015 . I f LULUCF emissions include the CH Emissions from Forest Fires, Emissions from and N O emissions reported for Non - CO 2 2 4 - Emissions from Grassland Fires, CO Fluxes from Settlement O N Dra ined Organic Soils, N O Fluxes from Forest Soils, Non 2 2 2 1 5 201 - Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 22 –

76 Soils , Coastal Wetlands Remaining Coastal Wetlands, Peatlands Remaining Peatlands, and CH Emissions from Land 4 . Converted to Coastal Wetlands, g Net CO flux is the net C stock change from the following categories: Forest Land Remaining Forest Land , Land Converted to 2 , Cropland Remaining Cropland , Land Converted to Cropland , Grasslan d Remaining Grassland , Land Converted Forest Land , Changes in Organic Soils Carbon Stocks, Changes in Urban Tree Carbon Stocks, Changes in Yard Trimmings and to Grassland Food Scrap Carbon Stocks in Landfills, Land Converted to Settlements , Wetlands Remaining Wetlands , and Land Converted to Wetlands . h The LULUCF Sector Total is the net sum of all emissions (i.e., sources) of greenhouse gases to the atmosphere plus removals CO (i.e., sinks or negative emissions) from the atmosphere. of 2 Notes: Totals may not sum du o independent rounding. Parentheses indicate net sequestration. e t Emissions calculated for the U.S. Inventory reflect current best estimates; in some cases, however, estimates are based on approximate methodologies, assumptions, and incomplete data. As new information becomes available in the future, the United States will continue to improve and revise its emission estimates. See Annex 7 of this report for further details on the U.S. process for estimating uncertainty associated with the emission estimates and for a more detailed discussion of the limitations of the current analysis and plans for improvement. Annex 7 also includes details on the uncertainty analysis performed for selected source categories. Completeness 1.8 ing CRF tables, serves as a thorough assessment of the anthropogenic sources This report, along with its accompany and sinks of greenhouse gas emissions for the United States for the time series 1990 through 2015. This report is intended to be comprehensive and includes the vast majority of em issions and removals identified as anthropogenic, consistent with IPCC and UNFCCC guidelines. In general, sources not accounted for in this Inventory are excluded imate and/or the sources because they are not occurring in the U.S., or because data are unavailable to develop an est 7 were determined to be insignificant in terms of overall national emissions per UNFCCC reporting guidelines. The United States is continually working to improve upon the understanding of such sources and seeking to find the equired to estimate related emissions. As such improvements are implemented, new emission sources are data r quantified and included in the Inventory. For a list of sources not included and more information, see Annex 5 and the respective source category sections in each chapter of this report. 1.9 Organization of Report In accordance with the revision of the UNFCCC reporting guidelines agreed to at the nineteenth Conference of the Parties (UNFCCC 2014), this Inventory of U.S. Greenhouse Gas Emissions and Sinks is seg regated into five sector - In addition, chapters on Trends in Greenhouse Gas Emissions and Other specific chapters, listed below in Table 1 - 6 . informatio n to be considered as part of the U.S. Inventory submission are included. Table 6 : IPCC Sector Descriptions - 1 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. 7 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 emissi ons Eq. The total national aggregate of below 0.05 per cent of the national total GHG emissions, and does not exceed 500 kt CO is 2 estimated emissions for all gases and categories considered insignificant shall remain below 0.1 percent of the national tota l GHG e missions.” 23 - 1 Introduction

77 Industrial Processes and Emissions resulting from industrial processes and product use of green house gases. Product Use Agriculture Anthropogenic emissions from agricultural activities except fuel combustion, which is addressed under Energy. Land Use, Land - Use land use, Emissions and removals of CO O from , and emissions of CH , and N 4 2 2 land - use change and forestry . Change, and Forestry Waste Emissions from waste management activities. Within each chapter, emissions are identified by the anthropogenic activity that is the source or sink of the greenhouse gas emissions being estimated Overall, the following organizational structure is (e.g., coal mining). consistently applied throughout this report: : Overview of emission trends for each IPCC defined sector Chapter/IPCC Sector Source category Description of source pathway and emission t rends. : Methodology : Description of analytical methods employed to produce emission estimates and identification of data references, primarily for activity data and emission factors. S : A discussion and quantification o f the uncertainty in emission eries Consistency Uncertainty and Time series consistency. - estimates and a discussion of time QA/QC and Verification : A discussion on steps taken to QA/QC and verify the emission estimates, beyond the scope of the overall U.S. QA/QC plan, and any key findings. alculations Rec A discussion of any data or methodological changes that necessitate a recalculation of previous : years’ emission estimates, and the impact of the recalculation on the emission estimates, if applicable. Planned Improvements : A discussion on any source - specific planned improvements, if applicable. Special attention is given to CO from fossil fuel combustion relative to other sources because of its share of 2 emissions and its dominant influence on emission trends. end - use sector (i.e., For example, each energy consuming residential, commercial, industrial, and transportation), as well as the electricity generation sector, is described individually. Additional information for certain source categories and other topics is also provided in several Anne Table 1 - 7 . xes listed in Table 1 - 7 : List of Annexes ANNEX 1 Key Category Analysis ANNEX 2 Methodology and Data for Estimating CO n Emissions from Fossil Fuel Combustio 2 2.1. Methodology for Estimating Emissions of CO from Fossil Fuel Combustion 2 2.2. Methodology for Estimating the Carbon Content of Fossil Fuels Methodology for Estimating Carbon Emitted from Non 2.3. Energy Uses of Fossil Fuels - ANNEX 3 Methodological Descriptions for Additional Source or Sink Categories 3.1. Methodology for Estimating Emissions of CH O, and Indirect Greenhouse Gases from Stationary , N 2 4 Combustion Methodology for Estimating Emissions of CH , 3.2. N from Mobile O, and Indirect Greenhouse Gases 2 4 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 CH ns from Coal Mining Emissio 4 3.5. Methodology for Estimating CH and CO Emissions from Petroleum Systems 2 4 Methodology for Estimating CH Emissions from Natural Gas Systems 3.6. 4 Methodology for Estimating CO N and 3.7. O Emissions from Incineration of Waste 2 2 8. Methodology for Estimating Emissions from International Bunker Fuels used by the U.S. Military 3. Methodology for Estimating HFC and PFC Emissions from Substitution of Ozone Depleting Substances 3.9. Methodology for Estimating CH 3.10. Emissions from Enteric Fermentation 4 3.11. Methodology for Estimating CH N O Emissions from Manure Management and 2 4 3.12. Methodology for Estimating N Emissions and Soil Organic C Stock Changes from O Emissions, CH 4 2 Agricultural Lands (Cropland and Grassland) Land and 3.13. Methodology for Estimating Net Carbon Stock Changes in Forest Land Remaining Forest Land Converted to Forest Land Emissions from Landfills Methodology for Estimating CH 3.14. 4 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 24 1

78 ANNEX 4 IPCC Reference Approach for Estimating CO Emissions from Fossil Fuel Combustion 2 ANNEX 5 Assessment of the Sources and Sinks of Greenhouse Gas Emissions Not Included ANNEX 6 Additional Information Global Warming Potential Values 6.1. 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 Chemical Formulas 6.7. ANNEX 7 Uncertainty 7.1. Overview 7.2. Methodology and Results 7.3. Reducing Uncertainty Improvements 4 . Planned 7. 7.5. Additional Information on Uncertainty Analyses by Source ANNEX 8 QA/QC Procedures Background 8.1. 8.2. Purpose Assessment Factors 8.3. 25 - 1 Introduction

79 2. Trends in Greenhouse Gas Emissions Recent Trends in U.S. Greenhouse Gas 2.1 Emissions and Sinks , 5 , total gross U.S. greenhouse gas emissions were 6 , 586.7 MMT ) or million metric tons , carbon dioxide (CO In 201 2 to Total U.S. emissions have increased by 3.5 percent from 1990 to 201 Eq. , and emissions de creased fro m 201 4 5 The decrease in total greenhouse gas emissions between 2014 and 2015 201 5 by 2.3 percent ( 153.0 MMT CO Eq.). 2 The emissions in CO decrease was driven in large part by a decrease in CO emissions from fossil fuel combustion. 2 2 from fossil fuel combustion was a result of multiple factors, including: ( 1 ) substitution from coal to natural gas consumption in the electric power sector ; ( 2 ) warmer winter conditions in 201 5 resulting in a de creased demand for ( heating fuel in the residential and commercial sectors ; and 3 ) a slight decrease in electricity demand. Since 1990, illustrate the 3 0.2 - 2 U.S. emissions have increased at an average annual rate of Figure percent. Figure 2 - 1 through since 1990. Overall, net emissions overall trend in total U.S. emissions by gas, annual changes, and absolute changes 11.5 in 201 5 were percent below 2005 levels as shown in Table 2 - 1 . Eq.) Gross 1 : Figure 2 - U.S. Greenhouse Gas Emissions by Gas (MMT CO 2 1 - 2 Trends

80 Figure 2 : Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the 2 - Previous Year 2 - 3 : Cumulative Change in Annual Gross U.S. Greenhouse Gas Emissions Relative to Figure Eq.) 1990 (1990=0, MMT CO 2 Overall, from 1990 to 201 5 , total emissions of CO percent), while total increased by 288.4 MMT CO Eq. ( 5.6 2 2 emissions of methane (CH ) decreased by 125.1 percent), and total emissions of nitrous oxide MMT CO 16.0 Eq. ( 2 4 During the same period, aggregate weighted emissions of (N O) decreased by 24.7 MMT CO percent). Eq. ( 6.9 2 2 hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SF ), and nitrogen trifluoride (NF ) rose 6 3 by Eq. ( 85.0 MMT CO percent). Despite being emitted in s maller quantities relative to the other principal 85.3 2 greenhouse gases, emissions of HFCs, PFCs, SF , and NF are significant because many of them have extremely 3 6 high global warming potentials (GWPs), and, in the cases of PFCs, SF , and NF , long atmospheric l ifetimes. 6 3 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. percent of total emissions in 201 5 8 As the largest contributor to U.S. greenhouse gas emissions, CO from fossil fuel combustion has accounted for 2 approximately 77 percent of GWP - weighted emissions for the entire time series since 1990. Emissions from this 5 and were responsible for most of the source category grew by 6.5 percent ( 309.4 MMT CO Eq.) from 1990 to 201 2 CO In addition, increase in national emissions during this period. emissions from fossil fuel combustion decreased 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 2 2

81 from 2005 MMT CO levels by Eq., a decrease of approximately 12 .1 percent between 2005 and 2015. From 697.2 2 4 to 201 5 , these emissions de creased by 2.9 percent ( 152.5 MMT CO 201 Eq.). Historically, changes in emissions 2 nant factor affecting U.S. emission trends. from fossil fuel combustion have been the domi emissions from fossil fuel combustion are influenced by many long Changes in CO term and short - term factors, - 2 and market trends hanges, including population and economic growth, energy price fluctuations , technological c On an annual basis, the overall consumption of fossil fuels energy fuel choices, and seasonal temperatures. and mix in the United States fluctuates primarily in response to changes in general economic conditions, overall energy the relative price of different fuels , weather, and the availability of non - fossil alternatives. For example, coal prices, consumption for electricity generation is influenced by the relative price of coal and a number of factors including , the ability to . Likewise, warmer winters alternative sources switch fuels, and longer terms trends in coal markets and electricity for result in a decreased demand for heating fuel will lead to a decrease in heating degree days and in the residential and commercial sector, which leads to a decrease in emissions from reduced fuel use. heat Energy - emissions also depend on the type of fuel or energy consumed and its C intensity. Producing a related CO 2 the CO unit of heat or electricity using natural gas instead of coal, for example, can reduce emissions because of the 2 lower C content of natural gas (see Table A - 39 in Annex 2.1 for more detail on the C Content Coefficient of different fossil fuels). A brief discussion of the year to year variability in fuel combustion emissions is provi ded below, beginning with 201 1 . Recent trends in CO emissions from fossil fuel combustion show a 3.9 percent decrease from 2011 to 2012, then a 2 2.6 percent and a 0.9 percent increase from 2012 to 2013 and 2013 to 2014, respectively, and a 2.9 percent decr ease from 2014 to 2015. Total electricity 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 by roughly 12 percent from 2011 to 2012, increased by 4 percent from 2012 to 2013, generate electricity dec reased stayed relatively flat from 2013 to 2014, and decreased by 14 percent from 2014 to 2015. The overall CO emission 2 eating trends from fossil fuel combustion also follow close ly changes in heating degree days over that time period. H 13 percent from 2011 to 2012, increased by 18 percent from 2012 to 2013, increased by 2 by degree days decreased CO percent from 2013 to 2014, and decreased by 10 percent from 2014 to 2015. The emission trends from overall 2 fossil fuel combustion also generally follow changes in overall petroleum use and emissions. Carbon dioxide emissions from petroleum decreased by 2.0 percent from 2011 to 2012, increased by 1.6 percent from 2012 to 2013, in creased by 0.8 percent from 2013 to 2014, and increased by 1.7 percent from 2014 to 2015. The increase in petroleum CO emissions from 2014 to 2015 somewhat offset emission reductions from decreased coal use in the 2 electricity sector from 2014 to 2015. 2 - 1 summarizes emissions and sinks from all U.S. anthropogenic sources in weighted units of MMT CO Eq., Table 2 while unweighted gas emissions and sinks i n kilotons (kt) are provided in Table 2 - 2 . Table 2 - 1 : Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO Eq.) 2 1990 2005 2011 2012 2013 2014 201 5 Gas/Source 5,569.5 5,123.0 6,131.8 CO 5,362.1 5,514.0 5,565.5 5,411.4 ₂ 5,746.9 5,049.8 Fossil Fuel Combustion 4,740.3 5,202.3 5,227.1 5,024.6 5,156.5 2,400.9 2,157.7 2,022.2 1,820.8 2,038.1 2,038.0 1,900.7 Electricity Generation a 1,887.0 1,707.6 1,696.8 Transportation 1,713.0 1,742.8 1,736.4 1,493.8 a 842.5 828.0 775.0 782.9 812.2 806.1 805.5 Industrial 338.3 357.8 325.5 282.5 329.7 345.4 319.6 Residential a Commercial 223.5 220.4 217.4 221.0 228.7 246.2 196.7 41.4 U.S. Territories 27.6 49.7 40.9 43.5 42.5 41.4 117.6 138.9 109.8 Energy Use of Fuels 123.6 119.0 125.5 106.7 Non - Iron and Steel Production & 101.5 68.0 61.1 55.4 53.3 58.6 48.9 Metallurgical Coke Production Natural Gas Systems 37.7 30.1 35.7 35.2 38.5 42.4 42.4 35.3 46.2 32.2 Cement Production 33.5 36.4 39.4 39.9 28.1 Petrochemical Production 21.3 27.0 26.3 26.5 26.4 26.5 14.6 14.2 13.3 Lime Production 11.7 14.0 13.8 14.0 4.9 6.3 9.3 Other Process Uses of Carbonates 8.0 10.4 11.8 11.2 3 - 2 Trends

82 9.2 9.3 9.4 10.0 9.6 10.8 Ammonia Production 13.0 8.0 12.5 10.6 10.4 10.4 10.6 10.7 Incineration of Waste 3.5 4.1 4.3 4.5 4.8 5.0 2.4 Urea Fertilization 1.4 4.1 4.0 4.2 4.5 4.3 Carbon Dioxide Consumption 1.5 4.3 3.9 6.0 3.9 3.6 3.8 4.7 Liming 3.9 4.2 3.9 3.7 3.6 3.6 Petroleum Systems 3.6 Soda Ash Production and 3.0 2.7 2.8 2.8 2.8 2.8 Consumption 2.8 6.8 4.1 3.3 3.4 3.3 2.8 2.8 Aluminum Production 2.2 Ferroalloy Production 1.4 1.7 1.9 1.8 1.9 2.0 1.8 1.7 1.5 Titanium Dioxide Production 1.7 1.6 1.7 1.2 1.9 1.3 1.2 1.5 1.3 1.3 1.3 Glass Production Urea Consumption for Non - 3.8 3.7 4.0 4.4 4.0 1.4 1.1 Agricultural Purposes Phosphoric Acid Production 1.3 1.2 1.1 1.1 1.0 1.0 1.5 0.6 1.0 1.3 1.4 1.0 0.9 Zinc Production 1.5 Lead Production 0.6 0.5 0.5 0.5 0.5 0.5 0.5 Silicon Carbide Production and 0.2 0.2 0.2 0.4 0.2 0.2 Consumption 0.2 Magnesium Production and + + + + + + + Processing Wood Biomass, Ethanol, and b 219.4 230.7 276.4 Biodiesel Consumption 276.2 299.8 307.1 291.7 c 99.8 International Bunker Fuels 113.1 111.7 105.8 103.2 110.8 103.5 655.7 780.8 680.9 672.1 666.1 658.8 659.1 CH ₄ 164.2 168.9 168.9 166.7 165.5 164.2 166.5 Enteric Fermentation Natural Gas Systems 194.1 159.7 154.5 156.2 159.2 162.5 162.4 134.3 119.0 120.8 179.6 116.6 115.7 Landfills 116.7 37.2 56.3 63.0 65.6 63.3 62.9 66.3 Manure Management 64.1 71.2 66.5 64.6 64.8 60.9 Coal Mining 96.5 55.5 46.0 48.0 46.4 44.5 43.0 39.9 Petroleum Systems 15.7 16.0 15.3 15.1 14.9 14.8 14.8 Wastewater Treatment 16.0 16.7 14.1 11.3 11.3 11.4 11.2 Rice Cultivation 6.6 7.4 7.1 Stationary Combustion 8.5 8.0 8.1 7.0 Abandoned Underground Coal Mines 7.2 6.6 6.4 6.2 6.2 6.3 6.4 1.9 2.1 2.1 Composting 0.4 1.9 1.9 2.0 a 5.6 2.8 2.3 2.2 2.1 2.1 2.0 Mobile Combustion Field Burning of Agricultural 0.2 0.3 0.3 0.3 0.3 Residues 0.2 0.3 0.2 0.1 + 0.1 0.1 0.1 0.2 Petrochemical Production + + + Ferroalloy Production + + + + Silicon Carbide Production and Consumption + + + + + + + Iron and Steel Production & + + + Metallurgical Coke Production + + + + + + Incineration of Waste + + + + + c 0.1 0.1 0.1 0.2 0.1 0.1 0.1 International Bunker Fuels 334.8 N O 335.5 359.5 361.6 364.0 340.7 335.5 ₂ 256.6 259.8 270.1 254.1 250.5 250.0 251.3 Agricultural Soil Management 11.9 20.2 21.3 21.4 Stationary Combustion 22.9 23.4 23.1 14.0 16.5 17.4 Manure Management 17.5 17.5 17.7 17.5 a 22.8 41.2 35.7 Mobile Combustion 20.4 18.5 16.6 15.1 12.1 11.3 10.9 10.5 10.7 10.9 11.6 Nitric Acid Production 3.4 4.4 4.8 4.8 4.9 4.9 5.0 Wastewater Treatment 5.5 7.1 10.2 Adipic Acid Production 15.2 3.9 5.4 4.3 4.2 4.2 N 4.2 O from Product Uses 4.2 4.2 4.2 4.2 ₂ 1.8 1.9 1.9 Composting 0.3 1.7 1.7 1.7 0.4 0.5 0.3 Incineration of Waste 0.3 0.3 0.3 0.3 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 4 - 5 2

83 + 0.1 0.2 0.2 0.2 0.2 0.2 Semiconductor Manufacture Field Burning of Agricultural Residues 0.1 0.1 0.1 0.1 0.1 0.1 0.1 c 1.0 1.0 0.9 0.9 0.9 0.9 International Bunker Fuels 0.9 166.7 120.0 154.3 155.9 159.0 173.2 HFCs 46.6 Substitution of Ozone Depleting d 99.7 145.3 150.2 Substances 154.6 161.3 168.5 0.3 HCFC 46.1 - 20.0 22 Production 8.8 5.5 4.1 5.0 4.3 0.2 0.2 0.2 0.2 0.3 0.3 0.2 Semiconductor Manufacture Magnesium Production and 0.0 + + 0.1 0.1 0.1 Processing 0.0 24.3 6.7 6.9 6.0 5.8 5.8 5.2 PFCs 2.8 3.2 3.4 3.0 2.8 3.2 3.2 Semiconductor Manufacture 3.4 3.5 2.9 3.0 2.5 2.0 Aluminum Production 21.5 Substitution of Ozone Depleting Substances + + + + + + 0.0 11.7 9.2 28.8 6.4 6.6 5.8 6.8 SF ₆ Electrical Transmission and Distribution 8.3 6.0 4.8 4.6 4.8 4.2 23.1 Magnesium Production and 2.7 2.8 1.6 1.5 1.0 0.9 Processing 5.2 0.5 0.7 0.4 Semiconductor Manufacture 0.4 0.7 0.7 0.4 0.6 0.5 0.7 + 0.6 0.5 0.6 NF ₃ 0.7 + 0.5 0.6 0.6 0.5 0.6 Semiconductor Manufacture 7,313.3 6,776.7 6,538.3 6,680.1 6,739.7 6,586.7 Total Emissions 6,363.1 e 23.0 19.9 26.1 19.2 19.7 19.7 LULUCF Emissions 10.6 f (779.8) (830.2) LULUCF (754.0) (769.1) Stock Change (782.2) (781.1) (778.7) Carbon g (753.8) (819.6) (731.0) (749.2) LULUCF Sector (763.0) (761.4) (758.9) Net Total 6,582.3 6,027.6 5,784.5 5,917.1 5,543.5 5,978.3 5,827.7 Net Emissions (Sources and Sinks) 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 CO Eq. 2 a a method update in this Inventory for estimating the share of gasoline used in on - road and non - road There was applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline or to industrial and commercial sectors for 2015, creating a break in the consumption from the transportation sect Planned Improvements section of Chapter 3.1 . the time series. The change is discussed further in b Emissions from Wood Biomass Ethanol Consumption are not included specifically in summing , and Biodiesel , Net carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the Energy sector totals. estimates for LULUCF. c from International Bunker Fuels are not included in totals. Emissions d Small amounts of PFC emissions also result from this source. e LULUCF emissions include the CH O emissions reported for Peatlands Remaining Peatlands , Forest Fires, and N 2 4 ; CH emissions from Drained Organic Soils Coastal Wetlands Remaining Coastal Wetlands , Grassland Fires, and 4 Land Converted to Coastal Wetlands ; and N O emissions from Forest Soils and Settlement Soils. 2 f LULUCF Carbon Stock Change ategories: Forest Land Remaining is the net C stock change from the following c Land Converted to Forest Land Cropland Remaining Cropland , Land Converted to Cropland , , Forest Land , Land Converted to Grassland , Grassland Remaining Grassland , Land Converted to , Wetlands Remaining Wetlands , , and Land Converted to Settlements . Refer to Table 2 - 8 for a Wetlands Settlements Remaining Settlements d source category . LULUCF breakout of emissions and removals for by gas an g and N The LULUCF Sector Net Total is the net sum of all CH O emissions to the atmosphere plus net carbon 2 4 stock changes. Table - 2 : Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (kt) 2 Gas/Source 1990 2005 2011 2012 2013 2014 201 5 CO 5,411,409 5,123,043 6,131,833 5,569,516 5,362,095 5,514,018 5,565,495 ₂ Fossil Fuel Combustion 4,740,343 5,746,942 5,227,061 5,024,643 5,156,523 5,202,300 5,049,763 1,820,818 2,400,874 2,157,688 Electricity Generation 2,022,181 2,038,122 2,038,018 1,900,673 a 1,742,814 1,713,002 Transportation 1,736,383 1,493,758 1,887,033 1,707,631 1,696,752 5 - 2 Trends

84 a 842,473 827,999 774,951 782,929 812,228 806,075 805,496 Industrial Residential 357,834 325,537 282,540 329,674 345,362 319,591 338,347 a 223,480 220,381 196,714 217,393 221,030 228,666 246,241 Commercial 27,555 49,723 40,874 43,527 41,365 41,380 U.S. Territories 42,467 Non Energy Use of Fuels 117,585 138,913 109,756 - 123,645 118,995 125,526 106,750 Iron and Steel Production & Metallurgical Coke 101,487 68,047 61,108 55,449 53,348 58,629 48,876 Production 37,732 30,076 35,662 35,203 38,457 42,351 42,351 Natural Gas Systems 35,270 Cement Production 46,194 32,208 33,484 36,369 39,439 39,907 Petrochemical Production 21,326 26,972 26,338 26,501 26,395 26,496 28,062 Lime Production 11,700 14,552 13,982 13,785 14,028 14,210 13,342 Other Process Uses of 8,022 6,339 9,335 Carbonates 4,907 10,414 11,811 11,236 Ammonia Production 13,047 9,196 9,292 9,377 9,962 9,619 10,799 Incineration of Waste 7,950 12,469 10,564 10,379 10,398 10,608 10,676 4,267 Urea Fertilization 3,504 4,097 2,417 4,504 4,781 5,032 Carbon Dioxide Consumption 1,472 1,375 4,083 4,019 4,188 4,471 4,296 Liming 4,667 4,349 3,873 5,978 3,907 3,609 3,810 3,876 Petroleum Systems 3,553 3,927 4,192 3,693 3,567 3,567 Soda Ash Production and Consumption 2,822 2,960 2,712 2,763 2,804 2,827 2,789 Aluminum Production 6,831 4,142 3,292 3,439 3,255 2,833 2,767 2,152 1,392 1,735 1,903 1,785 1,914 1,960 Ferroalloy Production Titanium Dioxide Production 1,195 1,755 1,729 1,528 1,715 1,688 1,635 1,535 1,928 1,299 1,248 1,317 1,336 1,299 Glass Production Urea Consumption for Non - Agricultural Purposes 3,784 3,653 4,030 4,407 4,014 1,380 1,128 1,342 Phosphoric Acid Production 1,529 1,171 1,118 1,149 1,038 999 Zinc Production 632 1,030 1,286 1,486 1,429 956 933 Lead Production 553 538 527 546 459 473 516 Production and Silicon Carbide 375 219 170 158 169 173 180 Consumption Magnesium Production and 3 3 2 Processing 2 2 3 1 Wood Biomass, Ethanol, and b 219,413 230,700 276,413 276,201 299,785 307,079 291,735 Biodiesel Consumption c International Bunker Fuels 103,463 113,139 111,660 105,805 99,763 103,201 110,751 CH 31,232 27,238 26,884 26,643 26,351 26,366 26,229 ₄ Enteric Fermentation 6,566 6,755 6,757 6,670 6,619 6,567 6,661 Natural Gas Systems 7,762 6,387 6,180 6,247 6,368 6,501 6,497 7,182 5,372 4,760 4,834 4,669 4,663 4,628 Landfills Manure Management 1,486 2,254 2,519 2,625 2,530 2,514 2,651 Coal Mining 3,860 2,565 2,849 2,658 2,584 2,593 2,436 1,858 Petroleum Systems 1,840 1,922 2,218 1,778 1,721 1,595 Wastewater Treatment 627 639 613 604 597 592 591 Rice Cultivation 641 667 564 453 454 456 449 265 Stationary Combustion 339 296 283 320 323 280 Abandoned Underground Coal Mines 288 264 257 249 249 253 256 Composting 15 75 75 77 81 84 84 a Mobile Combustion 226 113 91 85 82 80 87 Field Burning of Agricultural Residues 9 8 11 11 11 11 11 9 3 2 Production 3 3 5 7 Petrochemical Ferroalloy Production 1 + + 1 + 1 1 Silicon Carbide Production and + + + + Consumption 1 + + - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 6 2

85 Iron and Steel Production & Metallurgical Coke 1 + + + + + 1 Production + + + + + + Incineration of Waste + c International Bunker Fuels 5 5 4 3 3 3 7 O N 1,214 1,222 1,143 1,126 1,126 1,124 1,207 ₂ Agricultural Soil Management 861 872 906 853 841 839 843 Stationary Combustion 68 71 72 77 78 78 40 59 Manure Management 55 58 47 59 59 59 a Mobile Combustion 120 77 68 62 56 51 138 Nitric Acid Production 41 38 37 35 36 37 39 Wastewater Treatment 11 15 16 16 16 16 17 Adipic Acid Production 51 24 34 19 13 18 14 N ₂ O from Product Uses 14 14 14 14 14 14 14 Composting 1 6 6 6 6 6 6 2 1 1 Incineration of Waste 1 1 1 1 Semiconductor Manufacture + + 1 1 1 1 1 Field Burning of Agricultural + + + Residues + + + + c International Bunker Fuels 3 3 3 3 3 3 3 M M M M HFCs M M M Substitution of Ozone d Depleting Substances M M M M M M M HCFC + + - 22 Production 3 1 1 + + Semiconductor Manufacture + + + + + + + Magnesium Production and 0 Processing 0 + + + + + M M PFCs M M M M M M M M M M M M Semiconductor Manufacture Aluminum Production M M M M M M M Substitution of Ozone + + + Depleting Substances 0 + + + SF ₆ 1 1 + + + + + Transmission and Electrical Distribution 1 + + + + + + Magnesium Production and + Processing + + + + + + Semiconductor Manufacture + + + + + + + + + + + + + + NF ₃ Semiconductor Manufacture + + + + + + + + Does not exceed 0.5 kt. - Mixture of multiple gases M a - There was a method update in this Inventory - road applications. The for estimating the share of gasoline used in on road and non change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the tors for 2015, creating a break in the time series. The change is discussed transportation sector to industrial and commercial sec section of Chapter 3.1 . further in the Planned Improvements b , , and Biodiesel Consumption are not included specifically in summing Energy sector Emissions from Wood Biomass Ethanol are accounted for in the estimates for LULUCF. totals. Net carbon fluxes from changes in biogenic carbon reservoirs c Emissions from International Bunker Fuels are not included in totals. d Small amounts of PFC emissions also result from this source. Totals may not sum due to independent rounding. Parentheses indicat e negative values or sequestration. Notes: 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 - six year period of 1990 to 201 , total emissions in the Energy, Industrial Processes and Product Use, and Agriculture 5 sectors grew by 221.0 MMT CO 5.5 Eq. ( 4 .1 percent), 35.5 MMT CO Eq. ( Eq. ( 10.4 percent), and 27.0 MMT CO 2 2 2 Over the percent). percent), respectively. Emissions from the Waste sector decreased by 59.9 MMT CO Eq. ( 30.1 2 same period, estimates of net C sequestration for the Land Use, Land Use Change, and Forestry sector (magnitude - 7 - 2 Trends

86 Eq. ( of emissions plus CO 60.7 MMT CO removals from all LULUCF categories) increased by 7.4 percent 2 2 ). decrease in net C sequestration - 4 2 Figure : U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO 2 Eq.) Table : Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC 2 - 3 Sector (MMT CO Eq.) 2 1990 2005 2011 Chapter/IPCC Sector 2013 2014 201 5 2012 Energy 5,328.1 6,275.3 5,721.2 5,507.0 5,659.1 5,704.9 5,549.1 5,024.6 5,746.9 5,227.1 4,740.3 5,156.5 5,202.3 5,049.8 Fossil Fuel Combustion Natural Gas Systems 231.8 189.8 190.2 191.4 197.7 204.9 204.8 Non - 117.6 138.9 109.8 106.7 123.6 119.0 125.5 Energy Use of Fuels 66.5 64.1 71.2 96.5 64.6 64.8 60.9 Coal Mining Petroleum Systems 59.0 49.9 52.2 50.3 48.2 46.6 43.4 Stationary Combustion 27.6 28.4 28.0 20.4 30.9 31.5 30.1 a 46.9 38.6 25.1 22.6 20.6 18.6 17.1 Mobile Combustion 12.9 Incineration of Waste 8.4 10.9 10.7 10.7 10.9 11.0 Abandoned Underground Coal Mines 6.6 6.4 6.2 6.2 6.3 6.4 7.2 353.4 371.0 360.9 363.7 379.8 375.9 Industrial Processes and Product Use 340.4 Substitution of Ozone Depleting 0.3 99.8 145.4 150.2 154.7 161.3 168.5 Substances Iron and Steel Production & Metallurgical Coke Production 101.5 68.1 61.1 55.5 53.4 58.6 48.9 35.3 46.2 32.2 33.5 36.4 39.4 39.9 Cement Production Petrochemical Production 21.5 27.0 26.4 26.6 26.5 26.6 28.2 Lime Production 14.6 14.0 13.8 14.0 14.2 13.3 11.7 10.5 Nitric Acid Production 11.3 10.9 12.1 10.7 10.9 11.6 Other Process Uses of Carbonates 4.9 6.3 9.3 8.0 10.4 11.8 11.2 9.4 Ammonia Production 9.2 9.3 13.0 10.0 9.6 10.8 3.6 4.7 4.9 4.5 4.1 5.0 5.0 Semiconductor Manufacture 5.4 6.2 6.4 4.8 Aluminum Production 28.3 7.6 6.8 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 8 2

87 Carbon Dioxide Consumption 1.4 4.1 4.0 4.2 4.5 4.3 1.5 - 22 Production 46.1 20.0 8.8 5.5 4.1 5.0 4.3 HCFC 15.2 7.1 10.2 5.5 3.9 5.4 4.3 Adipic Acid Production O from Product Uses N 4.2 4.2 4.2 4.2 4.2 4.2 4.2 ₂ Electrical Transmission and 23.1 8.3 6.0 4.8 4.6 4.8 4.2 Distribution Soda Ash Production and Consumption 2.8 3.0 2.7 2.8 2.8 2.8 2.8 1.4 1.7 1.9 1.8 1.9 2.0 Ferroalloy Production 2.2 1.2 1.8 1.7 1.5 Titanium Dioxide Production 1.7 1.7 1.6 Glass Production 1.5 1.9 1.3 1.2 1.3 1.3 1.3 Urea Consumption for Non - Agricultural Purposes 3.8 3.7 4.0 4.4 4.0 1.4 1.1 Magnesium Production and Processing 5.2 2.7 2.8 1.7 1.5 1.1 1.0 1.3 1.2 1.1 1.1 1.0 1.0 Phosphoric Acid Production 1.5 0.6 1.0 1.3 Zinc Production 1.4 1.0 0.9 1.5 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 526.4 541.9 525.9 516.9 514.7 522.3 495.3 254.1 Agricultural Soil Management 259.8 270.1 256.6 250.5 250.0 251.3 Enteric Fermentation 164.2 168.9 168.9 166.7 165.5 164.2 166.5 Manure Management 51.1 72.9 80.4 83.2 80.8 80.4 84.0 Rice Cultivation 16.0 16.7 14.1 11.3 11.3 11.4 11.2 Urea Fertilization 3.5 4.1 4.3 4.5 4.8 5.0 2.4 3.9 4.7 4.3 3.9 6.0 3.6 3.8 Liming Field Burning of Agricultural 0.3 0.3 0.4 0.4 0.4 0.4 0.4 Residues 199.3 158.2 142.6 144.4 140.4 140.2 139.4 Waste 134.3 119.0 120.8 116.7 116.6 115.7 Landfills 179.6 19.1 20.4 20.1 19.9 19.8 19.7 19.7 Wastewater Treatment 3.5 3.5 3.7 3.9 4.0 4.0 0.7 Composting b 6,363.1 7,313.3 6,776.7 6,538.3 6,680.1 6,739.7 6,586.7 Total Emissions - Change, and Land Use, Land Use (761.4) (758.9) (749.2) (753.8) (763.0) Forestry (819.6) (731.0) (728.7) Forest Land (731.8) (733.8) (723.6) (733.5) (729.8) (784.3) 4.7 4.0 4.0 1.3 3.1 Cropland (0.7) 2.4 Grassland 0.4 0.9 9.9 0.8 0.4 25.3 13.8 Wetlands (4.0) (4.0) (3.9) (4.0) (4.1) (5.2) (3.9) (31.3) (30.4) (25.4) Settlements (28.3) (28.9) (20.5) (47.6) c 5,827.7 Net Emission (Sources and Sinks) 5,978.3 6,027.6 5,784.5 5,917.1 5,543.5 6,582.3 Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF. a There was a method update in this Inventory for estimating the share of gasoline used in on - road and non - road applications . in a shift in gasoline consumption from the The change does not impact total U.S. gasoline consumption. It mainly results transportation sector to industrial and commercial sectors for 2015, creating a break in the time series. The change is Planned Improvements section of Chapter 3.1 . discussed further in the b Total emissions without LULUCF. c Net emissions with LULUCF. Totals may not sum due to independent rounding. Parentheses indicate negative values or sequestration. Notes: Energy - related activities, primarily fossil fuel combustion, accounted for the vast majority of U.S. CO emissions for Energy 2 the period of 1990 through 201 5 . Emissions from fossil fuel combustion comprise the vast majority of energy - Figure related emissions, with CO b eing the primary gas emitted (see ). 2 - 5 Due to their relative importance, fossil 2 6 - , 5 2 In 201 ). fuel combustion - related CO Figure emissions are considered in detail in the Energy chapter (see 2 9 - 2 Trends

88 percent of the energy consumed in the United States (on a Btu basis) was produced through the 82 approximately The remaining 18 combustion of fossil fuels. percent came from other energy sources such as hydropower, biomass, nuclear, wind, and solar energy. as well as other greenhouse gas A discussion of specific trends related to CO 2 emissions from energy consumption is presented in the Energy chapter. - esponsible related activities are also r Energy Table 2 percent of total U.S. emissions of each gas, respectively). - 4 for CH and N O emissions ( 42 percent and 12 2 4 presents greenhouse gas emissions from the Energy chapter, by source and gas. 5 Eq.) : 201 5 - Energy Chapter Greenhouse Gas Sources (MMT CO Figure 2 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 10 2

89 Figure 2 : 201 5 U.S. Fossil Carbon Flows (MMT CO - Eq.) 6 2 4 : Emissions from Energy (MMT CO Table Eq.) - 2 2 Gas/Source 2005 2011 201 2 2013 2014 201 5 1990 CO 2 5,932.3 5,377.8 5,387.2 5,180.9 5,332.7 4,907.2 5,231.9 Fossil Fuel Combustion 5,746.9 5,227.1 5,024.6 5,156.5 5,202.3 5,049.8 4,740.3 Electricity Generation 2,400.9 2,157.7 2,022.2 2,038.1 1,820.8 2,038.0 1,900.7 a Transportation 1,887.0 1,707.6 1,696.8 1,742.8 1,736.4 1,493.8 1,713.0 a Industrial 828.0 806.1 775.0 782.9 812.2 842.5 805.5 Residential 357.8 338.3 325.5 282.5 329.7 345.4 319.6 a Commercial 223.5 220.4 196.7 221.0 228.7 246.2 217.4 U.S. Territories 49.7 27.6 40.9 43.5 42.5 41.4 41.4 Non - Energy Use of Fuels 138.9 109.8 106.7 117.6 123.6 119.0 125.5 Natural Gas Systems 30.1 37.7 35.7 35.2 38.5 42.4 42.4 Incineration of Waste 12.5 10.6 10.4 10.4 8.0 10.6 10.7 Petroleum Systems 3.9 3.6 4.2 3.9 3.7 3.6 3.6 b Biomass - Wood 206.9 195.2 194.9 211.6 215.2 217.7 198.7 c International Bunker Fuels 103.5 111.7 105.8 99.8 103.2 110.8 113.1 b Biofuels - Ethanol 22.9 72.9 72.8 74.7 76.1 78.9 4.2 b Biodiesel Biofuels - 0.9 0.0 8.3 8.5 13.3 14.1 13.5 CH 4 286.6 284.1 289.5 367.3 284.6 286.8 278.6 Natural Gas Systems 162.5 154.5 156.2 159.2 162.4 194.1 159.7 Petroleum Systems 64.1 71.2 66.5 96.5 64.6 64.8 60.9 Coal Mining 48.0 46.4 55.5 44.5 46.0 43.0 39.9 Stationary Combustion 7.4 8.0 7.1 6.6 8.5 8.1 7.0 Abandoned Underground Coal Mines 6.6 7.2 6.4 6.2 6.2 6.3 6.4 a Mobile Combustion 2.8 2.3 2.2 5.6 2.1 2.0 2.1 Incineration of Waste + + + + + + + c International Bunker Fuels 0.1 0.1 0.1 0.2 0.1 0.1 0.1 N O 2 56.4 53.6 44.4 41.7 40.3 38.6 42.1 Stationary Combustion 20.2 21.4 21.3 11.9 22.9 23.4 23.1 a Mobile Combustion 35.7 20.4 22.8 41.2 18.5 16.6 15.1 Incineration of Waste 0.3 0.3 0.5 0.4 0.3 0.3 0.3 11 - 2 Trends

90 c International Bunker Fuels 1.0 1.0 0.9 0.9 0.9 0.9 0.9 Total 6,275.3 5,507.0 5,721.2 5,328.1 5,659.1 5,704.9 5,549.1 + Does not exceed 0.05 MMT CO Eq. 2 a - road and non - There was a method update in this Inventory for estimating the share of gasoline used in on road applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the tors for 2015, creating a break in the time series. The change is discussed transportation sector to industrial and commercial sec 3.1 section of Chapter . further in the Planned Improvements b Biofuel Consumption are not included specifically in summing energy sector totals. Net Emissions from Wood Biomas s and carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF. c Emissions from International Bunker Fuels are not include d in totals. Totals may not sum due to independent rounding. Note: Table 2 - 5 based on the underlying U.S. Carbon dioxide emissions from fossil fuel combustion are presented in emissions Estimates of CO energy consumer data collected by the U.S. Energy Information Administration (EIA). 2 from fossil fuel combustion are calculated from the - use sectors” based on total consumption and se EIA “end appropriate fuel properties (any additional analysis and refinement of the EIA data is further explained in the Energy EIA’s fuel consumption data for the electric power secto r are comprised of electricity - only chapter of this report). and combined - - and - power (CHP) plants within the North American Industry Classification System (NAICS) 22 heat r category whose primary business is to sell electricity, or electricity and heat, to the public (nonutility powe producers can be included in this sector as long as they meet they electric power sector definition). EIA statistics for the industrial sector include fossil fuel consumption that occurs in the fields of manufacturing, agriculture, mining, and constructi on. 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. EIA’s fuel consumption data for the industrial sector consists of all fa cilities 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 EIA’s fuel consumption data for th activities in this sector). e residential sector consist of living quarters for private EIA’s fuel consumption data for the commercial sector consist of service - providing facilities and households. s that produce electricity equipment from private and public organizations and businesses (EIA includes generator and/or useful thermal output primarily to support the activities at commercial establishments in this sector). Table - 5 and Figure 2 - 7 summarize CO 2 emissions from fossil fuel combustion by end - use sector. Figure 2 - 8 further 2 describes the total emissions from fossil fuel combustion, separated by end use sector, including CH O in and N - 4 2 addition to CO . 2 Table 2 - 5 : CO E missions from Fossil Fuel Combustion by End - Use Sector (MMT CO Eq.) 2 2 End - Use Sector 1990 2005 2011 2012 2013 2014 201 5 a Transportation 1,740.1 1,746.9 1,711.9 1,700.6 1,717.0 1,891.8 1,496.8 1,696.8 1,713.0 1,736.4 1,742.8 Combustion 1,707.6 1,887.0 1,493.8 Electricity 3.7 4.1 4.3 3.9 4.0 4.7 3.0 a Industrial 1,355.0 1,399.3 1,399.6 1,375.7 1,407.0 1,529.2 1,564.6 805.5 806.1 775.0 Combustion 812.2 782.9 828.0 842.5 Electricity 549.6 593.2 624.7 592.8 594.7 686.7 736.6 Residential 1,003.9 1,080.1 1,116.2 1,007.8 1,064.6 931.4 1,214.1 Combustion 319.6 345.4 325.5 282.5 329.7 357.8 338.3 Electricity 684.3 734.7 790.7 725.3 734.9 856.3 593.0 a Commercial 909.4 934.7 958.4 897.0 925.5 755.4 1,026.8 246.2 228.7 Combustion 220.4 196.7 221.0 217.4 223.5 Electricity 663.1 706.0 738.0 700.3 704.5 538.0 803.3 b 41.4 41.4 40.9 43.5 42.5 U.S. Territories 27.6 49.7 Total 5,024.6 5,156.5 5,202.3 5,049.8 5,227.1 5,746.9 4,740.3 Electricity Generation 1,900.7 2,038.0 2,157.7 2,022.2 2,038.1 2,400.9 1,820.8 a - for estimating the share of gasoline used in on There was a method update in this Inventory road - road and non applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in tors for 2015, creating a gasoline consumption from the transportation sector to industrial and commercial sec . break in the time series. The change is discussed further in the Planned Improvements section of Chapter 3.1 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 12 2

91 b 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 electricity generation are allocated based on aggregate national - electricity consumption by each end - use sector. Totals may not sum due to independent rounding. 2 - 7 : 201 5 CO Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT Figure 2 Eq.) CO 2 2 - 8 : 201 5 Figure End - Use Sector Emissions of CO from Fossil Fuel Combustion (MMT CO 2 2 Eq.) The main driver of emissions in the Energy sector is CO from fossil fuel combustion. Electricity generation is the 2 largest emitter of CO , and e lectricity generators consumed 34 percent of U.S. energy from fossil fuels and emitted 2 38 percent of t he CO hanges in electricity demand and the carbon intensity from fossil fuel combustion in 201 5 . C 2 emissions. While emissions from the of fuels used for electricity generation have a significant impact on CO 2 electric power sector have increased by approximately 4 percent since 1990, the carbon intensity of the electric power sector, in terms of CO E q. per QBtu has significantly decreased by 16 percent during that same timeframe. 2 - 2 Figure . 9 This decoupling of electricity generation and the resulting emissions is shown below in 13 - 2 Trends

92 Figure : Electricity Generation (Billion kWh) and Emissions (MMT CO - Eq.) 9 2 2 Electricity generation emissions can also be allocated to the end - use sectors that are consuming that electricity, as Table 2 - 5 . The tran sportation end - use sector accounted for 1 , 740.1 MMT CO presented in Eq. in 201 5 or 2 emissions from fossil fuel combustion. percent of total CO approximately 34 The industrial end - use sector 2 accounted for 27 percent of CO use emissions from fossil fuel combustion. The res idential and commercial end - 2 sectors accounted for and 18 percent, respectively, of CO Both of these 20 emissions from fossil fuel combustion. 2 end - use sectors were heavily reliant on electricity for meeting energy needs, with electricity consumption fo r lighting, heating, air conditioning, and operating appliances contributing 68 and 73 percent of emissions from the residential and commercial end use sectors, respectively. Significant trends in emissions from energy source - categories over the twenty - year period from 1990 through 201 5 included the following: six • Total CO 5,049.8 emissions from fossil fuel combustion increased from 4 , 740.3 MMT CO Eq. in 1990 to 2 2 percent total increase over the twenty 4 year period. , 5 to 201 MMT CO From 201 Eq. in 201 5 – a 6.5 six - 2 these emissions creased by 152.5 MMT CO Eq. ( 2.9 de percent). 2 • Methane emissions from n atural gas systems and petroleum systems (combined here) decreased from 249.5 MMT CO Eq. in 1990 to 202.3 MMT CO . Eq. ( 47.2 MMT CO 5 Eq. or 18.9 percent) from 1 990 to 201 2 2 2 Natural gas systems CH percent) since 1990, largely due emissions decreased by 31.6 MMT CO Eq. ( 16.3 2 4 to a decrease in emissions from transmission, storage, and distribution. The decrease in transmission and storage emissions is largely due t o reduced compressor station emissions (including emissions from compressors and fugitives). The decrease in distribution emissions is largely attributed to increased use of pgrades at metering and plastic piping, which has lower emissions than other pipe materials, and station u Petroleum systems CH emissions de creased by regulating (M&R) stations. MMT CO 28.1 Eq. (or 15.6 2 4 percent) since 1990. This de crease is due primarily to de creases in emissions from associated gas venting . 6.8 • Carbon dioxide emissions from non - energy uses of fossil fuels in creased by 7.9 MMT CO Eq. ( 2 5 . Emissions from non percent) from 1990 through 201 energy uses of fossil fuels were 125.5 MMT CO - 2 2.3 emissions, approximately the same percent of total national CO 5 Eq. in 201 , which constituted 2 proportion as in 1990. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 14 2

93 percent) from • MMT CO Nitrous oxide emissions from stationary combustion increased by Eq. ( 94.0 11.2 2 1990 through 201 5 . 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. 26.1 MMT CO • Eq. ( 63.3 Nitrous oxide emissions from mobile combustion decreased by percent) from 2 1990 through 201 5 O national emission contr ol standards and emission control , primarily as a result of N 2 technologies for on - road vehicles. MMT • Carbon dioxide emissions from incineration of waste ( 10.7 MMT CO Eq. in 201 5 ) increased by 2.7 2 percent) from 1990 through 201 - Eq. ( 34.3 CO 5 , as the volume of plastics and other foss il carbon 2 containing materials in municipal solid waste grew. ) The decrease in CO 1 emissions from fossil fuel combustion was a result of multiple factors, including: ( 2 2 warmer 5 in 201 substitution from coal to natural gas consumption in the electric power sector ; ( winter conditions ) resulting in a ; and ( 3 ) a creased demand for heating fuel in the residential and commercial sectors slight decrease de in electricity demand. Industrial Processes and Product Use The Industrial Processes and Product Use (IPP U) chapter includes greenhouse gas emissions occurring from industrial processes and from the use of greenhouse gases in products. Greenhouse gas emissions are produced as the by - products of many non - energy - related industrial activities. For example, indu strial processes can chemically transform raw materials, which often release waste gases such as CO , 2 CH , and , and N Industrial processes also release HFCs, PFCs, SF O. These processes are show n in Figure 2 - 10 . 2 6 4 and other fluorinated compounds NF In addition to the use of HFCs and some PFCs as substitutes for ozone . 3 depleting substances (ODS), fluorinated compounds such as HFCs, PFCs, SF employed and , NF , and others are 3 6 emitted by a number of other industrial sources in the United States. These industries include aluminum production, HCFC - 22 production, semiconductor manufacture, electric power transmission and distribution, and magnesium 2 presents greenhouse gas emissions from industrial processes by source - 6 metal production an d processing. Table category. 15 - 2 Trends

94 Figure 2 : 201 5 Industrial Processes and Product Use Chapter Greenhouse Gas Sources - 10 Eq.) (MMT CO 2 - Table Eq.) 6 2 : Emissions from Industrial Processes and Product Use (MMT CO 2 2013 2005 2011 2012 2014 201 5 Gas/Source 1990 CO 208.8 191.7 174.3 171.0 172.9 179.3 170.7 2 Iron and Steel Production & Metallurgical Coke Production 68.0 61.1 55.4 101.5 53.3 58.6 48.9 Iron and Steel Production 66.0 54.9 59.7 99.0 51.5 56.6 46.0 Metallurgical Coke Production 2.5 2.0 1.4 0.5 1.8 2.0 2.8 Cement Production 36.4 46.2 32.2 33.5 39.4 39.9 35.3 Petrochemical Production 26.4 26.3 21.3 27.0 26.5 28.1 26.5 Lime Production 14.6 11.7 14.0 13.8 14.0 14.2 13.3 Other Process Uses of Carbonates 6.3 9.3 4.9 8.0 10.4 11.8 11.2 Ammonia Production 9.2 9.3 13.0 9.4 10.0 9.6 10.8 Carbon Dioxide Consumption 1.5 4.1 1.4 4.0 4.2 4.5 4.3 Soda Ash Production and Consumption 2.7 2.8 2.8 3.0 2.8 2.8 2.8 Aluminum Production 4.1 3.3 3.3 3.4 6.8 2.8 2.8 Ferroalloy Production 1.4 1.8 1.7 1.9 2.2 1.9 2.0 Titanium Dioxide Production 1.8 1.2 1.7 1.5 1.7 1.7 1.6 Glass Production 1.9 1.3 1.2 1.3 1.5 1.3 1.3 Urea Consumption for Non - Agricultural Purposes 3.7 4.0 3.8 4.4 4.0 1.4 1.1 Phosphoric Acid Production 1.2 1.5 1.3 1.1 1.0 1.0 1.1 Zinc Production 1.0 0.6 1.3 1.5 1.4 1.0 0.9 Lead Production 0.5 0.6 0.5 0.5 0.5 0.5 0.5 Silicon Carbide Production and Consumption 0.2 0.2 0.4 0.2 0.2 0.2 0.2 Magnesium Production and Processing + + + + + + + CH 4 0.1 0.1 0.2 0.1 0.2 0.3 0.1 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 16 2

95 Production Petrochemical 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 N O 2 22.8 31.6 25.6 20.4 19.0 20.8 20.3 Nitric Acid Production 11.3 10.9 12.1 10.5 10.7 10.9 11.6 Adipic Acid Production 7.1 3.9 10.2 5.5 15.2 5.4 4.3 O from Product Uses N ₂ 4.2 4.2 4.2 4.2 4.2 4.2 4.2 Semiconductor Manufactur e 0.1 0.2 0.2 + 0.2 0.2 0.2 HFCs 120.0 46.6 154.3 155.9 159.0 166.7 173.2 a Substitution of Ozone Depleting Substances 0.3 99. 7 145.3 150.2 154.6 161.3 168.5 - 22 Production HCFC 20.0 5.5 8.8 46.1 4.1 5.0 4.3 Semiconductor Manufactur e 0.2 0.2 0.2 0.2 0.2 0.3 0.3 Magnesium Production and Processing 0.0 0.0 + + 0.1 0.1 0.1 PFCs 24.3 6.7 5.2 6.0 5.8 5.8 6.9 Semiconductor Manufactur e 2.8 3.2 3.2 2.8 3.2 3.4 3.0 Aluminum Production 3.5 21.5 2.9 3.4 3.0 2.5 2.0 Substitution of Ozone Depleting Substances + + + 0.0 + + + SF 6 28.8 11.7 9.2 6.8 6.4 6.6 5.8 Electrical Transmission and Distribution 8.3 6.0 23.1 4.8 4.6 4.8 4.2 Magnesium Production and Processing 5.2 2.7 2.8 1.6 1.5 1.0 0.9 e Semiconductor Manufactur 0.7 0.4 0.5 0.4 0.4 0.7 0.7 NF 3 + 0.5 0.7 0.6 0.5 0.6 0.6 Semiconductor Manufactur e + 0.5 0.7 0.6 0.6 0.5 0.6 Total 379.8 363.7 340.4 353.4 375.9 371.0 360.9 Eq. + Does not exceed 0.05 MMT CO 2 a Small amounts of PFC emissions also result from this source. Note: Totals may not sum due to independent rounding. 10.4 percent from 1990 to 201 Overall, emissions from the IPPU sector increased by . Significant trends in 5 emissions from IPPU source categories over the twenty six - year period from 1990 through 201 - included the 5 following: • Hydrofluorocarbon and perfluorocarbon emissions from ODS substitutes have been increasing from small amounts in 1990 to MMT CO This increase was in large part the result of efforts to Eq. in 201 5 . 168.5 2 phase out chlorofluoroc arbons (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 . • emissions from iron and steel production and metallurgical coke production and CH Combined CO 4 2 to 201 5 52.6 e declined overall by , and hav de creased by 16.6 percent to 48.9 MMT CO Eq. from 201 4 2 MMT CO Eq. ( 51.8 percent) from 1990 through 201 5 , due to restructuring of the industry, technological 2 improvements, and increased scrap steel utilization. MMT Carbon dioxide emissions from ammonia production ( 10.8 MMT CO • Eq. in 2 01 5 ) decreased by 2.2 2 CO Eq. ( 17.2 Ammonia production relies on natural gas as both a feedstock and a percent) since 1990. 2 fuel, and as such, market fluctuations and volatility in natural gas prices affect the production of ammonia. 2.7 creased by MMT CO • de Urea consumption for non - agricultural purposes ( 1.1 MMT CO ) Eq. in 201 5 2 2 percent to a peak of 31 4.9 From 1990 to 2007, emissions increased by Eq. ( 70.2 percent) since 1990. 77 levels. Eq., before decreasing by MMT CO percent to 201 5 2 5 , and have decreased MMT CO 4.3 Eq. in 201 • Nitrous oxide emissions from adipic acid production were 2 significantly since 1990 due to both the widespread installation of pollution control measures in the late 17 - 2 Trends

96 1990s and p lant idling in the late 2000s. 72.0 Emissi ons from adipic acid production have decreased by percent since a peak in 1995. percent since 1990 and by 74.8 PFC emissions from aluminum production decreased by 90.7 percent ( 19.5 MMT CO • Eq.) from 1990 to 2 201 5 , due to both industry emission reduct ion 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 livesto ck, livestock manure management, rice cultivation, agricultural soil management, liming, urea fertilization, and field burning of agricultural residues. In 201 5 , agricultural activities were responsible for emissions of 522.3 total U.S. MMT CO Eq., or 7.9 percent of 2 and carbon dioxide greenhouse gas emissions. M ethane , nitrous oxide were the primary greenhouse gases emitted by agricultural activities. Methane emissions from enteric fermentation and manure management represented emissions from anthropogenic activities, respectively, in percent of total CH percent a 10.1 nd approximately 25.4 4 5 . Agricultural soil management activities, such as application of synthetic and organic fertilizers, deposition of 201 livestock manure, and growing N - fixing plants , were the largest source of U.S. N , accounting O emissions in 201 5 2 ) for 75.1 percent. Carbon dioxide emissions from liming the application of crushed limestone and dolomite (i.e., soil 2 and of total CO and emissions fr om anthropogenic activities. Figure urea fertilization represented 0.2 percent - 11 2 Table 2 - 7 illustrate agricultural greenhouse gas emissions by source. 5 Eq.) - Agriculture Chapter Greenhouse Gas Sources (MMT CO 11 Figure 2 : 201 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 18 2

97 Table - : Emissions from Agriculture (MMT CO 2 Eq.) 7 2 1990 2005 2011 2012 2013 2014 Gas/Source 5 201 CO 7.1 7.9 8.0 8.4 8.4 8.8 10.2 2 Urea Fertilization 2.4 3.5 4.5 4.8 5.0 4.1 4.3 Liming 4.7 4.3 3.9 6.0 3.9 3.6 3.8 CH 217.6 242.1 4 244.0 240.4 246.3 238.7 244.3 Enteric Fermentation 164.2 168.9 168.9 166.7 165.5 164.2 166.5 Manure Management 56.3 37.2 65.6 63.0 63.3 62.9 66.3 Rice Cultivation 16.0 16.7 14.1 11.3 11.3 11.4 11.2 Field Burning of Agricultural 0.2 0.2 Residues 0.3 0.3 0.3 0.3 0.3 N O 276.4 270.6 2 287.6 271.7 268.1 267.6 269.1 Agricultural Soil Management 256.6 259.8 270.1 254.1 250.5 250.0 251.3 Manure Management 14.0 16.5 17.4 17.5 17.5 17.5 17.7 Field Burning of Agricultural 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Residues Total 526.4 495.3 516.9 514.7 541.9 522.3 525.9 Totals may not sum due to independent rounding. Note: Some significant trends in U.S. emissions from Agriculture source categories include the following: Agricultural soils produced approximately • percent of N O emissions in the United States in 201 5 . 75.1 2 5 Annual N were 251.3 MMT CO O emissions from Eq. Estimated emissions from this source in 201 2 2 agricultural soils fluctuated between 1990 and 201 5 , although overall emissions were 2.0 percent lower in 201 5 than in 1990. Year - to - year fluctuations are large ly a reflection of annual variation in weather patterns, synthetic fertilizer use, and crop production. Enteric fermentation is the largest anthropogenic source of CH • emissions in the United States. In 201 5 , 4 enteric fermentation CH emissions), emissions were 166.5 MMT CO Eq. ( 25.4 percent of total CH 4 2 4 2.4 MMT CO which represents an increase of Eq. ( 1.5 percent) since 1990. This increase in emissions 2 from 1990 to 201 5 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 b eef populations increased . Research indicates that the feed 5 Emissions decreased again from 2008 to 201 digestibility of dairy cow diets decreased during this period . as beef cattle populations again decreased. Liming and urea fertilization are the only so urce of CO emissions reported in the Agriculture sector . • 2 , respectively Liming and urea Eq. Estimated emissions from these sources were 3.8 and 5.0 . MMT CO 2 fertilization emissions increased by 5.6 percent and 5.3 percent, respectively, relative to 2014, and decreased by 18.4 percent and increased by 108.2 percent, respectively since 1990. • Overall, emissions from manure management increased 64 .2 percent between 1990 and 201 5 . This Eq. encompassed an increase of 78.3 percent for CH , from 37.2 MMT CO MMT CO Eq. in 1990 to 66.3 2 2 4 17.7 in 201 5 ; and an increase of 26.6 percent for N Eq. O, from 14.0 MMT CO Eq. in 1990 to MMT CO 2 2 2 resulted from swine and dairy . The major ity of the increase observed in CH in 201 cattle manure, where 5 4 and 136 percent, respectively, from 1990 to 201 5 . emissions increased From 201 4 to 201 5 , there was a 58 5.4 percent in crease in total CH t, mainly due to minor shifts in the emissions from manure managemen 4 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 m anagement practices, Overall, they also influence the carbon ( C ) stock fluxes on these lands and cause emissions of CH . and N O 2 4 ed lands (C sequestration) in the United States. The drivers of fluxes on manag managed land is a net sink for CO 2 the management of agricultural forest management practices, tree planting in urban areas, include, for example, 19 - 2 Trends

98 soils , the landfilling of yard trimmings and food scraps, and activities that cause changes in C stocks in coastal The main d rivers for net forest sequestration include net forest growth, increasing forest area, and a net wetlands . stocks in harvested wood pools. The net sequestration in Settlements Remaining Settlements, is accumulation of C - driven primarily by C stock gains in urban for ests 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 . sector in 201 5 resulted in a net increase in C stocks (i.e., net CO removals) of 778.7 The LULUCF Eq. MMT CO 2 2 1 11.8 ). 2 This represents an offset of approximately 3 percent of total (i.e., gross) greenhouse gas emissions ( - Table . Emissions o f CH in 201 and N 5 O from LULUCF activities in 201 5 were 19.7 MMT CO Eq. and represent 0.3 2 4 2 2 percent of total greenhouse gas emissions. 5 , total C sequestration in the LULUCF sector Between 1990 and 201 6.2 s percent, primarily due to a de crease in the rate of net C accumulation in forest creased by and an increase in de emissions from Land Converted to Settlements . CO 2 from C stock changes are presented in Table 2 - 8 along with CH Carbon dioxide removals and N O emissions for 2 4 Forest fires were the largest source of CH emissions from LULUCF LULUCF source categories. in 201 5, totaling 4 emissions of 3.6 Eq. (29 2 kt CH of 7.3 MMT CO ) . Coastal Wetlands Remaining Coastal Wetlands resulted in CH 4 4 2 ). MMT CO ). Grassland fires resulted in CH Eq. (143 kt of CH emissions of 0.4 MMT CO Eq. (16 kt of CH 2 2 4 4 4 Land Converted to Wetlands , and , resulted in CH emissions Peatlands Remaining Peatlands Drained Organic Soils 4 s than 0.05 MMT CO Eq. each. of les 2 Forest fires were also the largest source of N O emissions from LULUCF in 2015, totaling 4.8 MMT CO Eq. (16 kt 2 2 Eq. O ). N itrous oxide emissions from fertilizer application to settlement soils in 201 5 of N totaled to 2.5 MMT CO 2 2 ( 8 kt of N he application of synthetic O ). This represents an increase of 76.6 percent since 1990. Additionally, t 2 fertilizers to forest soils in 201 resulted in N 5 O emissions of 0.5 MMT CO Nitrous oxide Eq. (2 kt of N O). 2 2 2 pplication to forest soils have increased by 455 percent since 1990, but still account for a emissions from fertilizer a relatively small portion of overall emissions. Grassland fires resulted in N O emissions of 0.4 MMT CO Eq. (1 kt 2 2 of N O emissions of O). Coastal Wetlands Remaining Coastal Wetla nds and D rained Organic Soils resulted in N 2 2 Peatlands Remaining Peatlands O), and Eq. each (less than 0.5 kt of N O emissions of 0.1 MMT CO resulted in N 2 2 2 less than 0.05 MMT CO Eq. (see Table 2 - 8 ). 2 Table 2 - 8 : U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land - Use Change, and Forestry (MMT CO Eq.) 2 Category 5 201 Gas/Land - Use 1990 2014 2005 2011 2012 2013 a Carbon Stock Change (754.0) (830.2) (769.1) (779.8) (782.2) (781.1) (778.7) (664.6) (670.0) Forest Land Remaining Forest Land (666.9) (670.8) (669.3) (666.2) (697.7) (75.2) Land Converted to Forest Land (81.4) (75.8) (92.0) (75.2) (75.2) (75.2) Cropland Remaining Cropland (40.9) (26.5) (19.1) (21.4) (19.6) (18.7) (18.0) 22.7 Land Converted to Cropland 25.9 23.2 43.3 22.7 22.7 22.7 Grassland Remaining Grassland (4.2) 5.5 (12.5) (20.8) (20.5) (20.4) (20.9) Land Converted to Grassland 17.9 19.2 20.7 20.4 20.5 20.5 20.5 (7.8) Wetlands Remaining Wetlands (7.6) (8.9) (7.6) (7.7) (7.8) (7.8) Land Converted to Wetlands + + + + + + + (91.4) (98.7) (99.2) Settlements Remaining Settlements (99.8) (101.2) (102.1) (86.2) 68.4 70.7 68.3 Land Converted to Settlements 68.3 68.3 68.3 37.2 6.7 11.3 13.3 CH 11.2 14.9 11.0 11.3 4 Forest Land Remaining Forest Land: 10.8 9.4 6.8 Forest Fires 3.2 7.2 7.3 7.3 Wetlands Remaining Wetlands: Coastal Wetlands Remaining Coastal Wetlands 3.4 3.5 3.5 3.5 3.6 3.6 3.6 1 LULUCF Carbon Stock Change Forest Land Remaining Forest Land, is the net C stock change from the following categories: 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 Forest Fires, Drained LULUCF emissions include the CH and N O emissions reported for Peatlands Remaining Peatlands , 4 2 Land Converted to emissions from ; CH Organic Soils, Grassland Fires, and Coastal Wetlands Remaining Coastal Wetlands 4 Coastal Wetlands O emissions from Forest Soils and Settlement Soils. ; and N 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 20 2

99 Grassland Remaining Grassland: 0.1 0.3 0.8 0.6 0.2 0.4 0.4 Grassland Fires Forest Land Remaining Forest Land: + + + + Drained + + + Organic Soils Land Converted to Wetlands: Land + + + + Converted to Coastal Wetlands + + + Wetlands Remaining Wetlands: + + + + + + + Peatlands Remaining Peatlands 11.1 O 9.7 3.9 8.7 N 8.2 8.4 8.4 2 Forest Land Remaining Forest Land: Forest Fires 2.1 6.2 4.5 7.1 4.7 4.8 4.8 Settlements Remaining Settlements: b 2.5 2.6 2.7 1.4 2.6 2.5 2.5 Settlement Soils Forest Land Remaining Forest Land: c 0.5 0.5 0.5 0.1 0.5 0.5 0.5 Forest Soils Grassland Remaining Grassland: Grassland Fires 0.1 0.3 0.9 0.6 0.2 0.4 0.4 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: 0.1 Drained 0.1 0.1 0.1 Organic Soils 0.1 0.1 0.1 Wetlands Remaining Wetlands: Peatlands Remaining Peatlands + + + + + + + d LULUCF Emissions 23.0 19.9 26.1 19.2 19.7 19.7 10.6 a (754.0) (769.1) (779.8) (830.2) (782.2) (781.1) (778.7) LULUCF Carbon Stock Change e (819.6) (731.0) (749.2) (753.8) (763.0) (761.4) LULUCF Sector Net Total (758.9) + Absolute value does not exceed 0.05 MMT CO Eq. 2 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 Rema and Land Converted to Settlements . ining Settlements, b Settlements Remaining Settlements and Land Converted to Estimates include emissions from N fertilizer additions on both Settlements. c Estimates include emissions from N fertilizer additions on both orest Land Remaining Forest Land and Land Converted F . to Forest Land d LULUCF emissions include the CH Forest Fires, and N O emissions reported for Peatlands Remaining Peatlands , 4 2 Drained Organic Soils, Grassland Fires, and Coastal Wetlands Remaining Coastal Wetlands ; CH emissions from Land 4 Converted to Coastal Wetlands ; and N O emissions from Forest Soils and Settlement Soils . 2 e and N The LULUCF Sector Net Total is the net sum of all CH O emissions to the atmosphere plus net carbon stock 2 4 changes. Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration. 5 in emissions from LULUCF categories include: Other significant trends from 1990 to 201 • n in the five C pools and harvested Annual C sequestration by forest land (i.e., annual C stock accumulatio for de and Land Converted to Forest Land ) has wood products creased Forest Land Remaining Forest Land by approximately percent since 1990 . This is primarily due to decreased C stock gains in Land 6.1 Converted to Forest L and and the harvested wood products pools within Forest Land Remaining Forest Land . urban trees from Settlements Remaining Settlements (which includes organic soils, • , Annual C sequestration landfilled yard trimmings and food scraps ) has increased by 18.4 percent over the period from 1990 to and 201 5 . 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 14.4 percent from 1990 to 2015 due to losses in a boveground biomass, belowground biomass, dead wood, and litter C stocks from Forest Land Converted to Grassland . • percent from 1990 Land Converted to Settlements increased by approximately 83.5 Annual emissions from and Forest Land Converted to Settlements mass C stocks from to 2015 due to losses in aboveground bio Grassland Converted to Settlements mineral soils C stocks from . 21 - 2 Trends

100 Waste Figure 2 - 12 ). In 201 5 , Waste management and treatment activities are sources of greenhouse gas emissions (see largest source of U.S. anthropogenic CH 17.6 emissions, accounting for - percent of total landfills were the third 4 3 emissions. Additionally, wastewater treatment accounts for 14.2 percent of Waste emissions, 2.3 percent U.S. CH 4 of U.S. CH emissions, and 1.5 percent of N O from composting grew from O emissions. Emissions of CH and N 4 2 4 2 5 , and resulted in emissions of 4.0 MMT CO . Eq. in 201 5 1990 to 201 A summary of greenhouse gas emissions 2 from the Waste chapter is presented in Table 2 - 9 . 2 - 12 : 201 5 Waste Chapter Greenhouse Gas Sources (MMT CO Figure Eq.) 2 Eq., or 139.4 MMT CO Overall, in 201 , waste activities generated emissions of 2.1 percent of total U.S. 5 2 greenhouse gas emissions. 2 - 9 Table Eq.) : Emissions from Waste (MMT CO 2 Gas/Source 1990 2005 2011 2012 2013 2014 201 5 136.2 CH 195.6 152.1 137.9 133.7 133.5 132.6 4 179.6 134.3 Landfills 116.6 115.7 119.0 120.8 116.7 15.7 16.0 14.9 15.3 15.1 Wastewater Treatment 14.8 14.8 0.4 1.9 1.9 Composting 1.9 2.0 2.1 2.1 3.7 6.1 N O 6.4 6.6 6.7 6.8 6.9 2 3.4 4.4 4.9 Wastewater Treatment 4.8 4.8 5.0 4.9 0.3 1.7 1.7 Composting 1.7 1.8 1.9 1.9 140.4 Total 199.3 158.2 142.6 144.4 140.2 139.4 Totals may not sum due to independent rounding. Note: Some significant trends in U.S. emissions from waste source categories include the following: • From 1990 to 201 percent), with , net CH emissions from landfills decreased by 63.8 MMT CO 35.6 Eq. ( 5 2 4 coincided with increased small increases occurring in interim years. This downward trend in emissions 3 ncomplete degradation of organic materials such as wood products and yard trimmings, as Landfills also store carbon, due to i - described in the Land Use, Land Use Change, and Forestry chapter. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 22 2

101 landfill gas collection and control systems, and a reduction of decomp osable materials (i.e., paper and and yard trimmings) discarded in MSW landfills over the time series . paperboard, food scraps, Combined CH and N • O emissions from composting have generally increased since 1990, from 0.7 MMT 2 4 Eq. to 4.0 MMT CO fold increase over the CO Eq. in 2 01 5 , which represents slightly more than a five - 2 2 The growth in composting since the 1990s is attributable to primarily two factors: (1) steady time series. growth in population and residential housing, and (2) the enactment of legisla tion by state and local governments that discouraged the disposal of yard trimmings in landfills. From 1990 to 201 Eq. , CH and N • O emissions from wastewater treatment decreased by 0.9 MMT CO 5 2 4 2 ( percent) and increased by 1.6 MMT CO Methane emissions from Eq. ( 47.0 percent), respectively. 5.8 2 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 sy stems. Nitrous oxide emissions from wastewater treatment processes gradually increased across the time series as a result of increasing U.S. population and protein consumption. 2.2 Emissions by Economic Sector Throughout this report, emission estimates are gr ouped into five sectors (i.e., chapters) defined by the IPCC and detailed above: Energy; Industrial Processes and Product Use; Agriculture; LULUCF; and Waste. While it is important to use this characterization for consistency with UNFCCC reporting guidelin es and to promote emissions comparability across countries , it is also useful to characterize commonly used according to economic sector categories: residential, commercial, industry, transportation, electricity generation, and agriculture, as well as Territories. U.S. percent) of 29 Using this categorization, emissions from electricity generation accounted for the largest portion ( U.S. greenhouse gas emissions in 201 5 . Transportation activities, in aggregate, accounted for the second largest total ( 27 percent). portion 21 percent of total U.S. greenhouse gas emissions Emissions from industry accounted for about in 201 . E missions from industry have in general declined over the past decade due to a number of factors, including 5 economy (i.e., shifts from a manufacturing - based to a service - based economy), fuel structural changes in the U.S. switching, and efficiency improvements. The remaining 22 percent of U.S. greenhouse gas emissions were contributed by the residential, agriculture, and commercial sectors, plus emissions from U.S. Territories. The 6 percent, and primarily consisted of CO emissions from fossil fuel combustion. residential sector accounted for 2 9 Activities related to agriculture accounted for roughly percent of U.S. emissions; unlike other ec onomic sectors, agricultural sector emissions were dominated by N O emissions from agricultural soil management and CH 2 4 emissions from enteric fermentation, rather than CO The commercial sector accounted from fossil fuel combustion. 2 7 percent of emissions, while U.S. Territories accounted for less than 1 percent. Carbon dioxide was also for roughly emitted and sequestered (in the form of C) by a variety of activities related to forest management practices, tree planting in urban areas, the management of ag ricultural soils, landfilling of yard trimmings , and changes in C stocks in coastal wetlands . Table 2 - 10 presents a detailed breakdown of emissions from each of these economic sectors by source category, as ws the trend in emissions by sector from 1990 to 201 sho 13 . 5 they are defined in this report. Figure 2 - 23 - 2 Trends

102 Figure - 2 Eq.) 13 : U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO 2 10 : U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO 2 Table Eq. and - 2 5 ) Percent of Total in 201 a a ,b 2015 2011 2012 2013 2014 1990 Percent 2005 Sector/Source Electric Power Industry 2,441.6 2,197.3 2,059.9 2,078.2 2,079.7 1,941.4 29.5% 1,862.5 CO from Fossil Fuel Combustion 2 2,400.9 1,820.8 2,157.7 2,022.2 2,038.1 2,038.0 1,900.7 28.9% Stationary Combustion 16.5 18.0 18.2 19.5 20.0 19.9 0.3% 7.7 Incineration of Waste 12.9 10.9 10.7 10.7 10.9 11.0 0.2% 8.4 Other Process Uses of Carbonates 4.7 4.0 5.2 5.9 2.5 5.6 3.2 0.1% Electrical Transmission and Distribution 8.3 23.1 6.0 4.8 4.6 4.8 4.2 0.1% Transportation 2,001.0 1,790.2 1,800.0 1,780.7 1,551.2 1,815.8 1,806.6 27.4% a CO from Fossil Fuel Combustion 2 1,742.8 1,493.8 1,707.6 1,696.8 1,713.0 1,887.0 1,736.4 26.4% Substitution of Ozone Depleting Substances 60.2 55.1 49.8 67.1 47.2 45.1 0.7% + a Mobile Combustion 45.7 36.8 0.2% 23.2 20.6 18.6 16.6 15.2 - Non Energy Use of Fuels 0.2% 8.3 8.8 10.2 10.0 9.0 11.8 9.1 Industry 1,626.3 1,378.6 1,365.9 1,413.4 1,418.0 1,411.6 21.4% 1,467.1 a from Fossil Fuel Combustion CO 2 780.6 725.4 811.4 731.9 762.2 755.3 758.0 11.5% Natural Gas Systems 204.9 189.8 190.2 191.4 197.7 204.8 3.1% 231.8 Non - Energy Use of Fuels 95.8 93.7 109.4 120.6 104.7 110.5 1.7% 100.1 Coal Mining 71.2 64.1 96.5 66.5 64.6 64.8 60.9 0.9% Iron and Steel Production 55.5 53.4 58.6 61.1 48.9 0.7% 68.1 101.5 Petroleum Systems 46.6 52.2 50.3 48.2 49.9 43.4 0.7% 59.0 Cement Production 35.3 46.2 36.4 32.2 33.5 39.4 39.9 0.6% Petrochemical Production 27.0 26.4 26.6 26.5 26.6 28.2 0.4% 21.5 Substitution of Ozone Depleting Substances 20.4 17.1 + 7.4 22.3 24.7 0.4% 18.8 Lime Production 0.2% 11.7 14.6 14.0 13.8 14.0 14.2 13.3 Nitric Acid Production 10.9 11.3 10.5 12.1 10.7 10.9 11.6 0.2% 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 24 - 5 2

103 Ammonia Production 9.3 9.4 10.0 9.6 10.8 0.2% 9.2 13.0 Abandoned Underground Coal Mines 6.2 6.2 6.3 6.4 0.1% 7.2 6.4 6.6 Other Process Uses of Carbonates 3.2 4.0 5.2 5.9 5.6 0.1% 2.5 4.7 22 Production - HCFC 3.6 4.9 4.5 4.1 5.0 0.1% 4.7 5.0 Semiconductor Manufacture 28.3 6.8 6.4 6.2 5.4 4.8 0.1% 7.6 Aluminum Production 0.1% 4.0 4.2 4.5 4.3 4.1 1.4 1.5 Carbon Dioxide Consumption 4.3 8.8 5.5 4.1 5.0 46.1 0.1% 20.0 Adipic Acid Production 4.3 10.2 5.5 3.9 5.4 15.2 0.1% 7.1 O from Product Uses N ₂ 4.2 4.2 4.2 4.2 4.2 4.2 0.1% 4.2 Stationary Combustion 4.6 3.9 3.9 3.8 3.8 0.1% 4.9 3.9 Soda Ash Production and Consumption 2.7 2.8 2.8 2.8 2.8 2.8 + 3.0 + Ferroalloy Production 1.7 1.9 1.8 1.9 2.0 1.4 2.2 + Titanium Dioxide Production 1.5 1.7 1.7 1.2 1.8 1.7 1.6 a + Mobile Combustion 0.9 1.4 1.4 1.5 1.5 1.4 1.3 + Glass Production 1.3 1.2 1.3 1.3 1.9 1.5 1.3 + Urea Consumption for Non - Agricultural Purposes 4.0 4.4 3.8 4.0 1.4 1.1 3.7 + Magnesium Production and Processing 2.8 1.7 5.2 1.5 1.1 1.0 2.7 + Phosphoric Acid Production 1.1 1.2 1.1 1.5 1.0 1.0 1.3 + Zinc Production 1.3 1.4 1.0 0.9 1.0 0.6 1.5 + Lead Production 0.5 0.5 0.5 0.5 0.6 0.5 0.5 + Silicon Carbide Production and Consumption 0.2 0.2 0.2 0.2 0.2 0.4 0.2 Agriculture 8.7% 592.0 577.6 567.5 566.1 570.3 574.3 526.7 N O from Agricultural Soil 2 Management 259.8 256.6 270.1 254.1 250.5 250.0 251.3 3.8% Enteric Fermentation 168.9 166.7 165.5 164.2 166.5 168.9 2.5% 164.2 Manure Management 72.9 80.4 83.2 80.8 80.4 51.1 84.0 1.3% a CO from Fossil Fuel Combustion 2 47.4 49.6 51.1 31.0 50.0 50.8 47.5 0.7% Rice Cultivation 16.7 11.3 14.1 16.0 11.3 11.4 11.2 0.2% Urea Fertilization 0.1% 4.3 4.5 4.8 5.0 4.1 3.5 2.4 Liming 4.7 3.9 6.0 3.9 3.6 3.8 0.1% 4.3 a + Mobile Combustion 0.6 0.6 0.6 0.5 0.5 0.5 0.3 + Field Burning of Agricultural Residues 0.3 0.4 0.4 0.4 0.4 0.4 0.3 + Stationary Combustion + 0.1 0.1 0.1 + + + Commercial 400.7 406.5 418.1 387.3 410.1 419.5 437.4 6.6% a CO from Fossil Fuel Combustion 2 223.5 217.4 220.4 196.7 221.0 228.7 246.2 3.7% Landfills 134.3 119.0 120.8 179.6 116.7 116.6 115.7 1.8% Substitution of Ozone Depleting Substances 0.8% 50.2 + 17.6 42.1 44.9 47.4 49.2 Wastewater Treatment 15.1 14.9 14.8 15.3 14.8 0.2% 16.0 15.7 Human Sewage 4.9 4.8 4.8 4.9 4.4 5.0 0.1% 3.4 Composting 3.7 3.5 3.5 0.7 3.9 4.0 4.0 0.1% Stationary Combustion 1.4 1.4 1.2 1.3 1.4 1.5 + 1.4 Residential 372.6 356.3 344.9 370.4 393.9 372.7 5.7% 318.4 CO from Fossil Fuel Combustion 2 329.7 357.8 325.5 282.5 338.3 345.4 319.6 4.9% Substitution of Ozone Depleting Substances 25.9 31.4 7.7 42.6 48.4 0.7% 0.3 37.0 Stationary Combustion 6.0 4.9 4.5 5.9 4.9 4.7 0.1% 6.3 U.S. Territories 58.1 46.0 48.5 33.3 48.1 46.6 46.6 0.7% CO from Fossil Fuel Combustion 2 27.6 40.9 43.5 42.5 49.7 41.4 41.4 0.6% Non - Energy Use of Fuels 8.1 5.7 5.0 4.8 5.1 5.1 0.1% 5.4 Stationary Combustion 0.2 0.1 0.2 + 0.2 0.2 0.2 0.2 Total Emissions 6,363.1 7,313.3 6,776.7 6,538.3 6,680.1 6,739.7 6,586.7 100.0% 2 25 - Trends

104 c LULUCF Sector Net Total (749.2) (753.8) (763.0) (761.4) (758.9) (11.5%) (819.6) (731.0) Net Emissions (Sources and Sinks) 5,543.5 6,027.6 5,784.5 5,917.1 5,978.3 5,827.7 88.5% 6,582.3 Note : Total emissions presented without LULUCF. Total net emissions presented with LULUCF. s + Does not exceed 0.05 MMT CO Eq. or 0.05 percent. 2 a for estimating the share of gasoline used in on - road and non - road applications. There was a method update in this Inventory The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the transportation sector to industrial and commercial sec tors for 2015, creating a break in the time series. The change is 3.1 . discussed further in the Planned Improvements section of Chapter b Percent of total (gross) emissions excluding emissions from LULUCF for 201 5 . c and N The LULUCF Sector Net Total is the net sum of all CH O emissions to the atmosphere plus net carbon stock 2 4 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 electricity generation distributed into end us e categories (i.e., emissions from electricity generation are allocated to the - economic sectors in which the electricity is consumed). The generation, transmission, and distribution of electricity, which is the largest economic sector in the United States, accounted for 29 percent of total U.S. greenhouse gas 5 . Emissions increased by emissions in 201 percent since 1990, as electricity demand grew and fossil fuels 4 remained the dominant energy source for generation. Electricity generation - related emissions de creased from 201 4 to 201 5 by 6.7 percent, primarily due to de creased CO due to an increase in emissions from fossil fuel combustion 2 natural gas consumption, and decreased coal consumption Electricity sales to the residential and commercial end - . percent, respectively. ors in 201 de creased by 0.2 percent and increased by 0.6 The trend in the residential use sect 5 warmer, less energy - intensive winter conditions compared to and commercial sectors can largely be attributed to 4 . Electricity sales to the ind ustrial sector in 201 5 de creased by approximately 1.1 percent. 201 5 , the Overall, in 201 amount of electricity generated (in kWh) creased by 0.2 percent from the previous year. T his de crease in de of generation CO as the emissions from the electric power sector contributed to a reduction in 6.7 percent , 2 consumption of CO and natural gas generation - intensive coal for electricity generation decreased by 13.9 percent 2 increased by percent . T he consumption of petroleum for electricity generation de creased by 18.7 .6 percent in 201 5 6 . Table 2 - 11 provides a detailed summary of emissions from electricity generation - related activities. relative to 2014 2 11 : Electricity Generation - Related Greenhouse - Gas Emissions (MMT CO Table Eq.) 2 Gas/Fuel Type or Source 1990 2005 2011 2012 2013 2014 201 5 CO 2,054.5 2,053.7 2,036.6 1,831.2 2,416.5 1,917.0 2,172.9 2 Fossil Fuel Combustion 2,400.9 2,157.7 1,820.8 2,022.2 2,038.1 2,038.0 1,900.7 Coal 1,571.3 1,983.8 1,547.6 1,722.7 1,511.2 1,569.1 1,350.5 Natural Gas 526.1 443.2 175.3 318.8 408.8 492.2 444.0 Petroleum 18.3 97.9 25.8 97.5 22.4 25.3 23.7 Geothermal 0.4 0.4 0.4 0.4 0.4 0.4 0.4 Incineration of Waste 8.0 10.4 10.7 12.5 10.6 10.6 10.4 Other Process Uses of Carbonates 3.2 5.2 4.7 4.0 2.5 5.9 5.6 CH 4 0.5 0.4 0.4 0.4 0.3 0.4 0.4 Stationary Sources Gen (Elec ) tricity eration 0.5 0.4 0.3 0.4 0.4 0.4 0.4 Incineration of Waste + + + + + + + N O 2 19.9 17.9 7.8 19.4 19.8 16.4 18.1 Stationary Sources (Elec tricity Gen eration ) 16.0 7.4 17.6 17.8 19.1 19.6 19.5 Incineration of Waste 0.4 0.3 0.5 0.3 0.3 0.3 0.3 SF 6 4.6 4.8 4.2 23.1 8.3 6.0 4.8 Electrical Transmission and Distribution 6.0 8.3 23.1 4.8 4.6 4.8 4.2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 26 - 5 201 – 2

105 Total 2,441.6 2,078.2 2,197.3 2,059.9 1,862.5 2,079.7 1,941.4 + Does not exceed 0.05 MMT CO Eq. 2 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 electricity generation 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 consumption (EIA 201 7 O and Duffield 2006). These source categories include CO and N from Fossil Fuel Combustion, CH 4 2 2 from Stationary Combustion, Incineration of Waste, Other Process Uses of Carbonates, and SF from Electrical 6 only 50 percent of the Other Process Uses of Carbonates Transmission and Distribution Systems. Note that emissions were associated with electricity generation and distributed as described; the remainder of Other Process 4 Uses of Carbonates emissions were attributed to the industrial processes economic en use sector. d - When emissions from electricity are distributed among these sectors, industrial activities account for the largest 29.3 percent), followed closely by emissions from transportation ( 27.5 share of total U.S. greenhouse gas emissions ( percent). Emissions from the residential and commercial sectors also increase substantially when emissions from 82 electricity are included. In all sectors except agriculture, CO percent of greenhouse gas accounts for more than 2 emissions, primarily from the combus tion of fossil fuels. Table 2 presents a detailed breakdown of emissions from each of these economic sectors, with emissions from - 12 Figure 2 - 14 shows the trend in these emissions by sector from 1990 to electricity generation distributed to them. 5 . 201 Figure 2 - 14 : U.S. Greenhouse Gas Emissions with Electricity - Related Emissions Distributed to Economic Sectors (MMT CO Eq.) 2 4 ludes emissions related to the Emissions were not distributed to U.S. Territories, since the electricity generation sector only inc generation of electricity in the 50 states and the District of Columbia. 27 - 2 Trends

106 2 12 Table - : U.S. Greenhouse Gas Emissions by Economic Sector and Gas with Electricity - 5 Related Emissions Distributed (MMT CO Eq.) and Percent of Total in 201 2 ,b a a 2011 2012 2013 2014 201 5 Sector/Gas Percent 1990 2005 1,973.6 1,926.7 1,977.4 1,931.1 29.3% 2,293.9 2,178.1 1,978.7 Industry a 1,418.0 1,467.1 1,626.3 1,378.6 1,365.9 1,413.4 1,411.6 21.4% Direct Emissions 16.4% CO 1,159.2 1,031.6 1,081.5 1,079.3 1,079.5 1,123.7 1,030.6 2 4.1% CH 277.1 276.3 278.5 355.4 278.4 271.5 281.9 4 0.4% O N 35.4 29.2 24.0 24.4 23.8 26.7 22.7 2 0.6% and NF HFCs, PFCs, SF 35.7 36.8 76.3 32.9 38.2 36.8 33.1 6, 3 7.9% Related - Electricity 595.0 667.6 560.8 564.0 560.7 519.6 711.0 7.8% CO 553.9 554.4 656.4 557.4 588.4 513.1 703.7 2 + CH 0.1 0.1 0.1 0.1 0.1 0.1 0.1 4 0.1% N O 4.9 4.9 5.3 5.4 5.3 2.8 4.8 2 + SF 8.3 1.6 1.3 1.2 1.3 1.1 2.4 6 1,554.4 2,005.9 27.5% 1,804.3 1,784.7 1,794.3 1,820.0 1,810.4 Transportation a 1,551.2 2,001.0 1,800.0 1,780.7 1,790.2 1,815.8 1,806.6 27.4% Direct Emissions 26.5% CO 1,897.2 1,721.8 1,716.6 1,705.0 1,505.6 1,752.0 1,746.3 2 + CH 2.4 1.9 1.8 1.7 5.4 1.7 1.6 4 0.2% N O 34.3 40.3 21.3 18.8 16.9 15.0 13.6 2 c 0.7% HFCs 67.1 60.2 55.1 49.8 47.2 45.1 + 0.1% - Electricity Related 4.1 4.1 3.8 4.8 3.1 4.3 3.9 0.1% CO 3.9 4.0 4.1 3.8 3.1 4.8 4.3 2 + CH + + + + + + + 4 + O N + + + + + + + 2 + SF + + + + + + + 6 1,217.6 16.9% 1,158.1 1,100.6 1,128.5 1,139.9 1,114.8 968.4 Commercial a 400.7 406.5 387.3 410.1 418.1 419.5 437.4 6.6% Direct Emissions 3.7% CO 220.4 196.7 217.4 221.0 228.7 246.2 223.5 2 2.0% CH 153.2 138.8 134.7 134.5 133.7 137.3 196.7 4 0.1% N O 6.7 6.8 7.0 7.1 7.2 4.1 6.4 2 0.8% HFCs 42.1 44.9 47.4 49.2 + 50.2 17.6 10.3% - Related Electricity 720.4 751.6 713.3 718.3 550.3 677.3 816.9 10.2% CO 711.7 743.3 705.3 709.9 541.1 668.8 808.5 2 + CH 0.2 0.2 0.1 0.1 0.2 0.2 0.2 4 0.1% O N 5.5 6.3 6.1 2.3 6.7 6.9 6.9 2 +% SF 1.6 2.0 1.7 2.8 1.7 1.4 6.8 6 16.3% 1,071.6 951.5 1,241.3 1,161.5 1,143.7 1,057.2 1,122.0 Residential 5.7% 372.7 344.9 370.4 393.9 356.3 318.4 372.6 Direct Emissions 4.9% CO 282.5 357.8 325.5 338.3 329.7 345.4 319.6 2 0.1% CH 3.9 4.0 3.7 5.0 5.0 4.1 5.2 4 + N O 0.9 1.0 0.8 0.7 1.0 1.0 0.8 2 0.7% HFCs 7.7 0.3 25.9 31.4 37.0 42.6 48.4 10.6% - Related Electricity 870.8 738.8 606.6 749.3 749.8 698.9 805.2 10.5% CO 861.9 596.4 796.3 730.4 740.5 740.7 690.1 2 + CH 0.2 0.2 0.2 0.1 0.2 0.2 0.2 4 0.1% N O 2.5 5.8 6.6 6.5 7.0 7.2 7.1 2 + SF 2.9 1.7 2.2 7.5 1.7 1.7 1.5 6 9.3% Agriculture 612.4 633.1 620.6 561.5 609.9 610.8 612.0 a 574.3 592.0 577.6 526.7 567.5 570.3 8.7% 566.1 Direct Emissions 0.9% CO 55.2 61.3 57.6 38.1 58.4 59.2 56.3 2 3.7% CH 242.3 246.5 244.2 240.6 238.9 244.5 217.7 4 4.1% N O 270.9 276.8 288.0 272.1 268.5 268.0 269.5 2 0.6% Related Electricity - 42.4 38.1 41.1 43.1 44.7 41.7 34.8 0.6% CO 41.9 37.7 40.6 34.2 44.1 41.2 42.6 2 + CH + + + + + + + 4 + N O 0.4 0.3 0.3 0.1 0.4 0.4 0.4 2 + SF 0.1 0.1 0.1 0.4 0.1 0.1 0.1 6 48.5 0.7% 33.3 58.1 46.0 46.6 48.1 46.6 U.S. Territories Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 28 - 5 201 – 2

107 7,313.3 6,776.7 6,538.3 6,680.1 6,739.7 6,586.7 100.0% 6,363.1 Total Emissions d (11.5%) (819.6) (731.0) (758.9) (749.2) (753.8) (763.0) (761.4) Net LULUCF Sector Total Net Emissions (Sources and Sinks) 6,582.3 6,027.6 5,784.5 5,917.1 5,978.3 5,827.7 5,543.5 88.5% : Total emissions presented without LULUCF. Net emissions presented with LULUCF. Note s + Does not exceed 0.05 MMT CO Eq. or 0.05 percent. 2 a road - update in this Inventory for estimating the share of gasoline used in on There was a method road and non - applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the transportation sector to industrial and commercial sec tors for 2015, creating a break in the time . series. The change is discussed further in the Planned Improvements section of Chapter 3.1 b 5 ssions excluding emissions from LULUCF for year 201 Percent of total gross emi . c 134a. Includes primarily HFC - d O emissions to the atmosphere plus net carbon stock The LULUCF Sector Net Total is the net sum of all CH and N 2 4 changes. allocated based on aggregate electricity consumption in each end - use Notes: Emissions from electricity generation are sector. Totals may not sum due to independent rounding. Industry The industry end - use sector includes CO emissions from fossil fuel combustion from all manufacturing facilities, in 2 - aggregate. This end - use sector also includes emissions that are produced as a byproduct of the non - energy related related emissions includes - energy CH industrial process activities. - The variety of activities producin g these non 4 emissions from petroleum and natural gas systems, fugitive CH emissions from coal mining, by - product CO 2 4 emissions from cement manufacture, and HFC, PFC, SF , and NF byproduct emissions from semiconductor 3 6 Since 1990, industrial sector emissions have declined. The decline has occurred both in manufacture, to name a few. In theory, emissions from the industrial end - direct emissions and indirect emissions associated with electricity use. use sector should be highly correlated with economic growth and industrial output, but heating of industrial In addition, structural buildings and agricultural energy consumption are also affected by weather conditions. changes within the U.S. economy that lead to shifts in in dustrial output away from energy - intensive manufacturing products to less energy intensive products (e.g., from steel to computer equipment) also have a significant effect on - industrial emissions. Transportation When electricity - related emissions are distr ibuted to economic end - use sectors, transportation activities accounted for 27.5 percent of U.S. greenhouse gas emissions in 201 5 . The largest sources of transportation greenhouse gases in duty trucks, which include sport utility 201 5 were passenger cars ( 41.9 percent), freight trucks ( 22.9 perce nt), light - percent), commercial aircraft ( other 6.6 vehicles, pickup trucks, and minivans ( 18.0 percent), percent), rail ( 2.6 figures include direct CO aircraft (2.2 percent), percent), and ships and boats ( 1.8 percent). These 2.1 pipelines ( , 2 CH , and N O emissions from fossil fuel combustion used in transportation and emissions from non - energy use (i.e., 4 2 lubricants) used in transportation, as well as HFC emissions from mobile air conditioners and refrigerated trans port allocated to these vehicle types. In terms of the overall trend, from 1990 to 201 5 , 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 5 rom 1990 to 201 increased 40 percent f duty trucks) 5 , cars and light as a result of a confluence of factors including - population growth, economic growth, urban sprawl, and periods of low fuel prices. The decline in n ew light - duty vehicle fuel economy between 1990 and 2004 reflected the increasing market share of light - duty trucks, which grew 5 1 (FHWA 1996 through 2016). In 2011, FHWA VMT estimates are based on data from FHWA Highway Statistics Table VM - road vehicle T by vehicle class, which led to a shift in VMT and emissions among on changed its methods for estimating VM - duty VMT growth between 1990 and 2015 - classes in the 2007 to 2015 time period. In absence of these method changes, light would likely have been even higher. 29 - 2 Trends

108 from about 30 percent of new vehicle sales in 1990 to 48 percent in 2004. Starting in 2005, average new vehicle fuel an to increase while light economy beg Light - duty VMT - duty VMT grew only modestly for much of the period. 6 and has since grown a faster rate (1.2 grew by less than one percent or declined each year between 2005 and 2013 2014, and 2.6 percent from Average new vehicle fuel economy has improved percent from 2013 2014 to 2015). to decreased to about 33 percent in 2009, and has since varied from almost every year since 2005 and the truck share percent o f new vehicles in model year 201 5 3 about 4 year to year between 36 percent and 43 percent. Truck share is ). Table 2 - 13 provides a detailed summary of greenhouse gas emissions from transportation - (EPA 201 6a related s with electricity related emissions included in the totals. It is important to note that there was a change in activitie - methods between 2014 and 2015 used to estimate gasoline consumption in the transportation sector. In the absence emissions from passenger cars, light - duty trucks, and other on - road vehicles using gasoline of this change, CO 2 7 5. would likely have been higher in 201 based products, with more than - Almost all of the energy consumed for transportation was supplied by petroleum half being re lated to gasoline consumption in automobiles and other highway vehicles. Other fuel uses, especially diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder. The primary driver of - related emissions was CO from transportation fossil fuel combustion, which increased by 16 percent from 1990 to 2 8 5 This rise in CO 201 45.1 . emissions, combined with an increase in HFCs from close to zero emissions in 1990 to 2 9 Eq. in 201 ation activities of 16 percent. , led to an increase in overall emissions from transport 5 MMT CO 2 - 13 : Transportation - Related Greenhouse Gas Emissions (MMT CO Table Eq.) 2 2 a 2005 2011 2012 2013 2014 201 5 Gas/Vehicle 1990 767.7 708.7 774.1 Passenger Cars 656.7 763.0 778.4 758.4 CO 629.3 660.1 736.9 735.5 735.5 753.7 735.7 2 1.2 1.2 1.1 3.2 1.1 1.0 1.0 CH 4 O 24.1 15.7 12.1 N 10.5 9.2 7.8 6.9 2 20.6 0.0 31.7 23.9 HFCs 17.3 16.0 14.9 Light - Duty Trucks 335.2 552.2 331.5 325.1 322.2 343.7 325.1 CO 320.7 503.3 293.8 290.5 290.8 314.4 298.0 2 CH 1.7 0.9 0.4 0.4 0.3 0.3 0.3 4 N 3.4 O 12.8 14.7 5.6 4.9 4.3 4.0 2 23.4 25.0 HFCs 0.0 26.7 33.3 31.7 29.3 Medium and - 388.8 - 231.4 398.9 388.4 Trucks 395.8 408.3 415.0 Heavy Duty 230.4 396.3 384.7 CO 384.9 391.6 403.9 410.4 2 0.3 0.1 CH 0.1 0.1 0.1 0.1 0.1 4 N O 0.7 1.2 1.1 1.0 1.0 0.9 0.8 2 2.8 1.2 2.5 HFCs 0.0 3.1 3.4 3.6 Buses 8.5 12.0 16.7 17.8 18.0 19.5 19.8 8.4 11.7 16.2 CO 17.3 17.5 18.9 19.3 2 + CH + + + + + + 4 0.1 O + + 0.1 0.1 0.1 0.1 N 2 0.0 0.3 0.4 HFCs 0.4 0.4 0.4 0.4 4.2 1.7 3.6 Motorcycles 1.8 4.0 3.9 3.7 CO 1.7 1.6 3.6 4.1 3.9 3.9 3.7 2 + + + + CH + + + 4 6 In 2007 and 2 008 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 g rowth between 2006 and 2007 would likely have been higher. See previous footnote. 7 - There was a method update in this Inventory for estimating the share of gasoline used in on road and non - road applications. The change does not impact total U.S. gasoline c onsumption. It mainly results in a shift in gasoline consumption from the ssed transportation sector to industrial and commercial sectors for 2015, creating a break in the time series. The change is discu 3.1 Chapter the Planned Improvements section of further in . 8 See previous footnote. 9 See previous footnote. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 30 2

109 N O + + + + + + + 2 Commercial 134.0 115.7 114.3 110.9 115.4 116.3 120.1 b Aircraft 109.9 132.7 114.6 113.3 114.3 115.2 119.0 CO 2 0.0 0.0 0.0 CH 0.0 0.0 0.0 0.0 4 N O 1.0 1.2 1.1 1.0 1.1 1.1 1.1 2 c 78.3 59.7 34.2 32.1 34.7 35.2 40.6 Other Aircraft CO 77.5 59.1 33.9 31.8 34.4 34.9 40.2 2 CH 0.1 0.1 + + + + + 4 O 0.7 0.5 0.3 0.3 0.3 0.3 0.4 N 2 d 44.9 45.0 Ships and Boats 46.5 40.2 39.5 17.6 33.1 39.3 44.3 44.3 45.5 CO 38.7 16.9 32.3 2 CH + + + + + + + 4 0.6 0.6 0.8 0.7 0.7 0.5 0.6 N O 2 HFCs 0.0 0.1 0.1 0.1 0.1 0.1 0.1 Rail 51.3 46.6 45.6 46.7 48.5 46.7 38.9 38.5 50.3 44.7 43.4 44.2 45.7 43.6 CO 2 CH 0.1 0.1 0.1 0.1 0.1 0.1 0.1 4 O 0.4 0.3 0.3 0.3 0.3 0.4 0.3 N 2 0.0 0.5 1.5 1.8 HFCs 2.4 2.7 2.1 Other Emissions from Electricity e 0.1 + + + + + + Generation f 36.0 32.4 Pipelines 40.5 46.2 39.4 38.0 38.1 CO 36.0 32.4 38.1 46.2 39.4 38.0 40.5 2 11.8 10.2 9.0 8.3 8.8 9.1 10.0 Lubricants 8.3 11.8 10.2 9.0 CO 8.8 9.1 10.0 2 1,554.4 2,005.9 1,804.3 1,784.7 1,794.3 1,820.0 1,810.4 Total Transportation International Bunker g 114.2 112.8 106.8 104.5 100.7 104.2 111.8 Fuels h 4.1 22.4 71.5 71.5 73.4 74.9 Ethanol CO 75.9 2 h 0.0 0.9 8.3 8.5 13.5 Biodiesel CO 14.1 13.3 2 + Does not exceed 0.05 MMT CO Eq. 2 a There was a method update in this Inventory for estimating the share of gasoline used in on - road and non - road applications. The change does not impact total U.S. gasoline consumption. It mainly results in a shift in gasoline consumption from the transportation sector to industrial and commercial sec tors for 2015, creating a break in the time series. The change is discussed further in the Planned Improvements section of Chapter 3.1 . b Consists o f emissions from jet fuel consumed by domestic operations of commercial aircraft (no bunkers). c Consists of emissions from jet fuel and aviation gasoline consumption by general aviation and military aircraft. d iated with fluctuations in reported fuel consumption, and Fluctuations in emission estimates are assoc may reflect issues with data sources. e Other emissions from electricity generation are a result of waste incineration (as the majority of - to - municipal solid waste is combusted in “trash tricity generation plants), electrical steam” elec transmission and distribution, and a portion of Other Process Uses of Carbonates (from pollution control equipment installed in electricity generation plants). f CO es, but not electricity. While the operation of estimates reflect natural gas used to power pipelin 2 pipelines produces CH and N O, these emissions are not directly attributed to pipelines in the U.S. 2 4 Inventory. g tivities; Emissions from International Bunker Fuels include emissions from both civilian and military ac these emissions are not included in the transportation totals. h Ethanol and biodiesel CO and estimates are presented for informational purposes only. See Section 3.10 2 - Use Change, and Forestry (see Chapter 6), in line with IPCC the estimates in Land Use, Land methodological guidance and UNFCCC reporting obligations, for more information on ethanol and b iodiesel. Notes: Passenger cars and light - duty trucks include vehicles typically used for personal travel and less duty trucks include vehicles larger than 8,500 lbs. HFC emissions than 8,500 lbs; medium - and heavy - - primarily reflect HFC t sum due to independent rounding. 134a. Totals may no 31 - 2 Trends

110 Commercial The commercial 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 Energy consumption of natural gas and petroleum products, primarily for heating and cooking needs. related - emissions from the residential and commercial sectors have generally been increasing since 1990, and are often correlated w ith short - term fluctuations in energy consumption caused by weather conditions, rather than prevailing economic conditions. Landfills and wastewater treatment are included in this sector, with landfill emissions decreasing since 1990 and wastewater treatme nt 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 la rgely due to the direct consumption of natural gas and petroleum products, primarily for heating and cooking needs. Emissions from the residential sectors have generally been increasing since 1990, and are often correlated with short - term fluctuations in - e nergy consumption caused by weather conditions, rather than prevailing economic conditions. In the long term, this sector is also affected by population growth, regional migration trends, and changes in housing and building attributes (e.g., size and insul ation). 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 201 5 , agricultural soil management was the largest s ource of N emissions in the United O emissions, and enteric fermentation was the largest source of CH 4 2 States. This sector also includes small amounts of CO emissions from fossil fuel combustion by motorized farm 2 equipment like tractors. The agriculture s ector is less reliant on electricity than the other sectors. Box 2 - 1 : Methodology for Aggregating Emissions by Economic Sector In presenting the Economic Sectors in the annual Emissions and Sinks Inventory of U.S. Greenhouse Gas , the Discussing greenhouse gas Inventory expands upon the standard IPCC sectors common for UNFCCC reporting. specific sectors improves communication of the report’s findings. emissions relevant to U.S. - emissions from the combustion of fossil fuels included in the omic sector, CO In the Electricity Generation econ 2 EIA electric utility fuel consuming sector are apportioned to this economic sector. Stationary combustion emissions , CH and N O are also based on the EIA electric utility sec tor. Additional sources include CO of CH and N O 4, 2 2 2 4 from waste incineration, as the majority of municipal solid waste is combusted in “trash to - steam” electricity - generation plants. Transmission and The Electricity Generation economic sector also includes SF from Electrical 6 from Other Process Uses of Carbonates (from pollution control equipment Distribution, and a portion of CO 2 installed in electricity generation plants). In the Transportation economic sector, the CO emissions from the combustion of fossil fuels included in the EIA 2 transportation fuel consuming sector are apportioned to this economic sector (additional analyses and refinement of the EIA data is further explained in the Energy chapter of this report). Emissions of CH O from Mobile and N 2 4 Com bustion are also apportioned to this economic sector based on the EIA transportation fuel consuming sector. Substitution of Ozone Depleting Substances emissions are apportioned based on their specific end uses within the - source category, with emissions fro m transportation refrigeration/air - conditioning systems to this economic sector. Finally, CO Energy Uses of Fossil Fuels identified as lubricants for transportation vehicles are emissions from Non - 2 included in the Transportation economic sector. emissions from the combustion of fossil fuels included in the EIA For the In dustry economic sector, the CO 2 industrial fuel consuming sector, minus the agricultural use of fuel explained below, are apportioned to this The CH O emissions from and N stationary and mobile combustion are also apportioned to this economic sector. 2 4 economic sector based on the EIA industrial fuel consuming sector, minus emissions apportioned to the Agriculture - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 32 2

111 economic sector described below. s emis sions are apportioned based on Substitution of Ozone Depleting Substance - uses within the source category, with most emissions falling within the Industry economic sector. their specific end related emissions from sources with methods considered within the IPCC IPPU sector have - Additionally, all process This includes the process - related emissions (i.e., emissions from the actual been apportioned to this economic sector. process to make the material, not from fuels to power the plant) from such activities as Cement Production, Iron and on and Metallurgical Coke Production, and Ammonia Production. Additionally, fugitive emissions Steel Producti from energy production sources, such as Natural Gas Systems, Coal Mining, and Petroleum Systems are included in A portion of CO om Other Process Uses of Carbonates (from pollution control the Industry economic sector. fr 2 equipment installed in large industrial facilities) are also included in the Industry economic sector. Finally, all remaining CO emissions from Non - Energy Uses of Fossil Fuels are assumed to be industrial in nature (besides the 2 lubricants for transportation vehicles specified above), and are attributed to the Industry economic sector. ndustrial fuel consuming sector surveys, additional data is used to As agriculture equipment is included in EIA’s i extract the fuel used by agricultural equipment, to allow for accurate reporting in the Agriculture economic sector from all sources of emissions, such as motorized farming equipment. Ener gy consumption estimates are obtained from Department of Agriculture survey data, in combination with separate EIA fuel sales reports. This supplementary data is used to apportion some of the CO O emissions from fossil fuel combustion, and CH and N 4 2 2 ions from stationary and mobile combustion, to the Agriculture economic sector. The other emission sources emiss included in this economic sector are intuitive for the agriculture sectors, such as N O emissions from Agricultural 2 from Enteric Fermentat ion, CH Soils, CH and N from Rice Cultivation, CO O from Manure Management, CH 4 2 2 4 4 emissions from Liming and Urea Application, and CH emissions from and N O from Forest Fires. Nitrous oxide 2 4 the IPCC) are also included in the the Application of Fertilizers to tree plantations (termed “forest land” by Agriculture economic sector. The Residential economic sector includes the CO emissions from the combustion of fossil fuels reported for the 2 EIA residential sector. Stationary combustion emissions of CH based on the EIA residential fuel and N O are also 2 4 Substitution of Ozone Depleting Substances are apportioned based on their specific end consuming sector. uses - within the source category, with emissions from residential air - N itrous conditioning systems to this economic sector. emissions from the Application of Fertilizers to developed land (termed “settlements” by the IPCC) are also oxide included in the Residential economic sector. The Commercial economic sector includes the CO emissions from the combustion of fossil fuels reported in the 2 and N EIA commercial fuel consuming sector data. Emissions of CH O from Mobile Combustion are also 4 2 apportioned to this economic sector based on the EIA transportation fuel consuming sector. Su bstitution of Ozone Depleting Substances emissions are apportioned based on their specific end - uses within the source category, with - conditioning systems apportioned to this economic sector. emissions from commercial refrigeration/air Public works sources including direct CH from Landfills and CH O from Wastewater Treatment and Composting are also and N 2 4 4 included in this 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 consumption, because energy - related activities are ons; (2) emissions per unit of fossil fuel consumption, because almost all energy related the largest sources of emissi - emissions involve the combustion of fossil fuels; (3) emissions per unit of electricity consumption, because the — utilities and non - utilities c ombined — was the largest source of U.S. greenhouse gas electric power industry 5 emissions in 201 ; (4) emissions per unit of total gross domestic product as a measure of national economic activity; or (5) emissions per capita. Table - 14 provides data on various statistics related to U.S. greenhouse gas emissions normalized to 1990 as a 2 baseline year. These values represent the relative change in each statistic since 1990. Gr eenhouse gas emissions in the United States have grown at an average annual rate of 0. percent since 1990. Since 1990, this rate is slightly 2 slower than that for total energy and for fossil fuel consumption, and much slower than that for electricity consu mption, overall gross domestic product (GDP) and national population (see Table 2 - 14 and Figure 2 - 15 ). These trends vary relative to 2005, when greenhouse gas emissions, total energy and fossil fuel consumption began . annual rate of 1 decreased at an average percent since 0 to peak. Greenhouse gas emissions in the United States have 33 - 2 Trends

112 Total energy and fossil fuel consumption have also decreased at slower rates than emissions since 2005, while . 2005 electricity consumption, GDP, and national population continued to increase. - : Recent Trends in Various U.S. Data (Index 1990 = 100) 14 Table 2 Avg. Annual Avg. Annual Change Change a a 2011 2012 2013 2014 201 5 Variable since 1990 1990 2005 since 2005 b - 1.0% 115 107 103 105 106 104 0.2% Greenhouse Gas Emissions 100 c - 0.2% 115 118 115 100 112 117 115 0.6% Energy Consumption c - 0.7% 110 0.4% Fossil Fuel Consumption 100 119 110 107 110 111 c 0.3% Electricity Consumption 134 137 135 136 138 137 1.3% 100 d 1.4% 171 GDP 159 168 100 174 178 183 2.5% e 0.8% 1.0% Population 100 128 118 125 126 126 127 a Average annual growth rate b GWP - weighted values c - ) 7 weighted values (EIA 201 - Energy content d Gross Domestic Product in chained 2009 dollars (BEA 201 7 ) e U.S. Census Bureau (201 6 ) 15 - 2 Figure : U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product ), U.S. Census Bureau (201 report. Source: BEA (201 ), and emission estimates in this 6 7 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 34 2

113 2.3 Indirect Greenhouse Gas Emissions (CO, , NMVOCs, and SO NO ) x 2 10 request that information be provided on indirect greenhouse gases, The reporting requirements of the UNFCCC These gases do not have a direct . , NMVOCs, and SO global warming effect, but indirectly which include CO, NO 2 x affect terrestrial radiation absorption by influencing the formation and destruction of tropospheric and stratospheric , by affecting the absorptive characteristics of the atmosphere. Addit ionally, some of ozone, or, in the case of SO 2 these gases may react with other chemical compounds in the atmosphere to form compounds that are greenhouse Carbon monoxide is produced when carbon - containing fuels are combusted incompletely. gases. Nitrogen oxides created by lightning, fires, fossil fuel combustion, and in the stratosphere from N O. Non - (i.e., NO and NO ) are 2 2 which include hundreds of organic compounds that participate in — methane volatile organic compounds — are emitted uene, ethane, and many others) atmospheric chemical reactions (i.e., propane, butane, xylene, tol industrial consumption of organic solvents. In the - primarily from transportation, industrial processes, and non is primarily emitted from coal combustion for electric power generation and the me tals industry. United States, SO 2 - containing compounds emitted into the atmosphere tend to exert a negative radiative forcing (i.e., cooling) Sulfur and therefore are discussed separately. sors for tropospheric ozone is their role as precur One important indirect climate change effect of NMVOCs and NO x formation. They can also alter the atmospheric lifetimes of other greenhouse gases. Another example of indirect interaction of CO with the hydroxyl radical — the major greenhouse gas formation into greenhouse gases is the atmosphe emissions — to form CO ric sink for CH . Therefore, increased atmospheric concentrations of CO limit the 2 4 number of hydroxyl molecules (OH) available to destroy CH . 4 11 O (EPA 2015), , NMVOCs, and S Since 1970, the United States has published estimates of emissions of CO, NO x 2 Table 2 - 15 shows that fuel combustion accounts for the majority of which are regulated under the Clean Air Act. Industrial processes — such as the manufacture of chemical and allied emissions of these indirect greenhouse gases. products, metals processing, and industrial uses of solvents — are also significant sou rces of CO, NO , and x NMVOCs. 2 - : Emissions of NO Table , CO, NMVOCs, and SO 15 (kt) 2 x Gas/Activity 1990 2005 2011 2012 2013 2014 201 5 NO 21,790 17,443 12,482 12,038 11,387 10,810 9,971 x Mobile Fossil Fuel Combustion 10,295 6,871 7,294 10,862 6,448 6,024 5,417 Stationary Fossil Fuel Combustion 10,023 5,858 3,807 3,655 3,504 3,291 3,061 Oil and Gas Activities 139 321 622 663 704 745 745 Industrial Processes and Product Use 592 424 424 572 452 443 434 Forest Fires 239 172 80 276 185 188 188 Waste Combustion 128 82 73 82 91 100 100 Grassland Fires 21 54 5 13 27 27 39 Agricultural Burning 6 6 7 7 7 8 8 Waste 2 1 + 2 2 2 2 CO 75,570 54,119 52,586 132,877 48,620 46,922 44,954 Mobile Fossil Fuel Combustion 119,360 58,615 38,305 36,153 34,000 31,848 29,881 Forest Fires 8,486 6,136 2,832 9,815 6,655 6,642 6,642 Stationary Fossil Fuel Combustion 3,884 4,648 5,000 4,170 4,027 3,741 3,741 Waste Combustion 978 1,403 1,003 1,318 1,632 1,947 1,947 Industrial Processes and Product Use 1,557 1,229 1,246 4,129 1,262 1,273 1,273 Oil and Gas Activities 302 318 610 666 723 780 780 10 See < http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf >. 11 om Field Burning of Agricultural Residues were estimated separately, and therefore not and CO emission estimates fr NO x taken from EPA (201 ). 6b 35 - 2 Trends

114 Grassland Fires 358 84 894 657 217 442 442 Agricultural Burning 178 191 234 232 240 239 239 Waste 7 5 6 1 8 9 9 NMVOCs 13,154 11,464 11,726 20,930 11,202 10,935 10,647 Industrial Processes and Product Use 3,793 5,849 7,638 3,929 3,861 3,723 3,723 Mobile Fossil Fuel Combustion 5,724 4,243 4,562 10,932 3,924 3,605 3,318 Oil and Gas Activities 510 2,517 554 2,651 2,786 2,921 2,921 Stationary Fossil Fuel Combustion 539 716 912 599 569 507 507 Waste Combustion 222 241 81 94 108 121 121 Waste 114 38 673 45 51 57 57 Agricultural Burning NA NA NA NA NA NA NA SO 2 4,357 5,874 5,876 3,448 20,935 13,196 5,877 Stationary Fossil Fuel Combustion 11,541 5,008 18,407 5,006 5,005 3,640 2,756 Industrial Processes and Product Use 604 831 1,307 604 604 496 496 Mobile Fossil Fuel Combustion 390 180 108 108 108 93 93 and Gas Activities Oil 619 142 142 142 95 70 793 Waste Combustion 25 38 15 15 32 32 15 Waste 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 2015) except for estimates from Field Burning of Agricultural Residues. - 2 3 : Sources and Effects of Sulfur Dioxide Box Sulfur dioxide (SO ) emitted into the atmosphere through natural and anthropogenic processes affects the earth's 2 radiative budget through its photochemical transformation into sulfate aerosols that can (1) scatter radiation from the sun back to space, thereby reducing the radiation reachin g the earth's surface; (2) affect cloud formation; and (3) affect atmospheric chemical composition (e.g., by providing surfaces for heterogeneous chemical reactions). The indirect effect of sulfur - derived aerosols on radiative forcing can be considered in two parts. The first indirect effect is the aerosols’ tendency to decrease water droplet size and increase water droplet concentration in the atmosphere. The second indirect effect is the tendency of the reduction in cloud droplet size to affect precipitation by increasing cloud lifetime and thickness. Although still highly uncertain the radiative forcing estimates from both the first and 2013 ). the second indirect effect are believed to be negative, as is the combined radiative forcing of the two (IPCC Sulfur dioxide is also a major contributor to the formation of regional haze, which can cause significant increases in Once SO is emitted, it is chemically transformed in the atmosphere and acute and chronic respiratory diseases. 2 th as the primary source of acid rain. Because of these harmful effects, the United States has returns to the ear regulated SO emissions in the Clean Air Act. 2 Electricity generation is the largest anthropogenic source of SO emissions in the United States, accounting for 59 .2 2 92 percent). Coal combustion contributes nearly all of those emissions (approximately Sulfur percent in 201 5 . dioxide emissions have decreased in recent years, primarily as a result of electric power generators switching from - high - oal and installing flue gas desulfurization equipment. sulfur c sulfur to low - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 36 2

115 3. Energy - related activities were the primary sources of U.S. anthropogenic greenhouse gas emissions, accounting for Energy 73 percent of total greenhouse gas emissions on a carbon dioxide (CO included ) equivalent basis in 2015 . 84.2 This 2 , methane (CH 42 , and 12 percent of the nation's CO - 97 Energy ), and nitrous oxide (N O) emissions, respectively. , 2 4 2 related CO equivalent emissions alone con stituted 79.4 percent of national emissions from all sources on a CO 2 2 related activities represented a much smaller portion of total - CO emissions from energy basis, while the non - 2 national emissions ( 4.8 percent collectively). Emissions from fos sil fuel combustion comprise the vast majority of energy - related emissions, with CO being the 2 32,381 Figure 3 - 1 ). Globally, approximately primary gas emitted (see million metric tons (MMT) of CO were 2 added to the atmosphere through the combustion of fossil fuels in 2014 , of which the United States accounted for 74 approxi mately 16 percent. Due to their relative importance, fossil fuel combustion - related CO emissions are 2 considered separately, and in more detail than other energy related emissions (see Figure 3 - 2 ). Fossil fuel - largest source of N and N combustion also emits CH O. Stationary combustion of fossil fuels was the second O 2 4 2 emissions in the United States and mobile fossil fuel combustion was the fourth large st source. Figure 3 - 1 : 2015 Energy Chapter Greenhouse Gas Sources (MMT CO Eq.) 2 73 Estimates are presented in units of million metric tons of carbon dioxide equivalent (MMT CO Eq.), which weight each gas 2 by its global warming potential, or GWP, value. See section on global warming potentials in the Executive Summary. 74 Global CO Emissions from Fossil emissions from fossil fuel combustion were taken from International Energy Agency CO 2 2 Fuels Combustion – Highlights < - - emissions - from - fuel https://www.iea.org/publications/freepublications/publication/co2 highlights 2016 > IEA ( 2016.html - ). combustion - 1 - 3 Energy

116 3 2 : 201 5 U.S. Fossil Carbon Flows (MMT CO Figure Eq.) - 2 activities other than fuel combustion, such as the production, transmission, storage, and distribution - Energy related These emissions consist primarily of fugitive CH from natural gas of fossil fuels, also emit greenhouse gases. 4 Table 3 systems, petroleum systems, and coal mini 1 summarizes emissions from the Energy sector in units of ng. - 2 Eq., while unweighted gas emissions in kilotons (kt) are provided in Table 3 - . Overall, emissions due to MMT CO 2 75 related activities were energy MMT CO 5,549.1 Eq. in 2015 , - an increase of 4.1 percent since 1990. 2 , and N 1 : CO Table , CH - Eq.) O Emissions from Energy (MMT CO 3 2 2 4 2 2012 Gas/Source 1990 2005 2011 2013 2014 2015 4,907.2 5,932.3 5,231.9 5,387.2 5,180.9 5,332.7 5,377.8 CO 2 a 4,740.3 5,746.9 5,227.1 5,024.6 5,156.5 5,202.3 5,049.8 Fossil Fuel Combustion 2,157.7 Electricity Generation 1,820.8 2,400.9 2,022.2 2,038.1 2,038.0 1,900.7 Transportation 1,493.8 1,887.0 1,707.6 1,696.8 1,713.0 1,742.8 1,736.4 775.0 Industrial 842.5 828.0 782.9 812.2 806.1 805.5 Residential 338.3 357.8 325.5 282.5 329.7 345.4 319.6 Commercial 217.4 223.5 220.4 196.7 221.0 228.7 246.2 41.4 U.S. Territories 27.6 49.7 40.9 43.5 42.5 41.4 Non Energy Use of Fuels 117.6 138.9 109.8 106.7 123.6 119.0 125.5 - Natural Gas Systems 37.7 30.1 35.2 38.5 42.4 42.4 35.7 8.0 12.5 10.6 10.4 10.4 10.6 10.7 Incineration of Waste Petroleum Systems 3.9 4.2 3.9 3.6 3.7 3.6 3.6 b - Wood 215.2 206.9 195.2 Biomass 194.9 211.6 217.7 198.7 b 103.5 113.1 111.7 International Bunker Fuels 105.8 99.8 103.2 110.8 b - Ethanol 72.9 4.2 22.9 Biofuels 72.8 74.7 76.1 78.9 b Biodiesel Biofuels 0.0 0.9 8.3 8.5 13.5 13.3 14.1 - CH 367.3 286.6 289.5 284.1 284.6 286.8 278.6 4 194.1 159.7 154.5 Natural Gas Systems 156.2 159.2 162.5 162.4 Coal Mining 96.5 64.1 71.2 66.5 64.6 64.8 60.9 55.5 46.0 48.0 46.4 44.5 43.0 39.9 Petroleum Systems Stationary Combustion 8.5 7.4 7.1 6.6 8.0 8.1 7.0 75 equivalent values based Following the revised reporting requirements under the UNFCCC, this Inventory report presents CO 2 IPCC Fourth Assessment Report n the o (AR4) GWP values. See the Introduction chapter for more information. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 2 3

117 Abandoned Underground 7.2 Coal Mines 6.6 6.4 6.2 6.2 6.3 6.4 a 5.6 2.8 2.3 Combustion 2.1 2.1 2.0 Mobile 2.2 + + + + + + + Incineration of Waste b 0.2 0.1 0.1 0.1 0.1 0.1 0.1 International Bunker Fuels N 53.6 56.4 44.4 O 42.1 41.7 40.3 38.6 2 Stationary Combustion 11.9 20.2 21.3 21.4 22.9 23.4 23.1 a Mobile 41.2 35.7 22.8 20.4 18.5 16.6 15.1 Combustion 0.3 0.4 0.3 0.5 0.3 0.3 0.3 Incineration of Waste b International Bunker Fuels 1.0 1.0 0.9 0.9 0.9 0.9 0.9 Total 5,328.1 6,275.3 5,721.2 5,507.0 5,659.1 5,704.9 5,549.1 + Does not exceed 0.05 MMT CO Eq. 2 a road and non - road In 2016, FHWA changed its methods for estimating the share of gasoline used in on - series inconsistency between 2015 and previous years in this Inventory. applications, which created a time The - Improvements under CO 3.1 from Fossil Fuel sections of Chapter method updates are discussed in the Planned 2 CH and N and O from Mobile Combustion. Combustion 4 2 b These values are presented for informational purposes only, in line with IPCC methodological guidance and CC reporting obligations, and are not included in the specific energy sector contribution to the totals, and are UNFC already accounted for elsewhere. 3 - 2 : CO , CH Table , and N Energy (kt) O Emissions from 2 2 4 Gas/Source 1990 2005 2011 2012 2013 2014 2015 5,387,235 CO 4,907,164 5,932,326 5,180,851 5,332,717 5,377,822 5,231,882 2 a 4,740,343 5,746,942 5,227,061 5,024,643 5,156,523 5,202,300 5,049,763 Fossil Fuel Combustion Non - Energy Use of Fuels 117,585 138,913 109,756 106,750 123,645 118,995 125,526 37,732 30,076 35,662 35,203 38,457 42,351 42,351 Natural Gas Systems Incineration of Waste 7,950 12,469 10,564 10,379 10,398 10,608 10,676 3,693 3,567 3,567 Petroleum Systems 3,553 3,927 4,192 3,876 b Biomass 215,186 206,901 195,182 194,903 211,581 217,654 198,723 - Wood b 113,139 111,660 105,805 103,463 103,201 110,751 International Bunker Fuels 99,763 b Biofuels 4,227 22,943 72,881 - 74,743 76,075 78,934 Ethanol 72,827 b Biodiesel Biofuels 0 856 - 8,470 13,462 13,349 14,077 8,349 CH 14,693 11,464 11,581 11,364 11,385 11,473 11,145 4 Natural Gas Systems 6,387 6,180 6,247 6,368 6,501 6,497 7,762 2,658 Coal Mining 2,565 2,849 3,860 2,584 2,593 2,436 Petroleum Systems 2,218 1,840 1,922 1,778 1,721 1,595 1,858 339 296 283 265 320 323 280 Stationary Combustion Abandoned Underground 249 264 257 Coal Mines 288 249 253 256 a Combustion Mobile 226 113 91 87 85 82 80 Incineration of Waste + + + + + + + b 7 5 5 International Bunker Fuels 3 3 3 4 N O 180 189 149 141 140 135 129 2 40 68 71 Stationary Combustion 77 78 78 72 a Combustion 138 120 77 68 Mobile 62 56 51 Incineration of Waste 2 1 1 1 1 1 1 b 3 3 3 3 International Bunker Fuels 3 3 3 + Does not exceed 0.5 kt a In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - road applications , which created a time series inconsistency between 2015 and previous years in this Inventory. The method updates are discussed in the Planned - Imp rovements sections of Chapter 3.1 under CO from Fossil Fuel Combustion and CH O from Mobile Combustion. and N 2 2 4 b These values are presented for informational purposes only, in line with IPCC methodological guidance and UNFCC C reporting obligations, and are not included in the specific energy sector contribution to the totals, and are already accounted for els ewhere . - 1 : Methodological Approach for Estimating and Reporting U.S. Emissions and Sinks Box 3 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 sinks presented in this 3 - 3 Energy

118 report and - accepted this chapter, are organized by source and sink categories and calculated using internationally methods provided by the Intergovernmental Panel on Climate Change (IPCC). Additionally, the calculated emissions and sinks in a given year for the Unit ed States are presented in a common manner in line with the The use of UNFCCC reporting guidelines for the reporting of inventories under this international agreement. consistent methods to calculate emissions and sinks by all nations providing their inven tories to the UNFCCC ensures that these reports are comparable. In this regard, U.S. emissions and sinks reported in this I nventory are comparable to emissions and sinks reported by other countries. Emissions and sinks provided in this Inventory do not pre clude alternative examinations, but rather, this Inventory presents emissions and sinks 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 IPCC methods used to calculate emissions and sinks, and the manner in which those calculations are conducted. Box 3 - 2 : Energy Data from EPA’s Greenhouse Gas Reporting Program On October 30, 2009, the U.S. Environmental 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 nhouse Gas Reporting Program (GHGRP). 40 CFR Part 98 applies to direct 40 CFR Part 98 is referred to as the Gree greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers, and facilities that inject CO underground for 2 el, except for certain suppliers of fossil fuels and sequestration or other reasons. Reporting is at the facility lev industrial greenhouse gases. 40 CFR part 98 requires reporting by 41 industrial categories. Data reporting by facility. In general, the affected facilities included the reporting of emissions from fuel combustion at that affected Eq. per year. threshold for reporting is 25,000 metric tons or more of CO 2 EPA’s GHGRP dataset and the data presented in this Inventory report are complementary . The GHGRP dataset continue to be an important resource for the In ventory, providing not only annual emissions information, but also s 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 categories of emissions , along with enhancing application ways that can highlight differences across regions and sub - of QA/QC procedures and assessment of uncertainties. y estimates and uses annual GHGRP data in a number of categor EPA continues to analyze the data on an annual basis, as applicable, for further es presented in this Inventory consistent with use to improve the national estimat 76 Box 3 - 4 IPCC guidance . (see, also, A s indicated in the respective P lanned I mprovements sections for source ) categories in this chapter, EPA is considering further use of facility - level GHGRP data to improve the national estim at es 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 UNFCCC reporting guidelines. In line with the GHGRP may differ from those used in the Inventory in meeting the I UNFCCC reporting guidelines, the nventory 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. Fu rther information on the reporting categorizations in EPA’s GHGRP and specific data caveats associated with monitoring methods in 77 EPA’s GHGRP has been provided on the GHGRP website. EPA presents the data collected by its GHGRP through a data publication tool that allows data to be viewed in 78 several formats including maps, tables, charts and graphs for individual facilities or groups of facilities . One area where the GHGRP fuel consumption activity data is used in the Energy sector is in disaggregating use industrial end - sector emissions in the category of CO Emissions from Fossil Fuel Combustion, for use in 2 . The industrial end Common Reporting F ormat (CRF) tables reporting emissions in - use sector activity data 76 . > nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf See < http://www.ipcc - 77 See < cription+of+Data+for+Certain+Sources+and+Processes>. http://www.ccdsupport.com/confluence/display/ghgp/Detailed+Des 78 See . - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 4 3

119 aggregated collected for the Inventory (EIA 2017a) represent - use sector. EPA’s GHGRP data for the industrial end industrial end - use sector. collects industrial fuel consumption activity data by individual categories w ithin the - Therefore, the GHGRP data is used to provide a more detailed breakout of total emissions in the industrial end use sector within that source category. Fossil Fuel Combustion (IPCC Source 3.1 Category 1A) combustion of fossil fuels for energy include the gases CO Emissions from the , CH is , and N O. Given that CO 2 2 4 2 the primary gas emitted from fossil fuel combustion and represents the largest share of U.S. total emissions, CO 2 sed at the beginning of this section. Following that is a discussion emissions from fossil fuel combustion are discus of emissions of all three gases from fossil fuel combustion presented by sectoral breakdowns. Methodologies for of CH and N O emissions from estimating CO from fossil fuel combustion also differ from the estimation 4 2 2 Thus, three separate descriptions of methodologies, uncertainties, stationary combustion and mobile combustion. Total CO , CH recalculations, and planned improvements are provided at the end of this section. , and N O 2 2 4 Table s from fossil fuel combustion are presented in 3 - 3 and Table 3 - 4 . emission 3 - 3 : CO , CH Eq.) , and N Table O Emissions from Fossil Fuel Combustion (MMT CO 2 2 2 4 Gas 1990 2005 2011 2012 2013 2014 2015 CO 2 5,202.3 5,746.9 4,740.3 5,227.1 5,024.6 5,156.5 5,049.8 CH 4 10.2 14.1 9.3 8.8 10.1 9.0 10.1 N O 2 53.1 44. 1 41. 7 56.0 41. 4 40.0 38. 2 Total 5,813. 4,807.6 5,097. 1 0 5,280. 5 5,075.2 5,208.1 5,252.4 Totals may not sum due to independent rounding Note: 3 - 4 : CO O Emissions from Fossil Fuel Combustion (kt) , CH , and N Table 2 4 2 Gas 1990 2005 1 201 201 2 201 3 201 4 201 5 CO 2 5,746,942 4,740,343 5,024,643 5,156,523 5,202,300 5,049,763 5,227,061 CH 4 405 361 408 565 374 352 405 O N 2 188 178 148 140 139 128 134 from Fossil Fuel Combustion CO 2 arbon dioxide is the primary gas emitted from fossil fuel combustion and represents the largest share of U.S. total C C . arbon dioxide emissions from fossil fuel combustion are presented in Table 3 - 5 greenhouse gas emissions. In percent relative to the previous year. 201 , CO The emissions from fossil fuel combustion decreased by 2.9 5 2 decrease in CO emissions from fossil fuel combustion was a result of multiple factors, including: (1) substitution 2 from coal to natural gas consumption in the electric power sector; (2) warmer winter conditions in 2015 resulting in a decreased demand for heating fuel in the residential and commercial sectors; and (3) a slight decrease in electricity demand. percent above In 201 5 , CO emissions from fossil fuel combust ion were 5,049.8 MMT CO 6.5 Eq., or 2 2 79 ). emissions in 1990 (see Table 3 - 5 79 An additional discussion of fossil fuel emission trends is presented in the Trends in U.S. Greenhouse Gas Emissions chapter. 5 - 3 Energy

120 Table 3 : CO - Emissions from Fossil Fuel Combustion by Fuel Type and Sector (MMT CO 5 2 2 Eq.) 1990 2005 2011 2012 2013 2014 2015 Fuel/Sector Coal 1,813.9 1,592.8 1,652.6 2,112.3 1,423.3 1,718.4 1,653.8 Residential 0.8 3.0 NO NO NO NO NO Commercial 5.8 4.1 3.9 12.0 3.8 9.3 2.9 Industrial 115.3 155.3 82.0 74.1 75.6 65.9 75.7 Transportation NE NE NE NE NE NE NE Electricity Generation 1,722.7 1,983.8 1,547.6 1,511.2 1,571.3 1,569.1 1,350.5 U.S. Territories 3.0 3.4 3.4 2.8 4.0 4.0 0.6 Natural Gas 1,166.7 1,000.3 1,291.5 1,352.6 1,391.2 1,463.6 1,422.0 Residential 277.9 262.2 254.7 224.8 238.0 266.2 252.8 Commercial 170.5 162.9 142.1 156.9 179.1 189.3 175.4 Industrial 468.4 388.5 417.3 434.8 408.9 467.5 451.9 Transportation 38.9 33.1 36.0 41.3 47.0 40.3 38.8 Electricity Generation 318.8 175.3 492.2 444.0 443.2 526.1 408.8 U.S. Territories 3.0 NO 3.0 1.3 1.4 2.6 3.0 a Petroleum 2,467.6 2,121.3 2,078.8 2,127.3 2,162.5 2,021.2 2,111.1 Residential 94.9 97.4 70.9 57.7 63.4 67.5 66.8 Commercial 51.3 63.3 44.1 38.0 35.6 67.9 35.7 Industrial 262.1 278.3 275.7 274.1 284.6 272.2 324.2 Transportation 1,854.0 1,457.7 1,668.8 1,655.4 1,666.0 1,702.5 1,697.6 Electricity Generation 97.9 97.5 25.8 18.3 22.4 25.3 23.7 U.S. Territories 34.3 45.4 26.9 36.0 37.5 36.6 34.3 b Geothermal 0.4 0.4 0.4 0.4 0.4 0.4 0.4 Total 5,746.9 4,740.3 5,227.1 5,024.6 5,156.5 5,202.3 5,049.8 + Does not exceed 0.05 MMT CO Eq. 2 stimated) NE (Not E ccurring) NO (Not O a In 2016, FHWA changed its methods for estimating the share of - road and non - road gasoline used in on applications, which created a time - series inconsistency between 2015 and previous years in this Inventory. The method changes resulted in a decrease in the estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. The method updates are discussed further in the Planned Improvements section below under CO from Fossil Fuel 2 Combustion. b - related CO Although not technically a fossil fuel, geothermal energy emissions are included for reporting 2 purposes. Note: Totals may not sum due to independent rounding. Trends in CO emissions from fossil fuel combustion are influenced by many long - term and short - term factors. On a 2 year - to - year basis, the overall demand for fossil fuels in the United States and other countries generally fluctuates in response to - fossil changes in general economic conditions, energy prices, weather, and the availability of non 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 consumption patterns, however, tend to be more a function of aggregate societal - trends that affect the scale of consumption (e.g., population, number of cars, size of houses, and number of houses), the ef ficiency 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). d on the source of energy and its carbon (C) intensity. The amount of C in fuels emissions also depen arbon dioxide C varies significantly by fuel type. For example, coal contains the highest amount of C per unit of useful energy. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 6 3

121 80 Petroleum has roughly 75 percent of the C per unit of energy as coal, and natural gas has only about 55 percent. - 6 shows annual changes in emissions during the last five years for coal, petroleum, and n atural gas in 3 Table selected sectors. Emissions and Total 201 Table 3 - 5 Emissions from Fossil Fuel 6 : Annual Change in CO 2 Eq. and Percent) Combustion for Selected Fuels and Sectors (MMT CO 2 2013 to 2014 Type Sector 2011 to 2012 2012 to 2013 Fuel 2014 to 2015 Total 2015 Electricity Generation Coal 0.1% 211.5 12.3% 60.1 4.0% - 2.2 - - - 218.7 - 13.9% 1,350.5 - Electricity Generation Natural Gas - 48.3 - 9.8% - 0.8 20.4% - 0.2% 82.9 18.7% 526.1 83.5 Electricity Generation Petroleum - 7.5 - 29.0% 4.1 22.3% 2.9 12.8% - 1.6 - 6.4% 23.7 Residential Natural Gas 4.4% 29.8 - 11.7% 41.4 18.4% - 11.6 - 25.1 - 9.0% 252.8 Commercial Natural Gas 5.7% - - 8.0% 22.3 14.2% 10.2 - 13.9 - 7.4% 175.4 13.6 Industrial Coal - - 9.7% 1.7 2.3% - 0.1 7.9 - 0.1% - 9.8 - 12.9% 65.9 Industrial Natural Gas 0.2% 467.5 17.5 4.2% 17.1 3.9% 16.5 3.7% - - 0.9 a a All Fuels All Sectors 202.4 - 3.9% 131.9 2.6% 45.8 0.9% - 152.5 - - 5,049.8 2.9% a Includes fuels and sectors not shown in table. Note: Totals may not sum due to independent rounding. As shown in Table 3 - 6 , recent trends in CO emissions from fossil fuel combustion show a 3.9 percent decrease 2 from 2011 to 2012, then a 2.6 percent and a 0.9 percent increase from 2012 to 2013 and 2013 to 2014, respectively, n remained relatively flat over that time and a 2.9 percent decrease from 2014 to 2015. Total electricity generatio period but emission trends generally mirror the trends in the amount of coal used to generate electricity. The by roughly 1 2 percent from 2011 to 2012, in creased by 4 consumption of coal used to generate electricity decreased percent from 2012 to 2013, stayed relatively flat from 2013 to 2014, and decreased by 1 4 percent from 2014 to 2015. The overall CO emission trends from fossil fuel combustion also follow closely changes in heating degree 2 days over that time period. H eating degree days decreased by 13 percent from 2011 to 2012, increased by 18 percent from 2012 to 2013, increased by percent from 2013 to 2014 and decreased by 10 percent from 2014 to 2 2015. A decrease in heating degree days leads to decreased demand for heating fuel and electricity for heat in the residential and commercial sector . The overall CO emission trends from fossil fuel , primarily in winter months 2 combustion also generally follow changes in overall petroleum use and emissions. CO missions from petroleum e 2 2.0 decreased by percent percent from 2011 to 2012, increased by 1.6 percent from 2012 to 2013, increased by 0.8 from 2013 to 2014, and increased by 1.7 emissions from percent from 2014 to 2015. The increase in petroleum CO 2 o 2015 somewhat offsets emission reductions from other sources like decreased coal use in the electricity 2014 t sector. In the United States, 82 percent of the energy consumed in 2015 was produced through the combustion of fossil s, and petroleum (see Figure 3 - 3 and Figure 3 - 4 ). The remaining portion was supplied fuels such as coal, natural ga 9 percent) and by a variety of renewable energy sources ( 10 percent), primarily by nuclear electric power ( 81 2017a EIA Specifically, petroleum supplied the largest share of ). hydroelectric power, wind energy and biofuels ( 37 percent of total U.S. energy consumption in 2015. Natural gas and coal domestic energy demands, accounting for 29 percent and followed in order of energy demand importance, accounting for approximately 16 percent of total U.S. energy consumption, respectively. - use sector and Petroleum was consumed primarily in the transportation end - se sectors the vast majority of coal was used in electricity generation. Natural gas was broadly consumed in all end u (see Figure 3 except transportation 5 ) (EIA 2017a ). - CO emissions from fossil fuel emissions from motor gasoline consumption are 23 percent of the CO 2 2 combustion. The majority of gasoline is used in the transportation sector. Note that a method update in the current Inventory impacted the allocation of gasoline between the transportation, commercial, and industrial sectors i n 2015 , 80 of all coal, natural gas, and petroleum fuels combusted in the United States. Based on national aggregate carbon content 81 Renewable energy, as defined in EIA’s energy statistics, includes the following energy sources: hydroelectric power, . geothermal energy, biofuels, solar energy, and wind energy 7 - 3 Energy

122 82 The recent trend however overall gasoline use, trends, and CO emissions were not impacted. since 2012 is an 2 increase in total gasoline use and CO emissions (see Figure 3 - 6 ). 2 Figure 3 - 3 5 U.S. Energy Consumption by Energy Source ( P ercent) : 201 Figure 3 4 : U.S. Energy Consumption (Quadrillion Btu) - 82 In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - road applications, which created a time series inconsistency between 2015 and previous years in this Inventory. The method changes resulted in a - decrease in the e stimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. The method updates are discussed further in the Planned Improvements section rom Fossil Fuel Combustion below f CO under . 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 8 3

123 CO 3 - 5 : 201 5 Figure Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT 2 CO Eq.) 2 3 6 Eq.) Figure - (MMT CO : U.S. Gasoline Consumption (Tbtu) and CO Emissions by Sector 2 2 During the Fossil fuels are generally combusted for the purpose of producing energy for useful heat and work. and smaller amounts of other gases, combustion process, the C stored in the fuels is oxidized and emitted as CO 2 9 - 3 Energy

124 83 - , CO, and NMVOCs. including These other C containing non CH CO gases are emitted as a byproduct of 2 4 incomplete fuel combustion, but are, for the most part, eventually oxidized to CO in the atmosphere. Therefore, it is 2 assumed all of the C in fossil fuels used to produce energy is eventually converted to atmospheric CO . 2 Box 3 - 3 : Weather and Non - Fossil Energy Effects on CO from Fossil Fuel Combustion Trends 2 ter of the year in particular, caused a significant In 2015, weather conditions, and a warm first and fourth quar decrease in demand for heating fuels and is reflected in the decreased residential emissions from 2014 to 2015. The United States in 2015 also experienced a warmer winter overall compared to 2014, as heat ing degree days decreased ( 10.2 percent). Warmer winter conditions compared to 2014 resulted in a decrease in the amount of energy required Cooling for heating, and heating degree days in the United States were 9.7 percent below normal (see Figure 3 - 7 ). which and despite this increase in cooling degree days, , percent degree days increased significantly , by 14.6 leads to increased demand for air conditioning in the residen tial and commercial sector, residential typically electricity demand decreased slightly. Summer conditions were significantly warmer in 2015 compared to 2014, 84 percent above normal 22.5 (see Figure 3 - 8 ) (EIA 2017a ). with cooling degree days - 3 : Annual Deviations from Normal Heating Degree Days for the United States 7 Figure (1950 201 5, Index Normal = 100 ) – 83 See the sections entitled Stationary Combustion and Mobile Combustion in this chapter for information on non - CO gas 2 emissions from fossil fuel combustion. 84 Degree days are relative measurements of outdoor air temperature. H eating degree days are deviations of the mean daily degrees Fahrenheit, while cooling degree days are deviations of the mean daily temperature above 65 temperature below 65 gy demand and related emissions than do . Heating degree days have a considerably greater effect on ener degrees Fahrenheit cooling degree days. Excludes Alaska and Hawaii. Normals are based on data from 1971 through 2000. The variation in these 14 percent for heating and cooling deg  10 percent and  normals during this time period was ree days, respectively (99 percent confidence interval). - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 10 3

125 8 : Annual Deviations from Normal Cooling Degree Days for the United States - 3 Figure 201 ) 5, Index Normal = 100 (1950 – 85 were brought online in 2015 , the utilization (i.e., ca pacity factors) Although no new U.S. nuclear power plants of creased in existing plants in 201 5 remained high at 92 percent. Electricity output by hydroelectric power plants de percent. In recent years, the wind power sector has been showing strong growth, such that, 201 5 4 by approximately on the margin, it is becoming a relatively important electricity source. Electricity generated by nuclear plants in 3 times as much of the energy generated in the United States from hydroelectric plants 201 5 provided more than , and wind power capacity factors since 1990 are shown in EIA 2016a ). Nuclear, hydroelectric ( Figure 3 - 9 . 85 The capacity factor equals generation divided by net summer capacity. Summer capacity is defined as "The maximum output - ti that generating equipment can supply to system load, as demonstrated by a mul hour test, at the time of summer peak demand (period of June 1 through September 30)." Data for both the generation and net summer capacity are from EIA (2016a). 11 - 3 Energy

126 Figure - : Nuclear, Hydroelectric, and Wind Power Plant Capacity Factors in the United 3 9 201 States (1990 5, Percent – ) Fossil Fuel Combustion Emissions by Sector In addition to the CO emitted from fossil fuel combustion, CH O are emitted from stationary and mobile and N 2 2 4 3 - 7 provides an overview of the CO , CH combustion as well. , and N Table O emissions from fossil fuel 2 2 4 combustion by sector. O Emissions from Fossil Fuel Combustion by Sector (MMT CO 7 Table , CH 3 - : CO , and N 2 2 4 2 Eq.) 2005 2011 1990 2013 2014 2015 2012 End Use Sector - Electricity Generation 2,057.7 3 2,417. 2,175. 7 1,828.5 4 2,058. 0 1,920.6 2,040. CO 2 1,820.8 2,157.7 2,022.2 2,038.1 2,038.0 1,900.7 2,400.9 CH 4 0.4 0.4 0.4 0.4 0.4 0.3 0.5 O N 2 7.4 17.6 17.8 19.1 19.6 19.5 16.0 a Transportation 1,925.6 1,732.7 1,540.6 1,719.3 1,733.6 1,761.5 1,753.5 CO 2 1,887.0 1,493.8 1,707.6 1,696.8 1,742.8 1,736.4 1,713.0 CH 4 5.6 2.3 2.2 2.8 2.1 2.1 2.0 N O 2 22.8 35.7 41.2 20.4 18.5 16.6 15.1 a Industrial 832.6 778.9 786.9 816.2 810.0 809.3 847.4 CO 2 842.5 775.0 782.9 812.2 806.1 805.5 828.0 CH 4 1.7 1.8 1.5 1.5 1.5 1.5 1.5 N O 2 2.4 3.1 2.4 2.4 2.4 2.4 2.9 Residential 344.6 330.4 287.0 335.6 351.3 324.3 362.8 CO 2 329.7 357.8 338.3 325.5 282.5 345.4 319.6 CH 4 5.0 5.2 4.0 3.7 5.0 3.9 4.1 N O 2 0.9 0.8 0.7 1.0 1.0 0.8 1.0 a Commercial 224.9 218.8 221.7 197.9 222.4 230.0 247.8 CO 2 217.4 220.4 196.7 223.5 221.0 228.7 246.2 CH 4 1.1 1.1 1.0 0.9 1.0 1.0 1.2 N O 2 0.3 0.4 0.3 0.4 0.3 0.3 0.3 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 12 3

127 b U.S. Territories 49.9 43.7 42.6 41.5 41.5 27.7 41.0 Total 5,813. 5,280. 5 5,075.2 5,208.1 5,252.4 5,097. 0 4,807.6 1 a In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - road - applications, which created a time series inconsistency between 2015 and previous years in this Inventory. The method changes resulted in a decrease in th e estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. The under CO from Fossil Fuel below method updates are discussed further in the Planned Improvements section 2 Combustion. b U.S. Territories are not apportioned by sector, and emissions are total greenhouse gas emissions from all fuel combustion sources. : Totals may not sum due to independent rounding. Note by electricity s Emissions from fossil fuel combustion - generation are allocated based on aggregate national electricity consumption by each end use sector. , gases emitted from stationary combustion include the greenhouse gases CH O and the Other than CO and N 2 2 4 86 indirect greenhouse , CO, and NMVOCs. gases NO Methane and N O emissions from stationary combustion 2 x sources depend upon fuel characteristics, size and vintage, along with combustion technology, pollution control maintenance practices. N emissions itrous oxide equipment, ambient environmental conditions, and operation and fuel mixes and combustion temperatures, as well as the - from stationary combustion are closely related to air Methane emissions from stationa ry combustion characteristics of any pollution control equipment that is employed. content of the fuel and combustion efficiency. are primarily a function of the CH 4 , including CH , N O, and indirect greenhouse gases Mobile combustion produces greenhouse gases other than CO 2 4 2 emissions from mobile , CO, and NMVOCs. As with stationary combustion, N including NO O and NO x x 2 combustion are closely related to fuel characteristics, air - fuel mixes, combustion temperatures, and the use of N pollution control equipment. O from mobile sources, in particular, can be formed by the catalyt ic processes used 2 to control NO Carbon monoxide emissions from mobile combustion are , CO, and hydrocarbon emissions. x combustion emission controls. Carbon significantly affected by combustion efficiency and the presence of post - st when air - fuel mixtures have less oxygen than required for complete combustion. monoxide emissions are highe These emissions occur especially in idle, low speed, and cold start conditions. Methane and NMVOC emissions from motor vehicles are a function of the CH content of the motor fuel, the amount of hydrocarbons passing 4 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 electricity generation to the sectors in which it is used. Four end - use sectors were defined: industrial, transportation, residential, and commercial. In the table below, electricity generation emissions have been distributed to each end - use sector ba sed upon the sector’s share of national electricity consumption, with the exception of CH and N O from 4 2 87 transportation. use Emissions from U.S. Territories are also calculated separately due to a lack of end - - specific consumption data. that emissions from combustion sources are distributed across the four end - This method assumes The results of this alternative method are use sectors based on the ratio of electricity consumption in that sector. 3 8 . Table presented in - - 8 : CO , CH Table 3 , and N Use Sector O Emissions from Fossil Fuel Combustion by End - 2 2 4 (MMT CO Eq.) 2 201 5 201 4 1990 2005 201 1 201 2 201 3 End Use Sector - a Transportation 1,737.7 1,723.2 1,737.0 1,543.7 1,930.4 1,757.3 1,765.6 CO 2 1,711.9 1,891.8 1,496.8 1,700.6 1,717.0 1,746.9 1,740.1 CH 4 5.6 2.8 2.3 2.0 2.2 2.1 2.1 N O 2 22.9 35.8 41.2 20.4 18.5 16.6 15.2 a Industrial 1,574.2 1,416.6 1,409.0 1,364.6 1,408.8 1,385.0 1,537.0 CO 2 1,375.7 1,399.6 1,407.0 1,564.6 1,529.2 1,355.0 1,399.3 86 Sulfur dioxide (SO ) emissions from stationary combustion are addressed in Annex 6.3. 2 87 O. The methodology used to calculate these Separate calculations were performed for transportation - related CH and N 4 2 emissions are discussed in the mobile combustion section. 13 - 3 Energy

128 CH 4 2.0 1.6 1.6 1.6 1.6 1.6 1.9 N O 2 7. 5 7.7 7.8 8.0 8.1 8.0 5.9 Residential 1,224.9 940.2 1,127.7 1,018.8 1,077. 5 1,093.3 1,015.7 CO 2 1,064.6 931.4 1,116.2 1,007.8 1,080.1 1,003.9 1,214.1 CH 4 4.2 4.2 3.9 5.4 5.1 5.2 4.0 N O 2 3.4 6.6 7.8 0 8. 7.3 7.9 7.1 a Commercial 759.1 904.5 966.0 943.0 917.9 1,033.7 933.6 CO 2 934.7 755.4 958.4 897.0 1,026.8 909.4 925.5 CH 4 1.2 1.2 1.1 1.2 1.1 1.2 1.3 O N 2 5.7 2.5 6.3 6.4 6.9 7.1 7.2 b U.S. Territories 49.9 27.7 41.0 43.7 42.6 41.5 41.5 Total 1 4,807.6 5,280. 5,075.2 5,813. 5,208.1 5,252.4 5,097. 0 5 a - road and non - In 2016, FHWA changed its methods for estimating the share of gasoline used in on road - series inconsistency between 2015 and previous years in this Inventory. applications, which created a time The method changes resulted in a decrease in th e estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. The method updates are discussed further in the Planned Improvements section below under CO 2 from Fossil Fuel Combustion. b U.S. Territories are not apportioned by sector, and emissions are total greenhouse gas emissions from all fuel combustion sources. Emissions from fossil fuel combustio n by Totals may not sum due to independent rounding. : Note s - electricity generation are allocated based on aggregate national electricity consumption by each end use sector. Stationary Combustion The direct combustion of fuels by stationary sources in the electricity generation, industrial, commercial, and Table 3 - re presents CO sidential sectors represent the greatest share of U.S. greenhouse gas emissions. emissions 9 2 The CO from fossil fuel combustion by stationary sources. emitted is closely linked to the type of fuel being 2 combusted in each sector (see Methodology section of CO from Fossil Fuel Combustion). Other than CO , gases 2 2 Table emitted from stationary c ombustion include the greenhouse gases CH and N 3 O. - 10 and Table 3 - 11 present 2 4 88 CH and N O emissions from the combustion of fuels in stationary sources. Methane and N O emissions from 2 2 4 stationary combustion sources depend upon fuel characteristics, combustion technology, pollution control pment, ambient environmental conditions, and operation and maintenance practices. emissions equi Nitrous oxide fuel mixes and combustion temperatures, as well as the - from stationary combustion are closely related to air equipment that is employed. Methane emissions from stationary combustion characteristics of any pollution control are primarily a function of the CH The CH content of the fuel and combustion efficiency. and N O emission 2 4 4 2008 Inventory to utilize the facility - specific technology estimation methodology was revised for the 1990 through and N O (see Methodology section for CH and fuel use data reported to EPA’s Acid Rain Program (EPA 2016a) 4 2 tationary C ombustion). Table 3 - 7 presents the corresponding direct CO from , CH S , and N O emission s from all 2 4 2 - use sectors . sources of fuel combustion , without allocating emissions from electricity consumption to the end - Table 3 : CO Emissions from Stationary Fossil Fuel Combustion (MMT CO Eq.) 9 2 2 1990 201 2005 201 1 201 2 3 201 4 201 5 Sector/Fuel Type 2,400.9 2,157.7 2,022.2 1,820.8 2,038.1 2,038.0 1,900.7 Electricity Generation 1,547.6 1,983.8 1,722.7 1,511.2 1,571.3 1,569.1 1,350.5 Coal Natural Gas 175.3 318.8 408.8 492.2 444.0 443.2 526.1 Fuel Oil 97.5 97.9 25.8 18.3 22.4 25.3 23.7 Geothermal 0.4 0.4 0.4 0.4 0.4 0.4 0.4 Industrial 806.1 842.5 828.0 775.0 782.9 812.2 805.5 155.3 115.3 82.0 74.1 75.7 75.6 65.9 Coal 434.8 451.9 468.4 467.5 Natural Gas 408.9 388.5 417.3 88 lues for CH , the va 11 - 3 Table and 10 - Table erritories cannot be disaggregated by gas in ce emission estimates for U.S. T Sin 3 4 and N erritory emissions. T O exclude U.S. 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 14 3

129 a 278.3 324.2 275.7 274.1 284.6 262.1 272.2 Fuel Oil Commercial 223.5 220.4 196.7 221.0 228.7 246.2 217.4 9.3 5.8 4.1 3.8 2.9 12.0 Coal 3.9 142.1 162.9 170.5 156.9 Natural Gas 179.1 189.3 175.4 a 63.3 51.3 44.1 Fuel Oil 38.0 35.6 67.9 35.7 338.3 357.8 325.5 282.5 329.7 345.4 319.6 Residential Coal 3.0 0.8 NO NO NO NO NO 262.2 254.7 224.8 238.0 266.2 277.9 252.8 Natural Gas 97.4 94.9 70.9 57.7 63.4 67.5 66.8 Fuel Oil U.S. Territories 27.6 49.7 40.9 43.5 42.5 41.4 41.4 Coal 0.6 3.0 3.4 2.8 4.0 4.0 3.4 Natural Gas NO 1.3 1.4 2.6 3.0 3.0 3.0 Fuel Oil 26.9 45.4 36.0 37.5 36.6 34.3 34.3 3,859.9 3,519.4 3,327.9 3,443.5 3,246.6 3,459.5 3,313.4 Total + Does not exceed 0.05 MMT CO Eq. 2 ( Not O ccurring ) NO a - road and non - road In 2016, FHWA changed its methods for estimating the share of gasoline used in on - the current Inventory. applications, which created a time series inconsistency between 2015 and previous years in The method changes resulted i n a decrease in the estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. The method ology section below under CO updates are discussed further in the Planned Improvements from Fossil Fuel Combustion. 2 Note: Totals may not sum due to independent rounding. Table 3 - 10 : CH Emissions from Stationary Combustion (MMT CO Eq.) 2 4 Sector/Fuel Type 1990 2005 2011 2012 2013 2014 2015 Electric Power 0.4 0.5 0.3 0.4 0.4 0.4 0.4 Coal 0.3 0.3 0.3 0.2 0.2 0.2 0.2 Fuel Oil + + + + + + + Natural gas 0.2 0.2 0.1 0.1 0.2 0.2 0.2 Wood + + + + + + + Industrial 1.8 1.5 1.5 1.5 1.5 1.7 1.5 Coal 0.4 0.2 0.2 0.2 0.2 0.2 0.3 Fuel Oil 0.1 0.1 0.2 0.1 0.1 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 0.9 1.0 0.9 0.9 0.9 Commercial 1.1 1.0 0.9 1.0 1.1 1.2 1.0 Coal + + + + + + + Fuel Oil 0.2 0.2 0.2 0.1 0.1 0.2 0.1 Natural gas 0.4 0.4 0.4 0.4 0.4 0.4 0.3 Wood 0.5 0.5 0.5 0.4 0.5 0.5 0.5 Residential 4.1 5.2 4.0 3.7 5.0 5.0 3.9 Coal NO NO NO NO NO 0.1 0.2 Fuel Oil 0.3 0.3 0.2 0.2 0.2 0.2 0.3 Natural Gas 0.6 0.6 0.5 0.5 0.6 0.6 0.6 Wood 3.1 3.2 3.0 4.1 4.1 3.1 4.1 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 6.6 7.4 8.5 7.1 8.0 8.1 7.0 Eq. + Does not exceed 0.05 MMT CO 2 NO ( Not O ccurring ) Note: Totals may not sum due to independent rounding. 15 - 3 Energy

130 Table - : N 3 O Emissions from Stationary Combustion (MMT CO 11 Eq.) 2 2 1990 201 2005 201 1 201 2 201 3 201 4 5 Sector/Fuel Type Generation Electricity 19.5 17.8 19.1 7.4 19.6 17.6 16.0 Coal 6.3 11.5 10.2 12.1 12.4 11.0 11.6 Fuel Oil 0.1 + + + + 0.1 + Natural Gas 1.0 6.1 7.5 7.0 7.2 8.4 4.3 Wood + + + + + + + Industrial 2.4 3.1 2.4 2.4 2.4 2.9 2.4 Coal 0.5 0.7 0.4 0.4 0.4 0.4 0.3 Fuel Oil 0.3 0.3 0.4 0.3 0.4 0.5 0.5 Natural Gas 0.2 0.2 0.2 0.2 0.3 0.2 0.2 Wood 1.5 1.5 1.5 1.6 1.5 1.6 1.5 Commercial 0.3 0.4 0.3 0.3 0.3 0.4 0.3 Coal + + + + + + 0.1 Fuel Oil 0.1 0.2 0.1 0.1 0.2 0.1 0.1 Natural Gas 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Wood 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Residential 0.9 1.0 0.8 0.7 1.0 1.0 0.8 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 0.5 0.5 0.7 0.7 0.7 0.5 0.5 U.S. Territories 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Coal + + + + + + + Fuel Oil 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Natural Gas + + + + NO + + Wood NO NO NO NO NO NO NO Total 23.4 21.4 22.9 21.3 23.1 11.9 20.2 Eq. + Does not exceed 0.05 MMT CO 2 ( O ccurring ) NO Not Note: Totals may not sum due to independent rounding. Electricity Generation The process of generating electricity is the single largest source of CO emissions in the United States, representing 2 35 emissions from all CO O emissions sources across the United States. Methane and N percent of total CO 2 2 2 accounted for a small portion of emissions from electricity generation, representing less than 0.1 percent and 1.0 Electricity generation also accounted for the largest share of CO percent, respectively. emissions from fossil fuel 2 37.6 percent in 201 5 . ion, approximately Methane and N combust O from electricity generation represented 4.9 and 2 51.0 CH percent of total and N O emissions from fossil fuel combustion in 2015, respectively. 2 4 reased by 4 percent since 1990, the carbon intensity of the While emissions from the electric power sector have inc E electric power sector, in terms of CO q. per QBtu has significantly decreased by 16 percent during that same 2 timeframe. This decarbonization of the electric power sector is a result of several key drivers. Coal - fired electricity generation (in kilowatt - hours [kWh]) decreased from almost 54 percent of generation in 1990 to 34 percent in 2 015. This generation corresponded with an increase in natural gas and renewable energy generation, largely from wind 11 percent of electric power generation in 1990, and and solar energy. Natural gas generation (in kWh) represented increased over the 26 ye ar period to represent 32 percent of electric power generation in 2015. This decoupling of electricity generation and the resulting emissions is shown below in Figure 3 - 10 . Decreases in natural gas costs and the associated increase in natural gas generation, particularly between 2005 and the main driver of the decrease in electric power sector carbon intensity. During this time period, the cost 2015, was of natural gas (in $/MMBtu) decreased by 51 percent while the cost of coal (in $/MMBtu) increased by 91 percent (EIA 2017a) . Between 1990 and 2015, renewable energy generation (in kWh) from solar and wind energy have helped drive the decreases in the carbon also 15, which 0.1 percent in 20 5 percent in 1990 to increased from - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 16 3

131 intensity of the electricity supply in the United States. This decrease in carbon intensity occurred even as total electricity retail sales increased 39 percent, from 2,713 billion kWh in 1990 to 3,75 9 billion kWh in 2015. and Emissions (MMT CO (Billion kWh) Eq.) Figure 3 - 10 : Electricity Generation 2 use sectors for lighting, Electricity was consumed primarily in the residential, commercial, and industrial end - (see Figure 3 - 11 ). heating, electric motors, appliances, electronics, and air conditioning Use Sector - : Electricity Generation Retail Sales by End (Billion kWh) Figure 3 - 11 17 - 3 Energy

132 The electric power industry includes all power producers, consisting of both regul ated utilities and non , - utilities (e.g. generators, and other small power producers). For the underlying - independent power producers, qualifying co energy data used in this chapter, the Energy Information Administration (EIA) places electric power generation into The electric three functional categories: the electric power sector, the commercial sector, and the industrial sector. power sector consists of electric utilities and independent power producers whose primary business is the production of e lectricity, while the other sectors consist of those producers that indicate their primary business is something 89 other than the production of electricity. , were reliant on electricity for Table 3 - 8 The industrial, residential, and commercial end use sectors, as presented in - The residential and commercial end - use sectors were especially reliant on electricity meeting energy needs. ng, heating, air conditioning, and operating appliances. In 2015, e lectricity sales to the consumption for lighti - end and sales to the commercial end use sector decreased by 0.2 percent use sector increased by 0.6 - residential The trend in the residential s ector can largely be attributed to warmer , less percent, respectively. energy - intensive winter conditions while the trend in the commercial sector can largely be attributed to a growing economy compared to 201 Electricity sales to the industrial sector in 201 5 decreased approx imately 0.7 . percent. Overall, in 201 5 , the 4 amount of electricity generated (in kWh) and the amount of electricity consumed (in kWh) decreased approximately 0.4 percent and 0.1 percent , respectively, relative to the previous year, while CO emissions from the electric power 2 in the decreased by 6.7 percent. sector decrease in CO emissions was a result of a significant decrease The 2 d consumption of coal and increase in the consumption of natural gas for electricity generation by 13.9 percent an 5 decrease in the consumption of 6.6 for electricity generation by petroleum 18.7 percent, respectively, in 201 , and a percent. Industrial Sector percent of CO , CH O, , and N , and N O, emissions accounted for 16 , 16 , and 6 Industrial sector CO , CH 2 2 2 4 4 2 ssions from fossil fuel combustion, respectively. from the , CH , and N emi O emissions resulted Carbon dioxide 4 2 direct consumption of fossil fuels for steam and process heat production. end - use The industrial rom EIA, includes activities such as sector, per the underlying energy consumption data f manufacturing, construction, mining, and agriculture. The largest of these activities in terms of energy consumption is manufacturing, of which six industries — Petroleum Refineries, Chemicals, Paper, Primary Metals, Food , and and EIA 2009b). Nonmetallic Mineral Products represent the vast majority of the energy use ( EIA 2017a — In theory, emissions from the industrial sector should be highly correlated with economic growth and industrial output, but heating of industrial b uildings and agricultural energy consumption are also affected by weather 90 conditions. In addition, structural changes within the U.S. economy that lead to shifts in industrial output away from energy intensi ve products (e.g., from steel to computer - - intensive manufacturing products to less energy equipment) also have a significant effect on industrial emissions. From 2014 to 2015, total industrial production and manufacturing output increased by 1.9 percent (FRB 2016). Over this period, output increased across production indices for Food, Petroleum Refineries, Chemicals, and Nonmetallic Mineral Products, and decreased slightly for Primary Metals and Paper (see Figure 3 - 12 ). Through EPA’s Greenhouse Gas Reporting Program (GHGRP), specific industrial sector trends can be discerned from the overall total EIA industrial fuel consumption data used for these calculations. For example, from 2014 to 2015 , the underlying EIA data showed decreased consumption of coal, and relatively flat The use of natural gas in the industrial sector. GHGRP data highlights that several industries contributed to these 91 trends including chemical manufacturing ; pulp, paper and print ; and food processing, beverages and tobacco , . 89 Utilities primarily generate power for the U.S. electric grid for sale to retail customers. Nonutilities produce electricity for le electricity market (e.g., to utilities for distribution and resale their own use, to sell to large consumers, or to sell on the wholesa to customers). 90 Some commercial customers are large enough to obtain an industrial price for natural gas and/or electricity and are or in U.S. energy statistics. These misclassifications of large commercial use sect - consequently grouped with the industrial end - use sector to appear to be more sensitive to weather conditions. customers likely cause the industrial end 91 Further details on industrial sector combustion emissions are provided by EPA’s GHGRP. See . - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 18 3

133 3 12 : Industrial Production Indices (Index 2007=100) - Figure 64 percent) and the overall U.S. economy ( 83 percent) from 1990 to 2015, Despite the growth in industrial output ( percent over the same time CO emissions from fossil fuel combustion in the industrial sector decreased by 4.4 2 A number of factors are believed to have caused this disparity bet ween growth in industrial output and series. decrease in industrial emissions, including: (1) more rapid growth in output from less energy - intensive industries relative to traditional manufacturing industries, and (2) energy - intensive industries such as steel are employing new , and methods, such as electric arc furnaces, that are less carbon intensive than the older methods. In 201 5 , CO , CH 2 4 use sector totaled - 1,364.6 N within the industrial end O emissions from fossil fuel combustion and electricity use 2 201 E q., a 3.2 percent decrease from MMT CO 4 emissions. 2 Residential and Commercial Sectors 5 emissions from fossil fuel percent of CO Residential and commercial sector CO and emissions accounted for 6 2 2 O emissions from fossil fuel combustion, and N 13 percent of CH combustion, CH emissions accounted for 43 and 4 2 4 O emissions from fossil fuel combustion, respectively. percent of N Emissions emissions accounted for 2 and 1 2 from these sectors were largely due to the direct consumption of natural gas and pe troleum products, primarily for 19 - 3 Energy

134 heating and cooking needs. use Coal consumption was a minor component of energy use in both of these end - , and N In , CO 201 , CH 5 within the residential O emissions from fossil fuel combustion and electricity use sectors. 2 4 2 - an were 1,015.7 MMT CO Eq. and 917.9 MMT CO Eq., respectively. Total CO d commercial end , use sectors 2 2 2 O emissions from fossil fuel combustion and electricity use within the residential and commercial end - , and N CH 4 2 decreased by 7.1 and 2.7 perc ent from 201 4 to 201 5 , respectively. use sectors end use sectors have generally been increasing since 1990, and are - Emissions from the residential and commercial - term fluctuations in energy consumption caused by weather conditions, rather t han often correlated with short In the long prevailing economic conditions. term, both sectors are also affected by population growth, regional migration trends, and changes in housing and building attributes (e.g., size and insulation). In 201 5 consumption represent ed 79 and 71 percent of the direct fossil fuel , combustion emissions from natural gas emissions from the residential and commercial sectors , respectively . Natural gas combustion CO CO emissions 2 2 5 from the residential and commercial sectors in 201 creased by 9.0 percent and 7.4 percent from 201 4 levels, de respectively. U.S. Territories Emissions from U.S. Territories are based on the fuel consumption in American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and other U.S. Pacific Islands. As describe d in the Methodology section of CO from 2 F ossil F uel C ombustion, this data is collected separately from the sectoral - level data available for the general ns are not calculations. As sectoral information is not available for U.S. Territories, CO , CH O emissio , and N 2 2 4 presented for U.S. Territories in the tables above, though the emissions will include some transportation and mobile combustion sources. Transportation Sector and Mobile Combustion This discussion of transportation emissions follows the alternat ive method of presenting combustion emissions by Table allocating emissions associated with electricity generation to the transportation end - use sector, as presented in - 3 emissions from all transportation sources (i.e., excluding O N , and , CH 3 CO 8 . Table presents direct - 7 2 4 2 use sector). emissions allocated to electricity consumption in the transportation end - use sector and other mobile combustion accounted for The transportation end MMT CO - Eq. in 201 5 , which 1,757.3 2 34 percent of CO O emissions from fossil emissions, 22 percent of CH percent of N 40 emissions, and represented 4 2 2 92 Fuel purchased in the United States for international aircraft and marine travel fuel combustion, respectively. 111.8 MMT CO ; these emissions are recorded as international bunkers and Eq. in 201 5 accounted for an additional 2 are not included in U.S. totals according to UNFCCC reporting protocols. - Transportation End Use Sector 5 From 1990 to 201 14 percent due, in large part, to , transportation emissions from fossil fuel combustion rose by 93 duty motor vehicles (passenger The number of vehicle miles traveled (VMT) by light - increased demand for travel. 94 - duty trucks) increased 40 percent from 1990 to 201 5 , cars and light as a result of a confluence of factors including population growth, economic growth, urban sprawl , and low fuel prices. periods of 92 Note that these totals include CO , CH and N O emissions from some sources in the U.S. Territories (ships and boats, 4 2 2 and N ational boats, non transportation mobile sources) and CH recre - O emissions from transportation rail electricity. 2 4 93 In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - road applications, which created a time serie s inconsistency between 2015 and previous years in this Inventory. The method changes resulted in a - decrease in the estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercia l and industrial sectors of th is Inventory for 2015. The method updates are discussed further in the Planned Improvements section below under CO from Fossil Fuel Combustion . 2 94 (FHWA 1996 through 2016) FHWA estimates are based on data from FHWA Highway Statistics Table VM VMT 1 . In 2011, - road vehicle - changed its methods for estimating VMT by vehicle class, which led to a shift in VMT and emissions among on - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 20 3

135 From 201 4 5 , CO emissions from the transportation end - use sector de creased by 0.4 percen t , though this to 201 2 downward trend is likely due to a methodology update in 2015 that on - road motor gasoline consumption decreased , 96 95 . These methodological updates, however, did not impact overall estimates relative to the previous methodology gasoline consumpt ion (discussed in the upfront section of this Chapter), which increased from 2014 to 2015 and remains largely influenced by transportation sector trends (e.g., VMT growth) as the majority of motor gasoline in (95 percent). From 2014 to 2015, there were increases in on 2015 was consumed in the transportation sector - road - duty trucks, jet fuel use in general aviation and commercial aircraft, and - and heavy distillate fuel oil use in medium - fuel use in ships and non . recreational boats sions increased slightly between 2014 and 2015, but have decreased 1 5 percent since 2007 Commercial aircraft emis 97 . Decreases in jet fuel emissions (excluding bunkers) since 2007 are due in part to improved (FAA 2017) ng, improvements in aircraft and engine technologies to operational efficiency that results in more direct flight routi reduce fuel burn and emissions, and the accelerated retirement of older, less fuel efficient aircraft. Almost all of the energy consumed for transportation was supplied by petroleum - th more than based products, wi half being related to gasoline consumption in automobiles and other highway vehicles. Other fuel uses, especially diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder. The primary driver of - relat ed emissions was CO Annex 3.2 presents the total emissions from from fossil fuel combustion. transportation 2 , N all transportation and mobile sources, including CO , and HFCs . O, CH 2 2 4 Transportation Fossil Fuel Combustion CO Emissions 2 Domestic transportation CO emissions increased by 16 percent (24 3 . ) between 1990 and 2015, an MMT CO 3 2 2 98 Among domestic transportation sources, annualized increase of 0.6 percent. in 2015, light - duty vehicles (including passenger cars and light - duty trucks) represented 59 percent of CO emissions from fossil fuel combustion, medium - 2 Table and heavy - duty trucks and buses 25 percent, commercial aircraft 7 percent, and other sources 9 percent. See for a detailed breakdown of transportation CO 12 emissions by mode and fuel type. 3 - 2 Almost all of the energy consumed by the transportation sector is petroleum based, including motor gasoline, diesel - and residual oil. Carbon dioxide fuel, jet fuel, emissions from the combustion of ethanol and biodiesel for transportation purposes, along with the emissions associated with the agricultural and industrial processes involved 99 ofuel, are captured in other Inventory sectors. in the production of bi Ethanol consumption from the transportation 13.1 billion gallons in 201 sector has increased from 0.7 billion gallons in 1990 to , while biodiesel consumption 5 has increased from 0.01 billion gallons in 2001 to billion gallons in 201 5 . 1.5 For further information, see S ection 3.10 on biofuel consumption at the end of this chapter and Table A - 9 5 in Annex 3.2. classes in the 2007 to 2015 time period. In absence of these method changes, light - duty VMT growth between 1990 and 2015 would lik ely have been even higher. 95 In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - road applications, which created a time - series inconsistency between 2015 and previous years in this Inventory. The method changes resulted in a l decrease in the estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercia and industrial sectors of this Inventory for 2015. The method updates are discussed further in the Planned Improve ments section CO from Fossil Fuel Combustion . below under 2 96 Note that this value does not include lubricants. 97 Commercial aircraft, as modeled in FAA’s AEDT , consists of passenger aircraft, cargo, and other chartered (FAA 2017) flights. 98 In 2016, FHWA c hanged its methods for estimating the share of gasoline used in on - road and non - road applications, which created a time - The method changes resulted in a series inconsistency between 2015 and previous years in this Inventory. r gasoline consumption for the transportation sector and a subsequent increase in the commercial decrease in the estimated moto The method updates are discussed further in the Planned Improvements section and industrial sectors of this Inventory for 2015. below under CO from Fossil Fu el Combustion . 2 99 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 a re accounted for in the estimates for Land Use, Land - Use Change, and Forestry (see Chapter 6). More information and additional - fuel analyses on biofuels are available at https://www.epa.gov/renewable - EPA's Renewable Fuels Standards website. See < - >. program standard 21 - 3 Energy

136 Carbon dioxide - duty truck s totaled 1,033.7 MMT CO emissions from passenger cars and light in 201 5 , an increase 2 100 of percent ( 83.7 MMT CO - ) from 1990 9 due, in large part, to increased demand for travel as fleet - wide light 2 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 201 5 ). Carbon dioxide emissions from passenger 101 in 2004, and since then have declined about 1 1,180.5 MMT CO duty trucks peaked at 2 percent. cars and light - 2 The decline in new light - d uty 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 vehicl e sales in 1990 to 48 percent in grew only 2004 . Starting in 2005, average new vehicle fuel economy began to increase while light - duty VMT modestly for much of the period. Light - duty VMT grew by less than one percent or declined each year between 102 and has since grown at a faster rate (1.2 percent from 2013 2005 and to 2014, and 2.6 percent from 2014 to 2013 . Average new vehicle fuel economy has improved almost every year since 2005, and the truck share decreased 2015 to about 33 percent in 2009, and has si nce varied from year to year between 36 and 43 percent. Truck share is about 4 3 percent of new vehicles in model year 201 5 (EPA 201 6c ). . Medium and heavy - duty truck CO This increase was largely emissions increased by 7 8 percent from 1990 to 201 5 - 2 - - and heavy due to a substantial growth in medium duty truck VMT, which increased by 95 percent between 1990 103 MMT and 201 5 . Carbon dioxide from the domestic operation of commercial aircraft increased by 8 percent ( 9.1 104 . emissions Across all categories of aviation, excluding international bunkers, CO 5 CO 0 to 201 ) from 199 2 2 105 percent ( percent ( 28.2 MMT CO ) ) between 1990 and 201 5 . 15 This includes a 58 decreased by 20.3 MMT CO 2 2 decrease in CO emissions from domestic military operatio ns. 2 and in Transportation sources also produce CH and N 14 O; these emissions are included in Table 3 - 13 and Table 3 - 2 4 the CH and N O from Mobile Combustion section. Annex 3.2 presents total emissions from all transportation and 4 2 mobile sources, including CO O, and HFCs. , CH , N 4 2 2 100 road applications, which In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - created a time The method changes resulted in a - series inconsistency between 2015 and previous years in this Inventory. dec rease in the estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. The method updates are discussed further in the Planned Improvements section bel ow under CO from Fossil Fuel Combustion . 2 101 See previous footnote . 102 VMT estimates are based on data from FHWA Highway Statistics Table VM - 1 (FHWA 1996 through 2016). In 2007 and duty VMT decreased 3.0 percent and 2.3 percent, respectively. 2008 light - 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 - r estimating VMT by vehicle class, which led to a shift in VMT and emissions among on oad vehicle classes in the 2007 to 2015 time period. In absence of these method changes, light - duty VMT growth between 2006 and 2007 would likely have been higher. 103 While FHWA data shows consistent growth in medium - and heavy - duty truck VMT over the 1990 to 2015 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 la rge changes in VMT by vehicle class, thus leading to a shift in VMT and emissions among on - road vehicle classes in the 2007 to 2015 time period. During the time period prior to the method change (1990 to 2006), VMT for medium - 51 percent. sed by and heavy - duty trucks increa 104 Commercial aircraft, as modeled in FAA’s AEDT, consists of passenger aircraft, cargo, and other chartered flights. 105 n national Includes consumption of jet fuel and aviation gasoline. Does not include aircraft bunkers, which are not included i emission totals, in line with IPCC methodological guidance and UNFCCC reporting obligations. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 22 3

137 - 3 - 13 : Sales - Weighted Fuel Economy of New Passenger Cars and Light Figure Duty Trucks, 1990 – 201 5 (miles/gallon) 6c Source: EPA (201 ) - (Percent) 5 Sales of New Passenger Cars and Figure 3 14 : 201 Light - Duty Trucks, 1990 – : EPA (201 ) Source 6c 23 - 3 Energy

138 Table 3 : CO - Emissions from Fossil Fuel Combustion in Transportation End - Use Sector 12 2 (MMT CO Eq.) 2 a , b a a a a 2011 2012 2013 2014 2005 2015 1990 Fuel/Vehicle Type b 1,183.7 1,068.8 1,064.7 1,065.6 1,0 96.1 1,0 70.5 983.5 Gasoline 621.4 655.9 Passenger Cars 732.8 731.4 731.4 7 49.5 7 31.3 - 309.1 477.2 280.4 277.4 277.7 29 9.9 28 3 .2 Light Duty Trucks c - Duty Trucks 38.7 34.8 38.9 - 38.7 39.5 40. 8 39. 3 Medium and Heavy 0.3 0.4 0.7 0.8 0.8 0.9 0.9 Buses Motorcycles 1.6 3.6 4.1 3.9 3. 9 3.7 1.7 d 13.9 12.4 12.3 12.3 1.1 12. 2 Recreational Boats 12.2 b (Diesel) Distillate Fuel Oil 433.9 457.5 430.0 4 47.7 460.7 262.9 427.5 7.9 4.2 4.1 4.1 4.1 4.1 4. 3 Passenger Cars 12.9 - 11.5 25.8 13.0 12.9 13.9 13.9 Duty Trucks Light c and Heavy - Duty Trucks Medium 190.5 360.2 344.4 344.4 350.0 361. 2 369. 4 - 8.0 10.6 14.4 Buses 15.4 15.5 16.9 17. 3 Rail 45.5 40.4 39.5 40.1 41.6 39.9 35.5 2.0 3.2 3.6 3.7 3.7 3.8 3.9 Recreational Boats e - Boats Non 7.5 8.0 Recreational 7.5 7.5 6.2 12.0 Ships and 10.1 f 11.7 9.4 International Bunker Fuels 7.9 6.8 5.6 6.1 8.4 Jet Fuel 184.2 189.3 146.6 143.4 147.1 148.6 157.7 g 114.6 109.9 132.7 Commercial Aircraft 113.3 114.3 115.2 119.0 19.4 11.6 35.0 12.1 11.0 15.4 14.7 Military Aircraft General Aviation Aircraft 39.4 37.3 20.4 18.0 21.8 18.0 24.0 f International Bunker Fuels 38.0 60.1 64.8 64.5 65.7 69.4 71.8 International Bunker Fuels from Commercial Aviation 30.0 55.6 61.7 61.4 62.8 66.3 68.6 Aviation Gasoline 3.1 2.4 1.9 1.7 1.5 1.5 1.5 General Aviation Aircraft 3.1 2.4 1.9 1.7 1.5 1.5 1.5 15.8 Residual Fuel Oil 19.3 19.4 22.6 15.1 5.8 4.2 e 19.4 22.6 19.3 Ships and Boats 15.8 15.1 5.8 4.2 f 38.9 53.7 43.6 International Bunker Fuels 34.5 28.5 27.7 30.6 j Natural Gas 38.8 36.0 33.1 38.9 40.3 41.3 47.0 Passenger Cars + + + + + + + Duty Trucks + + + - + + + + Light - and Heavy - Duty Trucks + + + + + + + Medium 0.8 0.6 0.8 + 0.8 0.8 0.8 Buses h Pipeline 36.0 32.4 38.1 40.5 46.2 39.4 38.0 j LPG 2.3 1.7 1.4 2.7 2.9 3.0 2.1 Passenger Cars + + + + + + 0.1 0.4 Light Duty Trucks 0.2 0.3 - 0.2 0.3 0.6 0.9 c - and Heavy - Duty Trucks 1.3 1.1 Medium 1.4 1.8 2.1 1.9 1.7 Buses 0.1 0.2 0.3 0.4 0.3 0.3 0.1 3.9 4.7 4.3 Electricity 4.0 4.1 3.7 3.0 3.9 4.7 4.3 Rail 4.0 4.1 3.7 3.0 1,496. k Total 8 1,891.8 1,711.9 1,700.6 1,717.0 1,7 46 . 9 1,7 40.1 1,600. f Total (Including Bunkers) 3 2,004.9 1,823.6 1,806.4 1,816.8 1,8 501.1 1,8 50 . 9 i - Ethanol 71.5 4.1 22.4 71.5 Biofuels 73.4 74.9 75.9 i 13.5 Biodiesel Biofuels 0.0 0.9 8.3 8.5 - 13.3 14.1 + Does not exceed 0.05 MMT CO Eq. 2 a In 2011 FHWA changed its methods for estimating vehicle miles traveled (VMT) and related data. These methodological - elbase. type to one that is based on whe changes included how vehicles are classified, moving from a system based on body 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 road vehicle classes. - n o - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 24 3

139 b Gasoline and diesel highway vehicle fuel consumption estimates are based on data from FHWA Highway Statistics Table 21, 2016, FHWA changed its methods for estimating the share of - 27 , and VM - 1 (FHWA 1996 through 2016) . In MF - MF road and non - road applications, which created a time - series inconsistency between 2015 and previous gasoline used in on - The method changes resulted in a decrease in the es timated motor gasoline consumption for the years in this Inventory. The transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. fr Planned Improvements CO section below under om Fossil Fuel Combustion . method updates are discussed further in the 2 - 1 is used to estimate the share of consumption between each on - road vehicle class . These fuel Data from Table VM consumption estimates are combined with estimates of fuel shares by vehicle type from DOE’s TEDB Annex Tables A.1 th (DOE 1993 through 2016) . TEDB data for 2015 has not been published yet, therefore 2014 data is used as a rough A.6 proxy. c - Includes medium - and heavy duty trucks over 8,500 lbs. d , EPA incorporated the NONROAD2008 model into MOVES2014. The curren t Inventory uses the NONROAD In 2014 component of MOVES2014a for years 1999 through 2015. e Note that large year over year fluctuations in emission estimates partially reflect nature of data collection for these sourc es. f Official estimates exclude emissions fro m the combustion of both aviation and marine international bunker fuels; however, - estimates including international bunker fuel related emissions are presented for informational purposes. g Commercial aircraft, as modeled in FAA’s AEDT, consists of passeng er aircraft, cargo, and other chartered flights. h emissions from natural gas powered pipelines transporting natural gas. Pipelines reflect CO 2 i of this chapter and Ethanol See Section 3.10 estimates are presented for informational purposes only. the and biodiesel estimates in Land Use, Land - Use Change, and Forestry (see Chapter 6) , in line with IPCC methodological guidance and UNF CCC reporting obligations, for more information on ethanol and biodiesel . j Transportation sector natural gas and LPG consumption are based on data from EIA (2016). In previous Inventory years, data from DOE TEDB was used to estimate each vehicle class’s s hare 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 le class’s share of the total natural gas Alternative Fuel Vehicle Data (Browning 201 7 ) is now used to determine each vehic These changes were first incorporated in this year’s Inventory and apply to the 1990 and LPG consumption. 2015 time to period. k Includes emissions from rail electricity. : This table does not include emission s - transportation mobile sources, such as agricultural equipment and Note s from non construction/mining equipment; it also does not include emissions associated with electricity consumption by pipelines or e does not include CO lubricants used in transportation. In addition, this tabl emissions from U.S. Territories, since these are 2 covered in a separate chapter of the Inventory. Totals may not sum due to independent rounding. Mobile Fossil Fuel Combustion CH and N O Emissions 2 4 Mobile combustion includes emissions of CH O from all transportation sources identified in the U.S. and N 2 4 106 mobile sources also include non - Inventory with the exception of pipelines and electric locomotives; transportation sources such as construction/mining equipment, agricultural - road, and equipment, vehicles used off 107 other sources (e.g., snowmobiles, lawnmowers, etc.). Annex 3.2 includes a summary of all emissions from both 106 Emissions of CH from natural gas systems are reported separately. More information on the methodology used to calculate 4 these emissions are include d in this chapter and Annex 3.4. 107 sections of the CO See the methodology sub - O from Mobile Combustion and N from Fossil Fuel Combustion and CH 4 2 2 sections of this chapter. Note that N emissions are O and CH emissions are reported using different categories than CO CO . 2 4 2 2 reported by end - use sector (Transportation, Industrial, Commercial, Residential, U.S . Territories), and generally adhere to a top - transportation sources (e.g., lawn and garden equipment, farm down approach to estimating emissions. CO - emissions from non 2 equipment, construction equipment) are allocated to their respective end - use sector (i.e., construction equipment CO emissions 2 emissions are reported are included in the Industrial end - use sector instead of the Transportation end - use sector). CH O and N 4 2 using the “Mobile Combustion” category, which includes non - transportation mobile sources. CH and N O emission estimates 4 2 type. These VMT) and emissions factors by source and technology are bottom - up estimates, based on total activity (fuel use, reporting schemes are in accordance with IPCC guidance. - emissions from non For informational purposes only, CO 2 transportation mobile sources are presented separately from their overall end use sector in Annex 3.2. - 25 - 3 Energy

140 transportation and mobile sources. 3 and Table 3 - 14 provide mobile fossil fuel CH - and N Table O emission 13 4 2 108 Eq. estimates in MMT CO 2 emissions ( 0.3 percent) but was the fourth portion of national CH Mobile combustion was responsible for a small 4 109 4.5 percent). From 1990 to 201 5, largest source of U.S. N mobile source CH O emissions ( emissions declined by 4 2 - percent, to 2.0 MMT CO 6 Eq. ( 80 kt CH road vehicles ), due largely to control technologi es employed in on 4 4 2 Mobile source 1990s to reduce CO, NO , NMVOC, and CH - emissions. since the mid O decreased emissions of N 2 x 4 110 percent, to 15.1 MMT CO Eq. ( 51 kt N O). by 63 Earlier generation control technologies initially resulted in 2 2 higher N O emissions, causing a 28 percent increase in N O emissions from mobile sources between 1990 and 1997. 2 2 - generation emission control technologies have reduced N O output, resulting in a 71 percent Improvements in later 2 O emissions were O emissio ns from 1997 to 2015 ( Figure 3 - 15 decrease in mobile source N . Overall, CH and N ) 4 2 2 - predominantly from gasoline - duty trucks. fueled passenger cars and light Figure 3 - 15 : Mobile Source CH and N Eq.) O Emissions (MMT CO 2 4 2 Table 3 - 13 : CH Emissions from Mobile Combustion (MMT CO Eq.) 2 4 a Fuel Type/Vehicle Type 1990 2005 2011 2012 2013 2014 2015 b 1.4 1.4 Gasoline On - Road 1.5 5.2 2.2 1.7 1.6 Passenger Cars 1.2 1.2 1.1 1.1 1.0 1.0 3.2 - 1.7 0.9 0.4 Light 0.4 0.3 0.3 0.3 Duty Trucks Medium - and Heavy - Duty 0.1 Trucks and Buses 0.1 0.1 0.3 0.1 0.1 0.1 Motorcycles + + + + + + + b + - Road + Diesel On + + + + + + Passenger Cars + + + + + + Light - Duty Trucks + + + + + + + 108 T See Annex 3.2 for a co mplete time series of emission estimates for 1990 through 2015. 109 In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - road applications. These method changes created a time series inconsistency in this Inventory between 2015 and previous years in CH and N - O estimates 2 4 for agricultural, construction, commercial, and industrial non - road mobile sources. The method updates are discussed further in . and N O from Mobil e Combustion the Planned Improvements section of below under CH 4 2 110 See above. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 26 3

141 Medium - - Duty and Heavy + Trucks and Buses + + + + + + + + + + Road + + Alternative Fuel On - + c 0.6 0.4 0.5 0.5 0.6 Road 0.6 0. 5 - Non + + + Ships and Boats + + + + d 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Rail 0.1 0.1 + + Aircraft + + + e 0.2 0.1 0.2 0.2 0.2 0.2 0.2 Agricultural Equipment Construction/Mining f 0.1 0.1 0.1 0.1 Equipment 0.1 0.1 0.1 g 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Other 5.6 2.8 2.3 2.2 2.1 2.1 2.0 Total Eq. + Does not exceed 0.05 MMT CO 2 a See Annex 3.2 for definitions of on - road vehicle types. b vehicle mileage estimates are based on data from FHWA Highway Statistics Table Gasoline and diesel highway 1 (FHWA 1996 through 2016). These mileage estimates are combined with estimates of fuel shares by - VM 2016). TEDB data for 2015 vehicle type from DOE’s TEDB Annex Tables A.1 through A.6 (DOE 1993 through has not been published yet, therefore 2014 data is used as a proxy. c In 2016, FHWA changed its methods for estimating the share of gasoline used in on road and non - road - These method changes created a time - applications. nconsistency in this Inventory between 2015 and series i previous years in CH and N O estimates for agricultural, construction, commercial, and industrial non - road 4 2 The method updates are discussed further in the mobile sources. Planned Improvements section below under CH 4 and N . O from Mobile Combustion 2 d Rail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption data for 2014 and 2015 are not available yet, therefore 2013 data is used as a proxy. e - Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off road in agriculture. f Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are - road in construction. used off g “Other” includes snowmobiles and other recreational equipment, logging equipment, lawn and garden equipment, railroad equipment, airport equipment, commercial equipment, and industrial equipment, as well as fuel consumption from trucks - road for commercial/industrial purposes. that are used off Note s : In 2011, FHWA changed its methods for estimating vehicle miles traveled (VMT) and related data. These - type to one methodological changes included how vehicles are classified, moving from a system based on body 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. tals may not sum due to To independent rounding. Table 3 - 14 : N Eq.) O Emissions from Mobile Combustion (MMT CO 2 2 a Fuel Type/Vehicle Type 1990 2005 2011 2012 2013 2014 2015 b - 37.5 31.3 18.4 16.1 14.1 12.3 10.8 Gasoline On Road 15.7 12.1 10.5 9.2 7.8 6.9 Passenger Cars 24.1 - Duty Trucks Light 14.7 5.6 4.9 4.3 4.0 3.4 12.8 - - and Heavy Medium Duty Trucks and Buses 0.9 0.7 0.7 0.6 0.5 0.5 0.5 + + + + Motorcycles + + + b Road Diesel On - 0.4 0.3 0.4 0.2 0.4 0.4 0.4 Passenger Cars + + + + + + + Light - Duty Trucks + + + + + + + - and Heavy - Duty Medium Trucks and Buses 0.3 0.4 0.2 0.4 0.4 0.4 0.4 Alternative Fuel On Road + + 0.1 - 0.1 0.1 0.1 0.1 c Road 4.1 Non 3.5 - 4.0 3.9 4.0 3.8 3.9 Ships and Boats 0.6 0.8 0.6 0.7 0.5 0.6 0.7 d 0.3 0.3 0.3 0.3 Rail 0.3 0. 4 0.3 Aircraft 1.7 1.8 1.4 1.3 1.4 1.4 1.5 e Agricultural Equipment 0.4 0.4 0.4 0.2 0.4 0.4 0.4 27 - 3 Energy

142 Construction/Mining f 0.6 0.3 0.5 0.6 Equipment 0.6 0.6 0.6 g Other 0.4 0.6 0.6 0.6 0.6 0.6 0.6 Total 41.2 35.7 22.8 20.4 18.5 16.6 15. 1 Eq. + Does not exceed 0.05 MMT CO 2 a - road vehicle types. See Annex 3.2 for definitions of on b Gasoline and diesel highway vehicle mileage estimates are based on data from FHWA Highway Statistics Table VM - (FHWA 1996 through 2016) . These mileage estimates are combined with estimates of fuel shares by vehicle type 1 from DOE’s TEDB Annex Tables A.1 through A.6 (DOE 1993 through 2016) . TEDB data for 2015 has not been published yet, therefore 2014 data is used as a proxy. c In 2016, FHWA changed its m ethods for estimating the share of gasoline used in on - road and non - road applications. These method changes created a time - series inconsistency in this Inventory between 2015 and previous years in CH - road mobile and N ercial, and industrial non O estimates for agricultural, construction, comm 2 4 O and N sources. The method updates are discussed further in the Planned Improvements section below under CH 2 4 . from Mobile Combustion d Rail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption 2015 are 2014 and not available yet, therefore 2013 data is used as a proxy. data for e Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off - road in agriculture. f Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are used - road in construction. off g “Other ” includes snowmobiles and other recreational equipment, logging equipment, lawn and garden equipment, railroad equipment, airport equipment, commercial equipment, and industrial equipment, as well as fuel consumption - road for commercial/industrial purposes. from trucks that are used off 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 201 time period. This resulted in large changes in VMT and fuel consumption data by vehicle class, thus 5 leading to a shift in emiss ions among on - road vehicle classes. Totals may not sum due to independent rounding. CO from Fossil Fuel Combustion 2 Methodology The methodology used by the United States for estimating CO emissions from fossil fuel combustion is 2 conceptually similar to the approach recommended by the IPCC for countries that intend to develop detailed, sectoral - based emission estimates in line with a Tier 2 method in the 2006 IPCC Guidelines for National 111 Greenhouse Gas Inventories The use of the most recently published calculation methodologies by (IPCC 2006). the IPCC, as contained in the 2006 IPCC Guidelines , is considered to improve the rigor and accuracy of this Inventory and is fully in line with IPCC Good Practice Guidance. A detailed desc ription of the U.S. methodology is presented in Annex 2.1, and is characterized by the following steps: Determine total fuel consumption by fuel type and sector 1. . Total fossil fuel consumption for each year is estimated by aggregating consumption data by en - use sector (e.g., commercial, industrial, etc.), primary d 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 EI A 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 112 nal Energy Statistics ( EIA 201 7b ) and Jacobs (2010). separately from EIA’s Internatio 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 111 The IPCC Tier 3B methodology is used for estimating emissions from commercial aircraft. 112 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 41.5 MMT CO Eq. in 2015. 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 28 3

143 pres ented "top down" — that is, energy consumption for fuel types and categories are estimated from energy The resulting quantities are production data (accounting for imports, exports, stock changes, and losses). The data collected in the United States by EIA on an annual basis referred to as "apparent consumption." are predominantly from mid - stream or conversion energy consumers such as and used in this Inventory These annual surveys are supplemented with end use energy refiners and electric power generators. - onsumption surveys, such as the Manufacturing Energy Consumption Survey, that are conducted on a c These consumption data sets help inform the annual surveys to arrive at periodic basis (every four years). 113 the national total and sectoral breakdowns for that total. 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 114 (NCV) (i.e., lower heating values). Subtract uses accounted for in the Industrial Processes and Product Use chapter . Portions of the fuel 2. egories — coking coal, distillate fuel, industrial other coal, petroleum consumption data for seven fuel cat 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 acti vity. To make these adjustments, additional data were collected from AISI (2004 through 201 6 2 ), U.S. ), Coffeyville (201 2001 through USAA (2008 through 2016), EIA ( 2017a, 2016a, 2016b ) , Census Bureau ( USGS 2011), 5a ), (USGS 2016a), USGS (2014 through 2016a), USGS (2014 through 2016b), USGS (1991 through 201 (199 5 through 201 3 ), USGS (1995, 1998, 2000 , 200 1 ), USGS (20 16b ), USGS (200 16c ), USGS (201 5a) , USGS (1991 through 201 3 ), USGS (2016d), USGS (2015b), USGS (2014), USGS (1996 through 2013), 115 USGS (199 1 through 201 5 b), USGS (2015 and 2016), USGS (1991 through 2015c) . Fossil fuel consumption estimates are adjusted . Adjust for conversion of fuels and exports of CO 3. 2 116 . downward to exclude fuels created from other fossil fuels and exports of CO natural gas is Synthetic 2 created from industrial coal, and is currently included in EIA statistics for both coal and natural gas. 117 Since October 2000, Therefore, synthetic natural gas is subtracted from energy consumption statistics. Since this CO the Dakota Gasification Plant ha is not emitted to Canada by pipeline. s been exporting CO 2 2 to the atmosphere in the United States, the associated fossil fuel burned to create the exported CO is 2 is the total fossil fuel burned at subtracted from fossil fuel consumption statistics. The associated fossil fuel generated CO that capture system multiplied by the fraction of the plant’s total site - the plant with the CO 2 2 is recovered by the capture system. To make these adjustments, additional data for ethanol and biodiesel and data were collected from EIA ( 2017a ), data for synthetic natural gas were collected from EIA ( 2016a ), exports were collected from the Eastman Gasification Services Company (2011), Dakota for CO 2 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 113 Emissions from Fossil Fuel Combustion in Annex 4 for a comparison of See IPCC Reference A pproach for Estimating CO 2 U.S. estimates using top - down and bottom - up approaches. 114 A crude convention to convert between gross and net calorific values is to multiply the heat content o f 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. 115 See sections on Iron and Steel Pr oduction 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 Process es and Product Use chapter. 116 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. 117 Annex 2.1. These adjustments are explained in greater detail in 29 - 3 Energy

144 obtained from the bottom - up analysis. As the total distillate and motor gasoline consu mption 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 201 6 ), Benson - ), EIA (2007), EIA (1991 through 201 (2002 through 2004), DOE (1993 through 201 ), EPA ( 2016d ), and 6 6 118 6 ). FHWA (1996 through 201 Adjust for fuels consumed for non - energy use s . U.S. aggregate energy statistics include consumption of 5. - These are fossil fuels that are manufactured into plastics, asphalt, energy purposes. fossil fuels for non Depending on the end - use, this can result in storage of som e or all of the C lubricants, or other products. As the emission pathways of C used for non - energy purposes are contained in the fuel for a period of time. vastly different than fuel combustion (since the C in these fuels ends up in products instead of being 3.2 are estimated separately in Section – combusted), these emissions Carbon Emitted and Stored in Products from Non - Therefore, the amount of fuels used for non - e nergy Energy Uses of Fossil Fuels. EIA - fuel consumption was provided by purposes was subtracted from total fuel consumption. Data on non 2017a ) ( . 6. Subtract consumption of international bunker fuels. According to the UNFCCC reporting guidelines emissions from international transport act ivities, 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 - use sector, how ever, so emissions from jet fuel) as part of consumption by the transportation end 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 119 the Under Secretary of Defense (Installations and Environment) and the Defense The Office of content). ) Logistics Agency Energy (DLA Energy) of the U.S. Department of Defense (DoD) ( DLA Energy 201 6 supplied data on military jet fuel and marine fuel use. Commercial jet fuel use wa s obtained from FAA DOC (1991 through (201 7 ); residual and distillate fuel use for civilian marine bunkers was obtained from 2 for 1990 through 2001 and 2007 through 201 ) 6 , and DHS (2008) for 2003 through 2006. 4 01 Consumption of these fuels was subtracted use - from the corresponding fuels in the transportation end sector. Estimates of international bunker fuel emissions for the United State s are discussed in detail in 3.9 – International Bunker Fuels. Section Total C was estimated by multiplying the amount of fuel Determine the total C content of fuels consumed. 7. This total C estimate defines the maximum amount of C that consumed by the amount of C in each fuel. otentially be released to the atmosphere if all of the C in each fuel was converted to CO could p . The C 2 content coefficients used by the United States were obtained from EIA’s Emissions of Greenhouse Gases in (EIA 2009a), and an EPA analys is of C content coefficients used in the GHGRP the United States 2008 A discussion of the methodology used to develop the C content coefficients are presented in (EPA 2010). Annexes 2.1 and 2.2. emissions are the product of the adjusted energy consumption (from Estimate CO Emissions. Total CO 8. 2 2 the previous methodology steps 1 through 6), the C content of the fuels consumed, and the fraction of C that is oxidized. coal, and natural gas The fraction oxidized was assumed to be 100 percent for petroleum, based on guidance in IPCC (2006) (see Annex 2.1). Allocate transportation emissions by vehicle type. This report provides a more detailed accounting of 9. 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 118 Bottom - up g asoline and diesel highway vehicle fuel consumption estimates are based on data from FHWA Highway MF 21, - Statistics Table MF - 27 , and VM - 1 (FHWA 1996 through 2016) . In 2016, FHWA changed its methods for estimating the - share of gasol - road and non - road applications, which created a time ine used in on series inconsistency between 2015 and previous years in this Inventory. The method changes resulted in a decrease in the estimated motor gasoline consumption for the The method transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for 2015. section below. Planned Improvements updates are discussed further in the 119 led discussion. See International Bunker Fuels section in this chapter for a more detai - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 30 3

145 allocate emissions by fuel type calculated for the transportation end - Heat contents and densities use sector. 120 EIA ( ) and USAF (1998). d from 2017a were obtaine For on - road vehicles, annual estimates of combined motor gasoline and diesel fuel consumption by • ); for each vehicle category, the 6 vehicle category were obtained from FHWA (1996 through 201 percent gasoline, diesel, and other (e.g., CNG, LPG) fuel consumption are estimated using data from 121 , 122 ). DOE (1993 through 201 6 - road vehicles, activity data were obtained from AAR (2008 through 201 6 ), APTA (2007 For non • 6 ), APTA (2006), BEA ( 2016 ), Benson (2002 through 2 004), DOE (1993 through 201 6 ), through 201 ), DOC (1991 through 201 6 ), 6 DLA Energy (201 DOT (1991 through 201 6 ), EIA (2009a), EIA 123 6d , EIA (20 13 ), EIA (1991 through 201 6 ), EPA (201 ) ) , ( and Gaffney (2007). 2017a • For jet fuel used by aircraft, CO emissions from commercial aircraft were developed by the U.S. 2 Federal Aviation Administration (FAA) using a Tier 3B methodology, consistent (see IPCC (2006) Annex 3.3). emissions from other aircraft were calculated directly based on reporte d Carbon dioxide 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 fue l 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 Combus tion - 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. level combustion emissions through EPA’s - The availability of facility Greenhouse Gas Repo rting Program (GHGRP) has 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. HGRP 2010 through 2015 reporting years, facility - level fossil fuel combustion emissions reported through For G GHGRP were categorized and distributed to specific industry types by utilizing facility - reported NAICS EPA’s codes (as published by the U.S. Census Bur eau). 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 120 For a more detailed description of the data sources used for the analysis of the transportation end use sector see the Mobile Combustion (excluding CO .8, ) and International Bunker Fuels sections of the Energy chapter, Annex 3.2, and Annex 3 2 respectively. 121 road vehicle class. Since Data from FHWA’s Table VM - 1 is used to estimate the share of fuel consumption between each on - - . t Inventory the curren VM are used as a proxy for 1 data for 2015 has not been published yet, fuel consumption shares from 2014 These fuel consumption estimates are combined with estimates of fuel shares by vehicle type from DOE’s TEDB Annex Tables (DOE 1993 through 2016) . TEDB data for 2015 has not been published yet, therefore 2014 data is used as a A.1 through A.6 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 incorpor ated for the 1990 through 2010 Inventory and apply to the 2007 through 2015 time period. This resulted in large changes road vehicle classes. in VMT and fuel consumption data by vehicle class, thus leading to a shift in emissions among on - 122 Inventory years, Transportation sector natural gas and LPG consumption are based on data from EIA (2016). In previous 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 incl ude estimates for natural gas use by medium and heavy duty trucks or LPG use by passenger cars, EIA Alternative Fuel Vehicle Data (Browning 201 7 ) is now used to determine each vehicle class’s share of the total natural gas and Inventory and apply to the 1990 to 2015 time period. LPG consumption. These chang es were first incorporated in the current 123 , EPA incorporated the NONROAD2008 model into MOVES2014. The current Inventory uses the NONROAD In 2014 . component of MOVES2014a for years 1999 through 2015 31 - 3 Energy

146 statistics to meet the UNFCCC reporting guidelines. The IPCC has provide d guidance on aligning facility level - 124 reported fuels and fuel types published in national energy statistics, which guided this exercise. effort represents an attempt to align, reconcile, and coordinate the As with previous Inventory reports, the current 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 125 UNFCCC along with this report. T he efforts in reconciling fuels focus on standard, common fuel types (e.g., natural gas, distillate fuel oil, etc.) where the fuels in E IA’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 I nventory repor ts 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. A nalyses were conducted linking GHGRP facility - level reporting with the information published by EIA in its MECS data in order to through 201 disaggregate the full 1990 time series in the CRF tables. It is believed that the current analysis has led 5 to impr ovements 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 Fossil fuels are the dominant source of energy in the United States, and CO is the dominant greenhouse gas emitted 2 - emissions are impacted by not only lower levels of energy as a product from their combustion. Energy related CO 2 consumption but also by lowering the C intensity of the energy sources employed (e.g., fuel switching from coal to natural gas). dent upon the C content of the The amount of C emitted from the combustion of fossil fuels is depen fuel and the fraction of that C that is oxidized. Fossil fuels vary in their average C content, ranging from about 53 126 Eq./QBtu for natural gas to upwards of 95 MMT CO Eq./QBtu for coal and petroleum coke. MMT CO In 2 2 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 sour ces employed to meet demand. Table - 15 provides a time series of the C intensity for each sector of the U.S. economy. The time series 3 incorporates only the energy consumed from the direct combustion of fossil fuels in each sector. For the purposes of following reporting guidelines and maintaining the focus of this section, renewable energy and nuclear electricity and consumption are not included in the totals shown in Table 3 - 15 in order to focus attention on fossil fuel For example, the C intensity for the residential sector does not include the combustion as detailed in this chapter. energy from or emissions related to the con sumption of electricity for lighting. Looking only at this direct consumption of fossil fuels, the residential sector exhibited the lowest C intensity, which is related to the large The C inten sity of the commercial sector has percentage of its energy derived from natural gas for heating. 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 thu s 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 CO Eq./EJ), which were 2 the primary sources of energ y. Lastly, the electricity generation sector had the highest C intensity due to its heavy reliance on . coal for generating electricity 124 S ee 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 < http://www.ipcc - c/tb/TFI_Technical_Bulletin_1.pdf nggip.iges.or.jp/publi >. 125 See . 18 126 One exajoule (EJ) is equal to 10 joules or 0.9478 QBtu. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 32 3

147 Table 3 15 : Carbon Intensity from Direct Fossil Fuel Combustion by Secto r (MMT CO - 2 Eq./QBtu) Sector 1990 2005 2011 2012 2013 2014 2015 a 57.4 56.6 55.7 Residential 55.5 55.3 55.4 55.6 a Commercial 56.5 59.1 57.5 56.1 55.8 55.5 57.3 a 64.3 64.3 62.4 62.0 61.8 61.4 61.2 Industrial a Transportation 71.5 71.1 71.4 71.5 71.4 71.5 71.5 b 82.9 87.3 85.8 Electricity Generation 79.9 81.3 81.2 78.1 c 73. 72.3 U.S. Territories 73.0 72.3 5 72.9 72.2 71.9 c 70.8 73.0 73.5 72.0 70.9 70.9 69.7 All Sectors a Does not include electricity or renewable energy consumption. b Does not include electricity produced using nuclear or renewable energy. c Does not include nuclear or renewable energy consumption. Note: Excludes non - energy fuel use emissions and consumpti on. fairly constant, as For the time period of 1990 through about 20 08 , the C intensity of U.S. energy consumption was over that time Starting in . the proportion of fossil fuels used by the individual sectors did not change significantly 2008 the C intensity has decreased, reflecting the shift from coal to natural gas in the electricity sector during that Per capita energy consumption fluctuated little from 1990 to 2007, but in 2015 was approximately 10.2 time period. - below levels in 1990 (see Figure 3 16 ). percent To differentiate these estimates from those of Table 3 - 15 , the C and described below includes nuclear and renewable energy EIA data to intensity trend shown in Figure 3 - 16 economy - provide a comprehensive Due to a general shift from a wide picture of energy consumption. efficiency, energy based economy to a service - based economy, as well as overall increases in - manufacturing consumption and energy - related CO emissions per dollar of gross domestic product (GDP) have both declined since 2 1990 ( BEA 2017 ) . Emissions Per Capita and Per Figure 3 - 16 : U.S. Energy Consumption and Energy - Related CO 2 Dollar GDP ), EPA (2010), and 2017a C intensity estimates were developed using nuclear and renewable energy data from EIA ( consumption data as discussed above and presented in Annex 2.1. fossil fuel 33 - 3 Energy

148 Uncertainty and Time - Series Consistency emitted is directly related to the amount of For estimates of CO from fossil fuel combustion, the amount of CO 2 2 fuel consumed, the fraction of the fuel that is oxidized, and the carbon content of the fuel. Therefore, a careful accounti ng of fossil fuel consumption by fuel type, average carbon contents of fossil fuels consumed, and - based products with long - term carbon storage should yield an accurate estimate of CO production of fossil fuel 2 emissions. es in the consumption data, carbon content of fuels and products, and carbon Nevertheless, there are uncertainti 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 CO See, for example, emission estimates is believed to be relatively small. 2 Marland and Pippin (1990). Although statistics of total fossil fuel and other energy consum ption are relatively accurate, the allocation of this consumption to individual end use sectors (i.e., residential, commercial, industrial, and transportation) is less - certain. For example, for some fuels the sectoral allocations are based on price rates ( i.e., tariffs), but a commercial establishment may be able to negotiate an industrial rate or a small industrial establishment may end up paying an Also, the deregulation of the natural gas industry and the industrial rate, leading to a misallocation of emissions. 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 CO emission esti mate from energy - related fossil fuel combustion, the amount of fuel used in 2 non energy production processes were subtracted from the total fossil fuel consumption. The amount of CO - 2 emissions resulting from non - energy related fossil fuel use has been calcu lated separately and reported in the Carbon These factors all contribute to the Emitted from Non - Energy Uses of Fossil Fuels section of this report (Section 3.2 ) . uncertainty in the CO estimates. - Detailed discussions on the uncertainties associated with C emitted from Non 2 Energy Uses of Fossil Fuels can be found within that section of this chapter. Various sources of uncertainty surround the estim ation of emissions from international bunker fuels, which are Section 3.9 – subtracted from the U.S. totals (see the detailed discussions on these uncertainties provided in Another source of uncertainty is fuel consumption by U.S. T International Bunker Fuels). The United erritories. States does not collect energy statistics for its territories at the same level of detail as for the fifty states and th e 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 CO emissions 2 up from the transport ation end - use sector to individual vehicle types and transport modes. In many cases, bottom - - type estimates from EIA. estimates of fuel consumption by vehicle type do not match aggregate fuel Further research is planned to improve the allocation into det use sector emissions. - ailed transportation end - recommended - The uncertainty analysis was performed by primary fuel type for each end use sector, using the IPCC Approach 2 uncertainty estimation methodology, Monte Carlo Stochastic Simulation technique, with @RISK software. For this uncertainty estimation, the inventory estimation model for CO from fossil fuel combustion was 2 integrated with the relevant variables from the inventory estimation model for International Bunker Fuels, to e the interaction (or endogenous correlation) between the variables of these two models. realistically characteriz About 120 input variables were modeled for CO related Fossil Fuel Combustion (including about 10 from energy - 2 for non - energy fuel consumption and about 20 for Internat ional Bunker Fuels). - In developing the uncertainty estimation model, uniform distributions were assumed for all activity related input 127 Triangular distributions were assigned for variables and emission factors, based on the SAIC/EIA (2001) report. 127 e input variables as a combination of SAIC/EIA (2001) characterizes the underlying probability density function for th ). uniform and normal distributions (the former to represent the bias component and the latter to represent the random component distribution was more appropriate to However, for purposes of the current uncertainty analysis, it was determined that uniform characterize the probability density function underlying each of these variables. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 34 3

149 the oxi dization factors (or combustion efficiencies). The uncertainty ranges were assigned to the input variables 128 based on the data reported in SAIC/EIA (2001) and on conversations with various agency personnel. nput variables were typically asymmetric around their inventory The uncertainty ranges for the activity - related i Bias (or systematic uncertainties) estimates; the uncertainty ranges for the emissions factors were symmetric. d with these variables (SAIC/EIA associated with these variables accounted for much of the uncertainties associate 129 For purposes of this uncertainty analysis, each input variable was simulated 10,000 times through Monte 2001). Carlo sampling. 3 - 16 The results of the Approach 2 quantitative uncertainty analysis are summarized in Fossil fuel Table . 5, emissions in 2015 were estimated to be between 4,944.0 and 286.9 MMT CO combustion CO Eq. at a 95 percent 2 2 This indicates a range of 2 percent below to 5 per cent above the 2015 emission estimate of 5, 049.8 confidence level. MMT CO Eq. 2 16 - Emissions from Energy - 3 Table : Approach 2 Quantitative Uncertainty Estimates for CO 2 (MMT CO R Eq. and Percent) elated Fossil Fuel Combustion by Fuel Type and Sector 2 a 2015 Emission Estimate Uncertainty Range Relative to Emission Estimate (%) (MMT CO Fuel/Sector Eq.) (MMT CO Eq.) 2 2 Lower Upper Upper Lower Bound Bound Bound Bound b 1,423.3 1,374.3 1,555.7 - 3% 9% Coal NE NE NE NE NE Residential 2.9 2.8 3.4 Commercial - 5% 15% 62.7 Industrial 65.9 - 5% 16% 76.3 Transportation NE NE NE NE NE 1,350.5 1,298.2 1,479.1 Electricity Generation - 4% 10% U.S. Territories 4.0 3.5 4.8 - 12% 19% b 1,463.6 1,447.1 Natural Gas 1,531.1 - 1% 5% 7% Residential 252.8 245.7 270.5 3% - Commercial 175.4 170.5 187.7 - 3% 7% - 453.3 500.8 467.5 3% 7% Industrial 38.8 37.7 41.5 Transportation - 3% 7% 510.9 Electricity Generation 526.1 - 3% 5% 553.0 U.S. Territories 2.6 3.5 - 12% 17% 3.0 b 2,162.5 2,032.6 Petroleum - 6% 6% 2,289.5 Residential 63.1 70.4 66.8 - 6% 5% 71.5 Commercial 67.9 64.0 - 6% 5% 321.0 Industrial 272.2 219.4 18% - 19% Transportation 1,697.6 1,587.0 1,808.2 - 7% 7% Electric Utilities 23.7 22.5 25.6 - 5% 8% 34.3 31.8 38.1 U.S. Territories - 7% 11% b 5% 5,049.4 4,943.6 5,286.5 - 2% Total (excluding Geothermal) Geothermal NE NE 0.4 NE NE b,c Total (including Geothermal) 5,049.8 4,944.0 5,286.9 - 2% 5% NE (Not Estimated) a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. 128 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, f or purposes of assigning uncertainty estimates to the sub - fuel type categories within each end - use sector in the current reported uncertainty estimates were extrapolated. - uncertainty analysis, SAIC/EIA (2001) 129 Although, in general, random uncertainties are the main focus of statistical uncertainty analysis, when the uncertainty s of estimates are elicited from experts, their estimates include both random and systematic uncertainties. Hence, both these type tainties are represented in this uncertainty analysis. uncer 35 - 3 Energy

150 b The low and high estimates for total emissions were calculated separately through simulations and, hence, the low and - high emission estimate s 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 CO emissions 2 from geothermal production. to the entire time series to ensure time - series consistency from 1990 Methodological recalculations were applied through 2015 with one recent notable exception related to estimating the share of motor gasoline used in various sectors, which is discussed in the Planned Improvements section below. De tails on the emission trends through time are described in more detail in the Methodology section, above. QA/QC and Verification - specific QA/QC plan for CO A source from fossil fuel combustion was developed and implemented. This effort 2 included a general ( Tier 1 ) analysis, as well as portions of a category - specific ( Tier 2 ) analysis. The Tier 2 procedures that were implemented involved checks specifically focusing on the activity data and methodology used emissions from fossil fuel comb ustion in the United States. Emission totals for the different for estimating CO 2 sectors and fuels were compared and 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 use sectors, 2011 through relative to the previous Inventory. EIA revised 201 4 natural gas consumption in all end - 201 consumption in natural gas 2014 Liquefied Petroleum Gas (LPG) consumption in all end - use sectors, 4 and coal the , 2014 coal consumption in the commercial sector, and 2013 distillate fuel consumption in sector electric power the industrial and transportation sectors . In 2016, EIA revised 2014 heat contents for coal, coal coke, and natural gas. The Federal Highway Administration (FHWA) updated 2014 motor gasoline consumption in the transportation mption and emissions sector, which also resulted in revisions to the commercial and industrial sector gasoline consu emissions from fossil fuel combustion for the transportation sector for 2014. This change resulted in 2014 CO 2 increasing by 0.3 percent and decreasing by roughly 1 percent for the industrial and commercial sectors. data for U.S. Territories was revised throughout the time series for petroleum, and in 2013 and Energy consumption 2014 for coal and natural gas (EIA 2017b). Eq. (less than Overall, these changes resulted in an average annual decrease of MMT CO percent) in CO 0.4 0.1 2 2 emis sions from fossil fuel combustion for the period 1990 through 201 4 , relative to the previous Inventory . Planned Improvements To reduce uncertainty of CO from fossil fuel combustion estimates for U.S. Territories , efforts will continue to 2 work with EIA and other agencies to improve the quality of the U.S. T erritories data. This improvement is not all - inclusive, and is part of an ongoing analysis and efforts to continually improve the CO from fossil fuel combustion 2 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 chara cterize 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 estim ate total, national U.S. emissions. In addition, and unlike the reporting requirements for this chapter under the UNFCCC reporting level fuel combustion emissions reported under the GHGRP may also include industrial - guidelines, some facility 130 In line with UNFCCC reporting guidelines, fuel combustion emissions are included in this ons. process emissi 130 http://unfccc.int/resource/docs/2006/sbsta/eng/09.pdf>. See < - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 36 3

151 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 b e useful to improve the emission estimates for the CO from fossil 2 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 consu mption data from EIA’s Manufacturing Energy Consumption Survey (MECS) , with updated data for 2014. Additional work will look at CO level reported data, and maintaining emissions from biomass to ensure they are separated in the facility - 2 tional energy statistics provided by EIA. In implementing improvements and integration of data consistency with na from EPA’s GHGRP, the latest guidance from the IPCC on the use of facility - level data in national inventories will 131 continue to be relied upon. An ongoing planne d improvement is to develop improved estimates of domestic waterborne fuel consumption. The I nventory 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. Domes tic 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 fe asibility of using domestic marine activity data to improve the estimates will continue to be investigated. EPA received a comment from FHWA that the trend of decreasing electricity use in the transportation sector does not match up with increased sales - in hybrid vehicles. Electricity data is allocated between of electric and plug economic sectors based on electricity sales data provided by the industry through EIA reports. The data for or railroads and railways. Electricity used to charge electricity used in transportation only includes electricity used f electric vehicles would fall under other sectors like residential and commercial use associated with home and public eaking out electricity use d charging stations. As a planned improvement we will look into the possibility of br to charge electric vehicles and report that electricity use under the transportation sector. Lastly, an additional planned improvement is to evaluate and potentially update the methodology for allocating motor gasoline consumpti on across sectors to improve accuracy and create a more consistent time series. Sectoral motor gasoline CO estimates in this Inventory are based on the annual publications of FHWA Highway Statistics 2 - 21, MF - 24, and VM - 1 as well as data from EIA a s discussed in the Methodology section above . In Tables MF 132 - road and non - road applications. 2016, FHWA changed its methods for estimating the share of gasoline used in on While these method changes did not impact overall gasoline consumption CO estimates in this Inventory (see 2 - series discussion of overall gasoline trends in the upfront section of this Chapter ), they did create a time inconsistency , and between 2015 and previous years in the allocation of gasoline among transportation, industrial commercial applications. The method changes resulted in a decrease in the estimated motor gasoline consumption for the transportation sector and a subsequent increase in the commercial and industrial sectors of this Inventory for Among other updat - road recreational 2015. es, FHWA included lawn and garden equipment as well as off - road gasoline consumption for the first time; these non - equipment in its estimates of non sources are road CO 2 included in the industrial and commercial sectors in this Inventory for 2015. While the effects of FHWA’s method changes are still being researched and cannot currently be isolated them from underlying trends, it is estimated estimate that, in absence of the changes, CO tion would have emissions from fossil fuel combustion in transporta 2 percent from 2014 to 2015, instead of the modest decline shown in the time series in two likely increased by roughly this Inventory; industrial CO from fossil fuel combustion would have likely decreased by roughly two percent 2 destly decreasing; and commercial CO by from fossil fuel combustion would have likely decreased instead of mo 2 roughly four percent instead of increasing by eight percent. For 1990 through 2015 trends, it is estimated that, in absence of the changes, CO by emissions from fossil fuel combustion in transportation would have likely increased 2 roughly percent , instead of 16 percent shown in the time series in this Inventory; industrial CO 19 from fossil fuel 2 percent inste decreased combustion by roughly 6 would have likely ad of 4 percent ; and commercial CO from fossil 2 fuel combustion would have likely increased by a few percent instead of 13 percent. EPA received a comment on 131 See < http://www.ipcc nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf >. - 132 The previous and new FHWA methodologies for estimating non - road gasoline consumption are described in Off - Highway and Public - Use Gasoline Consumption Estimation Models Used in the Federal Highway Administration , Publication Number PL - - 17 012. FHWA - 37 - 3 Energy

152 the draft version of this Inventory suggesting that the FHWA method updates be retrospectively ap plied for 2014 EPA plans to conduct further research to better understand FHWA’s method updates and to and earlier years. consult EIA and FHWA regarding potential improvements to the method for gasoline sectoral allocation across the time series . and N O from Stationary Combustion CH 4 2 Methodology O emissions from stationary combustion were estimated by multiplying fossil fuel and wood Methane and N 2 consumption data by emission factors (by sector and fuel type for industrial, residential, commercial, and U.S. Territories; and by fuel and technology type for the electric power sector). T he electric power sector utilizes a Tier 2 methodology, whereas all other sectors utilize a Tier 1 methodology. The activity data and emission factors used are the following subsections. described in Industrial, Residential, Commercial, and U.S. Territories National coal, natural gas, fuel oil, and wood consumption data were grouped by sector: industrial, commercial, erritories. T residential, and U.S. For the CH mates, wood consumption data for the United States was and N O esti 2 4 EIA 2017a ). Fuel consumption data for coal, natural gas, and fuel oil obtained from EIA’s Monthly Energy Review ( EIA 2017a ). B ecause the United for the United States were also obtained from EIA’s Monthly Energy Review ( States does not include territories in its national energy statistics, fuel consumption data for territories were provided 133 International Energy Statistics (EIA 201 7b ) and Jacobs (2010). separately by EIA’s Fuel consumption for the indust rial sector was adjusted to subtract out construction and agricultural use, which is reported under mobile 134 ) Construction and agricultural fuel use was obtained from EPA (201 6c sources. and FHWA (1996 through 2016) . Estimates for wood biomass consumption f or fuel combustion do not include wood wastes, liquors, municipal solid - waste, tires, etc., that are reported as biomass by EIA. Tier 1 default emission factors for these three end use sectors 2006 IPCC Guidelines for National Greenhou (IPCC 2006). U.S. se Gas Inventories were provided by the T erritories’ emission factors were estimated using the U.S. emission factors for the primary sector in which each fuel was combusted. Electric Power Sector The electric power sector uses a Tier 2 emission estimation methodology as fuel consumption for the electricity - technology type was obtained from EPA’s Acid Rain Program Dataset (EPA 201 6a ). generation sector by control - - and fuel This combustion technology use data was available by facility from 1996 to 201 5. The T ier 2 emission factors used were taken from IPCC ( 2006 ) , which in turn are based on emission factors published by EPA. ) total energy consumption estimates, the Since there was a difference between th e EPA (201 6a ) and EIA (201 7a from EIA (201 ) was apportioned to each combustion technology type and fuel 6a remaining energy consumption combination using a ratio of energy consumption by technology type from 1996 to 5 . 201 Energy consumption estimates were not available from 1990 to 1995 in the EPA (201 6a ) datase t, and as a result, consumption was calculated using total electric power consumption from EIA (201 7a ) and the ratio of combustion technology and fuel types from EPA (201 6a ). The consumption estimates from 1990 to 1995 were estimated by applying the 1996 c onsumption ra tio by combustion technology type to the total EIA consumption for each year from 1990 to 1995. Emissions were estimated by multiplying fossil fuel and wood consumption by technology - and fuel - specific Tier 2 IPCC emission factors. 133 U.S. Territories data also include combustion from mobile activities because data to allocate territories’ energy use were and N unavailable. For this reason, CH O emissions from combustion by U.S. Territories are only included in the stationary 2 4 ion totals. combust 134 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 ces. sour - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 38 3

153 Lastly, the re were significant differences between wood biomass consumption in the electric power sector between 7a nd ) datasets. The higher wood biomass consumption from EIA (201 7a ) in the electric ) a EIA (201 the EPA (201 6a tial, commercial, and industrial sectors according to their percent share of power sector was distributed to the residen ). 7a wood biomass energy consumption calculated from EIA (201 More detailed information on the methodology for calculating emissions from stationary combustion, including emission factors and activity data, is provided in Annex 3.1. Series Consistency Uncertainty and Time - Methane emission estimates from stationary sources exhibit high uncertainty, primarily due to difficulties in eplaces and wood stoves). The estimates of CH and N calculating emissions from wood combustion (i.e., fir O 4 2 emissions presented are based on broad indicators of emissions (i.e., fuel use multiplied by an aggregate emission stion technology and type of factor for different sectors), rather than specific emission processes (i.e., by combu emission control). use sector, using the IPCC - recommended An uncertainty analysis was performed by primary fuel type for each end - @RISK Approach 2 uncertainty estimation methodology, Monte Carlo Stochastic Simulation technique, with software. The uncertainty estimation model for this source category was developed by integrating the CH and N O stationary 2 4 source inventory estimation models with the model for CO from fossil fuel combustion to realistically characterize 2 About 55 input variables ction (or endogenous correlation) between the variables of these three models. the intera emissions from fossil were simulated for the uncertainty analysis of this source category (about 20 from the CO 2 del and about 35 from the stationary source inventory models). fuel combustion inventory estimation mo - related input In developing the uncertainty estimation model, uniform distribution was assumed for all activity 135 variables and N O emission factors, based on the SAIC/EIA (2001) report. For these variables, the uncertainty 2 136 ranges were assigned to the input variables based on the data reported in SAIC/EIA (2001). However, the CH 4 emission factors differ from those used by EIA. These factors and uncertainty ranges are based on IPCC default unc ertainty estimates (IPCC 2006). Stationary The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3 - 17 . including Eq. at MMT CO 4. combustion CH emissions in 2015 ( 16.5 biomass) were estimated to be between and 5 4 2 fidence level. This indicates a range of 36 percent below to 136 a 95 percent con percent above the 2015 emission 137 7.0 MMT CO biomass) were estimated Eq . estimate of Stationary combustion N O emissions in 2015 ( including 2 2 Eq. at a 95 percent confidence level. This indicates a range of percent to be between 18.2 and 34. 7 MMT CO 22 2 percent above the 2015 emission estimate of 23. 1 MMT 50 CO below to Eq. 2 135 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 ran dom co mponent). 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. 136 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, - use sector in the current for purposes of assigning uncertaint y estimates to the sub - fuel type categories within each end uncertainty analysis, SAIC/EIA (2001) - reported uncertainty estimates were extrapolated. 137 he nearest integer values and the high The low emission estimates reported in this section have been rounded down to t emission estimates have been rounded up to the nearest integer values. 39 - 3 Energy

154 Table 3 17 : Approach 2 Quantitative Uncertainty Estimates for CH - and N O Emissions from 2 4 Related Stationary Combustion, Including Biomass (MMT CO Eq. and Percent) - Energy 2 a Emission Estimate 2015 Uncertainty Range Relative to Emission Estimate Source Gas (MMT CO Eq.) (MMT CO (%) Eq.) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound CH 4 Stationary Combustion 7.0 4.5 16.5 36% +136% - N O 2 +50% 23.1 18.2 34.7 - 22% Stationary Combustion 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 CH and N O are greater than those associated with 2 4 estimates of CO from fossil fuel combustion, which mainly rely on the carbon content of the fuel combusted. 2 and N O estimates are due to the fact that emissions are estimated based on emission Uncertainties in both CH 4 2 factors representing only a limited subset of combus tion 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. series to ensure time Methodological recalculations were applied to th e entire time series consistency from 1990 - through 201 5 as discussed below Details on the emission trends through time are described in more detail in the . Methodology section, above. consistent with Volume 1, For more information on the general QA/QC process applied to this source category, , see section in the introduction of the 2006 IPCC Guidelines QA/QC and Verification Procedures Chapter 6 of the IPPU Chapter. QA/QC and Verification A source - specific QA/QC plan for stationary combustion was developed and implemented. This effort included a Tier 1 ) analysis, as well as portions of a category - specific ( Tier 2 general ( analysis. The Tier 2 procedures that were ) implemented involved chec ks specifically focusing on the activity data and emission factor sources and O, and the indirect greenhouse gases from stationary combustion in the methodology used for estimating CH , N 4 2 Emission totals for the different sectors and fuels w ere compared and trends were investigated. United States. Recalculations Discussion Methane and N O emissions from stationary sources (excluding CO ) across the entire time series were revised due 2 2 revised data from EIA (201 7a ) , EIA (2017b), and EPA (201 6a ) relative to the previous Inventory. The historical data changes resulted in an average annual Eq. (less than 0.1 percent) in CH increase of less than 0.1 MMT CO 4 2 decrease of less than 0.1 MMT CO Eq. ( less than 0.1 emissions, and an average annual percen t) in N O emissions 2 2 4 . from stationary combustion for the period 1990 through 201 Planned Improvements and N Several items are being evaluated to improve the CH O emission estimates from stationary combustion and 2 4 to reduce uncertainty . Efforts will be taken to work with EIA and other agencies to improve the for U.S. Territories T erritories data. Because these data are not broken out by stationary and mobile uses, further quality of the U.S. research will be aimed at trying to allocate consumption appropriat ely. In addition, the uncertainty of biomass emissions will be further investigated since it was expected that the exclusion of biomass from the estimates would Th ese improvements are reduce the uncertainty; and in actuality the exclusion of biomass increases the uncertainty. not all - inclusive, but are part of an ongoing analysis and efforts to continually improve these stationary combustion estimates from U.S. Territories . icultural use, which is reported Fuel use was adjusted for the industrial sector to subtract out construction and agr also include emissions from sources that may be captured as O and N CH . Mobile source under mobile sources 4 2 – 201 - Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 5 40 3

155 on part of the commercial sector. Future research will look into the need to adjust commercial sector fuel consumpti to account for sources included elsewhere. and N O from Stationary Combustion category involve continued research into the Future improvements to the CH 4 2 using CH data reported and N , O from stationary combustion data availability of , for example from other sources 2 4 under EPA’s GHGRP. In examining data from EPA’s GHGRP that would be useful to improve the emission estimates for CH and N O from Stationary Combustion category, particular attention will be made to ensure time 2 4 cility - level reporting data from EPA’s GHGRP are not available for all Inventory years series consistency, as the fa as reported in this Inventory. In implementing improvements and integration of data from EPA’s GHGRP, the latest 138 level data - in national inventories will be relied upon. guidance from the IPCC on the use of facility CH and N O from Mobile Combustion 2 4 Methodology O emissions from mobile combustion were calculated by multiplying emission factors by and N Estimates of CH 2 4 measures of activity for each fuel and vehicle type (e.g., light - duty gasoline trucks). Activity data included vehicle miles traveled (VMT) for on - road vehicles and fuel consumption for non - road mobile sources. The activity data and emission factors used are described in the subsections that follow. A compl ete discussion of the methodology used to estimate CH O emissions from mobile combustion and the emission factors used in the calculations is provided and N 2 4 in Annex 3.2. On - Road Vehicles - O emissions from gasoline and diesel on ad vehicles are based on VMT and emission and N ro Estimates of CH 4 2 factors by vehicle type, fuel type, model year, and emission control technology. Emission estimates for alternative 139 fuel vehicles (AFVs) are based on VMT and emission factors by vehicle and fuel type. ctors for gasoline and diesel on road vehicles utilizing Tier 2 and Low Emission Vehicle (LEV) Emission fa - technologies were developed by ICF (2006b); all other gasoline and diesel on - road vehicle emissions factors were ved from EPA, California Air Resources Board (CARB) and developed by ICF (2004). These factors were deri Environment Canada laboratory test results of different vehicle and control technology types. The EPA, CARB and covers three separate Environment Canada tests were designed following the Federal Test Procedure (FTP), which 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 re presents 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 t o 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 running emissions for each vehicle class emissions. These were then recombined based upon the ratio of start to from MOBILE6.2, an EPA emission factor model that predicts gram per mile emissions of CO , CO, HC, NO , and 2 x 140 PM from vehicles under various conditions, to approximate average driving characteristics. Emission fact ors for AFVs were first developed by ICF (2006a) after examining Argonne National Laboratory’s GREET 1.7 – Transportation Fuel Cycle Model (ANL 2006) and Lipman and Delucchi (2002). These sources describe AFV emission factors in terms of ratios to convention al vehicle emission factors. Ratios of AFV to conventional vehicle emissions factors were then applied to estimated Tier 1 emissions factors from light - duty - Emissions factors for heavy - gasoline vehicles to estimate light duty AFVs. in relation to duty AFVs were developed 138 See < http://www.ipcc - nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf >. 139 Alternative fuel and advanced technology vehicles are those tha t 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. 140 Additional information regarding the model can be found online at < https:/ /www.epa.gov/moves/description - and - history - - model >. - factor mobile - highway - vehicle - emission 41 - 3 Energy

156 gasoline heavy - A complete discussion of the data source and methodology used to determine duty vehicles. emission factors from AFVs is provided in Annex 3.2. Annual VMT data for 1990 through 2015 were obtained from the Federal Highway Admi nistration’s (FHWA) Highway Performance Monitoring System database as reported in Highway Statistics (FHWA 1996 through 141 2016). - specific vehicle categories VMT estimates were then allocated from FHWA’s vehicle categories to fuel using the calculated shares of vehicle fuel use for each vehicle category by fuel type reported in DOE (1993 through 2016) and information on total motor vehicle fuel consumption by fuel type from FHWA (1996 through 2016). VMT for AFVs were estimated based on Browning (2017). The ag e distributions of the U.S. vehicle fleet were obtained from EPA (201 , 2000), and the average annual age - specific vehicle mileage accumulation of U.S. 6b vehicles were obtained from EPA (20 16b ). Control technology and standards data for on - road vehicles we re 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 road mobile sources, fuel consumption data were employed as a measure of To estimate emissions from non - activity, and multiplied by fuel O and CH per kilogram of fuel specific emission factors (in grams of N - 2 4 142 consumed). ata were obtained from AAR (2008 through 2016), APTA (2007 through 2016), APTA Activity d (2006), BEA (1991 through 2015), Benson (2002 through 2004), DHS (2008), DLA Energy (2015), DOC (1991 through 2015), DOE (1993 through 2015), DOT (1991 through 2016), EIA (2002, 2007, 2016a), EIA (2007 through 2016), EIA (1991 through 2016), EPA (2016b), Esser (2003 through 2004), FAA (2017), FHWA (1996 through , 143 144 2016), - road modes were taken Gaffney (2007), and Whorton (2006 through 2014). Emission factors for non (2006) and Browning (2009). from IPCC Uncertainty and Time - Series Consistency A quantitative uncertainty analysis was conducted for the mobile source sector using the IPCC - recommended Approach 2 uncertainty estimation methodology, Monte Carlo Stochastic Simulation technique, using @RISK software. The uncertainty analysis was performed on 2015 estimates of CH and N O emissions, incorporating 4 2 probability distribution functions associated with the major input variables. For the purposes of this analysis, the ty was modeled for the following four major sets of input variables: (1) VMT data, by on - road vehicle and uncertain 141 - 1. In 2011, FHWA changed its methods for estimating data in The source of VMT is FHWA Highway Statistics Table VM - 1 table. These methodological changes inclu the VM - type ded how vehicles are classified, moving from a system based on body to one that is based on wheelbase. These changes were first incorporated for the 1990 through 2010 Inventory and apply to th e 2007 through 2015 time period. This resulted in large ch anges 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 b y “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 passenge r vehicle category in this Inventory. These changes are reflected in a lar ge drop in light - truck emissions between 2006 and 2007. 142 The consumption of international bunker fuels is not included in these activity data, but is estimated separately under the International Bunker Fuels source category. 143 methods for estimating the share of gasoline used in on road and non These In 2016, FHWA changed its road applications. - - O estimates method changes created a time series inconsistency in this Inventory between 2015 and previous years in CH - and N 2 4 for agricultural, construction, commercial, and industrial non - road mobile sources. The method updates are discussed further in the Planned Improvements section below . 144 This Inventory uses FHWA’s Agriculture, Construction, and Commercial/Indu strial MF - 24 fuel volumes along with the MOVES NONROAD model gasoline volumes to estimate non - road mobile source CH 0 emissions for these categories. and N 2 4 For agriculture, the MF - 24 gasoline volume is used directly because it includes both off - road truck s and equipment. For - - road trucks only; 24 volumes represented off construction and commercial/industrial gasoline estimates, the 2014 and older MF spective therefore, the MOVES NONROAD gasoline volumes for construction and commercial/industrial are added to the re categories in the Inventory. Beginning in 2015, this addition is no longer necessary since the FHWA updated its methods for road gasoline consumption. Among the method updates, FHWA now incorporates MOVES - road and non - estimating on nt gasoline volumes in the construction and commercial/industrial categories NONROAD equipme - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 42 3

157 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 road vehicle and equ ipment type, and (4) emission factor data, by non type. Uncertainty analyses were not conducted for NO , CO, or NMVOC emissions. Emission factors for these gases have x been extensively researched since emissions of these gases fro m 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 – Uncertainty Analysis of Emission Estimates. However, a much higher level of uncertainty is associated with 3.8 and N O emission factors due to limited emission test data, and because, unlike CO emissions, the emission CH 2 2 4 and N pathways of CH O ar e highly complex. 2 4 Mobile combustion CH emissions from all mobile sources in 201 5 were estimated to be between 1.6 and 2.5 MMT 4 percent above the Eq. at a 95 percent confidence level. CO This indicates a range of 18 percent below to 27 2 5 e mission estimate of 2.0 MMT CO corresponding 201 Eq. Also at a 95 percent confidence level, mobile 2 combustion N O emissions from mobile sources in 201 5 were estimated to be between 13.2 and 17.9 MMT CO 2 2 13 19 percent above the corresponding 201 5 percent below to emission estimate of 15.1 Eq., indicating a range of Eq. MMT CO 2 2 Quantitative Uncertainty Estimates for CH 3 18 : Approach Table - O Emissions from and N 2 4 Mobile Sources (MMT CO Eq. and Percent) 2 a a Uncertainty Range Relative to Emission Estimate 2015 Emission Estimate Source Gas (MMT CO (%) Eq.) (MMT CO Eq.) 2 2 Upper Lower Upper Lower Bound Bound Bound Bound - Mobile Sources 2.0 1.6 2.5 CH 18 % + 27 % 4 N O 15.1 13.2 17.9 - Mobile Sources % + 19 % 13 2 a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. This uncertainty analysis is a continuation of a multi year process for developing quantitative uncertainty estimates - for this source category using the IPCC Approach 2 uncertainty analysis. As a result, as new information becomes available, uncertainty characterization of input variables may be improved and revised. For additional information regarding uncertainty in emi and N O please refer to the Uncertainty Annex. ssion estimates for CH 2 4 QA/QC and Verification A source - specific Quality Assurance/Quality Control plan for mobile combustion was developed and implemented. This plan is based on the IPCC - recommended QA/QC Pl an. The specific plan used for mobile combustion was updated prior to collection and analysis of this current year of data. This effort included a general (Tier 1) analysis, as well as portions of a category specific (Tier 2) analysis. The Tier 2 procedure s focused on the emission factor and - activity data sources, as well as the methodology used for estimating emissions. These procedures included a qualitative assessment of the emission estimates to determine whether they appear consistent with the most rec ent activity data and emission factors available. A comparison of historical emissions between the current Inventory and the previous Inventory was also conducted to ensure that the changes in estimates were consistent with the changes in activity data and emission factors. Recalculations Discussion - road CH ncrease to and N Several updates were made to on O emissions calculations this year resulting in a net i 4 2 relative to the previous Inventory. First and N duty trucks O emissions from mobile combustion CH several lig ht - 2 4 were re - characterized as heavy - duty vehicles based upon gross vehicle weight rating (GVWR) and confidential sales dards each vehicle type was assumed to have met were data Second , which emission stan re - examined using . confidential sales data . Also, in previous Inventories , non - plug - in hybrid electric vehicles (HEVs) were considered alternative fueled vehicles and therefore were not included in the engine technology breakouts. For this Inventory, HEVs are now classified as gasoline vehicles ac ross the entire time series. PHEVs (plug - in hybrid electric vehicles) electric vehicles. are as Estimates of alternative fuel vehicle continue to be considered alternative fuel vehicles , 43 - 3 Energy

158 mileage for the last ten years were revised to reflect updates made to Energy Information Administration (EIA) data on alternative fuel use and vehicle counts. The energy economy ratios (EERs) in the alternative fuel vehicle analysis his inventory. EERs are the ratio of the gasoline equivalent fuel economy of a given were also updated in t These were updated to reflect the latest GREET technology to that of conventional gasoline or diesel vehicles. ). Overall, these changes resulted in an average annual model released in 2016 ( crease of 0. 02 MMT 6 ANL 201 in Eq. ( 1 CO percent) in CH emissions and an average annual in crease of 0.5 MMT CO Eq. ( 2 percent) in N O 4 2 2 2 4 emissions from mobile combustion for the period 1990 through 201 , relative to the previous report. series consistency from 1990 - Methodological recalculations were applied to the entire time series to ensure time through 2015 with two recent notable exceptions. First, an update to the method for estimating on - road VMT created - road CH O for the time periods 1990 to 2006 and 2007 to 2015. Second, an update to and N an inconsistency in on 2 4 the method for estimating share of motor gasoline used in on - road and non - road applications created an inconsistency in non - road CH and N 2014 and 2015 (discussed in more detail in the O for the time periods 1990 to 2 4 Planned Improvements section below). Details on the emission trends and methodological inconsistencies through time are described in more detail in the Methodology section, above. Improvements Planned While t he 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 u pdate our method for estimating mot or 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, CH road and N - O estimates for gasoline - powered non 2 4 sources in this Inventory are based on a variety of inputs, including FHWA Highway Statistics Table MF - road 24. In 2016, FHWA changed its methods for estimating the share of gasoline used in on - road and non - 145 between 2015 n this Inventory i hese method changes applications. series inconsistency T - c reate d a time and previous years in CH and N - O estimates for agricultural, construction, commercial, and industrial non 2 4 . While we are still researching the effects of FHWA’s method changes and cannot road mobile sources currently isolate them from underlying trends, we estimate that, in absence of the changes, CH and N O 4 2 emissions from non - road mobile sources would have likely increased by roughly 0.04 MMT CO Eq. 2 (roughly two percent increase ) from 2014 to 2015, instead of the 0.06 MMT CO Eq decrease (three 2 shown in the time series in this Inventory; and would have increased by roughly 0.9 percent decrease) MMT CO (63 percent Eq (roughly 72 percent increase) from 1990 to 2015 instead of 0.8 MMT CO Eq 2 2 increase). EPA received a comment on the draft version of this Inventory suggesting that we retrospectively apply the FHWA method updates for 2014 and earlier years. EPA plans to conduct further research to s better understand FHWA’s method update and to consult EIA and FHWA regarding pote ntial improvements to our method for estimating gasoline consumption for non - road mobile sources across the time series . Continue to explore potential improvements to estimates of domestic waterborne fuel consumption for • future Inventories. The Inventory e stimates 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 Third IMO GHG international sources for domestic bunker fuels, including data from studies such as the Study 2014 , continues to be explored. • Continue to examine the use of EPA’s MOVES model in the development of the Inventory estimates, including use for uncertainty analysis. Although the Inventory uses some of the underlying data from 145 The previous and new FHWA methodologies for estimating non - road gasoline are described in Off - Highway and Public - Use , Publication Number FHWA - 012. PL - 17 - Gasoline Consumption Estimation Models Used in the Federal H ighway Administration - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 44 3

159 MOVES, such as vehicle age distributions by model year, MOVES is not used directly in c alculating and N O mobile source emissions. The use of MOVES is currently being evaluated to develop new CH 2 4 - road and non road sources (including LPG and LNG non - road equipment), which emissions factors for on - . Oth er approaches for updating CH O emissions factors, and N may be integrated into future inventories 2 4 including use of the latest GREET model, are also being considered. Carbon Emitted from Non Energy Uses of - 3.2 Fossil Fuels (IPCC Source Category 1A) In addition to being combusted for energy, fossil f energy uses (NEU) in the United uels are also consumed for non - The fuels used for these purposes are diverse, including natural gas, liquefied petroleum gases (LPG), asphalt States. (manufactured from heavy oil), and coal (a viscous liquid mixture of heavy crude oil distillates), petroleum coke (metallurgical) coke (manufactured from coking coal). energy applications of these fuels are equally The non - ; reducing diverse, including feedstocks for the manufacture of plastics, rubber, synthetic fibers and other materials agents for the production of various metals and inorganic products; and non energy products such as lubricants, - waxes, and asphalt (IPCC 2006). emissions arise from non - energy uses via several pathways. Emissions may occur during the Carbon dioxide - Additionally, derived feedstocks. manufacture of a product, as is the case in producing plastics or rubber from fuel emissions may occur during the product’s lifetime, such as during solvent use. Overall, throughout the time series about 6 2 percent of the total C consumed for non - energy purposes was stored in products, and and across all uses, not released to the atmosphere; the remaining 38 percent was emitted. There are several areas in which non - er parts of this Inventory. energy uses of fossil fuels are closely related to oth For example, some of the NEU products release CO at the end of their commercial life when they are combusted 2 after disposal; these emissions are reported separately within the Energy chapter in the Incineration of Waste source category. - energy uses and the fossil - In addition, there is some overlap between fossil fuels consumed for non derived CO emissions accounted for in the Industrial Processes and Product Use chapter, especially for fuels used 2 as reducing agents. To avoid do uble - counting, the “raw” non - energy fuel consumption data reported by EIA are There are also net exports of petrochemicals that are not completely modified to account for these overlaps. accounted for in the EIA data, and the I nventory calculations adjust for the effect of net exports on the mass of C in non - energy applications. 25.5 As shown Table 3 - 19 , fossil fuel emissions in 2015 from the non - energy us es of fossil fuels were 1 in MMT CO Eq., which constituted approximately 2 2015 percent of overall fossil fuel emissions. In , the consumption of 2 fuels for non - energy uses (after the adjustments described above) was 4, 985.4 TBtu , an increase of 11.3 percent Table 3 since 1990 ( see Eq.) of the C in these fuels was stored, while the - 20 ). About 5 8.4 MMT ( 2 14.0 MMT CO 2 remaining MMT C ( 1 25.5 MMT CO 34.2 Eq.) was emitted. 2 Emissions from Non 3 19 : CO Eq. and Table - - Energy Use Fossil Fuel Consumption (MMT CO 2 2 P ercent) Year 1990 2005 2011 2012 2013 2014 2015 324.2 Potential Emissions 312.1 377.5 318.0 312.9 328.9 339.5 194. 5 238.6 208.2 C Stored 206.1 205.3 205.2 214.0 Emissions as a % of Potential 38% 37% 35% 34% 38% 37% 37% 119.0 117. 6 138.9 109.8 106.7 123.6 Emissions 125.5 Methodology C stored in products was to determine the aggregate quantity of fossil fuels consumed for The first step in estimating non The C content of these feedstock fuels is equivalent to potential emissions, or the product of energy uses. - 45 - 3 Energy

160 consumption and the fuel - B oth the non - energy fuel consumption and C content data were specific C content values. 6 ) (see Annex 2.1). Consumption of natural gas, LPG, pentanes plus, naphthas, other supplied by the EIA (2013, 201 ts that are not reflected in the raw oils, and special naphtha were adjusted to account for net exports of these produc data from EIA. Consumption values for industrial coking coal, petroleum coke, other oils, and natural gas in Table 20 and Table 3 - 21 have been adjusted to subtract non - energy uses that are included in the 3 - source categories of the , 147 146 Consumption values were also adjusted to subtract net exports Industrial Processes and Product Use chapter. of intermediary chemicals. energy uses, the quantity of C stored was estimated by multiplying the For the remaining non potential emissions by - a storage factor. For several fuel types — petrochemical feedstocks (including natural gas for non - fertilizer uses, LPG, • oil, pentanes plus, naphthas, other oils, still gas, special naphtha, and industrial other coal), asphalt and road — U.S. data on C stocks and flows were used to develop C storage factors, calculated lubricants, and waxes - energy products to (b) the total C content of the fuel as the ratio of (a) the C stored by the fuel’s non A lifecycle approach was used consumed. in the development of these factors in order to account for losses in the production process and during use. Because losses associated with municipal solid waste sector under the Incineration of Waste source management are handled separately in the Energy category, the storage factors do not account for losses at the disposal end of the life cycle. • For industrial coking coal and distillate fuel oil, storage factors were taken from IPCC (2006), which in turn draws from Marland and Rotty (1984). • aining fuel types (petroleum coke, miscellaneous products, and other petroleum), IPCC does not For the rem provide guidance on storage factors, and assumptions were made based on the potential fate of C in the respective NEU products. 3 - 20 : Adjusted Consumption of Fossil Fuels for Non - Energy Uses (TBtu) Table Year 1990 2005 2011 2012 2013 2014 2015 Industry 4,215.8 5,110.7 4,470.1 4,377.3 4,621.1 4,597.8 4,760.0 0 . 0 80.4 60.8 Industrial Coking Coal 132.5 119.3 48.8 121.8 Industrial Other Coal 8.2 11.9 10.3 10.3 10.3 10.3 10.3 Natural Gas to Chemical Plants 281.6 260.9 297.1 292.7 297.0 305.1 302.3 Asphalt & Road Oil 1,170.2 1,323.2 859.5 826.7 783.3 792.6 831.7 2,109.7 2,157.7 LPG 1,120.5 1,610.0 1,865.6 1,887.3 2,062.9 Lubricants 186.3 160.2 141.8 130.5 138.1 144.0 156.8 Pentanes Plus 117.6 95.5 26.4 40.3 45.4 43.5 78.4 Naphtha (<401 °F) 326.3 679.5 469.4 498.8 435.2 417.9 432.3 Other Oil (>401 °F) 662.1 499.4 368.2 267.4 209.1 236.2 216.8 Still Gas 36.7 67.7 163.6 166.7 164.6 162.2 160.6 Petroleum Coke 27.2 105.2 0.0 0.0 0.0 0.0 0.0 Special Naphtha 100.9 60.9 21.8 14.1 96.6 104.5 97.0 Distillate Fuel Oil 7.0 11.7 5.8 5.8 5.8 5.8 5.8 Waxes 33.3 31.4 15.1 16.5 14.8 12.4 15.3 Miscellaneous Products 137.8 112.8 164.7 161.6 171.2 182.7 188.9 Transportation 176.0 151.3 133.9 123.2 130.4 136.0 148.1 146 These source categories include Iron and Steel Production, Lead Production, Zinc Production, Ammonia Manufacture, Carbon Black Manufac ture (included in Petrochemical Production), Titanium Dioxide Production, Ferroalloy Production, Silicon Carbide Production, and Aluminum Production. 147 Some degree of double counting may occur between these estimates of non - energy use of fuels and proces s 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 t han disaggregated petrochemical production) as currently collected through EPA ’s GHGRP and by both fuel type and particular industries (e.g. , . used for the petrochemical production category - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 46 3

161 Lubricants 176.0 151.3 133.9 123.2 130.4 136.0 148.1 U.S. Territories 123.2 75.6 72.0 82.4 77.3 77.3 85.6 4.6 1.0 1.0 0.7 1.0 1.0 1.0 Lubricants 84.9 118.6 74.6 71.0 Other Petroleum (Misc. Prod.) 76.2 76.2 81.4 4,833.9 5,385.2 4,679.7 4,572.5 4,811.1 4,985.4 Total 4,477.4 3 Table 21 : 201 5 Adjusted Non - Energy Use Fossil Fuel Consumption, Storage, and Emissions - Adjusted Carbon Non Carbon Energy Content Carbon Potential Carbon Storage - a Carbon Stored Coefficient Emissions Factor Emissions Use (MMT (MMT CO (MMT C) C/QBtu) (MMT C) (MMT C) (TBtu) Eq.) 2 Sector/Fuel Type 4,760.0 NA 88.1 NA 57.9 30.1 110.5 Industry 0.4 31.00 3.8 0.10 3.4 12.5 Industrial Coking Coal 121.8 10.3 25.82 0.3 0.65 0.2 0.1 0.3 Industrial Other Coal Natural Gas to 2.9 302.3 14.47 4.4 1.4 5.3 Chemical Plants 0.65 20.55 17.1 1.00 17.0 0.1 0.3 Asphalt & Road Oil 831.7 2157.7 17.06 36.8 0.65 24.6 12.2 44.6 LPG 20.20 3.2 Lubricants 156.8 0.3 2.9 10.5 0.09 1.8 Pentanes Plus 78.4 19.10 1.5 0.65 1.0 0.5 417.9 18.55 7.8 0.65 5.2 2.6 9.4 Naphtha (<401° F) 20.17 4.4 0.65 2.9 5.3 Other Oil (>401° F) 216.8 1.4 17.51 2.8 0.65 1.9 0.9 3.4 Still Gas 162.2 27.85 + 0.30 + + + + Petroleum Coke 97.0 19.74 1.9 0.65 Special Naphtha 1.3 0.6 2.3 0.1 5.8 0.1 0.50 0.1 20.17 0.2 Distillate Fuel Oil 12.4 19.80 0.2 0.58 Waxes 0.1 0.4 0.1 + 188.9 20.31 3.8 + Miscellaneous Products 3.8 14.1 NA 148.1 NA 3.0 2.7 0.3 10.0 Transportation 148.1 20.20 3.0 0.09 0.3 2.7 10.0 Lubricants NA 1.5 NA 0.2 77.3 1.4 5.1 U.S. Territories 1.0 20.20 + 0.09 Lubricants + + 0.1 Other Petroleum (Misc. 0.2 76.2 20.00 1.5 1.4 5.0 Prod.) 0.10 58.4 92.6 34.2 125.5 Total 4,985.4 + Does not exceed 0.05 TBtu , MMT C, MMT CO Eq. 2 Not A pplicable ) ( NA a To 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 3 - 19 ). Table 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 e nergy 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 communication s. 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 Resource Conservation and Recovery 2009), Inventory, 1998 ( EPA 2000b), Biennial Reporting System (EPA 2000a, ), pesticide sales and use estimates (EPA 1998, 1999, 2002, , 2016c (EPA 2013b, 2015b Act Information System 47 - 3 Energy

162 2004, 2011), and the Chemical Data Access Tool (EPA 2012); the EIA Manufacturer’s Energy Consu mption Survey (MECS) (EIA 1994, 1997, 2001, 2005, 2010, 2013); the National Petrochemical & Refiners Association (NPRA Census ); Bank of Canada (2012, 2013, 2014 , 2016 ); Financial Bureau (1999, 2004, 2009 2002); the U.S. , 2014 through NEGI (2006); the United States International Trade Commission (1990 Planning Association (2006); I 6 ); Gosselin, Smith, and Hodge (1984); EPA’s Municipal Solid Waste (MSW) Facts and Figures (EPA 2013a; 201 ); the Rubber Manufacturers’ Association (RMA 2009, 2011, 2014 2014a ); the International Institute of , 2016b , 2016 , 2003, 2005, 2007, 2009, Synthetic Rubber Products (IISRP 2000, 2003); the Fiber Economics Bureau (FEB 2001 2010, 2011, 2012, 2013); the EPA Chemical Data Access Tool (CDAT) (EPA 2014b); the American Chemistry ) through 2011, 2013, 2014, Counci l (ACC 2003 ; and the Guide to the Business of Chemistry (ACC 2015a, 2016b . Specific data sources are listed in full detail in Annex 2.3. 2012, 2015b, 2016a) - Uncertainty and Time Series Consistency ucted to quantify the uncertainty surrounding the estimates of emissions and An uncertainty analysis was cond storage factors from non This analysis, performed using @RISK software and the IPCC - recommended energy uses. - 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. - - specific storage factors for (1) feedstock materials energy use analysis is based on U.S. As noted above, the non (natural gas, LPG, pentanes plus, naphthas, other oils, still gas , special naphthas, and other industrial coal), (2) Table Table 3 - 20 and asphalt, (3) lubricants, and (4) waxes. For the remaining fuel types (the “other” category in 3 - 21 ), the storage factors were taken directly from IPCC (2006) , where avail able, and otherwise assumptions were made based on the potential fate of carbon in the respective NEU products . To characterize uncertainty, five separate analyses or expert analyses were conducted, corresponding to each of the five categories. In all cases, statistical 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. ach 2 quantitative uncertainty analysis are summarized in Table 3 - 22 The results of the Appro Table (emissions) and 3 23 (storage factors). - - energy uses of fossil fuels in 201 5 was estimated to be between Carbon emitted from non 94.4 and 171.54 MMT CO Eq. at a 95 percent confidence level. This indicates a range of 2 5 percent below to 37 2 percent above the 201 emission estimate of 125.5 5 The uncertainty in the emission estimates is a Eq. MMT CO 2 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 CO Emissions from Non - 2 Energy Uses of Fossil Fuels (MMT CO Eq. and Percent) 2 a 201 5 Emission Estimate Uncertainty Range Relative to Emission Estimate Gas Source (MMT CO Eq.) (MMT CO Eq.) (%) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound CO Feedstocks 2 7 2.5 48.9 123.4 - 3 3 % 70 % Asphalt CO 2 - 8 0.3 0.1 0.6 % 5 23 % 1 CO Lubricants 2 20.6 17.0 23.9 - 17 % 16 % Waxes CO 2 0. 0. 4 0. 3 2 7 - 9 % 74 % CO Other 2 31.8 18.7 34.0 - 41% 7% Total CO 2 25% 37% 125.5 94.4 171.5 - 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. 201 - 5 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 48 3

163 Table 3 : Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non - - 23 Energy Uses of Fossil Fuels (Percent) a 201 5 Storage Factor Uncertainty Range Relative to Emission Estimate Gas Source (%) (%) (%, Relative) Lower Upper Lower Upper Bound Bound Bound Bound Feedstocks CO 2 7 % 5 3 % 72% 6 - 20 % 8 % Asphalt CO 2 99 .1 % 99.6 99.8 % % - 0.5 % 0.3 % Lubricants CO 2 4% 18 % - 57% 9% 91 % Waxes CO 2 58% 49% 71 % - 1 5 % 22% Other CO 2 6 6 % 43 % - 8 % 572 % % 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). Table 3 - 23 , feedstocks and asphalt contribute least to overall storage factor uncertainty on a percentage basis. In — appears to have tight flows — Although the feedstocks category the largest use category in terms of total carbon 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 Rather (e.g., plastics production), and eleven that result in emissions (e.g., volatile organic compound emissions). 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. pplied to the entire time series to ensure time - Methodological recalculations were a series consistency from 1990 . 5 as discussed below through 201 Details on the emission trends through time are described in more detail in the Methodology section, above. QA/QC and Verification - specif A source uality Assurance/Quality Control plan for non - energy uses of fossil fuels was developed and ic Q implemented. This effort included a Tier 1 analysis, as well as portions of a Tier 2 analysis for non - energy uses involving petrochemical feedstocks and for imp orts and exports. The Tier 2 procedures that were implemented involved checks specifically focusing on the activity data and methodology for estimating the fate of C (in terms of storage and emissions) across the various end - Emission and storage totals for the different uses of fossil C. 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 for 201 4 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 throu gh 2009, outputs exceeded inputs. In 2010 and 2011, inputs exceeded outputs, and in 2012, outputs slightly exceeded A portion of this discrepancy has been reduced and two inputs. In 2013 through 2015, inputs exceeded outputs. g portion (see Planned Improvements, below). to address the remainin strategies have been developed 49 - 3 Energy

164 Recalculations Discussion A number of updates to historical production values were included in the most recent Monthly Energy Review; these have been populated throughout this document. The categorization of hazardous waste data have been revised in accordance with EPA’s RCRAinfo Form Code List to correct the total number of tons burned for each waste type for years 2001 through 2014. The quantity of tons burned associated with the waste types Aqueous Waste, Organic Liquids and Sludges, Organic Solids, and Inorganic Solids has been revised to reflect the correct categorizations, resulting in increased historical emissions from hazardous waste incineration those years. Planned Improvements There are several improvements planned for the future: • Analyzing the fuel and feedstock data from EPA’s GHGRP to better disaggregate CO emissions in NEU 2 model and CO process emissions from petrochemical production. 2 • More accurate accounting of C in EPA has worked with EIA to determine the petrochemical feedstocks. cause of input/output discrepancies in the C mass balance contained within the NEU model. In the future, First, accounting of C in two strategies to reduce or eliminate this discrepancy will continue to be pursued. 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 inp ut calculation in estimating emissions will be reconsidered. Alternative approaches that rely more substantially on the bottom - up C output calculation will be considered instead. • Response to potential changes in NEU input data. In 2013 EIA initiated imple mentation of new data reporting definitions for Natural Gas Liquids (NGL) and Liquefied Petroleum Gases (LPG); the new definitions may affect the characterization of the input data that EIA provides for the NEU model and may EIA also obtains and applies proprietary r changes to the NEU methodology. therefore result in the need fo The data for LPG inputs that are not directly applied as NEU input data because the data are proprietary. - potential use of the proprietary data (in an aggregated, non rm) as inputs to the NEU model proprietary fo will be investigated with EIA. • 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). energy uses. Annual consumption for several fuel Reviewing the trends in fossil fuel consumption for non - • 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 petroleu m 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 T year’s worth of solvent composition data. he 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 defi nitions for volatile organic compounds, the source of emissive C in solvents. Additional sources of in order to update the C content assumptions. investigated solvents data will be Updating the average C content of cleansers (soaps and detergents) was rese • arched; although production and consumption data for cleansers are published every 5 years by the Census Bureau, the composition (C - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 50 3

165 content) of cleansers has not been recently updated. Recently available composition data sources may average C content for this category. facilitate updating the • 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 Potential revisions to the plastics methodology to account for the ob tained from personal communications. recent changes in published data will be investigated. Although U.S. - specific storage factors have been developed for feedstocks, asphalt, lubricants, and waxes, • - energy fuel types (industrial coking coal, default values from IPCC are still used for two of the non distillate oil), and broad assumptions are being used for miscellaneous products and other petroleum. Over the long term, there are plans to im prove 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 partic ular, the carbon • black abraded and stored in tires. 3 - Box 6 : Reporting of Lubricants, Waxes, and Asphalt and Road Oil Product Use in Energy Sector IPCC (2006) sions from the first use of fossil fuels as a product provides methodological guidance to estimate emis for primary purposes other than combustion for energy purposes (including lubricants, paraffin waxes, bitumen / 148 In this Inventory, C storage asphalt, and solvents) under the Industrial Processes and Product Use (IPPU) sector. and C emissions from product use of lubricants, waxes, and asphalt and road oil are reported under the Energy sector 149 - Energy Uses of Fossil Fuels source category (IPCC Source Category 1A). in the Carbon Emitted from Non The emiss ions 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 - s pecific 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 ricants, waxes, asphalt and road oil (see amount of fossil fuels used for non - energy uses, including inputs of lub - ). For those inputs, U.S. country specific data on C stocks and flows are used to develop S ection 3.2 , Table 3 - 21 - carbon storage factors, which are calculated as the ratio of the C stored by the fossil fuel non y products to the energ 150 total C content of the fuel consumed, taking into account losses in the production process and during product use. specific methodology to reflect national circumstances starts with the aggregate amount of fossil fuels The country - - energy uses and applies a C balance calculation, breaking out the C emissions from non - energy use of used f or non lubricants, waxes, and asphalt and road oil. Due to U.S. national circumstances, reporting these C emissions ing artificial adjustments to allocate both the C inputs and C outputs of the separately under IPPU would involve mak - non These artificial adjustments would also result in the C emissions for lubricants, waxes, energy use C balance. 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 e entire calculation of Energy Uses of Fossil Fuels source category for the Carbon Emitted from Non calculation, th - 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 nsumed during non - they were co energy related industrial activity. Emissions from uses of fossil fuels as feedstocks 148 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). 149 Non - methane volatile organic compound (NMVOC) emissions from solvent use are reported separately in the IPPU sector, followin g Chapter 5 of the 2006 IPCC Guidelines . 150 Methodology and Data for Data and calculations for lubricants and waxes and asphalt and road oil are in Annex 2.3 – Estimating CO Emissions from Fossil Fuel Combustion. 2 51 - 3 Energy

166 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. Incineration of Waste (IPCC Source 3.3 Category 1A1a) 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 ; Goldstein the scope of materials included in the definition of solid waste (EPA 2000 es 2001 ; Kaufman et al. 2004 ; Simmons et al. 2006 and Mad van Haaren et al. 2010). In the context of this section, t ; l as tires. In the United States, incineration of MSW waste includes all municipal solid waste (MSW) as wel scrap occur at waste - to energy facilities or industrial facilities where useful energy is recovered, and thus - tends to scrap tires are combusted for y, emissions from waste incineration are accounted for in the Energy chapter. Similarl , pulp and paper mills, and cement kilns . Incineration of waste results energy recovery in industrial and utility boilers in conversion of the organic inputs to CO emitted is of f ossil origin, . According to IPCC guidelines, when the CO 2 2 it is counted as a net anthropogenic emission of CO to the atmosphere. Thus, the emissions from waste incineration 2 are calculated by estimating the quantity of waste combusted and the fraction of the waste that is C derived from fossil sourc es. Most of the organic materials in municipal solid wastes are of biogenic origin (e.g., paper, yard trimmings), and Land Use, Land - Use Change, and Forestry c hapter. However, some have their net C flows accounted for under the — synthetic rubber, synthetic fibers, and carbon black in scrap tires — are of fossil origin. plastics, components Plastics in the U.S. waste stream are primarily in the form of containers, packaging, and durable goods. Rubber is - durable goods, such as clothing and footwear. Fibers in found in durable goods, such as carpets, and in non scrap tires (which municipal solid wastes are predominantly from clothing and home furnishings. As noted above, synthetic and are included in the rubber and carbon black) are also considered a “non - hazardous” waste contain 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 mass C - 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 l to data for the amount of MSW incinerated in 2015 were not available, so data for 2015 was assumed to be equa 2014. CO emissions from incineration of waste rose 34 percent since 1990, to an estimated 10.7 MMT CO Eq. 2 2 (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 CH e 1993; IPCC and N O emissions (De Soet 2 4 Eq. (less 2006). Methane emissions from the incineration of waste were estimated to be less than 0.05 MMT CO 2 than 0.5 kt CH ) in 2015, and have decreased by 32 percent since 1990. Nitrous oxide emissions from the 4 incineration of waste were es O) in 2015, and have decreased by 32 percent Eq. (1 kt N timated to be 0.3 MMT CO 2 2 since 1990. 3 - 24 : CO Eq.) , CH , and N CO O Emissions from the Incineration of Waste ( MMT Table 2 2 4 2 2015 Gas/Waste Product 1990 2005 2014 2011 2012 2013 8.0 12.5 10.6 10.4 10.4 10.6 10.7 CO 2 5.6 6.9 5.8 5.7 5.8 5.9 5.9 Plastics 0.3 1.6 1.4 1.3 1.2 1.2 1.2 Synthetic Rubber in Tires 1.5 2.0 1.7 Carbon Black in Tires 0.4 1.4 1.4 1.5 Rubber in Synthetic MSW 0.9 0.8 0.7 0.7 0.7 0.7 0.7 0.8 1.2 1.1 1.1 1.3 1.3 1.3 Synthetic Fibers + + + CH + + + + 4 0.3 O 0.5 0.4 N 0.3 0.3 0.3 0.3 2 11.0 Total 8.4 12.8 10.9 10.7 10.7 10.9 Eq. + Does not exceed 0.05 MMT CO 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 52 3

167 Table - 25 : CO 3 , CH , and N ) O Emissions from the Incineration of Waste ( kt 2 4 2 1990 2005 2011 2012 2013 2014 2015 Gas/Waste Product CO 7,950 12,469 10,564 10,379 10,398 10,608 10,676 2 5,588 6,919 5,757 Plastics 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 854 766 712 MSW 706 729 729 729 1,166 Synthetic Fibers 1,227 1,070 816 1,284 1,313 1,313 CH + + + + + + + 4 O 1 1 N 1 2 1 1 1 2 kt + Does not exceed 0. 5 Methodology from the incineration of waste include CO generated by the incineration of plastics, synthetic Emissions of CO 2 2 , as well as the incineration of synthetic rubber and carbon black in scrap tires. fibers, and synthetic rubber in MSW basis based on the data available. These The emission estimates are calculated for all four sources on a mass - emissions were estimated by multiplying the mass of each material incinerated by the C content of the material and tic the fraction oxidized (98 percent). Plastics incinerated in municipal solid wastes were categorized into seven plas 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 , and synthetic fibers . contain several ty pes of synthetic rubber, carbon black Each type of synthetic rubber has a were calculated based on the amount of discrete C content, and carbon black is 100 percent C. Emissions of CO 2 er and carbon black content of scrap tires. More detail on the scrap tires used for fuel and the synthetic rubb Annex 3.7 methodology for calculating emissions from each of these waste incineration sources is provided in . aste, data on the quantity of For each of the methods used to calculate CO emissions from the incineration of w 2 in product combusted and the C content of the product are needed. For plastics, synthetic rubber, and synthetic fibers discarded as municipal solid waste (i.e., the quantity generated minu s the MSW , the amount of specific materials quantity recycled) was taken from Municipal Solid Waste Generation, Recycling, and Disposal in the United States: (EPA 2000 through 2003, 2005 through 201 4 ) , and Advancing Sustainable Materials Facts and Figures ng Trends in Material Generation, Recycling and Disposal in the United Management: Facts and Figures: Assessi ( EPA 2015, 2016) and States detailed unpublished backup data for some years not shown in the reports (Schneider 2007). that in 2014, due to the lack of For 2015, the amount of MSW incinerated was assumed to be equal to 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 2012 through 2015 , so th e s e value s w ere assumed to e qual to for For synthetic rubber and carbon black in scrap tires, information was obtained from the 2011 value (Shin 2014). Management Summary for 2005 through 2015 data (RMA 20 16 ). Average C contents for the U.S. Scrap Tire synthetic rubber in municipal solid wastes were calculated from 1998 and 20 “Other” plastics category and 02 content for 1990 through 1998 is based on the 1998 value; C production statistics: 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 interne t site (RMA 20 The 12a ). 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 CO 2 emissi ons) was reported in EPA’s life cycle analysis of greenhouse gas emissions and sinks from management of This percentage is multiplied by the carbon stored to estimate the amount of carbon solid waste (EPA 2006). emitted. Incineration of waste, including M SW, also results in emissions of CH and N O. These emissions were calculated 4 2 N as a function of the total estimated mass of waste incinerated and emission factor . As noted above, CH O and s 2 4 emissions are a function of total waste incinerated in each year; for 1990 through 2008, these data were derived from - 3 53 Energy

168 the information published in BioCycle Data for 2009 and 2010 were interpolated between (van Haaren et al. 2010). inerated was not 2008 and 2011 values. Data for 2011 were derived from Shin (2014). Data on total waste inc BioCycle data set for 2012 through 2015 , so th e s e value w ere assumed to equal the 2011 BioC ycle available in the s data set value. 3 - 2 6 Table provides data on municipal solid waste discarded and percentage combusted for the total waste stream. emissions per quantity of municipal solid waste combu sted are default The emission factors of N O and CH 4 2 - technolo gy type and were taken from the default continuously fed stoker unit MSW incineration emission factors for IPCC ( 2006). - 26 Table Municipal Solid Waste Generation (Metric Tons) and Percent Combusted 3 : (BioCycle dataset) Incinerated (% of Waste Incinerated Year Discards) Waste Discarded 1990 13.0 % 30,632,057 235,733,657 25,973,520 10.0 % 2005 259,559,787 273,116,704 20,756,870 7.6% 201 1 a 20,756,870 7.6% 273,116,704 2012 a 7.6% 20,756,870 2013 273,116,704 a 20,756,870 7.6% 2014 273,116,704 a 7.6% 2015 20,756,870 273,116,704 a 2011 value. Assumed equal to Source: van Haaren et al. (2010) Series - Consistency Uncertainty and Time 2 Monte Carlo analysis was performed to determine the level of uncertainty surrounding the estimates A n Approach emissions and N O emissions from the incineration of waste (given the very low emissions for CH , no of CO 2 2 4 uncertainty estimate wa s derived). IPCC 2 analysis allows the specification of probability density Approach mirrors the calculation of the I functions for key variables within a computational structure that nventory estimate. ration variables (i.e., plastics, synthetic rubber, and textiles Uncertainty estimates and distributions for waste gene generation) were obtained through a conversation with one of the authors of the Municipal Solid Waste in the e not available directly from the United States reports. Statistical analyses or expert judgments of uncertainty wer information sources for the other variables; thus, uncertainty estimates for these variables were determined using assumptions based on source category knowledge and the known uncertainty estimates for the waste generation variables. The uncertainties in the waste incineration emission estimates arise from both the assumptions applied to the data and from the quality of the data. Key factors include MSW incineration rate; fraction oxidized; missing data on ; average C content of waste components; assumptions on the synthetic/biogenic C ratio; and waste composition O emissions. The highest levels of uncertainty surround the variables that are combustion conditions affecting N 2 based on assumptions (e.g., percent of clothing and footwear composed of synthetic rubber); the lowest levels of uncertainty surround variables that were determined by quantitative measurements (e.g., combustion efficiency, C content of C black). Approach 2 quantitative uncertainty analy sis are summarized in Table 3 - 27 . Waste incineration The results of the CO dence level. emissions in 201 5 were estimated to be between 9.6 and 12.1 MMT CO Eq. at a 95 percent confi 2 2 This indicates a range of 10 percent below to 13 percent above the 2015 emission estimate of 10.7 MMT CO Eq. 2 Also at a 95 percent confidence level, waste incineration N O emissions in 201 5 were estimated to be between 0.2 2 emission estimate and 1.3 MMT CO 2015 Eq. This indicates a range of 51 percent below to 330 percent above the 2 MMT CO 0.3 of Eq. 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 54 3

169 2 Quantitative Uncertainty Estimates for CO Table 27 : Approach - O from the and N 3 2 2 Incineration of Waste (MMT CO Eq. and Percent) 2 a 201 5 Emission Estimate Uncertainty Range Relative to Emission Estimate Eq.) Source Gas (MMT CO (%) (MMT CO Eq.) 2 2 Lower Upper Upper Lower Bound Bound Bound Bound Incineration of Waste CO 10.7 9.6 12.1 - 10% +13% 2 Incineration of Waste N 51% O 0.3 0.2 1.3 - +330% 2 a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval. Methodological recalculations were applied to the entire time series to ensure time - series consistency from 1990 5 as discussed below through 201 . Details on the emission trends through time are described in more detail in the Methodology section, above. QA/QC and Verification - specific Q uality Assurance/Quality Control plan was implemented for incineration of waste. This effort A source included a general ( Tier 1 ) analysis, as well as portions of a category - specific ( Tier 2 ) analysis. The Tier 2 procedures that were implemented involved checks specifically focusing on the activi ty data and specifically focused on the emission factor and activity data sources and methodology used for estimating emissions from incineration of waste. Trends across the time series were analyzed to determine whether any corrective actions were needed. Actions were taken to streamline the activity data throughout the calculations on incineration of waste. Recalculations Discussion For the current Inventory, emission estimates for 2014 have been updated based on Advancing Sustainable Materials Management : 2014 Fact Sheet ( EPA 2016). The data used to calculate the percent incineration was not Bio updated in the current Inventory. C ycle has not released a new State of Garbage in America Report since 2010 (with 2008 data), which used to be a semi - cation which publishes the results of the nation - wide MSW annual publi survey. The results of the survey have been published in Shin (2014). This provided updated incineration data for 2011, so the generation and incineration data for 2012 through 2015 are assumed equi valent to the 2011 values. The data for 2009 and 2010 were based on interpolations between 2008 and 2011. Advancing Sustainable Materials Management: A transcription error in 2013 plastics production data from EPA’s s in Material Generation, Recycling and Disposal in the United States Facts and Figures 2013: Assessing Trend dentified and corrected. The amount of HDPE discarded in 2013 was misreported and the value (EPA 2015) was i has been updated. This update results in updated emission estimate for the CO from Plastics for 2013. 2 Previously, the carbon content for synthetic fiber was assumed to be the weighted average of carbon contents of four fiber types based on 1999 fiber production data. This methodology has been (polyester, nylon, olefin, and acrylic) updated. A weighted average for the carbon content of synthetic fibers based on production data from 1990 through 2015 was developed for each year based on the amount of fiber produced. For each year, the weighted average p the amount of carbon emitted. This methodology update affects the synthetic carbon content was used to develo fiber CO estimates. 2 Planned Improvements The availability of facility - level waste incineration data through EPA’s Greenhouse Gas Reporting Program lp better characterize waste incineration operations in the United States. This will be examined to he GHGRP ) ( characterization could include future improvements as to the operations involved in waste incineration fo r energy, or. Additional examinations will be necessary as, unlike whether in the power generation sector or the industrial sect 55 - 3 Energy

170 151 the reporting requirements for this chapter under the UNFCCC reporting guidelines, level waste some facility - incineration emissions reported under EPA’s GHGRP may also include industrial process emi ssions. 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 from EPA’s GHGRP that would be useful to improve the emission estimates for the waste examining data - 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 CO emissions from the biomass component of waste are separated in 2 - the facility level reported data, and on maintaining consistency with national waste generation and fate statistics implementing improvements and currently used to estimate total, national U.S. greenhouse gas emissions. In integration of data from EPA’s GHGRP, the latest guidance from the IPCC on the use of facility - level data in 152 national inventories will be relied upon. GHGRP data is available for MSW combustors, which contains the CO , CH , and N information on O emissions from MSW combustion, plus the fraction of the emissions that are 4 2 2 biogenic. To calculate biogenic versus total CO emissions, a default biogenic fraction of 0.6 is used. The biogenic 2 rent input data and assumptions to verify the current MSW emission fraction will be calculated using the cur 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 Bio C ycle (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 Advancing 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 t he Environmental Research & Education Foundat ion (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 m ade to be consistent with calculations in other waste categories including landfilling and composting. to improve the transparency in the current reporting of waste Additional improvements will be conducted incineration. ste incineration is included within the overall calculations for the Currently, hazardous industrial wa Waste incineration activities that do not C mitted from N on - E nergy U ses of F ossil F uels source category. E arbon s are not included in this analysis and include energy recovery will be examined. Synthetic fibers within scrap tire 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 wi ll also be Fiber Economics Bureau explored. 3.4 Coal Mining (IPCC Source Category 1B1a) Three types of coal mining - related activities release CH to the atmosphere: underground mining, surface mining, 4 and post mining (i.e., coal - handling) activities. While surface mines account for the majority of U.S. coal - production, u nderground coal mines contribute the largest share of CH ) emissions (see Table 3 - 29 and Table 3 - 30 4 content of coal in the deeper unde rground coal seams. In 2015, 305 underground coal mines due to the higher CH 4 and 529 surface mines were operating in the United States. In recent years the total number of active coal mines in the United States has declined. In 2015, the United States was the second largest co al producer in the world (812 MMT), after China (3,527 MMT) and followed by India (691 MMT) (IEA 2016). 151 /sbsta/eng/09.pdf See < http://unfccc.int/resource/docs/2006 >. 152 nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf - http://www.ipcc See < >. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 56 3

171 Table 3 28 : Coal Production (kt) - Surface Total Underground Year Number of Mines Number of Mines Number of Mines Production Production Production 1,683 384,244 1,656 3,339 931,052 1990 546,808 691,448 2005 789 586 1,398 1,025,846 334,398 2011 508 313,529 788 684,807 1,296 998,337 610,307 2012 310,608 719 488 1,207 920,915 2013 395 309,546 637 581,270 1,032 890,815 2014 345 321,783 613 583,974 958 905,757 305 278,342 529 534,127 834 812,469 2015 Underground mines liberate CH from ventilation systems and from degasification systems. Ventilation systems 4 pump air through the mine workings to dilute noxious gases and ensure worker safety; these systems can exhaust significant amounts of CH to the atmosphere in low concentrations. Degasification systems are wells drilled from 4 the surfac e or boreholes drilled inside the mine that remove large, often highly concentrated volumes of CH before, 4 Some mines recover and use CH generated from during, or after mining. ventilation and degasification systems, 4 thereby reducing emissions to the atmo sphere. as the overburden is removed and the coal is exposed to the atmosphere. CH Surface coal mines liberate CH 4 4 emissions are normally a function of coal rank (a classification related to the percentage of carbon in the coal) and ines typically produce lower - rank coals and remove less than 250 feet of overburden, so their depth. Surface coal m level of emissions is much lower than from underground mines. - mining activities, as the coal is processed, transported, and stored for use. In addition, CH is released during post 4 emissions in 2015 were estimated to be 2,436 kt (60.9 MMT CO Eq.), a decline of 37 percent since 1990 Total CH 2 4 Table (see - 29 a nd Table 3 - 30 ). Of this amount, underground mines accounted for approximately 73 percent, 3 surface mines accounted for 14 percent, and post - mining emiss ions accounted for 13 percent. Table 3 - 29 : CH Emissions from Coal Mining (MMT CO Eq.) 2 4 1990 Activity 2011 2012 2013 2014 2015 2005 Underground (UG) Mining 74.2 42.0 50.2 47.3 46.2 46.4 44. 6 Liberated 80.8 59.7 71.0 65.8 6 4.5 63.1 60.5 Recovered & Used 6 ( .6) (1 7.7 ) ( 20.8 ) (1 8.5 ) (18.3) (16.7) (15.9 ) 10.8 1 1.9 Surface Mining 11.6 10.3 9.7 9.6 8.7 Mining (UG) 9.2 7. 6 Post 6.9 6 .7 6.6 6.7 5.8 - 5 Mining (Surface) 2.3 2. 6 2. - 2.2 2.1 2.1 1.9 Post Total 96.5 64.1 71.2 66.5 64.6 64.8 60.9 Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values. 3 30 : CH Emissions from Coal Mining (kt) Table - 4 1990 2005 2011 Activity 2012 2013 2014 2015 UG Mining 2,968 1,682 2,008 1, 891 1,849 1,854 1,783 3,234 2,390 2,8 39 2,6 31 2,580 2,523 2,421 Liberated (266) (708) (8 ) ( 740 ) (730) (668) (638 ) Recovered & Used 31 430 475 465 Surface Mining 388 386 347 410 Post - Mining (UG) 368 306 27 6 26 8 263 270 231 Mining (Surface) 93 103 10 1 - 8 9 84 84 75 Post Total 3,860 2,5 65 2,849 2,658 2,584 2,593 2,436 Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values. 57 - 3 Energy

172 Methodology : emissions from coal mining consists of two steps The methodology for estimating CH 4 Estimate emissions from underground mines. These emissions have two sources: ventilation systems and • They are estimated using mine - specific data, then summed to determine total CH degasification systems. 4 liberated. The CH mate of net recovered and used is then subtracted from this total, resulting in an esti 4 . emissions to the atmosphere • Estimate CH the methodology for emissions from surface mines and post - mining activities. Unlike 4 underground mines, which uses mine - specific data, the methodology for estimating emissions from surface specific mines and pos t - mining activities consists of multiplying basin - specific coal production by basin - gas content and an emission factor. Emitted from Underground Mines Step 1: Liberated and CH Estimate CH 4 4 from ventilation systems and from degasification systems. Some mines recover Underground mines generate CH 4 the generated CH , thereby reducing emissions to the atmosphere. Total CH and use emitted from underground 4 4 mines equals the CH liberated from ventilation systems, plus the CH liberated from degasification systems, minus 4 4 the CH recovered and used. 4 Step 1.1: Estimate CH Liberated from Ventilation Systems 4 liberated from ventilation systems, EPA uses data collected through its Greenhouse Gas Reporting To estimate CH 4 153 (subpart FF, “Underground Coal Mines”), data provided by the U.S. Mine Safety and Health Program (G HGRP) gas specific level (e.g., state other sources on a site Administration (MSHA), and occasionally data collected from - production ). Since 2011, the natio n’s “gassiest” underground coal mines — those that liberate more than data bases per year (about 14,700 MT 36,500,000 actual cubic feet of CH CO E q.) — have been required to report to EPA’s 2 4 154 6 ). GHGRP (EPA 201 Mines that report to the GHGRP must report quarterly measurements of CH emissions 4 to EPA; they have the option of recording their own measurements, or using the from ventilation systems spections of all mines in the United S tates measurements taken by MSHA as part of that agency’s quarterly safety in 155 concentrations. with de tectable CH 4 Since 2013, ventilation emission estimates have been calculated based on both GHGRP data submitted by The underground mines, and on quarterly measurement data obtained directly from MSHA for the remaining mines. the average daily emissions rate for the reporting year quarter. quarterly mea surements are used to determine Because not all mines report under the GHGRP, the emissions of the mines that do not report must be calculated quality assurance tool for validating GHGRP data. using MSHA data. The MSHA data also serves as a Estimate CH Step 1.2: Liberat ed from Degasification Systems 4 Particularly gassy underground mines also use degasification systems (e.g., wells or boreholes) to remove CH 4 before, during, or after mining. Th is CH six can then be collected for use or vented to the atmosphere. Twenty - 4 mines used degasification systems in 2015, and the CH removed through these systems was reported to EPA’s 4 GHGRP (EPA 2016). Based on the weekly measurements reported to EPA’s GHG RP, degasification data summaries for each mine were added together to estimate the CH liberated from degasification systems. Sixteen of 4 153 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). 154 3 underground coal mines reported to Underground coal mines report to EPA under Subpart FF of the GH GRP. In 201 5 , 12 the program. 155 . Readings below readings with concentrations of greater than 50 ppm (parts per million) CH MSHA records coal mine CH 4 4 detectable. - this threshold are considered non - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 58 3

173 the 26 mines with degasification systems had operational CH projects (see step 1.3 below) , and recovery and use 4 156 vented CH GHGRP from degasification systems to the atmosphere. reports show the remaining mines ten 4 - mining wells are attributed to the mine as emissions in the year in Degasification volumes for the life of any pre 157 EPA’s which the well is mined through. GHGRP does not require gas production from virgin coal seams (coalbed methane) to be reported by coal mines under subpart FF. Most pre - mining wells drilled from the surface program (subpart W, are considered coalbed methane wells and are reported under another subpart of the As a result, for the 10 mines with degasification systems that include pre - “Petroleum and Natural Gas Systems”). mining wells, GHGRP information was supplemented with historical data from state gas well production databases ( GSA 2016; WVGES 2016), as well as with mine - specific information regarding the dates on which DMME 2016; - mining wells are mined through (JWR 2010; El Paso 2009). the pre Degasification information reported to EPA’s GHGRP by underground coal mines was the primary sou rce of data used to develop estimates of CH liberated from degasification systems. Data reported to EPA’s GHGRP were used 4 to estimate CH liberated from degasification systems at 21 of the 26 mines that employed degasification systems in 4 2015. For the oth er five mines (all with pre - mining wells from which CH along was recovered), GHGRP data — 4 with supplemental information from state gas production — databases (DMME 2016; GSA 2016; WVGES 2016) or one mine, due to a lack of mine - provided liberated from degasification systems. F were used to estimate CH 4 CH liberated was based on both an information used in prior years and a GHGRP reporting discrepancy, the 4 - provided CH recovery and use rates and state gas sales records . estimate from historical mine 4 Step 1.3: Estimate CH or Recovered from Ventilation and Degasification Systems , and Utilized 4 Destroyed ) (Emissions Avoided recovery and use projects mines CH Sixteen had in place in 2015. Fourteen of these mines sold the recovered CH 4 4 to a pipeline , including one that also used CH recovered to fuel a thermal coal dryer. In addition, one mine used 4 CH for electrical power generation , and one used recovered CH . to heat mine ventilation air 4 4 recovered from the 2015; for those mines, estimates of CH Ten of the 16 mines deployed degasification systems in 4 systems were exclusively based on GHGRP data. Based on weekly measurements, the GHGRP degasification destruction data summaries for each mine were added together to estimate the CH recovered and u sed from 4 degasification systems. All 10 mines with degasification systems used pre - mining wells as part of those systems, but only four of them - mining wells in 2015. GHGRP and supplemental data were used to estimate CH intersected pre recovered and used 4 at two of these four mines; supplemental data alone (GSA 2016) were used for the other two mines, which reported to EPA’s GHGRP as a single entity. Supplemental information was used for these four mines because estimating recovery and use from pre CH mini ng wells requires additional data (not reported under subpart FF of EPA’s - 4 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 . GHGRP information was not used to estimate CH d at two mines. At one of these mines, a recover ed and use 4 portion of reported CH vented was applied to an ongoing mine air heating project. Because of a lack of mine - 4 provided information used in prior years and a GHGRP reporting discrepancy, the 2015 recovered and used CH 4 - mining wells at the other mine was based on an estimate from historical mine - provided from pre recovery and CH 4 use rates. Emissions recovered and used from the active mine degasification system were estimated based on a state gas production data information system. In 201 5 , one mine destroyed a portion of its CH emissions from ventilation systems using thermal oxidation 4 technology. The amount of CH recovered and destroyed by the project was determined through publicly - available 4 emission reduction project information ( ACR 2016 ). 156 Several of the mines venting CH from degasification systems use a small portion the gas to fuel gob well blowers in remote 4 locations where electricity is not available. However, this CH use is not considered to be a formal recovery and use project. 4 157 working face intersects the borehole or well when coal mining development or the well is “mined through” . A 59 - 3 Energy

174 Step 2: Emitted from Surface Mines and Post - Mining Activities Estimate CH 4 mining are - Mine emissions from surface coal mines or for post - specific data not available for estimating CH 4 activities. For surface mines, basin - specific coal production obtained from the Energy Information Admini stration’s (EIA 2016) was multiplied by basin - specific CH Annual Coal Report contents (EPA 1996, 2005) and a 150 percent 4 to CH emission factor (King 1994; Saghafi from over - and under - burden) (to account for estimate CH emissions 4 4 mining - - activities, basin 2013). For post specific coal production was multiplied by b asin - specific gas contents and a mid - emission factor for CH range 32.5 percent desorption during coal transportation and storage (Creedy 1993). 4 Basin specific in situ gas content data were comp iled from AAPG (1984) and USBM (1986). - - Series Consistency Uncertainty and Time A quantitative uncertainty analysis was conducted for the coal mining source category using the IPCC - recommended Approach 2 uncertainty estimation methodology. Because emission estimates from underground ventilation systems were based on actual measurement data from EPA’s GHGRP or from MSHA, uncertainty is relatively low. A degree of imprecision was introduced because the ventilation air measurements used were not contin uous but rather quarterly instantaneous readings that were used to determine the average daily emissions rate for the quarter. Additionally, the measurement equipment used can be expected to have resulted in an average of 10 percent overestimation of annua l CH emissions (Mutmansky & Wang 2000). GHGRP data were used for a 4 significant number of the mines that reported their own measurements to the program beginning in 2013; however, the equipment uncertainty is applied to both GHGRP and MSHA data. Estimates of CH because of the recovered by degasification systems are relatively certain for utilized CH 4 4 availability of GHGRP data and gas sales information. Many of the recovery estimates use data on wells within 100 feet of a mined area. However, uncertainty exists concerning the radius of influence of each well. The number of wells counted, and thus the avoided emissions, may vary if the drainage area is found to be larger or smaller than estimated. monitoring of mines that re EPA’s GHGRP requires weekly CH port degasification systems, and continuous CH 4 4 monitoring is required site. Since 2012, GHGRP data have been used to estimate CH - on or off - for utilized CH 4 4 emissions from vented degasification wells, reducing the uncertainty associated with prior MSHA e stimates used for this subsource. Beginning in 2013, GHGRP data were also used for determining CH recovery and use at mines 4 without publicly available gas usage or sales records, which has reduced the uncertainty from previous estimation based on information from coal industry contacts. methods that were In 2015 a small level of uncertainty was introduced with using estimated rather than measured values of recovered methane from two of the mines with degasification systems. An increased level of uncertain ty was applied to these two subsources, but the change had little impact on the overall uncertainty. Surface mining and post - mining emissions are associated with considerably more uncertainty than underground accurate emission factors from field measurements. However, since mines, because of the difficulty in developing underground emissions constitute the majority of total coal mining emissions, the uncertainty associated with underground emissions is the primary factor that determines overall uncertainty. The results of the Approach 2 Table 3 - 31 . Coal mining CH emissions in 2015 were estimated quantitative uncertainty analysis are summarized in 4 to be b etween 53.3 and 70.8 MMT CO Eq. at a 95 percent confidence level. This indicates a range of 12.5 percent 2 below to 16.2 percent above the 2015 emission estimate of 60.9 MMT CO Eq . 2 titative Uncertainty Estimates for CH Emissions from Coal 31 Table 3 - : Approach 2 Quan 4 Mining (MMT CO Eq. and Percent) 2 a Uncertainty Range Relative to Emission Estimate 2015 Emission Estimate Source Gas (MMT CO Eq.) (MMT CO Eq.) (%) 2 2 Lower Upper Upper Lower Bound Bound Bound Bound +16.2% 70.8 12.5% - Coal mining CH 53.3 60.9 4 a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 60 3

175 Methodological recalculations were applied to the entire time 2015 . series to ensure consistency from 1990 through m ethodology section. Details on the emission trends through time are described in more detail in the Recalculations Discussion For the current Inventory, revisions were made to the 2013 and 2014 undergro und liberated and recovered emissions. From 2013 to 2015, recovered emissions reported to EPA’s GHGRP for a large mine located in Virginia were inaccurate and could not be used. For the 1990 through 2013 and 1990 through 2014 Inventories, EPA estimated rec year historical average. In preparing the current overed emissions for this mine based on a five - Inventory, EPA was able to utilize the Virginia Division of Gas and Oil Data Information System (DGO DIS) to estimate recovered degasification emissions for th e Virginia mine based on published well production. The well production data was more accurate than the reported values in 2013, 2014, and 2015; thus 2013 and 2014 were s until the GHGRP revised using the 2015 methodology. The DGO DIS will continue to be used in future year reported values can be verified for this mine. 3.5 Abandoned Underground Coal Mines (IPCC Source Category 1B1a) Underground coal mines contribute the largest share of coal mine methane (CMM) emissions, with active g source of underground emissions. However, mines also continue to release CH after underground mines the leadin 4 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 buildu p of CH that may find its 4 way to surface structures through overburden fractures. As work stops within the mines, CH liberation decreases 4 but it does not stop completely. Following an initial decline, abandoned mines can liberate CH at a near - steady rate 4 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 issions can occur when CH em migrates to the surface through cracks and fissures in the strata overlying the coal 4 mine. The following factors influence abandoned mine emissions: • Time since abandonment; • Gas content and adsorption characteristics of coal; • CH flow capacity of the mine; 4 • Mine flooding; • Presence of vent holes; and • Mine seals. Eq. from 1990 through 201 Annual gross abandoned mine CH , emissions ranged from 7.2 to 10.8 MMT CO 5 2 4 from year to year. Fluctuations were due varying, in general, by less than 1 percent to approximately 19 percent mainly to the number of mines closed during a given year as well as the magnitude of the emissions from those Gross abandoned mine emissions peaked in 1996 (10.8 MMT CO mines when active. Eq.) due to the large numb er 2 158 gassy mine of closures from 1994 to 1996 (7 2 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, t here have been 2015 fewer than twelve gassy mine closures each year. There were six gassy mine closures in 2015 . In , gross 33 ). slightly from 8.7 to 9.0 MMT CO Gross Eq. (see Table 3 - 32 and Table 3 - increased abandoned mine emissions 2 of recovered and used at 40 mines, resulting in net emissions in 2015 emissions are reduced by CH 6.4 MMT CO 2 4 Eq. 158 ). per day (100 mcfd A mine is considered a “gassy” mine if it emits more than 100 thousand cubic feet of CH 4 61 - 3 Energy

176 Table 3 : CH - Emissions from Abandoned Coal Mines (MMT CO 32 Eq.) 2 4 2012 Activity 2011 2013 2014 2015 1990 2005 7.2 8.4 9.3 8.9 8.8 8.7 9.0 Abandoned Underground Mines 1. 8 2. 9 + 2. Recovered & Used 2.6 2.4 2.6 7 Total 7.2 6.6 6.4 6.2 6.2 6.3 6.4 + Does not exceed 0.05 MMT CO Eq. 2 Note: Totals may not sum due to independent rounding. Table 3 - 33 : CH Emissions from Abandoned Coal Mines (kt ) 4 1990 2005 Activity 2012 2013 2014 2015 2011 Abandoned Underground Mines 288 334 373 358 353 350 359 97 Recovered & Used 102 104 + 70 116 109 Total 264 257 249 249 253 256 288 + Does not exceed 0.5 kt Note: Totals may not sum due to independent rounding. Methodology emissions from an abandoned coal mine requires predicting the emissions of a mine from the time Estimating CH 4 of abandonment through the inventory year of interest. The flow of CH from the coal to the mine void is primarily 4 dependent on the mine’s emissions when active and the extent to which the mine is flooded or sealed. The CH 4 emission rate before abandonment reflects the gas content of the coal, the rate of coal mining, and th e flow capacity - free conventional gas well reflects the gas content of of the mine in much the same way as the initial rate of a water the producing formation and the flow capacity of the well. A well or a mine which produces gas from a coal seam and the s urrounding 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 ion of a reservoir is commonly modeled by mathematical equations and mapped imposed on the reservoir. The deplet 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 CH emissions over time for a given abandoned mine, it is necessary to apply a decline function, 4 initiated upon abandon ment, 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 CH leaves the system, the 4 reservoir pressure (Pr) declines as described by th e 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 flow rate is determined by the laws of gas flow through porous media, interest (atmospheric to 30 psia). The CH 4 - time equation can be generated that can be used to predict future emission s. This such as Darcy’s Law. A rate decline through time is hyperbolic in nature and can be empirically expressed as: (−1/푏) 푞 = 푞 (1+ 푏퐷 푡) 푖 푖 where, q = Gas flow rate at time t in million cubic feet per day (mmcfd) q = Initial gas flow rate at time zero (t ), mmcfd o i b = The hyperbolic exponent, dimensionless D = Initial decline rate, 1/y ea r i = (years) Elapsed time from t t o This equation is applied to mines of various initial emission rates that have similar initial pressures, permeability and 4). adsorption isotherms (EPA 200 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 62 3

177 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 U.S. mines prone to flooding will become completely flooded within eight years and therefore assumed that most will no longer have any measurable CH emissions. Based on this assumption, an average decline rate for flooded 4 mines was established by fitting a decline curve t o 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 fitting algorithm, emissions dat a were matched to the exponential the five basins. Using a least squares, curve - specific equations as was done with the vented, equation shown below. There was not enough data to establish basin - non - flooding mines (EPA 2004). (− 퐷푡) 푒 푞 = 푞 푖 where, q = Gas flow rate at time t in mmcfd Initial gas flow rate at time zero (t = q ), mmcfd o i D = Decline rate, 1/y ea r t = Elapsed time from t (years) o Seals have an inhibiting effect on the rate of flow of CH into the atmosphere compared to the flow rate that would 4 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 in itial 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 × (1 ealed mine / emission [initial emissions from s 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 min es (EPA 2004). For active coal mines, those mines producing over 100 thousand cubic feet per day (mcfd) account for about 98 percent of all CH emissions. This same relationship is assumed for abandoned mines. It was determined that the 4 closed after 1972 produced emissions greater than 100 mcfd when active. Further, the status 524 abandoned mines of 302 of the 524 mines (or 58 percent) is known to be either: 1) vented to the atmosphere; 2) sealed to some degree (either earthen or concrete seals); or, 3) flo oded (enough to inhibit CH flow to the atmosphere). The remaining 42 4 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 2015, grouped by Class according to Post - Abandonment State Total Total Mines Unknown Sealed Basin Known Vented Flooded 118 Central Appl. 26 52 40 143 261 Illinois 34 3 14 51 30 81 Northern Appl. 46 22 16 84 39 123 16 Warrior Basin 0 0 16 16 0 Western Basins 28 3 2 33 10 43 Total 1 48 5 4 100 302 2 22 5 24 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. 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 1972 mines are governed by the same physical, geologic, and hydrologic constraints that apply is assumed that pre - 1971 mines; thus, their emissions may be characterized by the same decline curves. - to post s. In emissions from coal mining came from seventeen counties in seven state During the 1970s, 78 percent of CH 4 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 63 - 3 Energy

178 decade) and applied to the other specific decline curve five states with gassy mine closures. As a result, basin - equations were applied to the 145 gassy coal mines estimated to have closed between 1920 and 1971 in the United - c, initial emission rates were used based on average States, representing 78 percent of the emissions. State specifi coal mine CH emissions rates during the 1970s (EPA 2004). 4 Abandoned mine emission estimates are based on all closed mines known to have active mine CH ventilation 4 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 abandon ment, 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 rates used reflect ventilation emissions only for pre - 1990 available for years prior to 1990, thus the initial emission closures. CH vented to determine the total CH liberation degasification amounts were added to the quantity of CH 4 4 4 rate for all mines that closed between 1992 and 2015. Since the sample of gassy m ines is assumed to account for 78 - 1972 and 98 percent of the post - 1971 abandoned mine emissions, the modeled results were percent of the pre multiplied by 1.22 and 1.02 to account for all U.S. abandoned mine emissions. From 1993 through 2015, emission tot emissions als were downwardly adjusted to reflect abandoned mine CH 4 avoided from those mines. The Inventory totals were not adjusted for abandoned mine reductions from 1990 through 1992 because no data was reported for abandoned coal mining CH recovery p rojects during that time. 4 Uncertainty and Time Series Consistency - A quantitative uncertainty analysis was conducted to estimate the uncertainty surrounding the estimates of emissions from abandoned underground coal mines. The uncertainty analysis describ ed 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, emiss ions from this source category are likely to fall. As discussed above, the parameters for which values must be estimated for each mine in order to predict its decline curve are: 1) the coal's adsorption isotherm; 2) CH flow capacity as expressed by perme ability; and 3) pressure at 4 abandonment. Because these parameters are not available for each mine, a methodological approach to estimating emissions was used that generates a probability distribution of potential outcomes based on the most likely value and the probable range of values for each parameter. The range of values is not meant to capture the extreme values, but rather values that represent the highest and lowest quartile of the cumulative probability density function of each parameter. Once the lo w, mid, and high values are selected, they are applied to a probability density function. The results of the Approach 2 quantitative uncertainty analysis are summarized in bandoned 3 - 35 . A nnual a Table Eq. at a 95 percent coal mine CH emissions in 201 5 were estimated to be between 5.2 and 7.9 MMT CO 2 4 18 5 4 6.4 emission estimate of confidence level. This indicates a range of percent above the 201 percent below to 2 specific data is CO Eq. One of the reasons for the relatively narrow range is that mine - MMT available for use in the 2 he largest degree of methodology for mines closed after 1972 . Emissions from mines closed prior to 1972 have t - specific CH un certainty because no mine liberation rates exist. 4 Table 3 - 35 : Approach 2 Quantitative Uncertainty Estimates for CH Emissions from 4 CO Abandoned Underground Coal Mines ( MMT Eq. and Percent) 2 a 5 Uncertainty Range Relative to Emission Estimate Emission Estimate 201 rce Gas Sou (MMT CO Eq.) (MMT CO (%) Eq.) 2 2 Lower Upper Upper Lower Bound Bound Bound Bound Abandoned Underground 7.9 +24% 18% - 5.2 CH 6.4 4 Coal Mines a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval. series consistency from 1990 - series to ensure time Methodological recalculations were applied to the entire time trends through time are described in more detail in the Methodology section, through 2015. Details on the emission above. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 64 3

179 3.6 Petroleum Systems (IPCC Source Category 1B2a) Methane emissions from petroleum systems are primarily associated with onshore and offshore crude oil production, is released to the atmosphere as fugitive transportati on, and refining operations. During these activities, CH 4 emissions, vented emissions, emissions from operational upsets, and emissions from fuel combustion. Fugitive and vented CO emissions from petroleum systems are primarily associated with crude oil production and refining 2 emissions from petroleum systems in 2015 operations but are negligible in transportation operations. Total CH 4 MMT CO Eq. were 39.9 Eq. (1, 595 kt). Total CO em issions from petroleum systems in 2015 were 3.6 MMT CO 2 2 2 (3,567 kt). Production Field Operations. Production field operations account for approximately 98 percent of total CH 4 emissions from petroleum systems. The predominant sources of emissions from p roduction field operations are pneumatic controllers, offshore oil platforms, associated gas venting and flaring, gas engines, chemical injection hydraulically fractured oil well completions, and fugitives from oil wellheads. These source pumps, oil tanks, s alone emit around 95 percent of the production field operations emissions. The remaining emissions are distributed among around 20 additional activities. emissions from production field operations have decreased by Since 1990, CH 29 percent, due to a l arge decrease in 4 associated gas venting . Production segment methane emissions have decreased by around 8 percent from 2014 levels, primarily due to decreases in emissions from associated gas venting and flaring, and in the number of hydraulically fractured oil well s that were complet ed in 2015 compared to 2014 . Vented CO associated with production field operations account for approximately 99 percent of the total CO 2 2 emissions from production field operations, while fugitive and process upsets together acc ount for approximately 1 percent of the emissions. The principal sources of CO emissions are oil tanks, pneumatic controllers, chemical 2 injection pumps, and offshore oil platforms. These four sources together account for slightly over 97 percent of the no n emissions from production field operations, while the remaining 3 percent of the emissions is - combustion CO 2 distributed among around 20 additional activities. Due to the activity data source for CO from flaring, it is not 2 possible to develop separate e stimates for flaring occurring in natural gas production and flaring occurring in oil production. Total CO emissions from flaring for both natural gas and oil were 18.0 MMT CO in 2015 and are 2 2 included in the Natural Gas Systems estimates. Crude Oil Tr ansportation. Crude oil transportation activities account for less than 1 percent of total CH emissions 4 from the oil industry. Venting emissions, including from tanks, truck loading, rail loading, and marine vessel loading operations account for 89 percen t of CH emissions from crude oil transportation. Fugitive emissions, almost 4 entirely from floating roof tanks, account for approximately 11 percent of CH emissions from crude oil 4 transportation. Since 1990, CH emissions from transportation have increased by 28 percent. However, because emissions from 4 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 transporta tion in 2015 increased by approximately 2 percent from 2014 levels. Crude Oil Refining. Crude oil refining processes and systems account for approximately 2 percent of total CH 4 . This low share is due to the fact that most o f the CH emissions from the oil industry in crude oil is removed or 4 escapes before the crude oil is delivered to the refineries. There is an insignificant amount of CH in all refined 4 0 around 5 products. Within refineries, incomplete combustion accounts for percent of the CH while emissions, 4 vented and fugitive emissions account for approximately 34 and 15 percent, respectively. Flaring accounts for 82 percent of combustion CH emissions. Refinery system blowdowns for maintenance and process vents are the 4 primary venting contributors (97 percent). Most of the fugitive CH emissions from refineries are from equipment 4 leaks and storage tanks (87 percent). emissions from re Methane fining of crude oil have increased by approximately 7 percent since 1990; however, . Since similar to the transportation subcategory, this increase has had little effect on the overall emissions of CH 4 ated between 24 and 28 kt. emissions from crude oil refining have fluctu 1990, CH 4 65 - 3 Energy

180 Flare emissions from crude oil refining accounts for slightly more than 77 percent of the total CO emissions in 2 petroleum systems. Refinery CO emissions de creased by approximately 7 percent from 1990 to 2015. 2 3 - : CH Table Emissions from Petroleum Systems (MMT CO 36 Eq.) 4 2 1990 2005 2011 2012 2013 2014 2015 Activity Production Pneumatic controller venting 19.1 17.1 16.1 14.8 17.8 18.5 18.6 Offshore platforms 4.7 4.7 4.7 4.6 4.7 4.7 5.3 Associated gas venting and flaring 17.1 14.4 16.2 14.7 9.0 6.0 3.7 Tanks 1.6 6.2 2.0 1.2 1.4 1.9 2.0 Gas Engines 2.1 1.7 1.9 2.1 2.1 2.3 2.3 Other Sources 4.9 5.3 7.1 8.0 8.3 8.9 7.7 Production Total 43.6 . 45 .2 47.2 45.6 42.2 39.0 54 7 Crude Oil Transportation 0.1 0.1 0.2 0.2 0.2 0.2 0.2 Refining 0.7 0.7 0.6 0.6 0.6 0.6 0.7 Total 5 5 . 5 46 .0 48.0 46.4 44.5 43.0 39.9 Note : Totals may not sum due to independent rounding. 3 - 37 : CH Emissions from Petroleum Systems (kt) Table 4 Activity 1990 2005 2011 2012 2013 2014 2015 Production Field Operations Pneumatic controller venting 76 4 683 643 593 714 740 745 Offshore platforms 188 211 185 188 188 188 188 Associated gas venting and flaring 685 647 587 361 239 148 576 Tank venting 74 82 48 55 64 80 248 Gas Engines 70 78 85 86 90 90 82 Other Sources 334 285 213 357 309 195 319 Production Field Operations 2,18 8 1,808 1,888 1,689 1,825 1,746 1,561 7 5 5 6 7 8 8 Crude Oil Transportation Refining 24 27 28 27 26 24 26 Total 2, 218 1,840 1,922 1,858 1,778 1,721 1,595 : Totals may not sum due to independent rounding. Note - Table : CO 3 Emissions from Petroleum Systems (MMT CO ) 38 2 2 Activity 1990 2005 2011 2012 2013 2014 2015 0. Production 0. 4 0. 3 0. 4 5 0. 6 0. 6 0. 6 Crude Refining 3.2 3.6 3.8 3.4 3.1 2.9 2.9 3.6 3.9 4 . 2 3 . 9 Total 3. 7 3. 6 3. 6 Note: Totals may not sum due to independent rounding. Table 3 - 39 : CO Emissions from Petroleum Systems (kt) 2 2012 Activity 1990 2005 2011 2013 2014 2015 Production 391 338 395 473 550 640 640 3,162 3,589 3,797 Crude Refining 3,404 3,143 2,927 2,927 3,567 3,567 Total 3, 553 3, 927 4,192 3,876 3,693 due to independent rounding. Note: Totals may not sum - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 66 3

181 Methodology The estimates of CH emissions from petroleum systems are largely based on GRI/EPA 1996, EPA 1999, 4 DrillingInfo, and GHGRP data (2010 through 2015). Petroleum Systems includes emission estimates for activities occurring in petroleum systems from the oil wellhead through crude oil refining, including activities for crude oil further production field operations, crude oil transportation activities, and refining operations. Annex 3.5 provides detail on the emissio n estimates for these activities. The estimates of CH emissions from petroleum systems do not 4 include emissions downstream of oil refineries because these emissions are considered to be negligible. 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 Researc h d), (EPA/GRI 1996a - Estimates of Methane Emissions from the U.S. Oil Industry Institute and EPA (EPA 1999), DrillingInfo (2015), consensus of industry peer review panels, Bureau of Ocean Energy Management (BOEM) (formerly Bureau of Ocean Energy Management , Regulation, and Enforcement [BOEMRE]) reports (BOEMRE 2004; BOEM 2011), analysis of BOEMRE data (EPA 2005; BOEMRE 2004), and the GHGRP (2010 through 2015). The emission factors for pneumatic controllers and chemical injection pumps were developed using GHGRP data for reporting year 2014. The emission factors for tanks, and associated gas venting and flaring were developed using GHGRP data for reporting year 2015. Emission factors for hydraulically fractured (HF) oil well completions (controlled and uncon trolled) were developed using DrillingInfo data analyzed for the 2015 NSPS OOOOa proposal (EPA 2015a). For offshore oil production, two emission factors were calculated using data col lected for all federal offshore platforms (EPA 2015b; BOEM 2014), one for oil platforms in shallow water, and one for oil platforms in deep water. For all sources other than associated gas venting and flaring , emission factors are held constant for the period 1990 through 2015 , and trends in emissions reflect changes in activit y levels . For associated gas venting and 2015 and the 2011 emission factors were applied flaring, year - specific emission factors were developed for 2011 - back to 1990 ities. Emission factors from EPA 1999 are used for all other production and transportation activ . References for activity data include DrillingInfo (2016), the Energy Information Administration annual and monthly reports (EIA 1990 through 2016), (EIA 1995 through 2016a, 2016b), Methane Emissions from the Natural Gas Institute and EPA (EPA/GRI 1996a Industry by the Gas Research d), Estimates of Methane Emissions from the U.S. - Oil Industry (EPA 1999), consensus of industry peer review panels, BOEMRE and BOEM reports (BOEMRE 2004; BOEM 2011), analysis of BOEMRE data (EPA 2005; BOEMRE 2004), the Oil & Gas Journal (OGJ 2016), the Interstate Oil and Gas Compact Commission (IOGCC 2012), the United States Army Corps of Engineers, (1995 through 2016), and the GHGRP (2010 through 2015). For many sources, complete activity data were not available for all ye ars 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 1996, and/o r GHGRP data. For floating roof tanks, the activity data were held constant from 1990 through 2015 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, suc h as from 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 activity data and are taken from BOEM number of platforms in deep water are used as d atasets (BOEM 2011a, b, c). A complete list of references for emission factors and activity data by emission source provided in Annex 3.5. is For petroleum refining activities, 2010 to 2015 emissions were directly ob tained from EPA’s GHGRP. All U.S. refineries have been required to report CH emissions for all major activities starting with emissions that and CO 2 4 occurred in 2010. The national total of these emissions for each activity was used for the 2010 to 2015 em issions. The national emission total for each activity w as divided by refinery feed rates for those inventory years to develop used to estimate national emissions for each refinery activity , which w as an average activity - specific emission facto r to 2009 based on national refinery feed rates for each year (EPA 2015c). from 1990 67 - 3 Energy

182 In this year’s Inventory, EPA has held constant the CO values from the previous Inventory (developed using the 2 to the CO estimates. See Planned methodology as described in this section) as it assesses improvements 2 The methodology for estimating CO emissions from petroleum systems includes calculation of Improvements. 2 nd vented, fugitive, and process upset emissions sources from 26 activities for crude oil production field operations a three activities from petroleum refining. Generally, emissions are estimated for each activity by multiplying CO 2 CH , as the CO emission factors by the corresponding content of gas relates to the CH . The content of gas data 4 2 4 production field operation are generally estimated by multiplying the CH s emission factors for CO emission 2 4 content and CH factors by a conversion factor, which is the ratio of CO content in produced associated gas. One 4 2 exception to this methodology are emission factors for offshore oil production (shallow and deep water), which were .e., flares, derived using data from BOEM (EPA 2015b; BOEM 2014). For the three petroleum refining activities (i asphalt blowing, and process vents); the CO emissions data for 2010 to 2014 were directly obtained from the 2 GHGRP. The 2010 to 2013 CO emissions data from GHGRP along with the refinery feed data for 2010 to 2013 2 were used to derive CO emiss ion factors (i.e., sum of activity emissions/sum of refinery feed) which were then 2 In this year’s Inventory, EPA has emissions for 1990 to 2009. applied to the annual refinery feed to estimate CO 2 developed using the methodology as described in this held constant the CO values from the previous Inventory ( 2 paragraph) as it assesses improvements to the CO estimates. See Planned Improvements. 2 Series Consistency - Uncertainty and Time The most recent uncertainty analysis for the petroleum systems emission est imates in the Inventory was conducted to 2009 Inventory that was released in 2011 for the 1990 has not yet been updated . Since the analysis was last , and to reflect improved data and cha nges in conducted, several of the methods used in the Inventory have changed industry practices and equipment. In addition, new studies and other data sources such as those discussed in the sections below offer improvement to understanding and quantifying the uncertainty of some emission source EPA is estimates. preparing a draft update to the uncertainty analysis conducted for the 2011 Inventory to reflect the new information and will seek stakeholder feedback on the draft analysis as part of the development of the next (i.e., ion, please see the Planned Improvements section. 1990 through 2016) Inventory. For more informat Table 3 - 40 below, To develop the values in EPA has applied the uncertainty percentage ranges calculated previously to the updated 201 5 emission estimates . To develop the uncertainty percentage ranges, EPA used the IPCC - recommended Approach 2 methodology (Monte Carlo Simulation technique) . The @RISK software model was used to quantify the uncertainty associated with the emission estimates using the 7 highest - emitting sources (“top 7 sources”) for the year 2009 The @RISK analysis provides for the specification of probability density function s for . key variables within a computational structure that mirrors the calculation of the Inventory estimate. The IPCC guidance no tes that in using this method, “ some uncertainties that are not addressed by statistical means may exist, ng from omissions or double counting, or other conceptual errors, or from incomplete including those arisi stimates developed from models.” understanding of the processes that may lead to inaccuracies in e As a result, the understanding of the uncertainty of emission estimates f or this category evolves and improves as the underlying methodologies and datasets improve. uncertainty ranges applied may not reflect the uncertainty associated with the Given the recent revisions, the . recently revised emission factors and activity data sources The results presented below provide with 95 percent certainty the range within which emissions from this source category are likely to fall for the year 2015 , based on the previously conducted uncertainty assessment using the hodology. The results of the Approach 2 quantitative uncertainty analysis are summarized recommended IPCC met 99.3 3 - 40 . Petroleum systems CH in emissions in 201 5 were estimated to be b etween 30.3 and Table MMT CO 2 4 emissions were estimated to be between 2.7 and 8.9 Eq., while CO MMT CO , Eq. at a 95 percent confidence level 2 2 based on previously calculated uncertainty . This indicates a range of 24 percent bel ow to 149 percent above the and CO Eq. for CH 2015 emission estimates of 39.9 and 3.6 , respectively. MMT CO 2 2 4 3 - Table : Approach 2 Quantitative Uncertainty Estimates for CH Emissions from 40 4 Petroleum Systems (MMT CO Eq. and Percent) 2 a Emission Estimate 201 5 Uncertainty Range Relative to Emission Estimate Gas Source b Eq.) (MMT CO (MMT CO (%) Eq.) 2 2 Lower Upper Lower Upper - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 68 3

183 Bound Bound Bound Bound 30.3 + 24% 149% Petroleum Systems - 99.3 CH 39.9 4 + 149% Petroleum Systems 24% - 8.9 2.7 CO 3.6 2 a Range of 201 5 relative uncertainty, based on 1995 base year activity factors, for a 95 percent confidence interval . b from other All reported values are rounded after calculation. As a result, lower and upper bounds may not be duplicable rounded values as shown in table. New 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 data sets were collected in the 1990s. To through 2015, for sources with new data, EPA reviewed available develop a consistent time series for 1990 , regulations, voluntary actions) information on factors that may have resulted in changes over the time series (e.g. der feedback on trends as well. For most sources, EPA developed annual data for 1993 2010 and requested stakehol - 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 yea can be found in the Recalculation Discussion below, r’s Inventory with additional detail provided in the 2017 Production Memo . For information on other sources, please see the Methodology Discussion above. QA/QC and Verification Discussion The petroleum system 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 and input, documentation, and calculation. QA/QC checks are consistently conducted performed for data gathering to minimize human error in the model calculations. EPA performs a thorough review of information associated with the Natural Gas STAR Program to assess whether the new studies, GHGRP data, regulations, public webcasts, and The EPA has a multi step data assumptions in the Inventory are consistent with current industry practices. - verification process for GHGRP data, including automatic checks during data lyses on entry, statistical ana - completed reports, and staff review of the reported data. Based on the results of the verification process, the EPA 159 follows up with facilities to resolve mistakes that may have occurred. As in previous years, EPA conducted early engagement and c ommunication with stakeholders on updates prior to greenhouse gas held data for oil public review. In December 2016 and January 2017, EPA stakeholder webinars on ing stakeholder and gas. In early 2017, EPA released memos detailing updates under consideration and request feedback. In February 2017, EPA released a public review draft of the Inventory. Stakeholder feedback received through these processes is discussed in the Recalculations Discussion and Planned Improvements sections below. several studies have measured emissions at the source level and at the national or regional level and In recent years, 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 In general, there are two major types of studies related to oil and gas greenhouse gas identifying areas for updates. data: studies that focus on measurement or quantifica of emissions from specific activities, processes and tion 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 sca les . 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 indi cations on estimates. potential over - and under - 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 s et, such as EDGAR) . In an nventory with measurement results that may be at other compare the national - level I effort to improve the ability to scales , a team at Harvard University along with EPA and other coauthors developed a gridded inventory of U.S. 1 degree x 0.1 degree ropogenic methane emissions with 0. anth spatial resolution, monthly temporal resolution, and 159 >. https://www.epa.gov/sites/produ ction/files/2015 - 07/documents/ghgrp_verification_factsheet.pdf See < 69 - 3 Energy

184 160 detailed scale - The gridded methane i nventory is designed to be consistent with dependent error characterization. Inventory of U.S. G - 2014) estimates for the year 2012, the U.S. EPA reenhouse Gas Emissions and Sinks (1990 161 which presents national totals for different source types. Recalculations Discussion The EPA received information and data related to the emission estimates through GHGRP reporting, the annual formal public notice periods, stakeholder feedback on updates under consideration, and new studies. Inventory In January 2017, the EPA released a draft memorandum, - Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990 Revisions under Consideration for Natural Gas and Petroleum Systems Production Emissions 2014: , referred to below as 2017 Production Memo , that discussed changes under consideration for that segment, and requested 162 stakeholder feedback on those changes. The EP A thoroughly evaluated relevant information available, and made updates to the production segment methodology for the Inventory including revised well count, equipment count, and pneumatic controller activity data, and revised activity and emissions data f or tanks and associated gas venting and flaring. In addition, as the updates to emission factors resulted in calculation of net emissions (already taking into account any reduced emissions) for sources in petroleum production, EPA removed Gas STAR reducti ons from the calculations. emissions, compared to the previous The combined impact of revisions to 2014 petroleum production segment CH 4 to 43 MMT CO Eq. (2 Inventory, is a decrease from 68 5 MMT CO Eq., or 37 percent). 2 2 The recalculations resulted in an average increase in emission estimates across the 1990 through 2014 time series, , or 6 compared to the previous Inventory, of 1 percent. The recalculations resulted in increases in the Eq MMT CO . 2 emission estimate in early ye ars of the time series, primar ily due to recalculations related to associated gas venting and flaring, and decreases in the emission estimate in later years of the time series, primarily due to recalculations for pneumatic controllers. In the current Inventory, EPA has held cons tant the CO values from the previous Inventory (developed using the 2 methodology as described in this section) as it assesses improvements to the CO estimates. See Planned 2 Improvements. Production This section references the memorandum, Inventory of U.S. 2015: - Greenhouse Gas Emissions and Sinks 1990 Revisions for Natural Gas and Petroleum Systems Production Emissions 2017 Production Memo ) , available at ( - gas - and - petroleum - systems - ghg - inventory - additional - information - 1990 - https://www.epa.gov/ghgemissions/natural - 2015 . This memorandum contains further details and documentation of recalculations. ghg C Well ounts ore recent version and improved data processing EPA has used a m of the DrillingInfo data set to update well counts data in the Inventory. For more information, see the 2017 Production Memo . This update , which addressed a double - counting issue in last year’s data set, res ulted in a decrease of 3 7 percent in oil well counts on average over the time series. Stakeholder feedback on the public review draft of the Inventory and on the 2017 Production Memo support the update to well counts data as it improves consistency with ot her recently published sources of well count data. 160 See < https://www.epa.gov/ghgemissions/gridded 2012 - methane - emissions >. - 161 - https://www.epa.gov/ghgemissions/us - greenhouse - gas - inventory See < report - 1990 - 2014 >. 162 1990 - See < https://www.epa.gov/gh gemissions/updates - under - consideration - petroleum - and - natural - gas - systems - - 2015 >. inventory - ghg - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 70 3

185 Table 3 41 : Oil Well Count Data - Oil Well Count 1990 2005 2011 2012 2013 2014 2015 Number of Oil Wells 572,639 481,340 540,743 564,348 580,960 598,627 586,896 904,675 764,371 838,899 867,375 884,652 898,268 NA Previous Estimate Percent Change in Counts - 37 % - 37 % - 36% - 35% - 34% - 33% NA NA ( Not Applicable ) Tanks EPA developed emission estimates for oil tanks using GHGRP data and a throughput - based approach. For more information, please see the 2017 Production Memo . Using 2015 GHGRP data, EPA developed a value for the tion (MMbbl) sent to tanks (62.7 percent ), and the fraction of petroleum sent to tanks fraction of petroleum produc large tanks with flares (55.5 percent ) , large tanks with VRU (20.1 percent ) , that is in each tank category: uncontrolled large tanks percent ) , small tanks with flares (1.9 percent ) , and small tanks without flares (17.6 (4.8 percent ) for 2015 . The fraction of petroleum production sent to tanks (62.7 percent ) was held constant throughout the 1990 through 2015 time series. The percentages of petroleum production sent to tanks that was sent to large tanks (93.2 ) and small tanks (6.8 percent ) were also held constant throughout the 1990 through 2015 time percent series. The 2015 fraction of tank in each control category was applied to for the years 2011 to 2015. For throughput uncontrolled categories. EPA then linearly 1990, it was assumed that all throughput was sent to tanks in the interpolated from 1990 to 2011 for each category. Category - specific emission factors developed from 2015 GHGRP data were applied for every year of the time series. EPA also developed a n emission factor for malfunc tioning separator v es. In the GHGRP, only large tanks report malfunctioning dump valve emissions . EPA has dump val applied the emission factor to all throughput in the large tank categories for each year of the time series. Stakeholder feedback on the publ ic review draft of the Inventory and on the 2017 Production M emo support the use of GHGRP data to calculate tank emissions and in particular the throughput approach, but recommended enhanced screening of GHGRP data. One stakeholder suggested that the tanks data underestimate tank emissions: the stakeholder suggested that the emission data and control efficiencies reported to GHGRP for this source may be k dump inaccurate and that the methods and data do not take into account the full volume of emissions from stuc valves and other malfunctions. Another stakeholder noted that aerial survey observations should not be presumed to indicate an underestimation of tank emissions in EPA’s GHGRP. Data are currently unavailable to assess nventory data on stuck dump valves, or to use aerial observations to inform Inventory malfunctions, to assess the I estimates for this source. See Planned Improvements. on average e in calculated emissions of 45 percent , The overall impact of the change is a decreas over the time series with smaller decreases in earlier years and larger decreases in recent years . 3 - 42 : Table MMbbl ) by Category and National Emissions National Tank Activity Data ( (Metric Tons CH ) 4 Activity Data/Emissions 1990 2005 2011 2012 2013 2014 2015 716 ) 0 470 MMbbl 822 946 1,106 1,197 Large Tanks w/ Flares ( Large Tanks w/ VRU ( MMbbl ) 0 171 260 298 343 401 434 Large Tanks w/o Control ( MMbbl ) 1,569 465 227 261 300 350 379 Small Tanks MMbbl ) 0 16 24 28 32 38 41 w/ Flares ( Small Tanks w/o Flares MMbbl) 115 65 ( 73 84 98 106 64 Total Emissions (MT) 248,325 81,604 80,474 48,100 55,259 63,605 74,305 Previous Estimated Emissions 250,643 187,872 220,021 (MT) 278,638 330,049 396,275 NA Percent Change in Emissions - 1 % - 57 % - 78% - 80% - 81% - 81% NA Not Applicable ) NA ( 71 - 3 Energy

186 Equipment Counts (Fugitive Sources) GHGRP for R Additional reporting to eporting Year (RY) EPA’s 2015 improved EPA’s allocation of GHGRP equipment counts between natural gas and petroleum for certain equipment leak category sources. EPA used the 2015 reporting data to develop improved counts of equipment per well. For more information, please see the 2017 using 2015 GHGRP data and applied those to Production Memo of equipment . EPA developed per well counts national oil well counts for years 2011 through 2015. The per well counts for 1990 through 1992 were retained from for 1993 through 2010 were developed by linear interpolation. Overall, the change nventories, and counts I previous decreased calculated emissions over the time series by around 14 percent, with the largest changes in light crude of the Inventory and on the separators. Stakeholder feedback on the public review draft 2017 Production Memo supported the use of updated GHGRP activity data. One stakeholder suggested that the approach of applying the GHGRP average equipment counts to all wells in the United States may not appropriately characterize production population that does not report to GHGRP, which may have higher or lower equipment counts per well. Data are currently unavailable to assess any differences between these populations. See Planned Improvements. - 43 : National Equipment Counts for Fugitive Sources and National Emissions Table (Metric 3 Tons CH ) 4 1990 2005 2011 2012 2013 2014 2015 Activity Data/Emissions Separators (Heavy Crude) 12,561 18,224 21,288 22,218 22,872 23,567 23,105 (Counts) Separators (Light Crude) 165,859 193,744 202,202 114,315 208,154 214,484 210,281 (Counts) Heater/Treaters (Light Crude) 87,010 112,277 126,575 132,100 135,989 140,124 (Counts) 137,378 Headers (Heavy Crude) 26,000 31,777 33,165 34,141 35,179 34,490 (Counts) 14,937 Headers (Light Crude) (Counts) 46,307 80,603 98,514 102,814 105,840 109,059 106,922 Total Emissions (MT) 26,428 37,486 43,504 45,403 46,740 48,161 47,218 Previous Estimated Emissions 28,420 45,244 54,139 55,977 57,092 57,970 NA (MT) 7 % Percent Change in Emissions - 17 % - 20% - - 19% - 18% - 17% NA NA ( Not Applicable ) Pneumatic Controllers and Chemical Injection Pumps The changes to pneumatic controller and chemical injection pump equipment counts result from the changes in oil GHGRP, which well counts described above and from the improved estimate of the counts of oil wells in EPA’s nts of controllers and pumps per oil well. The total per well counts of pneumatic improved the activity factors of cou controllers and pumps were updated using year 2015 GHGRP data. These per well counts were applied to years specific 2011 through 2015. For years 2011 through 2015, GHGRP year data on fractions of pneumatic controllers - in each category (high bleed “HB”, low bleed “LB”, and intermittent “IB”) were applied to the counts of pneumatic controllers. The 1990 through 1992 per well counts of controllers in each category and pumps were retained for - 1990 through 1992 and then the per well counts of pneumatic controllers in each category for 1993 through 2010 were developed by linearly interpolating from 1992 through 2011. Category specific emissions factors developed - entory from year 2014 GHGRP activity data were applied throughout the time series. The for the previous Inv recalculations resulted in large decreases in total national counts, but only minor changes in the annual fractions of controllers in each category. Overall, the change decreased calculated emissions over the time series by around 32 percent for pneumatic controllers, and 38 percent for chemical injection pumps. Table 3 - 44 : Pneumatic Controller and Chemical Injection Pump Na ti onal Equipment Counts and National Emissions (Metric Tons CH ) 4 Activity Data/Emissions 1990 2005 2011 2013 2014 2015 2012 Pneumatic Controllers High Bleed (Counts) 163,674 93,305 53,512 37,315 22,199 19,240 19,458 276,586 303,965 266,370 247,337 Low Bleed (Counts) 176,360 170,765 154,349 Intermittent Bleed 383,375 414,074 409,626 (Counts) 0 151,502 240,801 251,394 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 72 3

187 Previous High Bleed 163,225 160,475 103,061 76,469 50,241 43,211 NA (Counts) Previous Low Bleed 460,289 495,938 494,211 337,406 300,940 NA (Counts) 303,132 Previous Intermittent Counts - 284,053 533,112 Bleed 806,207 868,079 NA 599,859 593,342 Total Emissions (MT) 683,107 643,177 713,734 740,188 745,330 764,189 Previous Estimated 1,211,263 1,348,290 1,334,230 762,095 1,511,099 1,569,471 NA Emissions (MT) Percent Change in Emissions 0% - 44 % - 52% - 56% - 53% - 53% N A Chemical Injection Pumps Chemical Injection (Counts) 32,337 51,234 43 , 151 49,091 Pumps 52,742 54,346 53,281 Previous Pumps (Counts) 32,236 89,796 123,100 125,552 127,484 NA 119,058 77,636 49,001 65,388 74,389 Total Emissions (MT) 79,922 82,352 80,738 Previous Estimated 180,413 Emissions (MT) 48,849 136,071 186,537 190,253 193,181 NA Percent Change in 0% - 52 % - 59% - 58% - 58% - 57% NA Emissions ( NA ) Not Applicable a nd Flaring Associated Gas Venting EPA developed a new estimate for associated gas venting and flaring, replacing its previous estimates for stripper 2017 Production Memo . EPA developed a total percentage of oil well venting. For more information, please see the wells that vent and flare from 2015 GHGRP data (12 percen t), and applied that value to total national oil well counts the full time series. EPA then applied the GHGRP year - for specific split of that 12 percent between venting wells and flaring wells for years 2011 to 2015, and applied the 2011 split to each year from 1990 to 2011. Emission factors developed from year 2015 GHGRP data were applied for the full time series. EPA then removed the “stripper well” line item that had been included in previous inventories as those emissions are included in the Stakeholder feedback on the public review draft of the updated es timates for associated gas venting and flaring. 2017 Production Memo Inventory and on the support the use of GHGRP data to calculate emissions from this source and one stakeholder suggested that annua l updates to use GHGRP would be appropriate as this activity can vary , 1990 through 2010) associated gas significantly from year to year. Likewise, the stakeholder noted that past (e.g. venting and flaring likely varied significantly from year to year and from region to region. Data are not presently - to - year variation prior to 2011 into account. See Planned Improvements. available to take any year 3 Table 45 : Associated Gas Well Venting and Flaring National E missions (Metric Tons CH - ) 4 Source 1990 2005 2011 2012 2013 2014 2015 Associated Gas Well Venting 511,701 574,851 482,816 214,665 89,333 42,518 Emissions (MT) 608,758 Associated Gas Well Flaring Emissions (MT) 76,176 64,031 71,933 104,513 146,292 149,694 105,706 Previous Estimated Emissions from Stripper Wells (MT) 16,353 14,491 14,651 14,799 14,799 14,799 NA ( Not Applicable ) NA Gas S TAR Reductions i n Petroleum Systems Production Segment In the previous - time and ongoing reductions reported to the Natural Gas STAR Inventory, EPA included one 163 grouped together as “Other” line items in the petroleum systems production segment. The reductions Program resulted in a 1 to 5 percent decrease from potential produc tion segment emissions in years 1998 forward, and less 163 - star - gas >. program See < https://www.epa.gov/natural - 73 - 3 Energy

188 than 1 percent decrease in emissions for earlier years. Year 2008 was most impacted by Gas STAR with a 5 percent decrease from potential emissions. Most of the reductions that year resulted from implem enting artificial lift time reductions; 56 percent of year total) and recovering technologies that reduce well venting potential (one - casinghead gas (ongoing reduction; 40 percent of year total). These practices are generally reflected in the GHGRP data se ts used to calculate emissions from associated gas venting/flaring. EPA has revised the methodology for these sources in the 2017 Inventory to take into account control practices in the EFs and calculate net emissions account in calculations. Stakeholder feedback on directly, therefore Gas STAR reductio ns are no longer taken into support using Gas STAR reductions data 2017 Production Memo the public review draft of the Inventory and on the only where potential emissions are calculated, and removing them wh ere they create potential double - counting of Table 3 - 46 below shows Gas STAR data used in the previous Inventory for the production segment, and reductions. the production emissions calculated using a net emissions approach in this Inventory. - Table : Production Segment Gas STAR Reductions Update (Metric Tons CH 46 ) 3 4 Source 1990 2005 2011 2012 2013 2014 2015 Production Emissions 2,188,053 1,807,524 1,888,190 1,745,616 1,688,560 1,561,099 1,824,787 Previous Production Potential Emissions 1,519,486 1,957,071 2,262,521 2,346,709 2,586,398 2,725,082 N A Previous Production A N Gas STAR Reductions (154) (35,774) (30,856) (44,872) (45,013) (30,856) Previous Production 1,519,332 1,921,298 A 2,217,649 2,301,696 N 2,555,542 2,694,227 Net Emissions NA (Not Applicable) Note: Parentheses indicate negative values. Transportation Recalculations due to updated activity data for quantity of petroleum transported by barge or tanker in the transportation segment have resulted in an average increase in calculated emissions over the time series from this segment of less than 0.01 percent. Refining Recalculations due to updated data, including resubmitted GHGRP data, in the refining segment have resulted in an avera ge increase in calculated emissions over the time series from this segment of less than 0.01 percent. P lanned Improvements Plans for 2018 Inventory (1990 through 2016) and Future Inventories CO Data Update 2 In the current Inventory, EPA has held constant values from the previous Inventory as it assesses the CO 2 improvements to the CO estimates. EPA is reviewing CO data from EPA’s GHGRP and considering updates that 2 2 improve consistency of data sources and methods between the emission estimates (which have been updated in CH 4 recent years) and the CO emission estimates in Petroleum Systems. EPA has conducted a preliminary assessment 2 of the CO data and will seek stakeholder feedback on the draft assessment and options for updates . Using GHGRP 2 data to upd ate CO from Petroleum Systems. The update could could result in an increase in the estimate of CO 2 2 result in a shift in where CO from onshore production flaring for both from flaring is estimated — currently, CO 2 2 Natural Gas and Petroleum Systems is inclu ded in Natural Gas Systems. GHGRP data would allow for an estimate for CO from associated gas specifically from associated gas flaring. The 2015 GHGRP reported total of CO 2 2 . Scaling up to the national level using CH venting and flaring is around 10 MMT CO the same method as for 2 4 related CO - (based on oil well counts) would result in a significantly higher estimate. Similarly, scaling up tank 2 calculations would result in a significant CH emissions to the national level using the same method as used for 4 increase in emissions from that category. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 74 3

189 Uncertainty he most recent uncertainty analysis for the petroleum systems emission As noted in the Uncertainty discussion, t Since the estimates in the Inventory was conducted for the 1990 to 2009 Inventory that was released in 2011. to reflect improved data and analysis was last conducted, several of the methods used in the Inventory have changed changes in In addition, new studies and o ther data sources such as those industry practices and equipment. discussed in the sections below offer improvement to understanding and quantifying the uncertainty of some EPA is preparing a draft emission source estimates. update to the uncertainty analysis conducted for the 2011 Invent and will seek stakeholder feedback on the draft analysis as part of the ory to reflect the new information development of the next (i.e., 1990 through 2016) Inventory. Oil Wells Abandoned seeking emission factors and national activity Abandoned wells are not currently included in the Inventory. EPA is Stakeholder comments supported including this source category in future data available to calculate these emissions. Inventories , but noted that current ly data a re limited, and suggested reviewing data that wi ll become available in the future. EPA has identified studies with data on abandoned wells (Townsend ; Kang et al. 2016 ; Small et al. 2016 - Brandt et al. 2014), and is considering including an estimate for this source in future I nventories. A preliminary est imate, based on the national emission estimate from Townsend - Small et al. (2016) , the range of abandoned w ell counts in Townsend - Small et al. (2016) and Brandt et al. (2014) , and the current split between oil and gas wells in the total producing wells popu lation for 1990, is around 2.6 to 3.4 MMT CO seeks stakeholder feedback Eq . EPA 2 on abandoned wells. Upcoming Data, and Additional Data that Could Inform the Inventory continue to review data available from the GHGRP, in particular new data on hydraulically fractured oil EPA will well completions and workovers and new well specific information, available in 2017 for the first time. EPA will - consider revising its method s to take into account the new GHGRP data. EPA wil l 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. e used to update the Inventory . EPA wil l also continue to EPA continues to track studies that contain data that may b - down and bottom - up estimates, and which could lead to improved assess studies that include and compare both top understanding of unassigned high emitters or “superemitters,” (e.g. , identification of emission sources and information on fr equency 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 th at 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. See, for example, Equipment Counts in Recalculations Discussion. • 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. Abandoned well a ctivity and emissions data. See above section in Planned Improvements. • EPA will continue to seek available data on the se 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 CO O is produced from both naturally - occurring C 2 2 reservoirs and from industrial sources such as natural gas processing plants and ammonia plants. In the Inventory, - emissions from naturally are estimated based on the specific application. produced CO 2 75 - 3 Energy

190 In the Inventory, CO - EOR indus trial and commercial applications (e.g., food processing, that is used in non 2 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 used i n EOR operations is - occurring CO 2 emitted from natural gas processing and assumed to be fully sequestered. Additionally, all anthropogenic CO 2 ammonia plants is assumed to be emitted to the atmosphere, regardless of whether the CO is captured or not. These 2 emissions ar e 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 . Th storage of CO e methodology is based on the principle that the carbon capture and storage system should be 2 handled in a complete and consistent manner across the entire Energy sector. The approach accounts for CO 2 captured at natural and industrial sites as well as emis sions from capture, transport, and use. For storage specifically, specific evaluations. However, - a Tier 3 methodology is outlined for estimating and reporting emissions based on site IPCC (IPCC 2006) notes that if a national regulatory process exists, emis sions information available through that process may support development of CO emission estimates for geologic storage. 2 In the United States, facilities that produce CO for various end - use applications (including capture facilities such as 2 al plants and ammonia plants), importers of CO , exporters of CO , facilities that conduct geologic acid gas remov 2 2 sequestration of CO and facilities that inject CO underground, are required to report greenhouse gas data annually , 2 2 onducting geologic sequestration of CO to EPA through its GHGRP. Facilities c are required to develop and 2 implement an EPA approved site - specific monitoring, reporting and verification plan, and to report the amount of - sequestered using a mass balance approach. CO 2 Currently a vailable GHGRP d - level annual quantities of ata relevant for this inventory estimate consists of national CO For 2015, data from EPA’s GHGRP (Subpart captured and extracted for EOR applications for 2010 to 2015. 2 PP) were unavailable for use in the current Inventory report due data confidentiality reasons. A linear trend to 2014) to estimate 2 015 extrapolation was performed based on previous GHGRP reporting years (2010 through EPA will continue to evaluate the availability of additional GHGRP data and other opportunities for emissions. improving the emission estimates. These estimates indicate that the amount of CO captured and extracted from natural and industrial sites for EOR 2 applications in 2015 is 61.0 MMT CO specific monitoring Eq. (60,988 kt) (see Table 3 - 47 and Table 3 - 48 ). Site - 2 and reporting data for CO injection sites (i.e., EOR operations) were not readily available, therefore, the quantity of 2 captured and extr CO acted from industrial captured and extracted is noted here for information purposes only; CO 2 2 and commercial processes is assumed to be emitted and included in emissions totals from those processes . Table - 47 : Quantity of CO ) Capture d and Extract ed for EOR Operations (MMT CO 3 2 2 2014 2012 2015 Stage 1990 2005 2013 2011 Capture Facilities 4.8 6.5 9.9 9.3 12.2 13.1 13.5 20.8 28.3 48.4 48.9 47.0 46.2 47.5 Extraction Facilities Total 25.6 34.7 58.2 58.1 59.2 59.3 61.0 independent rounding. Note: Totals may not sum due to Table 3 - 48 : Quantity of CO Capture d and Extract ed for EOR Operations (kt) 2 Stage 1990 2005 2011 2012 2013 2014 2015 4,832 6,475 9,877 9,267 12,205 13,093 13,483 Capture Facilities Extraction Facilities 20,811 28,267 48,370 48,869 46,984 46,225 47,505 58,136 25,643 34,742 58,247 Total 59,189 59,318 60,988 Note: Totals may not sum due to independent rounding. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 76 3

191 3.7 Source Category Natural Gas Systems (IPCC 1B2b) The U.S. natural gas system encompasses hundreds of thousands of wells, hundreds of processing facilities, and over a million miles of transmission and distribution pipelines. Overall, natural gas systems emitted 162.4 MMT 6 Eq. (6, 4 9 7 kt) of CH CO in 2015, a 1 percent decrease compared to 1990 emissions, and a 0.1 percent de crease 4 2 Table 3 compared to 2014 emissions (see 49 , Table 3 - 50 , and Table 3 - 51 ) and 42.4 MMT CO Eq. (42,351 kt) of - 2 non - combustion CO in 2015, a 12 percent increase compared to 1990 emissions. 2 The 1990 to 2015 trend in CH is not consistent across segments. Overall, the 1990 to 2015 decrease in CH 4 4 ssions from distribution (75 percent decrease ), transmission and emissions is due primarily to the decrease in emi decrease), and processing storage (42 percent decrease) segments. Over the same time period, the (48 percent production segments saw increased methane emissions of 5 1 percent. Natural gas sys tems also emitted 42.4 MMT CO in 2015, a 12 percent increase compared to 1990 emissions. The 1990 Eq. (42,351 kt) of non - combustion CO 2 2 to 2015 increase in CO is due primarily to flaring; the volume of gas flared increased 93 percent from 1990. 2 Methane emissions from natural gas systems include those resulting from normal and non - combustion CO 2 operations, routine maintenance, and system upsets. Emissions from normal operations include: natural gas engine and turbine uncombusted exhaust, bleed and discha rge 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 four major stages of the natural gas system. Each of the stages is described and the combustion CO emissions are discussed. - and non different factors affecting CH 4 2 Production (including g athering and boosting). In the production stage, wells are used to withdraw raw gas from such as underground formations. Emissions arise from the wells themselves, and well - site gas treatment equipment dehydrators and separators. Gathering and boosting emi ssion sources are not reported under a unique segment, but sources are included within the production sector. The gathering and boosting 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 from facilities (custody transfer points are typically used to segregate sources between each segment). Emissions emissions and 44 percent of non production (including gathering and boosting) account for 6 percent of CH - 6 4 combustion CO emissions from natural gas systems in 2015. Emissions from gathering stations, pneumatic 2 controllers, liquids unloading, and off shore platforms account for most of the CH emissions in 2015. Flaring 4 emissions account for most of the non - combustion CO emissions. Due to the aggregated activity data source for 2 CO from flaring, it is not possible to develop separate estimates for fla ring occurring in natural gas production and 2 emissions from flaring (onshore and offshore) for both natural gas flaring occurring in oil production. Total CO 2 emissions in 2015 and are included in the Natural Gas Systems estimates. Methane and oil were 18.0 MMT CO 2 from production increased by 5 1 percent from 1990 to 2015, due primarily to increases in emissions from gathering and boosting stations ( driven by an increase in the number of stations), increases in emissions from pneumatic controllers (due to an increase in the number of controllers, particularly in the number of intermittent bleed controllers), and chemical injection pumps (due to an increase in the number of pumps). Carbon dioxide emissions from production increased 88 percent from 1990 to 20 15 due primarily to increases in flaring . Processing . In this stage, natural gas liquids and various other constituents from the raw gas are removed, resulting emissio ns from compressors, in “pipeline quality” gas, which is injected into the transmission system. Fugitive CH 4 combustion CO including compressor seals, are the primary emission source from this stage. Most of the non - 2 emissions come from acid gas removal (AGR) units, which are designed to remove CO from natural gas. 2 Processing plants account for 7 percent of CH emissions and 56 percent of non - combustion CO emissions from 4 2 emissions from processing decreased by 48 percent from 1990 to 2015 as emissions natural gas systems. Methane Carbon dioxide leaks and venting) and equipment leaks decreased. emissions from processing from compressors ( decreased by 15 percent from 1990 to 2015, due to a decrease in acid gas removal emissions. e, large diameter pipelines that transport Transmission and Storage. Natural gas transmission involves high pressur gas long distances from field production and processing areas to distribution systems or large volume customers 77 - 3 Energy

192 such as power plants or chemical plants. Compressor station facilities are used to move the gas throu ghout the U.S. emissions from these compressor stations, and venting from pneumatic transmission system. Fugitive CH 4 controllers account for most of the emissions from this stage. Uncombusted engine exhaust and pipeline venting are ssions from transmission. Natural gas is also injected and stored in underground formations, also sources of CH emi 4 or liquefied and stored in above ground tanks, during periods of low demand (e.g., summer), and withdrawn, (e.g., winter). In 2015, emissions from the Aliso Canyon processed, and distributed during periods of high demand leak event in Southern California contributed 2.0 MMT CO Eq. to transmission and storage emissions, around 5 2 percent of total emissions for this segment. Compressors and dehydrators are the primar y contributors to emissions 1 percent of emissions from the transmission and storage sector account for approximately 2 Methane from storage. emissions from natural gas systems, while CO emissions from transmission and storage account for less than 1 2 - perce emissions from natural gas systems. CH nt of the non emissions from this source decreased combustion CO 2 4 by 42 percent from 1990 to 2015 due to reduced compressor station emissions (including emissions from compressors and fugitives). CO ransmission and storage have decreased by 37 percent from 1990 emissions from t 2 to 2015, also due to reduced compressor station emissions. Distribution pipelines take the high - pressure gas from the transmission system at “city gate” stations, Distribution. reduce the press ure and distribute the gas through primarily underground mains and service lines to individual end users. There were 1,274,976 miles of distribution mains in 2015, an increase of over 330,000 miles since 1990 7 percent of CH emissions, which account for emissions (PHMSA 2016a; PHMSA 2016b). Distribution system 4 n 1 percent of non - emissions, result mainly from fugitive from natural gas systems and less tha combustion CO 2 r emissions than other pipe emissions from pipelines and stations. An increased use of plastic piping, which has lowe materials, has reduced and CO both CH emissions from this stage, as have station upgrades at metering and 4 2 emissions in 2015 were 75 percent lower than 1990 levels regulating (M&R) stations. Distribution system CH 4 m 43.5 MMT CO Eq. to 11.0 MMT CO (changed fro Eq.), while distribution CO emissions in 2015 were 72 2 2 2 emission from this segment are less than 0.1 MMT CO percent lower than 1990 levels (CO Eq. across the time 2 2 series). emissions for the four major stages o f natural gas systems are shown in MMT CO Total CH Eq. ( Table 3 - 49 ) and 4 2 provides additional information on how the estimates in kt ( 50 ). Table 3 - 51 - Table 3 - 47 were calculated. Table 3 3 - 51 shows the calculated potential CH release (i.e., potential emissions before any controls are applied) from Table 4 each stage, and the amount of CH that is est imated to have been flared, captured, or otherwise controlled, and 4 therefore not emitted to the atmosphere. Subtracting the value for CH that is controlled, from the value for 4 , results in the total net emissions values. calculated potential release of CH More disaggregated information on 4 and CO potential emissions and emissions is available in Annex 3.6. See Methodology for Estimating CH 4 2 Emissions from Natural Gas Systems . a 3 - 49 : CH Emissions from Natural Gas Systems (MMT CO Table Eq.) 2 4 Stage 1990 2005 2011 2012 2013 2014 2015 Field Production 70.6 95.2 106.6 106.9 106.3 108.2 104.5 Processing 11.1 10.9 11.1 21.3 11.7 10.1 10.1 Transmission and Storage 30.7 28.8 58.6 27.9 30.8 32.0 33.7 Distribution 43.5 22.1 11.1 11.3 11.2 11.2 11.0 Total 194.1 159.7 154.5 156.2 159.2 162.5 162.4 a These values represent CH emitted to the atmosphere. CH that is captured, flared, or otherwise controlled 4 4 (and not emitted to the atmosphere) has been calculated and removed from emission totals. Note: Totals may not sum due to independent rounding. a 3 - 50 : CH Emissions from Natural Gas Systems (kt) Table 4 2015 Stage 1990 2005 2011 2012 2013 2014 4,264 4,327 Field Production 2,826 3,808 4,178 4,274 4,253 Processing 853 466 405 406 434 446 445 Transmission and Storage 2,343 1,230 1,152 1,116 1,232 1,282 1,349 Distribution 1,741 884 444 451 449 446 439 Total 7,762 6,387 6,180 6,247 6,368 6,501 6,497 a that is captured, flared, or otherwise controlled (and emitted to the atmosphere. CH These values represent CH 4 4 not emitted to the atmosphere) has been calculated and removed from emission totals. independent rounding. Note: Totals may not sum due to - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 78 3

193 Table 51 : Calculated Potential CH 3 and Captured/Combusted CH - from Natural Gas 4 4 Eq.) Systems (MMT CO 2 1990 2005 2011 2012 2013 2015 2014 a 194.1 179.1 172.4 174.5 Calculated Potential 176.6 180.5 181.1 Field Production 101.0 112.4 114.7 114.4 116.7 115.8 70.6 Processing 10.1 10.1 21.3 11.7 11.1 11.1 10.9 Transmission and Storage 43.1 37.3 37.3 39.1 40.4 42.2 58.6 Distribution 23.3 12.6 12.4 12.2 12.2 12.0 43.5 Captured/Combusted 18.0 19.4 17.9 18.3 17.4 0.0 18.6 Field Production 7.9 7.8 8.1 8.6 9.2 0.0 5.8 Processing 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Transmission and Storage 0.0 12.4 8.5 9.4 8.3 8.4 8.5 Distribution 1.2 1.5 0.0 1.1 1.0 1.0 1.0 Net Emissions 194.1 159.7 154.5 156.2 159.2 162.5 162.4 Field Production 95.2 104.5 70.6 106.9 106.3 108.2 106.6 Processing 21.3 11.7 10.1 10.1 10.9 11.1 11.1 Transmission and Storage 27.9 58.6 30.7 28.8 30.8 32.0 33.7 Distribution 11.0 11.2 43.5 22.1 11.2 11.1 11.3 a In this context, “potential” means the total emissions calculated before voluntary reductions and regulatory controls are applied. Note: Totals may not sum due to independent rounding. - 52 : Non - combustion CO Emission Table s from Natural Gas Systems (MMT ) 3 2 1990 Stage 2005 2011 2012 2013 2014 2015 9.9 8.3 14.1 Field Production 13.7 16.6 18.6 18.6 Processing 27.8 23.7 21.7 21.5 21.5 21.8 23.7 Transmission and Storage 0.1 + + + + + + Distribution 0.1 + + + + + + Total 37.7 38.5 30.1 35.7 35.2 42.4 42.4 + Does not exceed 0.1 MMT CO Eq. 2 Note: Totals may not sum due to independent rounding. - 53 : Non - combustion CO Table Emissions from Natural Gas Systems (kt) 3 2 1990 2005 Stage 2011 2012 2013 2014 2015 13,684 18,585 16,649 14,146 Field Production 9,857 8,260 18,585 Processing 27,763 23,713 21,746 23,713 21,469 21,756 21,466 39 39 37 35 Transmission and Storage 36 62 43 14 14 14 14 15 Distribution 50 27 35,662 Total 30,076 37,732 42,351 42,351 38,457 35,203 Note: Totals may not sum due to independent rounding. Methodology The methodology for natural gas emission estimates presented in this public review draft of the Inventory involves and CO the calculation of CH emissions for over 100 emissions sources, and then the summation of emissions for 2 4 each natural gas segment. The approach for calculating emissions for natural gas systems generally involves the application of emission factors to activity data. For many sources, the approach uses technology - specific emission factors or emission factors that vary over time and ta ke into account changes to technologies and practices, which are used to calculate net 79 - 3 Energy

194 emissions directly. For others, the approach uses what are considered “potential methane factors” and reduction data to calculate net emissions. Emission Factors. Key re ferences for emission factors for CH emissions from the and non - combustion - related CO 2 4 U.S. natural gas industry include the Gas Research Institute (GRI) and EPA (EPA/GRI 1996) , the Greenhouse Gas Reporting Program (GHGRP), and others . The EPA/GRI study developed over 80 CH emission factors to characterize emissions from the various components 4 within the operating stages of the U.S. natural gas system. The EPA/GRI study was based on a combination of process engineering studies, collection of activity dat a, and measurements at representative gas facilities conducted in the early 1990s. Methane compositions from the Gas Technology Institute (GTI, formerly GRI) Unconventional g gross production for oil and Natural Gas and Gas Composition Databases (GTI 2001) are adjusted year to year usin gas supply National Energy Modeling System (NEMS) regions from the EIA. Therefore, emission factors may vary from year to year due to slight changes in the CH composition for each NEMS oil and gas supply module region. 4 emissions. Data from were also used to calculate non - combustion CO The e mission factors used to estimate CH 4 2 GTI 2001 were used to adapt the CH emission factors into non - combustion related CO emission factors. 4 2 Additional information about CO content in tran smission quality natural gas was obtained from numerous U.S. 2 combustion CO - emission factors. transmission companies to help further develop the non 2 . In the were used to develop emission factors for several sources in the Inventory GHGRP Subpart W data duction segment, GHGRP data were used to develop emission factors for gas well completions and workovers pro (refracturing) with hydraulic fracturing, pneumatic controllers and chemical injection pumps, condensate tanks, and g segment, for recent years of the times series, GHGRP data were used to develop liquids unloading. In the processin emission factors for fugitives, compressors, flares, dehydrators, and blowdowns/venting. In the transmission and storage segment, for recent years of the times series, GHGRP d ata were used to develop factors for pneumatic controllers. Other data sources used for emission factors include Marchese et al. for gathering stations, Zimmerle et al. for transmission and storage station fugitives and compressors, and Lamb et al. for re cent years for distribution pipelines and meter/regulator stations. See Annex 3.6 for more detailed information on the methodology and data used to calculate CH and non - 4 combustion CO emissions from natural gas systems. 2 Activity Data. Activity data were taken from various published data sets, as detailed in Annex 3.6. Key activity data sources include data sets developed and maintained by DrillingInfo, Inc.; U.S. Department of the Interior’s Bureau of Ocean Energy Management, Regulation and Enforcement (BOEMRE, previously Minerals and Management Service); U.S. Department of Energy’s Energy Information Administration (EIA) and Federal Energy Regulatory Gas Journal; U.S. Commission (FERC); the Natural Gas STAR Program annual emissions savings data; Oil and Department of Transportation’s Pipeline and Hazardous Materials Safety Administration; EPA’s Greenhouse Gas Reporting Program; the Wyoming Conservation Commission; and the Alabama State Oil and Gas Board. activity data are not available. For these sources, either 2014 data was used as a For a few sources, recent direct proxy for 2015 data, or a set of industry activity data drivers was developed and used to calculate activity data over the time series. Drivers include statistics on gas pr oduction, number of wells, system throughput, miles of various kinds of pipe, and other statistics that characterize the changes in the U.S. natural gas system infrastructure and operations. More information on activity data and drivers is available in Ann ex 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, emissions are calculated directly by applying emission factors to activity data. Howeve r, for certain sectors, some sources are calculated using potential emission factors, and the emissions is step of deducting CH potential estimates to develop net CH that is not emitted from the total CH 4 4 4 applied. To take into account use of such techno logies and practices that result in lower emissions, 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 de hydrator vents. Voluntary reductions included in the Inventory are those reported to Natural Gas STAR. values from the previous Inventory (developed using the In the current Inventory, EPA has held constant the CO 2 methodology as described in this secti on) as it assesses improvements to the CO estimates. See Planned 2 Improvements. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 80 3

195 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 ) an hour, and continued lea king until it was permanently sealed in February of ) of methane (CH 4 164 d for 2015, using the estimate of the leak published 2016. An emission estimate from the leak event was include only include those by the California Air Resources Board (99,650 MT for the duration of the leak), adjusted to emissions that occurred in 2015 (2016 emissions will be included in next Inventory). The 2015 emission the estimate of 78,350 MT CH was added to the 2015 estimate of fugitive emissions from storage wells, calculated 4 with an emission fa ctor approach, resulting in total emissions from storage wells in 2015 of 92,590 MT CH . For 4 Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990 - 2015: Update for more information, please see 165 . Storage Segment Emissions - onsistency Uncertainty and Time Series C The most recent uncertainty analysis for the natural gas and petroleum systems emission estimates in the Inventory to 2009 Inventory report that was released in 2011 and has not yet been updated . Since was conducted for the 1990 was last conducted, several of the methods used in the Inventory have changed to reflect improved data the analysis and changes in industry practices and equipment. In addition, new studies (e.g., Lamb, et al. 2015; Lyon, et al. 2015; t al. 2015) and other data sources such as those discussed in the sections below Marchese, et al. 2015; Zimmerle, e EPA is offer improvement to understanding and quantifying the uncertainty of some emission source estimates. update to the uncertainty analysis conducted for the 2011 Inventory to reflect the new information preparing a draft and will seek stakeholder feedback on the draft analysis as part of the development of the next (i.e., 1990 through 2016) Inventory. For more information, please see the Planned Improvements section. velop the values in To de 3 - 54 below, EPA has applied the uncertainty percentage ranges calculated Table previously to the updated 201 5 emission estimates . To develop the uncertainty percentage ranges, EPA used the IPCC - recommended Approach 2 methodology (Monte Carlo Simulation technique). The @RISK software model - emitting sour ces was used to quantify the uncertainty associated with the emission estimates using the 12 highest (“top 12 sources”) for the year 2009. The @RISK analysis provides for the specification of probability density functions for key variables within a computational structure that mirrors the calculation of the inventory estimate. The 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. Given the recent revisions, the uncertai nty ranges applied may not reflect the uncertainty associated with the recently revised emission factors and activity data sources. The results presented below provide with 95 percent certainty the range within which emissions from this source category are likely to fall for the year 201 5 , based on the previously conducted uncertainty assessment using the recommended IPCC methodology. The results of the Approach 2 quantitative uncertainty analysis are summarized in 3 - 54 . Natural gas systems CH emissions in 201 5 were estimated to be between 131.6 and 211.2 Table MMT CO 2 4 - Eq. at a 95 percent confidence level, based on previously calculated uncertainty. Natural gas systems non energy CO Eq. at a 95 percent confidence level. emissions in 2014 were estimated to be between 34.3 and 55.1 MMT CO 2 2 Table 3 - 54 : Approach 2 Quantitative Uncertainty Estimates for CH and Non - energy CO 2 4 Emissions from Natural Gas Systems (MMT CO Eq. and Percent) 2 a Uncertainty Range Relative to Emission Estimate 201 Emission Estimate 5 Gas Source b (MMT CO (%) Eq.) (MMT CO Eq.) 2 2 164 For more information on the Aliso Canyon event, and the measurements conducted of the leak, please see Ensuring Safe and , available Final Report of the Interagency Task Force on Natural Gas Storage Safety Reliable Underground Natural Gas Storage, at . 165 - and - gas - . 81 - 3 Energy

196 Lower Lower Upper Upper b b b b Bound Bound Bound Bound 131.6 Natural Gas Systems CH - 162.4 19% 211.2 +30% 4 c 42.4 CO +30% Systems Natural Gas 19% - 55.1 34.3 2 a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo Simulation analysis conducted for the year 2009. 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 - 49 and Table 3 - 50 . c An uncertainty analysis for the non - energy CO emissions was not per formed. The relative uncertainty estimated (expressed 2 as a percent) from the CH energy CO - uncertainty analysis was applied to the point estimate of non emissions 2 4 2015, for sources with new data, EPA through reviewed available To develop a consistent time series for 1990 regulations, voluntary actions) information on factors that may have resulted in changes over the time series (e.g. , and requested stakeholder feedback on trends as well. For most sources, EPA developed annual data for 1993 - 2010 by interpolating activity data or emission factors or both between 1992 and 2011 data points. Information on t ime - series consistency for sources updated in this public review draft can be found in the Recalculation Discussion below, with additional detail provided in the 2017 Production and Processing memos. For detailed information, please see Annex 3.5. nd Verification Discussion QA/QC a 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 med for data gathering and input, documentation, and calculation. QA/QC checks are consistently conducted perfor to minimize human error in the model calculations. EPA performs a thorough review of information associated with public webcasts, and the Natural Gas STAR Program to assess whether the new studies, GHGRP data, regulations, The EPA has a multi - step data assumptions in the Inventory are consistent with current industry practices. entry, statistical analyses on verification process for GHGRP data, including automatic checks during data - completed reports, and staff review of the reported data. Based on the results of the verification process, the EPA 166 follows up with facilities to resolve mistakes that may have occurred. As in previous years, EPA conducted early engagement and communication with stakeholders on updates prior to greenhouse gas data for oil public review. In December 2016 and January 2017, EPA held stakeholder webinars on sideration and requesting stakeholder and gas. In early 2017, EPA released memos detailing updates under con feedback. In February 2017, EPA released a public review draft of the Inventory. Stakeholder feedback received through these processes is discussed in the Recalculations Discussion and Planned Improvements sections bel ow. 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, i ncluding 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 measu rement or quantification of emissions from specific activities, processes and inverse modeling to estimate the level of overall emissions needed to equipment, and studies that use tools such as 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 provide general indications on many cases, incorporated data from these data sources. The second type of study can - and under - estimates. potential over 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 an . ) . In an d not another data set, such as EDGAR 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 i nventory of U.S. 166 See < https://www.epa.gov/sites/production/files/2015 - 07/documents/ghgrp_verification_factsheet.pdf >. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 82 3

197 anthropogenic methane emissions with 0.1° x 0.1° spatial resolution, monthly temporal resolution, and detailed 167 scale The gridded methane inventory is designed to be consistent with the 2016 - dependent error characterization. estimates for the year - 2014) .S. Greenhouse Gas Emissions and Sinks 2012, which presents Inventory of U (1990 168 national totals for different source types. Recalculations Discussion The EPA received information and data related to the emission estimates through the annual GHGRP reporting, Inventory formal public notice periods, stakeholder feedback on updates under consideration, and new studies. In January 2017, the EPA released draft memoranda that discussed the changes under consideration and requested 169 older feedback on those changes. stakeh thoroughly The EPA evaluated relevant information available, and made several updates in the Inventory, including revisions to production segment activity and emissions data, gathering and boosting facility emissions, and processing segment activity and emissions data. CH estimates The impact of all revisions to natural gas systems is a decrease of 13.6 MMT CO percent, Eq. , or 8 2 4 this Inventory. O the 2014 value in ver the time series, the comparing the 2014 value from the previous Inventory to 9 average change is a decrease of 1 MMT CO .0 Eq. , or percent. 7 2 In the current Inventory, EPA has held constant the CO values from the previous (i.e., 1990 through 2014) 2 this section) as it assesses improvements to the CO Inventory (developed using the methodology as described in 2 estimates. See Planned Improvements. Production This section references the Inventory production segment supporting memoranda: “Revisions to Natural Gas and 170 2017 P roduction Memo ). Petroleum Production Emissions,” (the This memorandum contains further details and documentation of recalculations. Overall, Recalculations for the production segment (including gathering and boosting facilities) resulted in a small decrease in the 2014 CH emission estimate, from 109.0 MMT CO Eq. in the previous Inventory, to 108.2 MMT 4 2 CO Eq., Eq. in this Inventory, or 1 percent. Over the time series, the average change is a decrease of 11 MMT CO 2 2 or 11 percent. Tanks based approach. For condensate tanks using GHGRP data and a throughput - EPA developed emission estimates for . Using 2015 GHGRP data, EPA developed a value for the more information, please see the 2017 Production Memo 79 the split between large tanks (77.1 percent) ), percent fraction of condensate 4 produced (MMbbl) sent to tanks ( . and small tanks (22.9 percent), and the fraction of sent to large and small tanks that is in each tank condensate sate is sent to tanks with flares, 13.4 percent is control category. For large tanks, for 2015, 69.0 percent of conden tanks with VRU, and 17.6 percent is sent to uncontrolled large tanks . For small tanks, 33.5 percent of sent to small tanks with condensate goes to and 66.6 percent goes to small tanks without flar es. To develop the time flares, series, t he fraction of condensate production sent to tanks (79.4 percent ) was held constant throughout the 1990 to ) 2015 time series. The percentages of production sent to tanks that was sent to large tanks (77.1 percent condensate 167 See < https://www.epa.gov/ghgemissions/gridded - 2012 - methane - emissions >. 168 invento See < https://www.epa.gov/ghgemissions/us - greenhouse - gas - ry - report - 1990 - 2014 >. 169 See Revisions under Consideration for Natural Gas and Petroleum Systems Production Emissions , and Revisions under Consideration for Natural Gas Systems Processing Segment Emissions , available at < https://www.epa.gov/ghgemissions/upd ates - under - consideration - petroleum - and - natural - gas - systems - 1990 - 2015 - ghg - inventory >. 170 additional - - - information - See < https://www.epa.gov/ghgemissions/natural - gas - and - petroleum - systems - ghg - inventory 1990 >. ghg - 2015 83 - 3 Energy

198 and small tanks (22.9 to 2015 time series. The 2015 fraction of percent ) were also held constant throughout the 1990 control category was applied to for the years 2011 to 2015. throughput in each 2015 For large tanks, it was assumed percent 50 percent went to tanks with that in 1990, of condensate to large tanks went to tanks without controls, 50 percent went to tanks with VRUs. The previous Inventory applied an assumption that 50 percent of flares, and that 0 and 50 percent to controlled tanks (VRU or flares). For small tanks, it was condensate went to uncontrolled tanks in 1990 This assumption was applied all throughput was sent to tanks in the uncontrolled categor y . assumed that the 2015 GHGRP data. For both large and small because of the relatively limited use of controls at small tanks in control category. Category - specific emission factors tanks, EPA linearly interpolated from 1990 to 2011 for each developed from 2015 GHGRP data were applied for every year of the time series. EPA also develop ed an emission EPA’s GHGRP, only large tanks report malfunctioning dump valves. factor for malfunctioning dump values. In EPA has applied the emission factor to all throughput in the large tank categories for each year of the time series. Stakeholder feed M emo support the use back on the public review draft of the Inventory and on the 2017 Production of GHGRP data to calculate tank emissions and in particular the throughput approach, but recommended enhanced d that the tanks estimate underestimates tank emissions: the screening of GHGRP data. One stakeholder suggeste GHGRP for this source may EPA’s stakeholder suggested that the emission data and control efficiencies reported to lume of emissions from stuck dump be inaccurate and that the methods and data do not take into account the full vo valves and other malfunctions. Another stakeholder expressed that aerial survey observations should not be presumed to indicate an underestimation of tank emissions in GHGRP. Data are currently unavailable to assess malf unctions, to assess the Inventory data on stuck dump valves, or to use aerial observations to inform Inventory estimates for this sources. See Planned Improvements. 86 percent on av erage over the time series. The overall impact of the change is a decrease in calculated emissions of 3 Table National Tank Activity Data ( MMbbl ) by Category and National Emissions : 55 - ) (Metric Tons CH 4 Activity Data/Emissions 1990 2005 2011 2012 2013 2014 2015 34 44 80 ) 99 118 124 126 MMbbl Large Tanks w/ Flares ( Large Tanks w/ VRU ( MMbbl ) - 7 16 19 23 24 24 Large Tanks w/o Control MMbbl 18 ) 20 25 ( 30 34 32 32 MMbbl ) - 5 Small Tanks w/ Flares ( 12 14 17 18 18 Small Tanks w/o Flares 34 20 36 15 35 23 ( 28 MMbbl ) Total Emissions (MT) 10,819 15,707 15,037 18,440 22,160 23,189 23,506 Previous Potential Emissions 93,224 119,191 229,284 (MT) 259,121 312,185 303,711 NA Previous Regulatory Reductions 31,908 61,381 69,369 83,575 81,306 NA (MT) - 189,752 87,283 167,902 93,224 228,610 222,405 NA Previous Net Emissions (MT) Percent Change in Emissions - 8 3 % - 88 % - 91 % - 90 % - 90 % - 90 % NA NA ( Not Applicable ) Well Counts of the DrillingInfo data set to update well counts EPA has used a more recent version and improved data processing data in the Inventory. For more information, see the 2017 Production Memo , which addressed a double - . This update resulted in a decrease of percent in gas well cou nts on average over the time counting issue in last year’s data set, 6 2017 Production Memo support series. Stakeholder feedback on the public review draft of the Inventory and on the . the update to well counts data as it improves consistency with other recently published sources of well count data 3 - 56 : Gas Well Count Data Table Gas Well Count 1990 2005 2011 2012 2013 2014 2015 440,371 202,628 355,234 438,672 431,926 433,941 421,893 Number of Gas Wells Previous Estimated Number of Gas Wells 218,709 373,903 463,198 460,588 454,491 456,140 NA NA % Percent Change - 7 % - 5 % - 5 % - 5 % - 5 % - 5 ) Not Applicable ( NA - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 84 3

199 Equipment Counts (Fugitive Sources) ) 2015 improved EPA’s allocation of GHGRP equipment eporting Year (R Additional reporting to GHGRP for R Y counts between natural gas and petroleum for certain equipment leak category sources. EPA used the 2015 reporting data to develop improved counts of equipment per well. For more information, please see the 2017 Production Memo . EPA developed per well counts of equipment using 2015 GHGRP data and applied those to national gas well counts for years 2011 through 2015. The per well counts for 1990 through 1992 were retained from previous for 1993 through 2010 were developed by linear interpolation. Overall, the change decreased I nventories, and counts percent, with the largest decreases in meters/piping (20 calculated emissions over the time series by around 1 3 percent). (2 percent), and dehydrato rs ( 19 Stakeholder feedback on the public review percent) and compressors 1 support the use of updated GHGRP activity data. One 2017 Production Memo draft of the Inventory and on the stakeholder highlighted a discrepancy in well count data reported under different categories in GHGRP. EPA will changes update next year’s Inventory with resubmitted data, which may equipment counts per well result in minor in One stakeholder suggested that the approach of applying GHGRP average equipment counts to all wells for 2015. in the United States may not appropriately characterize the production population that does not report to EPA’s GHGRP, which may have higher or lower equipment counts per well. Data are currently unavailable to assess any differences between these populations. See Planned Improvements. - 57 : National Equipment Counts for Fugitive Sources and National Emissions (Metric Table 3 ) Tons CH 4 Activity Data/Emissions 1990 2005 2011 2012 2013 2014 2015 Separators (Counts) 119,216 235,387 301,706 300,542 295,920 297,301 289,046 90,551 Heaters (Counts) 75,389 90,901 48,573 89,158 89,574 87,087 Dehydrators (Counts) 28,250 19,486 11,727 11,682 11,502 11,556 11,235 182,647 305,242 377,597 Meters/Piping (Counts) 376,141 370,356 372,084 361,753 Compressors (Counts) 17,575 28,423 34,473 34,340 33,812 33,969 33,026 138,777 248,898 300,327 Total Emissions (MT) 299,137 293,315 291,617 286,872 Previous Estimated Emissions 292,944 364,342 362,341 356,470 354,306 NA (MT) 153,106 Percent Change in Emissions - 9 % - 1 5 % - 1 8 % - 1 7 % - 8 % - 1 8 % NA 1 ( Not Applicable ) NA Pneumatic Controllers and Chemical Injection Pumps The changes to pneumatic controller and chemical injection pump equipment counts result from the changes in gas gas wells in GHGRP, which improved well counts described above and from the improved estimate of the counts of gas well. Total per well counts of pneumatic controllers the activity factors of counts of controllers and pumps per were updated using year 2015 GHGRP data. These per well counts were applied to and chemical injection pump - specific data on fractions of pneumatic years 2011 through 2015. For years 2011 through 2015, GHGRP year controllers in each category (high bleed “HB”, low bleed “LB”, and inte rmittent “IB”) were applied to the counts of pneumatic controllers. The 1990 through 1992 per well counts of controllers in each category and pumps were retained for 1990 through 1992 and then the per - well counts of pneumatic controllers in each category a nd pumps for 1993 through 2010 were developed by linear interpolating from 1992 through 2011. Category - specific emissions factors developed from year 2014 GHGRP data were applied throughout the time series. The recalculations using the latest GHGRP y data resulted in only minor changes in the annual fractions of controllers in each category, activit and small decreases in total calculated emissions. 3 - 58 : Pneumatic Controller and Chemical Injection Pump Table National Equipment Counts and National Emissions (Metric Tons CH ) 4 Activity Data/Emissions 1990 2005 2011 2012 2013 2014 2015 Pneumatic Controllers Low Bleed (Counts) 144,046 289,638 247,684 165,646 197,081 188,740 NA 28,956 39,218 21,672 High Bleed (Counts) 72,538 101,476 79,769 68,479 85 - 3 Energy

200 Intermittent Bleed 134,713 343,087 459,066 509,114 607,722 590,340 583,298 (Counts) Previous Low Bleed 138,223 276,586 239,734 144,443 226,280 NA (Counts) - Previous High Bleed 80,776 106,689 86,310 76,418 (Counts) 42,050 29,006 NA Previous Intermittent 526,908 (Counts) 360,379 484,942 150,013 645,408 579,633 NA Bleed Total Emissions (MT) 993,203 1,020,246 1,099,713 1,120,439 523,787 1,130,709 1,064,230 Previous Estimated Emissions 556,347 1,079,256 1,229,714 1,245,311 1,259,753 1,105,119 NA - 6 % - 8 % - 11% - 10% - 10% - 4% NA Percent Change Chemical Injection Pumps Chemical Injection 57,115 79,806 82,980 83,301 Pumps (Counts) 81,704 82,085 15,904 Previous Chemical Injection Pumps (Counts) 17,805 58,094 NA 84,538 84,061 82,948 83,249 Total Emissions (MT) 27,727 90,889 126,715 126,226 124,285 124,864 121,398 Previous Estimated Emissions 29,207 96,006 131,488 130,624 128,687 128,876 NA Percent Change in 5% - 5 % - 4 % - 3 % - 3 % - 3 % NA Emissions ( ) NA Not Applicable Liquids Unloading EPA updated its estimates for liquids unloading to use data from GHGRP. For more information, please see the from the previous Inventory that 56 2017 Production Memo . To develop this estimate, EPA retained the assumption iquids unloading (total percent of wells that vent for liquids unloading and wells percent of all gas wells conduct l - emitting systems)) over the time series (developed from that do not vent for liquids unloading (i.e., use of non API/ANGA 2012 ). EPA also retained the assumption that in 199 0, all of the 56 percent of wells with liquids unloading issues vent without plunger lifts. For the years 2011 to 2015, EPA applied the 2015 GHGRP fraction of - specific fractions of wells venting with gas wells that vent for liquids unloading (16.8 percent), and applied year plunger lifts and wells venting without plunger lifts. For years 1991 to 2010, EPA interpolated from the percentages of wells in each category for 1990 to 2011. For all years of the time series, EPA applied average EFs calculated from 11 to 2015 GHGRP data. The activity data assumptions and emission factors were developed and applied at the 20 national level, wherea the previous year’s Inventory calculated emissions with regional factors. Stakeholder s the Inventory and on the feedback on the public review draft of M emo support the use of GHGRP 2017 Production data to update this category, but suggested that the method be applied at the regional level. See Planned Improvements. an average decrease of 344,374 MT or 4 8 percent over The recalculation for liquids unloading emissions resulted in 6 the time series. The decrease in calculated emissions is much smaller in recent years (e.g., 1 0 percent for 201 7 percent for 1990 through 1995). through 2014), than earlier years of the time series (e.g., 5 3 - 59 : National Liquids Unloading Activity Data by Category and National Emissions Table (Metric Tons CH ) 4 Activity Data/Emissions 1990 2005 2011 2012 2013 2014 2015 Wells Venting w/o 34,265 73,436 28,366 31,309 113,978 33,646 Plunger Lifts (Counts) 28,634 Emissions (w/o Plunger) (MT) 352,138 226,881 87,638 96,729 105,862 103,951 88,464 Venting With Wells 26,237 45,535 Plunger Lifts (Counts) 42,307 - 38,219 39,176 42,166 Emissions (w/ Plunger) 120,673 112,114 109,376 121,076 (MT) - 75,085 130,313 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 86 3

201 Total Emissions (MT) 301,966 217,951 217,805 352,138 215,238 216,065 209,138 Previous Estimated 805,883 706,101 266,613 265,142 260,497 260,644 NA Emissions (MT) Percent Change in - 5 6 % - 57 % - 18% - 18% 17% - 17% - Emissions NA NA Not Applicable ) ( Boosting Episodic Emissions Gathering and EPA applied a factor developed in the Marchese study (37 metric tons CH per station) to calculate emissions from 4 gathering and boosting station episodic events , a source which was removed from the Inventory with last year’s up date to use GHGRP data (which only included production sites) for most categories, and Marchese et al. for gathering stations. To include an estimate for this source consistent with the estimate for gathering stations, EPA added information from the Marche se study . For more information, please see the 2017 Production Memo . This value was applied to all stations for each year of the time series. Stakeholder feedback on the public review draft of 2017 Production M the Inventory and on the support the use of future GHGRP data on gathering systems emo instead of the Marchese study data, and not including the estimate in this year’s Inventory. See Planned Improvements. National Gathering and Boosting Table - 60 : 3 Episodic Emission Activity Data (Number of Stations) and National Emissions (Metric Tons CH ) 4 2011 2012 Activity Data/Emissions 1990 2005 2015 2013 2014 4,549 2,565 2,968 4,246 4,638 5,034 5,276 Gathering Stations (counts) Total Emissions (MT) 94,905 109,816 195,212 157,102 168,313 171,606 186,258 Previous Estimated Emissions NA NA NA NA NA NA NA Percent Change in Emissions NA NA NA NA NA NA NA NA Not Applicable ) ( Gas STAR Reductions The production segment estimates include several sources that are calculated with potential emissions approaches, and therefore Gas STAR reductions are subtracted from the production segment estimates. Many of the activities oduction segment are cross - cutting and apply to more than one emissions source and reported to Gas STAR in the pr therefore cannot be assigned to one emissions source, but instead are included in an “other reductions” category. As many sources in production are now calculated with net f actor approaches, to address potential double - counting of reductions, a scaling factor is applied to the “other reductions” to reduce this reported amount based an estimate of the fraction of those reductions that occur in the sources that are now calculat ed using net emissions approaches. This fraction was developed by dividing the net emissions from sources with net approaches, by the total calculated (non - gathering) production segment emissions (without deducting the Gas STAR reductions). The fractions were recalculated this year to take into account that tanks are now calculated with net emissions approaches, and to address two minor errors in the previous calculation: 1). gas engine emissions , gathering pipeline emissions, and compressor start emission s were incorrectly included in the “potential emission” sources even though they have corresponding source - specific Gas STAR reductio ns and 2). reductions for 1990 to 1992 were not zeroed out, which would double count reductions if they are already include d in the GRI data set. The effect of these changes is a decrease in the fraction of emissions that are calculated with a potential emissions approach, and therefore a decrease in the fraction of “other reductions” that are applied in the production segmen t. The update results in a decrease in applied “other reductions” of an average of 1 MMT CO Eq. per year over the time series. 2 support using emo M Stakeholder feedback on the public review draft of the Inventory and on the 2017 Production ions data only where potential emissions are calculated, and removing them where they create Gas STAR reduct counting of reductions. - potential double 87 - 3 Energy

202 Table 3 : Production Segment Gas STAR “Other Reductions” Data (Metric Tons CH 61 ) and - 4 Scaling Factors (fraction) 1990 2005 2011 2012 2013 2014 2015 Total Calculated “Other NA 644,260 609,537 646,214 690,289 Reductions” 734,363 239,689 0.50 NA 0.28 0.27 0.27 0.28 Scaling Factor 0.30 “Other Reductions” 178,008 174,387 166,871 120,082 195,908 0 219,073 applied in Inventory Previous Total Calculated 656,509 3,517 621,786 251,938 658,463 “Other Reductions” 702,537 NA Previous Scaling Factor 1.00 0.62 0.44 0.45 0.46 0.49 NA Previous “Other Reductions” applied in Inventory 155,687 3,517 289,932 280,909 305,749 341,687 NA NA Not Applicable ) ( Processing This section references the memo Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990 2015: Updates for - 171 . Natural Gas Systems Processing Segment Emissions (2017 Processing Memo) Overall, recalculations for the processing segment resulted in a decrease of 12.8 MMT CO Eq., or 54 percent 2 comparing the 2014 value from the previous Inventory to this public review draft Inventory. Over the time series, the average change was a decrease of 2 9 percent. While stakeholders generally support the use of GHGRP data for the update, one stakeholder suggested using the Marchese et al. data set total average emi ssions per plant to adjust the source - specific GHGRP data upwards. Another stakeholder recommended the use of the GHGRP data and not the Marchese study. The Inventory update uses GHGRP data, unadjusted as described below. For more information, please se e the 2017 Processing Memo . Station Fugitives, Compressors, Flares a nd Dehydrators GHGRP data were used to update the estimates for station fugitives, compressors, flares, and dehydrators. For more information, see the 2017 Processing Memo. Linear interpolation was used to create time series consistency between ctors and activity factors (1990 through 1992) that generally rely on data from GRI/EPA earlier years’ emission fa 1996 and the GHGRP emission and activity factors for recent years. However, the plant fugitive emission factors in previous Inventories included plant fugitives but not compressor fugitives, and separate emission factors were applied for compressor emissions (including compressor fugitive and vented sources). There is also some overlap between those categories and the flare and dehydrator categories. Because of these dif ferences, the two sets of EPA’s GHGRP) cannot be directly compared. For the emission factors (GRI/EPA and factors calculated from - purpose of interpolating for the time series, EPA developed plant level emission factors for processing stations that include plant and compressor fugitive sources, compressor vented sources, flares, and dehydrators. The previous Inventory emission factors were used for 1990 through 1992; emission factors from EPA’s GHGRP were used for 2011 through 2015. Emission factors for 1993 through 2010 were developed through linear interpolation. EPA incorporated GHGRP average values of reciprocating and centrifugal compressors per processing plant, using year 2015 data. These values were applied for 2011 through 2015. GHGRP data for 2011 through 2015 were used to specific splits between centrifugal compressor seal types (wet versus dry seals). GHGRP year 2015 develop year - data were used to develop emission factors on a per plant basis for fugitives, flares, and dehydrators, and a per - - compr essor basis for compressors. Emission factors for dry seal centrifugal compressors were developed using GHGRP data supplemented with the previous Inventory emission factor for dry seal emissions. A stakeholder comment recommended using only 2013 through 2 015 GHGRP data to develop the emission factors for compressors and to continue to update emission estimates for this source using annual GHGRP data. EPA will consider this update for future Inventories. 171 1990 - - information See < https://www.epa.gov/ghgemissions/nat ural - gas - and - petroleum - systems - ghg - inventory - additional - 2015 >. ghg - - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 88 3

203 earlier years’ per plant compressor count estimates (1990 to 1992) In order to create time series consistency between and the most recent years’ per plant compressor count estimates (2011 to 2015) that were calculated using GHGRP data, compressor counts for the years 1993 through 2010 were calculated usin g linear interpolation between the data endpoints of 1992 and 2011. The overall impact of using revised emissions data and activity data from GHGRP is a decrease in emissions EPA’s sions decrease due to use of revised for fugitives and compressors. For the year 2014, the calculated CH emis 4 emission factors and activity data for processing plant fugitives, compressor venting, flares, and dehydrators is MMT CO Eq. approximately 16.5 2 Gas Engines and Turbines EPA’s T he estimates for gas engines and gas turbines w GHGRP. For more ere updated to incorporate data from information, please see the 2017 Processing Memo . GHGRP data were used to develop an updated value for million hours (MMHPhr) per plant for both gas engines and gas turbines. These values were applied to plant horsepower - counts for years 2011 to 2015. The previous estimates of MMHPhr per plant were reta ined for 1990 through 1992, and values for 1993 to 2010 were developed by linear interpolation between the 1992 and 2011 values. EPA retained the previous Inventory emission factor and applied it for all years of the time series. The recalculation for gas 1 1 9 0 MT, or engines resulted in an average increase in the estimate of percent over the time series. The 9 4, 15 percent over the time 657 recalculation for gas turbines resulted in an average decrease in the estimate of MT, or series. Blowdown Venting T estimate for blowdown venting was updated to incorporate data from GHGRP. For more information, please he . A per - plant emission factor was developed from 2015 GHGRP data, and applied to see the 2017 Processing Memo he previous emission factors were retained for 1990 through 1992, and plant counts for years 2011 through 2015. T values for 1993 through 2010 were developed by linear interpolation between the 1992 and 2011 values. The recalculation resulted in an average decrease in the estimate of 7,769 MT or 17 percent over the time series. 3 - 62 : CH Emissions from Processing Plants (Metric Tons CH Table ) 4 4 Activity Data/Emissions 1990 2005 2011 2012 2013 2014 2015 Plant Total Emissions (MT) 633,867 245,798 143,187 (Overlapping Sources) 143,341 153,160 157,063 156,252 Plant Fugitives (MT) NA NA 14,625 14,625 15,687 16,097 16,097 Reciprocating Compressors (MT) NA NA 64,413 69,089 70,896 70,896 64,413 Centrifugal Compressors 21,428 23,143 (Wet Seals) (MT) NA 22,767 NA 22,061 22,387 Centrifugal Compressors 6,787 NA 6,959 (Dry Seals) (MT) NA 7,936 8,260 9,165 Flares (MT) NA NA 19,776 19,776 21,212 21,767 21,767 NA NA 15,353 15,353 16,468 16,899 16,899 Dehydrators (MT) Gas Engines (MT) 137,102 168,297 211,002 211,002 226,322 232,241 232,241 3,424 Gas Turbines (MT) 3,861 3,883 3,883 4,165 4,274 4,274 16,494 12,267 13,134 13,134 14,088 14,456 14,456 AGR Vents (MT) Pneumatic Controllers (MT) 2,414 1,796 2,116 1,923 1,923 2,062 2,116 Blowdowns/Venting (MT) 59,507 34,586 32,251 32,251 34,593 35,497 35,497 Total Processing Emissions 434,390 445,648 444,837 (MT) 853,245 466,168 405,380 405,534 ) Not Applicable ( NA 89 - 3 Energy

204 Table - : Previous (last year’s) 1990 to 2014 Inventory Estimates for Processing Segment 3 63 ) Emissions (Metric Tons CH 4 1990 2005 2011 2012 2013 2014 2015 Activity Data/Emissions Previous Plant Total Emissions (Overlapping 633,867 621,625 761,618 Sources) (MT) 793,031 800,622 843,513 NA Previous Plants (MT) 42,295 31,457 33,681 33,681 36,126 37,126 NA Previous Recip. 324,939 327,869 420,871 442,077 445,551 473,829 Compressors (MT) NA Previous Centrifugal Compressors (Wet 240,293 229,237 236,115 Seals) (MT) 237,683 237,940 240,031 NA Previous Centrifugal Compressors (Dry Seals) (MT) - 6,483 36,835 43,755 44,889 54,117 NA Previous Kimray Pumps 5,005 3,712 4,764 (MT) 3,678 5,044 5,364 NA Previous Dehydrator Vents (MT) 22,662 22,866 29,352 30,831 31,073 33,045 NA Previous Gas Engines (MT) 137,102 138,338 177,578 186,526 187,991 199,923 NA 5,630 5,294 3,861 NA Previous Gas Turbines (MT) 5,253 3,896 5,001 16,494 12,267 13,134 13,134 14,088 14,478 NA (MT) Previous AGR Vents Previous Pneumatic 2,414 Controllers (MT) 1,796 1,923 1,923 2,062 2,119 NA Previous Blowdowns/Venting (MT) 59,507 44,259 47,387 47,387 50,827 52,235 NA - Total Potential Previous Emissions (MT) 853,245 822,180 1,006,640 1,047,252 1,060,884 1,117,897 NA Gas STAR Previous - NA Reductions (MT) (1,488) (155,501) (140,797) (140,368) (140,449) (140,744) Regulatory - Previous - (12,101) (16,316) Reductions (MT) (16,444) (17,488) (15,533) Previous - Total Net 654,578 903,697 850,739 Emissions (MT) 890,488 851,757 959,613 NA NA ( ) Not Applicable Note: Parentheses indicate negative values. Gas STAR Reductions in the Processing Segment EPA used new data from EPA’s GHGRP to calculate emission factors that account for adoption of control technologies and emission reduction practices. To develop estimates over the time series, EPA retained emission factors from the EPA/GRI study for early t ime series years (1990 - 1992), applied updated emission factors in recent years (e.g., 2011 forward), and used interpolation to calculate emission factors for intermediate years. This approach oluntary reductions (derived from Gas STAR data) results in net emissions calculated for each time series year. V and regulatory reductions (based on NESHAP implementation) are inherently taken into account with this reproach; therefore, it is no longer necessary to retain these “reduction” line items. EPA has removed the Gas STAR reductions for the processing segment. Over the 1990 to 201 4 time series, annual Gas STAR reductions averaged 2.1 Eq. MMTCO 84,583 MT CH , or 4 2 port using Stakeholder feedback on the public review draft of the Inventory and on the 2017 Processing memo sup Gas STAR reductions data only where potential emissions are calculated, and removing them where they create counting of reductions. - potential double Transmission and Storage storage segment, recalculations due to Although there were no methodological updates to the transmission and updated data (e.g., GHGRP station counts, the GHGRP split between dry and wet seal centrifugal compressors, and - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 90 3

205 GHGRP pneumatic controller data) impacted emission estimates, resulting in an average increase in calcul ated , or less than 0.01 percent. emissions over the time series from this segment of around 24 metric tons CH 4 Additional information on inclusion of the Aliso Canyon emissions can be found in the Methodology section above 172 Storage Memo and in the and not in the Recalculation Discussion section as it did not 2017 Transmission and involve recalculation of a previous year of the Inventory. Distribution Although there were no methodological updates to the distribution segment, recalculations due to updated dat a (e.g., GHGRP M&R station counts) impacted emission estimates, resulting in an average de crease in calculated emissions over the time series from this segment of around 664 metric tons CH , or 0.1 percent. 4 Planned Improvements Pla ns for 2018 Inventory (1990 through 2016) and Future Inventories CO Data Update 2 In this year’s Inventory, EPA has held constant the CO values from the previous Inventory as it assesses 2 improvements to the CO to improve ing estimates. EPA is review CO updates data from GH GRP and consider ing 2 2 consistency of data sources and methods between the CH emission estimates (which have been updated in recent 4 years) and the CO f the emission estimates in Natural Gas Systems. EPA has conducted a preliminary assessment o 2 data and will seek stakeholder feedback on the draft assessment and options for updates. Using GHGRP data to CO 2 update CO Systems , primarily due to a potential crease in the estimate of CO from Natural Gas could result in a de 2 2 flaring for both Natural Gas change in where CO from flaring is estimated -- currently, CO onshore production from 2 2 and Petroleum Systems is included in Natural Gas Systems. Uncertainty As noted in the Uncertainty discussion, t natural gas systems emission he most recent uncertainty analysis for the estimates in the Inventory was conducted for the 1990 to 2009 Inventory that was released in 2011. Since the to reflect improved data and analysis was last conducted, several of the methods used in the Inventory have changed anges in industry practices and equipment. In addition, new studies and other data sources offer improvement to ch is preparing a draft update understanding and quantifying the uncertainty of some emission source estimates. EPA to the uncertainty analysis co nducted for the 2011 Inventory to reflect the new information and will seek stakeholder feedback on the draft analysis as part of the development of the next (i.e., 1990 through 2016) Inventory. Gas Wells Abandoned included in the Inventory. EPA is seeking emission factors and national activity Abandoned wells are not currently in future data available to calculate these emissions. Stakeholder comments supported including this source category Inventories , but noted that currently data are limited, an d suggested reviewing data that will become available in the future. EPA has identified studies with data on abandoned wells (Townsend - Small et al. 2016, Kang et al. 2016, Brandt et al. 2014), and is considering including an estimate for this source in fut ure Inventories. A preliminary for emissions from abandoned gas wells estimate , based the national emission estimate values from Townsend - Small et al., the range of abandoned well counts in - Small et al. and Brandt. et al., and the 1990 split Townsend betwe en in oil and gas wells in total producing wells population, is around 0.9 to 1.2 MMT CO seeks e. EPA 2 stakeholder feedback on abandoned wells. 172 - - 1990 See . - 91 - 3 Energy

206 Anomalous Leak Events Inventory This Inventory includes an estimate for the 2015 portion of the Aliso Canyon leak event . Next year’s will incorporate the 2016 emission estimate for the event. of around 21,000 MT CH EPA seeks information on other 4 large emissions events. For example, in early 2017 a leaking underwater gas pipeline was identified in Alaska. Early estimates of the leak developed from February through March range from around 2 to 6 tons of CH emitted per 4 173 EPA will consider including an emission estimate for this leak in the 1990 - 2017 Inventory, to be published day. in 2019. tional Data that Could Inform the Inventory Upcoming Data, and Addi its continue to review data available from GHGRP, in particular new data on gathering and boosting EPA will stations, gathering pipelines, and transmission pipeline blowdowns and new well - ailable in specific information, av EPA will 2017 for the first time. EPA will consider revising its method to take into account the new GHGRP data. also assess GHGRP data for regional variation in liquids unloading practices and emissions and assess whether to apply GHGRP based emissions and activity factors at a regional level in the Inventory. Stakeholder comments on - the Inventory included a number of options for use of GHGRP processing segment data, including use of a 3 - year rolling average to develop emission factors. EPA wi ll consider other approaches for use of GHGRP data in the processing segment. 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), 174 industrial meters (distribution segment), an d sources within the gathering and storage segments. 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 or “superemitters,” ( e.g. identification of emission sources and , information on frequency of high emitters) as recommended in stakeholder comments. continues to seek new data that could be used to assess or update the estimates in the Inventory. For EPA also example, stakeho lder comments have highlighted areas where additional data that could inform the Inventory are limited or unavailable: currently Tank malfunction and control efficiency data. See Tanks in Recalculations Discussion. • • Activity data and emissions data for pr oduction facilities that do not report to GHGRP. See, for example, Equipment Leaks in Recalculations Discussion. • Natural gas power plant leak data. One stakeholder noted a recent study (Lavoie et al. 2017) that 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. • Abandoned well activity and emissions data. See above section in Planned Imp rovements. EPA will continue to seek available data on these and other sources as part of the process to update the Inventory. 173 See < http://dec.alaska.gov/spar/ppr/respons e/sum_fy17/170215201/170215201_index.htm > 174 - mitigate - and - quantify See < https://www.energy.gov/under - secretary - science - and - energy/articles/doe - announces - 13 - million - >. methane - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 92 3

207 3.8 Energy Sources of Indirect Greenhouse Gas Emissions related activities generate emissions of energy In addition to the main greenhouse gases addressed above, many - - CH volatile Total emissions of nitrogen oxides (NO ), carbon monoxide (CO), and non indirect greenhouse gases. x 4 5 organic compounds (NMVOCs) from energy are reported in Table 3 - 64 . - related activities from 1990 to 201 - Table : NO 3 , CO, and NMVOC Emissions from Energy - Related Activities (kt) 64 x 1990 2005 2011 Gas/Activity 2012 2013 2014 2015 NO 21,106 16,602 11,796 11,271 10,747 10,161 9,971 x Mobile Fossil Fuel Combustion 10,295 7,294 6,871 6,448 6,024 5,417 10,862 Stationary Fossil Fuel Combustion 5,858 3,807 3,655 3,504 3,291 3,061 10,023 Oil and Gas Activities 139 321 622 663 704 745 745 Waste Combustion 128 73 82 91 100 100 82 a International Bunker Fuels 1,704 1,956 1,398 1,139 1,138 1,225 1,553 CO 125,640 64,985 44,088 42,164 40,239 38,315 36,348 Mobile Fossil Fuel Combustion 119,360 58,615 38,305 36,153 34,000 31,848 29,881 Stationary Fossil Fuel Combustion 5,000 4,648 4,027 3,884 3,741 3,741 4,170 Waste Combustion 978 1,403 1,947 1,947 1,003 1,318 1,632 Oil and Gas Activities 723 610 666 318 780 780 302 a International Bunker Fuels 103 133 137 133 129 135 141 NMVOCs 12,620 7,191 7,759 7,558 7,357 7,154 6,867 Mobile Fossil Fuel Combustion 5,724 4,562 10,932 4,243 3,924 3,605 3,318 Oil and Gas Activities 2,786 554 510 2,517 2,651 2,921 2,921 Stationary Fossil Fuel Combustion 539 912 716 599 569 507 507 Waste Combustion 121 81 94 108 241 121 222 a International Bunker Fuels 54 57 51 46 41 42 47 a These values are presented for informational purposes only and are not included in totals. Totals may not sum due to independent rounding. Note: Methodology Emission estimates for 1990 through 201 5 were obtained from data published on the National Emission Inventory (NEI) Air Pollutant Em ission Trends web site (EPA 2016 ), and disaggregated based on EPA (2003). Emission e sources are held constant from 2013 for non - electric generating units (EGU) and non estimates for 2012 mobil and - 2011 in EPA (201 6 ). Emissions were calculated either for individual categories or for many categories 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 applications from various agencies. Activity data were used in conjunction with emission factors, which together relate the quantity of emissions to the activity. Emission factors are generally available from the EPA’s Compilation of Air Pollutant Emission Factors, AP - 42 (EPA 1997). The EPA currently derives the overall emission control efficiency of a source category from a variety of information sources, including published reports, the 1985 National Acid Precip itation and Assessment Program emissions inventory, and other EPA databases. Uncertainty and Time - Series Consistency Uncertainties in these estimates are partly due to the accuracy of the emission factors used and accurate estimates of activity data. A qua ntitative uncertainty analysis was not performed. - Methodological recalculations were applied to the entire time series to ensure time series consistency from 1990 Methodology section, . Details on the emission trends through time are described in more detail in the 5 through 201 above. 93 - 3 Energy

208 3.9 International Bunker Fuels (IPCC Source Category 1: Memo Items) Emissions resulting from the combustion of fuels used for international transport activities, termed international in national emission totals, but are reported separately based upon bunker fuels under the UNFCCC, are not included location of fuel sales. The decision to report emissions from international bunker fuels separately, instead of allocating them to a particular country, was made by the Intergovernmental N egotiating Committee in establishing 175 These decisions are reflected in the IPCC methodological the Framework Convention on Climate Change. guidance, including IPCC (2006), in which countries are requested to report emissions from ships or aircraft that art from their ports with fuel purchased within national boundaries and are engaged in international transport dep 176 separately from national totals (IPCC 2006). 177 Two transport modes are addressed under the IPCC definition of international bunker fuels: aviatio n and marine. , Greenhouse gases emitted from the combustion of international bunker fuels, like other fossil fuels, include CO 2 O for aviation transport modes. Emissions from ground and N CH and N O for marine transport modes, and CO 2 2 4 2 s — transport activitie — even when crossing international borders are allocated to the by road vehicles and trains country where the fuel was loaded into the vehicle and, therefore, are not counted as bunker fuel emissions. The 2006 IPCC Guidelines distinguish between different modes of air traffic. Civil aviation comprises aircraft used for the commercial transport of passengers and freight, military aviation comprises aircraft under the control of national armed forces, and general aviation applies to recreational and small cor porate aircraft. The 2006 IPCC Guidelines further define international bunker fuel use from civil aviation as the fuel combusted for civil (e.g., commercial) aviation purposes by aircraft arriving or departing on international flight segments. However, as mentioned above, and in keeping with the 2006 IPCC Guidelines , only the fuel purchased in the United States and off (i.e., departing) from the United States are reported here. The standard fuel used for civil used by aircraft taking - 178 pe jet fuel, while the typical fuel used for general aviation is aviation gasoline. aviation is kerosene - ty Emissions of CO from aircraft are essentially a function of fuel use. Nitrous oxide emissions also depend upon 2 e (i.e., take - off, climb, cruise, decent, and landing). Recent engine characteristics, flight conditions, and flight phas is emitted by modern engines (Anderson et al. 2011), and as a result, CH data suggest that little or no CH 4 4 emissions from this category are considered zero. In jet engines, N roduced by the oxidation of O is primarily p 2 atmospheric nitrogen, and the majority of emissions occur during the cruise phase. International marine bunkers comprise emissions from fuels burned by ocean - going ships of all flags that are engaged in international transport. Ocean - going ships are generally classified as cargo and passenger carrying, military (i.e., U.S. Navy), fishing, and miscellaneous support ships (e.g., tugboats). For the purpose of estimating greenhouse gas emissions, international bunker fuels are solely related to cargo and passenger carrying vessels, which is the largest of the four categories, and military vessels. Two main types of fuels are used on sea - going vessels: distillate diesel fuel and residual fuel se gas emitted from marine shipping. oil. Carbon dioxide is the primary greenhou Overall, aggregate greenhouse gas emissions in 2015 from the combustion of international bunker fuels from both 3 Eq., or 7.0 percent above emissions in 1990 (see Table aviation and marine activities were 111.8 MMT CO - 65 2 66 Table 3 - ). and Emissions from international flights and international shipping voyag es departing from the United States have increased by 88.8 percent and decreased by 40.6 percent, respectively, since 1990. The majority of these 175 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. 1c). 176 Note that the definition of international bunker fuels used by the UNFCCC differs from that used by the International Civil Aviation Organization. 177 nal Civil Aviation Most emission related international aviation and marine regulations are under the rubric of the Internatio 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). 178 Naphtha type jet fuel was used in the past by the military in turbojet and turboprop aircraft engines. - - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 94 3

209 emissions were in the form of CO (from marine transport modes) and N O were ; however, small amounts of CH 2 2 4 so emitted. al - , and N : CO 3 , CH Table 65 O Emissions from International Bunker Fuels (MMT CO Eq.) 2 4 2 2 1990 2005 2011 Gas/Mode 2012 2013 2014 2015 CO 110.8 103.5 113.1 111.7 105.8 99.8 103.2 2 Aviation 65.7 60.1 38.0 64.8 64.5 69.4 71.8 Commercial 30.0 55.6 61.7 61.4 62.8 66.3 68.6 Military 8.1 4.5 3.1 3.1 2.9 3.1 3.2 Marine 33.8 53.0 46.9 41.3 34.1 65.4 38.9 CH 0.2 0.1 0.1 0.1 0.1 0.1 0.1 4 a Aviation + + + + + + + Marine 0.1 0.1 0.1 0.2 0.1 0.1 0.1 N O 0.9 0.9 1.0 1.0 0.9 0.9 0.9 2 Aviation 0.6 0.6 0.4 0.6 0.6 0.7 0.7 Marine 0.2 0.4 0.4 0.3 0.5 0.2 0.3 100.7 Total 104.5 114.2 112.8 106.8 104.2 111.8 + Does not exceed 0.05 MMT CO Eq. 2 a emissions from aviation are estimated to be zero. CH 4 Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions. 3 Table : CO - , CH 66 , and N O Emissions from International Bunker Fuels (kt) 2 4 2 Gas/Mode 2005 2011 2012 1990 2013 2014 2015 103,463 113,139 111,660 105,805 CO 99,763 103,201 110,751 2 60,125 64,790 64,524 38,034 65,664 69,411 71,805 Aviation 65,429 53,014 46,870 41,281 Marine 34,099 33,791 38,946 4 7 5 5 CH 3 3 3 4 a 0 0 0 0 0 Aviation 0 0 Marine 7 5 5 4 3 3 3 N O 3 3 3 3 3 3 3 2 2 2 2 Aviation 1 2 2 2 Marine 2 1 1 1 1 1 1 a CH emissions from aviation are estimated to be zero. 4 Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions. Methodology Emissions of CO were estimated by applying C content and fraction oxidized factors to fuel consumption activity 2 data. This approach is analogous to that described under Section 3.1 – CO from Fossil Fuel Combustion. Carbon 2 content and fraction oxidized factors for jet fuel, distillate fuel oil, and residual fuel oil were taken directly from EIA versions were taken from and are presented in Annex 2.1, Annex 2.2, and Annex 3.8 of this Inventory. Density con Chevron (2000), ASTM (1989), and USAF (1998). Heat content for distillate fuel oil and residual fuel oil were taken from EIA (2016) and USAF (1998), and heat content for jet fuel was taken from EIA (2017). A complete description of the methodology and a listing of the various factors employed can be found in Annex 2.1. See Annex 3.8 for a specific discussion on the methodology used for estimating emissions from international bunker fuel use by the U.S. military. Emission estimates f or CH O were calculated by multiplying emission factors by measures of fuel and N 2 4 consumption by fuel type and mode. Emission factors used in the calculations of CH and N O emissions were 2 4 obtained from the 2006 IPCC Guidelines (IPCC 2006). For aircraft e missions, the following value, in units of grams O (IPCC 2006). For marine vessels of pollutant per kilogram of fuel consumed (g/kg), was employed: 0.1 for N 2 consuming either distillate diesel or residual fuel oil the following values (g/MJ), were employed and : 0.32 for CH 4 O. Activity data for aviation included solely jet fuel consumption statistics, while the marine mode 0.08 for N 2 included both distillate diesel and residual fuel oil. 95 - 3 Energy

210 Activity data on domestic and international aircraft fuel consumption were developed by the U.S. Federal Aviation - informed data from the FAA Enhanced Traffic Management System (ETMS) for Administration (FAA) using radar up 1990, 2000 through 2015 as modeled with the Aviation Environmental Design Tool (AEDT). This bottom - oach is built from modeling dynamic aircraft performance for each flight occurring within an individual appr calendar year. The analysis incorporates data on the aircraft type, date, flight identifier, departure time, arrival time, - world flight trajectories. To generate results departure airport, arrival ai rport, ground delay at each airport, and real informed aircraft data is correlated with engine and aircraft performance for a given flight within AEDT, the radar - data to calculate fuel burn and exhaust emissio ns. Information on exhaust emissions for in - production aircraft engines comes from the International Civil Aviation Organization (ICAO) Aircraft Engine Emissions Databank - up approach is in accordance with the Tier 3B method from the 2006 (EDB). This bottom IPCC Guidelines (IPCC 2006). estimates for 1990 and 2000 through 2015 are obtained from FAA’s AEDT model (FAA International aviation CO 2 - informed method that was used to estimate CO emissions for commercial aircraft for 1990, and 2017). The radar 2 2000 through 2015 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 2 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 (Ins tallations and Environment), DoD. Estimates of the percentage of each Service’s total operations that were international operations were developed by DoD. Military aviation bunkers included international operations, operations conducted from naval vessels at sea, and operations conducted from U.S. installations principally over international water in direct support of military operations at sea. Military aviation bunker fuel emissions were estimated using military fuel and operations data synthesized from u npublished data from DoD’s Defense Logistics Agency Energy (DLA Energy 201 6 ) . 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 - 67 . See Annex 3.8 for additional discussion of military data. Ac tivity 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 2015, 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 (2016). 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 Table the U.S. Navy. These fuel consumption estimates are presented in - 3 68 . Table 3 67 : Aviation Jet Fuel Consumption for International Transport (Million Gallons) - 1990 Nationality 2005 2011 2012 2013 2014 2015 U.S. and Foreign Carriers 3,222 5,983 6,634 6,604 6,748 7,126 7,383 U.S. Military 862 462 319 321 294 318 327 Total 4,084 6,445 6,953 6,925 7,042 7,445 7,711 Note: Totals may not sum due to independent rounding. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 96 3

211 Table 68 : Marine Fuel Consumption for International Transport (Million Gallons) - 3 1990 Fuel Type 2005 201 1 201 2 201 3 201 4 201 5 Residual Fuel Oil 4,781 3,881 2,537 3,463 3,069 2,466 2,718 Distillate Diesel Fuel & Other 617 444 393 280 235 261 492 U.S. Military Naval Fuels 522 471 382 381 308 331 326 3,081 3,536 Total 5,920 4,796 4,237 3,730 3,058 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 179 from domestic transport activities. For example, smaller aircraft on shorter routes often carry sufficient fuel to complete several fligh t 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 da ta 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 Forc e 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, There are uncertainties in aircraft operations and/or given fuel to other ships, aircra ft, governments, or other entities. 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 ba sed 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 voyages that would be This approach does not capture some while underway and fuel used while not underway. Conversely, emissions from fuel used while not underway preceding classified as domestic for a commercial vessel. an international voyage are reported as domestic rather than international as would be done for a commercial vessel. ere is uncertainty associated with ground fuel estimates for 1997 through 2001. Small fuel quantities may have Th 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 t ype, 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 f or 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 related to the various uncertainties has not been calculated, but is believed to be small. The uncertainties associated with future military bunker fuel emission estimates could be reduced through additional data collection. Although aggregate fuel consumption data have been used to estimate emissions from aviation, the recommended method for estimating emissions of gases other than CO 2006 IPCC Guidelines in the (IPCC 2006) is to use data by 2 ideally, movement data to better differentiate between specific aircraft type, number of individual flights and, domestic and international aviation and to facilitate estimating the effects of changes in technologies. The IPCC also 179 See uncertainty discussions under Carbon Dioxide Emissions from Fossil Fuel Combustion. 97 - 3 Energy

212 recommends that cruise altitude emissions be estimated separately using fuel cons umption data, while landing and 180 off (LTO) cycle data be used to estimate near ground level emissions of gases other than CO . - - take 2 ) data on marine vessel fuel consumption There is also concern regarding the reliability of the existing DOC (201 7 at U.S. customs stations due to the significant degree of inter - annual variation. reported series consistency from 1990 - Methodological recalculations were applied to the entire time series to ensure time through . Details on the emission trends through time are described in more detail in the Methodology section, 2015 above. QA/QC and Verification - A source specific QA/QC plan for international bunker fuels was developed and implemented. This effort included a Tier 1 analysis, as well as portions of a Tier 2 analysis. The Tier 2 procedures that were implemented involved checks specifically focusing on the activity data and emission factor sources and methodology used for estimating CO , CH , and N O from international bunker fuels in the United States. Emission totals for the different sectors and 2 4 2 fuels were compared and trends were investigated. No corrective actions were necessary. Planned Improvements data from a broader range of domestic and international sources for bunker fuels, The feasibility of including ing data from studies such as the Third IMO GHG Study 2014 includ (IMO 2014), is being considered. 3.10 Wood Biomass and Biofuels Consumption (IPCC Source Category 1A) The combustion of biomass fuels such as wood, charcoal, and wood waste and biomass based fuels such as ethanol , - biogas, and biodiesel generates CO In line with the in addition to CH O already covered in this chapter. and N 2 4 2 emissions from biomass reporting requirements for inventories submitted under the UNFCCC, CO combustion 2 emissions and are not directly included in the energy sector have been estimated separately from fossil fuel CO 2 contributions to U.S. totals. In accordance with IPCC methodological guidelines, any such emissions are calculated by accounting fo r net carbon (C) fluxes from changes in biogenic C reservoirs in wooded or crop lands. For a more Land Use, Land - Use Change, and Forestry chapter complete description of this methodological approach, see the 6 ), which accounts for the contribution of any resulting CO (Chapter emissions to U.S. totals within the Land Use, 2 Land Use Change, and Forestry sector’s approach. - In 2015 , total CO emissions from the burning of woody biomass in the industrial, residential, commercial, and 2 ) (see electricity generation secto rs were approximately 198.7 MMT CO Table Eq. ( 198,723 kt Tabl e 3 - 69 and 2 3 70 ). As the largest consumer of woody biomass, the industrial sector was responsible for 61.2 percent of the CO - 2 emissions from this source. The residential sector was the second largest emitter, constituting 22.4 percent of the total, while the commercial and electricity generation sectors accounted for the remainder. 180 U.S. aviation emission estimates for CO, NO d by EPA’s National Emission Inventory (NEI) Air , and NMVOCs are reporte x 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 e missions, which are more relevant for air quality evaluations. These estimates also include both domestic and international flights. Therefore, estimates reported un der the - Mobile Combustion section overestimate IPCC missions by including landing and , and NMVOC e defined domestic CO, NO x - off (LTO) cycles by aircraft on international flights, but underestimate because they do not include emissions from aircraft take on domestic flight segments at cruising altitudes. The estimates in Mobile Combustion are a lso likely to include emissions from - ocean going vessels departing from U.S. ports on international voyages. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 98 3

213 Tabl e 69 : CO 3 Emissions from Wood Consumption by End - Use Sec tor ( MMT CO - Eq.) 2 2 - Use Sector 1990 2005 201 1 2012 2013 2014 2015 End Industrial 135.3 136.3 122.9 123.1 124.4 121.6 125.7 Residential 59.8 44.3 46.4 43.3 59.8 59.8 44.5 Commercial 7.2 7.1 6.3 7.2 7.6 7.5 6.8 19.6 Electricity Generation 19.1 18.8 13.3 21.4 25.9 25.1 Total 215.2 206.9 195.2 194.9 211.6 217.7 198.7 Note: Totals may not sum due to independent rounding. 3 - 70 ) Emissions from Wood Consumption by End - Use Sector ( kt Table : CO 2 End Use Sector 1990 2005 201 1 2012 2013 2014 2015 - Industrial 135,348 136,269 122,865 125,724 123,149 124,369 121,564 59,808 44,340 46,402 43,309 59,808 59,808 44,497 Residential Commercial 6,779 7,218 7,131 6,257 7,235 7,569 7,517 13,252 19,074 18,784 Generation 19,612 21,389 25,908 25,146 Electricity Total 215,186 206,901 195,182 194,903 211,581 217,654 198,723 Note: Totals may not sum due to independent rounding. consumption in the United States. Ethanol is The transportation sector is responsible for most of the ethanol currently produced primarily from corn grown in the Midwest, but it can be produced from a variety of biomass a 90 percent gaso feedstocks. Most ethanol for transportation use is blended with gasoline to create line, 10 percent by volume ethanol blend known as E - 10 or gasohol. trillion Btu of ethanol, and as a result, produced In 2015 , the United States consumed an estimated 1,153.1 CO Ethanol emissions. approximately 78.9 MMT ) of CO 72 Eq. ( 78,934 kt ) (see Table 3 - 71 and Table 3 - 2 2 significantly due to the favorable economics of blending ethanol production and consumption has grown since 1990 into gasoline and federal policies that have encouraged use of renewable fuels . - 71 : CO Emissions from Ethanol Consumption ( MMT Table CO 3 Eq.) 2 2 End - Use Sector 1990 2005 201 1 2012 2013 2014 2015 a 4.1 22.4 Transportation 71.5 73.4 74.9 75.9 71.5 Industrial 0.1 0.5 1.1 1.1 1.2 1.0 1.2 + 0.1 0.2 0.2 0.2 0.2 1.8 Commercial Total 4.2 22.9 72.9 72.8 74.7 76.1 78.9 + Does not exceed 0.05 CO MMT Eq. 2 a See Annex 3.2, Table A - 9 5 for additional information on transportation consumption of these fuels. Note: Totals may not sum due to independent rounding. Table 3 : CO - Emissions from Ethanol Consumption ( kt ) 72 2 End - Use Sector 1990 2005 201 1 2012 2013 2014 2015 a 4,136 22,414 71,537 Transportation 71,510 73,359 74,857 75,946 Industrial 56 468 970 1,146 1,142 1,202 1,203 34 60 198 175 183 249 1,785 Commercial Total 4,227 22,943 72,881 72,827 74,743 76,075 78,934 a See Annex 3.2, Table A - 9 for additional information on transportation consumption of these fuels. 5 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 but it can be produced from a variety of biomass 2017a) . Biodiesel is currently produced primarily from soybean oil , feedstocks including waste oils, fats and greases. level blends (less Biodiesel for transportation use appears in low - level blends (between 6 and 20 percent) with dies - than 5 percent) with diesel fuel, high el fuel, and 100 percent biodiesel (EIA 2017b). 99 - 3 Energy

214 In 2015 trillion Btu of biodiesel , and as a result, produced , the United States consumed an estimated 190.6 14.1 MMT CO approximately Eq. ( 14,077 kt ) (see Table 3 - 73 and Table 3 - 74 ) of CO emissions. Biodiesel 2 2 production and consumption has grown since 2001 due to the favorable economics of blending significantly uraged use of renewable fuels (EIA 2017b). There was no biodiesel into diesel and federal policies that have enco measured biodiesel consumption prior to 2001 EIA (2017a). - 73 : CO Table Emissions from 3 Consumption ( MMT CO Eq.) Biodiesel 2 2 - Use Sector 1990 2005 201 1 2012 2013 2014 2015 End a Transportation 0.0 0.9 8.3 8.5 13.5 13.3 14.1 Total 0.0 0.9 8.3 8.5 13.5 13.3 14.1 + Does not exceed 0.05 MMT CO Eq. 2 a See Annex 3.2, Table A - 9 5 for additional information on transportation consumption of these fuels. Note: Totals may not sum due to independent rounding. - 74 : CO Table Emissions from 3 Consumption ( kt ) Biodiesel 2 - Use Sector 1990 2005 201 End 2012 2013 2014 2015 1 a Transportation 0 856 8,349 8,470 13,462 13,349 14,077 8,470 0 856 8,349 13,462 13,349 14,077 Total a information on transportation consumption of these fuels. See Annex 3.2, Table A - 9 5 for additional Totals may not sum due to independent rounding. Note: 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 - 75 ), 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 consu mption, 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) It was assumed that emission estimates. was then applied to the resulting quantities of woody biomass to obtain CO 2 the woody biomass contains black liquor and other wood wastes, has a moisture content of 12 percent, and is calculated by with 100 percent efficiency. The emissions from ethanol consumption were converted into CO 2 applying an emission factor of 18. 7 MMT C/QBtu (EPA 2010) to U.S. ethanol consumption estimates that were 2017a Table 3 - 76 ). The emissions from biodiesel consumption were calculated provided in energy units (EIA ) (see 20.1 MMT C/QBtu (EPA 2010 ) to U.S. biodiesel consumption estimates that were by applying an emission factor of 2017a ) (see Table 3 - 77 ). provided in energy units (EIA 3 : Woody Biomass Consumption by Sector (Trillion Btu) 75 Table - End - Use Sector 1990 2005 201 1 2012 2013 2014 2015 1,441.9 1,451.7 1,308.9 1,339.4 1,312.0 1,325.0 1,295.1 Industrial Residential 580.0 430.0 450.0 420.0 580.0 580.0 431.5 69.2 65.7 70.0 Commercial 60.7 70.2 73.4 72.9 Electricity Generation 128.5 185.0 182.2 190.2 207.4 251.3 243.9 Total 2,216.2 2,136.7 2,010.2 2,010.3 2,169.5 2,229.6 2,043.3 Note: Totals may not sum due to independent rounding. 3 - 76 : Ethanol Consumption by Sector (Trillion Btu) Table End - Use Sector 1990 2005 201 1 2012 2013 2014 2015 60.4 327.4 1,045.0 1,044.6 1,071.6 1,092.8 1,133.9 Transportation Industrial 0.8 6.8 16.7 16.7 17.6 14.4 15.0 2.9 0.5 0.9 Commercial 2.6 2.7 4.1 4.2 1,111.3 1,153.1 Total 61.7 335.1 1,064.6 1,063.8 1,091.8 Note: Totals may not sum due to independent rounding. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 100 3

215 Table 3 77 : Biodiesel Consumption by Sector (Trillion Btu) - End - Use Sector 1990 2005 201 1 2012 2013 2014 2015 Transportation 0.0 11.6 113.1 114.7 182.3 180.8 190.6 0.0 11.6 113.1 114.7 182.3 180.8 190.6 Total Note: Totals may not sum due to independent rounding. Uncertainty and Time Series Consistency - It is assumed that the combustion efficiency for woody biomass is 100 percent, which is believed to be an overestimate of the efficiency of wood combustion processes in the United States. Decreasing the combustion efficiency would decrease emission estimates. Additionally, the heat content applied to the consumption of wood y biomass in the residential, commercial, and electric power sectors is unlikely to be a completely accurate representation of the heat content for all the different types of woody biomass consumed within these sectors. Emission estimates from ethanol and biodiesel production are more certain than estimates from woody biomass consumption due to better activity data collection methods and uniform combustion techniques. Methodological recalculations were applied to the entire time series to ensure time consistency from 1990 - series through 2015 Details on the emission trends through time are described in more detail in the Methodology section, . above. Recalculations Discussion and ethanol values for 1990 through 2014 Wood consumption not revised relative to the previous Inventory , as were from EIA’s Monthly Energy Review (EIA 2017a ). Carbon dioxide emission there were no historical revisions om the estimates from biodiesel consumption were added for 1990 through 2015 to quantify biogenic emissions fr combustion of biodiesel. In previous Inventories, biodiesel consumption was subtracted from fossil fuel combustion estimates but not accounted for in this chapter. Planned Improvements - level combustion emissions through EPA’s Greenhouse Gas Reporting Program The availability of facility GHGRP ) will be examined to help better characterize the industrial sector’s energy consumption in the United ( woody biomass consumption by States, and further classify 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 t o 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 es, some facility level fuel combustion emissions reported under the GHGRP may also include industrial guidelin - 181 process emissions. In line with UNFCCC reporting guidelines, fuel combustion emissions are included in this d in the Industrial Processes and Product Use chapter, while process emissions are include chapter of this report. In examining data from EPA’s GHGRP that would be useful to improve the emission estimates for the CO from 2 biomass combustion category, particular attention will also be made to ensure t ime 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 CO emissions from biomass are separated in the facility - level reported data, and 2 maintaining consistency with national energy statistics provided by EIA. In implementing improvements and integration of data from EP A’s GHGRP, the latest guidance from the IPCC on the use of facility - level data in 182 national inventories will be relied upon. 181 See < http://unfccc.int/resource/docs/2006/sbsta/eng/09.pdf >. 182 - http://www.ipcc See < >. nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf 1 10 - 3 Energy

216 Currently emission estimates from biomass and biomass - based fuels included in this inventory are limited to woody and biodiesel. Other forms of biomass biomass, ethanol, based fuel consumption include biogas. An effort will be - EIA (2017a) natural gas made to examine sources of data for biogas including data from EIA for possible inclusion. al gas supply so no adjustments are needed to the natural gas fuel data already deducts biogas used in the natur consumption data to account for biogas. 3.1 , an additional planned improvement is to evaluate and potentially update our As per discussion in Section method for allocating motor gasoline consumption acros s 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 gasol distribution. ine’s sectoral - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 102 3

217 4. and Product Use Industrial Processes 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 4 1 . Figure - categories included in this chapter are presented in byproducts of various non Greenhouse gas emissions are produced as the energy - related industrial activities. That is, - these emissions are produced either from an industrial process itself, and are not directly a result of energy For example, raw materials can be chemical ly or physically transformed from one consumed during the process. state to another. This transformation can result in the release of greenhouse gases such as carbon dioxide (CO ), 2 methane (CH ), and nitrous oxide (N O). The processes included in this chapter include iron and steel pr oduction 2 4 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 num production, soda ash production and use, titanium dioxide production, CO consumption, production, alumi 2 ferroalloy production, glass production, zinc production, phosphoric acid production, lead production, silicon carbide production and consumption, nitric acid produc tion, and adipic acid production. In addition, greenhouse gases are often used in products or by end - consumers. These gases include industrial sources of man - made compounds such as hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride The present contribution of HFCs, PFCs, SF as well as N O. ), (SF ), nitrogen trifluoride (NF , and NF gases to the 3 3 2 6 6 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; SF is the most potent greenhouse gas the 6 Use of HFCs is growing rapidly since they are Intergovernmental Panel on Climate Change (IPCC) has evaluated. the primary substitutes for ozone depleting substances (ODSs), which are being phased - out under the Montreal Protocol on Substances that Deplete the Ozone Layer . Hydrofluorocarbons, PFCs, SF , and NF are employed and 3 6 emitted by a number of other industrial sources in the United States such as aluminum production, HCFC - 22 production, semiconductor manufacture, electric power transmission and distribution, and magnesium metal production and processing. Nitrous oxide is emitted by the production of adipic acid and nitric acid, semiconductor manufacturing, end - consumers in product uses through the administration of anesthetics, and by industry as a propellant in aerosol products. , IPPU generated emissions of equivalent (MMT CO million metric tons of CO 5 In 201 375.9 Eq.), or 5. 7 percent 2 2 Carbon dioxide emissions from all industrial processes were 170.7 of total U.S. greenhouse gas emissions. MMT CO Eq. ( 170,684 kt CO Methane emissions from industrial ) in 201 5 2 percent of total U.S. CO emissions. , or 3. 2 2 2 processes resulted in emissions of approximately 0.2 MMT CO , which was less than 1 Eq. ( 9 kt CH 5 ) in 201 4 2 Eq. ( emissions. Nitrous oxide emissions from IPPU were 20.3 MMT CO 68 kt N percent of U.S. CH O) in 201 5 , 2 2 4 5 6.1 percent of total U.S. N or O emissions. In 201 combined emissions of HFCs, PFCs, SF , and NF totaled 184.7 6 2 3 MMT CO Eq. Total emissions from IPPU in 201 5 were 10.4 percent more than 1990 emissions. Indirect 2 - Table 4 in kilotons (kt). 108 greenhouse gas emissions also result from IPPU, an d are presented in 1 - 4 Industrial Processes and Product Use

218 Figure 1 : 201 5 Industrial Processes and Product Use Chapter Greenhouse Gas Sources 4 - Eq.) (MMT CO 2 The increase in overal l 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). Emissions from mineral sources have either increased or no t changed significantly (e.g., glass (e.g., cement manufacturing) manufacturing) since 1990 but largely emissions from chemical economic cycles, while CO and CH follow 4 2 either decreased or not changed significantly. Hydrofluorocarbon emissions from the substitution of sources have , and NF from other OD S have increased drastically since 1990, while the emission trends of HFCs, PFCs, SF 6 3 sources are mixed. Nitrous oxide emissions from the production of adipic and nitric acid have decreased, while N O 2 emissions from product uses has rem ained nearly constant over time. Trends are explained further within each emission source category throughout the chapter. using IPCC Fourth Assessment Report Table 4 - 1 summarizes emissions for the IPPU chapter in MMT CO Eq. 2 (AR4) GWP values , following the requirements of the revised United Nations Framework Convention on Climate 183 Change (UNFCCC) reporting guidelines for national inventories (IPCC 2007). Unweighted native gas emissions 4 The source descriptions that follow in the chapter are presented in the order as in kt are also provided in Table . - 2 reported to the UNFCCC in the common reporting format tables, corresponding generally to: mineral products, chemical production, metal production, and emissions from the u ses of HFCs, PFCs, SF , and NF . 6 3 183 http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf ee < S >. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 2 4

219 4 Table 1 : Emissions from Industrial Processes and Product Use (MMT CO - Eq.) 2 3 201 1 201 2 Gas/Source 201 2005 201 4 201 5 1990 191.7 174.3 171.0 172.9 179.3 170.7 CO 208.8 2 Iron and Steel Production & 68.0 61.1 55.4 53.3 58.6 48.9 Metallurgical Coke Production 101.5 51.5 66.0 59.7 54.9 99.0 56.6 46.0 Iron and Steel Production 2.5 2.0 1.4 0.5 1.8 2.0 Metallurgical Coke Production 2.8 39.4 33.5 32.2 35.3 36.4 46.2 39.9 Cement Production 27.0 26.3 26.5 26.4 26.5 28.1 Petrochemical Production 21.3 11.7 14.6 14.0 13.8 Lime Production 14.0 14.2 13.3 10.4 4.9 6.3 9.3 8.0 11.8 11.2 Other Process Uses of Carbonates 13.0 9.2 9.3 9.4 10.0 9.6 10.8 Ammonia Production 1.5 1.4 4.1 4.0 4.2 Carbon Dioxide Consumption 4.5 4.3 Soda Ash Production and 2.8 2.8 2.7 2.8 2.8 3.0 2.8 Consumption 6.8 4.1 3.3 3.4 3.3 2.8 2.8 Aluminum Production 2.2 1.4 1.7 Ferroalloy Production 1.8 1.9 2.0 1.9 1.7 1.2 1.8 1.7 1.5 1.7 1.6 Titanium Dioxide Production 1.5 1.9 1.3 1.2 1.3 1.3 1.3 Glass Production Urea Consumption for Non - 4.0 3.7 4.0 4.4 1.4 1.1 Agricultural Purposes 3.8 1.5 1.3 1.2 1.1 1.1 1.0 1.0 Phosphoric Acid Production 0.6 1.0 1.3 1.5 1.4 1.0 0.9 Zinc Production 0.5 0.6 0.5 0.5 0.5 0.5 Lead Production 0.5 Carbide Production and Silicon 0.2 0.2 0.2 Consumption 0.4 0.2 0.2 0.2 Magnesium Production and + Processing + + + + + + 0.3 CH 0.2 0.2 0.1 0.1 0.1 0.1 4 + 0.1 0.1 0.1 0.2 0.2 Petrochemical Production 0.1 + + + + + + Ferroalloy Production + Silicon Carbide Production and + + + + + + Consumption + Iron and Steel Production & Metallurgical Coke Production + + + + + + + + + + + + + Iron and Steel Production + Metallurgical 0.0 0.0 0.0 0.0 0.0 0.0 Coke Production 0.0 31.6 22.8 25.6 20.4 19.0 20.8 20.3 N O 2 11.3 10.9 10.5 12.1 10.9 11.6 Nitric Acid Production 10.7 7.1 10.2 5.5 Adipic Acid Production 3.9 5.4 4.3 15.2 4.2 4.2 4.2 4.2 4.2 4.2 4.2 O from Product Uses N ₂ e + 0.1 0.2 0.2 0.2 0.2 0.2 Semiconductor Manufactur HFCs 120.0 154.3 155.9 159.0 166.7 173.2 46.6 Substitution of Ozone Depleting a 99.7 145.3 Substances 154.6 161.3 168.5 0.3 150.2 HCFC - 22 Production 46.1 20.0 8.8 5.5 4.1 5.0 4.3 Semiconductor Manufactur 0.2 0.2 0.2 0.2 e 0.2 0.3 0.3 Magnesium Production and 0.0 0.0 + + 0.1 0.1 0.1 Processing PFCs 24.3 6.7 6.9 6.0 5.8 5.8 5.2 2.8 3.2 3.4 e 3.0 2.8 3.2 3.2 Semiconductor Manufactur Aluminum Production 21.5 3.4 3.5 2.9 3.0 2.5 2.0 Substitution of Ozone Depleting a + 0.0 + Substances + + + + 5.8 SF 6.6 28.8 11.7 9.2 6.8 6.4 6 Electrical Transmission and 23.1 8.3 6.0 Distribution 4.8 4.6 4.8 4.2 3 - 4 Industrial Processes and Product Use

220 Magnesium Production and 5.2 2.7 2.8 1.6 1.5 1.0 0.9 Processing 0.4 e 0.7 0.4 0.4 0.5 0.7 0.7 Semiconductor Manufactur + 0.5 0.7 0.6 NF 0.6 0.5 0.6 3 + 0.5 0.7 0.6 0.6 0.5 0.6 Semiconductor Manufactur e 353.4 371.0 360.9 340.4 363.7 379.8 375.9 Total + Does not exceed 0.05 MMT CO Eq. 2 a Small amounts of PFC emissions also result from this source. independent rounding. Note: Totals may not sum due to - 2 : Emissions from Industrial Processes and Product Use (kt) Table 4 1990 2005 201 1 201 2 201 3 201 Gas/Source 201 5 4 CO 191,653 174,311 171,000 208,795 172,890 179,283 170,684 2 Iron and Steel Production & Metallurgical Coke Production 101,487 68,047 61,108 53,348 58,629 48,876 55,449 Iron and Steel Production 98,984 66,003 59,681 54,906 51,525 56,615 46,038 Metallurgical Coke Production 2,503 2,044 1,426 543 1,824 2,014 2,839 46,194 32,208 35,270 33,484 36,369 39,439 39,907 Cement Production 21,326 26,972 26,338 26,501 26,395 26,496 28,062 Petrochemical Production Lime Production 11,700 14,552 13,982 13,785 14,028 14,210 13,342 Other Process Uses of Carbonates 4,907 6,339 9,335 10,414 11,811 11,236 8,022 Ammonia Production 9,196 9,292 9,377 9,962 9,619 10,799 13,047 1,472 1,375 4,083 4,019 Carbon Dioxide Consumption 4,188 4,471 4,296 Soda Ash Production and Consumption 2,822 2,960 2,712 2,763 2,804 2,827 2,789 Aluminum Production 6,831 4,142 3,292 3,255 2,833 2,767 3,439 Ferroalloy Production 2,152 1,392 1,735 1,903 1,785 1,914 1,960 Titanium Dioxide Production 1,195 1,755 1,729 1,528 1,715 1,688 1,635 Glass Production 1,535 1,928 1,299 1,248 1,317 1,336 1,299 Urea Consumption for Non - 3,653 4,030 4,407 4,014 1,380 1,128 3,784 Agricultural Purposes 1,149 1,342 1,171 1,118 1,038 999 Phosphoric Acid Production 1,529 632 Zinc Production 1,030 1,286 1,486 1,429 956 933 516 553 538 527 459 473 Lead Production 546 Silicon Carbide Production and 375 219 170 Consumption 169 173 180 158 Magnesium Production and 1 3 3 2 2 2 3 Processing 4 3 12 CH 4 6 9 4 4 Petrochemical Production 3 2 3 3 5 7 9 + + + 1 Ferroalloy Production 1 1 1 Silicon Carbide Production and Consumption 1 + + + + + + Iron and Steel Production & 1 1 + + + + + Metallurgical Coke Production 1 + + Iron and Steel Production + + + 1 0 0 0 0 Metallurgical Coke Production 0 0 0 69 O 106 76 86 64 70 68 N 2 38 37 35 41 36 37 39 Nitric Acid Production Adipic Acid Production 51 24 34 19 13 18 14 14 14 14 14 14 14 14 N O from Product Uses ₂ 1 e + + 1 Semiconductor Manufactur 1 1 1 M HFCs M M M M M M Substitution of Ozone Depleting a Substances M M M M M M M + + + + HCFC - 22 Production 3 1 1 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 4 4

221 Semiconductor Manufactur + + + + + + + e Magnesium Production and 0 0 + + + + + Processing M M M PFCs M M M M M e M M M Semiconductor Manufactur M M M Aluminum Production M M M M M M M Substitution of Ozone Depleting a 0 + + Substances + + + + SF 1 1 + + + + + 6 Electrical Transmission and Distribution 1 + + + + + + Magnesium Production and Processing + + + + + + + + Semiconductor Manufactur e + + + + + + + + + + + + + NF 3 e + + + Semiconductor Manufactur + + + + + 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 2006 IPCC Guidelines for National Greenhouse Gas Inventories (2006 IPCC 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 s ource may not be currently occurring in the United States, data is not currently available for those emission sources (e.g., ceramics, non - metallurgical magnesium p roduction ), 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. fossil fuels consumed for non Box 3 - 6 Finally, as mentioned in the Energy chapter of report ( - energy uses for this ), alt and primary purposes other than combustion for energy (including lubricants, paraffin waxes, bitumen asph . 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; h owever , due to national circumstances regarding the allocation of energy EPA reports non - energy uses in the Energy chapter of this statistics and C balance data, Inventory. R eporting these non - energy use emissions under IPPU would involve making artificial adjustments to the non - s would also result in the C emissions for lubricants, waxes, energy use C balance. These artificial adjustment 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 balanc e and a less transparent approach, - Energy Uses of Fossil Fuels the entire calculation of C storage and C emissions is therefore conducted in the Non category calculation methodology, and reported under the Energy sector. For more information, see the Meth odology section for CO Energy Uses from Fossil Fuel Combustion and Chapter 3.2 Carbon Emitted from N on - 2 of the fuel consumption data for seven fuel categories — of Fossil Fuels. As stated in the Energy chapter, portions were ustrial other coal, petroleum coke, natural gas, residual fuel oil, and other oil — coking coal, distillate fuel, ind energy related - , as they were consumed during non . Emissions llocated to the IPPU chapter industrial activity rea 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 ational circumstances. n 5 - 4 Industrial Processes and Product Use

222 QA/QC and Verification Procedures IPPU sources, a detailed QA/QC plan was developed and implemented for specific categories For 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, Chap 2006 IPCC Guidelines that focus on annual procedures and checks to be used when ter 6 of the gathering, maintaining, handling, documenting, checking, and archiving the data, supporting documents, and files; and (2) source - category specific (Tier 2) proced ures 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. vity data and emission estimates are Examples of these procedures include : checks to ensure that acti consistent with historical trends ; that, where possible, consistent and reputable data to identify significant changes across sources; that interpolation or extrapolation techniques are co nsistent across sources are used and specified , units, sources; and that common datasets 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 fa ctors; and confirmed that estimates were calculated and reported for all applicable and able portions of the source categories for all years. Tier 1 QA/QC procedures - related QC (category - specific, Tier 2) have been performed for all IPPU and calculation procedures were performed for specific QC so urces. C onsistent with the IPCC 2006 Guidelines , additional category - (such as the comparison of reported consumption with modeled consumption more significant emission categories utes) or sources where significant methodological and data updates have using GHGRP data within ODS Substit taken place . The QA/QC implementation did not reveal any significant inaccuracies and all errors identified were category - specific QC procedures and documented and corrected. Application of these procedures, specifically 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. IPPU categories, activity data is obtained via aggregation of facility level data from EPA’s GHGRP , For most national commodity surveys conducted by U.S. Geologic Survey National Minerals Information Center, U.S. Department of Energy (DOE), U.S. Census B ureau, 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 e reliability of reported ) . T he uncertainty of the activity dat a is a function of th category The emission factors used plant - level production data and is influenced by the completeness of the survey response. ’s GHGRP and application of IPCC defaults. IPCC default factors are derived using are derived from EPA calculations that assume precise and efficient chemical reactions, or were based upon empirical data in published As a result, uncertainties in the emission coefficients can be attributed to, among other things, references. inefficiencies in the chemical reactions associated with each production process or to the use of empirically - derived emission factors that are biased; therefore, they may or may not represent U.S. national averages. A dditional assumptions are described within each source. The uncertainty analysis performed to quantify uncertainties associated with the 2015 emission estimates from IPPU - continues a multi year process for developing credible quantitative uncertainty esti mates 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 de velopment 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 il lustrative 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 uncertaint y analysis recognizes very significant connections among sources, a more comprehensive approach that accounts for all linkages will be identified as the uncertainty analysis moves forward . cal Approach for Estimating and Reporting U.S. Emissions and Sinks Methodologi Box 4 - 1 : In following the United Nations Framework Convention on Climate Change (UNFCCC) requirement under Article 4.1 to develop and submit national greenhouse gas emission inventories, the emissio ns and sinks presented in this accepted - report and this chapter, are organized by source and sink categories and calculated using internationally 2006 IPCC Guidelines for in the methods provided by the Intergovernmental Panel on Climate Change (IPCC) - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 6 4

223 Natio nal GHG Inventories . Additionally, the calculated emissions and sinks 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. T he use of consistent methods to calculate emissions and sinks by all nations providing their inventories to the UNFCCC ensures that these reports are comparable. In this regard, U.S. emissions and sinks reported in this Inventory are comparable to emission s and sinks reported by other countries. Emissions and sinks provided in this Inventory do not preclude alternative examinations, but rather, this Inventory presents emissions and sinks in a common format consistent with how countries are to report Invento ries under the UNFCCC. The report itself, and this chapter, follows this standardized format, and provides an explanation of the IPCC methods used to calculate emissions and sinks, and the manner in which those calculations are conducted. 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 Environmental Protection Agency 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 CO 2 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 f ossil fuels and industrial greenhouse gases. In general, the threshold for reporting is 25,000 metric tons or more of CO Eq. per year, but 2 reporting is required for all facilities in some industries. Calendar year 2010 was the first year for which data we re reported for facilities subject to 40 CFR part 98, though some source categories first reported data for calendar year 2011. The GHGRP dataset continue s EPA’s GHGRP dataset and the data presented in this Inventory are complementary. to be an important resource for the Inventory, providing not only annual emissions information, but also other annual information, such as can improve and refine national emission activity data and emission factors that estimates and trends over time . GHGRP data also allow E PA to disaggregate national inventory estimates in new ways that can highlight differences across regions and sub - categories of emissions , along with enhancing application number of category uses annual GHGRP data in a of QA/QC procedures and assessment of uncertainties. EPA use to improve the national further continues to analyze the data on an annual basis, as applicable, for estimates and estimat es presented in this Inventory consistent with IPCC guidance While many methodologies used in EPA’s . re consistent with IPCC GHGRP a i t 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 Inv entory is a comprehensive accounting of all emissions from source categories identified in the IPCC (2006) guidelines. Further information on the reporting categorizations in EPA’s GHGRP and specific data caveats associated with monitoring methods in EPA’s GHGRP has been provided on the 184 GHGRP website. 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 onfirm that a given data aggregation shields underlying CBI from public disclosure. to c EPA is 185 publishing only data values that meet these aggregation criteria. Specific uses of aggregated facility - level data are l sections. described in the respective methodologica 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, giv ing particular consideration to ensuring time series consistency and completeness. As stated previously in the Introduction chapter, t his year EPA has integrated 184 See < http://www. ipcc - nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf > . 185 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas Data, reporting>. - ghg Nove mber 25, 2014. See

224 186 GHGRP information for and also identi fied places where various Industrial Processes and Product Use categories 187 EPA plans to integrate additional GHGRP data in additional categories (see those categories Planned EPA has paid particular attention to ensuring time - series consistency for major Improvement sections for details). ve occurred from the incorporation of GHGRP data into these categories, consistent with 2006 recalculations that ha 188 Guidelines and IPCC Technical Bulletin on Use of Facility - Specific Data in National GHG Inventories. IPCC EPA verifies annual facility level reports through a multi - step process to identify potential errors and ensure that - 189 nd consistent . data submitted to EPA are accurate, complete, a The GHGRP dataset is a particularly important annual resource and will continue to be important for improving emissions estimates from Industrial Process and Product Use in future Inventory reports. Additionally, t he GHGRP has and will continue to enhance QA/QC procedures and assessment of uncertainties within the IPPU categories (see those categories for specific QA/QC rding the use of GHGRP data). details rega 4.1 Cement Production (IPCC Source Category 2A1) Cement production is an energy - and raw material - intensive process that results in the generation of carbon dioxide (CO chemical process itself. Emissions from fuels ) from both the energy consumed in making the cement and the 2 consumed for energy purposes during the production of cement are accounted for in the Energy chapter. During the cement production process, calcium carbonate (CaCO ) is heated in a cement kiln at a temperatu re range 3 of about 700 to 1000 degrees Celsius (1,292 to 1,832 degrees Fahrenheit) to form lime (i.e., calcium oxide or CaO) and CO emitted during cement production is in a process known as calcination or calcining. The quantity of CO 2 2 directly proportiona l to the lime content of the clinker. During calcination, each mole of limestone (CaCO ) heated 3 in the clinker kiln forms one mole of lime (CaO) and one mole of CO : 2 +ℎ푒푎푡 →퐶푎푂 +퐶푂 퐶푎퐶푂 3 2 Next, the lime is combined with silica ials to produce clinker (an intermediate product), with the - containing mater earlier byproduct CO being released to the atmosphere. The clinker is then allowed to cool, mixed with a small 2 190 e Portland cement. amount of gypsum and potentially other materials (e.g., slag, etc.), and used to mak Carbon dioxide emitted from the chemical process of cement production is the second largest source of industrial CO emissions in the United States. Cement is produced in 34 states and Puerto Rico. Texas, California, Missouri, 2 and Alabama were the five leading cement Florida, - producing states in 2015 and accounted for nearly 50 percent of total U.S. production (USGS 6b). Clinker production in 2015 increased approximately 1 percent from 2014 201 levels as c ement sales continued to increase compared to 2014. In 2015, U.S. in 2015 , but at a more moderate rate clinker production totaled 76,700 kilotons (EPA 2016). The resulting CO emissions were estimated to be 39.9 2 CO ). Eq. ( MMT 3 - 4 39, 907 kt ) (see Table 2 186 Adipic Acid Production, Aluminum Production, Carbon Dioxide Consumption, Cement Production, Electrical Transmission , HCFC - 22 Production, Lime Production, Magnesium Production and Processing, ODS Substitutes, Nitric Acid and Distribution . and Semiconductor Manufacture Production, Petrochemical Production, 187 Ammonia Production, Glass Production and Other fluorinated gas production . 188 http: //www.ipcc - nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf 189 https://www.epa.gov/sites/production/files/2015 - 07/documents/ghgrp_verification_factsheet.pdf >. See < 190 ement, which is produced using Approximately three percent of total clinker production is used to produce masonry c plasticizers (e.g., ground limestone, lime, etc.) and Portland cement (USGS 2011). Carbon dioxide emissions that result from the tegory. production of lime used to create masonry cement are included in the Lime Manufacture source ca - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 8 4

225 Eq. and kt) 4 : CO Table Emissions from Cement Production (MMT CO - 3 2 2 Year Eq. kt MMT CO 2 33.5 33,484 1990 46,194 2005 46.2 32,208 2.2 2011 3 35.3 2012 35,270 36,369 2013 36.4 39.4 39,439 2014 2015 39.9 39, 907 Greenhouse gas emissions from cement production increased every year from 1991 through 2006 (with the exception of a slight decrease in 1997), but decreased in the following years until 2009. Emissions from cement production were at their lowest levels in 2009 (2009 emissions are approximately 28 percent lower than 2008 emissions and 12 percent lower than 1990). Since 2010, emissions have increased by roughly 27 percent. In 2015, emissions from cement production increased by 1 percent from 2014 levels. Emissions since 1990 have increased by 19 percent. Emissions decreased significantly between 2008 and 2009, due d associated decrease in demand for construction materials. Emissions increased to the economic recession an slightly from 2009 levels in 2010, and continued to gradually increase during the 2011 through 2015 time period due to increasing consumption. Cement continues to be a critical component of the construction industry; therefore, the availability of public and private construction funding, as well as overall economic conditions, have considerable impact on the level of cement production. Methodology Carbon dioxide emissions were estimated using the Tier 2 methodology from the 2006 IPCC Guidelines . The Tier 2 methodology was used because detailed and complete data (including weights and composition) for carbonate(s) ous Tier 3 approach is impractical. Tier 2 specifies consumed in clinker production are not available, and thus a rigor the use of aggregated plant or national clinker production data and an emission factor, which is the product of the average lime fraction for clinker of 65 percent and a constant reflecting the mass of C released per unit of lime. O 2 U.S. Geological Survey (USGS) mineral commodity expert for cement has confirmed that this is a reasonable The assumption for the United States (Van Oss 2013a). This calculation yields an emission factor of 0.51 tons of CO per 2 ton of clinker produced, which was determined as follows: EF = 0.6 50 CaO × [(44.01 g/mole CO ) ÷ (56.08 g/mole CaO)] = 0.5 1 0 tons CO /ton clinker 2 clinker 2 During clinker production, some of the clinker precursor materials remain in the kiln as non calcinated, partially - calcinated, or fully calcinated cement kiln dust (CKD). The emissions attributable to the calcinated portion of the CKD are not accounted for by the clinker emission factor. The IPCC recommends that these additional CKD CO 2 ons should be estimated as two percent of the CO emissions calculated from clinker production (when data emissi 2 on CKD generation are not available). Total cement production emissions were calculated by adding the emissions from clinker production to the emissio ns assigned to CKD (IPCC 2006). Furthermore, small amounts of impurities (i.e., not calcium carbonate) may exist in the raw limestone used to of produce clinker. The proportion of these impurities is generally minimal, although a small amount (1 to 2 percent) magnesium oxide (MgO) may be desirable as a flux. Per the IPCC Tier 2 methodology, a correction for MgO is not used, since the amount of MgO from carbonate is likely very small and the assumption of a 100 percent carbonate source of CaO already yields an overestimation of emissions (IPCC 2006). ss 4 Table 4 ) were obtained from USGS (Van O The 1990 through 2012 activity data for clinker production (see - 2013b). Clinker production data for 2013 were also obtained from USGS (USGS 2014). The data were compiled by USGS (to the nearest ton) through questionnaires sent to domestic clinker and cement manufacturing plants, co. Following up on previous planned improvements, EPA has incorporated including the facilities in Puerto Ri 9 - 4 Industrial Processes and Product Use

226 clinker production data for 2014 and 2015 from EPA’s GHGRP (EPA 2016) to estimate emissions in these respective years. More details on how this change compares to USGS reported data ca n be found in the section on Series Consistency and Recalculations Discussion. Uncertainty and Time - Table 4 - 4 : Clinker Production (kt) Year Clinker 64,355 1990 88,783 2005 61,903 2011 2012 67,788 2013 69,900 2014 75, 800 2015 76, 700 Notes: Clinker production from 1990 through 2015 includes Puerto Rico. Data were obtained from USGS (Van Oss 2013a; USGS 2016 ), whose original data source was USGS and U.S. Bureau of Mines Minerals Yearbooks (201 4 data obtained from mineral industry surveys for cement in September 2015; 2015 data obtained from mineral industry surveys for cement in January 2016). Uncertainty and Time - S eries 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 CaCO likely consists of other carbonate and non , when a small percentage 3 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 fications. speci is reabsorbed when the cement is used for construction. As Additionally, some amount of CO 2 cement reacts with water, alkaline substances such as calcium hydroxide are formed. During this curing process, in the these compounds may react with CO atmosphere to create calcium carbonate. This reaction only occurs in 2 Because the amount of CO roughly the outer 0.2 inches of surface area. reabsorbed is thought to be minimal, it was 2 not estimated. assess this assumption by conducting a However, see Planned Improvements described below to re 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 y GHGRP Subpart H facilities for future Inventory reports. EPA verifies annual clinker production data required b 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 submit ted to EPA are accur ate, complete, and consistent. Based on the results of the verification process, the EPA follows up with facilities to resolve mistakes that may have 191 Facilities are also required to monitor and maintain records of monthly cli nker production. occurred. 4 The results of the Approach 2 quantitative uncertainty analysis are summarized in 5 . Table Based on the - uncertainties associated with total U.S. clinker production, the CO emission factor for clinker production, and the 2 emission factor for additional CO emissions from cement production were emissions from CKD, 2015 CO 2 2 estimated to be between 37. 5 and 42. 3 MMT CO This confidence level Eq. at the 95 percent confidence level. 2 indicates a range of approximately 6 percent below and 6 percent above the emission estimate of 39. 9 MMT C O 2 Eq. 191 - https://www.epa.gov/sites/production/files/2015 See < >. 07/documents/ghgrp_verification_factsheet.pdf - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 10 4

227 Emissions from Cement 4 - 5 : Approach 2 Quantitative Uncertainty Estimates for CO Table 2 Production (MMT CO Eq. and Percent) 2 a 2015 Emission Estimate Uncertainty Range Relative to Emission Estimate Gas Source Eq.) (MMT CO Eq.) (%) (MMT CO 2 2 Upper Lower Lower Upper Bound Bound Bound Bound CO Cement Production 6% 39. 9 37. 5 42. 3 - +6% 2 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 2015. Details on the emission trends through time are described in more detail in the Methodology section, above. More information on the consistency in clinker production data and emissions across the time series with the use of GHGRP clinker data for 2014 and 2015 can be fo und in the Recalculations Discussion section below. consistent with Volume 1, For more information on the general QA/QC process applied to this source category, QA/QC and Verification Procedures section in the int roduction of the , see 2006 IPCC Guidelines Chapter 6 of the IPPU Chapter. Recalculations Discussion During public review of the draft 1990 to 2015 Inventory, the EPA received additional expert judgment from USGS cement commodity expert regarding the country - specific emission factor used National Minerals Information Center emissions from clinker production. The emission factor was based on an average lime fraction for to estimate CO 2 better reflect known precision of the average clinker of 65 percent (0.6460 CaO) and was changed to 0.65 CaO to ton lime fraction. As a result, the associated emission factor changed from 0.5070 to 0.0510 per of CO of tons 2 . This updated emission factor was applied to recalculate the entire 1990 through 2015 time series. clinker USGS also informed EPA of a transcription error in historical clinker production data that was provided to EPA for previous reports. For the year 1996, clinker production was incorrectly provided as 75,706 thousand metric tons. The correct activit y data input for 1996 is 71,706 thousand metric tons. This data has been updated in this Inventory report and resulted in a decrease of the 1996 cement production emissions estimate by approximately 2.0 MMT CO 2 Eq., a decrease of 5 percent compared to the 1996 estimate included in previous Inventory reports. Finally, 2014 and 2015 clinker production data was available and has been incorporated into the (EPA 2016) - leve l data in national Inventory report. EPA relied upon the latest guidance from the IPCC on the use of facility specific QC process to compare activity data from GHGRP with existing data inventories and applied a category - series consistency of the emission estimates presented in the Inventory. For the - from USGS. This was to ensure time USGS and GHGRP clinker production data showed a difference of approximately 2 percent, while in year 2014, 2015 that difference decreased to less than 1 percent between the two sets of activity data. This difference resulted in GS data (USGS 2016a) by 0.7 MMT CO Eq. in 2014 and 0.0 MMT CO an increase of emissions compared to US 2 2 Eq. in 2015. Planned Improvements In response to comments from the Portland Cement Association (PCA) and UNFCCC expert technical reviews, EPA reported under EPA’s GHGRP that would be useful to improve the is continuing to evaluate and analyze data emission estimates for the Cement Production source category. EPA held a technical meeting with PCA in August 2016 to review inventory methods and available data from the GHGRP data set. Most c ement production facilities reporting under EPA’s GHGRP use Continuous Emission Monitoring Systems (CEMS) to monitor and report CO 2 emissions, thus reporting combined process and combustion emissions from kilns. In implementing further - egration of data from EPA’s GHGRP, the latest guidance from the IPCC on the use of facility improvements and int level data in national inventories will be relied upon, in addition to category specific QC methods recommended by 11 - 4 Industrial Processes and Product Use

228 192 2006 IPCC Guidelines EPA’s long - term improvemen t plan includes continued assessment and feasibility to use . additional GHGRP information, in particular disaggregating aggregated GHGRP emissions consistent with IPCC This longer term and UNFCCC guidelines to present both national process and combustion emissions streams. - planned analysis is still in development and has not been updated for this current inventory report. Finally, in additional response to from PCA during the public review of the draft Inventory in March 2017, feedback received have additional discussions with PCA to address additional longer - term improvements to review EPA plans to emissions from cement production to account for both organic material and methods and data used to estimate CO 2 um carbonate in the raw materia l, in additi on to carbonation that occurs later in the cement product magnesi lifecycle. EPA will work with PCA to identify data and studies on average MgO content of clinker produced in the United States, an average carbon content for organic materials in kiln feed in the Un ited States and CO reabsorption 2 rates via carbonation for various cement products. 4.2 Lime Production (IPCC Source Category 2A2) Lime is an important manufactured product with many industrial, chemical, and environmental applications. Lime production involv es three main processes: stone preparation, calcination, and hydration. Carbon dioxide (CO ) is 2 — mostly calcium carbonate (CaCO ) — is roasted at high generated during the calcination stage, when limestone 3 temperatures in a kiln to produce calcium oxide (CaO) and CO is given off as a gas and is normally . The CO 2 2 emitted to the atmosphere. 퐶푎퐶푂 →퐶푎푂 +퐶푂 3 2 Some of the CO generated during the production process, however, is recovered at some facilities for use in sugar 2 193 refining and precipitated calcium carbonate (PCC) production. Emissions from fuels consumed for energy purposes during the production of lime are accounted for in the Energy chapter. For U.S. operations, the term “lime” actually refers to a variety of chemical compounds. These include CaO, or ), or hydrated lime; dolomitic quicklime ([CaO•MgO]); and high - calcium quicklime; calcium hydroxide (Ca(OH) 2 dolomitic ]). •Mg(OH) hydrate ([Ca(OH) •MgO] or [Ca(OH) 2 2 2 - The current lime market is approximately distributed across five end use categories as follows: metallurgical uses, 7 percent; environmental uses, 31 percent; chemical and industrial uses, 22 percent; constr uction uses, 9 3 percent; and refractory dolomite, 1 percent (USGS 2016b). The major uses are in steel making, flue gas desulfurization systems at coal - fired electric power plants, construction, and water treatment, as well as uses in mining, pulp and paper scrubber, and there has been and precipitated calcium carbonate manufacturing. Lime is also used as a CO 2 experimentation on the use of lime to capture CO from electric power plants. 2 in 2015 — including Puerto Rico — was reported to be 18,279 kilotons Lime production in the United States - end 2015, there were 77 operating primary lime plants in the United States, including At year (Corathers 201 7 ). 194 Puerto Rico. Principal lime producing states are Missouri, Alabama, Kentucky, Ohio, Texas (USGS 2016a). U.S. li me production resulted in estimated net CO and 6 Eq. (13,342 kt) ( see Table 4 - emissions of 13.3 MMT CO 2 2 4 Table - 7 ). The trends in CO emissions from lime production are directly proportional to trends in production, 2 which are described below. 192 See < http://www.ipcc - nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf >. 193 with calcium hydroxid e. It is used as a filler and/or coating in the paper, food, and PCC is obtained from the reaction of CO 2 plastic industries. 194 s In 2015, 74 operating primary lime facilities in the United State reported to the EPA Greenhouse Gas Reporting Program. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 12 4

229 Table 4 : CO - Emissions from Lime Production (MMT CO 6 Eq. and kt) 2 2 MMT CO Eq. kt Year 2 11.7 11,700 1990 2005 14,552 14.6 14.0 13,982 2011 13.8 13,785 2012 2013 14.0 14,028 2014 14,210 14.2 2015 13,342 13.3 - 7 : Potential, Recovered, and Net CO Table Emissions from Lime Production (kt) 4 2 a Recovered Year Net Emissions Potential 11,959 259 11,700 1990 2005 15,074 522 14,552 407 13,982 2011 14,389 473 13,785 2012 14,258 467 14,028 2013 14,495 14,715 505 14,210 2014 422 13,342 2015 13,764 a For sugar refining and PCC production. Note: Totals may not sum due to independent rounding. In 2015, lime production decreased compared to 2014 levels (decrease of about 6 percent) at 18,279 kilotons, owing to decreased consumption by the U.S. nonferrous metallurgical industries (primarily copper) and steel industries (Corathers 2017; USGS 2016a) . Methodology - calcium and dolomitic lime produced were multiplied by their To calculate emissions, the amounts of high . The emission factor is the 2006 IPCC Guidelines respective emission factors using the Tier 2 approach from the tric ratio between CO product of the stoichiome and CaO, and the average CaO and MgO content for lime. The CaO 2 and MgO content for lime is assumed to be 95 percent for both high - calcium and dolomitic lime (IPCC 2006). The emission factors were calculated as follows: For high calc ium lime: - ) ÷ (56.08 g/mole CaO)] × (0.9500 CaO/lime) = 0.7455 g CO /g lime [(44.01 g/mole CO 2 2 For dolomitic lime: [(88.02 g/mole CO ) ÷ (96.39 g/mole CaO)] × (0.9500 CaO/lime) = 0.8675 g CO /g lime 2 2 Production was adjusted to remove the mass of chemica lly combined water found in hydrated lime, determined according to the molecular weight ratios of H ]) (IPCC 2006). These factors O to (Ca(OH) •Mg(OH) and [Ca(OH) 2 2 2 2 set the chemically combined water content to 24.3 percent for high - calcium hydrated lime, an d 27.2 percent for dolomitic hydrated lime. The 2006 IPCC G uidelines (Tier 2 method) also recommends accounting for emissions from lime kiln dust (LKD) typically not through application of a correction factor. LKD is a byproduct of the lime manufacturing process . recycled back to kilns grained material and is especially useful for applications requiring very - LKD is a very fine 13 - 4 Industrial Processes and Product Use

230 small particle size. Currently, data on Most common LKD applications include soil reclamation and agriculture. oduction is not readily available ime emission annual LKD pr to develop a country specific correction factor. L emissions from See the Planned estimates were multiplied by a factor of 1.02 to account for LKD (IPCC 2006). Improvements section associated with efforts to im prove uncertainty analysis and emission estimates associated with LKD. the amount of CO Lime emission estimates were further adjusted to account for captured for use in on - site 2 processes. All the domestic lime facilities are required to report these data to EPA under its GHGRP. The total - level annual amount of CO national captured for on - site process use was obtained from EPA’s GHGRP ( EPA 2 201 6 ) based on reported facility level data for years 2010 through 2015. The amount of CO captured/recovered for 2 - s ite process use is deducted from the total potential emissions (i.e., from lime production and LKD). The net lime on emissions are presented in Table 4 - 6 and Table 4 - 7 . GHGRP data on CO captured/recovered) removals (i.e., CO 2 2 was available only for 2010 through 2015. Since GHGRP data are not available for 1990 through 2009, IPCC “sp licing” techniques were used as per the 2006 IPCC G uidelines on time - series consistency (IPCC 2006, Volume 1, Chapter 5). and dolomitic Lime production data (by type, high - calcium - - quicklime, high - calcium - and dolomitic - hydrated, and dead - for 1990 through 2015 (see Table 4 - 8 ) were obtained from the U.S. Geological Survey burned dolomite) Corathers 2017 compiled by USGS to the nearest ton. Natural (USGS) (USGS 2016b; ) annual reports and are hydraulic lime, which is produced from CaO and hydraulic calcium silicates, is not manufactur ed in the United States (USGS 2011). Total lime production was adjusted to account for the water content of hydrated l ime by Table 4 - 9 converting hydrate to oxide equivalent based on recommendations from the IPCC, and is presented in (IPCC 200 ). The CaO and CaO•MgO co ntents of lime were obtained from the IPCC (IPCC 2006). 6 Since data for the individual lime types (high calcium and dolomitic) w ere not provided prior to 1997, total lime production for 1990 through 1996 was calculated according to the three year distributi on from 1997 to 1999. and Dolomitic Table - 8 : High - Calcium - and Dolomitic - Quicklime, High - Calcium - 4 - Hydrated, and Dead Burned - Dolomite Lime Production (kt) - Dolomitic Calcium - High Burned - Dead Dolomitic Calcium High - Quicklime Year Quicklime Hydrated Hydrated Dolomite 2,234 1,781 319 11,166 342 1990 474 2,990 2,220 200 2005 14,100 13,900 2,690 2,010 200 230 2011 13,600 2,790 2,000 253 200 2012 13,800 2,850 2,050 2013 260 200 200 2014 279 2,190 14,100 2,740 13,100 2,550 2,150 279 200 2015 Table - 9 : Adjusted Lime Production (kt) 4 Year High - Calcium Dolomitic 1990 12,466 2,800 2005 15,721 3,522 2011 15,367 3,051 15,060 3,167 2012 2013 15,297 3,252 15,699 2014 3,135 2,945 2015 14,670 Minus water content of hydrated lime. Note: - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 14 4

231 Uncertainty and Time - Series Consistency attributed to slight differences in the chemical composition of The uncertainties contained in these estimates can be recovery rates for on - site process use over the time series. Although the methodology lime products and CO 2 ies found in lime, such as iron accounts for various formulations of lime, it does not account for the trace impurit Due to differences in the limestone used as a raw material, a rigid specification of lime oxide, alumina, and silica. As a result, few plants produce lime with exactly the same properties. material is impossible. In addition, a portion of the CO emitted during lime production will actually be reabsorbed when the lime is 2 consumed, especially at captive lime production facilities. As noted above, lime has many different chemical, industrial, environmental, and construc In many processes, CO tion applications. reacts with the lime to create 2 calcium carbonate (e.g., water softening). Carbon dioxide reabsorption rates vary, however, depending on the For example, 100 percent of the lime used to produce precipit ated calcium carbonate reacts with CO application. ; 2 whereas most of the lime used in steel making reacts with impurities such as silica, sulfur, and aluminum compounds. Quantifying the amount of CO that is reabsorbed would require a detailed accounting of lime use in the 2 United States and additional information about the associated processes where both the lime and byproduct CO are 2 “reused” are required to quantify the amount of CO that is reabsorbed. Research conducted thus far has not yielded 2 195 the necessary informa reabsorption rates . However, some additional information on the tion to quantify CO 2 consumed on site at lime facilities has been obtained from EPA’s GHGRP. amount of CO 2 196 In some cases, lime is generated from calcium carbonate byproducts at pulp mills The and water treatment plants. In the pulping lime generated by these processes is included in the USGS data for commercial lime consumption. industry, mostly using the Kraft (sulfate) pulping process, lime is consumed in order to causticize a process li quor (green liquor) composed of sodium carbonate and sodium sulfide. The green liquor results from the dilution of the Kraft mills smelt created by combustion of the black liquor where biogenic carbon ) is present from the wood. C ( recover the calcium carbo nate “mud” after the causticizing operation and calcine it back into lime — thereby generating CO generation of lime could be considered a lime — for reuse in the pulping process. Although this re - 2 manufacturing process, the CO emitted during this process is mostly biogenic in origin, and therefore is not 2 included in the industrial processes totals (Miner and Upton 2002). In accordance with IPCC methodological guidelines, any such emissions are calculated by accounting for net C fluxes from changes in biogenic C reservoirs in wooded or crop lands (see the hapter). Land Use, Land - Use Change, and Forestry c In the case of water treatment plants, lime is used i n the softening process. Some large water treatment plants may recover their waste calcium carbonate and calcine it into quicklime for reuse in the softening process. Further research is necessary to determine the degree to which lime recycling is practice d by water treatment plants in the United States. that calcination emissions for LKD are around 2 percent. Another uncertainty is the assumption The National Lime Association (NLA) has commented that the estimates of emissions from LKD in the United States could be closer to 6 percent. They also note that additional emissions (approximately 2 percent) may also be generated through production of other byproducts/wastes (off - spec lime that is not recycled, scrubber sludge) at lime plants (Seeger 2013). ublic ly available on LKD generation rates , total quantities P not used in cement production , and types of other byproducts/wastes produced at lime facilities is limited. EPA initiated a dialogue with NLA to discuss data needs to generate a country - specific LKD fa ctor and is reviewing the information provided by NLA . NLA compiled and shared historical emissions information reported by member facilities and quantities for some waste products associated with generation of tot al calcined byproducts and LKD, as well as methodology and calculation worksheets that member facilities complete when reporting. There is uncertainty regarding the availability of data across the time series needed to generate a representative country specific LKD factor. U ncertainty of the activ ity - 195 Representatives of the National Lime Associat ion estimate that CO reabsorption that occurs from the use of lime may offset 2 emissions from calcination (Males 2003). as much as a quarter of the CO 2 196 Some carbide producers may also regenerate lime from their calcium hydroxide byproducts, which does not result in emissions of CO . In making calcium carbide, quicklime is mixed with coke and heated in electric furnaces. The regeneration of 2 lime in this process is done using a waste calcium hydroxide (hydrated lime) [CaC + Ca(OH) + 2H ], not calcium  C H O 2 2 2 2 2 CaO + H O] carbonate [CaCO  ]. Thus, the calcium hydroxide is heated in the kiln to simply expel the water [Ca(OH) + heat 2 3 2 . is released and no CO 2 15 - 4 Industrial Processes and Product Use

232 data is also of voluntarily reported plant - level production data . Further and completeness a function of the reliability research and data is needed to improve understanding of additional calcination emissions to consider revising the o ns that are based on IPCC g uidelines . More information can be found in the Planned current assumpti Improvements section below. Table 4 10 . The results of the Approach 2 quantitative uncertainty analysis are summarized in Lime CO emissions - 2 2.9 and 1 3.7 MMT CO for 2015 were estimated to be between 1 Eq. at the 95 percent confidence level. This 2 confidence level indicates a range of approximately 3 percent below and 3 pe rcent above the emission estimate of 13.3 MMT CO Eq. 2 4 Table - 10 : Approach 2 Quantitative Uncertainty Estimates for CO Emissions from Lime 2 Production (MMT CO Eq. and Percent) 2 a Emission Estimate Uncertainty Range Relative to Emission Estimate 2015 Source Gas (MMT CO (%) (MMT CO Eq.) Eq.) 2 2 Upper Lower Upper Lower Bound Bound Bound Bound CO Lime Production 2 3% 1 1 3.7 - +3% 2.9 13.3 a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. Methodological were applied to the entire time series to ensure consistency in emissions from 1990 approaches 5 . Details on the emission trends through time are described in more detail in the Methodology section, through 201 above. For more information on the general QA/QC process applied to this source category, consistent with Volume 1, 2006 IPCC Guidelines , see QA/QC and Verification Procedures section in the introduction of the Chapter 6 of the IPPU Chapter. Recalculations Discussion Updated data from Lisa Corathers (U.S. Geological Survey) (Corathers 2017) resulted in High - Calcium Quicklime production data changes for 2014 and Dolomi tic Quicklime production data changes for 2013 and 2014, as shown in Table 4 - 8 . Recovered emissions shown in Table - 7 were updated using aggregated GHGRP data from 2010 to 2015. This data 4 changed slightly from previous Inventory reports due to the adoption of new rounding technique to maintain consistency with other data sets. Bo th of these data updates resulted in changes to emissions estimates across the series (2011 to 2015) of less than 1 percent relative time to the previous Inventory (i.e., 1990 through 2014). Planned Improvements Future improvements involve finishing a review of data to improve current assumptions associated with emissions from production of LKD and other byproducts/wastes he Uncertainty and Time - Series Consistency as discussed in t section per comments from the NLA provided during the public review of th e draft 1990 to 2015 Inventory. In response to comments, EPA met with NLA on April 7, 2015 to outline specific information required to apply IPCC methods to develop a country - specific correction factor to more accurately estimate emissions from production of LKD. In response to this technical meeting, in January and February 2016, NLA compiled and shared historical emissions information reported by member facilities on an annual basis under voluntary reporting initiatives over ith generation of total calcined byproducts and LKD ( LKD reporting only 2002 through 2011 associated w ). This emissions information was reported on a voluntary basis consistent with NLA’s differentiated starting in 2010 level reporting protocol also recently provided. EPA needs - additional time to review the information facility provided by NLA and plans to work with them to address needs for EPA’s analysis, as there is limited information across the time nned series. Due to limited resources and need for additional QA of information, this pla improvement is still in process and has not been incorporated into this current Inventory report. As an interim step, EPA has updated the qualitative description of uncertainty to reflect the information provided by NLA. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 16 4

233 In addition, EPA plans to also review GHGRP emissions and activity data reported to EPA under Subpart S, in particular, review of aggregated activity data on lime production by type. Particular attention will be made to also - of the emissions estimates p resented in future I nventory reports, consistent with series consistency ensuring time , with the - level reporting data from EPA’s GHGRP This is required IPCC and UNFCCC guidelines. as the facility are not avai program's initial requirements for reporting of emissions in calendar year 2010, lable 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 197 be relied upon. Glass Production (IPCC Source Category 4.3 ) 2A 3 Glass production is an energy and raw - material intensive process that results in the generation of CO from both the 2 uels consumed for energy purposes energy consumed in making glass and the glass process itself. Emissions from f . sector during the production of glass are accounted for in the Energy Glass production employs a variety of raw materials in a glass - batch. These include formers, fluxes, stabilizers, and The major r aw materials (i.e., fluxes and stabilizers) which emit process - related carbon sometimes colorants. dioxide ( CO ) emissions during the glass melting process are limestone, dolomite, and soda ash. The main former in 2 Fluxes Other major formers in glass include feldspar and boric acid (i.e., borax). ). all types of glass is silica (SiO 2 Most commonly used flux materials are soda ash are added to lower the temperature at which the batch melts. o make glass more CO Stabilizers are used t ) and potash (potassium carbonate, K (sodium carbonate, Na O). 2 2 3 chemically stable and to keep the finished glass from dissolving and/or falling apart. Commonly used stabilizing agents in glass production are limestone (CaCO ), dolomite (CaCO MgCO ), alumina (Al O ), magnesia (MgO), 2 3 3 3 3 barium carbonate (B a C O ), strontium carbonate (SrCO ), lithium carbonate (Li (ZrO CO ) , and zirconia ) (OIT 3 2 2 3 3 2002). - house return of Glass makers also use a certain amount of recycled scrap glass (cullet), which comes from in lage or retention such as recycling or cullet broker services. glassware broken in the process or other glass spil - primarily limestone, dolomite and soda ash) release CO raw materials emissions The ( in a complex high 2 temperature chemical reaction during the glass melting process. This process is not direct ly comparable to the calcination process used in lime manufacturing, cement manufacturing, and process uses of carbonates (i.e., glass industry emissions (IPCC 2006). The limestone/dolomite use), but has the same net effect in terms of CO U.S. 2 ded into four main categories: containers, flat (window) glass, fiber glass, and specialty glass. The can be divi 20 09 tainer and flat glass (EPA The United States is one of the major majority of commercial glass produced is con ). global exporters of glass. Domestically, demand comes mainly from the construction, auto, bottling, and container industries. There are over 1,500 companies that manufacture glass in the United States, with the larges t being 198 Corning, Guardian Industries, Owens - Illinois, and PPG Industries. In 2015, 699 kilotons of limestone and 2,390 kilotons of soda ash were consumed for glass production (USGS 2015 c ; Willett 2017). Dolomite consumption data for glass manufacturing wa s reported to be zero for 2015. Use of limestone and soda ash in glass production resulted in aggregate CO emissions of 1.3 MMT CO Eq. (1,299 kt ) (see 2 2 . 4 Table ) . Overall, emissions have decreased 15 percent from 1990 through 2015 - 11 Emissions in 2015 decreased approximately 3 percent from 2014 levels while, in general, e missions from glass production have remained relatively constant over the time seri es with some fluctuations since 1990 . In general, these fluctuations were related to the behavior of the export market and the U.S. economy. Specifically, the extended 08 and 2010 resulted in downturn in residential and commercial construction and automotive industries between 20 reduced consumption of glass products, causing a drop in global demand for limestone/dolomite and soda ash, and a corresponding decrease in emissions. Furthermore, the glass container sector is one of the leading soda ash 197 See < http://www.ipcc - nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf >. 198 Excerpt from Glass & Glass Product Manufacturing Industry Profile, First Research. Available online at: - duct - Pro >. Manufacturing.html < http://www.firstresearch.com/Industry - Research/Glass - and - Glass 17 - 4 Industrial Processes and Product Use

234 consuming sectors in the United States. Some commercial food and beverage package manufacturers are shifting from glass containers towards lighter and more cost effective polyethylene terephthalate (PET) based containers, putting downward pressure on domestic consum ption of soda ash (USGS 1995 through 2015c). 4 - 11 : CO Eq. and Emissions from Glass Production ( MMT CO Table kt ) 2 2 Year MMT CO Eq. kt 2 1990 1.5 1,535 2005 1.9 1,928 2011 3 1, 299 1. 2012 1.2 1,248 1.3 1,317 2013 2014 1.3 1,336 1,299 1.3 2015 Note: Totals may not sum due to independent rounding Methodology ns were calculated based on the 2006 Carbon dioxide emissio Guidelines Tier 3 method by multiplying the IPCC quantity of input carbonates (limestone, dolomite, and soda ash) by the carbonate - based emission factor (in metric tons CO /metric ton carbonate): limestone , 0.43971; dolomite , 0.47732; and soda ash , 0.41492. 2 Consumption 5 of limestone, dolomite, and soda ash used for glass manufacturing were data for 1990 through 201 obtained from the USGS) Minerals Yearbook: Crushed Stone Annual Report (1995 through U.S. Geological Survey ( 2015b), 2015 preliminary data from the USGS Crushed Stone Commodity Expert (Willett 2017), the USGS Minerals Yearbook: Soda Ash Annual Report 5) (USGS 1995 through 2015c), USGS Mineral (1995 through 201 Industry Surveys for Soda Ash in January 2015 (USGS 20 15a) and the U.S. Bureau of Mines (1991 and 1993a), which are reported to the nearest ton. During 1990 and 1992, the USGS did not conduct a detailed survey of - limestone and dolomite consumption by end use. Consumption for 1990 was estimated by applying the 1991 percentages of total limestone and dolomite use constituted by the individual limestone and dolomite uses to 1990 total use. Similarly, the 1992 consumption figures were approximated by applying an average of the 1991 and 1993 mestone and dolomite use constituted by the individual limestone and dolomite uses to the percentages of total li 1992 total. Additionally, each year the USGS withholds data on certain limestone and dolomite end - uses due to confidentiality data. agreements regarding company proprietary For the purposes of this analysis, emissive end - uses that contained withheld data were estimated using one of the following techniques: (1) the value for all the withheld data points for limestone or dolomite use was distributed evenly to all withhe ld end - uses; or (2) the average percent of total limestone or dolomite for the withheld end - use in the preceding and succeeding years. There is a large quantity of limestone and dolomite reported to the USGS under the categories “unspecified – reported” and “unspecified – estimated.” A portion of this consumption is believed to be limestone or dolomite used for glass manufacturing. The quantities listed under the “unspecified” categories were, therefore, allocated to glass manufacturing according to the percen t limestone or dolomite consumption for glass manufacturing end use for that 199 year. Based on the 201 5 reported data, the estimated distribution of soda ash consumption for glass production compared 1995 through S ). (USG to total domestic soda ash consumption is 48 percent 2015c 199 This approach was recommended by USGS. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 18 4

235 Table 4 : Limestone, Dolomite, and Soda Ash Consumption Used in Glass Production ( kt ) - 12 1990 2005 2011 2012 2013 2014 2015 Activity Limestone 430 920 614 555 693 765 699 0 59 541 0 0 0 0 Dolomite Soda Ash 3,17 7 2,480 3,050 2,420 2,440 2,410 2,390 3,175 3,666 4,511 3, 094 2,975 3,133 3,089 Total Uncerta inty and Time - Series Consistency The uncertainty levels presented in this section arise in part due to variations in the chemical composition of limestone used in glass production. In addition to calcium carbonate, limestone may contain smaller amounts of magnesia, silica, and sulfur, among other minerals (potassium carbonate, strontium carbonate and barium carbonate, and dead burned dolomite). Similarly, the quality of the limestone (and mix of carbonates) used for glass manufacturing will depend on the type of glass being manufactured. The estimates below also account for uncertainty associated wit h activity data. Large fluctuations in reported The uncertainty resulting consumption exist, reflecting year to - year changes in the number of survey responders. - from a shifting survey population is exacerbated by the gaps in the time series of reports. The accuracy of distribution by end use is also uncertain because this value is reported by the manufacturer of the input carbonates , there has been no reported consumption of (limestone, dolomite and soda ash) and not the end user. For 201 5 dolomite for glas - been reported to USGS by dolomite manufacturers and not end ve s manufacturing. data ha Th ese users (i.e., glass manufacturers). There is a high uncertainty associated with this estimate, as dolomite is a major raw Add itionally, there is significant inherent uncertainty associated with material consumed in glass production. estimating withheld data points for specific end uses of limestone and dolomite. The uncertainty of the estimates for limestone and dolomite used in glass making is especially high. y, much of the limestone consumed in the Lastl United States is reported as “other unspecified uses;” therefore, it is difficult to accurately allocate this unspecified uses. quantity to the correct end sources of data on - Further research is needed into alternate and more complete - based raw material consumption by the glass industry. This year, EPA reinitiated dialogue with the USGS carbonate National Minerals Information Center Crushed Stone commodity expert to assess the current uncertainty ranges associated wi th quantities of carbonates consumed for glass production compiled and published in USGS reports. Table 4 - 13 . In 2015, glass The results of the Approach 2 quantitative uncertainty analysis are summarized in production CO emissions were estimated to be between 1. 2 and 1.4 MMT CO Eq. at the 95 percent confidence 2 2 percent above the emission estimate of 1.3 This indicates a range of approximately 4 percent below and 5 level. Eq. MMT CO 2 - Table : Approach 2 Quantitative Uncertainty Estimates for CO Emissions from Glass 4 13 2 Production (MMT CO Eq. and Percent) 2 a Uncertainty Range Relative to Emission Estimate 2015 Emission Estimate Gas Source (MMT CO Eq.) (MMT CO Eq.) (%) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound Glass Production CO 1.3 1. 2 1.4 - 4% +5% 2 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 consistency in emissions from 1990 through 201 5 . 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, , see QA/QC and Verification Procedures section in the introduction of the Chapter 6 of the 2006 IPCC Guidelines IPPU Chapter. 19 - 4 Industrial Processes and Product Use

236 Recalculations Discussion for 2014 were Limestone and dolomite consumption data revised relative to the previous Inventory based on the . obtained directly from the USG S Crush ed Stone Commodity expert, Jason Willett (Willett 2017) preliminary data In the previous Inventory (i.e., 1990 through 2014), preliminary data were used for 2014 updated for , which were the current Inventory. The published time series was reviewed to ensure time - se ries consistency. This update caused a decrease in 2014 emissions of less than 1 percent compared to 2014 emissions presented in the previous Inventory (i.e., 1990 through 2014) . Planned Improvements As noted in the lable activity data shows consumption of only limestone and previous reports, current publicly avai soda ash for glass manufacturing. While limestone and soda ash are the predominant carbonates used in glass , although in smaller manufacturing, there are other carbonates that are also consumed for glass manufacturing quantities. EPA has initiated review of available activity data on carbonate consumption by type in the glass industry from EPA’s Greenhouse Gas Reporting Program (GHGRP) reported annually since 2010, as well as USGS publications. EPA has initiated review of this activity data and anticipates to finalize assessment for future integration of data reported under EPA’s GHGRP in the spring of 2017 to improve the completeness of emission estimates and - facilitate category ume 1 of the specific QC per Vol 2006 IPCC Guidelines for the Glass Production source category. EPA’s GHGRP has an emission threshold for reporting, so the assessment will consider the completeness of carbonate consumption data for glass production in the United States. Parti cular attention will also be made to also ensuring time - series consistency of the emissions estimates presented in future I nventory reports, consistent with with the , IPCC and UNFCCC guidelines. This is required as the facility - level reporting data from EPA’s GHGRP 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 level data in national inventories data from EPA’s GHGRP, the latest guidance from the IPCC on the use of facility - 200 These planned improvements are ongoing and EPA will be relied upon. also initiate research into other may by the glass industry. sources of activity data for carbonate consumption Other Process Uses of Carbonates (IPCC 4.4 ) 4 Source Category 2A 201 ), dolomite (CaCO ) MgCO siderite Limestone (CaCO , are and other carbonates such as soda ash, magnesite, and 3 3 3 basic materials used by a wide variety of industries, including construction, agriculture, chemical, metallurgy, glass For production, and environmental pollution control. This section addresses only limestone and dolomite use . applications industrial heated sufficiently enough to calcine the , carbonates such as limestone and dolomite are material and generate CO as a byproduct. 2 →퐶푎푂 +퐶푂 퐶푎퐶푂 3 2 푀푔퐶 푂 → 푀푔푂 + 퐶 푂 2 3 Examples of such applications include limestone used as a flux or purifie r in metallurgical furnaces, as a sorbent in flue gas desulfurization (FGD) systems for utility and industrial plants, and as a raw material for the production of glass, lime, and cement. Emissions from limestone and dolomite used in other process sectors such as cement, lime, glass production, and iron and steel, are excluded from this section and reported under their respective source 200 - See < nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf >. http://www.ipcc 201 Limestone and dolomite are collectively referred to as limestone by the industry, and intermediate varieties are seldom distinguished. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 20 4

237 categories (e.g., Section 4.3 , Glass Production). Emission from soda ash consumption is reported under respective categories (e.g., Glass Manufacturing (IPCC Source Category 2A3) and Soda Ash Production and Consumption rgy purposes during these processes are (IPCC Source Category 2B7)). Emissions from fuels consumed for ene . accounted for in the Energy chapter Large Limestone is widely distributed throughout the world in deposits of varying sizes and degrees of purity. deposits of limestone occur in nearly every state in the United Sta tes, and significant quantities are extracted for In 2014, the leading limestone producing states are Texas, Missouri, Florida, Ohio, and industrial applications. ilarly, dolomite Kentucky, which contribute 43 percent of the total U.S. output (USGS 1995a through 2015). Sim deposits are also widespread throughout the world. Dolomite deposits are found in the United States, Canada, Mexico, Europe, Africa, and Brazil. In the United States, the leading dolomite producing states are Illinois, Pennsylvania, and Ne w York, which contribute 55 percent of the total 2014 U.S. output (USGS 1995a through 2015). limestone and 3,244 kt of dolomite were consumed for these emissive applications, excluding In 2015, 23,251 kt of emissions of 1 . estone and dolomite resulted in aggregate C ) glass manufacturing ( Usage of lim 1.2 Willett 2017b O 2 Eq. (1 1,236 kt) ( see MMT CO 4 - 14 and Table 4 - 15 ). While 2015 emissions have decreased 5 percent Table 2 verall emissions have 12 9 percent from 1990 through 2015. compared to 2014, o increased - 14 : CO Table Emissions from Other Process Uses of Carbonates (MMT CO 4 Eq.) 2 2 Other Magnesium Miscellaneous a Year Production Flux Stone Uses FGD Total 1.4 0.1 0.8 4.9 2.6 1990 2005 2.6 3.0 0.0 0.7 6.3 1. 5 5.4 0.0 2.4 9. 3 2011 0.0 1.1 5.8 2012 1.1 8.0 2013 2.3 6.3 0.0 1.8 10.4 11.8 1.8 2014 2.9 7.1 0.0 7.3 0.0 0.9 11.2 3.0 2015 a “Other 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. 4 - 15 : CO Table Emissions from Other Process Uses of Carbonates (kt) 2 Other Magnesium Miscellaneous a Total Production Flux Stone Uses FGD Year 64 2,592 1,432 1990 819 4,907 2,973 2005 2,649 6,339 0 718 2011 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 2,911 7,111 0 1,790 11,811 2014 2015 3,031 7,335 0 871 11,236 a “Other 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. Methodology Tier 2 method by multiplying the 2006 IPCC Guidelines Carbon dioxide emissions were calculated based on the quantity of limestone or dolomite consumed by the emission factor for limestone or dolomite calcination, 21 - 4 Industrial Processes and Product Use

238 respectively, Table 2.1 limestone: 0.43971 – /metric ton carbonate, and dolomite: 0.47732 metric ton metric ton CO 2 202 This methodology was used for flux stone, flue gas desulfurization systems, chemical CO /metric ton carbonate. 2 stone, mine dusting or acid water treatment, acid neutralization, and sugar refining. Flux stone used during the production of iron and steel was deducted from the Other Process Uses of Carbonates source category estimate and attributed to the Iron and Steel Production source category estimate. Similarly, limestone and dolomite consumption lime manufacturing are excluded from this categ ory and attributed to their for glass manufacturing, cement, and respective categories. Historically, the production of magnesium metal was the only other significant use of limestone and dolomite that 1 emissions. At the end of 200 produced CO , the sole magnesium production plant operating i n the United States 2 that produced magnesium metal using a dolomitic process that resulted in the release of CO emissions ceased its 2 b through 201 operations (USGS 1995 ). 2; USGS 2013 Consumption data for 1990 through 2015 of limestone and dolomite used for flux stone, flue gas desulfurization Table systems, chemical stone, mine dusting or acid water treatment, acid neutralization, and sugar refining (see ( - ) were obtai ned from the U.S. Geological Survey 16 4 Minerals Yearbook: Crushed Stone Annual Report USGS) (1995a through 2015), preliminary data for 2015 from USGS Crushed Stone Commodity Expert (Willett 2017b), American Iron and Steel Institute limestone and dolomite con sumption data (AISI 2016), and the U.S. Bureau of Mines (1991 and 1993a), which are reported to the nearest ton. The production capacity data for 1990 through 2015 of dolomitic magnesium metal also came from the USGS (1995b through 2012; USGS 2013) and the U.S. Bureau of Mines (1990 through 1993b). During 1990 and 1992, the USGS did not conduct a detailed survey of limestone and dolomite consumption by end use. Consumption for 1990 was estimated by applying the 1991 percentages of total - limestone and dolomi te use constituted by the individual limestone and dolomite uses to 1990 total use. Similarly, the 1992 consumption figures were approximated by applying an average of the 1991 and 1993 percentages of total ividual limestone and dolomite uses to the 1992 total. limestone and dolomite use constituted by the ind Additionally, each year the USGS withholds data on certain limestone and dolomite end - uses due to confidentiality - uses that contained agreements regarding company proprietary data. For the purposes of this analysis, emissive end d using one of the following techniques: (1) the value for all the withheld data points for withheld data were estimate limestone or dolomite use was distributed evenly to all withheld end - uses; (2) the average percent of total limestone or dolomite for the withheld end - use in the pr eceding and succeeding years; or (3) the average fraction of total limestone or dolomite for the end use over the entire time period. - There is a large quantity of crushed stone reported to the USGS under the category “unspecified uses.” A portion of this consumption is believed to be limestone or dolomite used for emissive end uses. The quantity listed for “unspecified uses” was, therefore, allocated to each reported end - use according to each end - use’s fraction of total 203 consumption in that year. Table 4 - 16 : Limestone and Dolomite Consumption (kt) 2014 2013 2015 Activity 1990 2005 2011 2012 6,737 7,022 4, 396 3,666 6,345 7,599 7,834 Flux Stone Limestone 5,804 3,165 2,531 3,108 4,380 4,243 4,590 933 3,857 1,865 Dolomite 559 1,965 3,356 3,244 12,326 FGD 3,258 6,761 13,185 14,347 16,171 16,680 2,610 1,835 1,632 5,548 3,973 4,069 1,980 Other Miscellaneous Uses Total 11,830 15,415 22,270 19,461 24,665 27,839 26,494 Uncerta inty and Time - Series Consistency The uncertainty levels presented in this section account for uncertainty associated with activity data. Data on limestone and dolomite consumption are collected by USGS through voluntary national surveys. USGS contac ts the mines (i.e., producers of various types of crushed stone) for annual sales data. Data on other carbonate consumption 202 IPCC Guidelines , Volume 3: Chapter 2. 2006 203 This approach was recommended by USGS, the data collection agency. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 22 4

239 are not readily available. - users and industry types. The producers report the annual quantity sold to various end and USGS estimates the the rest historical response rate for the crushed stone survey to be approximately 70 percent, year changes in the is estimated by USGS. to - Large fluctuations in reported consumption exist, reflecting year - number of survey responders. The uncertainty resulting from a shifting survey population is exacerbated by the gaps in the time series of reports. The accuracy of distribution by end use is also uncertain because this value is reported by the producer/mines and not the end user. Additionally, there i s significant inherent uncertainty associated with estimating withheld data points for specific end uses of limestone and dolomite. Lastly, much of the limestone consumed in the United States is reported as “other unspecified uses;” therefore, it is diffic ult to accurately allocate this unspecified quantity to the correct end - uses. This year, EPA re initiated dialogue with the USGS National Minerals Information Center Crushed Stone c ommodity expert to assess the current uncertainty ranges associated data with the limestone and dolomite consumption USGS. compiled and published by During this discussion, the expert confirmed that EPA’s range of uncertainty was still reasonable (Willett 2017a). lso arises in part due to variations in the chemical composition of limestone. In Uncertainty in the estimates a addition to calcium carbonate, limestone may contain smaller amounts of magnesia, silica, and sulfur, among other The exact specifications for limestone or dolomite used as flux stone vary with the pyrometallurgical minerals. process and the kind of ore processed. Table 4 The results of the Approach 2 quantitative uncertainty analysis are summarized in 17 . Carbon dioxide - emissions from other process uses of carbonates in 2015 were estimated to be between 9.9 and 13.2 MMT CO Eq. 2 This indicates a range of approximately 1 3 percent below an d 1 6 at the 95 percent confidence level. percent above the emission estimate of 11.2 MMT CO Eq. 2 Emissions from Other Table 4 - 17 : Approach 2 Quantitative Uncertainty Estimates for CO 2 Eq. and Percent) Process Uses of Carbonates (MMT CO 2 2015 Emission a Source Estimate Uncertainty Range Relative to Emission Estimate Gas (MMT CO Eq.) (MMT CO Eq.) (%) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound Other Process Uses 9.9 % 13.2 6 +1 CO 11.2 % - 1 3 2 of Carbonates a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. approaches were applied to the entire time series to ensure consistency in emissions from 1990 Methodological 5 Details on the emission trends through time are described in more detail in the Methodology section, . through 201 above. consistent with Volume 1, For more information on the general QA/QC process applied to this source category, QA/QC and V 2006 IPCC Guidelines , see Chapter 6 of the erification Procedures section in the introduction of the IPPU Chapter. Recalculations Discussion Limestone and dolomite consumption data by end - use for 2014 were updated relative to the previous Inventory sh Stone Commodity expert, Jason Willett . In the previous based on the preliminary data provided by USGS Cru Inventory (i.e., 1990 through 2014), preliminary data were used for 2014 which were updated for the current Inventory . The published time series was reviewed to ensure time - series consistency. This update caused a decrease in total limestone and dolomite consumption for emissive end uses in 2014 by approximately 2 percent , relative to . the previous report Planned Improvements Pending available resources, this section will integrate and present emiss ions from soda ash consumption for other chemical uses (non - glass production). Currently, in this document, these estimates are presented along with emissions from soda ash production (IPCC Category 2B7). This improvement is planned and will be implemented 23 - 4 Industrial Processes and Product Use

240 into the next Inventory report (i.e., 1990 to 2016). EPA also plans to continue the dialogue with USGS to assess uncertainty ranges for activity data used to estimate emissions from other process use of carbonates. Ammonia Production (IPCC Source 4.5 Categor y 2B1) ) Emissions of carbon dioxide (CO occur during the production of synthetic ammonia, primarily through the use of 2 natural gas, petroleum coke, or naphtha as a feedstock. The natural gas - , naphtha - , and petroleum coke - based processes produce CO ), the latter of which is used in the production of ammonia. The brine and hydrogen (H 2 2 electrolysis process for production of ammonia does not lead to process - based CO emissions. Emissions from fuels 2 consumed for energy purposes during the production of ammonia are accounted for in the Energy chapter. In the United States, the majority of ammonia is produced using a natural gas feedstock; however, one synthetic ammonia production plant located in Kansas is producing ammonia from petroleum coke feedstock. In some U.S. plants, some of the CO produced by the process is captured and used to produce urea rather than being emitted to 2 the atmosphere. There are approximately 13 companies operating 26 ammonia producing facilities in 17 states. of domestic ammonia production capacity is concentrated in the states of Louisiana (29 More than 55 percent percent), Oklahoma (20 percent), and Texas (6 percent) (USGS 2016). The primary There are five principal process steps in synthetic ammonia production from natural gas feedstock. , carbon monoxide (CO), and H to CO reforming step converts methane (CH ) in the presence of a catalyst. Only 2 4 2 30 to 40 percent of the CH feedstock to the primary reformer is converted to CO and CO in this step of the 2 4 process. The secondary r eforming step converts the remaining CH feedstock to CO and CO . The CO in the process 2 4 gas from the secondary reforming step (representing approximately 15 percent of the process gas) is converted to in the presence of a catalyst, water, and air in th e shift conversion step. Carbon dioxide is removed from the CO 2 process gas by the shift conversion process, and the hydrogen gas is combined with the nitrogen (N ) gas in the 2 process gas during the ammonia synthesis step to produce ammonia. The CO d in a waste gas stream with is include 2 other process impurities and is absorbed by a scrubber solution. In regenerating the scrubber solution, CO is 2 released from the solution. , including the primary a nd secondary reforming The conversion process for conventional steam reforming of CH 4 and the shift conversion processes, is approximately as follows: + 푁 24퐻 0. 88퐶퐻 푂 →0. 88퐶푂 +1. 26퐴푖푟 +1. +3퐻 2 2 2 2 4 퐻 + 3 퐻 → 2 푁 푁 3 2 2 To produce synthetic ammonia from petroleum coke, the petroleum coke is gasified and converte d to CO and H . 2 2 These gases are separated, and the H is used as a feedstock to the ammonia production process, where it is reacted 2 with N to form ammonia. 2 Not all of the CO produced during the production of ammonia is emitted directly to the atmosphere. Some of the 2 ammonia and some of the CO produced by the synthetic ammonia process are used as raw materials in the 2 production of urea [CO(NH ) ], which has a variety of agric ultural and industrial applications. 2 2 The chemical reaction that produces urea is: + 퐶푂 →푁퐻 퐶푂푂푁퐻 → 퐶푂(푁퐻 ) +퐻 푂 2푁퐻 3 2 2 2 2 4 2 Only the CO emitted directly to the atmosphere from the synthetic ammonia production process is accounted for in 2 dete rmining emissions from ammonia production. The CO that is captured during the ammonia production process 2 and used to produce urea does not contribute to the CO emission estimates for ammonia production presented in 2 sulting from the consumption of urea are attributed to the urea consumption or this section. Instead, CO emissions re 2 urea application source category (under the assumption that the carbon stored in the urea during its manufacture is released into the environment during its consumption or appli resulting from agricultural cation). Emissions of CO 2 applications of urea are accounted for in the Agriculture chapter. Previously, these emission estimates from the ction of the Land Use, se Cropland Remaining Cropland agricultural application of urea were accounted for in the - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 24 4

241 Land Use Change, and Forestry chapter. Emissions of CO resulting from non - agricultural applications of urea (e.g., 2 - use as a feedstock in chemical production processes) are accounted for in the Urea Consumption for Non Agricultural P urposes section of this chapter. were 10.8 MMT CO Eq. (10,799 kt), and are from ammonia production in 2015 Total emissions of CO 2 2 4 summarized in 18 and Table 4 - 19 . Ammonia production relies on natural gas as both a feedstock and a fuel, Table - market fluctuations and volatility in natural gas prices affect the produc tion of ammonia . Since 1990, and as such, decreased by 17 percent. emissions from ammonia production have Emissions in 2015 have increased by approximately 12 percent from the 2014 levels. 4 - 18 : CO Emissions from Ammonia Production (MMT CO Eq.) Table 2 2 Source 1990 2005 2011 2012 2013 2014 2015 Ammonia Production 13.0 9.2 9.3 10. 0 9.6 10.8 9.4 Total 13.0 9.2 9.3 9.4 10. 0 9.6 10.8 Table 4 - 19 : CO Emissions from Ammonia Production (kt) 2 2012 2015 Source 1990 2005 2011 2014 2013 13,047 9,196 9,292 Ammonia Production 9,377 9,962 9,619 10,799 619 Total 13,047 9,196 9,292 9,377 9,962 9 , 10,799 Methodology 2006 Carbon dioxide emissions from production of synthetic ammonia from natural gas feedstock is based on the IPCC Guidelines (IPCC 2006) Tier 1 and 2 method. A country - specific emission factor is developed and applied to national ammonia production to estimate emissions. The method uses a CO emission factor published by the 2 - based ammonia production European Fertilizer Manufacturers Association (EFMA) that is based on natural gas technologies that are similar to those employed in the Unite d States. This CO emission factor of 1.2 metric tons 2 CO /metric ton NH (EFMA 2000a) is applied to the percent of total annual domestic ammonia production from 2 3 natural gas feedstock. from ammonia production are then adjusted to account for the use of some of the CO produced Emissions of CO 2 2 from ammonia production as a raw material in the production of urea. The CO emissions reported for ammonia 2 production are reduced by a factor of 0.733 multiplied by total annual domestic urea production. This corresponds to a stoichiometric CO (IPCC /urea factor of 44/60, assuming complete conversion of ammonia (NH ) and CO to urea 2 2 3 2006; EFMA 2000b). All synthetic ammonia production and subsequent urea productio n are assumed to be from the same process — conventional catalytic reforming of natural gas feedstock, with the exception of ammonia production from petroleum coke feedstock at one plant located in Kansas. Annual ammonia and urea production are shown in Table specific data, 4 - 20 . The CO - emission factor for production of ammonia from petroleum coke is based on plant 2 wherein all carbon contained in the petroleum coke feed stock that is not used for urea production is assumed to be (Bark 2004). Ammonia and urea are assumed to be manufactured in the same emitted to the atmosphere as CO 2 manufacturing complex, as both the raw materials needed for urea production are produced b y the ammonia production process. The CO emission factor of 3.57 metric tons CO for the petroleum coke /metric ton NH 3 2 2 feedstock process (Bark 2004) is applied to the percent of total annual domestic ammonia production from petroleum coke feedstock. The /metric ton NH for production of ammonia from natural gas feedstock emission factor of 1.2 metric ton CO 2 3 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 CO /metric ton NH , with 1.2 metric ton 2 3 /metric ton NH CO as a typical value (EFMA 2000a). Technologies (e.g., catalytic reforming process, etc.) 3 2 tes for use of natural gas associated with this factor are found to closely resemble those employed in the United Sta feedstock to the catalytic as a feedstock. The EFMA reference also indicates that more than 99 percent of the CH 4 reforming process is ultimately converted to CO . As noted earlier, emissions from fuels consumed for energy 2 25 - 4 Industrial Processes and Product Use

242 es during the production of ammonia are accounted for in the Energy chapter purpos The total ammonia production . through 5 w ere obtained from American Chemistry Council ( 2016 ). For years before 2011, data for 2011 201 obtained from Coffeyville Resources (Coffeyville 2005, 2006, see - 20 ) were 4 a mmonia production data ( Table f Commerce (U.S. and the Census Bureau of the U.S. Department o 2007a, 2007b, 2009, 2010, 2011, and 2012) 1 ) as reported in Current Industrial Reports Fertilizer Materials Census Bureau 1991 through 1994 , 1998 through 201 - ammonia nitrate production from petroleum coke for years and Related Products annual and quarterly reports. Urea was obtained from Coffeyville Resources (Coffeyville 2005, 2006, 2007a, 2007b, 2009, 2010, 2011, 1 through 201 , and from CVR Energy, Inc. Annual Report (CVR 2012 ,2014, 2015, and 2016) for 2012, 2013, 2014, and and 2012) 2015. Urea production data for 1990 through 200 (USGS 8 were obtained from the Minerals Yearbook: Nitrogen 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 product ion statistics, and u rea , 2012, 2013 and 2014 were obtained from the Minerals Yearbook: Nitrogen production data for 2011 (USGS 2015, 2016). USGS urea production data for 2015 was not yet published and so 2014 data were used as a proxy for 2015. 4 - : Ammonia Production and Urea Production (kt) Table 20 Ammonia Urea Year Production Production 15,425 7,450 1990 2005 10,143 5,270 2011 10,325 5, 430 2012 10,305 5,220 5,480 2013 10,930 5,230 2014 10.515 2015 11,505 5,230 Uncertainty and Time - Series Consistenc y The uncertainties presented in this section are primarily due to how accurately the emission factor used represents an natural gas feedstock. average across all ammonia plants using Uncertainties are also associated with ammonia production estimates and the assumption that all ammonia production and subsequent urea production was from the — th the exception of one ammonia same process conventional catalytic reforming of natural gas feedstock, wi production plant located in Kansas that is manufacturing ammonia from petroleum coke feedstock. Uncertainty is - based ammonia also associated with the representativeness of the emission factor used for the petroleum coke ess. It is also assumed that ammonia and urea are produced at collocated plants from the same natural gas raw proc The uncertainty of the total urea production activity data , based on USGS Minerals Yearbook: Nitrogen material. is a function of th e reliabi data, lity of reported production data and is influenced by the completeness of the survey respons es. In addition, due to the fact that 2015 nitrogen data has yet to be published, 2014 is used as a proxy which may result in greater uncertainty. f rom ammonia production plants for purposes other than urea production (e.g., commercial sale, Recovery of CO 2 emissions from ammonia production, as data concerning the etc.) has not been considered in estimating the CO 2 disposition of recovered CO Such recovery may or may not affect the overall estimate of CO are not available. 2 2 emissions depending upon the end use to which the recovered CO is applied. Further research is required to 2 determine whether byproduct CO for application to end is being recovered from other ammonia production plants 2 uses that are not accounted for elsewhere. The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4 - 21 . Carbon dioxide emissions from ammonia production in 2015 were estimated to be between 9.9 and 1 1.7 MMT CO Eq. at the 95 2 This indicates a range of approximately 8 percent below and 8 percent above the emission percent confidence level. 10.8 f estimate o Eq. MMT CO 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 26 4

243 Table 4 21 : Approach 2 Quantitative Uncertainty Estimates for CO - Emissions from 2 Ammonia Production (MMT CO Eq. and Percent) 2 a Relative to Emission Estimate 2015 Emission Estimate Uncertainty Range Gas Source (MMT CO Eq.) (MMT CO Eq.) (%) 2 2 Upper Lower Lower Upper Bound Bound Bound Bound Ammonia Production CO 10.8 9.9 1 1.7 - 8% +8% 2 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 consistency in emissions from 1990 through 201 5 . Details on the e mission 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, 2006 IPCC Guidelines Chapter 6 of the , see QA/QC and Verification Procedures section in the introduction of the IPPU Chapter. Discussion Recalculations Production estimates for urea production for 2014 were updated relative to the previous Inventory using information Minerals Ye (USGS 2016). For the previous version of the Inventory arbook: Nitrogen obtained from the recent 2014 (i.e., 1990 through 2014), 2014 data were not published and so 2013 activity data was used as a proxy. This update for 2014 relative to the previous Inventory. resulted in a slight increase of emissions by approximately 2 percent Planned Improvements Future improvements involve continuing to evaluate and analyze data reported under EPA’s GHGRP to improve the emission estimates for the Ammonia Production source category, in particular new data from updated reporting 204 requirements finalized in October of 2014 (79 FR 63750) and December 2016 (81 FR 89188), that include - facility level ammonia production data, will be included in future reports if the data meets GHGRP CBI aggregation ia. Particular attention will be made to ensure time - series consistency of the emission estimates presented in criter future nventory reports, along with application of appropriate category - specific QC procedures consistent with I 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 - level data in national inventories will from EPA’s GHGRP, the latest guidance from the IPCC on the use of facility 205 Specifically, the planned improvements include assessing data to updat e the emission factors to be relied upon. include both fuel and feedstock CO capture and storage. Methodologies will also be emissions and incorporate CO 2 2 updated if additional ammonia production plants are found to use hydrocarbons other than natural gas for ammonia pro duction. Due to limited resources, this planned improvement is still in development and so is not incorporated into this Inventory. 204 See < https://www.epa.gov/ghgreporting/historical - rulemakings >. 205 - http://www.ipcc See < >. nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf 27 - 4 Industrial Processes and Product Use

244 4.6 Agricultural Urea Consumption for Non - Purposes Urea is produced using ammonia and carbon dioxide (CO as raw materials. All urea produced in the United States ) 2 are generated. There are is assumed to be produced at ammonia production facilities where both ammonia and CO 2 approximately 20 of these facilities operating in the United States. The chemical reac tion that produces urea is: 2푁퐻 + 퐶푂 →푁퐻 퐶푂푂푁퐻 → 퐶푂(푁퐻 ) +퐻 푂 4 2 2 2 2 2 3 emissions associated with urea consumed exclusively for non - This section accounts for CO agricultural purposes. 2 Carbon dioxide emissions associated with urea consumed for fe c hapter. rtilizer are accounted for in the Agriculture Urea is used as a nitrogenous fertilizer for agricultural applications and also in a variety of industrial applications. nts, resins, fillers, analytical reagents, The industrial applications of urea include its use in adhesives, binders, seala catalysts, intermediates, solvents, dyestuffs, fragrances, deodorizers, flavoring agents, humectants and dehydrating agents, formulation components, monomers, paint and coating additives, photosensitive agents, an d surface ) emissions from coal treatments agents. In addition, urea is used for abating nitrogen oxide (NO fired power plants - x and diesel transportation motors. from urea consumed for non - Emissions of CO ted to be 1.1 MMT CO agricultural purposes in 2015 were estima 2 2 - Table 4 - 22 and Table 4 Eq. (1,128 kt), and 23 . Data for 2015 on urea production data, urea are summarized in exports and imports are not yet published. Data for 2014 has been used as proxy for 2015. Net CO emissions from 2 urea consumption for non approximately 7 0 percent from 1990 . - agricultural purposes in 2015 have decreased by Agricultural Purposes (MMT CO - 4 Table Emissions from Urea Consumption for Non - 22 : CO 2 2 Eq.) 2012 1990 2005 Source 2011 2013 2014 2015 Urea Consumption 3.8 3.7 4.0 4.4 4.0 1.4 1.1 Total 3.8 3.7 4.0 4.4 4.0 1.4 1.1 4 - : CO Table Emissions from Urea Consumption for Non - Agricultural Purposes (kt) 23 2 1990 2005 2011 Source 2012 2013 2014 2015 4,407 3,653 4,030 Urea Consumption 3,784 4,014 1,380 1,128 Total 3,784 3,653 4,030 4,407 4,014 1,380 1,128 Methodology resulting from urea consumption for non - agricultural purposes are estimated by multiplying the Emissions of CO 2 - agricultural purposes by a factor representing the amount of amount of urea consumed in the United States for non CO used as a raw material to produce the urea . This method is based on the assumption that all of the carbon in 2 urea is released into the environment as CO during use, and consistent with the 2006 IPCC Guidelines . 2 The amount of urea consumed for non - agricultural purposes in the United States is esti mated by deducting the quantity of urea fertilizer applied to agricultural lands, which is obtained directly from the chapter ( see Agriculture Table 5 - 24 ) and is reported in Table 4 - 24 , from the total domestic supply of urea. In previous In ventory 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 tons of CO 3 A factor of 0.73 the amount of urea produced plus the sum of net urea imports and exports. per ton of 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 28 4

245 urea consumed is then applied to the resulting supply of urea for non - agricultural purposes to estimate CO 2 emissions from the amount of urea con non - agricultural purposes. The 0.733 tons of CO sumed for per ton of urea 2 emission factor is based on the stoichiometry of producing urea from ammonia and CO . This corresponds to a 2 /urea factor of 44/60, assuming complete conversion of stoichiometric CO (IPCC 2006; EFMA and CO to urea NH 2 3 2 2000). Urea production data for 1990 through 2008 were obtained from the Minerals Yearbook: Nitrogen (USGS 1994 (2011 ). through 2009). Urea production data for 2009 through 2010 were obtained from the U.S. Census Bureau The U.S. Census Bureau ceased collection of urea production statistics in 2011, therefore, u rea production data for 2011 , Minerals Yearbook: Nitrogen (USGS 2014 through 2016). Urea 2012, 2013 and 2014 were obtained from the 5 are not yet publicly available and so 201 4 data have production data for 201 been used as proxy. Urea import data for not yet publicly available and so 2014 data have been used as proxy. Urea import data 2015 are for 2013 and 2014 were obtained from the gen (USGS 2016). Urea import data for 2011 Minerals Yearbook: Nitro and 201 2 were taken fro m U.S. Fertilizer Import/Exports from the United States Department of Agriculture (USDA) Economic USDA suspended updates to this data Research Service Data Sets (U.S. Department of Agriculture 2012). Urea import data for the previous years were obtained from the U.S. Census Bureau after 2012. Current Industrial annual and quarterly reports for 1997 through 2010 (U.S. Census Reports Fertilizer Materials and Related Products 2011), The Fertilizer Institute (TFI 2002) for 1993 through 1996, and the United States Bureau 2001 through International Trade Commission Interactive Tariff and Trade DataWeb (U.S. ITC 2002) for 1990 through 1992 (see 4 - Table ). Urea export data for 2015 are not yet publicly available and so 2014 data have been used as proxy. Urea 24 export data for 2013 and 2014 were obtained from the (USGS 2016). Urea export data Minerals Yearbook: Nitrogen for 1990 through 2012 were taken from from USDA Economic Research Service U.S. Fertilizer Import/Exports Data Sets (U.S. Department of Agriculture 2012). USDA suspended updates to this data after 2012. Table 4 - 24 : Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports ( kt ) Urea Urea Urea Applied Urea Year Production Imports as Fertilizer Exports 1990 7,450 3,296 1,860 854 4,779 5,026 2005 5,270 536 2011 5, 430 5,587 5,860 207 2012 5,220 5,819 6,944 336 2013 6,470 5,480 6,141 335 3,510 5,230 6,520 339 2014 2015 5,230 6,862 3,510 339 inty and Time Series Consistency Uncerta - publicly - available data on the quantities of urea produced and consumed for non - agricultural There is limited - purposes. agricultural purposes is estimated based on a balance that Therefore, the amount of urea used for non s, urea exports, and the amount of urea used as fertilizer. relies on estimates of urea production, urea import The primary uncertainties associated with this source category are associated with the accuracy of these estimates as well as the fact that each estimate is obtained from a different data source. Because urea production estimates are no longer available from the USGS, there is additional uncertainty associated with urea produced beginning in 2011. There is also uncertainty associated with the assumption that all of the carbon in urea is released in to the environment as CO during use. 2 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4 - 25 . Carbon dioxide 1.2 emi - agricultural purposes were estimated to be between 1.0 and ssions associated with urea consumption for non MMT CO Eq. at the 95 percent confidence level. This indicates a range of approximately 12 percent below and 12 2 Eq. MMT CO 1.1 te of percent above the emission estima 2 29 - 4 Industrial Processes and Product Use

246 Table - : Approach 2 Quantitative Uncertainty Estimates for CO 4 Emissions from Urea 25 2 - Consumption for Non Eq. and Percent) Agricultural Purposes (MMT CO 2 a 2015 Emission Estimate Uncertainty Range Relative to Emission Estimate Gas Source (MMT CO (MMT CO Eq.) (%) Eq.) 2 2 Upper Upper Lower Lower Bound Bound Bound Bound Urea Consumption - Agricultural CO +12% 1.1 1.0 1.2 for Non - 12% 2 Purposes 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 consistency in emissions from 1990 through 201 5 . Details on the emission trends through time are described in more detail in the Methodology section, above. consistent with Volume 1, For more information on the general QA/QC process applied to this source category, Chapter 6 of the nd Verification Procedures QA/QC a section in the introduction of the , see 2006 IPCC Guidelines IPPU Chapter. Discussion Recalculations total urea production and estimates for urea exports and imports for 201 4 were updated Production estimates for Minera ls Yearbook: Nitrogen 6 (USGS 201 using information obtained from the ). This update, as well as the urea consumption update included below, resulted in a significant decrease in urea imports for 2014, resulting in a decrease of the 2014 emission estimate relative to the previous Inventory rep ort of approximately 66 percent. In addition, this update also resulted in an update to the urea export value for 2013. T he amount of urea consumed for agricultural purposes (used for calculating urea consumption for non - agricultural updated urea application purposes) in the Unit ed States for the years 2011 through 2014 was revised based on Th e estimates obtained from the Agriculture chapter (see Table 5 - 24 ). decrease in the s e update s resulted in a emission estimate relative to the previous Inventory report of approximately 5 percent in 201 3 and 66 percent in 2014, as previously described . As stated previously in the Methodology section, in previous Inventory reports th e 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, and has been moved to the Agriculture chapter for this report. 4.7 Nitric Aci d Production (IPCC Source Category 2B2) Nitrous oxide (N O) is emitted during the production of nitric acid (HNO ), an inorganic compound used primarily 2 3 to make synthetic commercial fertilizers. It is also a major component in the production of adipic acid — a feedstock for nylon — and explosives. Virtually all of the nitric acid produced in the United States is manufactured by the high - temperature catalytic oxidation of ammonia (EPA 1998). There are two different nitric acid production methods: strength nitric acid. The first method utilizes oxidation, condensation, and absorption to - weak nitric aci d and high strength acid (90 percent or greater - produce nitric acid at concentrations between 30 and 70 percent nitric acid. High nitric acid) can be produced from d ehydrating, bleaching, condensing, and absorption of the weak nitric acid. The basic process technology for producing nitric acid has not changed significantly over time. Most U.S. plants were built between 1960 and 2000. As of 2015, there were 34 active w eak nitric acid production plants, including one strength nitric acid production plant in the United States (EPA 2010; EPA 2016). - high - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 30 4

247 During this reaction, N O is formed as a byproduct and is released from reactor vents into the atmosphere. 2 fuels consumed for energy purposes during the production of nitric acid are accounted for in the Emissions from Energy chapter. ) with oxygen (O Nitric acid is made from the reaction of ammonia (NH ) in two stages. The overall reaction is: 3 2 +4퐻 4푁퐻 → 4 퐻푁푂 +8푂 푂 2 3 2 3 ). As such, the industry in the United (i.e., NO Currently, the nitric acid industry controls emissions of NO and NO x 2 - selective catalytic reduction (NSCR) and selective catalytic reduction (SCR) States uses a combination of non cess of destroying NO O. However, , NSCR systems are also very effective at destroying N technologies. In the pro 2 x NSCR units are generally not preferred in modern plants because of high energy costs and associated high gas built between 1971 and 1977 with NSCRs installed at temperatures. NSCR systems were installed in nitric plants - third of the weak acid production plants. U.S. facilities are using both tertiary (i.e., NSCR) and approximately one secondary controls (i.e., alternate catalysts). Nitrous oxide emissions from this sour ce were estimated to be 11.6 MMT CO O) in 2015 (see Eq. (39 kt of N 2 2 Table 26 ). Emissions from nitric acid production have decreased by 5 percent sin ce 1990, with the trend in the - 4 time series closely tracking the changes in production. Emissions have decreased by 20 percent since 1997, the highest year of production in the time series. 4 - 26 : N Table O Emissions from Nitric Acid Production (MMT CO Eq. and kt N O) 2 2 2 MMT CO O Eq. kt N Year 2 2 1990 12.1 41 38 11.3 2005 10.9 37 2011 2012 10.5 35 10.7 36 2013 10.9 37 2014 11.6 39 2015 Methodology Emissions of N and country O were calculated using the estimation methods provided by the 2006 IPCC Guidelines 2 specific methods from EPA’s GHGRP. The Tier 2 method was used to estimate emissions 2006 IPCC Guidelines a country - specific approach similar to the IPCC Tier 3 from nitric acid production for 1990 through 2009, and method was used to estimate N O emissions for 2010 through 2015. 2 2010 through 201 5 Process N O emissions and nitric acid production data were obtained directly from EPA’s GHGRP for 2010 through 2 - 2015 by aggregating reported facility level data (EPA 2016). 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 2015, there were 34 facilities that reported to EPA, including the known - single high strength nitric acid production facility in the United States (EPA 2016). All nitric acid (weak acid) facilities are required to calculate process emissi ons using a site - specific emission factor developed through annual performance testing under typical operating conditions or by directly measuring N O emissions using monitoring 2 206 equipment. The high - strength nitric acid facility also reports N O emissions associated with weak acid production 2 206 - Facilities must use standard methods, either EPA Method 320 or ASTM D6348 03 and must follow associated QA/QC specific QC of direct emission measurements. - edures consistent during these performance test consistent with category proc 31 - 4 Industrial Processes and Product Use

248 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: 207 EPA verifies annual facility - level GHGRP reports through a Nitr ic Acid Production of the regulation, Part 98. - step process (e.g., combination of electronic checks and manual reviews) to identify potential errors and multi accurate, complete, and consistent . Based on the results of the verification ensure that data submitted to EPA are 208 . process, the EPA follows up with facilities to resolve mistakes that may have occurred the e utilized to develop To calculate emissions from 2010 through 2015, GHGRP nitric acid production data ar 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 sociated emissions of nitric acid with and without abatement technologies are calculated. These percentages and as are the basis for developing the country specific weighted emission factors which vary from year to year based on the amount of nitric acid producti on with and without abatement technologies. 1990 through 2009 209 Using GHGRP data for 2010, country - specific N O emission factors were calculated for nitric acid production 2 with abatement and without abatement (i.e., controlled and uncontrolled emiss ion factors), as previous stated. The following 2010 emission factors were derived for production with abatement and without abatement: 3.3 kg N O/metric ton HNO produced at plants using abatement technologies (e.g., tertiary systems such as NSCR 3 2 systems) - O/metric ton HNO produced at plants not equipped with abatement technology. Country and 5.99 kg N 3 2 specific weighted emission factors were derived by weighting these emission factors by percent production with abatement and without abatement over time per iods 1990 through 2008 and 2009. These weighted emission factors were used to estimate N O emissions from nitric acid production for years prior to the availability of GHGRP data 2 (i.e., 1990 through 2008 and 2009). A separate weighted factor is included fo r 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. Based on the available data, it was assumed that emission factors for 2010 would be more representative of oper ating 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 level abatem application of facility - ent technologies, maintenance of abatement technologies, and also due to plant O abatement closures and start ups (EPA 2012, 2013; Desai 2012; CAR 2013). The installation dates of N - 2 lities reporting abatement technology use technologies are not known at most facilities, but it is assumed that faci have had this technology installed and operational for the duration of the time series considered in this report (especially NSCRs). specific weighted N O emission factors were used in conjunction with - annual production to estimate The country 2 N O emissions for 1990 through 2009, using the following equations: 2 퐸 ×퐸퐹 =푃 푖 푖 푤푒푖푔ℎ푡푒푑,푖 퐸퐹 ⌋ × ) 퐸퐹 푃 % ( + ) = ⌊ ( % 푃 퐸퐹 × 푢푛푐 푖 푐 푖 , 푡푒푑 ℎ 푤푒푖푔 푢푛푐 , 푖 퐶 , where, E = Annual N O Emissions for year i (kg/yr) i 2 nnual nitric acid production for year i (metric tons HNO ) = A P i 3 ) O/metric ton HNO O emission factor for year i (kg N EF = Weighted N 2 2 3 weighted,i 207 bin/text idx?tpl=/ecfrbrowse/Title40/40cfr98_main_02.tpl>. - See . 07/documents/ghgrp_verification_factsheet.pdf See < https://www.epa.gov/sites/production/files/2015 - 209 O process emissions, national production, and national share of nitric acid production with abatement and without National N 2 - ed from the GHGRP facility abatement technology was aggregat level data for 2010 to 2015 (i.e., percent production with and without abatement). - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 32 4

249 %P = Percent national production of HNO with N O abatement technology (%) c,i 3 2 ement technology (kg N = N O emission factor, with abat EF ) O/metric ton HNO 3 2 2 c %P = Percent national production of HNO O abatement technology (%) without N 2 unc,i 3 O emission factor, without abatement technology (kg N EF O/metric ton HNO ) = N 2 3 unc 2 i = year from 1990 through 2009 • For 2009: Weighted N O emission factor – 5.46 kg N . O/metric ton HNO 3 2 2 5.66 kg N O emission factor – • For 1990 through 2008: Weighted N O/metric ton HNO . 3 2 2 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 o f 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 7,200 1990 2005 6,710 7,600 2011 7,460 2012 2013 7,580 7,660 2014 2015 7,210 Uncertainty and Time - Series Consistency estimate N O emissions includes the share of U.S. nitric acid Uncertainty associated with the parameters used to 2 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. Whil e some information has been obtained through outreach with industry associations, limited information is available over the time series (especially prior to 2010) for a variety of facility level variables, including plant specific production levels, plant production technology (e.g., low, high pressure, etc.), and abatement technology type, installation date of abatement technology, and accurate destruction and removal efficiency rates. Production data prior to 2010 were obtained from National Census Bureau , which does not provide uncertainty estimates with their data. Facilities reporting to EPA’s GHGRP must measure production using equipment and practices used for accounting purposes. At this time EPA does not estimate uncertainty of the aggregated facilit y - level information. As noted in the Methodology section, 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 submitt ed to EPA are accurate, complete, and consistent. The annual production reported by each nitric acid facility under EPA’s GHGRP and then aggregated to estimate national N O emissions is assumed to have low uncertainty. 2 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4 - 28 . Nitrous oxide Eq. at e between 10.9 and 12.2 MMT CO emissions from nitric acid production were estimated were estimated to b 2 the 95 percent confidence level. This indicates a range of approximately 5 percent below to 6 percent above the 2015 emissions estimate of 11.6 MMT CO Eq. 2 33 - 4 Industrial Processes and Product Use

250 4 28 pproach 2 Quantitative Uncertainty Estimates for N - O Emissions from Nitric Table : A 2 Eq. and Percent) Acid Production (MMT CO 2 a 2015 Emission Estimate Uncertainty Range Relative to Emission Estimate Gas Source (MMT CO (MMT CO Eq.) Eq.) (%) 2 2 Lower Lower Upper Upper Bound Bound Bound Bound 6 Nitric Acid Production N O 11.6 10. 9 12.2 - 5% + % 2 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 consistency in emissions from 1990 through 2015. To maintain consistency across the time series and with the rounding approaches taken by other data sets, a new rounding approach was performed for the GHGRP Subpart V: Nitric Ac id data. This resulted in production data changes across the time series of 2010 to 2015, in which GHGRP data have been utilized. The results of this update have had an insignificant impact on the emissions estimates across the 2010 to 2015 time series. De tails 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, section in the introduction of the e QA/QC and Verification Procedures Chapter 6 of the , se 2006 IPCC Guidelines IPPU Chapter. Planned Improvements - term and long - term improvement to estimates and associated Pending resources, EPA is considering both near ort characterization of uncertainty. In the sh term, with 6 years of GHGRP data, EPA anticipates completing updates - of category - specific QC procedures to potentially also improve both qualitative and quantitative uncertainty estimates. Longer term, in 2020, EPA anticipates having information from E PA’s GHGRP facilities on the installation date of any N O abatement equipment, per recent revisions finalized in December 2016 to EPA’s 2 O emissions from nitric acid production over GHGRP. This information will enable more accurate estimation of N 2 the time series. 4.8 Adipic Acid Production (IPCC Source Category 2B3) Adipic acid is produced through a two - stage process during which nitrous oxide (N O) is generated in the second 2 stage. Emissions from fuels consumed for energy purposes during the production of ad ipic acid are accounted for in the Energy chapter. The first stage of manufacturing usually involves the oxidation of cyclohexane to form a cyclohexanone/ cyclohexanol mixture. The second stage involves oxidizing this mixture with nitric acid to produce ipic acid. Nitrous oxide is generated as a byproduct of the nitric acid oxidation stage and is emitted in the waste ad gas stream (Thiemens and Trogler 1991). The second stage is represented by the following chemical reaction: ( ) ) ( 퐶푂 푐푦푐푙표ℎ푒푥푎푛표푙 푐푦푐푙표ℎ푒푥푎푛표푛푒 (퐶퐻 + (퐶퐻 ) ) 퐶퐻푂퐻 + 푤퐻푁푂 5 5 3 2 2 ) ( + 푥푁 ) 퐶푂푂퐻 푎푑푖푝푖푐 푎푐푖푑 →퐻푂푂퐶(퐶퐻 푂 푂 +푦퐻 2 2 4 2 Process emissions from the production of adipic acid vary with the types of technologies and level of emission producing plants had N - a facility. In 1990, two major adipic acid O abatement technologies in controls employed by 2 place and, as of 1998, three major adipic acid production facilities had control systems in place (Reimer et al. 1999). ril 2006 and represented approximately two percent of production, did not One small plant, which last operated in Ap selective catalytic - O (VA DEQ 2009; ICIS 2007; VA DEQ 2006). In 2014, catalytic reduction, non control for N 2 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 34 4

251 reduction (NSCR) and thermal reduction abatement technologies were applie d as N O abatement measures at adipic 2 acid facilities (EPA 2016). Worldwide, only a few adipic acid plants exist. The United States, Europe, and China are the major producers. In production facilities (two in Texas and 2015, the United States had two companies with a total of three adipic acid one in Florida), all of which were operational (EPA 2016). The United States accounts for the largest share of global adipic acid production capacity (30 percent), followed by the European Union (29 percent) and Chin a (22 percent) (SEI 2010). Adipic acid is a white crystalline solid used in the manufacture of synthetic fibers, plastics, coatings, urethane foams, elastomers, and synthetic lubricants. Commercially, it is the most important of the aliphatic - four percent of all adipic acid produced in the acids, which are used to manufacture polyesters. Eighty dicarboxylic United States is used in the production of nylon 6,6; 9 percent is used in the production of polyester polyols; 4 lasticizers; and the remaining 4 percent is accounted for by other uses, percent is used in the production of p including unsaturated polyester resins and food applications (ICIS 2007). Food grade adipic acid is used to provide some foods with a “tangy” flavor (Thiemens and Trogler 1991). ous oxide emissions from adipic acid production were estimated to be 4.3 MMT CO Eq. (14 kt N O) in 2015 Nitr 2 2 (see 4 - 29 ). National adipic acid production has increased by approximately 40 percent over the period of 1990 Table through 2015, to approximately 1,055,000 metric tons (ACC 2016). Over the period 1990 throu gh 2015, emissions have been reduced by 72 percent due to both the widespread installation of pollution control measures in the late 1990s and plant idling in the late 2000s. In April 2006, the smallest of the four facilities ceased production of adipic id (VA DEQ 2009); furthermore, one of the major adipic acid production facilities was not operational in 2009 or ac 2010 (Desai 2010). All three remaining facilities were in operation in 2015. Very little information on annual trends in the activity data exis t for adipic acid. Table 4 - 29 : N O Emissions from Adipic Acid Production (MMT CO ) Eq. and kt N O 2 2 2 Year MMT CO O Eq. kt N 2 2 51 1990 15.2 2005 7.1 24 2011 10.2 34 5.5 19 2012 2013 3.9 13 2014 5.4 18 2015 4.3 14 Methodology Emissions are estimated using both Tier 2 and Tier 3 methods consistent with the 2006 IPCC Guidelines. Due to confidential business information, plant names are not provided in this section. Therefore, the four adipic acid - producing facilities will be referred to as Plants 1 through 4. Plant 4 was closed in April 2006. Overall, as noted above, the three plants that are currently operating facilities use abatement equipment. Plants 1 and 2 employ struction and Plant 3 employs thermal destruction. catalytic de 2010 through 201 5 All emission estimates for 2010 through 2015 were obtained through analysis of GHGRP data (EPA 2014 through level greenhouse gas emissions 2006 IPCC Guidelines T ier 3 method. Facilit y - 2016), which is consistent with the data were obtained from EPA’s GHGRP for the years 2010 through 2015 (EPA 2014 through 2016) and aggregated to national N O emissions. Consistent with IPCC Tier 3 methods, all adipic acid production facilities are required t o 2 calculate emissions using a facility specific emission factor developed through annual performance testing under - 35 - 4 Industrial Processes and Product Use

252 210 typical operating conditions or by directly measuring N More O emissions using monitoring equipment. 2 O emissions applicable to adipic acid information on the calculation, monitor ing and QA/QC methods for process N 2 211 EPA verifies production facilities under Subpart E can be found in the electronic code of federal regulations. through a multi - annual facility , combination of electronic checks and - level GHGRP reports step process (e.g. manual reviews) to identify potential errors and ensure that data submitted to EPA are accurate, complete, and 212 consistent. 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 engineer 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 n ot published (Desai 2010). These estimates were based on continuous process monitoring equipment installed at the two facilities. In 2009 and 2010, no adipic acid production occurred at Plant 1 per reporting to EPA’s GHGRP (EPA 2012; Desai 2011b). For the Plant 4, 1990 through 2009 N O emissions were estimated using the following Tier 2 equation from the 2 until shutdown of the plant in 2006 : 2006 IPCC Guidelines ( [ ]) 퐸 ×퐸퐹 × 1− 퐷퐹×푈퐹 = 푄 푎푎 푎푎 푎푎 where, E = N O emissions from adipic acid production, metric tons aa 2 = Quantity of adipic acid produced, metric tons Q aa EF = Emission factor, metric ton N O/metric ton adipic acid produced 2 aa DF = N O destruction factor 2 UF = Abatement system utility factor factor (i.e., N O emitted per unit of adipic acid produced), The adipic acid production is multiplied by an emission 2 O production in the which has been estimated, based on experiments that the reaction stoichiometry for N 2 O per metric ton of preparation of adipic acid, to be approximately 0.3 metric tons of N product (IPCC 2006). The 2 “N O destruction factor” in the equation represents the percentage of N O emissions that are destroyed by the 2 2 installed abatement technology. The “abatement system utility factor” represents the percentage of time that the abatemen t equipment operates during the annual production period. No abatement equipment was installed at the - Inolex/Allied Signal facility, which last operated in April 2006 (VA DEQ 2009). Plant specific production data for this facility were obtained across the time series from 1990 through 2006 from the Virginia Department of - Environmental Quality (VA DEQ 2010). 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 2011a). For 1990 through 2004, emissions were estimated using plant - specific production data and the IPCC factors as described above for Plant 4. Plant - ating national adipic acid level adipic acid production for 1990 through 2003 was estimated by alloc 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 tained and used for emission calculations (CW 2005). ob Chemical & Engineering News , “Facts and Figures” and Plant capacities for 1990 through 1994 were obtained from 96 were kept the “Production of Top 50 Chemicals” (C&EN 1992 through 1995). Plant capacities for 1995 and 19 same as 1994 data. The 1997 plant capacities were taken from Chemical Market Reporter , “Chemical Profile: Adipic Acid” (CMR 1998). The 1998 plant capacities for all four plants and 1999 plant capacities for three of the plants were obtaine d from Chemical Week , Product Focus: Adipic Acid/Adiponitrile (CW 1999). Plant capacities for 2000 for three of the plants were updated using , “Chemical Profile: Adipic Acid” (CMR Chemical Market Reporter 210 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. 211 See . 212 >. fication_factsheet.pdf 07/documents/ghgrp_veri - https://www.epa.gov/sites/production/files/2015 See < - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 36 4

253 held constant at 2001). For 2001 through 2003, the plant capacities year 2000 capacities. Plant for three plants were capacity for 1999 to 2003 for the one remaining plant was kept the same as 1998. National adipic acid production data (see - 30 ) from 1990 through 2015 were obtained from the American 4 Table Chemistry Council (ACC 2016). 30 : Adipic Acid Production (kt) Table 4 - kt Year 755 1990 865 2005 840 2011 950 2012 980 2013 1,025 2014 1,055 2015 Uncertainty and Time - Series Consistency Uncertainty associated with N O emission estimates includes the methods used by companies to monitor and 2 estimate emissions. While some information has been obtained through outreach with facilities, limited information is available over the time series on these methods, abatement tech nology destruction and removal efficiency rates and plant specific production levels. 4 31 - Table . Nitrous oxide The results of this Approach 2 quantitative uncertainty analysis are summarized in emissions from adipic acid production for 2015 were estimated to be between 4.1 and 4.5 MMT CO Eq. at the 95 2 mately 4 percent below to 4 percent above the percent confidence level. These values indicate a range of approxi Eq. 2015 emission estimate of 4.3 MMT CO 2 4 Table 31 : Approach 2 Quantitative Uncertainty Estimates for N O Emissions from Adipic - 2 Acid Production (MMT CO Eq. and Percent) 2 a 2015 Emission Estimate Uncertainty Range Relative to Emission Estimate Gas Source (%) (MMT CO Eq.) (MMT CO Eq.) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound Adipic Acid Production N O 4.3 4.1 4.5 - 4% +4% 2 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 consistency in emissions from 1990 through 2015. 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, , see QA/QC and V section in the introduction of the erification Procedures Chapter 6 of the 2006 IPCC Guidelines IPPU Chapter. 37 - 4 Industrial Processes and Product Use

254 Silicon Carbide Production 4.9 and Consumption (IPCC Source Category 2B ) 5 ) and methane (CH Carbon dioxide (CO ) are emitted from the production of silicon carbide (SiC), a material used 4 2 as an industrial abrasive. Silicon carbide is produced for abrasive, metallurgical, and other non - abrasive applications abrasive applications is not available and therefore in the United States. Production for metallurgical and other non - both CO and CH e stimates are based solely upon production estimates of silicon carbide for abrasive applications. 2 4 Emissions from fuels consumed for energy purposes during the production of silicon carbide are accounted for in the Energy chapter. To produce SiC, silica s and or quartz (SiO ) is reacted with C in the form of petroleum coke. A portion (about 35 2 , percent) of the carbon contained in the petroleum coke is retained in the SiC. The remaining C is emitted as CO 2 CH , or carbon monoxide (CO). The overall reaction is shown below (but in practice it does not proceed according 4 : to stoichiometry) ( ) + 3 퐶 → 푆푖퐶 + 2 퐶푂 푆푖푂 + 푂 → 2 퐶푂 2 2 2 Carbon dioxide is also emitted from the consumption of SiC for metallurgical and other non - abrasive applications. Markets for manu factured abrasives, including SiC, are heavily influenced by activity in the U.S. manufacturing sector, especially in the aerospace, automotive, furniture, housing, and steel manufacturing sectors. The U.S. approximately 50 percent) of SiC is used in metallurgical and Geological Survey (USGS) reports that a portion ( abrasive applications, primarily in iron and steel production (USGS 2006a). As a result of the economic other non - orts, particularly from China, downturn in 2008 and 2009, demand for SiC decreased in those years. Low cost imp combined with high relative operating costs for domestic producers, continue to put downward pressure on the production of SiC in the United States. However, demand for SiC consumption in the United States has recovered - grade silicon carbide was manufactured at a two facilities somew hat from its low in 2009 (USGS 2012a). Abrasive in 2015 (USGS 2016). Eq. (180 kt CO Carbon dioxide emissions from SiC production and consumption in 2015 were 0.2 MMT CO ) (see 2 2 4 - 32 and Table Table - 33 ). Approximately 51 percent of these emi ssions resulted from SiC production while the 4 remainder resulted from SiC consumption. Methane emissions from SiC production in 2015 were 0.01 MMT CO 2 and Eq. (0.4 kt CH Table 4 - 32 ) (see Table 4 - 33 ). Emissions have fluctuated in recent years, but 2015 emissions are 4 about 52 percent lower than emissio ns in 1990. Table - 32 : CO and CH 4 Emissions from Silicon Carbide Production and Consumption (MMT 2 4 CO Eq.) 2 2015 2014 Year 2013 1990 2005 2011 2012 0.4 0.2 0.2 0.2 0.2 0.2 0.2 CO 2 + + + CH + + + + 4 0.2 0.2 0.2 Total 0.2 0.2 0.2 0.4 + Does not exceed 0.05 MMT CO Eq. 2 4 33 : CO - and CH Table Emissions from Silicon Carbide Production and Consumption (kt) 4 2 Year 1990 2005 2011 2012 2013 2014 2015 170 375 219 CO 158 169 173 180 2 + + + + CH 1 + + 4 + Does not exceed 0.5 kt. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 38 4

255 Methodology 213 Emissions of CO from the production of SiC were calculated and CH using the Tier 1 method provided by the 2 4 2006 IPCC Guidelines . Annual estimates of SiC production were multiplied by the appropriate emission factor, as shown below : = 퐸퐹 ×푄 퐸 푠푐 푠푐,퐶푂2 푠푐,퐶푂2 1 푚푒푡푟푖푐 푡표푛 퐸퐹 = ) 퐸 × 푄 ( × 푠푐 , 퐶퐻 , 푠푐 4 푠푐 4 퐶퐻 푘푔 1000 where, E = CO emissions from p roduction of SiC, metric tons sc,CO2 2 EF = Emission factor for production of SiC, metric ton CO /metric ton SiC sc,CO2 2 = Quantity of SiC produced, metric tons Q sc E = CH emissions from production of SiC, metric tons 4 sc,CH4 EF /metric ton SiC = Emission factor for production of SiC, kilogram CH 4 sc,CH4 Emission factors were taken from the 2006 IPCC Guidelines : • 2.62 metric tons CO /metric ton SiC 2 /metric ton SiC • 11.6 kg CH 4 Emissions of CO from silicon carbide consumption for metallurgical uses were calculated by multiplying the 2 ) by the Minerals Yearbook: Silicon annual utilization of SiC for metallurgical uses (reported annually in the USGS carbon content of SiC (31.5 percent), which was d etermined according to the molecular weight ratio of SiC. Emissions of CO from silicon carbide consumption for other non - abrasive uses were calculated by multiplying the 2 - abrasive uses by the carbon content of SiC (31.5 perce nt). The annual SiC annual SiC consumption for non consumption for non - abrasive uses was calculated by multiplying the annual SiC consumption (production plus net imports) by the percent used in metallurgical and other non - abrasive uses (50 percent) (USGS 2006a) and then subtracting the SiC consumption for metallurgical use. Production data for 1990 through 2013 were obtained from the Minerals Yearbook: Manufactured Abrasives (USGS 1991a through 2015). Production data for 2014 and 2015 were obtained from the Minerals Industry Surveys: Abrasives (Manufactured) (USGS 2016). Silicon carbide production data obtained through the USGS National Minerals Information Center has been previously been r ounded to the nearest 5,000 metric tons to avoid disclosing bide consumption by major end use for 1990 through 2012 were obtained from company proprietary data . Silicon car ). In the previous r 34 the Minerals Yearbook: Silicon (USGS 1991b through 2013) (see Table 4 - eport, silicon carbide consumption data for 2013 and 2014 were not yet publicly available so 2012 data were used as proxy. In this year’s report, 2013 and 2014 activity data were available and the time series was recalculated to remove proxy data. Data 2015 silicon carbide consumption was not yet published by the USGS, resulting in the use of 2014 data as a for U.S. International Trade proxy. Net imports and exports for the entire time series were obtained from the (USITC) database updated from Commission ata provided by the U.S. Census Bureau (2005 through 2016). d 4 - 34 : Production and Consumption of Silicon Carbide (Metric Tons) Table Production Consumption Year 105,000 172,465 1990 2005 35,000 220,149 213 level GHGRP information to inform these estimates. The aggregated information EPA has not integrated aggregated facility - (e.g., activity data and emissions) associated with silicon carbide did not meet criteria to shield underly ing confidential business information (CBI) from public disclosure. 39 - 4 Industrial Processes and Product Use

256 136,222 2011 35,000 114,265 2012 35,000 35,000 134,055 2013 2014 35,000 140,733 35,000 153,475 2015 Uncertainty and Time - Series Consistency There is uncertainty associated with the emission factors used because they are based on stoichiometry as opposed to monitoring of actual SiC production plants. An alternative would be to calculate emissions based on the quantity of petroleum coke used during the production process rather than on the amount of silicon carbide produced. However, containing volatile , there is also uncertainty associated with the hydrogen - these data were not available. For CH 4 compounds in the petroleum coke (IPCC 2006). There is also uncertainty associated with the use or destruction of methane generated from the process in addition to uncertainty associated with levels of production, net imports, consumption levels, and the percent of total consumption that is attributed to metallurgical and other non - abrasive uses. The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4 - 35 . Silicon carbide production and consumption CO emissions from 2015 were estimated to b e between 9 percent below and 9 percent 2 above the emission estimate of 0.18 MMT CO Eq. at the 95 percent confidence level. Silicon carbide production 2 CH emissions were estimated to be between 9 percent below and 10 percent above the emission estimate of 0.01 4 MMT CO Eq. at the 95 percent confidence level. 2 : and CO Quantitative Uncertainty Estimates for CH Emissions from Table 4 - 35 Approach 2 4 2 Silicon Carbide Production and Consumption (MMT CO Eq. and Percent) 2 a Uncertainty Range Relative to Emission Estimate 2015 Emission Estimate Source Gas (MMT CO (%) Eq.) (MMT CO Eq.) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound Silicon Carbide Production 0.18 0.16 0.20 CO - 9% +9% 2 and Consumption Silicon Carbide Production CH + + + - 9% +10% 4 + Does not exceed 0.05 MMT CO Eq. 2 a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. Methodological entire time series to ensure consistency in emissions from 1990 approaches were applied to the through 2015. 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 sou consistent with Volume 1, rce category, Chapter 6 of the QA/QC and Verification Procedures , see 2006 IPCC Guidelines section in the introduction of the IPPU Chapter. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 40 4

257 4.10 Titanium Dioxide Production (IPCC Source ) 6 Category 2B ctured using one of two processes: the chloride process and the sulfate process. ) is manufa Titanium dioxide (TiO 2 - related carbon dioxide The chloride process uses petroleum coke and chlorine as raw materials and emits process ). Emissions from fuels consumed for energy purposes during (CO the production of titanium dioxide are accounted 2 for in the Energy chapter. The chloride process is based on the following chemical reactions: 2퐹푒푇푖푂 +7퐶푙 +3퐶 →2푇푖퐶푙 +2퐹푒퐶푙 +3퐶푂 3 2 3 4 2 푇푖퐶 2 + 2 푂 → 2 푇푖 푂 푙 푙 + 4 퐶 2 2 2 4 The sulfate proces s does not use petroleum coke or other forms of carbon as a raw material and does not emit CO . 2 The C in the first chemical reaction is provided by petroleum coke, which is oxidized in the presence of the chlorine and FeTiO produced in the United States has been produced using the (rutile ore) to form CO . Sinc e 2004, all TiO 2 2 3 chloride process, and a special grade of “calcined” petroleum coke is manufactured specifically for this purpose. The principal use of TiO is as a pigment in white paint, lacquers, and varnishes; it is also used as a pigment in the 2 manufacture of plastics, paper, and other products. In 2015, U.S. TiO production totaled 1,220,000 metric tons 2 ing TiO (USGS 2017). There were a total five plants produc in the United States in 2015. 2 Emissions of CO from titanium dioxide production in 2015 were estimated to be 1.6 MMT CO Eq. (1,635 kt CO ), 2 2 2 which represents an increase of 37 percent since 1990 (see 4 - 36 ). Compared to 2014, emissions from titanium Table dioxide production decreased by 3 percent in 2015 due to a 3 percent decrease in production. Table 4 - 36 : CO Emissions from Titanium Dioxide (MMT CO Eq. and kt) 2 2 Year MMT CO kt Eq. 2 1990 1,195 1.2 2005 1.8 1,755 2011 1.7 1,729 2012 1.5 1,528 1.7 1,715 2013 2014 1.7 1,688 2015 1.6 1,635 Methodology Emissions of CO production were calculated by multiplying annual national TiO from TiO production by chloride 2 2 2 specific emission factors using a Tier 1 approach provided in . The Tier 1 equation is 2006 IPCC Guidelines process - follows: as 퐸 = 퐸퐹 ×푄 푡푑 푡푑 푡푑 where, E = CO production, metric tons emissions from TiO 2 td 2 EF /metric ton TiO Emission factor (chloride process), metric ton CO = 2 td 2 produced Q Quantity of TiO = td 2 produced each year. For years prior to 2004, it was assumed that Data were obtained for the total amount of TiO 2 was produced using the chloride process and the sulfate process in the same ratio as the ratio of the total U.S. TiO 2 production capacity for each process. As of 2004, the last remaining sul fate process plant in the United States 41 - 4 Industrial Processes and Product Use

258 - closed; therefore, 100 percent of post 2004 production uses the chloride process (USGS 2005b). The percentage of metric tons production from the chloride process is estimated at 100 percent since 2004. An emission factor of 1.34 /metric ton TiO was applied to the estimated chloride - process production (IPCC 2006). It was assumed that all CO 2 2 TiO produced using the chloride process was produced using petroleum coke, although some TiO may have been 2 2 produced with graphi te or other carbon inputs. The emission factor for the TiO chloride process was taken from the 2006 IPCC Guidelines . Titanium dioxide 2 production data and the percentage of total TiO production capacity that is chloride process for 1990 through 2013 2 ) were obtained through the Table 4 (see 37 Minerals Yearbook: Titanium Annual Report (USGS 1991 through - 2014). Production data for 2014 and 2015 was obtained from the Minerals Commodity Summary: Titanium and 214 (USGS 2017). Titanium Dioxide Data on the percentage of total TiO production capacity that is chloride process 2 were not available for 1990 through 1993, so data from the 1994 USGS were used for t hese Minerals Yearbook years. Because a sulfate process plant closed in September 2001, the chloride process percentage for 2001 was estimated based on a discussion with Joseph Gambogi (2002). By 2002, only one sulfate process plant remained this plant closed in 2004 (USGS 2005b). online in the United States and 4 Table 37 : Titanium Dioxide Production (kt) - kt Year 1990 979 2005 1,310 2011 1,290 2012 1,140 2013 1,280 1,260 2014 2015 1,220 Uncertainty and Time - Series Consistency Each year, the U.S. Geological Survey (USGS) collects t itanium industry data for titanium mineral and pigment production operations. If TiO pigment plants do not respond, production from the operations is estimated on the 2 basis of prior year production levels and industry trends. Variability in response rates varies from 67 to 100 percent plants over the time series. pigment of TiO 2 some TiO may be produced using graphite or other carbon inputs, information and data regarding these Although 2 practices were not available. Titanium dioxide produced using graphite inputs, for example, may generate differing amounts of CO d as compared to that generated through the use of petroleum coke in per unit of TiO produce 2 2 production. While the most accurate method to estimate emissions would be to base calculations on the amount of reducing agent used in each process rather than on the amount of TiO produ ced, sufficient data were not available 2 to do so. production was process plant in the United States closed. Since annual TiO - As of 2004, the last remaining sulfate 2 not reported by USGS by the type of production process used (chloride or sulfate) prior to 2004 and only the percentage of total production capacity by process was reported, the percent of total TiO production capacity that 2 production to estimate the amount of TiO was attributed to the chloride process was multiplied by total TiO 2 2 produced usi ng the chloride process. Finally, the emission factor was applied uniformly to all chloride - process production, and no data were available to account for differences in production efficiency among chloride - process process TiO production, literature data - plants. In calculating the amount of petro leum coke consumed in chloride 2 214 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 d ata, emission factor) from these facilities did not meet criteria to shield underlying CBI from public disclosure. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 42 4

259 were used for petroleum coke composition. Certain grades of petroleum coke are manufactured specifically for use in the TiO chloride process; however, this composition information was not available. 2 Table - 38 . Titanium dioxide The results of the Approach 2 quantitative uncertainty analysis are summarized in 4 consumption CO emissio ns from 2015 were estimated to be between 1.4 and 1.8 MMT CO Eq. at the 95 percent 2 2 confidence level. This indicates a range of approximately 12 percent below and 13 percent above the emission estimate of 1.6 MMT CO Eq. 2 4 - 38 : Approach 2 Quantitative Uncertainty Estimates for CO Table Emissions from Titanium 2 Dioxide Production (MMT CO Eq. and Percent) 2 a Uncertainty Range Relative to Emission Estimate 2015 Emission Estimate Source Gas Eq.) (MMT CO (MMT CO Eq.) (%) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound Titanium Dioxide Production CO 1.6 1.4 1.8 - 12% +13% 2 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 consistency in emissions from 1990 through 2015. 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, , see Chapter 6 of the 2006 IPCC Guidelines QA/QC and Verification Procedures section in the introduction of the IPPU Chapter. Discussion Recalculations updated relative to the previous 4 w P 201 Inventory (i.e., 1990 through 2014) based on roduction data for as 2016 M inerals Commodity Summaries : Titanium and Titanium Dioxide (USGS 201 6) published data in the USGS . This resulted in a percent decrease in 201 4 CO 4 production relative to the previous report. emissions from TiO 2 2 Planned Improvements P lanned improvements include researching the significance of titanium - slag production in electric furnaces and synthetic - rutile production using the Becher process in t he United States. Significant use of these production processes will be included in future Inventory reports This planned improvement is still in development by the EPA . EPA ng aggregated facility - level and is not included in this report. continues to assess the potential of integrati t itanium GHGRP information for ioxide production facilities based on criteria to shield underlying CBI from d public disclosure . Pending available resources, EPA will also evaluate use of GHGRP data to improve category - specific QC consistent with both Volume 1, Chapter 6 of 2006 IPCC Guidelines and the latest IPCC guidance on the 215 use of facility level data in national inventories. - 4.11 Soda Ash Production and Consumption (IPCC Source Category 2B7 ) Carbon dioxide (CO ) is generated as a byproduct of calcining trona ore to produce soda ash, and is eventually 2 emitted into the atmosphere. In addition, CO may also be released when soda ash is consumed. Emissions from 2 215 See < >. nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf - http://www.ipcc 43 - 4 Industrial Processes and Product Use

260 on of soda ash are accounted for in the fuels consumed for energy purposes during the production and consumpti Energy sector. Calcining involves placing crushed trona ore into a kiln to convert sodium bicarbonate into crude sodium carbonate during trona - based production is based on the that will later be filtered into pure soda ash. The emission of CO 2 following reaction: ( ) ( ) + 5퐻 푂 +퐶푂 푆표푑푎 퐴푠ℎ →3 푁푎 • 푂 푇푟표푛푎 푁푎퐻퐶푂 2퐻 퐶푂 퐶푂 2푁푎 • 2 2 3 2 3 3 2 2 ) is a white crystalline solid that is readily soluble in water and Soda ash (sodium carbonate, Na strongly CO 2 3 alkaline. Commercial soda ash is used as a raw material in a variety of industrial processes and in many familiar consumer products such as glass, soap and detergents, paper, textiles, and food. (Emissions from soda ash used in glass production a re reported under Section 4.3 , Glass Production (IPCC Source Category 2A3). Glass production is its own source category and historical soda ash consumption figures have been adjusted to reflect this change.) After glass manufacturing, soda ash is used primarily to manufacture many sodium based inorganic chemicals, including - sodium bicarbonate, sodium chromates, sodium phosphates, and sodium silicates (USGS 2015b). I nternationally, two types of soda ash are produced, natural and synthetic. The United States produces only natural soda ash and is second only to China in total soda ash production. Trona is the principal ore from which natural soda ash is made. The Unite d States represents about one - fourth of total world soda ash output. Only two states produce natural soda ash: Wyoming and California. Of these two states, only net emissions of CO from Wyoming were calculated due to 2 216 specifics regarding the production pro cesses employed in the state. Based on preliminary 2015 reported data, the - estimated distribution of soda ash by end use in 2015 (excluding glass production) was chemical production, 58 percent; soap and detergent manufacturing, 13 percent; distributors, 11 percent; flue gas desulfurization, 9 percent; 217 other uses, 5 percent; water treatment, 3 percent; and pulp and paper production, 2 percent (USGS 2015b). U.S. natural soda ash is competitive in world markets because the majority of the world output of s oda ash is made synthetically. Although the United States continues to be a major supplier of world soda ash, China, which surpassed the United States in soda ash production in 2003, is the world’s leading producer. Eq. (1,714 kt emissions from the product ion of soda ash from trona were approximately 1.7 MMT CO In 2015, CO 2 2 CO ) in 2015. Total ). Soda ash consumption in the United States generated 1.1 MMT CO Eq. (1,075 kt CO 2 2 2 Table emissions from soda ash production and consumption in 2015 were 2.8 MMT CO ) (see Eq. (2,789 kt CO 2 2 Table ). - 4 - 3 9 and 40 4 Total emissions from soda ash production and consumption in 2015 decreased by approximately 1 percent from emissions in 2014, and have stayed approximately the same as 1990 levels. s have remained relatively constant over the time series with some fluctuations since 1990. In general, Emission these fluctuations were related to the behavior of the export market and the U.S. economy. The U.S. soda ash industry continued a trend of increased pro duction and value in 2015 since experiencing a decline in domestic and export sales caused by adverse global economic conditions in 2009. The annual average unit value of soda ash set a record high in 2012, and soda ash exports increased as well, accountin g for 55 percent of total production (USGS 2015b). 216 In California, soda ash is manufactured using sodium carbo nate - bearing brines instead of trona ore. To extract the sodium carbonate, the complex brines are first treated with CO in carbonation towers to convert the sodium carbonate into sodium 2 bicarbonate, which then precipitates from the brine solution. The p recipitated sodium bicarbonate is then calcined back into is recovered and recycled for use in the carbonation stage sodium carbonate. Although CO is generated as a byproduct, the CO 2 2 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 na was mined in Wyoming, the production numbers mined in Wyoming and shipped to Colorado. Prior to 2004, because the tro 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. 217 entages may not add up to 100 percent due to independent rounding. Perc - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 44 4

261 4 Table 3 9 : CO - Emissions from Soda Ash Production and Consumption Not Associated with 2 Glass Manufacturing (MMT CO Eq.) 2 Production Consumption Total Year 1.4 2.8 1990 1.4 2005 1.7 1.3 3.0 1.6 2011 1.1 2.7 2012 1.7 2.8 1.1 1.7 1.1 2.8 2013 2014 1.1 2.8 1.7 2015 1.7 1.1 2.8 Note: Totals may not sum due to independent rounding. 4 - 40 : CO Table Emissions from Soda Ash Production and Consumption Not Associated with 2 Glass Manufacturing (kt) Production Consumption Total Year 1990 1,431 1,390 2,822 2005 1,305 2,960 1,655 1,607 1,105 2011 2,712 1,097 1,665 2,763 2012 1,694 1,109 2013 2,804 1,685 1,143 2,827 2014 1,075 1,714 2015 2,789 Note: Totals may not sum due to independent rounding. Methodology kiln and chemically transformed into a crude soda During the production process, trona ore is calcined in a rotary ash that requires further processing. Carbon dioxide and water are generated as byproducts of the calcination process. Carbon dioxide emissions from the calcination of trona can be estimated based on the c hemical reaction shown above. Based on this formula, which is consistent with an IPCC Tier 1 approach, approximately 10.27 metric tons of trona are required to generate one metric ton of CO , or an emission factor of 0.0974 metric tons CO per 2 2 metric ton t rona (IPCC 2006). Thus, the 17.6 million metric tons of trona mined in 2015 for soda ash production (USGS 2015b) resulted in CO emissions of approximately 1.7 MMT CO Eq. (1,714 kt). 2 2 inor amounts in soap production, pulp Once produced, most soda ash is consumed in chemical production, with m and paper, flue gas desulfurization, and water treatment (excluding soda ash consumption for glass manufacturing). is usually emitted. In these applications, i As soda ash is consumed for these purposes, additional CO t is assumed 2 that one mole of carbon is released for every mole of soda ash used. Thus, approximately 0.113 metric tons of carbon (or 0.415 metric tons of CO ) are released for every metric ton of soda ash consumed. In future Inventories, 2 consistent with t he 2006 IPCC Guidelines for National Greenhouse Gas Inventories , emissions from soda ash consumption in chemical production processes will be reported under S 4.4 Other Process Uses of Carbonates ection (IPCC Category 2A4) (see Planned Improvements). for 1990 to 2015 were obtained Table 4 - 41 ) The activity data for trona production and soda ash consumption (see Minerals Yearbook for Soda Ash U.S. Geological Survey USGS) ( (1994 through 2015b) and USGS from the 45 - 4 Industrial Processes and Product Use

262 218 Mineral Industry Surveys for Soda Ash were collected (USGS 2016). Soda ash production and consumption data voluntary surveys of the U.S. soda ash industry. EPA will continue to analyze and assess by the USGS from opportunities to use facility level data from EPA’s GHGRP to improve the emission estimates for Soda Ash - 219 and UNFCCC gu idelines. Production source category consistent with IPCC 4 - 41 : Soda Ash Production and Consumption Not Associated with Glass Manufacturing Table (kt) a b Consumption Production Year 14,700 3,351 1990 17,000 3,144 2005 16,500 2,663 2011 17,100 2,645 2012 2013 17,400 2,674 2,754 2014 17,300 17,600 2,591 2015 a Soda ash produced from trona ore only. b Soda ash consumption is sales reported by producers which exclude imports. Historically, imported soda ash is less than 1 percent of the total U.S. consumption (Kostick 2012). - Series Consistency Uncertainty and Time Emission estimates from soda ash production have relatively low associated uncertainty levels in that reliable and - accurate data sources are available for the emission factor and activity data for trona based soda ash production. EPA plans to work with other entities to reassess the uncer tainty of these emission factors and activity data based on the most recent information and data. Through EPA’s GHGRP, EPA is aware of one facility producing soda ash from a liquid alkaline feedstock process. Soda ash production data was collected by the U SGS from voluntary surveys. A survey request was sent to each of the five soda ash producers, all of which responded, representing 100 percent of the total production data (USGS 2016). One source of uncertainty is the purity of the trona ore used for manuf acturing soda ash. The emission factor used for this estimate assumes the ore is 100 percent pure, and likely overestimates the emissions from soda ash manufacture. The average water - soluble sodium carbonate - bicarbonate content for ore mined in Wyoming ran ges from 85.5 to 93.8 percent (USGS 1995). For emissions from soda ash consumption, the primary source of uncertainty, however, results from the fact that these emissions are dependent upon the type of processing employed by each end use. Specific emission factors for each end - use are not available, - so a Tier 1 default emission factor is used for all end uses. Therefore, there is uncertainty surrounding the emission sumption and factors from the consumption of soda ash. Additional uncertainty comes from the reported con allocation of consumption within sectors that is collected on a quarterly basis by the USGS. Efforts have been made to categorize company sales within the correct end - use sector. The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4 - 42 . Soda Ash Production 5 percent and Consumption CO emissions for 2015 were estimated to be between 2.5 and 2.9 MMT CO Eq. at the 9 2 2 confidence level. This indicates a range of approximately 7 percent below and 6 percent above the emission estimate of 2.8 MMT CO Eq. 2 218 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 productio n of soda ash did not meet criteria to shield underlying confidential business information (CBI) from public disclosure. 219 nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf>. - See

263 Quantitative Uncertainty Estimates for C O 42 : Approach 2 - 4 Emissions from Soda Ash Table 2 Production and Consumption (MMT CO Eq. and Percent) 2 a 2015 Emission Estimate Uncertainty Range Relative to Emission Estimate Source Gas Eq.) (MMT CO Eq.) (MMT CO (%) 2 2 Upper Lower Lower Upper Bound Bound Bound Bound Soda Ash Production 2.9 +6% 2.8 2.5 CO - 7% 2 and Consumption a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. ensure consistency in emissions from 1990 Methodological approaches were applied to the entire time series to through 2015. 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, consisten t with Volume 1, 2006 IPCC Guidelines , see QA/QC and Verification Procedures section in the introduction of the Chapter 6 of the IPPU Chapter. Planned Improvements Soda ash consumed for other chemical uses will be extracted from the current soda ash consum ption emission estimates and included under those sources or Other Process Uses of Carbonates (IPCC Category 2A4) for the next Inventory report (i.e., 1990 through 2016). In addition, EPA plans to use GHGRP data for conducting category - 2006 IPCC Guidelines and the latest specific QC of emiss ion estimates consistent with both Volume 1, Chapter 6 of 220 level data in national inventories. - IPCC guidance on the use of facility Petrochemical Production (IPCC Source 4.12 8 ) Category 2B ) and methane ls results in the release of small amounts of carbon dioxide (CO The production of some petrochemica 2 ) emissions. Petrochemicals are chemicals isolated or derived from petroleum or natural gas. Carbon dioxide (CH 4 emissions from the production of acrylonitrile, carbon black, ethylene, ethylene dichloride, ethylene oxide, and , and CH emissions from the production of methanol and acrylonitrile are presented here and reported methanol 4 The petrochemical industry uses primary fossil fuels (i.e., nat under IPCC Source Category 2B8. ural gas, coal, petroleum, etc.) for non - fuel purposes in the production of carbon black and other petrochemicals. Emissions from indirect or direct process heat or fuels and feedstocks transferred out of the system for use in energy purposes (e.g., steam are currently accounted for in the Energy sector. production) Worldwide more than 90 percent of acrylonitrile (vinyl cyanide, C H N) is made by way of direct ammoxidation of 3 3 propylene with ammonia (NH This process is referred to as the SOHIO process ) and oxygen over a catalyst. 3 The primary use of acrylonitrile is as the raw after the Standard Oil Company of Ohio (SOHIO) (IPCC 2006). material for the manufacture of acrylic and modacrylic fibers. Other major uses include the production of plastics (acrylonit rile acrylonitrile [SAN]), nitrile rubbers, nitrile barrier resins, butadiene - styrene [ABS] and styrene - - The SOHIO process adiponitrile, and acrylamide. All U.S. acrylonitrile facilities use the SOHIO process (AN 2014). involves a fluidized bed reaction of - grade propylene, ammonia, and oxygen over a catalyst. The process chemical 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 byprod uct CO , carbon monoxide (CO), and water 2 220 I_Technical_Bulletin_1.pdf nggip.iges.or.jp/public/tb/TF - http://www.ipcc See < >. 47 - 4 Industrial Processes and Product Use

264 from the direct oxidation of the propylene feedstock, and produces other hydrocarbons from side reactions in the ammoxidation process. atic petroleum or coal - based Carbon black is a black powder generated by the incomplete combustion of an arom - feedstock at a high temperature. Most carbon black produced in the United States is added to rubber to impart The other major use of carbo n strength and abrasion resistance, and the tire industry is by far the largest consumer. black is as a pigment. The predominant process used in the United States is the furnace black (or oil furnace) In the furnace black process, carbon black oil (a heavy aromatic liquid) is continuously injected into the process. combustion zone of a natura - fired furnace. Furnace heat is provided by the natural gas and a portion of the l gas carbon black feedstock; the remaining portion of the carbon black feedstock is pyrolyzed to carbon black. The emissions are released fro resultant CO and uncombusted CH m thermal incinerators used as control devices, 4 2 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 o ther hydrocarbons (i.e., thermal black process), and by the open burning of carbon black feedstock (i.e., lamp black process); each of these process are used at only one U.S. plant each (EPA 2000). Ethylene (C of the plastics industry including polymers such as high, H ) is consumed in the production processes 4 2 low, and linear low density polyethylene (HDPE, LDPE, LLDPE); polyvinyl chloride (PVC); ethylene dichloride; ethylene oxide; and ethylbenzene. of ethane, propane, butane, Virtually all ethylene is produced from steam cracking naphtha, gas oil, and other feedstocks. The representative chemical equation for steam cracking of ethane to ethylene below: is shown 퐻 퐻 → 퐶 퐻 + 퐶 4 2 6 2 2 Small amounts of CH emissions are are also generated from the steam cracking process. In addition, CO and CH 4 2 4 also generated from combustion units. Ethylene dichloride (C Cl ) is used to produce vinyl chloride monomer, which is the precursor to polyvinyl H 2 4 2 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. Th e direct chlorination and oxychlorination reactions are shown below: (direct chlorination) 퐶푙 퐶 퐻 + 퐶푙 → 퐶 퐻 2 2 2 2 4 4 1 퐻 퐶 푂 + 2 퐻퐶푙 → 퐶 + 퐶푙 + 2 퐻 퐻 푂 (oxychlorination) 2 2 2 4 4 2 2 2 퐶 퐻 + 3 푂 → 2 퐶 푂 + 2 퐻 푂 (direct oxidation of ethylene during oxychlorination) 2 2 4 2 2 In addition to the byproduct CO produced from the direction oxidation of the ethylene feedstock, CO and CH 2 2 4 emissions are also generated from combustion units. Ethylene oxide (C O) is used in the manufacture of glycols, glycol ethers, alcohols, and a mines. Approximately H 2 4 70 percent of ethylene oxide produced worldwide is used in the manufacture of glycols, including monoethylene Ethylene oxide is produced by reacting ethylene with oxygen over a catalyst. The oxygen may be supplied to glycol. s through either an air (air process) or a pure oxygen stream (oxygen process). The byproduct CO the proces from 2 the direct oxidation of the ethylene feedstock is removed from the process vent stream using a recycled carbonate solution, and the recovered CO may be vented to the atmosphere or recovered for further utilization in other 2 sectors, such as food production (IPCC 2006). The combined ethylene oxide reaction and byproduct CO reaction is 2 The ethylene oxide process also exothermic and generates heat, which is recovered to produce steam for the process. produces other liquid and off - gas byproducts (e.g., ethane, etc.) that may be burned for energy recovery within the process. Al most all facilities, except one in Texas, use the oxygen process to manufacture ethylene oxide (EPA 2008). Methanol (CH OH) is a chemical feedstock most often converted into formaldehyde, acetic acid and olefins. It is 3 also an alternative transportation f uel, as well as an additive used by municipal wastewater treatment facilities in the a denitrification of wastewater. Methanol is most commonly synthesized from a synthesis gas (i.e., “syngas” – ) using a heterogeneous cat , CO, and CO mixture containing H There are a number of process techniques that alyst. 2 2 Worldwide, steam reforming of natural gas is the most common method; however, in can be used to produce syngas. Other syngas prod the United States only two facilities use steam reforming of natural gas. uction processes in the United States include partial oxidation of natural gas and coal gasification. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 48 4

265 Emissions of CO and CH from petrochemical production in 2015 were 28.1 MMT CO Eq. (28,062 kt CO ) and 2 2 4 2 44 43 Eq. (7 kt CH emissions from ), respectively (see Table 4 - 0.2 MMT CO and Table 4 - ). Since 1990, total CO 2 2 4 32 percent . Methane emissions from petrochemical (methanol and petrochemical production increased by 18 percent since 1990, given declining production. acrylonitrile) production have decreased by approximately Table 4 - 43 : CO and CH Eq.) Emissions from Petrochemical Production (MMT CO 2 4 2 Year 1990 2005 2011 2012 2013 2014 2015 CO 28.1 21. 3 27. 0 26.3 26.5 26.4 26.5 2 CH 0.2 0.1 + 0.1 0.1 0.1 0.2 4 21. 5 27. 0 26.4 Total 26. 6 26.5 26.6 28. 2 + Does not exceed 0.05 MMT CO Eq. 2 Note: Totals may not sum due to independent rounding. 4 - 44 : CO Table and CH Emissions from Petrochemical Production (kt) 4 2 Year 1990 2005 2011 2012 2013 2014 2015 26,496 326 28,062 CO 21, 26,395 2 6,972 26,338 26,501 2 CH 9 3 2 3 3 5 7 4 Methodology and CH Emissions of CO were calculated using the estimation methods provided by the 2006 IPCC Guidelines 2 4 and country - specific methods from EPA’s Greenhouse Gas Reporting Program (GHGRP). The 2006 IPCC Tier 1 method was used to estimate CO Guidelines emissions from production of acrylonitrile and and CH 4 2 221 - specific approach similar to the IPCC Tier 2 method was used to estimate CO methanol, and a country emissions 2 from carbon black, ethylene, ethylene oxide, and ethylene dichloride. The Tier 2 method for petrochemicals is a total emissions, but is not feedstock C mass balance method used to estimate total CO applicable for estimating CH 4 2 2006 IPCC Guidelines , the total feedstock C mass balance method (Tier 2) is based on emissions. As noted in the the assumption that all of the C input to the process is converted either into primary and secondary products or into CO . Further, the guideline states that while the total C mass balance method estimates total C emissions from the 2 process but does not directly provide an estimate of the amount of the total C emissions emitted as CO , or , CH 2 4 non - CH volatile organi c compounds (NMVOCs). This method accounts for all the C as CO Note, , including CH . 4 2 4 a subset of facilities reporting under EPA’s GHGRP use alternate methods to the C balance approach (e.g., ng approaches) to monitor CO Continuous Emission Monitoring Systems (CEMS) or other engineeri emissions and 2 these facilities are required to also report CH and N O emissions. Preliminary analysis of aggregated annual reports 2 4 shows that these emissions are less than 500 kt/year. EPA’s GHGRP is still reviewing this data to facilitate update of category - specific QC documentation and EPA plans to address this more completely in future reports. Carbon Black, Ethylene, Ethylene Dichloride and Ethylene Oxide 2015 2010 through were aggregated directly from EPA’s GHGRP dataset for 2010 Carbon dioxide emissions and national production In 2015, GHGRP data reported CO emissions of 3,260,000 metric tons from through 2015 (EPA GHGRP 2016). 2 carbon black production; 20,100,000 metric tons of CO from from ethylene production; 398,000 metr ic tons of CO 2 2 ethylene dichloride production; and 1,200,000 metric tons of CO from ethylene oxide production. These emissions 2 reflect application of a country - specific approach similar to the IPCC Tier 2 method and were used to estimate CO 2 221 - facility production . The GHGRP information for acrylonitrile and methanol level EPA has not integrated aggregated underlying CBI from aggregated information associated with production of these petrochemicals did not meet criteria to shield public disclosure. 49 - 4 Industrial Processes and Product Use

266 emission s 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 roduction data, etc.) to facilitate verification of reported emissions. supplemental emissions information (e.g., p Under EPA’s GHGRP, petrochemical production facilities are required to use either a mass balance approach or estimate facility level process CO CEMS to measure and report emissions for each petrochemical process unit to - 2 222 emissions. The mass balance method is used by most facilities and assumes that all the carbon input is converted . To apply the mas s into primary and secondary products, byproducts, or is emitted to the atmosphere as CO 2 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 etails on the greenhouse gas calculation , monitoring and QA/QC methods applicable to for their facility. More d petrochemical facilities can be found under Subpart X (Petrochemical Production) of the regulation (40 CFR Part 223 98). - level GHGRP reports thro ugh a multi step process (e.g. , combination of EPA verifies annual facility - and ensure that data submitted to EPA are electronic checks and manual reviews) to identify potential errors 224 accurate, complete, and consistent. 1990 through 2009 emission factor was calculated based on the types, an average national CO For prior years, for these petrochemical 2 GHGRP data and applied to production for earlier years in the time series (i.e., 1990 through 2010 through 201 5 2009) to estimate CO emissions from carbon black, ethylene, ethylene dichloride, and ethylene oxide. Carbon 2 emissions for petrochemical type dioxide emission factors were derived from GHGRP data by dividing annual CO 2 ined for “i” with annual production for petrochemical type “i” and then averaging the derived emission factors obta each calendar year 2010 through 2015. 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 so me operational efficiencies have been implemented at facilities over the time series. - specific CO emission factors that were calculated from the 2010 through 2015 GHGRP data The average country 2 are as follows: • 2.62 metric tons CO /metric ton carbon blac k produced 2 0.78 metric tons CO • /metric ton ethylene produced 2 • 0.040 metric tons CO /metric ton ethylene dichloride produced 2 • /metric ton ethylene oxide produced 0.44 metric tons CO 2 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 Guide to the Business of 1990 through 2009 were obtained from the American Chemistry Council’s (ACC’s) mistry (ACC 2002, 2003, 2005 through 2011). Annual production data for ethylene oxide were obtained from Che U.S. Chemical Industry Statistical Handbook for 2003 through 2009 (ACC 2014a) and from ACC’s Business ACC’s 222 CEMS, those CO emissions have been included in the aggregated A few facilities producing ethylene dichloride used CO 2 2 GHGRP emissions presented here. are from the combustion of For ethylene production processes, nearly all process emissions process off - gas. Under EPA’s GHGRP, Subpart X, ethylene facilities can report CO emissions from burning of process gases 2 using the o ptional combustion methodology for ethylene production processes, which is requires estimating emissions based on fuel quantity and carbon contents of the fuel. 2006 IPCC Guidelines (p. 3.57) which recommends This is consistent with the - including combustion emissions from fuels obtained from feedstocks (e.g., off under in the gases) in petrochemical production IPPU sector. In 2014, for example, this represented about 20 of the 80 reporting facilities. In addition to CO , these facilities are 2 required to report emissions of CH cilities using CEMS and N a O from combustion of ethylene process off - gas in flares. Both f 2 4 (consistent with a Tier 3 approach) and those using the optional combustion methodology are also required to report emissions of CH ysis of the and N . Preliminary anal O from combustion of petrochemical process - off gases and flares, as applicable 2 4 aggregated reported CH and N O emissions from facilities using CEMS and the optional combustion methodology suggests that 2 4 these annual emissions are less than 500 kt/yr so not significant enough to prioritize for inclusion in the report at thi s time. Pending resources and significance, EPA may include these emissions in future reports to enhance completeness. 223 See . 224 >. 07/documents/ghgrp_verification_factsheet.pdf - iles/2015 https://www.epa.gov/sites/production/f See < - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 50 4

267 of Chemistry for 1990 through 2002 (ACC 201 4b). As noted above, annual 2010 through 2015 production data for carbon black, ethylene, ethylene dichloride, and ethylene oxide, were obtained from EPA’s GHGRP. Acrylonitrile ed using the Tier 1 method in the Carbon dioxide and methane emissions from acrylonitrile production were estimat . and CH 2006 IPCC Guidelines Annual acrylonitrile production data were used with IPCC default Tier 1 CO 4 2 Emission factors used to estimate acrylonitrile emission factors to estimate emissions for 1990 through 2015. production emissions are as follows: • 0.18 kg CH /metric ton acrylonitrile produced 4 1.00 metric tons CO /metric ton acrylonitrile produced • 2 Business of Chemistry (ACC Annual acrylonitrile production data for 1990 through 2015 were obta ined from ACC’s 2016). Methanol Carbon dioxide and methane emissions from methanol production were estimated using Tier 1 method in the 2006 Annual methanol production data were used with IPCC default Tier 1 CO emission and CH IPCC Guidelines . 4 2 Emission factors used to estimate methanol production factors to estimate emissions for 1990 through 2015. emissions are as follows: • 2.3 kg CH /metric ton methanol produced 4 0.67 metric tons CO • /metric ton methanol produced 2 l production data for 1990 through 2015 were obtained from the ACC’s Business of Chemistry Annual methano (ACC 2016). 45 - : Production of Selected Petrochemicals (kt) Table 4 Chemical 1990 2005 2011 2013 2014 2015 2012 1,307 1,651 1,340 Carbon Black 1,280 1,230 1,210 1,220 Ethylene 16,542 23,975 25,100 24,800 25,300 25,500 26,900 Ethylene Dichloride 6,283 11,260 8,620 11,300 11,500 11,300 11,300 Ethylene Oxide 2,429 3,220 3,010 3,110 3,150 3,140 3,240 Acrylonitrile 1,214 1,325 1,055 1,220 1,075 1,095 1,050 Methanol 3,750 1,225 700 995 1,235 2,105 3,065 Uncertainty and Time - Series Consistency The CH and CO emission factors used for acrylonitrile and methanol production are based on a limited number of 2 4 - specific factors instead of default or average factors could increase the accuracy of the emission studies. Using plant estimates; however, such data were not available for the current Inventory report. The results of the quantitative uncertainty analysis for the CO emissions from carbon black production, ethylene, 2 Refer to the Methodolo gy section for ethylene dichloride, and ethylene oxide are based on reported GHGRP data. more details on how these emissions were calculated and reported to EPA’s GHGRP. There is some uncertainty in the applicability of the average emission factors for each petrochemical type across all prior years. While petrochemical production processes in the United States have not changed significantly since 1990, some operational efficiencies have been implemented at facilities over the time series. - Approach 2 quantitative uncertainty analysis are summarized in Table The results of the 46 . Petrochemical 4 production CO Eq. at the 95 percent emissions from 2015 were estimated to be between 26.7 and 29.4 MMT CO 2 2 confidence level . This indicates a range of approximately 5 percent below to 5 percent above the emission estimate Petrochemical production CH and 6 emissions from 2015 were estimated to be between 0.0 of 28.1 MMT CO Eq. 2 4 51 - 4 Industrial Processes and Product Use

268 0. 22 Eq. at the 95 percent confidence leve l. This indicates a range of approximately 57 percent below to MMT CO 2 2 MMT CO percent above the emission estimate of 0. Eq. 46 2 4 - 46 : Approach 2 Quantitative Uncertainty Estimates for CH Table Emissions from 4 Petroc Emissions from Carbon Black Production (MMT CO hemical Production and CO Eq. 2 2 and Percent) 2015 Emission a Estimate Source Uncertainty Range Relative to Emission Estimate Gas (MMT CO Eq.) (MMT CO Eq.) (%) 2 2 Upper Lower Lower Upper Bound Bound Bound Bound Petrochemical CO 28.1 26.7 +5% 29.4 - 5% 2 Production Petrochemical +4 % 6 CH % 0. 2 0.0 6 0. 22 - 5 7 4 Production a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. approaches Methodological were applied to the entire time series to ensure consistency in emissions from 1990 through 2015. 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. Recalculation Discussion CO were obtained from the EPA ’s GHGRP CBI aggregation analysis and updated from 2010 emissions data 2 through 2015. In addition, this update included adjusted rounding and altering of significant figures for these years. As a result of the rounding, some reported petrochemica ) increased and some l emissions in metric tons (MT CO 2 decrease d ) this across the 2010 to 2015 time series; however, when converted to million metric tons (MMT CO 2 annual emissions compared in its effect on total Eq. change) change became insignificant (less than 0.05 MMT CO 2 to the previous Inventory report. Planned Improvements - specific QC of activity data and EFs, along with further assessment of Improvements include completing category and N CH O emissions to enhance completeness in reporting of e missions from petrochemical production, pending 4 2 resources, significance and time series consistency considerations. Pending resources, a secondary potential improvement for this source category would focus on continuing to analyze the fuel and feedstock d ata 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 process ergy sector and CO energy use of fuels in the en counting may occur between CO - estimates of non 2 2 emissions from petrochemical production in this sector. Data integration is not feasible at this time as feedstock data from the Energy Information Administration ( 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). EPA, through its GHGRP, currently does not collect complete data on quantities of fuel consumed as feedstocks by petrochemical producers, on Recent revisions to reporting requirements finalized in 2014 and 2016 ly feedstock type. 225 ( 79 FR 63750; 81 FR 89188) may address this issue in future reporting years for the GHGRP data allowing for energy uses of fue easier data integration between the non - ls category and the petrochemicals category presented in reported petrochemical feedstocks but further QC is required. this chapter. EPA’s GHGRP has initiated analysis of 225 rulemakings - https://www.epa.gov/ghgreporting/historical - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 52 4

269 This planned improvement i s completed to report on progress in this currently under development and has not been current Inventory. 4.13 22 Production (IPCC Source HCFC - ) Category 2 B9a 23 or CHF Trifluoromethane (HFC - ) is generated as a byproduct during the manufacture of chlorodifluoromethane 3 22), which is primarily employed in refrigerat ion and air conditioning systems and as a chemical feedstock (HCFC - - 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 a nd 2007, U.S. production fluctuated but generally remained above 1990 levels. In 2008 and 2009, U.S. production declined markedly and has - 22 depletes stratospheric ozone, its production for non feedstock remained near 2009 levels since. Because HCFC - 226 s scheduled to be phased out by 2020 under the U.S. Clean Air Act. Feedstock production, however, is uses i permitted to continue indefinitely. - 22 is produced by the reaction of chloroform (CHCl ) and hydrogen fluoride (HF) in the presence of a HCFC 3 , (where x + y = 5), which reacts with catalyst, Sb . The reaction of the catalyst and HF produces SbCl Cl F y x 5 chlorinated hydrocarbons to replace chlorine atoms with fluorine. The HF and chloroform are introduced by submerged piping into a continuous - flow reactor that c ontains the catalyst in a hydrocarbon mixture of chloroform - - 21 (CHCl 22 F), HCFC and partially fluorinated intermediates. The vapors leaving the reactor contain HCFC 2 (CHClF ), HFC - 23 (CHF 21) and ), HCl, chloroform, and HF. The under - fluorinated intermediates (HCFC - 3 2 chloroform are then condensed and returned to the reactor, along with residual catalyst, to undergo further C - 22, HFC - 23, HCl and residual HF. The fluorination. The final vapors leaving the condenser are primarily HCF - 22, the HFC - 23 may be HCl is recovered as a useful byproduct, and the HF is removed. Once separated from HCFC released to the atmosphere, recaptured for use in a limited number of applications, or destroyed. Two faci lities produced HCFC - 22 in the United States in 2015. Emissions of HFC - 23 from this activity in 2015 were estimated to be 4.3 MMT CO Eq. (0.29 kt) ( see Table 4 - 47 ). This quantity represents a 15 percent decrease 2 from 2014 emissions and a 91 percent decrease from 1990 emissions . The de crease from 201 4 emissions and the in the HFC decrease from 1990 emissions primarily by changes were caused - 23 emission r ate (kg HFC - 23 - emitted/kg HCFC - 22 produced). The long term decrease in the emission rate is primarily attributable to six factors: 22 since - - 23 generated have ceased production of HCFC (a) five plants that did not capture and destroy the HFC 22 1990 lant that captures and destroys the HFC - 23 generated began to produce HCFC - (b) one p ; (c) one plant ; - 23 generated ; (d) the same plant implemented and documented a process change that reduced the amount of HFC - 23, primarily for destruction and secondarily for sale ; (e) another plant began destroying began recovering HFC - ; 23 HFC 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 - 47 : - 23 Emissions from HCFC - 22 Production (MMT CO 23) Eq. and kt HFC - HFC 2 kt HFC - Year MMT CO Eq. 23 2 46.1 3 1990 2005 20.0 1 2011 8.8 0.6 5.5 0.4 2012 2013 4.1 0.3 2014 5.0 0.3 0.3 2015 4.3 226 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] 53 - 4 Industrial Processes and Product Use

270 Methodology - 23 emissions for five of the eight HCFC - 22 plants that have operated in the United States since To estimate HFC methods comparable to the Tier 3 methods in the 2006 IPCC Guidelines (IPCC 2006) were used. Emissions , 1990 - 22 production facilities to EPA’s rts submitted by U.S. HCFC for 2010 through 2015 were obtained through repo GHGRP). EPA’s GHGRP mandates that all Greenhouse Gas Reporting Program ( - 22 production facilities HCFC report ir annual emissions of HFC - 23 from HCFC the 22 production processes and HFC - 23 destruct ion 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 emissi ons were aggregated and reported to EPA on an annual basis. 2006 For the other three plants, the last of which closed in 1993, methods comparable to the Tier 1 method in the were used. Emissions from these three plants have been calculated using the recommended IPCC Guidelines - 23/kg HCFC - emission factor for unoptimized plants operating before 1995 (0.04 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 - 2 3 to estimate their emissions of HFC - 23. Plants using thermal oxidation to abate their HFC - 23 emissions monitor the performance of their oxidizers to verify that the HFC - 23 is almost completely destroyed. 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 199 0 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, , 2009, 2010). To estimate 2010 through 2015 emissions, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008 22 production and HFC 23 emissions) reported through EPA’s GHGRP - - level data (including both HCFC facility - were analyzed. In 1997 and 2008, comprehensive reviews of plant - - 23 emis sions and HCFC - level estimates of HFC 22 production were performed (RTI 1997; RTI 2008). The 1997 and 2008 reviews enabled U.S. totals to be reviewed, updated, and where necessary, corrected, and also for plant level uncertainty analyses (Monte - Carlo - for 1990, 1995, 2000, 2005, and 2006. Estimates of annual U.S. HCFC - simulations) to be performed 22 production are presented in Table 4 - 48 . Table 4 - 48 : HCFC - 22 Production (kt) Year kt 1990 139 2005 156 2011 110 2012 96 2013 C 2014 C C 2015 C (CBI) Note: HCFC - 22 production in 2013 through 2015 is considered Confidential Business Information (CBI) as there were only two producers of HCFC - 22 in those years. - Uncertainty and Time Series Consistency - The uncertainty analysis presented in this level Monte Carlo Stochastic Simulation for section was based on a plant 2006. The Monte Carlo analysis used estimates of the uncertainties in the individual variables in each plant’s This analysis was based on the generation of 10,00 estimating procedure. 0 random samples of model inputs from the - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 54 4

271 probability density functions for each input. A normal probability density function was assumed for all normal probability density measurements and biases except the equipment leak estimates for one plant; a log - The simulation for 2006 yielded a 95 - percent fu nction was used for this plant’s equipment leak estimates. confidence interval for U.S. emissions of 6.8 percent below to 9.6 percent above the reported total. The relative errors yielded by the Monte Carlo Stochastic Simulation for 2006 were applied to the U.S. emission 5 . The resulting estimates of absolute uncertainty are likely to be reasonably accurate because (1) estimate for 201 lieved to have changed significantly since the methods used by the three plants to estimate their emissions are not be 5 2006, and (2) although the distribution of emissions among the plants may have changed between 2006 and 201 - 22 production and the HFC - 23 emission rate declined significantly), the two plants th at (because both HCFC contribute significantly to emissions were estimated to have similar relative uncertainties in their 2006 (as well as 2005) emission estimates. Thus, changes in the relative contributions of these two plants to total emissions are not likely to have a l arge impact on the uncertainty of the national emission estimate. . Approach 2 quantitative uncertainty analysis are summarized in Table 4 - 49 HFC The results of the - 23 emissions CO from HCFC 4.0 and 4.7 MMT 22 production were estimated to be between Eq. at the 95 percent confidence - 2 level. This indicates a range of approximately 7 percent below and 10 percent above the emission estimate of 4.3 MMT CO Eq . 2 4 - 49 : Approach 2 Quantitative Uncertainty Estimates for HFC - 23 Emissions from Table - 22 Production ( MMT CO Eq. and Percent) HCFC 2 a Emission Estimate Uncertainty Range Relative to Emission Estimate 201 5 Gas Source ( CO MMT Eq.) ( MMT CO Eq.) (%) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound +10% 7% HCFC - 22 Production HFC - 23 4.3 4.0 4.7 - a Range of emissions reflects a 95 percent confidence interval. QA/QC and Verification Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source - specific quality control measures for the HCFC - 22 Production category included the QA/QC requirements and verification procedures of the Greenhouse Gas Reporting - 22 producers are required to (1) Program. Under that Program, HCFC measure concentrations of HFC - 23 and HCFC - 22 in the product stream at least weekly using equipment and methods (e.g., gas chromatography) with an accuracy and precision of 5 percent or bett er at the concentrations of the - process samples, (2) measure mass flows of HFC 23 and HCFC - 22 at least weekly using measurement devices (e.g., flowmeters) with an accuracy and precision of 1 percent of full scale or better, (3) calibrate mass measurement evices at the frequency recommended by the manufacturer using traceable standards and suitable methods d published by a consensus standards organization, (4) calibrate gas chromatographs at least monthly through analysis of certified standards, and (5) docum ent these calibrations. EPA verifies annual facility - level reports from HCFC - 22 producers through a multi - step process (e.g. , a combination of electronic checks and manual reviews by staff) to identify potential errors and ensure that data submitted to EP A are accurate, complete, and consistent. Based on the results of the verification process, the EPA 227 follows up with facilities to resolve mistakes that may have occurred. 227 https://www.epa.gov/sites/production/files/2015 07/documents/ghgrp_verification_factsheet.pdf - 55 - 4 Industrial Processes and Product Use

272 4.14 Carbon Dioxide Consumption (IPCC ) Source Category 2B 10 Carbon dioxide (CO ) is used for a variety of commercial applications, including food processing, chemical 2 production, carbonated beverage production, and refrigeration, and is also used in petroleum production for enhanced oil recovery (EOR). Carbon dioxide used for EOR is injected underground to enable additional petroleum used in commercial applications other than EOR is assumed to be produced. For the purposes of this analysis, CO 2 to be emitted to the atmosphere. Carbon dioxide used in EOR applications is discussed in the Energy chapter under “Carbon Capture and Storage, including Enhanced Oil Recovery” and is not discussed in this section. - CO reservoirs, as a byproduct from the energy and industrial Carbon dioxide is produced from naturally occurring 2 production processes (e.g., ammonia production, fossil fuel combustion, ethanol production), and as a byproduct g CO from the production of crude oil and natural gas, which contain naturally occurrin as a component. Only CO 2 2 produced from naturally occurring CO reservoirs, and as a byproduct from energy and industrial processes, and 2 used in industrial applications other than EOR is included in this analysis. Carbon dioxide captured from biogenic sources (e.g., ethanol production plants) is not included in the Inventory. Carbon dioxide captured from crude oil and gas production is used in EOR applications and is therefore reported in the Energy chapter. crude oil and natural gas production. This CO Carbon dioxide is produced as a byproduct of is separated from the 2 crude oil and natural gas using gas processing equipment, and may be emitted directly to the atmosphere, or captured and reinjected into underground formations, used for EOR, or sold for other commercial uses. A further discussion of CO titled “Carbon Dioxide Transport, used in EOR is described in the Energy chapter in Box 3 - 7 2 al Storage.” Injection, and Geologic In 2015, the amount of CO produced and captured for commercial applications and subsequently emitted to the 2 atmosphere was 4.3 MMT CO ). This is a decrease of approximately 4 percent from Eq. (4,296 kt) (see Table 4 - 50 2 2014 levels and an increase of approximately 192 percent since 1990 . The 2015 emissions estimate is based on a linear extrapolation correlated with the trend found in the G HGRP data, as described in the Methodology section below. Emissions from CO Eq. and kt) Consumption (MMT CO Table 4 - 50 : CO 2 2 2 Year MMT CO Eq. kt 2 1990 1.5 1,472 2005 1.4 1,375 4,083 4.1 2011 4.0 4,019 2012 2013 4.2 4,188 4.5 2014 4,471 4.3 4,296 2015 Methodology Carbon dioxide emission estimates for 1990 through 2015 were based on the quantity of CO extracted and 2 transferred for industrial applications (i.e., non uses). Some of the CO - EOR end - produced by these facilities is used 2 for EOR and some is used in other commercial applications (e.g., chemical manufacturing, food production). It is assumed t production used in commercial applications other than EOR is eventually hat 100 percent of the CO 2 released into the atmosphere. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 56 4

273 2010 through 201 5 Greenhouse Gas Reporting Program For 2010 through 2014, data from EPA’s (GHGRP) (Subpart PP) were - - level estimate for use in the Inventory (EPA GHGRP from facility aggregated level reports to develop a national 2016). Facilities report CO extracted or produced from natural reservoirs and industrial sites, and CO captured 2 2 transferred to various end from energy and industrial processes and use applications to EPA’s GHGRP. This - analysis includes only reported CO transferred to food and beverage end - uses. EPA is continuing to analyze and 2 uses to enhance the complete assess integration of CO transferred to other end - ness of estimates under this source 2 category. Other end - uses include industrial applications, such as metal fabrication. EPA is analyzing the information reported to ensure that other end - use data excludes non - emissive applications and publication will not reveal confidential business information (CBI). Reporters subject to EPA’s GHGRP Subpart PP are also required to report the quantity of CO that is imported and/or exported. Currently, these data are not publicly available through the 2 GHGRP due to data co nfidentiality reasons and hence are excluded from this analysis. Facilities subject to Subpart PP of EPA’s GHGRP are required to measure CO extracted or produced. More details 2 on the calculation and monitoring methods applicable to extraction and production facilities can be found under 228 Subpart PP: Suppliers of Carbon Dioxide of the regulation, Part 98. The number of facilities that reported data to EPA’s GHGRP Subpart PP (Suppliers of Carbon Dioxide) for 2010 through 2014 is much higher (ranging from 44 to 48) than the number of facilities included in the Inventory for the 1990 to 2009 time period prior to the includes availability of GHGRP data (4 facilities). The difference is largely due to the fact the 1990 to 2009 data only CO transferred to end - use applications from naturally occurring CO reservoirs and excludes industrial sites. 2 2 For 2015, data from EPA’s GHGRP (Subpart PP) was unavailable for use in the current Inventory report due to data confidentiality reasons. A linear trend extrapolat ion was performed based on previous GHGRP reporting years - (2010 to 2014) to estimate 2015 emissions. This time series recalculation is consistent with Volume 1, Chapter 5 of the 2006 IPCC Guidelines . 1990 through 2009 For 1990 through 2009, data from EPA’ production data from s GHGRP are not available. For this time period, CO 2 - four naturally reservoirs were used to estimate annual CO emissions. These facilities were Jackson occurring CO 2 2 Dome in Mississippi, Brave and West Bravo Domes in New Mexico, and M cCallum Dome in Colorado. The for use in both EOR and in other commercial applications facilities in Mississippi and New Mexico produced CO 2 (e.g., chemical manufacturing, food production). The fourth facility in Colorado (McCallum Dome ) produced CO 2 for c New Mexico Bureau of Geology and Mineral Resources 2006) . ommercial applications only ( - EOR applications for the Carbon dioxide production data and the percentage of production that was used for non Jackson Dome, Mississippi facility were obtained from Adva nced Resources International (ARI 2006, 2007) for 1990 to 2000, and from the Annual Reports of Denbury Resources ( Denbury Resources 2002 through 2010) for production in units of MMCF CO 2001 to 2009 (see 4 - 51 ). Denbury Resources reported the average CO per Table 2 2 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 Dom e 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 2 014). 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. 228 bin/text - See . - 57 - 4 Industrial Processes and Product Use

274 Table 4 51 : CO - Production (kt CO ) and the Pe rcent Used for Non - EOR Applications 2 2 Year Jackson Dome, % Total CO McCallum West Bravo Bravo Dome, 2 NM Dome, CO - Non MS Dome, NM CO Production 2 a Production Production Production from Extraction EOR CO CO CO Production 2 2 2 - (kt) (% Non - (kt) (% Non (kt) (% Non - - (kt) (% Non and Capture EOR) EOR) Facilities (kt) EOR) EOR) 1990 1,344 (100%) 63 (1%) + 65 (100%) NA NA NA 2005 1,254 (27%) 58 (1%) + NA 63 (100%) 2011 NA NA NA NA 66,241 6% NA NA NA NA 66,326 6% 2012 2013 NA NA NA NA 68,435 6% NA NA NA 2014 NA 72,000 6% 2015 NA NA NA NA 72,569 6% + Does not exceed 0.5 percent. a Includes only food & beverage applications. were aggregated at the national level. NA (Not available). For 2010 through 2015, the publicly available GHGRP data Facility level data are not publicly available from EPA’s GHGRP. - - Series Consistency Uncertainty and Time There is uncertainty associated with the data reported through EPA’s GHGRP. Specifically, there is uncertainty associated with the amount of CO consumed for food and beverage applications given a threshold for reporting 2 under GHGRP applicable to those reporting under Subpart PP, in addition to the exclusion of the amount of CO 2 end - transferred to all other quantities that are being used for use categories. This latter category might include CO 2 - EOR industrial applications such as firefighting. Second, uncertainty is associated with the exclusion of non imports/exports data for CO suppliers. Currently t hese data are not publicly available through EPA’s GHGRP and 2 EPA verifies annual facility - - step process (e.g. , hence are excluded from this analysis. level reports through a multi combination of electronic checks and manual reviews by staff) to identify pot ential errors and ensure that data submitted to EPA are accurate, complete, and consistent. Based on the results of the verification process, the EPA 229 follows up with facilities to resolve mistakes that may have occurred. titative uncertainty analysis are summarized in Table 4 - The results of the Approach 2 quan . Carbon dioxide 52 consumption CO emissions for 2015 were estimated to be between 4.1 and 4.5 MMT CO Eq. at the 95 percent 2 2 confidence level. This indicates a range of approximately 5 percent below to 5 percent above the emission estimate of 4.3 MMT CO Eq. 2 Table 4 - 52 : Approach 2 Quantitative Uncerta inty Estimates for CO Emissions from CO 2 2 Eq. and Percent) Consumption (MMT CO 2 a Uncertainty Range Relative to Emission Estimate 2015 Emission Estimate Source Gas Eq.) (MMT CO (MMT CO Eq.) (%) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound Consumption CO 4.3 4.1 4.5 - 5% CO +5% 2 2 a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. 229 07/documents/ghg - https://www.epa.gov/sites/production/files/2015 See < >. rp_verification_factsheet.pdf - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 58 4

275 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990 through 2015. Details on the emission trends through time are described in more detail in the Methodology section, above. Volume 1, For more information on the general QA/QC process applied to this source category, consistent with , see QA/QC and Verification Procedures Chapter 6 of the 2006 IPCC Guidelines 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 guidel - level reporting data from EPA’s GHGRP, with the ines. This is required as the facility 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 230 when new data is available, are still These improvements, in addition to updating the time series be relied upon. in process and will be incorporated into future Inventory reports. Phosphoric Acid Production (IPCC Source 4.15 ) 10 Category 2B Phosphoric acid (H ) is a basic raw material used in the production of phosphate - based fer tilizers. Phosphoric PO 4 3 acid production from natural phosphate rock is a source of carbon dioxide (CO ) emissions, due to the chemical 2 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 ph osphate 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. The phosphoric acid production process involves chemical reaction of the calcium phosphate (Ca (PO ) ) 2 4 3 ) (EFMA SO PO ) and recirculated phosphoric acid (H component of the phosphate rock with sulfuric acid (H 4 3 2 4 2000). However, the generation of CO - sulfuric acid reaction, as shown below: is due to the associated limestone 2 퐶푎퐶푂 + 퐻 푂 +퐶푂 푆푂 2퐻 + 퐻 • 푂 →퐶푎푆푂 4 2 2 2 2 3 4 Total U.S. phosphate rock production sold or used in 2015 was 26.2 million metric tons (USGS 2017). Total imports of phospha te rock to the United States in 2015 were approximately 2.0 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 rock mining companies are vertically integrated with fertilizer being from Morocco (USGS 2017). All phosphate 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 2015 period, domestic production has decreased by nearly 47 percent. Total CO emissions from 2 phosphoric acid prod uction were 1.0 MMT CO Domestic consumption Eq. ( 999 kt CO ) in 2015 (see Table 4 - 53 ). 2 2 of phosphate rock in 2015 was estimated to have decreased by approximately 2 percent over 2014 levels, owing to 230 >. nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf - http://www.ipcc See < 59 - 4 Industrial Processes and Product Use

276 producers drawing from higher than average inventories and the closure of a mine in Florida. Domestic consumption ( also decreased because of lower phosphoric acid production USGS 2016). - 53 : CO Table 4 Eq. and kt) Emissions from Phosphoric Acid Production (MMT CO 2 2 Year MMT CO Eq. kt 2 1.5 1,529 1990 1,342 2005 1.3 2011 1.2 1,171 1.1 1,118 2012 2013 1.1 1,149 1.0 1,038 2014 2015 1.0 999 Methodology Carbon dioxide emissions from production of phosphoric acid from phosphate rock are estimated by multiplying the ) contained in the natural phosphate rock as calcium average amount of inorganic carbon (expressed as CO 2 carbonate by the amount of phosphate rock that is used annually to produce phosphoric acid, accounting for oduction and net imports for consumption. The estimation methodology is as follows: domestic pr 퐸 = 퐶 ×푄 푝푟 푝푎 푝푟 where, E = CO emissions from phosphoric acid production, metric tons 2 pa / = Average amount of carbon (expressed as CO C ) in natural phosphate rock, metric ton CO 2 pr 2 metric ton phosphate rock Quantity of phosphate rock used to produce phosphoric acid = Q pr emissions calculation methodology is based on the assumption that all of the inorganic C (calcium The CO 2 carbonate) content of the phosphate ro in the phosphoric acid production process and is ck reacts to produce CO 2 emitted with the stack gas. The methodology also assumes that none of the organic C content of the phosphate rock is converted to CO and that all of the organic C content remains in the phosphoric acid product. 2 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 reporte Table 4 - 54 d the annual amounts of phosphate rock exported and imported for consumption (see For the years 1990 through 1992, and 2005 through 2015, 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. prod uction in those states from 1993 to 2004 data. For the years 2005 through 2015, 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). Da ta 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 2015 were obtained from USGS Minerals Yearbook: Phosphate Rock (US GS 1994 through 2015b), and from USGS Minerals Commodity Summaries: Phosphate Rock (USGS 2016, 2017). From 2004 through 2015, 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 i n 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 inorganic , respectively (see Table C North Carolina and Idaho contains approximately 0.41 percent and 0.27 pe rcent ). 55 - 4 - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 60 4

277 Carbonate content data for phosphate rock mined in Florida are used to calculate the CO emissions from 2 f phosphate rock mined in Florida and North Carolina (80 percent of domestic production) and consumption o emissions from consumption carbonate content data for phosphate rock mined in Morocco are used to calculate CO 2 of imported phosphate rock. The CO lation is based on the assumption that all of the domestic emissions calcu 2 production of phosphate rock is used in uncalcined form. As of 2006, the USGS noted that one phosphate rock lable for this single producer in Idaho produces calcined phosphate rock; however, no production data were avai producer (USGS 2006). The USGS confirmed that no significant quantity of domestic production of phosphate rock is in the calcined form (USGS 2012). Table 54 : Phosphate Rock Domestic Co nsumption, Exports, and Imports (kt) 4 - 1990 Location/Year 2005 2011 2012 2013 2014 2015 U.S. Domestic Consumption 49,800 35,200 28,600 27,300 28,800 26,700 26,200 FL and NC 42,494 28,160 22,880 21,840 23,040 21,360 20,960 5,720 UT 7,306 7,040 and 5,460 5,760 5,340 5,240 ID Exports — FL and NC 6,240 0 0 0 0 0 0 Imports 2,630 3, 750 451 3,570 3,170 2,390 1,960 44,011 37,830 3 2,350 30,870 31,970 29,090 28,160 Total U.S. Consumption 55 : Chemical Composition of Phosphate Rock (Percent by Weight) Table 4 - North North Carolina Idaho Central Florida Florida (calcined) Composition Morocco (calcined) 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) 0.83 0.35 0.60 0.00 0.10 Inorganic Carbon (as CO ) 3.67 3.43 1.50 1.00 5.00 2 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 201 5 For previous years in the time series, . 2006, only total U.S. phosphate rock USGS provided the data disaggregated regionally; however, beginning in Regional production for 201 5 was estimated based on regional production data from production was reported. specific emission factors. There is uncertainty associated with the - previous years and multiplied by regionally 5 regional production data represents actual production in those regions. degree to which the estimated 201 Total U.S. phosphate rock production data are not considered to be a significant source of uncertainty because all the ers report their annual production to the USGS. Data for exports of phosphate rock domestic phosphate rock produc used in the emission calculation are reported by phosphate rock producers and are not considered to be a significant source of uncertainty. are based on international trade data collected by the U.S. Data for imports for consumption Census Bureau. These U.S. government economic data are not considered to be a significant source of uncertainty. An additional source of uncertainty in the calculation of CO horic acid production is the emissions from phosp 2 carbonate composition of phosphate rock the composition of phosphate rock varies depending upon where the ; material is mined, and may also vary over time. The Inventory relies on one study (FIPR 2003a) of chemical f the phosphate rock; limited data are available beyond this study. composition o Another source of uncertainty is the disposition of the organic carbon content of the phosphate rock. A representative of the Florida Institute of Phosphate Research ( FIPR ) indicated that in the phosphoric acid production process, the organic C content of the mined phosphate rock generally remains in the phosphoric acid product, which is what produces the color of the Organic carbon is therefore not inc luded in the calculation of CO emissions phosphoric acid product (FIPR 2003b). 2 from phosphoric acid production. A third source of uncertainty is the assumption that all domestically - produced phosphate rock is used in phosphoric acid production and used without first being calcined. Calcinat ion of the phosphate rock would result in conversion However, according to air permit information available to of some of the organic C in the phosphate rock into CO . 2 13). the public, at least one facility has calcining units permitted for operation (NCDENR 20 61 - 4 Industrial Processes and Product Use

278 Finally, USGS indicated that approximately 7 percent of domestically - produced phosphate rock is used to based chemicals, rather than phosphoric acid (USGS manufacture elemental phosphorus and other phosphorus - According to USGS, there is only one domestic producer of elemental phosphorus, in Idaho, and no data 2006). were available concerning the annual production of this single producer. Elemental phosphorus is produced by reducing phosphate rock with coal coke, and it is therefore assumed that 100 p ercent of the carbonate content of the in the elemental phosphorus production process. The calculation for CO phosphate rock will be converted to CO 2 2 emissions is based on the assumption that phosphate rock consumption, for purposes other than phosphoric acid emissions from 100 percent of the inorganic carbon content in phosphate rock, but none production, results in CO 2 from the organic carbon content . Approach 2 quantitative uncertainty analysis are summarized in Table 4 - 56 . 2015 p hosphoric acid The results of the 2 emissions were e stimated to be between 0. 8 and 1. MMT CO Eq. at the 95 percent confidence production CO 2 2 This indicates a range of approximately 19 percent below and 20 percent above the emission estimate of 1. 0 level. CO . Eq MMT 2 Table 4 - 56 : Approach 2 Quantitative Uncertainty Estimates for CO Emissions from 2 Phosphoric Acid Production (MMT CO Eq. and Percent) 2 a Uncertainty Range Relative to Emission Estimate 201 5 Emission Estimate Gas Source (MMT CO (MMT CO Eq.) (%) Eq.) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound Phosphoric Acid Production CO 1. 0 0. 8 1. 2 - 19% +20% 2 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 consistency in emissions from 1990 through 201 5 . Details on the emission trends through time are described in more detail in the Methodology section, above. ed to this source category, consistent with Volume 1, For more information on the general QA/QC process appli , see QA/QC and Verification Procedures Chapter 6 of the section in the introduction of the 2006 IPCC Guidelines IPPU Chapter. Recalculations Discussion Relative to the previous Inventory, the phosphate rock import data for 2011 through 2014 were revised based on updated data publicly available from USGS (USGS 2016). This revision resulted in a change in emission estimates ranging from approximately 2 to 5 percent across the time series of 2011 through 20 14 compared to the previous Inventory report. Planned Improvements EPA continues to evaluate potential improvements to the Inventory estimates for this source category, which include direct integration of EPA’s Greenhouse Gas Reporting Program ( GHGRP) data for 2010 through 2015 and the use of reported GHGRP data to update the inorganic C content of phosphate rock for prior years Confidentiality of CBI . continues to be assessed, in addition to the applicability of GHGRP data for the averaged inorganic C cont ent data (by region) from 2010 through 2015 to inform estimates in prior years in the required time series (i.e., 1990 through 2009) . In implementing improvements and integration of data from EPA’s GHGRP, the latest guidance from the 231 IPCC on the use of facility - level data in national inventories will be relied upon. This planned improvement is still in development by EPA and have not been implemented into the current inventory report. 231 >. nggip.iges.or.jp/public/tb/TFI_Technical_Bulletin_1.pdf - http://www.ipcc See < - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 62 4

279 4.16 Iron and Steel Production (IPCC Source Category 2C1) and Metallurgica l Coke Production step process that generates process - related emissions of Iron and steel production is a multi ( CO - ) carbon dioxide 2 methane ( CH as raw materials are refined into iron and then transformed into crude steel. ) and Emissions from 4 conventional fuels (e.g., natural gas, fuel oil) consumed for energy purposes during the production of iron and steel are accounted for in the Energy chapter. Iron and steel production includes six distinct production processes: coke production, sinter production, dir ect reduced iron (DRI) production, pig iron production, electric arc furnace (EAF) steel production, and basic oxygen furnace (BOF) steel production. The number of production processes at a particular plant is dependent upon the n. In addition to the production processes mentioned above, CO specific plant configuratio is also generated at iron 2 and steel mills through the consumption of process s (e.g., blast furnace gas, coke oven gas) used for byproduct various purposes including heating, annealing, and ele ctricity generation. Process byproduct s sold for use as synthetic natural gas are deducted and reported in the Energy chapter. In general, CO emissions are generated in 2 these production processes through the reduction and consumption of various carbon - con taining inputs (e.g., ore, scrap, flux, coke byproducts). In addition, fugitive CH from these processes but emissions can also be generated 4 also sinter, direct iron and pellet production. Currently, there are approximately 11 integrated iron and steel ste elmaking facilities that utilize BOFs to refine and produce steel from iron and more than 100 steelmaking facilities that utilize EAFs to produce steel primarily from - (USGS 2016) . In addition, there are 21 cokemaking facilities, of w hich 7 facilities are co recycled ferrous scrap 62 percent of the raw steel produced in the United located with integrated iron and steel facilities. Slightly more than even states: Alabama, Arkansas, Indiana, Kentucky, Mississippi, Ohio, and Tennessee (AI SI States is produced in s 2016a) . Total production of crude steel in the United States between 2000 and 2008 ranged from a low of 99,320,000 tons to a high of 109,880,000 tons (2001 and 2004, respectively). Due to the decrease in demand caused by the global economic downturn ( particularly from the automotive industry), crude steel production in the United States sharply decreased to 65,459,000 tons in 2009. In 2010, crude steel production rebounded to 88,731,000 tons as economic conditions improved and then continued to increas e to 95,237,000 tons in 2011 and 97,7 69 ,000 tons in 2012; crude steel production slightly decreased to 95,766,000 tons in 2013 and then slightly increased to 97,195,000 tons in 2014 (AISI 201 6 a) ; crude steel production decreased to 86,912,000 tons in 2015, a decrease of roughly 10 percent from and 2014 levels The United States was the fourth largest producer of raw steel in the world, behind China , Japan . 5 India 4.9 percent of world production in 201 , accounting for approximately (AISI 201 6 a). The majority of CO 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 s, with lesser amounts emitted from the byproduct carbon from pig iron used to produce steel. use of flux and from the removal of According to the , the production of metallurgical coke from coking coal is considered to be 2006 IPCC Guidelines an energy use of fossil fuel and the use of coke in iron and steel production is considered to be an industrial process source. Therefore, the 2006 IPCC Guidelines suggest that emissions from the production of metallurgical coke should be reported separately in the Energy sector, while emissions from coke consumption in iron and steel in the Industrial Processes and Product Use sector. However, the approaches and production should be reported emission estimates for both metallurgical coke production and iron and steel production are both presented here because and the relevant activity data is used to estimate e missions from both metallurgical coke production much of , some s (e.g., coke oven gas) of the metallurgical coke production iron and steel production. For example byproduct nd steel production process are consumed during iron and steel production, and some byproducts of the iron a process (e.g., blast furnace gas) are consumed during metallurgical coke production. Emissions associated with the Emissions associated with the use of consumption of these byproducts are attributed at the point of consumption. 63 - 4 Industrial Processes and Product Use

280 conventi onal fuels (e.g., natural gas, fuel oil) for electricity generation, heating and annealing, or other miscellaneous purposes downstream of the iron and steelmaking furnaces are reported in the Energy chapter. Metallurgical Coke Production Eq. (2,839 kt CO f 5 were 2.8 MMT CO Emissions of CO rom metallurgical coke production in 201 ) (see Table 2 2 2 57 and Table 4 - 58 ). Emissions increased in 2015 from 2014 levels and have increased overall since 1990. In the - 4 s a proxy. In previous Inventory, 2014 domestic coke production data were not published, so 2013 data was used a 5, this report, domestic coke production data for 2015 was available and so 2014 data were not used as proxy for 201 differing from the previous I nventory report. 2014 published domestic coke production data were also updated. Coke production in 2015 was 34 percent lower than in 2000 and 50 percent below 1990. Overall, emissions from 13 percent (0.3 MMT CO metallurgical coke production have increased by Eq.) from 1990 to 2015. 2 Table 57 : CO 4 - Emissions from Metallurgical Coke Production (MMT CO Eq.) 2 2 Gas 1990 2005 2011 2012 2013 2014 2015 CO 2.5 2.0 1.4 0.5 1.8 2.0 2.8 2 0.5 2.0 1.4 Total 2.5 1.8 2.0 2.8 4 - 58 : CO Emissions from Metallurgical Coke Production (kt) Table 2 1990 2005 2011 Gas 2012 2013 2014 2015 CO 2, 503 2,04 4 1,42 6 54 3 1,82 4 2,014 2,839 2 Iron and Steel Production Emissions of CO 86 and CH kt) and 0.00 from iron and steel production in 2015 were 4 6.0 MMT CO Eq. ( 46,038 2 4 2 MMT CO Eq. (0.3 kt), respectively (see Table 4 - 59 through Table 4 - 62 ), totaling approximately 46.0 MMT CO 2 2 Eq. Emissions decreased in 2015 and have decreased overall since 1990 due to restructuring of the industry, technological improvements, and increased scr ap steel utilization. Carbon dioxide emission estimates include emissions from the consumption of carbonaceous materials in the blast furnace, EAF, and BOF, as well as blast furnace gas and coke oven gas consumption for other activities at the steel mill. 4 levels . Overall, domestic pig iron In 201 5 , domestic production of pig iron decreased percent from 201 by 13 production has declined since the 1990s. Pig iron production in 201 5 was 47 percent lower than in 2000 and 49 MMT CO dioxide e missions from steel production have increase percent below 1990. Carbon 1 percent (0. 2 by d 2 percent ( 3 ve declined by 5 emissions from iron and steel production ha 1990, while overall CO 52.9 Eq.) since 2 Eq.) MMT CO from 1990 to 201 5 . 2 Eq.) CO : 59 - 4 Table Emissions from Iron and Steel Production (MMT CO 2 2 Source/Activity Data 2015 2014 2013 2012 2011 2005 1990 1.0 1.1 1.1 1.2 1.2 1.7 2.4 Sinter Production 45.6 17.5 18.4 10.9 11.9 18.6 11.7 Iron Production 1.1 1.0 1.2 1.2 Pellet Production 1.8 1.5 1.2 Steel Production 8.6 7.8 8.1 9.9 9.3 9.4 7.9 a Other Activities 41.2 35.9 29.7 31.7 28.7 27.9 24.3 5 6.0 5 9.7 5 4.9 5 1.5 6.6 46.0 Total 9 9.0 6 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. : 60 4 Table Emissions from Iron and Steel Production (kt) CO - 2 1990 2005 2011 2012 2013 2014 2015 Source/Activity Data 1,188 1,016 1,104 1,117 1,159 1,663 2,448 Sinter Production 8,629 1 11,696 11,935 10,918 18,376 17,545 45,592 Iron Production - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 64 4

281 1,503 1,179 1,219 1,146 1,126 964 Pellet Production 1,817 9,356 9,255 9,860 8,617 7, 845 8,082 7,933 Steel Production a 41,193 Other Activities 29,683 31,750 28,709 35,934 24,282 27,911 Total 66,003 98,984 59,681 54,906 51,525 5 46,038 6,615 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. 4 - 61 : CH Emissions from Iron and Steel Production (MMT CO Table Eq.) 2 4 Source/Activity Data 2005 2011 2012 2013 2014 2015 1990 Sinter Production + + + + + + + + + Total + + + + + Does not exceed 0.05 MMT CO + Eq . 2 4 - 62 : CH Table Emissions from Iron and Steel Production (kt) 4 Source/Activity Data 1990 2005 2011 2012 2013 2014 2015 Sinter Production 0.9 0.6 0.4 0.4 0.4 0.4 0.3 Total 0.9 0.6 0.4 0.4 0.4 0.4 0.3 Methodology 2006 IPCC Emission estimates presented in this chapter are largely based on Tier 2 methodologies provided by the . These Tier 2 methodologies call for a mass balance accounting of the carbonaceous inputs and outputs Guidelines during the iron and steel production process and the metallurgical coke production process. Tier 1 methods are used roduction ) for for certain iron and steel production processes (i.e., sinter production , pellet production and DRI p which available data are insufficient for utilizing a Tier 2 method. The Tier 2 methodology equation is as follows: 44 ( ) ( ) 푄 ×퐶 =[∑ 퐸 −∑ ×퐶 푄 ]× 푎 푎 퐶푂 푏 푏 2 12 푎 푏 where, c tons = E Emissions from coke, pig iron, EAF steel, or BOF steel production, metri 2 CO a = Input material a = Output material b b Q = Quantity of input material a , metric tons a C Carbon content of input = a , metric tons C/metric ton material material a = Quantity of output material b , metric tons Q b = C Carbon content of output material b , metric tons C/metric ton material b = Stoichiometric ratio of CO 44/12 to C 2 The Tier 1 methodology equations are as follows: 퐸 =푄 ×퐸퐹 푠,푝 푠,푝 푠 퐸 ×퐸퐹 =푄 푑,퐶푂 2 푑 푑,퐶푂 2 × = 퐸퐹 퐸 푄 푝 , 2 푝 , 퐶푂 2 퐶푂 푝 where, E = Emissions from sinter production process for pollutant p (CO ), metric ton or CH 4 s,p 2 Q = Quantity of sinter produced, metric tons s ), metric ton or CH (CO /metric ton sinter p EF p = Emission factor for pollutant s,p 2 4 Industrial Processes and Product Use 5 - 4 6

282 E = Emissions from DRI production proc ess for CO , metric ton d, 2 CO2 Q Quantity of DRI produced, metric tons = d , metric ton CO = EF Emission factor for CO /metric ton DRI 2 2 d, CO2 = Q Quantity of pellets produced, metric tons p = Emission factor for CO , metric ton CO /metric ton pellets produced EF 2 CO2 2 p , Metallurgical Coke Production Coking coal is used to manufacture metallurgical coke that is used primarily as a reducing agent in the production of iron and steel, but is also used in the production of other metals including zinc and lead (see Zinc Produ ction and Lead Production sections of this chapter). Emissions associated with producing metallurgical coke from coking coal are estimated and reported separately from emissions that result from the iron and steel production process. To estimate emission s from metallurgical coke production, a Tier 2 method provided by the 2006 IPCC Guidelines was utilized. carbon contained in materials produced during the metallurgical coke production process The amount of tar) is deducted from the amount of carbon contained in materials (i.e., coke, coke breeze, coke oven gas, and coal consumed during the metallurgical coke production process (i.e., natural gas, blast furnace gas, and coking coal). Light oil, which is produced during the metallurgical coke production proce ss, is excluded from the deductions due The amount of carbon contained in these materials is calculated by multiplying the material - to data limitations. ). specific carbon content by the amount of material consumed or produce Table 4 - 63 see The amount of coal tar d ( produced was approximated using a production factor of 0.03 tons of coal tar per ton of coking coal consumed. The amount of coke breeze produc ed was approximated using a production factor of 0.075 tons of coke breeze per ton of coking coal consumed (AISI 2008; DOE 2000). Data on the consumption of carbonaceous materials (other than for integrated steel mills only (i.e., steel mills with coking coal) as well as coke oven gas production were available - located coke plants). co Therefore, carbonaceous material (other than coking coal) consumption and coke oven gas production were excluded from emission estimates for merchant coke plants. Carbon conta ined in coke oven gas used for coke - oven underfiring was not included in the deductions to avoid double - counting. Table 4 - 63 : Material Carbon Contents for Metallurgical Coke Production Material kg C/kg 0.62 Coal Tar Coke 0.83 0.83 Coke Breeze Coking Coal 0.73 kg C/GJ Material 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. 2006 IPC C Guidelines i.e., provide a Tier 1 CH Although the emission factor for metallurgical coke production ( 4 0.1 g CH per metric ton of coke production) , it is not appropriate to use because CO emissions were estimated 2 4 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 or as CO level greenhouse emissions. This is consistent with a preliminary assessment of aggregated facility output - 2 gas CH emissions reported by coke production facilities under EPA’s GHGRP. The assessment indicates that CH 4 4 emissions from coke production are i nsignificant 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 and has not incorporated these emi ssions 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 In formation Administration (EIA), Quarterly Coal Report: October t hrough December (EIA 1998 through 201 6 a) (see Table 4 - 64 ). 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 65 - 4 Table 2008) (see The ). AISI (AISI 2004 through 201 6 a) and through personal communications with AISI ( - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 66 4

283 factor for the quantity of coal tar produced per ton of coking coal consumed was provided by AISI ( AISI The 2008 ). 1 of ntity of coke breeze produced per ton of coking coal consumed was obtained through Table 2 factor for the qua - Data on natural gas the report (DOE 2000). Energy and Environmental Profile of the U.S. Iron and Steel Industry consumption and coke oven gas production at merc hant coke plants were not available and were excluded from the Carbon contents for coking coal, metallurgical coke, coal tar, coke oven gas, and blast furnace emission estimate. The C content for coke breeze w as assumed to equal the C content 2006 IPCC Guidelines . gas were provided by the of coke. 4 Production and Consumption Data for the Calculation of CO - and CH 64 Emissions : Table 2 4 from Metallurgical Coke Production (Thousand Metric Tons) 1990 2005 2011 2012 2013 2014 2015 Source/Activity Data Metallurgical Coke Production 18,825 21,259 19,445 Coking Coal Consumption at Coke Plants 19,481 19, 321 17,879 35,269 25,054 15,167 13,989 13,764 13,898 13 ,748 12,479 Coke Production at Coke Plants Coal Breeze Production 2,645 1,594 1,458 1,412 1,461 1,4 49 1,341 58 638 583 565 584 0 536 Coal Tar Production 1,058 - 65 : Production and Consumption Data for the Calculation of CO Table Emissions from 4 2 3 Metallurgical Coke Production ( ) M illion ft 1990 2005 2011 2012 2013 2014 2015 Source/Activity Data Metallurgical Coke Production Coke Oven Gas Production 250,767 114,213 109,044 113,772 108,162 102,899 84,336 3,267 Natural Gas Consumption 2,996 3,175 599 3,247 3,039 2,338 Blast Furnace Gas Consumption 24,602 4,460 3,853 4,351 4,255 4,346 4,185 Iron and Steel Production , Emissions of CO from sinter production and pellet production were estimated by direct reduced iron production 2 multiplying total national sinter production and the total national direct reduced iron production by Tier 1 CO 2 4 - 66 ). Because estimates of sinter production , emission factors (see direct reduced iron production and pellet Table production were not available, production was assumed to equal consumption. Table 4 - 66 : CO and Emission Factors for Sinter Production , Direct Reduced Iron Production 2 Pellet Production Metric Ton /Metric Ton Material Produced CO 2 Sinter 0.2 Direct Reduced Iron 0.7 Pellet Production 0.03 Source: IPCC 2006 ) , Table 4.1. ( carbon contained in the produced To estimate emissions from pig iron production in the blast furnace, the amount of 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, and direct coal injection). The carbon contained in the pig iron, blast furnace gas, and blast furnace inputs was estimated by multiplying t he material specific C content by - Table each material type (see - 67 ). Carbon in blast furnace gas used to pre - heat the blast furnace air is combusted 4 to form CO as not during this process. Carbon contained in blast furnace gas used as a blast furnace input w 2 - counting. included in the deductions to avoid double 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 direct reduced iron, pig iron, and flux additions 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 o specific carbon - In each case, the carbon was calculated by multiplying material ven gas, fluxes, and pig iron. 67 - 4 Industrial Processes and Product Use

284 contents by each material type (see - 67 ). For EAFs, the amount of EAF anode consumed was approximated Table 4 by multiplying total EAF steel production by the amount of EAF anode consumed per metric ton of steel produced amount of flux (e.g., limestone and (0.002 metric tons EAF anode per metric ton steel produced [AISI 2008]). The Other Process Uses of Carbonates ” source category dolomite) used during steel manufacture was deducted from the “ to avoid double - counting. (IPCC Source Category 2A4) Carbon dioxide emissions from the consumption of blast furnace gas and coke oven gas for other activities occurring at the steel mill were estimated by multiplying the amount of these materials consumed for these purposes by the - specific carbon content (see Table 4 - 67 ) . material emissions associated with the sinter production, direct reduced iron production, pig iron production, Carbon dioxide steel production, and ot her steel mill activities were summed to calculate the total CO emissions from iron and steel 2 Table 4 - 59 and Table 4 - 60 ). production (see 4 - 67 : Material Carbon Contents for Iron and Steel Production Table Material kg C/kg Coke 0.83 Direct 0.02 Reduced Iron Dolomite 0.13 EAF Carbon Electrodes 0.82 0.83 EAF Charge Carbon Limestone 0.12 Pig Iron 0.04 0.01 Steel Material kg C/GJ 12.1 Coke Oven Gas Blast Furnace Gas 70.8 Source: IPCC ( 2006 ) , Table 4.3. Coke Oven Gas and Blast Furnace Gas, Table 1.3. The production process for sinter result , which are emitted via leaks in the production in fugitive emissions of CH s 4 equipment, rather than through the emission stacks or vents of the production plants. The fugitive emissions were 2006 IPCC Guidelines (see Table calculated by applying Tier 1 emission factors taken from the for sinter production - 68 ). Although the 1995 IPCC Guidelines (IPCC/UNEP/OECD/IEA 1995) provide a Tier 1 CH 4 emission factor 4 for pig iron production, it is not appropriate to use because CO emissions were estimated using the Tier 2 mass 2 balance methodology. The mass balance methodo logy makes a basic assumption that all carbon that enters the pig - containing output or as CO emissions; the iron production process either exits the process as part of a carbon 2 estimation of CH emissions is precluded. A preliminary analysis of facility - l evel emissions reported during iron 4 production further supports this assumption and indicates that CH emissions are below 500 kt CO Eq. and well 4 2 below 0.05 percent of total national emissions. The production of direct reduced iron also results in emissio ns of CH through the consumption of fossil fuels (e.g., natural gas , etc. ); however, these emission estimates are excluded 4 due to data limitations. Pending further analysis and resources, EPA may include these emissions in future reports to enhance completeness. EPA is still assessing the possibility of including these emissions in future reports and have not included this data into the current report. Table 68 4 - : CH Emission Factors for Sinter and Pig Iron Production 4 Material Produced Unit Factor Sinter 0.07 kg CH /metric ton 4 ( 2006 ) , Table 4.2 . Source: IPCC and pellet consumption data Sinter consumption 5 were obtained from AISI’s Annual for 1990 through 201 Statistical Report (AISI 2004 through 201 6 a) and through personal communications with AISI ( AISI 2008) (see Table 4 - 69 ). In general, direct reduced iron (DRI) consumpti on data were obtained from the U.S. Geological Survey ) and personal communication with Iron and Steel Scrap 5 (USGS 1991 through 201 ( USGS) Minerals Yearbook – the USGS Iron and Steel Commodity Specialist (Fenton 201 However, data for DRI consumed in EAF ). 5 s were not available for the years 1990 and 1991. EAF DRI consumption in 1990 and 1991 was calculated by multiplying the - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 68 4

285 total DRI consumption for all furnaces by the EAF share of total DRI consumption in 1992. Also, data for DRI available for the years 1990 through 1993. BOF DRI consumption in 1990 through consumed in BOFs were not 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) i n 1994. , direct reduced iron production and pellet production were The Tier 1 CO emission factors for sinter production 2 obtained through the - series data for pig iron production, coke, natural 2006 IPCC Guidelines (IPCC 2006). Time gas, fuel oil, sinter, and pellets consumed in the blast furnace; pig iron production; and blast furnace gas produced at l activities were obtained from the iron and steel mill and used in the metallurgical coke ovens and other steel mil 6 a) and through personal communications with AISI ( AISI AISI’s Annual Statistical Report (AISI 2004 through 201 - 2008) (see and Table 4 - 70 ). 69 Table 4 Data for EAF steel production, flux, EAF charge carbon, and natural gas consumption were obtained from AISI’s (AISI 2004 through 201 6 a) and through personal communications with AISI ( AISI 2006 tatistical Report Annual S b and AISI through 201 The factor for the quantity of EAF anode consumed per ton of EAF steel produced 6 2008). AISI Data for BO F steel production, flux, natural gas, natural ore, pellet , sinter was provided by AISI ( 2008). (AISI 2004 consumption as well as BOF steel production were obtained from AISI’s Annual Statistical Report AISI through 201 Data for EAF and BOF scrap steel, a) and through personal communications with AISI ( 6 2008). Minerals Yearbook – Iron and Steel Scrap (USGS pig iron, and DRI consumption were obtained from the USGS . Data on coke oven gas and blast furnace gas consumed at the iron and steel mill (other than in 5) 1991 through 201 Annual Statistical Report (AISI 2004 through 201 6 a) e EAF, BOF, or blast furnace) were obtained from AISI’s th 2008). AISI and through personal communications with AISI ( Natural btained from EIA’s Data on blast furnace gas and coke oven gas sold for use as synthetic natural gas were o Gas Annual (EIA 201 6b) . Carbon contents for direct reduced iron, EAF carbon electrodes, EAF charge carbon, 2006 IPCC Guidelines contents for natural The carbon limestone, dolomite, pig iron, and steel were provided by the . gas, f EIA 2016c ) and EPA ( EPA 2010). Heat contents for uel oil, and direct injection coal were obtained from EIA ( were obtained from EIA ( 1992, 2011 ) ; natural gas heat content was obtained fuel oil and direct injection coal EIA Annual Sta (AISI 2004 through 201 6 a). Heat contents for coke oven gas and AISI’s tistical Report from Table 37 of AISI’s Annual Statistical Report (AISI 2004 through 201 6 a) and blast furnace gas were provided in Table 37 of confirmed by AISI staff (Carroll 2016). - 69 : Production and Consumption Data for the Calculation of CO Table 4 and CH Emissions 4 2 from Iron and Steel Production (Thousand Metric Tons) Source/Activity Data 1990 2005 2011 2012 2013 2014 2015 Sinter Production 5,795 Sinter Production 12,239 8,315 5,941 5,583 5,521 5,079 Direct Reduced Iron Production Direct Reduced Iron Production 516 1,303 1,582 3,530 3,350 4,790 4,790 Pellet Production 32,146 Pellet Production 50,096 39,288 40,622 38,198 37,538 60,563 Pig Iron Production Coke Consumption 24,946 13,832 11,962 9,571 9,308 11,136 7,969 Pig Iron Production 49,669 37,222 30,228 32,063 30,309 29,375 25,436 Direct Injection Coal 1,485 2,573 2,604 2,802 2,675 2,425 Consumption 2,275 EAF Steel Production EAF Anode and Charge Carbon Consumption 1,127 1,257 1,318 1,122 1,127 1,116 67 Scrap Steel 42,691 46,600 50,500 Consumption 50,900 47,3 00 48,873 48,873 Flux Consumption 319 695 726 748 771 771 726 33,511 52,194 52,108 52,415 52,641 55,174 49,451 EAF Steel Production BOF Steel Production Pig Iron Consumption 47,307 34,400 31,300 31,500 29, 600 23,755 23,755 Scrap Steel 7,89 5,917 0 5,917 Consumption 14,713 11,400 8,800 8,350 69 - 4 Industrial Processes and Product Use

286 Flux Consumption 576 582 454 476 454 454 454 BOF Steel Production 42,705 34,291 36,282 34,238 43,973 33,000 29,396 4 - 70 : Production and Consumption Data for the Calculation of CO Table Emissions from 2 3 illion ft Iron and Steel Production ( unless otherwise specified) M 1990 Source/Activity Data 2011 2012 2013 2014 2015 2005 Pig Iron Production Natural Gas 59,844 59,132 62,469 Consumption 56,273 48,812 47,734 43,294 Fuel Oil Consumption 16,170 21,378 (thousand gallons) 19,240 17,468 16,674 9,326 163,397 Coke Oven Gas 22,033 16,557 17,772 Consumption 18,608 17,710 16,896 13,921 Blast Furnace Gas 1,439,380 1,299,980 1,063,326 1,139,578 1,026,973 Production 1,000,536 874,670 EAF Steel Production Natural Gas 8,751 Consumption 15,905 9,622 19,985 6,263 11,145 10,514 BOF Steel Production Coke Oven Gas 524 554 Consumption 568 568 524 386 3,851 Other Activities Coke Oven Gas 224,883 97,132 90,718 94,596 Consumption 89,884 85,479 70,029 Blast Furnace Gas 1,414,778 1,295,520 1,059,473 1,135,227 Consumption 1,022,718 996,190 870,485 Uncerta inty and Time Series Consistency - CO emissions from metallurgical coke production are based on material production and The estimates of 2 Uncertainty is associated with the total U.S. coking coal consumption data and average carbon contents. consumption, total U.S. coke production and materials consumed du ring 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. re is uncertainty associated with the fact that coal tar and coke breeze production were estimated based on coke The Since merchant coke plant data is production because coal tar and coke breeze production data were not available. of other carbonaceous materials consumed at coke plants, the mass balance equation for not included in the estimate from metallurgical coke production cannot be reasonably completed. Therefore, for the purpose of this analysis, CO 2 uncertainty parameters are applied to primary data in puts to the calculation (i.e., coking coal consumption and metallurgical coke production) only. The estimates of CO emissions from iron and steel production are based on material production and con sumption 2 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 There is uncertainty associated wi th the assumption that all coal used for associated IPCC default emission factors. 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 time series. Also for EAF steel production, there is uncertainty associated with the assumption that throughout the 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 eration); therefore, all consumption is attributed steel mill versus for energy generation (i.e., electricity and steam gen These data and carbon contents produce a relatively accurate estimate of CO to iron and steel production. 2 emissions. However, there are uncertainties associated with each. - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 70 4

287 The results of the Approach 2 q uantitative uncertainty analysis are summarized i Table 4 - 71 for metallurgical coke n emissions from metallurgical coke production and iron and steel Total CO production and iron and steel production. 2 MMT CO production for 2015 were estimated to be between 57.1 40.8 Eq. at the 95 percent confidence level. and 2 This indicates a range of approximately 17 percent below and 17 percent above th e emission estimate of 48.9 MMT CO Eq. Total CH emissions from metallurgical coke production and iron and steel production for 2015 were 4 2 estimated to be between 0.00 7 and 0.01 MMT CO This indicates a range of Eq. at the 95 percent confidence level. 2 appr oximately 19 percent below and 19 percent above the emission estimate of 0.009 MMT CO Eq. 2 - Table 4 71 : Approach 2 Quantitative Uncertainty Estimates for CO Emissions from and CH 4 2 d Metallurgical Coke Production (MMT CO Eq. and Percent) Iron and Steel Production an 2 a 2015 Emission Estimate Uncertainty Range Relative to Emission Estimate Gas Source (MMT CO Eq.) (MMT CO Eq.) (%) 2 2 Lower Upper Upper Lower Bound Bound Bound Bound & Iron Metallurgical Coke % +1 % CO 7 48.9 40.8 57.1 - 1 7 2 and Steel Production Metallurgical Coke & Iron CH + + + - 19% +19% 4 and Steel Production Eq. + Does not exceed 0.05 MMT CO 2 a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. approaches were applied to the entire time series to ensure consistency in emissions from 1990 Methodological 5 through 201 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 , see QA/QC a nd Verification Procedures section in the introduction of the 2006 IPCC Guidelines IPPU Chapter. Recalculations Discussion Updated data was obtained for 2014 direct reduced iron production (USGS 2015), 2014 process inputs for metallurgical coke production, outputs of U.S. meta llurgical coke production and direct reduced iron consumption (EIA 2016a). These revisions resulted in an increase of 2014 for BOF s teel p roduction CO emissions estimates 2 from metallurgical coke p and 2014 CO by 4 percent e missions estimates from i ron and st roduction roduction eel p 2 each compared to the previous inventory report. This year’s inventory now incorporates pellet production emission estimates as part of Iron and Steel production. The entire time ude this data to improve completeness and ensure series from 1990 to 2015 was recalculated to incl time - series consistency. With the inclusion of pellet production emissions estimates, total iron and steel production emissions increased between ranges of 0.79 and 2.03 MMT CO ntory (i.e., 1990 compared to the previous Inve 2 through 2014). Planned Improvements Future improvements involve improving activity data and emission factor sources for estimating CO and CH 4 2 emissions from pellet production. EPA will also evaluat e and analyz e data reported under EPA’s GHGRP to ies improve the emission estimates for th Iron and Steel Production process categor is and other . Particular attention will be made to ensure time - series consistency of the emissions estimates presented in future Inventory reports, level reporting data from EPA’s This is required as the facility - consistent with IPCC and UNFCCC guidelines. GHGRP, with the program's initial requirements for reporting of emissions in calendar year 2010, are not available ) as required for this Inventory. for all inventory years (i.e., 1990 through 2009 In implementing improvements and 71 - 4 Industrial Processes and Product Use

288 integration of data from EPA’s GHGRP, the latest guidance from t - he IPCC on the use of facility level data in 232 national inventories will be relied upon. Additional improvements include accou nting for emission estimates for the production of metallurgical coke to the Energy chapter as well as identifying the amount of carbonaceous materials, other than coking coal, consumed at ing the amount of coal used for direct injection Other potential improvements include identify merchant coke plants. and the amount of coke breeze, coal tar, and light oil produced during coke production. Efforts will also be made to identify information to better characterize emissions from the use of process gases and fu els within the Energy and This planned improvement is still in development and is not included Industrial Processes and Product Use chapters. in this current inventory report. 4.17 Ferroalloy Production (IPCC Source Category 2C2) ) ) and metha Carbon dioxide (CO are emitted from the production of several ferroalloys. Ferroalloys are ne (CH 2 4 composites of iron (Fe) and other elements such as silicon (Si), manganese (Mn), and chromium (Cr). Emissions ferroalloys from fuels consumed for energy purposes during the production of are accounted for in the Energy chapter. Emissions from the production of two types of ferrosilicon (25 to 55 percent and 56 to 95 percent silicon), silicon metal (96 to 99 percent silicon), and miscellaneous alloys (32 to 65 percent silicon) h ave been calculated. Emissions from the production of ferrochromium and ferromanganese are not included here because of the small number of manufacturers of these materials in the United States, and therefore, government information disclosure t the publication of production data for these production facilities. rules preven is emitted when metallurgical coke is oxidized Similar to emissions from the production of iron and steel, CO 2 temperature reaction with iron and the selected alloying el ement. Due to the strong reducing during a high - . A representative reaction equation for the environment, CO is initially produced, and eventually oxidized to CO 2 production of 50 percent ferrosilicon (FeSi) is given below: Fe + 2Si O O + 7C → 2FeSi + 7CO 2 2 3 While most of the carbon contained in the process materials is released to the atmosphere as CO , a percentage is 2 also released as CH and other volatiles. The amount of CH that is released is dependent on furnace efficiency, 4 4 ogy. operation technique, and control technol When incorporated in alloy steels, ferroalloys are used to alter the material properties of the steel. Ferroalloys are used primarily by the iron and steel industry, and production trends closely follow that of the iron and steel industry. Twelve com panies in the United States produce ferroalloys (USGS 2016a). and from ferroalloy production in 201 5 were 2.0 MMT CO Eq. (1,96 Emissions of CO kt CO 72 ) (see Table 4 - 0 2 2 2 Emissions of CH 4 73 ), which is a 9 percent reduction since 1990 Table - were from ferroalloy production in 201 5 . 4 0.01 MMT CO Eq. (0.5 kt CH ), which is a 19 percent decrease since 1990. 4 2 Table 4 72 : CO - and CH Emissions from Ferroalloy Production ( MMT CO Eq.) 2 2 4 Gas 1990 2005 2011 2012 2013 2015 2014 2.2 1.4 CO 1.9 1.8 1.9 2.0 1.7 2 CH + + + + + + + 4 2.2 1.4 1.7 Total 1.9 1.8 1.9 2.0 + Does not exceed 0.05 MMT CO Eq. 2 Table 4 - 73 : CO and CH ) Emissions from Ferroalloy Production ( kt 4 2 232 See . > al_Bulletin_1.pdf nggip.iges.or.jp/public/tb/TFI_Technic - http://www.ipcc < - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 72 4

289 1990 2005 2011 2012 2013 2014 2015 Gas 1,392 1,735 1,903 1,785 1,914 1,960 CO 2,152 2 1 1 CH + + 1 + 1 4 + Does not exceed 0.5 kt. Methodology 233 and CH Emissions of CO from ferroalloy production were calculated 2006 using a Tier 1 method from the 4 2 IPCC Guidelines - specific default emission factors provided by multiplying annual ferroalloy production by material emissions are as follows: and CH by IPCC (IPCC 2006). The Tier 1 equations for CO 2 4 ) ( =∑ ×퐸퐹 푀푃 퐸 푖 푖 퐶푂2 푖 where, E = CO emissions, metric tons 2 CO2 MP = Production of ferroalloy type i , metric tons i = Generic emission factor for ferroalloy type , metric tons CO EF /metric ton specific i 2 i ferroalloy product ( ) =∑ ×퐸퐹 퐸 푀푃 푖 퐶퐻4 푖 푖 where, E = CH emissions, kg H4 C 4 i MP = Production of ferroalloy type , metric tons i Generic emission factor for ferroalloy type i , kg CH /metric ton specific ferroalloy EF = i 4 product specific emission factors are not currently available. The Default emission factors were used because country - used to develop annual CO following emission factors were estimates: and CH 2 4 Ferrosilicon, 25 to 55 percent Si and Miscellaneous Alloys, 32 to 65 percent Si – 2.5 metric tons • /metric ton of alloy produced. CO /metric ton of alloy produced; 1.0 kg CH 4 2 /metric ton alloy produced; 1.0 kg CH – • 4.0 metric tons CO /metric ton of Ferrosilicon, 56 to 95 percent Si 4 2 alloy produced. • Silicon Metal – 5.0 metric tons CO /metric ton metal produced; 1.2 kg CH /metric ton metal produced. 4 2 oduced using petroleum coke in an electric arc It was assumed that 100 percent of the ferroalloy production was pr furnace process (IPCC 2006), although some ferroalloys may have been produced with coking coal, wood, other biomass, or graphite carbon inputs. The amount of petroleum coke consumed in ferroalloy production wa s calculated assuming that the petroleum coke used is 90 percent carbon (C) and 10 percent inert material ( Onder and Bagdoyan 1993) . Ferroalloy production data for 1990 through 201 (see Table 4 - 74 ) were obtained from the U.S. Geological Survey 5 ( USGS ) through the Minerals Yearbook: Silicon (USGS 199 6 through 201 3 ) and the Mineral Industry Surveys: Silicon (USGS 2014, 2015b, 2016b). The following data were available from the USGS publications for the time series: • Ferrosilicon, 25 to 55 percent Si: Annual production data were available from 1990 through 2010. • Ferrosilicon, 56 to 95 percent Si: Annual production data were available from 1990 through 2010. 233 EPA has not integrated aggregated facility - level GHGRP information to inform these estimates. The aggregated information derlying (e.g. , activity data and emissions) associated with production of ferroalloys did not meet criteria to shield un confidential from public disclosure. ) CBI business information ( 73 - 4 Industrial Processes and Product Use

290 • 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, USG S 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 produ ction 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. uction data (i.e., ferroalloy product production/total Ferroalloy product shares developed from the 2010 prod ferroalloy production) were used with the total silicon materials production quantity to estimate the production quantity by ferroalloy product type for 2011 through 2015 (USGS 2013, 2014, 2015b, 2016b ). 4 - 74 : Production of Ferroalloys (Metric Tons) Table Ferrosilicon Year Silicon Metal Misc. Alloys Ferrosilicon 25% - 95% - 55% 32 - 65% 56% 1990 321,385 109,566 145,744 72,442 NA 86,100 2005 123,000 148,000 2011 159,667 140,883 154,450 NA 2012 175,108 154,507 169,385 NA 2013 164,229 144,908 158,862 NA 2014 NA 176,161 155,436 170,404 2015 159,151 174,477 NA 180,372 - - NA regated along with ferrosilicon ( 25 55 % Si) Not Available for product type, agg Uncerta inty and Time - Series Consistency Annual ferroalloy production was reported by the USGS in three broad categories the 2010 publication: until 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 t he 2011 Minerals Yearbook , USGS started reporting all the ferroalloy production under a single category: t otal 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. E missions 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), information and data regarding these practi ces were not available. Emissions however - based from ferroalloys produced with wood or other biomass would not be counted under this source because wood 234 Even though emissions from ferroalloys produced with coking coal or graphite inp uts carbon is of biogenic origin. would be counted in national trends, they may be generated with varying amounts of CO per unit of ferroalloy 2 produced. The most accurate method for these estimates would be to base calculations on the amount of reducing agent used in the process, These data, however, were not available, rather than the amount of ferroalloys produced. and are also often considered confidential business information. Emissions of CH peration from ferroalloy production will vary depending on furnace specifics, such as type, o 4 Higher heating temperatures and techniques such as sprinkle charging will technique, and control technology. reduce CH emission emissions; however, specific furnace information was not available or included in the CH 4 4 estimates. 234 . Use Change, and Forestry chapter - in the Land Use, Land Emissions and sinks of biogenic carbon are accounted for - 5 201 – Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 74 4

291 The results of t Table - 75 . Ferroalloy production he Approach 2 quantitative uncertainty analysis are summarized in 4 and 2. en 1. 8 CO 2 MMT CO emissions from 2015 were estimated to be betwe Eq. at the 95 percent confidence 2 2 level. This indicates a range of approximately 12 percent below and 12 percent above the emission estimate of 2.0 Ferroalloy production CH Eq. MMT CO f approximately 12 emissions were estimated to be between a range o 4 2 Eq. percent below and 12 percent above the emission estimate of 0.01 MMT CO 2 4 - 75 : Approach 2 Quantitative Uncertainty Estimates for CO Table Emissions from 2 Ferroalloy Production (MMT CO Eq. and Percent) 2 a Uncertainty Range Relative to Emission Estimate 2015 Emission Estimate Source Gas (MMT CO Eq.) Eq.) (MMT CO (%) 2 2 Lower Upper Lower Upper Bound Bound Bound Bound CO 2. 2.0 1. 8 Ferroalloy Production 2 - 12% +12% 2 + CH + + Ferroalloy Production - 12% +12% 4 + Does not exceed 0.05 MMT CO Eq. 2 a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval. approaches were applied to the entire time series to ensure consistency in emissions from 1990 Methodological . through 201 5 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, 2006 IPCC Guidelines , see Chapter 6 of the section in the introduction of the QA/QC and Verification Procedures IPPU Chapter. Planned Improvements f improvements for more significant sources, EPA will continue to Pending available resources and prioritization o e and analyz e evaluat 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, p articular attention will be made to ensure time - series consistency of the emissions estimates presented in future Inventory reports, consistent with IPCC and UNFCCC guidelines. - level This is required as the facility 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 235 level data in national inventories will be relied upon. EPA is still assessing the possibility of use of facility - tory report and has not included these data sets into incorporating this planned improvement into the national inven the current inventory report. 4.18 Aluminum Production (IPCC Source Category 2C3 ) Aluminum is a light - weight, malleable, and corrosion - resistant metal that is used in many manufactured products, including ai . A s of recent reporting, the United States was the rcraft, automobiles, bicycles, and kitchen utensils eighth largest producer of primary aluminum, with approximately 3 percent of the world total production ( USGS 2016 ). The United States was also a major i mporter of primary aluminum. The production of primary aluminum — in 235 - >. etin_1.pdf http://www.ipcc nggip.iges.or.jp/public/tb/TFI_Technical_Bull See < 75 - 4 Industrial Processes and Product Use

292 addition to consuming large quantities of electricity — related emissions of carbon dioxide (CO results in process ) - 2 F ). ) and perfluoroethan and two perfluorocarbons (PFCs): perfluoromethane (CF e (C 6 2 4 ) is reduced Carbon dioxide is emitted during the aluminum smelting process when alumina (aluminum oxide, Al O 3 2 to aluminum using the Hall Heroult reduction process. The reduction of the alumina occurs through electrolysis in a - ral or synthetic cryolite (Na molten bath of natu ). The reduction cells contain a carbon (C) lining that serves as AlF 6 3 the cathode. Carbon is also contained in the anode, which can be a C mass of paste, coke briquettes, or prebaked C ion, most of this C is oxidized and released to the atmosphere as CO blocks from petroleum coke. During reduct . 2 from aluminum production were estimated to be Eq. (2,767 kt) in 2015 Process emissions of CO 2.8 MMT CO 2 2 4 - 76 ). (see The C anodes consumed during aluminum production consist of petroleum coke and, to a minor Table process emissions from aluminum production is extent, coal tar pitch. The petroleum coke portion of the total CO 2 energy use of petroleum coke, and is accounted for here and not under the CO considered to be a non from Fossil - 2 hese CO process Fuel Combustion source category of the Energy sector. Similarly, the coal tar pitch portion of t 2 emissions is accounted for here. - 76 Table 4 Emissions from Aluminum Production ( MMT CO Eq