United States
Environmental Protection
Agency
Air and Radiation
(6202 J)
EPA 430-R-93-003
April 1993
xvEPA
Anthropogenic Methane
Emissions in the United States:
Estimates for 1990
Report to Congress
Natural Gas Systems
PROPERTY OF
DIVISION
OF
nrv
Manure Management
-------
ANTHROPOGENIC METHANE
EMISSIONS IN THE UNITED STATES
Estimates for 1990
REPORT TO CONGRESS
Editor: Kathleen B. Hogan
U.S. Environmental Protection Agency
Office of Air and Radiation
April 1993
RBcyctod/RecyctaMe
Printed on pspsr that contains
at toast 50% racyctod fiber
-------
This document has been reviewed in accordance with the U.S. Environmental
Protection Agency's and the Office of Management and Budget's peer and
administrative review policies and approved for publication. Mention of trade
names or commercial products does not constitute endorsement or
recommendation for use.
-------
Table of Contents
FOREWORD ix
ACKNOWLEDGEMENTS x
EXECUTIVE SUMMARY ES-1
CURRENT EMISSIONS ES-1
FUTURE EMISSIONS ES-9
1. INTRODUCTION 1-1
1.1 BACKGROUND: The Importance of Methane 1-1
1.1.1 What is Methane? 1-2
1.1.2 Atmospheric Levels of Methane Are Rising 1-3
1.1.3 Methane and Global Climate Change 1-5
1.1.4 Stabilization of Global Methane Levels 1-7
1.2 OVERVIEW OF METHANE SOURCES 1-8
1.2.1 Anthropogenic Sources 1-8
1.2.2 Natural Sources 1-11
1.3 OVERVIEW OF REPORT 1-11
1.4 REFERENCES 1-12
2. METHANE EMISSIONS FROM THE NATURAL GAS SYSTEM 2-1
2.1 EMISSIONS SUMMARY 2-1
2.2 BACKGROUND 2-4
2.2.1 Stages of the Natural Gas System 2-4
2.2.2 Sources of Methane Emissions in the Natural Gas System 2-8
2.3 METHODOLOGY 2-10
2.3.1 Background 2-10
2.3.2 Steps Used to Estimate Emissions 2-11
2.4 CURRENT EMISSIONS 2-16
2.4.1 Field Production Facilities 2-16
2.4.2 Gas Processing Plants 2-23
2.4.3 Storage and Injection/Withdrawal Facilities 2-25
2.4.4 Transmission Facilities 2-28
2.4.5 Distribution Network 2-31
2.4.6 Compressor Engine Exhaust 2-33
2.4.7 Summary of Total Emissions from the Natural Gas System 2-36
2.4.8 Comparison with Previous Estimates 2-38
2.4.9 Uncertainties 2-38
2.5 FUTURE EMISSIONS 2-42
2.5.1 Current Operating Practices Scenario 2-42
2.5.2 Improved Technology and Operating Practices 2-48
2.6 LIMITATIONS OF THE ANALYSIS 2-49
2.7 REFERENCES 2-50
-------
Table of Contents (Continued)
3. METHANE EMISSIONS FROM COAL MINING 3-1
3.1 EMISSIONS SUMMARY 3-1
3.2 BACKGROUND 3-4
3.2.1 How Coalbed Methane Is Produced, Stored and Released 3-4
3.2.2 U.S. Mining Techniques 3-5
3.2.3 Methane Management Systems for Underground Mining 3-7
3.2.4 Post-Mining Emissions 3-8
3.3 METHODOLOGY 3-8
3.3.1 Emissions from Underground Mines -1988 3-9
3.3.2 Emissions from Surface Mines -1988 3-12
3.3.3 Post-Mining Emissions 3-14
3.3.4 Projections of Future Methane Emissions 3-14
3.4 CURRENT EMISSIONS 3-16
3.4.1 Overview 3-16
3.4.2 Basin-Specific Emissions 3-18
3.5 FUTURE EMISSIONS 3-22
3.5.1 Overview 3-22
3.5.2 Basin-Specific Estimates 3-24
3.6 LIMITATIONS OF THE ANALYSIS 3-27
3.6.1 Uncertainties in 1988 Emission Estimates 3-27
3.6.2 Uncertainties in Future Emission Estimates 3-28
3.7 REFERENCES 3-29
4. METHANE EMISSIONS FROM LANDFILLS 4-1
4.1 EMISSIONS SUMMARY 4-1
4.2 BACKGROUND 4-3
4.2.1 Landfill Refuse Management 4-3
4.2.2 Landfill Methane Production 4-4
4.2.3 Site-Specific Factors Affecting Methane Production 4-6
4.3 METHODOLOGY 4-8
4.3.1 Background 4-9
4.3.2 Steps Used to Estimate Emissions from Landfills 4-10
4.4 CURRENT EMISSIONS 4-21
4.5 FUTURE EMISSIONS 4-21
4.5.1 Background 4 23
4.5.2 Methodology 4 26
4.5.3 Future Emissions with Current Recovery Practices 4-31
4.5.4 Future Emissions with the Landfill Rule 4-31
4.5.5 Opportunities for Emission Reductions 4-33
4.6 LIMITATIONS OF THE ANALYSIS 4-33
4.7 REFERENCES 4-34
5. METHANE EMISSIONS FROM DOMESTICATED LIVESTOCK 5-1
5.1 EMISSIONS SUMMARY 5-1
5.2 BACKGROUND 5-2
5.3 METHODOLOGY 5-4
5.3.1 The Models Used to Evaluate Cattle 5-5
-------
Table of Contents (Continued)
5.3.2 Model Evaluation 5-6
5.3.3 Application of the Model 5-9
5.3.4 Methane Emissions from Other Animals 5-25
5.4 CURRENT EMISSIONS 5-26
5.4.1 U.S. Cattle Population 5-26
5.4.2 Cattle 5-27
5.4.3 Comparisons with Previous Cattle Emissions Estimates 5-28
5.4.4 Emissions from Other Animals 5-30
5.4.5 Uncertainties 5-30
5.5 FUTURE EMISSIONS 5-34
5.5.1 Future Emissions - "Current Practices" Scenario 5-34
5.5.2 Opportunities for Emission Reductions 5-41
5.6 LIMITATIONS OF THE ANALYSIS 5-43
5.7 REFERENCES 5-43
6. METHANE EMISSIONS FROM LIVESTOCK MANURE 6-1
6.1 EMISSIONS SUMMARY 6-1
6.2 BACKGROUND 6-2
6.2.1 The Fundamentals of Anaerobic Decomposition 6-2
6.2.2 Methane Producing Capacity of Livestock Manure 6-3
6.2.3 Factors Influencing Methane Production 6-4
6.3 METHODOLOGY 6-5
6.4 CURRENT EMISSIONS 6-16
6.4.1 Point Estimates 6-16
6.4.2 Range of Estimates 6-18
6.4.3 Comparison with Previous Estimates 6-19
6.5 FUTURE EMISSIONS 6-21
6.5.1 Extend Current Practices 6-24
6.5.2 Increased Use of Liquid Systems 6-24
6.5.3 Opportunities for Emission Reductions 6-26
6.6 LIMITATIONS OF THE ANALYSIS 6-28
6.7 REFERENCES 6-28
7. METHANE EMISSIONS FROM OTHER SOURCES 7-1
7.1 EMISSIONS SUMMARY 7-1
7.2 RICE CULTIVATION ' 7-2
7.2.1 Background 7-2
7.2.2 Methodology 7-3
7.2.3 Current Emissions 7-3
7.2.4 Future Emissions 7-4
7.2.5 Limitations of the Analysis 7-4
7.3 FUEL COMBUSTION 7-4
7.3.1 Background 7-4
7.3.2 Methodology 7-5
7.3.3 Current Emissions 7-7
7.3.4 Future Emissions 7-8
7.3.5 Limitations of the Analysis 7-9
Mi
-------
Table of Contents (Continued)
7.4 PRODUCTION AND REFINING OF PETROLEUM LIQUIDS 7-11
7.4.1 Background . 7-11
7.4.2 Methodology 7-11
7.4.3 Current Emissions 7-14
7.4.4 Future Emissions 7-15
7.4.5 Limitations of the Analysis 7-15
7.5 ADDITIONAL SOURCES OF METHANE EMISSIONS 7-15
7.5.1 Non-Fuel Biomass Burning 7-15
7.5.2 Industrial Processes and Wastes 7-16
7.5.3 Land-Use Changes 7-17
7.6 REFERENCES 7-18
iv
-------
List of Exhibits
Exhibit ES-1: Global Contribution to Radiative Forcing by Gas for 1990 ES-2
Exhibit ES-2: Contribution of Major Methane Sources to Total U.S.
Anthropogenic Emissions ES-6
Exhibit ES-3: U.S. Anthropogenic Emissions Summary ES-7
Exhibit ES-4: U.S. Contribution to Global Methane Emissions by Source ES-8
Exhibit ES-5: Estimates of Future Methane Emissions in the United States ES-10
Exhibit 1 -1: Estimated Sources and Sinks of Methane 1-4
Exhibit 1 -2: Measurements of Global Methane Concentrations 1-5
Exhibit 1-3: Global Contribution to Radiative Forcing by Gas for 1990 CO2-
Equivalent Basis Using IPCC 1991 GWPs for a 100-Year Time Horizon ... 1-6
Exhibit 1 -4: CO2 and Methane Reduction Comparison 1-8
Exhibit 2-1: The U.S. Natural Gas System 2-5
Exhibit 2-2: 1990 Natural Gas Production in the U.S. 2-7
Exhibit 2-3: Throughput at U.S. Gas Processing Plants 2-7
Exhibit 2-4: Emission Rates and Factors from Model Production Facilities 2-18
Exhibit 2-5: Emissions Rates and Factors from Model Gathering and Transmission
Systems 2-19
Exhibit 2-6: Emission Rates and Factors for Gathering Pipeline 2-20
Exhibit 2-7: Summary of Total Emissions from Production Facilities 2-22
Exhibit 2-8: Emission Rates and Factors for Model Processing Plants 2-24
Exhibit 2-9: Summary of Total Emissions From Processing Plants 2-25
Exhibit 2-10: Emission Rates and Factors from Model Injection/Withdrawal Plants .... 2-27
Exhibit 2-11: Summary of Total Emissions from Storage and Injection/Withdrawal
Plants in the U.S 2-28
Exhibit 2-12: Summary of Total Emissions from the Transmission System in the U.S. . . 2-30
Exhibit 2-13: Emission Rates and Factors for Distribution System Fugitive Emissions . . 2-32
Exhibit 2-14: Summary of Total Emissions from the Distribution System in the U.S. . . . 2-34
Exhibit 2-15: Summary of Total Emissions from Engine Exhaust in the U.S 2-36
Exhibit 2-16: Methane Emissions From the U.S. Natural Gas System (Tg/yr) 2-37
Exhibit 2-17: Comparison with Previous Emissions Estimates 2-38
Exhibit 2-18: Future U.S. Natural Gas Consumption Outlook (Tcf/yr) 2-44
Exhibit 2-19: Future Size of the U.S. Natural Gas System 2-47
Exhibit 2-20: Summary of Future Methane Emissions From the U.S. Natural Gas
System 2-48
Exhibit 3-1: Methane Emissions from U.S. Coal Mines 3-2
Exhibit 3-2: Stages in Coalification 3-4
Exhibit 3-3: Major U.S. Coal Basins and Coalbed Methane Resources 3-6
Exhibit 3-4: Diagram of Mine Degasification Approaches 3-9
Exhibit 3-5: Mine Degasification Approaches 3-10
Exhibit 3-6: Assumed Degasification Recovery Efficiencies 3-12
Exhibit 3-7: 1988 Estimated Degasification System Emissions 3-13
-------
List of Exhibits (Continued)
Exhibit 3-8: Average Methane Contents of Underground and Surface Coal, by Coal
Basin or State 3-15
Exhibit 3-9: Coal Production Forecasts 3-16
Exhibit 3-10: U.S. Methane Emissions, 1988 3-17
Exhibit 3-11: 1988 Emissions Summary 3-17
Exhibit 3-12: 1988 Coal Emissions 3-18
Exhibit 3-13: Comparison with Other Recent Emissions Estimates 3-18
Exhibit 3-14: 1988 Methane Emissions by Coal Basin (in Tg) 3-19
Exhibit 3-15: Coal Characteristics - Northern Appalachian Basin 3-19
Exhibit 3-16: Coal Characteristics - Central Appalachian Basin 3-20
Exhibit 3-17: Coal Characteristics - Black Warrior Basin 3-21
Exhibit 3-18: Coal Characteristics - Illinois Basin 3-;22
Exhibit 3-19: Coal Characteristics - Rockies and Southwest Basins . 3-;22
Exhibit 3-20: Emissions Growth Forecast 3-23
Exhibit 3-21: Degasification System Emissions 3-24
Exhibit 3-22: Methane Emissions from U.S. Coal Mining in 2000 and 2010 (Tg) 3-25
Exhibit 4-1: Materials Discarded in Municipal Solid Waste Landfills in 1990 (weight
basis) 4-5
Exhibit 4-2: Model of Relative Gas Composition in a Landfill 4-7
Exhibit 4-3: Distribution of Landfills by Total Waste in Place 4-13
Exhibit 4-4: Observed and Predicted Methane Emissions 4-16
Exhibit 4-5: Estimated Waste Landfilled Between 1960 and 1990 4-17
Exhibit 4-6: Landfill Size Distribution by Waste in Place 4-18
Exhibit 4-7: Waste Quantities Disposed in Industrial Landfills in 1985 4-19
Exhibit 4-8: National Emission Estimates for 1990 4-22
Exhibit 4-9: Comparison to Other Estimates of U.S. Methane Emissions from
Landfills 4-23
Exhibit 4-10: State Source Reduction Efforts (1991) 4-24
Exhibit 4-11: Future Landfill Waste Disposal Scenarios (106 Mg/Yr) 4-27
Exhibit 4-12: Projected Landfill Composition for 2000 4-28
Exhibit 4-13: Projected Landfill Waste Producing Methane in 2000 and 2010 4-30
Exhibit 4-14: National Emission Estimates for 2000 and 2010 with Current Recovery
Practices 4-32
Exhibit 5-1: Schematic of Ruminant and Monogastric Digestive Systems 5-4
Exhibit 5-2: Observed and Predicted Estimates of Metabolizable Energy (ME) ....... 5-7
Exhibit 5-3: Observed and Predicted Estimates of Methane Per Kilogram of Feed .... 5-8
Exhibit 5-4: Observed and Predicted Estimates of Methane Per ME 5-8
Exhibit 5-5: Comparisons with Moe and Tyrrell Equation Estimates 5-10
Exhibit 5-6: Representative Animal Characteristics: Heifers and Cattle Fed
for Slaughter 5-12
Exhibit 5-7: Representative Animal Characteristics: Dairy Cows and Beef Cows 5-12
Exhibit 5-8: Geographic Regions Used in the Analysis £.-13
Exhibit 5-9: Dairy Cow Diet Descriptions 5-16
Exhibit 5-10: Regional Estimates of Methane Emissions from Mature Dairy Cows Ei-17
vi
-------
List of Exhibits (Continued)
Exhibit 5-11: Regional Estimates of Emissions from Dairy Replacement Heifers:
0-12 Months 5-18
Exhibit 5-12: Regional Estimates of Emissions from Dairy Replacement Heifers:
12-24 Months 5-19
Exhibit 5-13: Regional Estimates of Methane Emissions from Beef Cows 5-20
Exhibit 5-14: Regional Estimates of Emissions from Beef Replacements: 0-12 Months . 5-21
Exhibit 5-15: Regional Estimates of Emissions from Beef Replacement Heifers:
12-24 Months 5-22
Exhibit 5-16: Regional Estimates of Emissions from Feedlot Fed Cattle:
Yearling System 5-23
Exhibit 5-17: Regional Estimates of Emissions from Feedlot Fed Cattle:
Weanling System 5-24
Exhibit 5-18: Methane Emissions Factors for Cattle by Region 5-25
Exhibit 5-19: Emissions Factors Used for Other Animals 5-26
Exhibit 5-20: Methane Emissions From U.S. Dairy Cattle 5-28
Exhibit 5-21: Methane Emissions From U.S. Beef Cattle 5-29
Exhibit 5-22: Comparison of Detailed Estimates by Johnson et al 5-31
Exhibit 5-23: Comparison of National Estimates by Johnson et al 5-32
Exhibit 5-24: Comparison of National Estimates by Byers 5-32
Exhibit 5-25: Methane Emissions from Other Animals 5-33
Exhibit 5-26: U.S. Per Capita Meat Consumption: 1975-1989 5-36
Exhibit 5-27: U.S. Meat Production and Trade: 1975-1989 5-37
Exhibit 5-28: Domestic Milk Production and Consumption: 1975-1989 5-38
Exhibit 5-29: Future U.S. Meat and Milk Production and Consumption Based on the
FAPRI Study 5-39
Exhibit 5-30: Future U.S. Meat and Milk Production and Consumption Based on the
EPA Study 5-40
Exhibit 5-31: Future U.S. Meat and Milk Production and Consumption Based on the
FAPRI and EPA Studies 5-40
Exhibit 5-32: Scenarios of Future Emissions 5-42
Exhibit 6-1: U.S. Animal Populations, Average Size, and VS Production 6-8
Exhibit 6-2: Maximum Methane Producing Capacity for U.S. Livestock Manure 6-9
Exhibit 6-3: Maximum Methane Producing Capacity Adopted For U.S. Estimates 6-9
Exhibit 6-4: Methane Conversion Factors for U.S. Livestock Manure Systems 6-13
Exhibit 6-5: Methane Conversion Factors for U.S. Livestock Manure Systems 6-14
Exhibit 6-6: Regions of the U.S. for Manure Management Characterization 6-15
Exhibit 6-7: Livestock Manure System Usage for the U.S 6-15
Exhibit 6-8: Methane Emissions by Animal Type (1990) 6-17
Exhibit 6-9: Methane Emissions by Manure Management System (1990) 6-17
Exhibit 6-10: Base, High, and Low Case Emission Estimate Assumptions 6-19
Exhibit 6-11: Base, High, and Low Case Emission Estimates for 1990 (Tg/Yr) 6-20
Exhibit 6-12: Comparison with Detailed Estimates by Lodman et al 6-22
Exhibit 6-13: Future U.S. Meat and Mild Production and Consumption Based on
FAPRI Study 6-23
Exhibit 6-14: Future U.S. Meat and Milk Production and Consumption Based on EPA
Study 6-23
Exhibit 6-15: Projected Range of Emissions for 2000 and 2010 (Tg/Yr) 6-25
vii
-------
List of Exhibits (Continued)
Exhibit 7-1: Data Used to Calculate Three-Year Average Hectare-Days 7-4
Exhibit 7-2: Quantity of Non-Methane VOCs and Ratio of Methane to Non-Methane
VOCs 7-I5
Exhibit 7-3: Data Used to Estimate Methane Emissions from Natural Gas
Consumption 7-7
Exhibit 7-4: Data Used to Estimate Emissions from Mobile Sources 7-8
Exhibit 7-5: Annual Methane Emissions from Stationary Combustion (Teragrams) .... 7-9
Exhibit 7-6: Annual Methane Emissions from Mobile Combustion (Teragrams) 7-10
viii
-------
FOREWORD
I am pleased to transmit the attached report, Anthropogenic Methane Emissions in the
United States. Estimates for 1990. the first of five methane-related reports prepared in
response to the Congressional mandate in the Clean Air Act Amendments of 1990. This
report provides estimates of methane emissions from the major sources in the United States.
Many researchers and policy-makers have found evaluation of methane emissions to
be frustrating due to the great uncertainties in the methane emissions from various sources.
These uncertainties arise from variability inherent in the nature of the processes and activities
that produce methane. This report should help remove much of this frustration. It presents
comprehensive and detailed methods that strongly reflect the underlying physical
characteristics for each of the major methane sources, couples the methods with recently
available information, and produces estimates that reduce some of the uncertainty and
characterize much of the remaining uncertainty.
This report is a large step forward in the methane area. It is not only a highly
informative volume summarizing the current state of knowledge of methane emissions in the
United States, but it also will make a major contribution to the international community as
other countries proceed to refine their estimates of methane emissions.
Quantifying the magnitude of methane emissions is an important preliminary step to
identifying opportunities for reducing these emissions. This report helps us to better identify
where methane emissions can be reduced, particularly where they can be reduced while
yielding benefits to energy and agricultural industries, as well as to the environment.
Paul M. Stolpman
Acting Director
Office of Atmospheric Programs
Office of Air and Radiation
be
-------
ACKNOWLEDGEMENTS
This report would not have been possible without the intensive and tireless efforts of a
number of people who contributed throughout the process of developing the report.
First, the lead authors of the main chapters with much thought and discussion
developed new and detailed methods for estimating U.S. emissions from the major methane
sources. These new methods incorporate large amounts of data on methane emissions,
much of which is only recently available, and help refine our understanding of the parts of
each system which contribute the majority of emissions. The lead authors are recognized as
follows:
Natural Gas Systems Michael Gibbs
Coal Mining Dina Kruger
Landfills Kathleen Hogan
Jonathan Woodbury
Domesticated Livestock Michael Gibbs
Mark Orlic
Livestock Manure Jonathan Woodbury
Kurt Roos
For some chapters, significant research was performed to assist in the development of
emissions estimates. This includes efforts of Lee Baldwin (University of California Davis) for
domesticated livestock and Andy Hashimoto (University of Oregon) and L.M. "Mac" Safley
(North Carolina State University) for livestock manure. Substantial analytical work was
performed in support of other chapters. This includes efforts by Tom Cantine, Mary
DePasquale, Michael Gibbs, Pradeep Hathiramani, Charlie Richman, and Jonathan Woodbury
of ICF Incorporated and Katherine Stenberg and Eric Taylor of the Bruce Company.
Useful comments were provided by people throughout industry and the U.S.
government and all comments were greatly appreciated. Comments of particular importance
were provided by Chuck Anderson (SEC Donohue), Don Augenstein (EMCON Associates),
Lee Baldwin (UC Davis), Bill Breed (DOE), Floyd Byers (TAMU), Gary Evans (USDA), Gerry
Finfinger (USBM), Andy Hashimoto (University of Oregon), Nelson Hay (AGA), Ray Huitric
(SWANA), Tony Janetos (NASA), Donald Johnson (Colorado State University), Carla Kertis
(USBM), Dave Kirchgessner (US EPA), Robert Lott (GRI), Edward Repa (NSWMA), LM. "Mac"
Safley (North Carolina State University), Susan Thorneloe (US EPA), and Ted Williams (DOE).
-------
EXECUTIVE SUMMARY
Methane is a large contributor to potential global warming, second to carbon dioxide
(Exhibit ES-1). Methane's overall contribution is large in part because it is a potent
greenhouse gas. Methane is twenty times more effective at trapping heat in the atmosphere
than carbon dioxide over a one hundred year time period.1 Furthermore, methane's
concentration in the atmosphere is changing at a rapid rate. Methane concentrations in the
atmosphere have more than doubled over the last two centuries and continue to rise
annually. These increases are largely due to increasing emissions from anthropogenic
(human related) sources, with anthropogenic emissions now constituting about seventy
percent of total emissions.
This report is one of a set of reports requested by Congress in Section 603 of the
Clean Air Act Amendments of 1990 to provide information on a variety of domestic and
international methane issues. This report provides estimates of methane emissions from the
major sources of anthropogenic methane emissions in the United States.
CURRENT EMISSIONS
Methane emissions are generated from a variety of complex geo-chemical, biological,
and energy systems. Emissions from these systems have large regional variations as a result
of different management practices, climates, and underlying physical conditions. The
emissions also vary temporally due to climatic factors, changes in management practices, and
the inherent variability of the emissions processes. Consequently, emissions can vary greatly
from place to place, day to day, season to season, and year to year. These variations in
emissions contribute to uncertainty in estimates of emissions from methane sources and
complicate attempts to verify emissions rates through direct measurement of the individual
sources.
Estimates of methane emissions from
the major sources in the United States have
been developed, along with the associated
uncertainties in the estimates. Because
comprehensive monitoring of methane
emissions from many of the major sources
In 1990 U.S. anthropogenic methane
emissions are estimated to have been
about 25 to 30 teragrams (Tg). Landfills
are the largest source, accounting for
about 36 percent of the total.
is not yet feasible, estimates are based on
experimental data, models, and engineering
analyses. The results are supported by
available measurement data. The current estimates include substantial uncertainty, and in
most areas this uncertainty is not expected to be resolved in the near term.
The characteristics and the current estimates for the major anthropogenic methane
sources in the United States are summarized as follows:
1 Methane is reported with a direct Global Warming Potential (GWP) of 11 over a one hundred year time frame
and with indirect effects that could be comparable in magnitude to its direct effect (IPCC 1992). The GWP reflects
the effect that releasing a kilogram of methane would have over a specified time horizon, relative to releasing a
kilogram of carbon dioxide.
ES-1
-------
Exhibit ES-1
Global Contribution to Integrated Radiative Forcing by Gas for 19901
Carbon Dioxide: 5696
N i tiro us
Ox i cle : 5%
CFCs : 1 \%
Methane : 18%
Estimated on a carbon dioxide equivalent basis using IPCC (1990) global warming
potentials (GWPs) for a 100-year time horizon. Anthropogenic emissions only.
1 This chart is used to present a general understanding of methane's contribution to future warming based on
the GWPs presented in IPCC (1990). However, these GWPs are continually being revised due to a variety of
scientific and methodological issues. It is likely that the contribution of CFCs presented will decrease and that
the contribution of other gases will be about the same or greater upon further investigation.
Landfills
Landfill gas, which is composed mainly of methane and carbon dioxide, results from
the anaerobic decomposition of degradable organic wastes. This process begins after the
waste has been in the landfill for a period of 10 to 50 days. Although the majority of the
methane is usually generated within 30 years of a landfill's completion, methane generation
can continue for 60 years or more. Emissions from landfills are affected by site-specific
factors such as:
• waste composition: degradable organic waste provides the material needed
for methane generation, while the presence of toxic wastes can inhibit methane
production;
moisture: moisture is necessary for methane production in landfills; and
size: the size of the landfill may influence methane emissions or the methane
emissions rate.
Landfills located in moist climates with highly degradable organic material and without a
history of toxic waste disposal have the highest emissions.
ES-2
-------
Emissions of methane from landfills in the United States are estimated to range from
about 8.1 to 11.8 Tg per year. The majority of these emissions result from the disposal of
wastes in municipal solid waste landfills (90 to 95 percent), with the remaining methane
emitted from the disposal of industrial wastes. The largest 20 percent of municipal waste
landfills produce about 80 percent of the methane. Landfills are the single largest source of
methane emissions in the United States, representing about 36 percent of U.S. anthropogenic
emissions. These emissions are also a significant portion of the 20 to 70 Tg estimated for
global landfill emissions, representing about 20 to 40 percent.
Domesticated Livestock
Methane is produced as part of the normal digestive processes of animals. Ruminant
animals (cattle, sheep, and goats) produce significant quantities of methane and account for
nearly all the methane emissions from domesticated livestock in the United States. Ruminant
animals have large "fore-stomachs" or rumens, in which microbial fermentation converts feed
into products that can be digested and utilized by the animal. It is this fermentation that
enables ruminant animals to eat coarse forages such as grasses and straws which
monogastric animals, including humans, cannot digest.
Methane is produced by rumen methanogenic bacteria and is exhaled or eructated by
the animal. The quantity of methane emitted is generally dependent upon the quantity and
type of feed consumed and the manner in which the feed is fermented in the rumen. The
emissions rate varies throughout the day, and can vary greatly among animals fed and
managed in a similar fashion. Mature animals with high feed intakes generally have the
largest emissions.
Methane emissions from domesticated livestock in the United States are estimated in
the range of 4.6 to 6.9 Tg per year. These emissions can be further divided among beef
cattle (69 percent of all livestock emissions), dairy cattle (26 percent) and other livestock
(5 percent). Managed livestock emissions represent about 21 percent of U.S. anthropogenic
methane emissions, which make it the second largest source. These emissions represent
about 7 percent of the 65 to 100 Tg estimated for world emissions of methane from
ruminants.
Coal Mining
Methane is formed during the coal formation process, and is stored within coal seams
and surrounding rock strata. When coal is mined, methane is released to the atmosphere. In
underground mines, methane is hazardous because it is explosive at low concentrations in air
(5 to 15 percent). Therefore, underground mines use ventilation and other degasification
systems to remove methane from mine working areas and this methane is usually vented to
the atmosphere. In surface mines, methane is emitted directly to the atmosphere as the rock
strata overlying the coal seam is removed.
The amount of methane released from a mine depends mainly upon the depth and
type of coal, with deeper mines generally emitting greater quantities of gas. Another
important factor is the mining method used. Emissions vary greatly from mine to mine and
can vary from day to day at an individual mine as a result of changes in specific geologic
and/or mining conditions.
ES-3
-------
Methane emissions from coal mining are estimated in the range of 3.6 to 5.7 Tg per
year. The majority of these emissions result from underground mining operations (about
70 to 80 percent). These emissions can be further divided between emissions from
ventilation systems where the methane is released at concentrations of less than one percent
in air (40 to 65 percent) and emissions from degasification systems where the methane is
emitted in concentrations between 30 and 95 percent. Coal mining emissions represent
about 17 percent of U.S. anthropogenic methane emissions, and about 13 percent of the 25
to 50 Tg estimated for global emissions of methane from coal mining.
Livestock Manure
Livestock manure contains un-digested organic material. When handled under
anaerobic conditions, microbial fermentation produces methane. The United States like many
developed countries manages the wastes from large concentrations of cattle, swine, and
poultry using liquid waste management systems that are conducive to anaerobic fermentation
and methane production. The emissions of methane from livestock manure are driven by the
quantity of manure produced, how it is handled, and the temperature at which it is handled.
The manure management system employed in particular is very important, with liquid-based
systems (such as lagoons) converting large portions of the available carbon to methane and
pasture systems converting fairly small portions. Emissions vary from system to system and
throughout the year.
Methane emissions from livestock manure are estimated in the range of 1.7 to 3.6 Tg
per year. A large portion of these emissions result from the management of wastes in liquid
and slurry systems (80 percent). Most of these emissions are from the management of dairy
cattle (30 percent) and swine (50 percent). Methane emissions from livestock manure
represent about 10 percent of U.S. anthropogenic methane emissions and about 10 percent
of the 20 to 30 Tg estimated for world emissions of methane from animal wastes.
Natural Gas Systems
Methane is the major component of natural gas. Any leakage or emission during the
production, processing, transmission, and distribution of natural gas contributes to total
methane emissions. Because natural gas is often found in conjunction with oil, leakage
during gas production from oil wells is also a source of emissions. Methane emissions from
natural gas systems are dependent upon a variety of operational or routine practices and the
state of existing facilities. Emissions can vary greatly from facility to facility.
Methane emissions from natural gas systems are estimated in the range of 2.2 to
4.3 Tg per year. These emissions result primarily from leakage (or fugitive emissions)
throughout all segments of the gas systems (38 percent). Emissions also result from the
exhaust of compressor engines as well as starts and stops of these engines (18 percent) arid
from the venting of pneumatic equipment used frequently in the transportation of gas
(21 percent). Other sources of emissions include routine maintenance, system upsets and
vents from dehydrator units (23 percent).
Methane emissions from natural gas systems represent about 10 percent of U.S.
anthropogenic methane emissions and about 9 percent of the 25 to 50 Tg estimated for
global emissions of methane from natural gas systems.
ES-4
-------
Other Sources
Methane is also produced from several other sources in the United States including
rice cultivation, combustion, petroleum production, industrial processes, and land use
change. Methane is produced during flooded rice cultivation by the anaerobic decomposition
of organic matter in the soil. These flooded soils are ideal environments for methane
production. Emissions of methane from U.S. rice cultivation are estimated to range from 0.1
to 0.7 Tg per year. These emissions represent about 1 percent of the total United States
emissions and about 0.5 percent of the global annual emissions of 20 to 150 Tg. Although
global emissions for rice cultivation are estimated to be large, this source is quite small in the
United States.
The process of fuel combustion is a recognized source of anthropogenic methane.
The majority of methane from this source is produced as a by-product of incomplete
combustion. For example, this may result even when methane is not an original component
of the fuel. Additionally, when methane is an original component of the fuel, and it is not fully
combusted, it may be emitted directly to the atmosphere. In general, the methane emissions
resulting from fuel combustion are much less than those associated with fuel production
activities, such as coal mining and oil and natural gas production, processing, transmission,
and distribution.
Total methane emissions from fuel combustion were estimated to range from 0.5 to
1.7 Tg in 1990. This represents approximately 3 percent of total U.S. methane emissions. Of
this total, stationary sources account for about 0.2 to 1.4 Tg per year, while mobile sources
account for the rest. Fuel wood combustion accounts for the majority of the stationary
source emissions.
Methane is emitted during the production, transportation and refining of petroleum.
Emissions from leaks (fugitive emissions) and equipment venting are found at oil wells, crude
oil treatment and storage facilities, and refineries. These facilities, which do not produce
natural gas for commercial sale, are not included in the natural gas systems estimate.
Total methane emissions from petroleum production and refining were estimated to
range from 0.1 to 0.6 Tg per year in the United States. Of this total, the majority is associated
with venting during oil production. This estimate, and global estimates of venting emissions
are particularly uncertain due to a lack of data.
Additional sources of anthropogenic methane emissions in the United States include
non-fuel biomass burning, treatment of industry wastewater, ammonia production, coke, iron,
and steel production, and land use changes. Emissions from these sources are believed to
be small relative to the other sources in this report. In addition, little information on methane
emissions from these sources is available upon which to base estimates.2 For now,
emissions from these sources are not estimated in this report.
Total emissions from anthropogenic sources in the U.S. for 1990 are estimated at
about 25 to 30 Tg per year. These estimates are summarized in Exhibit ES-2, Exhibit ES-3,
and Exhibit ES-4. Exhibit ES-2 displays the relative magnitudes of the individual sources.
2 Additional information will become available through efforts of EPA's Office of Research and Development on
emissions from wastewater treatment systems over the next several years.
ES-5
-------
Exhibit ES-2
Contribution of Major Methane Sources to Total U.S. Anthropogenic Emissions
Landfills. 36%
Domestic
Li vestock:
21%
Li vest urk
Manure 10%
Natural Gas
Systems• 10%
. ,,'"•' Other
Sources
Coal Mining 17%
Exhibit ES-3 provides a further characterization of the individual sources and indicates
which portions of the emissions from each source may be partially controllable. In general
(except for domesticated livestock) the emissions that may be partially controllable are
emissions from relatively large, gassy systems where there may be sufficiently available
methane to economically support recovery and utilization activities.
Up to 400 million Btu of fuel (equivalent to 400 bcf of natural gas or 16 million tons of
coal) could be available for recovery. In addition, the emissions from managed ruminants
may be partially reducible as the beef and dairy industries continue their trends of the past
decades toward increased production efficiency, producing an increasing quantity of product
with fewer animals. The potential for reducing emissions from these sources is examined in
Opportunities to Reduce Methane Emissions from Anthropogenic Sources in the United
States, another Report to Congress being prepared by EPA.
Finally, Exhibit ES-4 displays the contribution of U.S. emissions to the estimated global
emissions of methane from each source. This exhibit shows that the United States is a major
contributor for all sources other than rice cultivation.
ES-6
-------
Exhibit ES-3
U.S. Anthropogenic Emissions Summary
Source
Landfills3
Domesticated Livestock
Dairy Cattle
Beef Cattle
Other Animals
Total Domesticated Livestock
Coal Mining
Underground Coal Mines
Ventilation Systems
Degasffication Systems a
Surface Coal Mines
Post-Mining
Total Coal Mining
Natural Gas Systems
Fugitive Emissions
Pneumatic Devices
Engine Exhaust
Other
Total Natural Gas Systems'3
Livestock Manure
Liquid Based Systems
Solid Based Systems
Total Livestock Manure
Other Sources
Rice
Combustion
Oil Systems
Other0
Total Other Sources b
Total b
U.S. Emissions
(Tg)
8.1 to 11.8
1.2 to 1.8
3.2 to 4.8
0.2 to 0.4
4.6 to 6.9
2.3
0.5 to 1.8
0.2 to 0.8
0.5 to 0.9
3.6 to 5.7
0.7 to 1.9
0.4 to 1.1
0.3 to 0.6
0.4 to 1.9
2.2 to 4.3
1.4 to 2.3
0.3 to 1.3
1.7 to 3.6
0.1 to 0.7
0.5 to 1.7
0.1 to 0.6
Not Estimated
1.1 to 2.5
25 to 30
Partially
Controllable
s
'<
'
J
'
;
a Does not include methane recovered and used as an energy source. For landfills, about 1 .5 Tg was
recovered and flared or used for energy purposes. For coal mines, about 0.25 Tg was recovered and
sold to pipelines.
b The uncertainty in the total is estimated assuming that the uncertainty for each source is independent.
Consequently, the uncertainty range for the total is more narrow than the sum of the ranges for the
individual sources. For natural gas systems, total emissions are calculated assuming that some of the
uncertainty for each source is independent.
c Includes non-fuel biomass burning, wastewater from agricultural industries, ammonia production,
coke, iron, and steel production, and land use changes.
ES-7
-------
&
&
i
CO
CD
•9L
o
or
0)
O
T3
O
CO
0)
'
CO
Co
§'
CO
s
o
a
CD
Q.
Q>
CO
CO
CO
CU
I
CO
o
c
o
CD
CO
0)
O
O
O
I
CO
o
O
CD
£
o
_
o
CT
CD
55'
en
'
C
01
o
o
3.
5'
r+
O
O
o
CT
I s?
o :T
•o
o
(Q
(D
o'
I
(D
m
3
o
v>
(O
O
o
(D
IT
-------
FUTURE EMISSIONS
Methane emissions from sources in the United States have the potential to increase
over the next decades. Estimates were developed for the major sources for the year 2000
and the year 2010. Total annual emissions from anthropogenic sources are expected to grow
to between 27 and 35 Tg in the year 2000 and to between 29 and 39 Tg in the year 2010.
Exhibit ES-5 presents high and low estimates of future emissions by source. These future
projections include the following trends and assumptions:
Landfills. Emissions from landfills may increase by about 10 percent by 2000
and 15 percent by 2010 as the result of increased amounts of waste in place in
landfills. Although the cumulative amount of degradable waste in landfills is
increasing, the amount of waste disposed annually is not projected to increase
during this period. Future emissions could be reduced significantly if currently
proposed rules are implemented which would require collecting and flaring
landfill gas in order to reduce emissions of non-methane organic compounds.
Domesticated Livestock. Future emissions from domesticated livestock will be
driven by levels of beef and milk production and the production practices used.
Only modest increases in beef production are expected, based on a
continuation of recent trends in reduced per capita beef consumption. Milk
production could increase substantially, depending on the outcome of
international negotiations to promote free trade in milk products and other
agricultural commodities. Continuing increases in the productivity and
efficiency in the beef and dairy industries will help to offset increased emissions
due to increased levels of production.
Coal Mining. Growth in coal mining emissions over the next decades is likely
as U.S. coal production increases. In 1988, U.S. coal mines produced 961
million tons of coal. By 2000 it is forecast that production could range from
1,117 to 1,241 million tons, and in 2010, production could reach 1,364 to 1,560
million tons. Emissions growth may be moderated somewhat to the extent
that, as a result of the 1990 Clean Air Act Amendments, production shifts to
low sulfur surfaced mined coals, which tend to be less gassy. Underground
mining is expected to represent a significant portion of future coal production,
however. In addition, it is possible that U.S. underground mines will become
gassier in the future as mining heads toward deeper coal seams. The potential
impact of this trend is not reflected in these estimates, however.
Livestock Manure. Methane emissions from livestock manure may grow
substantially in the future. As the result of concern over groundwater and
surface water pollution, many states are now requiring farms to control manure
runoff. In many cases, farms will be switching from solid waste handling
systems to liquid treatment systems, such as lagoons, to comply with these
new requirements. Consequently, manure management may shift toward
practices with relatively high emissions rates.
• Natural Gas Systems. Methane emissions from the natural gas system will
increase in response to increases in the size of the natural gas system. Gas
production and consumption are expected to increase, as will the total miles of
ES-9
-------
pipeline and the number of other facilities. However, a variety of practices and
technologies are available that could offset potential increases in emissions.
Other Sources. Emissions from other sources are not expected to increase
substantially in the future. Although total fuel use is expected to increase,
improved efficiency of combustion and increasing use of emissions control
technologies will offset potential emissions increases. Oil production and
refining are expected to decline somewhat over the next several decades, so
these emissions are also expected to decline. Rice production in the U.S.,
which has been relatively steady in the recent past, is also expected to remain
unchanged in the next 20 years.
Exhibit ES-5
Estimates of Future Anthropogenic Methane Emissions in the United States
Methane Emissions Tg/yr
15
10 -
EstimateEDHagh Estimate
1990 2000 2010 1990 2000 2010 1990 2000 2010
Natural Gas, Coal Landfills
Methane Emissions Tg/yr
15 i-
10
1990 2000 2010 1990 2000 2010 1990 2000 2010
Domesticated
Livestock
Livestock
Manure
Other
Sources
ES-10
-------
REFERENCES
IPCC (Intergovernmental Panel on Climate Change). 1990. Climate Change: The IPCC
Scientific Assessment. Report Prepared for Intergovernmental Panel on Climate
Change by Working Group 1.
IPCC (Intergovernmental Panel on Climate Change). 1992. Climate Change 1992: The
Supplementary Report to the IPCC Scientific Assessment. Report Prepared for
Intergovernmental Panel on Climate Change by Working Group 1.
ES-11
-------
CHAPTER 1
INTRODUCTION
This report presents estimates of current and future emissions of methane from
anthropogenic (human related) sources in the U.S. This report is written in partial fulfillment
of Section 603 of the Clean Air Act Amendments of 1990, which requires that the EPA prepare
and submit to Congress a series of reports on domestic and international issues concerning
methane.
1.1 BACKGROUND: The Importance of Methane
Atmospheric concentrations of methane are increasing. These increases are highly
correlated with increases in global population and human-related activities that release
methane to the atmosphere. Assessing the current and potential future levels of methane
emissions from anthropogenic sources and the portion of such emissions that are
controllable is an important step towards developing emissions reduction strategies.
Reducing methane emissions from anthropogenic sources is one of the most effective means
of mitigating global warming in the near term for the following reasons:
• Methane (CH4) is one of the principal greenhouse gases, second only to
carbon dioxide (CO2) in its contribution to potential global warming. In fact,
methane is responsible for roughly 18 percent of the total contribution in 1990
of all greenhouse gases to "radiative forcing," the measure used to determine
the extent to which the atmosphere is trapping heat due to emissions of
greenhouse gases.1
• Methane concentrations in the atmosphere have been rising rapidly.
Atmospheric concentrations of methane are increasing at about 0.6 percent per
year (Steele et al. 1992) (in contrast to CO2, whose atmospheric concentrations
are increasing at about 0.4 percent per year)2 and have more than doubled
over the last two centuries (IPCC 1990a).
• Methane is a potent contributor to global warming. On a kilogram for
kilogram basis, methane is a more potent greenhouse gas than CO2 (about 60
times greater after 20 years, 22 times greater after 100 years, and 9 times
greater after 500 years).3
1 Global contribution to radiative forcing by gas is estimated on a carbon dioxide equivalent basis using IPCC
(1990a) global warming potentials for a 100-year time horizon, including direct and indirect effects of methane.
2 Based on measurements taken at Mauna Loa from 1970 to 1990 (Oakridge 1992).
3 Methane is reported with a direct Global Warming Potential (GWP) of 35 over a 20 year time frame, 11 over
100 years, and 4 over 500 years, and with indirect effects that could be comparable in magnitude to its direct effect
(IPCC 1992). The GWP reflects the effect that releasing a kilogram of methane would have over a specified time
horizon, relative to releasing a kilogram of carbon dioxide.
1-1
-------
• Reductions in methane emissions will produce substantial benefits in the
short-run. Methane has a shorter atmospheric lifetime than other greenhouse
gases - methane lasts around 11 years in the atmosphere, whereas CO2 lasts
about 120 years (IPCC 1992). Due to methane's high potency and short
lifespan, stabilization of methane emissions will have a rapid impact on
mitigating potential climate change.
• Methane stabilization is nearly as effective as limiting CO2 emissions to
1990 levels. In order to stabilize methane concentrations at current levels,
total anthropogenic methane emissions would need to be reduced by about 10
percent. This methane concentration stabilization would have roughly the
same effect on actual warming as maintaining CO2 emissions at 1990 levels
(Hogan et al. 1991).
• Significant portions of anthropogenic methane emissions are controllable.
For all the major U.S. sources of anthropogenic emissions, several well
demonstrated emissions reduction technologies are available. Moreover, in
contrast to the numerous sources of other greenhouse gasses, a few large and
gassy facilities often account for a large portion of methane emissions.
Therefore, applying emissions reductions strategies only to these gassiest
facilities would result in a substantial decrease in estimated current and future
methane emissions levels.
The unique characteristics of methane emissions demonstrate the significance of
promoting strategies to reduce the amount of methane discharged into the atmosphere.
Understanding the sources of methane emissions, and in particular the emissions from the
systems that are partially controllable, is the first step in identifying cost-effective options for
reducing emissions.
1.1.1 What is Methane?
Methane is a radiatively and chemically active trace greenhouse gas.4 Being
radiatively active, methane traps infrared radiation (IR or heat) and helps to warm the Earth. It
is currently second only to carbon dioxide in contributing to potential future warming. Being
chemically active, methane enters into chemical reactions in the atmosphere that increase not
only the abundance of methane, but also atmospheric concentrations of ozone5 and
stratospheric concentrations of water vapor, which are both greenhouse gases.
Methane is emitted into the atmosphere largely by anthropogenic sources, which
currently account for approximately 70 percent of the estimated 505 teragrams (Tg) of annual
4 A trace gas is a gas that is a minor constituent of the atmosphere. The most important trace gases
contributing to the greenhouse effect include water vapor, carbon dioxide, ozone, methane, ammonia, nitrous oxide,
and sulfur dioxide.
5 While methane does not contribute to the formation of urban smog, methane is a major concern in the
formation of ozone in the free troposphere.
1-2
-------
global methane emissions.6 Anthropogenic sources of methane emissions include: natural
gas and oil systems; coal mining; landfills; domesticated livestock; liquid and solid wastes;
rice cultivation; and biomass burning. Natural sources of methane, which currently account
for the remaining 30 percent of global emissions, include natural wetlands (e.g., tundra, bogs,
swamps), termites, wildfires, methane hydrates, and oceans and freshwaters.
The concentration of methane in the atmosphere is determined by the balance of the
input rate, which is increasing due to human activity, and the removal rate. The primary sink
(removal mechanism) for atmospheric methane is its reaction with hydroxyl (OH) radicals in
the troposphere. In this reaction, methane is converted into water vapor and carbon
monoxide, which is in turn converted into carbon dioxide (CO2). The atmospheric
concentration of OH radicals is determined by complex reactions involving methane, carbon
monoxide, non-methane hydrocarbons (NMHC), nitrogen oxides, and tropospheric ozone.
The size of an OH sink can vary and may actually decrease in response to increasing levels
of methane (IPCC 1992). A small amount of methane is also removed from the atmosphere
through oxidation in dry soils. Compared to removal by reaction with OH, this oxidation
mechanism is believed to be relatively small. There are no significant anthropogenic activities
that remove methane from the atmosphere. Exhibit 1 -1 presents a summary of methane
sources and sinks. Based on the balance of these sources and sinks, methane's
atmospheric lifetime is presently estimated to be about 11 years (IPCC 1992).
1.1.2 Atmospheric Levels of Methane Are Rising
The concentration of methane in the atmosphere has been steadily increasing. The
rise in methane concentrations has been well documented in recent studies and corroborated
by measurements from different locations and several monitoring groups.
Analyses of ice cores in Antarctica and Greenland have yielded estimates of
atmospheric methane concentrations of approximately 0.35 parts per million by volume
(ppmv) to 0.65 ppmv for the period between 10,000 and 160,000 years ago. Similar analyses
of air in ice cores have placed atmospheric methane concentrations at approximately
0.8 ppmv for the period between 200 and 2,000 years ago. The level of methane rose to
about 0.9 ppmv at the beginning of this century (IPCC 1990a).
Direct measurement of the global atmospheric methane concentration was begun in
1978. At that time the global atmospheric methane concentration was calculated to be 1.51
ppmv. In 1990, the level was approximately 1.72 ppmv - nearly double the concentration
level estimated for the beginning of this century (IPCC 1990a). A summary of the ice core
data and direct measurement data showing the increase in atmospheric methane
concentrations is provided in Exhibit 1 -2. In addition to ice core data and direct atmospheric
measurements, analysis of infrared solar spectra has shown that the atmospheric
concentration of methane increased by about 30 percent over the last 40 years (Rinsland et
al. 1985).
6 Portion of total methane emissions from anthropogenic sources is based on IPCC (1992). Total annual
methane emissions is based on Crutzen (1991).
1-3
-------
Exhibit 1-1
Estimated Sources and Sinks of Methane
(Tg CH4 per year)
Anthropogenic Sources:
Oil/Gas Systems1
Coal Mining
Landfills
Domesticated Livestock
Animal Wastes
Rice
Biomass Burning
Wastewater Treatment
Natural Sources:
Natural Wetlands
Termites
Oceans and Freshwaters
CH4 Hydrate
Destabilization
Total Natural and Anthropogenic: 2
Sinks:
Atmospheric removal
Removal by soils
Atmospheric Increase:
Global Estimate
50
40
30
80
25
60
40
25
115
20
15
0
505
470
30
32
Global Range
30-70
25-50
20-70
65- 100
20-30
20- 150
20-80
N/A
100 -200
10-50
5-45
0-5
400 - 61 0
420-520
15-45
28-37
1 1t is estimated that natural gas systems account for 25 to 50 Tg and oil
systems account for 5 to 20 Tg.
2 Source: Crutzen (1991). Estimation based on observation of atmospheric
sources and sinks rather than sum of individual emissions shown here.
Source: IPCC (1992)
Note: Estimates of global methane emissions are continually being
revised. Another EPA Report to Congress on International Methane
Emissions will contain new estimates of global methane emissions
from individual sources.
1-4
-------
Exhibit 1-2
Measurements of Global Methane Concentrations
1800-j
1300-
800-
O
300-
• Modem recoid
* Spto toe core
» Byrd toe core
• Dy» toe core
o Mortok toe core
A
O • • A
I
10s
I
10s
10* 10* 10* 101
Years Before Present (1990 A.D.)
10°
Annul atmospheric CH4 concentrations daring the past 160,000 yean
(derived Iron ice cores and the NOAA/CMDL fhsk MunpUng network).
Source: Oak Ridge (1990).
At present, the atmospheric abundance of methane is approximately 4,900 Tg (IPCC
1990a); this amount is thought to be increasing by about 30 to 40 Tg per year (Steele et al.
1992). Atmospheric methane concentrations are expected to continue to increase, although
global measurement programs indicate that the rate of increase appears to have slowed in
the last several years (Steele et al. 1992). However, given a continuation of the current
annual rate of increase of atmospheric methane of about 0.0095 to 0.0133 ppmv (Steele et al.
1992), the atmospheric concentration of methane would exceed 2.0 ppmv by the year 2020.
Recent models of expected future emissions and atmospheric processes indicate that without
controls, atmospheric concentrations could range from 3.0 ppmv to over 4.0 ppmv by the
year 2100 (USEPA 1989; IPCC 1992), although these scenarios should be reinvestigated
using the most recent information on methane concentration trends.
1.1.3 Methane and Global Climate Change
Methane's increasing concentration in the atmosphere has important implications for
global climate change. Methane is very effective at absorbing infrared radiation (IR) given off
1-5
-------
by the Earth's surface. By absorbing IR and inhibiting its release into space, the presence of
methane contributes to increased atmospheric and surface temperatures. This process is
commonly referred to as the "greenhouse effect."
A gram of methane is about 35 times more effective at warming the surface than a
gram of CO2 over a 20 year time frame (IPCC 1992). In addition to this direct radiative
forcing, methane's participation in chemical reactions in the atmosphere indirectly contributes
to global warming by influencing the amount of ozone in the troposphere and stratosphere,
the amount of hydroxyl in the troposphere, and the amount of water vapor in the
stratosphere. Methane's indirect effect on warming resulting from these chemical reactions
could be comparable in magnitude to its direct effect, although considerable uncertainty
remains (IPCC 1992)7 It has been estimated that approximately 18 percent of the
greenhouse effect is due to increasing atmospheric methane concentrations. The total
contribution to radiative forcing of all greenhouse gases in 1990 is shown in Exhibit 1-3.
Exhibit 1-3
Global Contribution to Integrated Radiative Forcing by Gas for 19901
Carbon Dioxide: 66%
N i trous
Ox ide : 5%
CFCs: 11%
Methane • 18%
Estimated on a carbon dioxide equivalent basis using IPCC (1990a) global warming
potentials (GWPs) for a 100-year time horizon. Anthropogenic emissions only.
1 This chart is used to present a general understanding of methane's contribution to future warming based on
the GWPs presented in IPCC (1990a). However, these GWPs are continually being revised due to a variety of
scientific and methodological issues. It is likely that the contribution of CFCs presented will decrease and that
the contribution of other gases will be about the same or greater upon further investigation.
7 The uncertainty in the GWPs for methane result largely from the indirect effects of methane in the atmosphere,
which have not been fully characterized, and from methodological issues in the GWP calculations. Some of these
uncertainties will be reduced over the next several years through the efforts of the Intergovernmental Panel on
Climate Change as well as others, including EPA's Office of Research and Development.
1-6
-------
Models of atmospheric chemical processes have indicated that increasing methane
concentrations result in net ozone production in the troposphere and lower stratosphere and
net ozone destruction in the upper stratosphere. The overall effect is that methane by itself
causes a net increase in ozone (Wuebbles and Tamaresis 1992).8
As the most abundant organic species in the atmosphere, methane plays an influential
role in determining the oxidizing capacity of the troposphere. Through reactions with
hydroxyl, 80 to 90 percent of methane destruction occurs in the troposphere (Cicerone and
Oremland 1988). Increasing methane levels could reduce hydroxyl, which would result in a
further increase in the methane concentration. A decrease in the oxidizing capacity of the
troposphere would increase not only the atmospheric lifetime of methane, but also the lifetime
other important greenhouse gases, and would permit transport of pollutants over long
distances, resulting in atmospheric changes even in remote regions (Wuebbles and
Tamaresis 1992).
Water vapor is one of the most important greenhouse gases. Stratospheric water
vapor concentrations should increase as concentrations of methane increase; methane
oxidation reactions roughly produce two moles of water vapor for each mole of methane that
is destroyed (Wuebbles and Tamaresis 1992). In addition to the impact on global warming,
increases in stratospheric water vapor concentrations as a result of increased methane
concentrations could contribute to the formation of polar stratospheric clouds (PSCs), which
have been identified as one factor that enables the chlorine and bromine from
chlorofluorocarbons (CFCs) and halon compounds to cause the severe seasonal loss of
stratospheric ozone over Antarctica (WMO 1990).
1.1.4 Stabilization and Further Reductions of Global Methane Levels
Since atmospheric methane has been increasing at a rate of about 30 to 40 Tg per
year, stabilizing global methane concentrations at current levels would require reductions in
methane emissions by approximately this same amount. Such a reduction represents about
10 percent of current anthropogenic emissions. This reduction is much less than the
percentage reduction necessary to stabilize the other major greenhouse gases: CO2 requires
a greater than 60 percent reduction; nitrous oxide requires a 70 to 80 percent reduction; and
chlorofluorocarbons require a 70 to 85 percent reduction (IPCC 1990b).
Because methane has a relatively short atmospheric lifetime as compared to the other
major greenhouse gases, reductions in methane emissions will help to ameliorate global
warming relatively quickly. Therefore, methane reduction strategies offer an effective means
of slowing global warming in the near term. Exhibit 1-4 compares the effect on future
temperature increases of stabilizing methane concentrations versus maintaining C02
emissions at 1990 levels. This exhibit illustrates that stabilizing atmospheric concentrations of
methane will have virtually identical effects on actual warming as capping C02 emissions at
1990 levels. The recent evidence that the rate of annual increase in methane emissions is
As described in IPCC (1990a), "Ozone plays an important dual role in affecting climate. While CO2and other
greenhouse gases are relatively well-mixed in the atmosphere, the climatic effect of ozone depends on its
distribution in the troposphere and stratosphere, as well as on its total amount in the atmosphere. Ozone is a
primary absorber of solar radiation in the stratosphere where it is directly responsible for the increase in temperature
with altitude. Ozone is also an important absorber of infrared radiation. The balance between these radiative
processes determines the net effect of ozone on climate."
1-7
-------
Exhibit 1-4
Carbon Dioxide and Methane Reduction Comparison1
Actual
Ttmpentun
Roughly MMnffca/ •wets on actual warning
- CO. •fntufons
-CH. ConcwMr»0
IPCC-BAU
CH. stabilization
CO2 capped at 1990
CH. and CO,
1988
2000
Assumes 3° equilibrium warming
Illlllllll Constitutes uncertainty range due to NO,
Figure 2. Benefits of methane stabilization where methane emissions are capped at 540 Tg/yr as compared
to capping CO, emissions at 1990 levels (and concentrations grow to over 500 pom by 2100)
1 Benefits of CH4 stabilization where CH4 emissions are capped at 540 Tg/yr as compared to
capping CO 2 emissions at 1990 levels (and concentrations grow to over 500 ppm by 2100).
Source: Hogan et al. (1991). ^
slowing (Steele et al. 1992) may mean that reductions on the order of 30 to 40 Tg could
reduce concentrations to the extent that they fell below the level of stabilization. This result
would also have large benefits for the global atmosphere.
1.2 OVERVIEW OF METHANE SOURCES
Methane emissions are generated from a variety of complex geo-chemical, biological,
and energy systems. Anthropogenic sources account for about 70 percent of annual global
methane emissions, while natural sources account for the remaining 30 percent, as shown in
Exhibit 1-1.
1.2.1 Anthropogenic Sources
Increases in methane concentrations are highly correlated with increases in global
population and human-related activities that release methane to the atmosphere. The U.S. is
one of the largest contributors of global anthropogenic methane emissions. The major U.S.
sources of anthropogenic methane emissions are: natural gas systems; coal mining; landlills;
1-8
-------
domesticated livestock; and, livestock manure. Compared to other nations on a source by
source basis, the U.S. is among the highest methane emitting nations for some of the
sources, but among the lowest for others. For example, the U.S. is estimated to have the
largest emissions from landfills - about one third of worldwide emissions from this source.9
In contrast, the U.S. accounts for a very small portion of emissions from rice cultivation - Asia
is responsible for 90 percent of emissions from this source. Similarly, while biomass burning
and wastewater treatment are significant anthropogenic sources in some developing
countries, they account for only a small portion of total U.S. emissions.
Natural Gas Systems and Oil Systems
Methane is the major component of natural gas. Leakage during the production,
processing, transmission, and distribution of natural gas contributes to total methane
emissions. Because natural gas is often found in conjunction with oil, gas leakage during oil
exploration and production is also a source of emissions. Very little data have been available
upon which to base estimates of methane emissions from oil and gas systems in the U.S. and
globally. Some authors have suggested that approximately 2 to 4 percent of the total global
natural gas production may be emitted. At this rate, total global emissions are estimated at
about 30 to 70 Tg per year (IPCC 1992). However, emissions rates vary significantly among
countries and regions. While substantial new information has been developed in the U.S and
elsewhere over the last year, further research is needed.
Coal Mining
Methane is formed during the coal formation process and is stored within coal seams
and surrounding rock strata. When coal is mined, methane is released to the atmosphere. In
underground mines, methane is hazardous because it is explosive at low concentrations in air
(5 to 15 percent). Therefore, underground mines use ventilation and other degasification
systems to remove methane from mine working areas; this methane is usually vented to the
atmosphere. In surface mines, methane is emitted directly to the atmosphere as the rock
strata overlying the coal seam is removed. Coal mining accounts for 25 to 50 Tg of global
methane emissions (IPCC 1992). The amount of methane released from a mine depends
mainly upon the type of coal and the depth of the coal seam - deeper coals and coals with a
higher carbon content generally hold more methane.
Landfills
Landfill gas, which is composed mainly of methane and carbon dioxide, results from
the anaerobic (in the absence of oxygen) decomposition of organic degradable wastes. This
process begins after the waste has been in the landfill for a period of 10 to 50 days and,
although the majority takes place within 30 years of a landfill's completion, methane
generation can continue for 60 years or more. Solid waste landfills account for 20 to 70 Tg of
global methane emissions (IPCC 1992). These emissions are concentrated in developed
countries, where a small number of large landfills account for the majority of emissions.
9 Most of the municipal solid waste generated in the U.S. is disposed of in sanitary landfills, whereas many
other countries incinerate waste or practice open dumping.
1-9
-------
Domesticated Livestock
Among the domesticated animals, ruminant animals (cattle, sheep, buffalo, goats, and
camels) produce significant quantities of methane as part of their normal digestive processes.
Ruminant animals are characterized by a large "fore-stomach" or rumen, in which microbial
fermentation converts feed into products that can be digested and utilized by the animal. The
microbial fermentation enables ruminant animals to utilize coarse forages that monogastric
animals, including humans, cannot digest. Methane is produced by rumen methanogenic
bacteria as a byproduct of normal rumen fermentation, and then is exhaled or eructated by
the animal. The amount of methane produced is dependent upon both animal type and
management practices. Global methane emission estimates from domesticated animals
range from 65 to 100 Tg (IPCC 1992).
Livestock Manure
Methane can be produced during the anaerobic decomposition of the organic material
in livestock manure. Many developed countries manage the wastes from large concentrations
of cattle, swine, and poultry using liquid waste management systems that are conducive to
anaerobic fermentation of the wastes and methane production. Global annual methane
emissions from animal wastes are estimated to be in the range of 20 to 30 Tg (IPCC 1992).
Other Sources: Rice Cultivation, Biomass Burning, and Wastewater Treatment
Methane is also produced from several other sources including, rice cultivation,
biomass burning, and wastewater treatment. Although global emissions for these sources are
estimated to be large, these sources are quite small in the U.S. Each source is discussed in
turn.
Methane is produced during flooded rice cultivation by the anaerobic decomposition
of organic matter in the soil. The flooded soils are ideal environments for methane production
because of their high levels of organic substrates, oxygen-depleting conditions, and moisture.
The level of emissions from rice cultivation varies according to several factors, including:
agricultural practices such as fertilization, water management, or double cropping systems;
soil characteristics such as soil type, acidity, redox potential, and temperature; and the
progression of the growing season. Methane emissions from flooded rice fields are estimated
to be from 20 to 150 Tg annually, and are likely to increase as the worldwide demand for rice
increases (IPCC 1992).
Biomass is burned as part of several agricultural practices, including: converting
forest and savannah ecosystems into cropping or pasture systems; returning nutrients to the
soil; reducing shrubs on rotational fallow lands; and removing crop residues. Biomass is also
burned to provide energy, for example as a cooking fuel. Incomplete combustion during
burning produces methane. The global contribution of biomass burning to methane emission
levels is relatively uncertain because of the lack of data on fire frequency, area burned, and
characteristics of fires. However, biomass burning is estimated to account for between 20
and 80 Tg of global methane emissions annually (IPCC 1992).
Wastewater treatment can produce methane emissions if organic constituents in the
wastewater are treated anaerobically and if the methane produced is released to the
atmosphere. Anaerobic methods are used to treat wastewater from domestic sewage, food
processing and other industrial facilities in some developing countries (Orlich 1990). In
1-10
-------
contrast, wastewater treatment in developed countries principally includes aerobic processes
or anaerobic processes in enclosed systems where methane is recovered and utilized.
Consequently, wastewater treatment in most developed countries has not been considered a
major source of methane emissions.10 Annual global methane emissions from wastewater
treatment lagoons are estimated to be about 25 Tg, although this is very uncertain (IPCC
1992).
1.2.2 Natural Sources
Natural sources of methane emissions include wetlands, termites, oceans and
freshwaters, and hydrates. The anaerobic environment found within natural wetlands is
conducive to the biological processes that result in methane formation. Current emissions
estimates from this source indicate a wide range of uncertainty, with estimates ranging from
100 to 200 Tg per year (IPCC 1992), with approximately half of these emissions resulting from
tropical wetlands and about one third coming from high latitude wetlands (IPCC 1990a). The
estimated range of methane emissions from termites is 10 to 50 Tg per year (IPCC 1992).
The wide range for annual emissions reflects the underlying uncertainty in the size of the
termite population and the amount of biomass they consume. Annual methane emissions
from oceans and freshwaters are estimated to be in the range of 5 to 45 Tg (IPCC 1992). No
adequate recent data is available to reduce the uncertainty of estimates from oceans and
freshwater. Finally, methane emissions from hydrates found in coastal sediments are believed
to be quite small, only 0 to 5 Tg (IPCC 1992). Emissions from natural sources could change
in response to a change in global climate. Although uncertainty exists in the effect that global
warming could have over these four sources, it has been suggested that emissions from
natural sources could rise significantly in response to global warming and other related
changes in the global climate.
1.3 OVERVIEW OF REPORT
Section 603 of the Clean Air Act Amendments of 1990 requires EPA to prepare and
submit to Congress a series of reports on domestic and international issues concerning
methane. The topics for the five required reports are: 1) Anthropogenic Methane Emissions
in the United States; 2) Options for Reducing Anthropogenic Methane Emissions in the United
States; 3) International Anthropogenic Methane Emissions; 4) Options for Reducing
International Anthropogenic Methane Emissions; and, 5) Methane Emissions from Natural
Sources. This report fulfills the requirement for the first topic.
This report presents emissions estimates for each of the major sources of human
related methane emissions in the U.S. Emission estimates are developed representing 1990
conditions and are projected to 2000 and 2010 accounting for important and identifiable
policies and trends.
The chapters of this report are as follows:
10 Some recent information indicates that wastewater managed in lagoons from industries such as the pulp and
paper industry may be emitting significant quantities of methane. Further efforts of EPA's Office of Research and
Development should clarify the contribution of this source.
1-11
-------
• Natural Gas Systems: including estimates ol accidental and intentional
releases of methane from field production, processing, transmission,
distribution, and engine exhaust.
• Coal Mining: including methane emitted from surface and underground
coal mines and methane emitted during the processing, transport, and
storage of extracted coal.
• Landfills: emissions occurring following the disposal of solid waste.
• Domesticated Livestock: including emissions from beef and dairy cattle
and other animals.
• Livestock Manure: including emissions from solid and liquid waste
management systems.
• Other Emissions: other human related sources of domestic methane
emissions including rice cultivation, stationary and mobile combustion,
production and refining of petroleum liquids, biomass burning, and
industrial processes and wastes.
1.4 REFERENCES
Cicerone, RJ. and R.S. Oremland. 1988. "Biogeochemical Aspects of Atmospheric Methane,"
Global Biogeochemical Cycles, vol. 2, p. 299-327. December 1988.
Crutzen, P.J. 1991. "Methane's Sinks and Sources" Nature No. 350. April 1991.
Hogan, K.B., J.S. Hoffman, and A.M. Thompson "Methane on the Greenhouse Agenda"
Nature Vol.354. November 21, 1991.
IPCC (Intergovernmental Panel on Climate Change). 1990a. Climate Change: The IPCC
Scientific Assessment. Report Prepared for Intergovernmental Panel on Climate
Change by Working Group 1.
IPCC (Intergovernmental Panel on Climate Change). 1990b. Methane Emissions and
Opportunities for Control. Workshop Results of the IPCC Response Strategies
Working Group. September 1990.
IPCC (Intergovernmental Panel on Climate Change). 1992. C//mate Change 1992; The
Supplementary Report to the IPCC Scientific Assessment. Report prepared for
Intergovernmental Panel on Climate Change by Working Group 1.
Lashof, D. 1989. 'The Dynamic greenhouse: Feedback processes that may influence future
concentrations of atmospheric trace gases and climate change." Climate Change No.
14 pp. 213-242. As cited in IPCC 1990.
1-12
-------
Oak Ridge. 1990. Oak Ridge National Laboratory/The Carbon Dioxide Information Analysis
Center. 1990. Trends '90. U.S. Department of Energy, Atmospheric and Climate
Research Division. Oak Ridge, Tennessee.
Oak Ridge. 1992. National Laboratory/The Carbon Dioxide Information Analysis Center. 1992.
Trends '91. U.S. Department of Energy, Atmospheric and Climate Research Division.
Oak Ridge, Tennessee.
Orlich, J. 1990. "Methane Emissions from Landfill Sites and Waste Water Lagoons." Paper
presented at the International Workshop on Methane Emissions from Natural Gas
Systems, Coal Mining, and Waste Management Systems. Workshop sponsored by
Environment Agency of Japan, U.S. Agency for International Development, and U.S.
Environmental Protection Agency. Washington DC. April, 1990.
Rinsland, C.P, J.S. Levine, and T. Miles. 1985. "Concentration of methane in the troposphere
deducted from 1951 infrared solar spectra." Nature No. 330 pp. 245-249. As cited in
IPCC (1990a).
Steele, L.P., E.J. Dlugokencky, P.M Lang, P.P Tans, R.C. Margin, and K.A. Masarie. 1992.
"Slowing down of the global accumulation of atmospheric methane during the 1980s."
Nature. Volume 358. July 23, 1992.
USEPA (U.S. Environmental Protection Agency). 1989. Policy Options for Stabilizing Global
Climate, Report to Congress. Office of Policy, Planning, and Evaluation. Washington,
D.C. 21P-2003.1. December 1990.
WMO (World Meteorological Institute). 1990. Scientific Assessment of Stratospheric Ozone:
1989. World Meteorological Organization Global Ozone Research and Monitoring
Project - Report No. 20. Geneva, Switzerland.
Wuebbles, D.J. and J.S. Tamaresis. 1992. The Role of Methane in the Global Environment
Paper prepared for submittal to the NATO Advanced Research Workshop, NATO-ASI
Book: Atmospheric Methane Lawrence Livermore National Laboratory March 19,
1992.
1-13
-------
CHAPTER 2
METHANE EMISSIONS FROM THE NATURAL GAS SYSTEM
U.S. Methane Emissions
from All Sources
Landfills
Other
Sources
Domestic
L i vested.
CoaI Mining
Natural Gas Systems
Annual Natural Gas
Methane Emissions
Global Emissions U S Emissions
Emissions Summary
Source
Field Production
Processing
Storage and Injection/ Withdrawal
Transmission
Distribution
Engine Exhaust
Total
1 990 Emissions
(Tg)
0.69 - 1.82
0.04 - 0.27
0.01 - 0.06
0.59 - 2.06
0.17 - 0.75
0.27 - 0.64
2.18 - 4.26 1
Partially
Controllable
/
/
/
/
% of 1990
Marketed Production
0.29%
0.02%
0.01%
0.28%
0.09%
0.11%
0.80%
1 The uncertainty in the total is estimated assuming that some of the uncertainty for each source is independent.
Consequently, the range for the total is more narrow than the sum of the ranges for the individual sources.
2.1 EMISSIONS SUMMARY
As the principal component of natural gas, methane is emitted from a wide variety of
components, processes, and activities that make up the U.S. natural gas system. The natural
gas system is large and complex, involving a myriad of activities, from gas production,
processing and transmission to distribution to residential, commercial and industrial
customers. Over a million miles of pipeline and thousands of facilities are operated and
maintained on an ongoing basis to supply and distribute this fuel.
2-1
-------
Over the past several years, several studies have been performed that provide a basis
for estimating methane emissions from the natural gas system. These studies include:
engineering analyses of model facilities; case studies of facility operations; detailed studies of
gas reported as "unaccounted for" by two major distribution systems; systematic
measurements of fugitive emissions from oil and gas production and processing facilities; and
a limited number of measurements from distribution system components. The estimates
presented in this study are based on the results of these analyses and previously-collected
data. To continue improving the basis for estimating methane emissions from this source, the
U.S. EPA in cooperation with the Gas Research Institute (GRI) has undertaken a research
program to collect additional emissions data. As the results of the EPA/GRI program becomes
available, the estimates of emissions from the natural gas system will be revised.
Based on the data available, methane emissions from the U.S. natural gas system in
1990 are estimated to range from 2.2 to 4.3 Tg/yr with a central estimate of about 3.0 Tg/yr.
These emissions are about 10 percent of total U.S. methane emissions, and are about
9 percent of the 25 to 50 Tg/yr global emissions from this source (IPCC, 1992). In 1990 the
emissions are estimated to have been less than about 1.5 percent of total marketed natural
gas. These estimates do not include emissions from oil production, processing, and
distribution, which are estimated at less than 0.6 Tg/yr in Chapter 7. Additionally, these
emissions estimates do not include combustion-related emissions associated with gas use by
customers, which are also estimated in Chapter 7.
To estimate methane emissions, the
U.S. natural gas system was divided into
the following major stages.
Field Production where raw
gas is withdrawn from
underground formations
using wells.
Studies performed during the past
several years provide a basis for
estimating methane emissions from the
U.S. natural gas system. Because more
work is warranted in some areas, a joint
EPA/GRI research program is collecting
data to improve the estimates.
Processing Plants where
constituents in raw gas are
removed (such as water, acid gas (CO2 and H2S) and non-methane
hydrocarbons) to upgrade the gas to pipeline quality specifications.
Storage and Injection/Withdrawal Facilities where processed gas is injected
and stored in underground formations, and subsequently withdrawn during
periods of high demand.
Transmission Facilities through which gas is transported long distances using
large diameter high pressure pipeline. Compressor engines are used to
pressurize the gas in the pipelines.
Distribution Facilities through which gas is delivered to customers at low
pressures using small diameter pipeline.
Compressor Engines are used throughout the gas system. Large reciprocating
and turbine compressors are used in the transmission stage. Reciprocating
compressor engines are also used in the production and processing stages.
2-2
-------
From each of these stages, methane emissions result from:
• Normal operations including compressor exhaust emissions, emissions from
pneumatic devices, and fugitive emissions (i.e., small chronic leaks from
components designed to store or convey gas and liquids);
• Routine maintenance including equipment blowdown and venting, well
workovers, and scraper (pigging) operations; and
• System upsets including emissions due to sudden, unplanned pressure
changes or mishaps.
Fugitive emissions across all stages are estimated to be the largest individual source
of emissions, accounting for about 38 percent of the estimated total. These emissions
originate throughout the entire system. Emissions associated with pneumatic devices are the
second largest individual source, accounting for approximately 20 percent of the total
estimated emissions. Pneumatic devices, used primarily in the production and transmission
stages, are designed to release small amounts of gas as part of their normal function.
Engine exhaust is the third largest
source of emissions. As part of their
normal operation, gas fired reciprocating
and turbine engines emit methane in their
exhaust gas. Together, fugitive emissions,
pneumatic devices and engine exhaust
account for nearly 75 percent of total
estimated emissions. Emissions from
routine maintenance activities and system
upsets are estimated to be relatively minor.
In 1990, emissions are estimated to be
less than 1.5 percent of gas marketed in
the U.S. In the future, emissions will
likely grow more slowly than gas
consumption. Additional research is
ongoing to resolve uncertainties in future
emissions rates.
Natural gas usage could increase significantly during the next decade, raising the
potential for increases in methane emissions to the atmosphere. However, improvements to
new and existing facilities and equipment (e.g., newer piping less susceptible to corrosion
and leaks and changes in operating practices) are expected to result in a reduction in the
portion of natural gas that is released as methane to the atmosphere. Taking into account
these changes in practices, emissions in the years 2000 and 2010 are estimated to be 2.4 to
5.0 Tg/yr and 2.4 to 5.4 Tg/yr, respectively. These emissions estimates reflect the potential
for gas consumption to increase from 16.8 Tcf in 1990 to about 20 to 24.5 Tcf by 2010.
Emissions are expected to increase more slowly than the rate of increased
consumption of gas. In 2000 and 2010, emissions are estimated at about 0.7 percent of
estimated total marketed production. Although this assessment does not consider that
certain types of emissions (e.g., emissions from system upsets) may become more frequent if
the system operates at closer to maximum capacity in the future, this omission is not
expected to bias the projection of future emissions significantly because emissions from such
sources are currently very small. Nevertheless, additional research would be useful for
quantifying these potential increases in emissions frequency.
These emissions estimates reflect a variety of uncertainties. Several of the emissions
estimates are based on engineering analyses of model facilities or case studies of operating
practices at several facilities. The representativeness of the facilities analyzed is difficult to
2-3
-------
assess. In cases where information was limited or unavailable, conservative emissions
estimates were used. Estimates of future emissions are limited by the uncertainty in future
gas demand and supply, as well as uncertainty in how emissions will change with changes in
demand and supply.
To improve the basis for estimating emissions, the EPA/GRI research program is
developing a database describing gas system activities and components. Emissions factors
for these activities and components are being developed by conducting emissions
measurements at selected representative facilities. The largest and most uncertain sources ol
emissions are being targeted in the program. In the preparation of this study, preliminary
results from the EPA/GRI program were used to estimate emissions from some sources.
Overall, the emissions estimates presented here are consistent with the latest information
available from that effort.
2.2 BACKGROUND
2.2.1 Stages of the Natural Gas System
Methane is emitted from a wide variety of sources and activities throughout the U.S.
natural gas system. The system itself is large and complex, encompassing hundreds of
thousands of wells, hundreds of processing facilities, hundreds of thousands of miles of
transmission pipelines, and over a million miles of distribution pipeline. Exhibit 2-1 presents a
schematic of the major stages of the natural gas system from production in the field to final
distribution to the consumer. Methane emissions from each stage are driven by factors that
are specific to the stage. Each stage is discussed in turn.
Field Production Facilities
Natural gas is produced in both gas and oil fields. Production facilities that produce
natural gas for commercial use can be categorized into three types of fields:
• Gas Fields with Gas Processing Plants. These fields withdraw gas from
underground formations. The gas, which is generally mixed with water and
other hydrocarbons (called condensate), is delivered to a centrally located gas
processing plant via gathering lines. The processing plant separates the
natural gas from the water, acid gas and other hydrocarbons and then typically
feeds the gas into the transmission system.
• Oil Fields with Gas Utilization Systems. These fields withdraw oil and casing
head gas from underground formations. The produced oil and gas are
generally mixed with water. The oil, water, and gas are usually separated in
the production field. The gas is delivered to a centrally located gas processing
plant via gathering lines. The processing plant separates the natural gas from
the other hydrocarbons and then typically feeds the gas into the transmission
system.
• Gas Fields without Gas Processing Plants. In select cases, gas fields produce;
gas that is of sufficient quality that it requires little or no processing prior to
injection into the transmission system. In these cases the gas processing step
is virtually eliminated.
2-4
-------
Exhibit 2-1
The U.S. Natural Gas System
Production
Transmission
Storage
Gathering
essing
Distribution
JU
Residential Commercial Industrial
Electric
Utility
2-5
-------
Gas is also withdrawn from oil fields that do not contain facilities for gathering or
processing the gas. In these circumstances the gas may be re-injected into the ground,
vented or flared. Because these fields do not produce gas for commercial use, they are not
included as part of the natural gas system covered here.
There were approximately 269,790 gas wells and 600,343 oil wells in the U.S. in 1990.
Total gross withdrawals of gas from these wells was about 21.5 Tcf (1 Tcf=1 trillion standard
cubic feet), 75 percent from gas wells and 25 percent from oil wells. Of this amount, about
2.5 Tcf was used for repressuring,1 and 0.4 Tcf was associated with non-hydrocarbon gases
or was vented or flared. Therefore, total marketed production of natural gas was about 18.6
Tcf (DOE, 1991 a). As shown in Exhibit 2-2, three states, Texas, Louisiana, and Oklahoma,
had combined gross withdrawals of 14.4 Tcf of gas, or nearly 70 percent of the national total.
These three states also accounted for 75 percent of total marketed production in 1990.
Natural Gas Processing Plants
Natural gas is usually processed in gas plants to produce products with specific
characteristics. Depending on the composition of the unprocessed gas, it is dried and a
variety of processes may be used to remove most of the heavier hydrocarbons, or
condensate, from the gas. The processed gas is then injected into the natural gas
transmission system and the heavier hydrocarbons are marketed separately.
As shown in Exhibit 2-3, total annual throughput at gas processing plants in the U.S.
was 14.6 Tcf in 1990. Plant capacity, however, was 24.9 Tcf, indicating that processing plants
in the U.S. were operating at approximately 59 percent of overall capacity.
Storage and Injection I Withdrawal Facilities
During periods of low gas demand (i.e., during summer in most areas), natural gas is
injected into underground storage reservoirs. These reservoirs are often depleted oil and gas
fields. During periods of high demand (i.e., winter in most areas) the gas is withdrawn from
the reservoir, processed if needed, and sent into the distribution network. These
injection/withdrawal facilities are generally referred to as "peak shaving" facilities because they
help to reduce the peak demand on the transmission system during periods of high demand.
The storage and injection/withdrawal facilities include a variety of processes and
equipment, including: compressors (to inject the gas into the ground); injection/withdrawal
wells; and separators and dehydrators to process the gas when it is withdrawn. Total
injection into storage was about 2.5 Tcf in 1990 (DOE, 1991 a). As discussed above, gas is
also re-injected for repressuring purposes in oil fields. This re-injection activity is not included
as part of the natural gas system.
Transmission Facilities
Transmission facilities are high pressure lines that transport gas from production fields,
processing plants, storage facilities, and other sources of supply over long distances to
distribution centers, or large volume customers. The U.S. natural gas transmission pipeline
1 The Prudhoe Bay oil field in Alaska accounted for the majority of the gas used for repressuring: nearly 2 tcf in
1988. There are no pipelines to transport this gas to the market.
2-6
-------
Exhibit 2-2
1 990 Natural Gas Production in the U.S.
(Tcf)
Location/State
California
Colorado
Louisiana
Oklahoma
Texas
All Others
TOTAL
Total
Gross
Withdrawals
0.446
0.269
5.304
2.258
6.907
6.306
21.490
Marketed
Production
0.363
0.243
5.242
2.258
6.343
4.113
18.562
Source: DOE (1991 a)
Exhibit 2-3
Throughput
Location
Louisiana
Texas
Oklahoma
All Other States
TOTAL
Sources:
a OGJ (1991)
b DOE (1991 a)
at U.S. Gas
Operating
Plants
1990"
75
301
101
257
734
Processing
Capacity
(Tcf/yr)
1990a
6.8
6.0
1.7
10.4
24.9
Plants
1990
Throughput
(Tcf/yr)b
4.2
3.9
1.1
5.4
14.6
network in 1990 was approximately 280,100 miles long (AGA, 1991b), connecting all states
except Alaska, Hawaii, and Vermont.2 In addition, some 89,500 miles of field and gathering
lines transport gas from individual wells to compressor stations, processing points, or main
Vermont receives gas via a pipeline from Canada. There is no gas pipeline to transport Alaskan gas to the
market.
2-7
-------
trunk pipelines. Natural gas flows primarily northeastward through this system from the major
production areas in Texas, Oklahoma, and Louisiana (DOE, 1991 a).
Transmission lines are usually buried. A variety of surface facilities support the overall
system including metering stations, maintenance facilities, and compressor stations located
along the pipeline routes. The pressure varies between systems depending on the grade of
steel, size of the pipe, and amount of gas transported.
Compressor stations are vital facilities in this stage as well as elsewhere in the natural
gas system. They include upstream scrubbers where the incoming gas is cleaned of particles;
and liquids before entering the compressors. Reciprocating engines and turbines are used to
drive the compressors which pressurize the gas. Compressor stations normally use pipeline
gas to fuel the compressors. They also use the gas to fuel electric power generators to meet
the station's electricity requirements.
Distribution Systems
Approximately 1.31 million miles of natural gas pipeline are used to distribute
processed natural gas to customers (AGA, 1991b). Distribution pipelines are extensive local
networks of generally small diameter, low pressure pipelines. Gas enters distribution
networks from transmission systems at "gate stations" where the pressure is reduced for
distribution within cities or towns.
Not all gas flows through the distribution system. Of the marketed production of 18.6
Tcf and net imports and supplemental supplies of 1.7 Tcf in 1990, about 16.8 Tcf was
delivered to customers via the distribution network. The difference of 3.5 Tcf (18.6 + 1.7 -
16.8) was associated with extraction losses3 (0.8 Tcf), lease and plant fuel use (1.2 Tcf),
pipeline fuel use (0.7 Tcf) and unaccounted for losses (0.8 Tcf). Of the natural gas delivered
to customers, 26 percent went to residential customers, 16 percent to commercial customers,
41 percent to industry, 17 percent to electric utilities (DOE, 1991 a).
For purposes of this analysis, the distribution system includes all facilities up to and
including the point at which gas is transferred to customers. With the exception of emissions
from natural gas fueled compressors used throughout the natural gas system, emissions
associated with the combustion of natural gas by customers are discussed separately in
Chapter 7. AH other emissions from customers' piping and equipment, such as fugitive
emissions, are assumed to be negligible.
2.2.2 Sources of Methane Emissions in the Natural Gas System
Methane emissions to the atmosphere from natural gas systems result from:
• Normal operations;
• Routine maintenance; and
• System upsets.
3 Extraction losses refer to the removal of liquids from the gas stream, and are not gas emissions.
2-8
-------
Each of these categories is described in turn.
Normal Operations
Normal operations are the day-to-day operations of a facility absent the occurrence of
abnormal conditions. Facilities emit methane during normal operations due to a wide variety
of operating practices and factors, including:
• Emissions from exhaust of reciprocating engines and turbines due to
incomplete combustion of natural gas used as fuel.
• Emissions from starting and stopping reciprocating engines and turbines.
• Emissions from pneumatic devices (gas-operated controls such as valves and
actuators). These emissions depend on the size, type, and age of the devices,
the frequency of their operation, and the quality of their maintenance.
• "Fugitive" emissions from system components. These emissions are
unintentional and usually continuous releases associated with leaks from the
failure of a seal or the development of a flaw, crack or hole in a component
designed to contain or convey gas. Connections, valves, flanges, instruments,
and compressor shafts can develop leaks from flawed or worn seals, while
pipelines can develop leaks from cracks or from corrosion.
Routine Maintenance
Routine maintenance includes regular and periodic activities performed in the
operation of the facility. These activities may be conducted frequently, such as launching and
receiving scrapers (pigs) in a pipeline, or infrequently, such as evacuation of pipes
("blowdown") for periodic testing or repair. In each case, the required procedures are
expected to release a particular amount of gas from the affected equipment. Releases also
occur during maintenance of wells ("well workovers") and during replacement or maintenance
of fittings.
System Upsets
System upsets are unplanned events in the system, the most common of which is a
sudden pressure surge. The potential for unplanned pressure surges is considered during
facility design, and facilities are provided with pressure relief systems to protect the
equipment from damage due to the increased pressure.
Relief systems vary in design. In some cases, gases released through relief valves
may be collected and transported to a flare for combustion or re-compressed and re-injected
into the system. In these cases, methane emissions associated with pressure relief events
will be small. In older facilities, relief systems may vent gases directly into the atmosphere or
may send gases to flare systems where complete combustion may not be achieved.
The frequency of system upsets varies with the facility design and operating practices.
In particular, facilities operating well below capacity are less likely to experience system
upsets and related emissions.
2-9
-------
Emissions associated with accidents
are also included under the category of
upsets. Occasionally, transmission and
distribution pipelines are accidentally
ruptured by construction equipment or other
activities. These ruptures not only result in
methane emission, but they can be
extremely hazardous as well.
Methane emissions to the atmosphere
from natural gas systems result from
normal operations, routine
maintenance, and system upsets.
Emissions may be continuous (e.g.,
fugitive emissions), periodic (e.g., from
routine maintenance), or irregular (e.g.,
from system upsets).
2.3 METHODOLOGY
2.3.1 Background
Although leakage from the natural gas system is addressed through strict safety
requirements, emissions rates have not been studied extensively or quantified. In particular,
when a leak is identified (e.g., through normal leak detection programs), emphasis is placed
on eliminating potential hazards, and the size of the leak is generally not estimated. Similarly,
methane emissions in engine exhaust or associated with the operation of pneumatic devices
have not been a concern, and consequently have not been quantified. Therefore, very little
emissions data have been developed for most of the emissions sources in the natural gas
system.
Because of this lack of emissions data, past studies such as Hitchcock and Wechsler
(1972), Abrahamson (1989) and Cicerone and Oremland (1988) have approximated methane
emissions from the natural gas system using estimates of "unaccounted for" gas, which is
defined as the difference between gas production and gas sales on an annual basis. Using
this aggregate approach, methane emissions have been estimated to be on the order of one
to three percent of annual gas consumption.
However, the applicability of "unaccounted for" gas estimates is very limited because
factors other than emissions account for the majority of the gas listed as "unaccounted for,"
including: meter inaccuracies, use of gas within the system itself, theft of gas (PG&E, 1990),
variations in temperature and pressure and differences in billing cycles and accounting
procedures between companies receiving and delivering the gas (INGAA, 1989).
Furthermore, because known releases of gas are not reflected in "unaccounted for" gas
estimates, such as emissions from compressor exhaust, the "unaccounted for" gas estimates;
cannot unambiguously be considered an upper or lower bound on emissions. Nevertheless,
the most comprehensive study of "unaccounted for" gas performed to date found that
emissions contribute only a small portion to the total unaccounted for quantity for a large
transmission and distribution system (PG&E, 1991).
There have been surveys that have used more reliable calculations of methane loss
instead of the "unaccounted for gas accounts." INGAA (1989) estimated that total U.S.
interstate pipeline methane loss was 0.13 percent of gas throughput transmitted by pipelines;
in 1988. This estimate was based on calculations of methane loss made by individual
pipelines in the categories of construction, transmission, pipeline affiliated production and
gathering, maintenance and storage. AGA (1989) estimated that 0.3 percent of total gas
consumed in the U.S. was emitted during transmission and distribution operations in 1988.
2-10
-------
For this study, a dis-aggregated
approach was taken in which each
emissions type is estimated for each stage
of the natural gas system based on a
working knowledge of operations and
events for each stage. Several recent
studies provide a basis for making these
estimates: Tilkicioglu and Winters (1989);
The reported quantity of "unaccounted
for" gas is not a good estimate of
emissions. This study is based on a dis-
aggregated assessment of emissions
based on a working knowledge of the
operations and emissions events for
each stage of the industry.
Tilkicioglu (1990); PG&E (1990) and SOCAL
(1992). In several important areas
additional analyses have recently been
prepared (Gibbs et al.. 1992; Kolb et al., 1992; and Radian, 1992a). Additional research is
ongoing to provide more data for estimating these emissions.
2.3.2 Steps Used to Estimate Emissions
Using the latest available data, the following general approach was used to estimate
emissions from each stage of the natural gas system:
1. One or more "model facilities" were defined for each stage. The model facilities
were selected on the basis of industry experience to be representative of the
diverse set of facilities in the U.S. system.
2. Each emissions type, except compressor engine exhaust, was estimated for
each model facility based on detailed data describing the facility and the
processes that lead to emissions.
3. Emissions factors for each model facility were estimated by dividing the
estimated emissions by an appropriate measure of the facility's size, such as
throughput in cubic feet per year or miles of pipeline. These emissions factors
then can be expressed as emissions per bcf of throughput per year, emissions
per well or emissions per mile of pipeline per year.
4. Average emissions factors were estimated for each stage by averaging the
emissions factors estimated for each of the model facilities in that stage.
5. National emissions were estimated by multiplying the average emissions
factors for each stage by the total applicable size of the national system, such
as bcf of throughput, number of wells, or miles of pipeline.
To estimate emissions from compressor engine exhaust, fuel use was estimated for
each stage by type of compressor: reciprocating or turbine. Emissions factors (i.e.,
emissions per amount of fuel used) were multiplied by fuel use to estimate emissions for each
engine type within each stage.
The accuracy of this general approach relies heavily on the representativeness of the
model facilities. Generally, model facilities were selected on the following criteria:
• Accessibility: for data collection and interviews with operators.
2-11
-------
• Size and process type: for the selection of the most common sizes and
process types found in the U.S.
• Location: for selection of the facilities in the regions that contribute a high
percentage to natural gas production in the U.S.
Age: for selection of the facilities which represent the "average age" within the
U.S.
Given the diversity of facilities and operating conditions encountered in the U.S. natural gas
system, and the relatively small number of model facilities studied to date, it is likely that
some situations are not adequately represented in the available data. Whether gaps in the
representativeness of the data cause the estimates to be biased upward or downward is not
evident. However, it appears that the available data provide a reasonable assessment of the
magnitude of emissions. The following sections describe the methods used to estimate each
type of emission.
Normal Operations
Normal operations emissions primarily result from pneumatic devices, the incomplete!
combustion of natural gas in reciprocating engines and turbines, the venting of natural gas
during engine starts and stops, and fugitive leaks.
Pneumatic Devices. Pneumatic devices are commonly used to regulate and control
gas pressures and flows throughout the natural gas system. These devices rely on
pressurized gas as an energy source for their operation. For example, the pressurized gas
may be used to maintain the position of an actuator. In most cases, the natural gas stream
itself is a suitable supply of pressurized gas, because by using the gas stream, a separate
source of pressurized gas is not needed. Emissions from a pneumatic device result when the
device is designed to release the pressurized gas it uses to the atmosphere.
Emissions are a function of the design and size of the device, the frequency of its
operation, and its age and state of repair. Emissions estimates for field production,
processing, and injection/withdrawal facilities were taken from Tilkicioglu (1990) and Radian
(1992a). Emissions from pneumatic devices on transmission systems were estimated using
measured rates from the PG&E (1990) and SOCAL (1992) unaccounted-for-gas studies.
Distribution systems were assumed to have negligible emissions in this category.
Engine Exhaust. Reciprocating engines and turbines are used throughout the natural
gas system to compress gas, generate electricity and perform other functions (such as pump
water). The exhaust from these engines is known to contain methane.
Total compressor engine exhaust is calculated separately for all stages by multiplying
the emissions factors for reciprocating engines and turbines by the corresponding estimates
of annual fuel use. Tilkicioglu (1990) reports an emissions factor of 508 kg of methane per
million of cubic feet (MMcf) of fuel used in reciprocating compressor engines (1,120 pounds
2-12
-------
per MMcf). EPA (1985) reports a methane emissions factor of approximately 587 kg/MMcf.4
A recent compilation of data on the composition of compressor exhaust for a joint EPA/GRI
study of methane emissions indicates that emissions are on the order of 513 kg/MMcf when
the available data from individual compressors are weighted by actual fuel usage (Campbell,
1991). Therefore, a representative emissions factor of 510 kg/MMcf is adopted here for
estimating these emissions.
Turbine engines exhaust much less methane per MMcf of fuel used than reciprocating
engines. Tilkicioglu uses an emissions factor of 11.8 kg/MMcf (26 pounds/MMcf). EPA
(1985) reports a value of 9.7 kg/MMcf and Campbell (1991) reports a value of 6.1 kg/MMcf.
For this study, an emissions factor of 9.0 kg/MMcf is used, which is roughly the average of
these reported values.
Of note is that natural gas is also used in external combustion equipment, such as
heaters, as part of the natural gas system. The methane emissions from these combustion
sources are considered as part of the combustion-related emissions presented in Chapter 7.
Engine Starts and Stops. In the process of starting and stopping reciprocating
engines and turbines, natural gas is generally vented from the equipment. The quantity of
gas vented is a function of the internal volume of the engine and the number of starts and
stops conducted annually. Generally, turbines are operated almost continuously, so that very
few starts and stops are conducted. Also, in some cases pressurized air is used to assist in
starting turbines, so that less gas is vented during engine starts in these cases.
Equipment Venting. Glycol dehydrators are the principal source of equipment venting
emissions. These dehydrators are used to remove water from natural gas through
continuous glycol absorption. Other compounds in the gas are also absorbed including
methane. The water-rich glycol is regenerated with heat, which drives the water out of the
glycol. The methane present in the glycol is driven out with the water in this process and is
vented to the atmosphere. Estimates of emissions resulting from the venting of glycol
dehydrators for production, processing and transmission were taken from Radian (1992a).
Fugitive Emissions. Fugitive emissions occur in all stages of the natural gas industry.
For production, processing, and storage facilities, fugitive emissions are primarily a function of
the number of components (e.g., connections and valves) installed and are estimated by:
• multiplying emissions factors by the number of installed components at model
facilities; and
• scaling the model facility estimates to the industry as a whole.
The model facilities were defined in Tilkicioglu and Winters (1989) and the emissions factors
were taken from Rockwell (1980), which quantified fugitive emissions at 11 U.S. oil and gas
facilities, including two gas processing plants.
4 EPA (1985) reports that hydrocarbon emissions from reciprocating natural gas pipeline compressor engines
are 1,400 pounds per MMcf of fuel used. This amount is reportedly 90 to 95 percent methane. Assuming a value of
92.5 percent, the emissions factor for methane is computed as: 1,400 x 0.925 = 1,295 pounds per MMcf, or about
587 kg per MMcf. Note that the EPA (1985) lists the emissions in units of carbon; however, Charles Urban (1992)
corrected EPA (1985), reporting that the emissions factors were estimated in units of methane.
2-13
-------
Although the Rockwell (1980) emissions factors were originally developed for total
hydrocarbons as opposed to methane, a recent re-analysis of the raw data from the study to
develop methane emissions factors indicates that applying the emissions factors to the
facilities defined in Tilkicioglu and Winters produces consistent estimates of fugitive methane
emissions (Gibbs et al.. 1992). Recently, however, an ongoing analysis for the American
Petroleum Institute (API) has indicated that fugitive emissions rates at oil and gas production
and processing facilities have declined since the Rockwell study was performed. Therefore,
the fugitive emissions factors were adjusted to reflect the lower rates found in the ongoing
API study.
Transmission pipelines also are known to have periodic fugitive leaks. These
emissions are generally short-lived due to the leak survey program and weekly air patrols
conducted by the pipeline. An emissions factor per mile of pipeline was estimated based on
the PG&E estimates of these emissions from the transmission portion of their pipeline system.
Fugitive emissions from distribution systems result from small chronic leaks in:
(1) buried pipelines, for example due to corrosion or leaking pipe joints; and (2) non-buried
facilities, such as pressure regulating equipment. Estimates for buried pipelines are based on
leak rates reported in PG&E (1990) and SOCAL (1992). Both PG&E and Southern California
Gas (SOCAL) undertook programs specifically to measure the rate of leaks detected in their
underground systems. A total of 20 leak rate measurements were conducted by PG&E, the
results of which were extrapolated to the entire PG&E system based on the number of leaks
reported in a year. SOCAL measured 40 leaks in their distribution system. Preliminary results
from that study are very similar to the results of the PG&E study. Consequently, the results of
the two studies were combined to estimate fugitive emissions from distribution pipeline leaks.
Fugitive emissions from non-buried distribution system facilities are based on initial
measurements from the EPA/GRI research program. These data consist of measurements at
28 distribution system gate stations and pressure regulating stations, as reported in Kolb
etal. (1992).
Other Miscellaneous Emissions. Most other normal operation emissions sources are
considered to be negligible. However, Tilkicioglu and Winters also estimated emissions from
gas vented at drip points along the transmission system. These are points in the pipeline
where accumulated liquids are periodically removed. The emissions factor for drip points was
estimated from data reported by PG&E in terms of the volume emitted per mile.
Routine Maintenance
Emissions from routine maintenance include emissions from maintenance of gathering
pipelines, well workovers, orifice fittings, station blowdowns, transmission station shutdown:;,
pipeline repair, and compressor blowdowns. For each of these, emissions result when
facilities or equipment are opened to the atmosphere.
In general, routine maintenance emissions are estimated by multiplying the frequency
with which the maintenance activity is performed by the quantity released per activity. The
frequency with which activities are performed was estimated based on interviews with station
operators and operational records reviewed at the model facilities. While some practices vary
across facilities, most maintenance activities are fairly standard. Therefore, the frequency
estimates are believed to be reliable even though only a small number of model facilities was
examined.
2-14
-------
The emissions per activity were estimated based on the physical dimensions of the
model facilities. Examples of these emissions estimates include:
• Well Workovers: The number of reported well workovers per year is multiplied
by an estimated leak rate during the workover. The leak rate is estimated
assuming that during well workovers, wells are open to the atmosphere for 48
hours, and two cubic feet of gas are emitted each hour.
Orifice Fitting Replacement: Each time orifice fittings are replaced on a well, a
20 foot long by 2 inch pipe is vented. Junior fittings, with a pressure of 30 PSI
are replaced twice a year. Low pressure senior fittings (30 PSI) and high
pressure (300 PSI) senior fittings are replaced once a year.
• Maintenance Slowdowns: Field production station blowdowns generally occur
on an as-needed basis, but it is assumed that they occur once a year resulting
in an emission equal to the station volume. Similarly, transmission stations are
shutdown once a year, and it is assumed conservatively that the entire station
is vented out.
Transmission pipeline repair: The information used to calculate emissions
includes the length of pipeline section (15 miles), frequency of repair, and
pressure of gas emitted. The emission due to pipeline repair from the entire
representative pipeline is the emission rate from the sample segment multiplied
by the number of segments in the pipeline.
System Upsets
Methane emissions from system upsets include emergency blowdowns, station
shutdowns, and accidental pipeline ruptures called "dig-ins." Information used to calculate
these emissions include interviews with plant operators and facility logs. Because upsets are
unplanned, these estimates are more uncertain than regularly scheduled routine maintenance
emissions or normal operations emissions. In general, the volume emitted is calculated from
the number of annual upsets and the volume of vented facilities and equipment.
Data reported in PG&E (1990) were used to estimate emissions factors for dig-ins on
transmission systems and distribution systems. Because transmission systems are generally
well marked, dig-ins are infrequent and do not contribute significantly to emissions. While
emissions from system upsets in distributions systems are generally considered small, dig-ins
are the primary source of these emissions.
Emissions also result from venting and flaring activities at production facilities.
Preliminary analyses indicate that most venting and flaring takes place at oil production
facilities that do not market gas. Consequently, these emissions are discussed in Chapter 7.
2-15
-------
2.4 CURRENT EMISSIONS
2.4.1 Field Production Facilities
Gas and Oil Wells
National emissions from gas and oil wells were estimated based on analyses of four
model facilities described in Tilkicioglu and Winters (1989) and Tilkicioglu (1990). Only oil
wells that produce gas for sale commercially are included (those that do not are included in
Chapter 7, Petroleum Production).
• Facility 1: Consists of 500 oil producing wells and has a natural gas field
capacity of 4.0 billion cubic feet per year (bcf/yr). Oil produced is pumped oul
of the wells and results in production of natural gas, which is fed to a low
pressure separation and gas transportation system.
Facility 2: Consists of approximately 300 oil producing and 21 gas producing
wells and has a natural gas field capacity of 6.8 bcf/yr. Separation facilities are
also included.
• Facility 3: Has a capacity of 116 bcf/yr and contains 6 high-flow wells and
separation equipment.
• Facility 4: Consists of 400 gas wells, which are divided into high pressure (225
wells) and low pressure (175 wells) systems, and has a natural gas field
capacity of 32.5 bcf/yr. The gas from these wells is of sufficient quality that it is
injected into a transmission pipeline with virtually no treatment.
Normal Operations: Fugitive emissions are the only normal operations emissions
estimated for gas and oil wells. These emissions were estimated for model Facility 3's gas
wells and related treatment equipment. The emissions from the gas wells were estimated at
6.34 Mg/yr, or about 1.06 Mg/well.5 Fugitive emissions from the treatment facility were
estimated at 14.36 Mg/yr per facility. For purposes of this analysis, it is assumed that one
model treatment facility is required per every six gas wells.
Estimates of fugitive emissions from gas-producing oil wells were not available. The
emissions per oil well were estimated based on the following:
• oil well/treatment facilities have about one-quarter the number of components
as gas well/treatment facilities (on a per-well basis); and
• the methane fraction of oil well streams is about one-third the value for gas well
streams.
Therefore, the fugitive emissions factor for gas-producing oil wells was estimated as 1/12 the
emissions factor used for gas well/treatment facilities. Taking into account the components in
both the wellhead and the separation facility, the resulting emissions factor for oil wells is
about 0.29 Mg/well.
5 Mg = megagram = 106 grams = 1,000 kilograms = 1 metric ton.
2-16
-------
These fugitive emissions factors are shown in Exhibit 2-4. However, preliminary results
from an ongoing study by API indicate that current emissions rates from U.S. oil and gas
facilities are only about 1/4 what they were in 1980 when the measurements were conducted
that underlay the fugitive emissions factors used to develop these estimates (Webb, 1992).
Consequently, the fugitive emissions factors listed in Exhibit 2-4 are believed to be too high
by a factor of 4, and are adjusted downward to estimate national emissions. These adjusted
emissions factors are shown in Exhibit 2-4 as the "Revised" Average Emissions Factors.
Routine Maintenance: Routine maintenance emissions were estimated at the model
Facilities 1, 2, and 4. Emissions at Facilities 1 and 2 are attributable solely to well workovers
which are performed annually. Emissions due to routine maintenance at Facility 4 result from
annual well workovers and yearly compressor station blowdowns. The emissions rates and
emissions factors for all three facilities are presented in Exhibit 2-4.
System Upsets: System upsets were estimated at model Facilities 1, 2, and 4. At
Facility 1, emissions due to system upsets resulted from the emission of methane from crude
oil overflow tanks. System upsets at Facility 2 were considered negligible, and at Facility 4,
emissions from system upsets resulted from station blowdowns due to an emergency three
times a year. These emission rates and factors are summarized in Exhibit 2-4. The average
emissions factors across the model facilities are also shown in the exhibit.
Gathering Facilities
Data were not available describing emissions from gathering lines. Consequently, the
emissions characteristics of transmission lines were used. Although gathering and
transmissions lines are similar, gathering lines are often operated at lower pressures and may
be in a more corrosive environment. Emissions from transmission lines are described more
fully below (see section 2.4.4).
Normal operations and system upset emissions were estimated by Tilkicioglu and
Winters (1989) and Tilkicioglu (1990) for model transmission system facilities (section 2.4.4
describes the model facilities). Additionally, estimates for pneumatic devices and routine
maintenance emissions were made using the results from PG&E (1990) and SOCAL (1992).
The SOCAL study also reported an emissions factor for fugitive emissions from packing seals
not considered in the Tilkicioglu and Winters study.
Normal Operations: Normal operations emissions include fugitive emissions,
emissions from pneumatic devices, non-exhaust engine emissions, meter scrubber emissions,
and pipeline scraper emissions. Fugitive emissions were estimated by Tilkicioglu and Winters
(1989) and SOCAL (1992). Tilkicioglu and Winters estimated fugitive emissions based on
leakage rates for various component types, resulting in an estimate of 0.49 Mg/mile. SOCAL
estimated additional fugitive emissions from packing seals of 1.05 Mg/mile. Consequently,
the total fugitive emissions factor used to estimate national emissions is 1.54 Mg/mile
(Exhibit 2-5).
Emissions factors for non-exhaust emissions from compressor engines and scrubber
and scraper operation emissions are also summarized in Exhibit 2-4. These emissions factors
were developed by Tilkicioglu and Winters for three model systems. SOCAL (1992) and
PG&E (1990) report emissions factors for pneumatic devices along transmission systems
which vary by about one magnitude. The average emissions factor of 0.73 Mg/mile is used in
the analysis, and is shown in Exhibit 2-6.
2-17
-------
8
*^
'5
a
n
UK
c
o
'•s
3
•a
o
&_
Q.
'S
•a
o
T S
CM c
*- E
•~ o
fi 2
•c tl
X CO
III ^
IU o
«
u
CO
LL.
•a
CO
CO
a>
a
DC
C
o
'to
.£2
Ul
"Revised"
Average
Emissions
Factors0
w
B C „
5>-S S
§8.!
> "E «
< S "-
•^ nj
^
>.
ff
0
a
u.
p>
|
o
a
u.
CM
Jj
S
u.
^v
1
o
if
Ik.
!!
£ S.
si
£f
•c
^
o
o
•
... "£•
5 ^
« 0)
CC "B
2
c*
if
£f
o ^
S -c
_. 0)
i.
il
a o>
u_ 5
•*• «c
^^
« >.
5 ^>
s
Emissions Type
= | =
fi f
15 S P
0 CO 0
= "S ^
CO CO CO
§£ $
II 1
8 co S
•^ j d
-"^ ^^^^
f « f
|^ 1
it t
SB 18
>- * ci
«*.
o
d
|| Routine Maintenance
i
co
CO
t3
^
f
,
§
o
0
co
UJ
UJ
z
o
o
o>
CO
CO
in
CO
1
Q.
E
®
&
C/5
s.
7;
E
10
c\i
co
CO
"CD
CD
^
•5
£>
to
CD
5.
•s
.Q
•*T
Gas Production
t<
CO
CD
CD
W
>< -S1
CD TJ
•S 3
fl> "tfl
CD _
C/5 Q.
CO <
CD 0)
2 c
1 1
1 §
1 °
^* t/1
co ±:
CO =1
±= CO
<. o>
f~ c
o
CO "O
O) co
" CO
E £1
0 CO
•t c
CO O
C CO
O CO
"co 'c
CO C
._ j,
CO E CO
>. • .1
•±= CD S
= ^ ai
o •*-* ^
? i o
£ § ~
• 1 5
« I 1
~ to
« CD 3
CO -° f
>. s •
^ T3 0
0 « S
S | £
-------
1
0)
I
CO
.0
'Jo
0)
E
w
5
JL-
i-
•o
c
a
o>
^
a>
in a
CN O
•"d "SI
•s •§
£ O
i2 s
|
2
fN
§
(Q
U.
•a
(Q
0)
Q)
%
DC
(0
c
o
'55
(0
i
UJ
lisf
s 8*^
8 — * o>
<•§"• =
UJ ^*^
•o „
3 £ •. "•"
Sill
5 E"-S
sr ~
^
4)
&
«
|
CO
X
(0
CM
Q)
«
w"
^
8,
w
if
« 0)
i.
0 "t
™ -^
IT I
s_
k. "•"
II
"1
m "t
S ^
Sf
U. ^
'S'
»• M
IE
"I
0 ^
S •£•
B Dl
= E
"
di
li
*i
*^»
l«
= s
a>
Q.
£
(A
I
in
»
UJ
Ht- 3; 0)
S «H -r
^ O 0
•fe UJ LU
S Z Z
%> LU UJ
•* Z Z
0 ^
1
TS LU LU
S z z
^
g fe o
•^ o
LU "? _
2 fe °
$ ^ S
z do
C\J T~
s 2 s
g 8 8
Z d °
LU co °!
* ** $
CO C
c o 11-
.0 M 2
5 CO 3
co •= cd
8. - 1
2 § i c ^
i 1 tz |
o £ >" 5
z
o
d
:
UJ
LU
z
in
o
d
<£>
in
CM
d
i>^
9
S
d
CO
CO
co
£.
CD
CO
Q.
E
S
s.
CO
CO
£
E
§
co"
CO
5
E
o
T~
CO
_£
I
CM
CM
CM
CO
-5
£
CO
N
55
a>
"
CO
a
o
to
CO
a
"co
o
"&
£
Q.
•o
>. §
"i »-
* §
•d '5 co"
< S co
* 2 E ° 1
1 f « c 'i
1 1 £ I *
• . 0 S C
CO CO -^ CO 'C
! 3 S § 3
01 <° <8 co E
at a> E £ ^
c c s Q.
-------
Exhibit 2-6
Emission Rates and Factors for Gathering Pipeline
Emissions Type
Pneumatic Devices
Routine
Maintenance
Total Miles
PG&E Study
Rate
(Mg/yr)
4,290
3,630
Factor
(Mg/mile)
1.35
1.14
3,189 miles
SOCAL Study
Rate
(Mg/yr)
455
1,829
Factor
(Mg/mile)
0.12
0.46
3,964 miles
Average
Emissions
Factor
(Mg/mile)
0.73
0.80
Sources: PG&E (1990) and SOCAL (1992)
Routine Maintenance: Estimates of emissions factors for routine maintenance are
shown in Exhibit 2-6 based on the SOCAL and PG&E studies. Routine maintenance activities;
considered include routine purge and blowdown activities.
System Upsets: System upset emissions were estimated by Tilkicioglu and Winters tor
three model systems (Exhibit 2-5). System upset included an emergency system shutdown in
model systems 1 and 2 and a compressor blowdown in model system 3. These emissions
are relatively small compared to other sources in the gathering systems.
Dehydrator Vent Emissions
Radian (1992a) estimated dehydrator vent emissions based on measurements of four
dehydrators in Louisiana and a simulation of four representative dehydrator types.
Dehydrators remove moisture (water and liquid hydrocarbons) from the gas stream by
passing it through a drying medium, which is commonly glycol. The glycol is regenerated by
heating it and driving off the water and other hydrocarbons as vapor. In most cases this
vapor is vented to the atmosphere, although controls are under consideration to reduce
emissions of aromatic hydrocarbons from these vents.
The uncontrolled venting of the vapors during glycol regeneration also releases
methane because, like the other hydrocarbons, some methane is absorbed by the glycol
during the drying process. Based on their analysis, Radian developed and emissions factor
of 5.57 Mg/dehydrator, assuming an average dehydrator size of 1 MMcf per day.
Pneumatic Device Emissions from Heaters, Separators, and Dehydrators
Radian (1991 a) summarized available estimates of emissions from pneumatic devices
used on heaters, separators, and dehydrators in production fields. Based on PG&E (1990),
Radian extended the work of Tilkicioglu and Winters (1989). PG&E obtained pneumatic
device emissions rates from manufacturer's reports and laboratory testing of various devices.
Emissions factors for these facilities were estimated as follows:
2-20
-------
• Emissions from pneumatic devices:
Heaters: 1.64 Mg/heater/year;
Separators: 1.06 Mg/separator/year;
Dehydrators: 4.27 Mg/dehydrator/year.
Summary of Total Emissions from Field Production Facilities
The estimated total emissions from production facilities in the U.S. is 1.08 Tg. This
estimate is obtained by multiplying the average emissions factors by the estimates of the
applicable size of the U.S. system for this stage of the industry. The activity factors used are
as follows.
total gross withdrawals of 21,490 bcf (DOE, 1991 a) is used to estimate
emissions from routine maintenance and system upsets from gas and oil well
facilities;
• gas wells (269,790) (DOE, 1991 a), treatment facilities (1 per 6 gas wells), and
oil wells (288,165) are used to estimate fugitive emissions from gas and oil
wells;
• miles of gathering pipeline (89,500) (AGA, 1991 b) are used to estimate all
emissions from gathering facilities; and
the number of heaters, separators, and dehydrators (Cowgill, 1992) are used to
estimate the pneumatic device-related emissions and the dehydrator vent
emissions.
Within the production stage, fugitive
emissions and pneumatic devices
account for about 75 percent of the
1.1 Tg/yr emissions.
To summarize all the computations,
Exhibit 2-7 is divided into four sections,
reflecting the four types of facility sizes
used. The top half of the exhibit shows the
calculations for those emissions based on
gross withdrawals and the numbers of gas
and oil wells and treatment facilities. The
bottom half shows the calculations for the
gathering facilities and the heaters, separators, and dehydrators. The total for each section is
shown, along with the grand total of 1.08 Tg/yr at the bottom of the exhibit. Across the four
sections, fugitive emissions and emissions from pneumatic devices account for the majority of
the emissions, about 75 percent.
6 The number of oil wells producing gas were estimated as follows. In 1990 there were an estimated 600,343 oil
wells, which is the average reported for December 31, 1989 and 1990 (OGJ, 1990 and 1991). Of these, 48 percent
(288,165) are estimated to be producing gas for commercial sale based on the Radian (1992a) review of Texas oil
leases. By using this percentage, it is assumed that the number of wells is proportional to the number of leases.
2-21
-------
Exhibit 2-7
Summary of Total Emissions from Production Facilities
Emissions Type
Normal Operations
Fugitive Emissions
Wellheads
Treatment Facilities
Oil Wells
Routine Maintenance
Well Workovers
Systems Upsets
Oil overflow tanks and
Station Slowdown
Total
Gross Withdrawals of 21,490 bcf
Average
Emissions Factor
(Mg/bcf)
0.02
1.33
NA
Total
Emissions
(Tg/yr)
<0.001
0.03
0.03
269,790 Gas Wells
44,965 Treatment Facilities
288,165 Oil Wells
Average
Emissions Factor
(Mg/Well)
0.27
3.59
0.07
NA
Total
Emissions
0'g/yr)
0.07
0.16
0.02
0.25
Emissions Type
Normal Operations
Pneumatic Devices
Heaters
Separators
Gas Dehydrators
Gathering Pipeline
Dehydrator Vents
Fugitive Emissions
Gathering Pipeline (with
packina seals)
Engine - Other*
Other6
Routine Maintenance
Gathering Pipes (Blow & Purge)
System Upsets
Gathering Pipelines
Total
Total U.S. Emissions from
Production Facilities
Gathering Pipeline
89,500 miles
Average
Emissions Factor
(Mg/mile)
0.73
1.54°
0.24
0.18
0.80
0.1
NA
Total
Emissions
(Tg/yr)
0.07
0.14
0.02
0.02
0.07
0.01
0.33
Process Units
54,250 Heaters
180,653 Separators
19,776 Dehydrators
Average
Emissions Factor
(Mg/Unit)
1.64*
1.06^
4.27 d
5.57 e
NA
Total
Emissions
(Tg/y)
0.09
0.19
0.08
0.11
0.47
1.08 Tg/yr
a Emissions from compressor exhaust are estimated separately in section 2.4.6. See text.
b Includes scraper operations
c The emissions factor combines Tilkicioglu and Winters (1989) estimate of 0.49 Mg/mile with SOCAL (1992)
estimate of 1.05 Mg/mile for packing seals. See Exhibit 2-5.
d Source: Radian (1991 a)
e Source: Radian (1992a)
2-22
-------
2.4.2 Gas Processing Plants
With the exception of emissions from glycol dehydrator vents, emissions from gas
processing plants were estimated based on analyses of model plants by Tilkicioglu and
Winters (1989) and Tilkicioglu (1990). Two model plants were analyzed by Tilkicioglu (1990),
one in the Rocky Mountain region and the other in Western Texas (Plants 1 and 2,
respectively). The two plants utilize the cryogenic process, the most common processing
method in the U.S., and they have capacities of 14 and 10 bcf/yr, respectively. The Rocky
Mountain plant was operating at a low utilization factor in 1988 of 50 percent (7 bcf/yr
throughput), while the Western Texas plant had a higher percent utilization of 88 percent (8
bcf/yr throughput).
Fugitive emissions were estimated by Tilkicioglu and Winters (1989) at a gas
processing plant with a capacity of approximately 3.7 bcf/year (Plant 3). The emissions were
estimated on the basis of component counts (valves, flanges, connectors, pressure relief
devices) and emission factors for each component type from Rockwell (1980).
The emissions estimates for the model plants are as follows (see Exhibit 2-8):
• Normal Operations: Emissions from normal operations include engine exhaust,
fugitive emissions, uncombusted methane in inlet flares, compressor start/stop
emissions pneumatic devices, and glycol dehydrator venting. Analyses of
Plants 1 and 2 indicate that emissions from engine starts/stops ranged from 0.8
to 3.2 Mg/bcf of throughput. An average rate of 2.0 Mg/bcf is used in the
analysis. Fugitive emissions were estimated at Plant 3 to be about 20 Mg/bcf.
As with production field fugitive emissions, ongoing analyses by API indicate
that the Rockwell (1980) emissions factors overstate current emissions rates.
For gas plants current emissions rates are about 1/20 the rate derived from
Rockwell (1980) (Webb, 1992). Consequently, the fugitive emissions factor for
gas plants was adjusted to reflect these latest estimates. These adjusted
emissions factors are shown in Exhibit 2-8 as the "Revised" Average Emissions
Factors. Other normal operation emissions are relatively minor.
• Routine Maintenance: Emissions from routine maintenance were estimated at
Plants 1 and 2. Normally, the plants are shut down once every four years to
conduct general maintenance and emissions result because gas is flared
before the plants are shut down. As a conservative estimate, it was assumed
that compressor scrubber vessels were vented out in order to replace valves or
to inspect the vessels once per year.
• System Upsets: Emissions from system upsets were estimated at Plants 1 and
2. System upsets, which can produce significant emissions during high
utilization periods, were assumed to be absorbed within Plant 1 without
triggering the relief systems because of the low plant utilization factors. Even
at Plant 2 these emissions were relatively minor.
In addition to these emissions estimated at the model plants, Radian (1992a) estimates
an emissions factor for glycol dehydrators at gas processing plant. This emissions factor is
the same value used for dehydrators in production fields.
2-23
-------
(0
CO
DC
c
0
i
IU
* a> wo c->
flllli
F^^ "S-
• S-?
slit
tf E IT
^ III ""*
m
(0
a
CM
2
Q.
(8
a
^ c-
£ -a
a ^
£ 5
«=s
1}
u- S
•—.
c| "Si
2
3°>
"" L
„_ ^
*^ "^
(0 Q)
CC ^5
Emissions Type
T3
I1
S 1 | 1 |
!n
in
T3
CD
s - 1 1 i
*•— *
•°
tn
in
LU LU LU LU ^
Z Z Z Z g
LU LU UJ LU m-
Z Z Z Z *
CVI 0 LU * UJ
m Z Q Z
"°. LU C\J LU
S ° 2 ^ Z
CO III Lil
o ° Z ° Z
CM _^ LU ._ LU
in Z Z
CO CO
m o W n
1 1 1 1 1
Q CO ^2 W g g)
1 p § 1 I 1
0 UJ CL Q O U.
5
d
LU
z
LU
z
§
d
CD
d
f^
d
II
|| Routine Maintenance
0
o
o
o
UJ
z
UJ
z
o
o
d
o
o
fi
in
a
a>
-S
•5
o
>.
I
«"
^
1
00
«o
|| Throughput
a)
i
CO
f
B)
E
O)
CO C
TT 0
csi g>
c o
o ,_
CD ID
U) ^
£ CO
O
2 c
I "i
% i
1 8
to o
E u>
To ~
® C
1 1
(0 ^
1 o
*— •«—
o c
s 1
-------
Total methane emissions from processing plants in the U.S. are estimated at 0.09
Tg/yr in 1990 by multiplying the emissions factors by the total U.S. throughput and population
of dehydrators (Exhibit 2-9). Emissions from the venting of glycol dehydrators account for
most of the emissions (about 45 percent). Compressor starts/stops and fugitive emissions
were also estimated to be important, with the other sources being negligible.
Exhibit 2-9
Summary of Total Emissions From Processing Plants
TOTAL U.S. THROUGHPUT = 14,610 bcf/yr
Dehydrators = 6,603 a
Emissions Type
Normal Operations
Engines - Start/Stops
Pneumatic devices
Dehydrator Vents
Other b
Fugitive Emissions
Routine Maintenance
System Upsets
Average
Emissions
Factor
(Mg/bcf)
2.00
0.00
5.57
(Mg/dehydrator)
0.07
1.005C
0.10
0.01
Total U.S. Emissions from Processing
Total U.S.
Emissions
(Tg/yr)
0.03
0.00
0.04
0.001
0.015
0.001
0.0001
, $m
a Source: Cowgill (1992)
b Includes inlet flare activity. Emissions from compressor exhaust are
estimated separately in section 2.4.6.
c The emissions factor from Exhibit 2-6 has been adjusted to reflect the
latest results of the ongoing API analysis. See text.
2.4.3 Storage and Injection/Withdrawal Facilities
Tilkicioglu (1990) analyzed five model injection/withdrawal facilities. All five facilities
are located in California and serve as "peak-shaving" facilities for two major population
centers. The five plants are as follows:
• Plant 1: This facility operates 24 hours per day, 7 days a week. Natural gas is
received from the transmission system and injected into the underground
reservoir using a total of 9 compressors, 6 of which are reciprocating engines
2-25
-------
and 3 are turbines. The average engine horsepower per compressor unit is
4,700 HP and the maximum injection pressure is 3,500 psi. Gas injection in
1989 was 49 bcf.
• Plant 2: Plant 2 operates in an identical manner as Plant 1. However, Plant 2
has 7 compressors, all of which are reciprocating engines. The average
engine horsepower per compressor unit is 4,100 HP and the maximum
injection pressure is 3,900 psi. Gas injection in 1989 was 24 bcf.
• Plant 3: This plant operates in a similar manner to Plants 1 and 2. As
opposed to Plants 1 and 2, which use large compressors, Plant 3 utilizes a
large number of small compressors - 10 reciprocating compressors with an
average engine horsepower per compressor unit of 1,400 HP. The maximum
injection pressure is 1,500 psi, significantly lower than Plants 1 and 2. Gas
injection in 1989 was 11 bcf.
• Plant 4: This plant is similar to Plant 3. Nine reciprocating engine
compressors with an average engine horsepower per compressor unit of 1,400
HP inject gas from the transmission system into the underground storage
reservoir at a maximum injection pressure of 1,500 psi. Gas injection in 1989
was 12 bcf.
• Plant 5: This plant has one 4,000 HP reciprocating engine compressor that
compresses gas for injection at a maximum of 1,600 psi. 1989 gas injection
was 11 bcf.
In addition to these five model plants, Tilkicioglu and Winters (1989) estimated fugitive
emissions at a large facility (Plant 6) with a capacity of 73 bcf/yr. The facility consists of
compressors, coolers, scrubbers, and injection wells.
Emissions estimates for these model facilities are summarized in Exhibit 2-10. As
shown in the exhibit, the emissions rate for engine starts/stops range from 0.4 to 10.4 Mg/bcf,
with an average rate of 3.7 Mg/bcf. This is the largest source of emissions for these facilities
(Emissions from compressor exhaust are considered separately below in section 2.4.6).
Emissions from pneumatic devices, fugitive emissions, and other normal operation sources
are minor.
Emissions from routine maintenance were calculated at Plants 1 through 5. For Plants
1 through 4, routine maintenance included a yearly station blowdown and work-over of the
wells. For Plant 5, routine maintenance involved a station blowdown of main headers and
compressor units twice a year and of wellhead separators 3 times a year. Emissions from
routine maintenance range from 0.17 Mg/bcf to 8.7 Mg/bcf, with an average rate of
3.5 Mg/bcf. This is the second largest source of emissions from these facilities.
System upset emissions were calculated at Plants 1 through 5. For all five plants,
emissions from system upsets resulted from a station blowdown due to an emergency once
every two years. The average emissions factor for this source is 1.4 Mg/bcf.
2-26
-------
CO
CO
a
"m
CO
•a
f»
Jv
±«
^B
^
C
o
CD
"£*
0 0)
*~ "O
N 0
Exhibit
Factors from A/
•a
CO
CO
CP
0
a.
c
_o
8—
E
UJ
£. o - ST
OJJs n O
CO CO »S A
i- o) O —
S e « 5*
Sjj11-^
(O
£
A
a
in
S
CO
a
*•
a
m
CO
a
CM
c
CO
Q.
^
**
C
CO
a
oS
« €
So,
u-5
*— •
^.
££•
J5 °>
tts
If
u- S
di «L.
S ^
S o>
S.
s^
2 A
Q ^
CO 0>
*S
^_ _
S >•
0 0)
OC s
S-
fil
CO O)
u- 5
S^
5 »
OC ^
^-
si
a o>
"s
tf^,
a*
£|
ll
co o>
"s
""^
^^
* >•
5f
^£
Emissions Type
^ s s &
« d o'd
LU UJ LU fc
Z-^ -^ -
*- o
LU LU LU *
z z z o>
^t- T- CM LU
o" d d Z
r^ o o LU
•* -r-' CM Z
^- IM
- O Q Mj
co ° ° Z
.^
i-: o o ^!
k z
So o %
CO
£' o ° z
*
So o yj
d 2
S o o g
§ 0 0 jg
* III
,-' o o ^
m ^
CO
a.
o
w ° •"
,1 ll ll '. ll
1 g I?* 2 oS £ §> E
gg^LU Q.T3 O LLLU
Z O
m
co'
LU
LU
z
h-
d
O)
m'
CO
CO
OJ
00
q
05
^
i^.
in
CM
CO
cvi
CO
co'
CM
(Routine
Maintenance
Tt
LU
LU
z
t"-
o
d
00
d
CM
h-
m
CM
CO
co'
in
8
S
0
O5
t^
00
d
m
8
|| System Upsets
t_
5:
•5
^3
CO
N.
k.
>.
^
Q
k_
>.
^
JD
CM
t
JD
^
**T
O
_Q
•^t
C\J
k_
->•
•G
(-*
JLJ
2
•*»•
IGas Injection/
Withdrawal (avg)
co'
•*
CM
C
o
t3
CD
in
_c
>«
CD
15
^
(0
o.
05
m
•D
o>
.^^
(C
E
"fn
05
isor exhaust are
£
Q.
£
8
2
CO
O
0
'in
in
e
LU
TD
10
STJ
Ia. Includes emissic
NE = Not Estimate
-------
Estimated emissions were relatively small, totaling 0.02 Tg/yr. Exhibit 2-11 summarizes
total methane emissions from injection/withdrawal plants in the U.S. The emissions were
estimated by multiplying the average emissions factors by total U.S. injection and withdrawals
in 1990. To be consistent with the manner in which the emissions factors were developed,
the average of the injection (2,450 bcf/yr) and withdrawal (1,949 bcf/yr) amounts, 2,200 bcf/yr,
was used in the estimates. Most of the emissions are associated with compressor start/stops
and routine maintenance. Fugitive emissions and other sources are negligible.
Exhibit 2-11
Summary of Total Emissions from Storage
and Injection/Withdrawal Plants in the U.S.
TOTAL U.S. THROUGHPUT = 2,200 bcf/yr
Emissions Type
Normal Operations
Engines
Start/Stops
Pneumatic devices
Other8
Fugitive Emissions
Routine Maintenance
System Upsets
Average
Emissions
Factor
(Mg/bcf)
3.7
0.02
0.04
0.27
3.5
1.4
Total U.S. Emissions from Storage
Total U.S.
Emissions
(Tg/yr)
0.01
<0.001
<0.001
<0.001
0.008
0.003
S:?:^.' : ":
a Includes emissions due to orifice changes. Emissions
from compressor exhaust are estimated separately.
2.4.4 Transmission Facilities
Emissions estimates for transmission facilities are based on several sources.
Tilkicioglu (1990) analyzed three cross-country transmission systems located in Arizona
(System 1), Southern California (System 2), and Northern California (System 3). The systems
cover approximately 1,200 miles and had a combined throughput of about 3,000 bcf/yr Due
to the great distances these transmissions systems traverse, only segments of each of the
systems were studied in detail. These segments included the mainline compressor and
metering stations. Emissions rates for the model segments were then applied to the entire
length of the pipeline to estimate emissions for the model pipelines. The margin for error in
such an extrapolation is relatively low due to the repetitiveness of components along the
length of the transmission system and to the uniformity of operating practices.
2-28
-------
The criteria for selecting model transmission system segments were:
• Uniformity of throughput, pressure, and pipeline size along the segment;
• Similarity of compressor stations in equipment, pressure, and horsepower;
• Lack of feed lines supplying additional quantities of gas to the mainline at
intermediate points; and
Lack of branch lines taking gas from the mainline for feed to different systems.
These three model facilities were used to estimate: (1) non-exhaust compressor
emissions due to compressor station blowdowns, compressor scrubber operations, and
compressor starts/stops; (2) scrubber operations at metering facilities; (3) scraper operations
on the pipelines; and (4) system upsets due to emergency system shutdowns.
Estimates by Tilkicioglu and Winters (1989) of fugitive emissions from transmission
facilities were based on the study by Pacific Gas and Electric (PG&E 1990) of a transmission
system in California (System 4). The model system consists of 3,189 miles of pipeline. Leaks
from pipeline corrosion and from inadequately sealed valves, fittings, and assemblies were
considered.
Estimates from pneumatic device venting and routine maintenance were made using
the volumes calculated by PG&E (1990) and SOCAL (1992) in their "unaccounted for gas"
studies. Routine maintenance emissions include station shutdowns, as well as periodic
servicing of metering stations and pipelines. The emissions from the venting of glycol
dehydrators were taken from Radian Corporation (1992a). The data used by Radian for
estimating the emissions factor for glycol dehydrators are the same as those used for
estimating the emissions factor for glycol dehydrators at field production facilities
(5.57 Mg/yr/dehydrator).
The emissions factors developed for
the four model facilities are listed above in
Exhibit 2-5.7 As shown in the exhibit,
fugitive emissions account for the largest
emissions.
Transmission system emissions are
estimated to be about 1 Tg/yr (not
including compressor exhaust
emissions). Fugitive emissions,
pneumatic devices, and routine
maintenance account for 80 percent of
the estimate.
In addition to these fugitive
emissions, SOCAL (1992) estimated packing
seal fugitive emissions of 1.05 Mg/mile. This
estimate from SOCAL is added to the
estimate shown in Exhibit 2-5 to estimate a
total fugitive emissions factor of 1.54 Mg/mile, which is used to estimate national emissions.
Emissions factors for pneumatic devices and routine maintenance are shown in Exhibit 2-6.
Both of these sources are relatively important for transmission systems.
7 As discussed above, the emissions factors for transmission systems were also applied to gathering facilities.
See text.
2-29
-------
Total methane emissions from transmission systems are estimated at about 1.04 Tg/yr
(Exhibit 2-12). Fugitive emissions, venting from pneumatics devices and emissions from
routine maintenance are the major sources. All emissions factors except for glycol dehydrator
vents were multiplied by the total transmission system mileage of 280,100 miles (AGA, 1991b).
For emissions from glycol dehydrators, the emissions factor was multiplied by Cowgill's
(1992) estimate of dehydrators at transmission facilities (6,097).
Exhibit 2-1 2
Summary of Total Emissions from the Transmission System in
the U.S.
Total U.S. Miles = 280,100
Glycol Dehydrators = 6,097
Emissions Type
Normal Operations
Engine - Non-exhaust3
Pneumatic devices
Other6
Dehydrator Vents
Fugitive Emissions
Routine maintenance
System upsets
Average
Emissions
Factor
(Mg/mile)
0.24
0.73
0.18
5.57 c
1.54d
0.80
0.10
Total U.S. Emissions from the Transmission System
Total U.S.
Emissions
(Tg/yr)
0.07
0.20
0.05
0.03
0.43
0.22
0.03
1.04
a Includes emissions from compressor station blowdowns, compressor
scrubber operations, and compressor starts. Emissions from compressor
engine exhaust are estimated separately in section 2.4.6.
b Includes scrubber operations at metering stations and pipelines.
c Emissions per dehydrator.
d This emissions factor combines the Tilkicioglu and Winters (1989) estimate
(Exhibit 2-5) with the SOCAL (1992) estimate of 1.05 Mg/mile for packing seals.
2-30
-------
2.4.5 Distribution Network
Distribution systems take ownership of the gas from the transmission system and
lower the pressure of the gas before delivering it to the consumer. The distribution system
consists of a network of small diameter, low pressure pipelines, metering stations and
regulating stations.
Gas enters the distribution networks at "gate stations" where the pressure is reduced
for distribution. The gas is carried by distribution mains throughout cities and towns. After a
further reduction in pressure, the gas is delivered to consumers through service pipelines.
Distribution mains and service pipelines are buried underground, while metering and
regulating stations, where the gas pressure is lowered, are located above ground or in
underground vaults.
Distribution mains are typically steel pipes, with some older systems having cast iron
pipes. Leakage from cast iron pipes occurs at joints, while leakage from steel pipes can
occur as a result of corrosion. Cathodic protection is used to reduce the rate of corrosion of
steel pipes. Service lines are used to run gas from distribution mains to customers. Plastic
pipes are now used to replace old service lines (e.g., steel and cast iron) and install new
lines. Leakage from plastic can occur from cracks in the material.
Normal Operations
Fugitive emissions from the distribution system were estimated in "unaccounted for"
gas studies undertaken by Pacific Gas and Electric Company (PG&E 1990) and by Southern
California Gas Company (SOCAL 1992).
• The PG&E distribution system serves 43 counties in California with
approximately 26,000 miles of service lines and 34,000 miles of distribution and
feeder lines, for a.total distribution system size of 60,000 miles. The system is
considered relatively new; approximately 77 percent of the lines (47,000 miles)
were installed after 1950. The system receipts were 857 bcf in the year of the
study.
• The SOCAL distribution system serves 13 counties in California with
approximately 41,000 miles of service lines and 41,000 miles of distribution and
feeder lines, for a total size of 82,000 miles. The system throughput was 1,162
bcf in 1991.
Each of the studies measured the leak rates of a statistically-selected sample of
leaks detected in their distribution pipelines. The PG&E study measured 20 leaks and the
SOCAL study measured 40 leaks. Total emissions from each system were estimated by
multiplying the leakage rates per leak by the number and duration of leaks detected annually.
To account for variations in leak rates by pipe type, separate estimates of emissions
factors per mile were prepared for plastic and non-plastic piping. Only one leak from plastic
piping was measured during the PG&E study, which produced an unusually high emissions
rate. Two measurements were performed during the SOCAL study which were more in line
with expectations regarding the relative leak rates of plastic and non-plastic piping.
Therefore, only SOCAL's data for plastic mains was used to estimate the plastic mains
2-31
-------
emissions factor. Exhibit 2-13 summarizes the estimates of the emissions factors and the
fugitive emissions obtained from the two studies.
Fugitive emissions from gate stations and regulating stations were estimated based on
data reported by Kolb et al. (1992). Kolb et al. surveyed distribution systems in 11 towns
using a highly sensitive mobile methane detector (McManus et al.. 1991; McManus et_al.,
1989). All 28 gate stations identified in these 11 towns had detectable levels of methane
emissions. At each of the 28 gate stations, the methane emissions rate was measured using
a tracer release technique (Lamb et al.. 1986). The measured emissions rates varied from
about 2 to 200 Mg/yr per gate station, with an average rate of about 32 Mg/yr.
During the surveys of the 11 towns, emissions were rarely detected at the regulating
stations. Measurements at several that had detectable emissions resulted in emissions rates
several magnitudes below the rates measured for the gate stations. Consequently, emissions
from these facilities are assumed to be negligible.
Exhibit 2-1 3
Emission Rates and Factors for Distribution System Fugitive
Emissions
Pipe Type
Plastic Pipe
Non-plastic Pipe
System Size
PG&E Study
Rate
(Mg/yr)
5,551
Factor
(Mg/mile)
0.13
Plastic: 18,753 miles
Non-plastic: 42,662
SOCAL Study
Rate
(Mg/yr)
652
11,285
Factor
(Mg/mile)
0.02
0.22
Plastic: 31 ,325
Non-plastic: 50,699
Average
Emissions
Factor
(Mg/mile)
0.02
0.18
Routine Maintenance
Emissions attributable to routine maintenance were examined by Tilkicioglu and
Winters (1989) based on PG&E (1990). Routine maintenance emissions consisted of the
purging of service meters and pipeline segments prior to repairs. The estimated emissions
and the emissions factor were 207 Mg per year and 0.0035 Mg per mile of pipeline.
System Upsets
Emissions attributable to system upsets were estimated in PG&E (1990). These
emissions were principally associated with dig-ins from outside sources, such as construction
crews. The estimated emissions and the emissions factor were about 1,900 Mg per year and
0.031 Mg per mile of pipeline.
2-32
-------
Summary of Total Emissions from the Distribution Network
In 1990 total methane emissions
from the U.S. distribution system were
approximately 0.33 Tg/yr, with almost 88
percent from fugitive leaks. Although
system upsets and additional subcategories
of normal operations (e.g., pneumatic
Emissions from distribution systems are
estimated at about 0.33 Tg/yr. fugitive
emissions from pipeline leaks and gate
stations comprise over €5 percent of
the estimate.
devices) were not estimated, it is anticipated
that these emissions are negligible. The
emissions factors used and the emissions
estimates are listed in Exhibit 2-14. The activity factors used are as follows.
• The national total of mains (836,700 miles) and services (474,038 miles) in 1990
was 1,310,738 miles (AGA, 1991b). This was used to estimate emissions from
routine maintenance and system upsets.
• The miles of plastic and non-plastic pipeline are estimated as 427,780 and
882,958 miles respectively (AGA, 1992). These were used to estimate fugitive
emissions from plastic and non-plastic pipeline.
The national estimate of gate stations was estimated using information from Kolb et al.
(1992) and from four major gas companies.
According to data provided by Brooklyn Union Gas, Washington Gas Light,
Consolidated Edison Company of New York and Peoples Gas/Chicago there is
an average of 545 miles of mains and service lines for every gate station.
• For the 11 towns studied by Kolb et al.. there was an average of 161 miles of
mains and service lines for every gate station.
The average of these two estimates gives an overall estimate of 353 miles of mains and
service lines for every gate station. Using this estimate with the national total of mains and
services (1,310,738 miles) gives a national total of 3,713 gate stations, and this is used to
estimate the fugitive emissions from gate stations.
2.4.6 Compressor Engine Exhaust
Engines are used throughout the entire natural gas industry, including production
fields, processing plants, injection/withdrawal facilities, and transmission facilities. The
emissions factors of 0.510 Mg/MMcf (reciprocating engines) and 0.009 Mg/MMcf (turbines)
were adopted for estimating the emissions from engine exhaust. These factors were
multiplied by the estimates of fuel used in the different stages of the industry.
Production
While published estimates of the amount of compressor fuel used in production fields
are not available, the total gas use in production fields was reportedly about 806,000 MMcf
(DOE, 1991 a). This total includes fuel used in heaters and other equipment in addition to the
2-33
-------
Exhibit 2-1 4
Summary of Total Emissions from the Distribution System
in the U.S.
Total U.S. Miles = 1,310,738
Total U.S. Plastic Miles = 427,780
Total U.S. Non-plastic Miles = 882,958
Gate Stations = 3,713
Emissions Type
Normal Operations
Engines - Non-exhaust3
Pneumatic devices
Other b
Fugitive Emissions
Plastic Pipe
Non-Plastic Pipe
Gate Stations
Routine Maintenance
System Upsets
Average
Emissions
Factor
(Mg/mile)
NE
NE
NE
0.02
0.18
32.0C
0.0035
0.031
Total U.S. Emissions from the Distribution System
Total U.S.
Emissions
(Tg/yr)
NE
NE
NE
0.01
0.16
0.12
0.004
0.04
0,33
a Emissions from engine exhaust are estimated separately in section 2.4.6.
b Includes scrubber operations at metering stations and pipelines.
c Emissions in Mg/yr per gate station.
NE = Not estimated - believed to be negligible
fuel used in compressors. Therefore, this value represents an upper-bound estimate of fuel
use for compressors in the production stage.
Additional information is available from Tilkicioglu (1990), which estimates the
compressor fuel use for a gas production site that injects gas directly into a transmission
system, without first going through a gas processing plant. Compressors used in production
fields will most likely be found at fields such as this model facility. Tilkicioglu estimated
72.7 MMcf of fuel used per bcf of gas produced. Given the nature of production fields, only
reciprocating compressors are used.
To estimate the national total of fuel used in compressors in production fields, the total
gas produced at these fields that inject directly into high pressure pipelines is needed. This
2-34
-------
quantity is estimated as the portion of gross withdrawals that is not treated in gas processing
plants, as follows:
Portion
of Gross
Withdrawals
= 100%-
1990 Gas Plant Throughput
Marketed Production - Extraction Loss - Lease/Plant Fuel
= 100% - 14.61 Tcf/ (18.56 Tcf - 0.78 Tcf - 1.24 Tcf)
= 11.7 percent.
Engine exhaust emissions are estimated
at about 0.4 Tg/yr. Storage,
transmission, and distribution account
for about 50 percent of the emissions,
with remainder divided about equally
among production and processing.
This approach assumes that
Marketed Production minus Extraction Loss
minus Lease/Plant Fuel is a reasonable
estimate of the total gas injected into U.S.
transmission systems. The portion that is
processed by gas plants is estimated using
the 1990 gas plant throughput. Using this
approach, the national estimate of gas
produced and injected directly into pipelines
would be 11.7 percent of gross withdrawals
(21,490 bcf), or 2,514 bcf. Using the fuel use factor of 72.7 MMcf of fuel per bcf of gas
produced yields an estimate of 182,800 MMcf used in compressors. This is about 20 to 25
percent of total reported plant fuel use. Using this estimate of fuel use and the emissions
factor of 0.51 Mg/bcf yields an emissions estimate of 0.093 Tg/yr (Exhibit 2-15).
Processing
For estimating the total fuel used at gas plants, it is assumed that 50 percent of plant
fuel reported to the DOE (1991 a) is for compressor engine fuel, and that virtually all the fuel is
used in reciprocating engines. Thus, the national fuel used for compressor engines is 50
percent of 428,657 MMcf, or 214,329 MMcf, and the national emissions are estimated as
0.11 Tg/yr (Exhibit 2-15).
Storage, Transmission and Distribution
Pipeline fuel of 659,816 MMcf was reported to DOE (1991 a) for 1990. This fuel is used
both in reciprocating engines and turbines. The total is divided among these two engine
types as follows:
Using the Gas Research Institute compressor database, Jones (1992)
estimated that 69 percent of all prime mover horsepower in natural gas utilities
are reciprocating engines, while the other 31 percent are turbines. Jones
estimated that, on the average, turbines use 1.28 times as much fuel per
horsepower as that used by reciprocating engines. Combining these two
values indicates that the annual fuel used by all compressor engines can be
split in the ratio of 63:37 between reciprocating engines and turbines.
Using the ratio of 63:37, the annual fuel use estimate for reciprocating engines
and turbines is 415,684 MMcf and 244,132 MMcf respectively.
2-35
-------
Exhibit 2-1 5
Summary of Total Emissions from Engine Exhaust in the U.S.
Emissions Type
Production
Reciprocating Engines
Processing
Reciprocating Engines
Storage, Transmission, & Distribution
Reciprocating Engines
Turbines
Average
Emissions
Factor
(Mg/MMcf)
0.51
0.51
0.51
0.009
Compressor
Fuel Use
(MMcf)
182,800
214,329
415,684
244,132
Total U.S. Emissions from Engine Exhaust
Total U.S.
Emissions
0"g/yr)
0.09
0.11
0.21
0.002
&41
Applying the emissions factors for each engine type to these fuel use estimates yields total
emissions from this stage of about 0.2 Tg/yr (Exhibit 2-15).
The total engine exhaust emissions from all the segments of the industry in 1990 were
approximately 0.41 Tg/yr. About 99 percent of these emissions were from reciprocating
engines. The emissions estimate for the storage, transmission and distribution stages of the
industry are comparable to the estimates by Jones (1992) for these stages. Jones did not,
however, estimate emissions from the production and processing stages.
2.4.7 Summary of Total Emissions from the Natural Gas System
The 1990 emission of methane from the U.S. natural gas system is approximately 2.97
Tg/yr (Exhibit 2-16). The single largest source of emissions is fugitive emissions, principally
from the transmission system, production facilities, and the distribution system. The estimate
for fugitive emissions from distribution systems is based on field measurements of both pipe
leakage and gate station leakage. Additional measurements will help strengthen these
estimates.
The second largest emissions source is estimated to be pneumatic devices. This
estimate is based on initial investigations under the EPA/GRl research program, and will also
be strengthened through additional study. Overall, emissions associated with normal
operations account for over 85 percent of total emissions. Routine maintenance and system
upsets contribute only minor methane emissions in all stages of the U.S. natural gas system.
2-36
-------
-e
^
£
^
0)
H
CO
CO
O
a
3
s i
f\i *
^0
^Q ^^
_^^ flJ
,2 |
o
11.
g
o
IS
CO
E
Ul
0)
CO
f
S
a
I-
a> to
£ 3
c ^s
uj ifi
c
1
j:
5
c
o
80
£
i
CD
Injection/
Withdrawal
D>
"5
CO
01
o
o
a
Field
oductlon
a
en
Q.
Emissions T;
SCO CM 1- CO K
T- T- V 1- O
d d r^ d d d
«?
T~ f}
| | '5°' 1
CM
0,
S
LU LU {$ °' i LU LU
Z Z c> « ' Z Z
C3,
_ _ ^
f § 8 ~ ^
_ — : -« O — ' fcO .M O
O ° CO ^ CO*" h» lO^^
S. 1, S. d, §,
*- ^ q T-
d z d ' d o o
V V 0 V
0.
8^8 §
8°S°5°' | 8° |
O O O OJ O O OT- w
o o P o v
S. S. S S-
x— «.
(Q CI^ ^** "tr** 'fr*'
i-Q CO ^" CO (O
3°' -S 85 „ g§g§
d^- og °SJ ' °5 05
S S- S. S- 2-
(0 CO
"> co c i;
w .9 ^ o w
^ i- P X ^
CD -^ O o LU g
.S- t3 ^ m CO
° E -a > £ {? •°i_
w 5 ?> « -£ -i, co
Si CO £ CO CD O) jr-
tC CD 3 C C •=
g a. o u. LU LU O
^
CM
CO
d
1
o
Q
V
tff
0°
o.
o
T- d
q
o.
o
d
V
OJ
dg
o,
outine Maintenance
DC
,_
d
I
CM"
s5
°' PJ
t
cf
8°
d ^_
q
o
d
V
T-
§
d
V
S5
dg
o.
1
co"
&f
oi oo
5.
S
i- d
f
d ^
CM
o,
_
!5
co d
CO
d r^
o.
I
*~ O5
IO
0.
g
go
d i-
o
o.
p
gd
C3_
eo^
o.
CO
c
o
CO
"E
Dtal 1990 Methane E
i-
s «
0 g
'Si
CO Q.
o" °-
i_ T3
CD C
J3 CO
0 O
" §
O to
If
Q. CD
E CD
O C
O ^
T3 t6
t3 w *"
s.-—
^ S.'cL
g o co
1 si
CO _Q CD
§2|
to o °>
2^?
a c w
O) JO . Q)
i -2 lr|j !>
73 CD "D g CD
s .y to §• c
i§lll
CD Z
-------
2.4.8 Comparison with Previous Estimates
The emissions estimates from this study are about 0.8 percent of 1990 marketed gas
production. Exhibit 2-17 compares this estimate with previously published values. Radian
(1992b) draws on much of the same data used in this study. As shown in the exhibit, Radian
estimates higher emissions from the production stage, and similar values from the other
stages of the industry. Radian (1992b) did not have access to the latest API study results,
which may account for the difference in the estimates for the production stage.
Barns and Edmonds (1990) estimate emissions at about 2.0 percent of marketed
production: 0.5 percent for production and processing systems and 1.5 percent for
transmission and distribution systems. The emissions estimate for production and processing
systems is based on the application of the Rockwell (1980) emissions factors to a model
facility. In this study, the Rockwell (1980) emissions factors were adjusted downward based
on results from an ongoing API study, which may account for the difference in these
estimates. The Barns and Edmonds 1.5 percent estimate for transmission and distribution
systems is based on reported "unaccounted for" gas values, which have been shown in
PG&E (1990) to be a large overestimate of emissions for the PG&E system.
AGA's (1989) survey estimated emissions from transmission and distribution facilities
at 0.3 percent of marketed production. This figure compares well with this study's estimate of
nearly 0.4 percent considering that the AGA (1989) estimate does not include emissions from
intrastate pipelines.
According to Abrahamson (1989) the leakage from gas production is approximately
0.13 percent of total dry gas production while methane emissions from transmission and
distribution is about 2.7 percent. These estimates are based partly on reported "unaccounted
for" gas quantities, and like the Barns and Edmonds estimates, were not developed from
detailed assessments of emissions processes.
2.4.9 Uncertainties
The estimates summarized in Exhibit 2-16 are based on the best available data.
However, as discussed above, only limited numbers of measurements have been performed,
and most of the estimates are extrapolated based on results for a small number of model
facilities. Although selected to be as representative as possible, the representativeness of the
model facilities may be lacking in some cases. Consequently, relatively large uncertainties
are associated with most of the estimates.
Objective data are not available for quantifying the uncertainties in the emissions
estimates. Consequently a subjective assessment of the uncertainty in the estimates was
performed using the following assumptions.
It is assumed that the point estimates are estimates of the mean values of the
true emissions. This assumption is reasonable because there are no
indications that the model facilities or emissions factors used are biased.
• The uncertainty in each emissions estimate is represented by a range around
the mean estimate. Ranges are specified so that it is "very likely" that the true
2-38
-------
w
01
s
« -^
iSf
to —
«2 TJ
C O
O 3
» 0
(A O
-1C V.
N E 0-
T •" -a
CM cq f>
as = «
fi -i *
7- > n
x £ 1
ui n. *
£ 0
11
S
01
E
0)
0)
CO
£•
^^
M
T3
C
c a
0 5
•= i 1
n
Q.
E
o
O
to
3
CO
C.
X
UJ
0)
c
"en
c
UJ
c
.S
.Q
to
O
CD
D)
CO
O
c/5
c
.g
CO
CO
Transmi
Ol
c
rocessi
D.
c
o
IJ
T3
O
ol
>.
T3
1
vp
5^
Q
^
d
-^
o^
o
d
M
-$
£
d
^P
S^
§
d
V
IS
o
>.
T3
cl
CO
!c
LU
Z
A
LU
Z
vO
s^
o
d
V
a
IS
d
^
d
S?
h-
d
|| Radian (1992b)
UJ
z
•vO
S?
in
S?
in
d
I Barns and
Edmonds (1990)
LU
z
o
3?
CO
d
LU
Z
LU
Z
a?
CO
O)
<
O
<
LU
Z
5?
r-
c\i
S?
d
Abrahamson
[ (1989) |
CO
o
CO
$
Q.
§
o
m tfi
1 i
>>
-------
emissions rate falls within the range. As such, the range is analogous to a 90
or 95 percent confidence interval about the mean estimate. However, data are
not available for quantifying objectively the level of confidence associated with
the subjective ranges.
The uncertainty in the estimates are driven by both the uncertainties in the
emissions factors and the uncertainties in the activity levels (e.g., miles of
pipeline). Generally, the uncertainties in the emissions factors are much larger
than the uncertainties in the activity levels because the emissions factors are
based on a small number of model facilities or measurements, and (with a few
exceptions) the activity levels come from published government and industry
statistics.
Subjective uncertainty ranges for the emissions factors were assigned based
on the amount of information available for estimating the values. For most of
the emissions factors, a very broad uncertainty range was assumed: from one-
third to three times the mean value. In other words, for an emissions factor of
6.0, the uncertainty range would be 2.0 to 18.0. This range spans an entire
magnitude, which indicates that the emissions factors are not known with great
precision. While measurements at individual facilities may fall outside this
subjective uncertainty range, this range is quite broad as a representation of
the uncertainty in the average emissions factor.
For pneumatic devices and fugitive emissions (not including packing seal
emissions), a slightly narrower uncertainty range of one-half to two times the
mean value was assumed. The emissions factors for fugitive emissions are
based on numerous measurements, and the principal uncertainty is the
representativeness of the model facilities. The emissions factors for pneumatic
devices are based on manufacturers specifications and measurements. Again,
the principal uncertainty is the representativeness of the model facilities.
A narrower uncertainty range is assumed for the compressor exhaust
emissions factor, two-thirds to 1.5 times the mean value because this
emissions factor is based on many hundreds of detailed measurements, and
the population of compressors is relatively well-defined.
A subjective uncertainty range of ±25 percent of the mean value is used for
the activity levels. This uncertainty range is probably larger than necessary to
ensure that it is very likely that the true value falls within the range. However,
this range is adopted to be conservative.
To estimate the uncertainty in the product of the uncertain emissions factor and
the uncertain activity level, it was assumed that the two sets of uncertainly are
independent and uncorrelated. This assumption is reasonable because the
emissions factors and the activity levels are derived separately from different
sources. The uncertainty in the product of the two values was estimated
numerically by simulating random draws from the uncertainty ranges for each
value, and multiplying them together. The 2.5 and 97.5 percentiles of the
distribution of the simulated product were used to describe the uncertainty.
2-40
-------
For the emissions factor, a log-normal distribution was used to simulate the
values, which is consistent with the asymmetrical nature of the uncertainty.
The lower end of the uncertainty range was assumed to be the minimum value
of the log-normal, and the upper end of the uncertainty range was assumed to
be the 97.5 percentile value of the distribution. The uncertainty in the activity
levels was assumed to be normally distributed, with the low value representing
the 2.5 percentile value and the high value representing the 97.5 percentile
value.
The results of the simulated uncertainties for each emissions factor-activity level pair
are shown in Exhibit 2-16 in parentheses below each estimate. As expected, the range of
uncertainty is quite large for each estimate. For example, the uncertainty range for the Field
Production dehydrator vents emissions is 0.04 to 0.33 Tg/yr, with a mean estimate of
0.11 Tg/yr.
To estimate the uncertainty in the total national emissions, the uncertainties in the
individual estimates that make up the national estimate must be combined. First, total
emissions were derived for each segment, including estimates of uncertainties for these totals.
Then, the industry segment totals were combined to estimate national emissions.
To sum the emissions within each segment, the estimates that are derived from
different studies are assumed to be independent, so that the uncertainties of each are not
correlated. However, many of the estimates within each stage are derived from the same
small number of model facilities. For these estimates, the uncertainties are assumed to be
perfectly correlated. To combine the uncertainties that are perfectly correlated, all the low
estimates are added to produce the low estimate for the total, and all the high estimates are
added to produce the high estimate for the total. This method preserves the wide uncertainty
range of the estimates derived from the same model facilities. For the estimates that are
independent, the uncertainty in the summation was simulated numerically. The uncertainty
distribution for each value was assumed to be log-normal.
The following estimates within each stage are assumed to have perfectly correlated
uncertainties:
• Field Production: fugitive emissions (not from packing seals), other engine
emissions, other emissions, routine maintenance emissions, and system upset
emissions.
• Processing: all emission sources.
• Injection/Withdrawal: all emission sources.
• Transmission: fugitive emissions (not from packing seals), other engine
emissions, other emissions, routine maintenance emissions, and system upset
emissions.
Distribution: routine maintenance emissions and system upset emissions.
Because most of the sources are assumed to have perfectly correlated uncertainties, nearly
the full uncertainty is preserved within each stage.
2-41
-------
To estimate the uncertainty in the national total, the uncertainties in the total estimates
for each stage were combined. Because separate model facilities and methods were used to
estimate emissions from each stage, these values were added assuming that they were
independent. The summation was performed using numerical simulation, again assuming
that the uncertainty distribution for each of the stage totals was log-normal. The resulting
estimate of the uncertainty range is about 2.2 to 4.3 Tg/yr, or about -25 percent and +45
percent of the mean estimate (see Exhibit 2-16).
This resulting estimate of uncertainty is reasonable for several reasons. The range is
quite broad, which is consistent with the fact that the underlying uncertainties in the
emissions factor estimates are also broad. The high estimate is about twice the low value.
The uncertainty is asymmetrical, which is expected because while some of the emissions
factors could be much larger than estimated (e.g., 200 percent higher than the mean value)
they cannot be much lower because they are known to be above zero. Finally, the
uncertainty in the individual emissions estimates and the estimates for each stage are broader
than the uncertainty in the final total. This characteristic is consistent with the fact that
although each estimate is very uncertain, a portion of the uncertainty from one estimate is
offset by the uncertainty in another estimate, thereby reducing the relative uncertainty in the
overall total.
2.5 FUTURE EMISSIONS
Future methane emissions from the U.S. Natural Gas System will be driven by
changes in the size of the national system and changes in operating practices and
technology. Changes in the size of the system will be driven by the future demand for natural
gas as well as the changes in the productivity of the natural gas system. An increase in the
size of the system in terms of total throughput, miles of pipeline or number of wells would
tend to increase emissions, although not necessarily proportionately. Future operating
practices will be driven by improved technology and the need for increased efficiency. The
use of improved alternative practices will reduce methane emissions per unit of the natural
gas delivered by the system.
Future methane emissions from the U.S. natural gas system have been forecasted for
the years 2000 and 2010 under three different demand scenarios assuming that the current
operating practices remain unchanged. Following the presentation of these scenarios, the
implications of the enhanced use of alternative practices are discussed.
2.5.1 Current Operating Practices Scenario
Under current operating practices, changes in future methane emissions will be
determined by the changes in the size of the gas system which in turn will be determined by
the changes in future consumption of natural gas and the productivity of the system. In 1990,
the AGA forecasted three scenarios for natural gas consumption in the U.S. from 1990
through 2010 (AGA, 1990). In all of the scenarios, supplies from the lower 48 states and
Alaska along with imports from Canada, Mexico, and other countries were projected to be
sufficient to meet demand at the assumed prices (AGA, 1991 a). Based on these energy
demand scenarios of the AGA, the following three cases of future methane emissions from
the U.S. Natural Gas system can be estimated.
2-42
-------
• Base Case. This scenario assumes a "business as usual" situation with no
major energy or environmental policy changes. Energy use is expected to
increase to a total of 99 quads from a 1989 level of 80 quads. Natural gas'
share of the energy market is expected to remain almost constant, changing
from 24 percent in 1989 to 24.3 percent and 23.3 percent in 2000 and 2010
respectively. Oil prices are expected to grow steadily by 4.2 percent annually,
reaching a price of $47/barrel by 2010. The field-acquisition price of natural
gas is projected to rise 5.0 percent annually to about $5.00/MMBtu from
today's price of below $2.00/MMBtu. Average end-user prices are expected to
grow only 2.5-4.0 percent annually, due to expected efficiency improvement in
the transmission and distribution system.
• Low Energy Case. In this scenario, strict energy conservation measures are
put into effect to curb primary energy demand. In addition, the following strong
environmental policy measures are taken:
after 1990, measures such as a "sulfur tax" are implemented to curb
sulfur dioxide emissions and reduce damage from acid rain; and
after 2000, measures such as a "carbon tax" are implemented to reduce
carbon dioxide emissions from the burning of fossil fuels and counter
potential greenhouse effects.
Total U.S. energy use is expected to grow to only 84 quads in 2010 under this
scenario, with a 25 percent share being natural gas. Natural gas is expected
to take an increasing share of the electricity generation, commercial cooling,
and fleet transportation markets under these environmental policies.
• High Energy Case. In this third scenario, the same environmental policies of
the Low Energy Case are put into effect. However, the energy conservation
policy is less strict, with 50 percent less conservation than the Low Energy
Case. Thus, U.S. energy demand grows to 91 quads in 2010 with natural gas
having a market share of 28 percent.
Given these forecasts of natural gas consumption, the range in emissions with current
practices can be estimated stage by stage assuming current practices remain in effect. The
emissions in 1990 are projected into the future using one of the following two ratios:
• The Domestic Provision Ratio (DPR). The domestic provision for a year is that
part of the total U.S. natural gas demand that is met using domestic sources.
Therefore, Domestic Provision = total consumption - imports. The Domestic
Provisions Ratio (DPR) for a target year is the ratio of the estimated Domestic
Provision for that year to the domestic provision for 1990 which is used as the
base year.
The Total Consumption Ratio (TCR). This is the ratio of the estimated total
consumption of natural gas in the target year to the total natural gas
consumption in the base year of 1990. The TCR reflects the change in total
consumption between the target year and the 1990 base year.
2-43
-------
Exhibit 2-18 compares the expected natural gas consumption, the domestic provision,
the DPR and the TCR under the three scenarios for the years 2000 and 2010. Natural gas
usage is expected to increase significantly during the next decade. The expected total
natural gas consumption by the residential, commercial, industrial, and utility generation
consumers in the year 2000 ranges from about 20.2 Tcf in the Low Energy Case to 22.3 Tcf in
the High Energy Case, with the Base Case at about 21.1 Tcf. In the year 2010, the range is
from 19.9 to 24.6 Tcf with a Base Case of 21.8 Tcf.8
Consequently, the DPR in the year 2000 ranges from a low of 1.13 to a high of 1.24
with a base estimate of 1.18 while in the year 2010, it ranges from a low of 1.11 to a high of
1.35 with a base estimate of 1.21. The TCR in the year 2000 ranges from a low of 1.20 to a
high of 1.33 with a base estimate of 1.25 while in the year 2010, it ranges from a low of 1.18
to a high of 1.46 with a base estimate of 1.30.
Exhibit 2-1 8
Future U.S. Natural Gas Consumption Outlook (Tcf/yr)
Users
Residential
Commercial
Industrial
Utilities
Total Consumption
Imports
Domestic Provision
Domestic Provision
Ratio (DPR)
Total Consumption
Ratio (TCR)
1990
4.39
2.68
6.97
2.79
16.83
1.53
15.30
1.000
1.000
2000
LOW
4.66
3.69
7.38
4.47
20.19
2.90
17.29
1.131
1.200
BASE
4.85
3.50
8.93
3.79
21.07
3.10
17.97
1.175
1.252
HIGH
4.76
3.88
8.54
5.15
22.33
3.40
18.93
1.238
1.327
2010
LOW
4.56
4.08
6.89
4.37
19.90
3.0
16.90
1.105
1.183
BASE
4.95
3.79
8.74
4.37
21.85
3.40
18.45
1.206
1.298
HIGH
4.76
4.56
10.00
5.24
24.56
3.90
20.66
1.351
1.460
Source: AGA 1990
These ratios are used to estimate the sizes of the stages of the natural gas system for
the years 2000 and 2010. The DPR is used to scale the size of the production and
processing stages as well as the volume of lease/plant fuel use; the TCR is used to scale the
size of the storage stage as well as the volume of pipeline fuel use. For the transmission and
distribution stages, annual estimated pipeline growth rates are used. Each stage is described
in turn.
8 These figures appear to cover the range from other studies. The Annual Outlook for Oil & Gas 1991 has a
DOE/EIA projection for 2010 which parallels the AGA's Low Energy Case for natural gas use, even though total US
energy use is larger and natural gas market share smaller (DOE, 1991 b).
2-44
-------
Production Stage
Assuming that all future imports of natural gas will be consumed by U.S. consumers,
U.S. production wells will be responsible for providing the domestic provisions in the years
2000 and 2010. There are diverse expert opinions on the future productivity of gas wells.
In a conservative estimate of fugitive emissions, the productivity of gas wells in 1990 is
considered to remain constant over the next 20 years These assumptions allow us to project
the number of gas wells and treatment facilities in 2000 and 2010 by multiplying the 1990
estimate by the DPR.
Gathering pipelines are constructed to collect gas from natural gas wells, and thus
grow and contract with the number and spacing of such wells. Greater drilling selectivity
along with increases in the number of infill wells and recompletions increased the number of
wells per mile of gathering from 1.9 in 1975 to 3 in 1990. For the purposes of extrapolation,
the 1986-1990 average of 2.75 wells/mile was used to estimate the total length of gathering
pipeline in 2000 and 2010 (AGA, 1991). This ratio was applied to the number of wellheads
and treatment facilities projected for the years 2000 and 2010.
While domestic oil production is expected to continue to decline slowly in the US over
the next two decades, it is assumed that higher gas prices will increase the proportion of oil
wells marketing the associated gas produced with their oil. Thus, the number of oil wells
producing gas is also projected using the DPR. The number of glycol dehydrators, heaters
and separators are also extrapolated using the DPR because these facilities are roughly
proportional to the number of wells.
Processing Stage
While the volume of gas processed is driven mainly by the demand for the liquid
products removed and thus is quite volatile, a future estimate can be made by projecting the
1990 volume into the future using the DPR. The DPR is used instead of the TCR because
imported gas is generally processed in the country of origin before shipping by pipeline or
tanker. The number of glycol dehydrators in this stage is also extrapolated using the DPR.
Storage Stage
Storage volumes and their emissions are assumed to increase with the volume of gas
delivered to consumers. Thus, methane emissions in 2000 and 2010 are estimated by
projecting the 1990 storage volumes by the TCR.
Transmission Stage
Transmission pipelines in the U.S. have been growing at a steady rate of 1400
miles/year over the last 20 years. According to experts in the field, although there will be
9 According to William Fisher at the Bureau of Economic Geology in Texas average well productivity should
increase over the next several decades, but increasing wellhead prices will probably lead to the drilling of more
marginal-quality gas prospects. The late 1970's and early 1980's witnessed a precipitous decline in the average
productivity of natural gas wells. During the late 1980's, however, the decline halted and average well productivity
stabilized. The AGA TERA model assumes that average productivity per gas well will resume its fall over the next two
decades.
2-45
-------
some regional differences, the overall growth rate over the next few decades should follow
the historical pattern.10 Thus, a constant growth rate was used to estimate the total length
of transmission pipelines in 2000 and 2010.
The number of glycol dehydrators in the transmission system were projected for the years
2000 and 2010 using the TCR assuming that the number of dehydrators increase in
proportion to the throughput of gas consumed.
Distribution Stage
The main trend over the last decade has been the growing popularity of using plastic
pipe for distribution mains and services. According to Watts (1990), 97 percent of new
services and 87 percent of new mains are plastic. In addition, for every two miles of main or
service mile added in the late 1980's, about one mile of existing line was replaced, usually
with plastic.
From 1980 to 1990, distribution mains grew by 13,500 miles/yr while from 1983 to
1990, services grew by 8,400 miles/yr (AGA 1984 and 1991b). Assuming that these growth
rates remain constant over the next 20 years and that the ratio of plastic to non-plastic use
and replacement remain as in the late 1980's, the 427,780 miles of plastic pipeline and
882,958 miles of non-plastic pipeline estimates of 1990 would become 736,210 miles of
plastic and 793,528 miles of non-plastic in the year 2000 and subsequently 1,044,640 miles of
plastic and 704,098 miles of non-plastic in the year 2010.
The number of gate stations are expected to increase with the increase in distribution
pipeline mileage assuming that the average ratio of 1 gate station per 353 miles remains
constant.
Compressor Engine Exhaust
Compressor engine lease/plant fuel use is assumed to increase over time with the
volume of gas produced and processed domestically. Thus, the 1990 volume of fuel use by
compressor engines at production and processing facilities is projected to 2000 and 2010
using the DPR.
Compressor engine pipeline fuel use is assumed to increase with the total volume of
gas consumed, and the 1990 fuel use for transmission, distribution, and storage facilities is
projected into the future using the TCR. Both of these projections also assume that there are
no major changes in engine efficiency and that the horsepower and fuel use proportion of
reciprocating engines to turbine engines does not change significantly over the next two
decades.
Summary of Total Future Emissions
Using these assumptions for the different stages, Exhibit 2-19 presents the estimated
size of the natural gas system under the three scenarios. Given current practices, activity
factors are not expected to change substantially over the next 20 years. Methane emissions
10 Conversations with Jeff Meyers and Art Eberle of Columbia Gas, Leonard Crook of ICF, and Brian White at
the AGA.
2-46
-------
E
0)
SL
>>
(0
w
(0
O
"5
^ . -5
o> ^
V «
CM Z
5 »:
£ 3
UJ Jg
*<
M—
O
0)
N
(0
0>
(^rf
3
U.
o
§
§
£ D)
*J
s!
m C
UJ
>.
P
CD
UJ
O
* g;
B) *-
2* 0)
^.S
s °»
g £
CD C
UJ
*s
0 g
UJ
8
O)
^~
<-<
c
D
CD
O)
CO
CO
IB* 8iS 1 §
is'l clss S s
£ g> C is. £2 o °° oo
fo-SjS $Eo{§ « Z
8*$ tfgtf ? »
?8s sis S g
8 a 5 ffi 8 s § a
C\I ™ CO T~" ^~"
8gg s°g § |
- (D ^^- ^^ ^^- ^T - uJ
CO irT ^O k.J" CO — J~ •* — rrT
sSg co-c^cS 2! S
Isl 852 $ 1
-as s-s " a"
CO "" CO <0 CM ^ i- CM
£$S 5P:g 8 g
°. S 00 CO CM co °l CO
>n r-T m • *»• «r o _..-
°£s £8" £ Si
§SS g|s£ g §
i*-- 05 *^ CM * is- in •*
OJ .^T CO — r O o,- of -
8 S 8 S « 5 g ^
_co
"5
CO
16 co
•50) o
co .£ co _
^Si gl 1 f
aj<22 co«» » ^
§Eo. fcgo .Q- S
co^co «Q.,
•*7
O
X3
1 Throughput
CD
5
o
CO
1 1
CO O
«» w"
" S
CM O
CM co"
is." O
CO
5 i
O ^,T
00 Cft
0
h^- en
o
l-~ 0
« ^~
rs. 05
CM
is, §
O) *-
P. 0"
" CM
lo"
CD
CO 1
O CD
ra .c
"O O5
>. Q.
s b-
Q c
No. of Gas
Transmissic
c
o
co
CO
CO
c
2
h=
§-. o „
00 Tf 00
|s, CO CO g J*
1^- Is ;-
^ co " r^
o co ^3- oo
r^. co co 0^ ^"
S3 It 5
^_- CD "- N.
§ 00 ° OO
si- 55 ft
" •* o O •*
^- CD "- |s.
§00 O OO
CO ^— CM t
is- o CM in co
^-" oo" CD" co" w-
^ m co en ^"
o in is. is.
o oo o oo
O CO -r- CM ^f
is. o CM m co
T-" 00" CO" CO" w-
rs. m co en *
01 in is. is.
§co o oo
CO T- CM tj-
is. o CM in co
•r-" 00" CO" CO" W-
h^ in co en *r
en m N. is.
O 00 O CO
o co oo in co
|s- O is. O5 i-
CD" •*" N." CM" |v--
CO is. CM 00 CO
OO •* ^t 00
CO
-CD
"c
"ST-B
-2 CD
E i
^-~ CD
•5T g -9-
^* i °- g
CO ~ CD 0 ri
CD E Q. = -2
1 T "5- co S
-S.9E «
.£ > w c S
CO m CO o CO
2 co E z O
c
o
.Q
03
Q
is
co co
co to
- O)
CM
Ǥ
co CD"
•5- ^
- 00
|S
co Q
CO ^O
- r^
$5o
PS
CM «
£>£
t—
? CM
CO CD
5»"
Z" °°
1 ,396,873
I 791,889
S CD
£ co-
co en
^-8
^.
"o 'CT'
^ >-
5-S
u
"m ^r
™ §-
U
Lease/Plant
Pipeline Fue
gl
•E to
cn^i
C X
UJ UJ
-------
for the future, therefore, can be forecasted for all the stages of the system by multiplying the
1990 emissions factors with the forecasted size of the natural gas system in years 2000 and
2010. Exhibit 2-20 presents the expected range of methane emissions for the future.
The resulting methane emissions from the U.S. natural gas system in the year 2000,
assuming the continued use of current practices, are estimated to range from 3.3 Tg/yr in the
Low Energy case to 3.5 Tg/yr in the High Energy estimate, with a Base Case estimate of
3.4 Tg/yr. In the year 2010 the methane emissions are estimated to be 3.3 Tg/yr in the Low
Energy case and 3.8 Tg/yr in the High Energy case, with a Base Case estimate of 3.5 Tg/yr.
Given an uncertainty of -25 percent and +45 percent in the underlying 1990 estimates of
methane emissions, each of these future methane emissions estimates have a range of about
-25 percent and +45 percent as well. Given the uncertainty in the future rate of expansion of
the natural gas system, and the impact that the expansion will have on emissions, the future
emissions estimates have a relatively wide range of uncertainty.
Exhibit 2-20
Summary of Future Methane Emissions From the U.S. Natural Gas System
(Tg/yr)
Stage
Production
Processing
Storage
Transmission
Distribution
Engine Exhaust
1990
1.08
0.09
0.02
1.04
0.33
0.41
2000 I 2010
Low
Energy
1.24
0.10
0.02
1.10
0.35
0.49
Total 2.97 3.30
JJ2.18 4.26)JL f2-43 4-74)
Base
Energy
1.30
0.10
0.02
1.10
0.35
0.51
3.38
(2.48 4.85)
High Low
Energy || Energy
1.37
0.10
0.03
1.10
0.35
0.54
1.22
0.09
0.02
1.15
0.37
0.48
3.49 3.33
(2,565.00) Jj2.45 4,78)
Base
Energy
1.33
0.10
0.03
1.15
0.37
0.52
3.50
(2.57 5.02)
High
Energy
1.49
0.11
0.03
1.16
0.37
0.59
3.75
(2.75 5.38)
2.5.2 Improved Technology and Operating Practices
Although methane emissions are expected to increase in the coming decade, as
demand for natural gas increases, emissions from the natural gas system may be offset by
the use of improved technology and operating practices. While total methane emissions in
1990 were estimated to be nearly one percent of total gas consumption, the fraction emitted
in the future may be somewhat less if alternative practices are adopted. Examples of these
practices are as follows.
Directed inspection and maintenance programs to reduce fugitive emissions
from surface facilities.
2-48
-------
• Improving combustion efficiency on the engines that power the natural gas
system with lean burn systems and better fuel monitoring to reduce methane in
emissions exhaust.
Using more turbine engines instead of reciprocating engines at compressor
stations on new pipeline construction.
Repairing leaking pipes (eg., cast iron) or replacing them with plastic.
• Replacing pneumatic controls that emit gas with non-emitting controls or
electric controls.
• Capturing or flaring methane routinely vented during normal operations or
maintenance.
The potential for using these technologies and practices, and the implications of their use for
emissions, are examined in EPA (1992).
2.6 LIMITATIONS OF THE ANALYSIS
The primary limitation of the analysis is that only several studies form the basis for the
emissions factor estimates. Many of the estimates are based on a small number of case
studies, and the representativeness of the available estimates is difficult to assess. For
example, the emissions rates for underground distribution pipelines are based on studies of
only two systems, PG&E and SOCAL. In cases where information was limited or not
available, conservative emission values were assigned.
The emissions estimates are also limited because they do not include all possible
sources, such as abandoned wells. No data on the number of abandoned wells or the
potential emissions rates from such wells were identified.
Estimates of future emissions are limited by the uncertainty in future gas demand and
how emissions will change with changes in demand. The future estimates presented are
based on the assumption that emissions will grow with the size of the gas industry, in terms
of sales, mileage, or other variables. However, emissions are estimated to grow at a rate less
than the rate of growth in demand. Changes in operating practices may tend to limit future
increases in emissions even as gas sales increase.
The Gas Research Institute and the U.S. EPA are conducting research and analyses to
improve the basis for estimating methane emissions from the U.S. natural gas system. Efforts
are being focused to reduce uncertainties in the main sources of emissions, including:
fugitive emissions from production, processing, transmission, and distribution systems;
emissions from pneumatic devices; and emissions from engine exhaust.
2-49
-------
2.7 REFERENCES
Abrahamson D. (1989). "Relative Greenhouse Effect of Fossil Fuels and the Critical
Contribution of Methane," presented to The Oil Heat Task Force.
AGA (American Gas Association). 1984. Gas Facts - 1983 Data AGA; Arlington, VA.
AGA. 1989. "Natural Gas Transmission and Distribution Methane Emissions," AGA,
Engineering Technical Note, Arlington, VA.
AGA. 1990. The Outlook lor Gas Energy Demand, 1990-2010 AGA; Arlington, VA.
AGA. 1991 a. The Gas Energy Supply Outlook, 1991-2010 AGA; Arlington, VA.
AGA. 1991b. Gas Facte - 7990 Data AGA; Arlington, VA.
AGA. 1992. Personal communication between Gordon Weynand, U.S. EPA, Washington D.C.
and Linda Martin, AGA on 2/10/92.
Barns, D.W. and J.A. Edmonds. 1990. "An Evaluation of the Relationship Between the
Production and Use of Energy and Atmospheric Methane Emissions," prepared for the
Office of Energy Research, U.S. Department of Energy, Washington, D.C.
Campbell, L 1991. "Methane Emissions from Reciprocating Engines/Gas Turbines Used in
Gas Industry," presented to the Advisory Committee on Methane Emissions from the
Natural Gas Industry, sponsored by the U.S. Environmental Protection Agency and the
Gas Research Institute, San Antonio, Texas, August 8, 1991.
Cicerone, R.J. and R.S. Oremland. 1988. "Biogeochemical Aspects of Atmospheric Methane,"
Global Biogeochemical Cycles, Vol. 2, No. 4, Dec.1988, pp. 299 - 327.
Cowgill, M.R. 1992. Presentation at the GRI/EPA Gas Industry Methane Emissions Advisors
Committee meeting in Austin, Texas, April 22-24, 1992
DOE (Department of Energy). 1991 a. Natural Gas Annual 1990. Energy Information
Administration, Office of Oil and Gas. Washington, DC. DOE/EIA-0131 (90)/1.
DOE. 1991b. Annual Outlook for Oil & Gas. Energy Information Administration, Office of Oil
and Gas. Washington, DC. DOE/EIA-0517(91).
DOE. 1990. Natural Gas Annual 1989. Energy Information Administration, Office of Oil and
Gas. Washington, DC. DOE/EIA-OL 3(89).
EPA (Environmental Protection Agency). 1985. Compilation of Air Pollutant Emission Factors
Volume I: Stationary Point and Area Sources, Office of Air and Radiation, Research
Triangle Park, North Carolina, AP-42, September 1985.
EPA (Environmental Protection Agency). 1992. Options for Reducing Methane Emissions from
Anthropogenic Sources in the United States, Global Change Division, Office of Air and
Indoor Air Programs, Washington, D.C. In Preparation.
2-50
-------
Fisher, W. (Bureau of Economic Geology) remainder of citation to be supplied
Gibbs, M.J., P.P. Hathiramani and M. Webb. 1992. Fugitive Methane Emissions from Oil and
Gas Production and Processing Facilities Emissions Factors Based on the 1980 API-
Rockwell Study, report prepared for the Global Change Division, Environmental
Protection Agency, Washington, D.C.
Hitchcock D.R. and A.E. Wechsler. 1972. "Biological Cycling of Atmospheric Trace Gases
prepared for the National Aeronautics and Space Administration, Washington, D.C.,
March 1972.
INGAA (Interstate Natural Gas Association of America), "Global Warming and Methane Loss
from Interstate Natural Gas Pipeline," Rate and Policy Analysis Department,
Washington, D.C.
IPCC (Intergovernmental Panel on Climate Change). 1992. Climate Change 1992. The
Supplementary Report to the IPCC Scientific Assessment. Prepared for the IPCC by
Working Group I.
Jones, D. L 1992. "Me.thane Emissions from Natural Gas Industry, Reciprocating Engines and
Gas Turbines" presented to the GRI Advisory Committee on Methane Emissions from
the Natural Gas Industry, sponsored by the U.S. Environmental Protection Agency and
the Gas Research Institute, Research Triangle Park, North Carolina, Feb 5, 1992.
Kolb, C.E, et al. 1992. Methane Emissions from Natural Gas Distribution Systems, prepared for
the U.S. Environmental Protection Agency and the Gas Research Institute, Aerodyne
Research, Inc., Billerica, Massachusetts.
Lamb, B.H., H. Westberg, and E. Allwine. 1986. "Isoprene Emission Fluxes Determined by
Atmospheric Tracer Technique," Atmos. Environ., 20:1.
McManus, J.B., P.L Kebabian, and C.E. Kolb. 1989. "Atmospheric Methane Measurement
Instrument Using a Zeeman-split He-Ne Laser," Appl. Opt., 28:5016-5023.
McManus, J.B., P.L. Kebabian, and C.E. Kolb. 1991. "Aerodyne Research Mobile Infrared
Methane Monitor," in Measurement of Atmospheric Gases, SPIE Proceedings,
1433:330-339.
Oil and Gas Journal (OGJ). 1991. "Worldwide Gas Processing Capacities as of 1991, and
Average Production," Oil and Gas Journal Vol 89, No. 29.
OGJ. 1990. "Worldwide Production and Refining Report" Oil and Gas Journal Dec., 1990.
OGJ. 1991. "Worldwide Production and Refining Report" Oil and Gas Journal Dec., 1991.
Pacific Gas & Electric Company (PG&E). 1990. Unaccounted for Gas Project Summary Volume
PG&E Research & Development; San Ramon, CA; GR 1-90/0067.1.
Radian Corporation. 1991 a. "Estimate of U.S. Methane Emissions," Draft Peer Review Report,
Vol.1.
2-51
-------
Radian Corporation. 1991b. "Results of the Glycol Waste Survey - Draft Report," (updated)
prepared for the Gas Research Institute.
Radian Corporation. 1992a. "Venting and Flaring Emissions from Production, Processing, and
Storage in the U.S. Natural Gas Industry," Updated Draft Report prepared for the U.S.
EPA and the Gas Research Institute.
Radian Corporation. 1992b "Methane Emissions from the Natural Gas Industry: Production
and Transmission Emissions," prepared for the Air and Waste Management
Association.
Rockwell International. 1980. Fugitive Hydrocarbon Emissions from Petroleum Production
Operations, prepared for the American Petroleum Institute; API Publication No. 4322.
Southern California Gas Company (SOCAL). 1992. Unaccounted for Gas Project Summary
Volume (in preparation) SOCAL Research & Development; Los Angeles, CA.
Tilkicioglu, B. H. 1990. Annual Methane Emissions Estimates of the Natural Gas Systems in the
U.S., Phase II.; Pipeline Systems Inc.
Tilkicioglu, B. H. and D. R. Winters. 1989. Annual Methane Emissions Estimates of the Natural
Gas and Petroleum Systems in the U.S.; Pipeline Systems Inc.
Urban C.W., South West Research Institute. 1992. Personal conversation with Jonathan
Woodbury, ICF on May 7, 1992.
Watts, J. 1990. "25th Distribution Piping Report," Pipeline & Gas Journal, Dec., 1990; pp. 14-
15.
Webb, M. 1992. "1991 API Fugitive Hydrocarbon Emissions Factors for Onshore Petroleum
Production Operations," Draft Report.
2-52
-------
CHAPTER 3
METHANE EMISSIONS FROM COAL MINING
U.S. Methane Emissions
from All Sources
Nat ur a I Gas
Systems
Other Soui ces
I Mini ng
Annual Coal Mine
Methane Emissions
GI oba I Em i -ss i on
U *5 Em i r,^ i orv3
Emissions Summary
Source
Underground Coal Mines
Ventilation Systems
Degasification Systems2
Surface Coal Mines
Post-Mining
Total (1988)
Total (1990)3
1 988 Emissions
(Tg)
2.1 1
0.5- 1.6
0.2 - 0.7
0.5 - 0.8
3.3 - 5.2
3.6 - 5.7
Partially
Controllable
/
1 No formal uncertainty range included in estimates (see discussion
in Methodology section).
2 Does not include an additional 0.25 Tg recovered from coal mines
in Alabama and Utah that was sold to pipelines instead of being
vented to the atmosphere.
3 The 1990 emissions estimate was extrapolated from the 1988
estimate; 1 988 is the latest year for which complete data is
available (see discussion in Emissions Summary section below).
3.1 EMISSIONS SUMMARY
In 1988, an estimated 3.5 to 5.4 teragrams (Tg) of methane was liberated by coal
mining. Of this amount, approximately 3.3 to 5.2 Tg (172.1 to 271.2 billion cubic feet or Bcf;
3-1
-------
4.9 to 7.8 billion cubic meters or Bern) was emitted to the atmosphere. The remaining 0.25
Tg (13.0 Bcf; 0.4 Bern) was used instead of being vented. Emissions from U.S. coal mines
are summarized in Exhibit 3-1.
Emissions were estimated for 1988, as opposed to 1990, due to data availability.
Because coal production was higher in 1990 by about 8 percent, it is likely that methane
emissions were also higher. Using a simple ratio of emissions to coal production between
1988 and 1990, results in a 1990 emissions estimate of 3.6 to 5.7 Tg.
Coal mining is currently the third largest source of methane emissions in the United
States, accounting for about 16 percent of total U.S. methane emissions in 1988. These
emissions had the same energy content as 9 to 13 million tons of coal (8 to 12 million metric
tons), which is approximately one percent of total U.S. coal production in 1988.1
U.S. coal mines accounted for an estimated 10 to 15 percent of emissions from coal
mining worldwide in 1988 (USEPA 1990a). The United States is estimated to be one of the
world's three largest emitters of methane from coal mining; the other two are the People's
Republic of China and the former Soviet Union.
Exhibit 3-1
Methane Emissions from U.S. Coal Mines
(in Teragrams)
Emissions Source
Underground Mines
Surface Mines
Post-Mining
Total U.S. Emissions
Methane Recovered1
1988
2.6 - 3.7
0.2 - 0.7
0.5 - 0.8
3.3 - 5.2
0.25
2000
3.0 - 4.8
0.2 - 0.8
0.5-1.0
3.7 - 6.5
0.25
2010
4.1 - 6.6
0.3 - 0.9
0.7- 1.2
5.0 - 8.7
0.25
1 Methane recovered represents the amount of methane
that is currently collected and used by U.S. coal mines
instead of being vented to the atmosphere. For 1988, this
amount is known. For 2000 and 201 0, it was assumed that
methane recovery would remain at 1 988 levels.
Methane emissions from coal mining are projected to increase significantly in 2000
and 2010 due to expected increases in coal production. Total methane liberations from coal
mining could range from 3.7 to 6.5 Tg in 2000 and 5.0 to 8.7 Tg in 2010.
One short ton equals 0.9 metric tons. All values in this chapter are represented in short tons.
3-2
-------
Most of the methane from U.S. coal mining is emitted by mines in the Northern
Appalachian, Central Appalachian and Black Warrior Basins. These coal basins contain many
of the gassiest underground mines in the United States. Several mines in these basins have
degasification systems in place that are recovering high-quality methane that could potentially
be used to generate electricity or sold to a pipeline. Currently, however, only a few mines in
the Black Warrior Basin and one mine in Utah are recovering and using methane from these
systems instead of venting it to the atmosphere.
The key sources for methane emissions from coal mining are: ventilation and
degasification systems at underground mines; surface mines; and post-mining emissions.
Methane emissions associated with coalbed methane recovery in non-mining areas were not
considered in this chapter, but are included in the emissions estimates for natural gas
production, processing, transmission and distribution (Chapter 2 of this report).
• Underground Mines
Underground mines accounted for more than 70 percent of total methane emissions
from coal mining in 1988. They will also contribute significantly to emissions in the future.
About 55 to 80 percent of the methane liberated by underground coal mines in the
U.S. in 1988 was emitted to the atmosphere from ventilation air shafts. Because this methane
is contained in air at very low concentrations (less than 1 percent), there are few uses for it.
Ventilation air streams will continue to represent a significant portion of methane emissions
from underground coal mines in the future.
In 1988, an estimated 0.7 to 1.8 Tg (36.5 to 93.9 Bcf; 1 to 2.7 Bern) of methane was
recovered by degasification systems at U.S. coal mines. These systems, which include
surface gob wells and in-mine boreholes, are in use at about 35 U.S. coal mines and they
recover methane in higher, useful concentrations. Six U.S. mines sold the methane produced
by degasification systems to local pipeline companies, and as a result about 0.25 Tg (13.0
Bcf; 0.4 Bern) of this methane was not emitted into the atmosphere.
Emissions from degasification systems at underground mines could increase
significantly in the future, possibly reaching 0.6 to 2.1 Tg in 2000 (31 to 109 Bcf; 1.0 to 3.3
Bern) and 0.9 to 2.9 Tg (46.9 to 151.2 Bcf; 1.4 to 4.6 Bern) in 2010. If key barriers to methane
recovery are removed, much of this gas could potentially be recovered profitably instead of
being emitted to the atmosphere.
• Surface Mining
Methane emissions per ton of coal mined are low for surface mined coals. Given the
large coal production at U.S. surface mines, however, this emissions source is significant. In
1988, surface mining emissions were an estimated 0.2 to 0.7 Tg.
• Post-Mining
Some methane remains in the coal after it has been mined and can be emitted during
transportation, storage, and handling of the coal. Post-mining emissions in the United States
are estimated to be approximately 25 to 40 percent of the in-situ methane content of the coal,
or about 0.5 to 0.8 Tg in 1988.
3-3
-------
3.2 BACKGROUND
Methane emissions from coal mines are heavily dependent on the geological
characteristics and history of the coalbed. Factors such as coal rank, depth, and permeability
affect the amount and distribution of methane in the coalbed and surrounding strata, which in
turn determine the quantity and rate of methane release during mining. In addition, the type
and rate of mining, as well as the geometry of the mine, have important implications for
methane release. Because methane is a safety hazard in underground mines, substantial
research has been undertaken to determine ways of predicting and controlling its emissions
into mine working areas.
3.2.1 How Coalbed Methane Is Produced, Stored and Released
Coal is formed over millions of years
as organic matter is transformed by complex
processes known as "coalification."
Coalification is controlled by chemical and
physical processes, temperature, pressure
and geologic history. Differing levels of
coalification produce different "rank" coals,
as shown in Exhibit 3-2.2 Coalification
results in both physical and chemical
changes, including methane generation.
Other byproducts of the coalification
process are water and carbon dioxide.
The amount of methane produced
increases throughout the coalification
process. Thus, higher ranked coals tend to
contain more methane than lower ranked
coals. Methane is stored in the coal itself
and can also be contained in the
surrounding strata. In addition, some of the
methane generated by coalification
generally escapes to the atmosphere as a
result of natural processes.
How Methane Is Stored in Coal
cr
o
u
trt
<
UJ
CE
O
Exhibit 3-2
Stages in Coalification
t
Graph i te
t
Anthraci te
t
Bi tuminous
t
Sub-Bitumi nous
t
L i gn i te
t
Peat
Large amounts of methane can be stored within the microstructure of coal. Methane
storage in coalbeds, mainly by adsorption onto internal coal surfaces, is a function of rank,
and present day pressure and depth of burial.3 In general, coals of increasing rank have
higher storage capacities. In addition, storage capacity increases almost linearly with
2 The term "rank* is used to designate differences in coal that are due to the progressive change from lignite to
anthracite. Higher rank coals contain more fixed carbon, less volatile matter, and less moisture.
3 Adsorption is the adhesion in an extremely thin layer of molecules to the surfaces of solid bodies with which they
are in contact.
3-4
-------
increasing pressure or depth. Therefore, at a given rank, deeper coals store more gas than
shallower ones.
Even high rank coals cannot store all of the methane generated during coalification,
however. The highest gas contents measured for anthracite coal in the United States, for
example, are only 10 to 12 percent of the total amount of methane that was generated during
coalification. The rest of the methane migrated out of the coal over time. Some of this gas
remains stored in the surrounding strata, and some has likely been emitted to the atmosphere
as a result of natural processes.
Factors Determining Methane Emissions
During mining, methane is liberated by
the mined coal seam as well as
surrounding coal seams and/or gas
bearing strata. The amount of methane
liberated can be many times higher
than the amount of methane contained
in the mined coal seam.
Methane is released when pressure
within a coalbed is reduced, either through
mining or through natural erosion or
faulting. Methane will migrate through coal
from zones of higher concentration to zones
of lower concentration until it intersects a
pathway, such as a cleat, joint system or
fracture. The size, spacing, and continuity
of such pathways determines the
permeability of the coal and largely controls
the flow of methane through the coal and to
the surface or the mine workings.
As pressure is reduced during mining, methane is liberated from the seam being
mined and from the strata above and below the mined seam. In addition to the rank and
depth of the coal, the amount of disturbance to the surrounding strata as a result of mining
activities will also have important implications for emissions. The amount of methane
liberated by mining activities can exceed the amount of gas contained in the mined coal by
as much as 3 to 9 times (Kissell et al. 1973).
3.2.2 U.S. Mining Techniques
Coal is produced in the United States in surface and underground mines, and the
choice between these mining types depends primarily on the thickness of the coalbed and its
depth from the surface. Coalbeds shallower than 60 meters are generally mined from the
surface, while deeper coalbeds are usually mined by underground methods. Given enough
coal thickness, surface mining methods can be applied to a depth of several hundred meters.
Each mining method has different implications for release of methane to the atmosphere.
The major U.S. coal basins are shown in Exhibit 3-3. In general, coal in the Western
basins is mined using surface methods, while most Eastern coals are mined using
underground methods.
Underground Mining
Underground mining accounted for about 40 percent of total U.S. coal production in
1988. Most underground mining occurs in the Eastern United States, primarily in the
3-5
-------
Northern and Central Appalachian Basins (including Pennsylvania, Virginia, West Virginia,
Ohio, Kentucky) and the Black Warrior Basin of Alabama.
Exhibit 3-3
Major U.S. Coal Basins and Coalbed Methane Resources
Western Washington
24 Tcf
Wind Rivei
2 Tcf ..
Greater Green
3-1 Tcf
Powder R i vei~
30 Tcf
Northern Appalachian
61 Tcf
Cent i a I AppnIac hia n
5 Tcf
Most U.S. underground mines are less than 300 meters deep, but several reach
depths of 600 to 700 meters. Methane can be emitted during mine construction, coal
production, and from abandoned mine workings. The bulk of the emissions tend to be
associated with coal production, however, and in particular with the caving of the roof and
floor rocks, which creates pathways for the gas to move into the mine workings from unmined
areas of the target coal seam and other strata.
Two underground mining methods
are commonly used in the United States:
room-and-pillar mining and longwall mining.
The choice between these methods
depends on geologic factors, such as depth
and terrain, and economic factors, such as
equipment cost. Longwall mines are
typically bigger and deeper than room-and-
pillar mines. They are also more expensive
to equip and operate, but generally have higher coal production rates. The higher
Longwall mining tends to liberate more
methane than room-and-pillar mining.
Thirty-six of the 50 gassiest
underground mines in the U.S. use
longwall mining methods.
3-6
-------
production, coupled with the more extensive caving typically associated with longwall mines,
tends to result in higher methane emissions.
Room-and-pillar mining is the most common underground mining technique in the
United States, although the number of longwall mines is growing. Mechanized longwall
mining was introduced in the U.S. during the 1960's, and today there are almost 100 longwall
mines in operation. Thirty-six of the 50 gassiest U.S. underground mines use longwall mining
methods.
Surface Mining
Surface mining, also called strip mining, is used to mine coal at shallow depths. In
essence, it involves large scale earth-moving; first the overburden on top of the coal is
excavated, and then the coal can be removed. Coal recovery rates at surface mines can
exceed 90 percent.
In 1988, 568 million tons of coal (511
million metric tons) was produced at surface
mines, mostly in subbituminous and lignite
mines in the Western United States. This
represented about 60 percent of total U.S.
coal production. The largest and fastest
growing U.S. surface mining region is the
Methane emissions from surface mines
are highly uncertain. Available
information indicates that emissions per
ton of coal mined are likely to be low
because these coals are generally low
ranked and buried close to the surface.
Powder River Basin of Wyoming and
Montana. Surface mines are also located in
the lignite fields of North and South Dakota
and Montana, and the Eastern bituminous coal basin in Illinois, Indiana, and Western
Kentucky.
Surface mines are not required to monitor methane emission levels because this
methane is emitted directly into the atmosphere and does not pose a safety hazard to miners.
Thus, few emission measurements are currently available. The U.S. Environmental Protection
Agency's Office of Research and Development has undertaken a field measurement study of
methane emissions from surface mines that should be completed in 1994.
Based on available information, it appears that methane emissions from surface mines
are low as compared to underground mines because the coals are typically lower ranked and
are buried at shallower depths. Given the magnitude of coal production from surface mining,
however, this emission source is not insignificant.
3.2.3 Methane Management Systems for Underground Mining
Methane is a serious safety threat in underground coal mines because it is highly
explosive in atmospheric concentrations of 5 to 15 percent. The U.S. Mine Safety and Health
Administration (MSHA), an agency of the U.S. Department of Labor, requires close monitoring
of methane levels and careful design of mine ventilation systems to ensure that methane
concentrations are kept below explosive levels in underground mines. In mine entries used
by personnel, methane levels cannot exceed 1 percent, and in certain designated areas of the
mine not frequented by mine personnel, methane levels cannot exceed 2 percent. If these
concentrations are exceeded, MSHA requires that coal production cease until the ventilation
3-7
-------
Many gassy U.S. underground mines
use degasification systems in addition
to ventilation to ensure safe mining
conditions. These systems produce
high quality methane that can be sold
to pipelines or used to generate
electricity.
system is able to reduce methane
concentrations to acceptable levels.
A variety of methane control
methods are employed in many U.S. mines
because of the hazard and because the
costs of elevated methane concentrations
can be high if coal production must be
suspended. Historically, ventilation has
been the main technique used for
controlling methane concentrations in coal
mines. In many mines, however, methane emissions into the mine workings cannot be
economically maintained at safe levels using ventilation alone and other degasification
systems are used. These systems can recover methane before, during or after mining and
keep it from migrating into the mine working areas.
Mine degasification systems are currently used primarily to improve safety and reduce
ventilation costs. However, these systems can recover methane with a high enough energy
content to warrant sale to pipelines or use for electricity generation. In addition to the
potential economic benefits associated with the sale of this gas, such projects have the
added advantage of reducing atmospheric methane emissions, a potent greenhouse gas.
Methane management methods are shown in Exhibit 3-4 and the key characteristics of these
systems are summarized in Exhibit 3-5.
Certain methane recovery techniques, particularly vertical drainage in advance of
mining, have also been used extensively to recover coalbed methane in non-mining areas.
These projects, which were encouraged in part by the Section 29 Unconventional Gas tax
credit, recovered almost 350 Bcf (9.9 Bern) of coalbed methane in 1990, primarily in the Black
Warrior Basin of Alabama and the San Juan Basin of Colorado and New Mexico. Because
these projects are not associated with mining, they did not affect methane emissions from this
source. Methane emissions from stand-alone coalbed methane production, processing and
transmissions are included in the discussion of natural gas systems (Chapter 2 of this report).
3.2.4 Post-Mining Emissions
Not all of the methane contained in coal is released during mining. Some methane
remains in the coal after it is removed from the mine and can be emitted over the following
days as the coal is transported, processed and stored. Depending on the characteristics of
the coal and the way it is handled after leaving the mine, the amount of methane released
during post-mining activities can be significant and can continue for days or even months.
The greatest releases occur when coal is crushed, sized, and dried in preparation for
industrial or utility uses (USEPA 1990b).
3.3 METHODOLOGY
Methane emissions are estimated for each major coal mining source, including both
ventilation and degasification systems at underground mines, surface mines, and post-mining
coal transport and handling. The reported emissions for 1988 were based on actual data
3-8
-------
where available and emissions were estimated where data were unavailable. Emission
estimates were prepared for 1988, as opposed to 1990, because this was the most recent
year for which ventilation emission data were published. Emissions were forecast for 2000
and 2010, based on projections of future coal production.
Exhibit 3-4
Diagram of Mine Degasification Approaches
(c) Vertical Gob Well
(d) Vertical Degasification Well
(e) Cross Measure and
Horizontal Boreholes
(f) Surface Equipment
3.3.1 Emissions from Underground Mines -1988
In 1988, methane emissions from U.S. underground mines included: (1) measured
methane emissions in the ventilation air at the gassiest underground mines (Trevits et al.
1991); (2) estimated ventilation emissions from mines for which measurements were not
made; and, (3) estimated degasification system emissions.
3-9
-------
Exhibit 3-5
Mine Degasification Approaches
Method
Ventilation
Vertical Wells in
Advance of Mining
Gob Wells
Horizontal Boreholes
Cross-Measure
Boreholes
Description
• Universal method to dilute and exhaust methane to the atmosphere.
• Sufficient, in many mines, to maintain safe mining conditions.
• In gassy mines, may be necessary to supplement with other methane
degasification systems.
• Pre-drains methane via surface wells before mining operations begin.
• Can recover large amounts of pipeline quality methane.
• Technology also used to produce gas from coal seams that are not
being mined.
• Can produce methane from multiple coal seams.
• Used in longwall mining to drain methane from portions of overlying
strata allowed to collapse after mining ("gob areas") via surface wells.
• Can recover large amounts of methane, sometimes contaminated with
mine air.
• Drilled from inside the mine to degasify the coal seam being mined either
years in advance of or shortly before mining.
• Methane is removed through an in-mine piping system.
• Can recover pipeline quality gas.
• Drilled from inside the mine to degasify the overlying or underlying coal
and rock strata.
• Methane is removed through an in-mine piping system.
• Gas can become contaminated with mine air during production.
• Used infrequently in the U.S.
Source: For more information, refer to Baker et al. 1988; Baker et al. 1986; Duel et al. 1988; Dixon
1987; USEPA 1990a; and USEPA 1990b.
Measured Ventilation Emissions
Methane emissions in ventilation air are available from MSHA for about 200 of the
gassiest U.S. underground coal mines. A database compiled from 1988 MSHA inspection
data by the USBM reports the emissions of methane from each mine with emissions
exceeding 100,000 cubic feet per day (2,857 cubic meters) in ventilation air (USBM 1991).
About one-third of all active U.S. underground mines are included in the USBM database.
The reported methane emissions were used for ventilation air estimates for those mines
included in the USBM database.
MSHA does not provide an uncertainty estimate for these measured ventilation
estimates. Accordingly, no uncertainty range was assumed in this report. However, the
uncertainty range associated with emissions reported at individual mines may be as high as
+/- 20 percent. This uncertainty is due to calculation errors in sampling, anemometer
accuracy, and the annualization of quarterly estimates (Niewiadonski 1992). Further research
is warranted to confirm the accuracy of these measurements.
3-10
-------
Estimated Ventilation Emissions
Methane emissions from ventilation systems were estimated for the underground
mines not included in the USBM database. These other mines were classified into three
categories: Active Mines with Detectable Methane Emissions; Active Mines with Non-
Detectable Methane Emissions; and Inactive or Abandoned Mines. Estimation methodologies
were developed based on information provided by USBM and MSHA about their
characteristics and regulatory treatment. The estimated ventilation emissions for these mines
represented less than 2 percent of measured ventilation emissions in 1988. This factor was
applied to the actual ventilation emissions for each coal basin.
• Active Mines with Detectable Emissions: In addition to the mines in the USBM
database, approximately 300 underground mines emitted detectable levels of
methane in 1988.4 MSHA has estimated that these mines emitted about 1 Bcf
(30 million cubic meters) in 1988 (Trevits et al. 1991).
• Active Mines with Non-Detectable Emissions: In 1988, about 1,400 U.S
underground mines had non-detectable methane emissions, implying a
methane concentration in the ventilation air of less than 0.1 percent. MSHA
requires that all active underground mines, even those with undetectable
methane concentrations, ventilate at a minimum rate of 3,000 cubic feet (86
cubic meters) of air per minute. Thus, methane emissions were estimated for
these mines by multiplying the statutory ventilation requirements by an
assumed methane concentration of 0.05 percent in air. These assumptions
resulted in estimated emissions of about 1 Bcf (30 million cubic meters) from
this category of mines in 1988.
• Inactive or Abandoned Mines: Emissions from inactive or abandoned mines
were not estimated, due to the absence of reliable data on the number of
inactive or abandoned mines and their emission levels. In most states
abandoned mines must be sealed and are not ventilated actively. There are
specific cases where abandoned mines are producing significant quantities of
gas.5 It is difficult to use these special cases to develop a national emissions
estimate, however. More research is warranted to confirm that emissions from
inactive and abandoned underground mines are low.
Degasification System Emissions
Specific information on methane emissions from the degasification systems in place at
U.S. coal mines is not currently available because coal mine owner/operators are not required
to report emissions from these systems. In fact, without close examination of the mine
ventilation plans provided to MSHA for each mine, it is difficult to confirm which mines have
degasification systems in place.
4 Personal communication with Jack Tisdale, MSHA, February 27, 1992. According to MSHA, the methane
measurement devices currently in use can detect concentrations of 0.1 percent in air.
5 Personal communication with Thomas Hite, president of Hite Operating Co., Feb. 20, 1992.
3-11
-------
Exhibit 3-6
Assumed Degasification Recovery Efficiencies
Coal Basin
Northern Appalachian
Central Appalachian
Black Warrior
Illinois
Western & Other
Low Case
30%
40%
40%
30%
40%
High Case
65%
65%
65%
65%
65%
Degasification system
emissions were estimated for mines
known or believed to have such
systems in place.6 Low and high
estimates were developed based on
information about likely coal mine
degasification strategies and on
conditions in various coal basins.
The assumptions used are
summarized in Exhibit 3-6, which
shows the percentage of methane
liberations assumed to be recovered
by degasification systems at mines in
different basins. Known recovery
factors were applied to those mines that reported the methane recovery from their
degasification systems (i.e., those mines that sold the gas to pipelines). The recovery factors
were applied to the measured ventilation emissions to estimate total emissions.
The degasification system emissions estimates are summarized in Exhibit 3-7. As
indicated, six mines currently sell methane from their degasification systems to local pipeline
companies instead of emitting it to the atmosphere. In addition, five mines are reported to be
developing systems that will enable them to sell their recovered methane to pipelines. The
remaining mines are venting the recovered methane to the atmosphere. The methane emitted
by mine degasification systems represents the most economically and technically attractive
opportunity for reducing methane emissions to the atmosphere associated with coal mining.
3.3.2 Emissions from Surface Mines -1988
Measurements of methane emissions from surface mines are currently unavailable,
although a field measurement study is underway to better quantify emissions from this
source.7 In the absence of measurements, emissions were estimated using reported
methane contents for the surface coals mined in each coal basin, as shown in Exhibit 3-8.
For each coal basin, the estimated methane content of the coal was multiplied by an
emission factor and by the basin's surface coal production. In the low case, an emission
factor of 1 was used; that is, it was assumed that only the methane actually contained in the
mined coal seams would be emitted. In the high case, however, it was assumed that actual
emissions would be 3 times greater than the methane content of the target coal seam due to
the release of methane from the surrounding strata.8
6 This list was developed based on discussions with USBM and MSHA officials, industry representatives and
literature review.
7 This study is being done by the U.S. EPA's Office of Research and Development (Kirchgessner et al. 1992a).
8 This assumption is consistent with the methodology developed by Environment Canada in their report on
greenhouse gas emissions. (Environment Canada 1992.) Preliminary results of the U.S. EPA study indicate that
the factor could be as high as five (Kirchgessner et al. 1992b).
3-12
-------
Exhibit 3-7
1 988 Estimated Degasification System Emissions
Company
Mine
Emissions
(mmcf)1
Northern Appalachian:
Bethenergy Mines
Cyprus Emerald Resources
United States Steel
Consolidation Coal
Consolidation Coal
Eastern Associated Coal
Consolidation Coal
Consolidation Coal
Consolidation Coal
Consolidation Coal
Consolidation Coal
Consolidation Coal
Cambria Slope #33
Emerald #1
Cumberland
Bailey
Loveridge #22
Federal #2
Arkwright
Humphrey #7
Osage #3
Blacksville #1
Blacksville #2
Robinson Run
515-2,235
580-2,510
1,000-4,340
655 - 2,845
1,235-5,355
1,375 - 5,965
595 - 2,575
925 - 4,000
565 - 2,440
485-2,100
1,330- 5,760
345 - 1 ,490
Central Appalachian:
Consolidation Coal
United States Steel
Consolidation Coal
Island Creek Coal
Island Creek Coal
Island Creek Coal
Island Creek Coal
Amonate
Shawnee
Buchanon #1
VP #1
VP#3
VP#5
VP #6
705- 1,965
75 - 205
2,630 - 7,320
1,970-5,490
1,605-4,475
2,095 - 5,830
1 ,920 - 5,355
Black Warrior:
Jim Walter Resources
Jim Walter Resources
Jim Walter Resources
Jim Walter Resources
U.S. Steel Mining
Blue Creek #4
Blue Creek #5
Blue Creek #7
Blue Creek #3
Oak Grove
2,575
3,045
1,495
2,280
2,850
Uses
Recovered
Methane
possible
possible
possible
possible
possible
X
X
X
X
X
3-13
-------
Exhibit 3-7
1 988 Estimated Degasif ication System Emissions
Company
Mine
Emissions
(mmcf)1
Illinois:
Old Ben Coal
Old Ben Coal
Pyro Mines
Old Ben #25
Old Ben #26
Pyro Wheatcroft #9
295 - 1 ,290
415 - 1,150
45 - 205
Western:
Mid-Continent Resources
Mid-Continent Resources
Cyprus Empire Coal
Western Fuels - Utah
Soldier Creek Coal
Dutch Creek Mine B
Dutch Creek Mine M
Eagle #5
Deserado
Soldier Canyon
1,365-3,795
1,265 - 3,525
25 -70
170-475
840
Uses
Recovered
Methane
X
1 One cubic foot = 0.028 cubic meters.
3.3.3 Post-Mining Emissions
The methane emitted during the post-mining transportation, storage, and handling of
coal has not been systematically measured or evaluated. Previous analyses have estimated
that 25 to 40 percent of the in-situ methane content of extracted coal would be released to
the atmosphere after the coal leaves the mine. British Coal, for example, estimates that post-
mining emissions are 40 percent of the in-situ content because their coals have low
permeability and the gas desorbs slowly (British Coal 1991). Similarly, Environment Canada
estimates that only 54 percent of the methane contained in their surface mined coals is
released during mining (Environment Canada 1992).
In the absence of actual measurements for U.S. coals, post-mining emissions have
been estimated to range from 25 to 40 percent. The low case estimate of 25 percent
represents a conservative assumption, while the high case is more consistent with experience
in other countries. For each coal basin, these emission factors were applied to the methane
contents reported for surface and underground coals (see Exhibit 3-8) and multiplied by 1988
coal production.
3.3.4 Projections of Future Methane Emissions
Methane emissions will change in the future as a result of changes in coal production
and shifts in production among coal basins. In addition, it is possible that emissions factors
will increase in the future if deeper and gassier coal seams are exploited. These estimates
were prepared using two coal production forecasts to reflect the range of possible production
levels. Given the difficulty of quantifying the potential increase in emission factors, however, it
3-14
-------
was assumed that basin-specific emission factors for underground and surface mines would
not change over time.
Exhibit 3-8
Average Methane Contents of Underground and Surface Coal,
by Coal Basin or State
Underground Coal
Basin
Northern Appalachian
Central Appalachian
Warrior
Piceance
San Juan
Illinois
Uinta
Green River
Pennsylvania Anthracite Fields
Cf/ton1
17.3
33.3
32.1
25.6
22.8
5.8
4.2
4.2
14.1
Surface Coal
Basin or State
Appalachian (including Warrior)
Illinois
Powder River
Arkoma
San Juan
Alaska
Arizona
California
Louisiana
North Dakota
Texas
Washington
Cf/ton1
5.0
3.9
0.3
10.9
1.5
0.3
1.6
3.9
0.3
03
0.3
0.3
1 one short ton = 0.9 metric tons
Source: USEPA 1990b
Projections of future U.S. coal production are key to estimating future methane
emissions from coal mining. In addition to projecting overall coal production, these forecasts
must account for shifts in production between surface and underground mines and also
between different coal producing areas.
Two coal production forecasts were used in this analysis, as shown in Exhibit 3-9.9
Both scenarios were developed after passage of the 1990 Clean Air Act Amendments and
reflect the expected impacts of that legislation on the coal industry. One possible impact of
this legislation is a shift from high sulfur coal, which is predominantly produced in eastern
underground mines, to lower-sulfur western surface mined coal. In addition, some coal use
may be displaced by natural gas, which does not emit SO2. Both of these shifts would tend
to result in lower future methane emissions from coal mining than might have been expected
in the absence of this legislation. The low case scenario assumes that energy efficiency
improvements, continued low gas prices, and the 1990 Clean Air Act Amendments will result
The "high" coal production forecast is based on the EIA Annual Energy Outlook, 1992 (EIA 1992). The "low"
coal production forecast is based on DOE/EIA's Supporting Analysis for the National Energy Strategy (DOE 1992)
3-15
-------
in lower coal production in the future than is anticipated under the high case scenario. For
both scenarios, coal production was forecast for individual states and by coal basin, as well
as being disaggregated by underground, surface or lignite mines.
Exhibit 3-9
Coal Production Forecasts
(0
o
u
01
c
o
2.000
1,500
1.000
500
Low Case
High Case
Underground
Surface
1988 2000 2010 2000 2010
The forecast coal production for each basin was multiplied by basin-specific emission
factors for underground and surface mining, which were developed from the 1988 emissions
estimates. These emission factors are included in Exhibits 3-14 to 3-18. Once the total
quantity of methane released from coal mining was calculated for each basin, the estimated
amount of methane that would be recovered and utilized - rather than vented to the
atmosphere - was subtracted. Because the quantity of methane recovered is highly
uncertain and dependent upon numerous factors including wellhead gas prices and disputes
over coalbed methane ownership, the amount of methane recovered from each basin in 1988
was used as the estimate for the amount recovered in 2000 and 2010.
3.4 CURRENT EMISSIONS
3.4.1 Overview
Total methane liberations from U.S. coal mines were 3.5 to 5.4 Tg in 1988. More than
90 percent of this methane (3.3 to 5.2 Tg) was emitted to the atmosphere. The remaining
0.25 Tg (13.0 Bcf; 0.4 Bern) was recovered from degasification systems at six U.S. mines and
3-16
-------
Exhibit 3-10
Current U.S. Methane Emissions
sold to pipelines. The amount of methane
emitted to the atmosphere corresponds to
approximately 172.1 to 271.2 Bcf (4.9 to 7.8
Bern). In 1988, U.S. methane emissions
from coal mining accounted for
approximately 10 to 15 percent of the
estimated global emissions from this source
(USEPA 1990a). Moreover, coal mining
represented approximately 17 percent of
U.S. methane emissions from all sources.
Coal mining's share of U.S. methane
emissions is shown in Exhibit 3-10, and
Exhibit 3-11 summarizes the methane
emissions estimates in 1988.
Underground mines accounted for
about 75 percent of U.S. methane emissions
from coal mining in 1988, as shown in
Exhibit 3-12. These mines liberated an estimated 2.6 to 3.7 Tg (135.2 to 192.4 Bcf; 3.9 to 5.5
Bern) of methane. Approximately 0.5 to 1.6 Tg (26 to 83.2 Bcf; 0.8 to 2.4 Bern) of methane is
estimated to have been emitted by mine degasification systems in 1988. This estimate is
highly uncertain because mines are not required to report their emissions from this source.
Only 0.25 Tg (13.0 Bcf; 0.4 Bern) of gas from degasification systems was recovered and sold,
however, and the rest was simply vented to the atmosphere.
Exhibit 3-11
1 988 Emissions Summary
Key Source
Underground Coal Mines:
Ventilation Systems
Degasification Systems1
Surface Coal Mines
Post-Mining
TOTAL
Estimated Emissions (Tg)
2.1
0.5- 1.6
0.2 - 0.7
0.5 - 0.8
3.3 - 5.2
1 Does not include an additional 0.25 Tg recovered from
coal mines in Alabama and Utah that is currently sold to
pipelines instead of being vented to the atmosphere.
Emissions from surface mines were 0.2 to 0.7 Tg in 1988, accounting for about 6 to 13
percent of U.S. emissions. Post-mining emissions from all coal production were estimated to
be approximately 0.5 to 0.8 Tg. Both the surface and the post-mining estimates were based
3-17
-------
on simple assumptions related to the
estimated methane content of the mined
coal, and both of these areas warrant further
study.
The results of other recent emission
estimates are compared with this analysis in
Exhibit 3-13. As this table shows, most of
the estimates are quite consistent with this
study. Previous analyses have generally
been based either on statistical models or
on very generalized assumptions. The
study contained in this report is one of the
first to develop underground emissions
estimates using available mine-by-mine data.
Exhibit 3-12
1988 Coal Emissions
Underground
Venti I at ion
Systems
Underground
Degas i f n:at i
Systems
Post-Mi rii ng
Surfa<:e Mines
Exhibit 3-1 3
Comparison with Other Recent Emissions Estimates
Study
Report to Congress
USEPA (1990b)
Kirchgessner et al.
(1992b)
DOE (1991)
Emissions
Estimate
3.3 - 5.2 Tg
5.4 - 8.6 Tg
3.5 Tg
4.1 Tg
Estimation
Year
1988
1988
1989
1990
Description of Method
Mine-by-mine for underground mines;
general assumptions for other sources.
Statistical analysis; designed to estimate
global emissions; not based on most recent
U.S. underground emissions data.
Estimate is for underground mining only -
not surface mining or post-mining activities;
based on statistical analysis; no uncertainty
estimate developed to date.
Uses general emission factors for all sources;
estimates are not disaggregated by source
(underground, surface, or post-mining).
3.4.2 Basin-Specific Emissions
Most of the methane liberated by coal mining in the United States is emitted by
underground mines in the Appalachian and Black Warrior basins, as shown in Exhibit 3-14.
The Northern and Central Appalachian coal basins are the major sources of U.S. methane
emissions from coal mining, largely because most of their coal is produced in underground
mines. Mines in the Black Warrior Basin of Alabama are also very gassy, but many of them
are currently selling the gas recovered from their degasification systems to pipelines, which
significantly reduces emissions.
3-18
-------
Exhibit 3-1 4
1 988 Methane Emissions bv Coal Basin tin To)
Basin
Northern Appalachian
Central Appalachian
Black Warrior
Illinois
Western
TOTAL
Underground Mines
Ventilation
0.7
0.6
0.3
0.2
0.2
2.1
Deaasification
0.2 - 0.8
0.2 - 0.6
< 0.11
0-0.1
0.1 - 0.22
0.7-1.8
Surface Mines
and
Post-Mining
0.2 - 0.3
0.4 - 0.7
0.04 - 0.1
0.1 - 0.2
0.1 - 0.2
1.0- 1.5
Total
1.1 - 1.8
1.2- 1.9
0.5 - 0.6
0.3 - 0.4
0.3 - 0.5
3.3 - 5.2
1 Does not include 0.23 Tg that was recovered and utilized rather than released to the
atmosphere.
2 Does not include 0.02 Tq that was recovered and utilized bv one Utah mine.
Northern Appalachian Basin
The Northern Appalachian Basin (NAB) is one of the oldest coal producing regions in
the United States. In 1988, approximately 166 million tons (150 million metric tons) of coal
was mined in the basin, over 58 percent of which was produced in underground mines. Most
of the underground mines in the NAB use room-and-pillar mining methods. There are also 23
longwall mines in the basin, and 12 mines are believed to have degasification systems in
place. Key characteristics of the NAB are shown in Exhibit 3-15.
Exhibit 3-1 5
Coal Characteristics - Northern Appalachian Basin
1988 Coal Production (million tons)
Mining Methods
Average Depth of Mining (ft.)
Coal Rank
Coal Age
Average Gas Content (cf/ton)
Estimated Emission Factor (cubic
feet/ton mined)
Underground Mined Coal
97
Room & Pillar, Longwall
800- 1,200
Bituminous
Pennsylvanian
140 - 460
450-780
Surface Mined Coal
69
Strip
< 500
Bituminous
Pennsylvanian
49
49 - 148
Sources: Kelafant et al. 1988a and 1988b; USDOE 1988; USEPA 1990b; Trevits et al. 1991.
Total methane emissions in the NAB were estimated to be 1.1 to 1.8 Tg (57.4 to 93.9
Bcf; 1.6 to 2.7 Bern) in 1988. Underground mines accounted for more than 80 percent of
these emissions (0.9 to 1.5 Tg). About 30 to 55 percent of the methane released by
3-19
-------
underground mines in the basin is produced by degasification systems. In 1988, these
emissions were estimated to be 0.2 to 0.8 Tg (10.4 to 41.7 Bcf; 0.3 to 1.2 Bern), all of which
was emitted to the atmosphere.
Central Appalachian Basin
The Central Appalachian Basin (CAB), like the NAB, is an old coal producing region.
In 1988, the CAB produced about 263 million tons (238 million metric tons) of coal, of which
68 percent was produced in underground mines. There are approximately 16 longwall mines
in the CAB. Of the seven mines which are believed to have degasification systems in place,
four use longwall mining methods. Key characteristics affecting methane emissions are
shown in Exhibit 3-16.
Exhibit 3-16
Coal Characteristics - Central Appalachian Basin
1988 Coal Production (million tons)
Mining Methods
Average Depth of Mining (ft.)
Coal Rank
Coal Age
Average Gas Content (cf/ton)
Estimated Emission Factor
(cubic feet/ton mined)
Underground Mined Coal
179
Room & Pillar, Longwall
1,500-2,500
Bituminous
Pennsylvanian
200 - 660
220-330
Surface Mined Coal
84
Strip
< 500
Bituminous
Pennsylvanian
49
49-148
Sources: Kelafant et al. 1988a and 1988b; USDOE 1988; USEPA 1990b; Trevits et al. 1991.
Total methane emissions in the CAB were estimated to be 1.2 to 1.9 Tg (62.6 to 99.1
Bcf; 1.8 to 2.8 Bern) in 1988. Underground mines released 0.8 to 1.1 Tg, with degasification
systems emitting about 28 to 52 percent of this gas (0.2 to 0.6 Tg). As in the NAB, all of this
gas is currently vented to the atmosphere. At least two coal companies in southwestern
Virginia are developing projects to recover methane for sale to a pipeline, however. If these
projects incorporate gob gas recovery, emissions could be reduced by 17 to 29 percent
depending on the level of gob gas emissions and the percentage of emissions recovered by
the projects.
Black Warrior Basin
The Black Warrior Basin (BWB) produced about 27 million tons (24 million metric tons)
of coal in 1988, of which 55 percent came from its underground mines. Many of the
underground mines in the basin are very deep (over 2000 feet). There are six longwall mines,
and five have degasification systems in place. Exhibit 3-17 summarizes coal characteristics
for the BWB.
3-20
-------
Exhibit 3-17
Coal Characteristics - Black Warrior Basin
1988 Coal Production (million tons)
Mining Methods
Average Depth of Mining (ft.)
Coal Rank
Coal Age
Average Gas Content (cf/ton)
Estimated Emission Factor
(cubic feet/ton mined)
Underground Mined Coal
15
Room & Pillar, Longwall
1,000-2,000
Bituminous
Pennsytvanian
300-500
2,500
Surface Mined Coal
12
Strip
< 500
Bituminous
Pennsylvanian
49
49- 148
Sources: Kelafant 1988a and 1988b; USDOE 1988; USEPA 1990b; Trevits et al. 1991.
Although coal production is not large, the BWB is one of the gassiest mining regions
in the country. Methane liberations were slightly less than 0.8 Tg (41.7 Bcf; 1.2 Bern) in 1988,
of which 96 percent was released by underground mines. Mine degasification systems
recovered about 0.23 Tg (12.0 Bcf; 0.3 Bern) of pipeline-quality methane; this quantity is
known with precision because the recovered gas is sold to local pipelines instead of being
released to the atmosphere. Thus, methane emissions from BWB mines were 0.5 Tg (25.6
Bcf; 0.7 Bern) in 1988.
Other U.S. Coal Basins
Approximately 504 million tons (454 million metric tons) of coal (52 percent of U.S.
production) is mined in other U.S. coal basins. Over 79 percent of this coal is produced in
surface mines, the largest of which are located in the western states of Montana and
Wyoming. The Illinois Basin has the highest emissions, largely because of its relatively
greater proportion of underground mines. There are only 15 longwall mines in other U.S.
basins (mostly in Illinois) and only nine mines outside of Appalachta and the BWB are
believed to have degasification systems in place. Exhibits 3-18 and 3-19 summarize coal
characteristics for the Illinois and Western basins.
Methane emissions from other U.S. basins totaled 0.7 to 0.9 Tg (36.5 to 46.9 Bcf; 1 to
1.3 Bern) in 1988. Surface mining represented approximately 0.2 Tg, and underground
mining accounted for the remaining 0.5 to 0.6 Tg. Mine degasification systems were
estimated to have emitted 0.1 to 0.2 Tg (5.2 to 10.4 Bcf; 0.2 to 0.3 Bern), all of which was
released to the atmosphere except for 0.02 Tg that was recovered and used by one Utah
Mine.
3-21
-------
Exhibit 3-1 8
Coal Characteristics - Illinois Basin
1988 Coal Production (million tons)
Mining Methods
Average Depth of Mining (ft.)
Coal Rank
Coal Age
Average Gas Content (cf/ton)
Estimated Emission Factor
(cubic feet/ton mined)
Underground Mined Coal
67
Room & Pillar, Longwall
500- 1,000
Bituminous
Pennsylvanian
30-150
160-190
Surface Mined Coal
63
Strip
< 500
Bituminous
Pennsylvanian
39
39-116
Sources: AAPG 1984; USDOE 1988; USEPA 1990b; Trevits et al. 1991.
Exhibit 3-1 9
Coal Characteristics - Rockies and Southwest Basins
1988 Coal Production (million tons)
Mining Methods
Average Depth of Mining (ft.)
Coal Rank
Coal Age
Average Gas Content
Estimated Emission Factor
(cubic feet/ton mined)
Underground Mined Coal
35
Room & Pillar, Longwall
500- 1,500
Subbituminous, Bituminous
Cretatious, Tertiary, others
226
410 - 570
Surface Mined Coal
340
Strip
< 200
Subbituminous, Bituminous
Cretatious, Tertiary, others
15
15-46
Sources: AAPG 1984; USDOE 1988; USEPA 1990b; Trevits et al. 1991.
3.5 FUTURE EMISSIONS
3.5.1 Overview
Methane emissions from coal mining are forecast to grow between 1988 and 2010 as
a result of projected increases in coal production. Exhibit 3-20 illustrates the projected
growth in emissions in 2000 and 2010. By 2000, it is expected that coal mining will represent
over 15 percent of total U.S. methane emissions, and by 2010, its share could reach 20
percent.
Underground mines will continue to be the largest source of methane emissions from
coal mining, representing 72 to 78 percent of total emissions in 2000 and 76 to 82 percent in
2010. The growth in emissions from underground mining may be mitigated somewhat by the
3-22
-------
likelihood that there will be a shift away from underground coal toward low-sulfur surface
mined coal as a result of the acid rain regulations promulgated under the Clean Air Act
Amendments. This possibility is reflected in the coal production forecasts used for this
analysis.
Exhibit 3-20
Emissions Growth Forecast
Global 1988 US 1988
Low Estimate
US 2000 US 20'10
1 High Est imate
In addition to overall growth in underground mining emissions, the estimates forecast
a significant growth in degasification system emissions. In 2000, it is estimated that mine
degasification systems will emit 0.6 to 2.1 Tg (31.2 to 109.5 Bcf; 1.0 to 3.3 Bern) of methane,
representing about 20 to 40 percent of underground emissions. By 2010, moreover,
degasification systems could emit an estimated 0.9 to 2.9 Tg (46.9 to 151.2 Bcf; 1.4 to 4.6
Bern). Under these scenarios, between 36 and 45 mines are assumed to be using
degasification systems in 2000 and between 40 and 54 in 2010, as compared to
approximately 35 mines today.10 As shown in Exhibit 3-21, most of the degasification
system emissions in 2000 are forecast to come from the Appalachian and Black Warrior
Basins.
10 The number of mines that might use degasification systems in the future was estimated by comparing the
number of mines assumed to use degasification systems in 1988 (32) to the total estimated degasification
emissions for that year.
3-23
-------
Both surface and post-mining
emissions are also expected to increase
slightly in the future, consistent with higher
coal production.
Exhibit 3-21
Degasification System Emissions
in 2000, by Basin
Northern Appalachian
Centi a I
Appalachian
Western
3.5.2 Basin-Specific Estimates
The Northern and Central
Appalachian coal basins will continue to be
the major sources of methane emissions
from U.S. coal mining in 2000 and 2010, as
shown in Exhibit 3-22. Mines in the Black
Warrior Basin are expected to be gassier as
well, but these mines are currently selling
some of their methane and it is expected
that this practice will continue.
Northern Appalachian Basin
Coal production in the Northern Appalachian Basin is projected to range from 143 to
164 million tons (129 to 149 million metric tons) in 2000, of which about 75 percent is
expected to come from underground mines. In 2010, the NAB is expected to produce 187 to
214 million tons (168 to 193 million metric tons) of coal, with a similar ratio of surface to
underground mining.
Total methane emissions in the basin are projected to be approximately 1.1 to 2.1 Tg
(57.4 to 109.5 Bcf; 1.6 to 3.1 Bern) in 2000 and 1.4 to 2.8 Tg in 2010 (73.0 to 146.0 Bcf; 2.1 to
4.1 Bern). Of these emissions, underground mining is projected to account for 0.9 to 1.8 Tg
in 2000 and 1.2 to 2.4 Tg in 2010 and surface mining to account for about 0.1 Tg in 2000 and
0.1 to 0.2 in 2010. Post-mining emissions are estimated to be approximately 0.06 Tg in 2000
and 0.08 Tg in 2010.
The use of degasification systems in underground mines is expected to increase in the
NAB. Degasification estimates range from 0.2 to 1.0 Tg (10.4 to 52.2 Bcf; 0.3 to 1.5 Bern) in
2000 and 0.3 to 1.3 Tg (15.6 to 67.8 Bcf; 0.4 to 1.9 Bern) in 2010. This represents about 22
to 55 percent of total methane emissions expected from the basin's underground mines and
about 30 to 45 percent of the nation's estimated degasification emissions.
Central Appalachian Basin
The Central Appalachian Basin is expected to produce 297 to 330 million tons of coal
in 2000, of which about 40 percent will be from underground production. Coal production in
2010 is expected to be 338 to 386 million tons. The proportion of underground mining is
projected to drop slightly, to just below 40 percent.
Total methane emissions are forecasted to range from 1.2 to 2.2 Tg (62.6 to 114.7 Bcf;
1.8 to 3.3 Bern) in 2000 and 1.4 to 2.6 Tg (73.0 to 135.6 Bcf; 2.1 to 3.8 Bern) in 2010.
Underground mining is expected to account for 0.7 to 1.2 Tg in 2000 and 0.8 to 1.5 Tg in
2010. Surface mining emissions are expected to reach 0.3 to 0.4 in both 2000 and 2010.
3-24
-------
I
o
5
Ol
1
i
.E
O)
•—
N S
O •••
£ O
£ o
•S yj
3
i
w
0
•w
CO
1
2
^*
CB
j*
U
S
0
^
1^1
0
^5
3R
CO
±J
t-
o
o c
111
«I|
Q.
CO
0)
i
•a
c
o
s>
CD
•o
c
Q
isfficatli
O)
fl)
Q
0
7
CD
To
«-
Q
^ cu S
O.
CO
0)
c
i
•o
c
a
g
O)
CD
•o
3D
C
o
«-
_o
CO
CO
O)
ci>
Q
c
o
2
I
c
CB
00
co
CM
«
£
d
CM
d
CO
T—
CO
d
^
en
d
,_
CM
T
CO
d
d
q
CM
d
00
d
1
N.
d
£-
CO
03
1
JC
o
Z
to
CM
*
^
^J
m
d
00
d
CM
d
d
d
CM
CM
CM
o>
d
i-
d
CO
d
CM
d
CO
d
<
d
c
.S
u
5
O
CM
•«—
i
d
CM
d
o
d
*CM
d
d
CO
d
d
^
d
CO
d
•^
d
S
d
T-
d
o
d
CO
d
•
m
d
1 Black Warrior
co
d
4
(0
d
CM
d
0
d
d
CO
o
d
d
d
m
d
d
•^
d
S
d
•>—
d
d
CO
d
>
CO
d
I
g
O
*
s
d
T(-
O
d
o
d
o
d
o
d
S
d
CM
O
d
o
d
S
d
8
d
5
d
o
d
o
o"
o
d
1
o
d
•b
3
O
Western Inter!
S
d
4
CO
o
d
§
d
CM
o
d
o
d
o
d
o
d
o
d
fe
o
I
s
d
c§
d
d
o
d
o
d
o
d
o
d
ID
"5
Q.
«
|| Northern Ore,
CO
•»-
i
CO
d
CM
d
d
CM
m
d
CM
d
CO
d
d
en
d
d
^__
d
d
^n
d
d
TT
d
•
CO
d
0)
CD
«;
Rockies - Soi
o
d
•
o
d
o
d
o
d
o
d
o
d
5
d
o
d
5
d
o
d
o
d
o
d
0
d
o
d
o
d
•
o
d
(0
CO
JO
II
|| Northwest - A
^
CO
i
o
in
c\i
O)
d
en
CM'
w
d
CO
CM
co
in
CO
1
CO
IS.
i
00
d
,_
c\i
«o
d
CM
1
cti
1
CD CD
CD CD
"S. Q.
ID in
O O
E E
IS 15
CD CD
o o
® ®
cz c
CD CD
C C
CO CO
£ £
CD CD
"D "O
O> CD
N N
•o n
c c
CO CO
"D "D
CD CD
CD CD
§ §
0 0
CD CD
Q) 03
0 0
T3 T3
E 1
D 3
(O (A
CO CO
CO CM
CM 0
d d
CD CD
T3 "0
3 3
U U
C C
o o
c c.
8 in
a>
0 0
Q Q
••- CO
-------
Post-mining emissions are expected to be about 0.3 Tg in 2000 and 0.4 Tg in 2010.
Emissions from degasification systems in the CAB are expected to increase significantly in the
future, reaching 0.2 to 0.6 Tg (10.4 to 31.3 Bcf; 0.3 to 0.9 Bern) in 2000 and 0.2 to 0.8 Tg
(10.4 to 41.7 Bcf; 0.3 to 1.2 Bern) in 2010. This represents 28 to 51 percent of projected
underground emissions from the basin and 28 to 33 percent of the total projected
degasification emissions. If project developments in Virginia are successful, it is possible that
several mines could sell the methane from their degasification systems instead of venting it to
the atmosphere.
Black Warrior Basin
Total coal production for the Black Warrior Basin is forecast to be 33 to 37 million tons
in 2000 and 51 to 59 million tons in 2010. Approximately 45 percent of total production is
expected to come from underground mines.
Methane emissions estimates for the Black Warrior Basin range from 0.6 to 0.7 Tg
(31.3 to 36.5 Bcf; 1.0 to 1.1 Bern) in 2000 and from 0.9 to 1.2 Tg (46.9 to 62.5 Bcf; 1.4 to 1.9
Bern) in 2010. Underground mining emissions are expected to account for 0.5 to 0.7 Tg in
2000 and 0.8 to 1.0 Tg in 2010, while both surface mining and post-mining emissions are
likely to remain below 0.1 Tg.
Degasification systems are expected to emit less than 0.1 Tg (5.2 Bcf; 0.2 Bern) of
methane in 2000 and 0.1 to 0.2 Tg (5.2 to 10.4 Bcf; 0.2 to 0.3 Bern) in 2010. This represents
up to 20 percent of methane emissions projected from the basin's underground mines and
up to 11 percent of total projected degasification emissions. Mines in the BWB are currently
selling the methane from their degasification systems to the local pipeline, and it is
anticipated that this practice will continue. For 2000 and 2010, it was estimated that mines in
the Warrior basin would recover 0.23 Tg of methane. This amount is not included in the
underground emissions estimates for this basin.
Other U.S. Coal Basins
The rest of the U.S. coal production for 2000 and 2010 is largely in the west and will
be primarily surface production. Total coal production in these basins in 2000 is estimated to
be 477 to 710 million tons, of which underground mining should contribute about 25 percent.
Production in 2010 is forecasted to be 789 to 902 million short tons; underground mining is
projected to account for about 32 percent of total production.
Total emissions for these basins are estimated to range from 1.0 to 1.5 Tg (52.2 to
78.2 Bcf; 1.5 to 2.2 Bern) in 2000 and from 1.5 to 2.3 Tg (78.2 to 119.9 Bcf; 2.2 to 3.4 Bern) in
2010. Of these emissions, underground mining emissions estimates range from 0.7 to 1.1 Tg
in 2000 and from 1.2 to 1.6 Tg in 2010. Surface mining emissions are expected to range from
0.1 to 0.2 Tg in both 2000 and 2010. As noted earlier, surface mining emissions from these
basins are low because coals in these basins are not very gassy.
Methane emissions from degasification systems in these basins are expected to be
low compared to the Eastern basins. Estimates range from 0.1 to 0.4 Tg (5.2 to 20.9 Bcf; 0.2
to 0.6 Bern) in 2000 and from 0.2 to 0.6 Tg (10.4 to 31.3 Bcf; 0.3 to 0.9 Bern) in 2010. Most
of these emissions are anticipated to come from mines in Colorado, Illinois and Utah. These
numbers represent 18 to 37 percent of expected underground emissions from these basins
and between 14 and 20 percent of projected U.S. degasification emissions. One Utah mine is
3-26
-------
currently selling methane to pipelines, and it is anticipated that these sales will continue.
Accordingly, it was estimated that approximately 0.02 Tg would continue to be recovered and
utilized from western mines. This amount is not included in the emissions estimates for
underground mines in 2000 and 2010.
3.6 LIMITATIONS OF THE ANALYSIS
The estimates presented in this analysis are shown as ranges to reflect the
uncertainties surrounding them. In most cases, these uncertainties stem from the lack of
available measured data and the need to make assumptions. In developing the assumptions,
a wide variety of experts were consulted to accurately reflect mining experience. The key
uncertainties are described below and are grouped according to their application to current
or future estimates.
3.6.1 Uncertainties in 1988 Emission Estimates
Degasification System Emissions
As mentioned previously, mines are not required to report the emissions from their
degasification systems, and it is not straightforward to even identify which mines have these
systems. This analysis thus includes two key uncertainties related to degasification system
emissions:
(1) EPA identified those mines with such systems in place based on published
information and discussions with mine personnel and government agencies
such as MSHA and USBM. The list of mines, which is shown in Exhibit 3-7, is
believed to be reasonably complete, but it is possible that some mines with
degasification systems were not included. To the extent that this is the case,
emissions will be underestimated.
(2) For those mines with degasification systems, EPA made assumptions regarding
the share of total emissions captured by the systems. Mines will use different
methane recovery methods and have different strategies regarding the
percentage recovery in degasification as compared to ventilation systems.
Based on discussions with industry personnel, it was assumed that a typical
mine would recover between 40 and 65 percent of total emissions in its
degasification system. These assumptions were applied to the reported
ventilation emissions to estimate total emissions from mines with degasification
systems in place. To the extent that the degasification strategy varies by mine
or by basin, emissions could be over- or under-estimated.
Surface Mining Emissions
Direct emission measurements are currently unavailable for surface mines. It is known
that surface mined coal has lower methane contents, however, and that emissions per ton of
coal mined are likely to be low. Given the large amount of surface mining in the United
States, these emissions should not be overlooked. In this analysis, it was assumed that total
emissions would range from one to three times the methane content of the mined coal. It is
3-27
-------
likely that methane emissions will follow similar patterns in surface and underground mines,
although it was assumed in the low case that there would be no significant methane
emissions from the surrounding strata. EPA's Office of Research and Development is
currently investigating surface mining emissions in more detail and should resolve many
uncertainties related to these emissions.
Post-Mining Emissions
Direct emission measurements are currently unavailable for methane emissions from
post-mining activities. The level of emissions will depend on coal characteristics (such as
permeability) and the manner in which it is handled. In these estimates it was assumed that
about 25 to 40 percent of the gas in coal would be released after the coal leaves the mine.
To the extent that any specific coal desorbs gas more rapidly or more slowly than the
average, the emissions could be over or underestimated.
3.6.2 Uncertainties in Future Emission Estimates
Coal Production
Future coal production levels represent one of the major uncertainties associated with
estimating future methane emissions. To the extent that coal production is over- or under-
estimated, emissions will be correspondingly too high or too low. In addition, as discussed
previously, methane emissions vary by basin because of mining methods, geologic
characteristics, and other factors. The estimates in this report were prepared on a basin-by-
basin basis, using coal production forecasts that included basin-specific estimates. To the
extent that coal production shifts among basins, methane emissions would be affected.
Impact of Clean Air Act Amendments of 7990
A related uncertainty is the impact of the 1990 Clean Air Act Amendments' provisions
on acid rain. Under this regulation, utilities are required to reduce SO2 emissions by 10
million tons by 2000 and future emissions are capped. In addition, NOX emissions are also
reduced. It is likely that utilities will comply with this act using a variety of measures,
including emission control technologies, coal switching (high to low sulfur), fuel switching
(coal to natural gas or oil), and demand reduction. With the exception of emission control
technologies, these actions would be expected to reduce coal demand in general and high
sulfur coal demand in particular, both of which would reduce methane emissions associated
with coal mining. The coal production forecasts used in this analysis reflect two possible
scenarios under the acid rain legislation. There is currently a great deal of uncertainty
regarding the most likely utility compliance choices, however, and significant surprises in
utility actions could affect future methane emissions from coal mining.
Future Emission Factors
Future estimates were prepared assuming that the 1988 emission factors for each coal
basin would not increase over time. This assumption is probably adequate for estimating
surface and post-mining emissions. For underground mines, however, it is possible that
future emissions will be underestimated because of the increasing gassiness of the mines.
Over time, underground mines in many coal basins may become gassier, as the less gassy
coal seams are depleted. To reflect this, the basin-specific emission factors should increase.
3-28
-------
Given the uncertainty in the likely rate of increase, however, and the lack of comprehensive
historical data on which to base such an assumption, it was conservatively assumed in this
analysis that underground mining would not become gassier over time,
Utilization of Recovered Methane
Because the profitability of coal mine methane recovery and utilization projects will
vary significantly based on numerous economic and other factors, it is difficult to estimate the
quantity of methane that would be recovered from each basin in 2000 and 2010. This chapter
assumes that methane recovery in each basin remains at 1988 levels. However, assuming
that real gas prices increase over the next twenty years and that a number of extant barriers
to coal mine methane projects - such as the legal ownership issues - are resolved, the
quantity of methane recovered could increase significantly (see coal mining chapter in the
Report to Congress on Options for Reducing Methane Emissions from Anthropogenic
Sources in the U.S.).
3.7 REFERENCES
AAPG (American Association of Petroleum Geologists). 1984. Coalbed Methane Resources
of the United States, AAPG Studies in Geology Series #17.
Baker, E.G., F. Garcia, and J. Cervik. 1988. Cost Comparison of Gob Hole and Cross-Measure
Borehole Systems to Control Methane in Gobs, U.S. Bureau of Mines, Rl 9151.
Baker, E.G., R.H. Grau and G.L. Finfinger. 1986. Economic Evaluation of Horizontal Borehole
Drilling for Methane Drainage from Coalbeds, U.S. Bureau of Mines, 1C 9080.
British Coal. 1991. Quantification of Methane Emissions from British Coal Mine Sources, report
produced for the Working Group on Methane Emissions, the Watt Committee on
Energy.
Dixon, C.A. 1987. "A Miner's Viewpoint," Proceedings of the 1987 Coalbed Methane
Symposium, pp.7-10, Tuscaloosa, Alabama.
Duel and Kim. 1988. Methane Control Research: Summary of Results, 1964-80, Bureau of
Mines Bulletin, B-687.
EIA (Energy Information Administration). 1992. Annual Energy Outlook 1992, 1992, U.S. Dept.
of Energy, Office of Integrated Analysis and Forecasting.
Environment Canada. 1992. Canada's Greenhouse Gas Emissions Estimates for 1990. Draft
April, 1992.
Hunt, A.M., and D.J. Steele. 1991. Coalbed Methane Technology Development in the
Appalachian Basin, Gas Research Institute Topical Report.
IPCC. (Intergovernmental Panel on Climate Change, US/Japan Working Group on Methane).
1992. Technological Options for Reducing Methane Emissions: Background
Document of the Response Strategies Working Group. Draft January 1992.
3-29
-------
Kelafant, J.R., and C.M. Boyer. 1988a. A Geologic Assessment of Natural Gas from Coal
Seams in the Central Appalachian Basin, Gas Research Institute Topical Report.
Kelafant, J.R., D.E. Wicks, and V.A. Kuuskraa. 1988b. A Geologic Assessment of Natural Gas
from Coal Seams in the Northern Appalachian Coal Basin, Gas Research Institute
Topical Report.
Kirchgessner, D.A., S.D. Piccot, and A. Chadha. 1992a. Estimation of Methane Emissions
from a Surface Coal Mine Using Open-Path FTIR Spectroscopy and Modeling
Techniques. Chemosphere. (In Press).
Kirchgessner, D.A., S.D. Piccot, and J.D. Winkler. 1992b. Estimate of Global Methane
Emissions from Coal Mines. U.S. Environmental Protection Agency 1992.
Kissell, F.N., C.M. McCulloch, and C.H. Elder. 1973. The Direct Method for Determining
Methane Content of Coalbeds for Ventilation Design, U.S. Bureau of Mines Information
Circular 7767, 17pp.
Niewiadonski, G. 1992. Personal communication. U.S. Mine Safety and Health
Administration. September 23, 1992.
Trevits, M.A., G.L Finfinger, and J.C. Lascola. 1991. "Evaluation of U.S. Coal Mine
Emissions," in Proceedings of the Fifth U.S. Mine Ventilation Symposium. Society for
Mining, Metallurgy and Exploration, Inc., Littleton, Colorado.
USDOE (U.S. Department of Energy). 1989. Coal Production 1988. Energy Information
Administration, Office of Coal, Nuclear, Electric and Alternative Fuels, Washington, DC.
DOE/E1A-0118(88).
USDOE (U.S. Department of Energy). 1991. Report to the Congress of the United States:
Limiting Net Greenhouse Gas Emissions in the United States (Volume 11 - Energy
Responses). Office of Environmental Analysis. Washington, DC. DOE/PE-XXXX
(Advance copy).
USDOE (U.S. Department of Energy). 1992. Energy Consumption and Conservation Potential:
Supporting Analysis for the National Energy Strategy, Energy Information
Administration, Office of Energy Markets and End Use, Washington, D.C., DOE/EIA
UC-98, p.20.
USEPA (U.S. Environmental Protection Agency). 1990a. Methane Emissions and
Opportunities for Control: Workshop Results of Intergovernmental Panel on Climate
Change. Office of Air and Radiation (ANR-445). Washington, DC. EPA/400/9-90/007.
USEPA (U.S. Environmental Protection Agency). 1990b. Methane Emissions From Coal
Mining: Issues and Opportunities for Reduction. Prepared by ICF Resources
Incorporated for Office of Air and Radiation, USEPA, Washington, DC.
3-30
-------
CHAPTER 4
METHANE EMISSIONS FROM LANDFILLS
U.S. Methane Emissions
from All Sources
Coa I M i n i ng
Natui a I Ga<3 System^.
Annual Landfill
Methane Emissions
GIobaI Emissions U 5 Em
Emissions Summary
Source
Large Municipal Landfills (152 Total)
Medium Municipal Landfills (1,137)
Small Municipal Landfills (4,744)
Industrial Landfills
1 990 Emissions
(Tg)
2.6 - 4.2
3.3 - 6.0
0.9- 1.5
0.6 - 0.9
Partially
Controllable
/
/
/
/
Total1'2 8.1 - 11.8
1 The uncertainty in the total is estimated assuming that some of the uncertainty for
each source is independent. Consequently, the uncertainty range for the total is more
narrow than the sum of the ranges for the individual sources.
2 The total does not include an additional 1 .5 Tg of methane recovered from landfills
that was flared or used as an energy source.
4.1 EMISSIONS SUMMARY
Landfills are the largest single anthropogenic source of methane emissions in the
United States. U.S. landfill methane emissions in 1990 are estimated to range from about 8.1
to 11.8 Tg/yr, or about 36 percent of total U.S. methane emissions. U.S. landfill emissions
account for about twenty to forty percent of the estimated 20 to 70 Tg/yr of global landfill
emissions (IPCC 1992).
4-1
-------
Municipal solid waste landfills account for about 90 to 95 percent of landfill emissions
of methane, and industrial landfills account for 5 to 10 percent. Although an estimated 6,000
landfills emit methane in the U.S., about 1,300 account for nearly all the methane emitted. Of
these, about 900 landfills account for 85 percent of the waste in landfills and 75 percent of
the methane emitted. The nineteen largest landfills account for about 25 percent of the waste
in landfills and 20 percent of the methane generated. Of the total methane generated by
landfills, about 10 percent is currently recovered for use as energy.
Methane is produced during the anaerobic decomposition of organic material in
landfills by bacteria. Methane production typically begins 1 to 2 years after waste placement
in a landfill and may last from 10 to 60 years, or longer. Unless this gas is collected by a
recovery system, it migrates through the landfill and most of it is emitted into the atmosphere.
A small portion of the methane (about 10 percent) may be oxidized (converted to carbon
dioxide and water) before it is emitted.
A number of factors influence the
amount of methane produced from a landfill
and cause different landfills to have different
levels of emissions. These factors include
the following:
Although an estimated 6,000 landfills
emit methane in the U.S., about 1,300
account for nearly all the methane
emitted. Of these, about 900 landfills
account for 85 percent of the waste in
landfills and 75 percent of the methane
emitted.
• Quantity of organic waste.
Because the organic material
in the waste sustains the
microorganisms that produce
methane, larger quantities of
organic material increase the methane producing capacity of a landfill.
Therefore, the quantity of paper, food wastes, yard wastes, and other organic
materials placed in a landfill is a dominant factor affecting emissions.
• Contact with oxygen. Methane is only produced under anaerobic conditions
where free oxygen is not in contact with the waste. In general, as the depth of
the landfill or the density of the waste increases, anaerobic conditions will exist
that promote methane production.
Other factors that influence methane production and cause differences in emissions among
landfills include nutrient availability, the presence of toxic compounds, landfill temperature,
moisture content, and pH. These factors have been addressed in this analysis by estimating
U.S. emissions based on information from about 100 landfills; however, substantial
uncertainty remains.
The amount of methane produced by U.S. landfills is expected to increase over the
next two decades. Although the rate at which waste is landfilled annually is not expected to
increase over this period because of current trends toward recycling and alternative disposal
methods, the aggregate amount of methane producing waste accumulated in landfills will
increase. However, an EPA landfill rule has been proposed that may significantly reduce
emissions, depending on which standards are adopted.
1 The proposed rule is the Standards of Performance for New Stationary Sources and Guidelines for Control of
Existing Sources: Municipal Solid Waste Landfills (USEPA 1991c)
4-2
-------
Without considering the possible impact of the proposed landfill rule, emissions in the
years 2000 and 2010 are estimated to be about 9 to 13 Tg/yr. If the 150 Mg/yr stationary
source standard for new landfills and the source guidelines for existing landfills are adopted,
emissions in the years 2000 and 2010 are estimated to be approximately 5 to 9 Tg/yr. If the
25 Mg/yr standard and guidelines are adopted, emissions in the years 2000 and 2010 are
estimated to be approximately 4 to 7 Tg/yr.
As can be seen by the relatively large emissions range, the estimates of methane
emissions remain uncertain. This uncertainty is due to a variety of factors. Most importantly,
there are very few measurements of actual methane emissions from typical landfills. Instead,
this and other analyses rely on data describing the amount of methane recovered from
landfills that are extracting methane for use or sale, and these data are an imperfect surrogate
for estimating rates of methane emissions. Additionally, U.S. landfills are not adequately
characterized in terms of the site-specific factors that affect methane production and
emissions to allow these factors to be considered in making the national estimates. Finally,
the amount of organic wastes expected to be disposed in the future is also uncertain.
4.2 BACKGROUND
Although municipal solid waste has been disposed in dumps for many decades, the
widespread use of sanitary landfill designs that promote methane generation and emission is
relatively recent. Anaerobic conditions present in sanitary landfills allow microbes to break
down organic material and produce methane in landfills. A variety of site-specific factors
determine the amount of methane generated and released from the landfill over time.
4.2.1 Landfill Refuse Management
Landfills currently receive over 70 percent of the solid waste generated in the U.S.
(USEPA 1990). Landfills evolved from unregulated "dumps" which traditionally received a
majority of the country's refuse. A dump typically was a hole in the ground where refuse was
deposited and then frequently burnt. This disposal method created numerous problems,
including: attraction of rats and flies; uncontrolled fires and air quality problems; and
groundwater contamination.
In response to these problems,
additional environmental controls were
placed on dumps which then became
known as sanitary landfills. Sanitary landfills
are designed to confine the refuse to the
smallest practical area and volume. At the
end of each day of operation or at more
frequent intervals, sanitary landfills are
covered with a 6 inch layer of earth.
Sanitary landfills have been used widely
since the early 1970s.
There are very few measurements of
methane emissions from landfills.
Consequently, analyses must rely on
data describing the amount of methane
recovered from landfills for use or sale,
which are an imperfect surrogate for
data on methane emission rates.
Most sanitary landfills are divided into individual cells. Each cell is effectively a
separate landfill that is developed at a given time. When a cell is filled, an additional cover
4-3
-------
consisting of clay and topsoil, referred to as a cap, is applied and the cell is closed. If the
landfill is not separated into cells, this final cap is applied after the closure of the entire
landfill. This placement of a cap limits the infiltration of moisture and allows the land to be
reclaimed and used for other purposes such as recreational parks.
Although sanitary landfills eliminated many of the problems associated with dumps,
other problems persist. Unless strict engineering principles are applied, groundwater
contamination can result. Landfills also produce large quantities of methane gas which create
both a safety hazard and a threat to the global environment (USEPA 1991 a).
4.2.2 Landfill Methane Production
Landfills produce methane because the organic material in the waste decomposes in
an environment free of molecular oxygen. Organic material is contained in yard waste,
household garbage, food waste, and paper. Because about 70 percent of the waste placed
in landfills is organic, the potential for methane production is great (USEPA 1992a).2 (See
Exhibit 4-1.)
Refuse decomposition is a natural process in which microorganisms derive energy
and material for growth by metabolizing organic material. In an anaerobic environment that
lacks free oxygen, the organic material is decomposed by anaerobic and facultative bacteria
(living in the absence or presence of oxygen). The end products of anaerobic decomposition
are methane, carbon dioxide, and stabilized organic material.3
The decomposition process can be described in terms of five stages: aerobic;
hydrolytic; acid forming; methanogenic; and stabilization. Although each stage is described
sequentially, different stages almost certainly occur simultaneously within a landfill.
Stage I: Aerobic. The aerobic stage begins when the refuse is generated and
while it remains in contact with free oxygen. Aerobic bacteria decompose the
organic material in the waste using molecular oxygen. The end-products of
aerobic decomposition are carbon dioxide, water, heat, and stabilized organic
material. The aerobic phase continues until the available oxygen is consumed.
In general, the aerobic stage may last from several days to several weeks. If
the landfill refuse is poorly compacted or remains uncovered, aerobic
conditions will continue near the surface of the landfill.
Stage II: Hydrolytic. In this stage, complex organic materials (carbohydrates,
proteins, lipids) in the refuse are broken down through the hydrolytic action of
enzymes. Enzymes are proteins formed by living cells that act as catalysis in
metabolic reactions. The enzyme cellulase is responsible for breaking down
carbohydrates such as cellulose and starch into simple sugars (e.g., glucose).
Simple organic acids are produced when the enzyme lipase breaks down fats
(lipids) into smaller chained fatty acids and the enzyme protease breaks down
2 The remaining 35% may not be completely inorganic. EPA's Office of Research and Development is
conducting research on the relative gas potential of various biodegradable waste streams.
3 Stabilized organic material is material that is not broken-down or decomposed further.
4-4
-------
Exhibit 4-1
Materials Discarded in Municipal Solid Waste Landfills in 1990 (weight basis)
ORGAN I C!
Yard Wastes I9R
Papei 3
Textiies 3% ^
Food Wastes
Otnei
MORGAN
Source: USEPA 1992a.
* The material may contain organics that break down very slowly (and thus do not significantly contribute to
methane generation).
proteins into amino acids. The amount and rate of breakdown can vary
substantially and depend on the enzymes present, the characteristics of the
refuse and environmental factors such as pH, temperature, and moisture.
Stage III: Acid Forming. Anaerobic and facultative bacteria reduce (ferment)
the simple sugars produced in Stage II to simple organic acids. Acetic acid is
the primary product of the breakdown of carbohydrates, though other organic
acids such as propionic acid and butyric acid can be formed. In addition,
metabolic hydrogen4 and carbon dioxide are produced. The organic acids,
along with metabolic hydrogen and carbon dioxide form a substrate5 for the
methane forming bacteria in Stage IV. Unlike the methane forming bacteria,
4 Metabolic hydrogen refers to hydrogen that is used in metabolic reactions but is not necessarily found in a
free molecular state as H 2
5 Substrate refers to the material that the bacteria use for growth and metabolism.
4-5
-------
the acid formers are fast growing and thrive at a broad range of temperature
and pH conditions. With acetic acid as an end product, the inputs and outputs
of the breakdown of a simple sugar molecule (glucose) in Stage III can be
represented as:
C6H12°6 + 2H2° ~> SCHjjCOOH + 2CO2 + 4H2
glucose + water —> acetic acid + carbon dioxide + metabolic
hydrogen
• Stage IV; Methanoqenic. Methane producing bacteria (methanogens) convert
the simple organic acids, metabolic hydrogen, and carbon dioxide from Stage
III into methane and carbon dioxide.6 Methanogens are strict anaerobes and
cannot tolerate the presence of molecular oxygen. Methanogens multiply
slowly and are very sensitive to temperature, pH, and substrate composition.
Methane production may not begin until one to two years after refuse
placement and may continue for ten to sixty years (USEPA 1991c). With acetic
acid, metabolic hydrogen and carbon dioxide as substrate, the inputs and
outputs of the methanogenic process can be represented as:
2CH3COOH —> 2CH4 + 2CO2
acetic acid —> methane + carbon dioxide
4H2 + C02 —> CH4 + 2H20
metabolic + carbon dioxide —> methane + water
hydrogen
Stage V: Stabilization. During this final phase, microbial action and gas
production cease as the degradable organic material is completely consumed.
The landfill is considered stabilized when all biological and chemical activity
within the landfill becomes negligible.
Exhibit 4-2 shows the relative levels of methane, carbon dioxide, and other gases during each
stage of the decomposition process. Maximum methane production may not be reached until
a number of years after the refuse is in place.
4.2.3 Site-Specific Factors Affecting Methane Production
Many factors influence the duration of each stage of the microbiological decom-
position process and the amount of methane generated per quantity of refuse. Conse-
quently, methane production may vary significantly from landfill to landfill and from area to
area within an individual landfill. The key factors influencing methane generation are: amount
of refuse in place; refuse characteristics; and moisture, In addition, temperature and pH also
affect methane production. These factors are discussed below.
6 The pathway of CO 2 and acetic acid reduction to CH4is complex and involves numerous steps. The
reactions shown below only indicate the grass inputs and outputs of this stage.
4-6
-------
Exhibit 4-2
Theoretical Model of Relative Gas Composition in a Landfill
Stage
100 -•
IV
a
P 80-
#
W 60 -
ct
n
•H
4-J
w 40 -
n
PH
a
o
u
20 -
w
ctf
0 ^
I II III IV V
\
I
02
A
(
^
r \
{
I x-^
A^-nT-
CH4
>'
C02
>
N2
>^
Gas
Product i on
Ends In
x^ N^
\ \
\ \
N \
\ \
\ \
\ \
\ N
\ \
\ \
\ \
\ N
\ \
\ \
\ \
Time After Placement
Stage 1: Aerobic Stage IV: Methanogenic
Stage II: Hydrolytic Stage V: Stabilization
Stage III: Acid Forming
Source: Emcon 1982.
Refuse in Place
The methane producing capacity of a landfill is directly related to the total quantity of
refuse in place in the landfill. Because the refuse is the substrate for the bacteria
decomposing the refuse, larger refuse quantities will generally produce more methane,
(assuming the composition of the refuse does not change).
Refuse Characteristics
Refuse composition affects methane production for several reasons, including:
• Organic content. The composition of the refuse determines the maximum
methane producing capacity of the refuse. The greater the organic content of
the refuse, the greater the methane producing capacity of the landfill. In
addition, the greater the energy content and degradability of the refuse, the
greater the methane producing capacity. For example, food wastes are highly
4-7
-------
biodegradable and will produce more methane per unit volume than will wood
waste or textile waste.
Nutrient availability. Bacterial growth depends on the availability of nutrients
such as nitrogen, phosphorus, sulfur, potassium, sodium, and calcium.
Deficiency in one or more of these nutrients will inhibit bacterial growth and
methane formation.
• Particle size. The density and consistency of the refuse affects the activity of
bacteria. Smaller particle sizes increase the surface area on which reaction
may occur, promoting greater bacterial activity and methane production.
Moisture
Moisture is essential for anaerobic decomposition (Loehr 1984). First, water is
necessary for bacterial cell growth and metabolism. In addition, water transports nutrients
and bacteria to other areas within the landfill. Some of the necessary moisture is supplied by
the incoming refuse which generally has a moisture content between 20 and 30 percent
(although this can vary greatly). Food and garden refuse, in particular, has a high moisture
content. Other factors that affect the moisture content include: surface water infiltration;
groundwater infiltration; water produced during the decomposition process; and liquid
additions (e.g. sludge).
Other Factors
Other factors influencing methane production in a landfill include the following:
• Temperature. Temperature affects the growth rate of the bacteria responsible
for methane formation. In general, methane production increases with rising
temperature in the landfill.
Under anaerobic conditions, most landfills exhibit temperatures between 29°C
and 38°C with the average temperature in the anaerobic zone around 35°C
(Gunnerson and Stuckey 1986). Soil covers and top layers of landfills insulate
the anaerobic area and help maintain these temperatures. At depths greater
than six to twelve feet the landfill temperature may become independent of
ambient air temperatures (Rettenberger and Tabasaran 1980; Stegman 1986;
Gunnerson and Stuckey 1986). Therefore, air temperature can be a less
important factor for deeper landfills.
• pH. Methane formation occurs within a pH range of 6.5 to 8.0; beyond this
range production ceases. Methanogens are most productive when the pH is
between 6.8 and 7.2. The pH of most landfills is in this range.
4.3 METHODOLOGY
Methane emissions from landfills are estimated by developing statistical relationships
between measured gas recovery rates and landfill waste quantities, and applying the
relationships to the population of U.S. landfills. A data set of 99 landfills was developed and
verified by contacting landfill operators. Data on 85 of the 99 landfills were used to estimate
4-8
-------
emissions. Data describing the population of landfills in the U.S. remains somewhat
uncertain, adding to the uncertainty in the final estimates.
4.3.1 Background
The preferred approach for estimating methane emissions from landfills in the U.S.
would be to measure the actual emissions from each of the approximately 6,000 operating
landfills and the many thousands of closed landfills. Unfortunately, such an approach is
highly impractical. As an alternative, previous efforts to estimate methane emissions from
landfills have taken one of two approaches:
1) Determine the emissions "potential" of a representative quantity of refuse
through theoretical considerations (e.g., carbon content) or laboratory
simulation. Scale these values to the national level by estimating the quantity
of refuse in all landfills (e.g., Bingemer and Crutzen 1987; Augenstein 1990).
2) Use the data that are available to determine the actual generation of methane
per quantity of refuse and multiply this value by the estimated quantity of refuse
in all landfills (e.g., USEPA 1992c).
Most previous estimates of methane emissions from landfills are based on the first approach
and rely on kinetic models of landfill gas formation or on laboratory simulation experiments.
These methods are not based on actual measured methane generation rates from landfills. A
limitation of this approach is that it is often difficult to extrapolate with accuracy from
theoretical or laboratory results to field conditions. The second approach, which relies on
actual field measurements, was chosen for this analysis because it relies on actual data rather
than theoretical results.
The data used for this analysis is provided by the many landfills that recover methane
for energy use from all or a portion of their refuse. An existing data base of such landfills and
their measured gas recovery rates was verified and supplemented through direct contact with
landfill operators. The modified data base includes verified information on 85 landfills that
may be viewed as representative of the landfills that contain the majority of the waste in place
in the United States.
A statistical model was developed from the verified data base that establishes the
relationship between quantity of refuse in place and methane production in the landfill. The
analysis builds upon analyses performed by the U.S. EPA Air and Energy Engineering
Research Laboratory (AEERL) which indicate that a relatively simple model can be used to
estimate methane generation rates. For 21 landfills, AEERL conducted site visits and
examined in detail the relationships among methane recovery and (1) refuse quantity; (2)
refuse characteristics (such as moisture content, temperature, and pH); and (3) landfill
characteristics (such as age, depth, volume, and surface area). Although many of these
factors were found to be correlated with observed gas recovery rates, AEERL chose refuse
quantity as the preferred variable for explaining variation in the observed gas data (USEPA
1992c) because it explains much of the variation and is readily available. The simple model
developed in this report, based on information from 85 landfills that are representative of the
larger "population" of landfills and vary in terms of depth, age, regional distribution, and other
factors, should provide a robust estimate of methane emissions from landfills and the
associated uncertainty in these emissions (Peer et al. 1992).
4-9
-------
4.3.2 Steps Used to Estimate Emissions from Landfills
To estimate methane emissions from landfills, it is necessary to collect information on
emissions and other characteristics from individual landfills. This information can be used to
develop a model that relates emissions to landfill characteristics. U.S. national emissions are
estimated by collecting data on the characteristics of the U.S. landfill population and applying
this information to the emissions model. These steps may be formulated as follows:
Model Development
• Collect and Verify Landfill Data including methane generation and waste
quantity data at representative landfills across the U.S.
Specify and estimate a statistical model using the data obtained in the previous
step. Relate methane production to the quantity of waste in place at each
landfill and estimate the statistical uncertainty of the model emission estimates.
Application of Model
Identify the number, size, and geographic location of landfills in the U.S. in
order to define "model" landfills that represent the over 6000 landfills in the U.S.
• Estimate methane production for each model landfill using the statistical model.
Evaluate the uncertainty of these estimates for each model facility and estimate
"low" and "high" production rates.
• Estimate National Emissions by multiplying the mean, low, and high production
estimates for each model landfill by the estimated number of these landfills in
the U.S. Adjust these estimates for the methane that is recovered or otherwise
not emitted to the atmosphere.
These steps were carried out as described below.
Mode/ Deve/opment -- Co//ect and Verify Landfill Data
Methane is recovered and combusted at many landfills in the U.S. either for safety
reasons or to utilize the energy. Because very little work has been performed to measure
methane generation from landfills without recovery systems, it is data from sites with recovery
systems that provide the best information on emissions from landfills. The data used in this
report are based on the following sources:
• Governmental Advisory Associates, Inc. (GAA 1991) developed a database of
landfills currently recovering landfill gas in the U.S. that includes the following
information on each landfill: the gas recovered; the methane content of the
gas; waste in place; landfill area; landfill area under the influence of the
recovery system; depth; age; and many other variables. For this study, the
4-10
-------
GAA data on gas recovery and waste under the influence of the recovery
system were used to develop the emissions model.7
Detailed information was also collected on six landfills not included in the GAA
database. Five of these six were from the AEERL study (USEPA 1992c) and
information on one was obtained directly from the landfill operator (USEPA
1992b).
All the data from these sources used in this study were verified by contacting the
landfill and/or gas recovery system owners and operators. This verification process was
important because the GAA data were not originally collected to provide a basis for
estimating methane production. Furthermore, the data verification process identified two
important areas where the GAA data could be misinterpreted for purposes of estimating
emissions in this study:
For about one third of the landfills in the GAA database the data on landfill
area and landfill area under the influence of the gas recovery system were
revised. In some cases, the original GAA data report the maximum design area
of the landfill and not the area currently containing waste.
For about 15 percent of the landfills the gas production data were revised to
better reflect the actual methane recovered from the landfill, as opposed to
system capacity or gas or power sales.
Because the amount of gas recovered is an under-estimate of the amount of gas
produced in the landfill, the owners and operators were also asked to estimate the gas
collection efficiency for their landfill. The collection efficiency is the ratio of the gas recovered
to the gas produced in the landfill and is always less than 1.0. Collection efficiency is not an
easily measured quantity, and consequently a subjective assessment was required.
To develop as accurate an estimate as possible for collection efficiency, the factors
affecting collection efficiency were discussed with the landfill owners and operators. These
factors include: the integrity and permeability of the cap, the well spacing, the landfill depth,
and the vacuum pulled on the wells. In cases where the landfill owners and operators could
not estimate the collection efficiency, a 75 percent value was assumed. This value is based
on the 70 to 80 percent collection efficiency that is generally reported for well designed and
operated landfill gas collection systems.8 Based on this information, the methane generation
rate at each landfill was estimated as:
Methane Generation = Methane Recovered / Gas Collection Efficiency (4.1)
7 The refuse under the influence of the recovery system was estimated using the ratio of the landfill area under
the influence of the recovery system to the total landfill area times the total refuse in place. This information was
verified for most of the GAA landfills by contacting the landfill operators.
8 About half of the owners and operators were able to report collection efficiency estimates. The 75 percent
default value was applied to about 25 percent of the landfills. For the remaining 25 percent, a judgment was made
to use a collection efficiency above or below the 75 percent default value. In many cases a collection efficiency of
90 to 95 percent was adopted because the owner or operator reported that frequent monitoring for elevated
methane levels revealed no leaks.
4-11
-------
To be useful for estimating methane emissions, the landfills in the data set must be
representative of landfills generally in the U.S. Because nearly all the landfills in the data set
currently recover landfill gas to produce energy, and because most landfills in the U.S. do not
recover gas, there was concern that these landfills may not be typical. Based on discussions
with industry experts, it was determined that virtually all landfills with 1 million megagrams or
more of welled waste in place can produce enough gas to make them candidates for gas
recovery. Once a landfill has at least this much waste, other factors tend to control whether a
recovery project is actually implemented, including: access to the landfill; the proximity of a
customer for the gas; the cost of connecting to the electric power grid to seli electricity; and
the price at which electricity can be sold.
Based on this information from industry experts, the 85 landfills in the data set with 1
million megagrams or more of welled waste are believed to be representative of landfills in
this size range. They are in the data set not because they are unusually gassy, but because
other factors allowed their owners and operators to proceed with gas recovery projects.
Consequently, the data on these landfills are appropriate for developing an emissions
estimate.
However, the 14 smaller landfills in the data set may not be representative; these
landfills may have gas recovery projects because they are unusually gassy. Consequently,
the data on these landfills are not appropriate for developing an emissions estimate. These
data are therefore excluded from this analysis.
The resulting data set used for this study includes 85 landfills with waste in place in
spanning a range from 1.2 million Mg to over 30 million Mg (see Exhibit 4-3). This range
represents over 60 percent of U.S. landfilled waste (USEPA 1987). The full data set is
presented in Appendix A of this chapter.
Model Development •• Specify and Estimate Statistical Model
The AEERL analyses show that waste in place explains the largest portion of the
observed variation in methane production (USEPA 1992c). Using the AEERL approach with
the 85 landfills in the data set developed for this study, the relationship between landfill welled
waste in place (W) and methane generation (CH4) for landfills with more than 1 million
megagrams of welled waste in place was estimated as
CH4 (m3/min) = 13.8 + 3.7W (106 Mg) (4.2)
4.4 11.3
R2 = 0.61 n = 85 Range: 1,215,400 Mg < w < 45,360,000 Mg
The figures below each coefficient estimate are the t-statistics for the 95 percent confidence
interval for the null hypothesis that the true coefficient values are zero. For this model, both
the intercept term and the coefficient for waste mass (W) are significantly different from zero
at the 95 percent confidence level.
Because the 85 landfills used to develop Model 4.2 all have over 1 million megagrams
of welled waste in place, Model 4.2 is not appropriate for estimating emissions from smaller
landfills with less waste. Consequently, an alternative approach was required to estimate the
emissions for the many landfills that fall into the smaller size categories. For this analysis,
methane generation for landfills with welled waste in place less than 1 million megagrams was
estimated by:
4-12
-------
Exhibit 4-3
Distribution of Landfills by Total Waste in Place
10 -
1-2 2-3 3-4 4-5 5-10 10-20
Total Refuse in Place (Million Mg)
Data set for this study developed from GAA (1991); Thorneloe (1992a); USEPA (1992b); USEPA (1992c).
calculating the average methane generation per megagram of waste for the 85
landfills used to estimate Model 4.2; and
averaging these 85 values to obtain the average methane generation rate per
megagram of waste.
This approach produces an average rate of methane generation per unit of waste that reflects
the diverse landfill characteristics represented in the data set. To be valid, this approach
implicitly assumes that with the exception of size, the relevant characteristics of small landfills
are similar to the characteristics of the larger landfills in the data set.9
Using this approach, the average methane generation per million megagrams of waste
is given by:
CH4 (m3/min) = 7.1 W (106 Mg) Std Dev = 0.40
(4.3)
17.8
Range: 1,215,600 Mg < W < 45,360,000 Mg
9 The "relevant" characteristics are those that influence gas production, such as waste composition, moisture
content, ability of the gas to migrate out of the waste, temperature, and similar factors.
4-13
-------
The figure below the coefficient estimate is the t-statistics for the 95 percent confidence
interval for the null hypothesis that the true coefficient value is zero. For this model, the
coefficient for waste mass (W) is significantly different from zero at the 95 percent confidence
level.
Models 4.2 and 4.3 behave reasonably for estimating national emissions. The
coefficients are significant with the expected sign, and for Model 4.2 the residuals (the
differences between the observations and the estimates produced with the model) do not
exhibit undesirable characteristics. However, additional data analysis indicated that about 30
percent of the landfills in the data set are located in arid regions (less than 25 inches of
rainfall per year), while only about 13 percent of the waste in the U.S. is generated (and
presumably disposed of) in arid regions.10 Because moisture can affect gas production
and emissions from a landfill, the over-representation of landfills in arid regions in the data set
could bias the estimates of national emissions downward.
To assess the importance of the landfills in arid areas, Models 4.2 was revised to
include a "Dummy" variable that indicates whether the landfill is in an arid region.11 Using
the Dummy variable, the influence of the arid landfills was found to be statistically significant
for Model 4.2, for which there are 26 landfills in arid regions. Consequently, Model 4.2 was
re-estimated including the Dummy variable, D:
CH4 (m3/min) = 8.22 + 5.54 W (106 Mg) - 2.09 D-W (4.4)
2.6 10.2 4.0
R2 = 0.67 n = 85 Range: 1,215,600 Mg < w < 45,360,000 Mg
The coefficient for the Dummy variable is significant, and has the expected sign (negative),
meaning that in this data set, landfills in arid regions have lower methane production.12
Similarly, average methane generation per ton of waste in Model 4.3 can be re-
estimated for non-arid and arid landfills. For the landfills in non-arid regions the results are:
CH4 (m3/min) = 7.66 W (106 Mg) Std Dev = 0.46 (4.5)
16.7 N = 59 Range: 1,215,600 Mg < W < 19,051,200 Mg
10 The rainfall for each landfill was estimated using the National Climatic Data Center (NCDC) Historical
Climatological Series Divisional Data. The climate normal period of 1951 to 1980 was used. The climate data
applicable to each landfill was ascertained based on the location of the landfill and detailed maps of the climate
divisions. The estimate of waste disposal in arid regions is based on an estimate of U.S. population in arid regions.
11 The "Dummy" variable is set to 1 when the landfill is in an arid region, and 0 otherwise.
12 As with all statistical analyses, the finding of a statistically-significant coefficient does not necessarily reveal
the causation mechanism of the statistically-significant effect. In this case, several alternative mechanisms may
explain the observed influence of rainfall on methane production, including: wetter climates lead to wetter wastes
which produce more methane; and drier climates result in drier wastes and soils which allow more oxygen to
infiltrate the landfill and oxidize the waste. Additionally, some of the "arid" landfills are in the South Coast Air Quality
Management District (SCAQMD) in California. The SCAQMD requires that steps be taken to limit gas production
and emissions, which may also be one mechanism leading to the observed statistical effect. Whatever the
causation mechanism, the inclusion of the Dummy variable in this analysis helps to compensate for a potential bias
in the data.
4-14
-------
For landfills in arid regions the results are:13
CH4 (m3/min) = 5.87 W (106 Mg) Std Dev = 0.82 (4.6)
7.2 N = 26 Range: 1,317,600 Mg < W < 45,360,000 Mg
Models 4.4, 4.5, and 4.6 are adopted for this analysis and Exhibit 4-4 shows how the
regression model fits the landfill data. The upper line represents methane production for
landfills in non-arid regions and the lower line represents methane production for landfills in
arid regions. The "boxes" correspond to the landfills in non-arid regions and the "pluses" to
landfills in arid regions. As shown in the exhibit, although there is variability in the data, the
models are a good representation of the relationship between methane produced and waste
in place.
Model Application -• Identify the Number and Size of Landfills in the U.S.
To apply the models to estimate national emissions, information is necessary on the
waste in place in landfills (W) in the U.S., the number of landfills, and their size distribution.
This information can be used to define "model" landfills representative of landfills in the U.S.
The number of landfills in arid regions is also required. For purposes of this analysis, the
estimate that 13 percent of waste is disposed in arid regions is assumed to apply uniformly to
all representative landfill types. Methane production from these model landfills can then be
estimated using models 4.4 to 4.6. National emissions will equal the sum of the emissions for
each model landfill times the estimated number of each model landfill in the U.S.
Estimate total waste in place in landfills in the U.S.
Current methods of landfill waste disposal became common during the early 1960s.
Waste was commonly handled before then using open burning. The landfills in the data set
used in this study generally first accepted waste in the early 1960s (GAA 1991). The total
waste in place in these landfills is the sum of the waste disposed annually in these landfills
over the last thirty years. Therefore, given the data used to develop models 4.4 to 4.6, and
the fact that the use of sanitary landfills became widespread in the 1960s, the proper estimate
of waste in place for 1990 is the sum of the quantity of waste landfilled during the 1960s,
1970s, and 1980s.
During the mid-1980s, the EPA Office of Solid Waste (EPAOSW) conducted a detailed
survey of the U.S. municipal landfill industry. Data were collected from over 1,000 landfills
across the United States (USEPA 1988a). The EPAOSW survey estimates that in 1986 about
190 million metric tons of waste were placed in landfills. This total is similar to the results of a
survey of state waste management officials performed for Biocycle, which estimates that
about 194 million metric tons of waste were landfilled in 1991 (Biocycle, 1992a). The 190
million metric tons of waste estimated by the EPAOSW survey consists of 156 million metric
tons of commercial/residential waste and 34 million metric tons of other waste. These values
are adopted as the average waste disposal quantity of the 1980s. Values for the 1960s and
1970s can be calculated from the following:
13 Using a one-tailed t-test, the coefficient in Model 4.5 is larger than the coefficient in Model 4.6 at a 95
percent confidence level. This result is consistent with the Dummy variable being statistically-significantly different
from zero in Model 4.4.
4-15
-------
X
Ul
"5
•D
O
V>
O
'55
10
I
UJ
0)
(0
TJ
V
a>
in
ro
(M
O
o
(N
O
O
IN
O
oo
CD
T5
1
T3
CD
CO
u
(0
TO
T3
03
O
C
CD
en
E
CD
O
CO
0>
N
'CO
"0
CO
CD
c
T5
(0
(O
OS
c:
O
en
CD
c
O
'to
D)
CD
jad
JQ
c
o
CD
.C
cu
o
-------
• Commercial and Residential Waste. EPA's Office of Solid Waste reported that
commercial and residential waste grew by 24 percent between 1965 and 1975
and by 26 percent between 1975 and 1985 (USEPA 1990).
Other Landfilled Waste. Other wastes include sewage sludges, industrial
waste, and construction and demolition debris. These wastes are assumed to
grow at a rate equal to population growth since 1960. Population growth
between 1965 and 1975 averaged about 1.1 percent and population growth
between 1975 and 1985 averaged about 1.0 percent.
Exhibit 4-5 presents the estimates of total waste placed in landfills over the last 30 years,
which is approximately 4,700 million metric tons by 1990. Based on the uncertainty in the
EPAOSW survey, it is assumed that the mean estimate of 4,700 million metric tons is
distributed normally with a standard deviation of 350 million metric tons. This implies a 95
percent confidence interval of about ± 15 percent about the mean estimate of 4,700 million
metric tons.
Exhibit 4-5
Estimated Waste Landfilled Between 1 960 and 1 990
Period
1960s
1970s
1980s
Total waste placed in landfills
between 1960-1990 c
A
B
c
Average Annual Quantity of Waste
(million Mg)
Commercial/Residential A
100
124
156
3,800
Disposed in
Other6
27
30
34
900
Landfills
Total
127
154
190
4,700
Based on USEPA (1990).
Other Waste includes sewage sludge, industrial process waste, and construction and demolition debris.
Rounded to the nearest 100 million Mg.
Estimate the number and size distribution of landfills
The EPAOSW survey estimated that 6,034 landfill facilities were active during 1986. In
addition, approximately 3,000 very small landfills are believed to have closed in the 1980s
prior to the 1986 EPAOSW Survey. Additionally, there may be tens of thousands of landfills
that closed prior to the 1980s
14
The survey also describes the distribution of landfill sizes in the U.S. but does not
include the 3,000 very small landfills that closed in the early 1980s. Exhibit 4-6 illustrates the
14 There are thought to be about 30,000 older closed landfills (Thorneloe 1992b). Field measurements of
urban methane concentrations indicate that older closed landfills are often significant sources of emissions in the
urban environment (Kolb et al. 1992). Omission of the older closed landfills from this analysis biases the estimates
of methane emissions downward.
4-17
-------
distribution of waste in place in 1986. The 3,000 very small landfills are assigned to Class 1
and are assumed to have very small quantities of waste in place and to represent only a
negligible fraction of the total waste in place. Classes 2 through 7 represent the landfills that
were open when the Survey was conducted. This distribution is assumed to represent the
distribution of waste in place for 1990.
Exhibit 4-6
Landfill Size Distribution by Waste in Place
Class
Range
Low (MG)
High (MG)
Number of
LandfillsA
Percent
of Total
Waste in
PlaceA
Waste in
Place
(106MG)B
•w0-
Avg/Fac
r
1 (Closed)
2
3
4
5
6
7
0
0
500,000
1,000,000
5,000,000
10,000,000
20,000,000
500,000
500,000
1,000,000
5,000,000
10,000,000
20,000,000
200,000,000
3,000
4,744
425
712
106
27
19
10.5%
6.6%
33.6%
15.1%
8.8%
25.4%
negligible
494
312
1,581
709
411
1,194
< 0.008
0.104
0.733
2.221
6.673
15.072
61.223
Total
9,034
100.0%
4,700
"W0" is the average waste in place per facility in that class.
A The distribution of the number of landfills and the percent of waste in place were derived from USEPA (1987). The number
of operating landfills (6,034) is from USEPA (1988a).
B The waste in place was computed by multiplying the percent of waste in place for each size class by the total waste in
place (4,700 million Mg) reported in Exhibit 4-5.
C The average waste in place per facility was computed for each size class as the waste in place divided by the number of
landfills.
The number and average sizes for each landfill class define the "model" landfills used
to represent the U.S. landfill population. These model landfills are used to estimate national
emission estimates.
Model Application • Estimate Emissions lor Each Model Landfill
Using Models 4.4 to 4.6 and the model landfill sizes, emissions can be estimated for
each model landfill as follows.
For each landfill size class, estimate emissions per landfill using the average
waste in place ("W0" in Exhibit 4-6) and Models 4.4 to 4.6 (depending on size).
• Estimate the uncertainty for each landfill class using the uncertainty in the
model estimates for Model 4.4 to 4.6 and the uncertainty in the total amount of
waste in place. These uncertainties can be estimated numerically using the
estimated distributions of the model estimates and total waste in place.
4-18
-------
Estimate National Emissions
Emissions of methane to the atmosphere will equal total methane production from
municipal landfills adjusted for the methane produced by industrial landfills, the methane
recovered, and the methane oxidized in the landfills before being released to the atmosphere.
These adjustments can be described as:
Net Methane Emissions = municipal landfill methane generation
plus industrial landfill methane generation
minus municipal methane recovery
minus industrial methane recovery
minus methane oxidation by soil
Municipal landfill methane generation is described by Models 4.4 to 4.6 and the above
described methodology. The above methodology and data, however, do not include the
small amount of waste managed in industrial waste landfills. Industrial waste landfills receive
nonhazardous waste from factories, processing plants, and other manufacturing activities.
Unlike municipal solid waste landfills, the organic content of the waste in industrial landfills is
only about 11 percent (USEPA 1988b). Because no information is available on methane
generation at industrial landfills, the approach used in this analysis is to assume that the
organic waste in industrial landfills produces methane at the same rate as the organic waste
in municipal landfills. Industrial landfill methane generation is calculated as follows:
• Estimate the amount of organic waste placed in industrial landfills per year.
EPAOSW estimates that in 1986 8.6 million megagrams of organic waste was
deposited in industrial landfills (see Exhibit 4-7).
Exhibit 4-7
Waste Quantities Disposed in Industrial Landfills in 1985
Industry Type
Waste Quantity Landfilled
(million Mg)
Percent of Total
ORGANIC WASTE
Pulp and Paper
Food Products
Leather Products
Total Organic Waste
5.33
3.26
0.01
8.60
6.8 %
4.2 %
<0.1 %
11.0%
INORGANIC WASTE
Electric Power Generation
Chemicals
Metals
Other
Total Inorganic Waste
48.48
8.51
4.59
8.02
69.60
62.0 %
10.9 %
5.9 %
10.3%
89.0 %
TOTAL WASTE
78.20
100.0 %
Source: USEPA (1988b).
4-19
-------
• Estimate the amount of municipal waste that would have the same quantity of
organic waste. Assuming that 65 percent of municipal waste is organic, then
13.2 million megagrams of municipal waste would also contain 8.6 million
megagrams of organic waste:
8.6 .10* Mg Organic Waste m 132 . 1£)6 Mg Munjcjpa, Wasf0
65%
• Calculate what fraction of the annual waste deposited in municipal landfills this
quantity represents. Assuming that about 190 million megagrams of total
municipal waste are placed in municipal landfills each year (see Exhibit 4-5),
then, the organic content of industrial landfill waste represents about 7 percent
the methane producing capacity of the municipal waste:
13.2 • 10" Mg __ 7
190 - 106 Mg
Therefore, this analysis assumes that industrial landfill methane generation equals about 7
percent of municipal landfill methane generation.
As previously described, some of the methane produced by landfills is collected and
either flared or utilized to provide energy. The data set used in this analysis indicates that in
1991 approximately 1.2 Tg/yr of methane were recovered (for energy use) nationally by
municipal solid waste landfills (GAA 1991; USEPA 1992c). In addition, a small number of
landfills are believed to recover and flare methane without energy recovery and were not
included in the GAA database. The amount of methane flared without energy recovery is not
known and in addition no information is available on methane recovery from industrial
landfills. This analysis assumes than an amount equal to 25 percent of the methane
recovered for energy use is recovered and flared without energy recovery. Not considering
the small portion of the methane that is not combusted in the flares or other energy utilization
devices, total methane recovery from landfills is assumed to equal 1.2 Tg/yr plus 0.3 Tg/yr, or
1.5 Tg/yr.
Finally, methane may be oxidized in the top layer of soil over the landfill. The landfills
that recover methane collect the gas through pumps and wells so that the gas does not pass
through the soil cover of the landfill. Methane traveling through a landfill without a recovery
system, however, may be oxidized in this layer of soil (Whalen, Reeburgh and Sandbeck
1990). The amount of oxidation that occurs is uncertain and depends on the characteristics
of the soil and the environment. For purposes of this report, it is assumed that 10 percent of
the produced methane is oxidized in the soil (Mancinelli and McKay, 1985).
With these adjustments, a national estimate and uncertainty range were calculated by:
• Multiplying the mean, high, and low emission estimates for each landfill class
by the number of landfills in that class.
• Summing the mean, high, and low estimates over all landfill classes to obtain
national mean, high, and low totals.
4-20
-------
Calculating national emissions by adjusting for the methane that is recovered
(1.5 Tg/yr), the amount produced by industrial landfills (7 percent), and the
amount oxidized in the soil above the landfills (10 percent).
4.4 CURRENT EMISSIONS
Landfills in the U.S. are estimated to emit between 8.1 Tg/yr and 11.8 Tg/yr with a
central estimate of 9.9 Tg/yr. This amount, which does not include approximately 1.5 Tg that
was recovered from landfills and either flared or used as an energy source, represents about
20 to 40 percent of the estimated world landfill emissions of 20 to 70 Tg/yr (IPCC 1992).
Based on this analysis, landfills are by far the largest anthropogenic source of methane
emissions in the U.S., accounting for about 36 percent of the estimated total annual U.S.
anthropogenic emissions of 25 to 30 Tg/yr. Exhibit 4-8 presents these estimates by landfill
size class. Exhibit 4-8 shows that:
Nineteen of the largest
landfills in the U.S. generate
about 20 percent of estimated
methane generated by
landfills. These nineteen
landfills receive about 25
percent of all landfill waste in
the country.
U.S. landfills are estimated to emit
between 8.1 and 11.8 Tg/yr of methane.
Landfills are by far the largest
anthropogenic source of methane
emissions in the U.S., accounting for
about 37 percent of the total annual
U.S. anthropogenic emissions.
The largest 900 landfills in the
U.S. receive about 85 percent
of all landfill waste and produce about 75 percent of all landfill methane.
About 1,300 of the 6,000 landfills in the U.S. generate almost all the methane
and receive almost all the waste.
These estimates imply that by recovering methane from the very largest landfills, the U.S.
could substantially reduce overall methane emissions to the atmosphere and in addition
provide a source of clean and reliable energy. Exhibit 4-9 compares the estimates provided
by this report with other previously published estimates. As shown in the exhibit, the
emissions estimates in the study fall in the middle of the range of previous estimates.
4.5 FUTURE EMISSIONS
Future emissions will be influenced by a variety of factors and trends including:
increases in solid waste generation; increases in the portion of the waste stream diverted
away from landfills through the use of recycling, composting, and other methods;
"regionalization" of landfilling operations, with a trend toward fewer, larger landfills; and
regulatory requirements that limit landfill gas emissions. A range of scenarios of these factors
was examined to estimate future emissions.
4-21
-------
Q
o>
O)
^™
s
CO
0)
Q
* |
"^* 0)
±£ UJ
.Q
"S 5
111 «
E
UJ
"c5
o
'•s
CO
o
E "^
UJ £L
M •?
CO I—
c c- •
0
b* ^^
CD 5
0> g "£
c ^
CO C i-
.C (Q Q)
"S ^*
SCO
£
15
E *?
IS ^J
• ^T
III Q)
^^ ^fc
__ .Q
I*,
*•
o
0
CO
Q.
'co
8 £
C O>
0) T
"O (0
1 fc
O CD
*t!
0 c
^ O
in -J
O)
C — i—
2 H h?
||f
^ ^
__
i2 .-— »*
« ^ ^
Si '
^ n GO
Z ^
cvJ
s3-
If
i5
co
Q> to
.y co
CO Q
co in in in oo co
in CD o o
oi i— CD d oo
O) O CM O T-
CD in oo i- o>
i- i- d i- o>
'is! II
« 8 | •§ T
CD ® c ,2< £
C "^ ^~ U
CO tO *A Cr}
[3 3 13 ^ C
CD ^ CL C .S
C 5 5 "
co ^- Si
£ E
CD UJ
^r ^^
8
"*
o
co"
"co
f~\
°
CD
08
in •*
c c
.2 .2
"co to
tr cr
UJ UJ
CD CD
CD CD
CO CO
Sw
^k
0 ^
^ o"
Q
• +
vJ <^^_ ^
CD . CD
•? D)
-= C
« re
c *-
O r—
c £
a o
jz. E
B w
0) Cl)
o c
« «
co £
*^ $
ra 2
•1 fe
E c
Q)
co o
li! _ £
CM Q) .2
Ix. *i
-r, ? re
^ t *^
c = o
Co (0
§ w £
co — »_
N- re o
H C3 —
o 0) ^
E T-: £ 52
Jgj o~
| |||
•ffi o §
C S 0 "
CD .2 C 0)
•D O Q) .C
•»c "D ~
5ilf
O) o
|| » |
< £ jE o
< m o
-------
Exhibit 4-9
Comparison to Other Estimates of U.S. Methane Emissions from Landfills
This Study
Bingemer and Crutzen (1 987)
Augenstein (1990)
USEPA(1992c)
Emission Estimate (Tg)
8-12
11-25
3-8
3 - 9.4A
Reasons for Differences
Uses an annual disposal rate
model and a higher methane
generation potential (500 g CH4
per kg of waste carbon) based on
theoretical stoichiometric
relationships.
Uses lower estimate of total waste
disposed since 1 960 by a factor of
approximately 1 .5.
Did not account for recovery
efficiencies so estimates should
be increased by 20 to 25%.
Based on small sample of 21
landfills.
A Adjusted to use total waste in place as developed in this report because reported estimate based on
illustrative waste volume. AEERL is refining this initial estimate based on use of an expanded data base
(110 sites) and additional field data.
4.5.1 Background
Landfill practices in the U.S. are undergoing changes that will affect landfill gas
emissions. In the last decade, increased public awareness of the hazards of waste disposal
has led to restrictions on siting and operation of landfills. In addition, both the Clean Air Act
(CAA) Amendments and the reauthorization of the Resource Conservation and Recovery Act
(RCRA) are expected to have a substantial impact on the municipal solid waste (MSW) landfill
industry. Compliance with the CAA and RCRA will likely increase the cost of waste disposal.
Some major implications of these expected changes are:
• increased Recycling. Recycling will become an increasingly cost effective
method of reducing the total waste stream. In 1991, approximately 14 percent
of U.S. waste was recycled. State percentages ranged from 3 percent in
Mississippi to 34 percent in Washington. Exhibit 4-10 provides information on
each state's recycling, composting and deposit/return laws. Most states now
have recycling goals of 20 - 50 percent, often as part of mandatory programs
(Biocycle 1992b).
• Increased Use of Alternative Disposal Methods. Alternative methods of waste
disposal such as composting and incineration will continue to grow, reducing
the portion of waste that is managed in landfills. EPAOSW (USEPA 1992a)
projects that the percentage of municipal solid waste composted will increase
4-23
-------
STATE
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
D.C.
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
State
Percent
Recycled
8
6
5
5
17
16
15
8
7
21
5
4
8
12
8
10
5
10
10
17
10
29
25
31
8
10
6
10
10
5
30
5
14
17
10
3
10
21
10
Exhibit
4-10
Source Reduction Efforts (1991)
Disposal
Bans
VB, T
VB, T, Y
B
VB, T, Y, M
VB
VB
VB, T
VB, T, Y
VB, T, Y, M
T
VB, T
Y
T
VB, T, Y
VB, Y
B, VB, T, Y, M
VB
VB, T, Y, M
VB
VB.Y
VB
VB, T, Y, M
VB, M
VB, T, Y
VB, T, M
VB, Y
Mandatory
Deposit/Return Law
VB, C
VB
c, B
c
VB
VB
c
T
VB
C
c
C, VB
VB
VB
VB
T
VB
C, VB
VB
VB
C, VB
VB
Number of
Composting
Programs
9
0
0
5
21
3
79
2
1
20
1
1
6
106
10
30
30
6
2
13
5
180
200
331
11
37
2
15
1
65
270
2
170
43
10
20
5
20
169
(continued)
4-24
-------
STATE
State
Percent
Recycled
Exhibit 4-10
Source Reduction Efforts (1991)
Disposal Mandatory
Bans Deposit/Return
(continued from previous
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Pet of Total
Waste
15
5
10
2
10
10
20
10
34
10
17
3
14A
VB
VB, T, Y, M
VB, T
VB, T, Y
VB, T, M
VB
B, VB, T, M
VB
VB, M
VB, T, Y
VB, T, Y, M
page)
VB
VB
VB
VB
c
VB
VB
VB
KEY B=mercury oxide or other batteries C=beverage containers
VB=vehicle batteries Y=yard waste
A This average is slightly lower than the 1990 figure of 14.9% calculated
Office of Solid Waste (USEPA 1992a).
Sources: Biocycle 1992a and 1992b.
Number of
Composting
Law Programs
11
0
3
0
8
2
9
36
12
4
213
2
M= motor oil T=tires
on a material flows basis by EPA's
from 2 percent in 1990 to 5.3 percent in 1995 and 7 percent in 2000, largely
due to bans on landfilling of yard wastes in many states. The percentage of
municipal solid waste combusted is projected to rise from 16 percent in 1990
to almost 21 percent in 2000.
• The construction of fewer landfills. The difficulty of siting new landfills along
with the increased cost of regulatory compliance may lead to fewer and
generally larger landfills than exist today.
• Increased closures of landfills. Due to the increasing regulatory costs, many
less efficient landfill operators may be forced to close, further decreasing the
number of landfills.
These changes will also affect methane production by landfills. Future landfill methane
production will be driven by the following factors:
The quantity of waste generated and placed in landfills. While recycling,
composting, and combustion will remove more and more municipal solid waste
from the waste stream before it reaches landfills, the quantity of municipal solid
waste generated is expected to increase.
4-25
-------
• The composition of the waste. The organic content of landfill waste is
expected to increase slightly, despite the increased recovery rates discussed
above. This is due to the increase in paper, wood, and other organic
components of municipal solid wastes.
• The manner in which landfills are managed; and
• The amount of methane that is recovered.
4.5.2 Methodology
Projecting the amount of waste placed in landfills is at best uncertain. New products
and innovations (e.g., the "paperless" office), changes in personal preferences (e.g., the use
of composite materials) can significantly affect the quantity and composition of the waste
stream. In spite of these difficulties, the EPA Office of Solid Waste (EPAOSW) developed
projections for municipal solid waste generation to 2000. The projections are based on
individual projections of the primary components of the waste stream, including: paper
products; metals; and yard wastes. Overall, EPAOSW projects that generation of municipal
solid waste will increase by 13.5 percent (by weight) between 1990 and 2000 (USEPA 1992a).
Because of the changing regulatory climate and expected growth of recycling,
projecting future waste quantities and composition is very difficult. Three different scenarios
were developed for this analysis and are described below.
Scenario 1: USEPA (1992a) projects that the quantity of municipal solid waste
disposed in landfills will decrease by 16.3 percent between 1990 and 2000.
EPAOSW's estimate actually includes a 13.5 percent increase in generation of
municipal solid waste over this period, with per capita waste increasing from 4.3
pounds per day in 1995 to 4.5 pounds per day in 2000 (USEPA 1992a). However, this
trend of higher generation is more than offset by increased rates of recycling (15
percent to 23 percent), composting (2 percent to 7 percent), and combustion (16
percent to 21 percent). Application of these trends to the 190 million Mg of waste
landfilled in 1990 results in an estimate of 159 million Mg for 2000. Assuming that the
rate at which waste is disposed in landfills decreases by half as much between 2000
and 2010 (because gains in recycling, composting and combustion slow), or by 8.2
percent, the figure for 2010 is 146 million Mg.
Scenario 2: In this scenario, the increased generation of solid waste is assumed to
just offset the gains in recycling, composting, and combustion. Therefore, the quantity
of waste landfilled in 1990,190 million Mg, remains constant each year through 2000
and 2010.
Scenario 3: This scenario adopts the EPAOSW rate of increase in generation of
municipal solid waste, but uses a lower projected increase for recycling, composting,
and combustion: a five percent gain to 2000 with no further increase to 2010.
Application of these assumptions results in an estimate of 205 million Mg of waste
landfilled in 2000 and 233 million Mg landfilled in 2010.
4-26
-------
The estimates for each scenario are summarized in Exhibit 4-11. The analysis of future
emissions of methane developed in the following sections of this report uses the mid-range
estimate provided by scenario 2 of 190 million Mg per year between 1990 and 2010. The
analysis can be modified to reflect more recent information as additional data become
available on disposal trends.
In addition to assumptions on the quantity of waste landfilled between 1990 and 2010,
the following sections describe waste composition, landfill management, and gas recovery
assumptions used to project future methane emissions from landfills.
Exhibit 4-11
Scenario 1
Scenario 2
Scenario 3
Future Landfill Waste Disposal
1990
190
190
190
Scenarios (106Mg/Yr)
2000
159
190
205
2010
146
190
233
Landfill Waste Composition
The organic composition of the waste is expected to remain near current levels.
Because recycling and recovery affect both organic and non-organic materials, recycling will
not significantly affect the organic fraction of landfill waste. In 1990 the organic content of
municipal solid waste after recovery was about 69 percent. By 2000 this percentage is
expected to decrease slightly to about 66 percent. Exhibit 4-12 shows the projected
composition of the municipal waste stream for 2000 and the projected composition after
recycling and recovery.
Landfill Waste Management
Although both the amount of waste and the organic content of the waste that enters
landfills is expected to remain stable over the next two decades, the trend toward fewer and
larger landfills will affect landfill emissions. Over at least the last ten years, the number of
active landfills has decreased and the average size of active landfills has increased. Between
the mid-1970s and mid-1980s, the amount of municipal solid waste increased by 25 percent
(USEPA 1990) while the number of active landfills has decreased from over 18,000 (USEPA
1991c) to about 6,000 (USEPA 1988a). The trend toward fewer and larger landfills is often
referred to as the regionalization of landfills.
The trend toward the regionalization of landfills should continue because larger
landfills are expected to have lower disposal costs per ton of waste than smaller landfills.
This results in large part from the cost of regulatory compliance and the difficulty of siting new
landfills (USEPA 1991c). Under proposed revisions to RCRA, the cost of regulatory
compliance could increase significantly and further increase the trend towards regionalization.
4-27
-------
Projected
Type of Waste
ORGANIC
Paper and Paper Board
Yard Trimmings
Food wastes
Wood
Textiles
Total Organic
INORGANIC
Glass
Metals
Plastics
Other
Total Inorganic
Total
Exhibit 4-12
Landfill Composition for 2000*
Composition of MSW Generated
(10° Mg and Percent of Total)
1990
67
(38%)
32
(18%)
12
(7%)
11
(6%)
5
(3%)
126
(72%)
12
(7%)
15
(8%)
15
(8%)
10
(5%)
51
(28%)
178
(100%)
2000
77
(38%)
30
(15%)
12
(6%)
15
(7%)
6
(3%)
140
(69%)
12
(6%)
15
(8%)
23
(11%)
12
(6%)
64
(31%)
201
(100%)
Composition of MSW After Recovery
(106 Mg and Percent of Total)
1990
47
(32%)
28
(19%)
12
(8%)
11
(7%)
5
(3%)
103
(69%)
10
(7%)
11
(8%)
15
(10%)
9
(6%)
45
(31%)
147
(100%)
2000
46
(33%)
15
(11%)
12
(9%)
13
(9%)
5
(4%)
92
(66%)
8
(6%)
10
(7%)
20
(14%)
11
(7%)
49
(34%)
141
(100%)
A Figures do not include other wastes that are placed in landfills including sewage sludge, industrial
process waste, and construction and demolition debris. The total amount and composition of these
other wastes are not expected to significantly affect the overall organic composition of landfills in the
future.
Source: USEPA (1992a)
Totals may not add due to
rounding.
The EPA Office of Solid Waste (EPAOSW) has projected that only landfills accepting
more than about 60,000 Mg per year will be economical to operate (USEPA 1991c). This
acceptance rate is assumed to affect all class 2 landfills and about 50 percent of class 3
landfills (USEPA 1987). Over the next twenty years EPAOSW projects that over 80 percent of
the waste now being received by landfills now receiving less than about 60,000 Mg per year
will be shifted to larger landfills. For the purposes of this report, the following assumptions
are made:
• By 2000 the average rate of waste disposal will decrease by 65 percent for size
class 2 and by 20 percent for size class 3. The average rate of waste disposal
4-28
-------
for size classes 4 to 7 will increase in proportion to their 1990 average disposal
rates. The number of landfills producing methane in each size class will remain
equal to the 1990 value.
By 2010 the average rate of waste disposal will decrease by 80 percent for size
class 2 and by 23 percent for size class 3. The average rate of waste disposal
for size classes 4 to 7 will increase in proportion to their 1990 waste disposal
rates. The number of landfills producing methane in each size class will remain
equal to the 1990 value (although some will likely be closed).
Waste in place producing methane for each landfill class is assumed to equal the
waste placed in each class over the preceding thirty years. Although it is believed that landfill
waste will produce methane over a greater period of time, the thirty year horizon is chosen
because the average age of the landfills used to estimate the emissions models (4.4 to 4.6) is
between 25 and 30 years. Using a longer time horizon with the estimates based on 25 to 30
year old landfills would be extrapolating outside the range of the data. To the extent that
waste produces methane after thirty years, the estimates presented in the report will
understate future methane emissions from landfills.
Based on these assumptions and the current annual waste disposal rates by landfill
size class based on the EPAOSW Survey, Exhibit 4-13 lists the projected current and future
waste disposal rates for each landfill class and the estimated waste in place producing
methane and the average amount of waste producing methane per landfill class for 1990,
2000, and 2010. These estimates will be used to estimate emissions for 2000 and 2010. As
with current estimates (See Exhibit 4-8), these estimates of total waste in place are assumed
to be normally distributed with a confidence interval of about ±15 percent about the mean.
Landfill Gas Recovery
Gas is recovered from landfills to protect human health and the environment and to
utilize the methane as a source of energy. Methane gas that is recovered and burned will not
be released to the atmosphere.15 Currently, over 100 landfills recover about 1.2 teragrams
or 64 billion cubic feet of methane gas per year and produce over 300 megawatts of power
(Thorneloe 1992a). In addition, another 0.3 Tg/yr is assumed to be recovered and flared
without energy recovery. Under the proposed landfill rule, the amount of methane recovered
from landfills could increase significantly, depending on the limits set in the rule. Because the
landfill rule has not been promulgated in final form, two landfill gas recovery scenarios are
defined to estimate future landfill recovery rates:
Extend Current Recovery Practices. Under this scenario, 1.5 Tg of methane is
assumed to be recovered annually, which is the estimate of current methane recovery
from landfills nationally.
Landfill Rule. The landfill rule is designed to reduce emissions of non-methane
organic carbons (NMOCs) to the atmosphere. Landfills that would emit more than a
specified quantity (the cutoff) of NMOCs will be required to collect and combust the
15 If the methane is combusted in a flare or generator engine, a small amount of methane will escape
combustion and be released to the atmosphere.
4-29
-------
<
^
0
OI
•D
C
(D
0
O
CM
c
Q)
CO
£
Q)
n s
^» ^™
' D)
^ C
•£•» "^
£ "D
X O
111 £
0)
«
CO
?
;c
^Q
c
•a
a>
V
'o
i
0
T—
°
o
o
CM
O
o>
o>
0-55
X ri«o
* > o>
< 0
CD *
0 I
« o -—
°- of
c ,c •*=
Q.
ID —
CD CD C
to •— 8
> ° CD
CC ^^
^.^
CO C"
= oy_ 5
? ^0
< O
m ^"
8 i
c _c ^
£ 3 °
"to T3 ^»
CO O
Set
CD CD C
«& ~ Q3
rr* >^«
•o^f
§ "gjo
— > ^3
< ^
8 i*
E°-3
°- 0)5
c: .9
QJ T ^3
OT T3 Z^
CO O
§ ct
CD CD C
l|s
5 CD ^
II
E c
^ CO
z — '
CD ^
.ti to
CO Q
in in in in in co
CM co T- oo en CM
p cq p p cq co
o o co' oS o iri
CM CO
o i- co co en -i- o
CM en •* CD in CM o
i- CM i- en in co r-
CM~ T-" in"
N^ -sO xP vP xP "\P
o^ o^ c>^ o^ o^ u^
t— t— N- en co ^~
c\i in r^ cb en co
CO t- CM
T— i — en co ^ T—
'd- tO ^t CO h- CM
O CO 1^- CM CO CO
CD CD CM CO CO h^
^
in T- co co o en o
en co m i^- T— N- o
T- CM en co m ^j- co
••-" T-" in"
u^ tf^ o^- 0s* 0s" 0s*
r^ co en co co en
co in to CD en r^
CO T- CM
^ co o o en -i—
o co CM en co co
i- h~ CM CO CM CO
CD CD c\i co in cvi
T- CO
•si- CM T- en i- -* o
en T— co o •>— en o
^~ co in r^ ^i~ •»— h~
T-~ i-^ •*"
xP xP xp xP xP xO
in co co ••- co ^j-
CD CD co in co in
i- CO T- CM
M- in CM co N en ^
TJ- CM i- O CM -i- CO
N- ^J" h^ T— O
xf CO"
..
CM co ^ m co h- *
0
o>
£ CD
3 CO
T) 0.
*- ^
E Q.
8 *
el
CO £
SS "co
§ I
~ .«;
CD "O
CD
CO OT
o °-
C*P o"
c O5
^ .E
CD CD
CD "5-
e-
CD CD
> to
O to
11
^ *•
a 2
jnj
CD §
to CT
CQ CD
* CO
og
3 CM C3
C __ OI
CO CD '"
CD Q S
£ Q.J
"to _g o>
cr co .-
co to E
_ JK n
O 5 ^
i «"1J-
•^ x Q-
CD CD g,
CO O (Q
£ ^ $
E CD ® C
^3 ^s ^3
^ Ml CT
"o ® 2 o
"D w "P o
O Q) •£-
Q. °-0 CD
_ CO O 13
CD CD O T3
0 •}=; CM J~
CO ™ ^
l"i §* ™
-r o 2 o
*- jjr- -C^
fo S -^ to"
5 Ol r~
Q) OJ C
* i?" «2
I- t5 E «
< §2 °
-------
landfill gas. Because methane will also be collected and burned, methane emissions
from these landfills will be greatly reduced.
The Regulatory Impact Analysis performed for the landfill rule identified the reduction
in methane emissions expected at alternative cutoff levels for both new and existing
landfills. Based on this analysis, the following assumptions are made to estimate
methane emissions for 2000 and 2010 at three proposed cutoff levels:
25 Mg/yr NMOCs. Methane recovery systems will be installed at the MSW
landfills accounting for 80 percent of methane production. These recovery
systems will collect 80 percent of the methane produced in these landfills.
100 Mg/yr NMOCs. Methane recovery systems will be installed at the MSW
landfills accounting for 70 percent of methane production. These recovery
systems will collect 80 percent of the methane produced in these landfills.
• 150 Mg/yr NMOCs. Methane recovery systems will be installed at the MSW
landfills accounting for 60 percent of methane production. These recovery
systems will collect 80 percent of the methane produced in these landfills.
4.5.3 Future Emissions with Current Recovery Practices
With these assumptions and applying Models 4.4 to 4.6, emissions for 2000 will range
between 8.8 and 12.7 Tg/yr. Emissions for 2010 will range between 9.5 and 13.4 Tg/yr.
Although the amount of waste and the organic composition remains the same as 1990 levels
between 1990 and 2010, 1990 emissions increase by 8 percent by 2000 and by 15 percent by
2010 because the total amount of waste in landfills that produces methane increases between
1990 and 2010. Exhibit 4-14 summarizes these results. As above, estimates are shown by
class of landfill size, and summed to estimate national emissions.
4.5.4 Future Emissions with the Landfill Rule
Alternative NMOC emissions cutoffs are being considered for the proposed landfill
rule. This analysis reports methane emission estimates for three NMOC emission cutoffs:
25 Mg/yr, 100 Mg/yr, and 150 Mg/yr. Adoption of the 25 Mg/yr NMOC emissions cutoff for
the landfill rule will dramatically reduce methane emissions to the atmosphere from landfills.
Methane emissions from landfills would be about 50 percent below current emissions by 2000
and 2010, or approximately 3.5 to 6.5 Tg per year. To meet the 25 Mg/yr NMOC cutoff by
2000, over 1,200 new and existing landfills will require gas recovery systems. Because the
collected methane can be used to generate electricity or sold as natural gas, revenues can
be generated that may offset the cost of these recovery systems.
Adoption of the less stringent standards will also reduce emissions. Under the
100 Mg/yr NMOC emissions cutoff rule methane emissions would be about 4.5 to 7.5 Tg/yr in
2000 and 2010. Over 800 landfills would require recovery systems. Under the 150 Mg/yr
NMOC emissions cutoff rule methane emissions would be about 5.0 to 9.0 Tg/yr in 2000 and
2010. Over 800 landfills would require recovery systems.
4-31
-------
CO
0)
o
o
m
: Recovery Pn
c
0)
O
2S
5
O
T- O
1 0
o
o
CM
w
o
*^
CO
CD
03
'•a
w
UJ
c
o
CO
CO
E
UJ
«.
CO
o
75
z
4
^"
^
O>
0)
1
7
Ul
C8
o
z
i/lethane per
Landfills
i3 per minute)
— 3
CO
» ^
(Q ®
E *
« 5
-1
0 UJ
W A>
« «
to
0
CB
Q.
C
Q)
To
CO
^
CO £
= O)
ffl _ I
T3 CO
^ 1
cSi „
1 -
CO ™ ^
0 0)
5l-i-
co < 5?
^ 0 CO
•C. U <+—*'
Z
-i r^ in in in in
CM CM CM CM
CM CM CM CM
cci cd cd cd
T- 1- T- in oo CD •
rt CD in CO CD CM
o co r- CM to co
o o CM co co in
T- t1^
in T- co oo o 05
C35 00 in h» T— h-
i— CM o> oo in ^
•^ ^
S? 6? S? S? S? 3?
?- co o CD co en
co in CD co Q !•*•
CO T- CM
^- in CM co Is- o>
Tf CM T- O CM i-
^
CM co ^t m CD r^
O) O CM t- O
in in o rj- N
^r -^ i- i- N
''"• T"
SO ''t CO O
in i^ en oo
o i- ci o od
So co en co
m co -i- r»
CM -r- 0 i- 0
T— ^
° >- To co
C Jl\ *^ O C3
mil
c tr ^ o *•
o 3 3 § 1
m C Q_ C _
c ^ ^ co
CO "
£ E
flt
w
Z
8
CO
in"
S
CD"
To
Q
1-
•
• j
9
9
CM
a
1
E
1
iU
0 CO i- CM i- O
TJ- en CM •* co r^-
O O N- CM i- CO
in o rr Tfr o oo
CM co CD r^- en co
O 0 TT i- O CM
i— m h* i^ o co
CO h» CJ) O i- O
o o m CM -i- co
i- o CD m co oo
o ^* co en oo ^
y— CO h~ O)
CM
CM CM CD Tf CO CO
o" iri Tt cd i- cd
CM in CM CD
1- 'Sf
en en en o> CT> en
r^ h- o o o o
T^ W c\i c\i cvi c\i
CD CD ^ ^ ^ ^~
co CD in in in m
r^ t^ in in in in
CM CM CM CM
CM CM CM CM
cd cd cci cd
m ^r co CM oo CM
CM 00 t- CO CO •*
0 CO 0 0 -^ -r-
d o co en o co
CM CO
o T- CD co en T-
CM en • *= '.5 o
j^: o *" « CM
03 (} ^ ^3
Q) W *— X O
C DC i O *-
S || | 8
co ^ ^ SB.
£ E
CD m
5 I!
Z
g
!"•*.
in"
S
CD"
15
0
1-
C v>
E E
i2*i:
"* V*
OC\J
*>l
5 "8
CO W
• s
§ «0
c ^"*
*- E
° »
CD CM
-O CO
E co
W H^
•t
§" c
i €'
5 O
Q)
co '£
v— c
CO
+ CO
(0
•c
i
|
^
§
»
„
jv °?
CO •«•
ii 5
>- UJ
"Si r~
(? •—
^T °
•U ^— '
0 0
K- C
4-«
S **~
^ 0)
< 0)
CO
< CD
-------
4.5.5 Opportunities for Emission Reductions
Profitable opportunities exist for reducing methane emissions from some landfills.
Proven technologies are available to recover and utilize the methane. As described in the
GAA database, over 100 landfills recover methane to generate electricity or to sell the gas to
an industrial user. The implementation of the landfill rule will greatly increase the amount of
gas that is recovered and utilized.
There are an abundance of landfills in which gas recovery is technically feasible but is
not undertaken because of economic, regulatory, or other barriers, as described below:
Economic barriers. The most common purchasers for landfill gas energy are
electric utilities. Because almost all landfills are located near power lines,
utilities provide a reliable and accessible customer. However, because utilities
are generally only required by law to pay the avoided cost of generating
electricity, the price received by the landfill is significantly below the market
price for power and is often insufficient to justify project development. This
barrier is common to virtually all "alternative" energy sources, including
cogeneration, biomass, solar, and wind.
• Regulatory barriers. Regulations exist that hinder the development of landfill
gas energy development. For example, energy recovery equipment must often
meet air emission standards that do not consider that the equipment is being
used to mitigate other harmful emissions. Despite these obstacles, the forecast
for landfill gas recovery is optimistic. The trend toward very large scale,
comprehensive waste management facilities may create new markets for landfill
gas developers. A detailed assessment of opportunities for landfill gas
recovery will be included in a separate report.
Other barriers. Landfill recovery technologies have been improving over the
past ten years, and some landfill operators may be unfamiliar with the latest
methods. Additionally, operators may be unaware of the opportunities for
selling electricity to utilities that are now available to independent power
producers. These and other informational barriers may constrain the wider
development of landfill gas energy projects. Another potential impediment
stems from the fact that many landfills are owned by municipalities; given their
numerous responsibilities, some municipalities may be unable to place a high
priority on developing a landfill methane power project.
4.6 LIMITATIONS OF THE ANALYSIS
The methane emission estimates presented in this chapter are uncertain for a variety
of reasons, including the following:
• Landfill and Waste Characteristics. The actual number and size of landfills and
other waste management facilities are not known with certainty. In particular,
many small and unregulated facilities may exist that are not included in these
estimates. Evidence indicates that small and long abandoned waste dumps
produce significant quantities of methane.
4-33
-------
• Time Horizon. This report assumes that landfill wastes produce methane over
a thirty year period. If the true period is significantly longer, then current and,
particularly, future emissions could be greatly understated.
Landfill Emission Data. The model used to estimate methane production is
based primarily on data from the largest landfills found in the U.S. The basis
for estimating emissions from small landfills needs to be improved.
• Landfill Emissions Measurements. There are very few measurements of
methane emissions from landfills. This analysis is based on data describing
methane recovered from landfills. The methane recovery information is an
imperfect surrogate for emissions measurements. If the landfills used in this
analysis are not representative of landfills as a whole, then the models
developed for this analysis may not accurately represent overall landfill
methane generation.
• Methane Oxidation. Little information is available on the amount of methane
oxidized by the soil cover over landfills. The 10 percent oxidation rate
assumed in this report is based on limited measurements. If the oxidation rate
is significantly higher, net methane emissions to the atmosphere could be
lower.
• Future Landfill Emission. Because projections of the quantity and composition
of waste landfilled over the coming decades are very uncertain, emission
estimates are very uncertain. If significantly more or less waste is landfilled or
the organic fraction of waste changes, the estimates of future emissions
presented in this chapter could be understated or overstated by a large
amount.
The EPA is currently undertaking studies to improve the basis for making emission
estimates. In particular, EPA ORD is working to develop field measurements and additional
data that will improve the ability to develop emission models. As additional data become
available, the estimates can be improved.
4.7 REFERENCES
Augenstein, D.C. 1990. Greenhouse Effect Contributions of United States Landfill Methane.
GRCDA 13th annual landfill gas symposium, Lincolnshire, IL
Bingemer, H.G. and P.J. Crutzen. 1987. The Production of Methane from Solid Wastes. Journal
of Geophysical Research. 92:2181-2187.
Biocycle. 1992a. "1992 Nationwide Survey: The State of Garbage in America." April 1992.
Biocycle. 1992b. "Solid Waste Legislation: The State of Garbage in America." May 1992.
Emcon Associates. 1982. Methane Generation and Recovery from Landfills. Ann Arbor
Science. Ann Arbor, Ml.
4-34
-------
GAA (Governmental Advisory Associates, Inc.). 1991. 1991-92 Methane Recovery From
Landfill Yearbook. Government Advisory Associates, Inc. New York. 1991.
Gunnerson, Charles G. and David C. Stuckey. 1986. Integrated Resource Recovery Anaerobic
Digestion: Principles and Practices for Biogas Systems. World Bank Technical Paper
Number 49.
IPCC. 1992. "Climate Change 1992 The Supplementary Report to the IPCC Scientific
Assessment. Report Prepared for IPCC Working Group I. (J.T. Houghton, B.A.
Callander and S.K. Varney, eds.). World Meteorological Organization and the United
Nation Environment Program.
Kolb, C.E. et al. 1992. Methane Emissions from Natural Gas Distribution Systems. Prepared
for the Global Change Division Office of Air and Radiation, U.S. Environmental
Protection Agency, Washington, D.C.
Loehr, Raymond C. Pollution Control for Agriculture. Second Edition. Academic Press, Inc.
Orlando, Florida. 1984.
Mancinelli, R.L and C.P. McKay. 1985. "Methane-Oxidizing Bacteria in Sanitary Landfills." In
Proceedings of the first symposium on biotechnical advances in processing municipal
wastes for fuels and chemicals, A.A. Antoxopoulos, Argonne National Laboratory.
Peer, R.I., S.A. Thorneloe, and D.L Epperson. 1992. "A Comparison of Methods for
Estimating Global Methane Emissions from Landfills," U.S. Environmental Protection
Agency Air and Energy Engineering Research Laboratory, Research Triangle Park, NC.
Accepted for Publication in Chemosphere (in press).
Rettenberger, G. and Tabasaran, O. 1980. Untersuchungen uber Entstehung Ausbereitung und
Erfassung von Zersetqungsgasen in Abfallblagerungen (Investigation on the origin,
movement and collection of landfill gas). Report No. 103 02 207. Umwelbuntesamt.
Berlin.
Stegman, R. 1986. Grunlagen der Deponieentgasung. Basisinformation uber Entstehung
Deponiegas (Principles of landfill gas extraction. Basic information on the genesis of
landfill Gas). Munich.
Thorneloe, S.A. 1992a. "Landfill Gas Recovery/Utilization - Options and Economics,"
presented at The Sixteenth Annual Conference by the Institute of Gas Technology on
Energy from Biomass and Wastes, Orlando, Florida. March 5, 1992.
Thorneloe, S.A. 1992b. Personal Communication. EPA Air and Energy Engineering Research
Laboratory, Research Triangle Park, NC. November 1992.
USEPA (United States Environmental Protection Agency). 1987. National Survey of Solid
Waste Municipal Landfills. Database supplied by DPRA, Inc. September 1987.
USEPA (United States Environmental Protection Agency). 1988a. National Survey of Solid
Waste (Municipal) Landfill Facilities. Washington, D.C. September 1988.
4-35
-------
USEPA (United States Environmental Protection Agency). 1988b. "Solid Waste Disposal in the
United States," Report to Congress, Vol. II, EPA Office of Solid Waste and Emergency
Response, Washington, D.C. EPA/530-SW-88-011, October 1988.
USEPA (United States Environmental Protection Agency). 1990. Characterization of Municipal
Solid Waste in the United States, 1960-2010. Washington, D.C.
USEPA (United States Environmental Protection Agency). 1991 a. Air Emissions from
Municipal Solid Waste Landfills - Background Information for Proposed Standards and
Guidelines. EPA Office of Air Quality Planning and Standards, Research Triangle Park,
NC. EPA 450/3-90-011 a, March 1991.
USEPA (United States Environmental Protection Agency). 1991b. Analysis of Factors
Affecting Methane Gas Recovery from Six Landfills. Air and Energy Engineering
Research Laboratory, Research Triangle Park, NC. EPA 600/2-91-055, September
1991.
USEPA (United States Environmental Protection Agency). 1991c. "Standards of Performance
for New Stationary Sources and Guidelines for Control of Existing Sources: Municipal
Solid Waste Landfills," Federal Register, May 30, 1991, pp. 24467-24528.
USEPA (United States Environmental Protection Agency). 1992a. "Characterization of
Municipal Solid Waste in the United States: 1992 Update." EPA Office of Solid Waste
and Emergency Response, Washington, D.C. EPA/530-R-019, July 1992.
USEPA (United States Environmental Protection Agency). 1992b. Personal communication
with landfill operator in Bristol, Wisconsin.
USEPA (United States Environmental Protection Agency). 1992c. Development of an
Empirical Model of Methane Emissions from Landfills. Air and Energy Engineering
Research Laboratory, Research Triangle Park, NC. EPA 600/R-92-037, March 1992.
Whalen, S.C., W.S. Reeburgh, and K.A. Sandbeck. 1990. "Rapid Methane Oxidation in a
Landfill Cover Soil." Applied and Environmental Microbiology. November 1990.
4-36
-------
4 «
0)
Q.
Q.
<
3
T3 O
5 o>
< co
cc
T3 03 f2
ill
!§.»
(» c
si
S 03
TJ £ ^
o —
e ±
O. O
T3 5 *^**
CD O £
oi21!
5SS
o o •2-
^ .o
o m
m
i- 00 O O
^— gQ {^ C^
CO ^ O O
^—
00
IO O
g 8
t>-" o"
8 8
O T O •*»-
O CO O OO
•«- O CO
CO O ^
^* ^^ lO
00 tO ^"
^3 ^"
O CO
8
CD
C\J
O
O
o
CVJ
IO
't
O h— O
O O O
incoo
O O
O O
oo
O
O
in
o
C0_
CM
8 8
o o
8 8
O CD
° CM
CO ^3
C7) I"**
f1^ ^3
O) O
^"
CO
o o
o o
o o
8 9
O CD
O" h-"
8 S
OvJ
T-CO
CM
O
c\j
IO
CO'—
-
10
^5
CO
i—
CD
CO
CO
9 2 8
CM" to" CM"
fv.
O
00
CO
O IO
IO CO
CM
•r-" O" T-~
O^
O
CO 1O O O> CO
*•" C\J O3 CD ^"
CO CD CO
O f1^ •» —
n CM"
CM" CM" CM"
oo o
o
-------
I
Q. *
a. *
<
cn
I
Q
S .o
tr
CD "jif
£ co g
18 8 2
"D S —
8.e-i
3 u- E
o?^
i: I m
Q. O ^
(D O IT
IP
O
® o "E
1 S'g
o o -^
CD
s
53
o
(0 (0
CD CD
O
000
-z.-z.~z.
o o o o o o
0000
z z z z
O O
Z Z
o o
o o o o
z z z z
O O O CO O O
O O 00 CO O O
o o 05 co o o
Op CM O 00 O)
CO CD CO O
TJ-" e\T in oo"
CM
2 <0 CM CM O O>
C. C CM CO CO CD
in co o
co o o
co t^ o
intooeo
ininT-i-
inT-eoeoo
O TJ- CM CO O CM •--cocMCMin
co *a- o oo r-- oo r*-
in •*!• CO Ol O> f- CO
* « oo r~ ~ •
m in CM m
CM in ^r co
CO O) O ~
CM
s«
<
CD
< Q Q Q
if
tcS
g o
O
0)
I
CD
E ^ ^
5
CO
c *"
coo
Ifli
= 5 -2 co
C .2
O ^
o co
" CO
? O
CO 4-*
» §
o
I?
en
|1
CD «J
Q. CO
m
~ o
Q. DC
CD
CD
0)
O)
CD
Q.
-------
V
c c
S. -o
a S
< 3
a
ra
a
5 o
'< c?
tr
•D CD CO
III
§5 «.
2 to J2
S 9 I
Ssf
3 U- E
sSl
ol o x-
•o g: ^
^ .9
O m
m
^ to
o (0
O CD
a>
15
53
oooooooooooooooooooooooooooo
ZZZZZZZZZZZZZZZZZZZZ2ZZZZZZZ
co"
eo
CM"
*-*-rCeoeoi-weoe«j t- If) O *~ O O
^-
CO
^* CO
O O
O
IO
CO CM O CM O"
T— O" O C-" O O" O O)~
oocoococo'*-o-<—
cDr>-eocj>^-cooo>
CO1-COO
CO
coT^oocMcoc6r>-coo>Ti-cooo)CNOcocor^coo
T-" co" •r-" o" co" co" T-~ co" co" co" CD" in" co" CM" CM" o" ^ co" w" CM" co"
CM co o
'tCD^t-
Cn^-CO
CO
CM
CNCNi
CDinO
CM"
-r-" co"
C\T
-r-" CM"
cooor^
CO'd-^-
CO"
COO5CNJCM
00 00 CO O) 00 O)
CO CO N- t*- LO Is- CO
CO CO
^^ W3
CO
^^
CO CO
^3 tO IO CJ
5s* S^ 0s* S~^ 5s* S*~
^3 ^3 O IO ^3 ^3
^ i— if) O CD O>
*-" csf in
*-
mo
lOO^
8 § R S
CM" 10" T-"
Ift CO CO CO
T-~ CM" i-"
T-
rs*-o
^-CXJ
o> o> to co oo 10 in
co" -r-" T-"
T!
CD
CO
2
o
CO
t)
O)
CD
a.
-------
V
_
«
<
XJ O
3 o>
< CD
cc
"g £ {2
"3 M 2
> § CO
CD
0:6-2-
E I?
CD U
°£
IJg-
lit
ooooooooooooo
ZZZZZZZZZ2ZZZ
0>
o o
z z
§ 8 8
i- O O
oo" •<-" 10
CO CO
co"
T-" CM CO 00" IO~ CJ
oo" CJ
CN CO
CD
-r- e\T eo" co"
co
T-1' C\I -r-
CO O) O)
•r-~ ^-" CM CO
-
O 00 U^ LO lO U")
co r**- co co r*"" co
CO
co
v
CO
co
§
° o
in in
C3
^p
c\T
c\T
10" •«-"
CO CO
oor^
oh-.
CM" CM"
CM
8
C
1 5
CD
§
T5
CD
co
U
CO
2
01
CO
o>
CD
s
_CO
1
JO
_>,
c§
<
S3
C\J
s
a.
LU
co
"0
c
co
<
a.
co
CD O)
o
1
D
-------
CHAPTER 5
METHANE EMISSIONS FROM DOMESTICATED LIVESTOCK
U.S. Methane Emissions
from All Sources
Li vested Manure
Other Sources
Landf i I Is ..
Natural Gas
Systems
Domestic Livestoc
Annual Domesticated Livestock
Methane Emissions
Global Emissions U S Emissions
Emissions Summary
Source
Dairy Cattle
Dairy Cows
Replacements
Beef Cattle
Beef Cows
Replacements
Bulls
Feedlot Fed Cattle
Other Animals
Total
1990 Emissions
(Tg)
0.9 - 1 .4
0.3 - 0.4
1.8-2.7
0.4 - 0.6
0.2 - 0.3
0.9-1.3
0.2 - 0.3
4.6 - 6.9
Partially
Controllable
'
/
*
5.1 EMISSIONS SUMMARY
Methane emissions from domesticated livestock in the U.S. in 1990 are estimated as
4.6 to 6.9 Tg, with a central estimate of 5.8 Tg. Dairy and beef cattle account for 95 percent
of these emissions. A firm foundation of scientific measurements and understanding of
methane formation and emission supports these estimates. The principal uncertainties in the
estimates of current emissions are associated with the large diversity of animal management
5-1
-------
practices found in the U.S., all of which cannot be characterized and evaluated precisely. The
uncertainty estimate is subjective, based on the sensitivity of the results to various
assumptions.
In the future, emissions may increase if beef and milk production increase. However,
due to declining per capita consumption, U.S. beef production has remained flat over the past
10 years, and may remain flat or decline in the future. While milk production is anticipated to
increase, improved efficiency in production may work to limit future increases in emissions
from the dairy sector.
Future relaxation of international trade restrictions may affect future U.S. milk and meat
production, so that methane emissions from this source would also be affected. For example,
if the U.S. becomes a significant exporter of milk products, domestic milk production could
increase by over 30 percent by the year 2010. Emissions from the dairy sector could
increase substantially under this scenario, although by an amount less than the increase in
production because production efficiency is also increasing over time.
Considering these various factors,
emissions are estimated to increase from
4.6 to 6.9 Tg/yr in 1990 to a range of 5.0 to
7.9 Tg/yr by 2000 and 4.8 to 8.2 Tg/yr by
2010. The high estimates in these ranges
Decades of research and measure-
ments provide a firm scientific basis for
estimating methane emissions from
livestock in the U.S.
assume substantial increases in milk
production for export and a small increase
in beef production associated with beef
maintaining its domestic market share of red meat consumption. The low estimates assume
that beef production declines by 2010 and that dairy production increases at the rate of
domestic consumption only. Around each of these emissions estimates is an uncertainty of
about ±20 percent, based on the uncertainty of the factors that form the basis of the 1990
emissions estimate.
5.2 BACKGROUND
Methane is produced as part of the normal digestive processes of animals. Referred
to as "enteric fermentation," emissions from these processes account for a significant portion
of the global methane budget, about 65 to 100 Tg annually (IPCC, 1992). Of domesticated
animals, ruminant animals (cattle, buffalo, sheep, goats and camels) are the major source of
methane emissions. Furthermore, cattle are the primary source of methane emissions from
enteric fermentation in the U.S., contributing 95 percent of the total emissions from this
source. For this reason, emissions from cattle are emphasized in this chapter with only brief
discussion of emissions from other domestic animals. Additionally, emissions from wild
ruminant animals such as deer, and wild non-ruminant herbivores such as rabbits, are not
considered because: (1) emissions from these animals are considered a natural source; and
(2) emissions from these animals are very small in the U.S.
1 Although non-ruminant animals produce only a small quantity of methane from enteric fermentation as
compared with ruminant animals, emissions from non-ruminant animal manure, especially swine manure, may be
significant. Methane emissions from livestock manure are discussed in Chapter 6 of this report.
5-2
-------
Ruminant animals are characterized by a large "fore-stomach" or rumen. Within the
rumen, microbial fermentation breaks down the feed consumed into soluble products that can
be utilized by the animal. Approximately 200 species and strains of microorganisms are
present in the anaerobic rumen environment, although only a small portion, about 10 - 20
species, are believed to play an important role in ruminant digestion (Baldwin and Allison
1983). The microbial fermentation that occurs in the rumen enables ruminant animals to
digest coarse plant material which monogastric animals, including humans, can not digest.
Exhibit 5-1 presents a schematic of the ruminant and monogastric digestive systems.
Methane is produced in the rumen by bacteria as a byproduct of the fermentation
process. This methane is exhaled or eructated by the animal. Non-ruminant herbivores such
as horses, mules, rabbits, pigs, and guinea pigs do not support this pre-gastric fermentation.
Some microbial fermentation does occur in the large intestines or ceca of these animals, but
the methane produced in this manner is quite small compared to the amount produced by
ruminant animals.
Cattle are the primary source of
livestock methane emissions in the U.S.
Cattle eructate or exhale methane as
part of their normal digestive processes.
A significant scientific literature exists
that describes the quantity of methane
produced by individual ruminant animals,
particularly cattle. This literature results
from decades of research evaluating feeding
practices for cattle and other ruminants.2
Over the past 30 years, hundreds of
methane measurements have been
performed on a wide variety of cattle diets typically used in the U.S. The USDA Ruminant
Nutrition Laboratory has been the primary focus of dairy animal evaluations, and the Colorado
State University Department of Animal Science has been the primary focus of beef animal
evaluations. The main purpose of these measurements was the development of scientifically-
based feeding standards for dairy and beef animals in the U.S. The standards are presented
in a series of National Research Council publications (e.g., NRC, 1989; NRC, 1984) and are
used routinely for determining feeding strategies.
At the same time that the experimental data on whole animals were being developed,
significant progress was made in understanding the microbiological processes involved in
ruminant digestion and animal growth. Detailed assessments of rumen fermentation at the
microbiological level had been performed and continue to be refined. A mechanistic
understanding of ruminant digestion and physiology has evolved that allows quantitative
models to be developed.
The understanding of ruminant digestion and physiology continues to evolve, so that
feeding efficiency and animal productivity continue to improve. Commercial products are
available that improve production by manipulating rumen digestion processes directly.
Increasingly sophisticated manipulations of digestion and other physiologic processes are
2 Calorimetry is the laboratory technique currently used to perform in-depth evaluations of alternative feeding
practices. This technique involves placing an animal in a climate-controlled confinement chamber for a period of
several days, and measuring the levels of inputs to (feed, oxygen, carbon dioxide) and outputs from the chamber
(feces, urine, milk, carbon dioxide, oxygen and methane). Because methane is produced as part of the normal
digestive process of cattle, methane is measured as part of this feed evaluation technique.
5-3
-------
Exhibit 5-1
Schematic of Ruminant and Monogastric Digestive Systems
Colon
Esophagus
PIG
Monogastric
Stomach
6.8 qt
Esophagus Reticulum Ab°m«um
COW
Ruminant
Rectum
Rumen
160 qt
Large intestine —(
40 qt
Source: After Ensminger (1983).
expected in the future. Nevertheless, there is currently a firm basis of scientific measurements
and mechanistic understanding for estimating methane emissions from ruminant animals in
the U.S.
5.3 METHODOLOGY
To estimate methane emissions from domesticated livestock in the U.S., emissions
factors were developed for representative animal types and then multiplied by applicable
animal populations. These values estimates were then summed to estimate total annual
methane emissions.
Given their population and size, cattle account for the majority of methane emissions
from livestock in the U.S. Because of the in-depth understanding of rumen digestion
processes for cattle typically found in the U.S., and because U.S. cattle production systems
are well characterized, it is possible to estimate methane emissions from cattle in the U.S.
with reasonable precision. Detailed analyses using mechanistic models of rumen digestion
and animal production were performed to estimate emissions factors for the diverse types of
5-4
-------
cattle and cattle feeding systems found in the U.S. A variety of diets and management
practices were defined and evaluated for each of five regions of the U.S. The emissions
factors developed based on these analyses are specific to U.S. cattle and are consequently
preferred for this assessment over previous estimates based on averages for groups of
countries (see, e.g., Crutzen et al.. 1986).
To estimate methane emissions from other animals, emissions factors were taken from
published literature. This approach is reasonable given that these animals do not contribute
significantly to total U.S. emissions and because the variability in emissions factors among
countries for the other animals is much smaller than the variability in emissions factors for
cattle.
5.3.1 The Models Used to Evaluate Cattle Production and Emissions
Mechanistic models of rumen digestion and animal production were used to develop
the cattle emissions factors in this study. The digestion model, originally described in
Baldwin et al. (1987a), explicitly models the fermentation of feed within the rumen, including
the creation and passage of the products of digestion, such as volatile fatty acids (VFAs),
amino acids, carbohydrates, fat, and protein. This model estimates the amount of methane
formed and emitted as the result of the microbial fermentation that takes place in the rumen.
The digestion model is linked to an animal production model that predicts growth,
pregnancy, milk production and other production variables as a function of the digestion
products predicted by the digestion model. By linking the digestion model with the animal
growth model, the combined modeling framework can be used to evaluate energetic
relationships in the animal. Specifically, the model framework was designed to investigate the
factors that cause variations in the relationship between feed input characteristics and animal
outputs including weight gain, lactation, heat production, pregnancy, and methane
production.
A validated mechanistic model of rumen
digestion and methane production was
used to estimate emissions for the
cattle feeding systems used in the U.S.
A total of 32 diets were simulated for 8
animal types in 5 regions.
To develop emissions factors for
U.S. cattle this mechanistic modeling system
is preferred to statistical relationships
among feed characteristics and methane
production (e.g., Moe and Tyrrell (1979) and
Blaxter and Clapperton (1965)). The
statistical models are only valid for the feed
types and feeding levels that were used to
develop the models. However, cattle in the
U.S. consume a wide variety of feeds at
various feed levels that differ regionally, temporally, and by production system. The available
statistical models are limited in their ability to estimate methane emissions from this wide
range of conditions. Alternatively, the strength of the mechanistic model is that it has been
validated for the wide range of feeding conditions encountered in the U.S., and consequently
can be used to estimate methane emissions with greater accuracy.
5-5
-------
5.3.2 Model Evaluation
Evaluation of the structure and performance of the rumen digestion model and the
ability of the model to predict the digestible energy (DE) and metabolizable energy (ME)
values of feeds were described in Baldwin et al. (1987a). The applicability of the digestion
model to lactating diary cows and the validation of the animal performance aspect of the
model framework were described in Baldwin et al. (1987b). Subsequently, the model was
revised and evaluated for growing cattle (Dimarco and Baldwin, 1989).
For purposes of estimating methane emissions in this report, the basic model
(described in Baldwin et al.. 1987b) was revised and additional analyses and evaluations of
the model were undertaken. The model was generalized to enable evaluations of a broader
range of animal weights and stages of maturity, including weaned calves, feeder cattle, and
lactating cows. The model was also revised to evaluate a wider range of diets, from poor
quality forages to high quality grain-based feedlot rations. The following major changes were
performed to provide the model with these capabilities:
• Initial conditions such as weight of body (largely muscle and skeleton), viscera,
carcass fat, and gut fill were scaled to body weight.
Parameter values for metabolic equations were scaled to metabolic body
weight3 as described by Kleiber (1961).
• Separate specifications for hemicellulose and cellulose were added because
their fermentation products differ, which can affect methane emissions.
Separate treatment of hemicellulose and cellulose is consistent with the
equation of Moe and Tyrrell (1979), which uses separate specifications for
predicting methane emissions from dairy cows.
Hydrolytic rate constants for hemicellulose and cellulose differ among grasses,
legumes, and silages. Because the original model only considered legumes,
new rate constants for grasses and silages were introduced.
To validate the model, a challenge data set of 35 different rations was developed from
a literature survey, including diets that range in quality from 1.6 to 3.1 Meal/kg of ME.4 A
total of 16 cattle diets and 19 sheep diets were evaluated and used to specify the model
parameters. Exhibits 5-2, 5-3, and 5-4 show graphs of observed and predicted values for the
16 cattle diets. Exhibit 5-2 shows that the observed and predicted ME values correspond
very well (correlation coefficient = 0.98). Exhibit 5-3 presents the observed and predicted
methane values per kilogram of feed (dry matter basis), also showing good agreement
between the model and the observed values, although the variation is relatively high
(correlation coefficient = 0.62).
3 Metabolic body weight is defined as empty body weight (EBW) raised to the 0.75 power: EBW° 75.
4 Meal/kg = megacalories per kilogram of feed on a dry matter basis. Low ME values (e.g., 1.6 Meal/kg) are
typical of low quality forages, such as a mature hay. High ME values (e.g., 3.1 Meal/kg) are typical of high grain
rations with less than 15 percent forage.
5-6
-------
Exhibit 5-2
Observed and Predicted Estimates of Metabolizable Energy (ME)
(Meal/kg)
T)
(D
3 1
21-
19-
1 7
I D. '
17 19 21 23 25
Observed
27
29
Diet Definitions:
A: 20% Alfalfa, 80% Corn-soybean meal concentrate
B: 40% Alfalfa, 60% Corn-soybean meal concentrate
C: 40% Alfalfa, 60% Corn-soybean meal concentrate
D: 40% Alfalfa, 60% Corn-soybean meal concentrate
E: 50% Alfalfa, 50% Corn-soybean meal concentrate
F: 60% Alfalfa, 40% Corn-soybean meal concentrate
G: 60% Alfalfa, 40% Corn-soybean meal concentrate
H: 69% Corn silage 80% Corn-soybean meal cone.
I: 41% Corn silage, 59% Corn-soybean meal cone.
J: 28% Corn silage, 72% Barley concentrate
K: Alfalfa hay
L: Clover hay
M: Timothy hay (early)
N: Timothy hay (mid-bloom)
O: Timothy hay (late)
P: Timothy hay (late)
Sources of the observed data: Colovos et al. (1949); Coppock et al. (1964); Flatt et al.
(1967); Moe and Tyrrell (1972b); Moe and Tyrrell (1977); Moe et al. (1973a); Moe et al.
(1973b); Tyrrell and Moe (1972); Wainman et al. (1979).
5-7
-------
Exhibit 5-3
Observed and Predicted Estimates of Methane Per Kilogram of Feed
(Meal/kg)
-o
OJ
u
,-1
T3
1)
P,
i_i
O
O
o
o
o
0
o
o
o
•n^J —
33 -
31 -
29 -
27 -
25 -
-
'"' '•*, _
-i t
_/
19 -
17 -
O
M
^'''
K ^-'
° N /' P L
B „.''' E F
.^
.^^
^ H
C ' °
I
J
17 O19 O21 O23 O25 O27 O29 O31 O33 O
Observed
See Exhibit 5-2 for diet definitions.
Exhibit 5-4
Observed and Predicted Estimates of Methane Per ME
(Percent)
17 O 95 -
15 056 -
" 1 1 O % -
i>
IX
7 095 -
5 O56 -
5
^ '
^^ L
^^ O
G ^^
^^ H
A I
J
Q% 7 096 9 O95 11 O55 13 O95 15 O56 17
Observed
O*
See Exhibit 5-2 for diet definitions.
5-8
-------
Because the modeled ME values are used to simulate the feed intakes of the animals,
the predicted methane production per ME is a better indicator of model performance than is
methane per kilogram of feed. Exhibit 5-4 shows the excellent agreement between the
observed and predicted estimates of methane production per ME (correlation coefficient =
0.85). The variations shown in the graph are similar in magnitude to those observed in
experimental measurements.
An additional evaluation was performed by comparing the model's methane emissions
estimates to those predicted by the Moe and Tyrrell (1979) equation, which was derived
statistically to fit an extensive data set of 404 measurements of methane emissions from dairy
cattle fed diets commonly found in the U.S. A comparison of 11 dairy cow diets shows that
this model is within 10 percent of the Moe and Tyrrell equation estimates (Exhibit 5-5).
Finally, validation of the model is provided by its estimates of metabolic functions. To
compute daily feed intake within the model, the NRC equations for daily ME requirements for
maintenance, growth, pregnancy, and lactation were applied using the model-estimated
ME/kg values for the feed. Given the daily feed intake, the model simulates the formation and
use efficiencies for absorbed nutrients for maintenance, protein and fat accretion, milk
synthesis, gestation, and other metabolic functions. If the model estimates were incorrect, the
errors in the simulation of daily animal performance would accumulate over time. When
summed over a 365 day simulation, even small errors in the simulation of input:output
relationships would be evident in the model predictions of growth, milk synthesis, or other
variables. In the preparation of this study, model estimates of weight gain and milk synthesis
were always within 2 percent of the values specified using the NRC equations, indicating that
feed intake levels and metabolic processes were modeled correctly.
Based on these analyses the modeling system used in this study has been
demonstrated to be valid for the range of diets and management systems typically
encountered for mature dairy cows, mature beef cows, mature bulls, and growing steers and
heifers in the U.S. The main outputs of the model, including the energy value of feeds,
methane production, and animal production can be estimated within the experimental and
animal variations observed in the literature.
5.3.3 Application of the Model
To apply the model, categories of animal types were defined to represent the different
sizes, ages, feeding systems, and management systems that are typically found in the U.S.
Representative diets were defined for each category of animal, reflecting the diversity of diets
that are found in various regions of the U.S. Each animal type within each region was then
evaluated using the model.
Representative Animal Types
The following animal types were defined for the cattle population:
• Dairy Animal Types:
Replacement heifers 0-12 months of age
Replacement heifers 12-24 months of age
Mature dairy cows (over 24 months of age)
5-9
-------
Exhibit 5-5
Comparisons with Moe and Tyrrell Equation Estimates
Lactation Diet
Model Estimates
ME Intake
(Meal)
Feed ME
(Meal/kg)
Methane
(kg)
Moe and
Tyrrell
Methane
(kg)
1. High quality alfalfa hay (19.6% CP)C
14,385
2.35
139
138
2. 75% alfalfa hay, 25% corn meal-SBMc
concentrate
16,156
2.48
133
104
3. 60% alfalfa hay, 40% corn-SBM concentrate
15,136
2.53
124
118
4. 50% alfalfa hay, 50% corn-SBM concentrate
15,719
2.61
116
105
5. 60% alfalfa hay, 40% corn-cottonseed meal
concentrate (15% CP)
16,582
2.56
130
128
6. 40% alfalfa hay, 5.5% SBM, 54.5% corn
15,426
2.73
93
101
7. 69% corn silage, 16% corn meal, 14% SBM,
1% mineral supplement
15,559
2.65
113
120
8. 40% alfalfa hay, 5.5% SBM, 54.5% ground
oats
15,851
2.61
109
120
9. 50% alfalfa hay, 50% barley-SBM concentrate
15,424
2.57
117
106
10. 40% timothy hay, 45% corn meal, 15% SBM,
cane molasses, mineral cone.
15,480
2.69
123
128
11. Early timothy hay (7.9% CP) supplemented
to 14.5% CP with cottonseed meal
12,088
2.41
136
138
a Diet during the dry period was grass hay (e.g., timothy) supplemented to 14.5% CP with a protein
supplement, such as cottonseed meal.
b CP = crude protein
c SBM = soybean meal
Simulation conditions included, initial empty body weight (EBW) of 550 kg; 305 day lactation period; 60 day
dry period; feed intake during lactation computed daily according to NRC (1989) equation as modified by
Mertens (1985) to restrict excess intake of NDF; feed intake during the dry period computed according to NRC
(1989) equation for maintenance of a pregnant cow plus an allocation for pregnancy according to Moe and
Tyrrell (1972a) plus an allocation for fat gain when cows end lactation at under 550 kg EBW. Except for diets 1
and 11, feed intake for fat gain during the dry period was less than 1 kg per day. Feed intakes for diets 1 and
11 were severely limited during lactation due to their high fiber content. Consequently, weight losses in early
lactation were severe and simulated cows were low in body fat at the end of the simulation.
5-10
-------
• Beef Animal Types:
Replacement heifers 0-12 months of age
Replacement heifers 12-24 months of age
Mature beef cows (over 24 months of age)
Heifers and steers grown for slaughter
Mature bulls.
Due to their small number, mature dairy bulls were not evaluated. Dairy calves that are not
kept as replacements are generally fed for slaughter. Consequently, these animals are
included in the total for heifers and steers grown for slaughter.
Exhibits 5-6 and 5-7 summarize the characteristics used to simulate the representative
animals. Dairy and beef replacement heifers age 0 to 12 months are simulated starting at 165
days of age through 365 days of age. In actuality, replacement heifers will consume a mixed
diet of forages and milk starting at about 60 to 90 days of age. This mixed diet will be
consumed through weaning at about 205 days. The choice of 165 days of age as the
starting point for the simulation is a compromise to reflect that although some forage is
consumed prior to weaning, most of the feed energy consumed prior to weaning is derived
from milk which is not fermented in the rumen. The overall estimate of methane emissions
from cattle in the U.S. is not sensitive to this choice of 165 days as the starting point for the
simulation because replacement heifers age 0 to 12 months are a very small contributor to
total emissions.
Dairy and beef replacement heifers age 12-24 months are simulated to be pregnant
and ready to give birth at about 24 months of age. Dairy replacements are simulated to grow
to a larger body weight by 24 months of age, reflecting the larger frame size of Holstein cows.
The steers and heifers that are fed in feedlots for slaughter are simulated from 165
days of age through final feeding in the feedlot. Two managements systems were simulated.
The Yearling System includes a 260 day stocker period, from 165 days of age to 425 days,
followed by a 140 day feedlot period. The Weanling System puts the calves on feed starting
at 165 days of age. A 257 day feeding period is simulated to bring the steers and heifers in
the Weanling System up to slaughter weight of 480 kg. The end points for feedlot feeding
were simulated to achieve low choice grades, at about 26 to 30 percent carcass fat, with the
Weanling System producing carcasses on the high side of this range.
Mature dairy cows are simulated to have a 305 day lactation period followed by a 60
day dry period. Pregnancy is simulated so that the cow gives birth at the end of the 60 day
dry period. The milk production per lactation is simulated to match the observed average
milk production in each of the five regions. Beef cows are simulated in a similar manner,
although with a shorter period of lactation and lower milk production. Although dry feed
intake for the beef calves is simulated starting at 165 days of age, a 205 day lactation period
is used for the beef cows recognizing, as discussed above, that calves are not fully weaned
until about 205 days of age.
Beef bulls are simulated to be about 650 kg of empty body weight, or about 725 kg of
live weight. They are very active during the 90 day breeding period, during which they lose
weight. They are less active during the remainder of the year, during which they slowly regain
the weight lost during the breeding season.
5-11
-------
Exhibit 5-6
Representative Animal Characteristics: Heifers and Cattle Fed for Slaughter
Animal Type
Initial
Weight
(kg)a
Final
Weight
(kg)
Initial Age
(days)
Final Age
(days)
Other
Replacement Heifers:
Dairy Replacement Heifers:
0-12 months
Dairy Replacement Heifers:
12-24 months
Beef Replacement Heifers:
0-12 months
Beef Replacement Heifers:
12-24 months
170
285
165
270
285
460
270
390
165
365
165
365
365
730
365
730
--
Pregnant
-
Pregnant
Feedlot Fed Cattle for Slaughter:
Yearling Systemb
Weanling System0
170
170
480
480
165
165
565
422
fed to 26-27%
carcass fat
fed to 29-30%
carcass fat
a All weights reported as empty body weight.
b Includes 260 day stocker period principally on forages and a 140 day feedlot period with a high grain ration.
c Includes a 257 day feeding period, initially at 30 to 50 percent concentrate (125 days), followed by 132 days
of a high grain ration.
Exhibit 5-7
Representative Animal Characteristics: Dairy Cows and Beef Cows
Animal Type
Dairy Cows
Beef Cows
Beef Bulls
Initial and
Final Weight
(kg)a
550
450
650
Lactation/Dry
Periods
(days)
305/60
205/160
NA
Milk Production/
Lactation
(kg)
5,570-7, 190b
1,400
NA
Other
Pregnant
Pregnant
NA
a All weights reported as empty body weight.
b Milk production per lactation varies by region.
5-12
-------
Representative Diets
A total of 32 different diets were defined to represent the diverse feeds and forages
consumed by cattle in the U.S. Fourteen diets were defined for dairy cattle, including 6 for
dairy cows and four each for replacement heifers 0-12 months and 12-24 months. The 18
beef cattle diets include 3 each for beef cows, replacements 0-12 months, Weanling System
heifers and steers, and Yearling System heifers and steers. Four diets were defined for beef
replacements 12-24 months, and 2 diets were defined for beef bulls. The diets were defined
to reflect the broad range of feeding practices found throughout the U.S., which for purposes
of analysis was divided into the five regions shown in Exhibit 5-8.
Exhibit 5-8
Geographic Regions Used in the Analysis
West
North
Central
Includes Alaska and
5-13
-------
Dairy Cow Diets. Six dairy cow diets were specified and evaluated. These six diets
were used in various percentages to represent the typical diets in each of the five regions
(see Exhibit 5-9) and include various amounts of alfalfa hay, corn silage or grass hay forages
and concentrates with corn, soybean meal, barley, or other grains. During the 60 day dry
period grass forage was simulated to be fed, supplemented to 14 percent crude protein in all
regions. As shown in the exhibit, the simulated diets vary in quality from 2.41 Meal/kg to 2.69
Meal/kg.
The specification of the extent to which each diet is used in each region was based on
comments from cattle experts in the regions and the availability of alfalfa, corn silage and the
various grains regionally. For example, in the West a barley concentrate is simulated
because it is used in the Northwest. Although there is uncertainty in the specification of the
diets and the appropriate percentages for each region, the estimates of emissions form dairy
cows is not overly sensitive to these assumptions because the estimates of emissions per
unit of metabolizable energy are similar for each of the diets.
Based on the percentages shown in Exhibit 5-9, average dairy cow diets were
simulated reflecting the appropriate mix of diets and the average milk production in each
region. The simulated methane emissions per mature cow range from about 109 kg/yr in the
North Central region to 126 kg/yr in the South Atlantic. The emissions are lower for the
higher quality diets as expected. Exhibit 5-10 summarizes the estimates.
Dairy Replacement Heifer Diets. Four diets were simulated for each of the two groups
of replacement heifers: 0-12 months and 12-24 months. The 0-12 month replacement diets
are primarily good quality grasses and silages. A limited amount of grain is also included.
The diets for the 12-24 month replacement heifers are similar, but of higher quality to support
both growth and pregnancy.
As with dairy cows, the extent to which each of the diets was used in each region was
specified. The resulting estimates of emissions per head for each region reflect the weighted
average emissions for each diet using the percentages assigned for each region. For 0-12
month replacements the emissions estimates vary from about 18.9 kg/yr in the North Central
to about 20.7 kg/yr in the West. For the 12-24 month replacements the estimates vary from
57.4 kg/yr in the North Central to about 61.7 kg/yr in the South Central. Despite the
variations in the use of diets among regions, the regional average emissions per head are
similar. Exhibits 5-11 and 5-12 summarize the estimates.
Beef Cow Diets. Three diets were simulated for beef cows, with varying quality of
forages and use of supplements. Annual emissions for the three diets vary from about 54 to
72 kg/head. The three diets are simulated to be used in different percentages in the five
regions, so that annual emissions across the regions vary from about 60 kg/head to 71
kg/head. Exhibit 5-13 summarizes these results.
Beef Replacement Heifer Diets. Seven diets were simulated for beef replacement
heifers: three for 0-12 months and four for 12-24 months. As with the dairy replacements
age 0-12 months, relatively high quality diets were used because calves are limited in their
ability to consume and digest large amounts of forages. High quality grasses and corn silage
were simulated, with some supplementation. The diets for the 12-24 month olds were similar
to the beef cow diets, including adequate energy for both growth and pregnancy.
5-14
-------
The emissions estimates per head for the 0-12 month old replacements vary from
about 19 to 24 kg. The emissions estimates for the 12-24 month old replacements vary from
about 61 to 68 kg, which are similar to the emissions rates simulated for the mature beef
cows. Exhibits 5-14 and 5-15 summarize these estimates.
Diets for Feedlot Fed Cattle. Six diets were simulated for feedlot fed cattle. The three
diets simulated for the Yearling System include the stocker period, during which primarily
forages are fed. The Weanling System diets start immediately with rations of forages mixed
with grain concentrate. The emissions from the Weanling System are lower than the
emissions from the Yearling System because: (1) cattle on the Weanling System reach
slaughter weight 140 days faster than those on the Yearling System; and (2) the cattle on the
Weanling System consume a higher proportion of grain rations which do not produce as
much methane as the forage-based diets.
Emissions per head vary from about 50 to 54 kg for the Yearling System and from
about 25 to 30 kg for the Weanling System for the diets examined. These emissions
estimates cover the entire simulation period for each system, which exceeds 365 days for
both of the systems. Because nearly all feedlot fed cattle are finished in three regions (North
Central, South Central and West), the proportions with which each of the diets is used within
each of the three regions were specified. The cattle that go through a stocker phase in other
regions prior to feedlot feeding are implicitly counted in the three feedlot regions.
For this analysis, the model emissions estimates were adjusted to reflect the use of
ionophores and hormone implants. In the Weanling System, ionophores and implants will
each improve feed efficiency by 5 to 10 percent throughout the entire simulation period, so
that emissions will be about 10 to 20 percent lower than the model estimates. Therefore, the
emissions estimates were reduced by 15 percent for these cattle. In the Yearling System,
ionophores and implants are not as prevalent throughout the stocker phase. Consequently, a
10 percent reduction was used.
It has also been reported that ionophores can reduce methane production in the
rumen directly by changing rumen fermentation patterns. Recent studies with growing steers
indicate that this effect is temporary, lasting only about 16 days (Johnson, 1992).
Consequently, this effect is minor and is not considered here.
Exhibits 5-16 and 5-17 summarize the estimates per head for each of the feedlot
regions. As shown in the exhibits, the emissions per head do not vary significantly among
the regions. However, the emissions per head from the Yearling System are about twice the
emissions per head from the Weanling System.
Diets for Bulls. Two bull diets were simulated: (1) 90 days of 50% forage:50%
concentrate prior to the breeding season, followed by 100% forage for the remainder of the
year; and (2) 100% forage throughout the year. For both diets total feed intake was about
3,700 kg/year, with ME intake of about 9,500 Meal/year and 69 percent digestibility (on an
energy basis). Methane emissions for both diets were about equal, at 100 kg/head/year.
Exhibit 5-18 summarizes the emissions factors per head for all the cattle considered in
this study. As shown in the exhibit, the emissions factors range from about 20 kg/head for
dairy replacement heifers to about 115 kg/head for mature dairy cows. Across regions, the
emissions factors vary by about ±15 percent or less for each animal type, reflecting the
variation in diets consumed in each region.
5-15
-------
^
o
S.
0
O) W
10 Q
sa +*
A 4)
IE Q
2 £
o
u
n
Q
1
Q
5
**
0)
c
s
CO
o
IO
Q
**
0
b
CO
CD
b
CM
•&
b
CD
b
»0
™ *—
j_ ^_
tl
life
*~ QJ
•^ ^~ 5i
(5 §• Tt
LU CO •?•
- c
^* QJ
"«' £ £
£>E 8 8
o E| S
*S o CO in
(0
0 « 10 |
*~
- 2
m* ^^
' ffi
W >• <±
rfc co c
(D J2 CD
S S 8
CD" -
S'g
"co E
E E «S
o o CD
u u co
>^ vP vp
s!£
cz
c *
^ § n ST
•*- ° o O
His
= o c
CO U (D
Q
000
in w o
Description
,-
c\i
s
c\i
r-
m
c\i
CO
C\j
CO
in
c\J
T—
CO
c\i
"3
"
o>
DC
CO
CO
CO
CO
N£.
n
CO
|| North Atlantic
<£
o
CO
CO
x^
^~
*
1 South Atlantic
o
in
^?
~
in
|| North Central
<£
CO
CO
CO
CO
TJJ
1
I South Central
^p
3)
CM
^
N.
0)
(D
C
CD
O
CD
Q.
O
O
O
I
3
C
g
O)
CD
o
*^
CO
CD
at
eg
CD
k_
(D
Q.
0>
1
r^
C
O
O)
CD
O
0>
_c
TO
CD
Ifl
CD
O
T3
CD
^
1
10
OT
OD
T3
O
CO
0>
r»
O
O
C
0>
CD
CD
^
o
f~
_ m
CO CO
E -c o
c B ^
CD Q. 'C
>> •§ -
br: "75
II ° c
5 II .Q
CD Q- a?
CO O CC
-------
Exhibit 5-10
Regional Estimates of Methane Emissions from Mature Dairy Cows
Statistics for the Average Animal Modeled
Feed consumed
per year (kg DM)
MEa consumed
per year (Meal)
Diet ME (Meal/kg)
Average feed
digestibility (%)b
Methane
emissions per
year (kg/cow)
Milk Production
per Cow per Year
(kg)
Methane
emissions per kg
of milk produced
(9/kg)
N. Atlantic
5735
15,224
2.65
68
117.5
6710
17.5
S. Atlantic
5460
13,421
2.46
66
126.5
6110
20.7
N. Central
5805
15,012
2.59
66
109.4
6830
16.0
S. Central
5182
12,975
2.50
64
114.8
5570
20.6
West
6032
15,190
2.52
66
119.3
7190
16.6
a ME = metabolizable energy
b Digestibility is reported as simulated digestible energy divided by gross energy
intake.
Note: Regional diets are weighted averages of the diets shown in Exhibit 5-9.
5-17
-------
Exhibit 5-11
Regional Estimates of Emissions from Dairy Replacement Heifers: 0-12 Months
Statistics for the Average Animal Modeled
Diet Description
Feed consumed per
year (kg DM)
ME0 consumed
(Meal)
Diet ME (Meal/kg)
Average feed
digestibility (%)d
Methane emissions
(kg/head)
Diet 1
Alfalfa hay
1116
2623
2.35
62
21.4
Diet 2
75% alfalfa
hay, 25%
concen.3
1080
2684
2.48
65
20.0
Diet 3
High quality
grass forage
(CP=18%r
967
2613
2.70
67
20.1
Diet 4
Corn silage
with protein
to 14% CP
904
2432
2.69
69
14
Regional Distribution of Diets (%)e
North Atlantic
South Atlantic
North Central
South Central
West
25%
33%
25%
15%
50%
25%
60%
67%
50%
85%
25%
15%
25%
Emissions
(kg)
19.5
20.5
18.9
20.3
20.7
a Concentrate of corn meal and soybean meal
b CP = crude protein
c ME = metabolizable energy
d Digestibility is reported as simulated digestible energy divided by gross energy intake.
e Regional distribution of diets shows the extent to which each of the four diets is used in each region.
The emissions estimates are the weighted average emissions using these percentages.
5-18
-------
Exhibit 5-12
Regional Estimates of Emissions from Dairy Replacement Heifers: 12-24 Months
Statistics for the Average Animal Modeled
Diet Description
Feed consumed per
year (kg DM)
MEC consumed
(Meal)
Diet ME (Meal/kg)
Average feed
digestibility (%)c
Methane emissions
(kg/head)
Diet 1
Alfalfa hay
3184
7419
2.33
62
63.0
Diet 2
75% alfalfa
hay, 25%
concen.
a
3018
7437
246
64
57.3
Diet 3
Grass forage
of declining
quality6
3172
7183
2.25
58
61.4
Diet 4
Corn silage
with protein
to 14% CPC
2540
6801
2.68
67
47.9
Regional Distribution of Diets (%)f
Emissions
(k9)
North Atlantic
25%
50%
25%
58.4
South Atlantic
25%
10%
45%
20%
58.7
North Central
33%
33%
33%
57.4
South Central
20%
80%
61.7
West
50%
25%
25%
61.2
a Concentrate of corn and cottonseed meal
b High quality grass forage for 100 days (ME=2.8 Meal/kg). Intermediate quality grass forage for 100
days (ME=2.5 Meal/kg). Lower quality grass forage for 165 days (ME=2.1 Meal/kg).
c CP = crude protein
d ME = metabolizable energy
e Digestibility is reported as simulated digestible energy divided by gross energy intake.
f Regional distribution of diets shows the extent to which each of the four diets is used in each region.
The emissions estimates are the weighted average emissions using these percentages.
5-19
-------
Exhibit 5-13
Regional Estimates of Methane Emissions from Beef Cows
Statistics for the Average Animal Modeled
Diet Description
Feed consumed per
year (kg DM)
MEd consumed
(Meal)
Diet ME (Meal/kg)
Average feed
digestibility (%)e
Methane emissions
(kg/head)
Diet 1
Pasture for
7 mos; mixed
hay for
5 mosa
3029
7370
2.43
63
63.4
Diet 2
Pasture of
varying
quality6
3172
7731
2.44
63
71.7
Diet 3
Pasture with
4 mos of
supplement0
2700
7047
2.61
65
53.7
Regional Distribution of Diets
Emissions
(kg)
North Atlantic
80%
20%
60.5
South Atlantic
20%
80%
70.0
North Central
60%
40%
59.5
South Central
10%
90%
70.9
West
10%
80%
10%
69.1
a Seven months of pasture declining in quality as the seasons progress. Five months
of mixed hay, grass with some legumes.
b Pasture quality varies with the seasons.
c Pasture with four months of supplementation using a mixed forage (80%) and
concentrate (20%) supplement.
d ME = metabolizable energy
e Digestibility is reported as simulated digestible energy divided by gross energy
intake.
f Regional distribution of diets shows the extent to which each of the three diets is
used in each region. The emissions estimates are the weighted average emissions
using these percentages.
5-20
-------
Exhibit 5-14
Regional Estimates of Emissions from Beef Replacements: 0-12 Months
Statistics for the Average Animal Modeled
Diet Description
Feed consumed per
year (kg DM)
MEC consumed (Meal)
Diet ME (Meal/kg)
Average feed
digestibility (%)c
Methane emissions
(kg/head)
Diet 1
Legume
pasture with
supplement8
984
2443
2.48
65
18.1
Diet 2
Very high
quality grass
(18% CP)b
1011
2614
2.58
68
27.2
Diets
Corn silage
supplemented
to 14% CP
922
2454
2.66
68
15.8
Regional Distribution of Diets (%)e
Emissions
(kg)
North Atlantic
50%
20%
30%
19.2
South Atlantic
50%
50%
22.7
North Central
33%
33%
33%
20.4
South Central
40%
60%
23.6
West
50%
50%
22.7
a Concentrate = 25% of ration
b CP = Crude protein
c ME = metabolizable energy
d Digestibility is reported as simulated digestible energy divided by gross energy intake
e Regional distribution of diets shows the extent to which each of the three diets is used in
each region. The emissions estimates are the weighted average emissions using these
percentages.
5-21
-------
Exhibit 5-15
Regional Estimates of Emissions from Beef Replacement Heifers: 12-24 Months
Statistics for the Average Animal Modeled
Diet Description
Feed consumed per
year (kg DM)
MEe consumed
(Meal)
Diet ME (Meal/kg)
Average feed
digestibility (%)f
Methane emissions
(kg/head)
Diet 1
Varying
quality grass
forage8
2454
6356
2.59
67
66.9
Diet 2
Varying
quality grass
forageb
2675
6524
2.49
66
71.0
Diets
Varying
quality grass
with winter
supplement0
2359
5990
2.54
66
56.5
Diet 4
Varying
quality grass
with winter
supplementd
2305
6000
2.60
67
54.8
Regional Distribution of Diets (%)9
Emissions
(kg)
North Atlantic
50%
50%
63.8
South Atlantic
50%
40%
10%
67.5
North Central
33%
33%
33%
60.8
South Central
80%
20%
67.7
West
33%
33%
33%
64.8
a 165 days of high quality grass followed by 200 days of intermediate quality grass.
b 120 days of high quality grass followed by 125 days of intermediate quality grass -- grass hay
provided for 120 days during winter
c 120 days of high quality grass followed by 125 days of intermediate quality grass -- medium quality
alfalfa with a corrrsoybean meal concentrate (25%) provided for 120 days during winter
d 120 days of high quality grass followed by 125 days of intermediate quality grass -- corn silage
supplemented to 14% CP provided for 120 days during winter
e ME = metabolizable energy
f Digestibility is reported as simulated digestible energy divided by gross energy intake.
g Regional distribution of diets shows the extent to which each of the three diets is used in each region.
The emissions estimates are the weighted average emissions using these percentages.
5-22
-------
Exhibit 5-16
Regional Estimates of Emissions from Feedlot Fed Cattle: Yearling System
Statistics for the Average Animal Modeled
Diet Description
Feed consumed (kg DM)
ME consumed (Meal)
Diet ME (Meal/kg)
Average feed digestibility (%)c
Methane emissions (kg/head)
Adjustment for ionophores
and hormone implants
Methane emissions (kg/head)
Diet 1
Diet 2
Diets
All diets include forages during the stocker
phase followed by high grain diets during
feedlot feeding8
2865
7588
2.65
67
50.0
90%
45.0
2775
7383
2.66
67
54.1
90%
48.7
2755
7366
2.67
68
52.9
90%
47.6
Regional Distribution of Diets (%)c
Emissions
(kg)
North Central
30%
20%
50%
47.0
South Central
100%
47.6
West
20%
50%
30%
47.6
a All three diets include a high quality mixed hay (legume and grass) for the first winter (90 days).
The three diets then include:
Diet 1: mixed pasture (legume and grass) to 425 days of age; 50% alfalfa:50% concentrate for 40
days; 10% alfalfa:90% concentrate for 100 days.
Diet 2: grass pasture to 425 days of age; 50% alfalfa:50% concentrate for 40 days;
10% alfalfa:90% concentrate for 100 days.
Diet 3: grass pasture to 425 days of age; 70% corn silage:30% concentrate for 40 days;
10% a(falfa:90% concentrate for 100 days.
b ME = metabolizable energy
c Digestibility is reported as simulated digestible energy divided by gross energy intake.
d Regional distribution of diets shows the extent to which each of the four diets is used in each
region. The emissions estimates are the weighted average emissions using these percentages.
Only the three regions with feedlots are shown.
5-23
-------
Exhibit 5-17
Regional Estimates of Emissions from Feedlot Fed Cattle: Weanling System
Statistics for the Average Animal Modeled
Diet Description
Feed consumed (kg DM)
MEb consumed (Meal)
Diet ME (Meal/kg)
Average feed digestibility (%)c
Methane emissions (kg/head)
Adjustment for ionophores
and hormone implants
Methane emissions (kg/head)
Diet 1
Diet 2
Diets
All diets include mixed rations with increasing
amounts of high grain concentrates3
1935
5232
2.70
68
31.2
85%
26.5
1763
5184
2.94
71
25.3
85%
21.5
1742
5059
2.90
71
25.4
85%
21.6
Regional Distribution of Diets (%)c
Emissions
(kg)
North Central
20%
20%
60%
22.6
South Central
50%
50%
24.0
West
40%
30%
30%
23.5
a The following diets were simulated:
Diet 1: 60% alfalfa:40% concentrate for 125 days; 10% atfalfa:90% concentrate for 132 days.
Diet 2: 50% alfalfa:50% concentrate for 125 days; 10% alfalfa:90% concentrate for 132 days.
Diet 3: 69% corn silage:31% concentrate for 125 days; 10% alfalfa:90% concentrate for 132 days.
b ME = metabolizable energy
c Digestibility is reported as simulated digestible energy divided by gross energy intake.
d Regional distribution of diets shows the extent to which each of the four diets is used in each
region. The emissions estimates are the weighted average emissions using these percentages.
Only the three regions with feedlots are shown.
5-24
-------
Exhibit 5-18
Methane Emissions Factors for Cattle by Region
Statistics for the Average Animal Modeled
(kg/head)
Dairy Replacement
Heifers: 0-12 Months
Dairy Replacement
Heifers: 12-24 Months
Dairy Cows
Beef Replacement
Heifers: 0-12 Months
Beef Replacement
Heifers: 12-24 Months
Beef Cows
Yearling System
Heifers and Steers
Weanling System
Heifers and Steers
Bulls6
N. Atlantic
19.5
58.4
117.5
19.2
63.8
61.5
NAa
NA
100.0
S. Atlantic
20.5
58.7
126.5
22.7
67.5
70.0
NA
NA
100.0
N. Central
18.9
57.4
109.4
20.4
60.8
59.5
47.0
22.6
100.0
S. Central
20.3
61.7
114.8
23.6
67.7
70.9
47.6
24.0
100.0
West
20.7
61.2
119.3
22.7
64.8
69.1
47.6
23.5
100.0
a NA = not applicable
b Emissions from bulls are estimated nationally with a single emissions factor.
5.3.4 Methane Emissions from Other Animals
The method for estimating methane emissions from other animals consists of:
• selecting a methane emissions factor in kilograms per head per year for each
animal; and
multiplying the emissions factor by the animal population in the U.S.
Average emissions factors for each of the major animals have been published by Crutzen et
al. (1986). These emissions factors consider typical animal sizes, feed intakes, and feed
characteristics for developed and developing countries. The emissions factors for goats,
sheep, pigs, and horses developed by Crutzen et al. for developed countries are used in this
analysis. The emissions factors and the assumptions upon which they are based are listed in
Exhibit 5-19.
5-25
-------
Exhibit 5-1 9
Emissions Factors Used for Other Animals
Animal
Sheep
Goats
Pigs
Horses
Mean Body
Weight
(kgs)
60
NR
NR
550
Gross Energy
Intake
(Meal/day)
4.8
3.3
9.1
26.3
Energy Intake
Released as
Methane (%)
6
6
0.6
2.5
Emissions
Factor
(Kg/head/yr)
8
5
1.5
18
NR = not reported.
Source: Crutzen et al. (1986).
5.4 CURRENT EMISSIONS
5.4.1 U.S. Cattle Population
To apply the emissions factors for cattle developed in the previous section, the U.S.
cattle population must be estimated for each of the representative animal types. For all but
the feedlot fed cattle, the population estimates are available from publications for 1990:5
0-12 months: 4.2 million;
12-24 months: 4.2 million;
• Dairy Replacement Heifers:
• Dairy Replacement Heifers:
• Dairy Cows: 10.1 million;
Beef Replacement Heifers: 0-12 months: 5.5 million;
Beef Replacement Heifers: 12-24 months: 5.5 million;
Beef Cows: 33.5 million; and
Beef Bulls: 2.2 million.
These data, totaling 65.2 million head, are expressed on an annual basis, and can be
interpreted as representing the average number of head in each category for the year.
However, as described above the feedlot fed cattle are simulated for periods exceeding a
year. Consequently, the appropriate population to use in conjunction with the emissions
factors was derived from annual slaughter statistics.
USDA (1992a) reports that 22.5 million head of cattle were marketed from the feedlots
in the 13 major feedlot states in 1990. National statistics of marketed cattle are not available.
However, CF Resources (1991) reports that on average the 13 states accounted for about
5 For example, Schoeff and Castaldo (1991) present data nationally and by region. Various USDA publications
show similar national data.
5-26
-------
85.5 percent of all cattle on feed nationally. Assuming that the 13 states also account for
85.5 percent of all cattle marketed from feedlots, the national total is estimated as: 22.5
million * 0.855 = 26.3 million.
To maintain this level of fed cattle marketing, 26.3 million calves must be born
annually. Combining this estimate with the replacements (4.2 + 5.5 = 9.7 million) and the
losses associated with deaths and veal slaughters (4.5 million) yields a total calf crop
estimate of 40.5 million. This figure corresponds reasonably well to USDA (1992b) estimates.
This figure can also be checked by estimating the total annual cattle slaughter. Using the
culls rates for beef and dairy (9.7 million) and the estimate of marketed cattle from feedlots
produces an estimate of 36 million, which again is in good agreement with USDA (1992b)
estimates.
Finally, this estimate can be checked by estimating the average annual population of
cattle that are or will be fed in feedlots. Based on discussions with industry representatives, it
was estimated that about 70 to 90 percent of the cattle marketed from feedlots were grown
using the Yearling System, with the remaining using the Weanling System. Using an 80
percent estimate implies that the average annual cattle population for this segment of the
industry is:
[80% x (565 days) + 20% x (422 days)] * 365 days x 26.3 million = 38.6 million.
Adding this estimate to the population of the other segments yields a total annual average
population of 103.8 million. This estimate is well within the range of USDA (1992b) estimates
of population, which fluctuate during the year. For example, USDA (1992b) estimates the
January 1, 1990 cattle population at 98.2 million and the July 1, 1990 cattle population at
107.4 million. Based on the consistency checks, this estimate of 26.3 million is the
appropriate annual figure to use with the emissions factors for feedlot fed cattle. The
remainder of this section summarizes the emissions estimates based on the emissions factors
and the animal populations.
5.4.2 Cattle
Using the emissions factors developed above and regional dairy cattle populations,
national emissions from the dairy industry are estimated at about 1.5 Tg in 1990. Dairy cows
account for the majority of these emissions, nearly 1.2 Tg. The North Central region, with
over 40 percent of the nation's dairy cows has the largest emissions. Exhibit 5-20
summarizes the emissions estimates.
Emissions from the beef industry are estimated at about 4.0 Tg in 1990. Beef cows
account for the majority of this emissions estimate, approximately 2.2 Tg. Cattle that are
feedlot fed account for the next largest source, about 1.1 Tg. This emissions estimate
includes both the Weanling and Yearling Systems, and consequently includes emissions
during the stocker phase as well as the time spent in feedlots. The emissions estimates for
bulls and replacements are relatively smaller.
With about one-third of the nation's beef cows, the South Central region has the
largest beef cow emissions estimate. The North Central region, however, accounts for the
largest share of emissions from feedlot fed cattle, and has about 30 percent of the beef cows
as well. Exhibit 5-21 summarizes these estimates.
5-27
-------
Exhibit 5-20
Methane Emissions From U.S. Dairy Cattle
Region/Animal Type
North Atlantic
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
South Atlantic
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
North Central
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
South Central
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
West
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
National Total
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
Total
Emissions Factor
(kg/head/yr)
19.5
58.4
117.5
20.5
58.7
126.5
18.9
57.4
1094
20.3
61.7
114.8
20.7
61.2
119.3
19.6
58.8
114.6
80.4
Population
(000 Head)
712
712
1,795
268
268
710
1,987
1,987
4,497
405
405
1,156
833
833
1,972
4,205
4,205
10,130
18,540
Emissions
(Tg/yr)
0.014
0.042
0.211
0.005
0.016
0.090
0.038
0.114
0492
0.008
0.025
0.133
0.017
0.051
0.235
0.082
0.247
1.161
1.490
5.4.3 Comparisons with Previous Cattle Emissions Estimates
Total emissions trom dairy and beef cattle in the U.S. are estimated to be about 5.5 Tg
per year for 1990. Compared with previous estimates, this value is in the middle of the range.
Crutzen et al. (1986) estimated an average emissions factor of 55 kg per head for cattle in
developed countries. Using Crutzen's emissions factor and the total U.S. cattle population of
about 100 million, the 1990 emissions would be about 5.5 Tg per year.
A detailed estimate of U.S. cattle emissions was prepared by Johnson et al. (1991)
showing total emissions for the beef and dairy sectors of 6.0 Tg per year. While this estimate
is only about 10 percent higher than the estimate in this study, the main factors causing
Johnson's estimates to be higher are:
• Johnson et al. estimate a 12 percent higher emissions factor for dairy cows
because they estimate a 16 percent higher feed energy intake. This difference
accounts for nearly one-half of the difference in the total estimates.
5-28
-------
Exhibit 5-21
Methane Emissions From U.S. Beef Cattle
Region/Animal Type
North Atlantic
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
South Atlantic
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
North Central
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
Weanling System Steers/Heifers6
Yearling System Steers/Heifers
South Central
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
Weanling System Steers/Heifers
Yearling System Steers/Heifers
West
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
Weanling System Steers/Heifers
Yearling System Steers/Heifers
Bulls: Nationally
National Total
Replacements 0-12 months
Replacements 12-24 months
Mature Cows
Weanling System Steers/Heifers
Yearling System Steers/Heifers
Bulls
Totald
Emissions Factor
(kg/head/yr)
19.2
63.8
61.5
22.7
67.5
70.0
20.4
60.8
59.5
22.6
47.0
23.6
67.7
70.9
24.0
47.6
22.7
64.8
69.1
23.5
47.6
100.0
22.3
65.0
66.7
23.1
47.3
100.0
47.5
Population
(000 Head)3
87
87
337
594
594
3,418
1,546
1,546
10,592
2,963
11,852
2,079
2,079
12,359
1,164
4,656
1,229
1,229
6,772
1,133
4,532
2,200
5,535
5,535
33,478
5,260
21,040
2,200
85,398C
Emissions
fTg/yr)
0.002
0.006
0.021
0.013
0.040
0.239
0.032
0.094
0.630
0.067
0.557
0.049
0.141
0.876
0.028
0.222
0.028
0.080
0.468
0.027
0.216
0.220
0.124
0.360
2.234
0.122
0.994
0.220
4.054
a Population for slaughter steers and heifers in each region is the number slaughtered annually.
b The emissions from Yearling and Weanling System steers and heifers are assigned to the regions in which
they are managed in feedlots.
c The national population is estimated using the average annual population of Yearling and Weanling System
cattle: 38.65 million. See text.
d Total may not add due to rounding.
5-29
-------
• Johnson et al. estimate a higher emissions factor tor the heifers and steers that
are grown for feedlot feeding and slaughter. This differences account for about
one-third of the difference in the total estimates.
These factors account for about 80 percent of the difference between the estimates. There
are relatively small differences in the estimates of the portion of feed energy that is converted
to methane, and these differences do not contribute significantly to the differences in the
estimates of total emissions. Exhibit 5-22 compares the details of the estimates by Johnson
et al. with the details of the estimates in this study for each segment of the animal population.
Exhibit 5-23 summarizes the national estimates from each study.
Byers (1990) estimated methane emissions from U.S. cattle at about 4.0 Tg per year.
Exhibit 5-24 shows that the estimates by Byers are lower than the estimates in this report for
every animal category. The reasons for the differences in the estimates are not known
because details are not provided in Byers (1990).
5.4.4 Emissions from Other Animals
Methane emissions from other animals are estimated to be about 0.3 Tg per year. As
expected, these emissions are very small as compared with the emissions from dairy and
beef cattle. Exhibit 5-25 presents the emissions factors and animal populations used. Pigs,
sheep, and horses contribute equally to national emissions.
5.4.5 Uncertainties
Although the emissions estimates are presented as point estimates, there are a variety
of factors that make the emissions estimates uncertain. First, animal population and
production statistics are uncertain. In particular, estimates of the population of beef cows
managed under extensive range conditions are uncertain. Second, methane emissions from
cattle are influenced by the characteristics of the feed consumed. The diets analyzed using
the rumen digestion model are broad representations of the types of feeds consumed within
each of the major regions. Consequently, uncertainty is added due to the inability to
represent the full diversity of feeding strategies that are used. Finally, the rumen digestion
model was validated using experimental data that is itself uncertain. The estimates from the
rumen digestion model can be no better than the underlying experimental data upon which it
is based.
Data do not exist upon which to base an objective assessment of the uncertainties in
the estimates. A subjective assessment of the uncertainty in the estimate of total methane
emissions from cattle is as follows:
Biases in the estimates are not anticipated. It is assumed that the point
estimates are mean values.
• Uncertainty is defined as a range about the mean. The confidence level for the
range defines the likelihood that the true value (in this case the true emissions
rate) falls within the specified range. Because this assessment of uncertainty is
subjective, an objective quantification of the confidence level is not possible.
5-30
-------
CO
c
V)
c
o
a
V)
CM •K
CM CO
3 i
£ (0
a iu
^ -Q
1
D
"5
c
o
CD
0.
o
o
>*
W
I
CO
"5
S
c
O
CO
SL
O
A
CO
CO
a
IS
UJ
JO ^
X .^
o —
„ i^> ^"
o F?g
0 LU S
>*eo *CO
t> CD "cc
|— C *±CO
LU £ O
*—
S? T3
CO .CD
D "-
cfo
Q- C.
||
.Q _
I .i£
O "-"
>. •*
"B EPS
CD C-'
* jit O
'
>,« g
S 21 .2- "s
o c 2TO
LU £ O
*~
Is
ti(D~~
Q. o
<2 CD
E Q.
S
IO
CO
CO
CO
IO
8
in
CO
CO
to
O
CD
CO
CO
co"
CD
CM
CO
••-
10
in
CD
CO
Is-
CO
CO
CO
O
CD
CO
CO
T—
O
CO
CO
CM
m
CO
CO
o
to
o
O
'(0
o
CM
CO
CM
00
in
^
c\l
in
to
CO
Is-
o
o
O
Q
CO
a>
CO
o
CM
IO
8
f-
o>
12-24 mo.
replacements
CO
10
m
CO
Is:
••~
S
CO
en
o>
05
0 1
E 1
S^
CNJ g-
o
I
CM
CO
CM
CO
^ —
m
CO
CO
CM
CM
at
CO
p
o
CO
05
CM
IO
CM
CM
6
5
CM
h-
10
^.
CO
in
*
CO
CO
"D
to
CD
6
"o
5
05 co
CO T3 jj,
CO c C
C CD •£
O £ CO
11 |
to f?
2 to" —
x S- ?
CD >- .5£
.<2 to T,
52 ro
2m CO
"D "^ "^ CZ S
5 * E E
>• | - 1 2
•§ ^ o " 5
O ^ 73 o_ ST\
w QJ ^^ ~*
^ | B^ «
P t = O c
— t 3 tJ (0 QJ
C C g 05 g
T3 "O **~ o ®
td r- —
w ° s~
ar 2 £ is w
*" -~ 0> T ^*
5 | is g- ¥
o o OT ^ -a
*~ _ to" -C jg
! 1 il i
•^ ._ o E .a
CD CD CD •"* CD
a. Q- TJ n -a
Oj tO U .£ .2
iS o g « g
*C W (0 n
1 1 ll 1
I 1 ll ll
(O ^3 0 E "D T3
-------
Exhibit 5-23
Comparison of National Estimates by Johnson et ai.
(Tg/year)
Animal Type
Johnson et al.
This Study
Comments
Beef Cows
2.3
2.2
Johnson et al. have a higher feed intake and
lower methane conversion
Dairy Cows
1.4
1.2
Johnson et al. have a higher feed intake and
lower methane conversion
12-24 mo. replacements
0.6
0.6
Estimates are similar
Bulls
0.2
0.2
Johnson et al. have a higher feed intake and
lower methane conversion
All Others8
1.5
1.3
Johnson et al. have a higher feed intake and
lower methane conversion
Totalc
6.0
5.5
a All others include calves, stockers, and feedlot cattle.
b Totals may not add due to rounding.
Exhibit 5-24
Comparison of National Estimates by Byers
(Tg/year)
Animal Type
Beef Cows
Dairy Cows
Beef Replacements 12-24 months
Dairy Replacements 12-24 months
Bulls
Calves
Stockers and Feedlot Cattle
Total8
Byers
1.81
0.94
0.16
0.15
0.12
0.06
0.71
3.96
This Study
2.23
1.16
0.36
0.25
0.22
0.21
1.12
5.54
a Totals may not add due to rounding.
5-32
-------
Exhibit 5-25
Methane Emissions from Other Animals
Animal
Sheep
Goats
Pigs
Horses
Total
Emissions Factor
(kg/head/yr)
8
5
1.5
18
--
Population
(000 head)
11,364
1,900
53,852
5,215
--
Emissions
(Tg/yr)
0.09
0.01
0.08
0.09
0.27
Sources: Emissions factors from Crutzen et al. (1986).
Populations from FAO (1991).
Nevertheless, the range developed here was selected to represent the range
into which of the mean emissions rate is very likely to fall. As such, the range
is analogous to an objectively-estimated uncertainty range, such as a 90 or 95
percent confidence interval.
Animal populations are taken from industry and government statistics. In the
U.S. these data are scrutinized closely and are checked against independent
production statistics, so a relatively narrow uncertainty range is warranted. A
range of ±5 percent is assumed as the uncertainty range for animal
populations. This range is consistent with the intra-annual fluctuations in the
cattle population estimates made by USDA each year.
Feed characteristics are not routinely reported and were derived based on
aggregate feed production statistics and typical grazing resources by region,
subject to constraints imposed by the simulations of animal performance. The
variations in the emissions factors by region, which reflect the variations in feed
consumption by region, are all less than about 10 percent of the national
average emissions factor for each animal type. This result shows that although
there is diversity in feed type consumed across the regions, the emissions
factor estimates are not overly sensitive to this diversity. To be conservative, it
is assumed that the feed data specifications result in an uncertainty of ±15
percent of the final emissions estimates.
There is uncertainly in the estimates of the portions of the feedlot fed cattle that
are managed using the Weanling and Yearling Systems. Changing the
adopted value of 80 percent in the Yearling System by 10 percent (e.g., to 70
or 90 percent) causes the emissions estimate to change by less than 0.1 Tg.
Therefore, this assumption does not contribute significant uncertainty to the
total emissions estimate.
The uncertainty in the experimental data and the model based upon the data
contribute to the uncertainty in the estimates. However, the physiology of
5-33
-------
methane production in cattle constrains the range of this uncertainty. Although
individual measurements vary, average methane production rates as a
percentage of ME or gross energy (GE) intake for normal well-fed cattle on
forage-based diets usually fall within a fairly narrow range of ±15 percent of a
mean value (e.g., 6.0 ±1.0 percent of GE). It is also known that high grain
diets have lower average methane production rates, and similarly typically fall
within a fairly narrow range. Because the model emissions estimates fall in the
expected ranges relative to simulated ME and GE intakes (which are verified
through observations of animal performance), the uncertainty in these
estimates should be no larger than the observed variability . Therefore a range
of ±15 percent is adopted for the uncertainty contributed by the model.
These sources of uncertainty compound to result in an overall uncertainty of about ±20
percent in the estimate of methane emissions from cattle.6 Although the uncertainty in the
estimate of emissions from Other Animals is probably larger (because a less detailed analysis
was performed), the uncertainty in the cattle estimates drive the overall uncertainty because
cattle account for the overwhelming majority of emissions. Therefore, this uncertainty range
of ±20 percent is applied to the mean national estimate of 5.8 Tg/yr to produce a low
estimate of 4.6 Tg/yr and a high estimate of 6.9 Tg/yr.
5.5 FUTURE EMISSIONS
5.5.1 Future Emissions - "Current Practices" Scenario
Future methane emissions from cattle and other animals in the U.S. will be driven by
future levels of production and the production methods used. Based on past trends, both the
beef and dairy industries are improving in productivity and efficiency. In the dairy industry,
milk production per cow continues to increase. The beef industry has recently initiated efforts
to reduce fat production and orient beef marketing toward a value-based system. These
initiatives may reduce methane emissions per unit of production in the future. However, for
this "Current Practices" scenario, the implications of potential changes in production practices
for future emissions are not considered. Only changes in the level of production drive the
estimates of future emissions for this scenario.
Future levels of milk and meat
production will be driven principally by
domestic demand. However, the ability of
U.S. suppliers to compete with other
producers in both the domestic market and
improvements in productivity and
efficiency in the dairy and beef
industries can help offset potential
increases in emissions due to increased
levels of production in the future.
foreign markets may also play a role.
Because international trade in agricultural
commodities is highly influenced by national
and international trade policies and agreements, future U.S. production, and hence future U.S.
emissions, will be influenced by potential changes in these policies.
6 The uncertainties are compounded assuming each source of uncertainty is normally distributed and
independent. The resulting range in the mean emissions estimate is a subjective range based on the subjective
assessment of the individual uncertainties discussed in the text.
5-34
-------
Over the past 15 years strong trends have been evident in the domestic production
and consumption of milk and meat products. Domestic per capita meat consumption has
increased by about 0.9 percent per year during the period 1975 to 1990 (see Exhibit 5-26).
During this period, the market share of the various meats has changed substantially:
• per capita consumption of beef, veal, lamb, and mutton has decreased nearly
25 percent;
• per capita consumption of pork initially increased by about one third, and then
leveled out and declined; and
per capita consumption of poultry increased by about 75 percent.
Because the total U.S. population has increased, U.S. beef production has remained fairly flat
during this period despite substantial reductions in per capita consumption. Pork production
increased by about 34 percent during this period, and poultry production more than doubled.
Although international trade in meat fluctuates from year to year, these fluctuations were not
an important factor in the overall production trends during this period (see Exhibit 5-27).
Domestic milk production and consumption increased substantially during the 1975 to
1989 period. Production and domestic consumption rose by about 25 percent, while per
capita consumption rose by about 8.4 percent. Net milk imports are small compared to
domestic production and consumption, and consequently have not affected these trends (see
Exhibit 5-28).
In the absence of significant changes in international trade agreements, future
production and consumption of milk and meat products in the U.S. are anticipated to
continue along these trends. The Food and Agricultural Policy Research Institute (FAPRI
1991) presents the results of one assessment of future U.S. production and consumption
through 2001 under this assumption of no changes in trade policies. The FAPRI assessment
indicates that:
• beef production will grow slowly (about 0.5 percent per year) as exports grow
to offset the continued decline in per capita consumption;
• pork production will fluctuate resulting in a slight increase in total pork
production as per capita consumption declines net imports are reduced;
poultry production will increase by nearly 40 percent as per capita
consumption continues to increase rapidly; and
milk production will increase due to the increase in population and the slight
increase in consumption per capita as net imports are constant.
Exhibit 5-29 shows these FAPRI results, and an extrapolation of the estimates to the year
2010. As shown in the exhibit, by 2010 beef production may be essentially the same as
1990, indicating that methane emissions from cattle will be essentially unchanged assuming
that production practices remain unchanged. Because milk production is expected to
increase by about 18 percent by 2010, methane emissions from the dairy sector may increase
by this amount, again assuming no change in production practices.
5-35
-------
Exhibit 5-26
U.S. Per Capita Meat Consumption: 1975-1989
(Pounds of Retail Product Equivalent)
Pounds
80
60
40
20
0
O^^___^ PouJ
t ^-— " " " ^
^^-^^^^^ ' ~~— — p
.r"^ ~~
----
o — o
1975 1980 1985
Year Beef Veal Lamb/ Pork P
Mutton
1975 89.4 3.5 1.8 43.4
1980 77.3 1.5 1.5 57.8
1985 79.6 1.8 1.4 52.2
1989 69.5 1.2 1.5 52.2
1975-89 Change -22.3% -64.3% -15.9% 20.2%
Annual Change -1.8% -7.1% -1.2% 1.3%
Source: USDA (1990).
tryo
Beef
~~~0
orl
Q
O
1900
oultry Total
49.4 187.5
61.6 199.7
71.0 206.1
86.7 211.1
75.5% 12.6%
4.1% 0.9%
All data in equivalent retail product. Carcass weights converted to retail weights for
red meats based on data in AMI (1991). Poultry consumption includes chicken
and turkey.
5-36
-------
Exhibit 5-27
U.S. Meat Production and Trade: 1975-1989
Production Relative to 1 975
1975
250
200
Year3
1975
1980
1985
1989
1975-89
Change
Annual
Change
150
100
50
0
= 100
^^'-^'"'
a^"""' __^
^oultry ( ,
Purl;
Beef
1
__^_ Veal/Lamb/ Mutt on|
1Q75 IQSO 1985
Beef Veal Lamb/Mutton Pork
Prod'n Trade Prod'n Trade Prod'n Trade Prod'n 1
23,975 1,768 873 24 411 25 11,779
21,643 1,917 400 19 318 34 16,617
23,728 1,795 515 19 359 29 14,807
23,087 1,239 355 1 347 59 15,759
-3.7% -59% -16% 34%
-0.3% -6.2% -1.2% 2.1%
1QQC
Poultry
rade Prod'n Trade
286 10,526 (202)
227 14,541 (695)
58 17,340 (465)
17 22,279 (878)
112%
5.5%
a All production and trade data in millions of pounds. All red meat data in carcass equivalent. All poultry
data in retail equivalent. Poultry production includes chicken and turkey. All trade data are net imports. The
negative values for poultry indicate net exports. The graph of production is shown relative to 1975 (i.e., with
1975=100) because the red meat data are in carcass equivalent and the poultry data are in retail equivalent.
Source: USDA (1990)
5-37
-------
Exhibit 5-28
Domestic Milk Production and Consumption: 1975-1989
Per CapitJ
Production Consumption
C Billioi
150
140
130
120
110
100
i Pounds.) ( Po
Production
.&
/ J
.s ^Q~^^ P
-------
Exhibit 5-29
Future U.S. Meat and Milk Production and Consumption Based on the FAPRI Study
Period
1990to
2000
2000 to
201 Oa
Beef
Per Capita
Consum'n
-10%
-10%
Prod'n
5.0%
-7.0%
Pork
Per Capita
Consum'n
-6.0%
-6.0%
Prod'n
4.3%
-2.9%
Poultry
Per Capita
Consum'n
26%
24%
Prod'n
37%
28%
Milk
Per Capita
Consum'n
2.2%
2.2%
Prod'n
12%
5.6%
Source: Estimates for 1990 to 2000 from FAPRI, 1991.
a Estimates for 2010 meat production and consumption developed assuming that: reductions in per capita
consumption of beef and pork continue at the 1990 to 2000 rate; total per capita meat consumption increases at
the 1990 to 2000 rate; per capita poultry consumption increases at the rate needed to account for the reduction
in red meat consumption and the increase in total meat consumption; population increases by 3.3 percent from
2000 to 2010; and net imports are constant from 2000 to 2010. Per capita milk consumption is assumed to
increase at the 1990 to 2000 rate.
There are, however, possible changes in international trade policies that would cause
U.S. production of milk products to increase at a rate greater than the rate shown in
Exhibit 5-29. Additionally, efforts are underway in the beef industry to stop the loss of beef's
market share of total meat consumption in the U.S. These factors could lead to higher U.S.
beef and milk production in the future than indicated in Exhibit 5-29, and consequently higher
methane emissions.
As an alternative scenario of future meat and milk production, Exhibit 5-30 displays the
estimates of milk and meat production and consumption in the U.S. published by U.S. EPA as
part of its evaluation of policies for stabilizing global climate (USEPA 1989). This scenario,
based on the results of an international agricultural model that incorporates demand and
supply functions for the major agricultural commodities, including grains, meat, and milk
products, trade policies and domestic agricultural pricing and supply policies, indicates the
following:
per capita consumption of beef may increase from 1990 to 2000 and then
decrease from 2000 to 2010;
• per capita consumption of milk may increase throughout the period 1990 to
2010; and
• milk exports may increase so that the U.S. becomes a significant net exporter
of milk products.
5-39
-------
^
TJ
cl
a.
• ••
in
Q)
0
TJ
0)
(0
a
m
c
o
E
3
CO
O £
co o
10 °
m p
~ (0
UJ O
|
TJ
O
£
£
1
TJ
CO
CO
CD
^~
CO
D
0)
2
"5
•M
Jtf
S
V
3
Q
a.
a
££,
a
„.
4>
m
§
•5
1
a
5 ^*
a.E
CD 3
Og
15
Q
Productii
• c
*i
O «
ao
c
o
S
T)
O
a
CO r-
Per Caplt
Consum'l
o
t>
•o
o
a
CO c
ll
U «
c
d> O
0.0
TJ
—
cu
O,
&
00
w
e
CO
1
CO
ss
12
in
cd
^
in
g
to
1
o
o
s
$
(O
0
o
1
CM
0
c\i
in
g
c\i
co
i
o
o
*"*
o
s
S i5
Q. CD
•-- **
5 S.
CO ^v.
to r^>
CD O
^ CO
II
s?
CD ffl
£ c
" 0
"'H
CD c
18 i
CD CO
11
|1
D
o 1
°-£
•Q
§ g
C ffl
o c.
Q^ »J
E £ ffl
^ ^ ^55
o *•- o
O c
ST Q o Jo
?- ij " ^
< ? *" 2
a. £ -a °-
UJ oj Q> ••-
™ |lg
E § c o
0 CO CD Q.
fc ffl« S
1 ll?
m O •{= T3
"D « S
> co ffl
jC CD CD • -
E ffl o
"° s tj S.
ffl o •- ^
•S °- c m
a> TJ .2 S
f S£8
o ^ t *g
w 6 io ®
CD Q. S E
•w ,_ o
P 0 ° «
'^ S c2
uj 2 3 c
wo —
P O_ ®
O "^ W ffl
^ ."2 .« CD
o ffl y
c/) co o .5
(0
CO
•o
3
85
Q.
UJ
•o
CD
E
Q.
0)
4^
O
Tl
w
CO
ffl
o
co Q-
10 |
£ M
a c
I t
c
CO
c
_o
*^
a
•o
S
a
je
I
•o
C
CO
n
0)
S
(A
D
3
3
IL
0
0
II
r-
•i
0
0
H
Ji
t!
0
H
1
} 0
) II)
1 m
\
1
3 0
D V)
v n
A
A 01
ft (X
0) H
U H
a.
< p,
t, <
W IK
,
! ,
i i
t T- y
0
\ n.
^
J 6
^ '
1
•H '
i
n
i
i
i
i -
I
\. •
\
V
A
\
V
V
1 . 1 1 1 . . .
o o o o a r
o m o ID o if
n IN N * *
* k,
2 :ifl: "
\ i
\
\ t j
V \ It
^ ^ M ^
\ \ 111
\ \ I U
v \ V
s\ »
V <(l f
\ \
\ k i
\ / \
V \
\/ "
V
i . i . t . i
0 0 0 0 0 C
o n o n o »
m « N H H
0
0
(N
0
0
0
IN
0
Oi
H
0
00
Oi
^
o
^
a
H
0
(N
0
G
0
D
0
0
00
0
H
0
h
) ^
)
-------
This scenario differs from the FAPRI scenario principally because it shows an initial increase
in per capita beef consumption and a substantial increase in milk exports. In order for beef
consumption per capita to increase, the recent trend in declining consumption would have to
be reversed. In order for milk exports to increase substantially, trade agreements that lift
restrictions on world trade in milk products would have to be implemented.
Given the diverse factors that will affect future production and consumption of milk
and meat products in the U.S., these two scenarios are adopted as a range of the potential
future. Exhibit 5-31 summarizes these estimates of future production trends. In the absence
of changes in production methods, using these scenarios, emissions in 2000 may range from
about 5.0 to 7.9 Tg/yr and emissions in 2010 may range from 4.8 to 8.2 Tg/yr (see
Exhibit 5-32). The uncertainty in the future level of production accounts for about 25 percent
of the uncertainty in 2010 emissions, with the ±20 percent uncertainty (estimated above)
accounting for the remainder of the range.
5.5.2 Opportunities for Emission Reductions
Future methane emissions could be reduced substantially if any of a variety of
emissions reduction opportunities are implemented over the next decades. Strategies that
improve animal productivity, especially through improved nutrition, will result in a reduction in
methane emissions per unit of product produced. Strategies such as improved feed
characteristics, selective breeding and other intensive management techniques have been
shown to greatly increase productivity. When coupled with saturated market conditions and
widespread adoption, improvements such as these can result in absolute reductions of
methane emissions.
A variety of techniques are available or under development that can increase animal
productivity and reduce methane emissions per unit product in the U.S. in the near term.
Both the dairy and beef industries have efforts underway to improve productivity and
efficiency. Examples of these techniques include the following:
• Improvement in Beef Cattle - Considerable opportunity exists to improve the
genetic characteristics of beef animals in the U.S. Research is needed to
develop objective measures of heritable desirable traits. Additionally, the
marketing system needs to provide incentives to invest in improved genetics.
Recently, the beef industry has initiated considerable efforts in this area, and
progress is anticipated. These advances will help to offset potential future
increases in emissions associated with increased levels of production.
Targeted Mineral/Protein Supplements -- The use of supplements is a well
known technique for correcting specific deficiencies in minerals and protein
among grazing and fed animals in the U.S. Reproductive efficiency among
some beef cows may be enhanced substantially. Current implementation is
limited by lack of data on critical deficiencies and the existing marketing/pricing
arrangements in the industry.
bST - Bovine somatotropin (bST) is a naturally-occurring product of the cow's
pituitary gland. During the past 10 years it has become possible to produce
large quantities of bST using recombinant DMA techniques. bST has been
5-41
-------
Exhibit 5-32
Scenarios of Future Emissions
(Tg/year)
Animal
Type
Beef Cattle
Dairy
Cattle
Others0
Totald
Range6
1990
Emissions
4.0
1.5
0.3
5.8
4.6-6.9
2000 Emissions
Low8
4.2
1.7
0.3
6.2
5.0-7.4
Highb
4.6
1.7
0.3
6.6
5.3-7.9
2010 Emissions
Low*
3.9
1.8
0.3
6.0
4.8-7.2
Highb
4.5
2.0
0.3
6.8
5.4-8.2
a Low scenario based on estimates in Exhibit 5-29.
b High scenario based on estimates in Exhibit 5-30.
c Emissions from pigs assumed to change with pork production.
Emissions from sheep and goats assumed to change with beef
production. Emissions from horses assumed to remain constant.
d Total may not add due to rounding.
e A range of ±20 percent is used.
found to be effective in increasing milk production in dairy cows by about 10 to
25 percent per lactation, promoting feed efficiency in fed steers, and
repartitioning growth to lean tissues. To date, bST has been approved for
commercial use in 8 countries, and is undergoing review in the U.S.
Anabolic Steroids - Steroid implants are a proven commercialized technique
for promoting feed efficiency and repartitioning growth to lean tissues in beef
production. Implants are used throughout the U.S. beef industry, although use
could be increased in some segments.
Milk Marketing - Eliminating surplus milk production will reduce methane
emissions. Additionally, changing the pricing systems to reduce incentives for
surplus fat production would potentially lead to modifications in feeding
practices that would, as a side benefit, reduce methane emissions per amount
of milk produced.
Beef Marketing: U.S. - Re-orienting the beef marketing system to reduce the
amount of trimmable fat produced will reduce methane emissions. Significant
efforts are under way by the beef industry to emphasize "value-based
marketing" which will have this effect.
5-42
-------
In addition to these near term reduction strategies, several very long term options may
become available as the result of ongoing research, including:
Transgenic Manipulation - Over the very long term it will be possible to transfer
desirable genetic traits among species. This technique holds great promise for
improving the efficiency of production among large ruminants.
• Twinning - Techniques are under development to promote the production of
healthy twins from cattle (e.g., inhibin vaccines). When combined with
adequate nutrition for the mother and offspring, twinning can substantially
reduce the number of mother cows required to produce calves, thereby
substantially reducing methane emissions per product produced.
• Bioengineering of Rumen Microbes - Efforts are under way to develop rumen
microbes that can utilize feed more efficiently, thereby enhancing animal
performance and reducing methane emissions. This option is considered very
long term in nature.
5.6 LIMITATIONS OF THE ANALYSIS
The large diversity in animal management conditions used in the U.S. limits the ability
of this analysis to estimate methane emissions from livestock. In particular, there are a large
number of small producers involved in the beef industry that use a variety of feeding
strategies. The adequacy of this wide range of diets strategies has not been well studied,
and there may be undetected sub-clinical mineral or other deficiencies that would increase
methane production to levels above those estimated in this study.
Another limitation of the analysis is that to date there have been no direct
measurements of methane production by grazing animals because whole animal calorimetry
has been the principal method of measuring methane in the past. While these measurements
are lacking, it appears that calorimetry measurements of stall-fed animals apply to grazing
animals because the energetic relationships developed from stall-fed experiments predict
growth and lactation of grazing animals well. Consequently, there is confidence that the stall-
fed measurements are adequate for making these estimates. Nevertheless, efforts currently
under way to measure methane from grazing animals using a non-intrusive technique may not
only improve these estimates of methane emissions but may also provide the first direct
measurements of digestion characteristics of grazing animals.
5.7 REFERENCES
AMI (American Meat Institute). 1991. Meat Facts 1991. Washington, D.C.
Baldwin, R.L, and M.J. Allison. 1983. Rumen metabolism. Journal of Animal Science
57:461-477.
Baldwin, R.L, J. France, D.E. Beever, M. Gill, and J.H.M. Thornley. 1987b. Metabolism of the
lactating cow. III. Properties of mechanistic models suitable for evaluation of energetic
relationships and factors involved in the partition of nutrients. Journal of Dairy
Research 54:133-145.
5-43
-------
Baldwin, R.L, J.H.M. Thornley, and D.E. Beever. 1987a. Metabolism of the lactating cow. II.
Digestive elements of a mechanistic model. Journal of Dairy Research 54:107-131.
Blaxter, K.L., and J.L Clapperton. 1965. Prediction of the amount of methane produced by
ruminants. British Journal of Nutrition 19:511 -522.
Byers, P.M. 1990. Beef Production and the Greenhouse Effect. The Role of Methane from
Beef Production in Global Warming. Texas A&M University College of Agriculture and
Life Sciences Department of Animal Science. College Station, Texas.
CF Resources. 1991. 1991 Cattle Industry Reference Guide. Englewood, Colorado.
Colovos, N.F. et al. 1949. The nutritive value of timothy hay at different stages of maturity as
compared with second cutting clover hay. Journal of Dairy Science 32:659-664.
Coppock, C.E. et al. 1964. Effect of hay to grain ratio on utilization of metabolizable energy
for milk production by dairy cows. Journal of Dairy Science 47:1330-1338.
Crutzen, P.J., I. Aselmann and W. Seller. 1986. Methane Production by Domestic Animals,
Wild Ruminants, Other Herbivorous Fauna, and Humans. Tellus 386:271 -284.
Dimarco, O.N. and R.L. Baldwin. 1989. Implementation and Evaluation of a Steer Growth
Model. Agricultural Systems 29:247-265.
Ensminger, M.E. 1983. Animal Science. The Interstate Printers and Publishers. Danville,
Illinois.
FAPRI (Food and Agricultural Policy Research Institute). 1991, Summary of FAPRI Baseline.
November 1991.
FAO (Food and Agriculture Organization). 1991. 1990 FAO Production Yearbook. United
Nations. Volume 44. FAO, Rome.
Flatt. W.P. et al. 1967. Energy utilization by high producing dairy cows. II. Summary ol
energy balance experiments with lactating holstein cows. European Association of
Animal Production 12:235-251.
Johnson, D.E. 1992. Personal communication. Colorado State University, Fort Collins,
Colorado.
Johnson, D.E., M. Branine. G.M. Ward, B. Carmean, and D. Lodman. 1991. Livestock
Methane Emissions: Variation, Comparative Warming Perspectives and Amelioration
Potential. Proceed/ngs of the Southwest Nutrition and Management Conference.
University of Arizona College of Agriculture Department of Animal Sciences. Tempe,
Arizona.
Kleiber, M. 1961. The Fire of Life. An Introduction to Animal Energetics. Robert E. Kleiber
Publishing Company, Malabar, Florida.
5-44
-------
Mertens, D.R. 1985. Factors influencing feed intake in lactating cows: From theory to
application using neutral detergent fiber. Proceedings of the Georgia Nutrition
Conference. Atlanta, Georgia, pp. 1-20.
Moe, P.W., and H.F. Tyrrell. 1972a. Metabolizable energy requirements of pregnant dairy
cows. Journal of Dairy Science 55:480-483.
Moe, P.W., and H.F. Tyrrell. 1972b. Net energy value for lactation of high- and low-protein
diets containing corn silage. Journal of Dairy Science 55:318-324.
Moe, P.W., and H.F. Tyrrell. 1977. Effect of feed intake and physical form on energy value of
corn in timothy hay for lactating cows. Journal of Dairy Science 60:752-758.
Moe, P.W., and H.F. Tyrrell. 1979. Methane production in dairy cows. Journal of Dairy
Science 62:1583-1586.
Moe, P.W., H.F. Tyrrell, and N.W. Hooven. 1973a. Energy balance measurements with corn
meal and ground oats for lactating cows. Journal of Dairy Science 56:1149-1153.
Moe, P.W., H.F. Tyrrell, and N.W. Hooven. 1973b. Physical form and energy value of corn
grain. Journal of Dairy Science 56:1298-1304.
NRC (National Research Council). 1984. Nutrient Requirements of Beef Cattle. National
Academy Press, Washington, D.C.
NRC (National Research Council). 1989. Nutrient Requirements of Dairy Cattle. National
Academy Press, Washington, D.C.
Schoeff, R.W. and D.J. Castaldo. 1991. Market Data 1990. Feed Management 42:10-33.
Tyrrell, H.F. and P.W. Moe. 1972, Net energy for lactation of a high and low concentrate
ration containing corn silage. Journal of Dairy Science 55:1106-1112.
USDA (U.S. Department of Agriculture). 1990. Agricultural Statistics 1990. U.S. Government
Printing Office, Washington, D.C.
USDA (U.S. Department of Agriculture). 1992a. Cattle an Feed. National Agricultural
Statistics Service, USDA, Washington, D.C. April 1992.
USDA (U.S. Department of Agriculture). 1992b. Cattle. National Agricultural Statistics
Service, USDA, Washington, D.C. February 1992.
USEPA (U.S. Environmental Protection Agency). 1989. Policy Options for Stabilizing Global
Climate. Report to Congress. Office of Policy, Planning, and Evaluation. Washington,
D.C.
Wainman, F.W., J.S. Smith and P.J.S. Dewey. 1979. The predicted and observed
metabolizable energy values of mixtures of maize silage and barley fed to cattle.
European Association of Animal Production 26:55-58.
5-45
-------
CHAPTER 6
METHANE EMISSIONS FROM LIVESTOCK MANURE
U.S. Methane Emissions
from All Sources
Domest ic Li ve<;tocl
Nfeitura I Gas,-"'
Systems
LandfI I is
Annual Livestock Manure
Methane Emissions
GlobaI Emissions
EMISSIONS SUMMARY
Source
By Manure Management System
Liquid Based
Solid Based
By Animal Type
Dairy Cattle
Swine
Other
Total Emissions (1990)
1 990 Emissions (Tg)
1.4-2.3
0.3-1.3
0.6- 1.0
0.8 - 1.4
0.3- 1.2
1.7-3.6
Partially
Controllable
/
/
/
6.1 EMISSIONS SUMMARY
Methane is produced during the anaerobic decomposition of the organic material in
livestock and poultry manure. Methane emissions from livestock and poultry manure in the
U.S. in 1990 are estimated to be in the range from 1.7 to 3.6 Tg/yr with a central estimate of
2.3 Tg/yr, or about 10 percent of total U.S. methane emissions. U.S. livestock manure
emissions are about 10 percent of the 20 to 30 Tg/yr estimated global annual emissions from
livestock manure.
Of the total 1990 U.S. emissions of 2.3 Tg/yr, two animal groups account for about
1.85 Tg/yr or about 80 percent of total emissions:
• Swine account for about 1.12 Tg/yr or about 50 percent; and
6-1
-------
Dairy cattle account for about 0.73 Tg/yr or about 30 percent.
Liquid based systems (anaerobic lagoons, liquid/slurry storage and pit storage) handle
about 10% of total manure and account for 1.9 Tg/yr, or about 80 percent of total emissions.
Solid based systems (pasture/range, drylots, solid storage, daily spread) handle about 90
percent of the manure (Safley et al., 1992) and account for 0.4 Tg/yr, or about 20 percent of
total emissions. Although solid based systems handle most of the manure, methane
production is small because the methane producing potential of solid based systems is low.
Methane emissions from livestock
and poultry manure could increase
significantly during the next decade. As the
demand for animal products increase, the
number of animals and the amount of
manure produced will also increase. In
addition, the use of livestock manure
management systems that promote methane
production (e.g., anaerobic lagoons) will
likely increase substantially because of
concerns over applying manure during times when crops cannot utilize the nutrient value of
the manure. Emission estimates for 2000 range between 1.9 and 5.7 Tg/yr, and emission
estimates for 2010 range between 1.9 and 6.0 Tg/yr.
Swine and dairy cattle account for the
majority of methane emissions from this
source. Emissions could increase
significantly in the next 20 years as the
use of liquid-based manure manage-
ment systems increases.
6.2 BACKGROUND
Manure decomposition is a process in which microorganisms derive energy and
material for cellular growth by metabolizing organic material in the manure. When
decomposition occurs without oxygen present (anaerobically), methane is an end-product of
the process. This section will describe the fundamentals of anaerobic decomposition; the
methane producing capacity of livestock manure; and the factors that influence methane
production from livestock manure.
6.2.1 The Fundamentals of Anaerobic Decomposition
Livestock manure is primarily composed of organic material and water. Under
anaerobic conditions, the organic material is decomposed by anaerobic and facultative (living
in the presence or absence of oxygen) bacteria. The end products of anaerobic
decomposition are methane, carbon dioxide, and stabilized organic material.
The anaerobic decomposition process can be represented in three stages: hydrolytic;
acid forming; and methanogenic. Carbohydrates decomposition can be illustrated as
follows:1
1 This discussion focuses on the decomposition of carbohydrates because carbohydrate decomposition
accounts for the majority of the methane produced from livestock manure and because the process of methane
production from the decomposition of carbohydrates is best understood. By weight, the volatile solids portion of
cattle and swine manure is approximately 40 percent carbo-hydrate, 15 to 20 percent protein, and up to 10 to 20
percent fat with the remainder composed of other material (Hrubant, Rhodes, and Sloneker, 1978).
6-2
-------
Stage 1: Hvdrolytic. In the first stage, complex organic materials in the manure
substrate are broken down through the hydrolytic action of enzymes. Enzymes
are proteins formed by living cells that act as catalysts in metabolic reactions.
The amount and rate of breakdown can vary substantially and depend on the
enzymes present, the characteristics of the manure, and environmental factors
such as pH and temperature.
Stage 2: Acid Forming. Anaerobic and facultative bacteria reduce (ferment) the
simple sugars produced in Stage 1 to simple organic acids. Acetic acid is the
primary product of the breakdown of carbohydrates, though other organic
acids such as propionic acid and butyric acid can be formed. In addition,
metabolic hydrogen and carbon dioxide are produced. With acetic acid as an
end product, the breakdown of a simple sugar molecule (glucose) in Stage 2
can be represented as:
C6H12°6 + 2H2° —> 2CH3COOH + 2CO2 + 4H2
glucose + water acetic acid carbon dioxide metabolic
hydrogen
Stage 3: Methanogenic. Methane producing bacteria (methanogens) convert
the simple organic acids, metabolic hydrogen, and carbon dioxide from Stage
2 into methane and carbon dioxide. Methanogens are strict anaerobes and
cannot tolerate the presence of molecular oxygen. Methanogens multiply
slowly and are very sensitive to temperature, pH, and substrate composition.
With acetic acid, metabolic hydrogen and carbon dioxide as substrate, the
reactions producing methane can be expressed as:
2CH3COOH —> 2CH4 + 2CO2
acetic acid —> methane + carbon dioxide
4H2 + C02 —> CH4 + 2Hf
metabolic + carbon dioxide —> methane + water
hydrogen
6.2.2 Methane Producing Capacity of Livestock Manure
In general, livestock manure is highly conducive to methane generation due to its high
organic content and the presence of useful bacteria. However, the specific methane
producing capacity of livestock manure depends on the specific composition of the manure
which in turn depends on the composition and digestibility of the animal diet. The greater the
energy content and digestibility of the feed, the greater the methane producing capacity of
the resulting manure. For example, feedlot cattle eating a high energy grain diet produce a
highly biodegradable manure with a high methane producing capacity. Range cattle eating a
low energy forage diet produce a less biodegradable manure with only half the methane
producing capacity of feedlot cattle manure.
In principal, the ultimate methane producing capacity of a quantity of manure can be
predicted from the gross elemental composition of the manure. In practice, however,
insufficient information exists to implement this approach and the methane producing capacity
is determined through direct laboratory measurement. The methane producing capacity of
6-3
-------
livestock manure is generally expressed in terms of the quantity of methane that can be
produced per kilogram of volatile solids (VS) in the manure.2 This quantity is commonly
referred to as B0 with units of cubic meters of methane (CH^ per kilogram VS (m3 CH4 / kg
VS). Representative B0 values for a number of livestock manure types are discussed in
section 6.3 of this chapter.
6.2.3 Factors Influencing Methane Production
While a particular quantity of manure may have a certain potential to produce methane
based on its volatile solids content, the management of the livestock manure and the
environment in which the manure is managed are the major factors influencing the amount of
methane actually produced during manure decomposition.
The characteristics of the manure
management systems and environmental
conditions can be expressed in a methane
conversion factor (MCF) which represents
the extent to which the potential for emitting
methane is actually realized. Manure
systems and climate conditions that
promote methane production will have an
MCF near 1 and manure systems and
climate conditions that do not promote
methane production will have an MCF near 0.
MCF are:
The methane producing potential of
manure has been measured extensively
in the laboratory. The portion of this
potential that is realized is controlled by
the manner in which the manure is
managed.
The primary characteristics determining the
Livestock Manure Management System Factors
• Contact with Oxygen. Under aerobic conditions where oxygen is in contact
with the manure, there is no potential for methane production.
• Water Content. Liquid based systems promote an oxygen-free environment
and anaerobic decomposition. In addition, water is required for bacterial cell
production and metabolism and acts as a buffer to stabilize pH. Moist
conditions increase the potential for methane production.
pH. Methane producing bacteria are sensitive to changes in pH. The optimal
pH is near 7.0 but methane can be produced in a range between 6.6 and 8.0.
• Nutrients. Bacterial growth depends on the availability of nutrients such as
nitrogen, phosphorus, and sulfur. Deficiency in one or more of these nutrients
will inhibit bacterial growth and methane formation. Animal diets typically
contain sufficient nutrients to sustain bacterial growth. Therefore, nutrient
availability is not a limiting factor in methane production under most
circumstances.
2 Volatile solids (VS) are defined as the organic fraction of the total solids (TS) in manure that will oxidize and
be driven off as gas at a temperature of 600°C. Total solids (TS) are defined as the material that remains after
evaporation of water at a temperature between 103° and 105°C.
6-4
-------
Climate Factors
Temperature. Methanogenesis in livestock manure has been observed
between 4° C and 75° C. Temperature is one of the major factors affecting the
growth of the bacteria responsible for methane formation (Chawla, 1986). The
rate of methane production generally increases with rising temperature.
• Moisture. For non-liquid based manure systems, the moisture content of the
manure is determined by rainfall and humidity. The moisture content of the
manure will determine the rate of bacterial growth and decomposition. Moist
conditions promote methane production.
These factors can be combined into the following expression for estimating realized
methane emissions from livestock manure:
Realized Emissions = BQ • MCF (6.1)
where B0 = the maximum methane producing capacity of the manure
determined by animal type and diet (m3 CH4 / kg VS).
MCF = Methane Conversion Factor (MCF) that represents the extent to
which the B0 is realized for a given livestock manure
management system and environmental conditions. Note: 0 <
MCF * 1.
6.3 METHODOLOGY
Methane emissions from livestock manure depend on the type of manure, the
characteristics of the manure management system, and the climatic conditions in which the
manure decomposes. While limited data are available on which to base emission estimates,
a study recently prepared for the USEPA provides an adequate basis for making initial
estimates (Safley et al., 1992). Additional analysis is ongoing to provide additional data for
estimating these emissions.
Based on the Safley et al. (1992) approach, emission estimates were developed by:
identifying the manure management systems in use in the United States and
their methane producing potential;
• estimating the amount and type of manure managed by each system; and
• estimating emissions by multiplying the amount of manure managed in each
system by the estimated emission rate per unit of manure in the system.
Information was obtained from a variety of sources, including:
the U.S. Census of Agriculture;
USDA agriculture statistics;
6-5
-------
• livestock manure management experts throughout the U.S.; and
• scientific literature.
Total emissions will equal the quantity of volatile solids managed in each system times
emissions per kilogram of volatile solids (VS) for that system. Safley et al. (1992) used the
following procedure to estimate total emissions:
• Collect data on: (1) the populations of the major animal types in each state of
the U.S. (A/); and (2) their typical animal mass (TAM).
• Collect information on the characteristics of the manure produced by each of
the animal populations in each state, including: (1) the amount of volatile
solids (VS) produced; and (2) the methane producing capacity (B0) of the
manure. The amount of volatile solids produced depends on the number of
animals in the state and their mass:
VS,, = Nlk • TAM, • vs, (6.2)
where:
A/I k = number of animals of type / in state k.
T/\Mj = typical animal mass in kilograms of animal /'; and
vSj = the average annual volatile solids production per unit of
animal mass (kilograms per kilogram) for animal /.
Identify the livestock manure management systems used in each state and the
percentage of manure managed by each (WS%).
• Estimate the methane producing potential (MCF) of each manure management
system in each state based on the average monthly temperature in the state.
• Estimate methane emissions for each animal and manure system in each state
(TM) by multiplying the amount of volatile solids (VS) produced by the methane
producing capacity of the manure (B0) times the methane producing potential
(MCF) of the manure system in each state.
TMijk = VSlk - Boi • MCFjk • WS%ijk (6.3)
where:
VSj k = total volatile solids produced (kg/yr) for animal; in
state k;
B01 = maximum methane producing capacity per
kilogram of VS for animal /;
MCF• k = methane conversion factor for each manure
system / in state k;
WS%j j k = percent of animal /'s manure managed in manure
system / in state k.
Estimate total annual methane emissions (TM) for animal / as the sum of
annual emissions over all applicable manure management systems / and
states k:
6-6
-------
/ K
Estimate total annual methane emissions from all animals (TM) as the sum over
all animal types / as follows:
TM = ^2 ™i (6.5)
These equations show that methane emissions are driven by four main factors: the
quantity of VS produced; the B0 values for the manure; the MCFs for the manure
management systems; and the portion of the manure handled by each manure management
system (WS%). The following sections describe the data collected to implement this method.
Volatile Solids Production (VS)
Methane emissions from livestock manure are directly related to the amount of volatile
solids (VS) produced. The data required to estimate total VS production are the number of
animals (N-), average size (TAM-), and average VS production per unit of animal size (vSj).
In the U.S., considerable data are available to allow the populations of animals to be
analyzed by: species, production system, and (for cattle) age. Six main categories of
animals were defined: feedlot beef cattle;3 other beef cattle; dairy cattle; swine; poultry; and
other. These main categories were further divided into 20 subcategories. For each
subcategory, VS production was estimated using data on: the animal population; the typical
animal mass (TAM); and the VS production per unit of animal mass. Exhibit 6-1 lists the data
obtained for the 20 subcategories. The cattle populations and weights are equal to those
used in Chapter 5.4
Maximum Methane Producing Capacity (B0)
The maximum amount of methane that can be produced per kilogram of VS (B0)
varies by animal type and diet. Measured B0 values for beef manure range from 0.17 cubic
meters of methane per kilogram of VS (m3/kg-VS) for a corn silage diet to 0.33 m3/kg-VS for a
corn-based high energy diet that is typical of feedlots. Exhibit 6-2 summarizes these values.
Appropriate B0 values were selected depending on the typical diet of each animal type
and category. For animal types without B0 measurements, the B0 was estimated based on
similarities with other animals and the authors' experience. Ruminants for which there were
no literature values were assumed generally to have the same values as cattle, except for
sheep, which were assumed to have B0 values 10 percent higher than cattle (Jain et al.
1981). Exhibit 6-3 lists the values selected for the analysis.
3 Feedlot cattle are animals fed a ration of grain, silage, hay and protein supplements for the slaughter market
(ASB, 1991).
4 Chapter 5 reports cattle weights on an empty body weight basis. These values were converted to live weight
for purposes of making estimates in this chapter.
6-7
-------
Exhibit 6-1
U.S. Animal Populations, Average Size, and VS Production
Animal Type
Feedlot Beef Cattle
Other Beef Cattle
Dairy Cattle
Swine
Poultry0
Other
Steers/Heifers
Calves
Heifers
Steers
Cows
Bulls
Total
Heifers
Cows
Total
Market
Breeding
Total
Layers
Broilers
Ducks
Turkeys
Sheep
Goats
Donkeys
Horses and Mules
Population A'B
Ni
10,088,000
36,040,000
5,535,0000
2,162,000
33,478,000
2,200,000
79,205,000
4,205,000
10,130,000
14,335,000
48,259,000
7,040,000
55,299,000
355,469,000
951,914,000
7,000,000
53,783,000
10,639,000
2,396,000
4,000
2,405,000
Typical
Animal
Mass
0~AM|)C
Kg
415
180
360
360
500
720
410
610
46
181
1.6
0.7
1.4
3.4
70
64
300
450
Manure per day0
(kg/day per 1000 kg mass)
Total
Manure
58
58
58
58
58
58
86
86
84
84
64
85
107
47
40
41
51
51
Volatile
Solids
vsi
7.2
7.2
7.2
7.2
7.2
7.2
10
10
8.5
8.5
12
17
18.5
9.1
9.2
9.5
10
10
A Population data for swine, poultry, and sheep from ASB (1989a-f). Goat and horse population data from Bureau
of Census (1987). Population data for cattle from Chapter 5. Population data as of January 1, 1988 for poultry,
and sheep and as of December 1, 1987 for swine, goats, and horses. Cattle populations represent an average
for 1990.
B Broiler/turkeypopulations estimated yearly based on number of flocks per year (North 1978; Carter 1989).
c Source: Taiganides and Stroshine (1971).
D Source: ASAE (1988).
6-8
-------
Animal
Type
Beef
Beef
Beef
Beef
Beef
Dairy
Dairy
Dairy
Dairy
Horse
Poultry
Poultry
Poultry
Poultry
Swine
Swine
Swine
Swine
Swine
Swine
Swine
Swine
Exhibit 6-2
Maximum Methane Producing Capacity
for U.S. Livestock Manure
B_
Diet (m3CH4u/kg-VS)
7% corn silage, 87.6% corn
Corn-based high energy
91 .5% corn silage, 0% corn
58-68% silage
72% roughage
Roughage, poor quality
Grain-based ration
Barley-based ration
Corn-based high energy
Corn-based high energy
Corn-based high energy
Corn-based high energy
Corn-based high energy
Corn-based high energy
0.29
0.33
0.17
0.23
0.33
0.24
0.17
0.14
0.10
0.33
0.39
0.36
0.24
0.24
0.36
0.48
0.32
0.52
0.48
0.47
0.44
0.45
Reference
Hashimoto et al. (1981)
Hashimoto et al. (1981)
Hashimoto et al. (1981)
Hill (1984)
Chen, et al. (1980)
Morris (1976)
Bryant et al. (1976)
Hill (1984)
Chen, etal. (1988)
Ghosh (1984)
Hill (1982)
Hill (1984)
Webb&Hawkes (1985)
Hawkes & Young (1 980)
Summers & Bousfield (1980)
Hashimoto (1 984)
Hill (1984)
Kroeker et al. (1984)
Stevens & Schulte (1 979)
Chen (1983)
lannotti et al. (1979)
Fischer et al. (1975)
Exhibit 6-3
Cattle:
Swine:
Poultry
Sheep:
Goats:
Horses
Maximum Methane Producing
Animal Type, Category
Beef in Feedlots
Beef Not in
Feedlots
Dairy
Breeder
Market
Layers
Broilers
Turkeys
In Feedlots
Not in Feedlots
and Mules:
Capacity Adopted For
Maximum Potential
Emissions (B^
0.33
0.17
0.24
0.36
0.47
0.34
0.30
0.30
0.36
0.19
0.17
0.33
U.S. Estimates
Reference
Hashimoto et al. (1981)
Hashimoto et al. (1981)
Morris (1976)
Summers & Bousfield (1 980)
Chen (1983)
Hill (1 982 & 1984)
Safley et al. (1992)
Safley etal. (1992)
Safley etal. (1992)
Safley etal. (1992)
Safley etal. (1992)
Ghosh (1984)
6-9
-------
Manure Management Systems Definitions
A variety of manure management practices are in use throughout the U.S. The
following is a brief description of the major livestock manure management systems in use.
PASTURE/RANGE
DAILY SPREAD
SOLID STORAGE
DRYLOT
DEEP PIT STACKS
LITTER
PADDOCK
Liquid/Slurry
ANAEROBIC LAGOON
Animals that are grazing on pasture are not on any true manure
handling system. The manure from these animals is allowed to lie as is,
and is not managed at all.
With the daily spread system the manure is collected in solid form, with
or without bedding, by some means such as scraping. The collected
manure is stored until applied to fields on a regular basis.
In a solid storage system the solid manure is collected as in the daily
spread system, but this collected manure is stored in bulk for a long
period of time (months) before any disposal.
In dry climates animals may be kept on unpaved feedlots where the
manure is allowed to dry until it is periodically removed. Upon removal
the manure may be spread on fields.
With caged layers the manure may be allowed to collect in solid form in
deep pits (several feet deep) below the cages. The manure in the pits
may only be removed once a year. This manure generally stays dry.
Broilers and young turkeys may be grown on beds of litter such as
shavings, sawdust, or peanut hulls, and the manure/litter pack is
removed periodically between flocks. This manure will not generally be
as dry as with deep pits, but will still be in solid form.
Horses are frequently kept in paddocks where they are confined to a
limited area, but not entirely confined to their stalls. This manure will be
essentially the same as manure on pasture or drylot.
These systems are generally characterized by large concrete lined tanks
built into the ground. Manure is stored in the tank for six or more
months until it can be applied to fields. To facilitate handling as a
liquid, water usually must be added to the manure, reducing its total
solids concentration to less than 12 percent. Slurry systems may or
may not require addition of water.
Anaerobic lagoon systems are generally characterized by automated
flush systems that use water to transport the manure to treatment
lagoons that are usually greater than six feet deep. The manure resides
in the lagoon for periods ranging from 30 days to over 200 days
depending on the lagoon design and other local conditions. The water
from the lagoon is often recycled as flush water. Periodically the lagoon
water may be used for irrigation on fields with the treated manure
providing fertilizer value.
6-10
-------
PIT STORAGE Liquid swine manure may be stored in a pit while awaiting final disposal.
The pits are often constructed beneath the swine building. The length
of storage time varies, and for this analysis is divided into two
categories: less than one month or greater than one month.
Methane Conversion Factors (MCFs)
The extent to which the maximum methane producing capacity (B0) is realized for a
given livestock manure management system and environmental conditions is defined as the
Methane Conversion Factor (MCF) for the manure system. For example, a manure system
that produces no methane emissions will have an MCF of 0. A manure system that achieves
full potential methane emissions would have an MCF of 1.
To assess the MCF values for a wide range of livestock manure management systems,
two broad classifications of livestock manure handling systems can be defined based on the
total solids content of the manure:
• Solid systems have a total solids content greater than about 20 percent.
Liquid/slurry systems have a total solids content less than 20 percent.
Manure as excreted may have a total solids content from 9 to 30 percent (Taiganides
1987). This solids content may be modified by adding an absorbent bedding material to
increase the total solids content for easier handling. Alternatively, water may be added to
lower the total solids to allow for liquid transport and handling.
These classifications of systems are particularly important to the potential for methane
production from the manure. Liquid and slurry systems will typically cause anaerobic
conditions to develop, which result in methane production. Solid systems promote conditions
that limit methane production even if anaerobic conditions may exist.
Safley et al. (1992) reviewed the literature to investigate the appropriate range of MCF
values for U.S. manure management systems. Although some data were available, MCF
values were estimated for many systems. To improve the MCF estimates, the U.S.
Environmental Protection Agency is sponsoring analysis to better estimate the MCF for
several key livestock manure systems. Preliminary findings from this analysis indicates that:
The estimated MCF value of dry in situ pasture, range, paddock, and solid
storage manure is 1 to 2 percent. The estimated MCF for drylot manure is 1 to
5 percent. However, the analysis has not yet considered the effect of moisture
or emissions that may result when the manure is washed into streams, rivers,
and lakes or incorporated into the soil (Hashimoto 1992).
The MCF value liquid/slurry and pit storage varies greatly by temperature and
is on the order of 10 percent at 10°C to 65 percent at 30°C (Hashimoto 1992).
The MCF value for daily spread is less than 1 percent (Hashimoto 1992).
The MCF value for anaerobic lagoons is on the order of 90 percent. This
estimate is based on continuous methane measurements taken over a two and
one-half year period at a North Carolina dairy farm (Safley 1991).
6-11
-------
The MCF values for each system are listed in Exhibit 6-4. The MCF for an individual state will
depend on the average monthly temperature and are calculated by:
• estimating the average monthly temperature in each climate division;5
• estimating the MCF value for each month using the average temperature data
and the MCF values listed in Exhibit 6-4;
• estimating the annual MCF by averaging the monthly division estimates; and
• estimating the state-wide MCF by weighting the average MCF for each division
by the fraction of the state's dairy population represented in each division.6
Exhibit 6-5 summarizes the MCF estimates by for each state.
Livestock Manure Management System Usage (WS%)
Methane emissions are driven by four
main factors: the quantity of VS
produced; the B0 values for the
manure; the MCFs for the manure
management systems; and the portion
of the manure handled by each manure
management system.
Livestock manure management
system usage in the United States was
determined by obtaining information from
Extension Service personnel in each state.
The U.S. was divided into eleven geographic
regions based on similarities of climate and
livestock production as shown in Exhibit 6-6.
For states that did not provide information,
the regional average manure system usage
was assumed. Some states did not give
data for all animal types and a regional
average was used in these cases.
Exhibit 6-7 lists the percentage of manure managed by the major systems in the
United States. The important manure management characteristics in the U.S. are:
• Approximately one-third of dairy manure is managed as a liquid and
approximately one-third is spread directly to cropland.
• Seventy-five percent of swine manure is managed as a liquid.
• Poultry manure is primarily managed by deep pit stacking or litter and is
included in "other systems" in Exhibit 6-7.
5 The average temperature in each climate division of each state was calculated for the normal period of 1951
to 1980 using the National Climatic Data Center (NCDC) time-bias corrected Historical Climatological Series
Divisional Data (NCDC 1991).
6 The dairy population in each climate division were estimated using the dairy population in each county
(Bureau of the Census 1987) and detailed county and climate division maps (NCDC 1991). Using the dairy
population as a weighting factor may slightly over or underestimate the MCFs for other livestock populations.
6-12
-------
Exhibit 6-4
Methane Conversion Factors for U.S. Livestock Manure Systems
MCFs based on
laboratory measurement
Pasture, Range, PaddocksA
Liquid/Slur^
Pit Storage < 30 daysA
Pit Storage > 30 daysA
Drylot8
Solid StorageA
Daily SpreadA
MCF measured by
long term field monitoring
MCF at 30°C
2%
65%
33%
65%
5%
2%
1 %
MCF at 20°C
1 .5 %
35%
18%
35%
1 .5%
1.5%
0.5 %
MCFat10°C
1 %
10%
5%
10%
1 %
1 %
0.1 %
Average Annual MCF
Anaerobic Lagoons0 90 %
MCFs estimated by Safley et al.
Average Annual MCF
LitterD 10%
Deep Pit Stacking0 5 %
A Hashimoto (1992)
B Based on Hashimoto (1992).
C Safley et al. (1992) and Safley and Westerman (1992).
D Safley et al. (1992).
6-13
-------
Methane
Exhibit 6-5
Conversion Factors for U.S. Livestock Manure Systems
Pasture,
Range &
State Paddocks Drylot
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Other Systems: Pit Storage for
1.4%
1.4%
1.3%
1.2%
0.9%
0.9%
1.2%
1.5%
1.4%
0.8%
1.1%
1.0%
0.9%
1.1%
1.2%
1.4%
0.8%
1.1%
0.9%
0.8%
0.8%
1.4%
1.1%
0.7%
1.0%
1.2%
0.8%
1.0%
1.2%
0.9%
1.3%
0.7%
1.0%
1 .4%
1.1%
0.9%
1.0%
1.3%
0.8%
1.3%
1 .4%
0.9%
0.8%
1.2%
1.0%
1.2%
0.8%
0.8%
1.9%
1.9%
1 .8%
1.4%
1.0%
1.0%
1 .4%
2.4%
1.8%
0.8%
1.3%
1.2%
1.1%
1.5%
1.5%
2.1%
0.8%
1.2%
1.0%
0.9%
0.8%
1.9%
1.4%
0.8%
1.1%
1.4%
0.8%
1.1%
1.3%
0.9%
1.5%
0.7%
1.1%
1.9%
1.1%
1.0%
1.1%
1.7%
0.9%
1.6%
2.1%
1.0%
0.8%
1.4%
1.0%
1.3%
0.8%
0.8%
Solid Daily
Storage Spread
1.4%
1.4%
1.3%
1.2%
0.9%
0.9%
1 .2%
1.5%
1 .4%
0.8%
1.1%
1.0%
0.9%
1.1%
1.2%
1.4%
0.8%
1.1%
0.9%
0.8%
0.8%
1.4%
1.1%
0.7%
1 .0%
1.2%
0.8%
1 .0%
1.2%
0.9%
1.3%
0.7%
1.0%
1 .4%
1.1%
0.9%
1.0%
1 .3%
0.8%
1.3%
1 .4%
0.9%
0.8%
1.2%
1.0%
1.2%
0.8%
0.8%
0.4%
0.4%
0.4%
0.3%
0.2%
0.2%
0.3%
0.6%
0.4%
0.2%
0.3%
0.3%
0.2%
0.3%
0.3%
0.5%
0.2%
0.3%
0.2%
0.2%
0.2%
0.4%
0.3%
0.2%
0.2%
0.3%
0.2%
0.3%
0.3%
0.2%
0.3%
0.2%
0.2%
0.4%
0.2%
0.2%
0.2%
0.4%
0.2%
0.3%
0.5%
0.2%
0.2%
0.3%
0.2%
0.3%
0.2%
0.2%
less than 30 days is assumed to have an MCF equal to 50% of the
Liquid/Slurry. Pit Storage for more than 30 days is assumed to have an MCF equal to liquid/slurry.
lagoons are assumed to have an MCF of 90%; litter and deep pit stacks an MCF of 10%.
Liquid/
Slurry
29.0%
28.9%
27.6%
21 .9%
18.2%
18.5%
22.6%
38.6%
29.0%
15.5%
22.8%
21 .5%
20.7%
24.7%
23.8%
32.5%
15.5%
21.0%
18.1%
17.0%
18.0%
29.3%
24.1%
15.8%
20.8%
22.1%
16.3%
20.6%
21.3%
18.1%
24.5%
16.8%
20.2%
28.7%
16.2%
18.7%
18.7%
27.3%
19.1%
24.8%
31.7%
17.4%
16.6%
22.5%
15.5%
21 .4%
17.0%
15.9%
MCF for
Anaerobic
-------
Exhibit 6-6
Regions of the U.S. for Manure Management Characterization
North East Connecticut, Maine, Massachusetts, *New Hampshire, New Jersey, *New York,
Pennsylvania, Rhode Island, Vermont.
South East *Delaware, *Florida, *Qeorgia, Maryland, *North Carolina, *South Carolina,
*Virginia, *West Virginia.
Plains *Colorado, *Kansas, *Montana, *Nebraska, *North Dakota, *South Dakota,
Wyoming.
South *Alabama, *Arkansas, Kentucky, *Louisiana, *Mississippi, *Tennessee
South West *New Mexico, *Oklahoma, Texas.
Mid West "Illinois, "Indiana, Michigan, *Ohio, "Wisconsin, *lowa, "Minnesota, "Missouri.
North West "Idaho, "Oregon, "Washington
Far West "Arizona, Nevada, "Utah
Pacific West "California
North Pacific "Alaska
Pacific Islands "Hawaii
States that have supplied estimates of their percent use of manure management.
Exhibit 6-7
Livestock Manure System Usage for the U.S.
Animal
Non-Dairy Cattle
Dairy
Poultry6
Sheep
Swine
Other Animals c
Anaerobic
Lagoons
0%
10%
5%
0%
25%
0%
Liquid/Slurry
and Pit
Storage
1%
23%
4%
0%
50%
0%
Daily
Spread
0%
37%
0%
0%
0%
0%
Solid
Storage
& Drylot
14%
23%
0%
2%
18%
0%
A Includes liquid/slurry storage and pit storage.
B Includes chickens, turkeys, and ducks.
C Includes goats, horses, mules, and donkeys.
Pasture,
Range &
Paddock
84%
0%
1%
88%
0%
92%
Litter,
Deep Pit
Stacks and
Other
1%
7%
90%
10%
6%
8%
Totals may not add due to rounding.
Source: Safleyetal. (1992).
6-15
-------
6.4 CURRENT EMISSIONS
Detailed estimates of methane emissions from the anaerobic decomposition of
livestock manure in the U.S. were calculated using the previously described data on volatile
solids (VS) production, maximum methane producing capacity (B0), manure system
definitions, methane conversion factors (MCFs), and manure system usage (WS%). Because
several animal populations are estimates as of 1987, national emissions estimates for 1990
are based on the change in animal production and U.S. population between 1987 and 1990
for several animal types. In addition, "high" and "low" case emission estimates are presented
to indicate the uncertainty of the point estimates.
6.4.1 Point Estimates
Livestock and poultry manure in the United States emitted 2.3 Tg of methane to the
atmosphere in 1990, or about 10 percent of the world's total emissions of about 25 Tg/yr.
Exhibit 6-8 summarizes the estimated contribution of the major animal groups for 1990, and
shows the growth rates used to convert from 1987 to 1990 emissions for swine, poultry, and
other livestock.7 Of the total 2.3 Tg/yr, two animal groups account for 1.85 Tg or about 80
percent of the total:
Swine manure produced 1.12 Tg/yr or about 50 percent of the U.S. total
emissions; and
Dairy cattle manure produced 0.73 Tg/yr or about 30 percent of the U.S. total
emissions.
These estimates also account for the fact that there are 22 projects currently
recovering methane from dairy, swine, and poultry manure management facilities (ICF, 1992).
A total of 15 digesters are producing fuel from the manure of about 9,000 dairy cows, or
about 0.09 percent of the national dairy cow population. Five digesters are producing fuel
from the manure of about 33,000 hogs, or about 0.06 percent of the national total. Finally, 2
digesters are being used to produce fuel from the manure of about 400,000 head of poultry,
or about 0.03 percent of the national total. Assuming that these projects are replacing the
average manure management systems for these animals, a total of about 0.001 Tg of
emissions are prevented from these 22 systems.
The portions of the U.S. methane emissions from the different livestock manure
management systems are shown in Exhibit 6-9. Of the total emissions:
• Liquid based systems (anaerobic lagoons plus liquid/slurry and pit storage)
account for about 80 percent of total emissions. Because liquid based
systems are often used for confined and energy intensive livestock operations,
they may provide an opportunity for emissions reduction by capturing the
methane for use as an on-farm energy source. The USEPA is currently
7 The conversion from 1987 emissions to 1990 emissions assumes that methane emissions per unit of swine
produced remained constant between 1987 and 1990 and that per capita emissions for other livestock manure
remained constant.
6-16
-------
Exhibit 6-8
Methane Emissions by Animal Type (1 990)
Animal Type
BeefA
Dairy A
SwineA
Poultry8
Other0
Total
Change in
Production
1987 to 1990
(NA)
(NA)
7%
4%
3%
Methane (Tg/yr)
1987
(NA)
(NA)
1.05
0.22
0.02
1990
0.17
0.73
1.12
0.23
0.02
2.28
Notes: 1987 estimates based on Safley et al. (1992) and Hashimoto (1992).
1990 estimates based on 1987 values adjusted for the change in production between 1987 and 1990.
Totals may not add due to rounding.
A Production data based on AMI (1991).
B Poultry includes broilers, layers, turkeys, and ducks. Production data based on USDA (1990).
C Other includes sheep, goats, horses, mules, and donkeys. Production data based on the change in U.S.
population between 1987 and 1990.
NA Not applicable.
Exhibit
Methane Emissions by Manure
System Type
Pasture/Range
Anaerobic Lagoon
Liquid/Slurry and Pit Storage
Drylot
Solid Storage
Daily Spread
Other
Total
Source: Based on Safley et al. (1992) and Hashimoto (1992).
6-9
Management System (1 990)
Emissions
0"g/yr)
0.12
1.42
0.44
0.03
<0.01
<0.01
0.26
2.28
Emissions
(Percent)
5%
62%
19%
1 %
0%
0%
11 %
100%
6-17
-------
assessing the economic and technical feasibility of these opportunities in
several key U.S. states, including: California, Iowa, Illinois, North Carolina, and
Texas.
Solid based systems (pasture/range, drylots, solid storage, daily spread, and
other) account about 20 percent of total emissions. Although most manure is
managed as a solid, solid systems make a small contribution to overall
emissions because of their low methane conversion factors (MCFs),
6.4.2 Range of Estimates
The estimates presented above should be regarded with some caution since some of
the data used to make these estimates are uncertain, in particular:
• The estimated MCF values for pasture, range, drylots, solid storage, and
paddocks are very uncertain. The MCF estimates used in this report are based
on dry manure. This may understate the MCF for regions of the U.S. with
significant rainfall. Because such a large fraction of livestock manure is
managed in these systems, this creates uncertainty in the emissions estimate.
• The methane producing potential of liquid/slurry and pit storage manure
systems may be greater than assumed in this study. Because of the
widespread use of these systems, total emissions may be underestimated.
At this time, insufficient information exists to provide a statistical confidence limit for the
emission estimates presented above. The greatest uncertainty in the emission estimates
results from the methane conversion factor assumptions for the various manure management
systems. While assumptions concerning other factors are somewhat uncertain (i.e., methane
producing capacity of the manure (B0), animal populations and manure quantities, manure
system usage), their contribution to the overall uncertainty is likely to be less than the MCF
estimates.
To capture the uncertainty in these estimates, "high" and "low" case emission
estimates have been defined as follows:
• High Case. The MCF for liquid/slurry, pit storage, litter and deep pit stacking
systems is assumed to be double the base case. The MCF for solid systems
(except litter and deep pits) is assumed to be five times the base case.
• Low Case. The MCFs for each of the major solid systems (pasture/range, solid
storage, and drylots) are assumed to be 80 percent of the base case. The
MCF for liquid/slurry and pit storage is assumed to be 90 percent of the base
case. The MCFs for litter and deep pits are assumed to be half the base case.
The MCF for anaerobic lagoons are estimated using a lagoon methanogenesis
model prepared for USEPA.8
8 The model estimates methane production based on loading rates, lagoon characteristics and climate. The
model estimates are "conservative" because the model focuses on the amount of methane that can be recovered
reliably for use as an energy source.
6-18
-------
Exhibit 6-10 lists the MCF assumptions used to estimate the low and high cases.
For the 1990, the range of emissions implied by these cases is about 1.7 Tg/yr to
3.6 Tg/yr. Exhibit 6-11 presents the High Case and Low Case estimates along with the Base
Case estimates.
Exhibit 6-10
Base, High, and Low Case Emission Estimate Assumptions
Management System
Pasture, Range, Paddock, Drylot, Daily Spread
Liquid/Slurry, Pit Storage
Litter, Deep Pits
Anaerobic Lagoons
MCF
High Case
Five Times
Base Case
Two Times
Base Case
Two Times
Base Case
Same as
Base Case
Low Case
80 percent of
Base Case
90 percent of
Base Case
50 percent of
Base Case
Model Estimates
40 to 1 00 percent
of Base Case
6.4.3 Comparison with Previous Estimates
As discussed above, the emissions estimates in this study are generally based on the
method, data and assumptions presented in Safley et al. (1992). These estimates differ from
Safley et al. in the following respects:
Emissions from dry systems (principally pasture/range and drylot conditions)
were revised downward reflecting the results of measurements by Hashimoto
(1992).
The temperature sensitivity of emissions from liquid/slurry systems was added
based on the results of measurements by Hashimoto (1992), which increased
the estimate of emissions from these systems.
The cattle populations and weights were revised to be consistent with the data
presented in Chapter 5. These modifications reduced the emissions estimate
for dairy and beef cattle by about 0.1 Tg/yr.
Overall, the U.S. emissions estimate presented here is about 1.6 Tg/yr less than the estimate
presented in Safley et al. (1992).
6-19
-------
Exhibit 6-11
Base, High, and Low Case Emission Estimates for
1990 (Tg/Yr)
By Manure Management System
Solid Systems
Pasture/Range
Drylot
Solid Storage
Other Solid Systems A
Total Solid Systems0
Liquid Systems
Liquid/Slurry Storage
Pit Storage
Anaerobic Lagoon
Total Liquid Systems'3
Total0
Base Case
0.12
0.03
<0.01
0.26
0.41
0.21
0.23
1.42
1.87
2.28
High Case
0.59
0.15
0.02
0.54
1.30
0.42
0.46
1.42
2.30
3.60
Low Case
0.09
0.02
<0.01
0.15
0.26
0.19
0.21
1.04
1.44
1.70
Bv Animal Tvoe
Beef
Dairy
Swine
Poultry8
Other0
Total0
Base Case
0.17
0.73
1.12
0.23
0.02
2.28
High Case
0.67
1.04
1.43
0.38
0.09
3.60
Low Case
0.13
0.56
0.85
0.14
0.02
1.70
A Other solid systems include litter, deep pit stacking, and paddocks.
B Includes broilers, layers, turkeys, and ducks.
C Includes sheep, goats, horses, mules, and donkeys.
D Totals may not add due to rounding.
6-20
-------
Lodman et al. (1992) developed estimates independently from the data and
assumptions presented in Safley et al. Exhibit 6-12 summarizes the major components of the
estimates. As shown in the exhibit, Lodman et al. did not estimate emissions from swine and
other livestock, including poultry, horses, sheep and goats. Upon comparing the cattle
estimates, the following is revealed:
• Both studies have about the same cattle populations. This study groups the
dairy calves with the Other Cattle, while Lodman et al. include them with the
dairy population.
This study uses slightly higher values for the rate of VS production per 1000 kg
of animal live weight. As a consequence, this study's estimate of VS
production from cattle is about 5 to 10 percent greater than the estimate by
Lodman et al.
• This study's emissions estimate for feedlot cattle is less than the Lodman et al.
estimate because a lower MCF is used: 1.8 percent versus 5.0 percent. This
study's estimate for Other Cattle is larger than the Lodman et al. estimate
because of a higher MCF: 1.5 percent versus 1.0 percent.
• Differences in the Dairy Cattle estimates arise because this study uses a larger
average MCF (16 percent), as contrasted with 10 percent used by Lodman et
al. The higher MCF for Dairy Cattle manure in this study is driven by the MCF
values for anaerobic lagoons and liquid slurry (which are based on field and
laboratory measurements, respectively). As a contrast, the emissions estimates
for dry manure management systems are similar between the two studies.
6.5 FUTURE EMISSIONS
Future methane emissions from cattle and other livestock manure in the U.S. will be
driven by the future levels of animal production and the manure management systems used.
Trends in animal production were discussed in Section 5.5 of this report. This chapter uses
the same two scenarios for future animal production developed in Chapter 5. In summary,
these scenarios are:
FAPRI Scenario. FAPRI (1991) presents the results of one assessment of
future U.S. production and consumption through 2001. The FAPRI assessment
indicates that beef production will grow slowly; that pork production will
fluctuate resulting in a slight increase in total pork production; that poultry
production will increase dramatically; and that milk production will continue to
grow. Exhibit 6-13 shows these FAPRI results and an extrapolation to the year
2010.
• EPA Scenario. EPA developed estimates of milk and meat production and
consumption in the U.S. as part of its evaluation of policies for stabilizing
global climate (EPA 1989). The EPA estimates indicate that: per capita
consumption of beef may increase from 1990 to 2000 and then decrease from
2000 to 2010; per capita consumption of milk may increase throughout the
period; and milk exports may increase so that the U.S. becomes a significant
net exporter of milk products. Exhibit 6-14 presents the EPA estimates.
6-21
-------
CO
«
c
CO
£
T3
o
_l
£
CO
X =
UJ CO
"Z
Q
£
1
C
I
^
CO
Q.
O
u
I
ffi
y>
c
*co
«
by Lodman
CO
0>
*•*
CO
1
UJ
T3
CD LL
=50
I5
""•»
x £>
o t
>_ "3
>.*:
groo
off
^5
2£
0)
N
co 3i
ro*.
2
"•*.
ci"^
£^
"D
CD UL
P
-r "3
ot
,_ "3
>.^
^0
c^f
55
^£
CD
IM
co "3
0)&
>
<
QCtfT
o o
Q- C-
(Q g)
-^ £
^^
S?
CO
•*
o
d
CM
r^
d
m
5
CO
d
.0
xp
tf~
m
8
d
0
*—
m
CD
d
O)
°
(IS
CO
d
CD
S
o O
"g-5
CD CD
LL m
s?
S)
CO
d
CM
f-.
R
d
N.
s
M-
R
T3
2
r-
&
d
Eg
in
CD
d
CM
CO
u
o
£
;- CD
11
O O
5^
O
CD
*—
CO
h-
d
g>.
CM
8
In
m
CO
M-
^0
5--
O
s
•<•
d
m
CM
. -CD
CM m
1 — CD
0 d
CO
in
u
CO
00
.5-1
S3
S1*-
o
8
CM
^
m
00
d
CO
CO
CO
in
m
Estimated
5
CD
CO
i£
?r
o>
m
CM
d
•*
, -C
^- in
O) 00
d W
at
CO
CO
8
T—
Estimated
B
CD
6
(D '
0
X
CO
d
II
m
CO
CO
•i-
o
•*
X
°9 c
CO =
£i co
CD
f 1 8
T3 CO (0
Q) i—
* . w . . o
I! CO CD CO ^ oj ^
O > ^ > 1C ^i r^
lo8-B^B^^
» c^» g_ jt *
"BlSlaaSs
5,0-5 2 c §>"* «
coSo^'c — — o
^l-l-K-S-^O)^
o»oo» ss?_
^ CD !£ CD "D 0 T
cccgcgo-a,
o«S5co|S15
"g ® SKS § f -S en
coE4SE-scOj>-
t8?S=JeS
S<5|gvB0-5r,-g
cm j^co js o ^
O C C. pii _
~t«t^EE«
•Sj2o^d^o§
^^•^*^rI(^-=-
Q-ei-no'tt- «J
a§i;ssi-
Im°fco0|°°^
i«S«|«S§
s,§l§ls§°
«^3^ COT) S «
5 SS &8 S |S S
•5mo_mo5v-a:
CD JD o°o CDs- cnr:
-------
>,
•o
3
CO
£
Q.
<
u.
c
o
•o
a>
CO
CD
m
.0
4^
Q.
E
3
CO
CO O
v o
» TJ
s: c
A CD
£ C
X O
UJ ^
•o
o
D!
-^
i
•o
(0
(D
a>
.
(O
3
0)
3
u.
^£
ii
^<
o
a.
*t:
^
D_
<
"S
0)
m
c
o
'6
1
o_
c
ro .0
."G ^3
Q. CL
ro c
o i
v_ CO
CD C
Q. o
0
Production
c
« o
*^ *^
'a CL
ro c
o i
»_ CO
CD C
Q. O
0
eduction
a
Per Capita
Consumption
£
o
ti
T3
O
cL
c
ro o
*~* '~
'0.0.
m p
0 §
fc g
Q. 0
O
•%
W
'L
CD
d?
CM
v.O
d^
CM
CM
d?
h~
CO
d?
co
CM
vp
d^
CO
^
xp
d^
O
CD
ss
o
!(-.
U 1
v.O
d?
0
1
O
O
o
CM
O
o
en
O)
d5
CO
in
.0
d?
CM
CM
S?
oo
CM
d?
^t
CM
S?
a>
CM
3?
o
CD'
S?
q
rs^
V
ft
1
"""i"
D
O
O
CM
O
*~"
O
o
o
OJ
-£ c „
c5 oo
a§«S
T5 'S 3 §
«E&i
s§"i
^« L_ U. *w
S 8 -2 -22
0 "S.c
= J5E.2
0^3-=
E. 0 S Pf
c Q. S 1
E _T H 3
D ro o co
co ~ *- c
c Q. re o
O CD Q) O
o o F ^
November 1991.
reductions in per capita
3 1 990 to 2000 rate; per
md the increase in total i
to 20 10. Per capita mil}
CD" ^ re o
c. «-; ** c 2
=5 ^ ro .2 CM
1 ^^ E 1
OQ C CO § 0
s I S § r
§ 3 « | §
£ a c S ?
**«r ^ »u it
O T3 O CD O
>. 0) .= C O
C. CL CL c m
«f~L ~ __ W
_ 5 E ^ is
E CD D pj ro
c S co co
I Ǥn
S «- o p
^ .2 to » t
1 tils
Z^ D ro CD
_ CO •-* "- T3
DC C Q. d) C
o. o ro x: ro
< 00*-..
M 1-1 »_ >- O
"• E oj £ •^>
E re o-t; S
0 — C ^
^ §S §2
<-^ .— O LJ
§ §-«§ -
S ^ oj ro o CD
o 050^^
4»* r\ _. __-„ r-
o !^o 12 o g
S to S-g*: S
O5 CD CM CD «- CM
I E2 g S o
2 °S o %Z
g Q-CM^ S ag
1 S 0 0^ 2 m
.i ^ •= .c £ n "
*T* W W *^ ^ >S +-•
•S 5P, 2 to ro -0 to
m ^ to ro co co ro
^ C CD Q) 0) Q)
Q) .2 •= ^ CO CO CO
o o to c ro ro co
5 £ i2 E £ £ £
O o 0 o o
co < OQ o .£.£.£
to
J5
£
X
UJ
^
if
>.
•a
3
co
<
Q.
UJ
0
•D
CO
CO
CD
m
o
s.
E
3
CO
o
O
^
W
CD
C
0
1
a
;>,
£_^
O
Q.
\°
Pff
Sir?
W
a.
jf
i
•o
c
CD
*•
CD
0)
5
CO
s
= L
3
1 1
IL
"85
0)
m
o
'•5
T3
2
d?
CO
cr
'T~°
Q.
ro c
.« -
O-E
ro 3
O co
CD §
CL 0
c
o
C
D
T3
O
^
a.
c
ro o
.•*; *=
CL Q.
ro c
0 §
>_ CO
CD C
a. o
O
c
o
'•5
D
T3
O
^
CL
C
ro .0
Q. Q.
31
i_ CO
CD C
a. o
O
c
o
o
T3
E
a.
c
ro .0
£ "+3
CL CL
<5 i
i5 c
Q. 0
0
n
.g
a>
a.
d?
in
CO
d*
co
xp
d^
1-
co
S?
m
-sP
cr-
m
CO
sS
O^
in
xp
d^
in
CD
8
0
CM
O
4—1
o
O)
O).
T—
d?
Ti-
co
d?
00
o
d?
oo
CM
d?
Tj-
CM
d?
O
CM
1
d?
CO
in
1
d?
o
«V
d?
CO
in
o
M
D
O
3
M
•5
CD
.O
•fc
(_)
u
u
o
i_
c
o
O
o
•c
c
a
a
C
"ro
c
O
o
Stabilizi
_o
CO
o
S
0
>
o
"o
5T
oo
0)
<
JU
i- fS
E *-:
0 Q
__ ^
o> 2
= c?
CD •—
'•g
^c ro
E •£
"° S
c o
"'•s
i> 3
cu re
f fi
£ -o
« ro
D ro
o d)
S 1
UJ re
. Q.
CD -
O >,
*- O
3 ~
0 0
CO Q.
o> re
CL CD
Ii
eo 5?
ro h.
as d>
ro re
E -
re -Q
w £
a) re
w t
S 3
o g
.E o
O (u
C E
Q._C
•D Q)
m ^
re c
c co
•go
o
ro
o o
CL o ;>
ro £
|i
CO CD
co •£
ro ro
° •« "D
Q) ra C
> co re
OJ Q) . .
"
£2
TJ o c
S|8
•£ | to
. o co {£
s-e-§E
ofeoro
w « ^2
CO CD ^ .-
S« o~
ro ro
-------
The level of production determines how much manure is produced; the systems used
to manage the manure will determine how much methane is produced. Unlike milk and meat
production, forecasts of future livestock manure management practices are not available. For
the purposes of this report, two scenarios of future livestock manure management practices
were developed as follows:
Extend Current Practices. An increase in milk and meat production results in
larger quantities of manure with no change in livestock manure management
practices.
• Increased Use of Liquid Systems. An increase in milk and meat production
and a shift in livestock manure system usage for dairy and swine towards liquid
systems.
Future livestock manure management practices have the greatest effect on projected
emissions. The relatively small changes in livestock manure production only have a small
influence on the forecasts. In addition, a trade-off in meat production between beef and
poultry has only a small effect on emissions because similar manure management systems
are used for both (i.e. dry systems). Changes in dairy and swine production have the
greatest influence on methane emissions because dairy and swine farms generally utilize
liquid based manure management systems.
6.5.1 Extend Current Practices
The Extend Current Practices Scenario assumes that current manure management
practices continue to be used in 2000 and 2010. Changes in methane emissions from
livestock manure then are driven by the changes in the production of milk and meat. This
assumes that manure production per unit of milk and meat produced remains unchanged in
the future.
With these assumptions and using the animal production estimates presented in
Exhibits 6-13 and 6-14 and the range of MCF values presented in Exhibit 6-10, emissions for
2000 will range between 1.9 Tg/yr and 4.1 Tg/yr. Emissions for 2010 will range between 1.9
Tg/yr and 4.4 Tg/yr. The differences in animal production have only a small effect on the
estimates because decreased beef production is offset by increased poultry production.
Exhibit 6-15 summarizes these results.
6.5.2 Increased Use of Liquid Systems
The assumption that manure management practices remain the same, however, may
not be valid. In response to growing concerns over ground and surface water pollution,
many states are requiring farms to control runoff from corrals and other areas where livestock
manure accumulate. In many cases, these requirements will lead to the increased use of
liquid based livestock manure systems such as anaerobic lagoons.
6-24
-------
Projected
*£££ Anlmar
Exhibit 6-1 5
Range of Emissions for 2000 and 2010 (Tg/Yr)
Fype 1990
2000
2010
'•••< »EXJ©f?Q C**dfT©ftt " i^PSIC'tiCIKJ ' """""•••";"" • "' : • • • ' -
FAPRI Scenario
EPA Scenario
Beef
Dairy
Swine
Poultry A
Other8
Total
Beef
Dairy
Swine
PoultryA
Other8
Total
FAPRI Scenario
EPA Scenario
Beef
Dairy
Swine
PoultryA
Other8
Total
Beef
Dairy
Swine
PoultryA
Other8
Total
0.13 - 0.67
0.56- 1.04
0.85-1.43
0.14 - 0.38
0.02 - 0.09
1.70-3.60
0.13 - 0.67
0.56 - 1.04
0.85 - 1.43
0.14 - 0.38
0.02 - 0.09
1.70-3.60
Increased Use of liquid
0.13 - 0.67
0.56-1.04
0.85-1.43
0.14 - 0.38
0.02 - 0.09
1.70-3.60
0.13 - 0.67
0.56- 1.04
0.85- 1.43
0.14 - 0.38
0.02 - 0.09
1.70-3.60
0.14
0.62
0.89
0.19
0.02
1.85
0.15
0.63
0.98
0.16
0.02
1.93
Systems
0.14
1.14
1.67
0.19
0.02
3.14
0.15
1.15
1.84
0.16
0.02
3.32
-0.71
- 1.16
- 1.49
-0.52
-0.09
-3.96
-0.77
-1.18
- 1.64
-0.44
-0.09
-4.12
-0.71
-1.69
-2.38
-0.52
-0.09
-5.39
-0.77
-1.72
-2.62
-0.44
-0.09
-5.65
0.13
0.66
0.86
0.24
0.02
1.90
0.15
0.74
0.96
0.20
0.02
2.06
0.13
1.20
1.62
0.24
0.02
3.20
0.15
1.34
1.80
0.20
0.02
3.51
-0.66
- 1.23
- 1.44
-0.67
-0.09
-4.08
-0.76
- 1.37
- 1.61
-0.57
-0.09
-4.39
-0.66
- 1.79
-2.31
-0.67
-0.09
-5.51
-0.76
-2.00
-2.57
-0.57
-0.09
-5.99
A Includes broilers, layers, turkeys, and ducks.
B Includes goats, horses, mules and donkeys.
Ranges for each cell based on
the range of MCF values listed in Exhibit 6-1 0.
6-25
-------
For example, the Texas Water Commission has instituted regulations to assure zero
discharge of manure or wastewater from concentrated animal feeding operations.9 Because
daiiy operations in Texas often utilize liquid based manure handling systems, it is anticipated
that the primary method for dairies to comply with these regulations will be to utilize
anaerobic lagoon systems. Because liquid systems produce significantly more methane than
solid systems, a shift towards increased use of liquid systems will result in significantly higher
emissions in the future. The shift towards liquid based systems likely will be greatest for
dairies and swine operations that already utilize liquid based systems. Beef, poultry, and!
other livestock operations that do not now utilize liquid based systems are not anticipated to
shift to liquid systems.
For dairy and swine operations under this scenario, the following assumptions are
made for manure management practices by the year 2000:
• Dairy manure that is currently managed in liquid/slurry systems will be
managed in lagoons; dairy manure that is currently managed in lagoons will
continue to be managed in lagoons. All other dairy manure will continue to be
managed with current practices.
• Swine manure that is currently managed in pits will be managed in lagoons;
swine manure that is currently managed in lagoons will continue to be
managed in lagoons. All other swine manure will continue to be managed with
current practices.
These assumptions imply that the fraction of dairy manure managed in lagoons will increase
from 11 percent currently to 32 percent by the year 2000; swine manure managed in lagoons
will increase from 29 percent currently to 73 percent by the year 2000.
With these assumptions and using the animal production estimates presented in
Exhibit 6-13 and 6-14 and the range of MCF values presented in Exhibit 6-10, emissions for
2000 will range between 3.1 Tg/yr and 5.7 Tg/yr. Emissions for 2010 will range between
3.2 Tg/yr and 6.0 Tg/yr. This represents about an 65 to 90 percent increase from current
emissions. Clearly, if liquid based systems are adopted on a wide scale, methane emissions
from livestock manure could increase significantly. Exhibit 6-15 summarizes these results for
the two production and manure system scenarios.
6.5.3 Opportunities for Emission Reductions
Although methane emissions are expected to increase in the coming decades, a
number of opportunities are available to reduce emissions. In many cases, the methane
produced by livestock manure can be collected and used as an on-farm energy source.
Opportunities for recovering methane will be greatest when the manure is managed in a
concentrated form and where temperatures are warm. In particular, the USEPA is currently
identifying profitable options for recovering methane from large dairy farms in Texas and
California and from large hog operations in North Carolina. Examples of these options
include:
9 Concentrated feeding operations include operations with more than 200 mature dairy cattle (whether milked
or dry); 1,000 feed and/or slaughter cattle; or 1,000 swine.
6-26
-------
Covered lagoons. Lagoons are commonly used to store and treat livestock
manure. The manure decomposes anaerobically in the lagoon and produces
methane. By placing a floating cover over the lagoon, methane gas can be
collected and utilized as an energy source. Several successful covered
lagoons are operating throughout the U.S. Lagoon recovery systems are most
applicable in warm climates and where livestock manure are managed as a
liquid (e.g., large dairy and hog farms).
Although methane emissions are
expected to increase in the coming
decades, a number of opportunities are
available to reduce emissions. In
many cases, the methane produced by
livestock manure can be collected and
used as an on-farm energy source.
• Plug flow digesters. Plug flow
digesters utilize solid manure
(undiluted with water) to
produce methane. An
expandable cover is placed
over a trough and manure is
added at one end of the
trough daily. Each day's
"plug" of manure slowly
pushes the mass of manure
down the trough. The
manure in the trough decomposes anaerobically and produces methane which
is collected and utilized. The decomposed, or stabilized, manure is removed at
the other end of the trough. The amount of methane produced depends on
the quantity of manure and the average retention time in the trough. Plug flow
digesters can be utilized in warm and cold climates.
• Mixed tank digesters. Mixed tank digesters also are commonly used in the
U.S. and world to produce methane gas from livestock manure. Mixed tank
digesters receive a continuous, daily flow of manure. The manure is mixed
periodically inside the tank where it remains for on average 20 to 30 days
before being removed. The manure decomposes anaerobically within the
closed tank and produces methane which is collected. In order to achieve
optimal gas production, hot water may be circulated through the tank to
increase the digester temperature. Mixed tank digesters can be utilized in most
climates.
The methane collected can be utilized in a number of ways, depending on the needs
of the farmer. Energy use options include:
• refrigeration of milk and hot water to wash dairy cows;
• heat for piglets and growing pigs;
• on-farm electricity generation for on-farm use; and
• on-farm electricity generation for sale to utilities.
The amount of methane emission reductions that could be achieved depend on a
number of factors, including: livestock manure management practices; energy prices; and
the cost of the recovery and utilization equipment. A detailed assessment of these
opportunities will be included in a separate report.
6-27
-------
6.6 LIMITATIONS OF THE ANALYSIS
The methane emission estimates presented in this chapter are uncertain for a variety
of reasons, including:
The estimates of the methane produced by pasture, range, solid storage and
drylot manure is uncertain. Assumptions regarding methane emission from
manure in pastures have a large influence on the overall emissions estimate
because a large portion of livestock manure is found in pastures.
• Limited data are available to assess the methane producing potential of
livestock manure systems under the conditions in which it is found throughout
the U.S.
The USEPA is currently undertaking studies to improve the basis for making emission
estimates. As additional data become available, the estimates can be improved.
6.7 REFERENCES
AMI (American Meat Institute). 1991. Meat Facts. American Meat Institute. Washington, D.C.
ASAE (American Society of Agricultural Engineers). 1988. Manure Production and
Characteristics. ASAE Data: ASAE D384.1. American Society of Agricultural Engineers.
St. Joseph, Ml.
ASB (Agriculture Statistics Board). 1989a. Cattle. Released: February 8, 1989. Agricultural
Statistics Board. ERS-NASS, USDA, P.O. Box 1608, Rockville, MD 20850. 15 pp.
ASB (Agriculture Statistics Board). 1989b. Cattle on Feed. Released: January 26, 1989.
Agricultural Statistics Board. ERS-NASS, USDA, P.O. Box 1608, Rockville, MD 20850.
14pp.
ASB (Agriculture Statistics Board). 1989c. Hogs and Pigs. Released: January 6, 1989.
Agricultural Statistics Board. ERS-NASS, USDA, P.O. Box 1608, Rockville, MD 20850.
20pp.
ASB (Agriculture Statistics Board). 1989d. Layers and Egg Production, 1988 Summary.
January, 1989. Agricultural Statistics Board. ERS-NASS, USDA, P.O. Box 1608,
Rockville, MD 20850. 40 pp.
ASB (Agriculture Statistics Board). 1989e. Poultry, Production and Value, 1988 Summary.
April, 1989. Agricultural Statistics Board. ERS-NASS, USDA, P.O. Box 1608, Rockville,
MD 20850.
ASB (Agriculture Statistics Board). 1989f. Sheep and Goats. Released: February 8, 1989.
Agricultural Statistics Board. ERS-NASS, USDA, P.O. Box 1608, Rockville, MD 20850.
11 pp.
6-28
-------
Bryant, M. P., V. H. Varel, R. A. Frobish, and H. R. Isaacson. 1976. 347 pp. In: H. G.
Schlegel (ed.). Seminar on Microbial Energy Conversion. E. Goltz KG. Gottingen,
Germany.
Bureau of Census. 1987. Census of Agriculture. United States Department of Commerce.
U.S. Government Printing Office. Washington, DC 20402.
Carter, T. A. 1989. Personal communication with Dr. Thomas A. Carter. Extension Professor
of Poultry Science. Poultry Science Department. North Carolina State University. Box
7608. Raleigh, NC 27695-7608.
Chawla, O.P. 1986. Advances in Biogas Technology. Indian Council of Agricultural Research:
New Delhi.
Chen, T. H., D. L. Day, and M. P. Steinberg. 1988. Methane production from fresh versus dry
dairy manure. Biological Wastes. 24:297-306.
Chen, Y. R. 1983. Kinetic analysis of anaerobic digestion of pig manure and its implications.
Agricultural Wastes. 8:65-81.
Fischer, J. R., D. M. Seivers, and D. C. Fulhage. 1975. Anaerobic digestion in swine wastes.
pp. 307-316. In: W. J. Jewell (ed.). Energy, Agriculture and Waste Management. Ann
Arbor Science. Ann Arbor, Ml.
Ghosh, S. 1984. Methane production from farm waste, pp. 372-380. In: M. M. EI-Halwagi
(ed.). Biogas Technology, Transfer and Diffusion. Elsevier. New York.
Hashimoto, A. G., V. H. Varel, and Y. R. Chen. 1981. Ultimate methane yield from beef cattle
manure; effect of temperature, ration constituents, antibiotics and manure age.
Agricultural Wastes. 3:241 - 256.
Hashimoto, A. G. 1984. Methane from swine manure: effect of temperature and influent
substrate composition on kinetic parameter (k). Agricultural Wastes. 9:299-308.
Hashimoto, A. G. 1992. Personal communication with Dr. Andrew Hashimoto. Professor and
Department Chairman. Bioresource Engineering Department. Oregon State
University. Corvallis, OR. July 1992.
Hawkes, F. R. and B. V. Young. 1980. Design and operation of laboratory- scale anaerobic
digesters: operating experience with poultry litter. Agricultural Wastes. 2:119-133.
Hawkes, F. R. and B. V. Young, 1980. Design and operation of laboratory-scale anaerobic
digesters: operating experience with poultry litter. Agricultural Wastes. 2:119-133.
Hill, D. T. 1982. Design of digestion systems for maximum methane production.
Transactions of the ASAE. 25(1) :226-230.
Hill, D. T. 1984. Methane productivity of the major animal types. Transact/ens of the ASAE.
27(2):530-540.
6-29
-------
Hrubant, G.R., R.A. Rhodes, and J.H. Sloneker, "Specific Composition of Representative
Feedlot Wastes: A Chemical and Microbial Profile," SEA-NC-59. Northern Regional
Research Center, U.S. Department of Agriculture, Peoria, Illinois, 1978.
lannotti, E. L, J. H. Porter, J. R. Fischer, and D. M. Sievers. 1979. Developments in Industrial
Microbiology. 20(49): 519-520.
ICF Incorporated. 1992. U.S. Anaerobic Farm Digester Study. Prepared for the Global
Change Division of the Office of Air and Radiation, U.S. Environmental Protection
Agency, Washington, D.C.
Jain, M. K., R. Singh, and P. Tauro. 1981. Anaerobic digestion of cattle and sheep waste.
Agricultural Wastes. 3:65-73.
Kroeker, E. J., D. D. Schulte, A. B. Sparling, and J. T. Chieng. 1984. Anaerobic treatment
process stability. Journal of the Water Pollution Control Federation. 51:718-727.
Lodman, D.W., et al. 1992. Estimates of methane emissions from manure of U.S. cattle.
Chemosphere, in press.
Morris, G. R. 1976. Anaerobic fermentation of animal wastes: a kinetic and empirical design
fermentation. M. S. Thesis. Cornell University.
NCDC (National Climatic Data Center) 1991. Historical Climatological Series Divisional Data.
National Oceanic and Atmospheric Administration. Ashville, NC.
North, M. O. 1978. Commercial Chicken Production Manual. AVI. Westport, Connecticut.
Safley, LM. 1991. Personal communication with Dr. Lawson Safley. Professor of Biological
and Agricultural Engineering. North Carolina State University. Raleigh, North Carolina,
January 1991.
Safley, L.M., M.E. Casada, J.W. Woodbury, and K.F. Roos (1992). "Global Methane Emissions
from Livestock and Poultry Manure." EPA/400/1091/048. U.S. Environmental
Protection Agency. Washington, D.C. February 1992.
Safley, L.M., Jr. and P.W. Westerman 1992. "Performance of a Low Temperature Lagoon
Digester." Bioresource Technology. 41:167-175.
Stevens, M. A. and D. D. Schulte. 1979. Low temperature digestion of swine manure.
Journal of the Environmental Engineering Division, ASCE. 105(EE1): 33-42.
Summers, R. and S. Bousfield. 1980. A detailed study of piggery-waste anaerobic digestion.
Agricultural Wastes. 2:61 -78.
Taiganides, E. P. 1987. Animal waste management and wastewater treatment, pp. 91-153,
In: D. Strauch (ed.). Animal Production and Environmental Health. Elsevier. New
York.
6-30
-------
Taiganides, E. P. and R. L. Stroshine. 1971. Impacts of farm animal production and
processing on the total environment, pp. 95-98. In: Livestock Waste Management and
Pollution Abatement. The Proceedings of the International Symposium on Livestock
Wastes, April 19-22, 1971, Columbus, Ohio. ASAE. St. Joseph, Ml.
USDA (United States Department of Agriculture). 1990. Agricultural Statistics 1990. U.S.
Department of Agriculture. Washington, DC.
Webb, A. R. and F. R. Hawkes. 1985. Laboratory scale anaerobic digestion of poultry litter:
gas yield-loading rate relationships. Agricultural Wastes. 13:31-49.
6-31
-------
CHAPTER 7
METHANE EMISSIONS FROM OTHER SOURCES
U.S. Methane Emissions
from All Sources
Oomestic
Natural Gas
Systems
Landf ills
Coa I M i n i ng
Live^.tocl Manure
Othei SOUP ces
Annual Methane Emissions
From Other Sources
GI oboI Emiss1on1*
Emissions Summary
Source
Rice Cultivation
Fuel Combustion
Stationary Combustion
Mobile Combustion
Oil Systems
Other Sources
Non-fuel Biomass Burning
Industrial Processes and Wastes
Land Use Changes
Total
Emissions
(Tg/yr)
0.1 - 0.7
0.3- 1.4
0.2 - 0.4
0.1 - 0.6
No Estimate
Provided
1.1 -2.51
Partially
Controllable
/
/
•
1 The uncertainty in the total is estimated assuming that the uncertainty for
each source is independent. Consequently, the uncertainty range for the
total is more narrow than the sum of the ranges for the individual sources.
7.1 EMISSIONS SUMMARY
There are several other important anthropogenic sources of methane emissions, in
addition to those discussed in previous chapters. These other sources include:
7-1
-------
• rice cultivation;
• fuel combustion;
• production and refining of petroleum liquids;
• non-fuel biomass burning;
• various industrial processes and waste streams; and
• land-use changes.
These additional sources of methane emissions are relatively minor contributors to
overall anthropogenic methane emissions in the United States, though some of the sources,
such as rice cultivation, biomass burning, and wastewater treatment are of great importance
globally. With limited exception, little detailed research has been conducted on these sources
in the U.S.. For rice cultivation, fuel combustion, and production and refining of petroleum
liquids, estimates have been made of methane emissions, however, they are associated with
a large degree of uncertainty. For the remainder of these sources, no estimates have yet
been completed.
The first three sections of this chapter present methane emissions from rice cultivation,
fuel combustion, and production and refining of petroleum liquids. Due to limited information
and their relatively marginal significance as sources in the United States, discussions of
methane emissions from non-fuel biomass burning, industrial processes and waste streams,
and land-use changes are limited to brief descriptions in the final section.
7.2 RICE CULTIVATION
7.2.1 Background
Flooded rice fields generate methane through the anaerobic decomposition of organic
matter in the fields. The methane produced under these anaerobic conditions is released into
the atmosphere primarily via the rice plants themselves during the growing season. The rice
fields that generate significant amounts of methane are irrigated or rainfed, with less than one
meter of floodwater depth. Upland rice fields, which are not flooded, and deep-water, floating
rice fields are not believed to produce significant amounts of methane.
On a global level, flooded rice cultivation could constitute the single largest
anthropogenic source of methane emissions. Estimates of world-wide annual methane
emissions from flooded rice fields range from 20 to 150 Tg with an estimated mid-point of
about 60 Tg (IPCC 1992). These estimates are based on limited experimental data developed
in the U.S., Spain, Italy, India, China, and the Philippines. Importantly, the available estimates
are based on observations in "undisturbed plots," without the normal disturbances that would
occur from walking through the fields for planting, fertilizing, and weeding. These
disturbances could liberate substantially more methane since only about twenty percent of
the methane generated in a rice field is released to the atmosphere under undisturbed
conditions. Emissions from rice are about 10 percent of total world emissions and 20 percent
of emissions related to human activities.
Global emissions are uncertain because the production and emission of methane from
rice fields are the result of complex processes and emissions can vary greatly from field to
field. Based upon research to date, scientists have identified a variety of factors that affect
methane emissions. These factors include:
7-2
-------
soil type;
temperature;
oxidation-reduction (redox) potential;
PH;
• management;
• water management technique; and
• cultivar type.
Further research is needed before the exact relationships between these factors and methane
emissions are determined.
Also contributing to the difficulty of estimating emissions, methane emissions from rice
fields vary substantially diurnally and seasonally. Seasonal variations are related to the
cropping cycle with peak methane emissions during the tillering, reproductive, and flowering
stages of rice cropping. In light of the many factors influencing rice field methane emissions,
the emissions rates can be expected to vary both within and between countries.
7.2.2 Methodology
The report Estimation of Greenhouse Gas Emissions and Sinks (OECD 1991)
recommends a methodology for estimating emissions of methane from rice cultivation. This
methodology estimates emissions based on: rice ecology, growing season length, and a
range of methane emissions rates. Rice ecology refers to the type of rice field (i.e., flooded
rice fields), the area under cultivation, and the number of crops per year. Growing season
length is the duration of each crop cycle. These two factors are used to arrive at a total
number of hectare-days cultivated per year. Ideally, to avoid unrepresentative results based
upon fluctuations in economic or climatic conditions, at least three years of data on rice
cultivation should be used to develop an average estimate.
Due to the large variability in emissions, the use of a range of daily methane emissions
rates per hectare-day cultivated is recommended. Also, use of daily emissions rates (derived
from whole season measurements, rather than seasonal rates) is recommended, to allow for
variability in growing season lengths. Total annual methane emissions are derived by
multiplying the methane emissions rates by the total number of hectare-days cultivated.
The rice ecology and growing season length for the U.S. for years 1989 to 1991 are
presented in Exhibit 7-1. Methane emission rates for rice cultivation in the U.S. were derived
from studies performed in the U.S. by Sass et al. (1990) and Cicerone et at. (1983). The
range of emissions rates is 0.067 to 0.42 g CH4/m2/day.
7.2.3 Current Emissions
The estimated range of methane emitted from flooded rice fields in the United States
in 1990 is about 0.1 to 0.7 teragrams (Tg). This range represents approximately one percent
of the total anthropogenic methane emissions in the United States in 1990 and less than one
percent of the world's estimated 20 to 150 Tg (IPCC 1992) of methane emissions from rice
cultivation.
7-3
-------
Exhibit 7-1
Data Used to Calculate Three-Year Average Hectare-Days
Total Hectares Harvested
Cropping Cycle Length
Number of Cropping Cycles
Total Hectare-Days
Three-Year Average Hectare-Days
1989: 1,087,000
1990: 1,142,000
1991: 1,113,000
1 53 days
1
1989: 166,311,000
1990: 174,726,000
1991: 170,289,000
170,442,000
Source: USDA (1991); Matthews et al. (1991).
7.2.4 Future Emissions
Rice production and land area devoted to production are generally expected to remain
fairly constant in the U.S. over the next several decades as they have for the last 15 years.
Accordingly, annual methane emissions are expected to remain in the range of 0.1 to 0.7 Tg.
7.2.5 Limitations of the Analysis
Estimation of methane emissions from rice cultivation remains very uncertain.
Research on the processes that produce and emit methane from flooded rice fields is not
complete and the processes involved are not yet fully understood.
7.3 FUEL COMBUSTION
7.3.1 Background
The process of fuel combustion is a recognized source of anthropogenic methane.
Methane emissions related to combustion result, for the most part, from the incomplete
combustion of fuel. Methane may be a component of the fuel that is not totally combusted
and thus emitted into the atmosphere. In cases in which methane is not a component of the
fuel, methane may be created in the combustion process. Venting associated with starting
and stopping gas-fired turbines and evaporation that accompanies energy use may also
result in methane emissions. In general, the methane emissions resulting from fuel
combustion are much less than those associated with fuel production activities, such as coal
mining and natural gas production, processing, transmission, and distribution.
7-4
-------
Fuel consumption activities that result in the emission of methane from fuel
combustion may be divided into two source groups: stationary sources and mobile sources.
Stationary sources include:1
• wood-fired equipment;2
• coal-fired equipment;
• natural gas-fired equipment; and
• oil-fired equipment.
Mobile sources of combustion related to methane contribute a lesser amount of
methane than stationary sources and include:
• highway and off-highway vehicles;
• aircraft;
• railway transportation; and
• agricultural, industrial, and construction machinery.
7.3.2 Methodology
The methodologies for estimating methane emissions from stationary and mobile
sources are presented independently.
Stationary Combustion
Estimation of methane emissions from stationary fuel combustion ideally involves the
use of three types of information (OECD 1991):
• national energy data, by source sector (e.g., energy used in industry sector,
energy used in agriculture sector);
• emissions factors per unit of energy use, including consideration of type of fuel
used, and type and vintage of technology; and
• technology splits for energy data (i.e., type and prevalence of technologies
employed in each energy sector).
Emissions factors are applied to energy use by source sector, as modified by information on
technologies employed in each energy use sector.
1 Methane is also emitted from two other stationary sources: gas-fired pipeline compressors used by the natural
gas industry, and non-fuel biomass burning. Emissions from compressors in the natural gas industry are
discussed in Chapter 2 and methane emissions from non-fuel biomass burning are discussed in section 7.5.
2 The industrial sector accounts for about 65 percent of wood combustion -- primarily the paper, lumber and
wood products industries. Within the industrial sector, waste wood serves as fuel for a variety of wood energy
conversion systems including boilers, cogenerators, kilns, dryers, and gasifiers. The residential sector accounts for
about 35 percent of wood combustion; wood is burned in free-standing stoves, fireplaces, central heating
equipment for houses and large boilers for apartment buildings. The utility sector accounts for less than 1 percent
of wood combustion (EIA 1987).
7-5
-------
Due to lack of estimates on fuel use by source category necessary to develop
emissions factors for each energy source category, a different approach was used for
estimating emissions for oil and coal fired equipment and for wood combustion. This
approach is based on the methodology used in Preliminary Estimates of Greenhouse Gas
Emissions and Sinks for the United States, 1988 (U.S. Government 1991). Methane emissions
from stationary combustion from these sources are estimated based upon total non-methane
volatile organic compound (VOC) emissions by source category, and the ratio of methane to
non-methane VOC emissions specific to each source category. These data are shown in
Exhibit 7-2.
Exhibit 7-2
Quantity of Non-Methane VOCs and
Ratio of Methane to Non-Methane VOCs
Source
Oil Fired Equipment
Coal Fired Equipment
Wood Fired Equipment
(Industrial use)
Wood Fired Equipment
(Residential use)
Non-Methane VOCs
(Tg)
0.012
0.056
0.485
0.261
Ratio of Methane to
Non-methane VOCs
0.05 to 0.10
0.05 to 1.0
0.2
2
Source: Non-methane VOCs are from USEPA (1991); Ratio of methane to
non-methane VOCs for coal and oil fired equipment are from U.S.
Government (1991).
Emissions from wood fired equipment are based on USEPA (1985). For
industrial wood combustion, the mean methane to non-methane VOC ratio
is based on wood combustion in boilers. For residential wood combustion,
the mean ratio is based on residential wood stoves. Wood combustion by
the utility sector is assumed to be less than 1 percent and is not accounted
for here (see footnote 2 of this report).
To estimate methane emissions from natural gas combustion, natural gas
consumption data was disaggregated by end use sector (residential, commercial, industrial,
and utility) and by technology (boiler or non-boiler). Methane emissions factors shown in
USEPA (1985) were then applied to annual consumption to calculate emissions from each
source. Because Chapter 2 of this report already includes estimates for methane emissions
from natural gas production, processing, and transportation, these emissions are omitted from
gas-fired combustion estimates shown in this section. The data used to estimate methane
emissions from natural gas combustion are shown in Exhibit 7-3.
7-6
-------
Exhibit 7-3
Data Used to Estimate Methane Emissions from Natural Gas Consumption
Technology and End-Use Sector
Boilers
Residential
Commercial
Industrial
Utility (94% of all utility)
Plant (50% of plant & lease)
Non-boilers
Utility Recip. Engines
Utility Turbines
Total3
1 990 Consumption
(mmcf)1
4,390,591
2,679,687
6,969,543
2,618,984
617,665
16,717
150,452
17,443,639
Emissions Factor
(Ib/mmcf)2
2.7
2.7
3.0
0.3
3.0
1260.0
37.8
1 Source: Consumption data by end use sector is from EIA (1991b). Information
on technology splits based on personal communication with DOE (5/92).
2 Source: USEPA (1985).
3 Consumption data for pipeline fuel and one-half of industrial plant and lease fuel
are not included in this total, but are addressed in Chapter 2.
Mobile Combustion
Estimates of emissions from mobile combustion are based upon measures of
transportation activity, by vehicle type and vintage, and by available emissions factors for
each vehicle type. The total distances travelled by vehicle and model type were multiplied by
the relevant methane emissions factor to arrive at estimated total methane emissions. Due to
lack of information, the estimates do not include emissions from aircraft in cruise mode and
emissions from alternative motor vehicle fuels. The data used to estimate emissions from
mobile combustion is presented in Exhibit 7-4.
7.3.3 Current Emissions
Total methane emissions from fuel combustion were estimafed to range from 0.5 to 1.7
Tg in 1988. This represents approximately 3 percent of current total U.S. methane emissions
from anthropogenic sources.
Stationary Combustion
Of the estimated 0.5 to 1.7 Tg of total methane emissions from combustion, stationary
sources accounted for 0.3 to 1.4 Tg. Exhibit 7-5 presents the low, high, and mean estimates
of methane emissions by fuel source. Wood combustion contributes the greatest amount of
methane emissions.
7-7
-------
Exhibit 7-4
Data Used to Estimate Emissions from Mobile Sources
Vehicle Type
Activity Level
Emission Factor
Highway Vehicles
Passenger Cars
Light Trucks
Heavy Trucks
Motorcycles
vehicle km traveled
2,438,230,330,000
663,100,000,000
170,600,000,000
15,400,000,000
g/km1
0.047
0.151
0.137
0.240
Other Mobile Sources
Boats
Locomotives
Farm Equipment
Other Off Road
Jet and Turboprop Aircraft
Gasoline (Piston) Aircraft
kg fuel
14,380,000,000
10,350,000,000
12,340,000,000
12,240,000,000
41,710,000,000
975,100,000
g/kg fuel
0.230
0.250
0.450
0.180
0.087
2.640
Source: Emission factors are derived from OECD (1991). Highway Vehicles
were initially broken down by both vehicle class and pollution control
technology, each having a separate emissions coefficient to account for
differences in emission characteristics. The emission factors presented here are
a weighted average of these factors, calculated by dividing total emissions for
each class of highway vehicle by total kilometers traveled. Vehicle miles
traveled are from USDOT (1992).
Mobile Combustion
Mobile sources were estimated to contribute a lesser amount of methane,
approximately 0.3, with a plausible range from 0.15 to 0.4 Tg. Exhibit 7-6 presents the
estimates for different mobile sources. Highway vehicles constitute the largest mobile
combustion source of methane, emitting an estimated 0.25 Tg. Passenger cars accounted for
over half of total highway vehicle methane emissions. Non-highway sources were estimated
to contribute a total of 0.02 Tg.
7.3.4 Future Emissions
Methane emissions from mobile combustion in the U.S. are not expected to increase
significantly over the next few decades. This is due to the use of cleaner burning
7-8
-------
Exhibit 7-5
Annual Methane Emissions from
Stationary Combustion
(Teragrams)
Source
Coal
Fuel Oil
Wood (Industrial)
Wood (Residential)
Natural Gas
Total
Estimated Methane Emissions
Low
0.0028
0.0006
0.0485
0.261 1
0.0158
0.3
High
0.0560
0.0012
0.1940
1.0440
0.0630
1.4
Mean
0.0290
0.0009
0.0970
0.5222
0.0315
0.7
For coal and fuel oil combustion, the mean emissions
estimate is calculated from the average of the low and high
emissions factors shown in Exhibit 7-2.
For wood and natural gas combustion, the mean
estimate is calculated from the emissions factors
Exhibits 7-2 and 7-3. Low and high
emissions
shown in
emissions estimates
are calculated assuming that the uncertainty range for the
estimate is from 1/2 to 2 times the mean emissions
estimate.
technologies and the increased prevalence of sophisticated tail-pipe technologies. The
implementation of these technologies are assumed to offset possible increases in emissions
due to projected population growth over the next two decades.
Stationary combustion methane emissions over the next few decades may increase or
decrease, primarily depending on the projected consumption levels for each fuel source. U.S.
consumption of coal and natural gas is projected to increase over the next twenty years.
Accordingly, methane emissions from these sources are likely to increase in the future. In
contrast, fuel oil consumption, and accordingly methane emissions from this source, is
projected to decrease over the next twenty years (EIA 1991 a).
7.3.5 Limitations of the Analysis
Estimates of methane emissions from energy combustion have a high degree of
uncertainty largely due to lack of emissions data. Estimations for oil and coal-fired and wood
stationary combustion sources are limited by the lack of accurate data on fuel use in each of
the various source categories. Without fuel use data, the estimate is derived from the
7-9
-------
Exhibit 7-6
Annual Methane Emissions from
Source
Highway Vehicles
Passenger cars
Light trucks
Heavy trucks
Motorcycles
SUBTOTAL
Other Mobile Sources
Boats
Locomotives
Farm Equipment
Construction, Industrial, and
Snowmobiles
Jet and Turboprop Aircraft
Gasoline (Piston) Aircraft
SUBTOTAL
Total All Mobile Sources
Mobile Combustion
Estimated Methane
Emissions (Tg)
0.115
0.101
0.027
0.004
0.247
0.0059
0.0026
0.0056
0.0022
0.0036
0.0026
0.023
0.270
Methane emissions from mobile sources are estimated to have an
uncertainty factor of +/- 50 percent, giving a range of about 0.15
to 0.4 Tg.
percentage of methane emissions relative to non-methane VOC emissions. These
percentages have broad ranges that result in very imprecise estimates.
The estimations of methane emissions from mobile combustion sources are more
accurate. This is due to the availability of data on distances travelled by vehicle type, vintage
and fuel used, and associated methane emissions factors. The methane emissions attributed
to aircraft are probably an underestimate, as they include emissions from take off and landing
only. Emissions estimates from aircraft in cruise mode were not available. Though somewhat
more accurate than estimates for stationary sources, the estimates from mobile combustion
are still given with a range of uncertainty of plus or minus 50 percent.
7-10
-------
7.4 PRODUCTION AND REFINING OF PETROLEUM LIQUIDS
7.4.1 Background
Several activities conducted during the production and refining of petroleum products
produce methane emissions. Tilkicioglu and Winters (1989) identified the major sources of
these emissions as:
• fugitive emissions from oil wells and related production field treatment and
separation equipment;
• emissions during the routine maintenance of production field equipment;
• emissions from fixed roof and floating roof crude oil storage tanks;
• emissions from refinery processes; and
emissions from crude oil tanker loading and unloading.
Additionally, venting and flaring of gas during oil and gas production is a source of
methane emissions. In Chapter 2 the main sources of venting from the gas industry were
evaluated, including emissions from pneumatic devices, glycol dehydrators, heaters, and gas
plants. For purposes of this analysis, venting and flaring emissions from wellheads are
included as part of oil system emissions because preliminary analyses indicate that a large
majority of emissions from wellhead venting and flaring originate from oil wells that do not
market gas (Radian 1992). Fuel combustion-related emissions, e.g., emissions from plant and
lease fuel used during oil production or refinery fuel, are included in the stationary
combustion emissions estimate in this chapter.
Activities downstream of oil refineries, such as gasoline storage and distribution are
expected to have negligible methane emissions because refined products contain virtually no
methane. Measurements conducted near such facilities confirm this expectation (Blake 1990).
Consequently, only production, crude oil transportation, and refining activities are included in
this assessment. The methodology and emissions estimates for each of these sources are as
follows.
7.4.2 Methodology
Production Field Fugitive Emissions
Tilkicioglu and Winters (1989) estimated methane emissions per oil well using a model
oil production facility with 60 oil wells, heaters, separators, a surge tank, and related piping.
The facility had a production capacity of 40,000 barrels per day (bpd), and had an average of
about 47 components per well, including all equipment at the facility. The relatively high
production rate per well at this model facility does not have an adverse impact on the
emissions estimate because fugitive emissions are driven by the number of components at
the facility, which is relatively insensitive to the production rate.
Using the emissions factors published in Rockwell (1980), fugitive emissions from the
model facility were estimated at 76.7 kilograms per well per year (kg/well/yr). Recent analyses
7-11
-------
indicate that current fugitive emissions rates from oil facilities are lower than the rates implied
by the 1980 emissions factors (Webb 1992). Consequently, this emissions factor is probably
an over estimate of current emissions.
In Chapter 2, fugitive emissions from oil wells that produce gas were estimated at
72 kg/well/yr. This estimate is also likely to be an over estimate for oil wells that do not
produce gas, because these wells have lower gas flow rates, and probably fewer gas-related
components. For purposes of this analysis, this emissions factor of 72 kg/well/yr is used,
recognizing that it probably overstates the emissions from this source.
As discussed in Chapter 2, there were about 597,320 oil wells in the U.S. in 1990. Of
these, emissions from 48 percent, or 288,165, were included as part of the gas systems
analysis because they also produce gas. Therefore, this analysis uses the remaining 309,155
wells to estimate total fugitive emissions as:
309,155 wells x 72 kg/well/yr = 22.3 million kilograms per year.
Production Field Routine Maintenance Emissions
Tilkicioglu and Winters (1989) estimated the emissions from routine maintenance at the
model production facility to be 0.15 kg/well/yr. These emissions are associated with repairing
and maintaining valves, piping, and other equipment. Based on 309,155 wells, the national
emissions are estimated as:
309,155 wells x 0.15 kg/well/yr = 46,000 kilograms per year.
Crude Oil Storage Facility Emissions
Crude oil storage tanks emit hydrocarbons, including methane. The processes that
contribute to these emissions have been studied in depth and generally fall into two
categories:
• Breathing losses principally include the emissions around roof seals and joints
while the tank is in use. Wind speed is known to influence these emissions
rates.
• Working losses refer to the vapor emissions that occur when tanks are emptied
and filled. The vapor in the space above the liquid in the tank is often released
to some extent. These losses are driven by the frequency with which the tanks
are emptied and filled.
Fugitive emissions from the piping and other equipment at storage facilities also contribute to
methane emissions. Emissions from tank maintenance and repair are negligible and are not
considered here.
There are significant uncertainties in estimating crude oil storage tank emissions
because a good census of tank characteristics that influence emissions is not available,
including data on tank size, turnover rate (frequency of filling and emptying), roof
construction, and condition of seals. Tilkicioglu and Winters (1989) estimated emissions
7-12
-------
based on a model tank farm facility with fixed roof and floating roof tanks. Emissions factors
developed for the model facility were applied to published crude oil storage capacity data.
For breathing losses from floating roof tanks, average wind speed data were used to adjust
the model emissions rates by region of the U.S. The resulting emissions estimates are:
breathing losses: 889,000 kilograms per year;
working losses: 991,000 kilograms per year; and
fugitive emissions: 17,000 kilograms per year
for a total national emissions of 1.9 million kilograms per year. About 65 percent of the
emissions are estimated to be associated with fixed roof tanks. Floating roof tanks account
for about 35 percent, and fugitive emissions from related equipment and piping account for
less than 1 percent.
Refineries
Tilkicioglu and Winters estimated methane emissions from refineries. Two main
sources were considered:
Atmospheric distillation is typically the first stage of crude oil processing, where
the hydrocarbon components are separated into fractions by distillation and
steam stripping. Based on a model unit with a capacity of 30,000 bpd,
emissions from this source were estimated to be negligible.
• Waste gas streams containing methane are produced by several refining
processes. Radian (1980) identified heater flue gas as the primary source of
methane emissions from waste gas streams based on measurements at 10
refineries. These data were extrapolated to total U.S. refining capacity to
estimate methane emissions at 10.4 million kilograms per year.
Other sources, such as routine maintenance and system upsets are not expected to produce
significant emissions from refineries.
Marine Vessel Operations
The loading and unloading of crude oil tankers is known to release hydrocarbons,
including methane. The rate of emissions is influenced by tanker designs and emissions
control measures that are taken. Based on API Publication 2514, Atmospheric Hydrocarbon
Emissions from Marine Vessel Transfer Operations. Tilkicioglu and Winters estimated national
methane emissions associated with: (1) the loading and unloading of domestically-produced
crude oil transported by tanker; and (2) the unloading of foreign-produced crude transported
by tanker.
The quantity of domestic crude oil transported by tanker was estimated as Alaskan
crude oil production less Alaskan refinery crude oil utilization, plus 10 percent of non-Alaskan
crude oil production. Crude oil imports by tanker were estimated as total imports less
imports form Canada. The emissions factor for hydrocarbons of 1.0 pound per 1000 gallons
of crude oil was adjusted to reflect a methane content of 20 percent. The resulting methane
emissions estimate for this source is 6.1 million kilograms per year.
7-13
-------
Venting and Flaring
Venting and flaring activities commonly refer to the disposition of gas that cannot be
contained or otherwise handled. Usually, these activities are associated with oil and gas
production activities. For example, an oil well may produce an amount of gas that is too
small to market. This gas may be re-injected into the underground formation, or flared if re-
injection is not feasible. If flaring were not feasible, this gas may be vented. Whether vented
or flared, this gas would be considered part of the vented and flared quantity.
More recently, Radian (1992) expanded the definition of venting and flaring to include
gas releases from any equipment that is designed to release gas. For example, emissions
from a pneumatically-operated device that is powered by a pressurized gas stream would be
included as venting emissions under Radian's new definition. Similarly, gas released during
turbine engine start-up would also be considered a venting emission. For purposes of this
study, routine venting associated with the operation of equipment in the production,
processing, transmission, and distribution of gas is included in the emissions estimates for
gas systems (see Chapter 2). In this section, the traditional concept of venting and flaring is
used.
Venting and flaring activities release methane because the vented gas typically has a
high methane content and because flares do not always destroy 100 percent of the methane
in the gas. Most oil and gas producing states have regulations that restrict gas venting
during oil and gas production. Consequently, large scale venting is not common in the U.S.
The data for estimating how much gas is vented and how much is flared are very
poor. While U.S. venting and flaring data are published by the Department of Energy and
others annually, these data are fraught with numerous problems (Radian, 1992). Based on a
very preliminary assessment of the factors that contribute to venting and flaring, and an
assessment of venting and flaring data in Texas, Louisiana, and Oklahoma, Radian (1992)
estimated that methane emissions are about 4 percent of the total venting and flaring quantity
reported each year. A second estimate, by Barns and Edmonds (1990) estimated a factor of
20 percent based on conversations with experts in each of the oil producing states.
Using these estimates and the venting and flaring reported for 1990 results in
estimates of 92.5 to 462 million kilograms per year. This wide range reflects the considerable
uncertainty in the estimate for this emissions source. Additional research is needed to
improve the basis for estimating these emissions.
7.4.3 Current Emissions
Total emissions from oil production and refining are estimated as the sum of the
emissions from the above categories. Emissions from other activities are negligible (e.g.,
fugitive emissions from crude oil pipelines) or counted elsewhere (e.g., engine and heater fuel
is included in stationary source combustion).
Total emissions, excluding venting and flaring, are therefore about 40.7 million
kilograms per year, or 0.04 Tg/yr. Because this estimate is based on several model facilities,
actual emissions could be higher or lower. An uncertainty range of 1/4 to 4 times the
estimate is adopted, for a range of 0.01 to 0.16 Tg/yr. Adding the low and high venting and
7-14
-------
flaring values to these estimates produces a range of 0.10 to 0.62 Tg/yr. The high estimate is
driven by the high venting and flaring assumption from Barns and Edmonds (1990).
7.4.4 Future Emissions
Because domestic oil production is expected to continue to decline for the next 20
years, emissions from this source are also expected to decline. Additionally, as gas prices
increase, venting and flaring from oil wells may also decline. To reflect these trends, it is
assumed that emissions from this source decline by 10 percent by the year 2000, and by an
additional 10 percent by the year 2010.
7.4.5 Limitations of the Analysis
The estimate of emissions from petroleum production and refining is limited by a
general lack of data needed to describe emissions from venting and flaring activities.
Published venting and flaring data are based in many cases on accounting reports that
"balance" estimates of gas produced and utilized, with the difference being allocated to
"vented and flared." Consequently, the estimates of these emissions, which are the largest
component of the emissions from this source, are extremely uncertain.
7.5 ADDITIONAL SOURCES OF METHANE EMISSIONS
There are other of sources of methane emissions in the United States. Emissions
levels for these additional sources are generally perceived to be relatively small. However,
insufficient research has been conducted to estimate their contribution to increases in
atmospheric methane. This section provides brief descriptions of the additional sources.
7.5.1 Non-Fuel Biomass Burning
Biomass refers to organic material, both above and below ground, living and dead.
Methane is a by-product of biomass burning, resulting from incomplete combustion. On a
global level the burning of biomass, often for land-clearing purposes in tropical or sub-tropical
countries, is an important source of methane. Cicerone and Oremland (1988) have estimated
that biomass burning accounts for an annual 55 Tg of methane emissions, or over 10 percent
of global methane emissions.
The most important category of non-fuel biomass burning in the United States is waste
combustion.3 USEPA (1985) estimates that there are 150 municipal solid waste combustion
plants in the United States, in addition to an undetermined number of industrial and
commercial refuse incinerators. Sewage sludge combustion is another type of waste
combustion. USEPA (1985) estimates that there are 200 sewage sludge incinerators
3 Methane emissions estimates in section 7.3, above, include consideration of biomass fuel combustion. As
noted in Exhibit 7.5, the median estimate of total methane emissions from wood fuel combustion is about 0.7
teragrams.
7-15
-------
operating in the United States. Agricultural wastes may also be disposed of through
combustion. The practice of burning agricultural wastes is not common in the United States,
though some burning of field crops, firing for rangeland improvement and burning of logging
residues does take place in Western states.
Prescribed burning, as a method of forest management and not land clearing, is
practiced in the United States and is an instance of non-fuel biomass burning. However,
because prescribed burning is believed to replace or reduce natural forest fires, this practice
is not considered to be a net source of methane emissions. The burning of biomass for land
clearing purposes, though an important source of methane emissions in tropical and sub-
tropical countries, is rare in the United States.
Structural fires constitute another type of non-fuel biomass burning, though the
contribution of structural fires to methane emissions is not well researched. Methane
emissions from this particular source are thought to be minimal.
In summary, non-fuel biomass burning is not believed to be a significant source of
methane emissions in the United States. In general, methane emissions from non-fuel
biomass burning are not well documented. While methane emissions factors exist for some
forms of waste combustion, no national figures are available to measure total methane
emissions from this source category.
7.5.2 Industrial Processes and Wastes
Certain industrial processes and wastes generate methane. These processes and
wastes include wastewater from agricultural industries, the production of synthetic ammonia,
and the production of coke, iron, and steel (Beck et al. 1992).
Wastewater from Agricultural Industries
Methane is produced in anaerobic lagoons used to treat wastewater (Orlich 1990).
Wastewater treatment in anaerobic lagoons is a more common practice in developing
countries than in the United States and other developed countries where sewage is either
treated aerobically or the gas is recovered for energy purposes.
Though not a common practice, some agricultural industries in the United States do
produce wastewater streams in which anaerobic decomposition of organic matter may
produce methane. This phenomenon is associated with agricultural industries because of the
high organic load of their wastewater streams. Food product-related facilities such as
distilleries, breweries, fruit canneries, and starch plants are examples. Very little data exist on
which to base an estimate of methane emissions from this source in the United States.4
4 Some recent information indicates that wastewater managed in lagoons from industries such as the pulp and
paper industry may be emitting significant quantities of methane. Further efforts of EPA's Office of Research and
Development should clarify the contribution of methane from this source.
7-16
-------
Ammonia Production
The production of synthetic ammonia results in a limited amount of methane
emissions. Anhydrous ammonia is produced through a reaction of hydrogen with nitrogen,
and then compression and cooling of the resultant gas. Nitrogen is obtained from air, while
the source of hydrogen is natural gas (containing primarily methane), naphtha, or the
electrolysis of brine at chlorine plants. Approximately 98 percent of synthetic ammonia in the
United States is produced using natural gas.
Various releases from the ammonia production process result in methane emissions,
including fuel combustion emissions. Fuel combustion for all industrial uses is included in
methane emissions estimates shown in section 7.3. Due to limited information, no attempt to
project non-fuel combustion emissions from ammonia production in the United States is
presented here.
Coke, Iron and Steel Production
The production of coke, iron and steel results in some release of methane to the
atmosphere. The emissions occur both as a byproduct of the production process and from
the combustion of fuel to heat the furnaces. Methane emissions from fuel combustion for all
industrial uses are included in section 7.3. No estimates of emissions from coke, iron and
steel production are presented here because of insufficient information.
7.5.3 Land-Use Changes
Alterations in land-use can result in increases or decreases in methane emissions by
affecting natural sources or sinks of methane. In the United States, the contribution of land-
use changes to methane emissions remains largely unquantified and no estimates of net
changes in methane emissions are provided for these sources. Three major land-use
changes have been identified as affecting methane emissions in the United States: wetland
drainage, flooding of lands, and the conversion of grasslands to cultivated lands. Drainage of
wetlands results in a net decrease in methane emissions, and is not included in this
discussion of methane sources. Flooding of lands and conversion of grasslands to cultivated
lands are briefly described below.
Flooding of Lands
Flooding of dry land areas, such as results from dams and other anthropogenic water
diversion projects, results in net methane emissions to the atmosphere. The methane
emissions are due to the anaerobic decomposition of vegetation and soil carbon existent
from flooded land, and other organic matter that may accumulate and decompose in the
floodwater. The amount of methane emitted from flooding of lands will vary greatly
depending upon the depth of the floodwater, the nature and duration of the flooding,
vegetation type, soil type, and temperature. Because methane emissions will vary with
temperature, emissions rates will fluctuate seasonally. Other factors held equal, the temperate
climate that predominates in the United States results in lower methane emissions than would
be produced from flooded lands in countries with warmer climates.
Little research has been conducted on methane emissions from flooded lands. No
estimates of United States methane emissions from flooded lands are known to exist.
7-17
-------
Conversion of Grasslands to Cultivated Lands
Recent research by Mosier et al. (1991) conducted in Colorado has illustrated that
conversion of grasslands to cultivated lands may result in a reduction in the net uptake of
methane from these lands, and thus a net increase in release of methane to the atmosphere.
The reduction in methane uptake is associated with nitrogen fertilization of naturaf
ecosystems. Additional research is needed to determine the pre-conversion rate of methane
uptake in order to estimate the net effect on methane emissions of this land-use change.
7.6 REFERENCES
Barns, D. W. and J. A. Edmonds. 1990. An Evaluation of the Relationship Between the
Production and Use of Energy and Atmospheric Methane Emissions Prepared for U.S
Department of Energy (DOE//NBB-0088P).
Beck, L.L., S.D. Piccot, and D.A. Kirchgessner 1992. "Methane Emissions from Industrial
Sources" in The Global Methane Cycle: Its Sources, Sinks, Distributions and Role in
Climatic Changes. In Press.
Blake, D. 1990. Personal communication. U.C. Irvine, Irvine, California.
Cicerone, R.J., J.D. Shetter, and C.C. Delwiche. 1983. Seasonal variation of methane flux
from a California rice paddy. Journal of Geophysical Research 88:7203-7209.
Cicerone, R. and R. Oremland. 1988. "Biogeochemical Aspects of Atmospheric Methane,"
Global Biogeochemical Cycles, 2, 299-327, December 1988.
EIA (Energy Information Administration). 1987. Estimates of Biofuels Consumption in the U.S.
During 1987. EIA Service Report. U.S. Department of Energy. Washington, D.C.
CNEAF/NAFD 89-03.
EIA (Energy Information Administration). 1991 a. Annual Energy Outlook. Office of Energy
Markets and End Use. U.S. Deparment of Energy. Washington, D.C. DOE/EIA-
0383(91) March 1991.
EIA (Energy Information Administration). 1991b. Natural Gas Annual 1990. Office of Oil and
Gas, U.S. Deparment of Energy. Washington, D.C. DOE/EIA-0131 (90)/1. December
1991.
IPCC (Intergovernmental Panel on Climate Change). 1992a. Climate Change 1992: The
Supplementary Report to the IPCC Scientific Assessment. Report prepared for
Intergovernmental Panel on Climate Change by Working Group 1.
Matthews, E., I. Fung, and J. Lerner. 1991. "Methane Emissions from Rice Cultivation:
Geographic and Seasonal Distribution of Cultivated Areas and Emissions," Global
Biogeochemical Cycles 5:3-24.
7-18
-------
Mosier, A., D. Schimel, D. Valentine, K. Bronson, and W. Parton. 1991. "Methane and Nitrous
Oxide Fluxes in Native, Fertilized and Cultivated Grasslands," Nature, 350, 330-332,
1991. Cited in US Government, 1991.
OECD (Organization for Economic Cooperation and Development). 1991. Estimation of
Greenhouse Gas Emissions and Sinks, Final Report from the OECD Experts Meeting,
18-21 February 1991, Prepared for Intergovernmental Panel on Climate Change,
August 1991.
Orlich, J. 1990. "Methane Emissions from Landfill Sites and Waste Water Lagoons." Paper
presented at the International Workshop on Methane Emissions from Natural Gas
Systems, Coal Mining, and Waste Management Systems. U.S. Environmental
Protection Agency. Washington DC. April 1990.
Radian Corporation. 1992. "Venting and Flaring Emissions from Production, Processing, and
Storage in the U.S. Natural Gas Industry." Preliminary Draft Report prepared for the
U.S. EPA and the Gas Research Institute.
Rockwell International. 1980. Fugitive Hydrocarbon Emissions from Petroleum Production
Operations. Prepared for the American Petroleum Institute, API Publication no. 4322.
Sass, R.L, F.M. Fisher, P.A. Harcombe, and FT. Turner. 1990. Methane production and
emissions in a Texas rice filed. Global Biogeochemical Cycles 4:47-68.
Schutz, H., A. Holzapfel-Pschorn, R. Conrad, H. Rennenberg, and W. Seller. 1989. "A 3-Year
Continuous Record of the Influence of Daytime, Season, and Fertilizer Treatment on
Methane Emission Rates from an Italian Rice Paddy," Journal of Geophysical
Research 94:16405-16416.
Tilkicioglu, B. H. and D. R. Winters. 1989. Annual Methane Emissions Estimates of the Natural
Gas and Petroleum Systems in the U.S. Pipeline Systems Inc.
USDA (U.S. Department of Agriculture). 1991. Crop Production 1990 Summary. USDA,
Washington, D.C.
USDOT (U.S. Department of Transportation). 1992. Transportation Statistics, 1990.
Washington, D.C.
USEPA (U.S. Environmental Protection Agency). 1985. Compilation of Air Pollutant Emission
Factors. Office of Air and Radiation. Research Triangle Park, North Carolina. AP-42.
September 1985.
USEPA (U.S. Environmental Protection Agency). 1991. National Air Pollutant Emission
Estimates: 1940-1989. Office of Air Quality Planning and Standards. Research
Triangle Park, North Carolina. EPA-450/4-91 -004. March 1991.
U.S. Government. 1991. Preliminary Estimate of Greenhouse Gas Emissions and Sinks for the
United States, 1988. Submitted to the Intergovernmental Panel on Climate Change,
October 1991.
Webb, M. 1992. Personal communication. STAR Environmental, Torrance, California.
7-19
------- |