EPA-600 /R-96"080b
June 19S6
METHANE EMISSIONS FROM
THE NATURAL GAS INDUSTRY
VOLUME 2: TECHNICAL REPORT
FINAL REPORT
Prepared by:
Matthew R. Harrison
Lisa M. Campbell
Theresa M. Shires
R. Michael Cowgill
Radian International LLC
8501 N. Mopac Blvd.
P.O. Box 201088
Austin, TX 78720-1088
DCN: 650-049-20-01
For
GRI Project Manager: Robert A. Lott
GAS RESEARCH INSTITUTE
Contract No. 5091-251-2171
8600 West Bryn Mawr Avenue
Chicago, EL 60631
and
EPA Project Manager: David A. Kirchgessner
U.S. ENVIRONMENTAL PROTECTION AGENCY
Contract No. 68-D1-0031
National Risk Management Research Laboratory
Research Triangle Park, NC 27711
-------
FOREWORD
The U. S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
EPA REVIEW NOTICE
This report has been peer and administratively reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
ia
-------
DISCLAIMER
LEGAL NOTICE: This report was prepared by Radian International LLC as an account of
work sponsored by Gas Research Institute (GRI) and the U.S. Environmental Protection Agency
(EPA). Neither EPA, GRI, members of GRI, nor any person acting on behalf of either:
a. Makes any warranty or representation, express or implied, with respect to the accuracy,
completeness, or usefulness of the information contained in this report, or that the use of
any apparatus, method, or process disclosed in this report may not infringe privately
owned rights; or
b. Assumes any liability with respect to the use of, or for damages resulting from the use of,
any information, apparatus, method, or process disclosed in this report.
NOTE: EPA's Office of Research and Development quality assurance/quality control (QA/QC)
requirements are applicable to some of the count data generated by this project. Emission data
and additional count data are from industry or literature sources, and are not subject to
EPA/ORD's QA/QC policies. In all cases, data and results were reviewed by the panel of experts
listed in Appendix D of Volume 2.
ii
-------
RESEARCH SUMMARY
Title
Contractor
Methane Emissions from the Natural Gas Industry,
Volume 2: Technical Report
Final Report
Radian International LLC
Principal
Investigators
Report Period
Objective
Technical
Perspective
GRI Contract Number 5091 -251 -2171
EPA Contract Number 68-D1-0031
Matthew R. Harrison
Lisa M. Campbell
Terri M. Shires
R. Michael Cowgill
March 1991 - June 1996
Final Report
This report describes the results of a study to quantify the annual methane
emissions from the natural gas industry.
The increased use of natural gas has been suggested as a strategy for
reducing the potential for global warming. During combustion, natural gas
generates less carbon dioxide (C02) per unit of energy produced than either
coal or oil. On the basis of the amount of C02 emitted, the potential for
global warming could be reduced by substituting natural gas for coal or oil.
However, since natural gas is primarily methane, a potent greenhouse gas,
losses of natural gas during production, processing, transmission, and
distribution could reduce the inherent advantage of its lower C02 emissions.
To investigate this, Gas Research Institute (GRI) and the U.S. Environmental
Protection Agency's Office of Research and Development (EPA/ORD)
cofunded a major study to quantify methane emissions from U.S. natural gas
operations for the 1992 base year. The purpose of this study was to provide
emissions data that could be used to construct global methane budgets and to
determine the relative impact of natural gas on global warming versus the
impact from coal and oil.
This summary report is volume 2 of a multi-volume set of reports that fully
describe the project.
in
-------
Results The national emissions for the base year are 314 ± 105 Bscf (± 33%), which
is equivalent to 1.4 ± 0.5% of gross natural gas production. The overall
program also showed that the percentage of methane emitted for an
incremental increase in natural gas sales would be significantly lower than
the baseline case.
On an industry segment basis, the production segment emits 84.4 Bscf, gas
processing plants emit 36.4 Bscf, transmission and storage facilities emit
116.5 Bscf, and distribution systems emit 77.0 Bscf. The report also shows
that the largest type of methane emissions is fugitives, which accounts for
195.2 Bscf from all segments combined.
The program reached its accuracy goal and provides an accurate estimate of
methane emissions that can be used to construct U.S. methane inventories
and analyze fuel switching strategies.
Technical The techniques used to determine methane emissions were developed to be
Approach representative of annual emissions from the natural gas industry. However, it
is impractical to measure every source continuously for a year. Therefore,
emission rates for various sources were determined by developing annual
emission factors for sources in each industry segment and extrapolating these
data based on activity factors to develop a national estimate, where the
national emission estimate is the product of the emission factor and activity
factor.
The development of specific emission factors and activity factors for each
industry segment are presented in a separate report.
Project For the 1992 base year the annual methane emissions estimate for the
Implications U.S. natural gas industry is 314 Bscf ± 105 Bscf (± 33%). This is equivalent
to 1.4% ± 0.5% of gross natural gas production. Results from this program
were used to compare greenhouse gas emissions from the fuel cycle for
natural gas, oil, and coal using the global warming potentials (GWPs)
recently published by the Intergovernmental Panel on Climate Change
(IPCC). The analysis showed that natural gas contributes less to potential
global warming than coal or oil, which supports the fuel switching strategy
suggested by the IPCC and others.
In addition, results from this study are being used by the natural gas industry
to reduce both operating costs and emissions. Some companies are also
participating in the Natural Gas-Star program, a voluntary program sponsored
by EPA's Office of Air and Radiation in cooperation with the American Gas
Association to implement cost-effective emission reductions and to report
iv
-------
reductions to EPA. Since this program was begun after the 1992 baseline
year, any reductions in methane emissions from this program are not
reflected in this study's total emissions.
Robert A. Lott
Senior Project Manager, Environment and Safety
-------
TABLE OF CONTENTS
Page
1.0 SUMMARY 1
2.0 INTRODUCTION 3
3.0 METHODS 9
3.1 Emission Source Characterization 9
3.1.1 General Industry Description 9
3.1.2 Operating Mode 17
3.1.3 Emission Types 19
3.2 Emission Estimation Technique 20
3.2.1 Measurement Techniques for Steady Emissions 20
3.2.2 Calculation Approach for Unsteady Emissions 23
3.3 General Extrapolation Methodology 24
3.3.1 Sampling Approach 25
3.3.2 Redefining the Emission Factor 26
3.4 Accuracy 27
3.4.1 Precision 27
3.4.2 Bias 28
3.5 Quality Assurance and Quality Control Approach 28
3.5.1 Overview 29
3.5.2 Definitions 30
3.5.3 Quality Control 31
3.5.4 Quality Assurance 33
4.0 DETAILED RESULTS 37
4.1 Emission Type Summary 37
4.2 Fugitive Emissions 39
4.2.1 Equipment Leaks 41
4.2.2 Underground Pipeline Leaks 50
4.3 Vented Emissions 52
4.3.1 Pneumatic Devices 52
4.3.2 Blow and Purge 55
4.3.3 Dehydrator Glycol Pumps 56
4.3.4 Dehydrator Vents 57
4.3.5 Chemical Injection Pumps 58
4.4 Combusted Emissions 59
4.5 Largest Sources by Industry Segment 61
vi
-------
TABLE OF CONTENTS
(continued)
4.6 Equipment Emissions 68
4.7 Accuracy Results 72
5.0 ANALYSIS AND CONCLUSIONS 74
5.1 Impact of Natural Gas Use on Global Warming 75
5.2 Comparison to Previous Estimates 78
5.3 Current and Future Industry Emissions 80
5.3.1 Industry Practices to Reduce Methane Emissions 80
5.3.2 Incremental Increases in System Throughput 82
5.4 Lessons Learned for Future Studies 83
5.4.1 Sampling/Statistical Methods 83
5.4.2 Measurement Methods 86
5.4.3 Significant Sources 90
6.0 REFERENCES 91
APPENDIX A - Summary Table of Emissions Sources A-l
APPENDIX B - Effect of Methane Emissions on Global Warming B-l
APPENDIX C - Conversion Table C-l
APPENDIX D - Project Reviewers D-l
vii
-------
LIST OF FIGURES
Page
3-1 Gas Industry Flow Chart 11
3-2 Gas Industry Boundaries 14
3-3 Transmission and Storage Stations 16
3-4 Distribution Segment Equipment 18
4-1 Emissions by Type 38
4-2 Major Contributors to Fugitive Emissions from the Natural Gas Industry 39
4-3 Major Contributors to Fugitive Emissions—By Segment Facilities 40
4-4 Contributions to Vented Emissions 52
4-5 Summary of Methane Emissions 61
4-6 Production Segment Largest Sources 64
4-7 Gas Processing Segment Largest Sources 65
4-8 Transmission and Storage Largest Sources 66
4-9 Distribution Largest Sources 67
5-1 Contribution of Major Methane Sources to Total U.S. Anthropogenic Emissions 74
B-l Breakeven Percentage—Natural Gas Compared with Coal B-10
viii
-------
LIST OF TABLES
Page
2-1 U.S. Gas Industry Methane Emissions Study Report List 5
3-1 Industry Characterization 12
3-2 Emission Characterizations 21
4-1 United States Natural Gas Industry Largest Methane Emissions Sources 38
4-2 Example of National Emissions Estimation for Gas Storage Facilities 43
4-3 Summary of Methane Emissions 61
4-4 Emissions by Type 63
4-5 Production Segment Largest Sources 64
4-6 Gas Processing Segment Largest Sources 65
4-7 Transmission and Storage Segment Largest Sources 66
4-8 Distribution Segment Largest Sources 67
4-9 Emissions by Equipment 69
4-10 Estimated Equipment Emission Factors 71
5-1 Equivalent C02 Emissions 78
B-l Global Warming Potential of Methane B-5
B-2 Amount of C02 and Methane Remaining in Atmosphere with Time B-7
B-3 Breakeven Percentage (BP) for Coal and Oil for Various GWP Integration Intervals . B-9
B-4 Incremental Changes in Emissions Resulting from Increased Gas Sales
- Expected Gas B-l2
ix
-------
LIST OF TABLES
(Continued)
Page
B-5 Incremental Changes in Emissions Resulting from Increased Gas Sales -
Upper Limit Case B-12
B-6 Sources of C02 Equivalents for Each Fuel Type B-16
B-7 C02 Equivalent Emissions from Natural Gas B-19
B-8 C02 Equivalent Emissions from Coal B-20
B-9 1990 Methane and C02 Emissions from Crude Production Through Refined
Product Transportation B-21
B-10 Properties of Fuel Oils B-23
B-ll C02 Equivalent Emissions from Fuel Oil B-24
B-12 Equivalent C02 Emissions for Natural Gas, Oil, and Coal B-24
x
-------
1.0 SUMMARY
This report summarizes the results of a project sponsored by Gas Research
Institute (GRI) and U.S. Environmental Protection Agency's Office of Research and
Development (EPA/ORD) to quantify methane emissions from the natural gas industry. The
project was initiated to evaluate whether the suggested strategy of increasing the use of natural
gas to reduce global warming was valid in light of methane emitted from the industry. It also had
the purpose of determining the gas industry's contribution of methane to the global inventory of
greenhouse gas emissions.
During combustion, natural gas generates less carbon dioxide (C02) per unit of
energy produced than either coal or oil. On the basis of the amount of C02 emitted, the potential
for global warming could be reduced by substituting natural gas for coal or oil. However, since
natural gas is primarily methane, a potent greenhouse gas, losses of natural gas during
production, processing, transmission, and distribution could reduce the inherent advantage of its
lower C02 emissions.
To investigate this, GRI and EPA/ORD cofunded a major study to quantify
methane emissions from U.S. natural gas operations for the 1992 base year. The results of this
study can be used to construct global methane budgets and to determine the relative impact on
global warming of natural gas versus coal and oil.
For the 1992 base year the annual methane emissions estimate for the U.S. natural
gas industry is 314 Bscf ±105 Bscf (± 33%).* This is equivalent to 1.4% ± 0.5% of 1992 gross
natural gas production. The project reached it accuracy goal of determining emissions within
± 0.5% of production, and provides an accurate methane emissions estimate that can be used in
fuel switching analyses. The program also showed that the percentage of methane emitted for an
'Readers more comfortable with metric units will find a conversion table in Appendix C.
1
-------
incremental increase in natural gas production would be significantly lower than the baseline
case.
Results from this program were used to compare greenhouse gas emissions from
the fuel cycle for natural gas, oil, and coal using the global warming potentials (GWPs) recently
published by the Intergovernmental Panel on Climate Change (IPCC).1 The analysis showed that
natural gas contributes less to potential global warming than coal or oil, which supports the fuel
switching strategy suggested by the IPCC and others. Even across a wide range of assumptions
on factors affecting the global warming potentials, natural gas production and use in the United
States contributes less to global warming than coal or oil.
The results are currently being used by the natural gas industry to reduce
operating costs while reducing emissions. This has led to the development of a voluntary
program, the Gas-Star program, sponsored by EPA in cooperation with the American Gas
Association (A.G.A.).2 As part of this voluntary program, participating companies implement
cost-effective emission reductions and report the reductions to EPA.
2
-------
2.0 INTRODUCTION
The increased use of natural gas has been suggested by IPCC and EPA as a
strategy for reducing global warming.1,3 During combustion, natural gas generates less carbon
dioxide (C02) per unit of energy produced than either coal or oil. On the basis of the amount of
C02 emitted, global warming could be reduced by substituting natural gas for coal. However,
since natural gas is primarily methane, a potent greenhouse gas, losses of natural gas during
production, transmission, and distribution could reduce the inherent advantage of its lower C02
emissions. For this reason, GRI and EPA jointly funded and managed a program to estimate
methane emissions from the U.S. natural gas industry for the 1992 base year. The objective of
this comprehensive program was to quantify methane emissions from the gas industry starting at
the wellhead and ending immediately downstream of the customer's meter. The accuracy goal of
the project was to determine these emissions to within 0.5% of natural gas production based on a
90% level of confidence. This is equivalent to an accuracy goal of ±111 billion standard cubic
feet (Bscf) per year for the 1992 base year.
The methane emissions program was conducted in three phases: scoping,
methods development, and implementation phase. During the scoping phase of the program, the
methane emissions from each source in the gas industry were quantified on the basis of available
data and engineering judgement. These initial estimates were used to set priorities for data
collection according to the relative importance of their contribution to emissions or the
uncertainty in emissions.
In the second phase of the program, methods were developed to measure and/or
calculate methane emissions from the variety of sources that make up the gas industry. These
methods were validated through tests designed to quantify the accuracy of the measurement
approach (i.e., proof of concept tests), and through industry review of the analytical methods.
However, emissions could not be measured or calculated from each piece of equipment (e.g.,
every glycol dehydrator, compressor engine, etc.) in the industry because of the vast amount of
3
-------
equipment. Therefore, a major task in the second phase was to develop defensible techniques for
extrapolating a limited amount of data collected for each source category to other sources in the
category in order to develop a national emissions estimate.
The third phase of the program focused on collecting data needed to define
emissions from all sources and extrapolating these data to estimate nationwide methane
emissions. Data collection in the third phase of the program concentrated on high priority
sources (i.e., sources with large emissions and/or large uncertainties). An Advisory Committee
consisting of industry representatives, project sponsors, and other interested parties including
scientists, government policy analysts, and environmentalists provided guidance and peer review
for all phases of the program. In addition, Gas Industry Review Panels for each segment of the
gas industry provided more detailed technical review of the project to ensure that the
methodologies and assumptions used in the study were consistent with industry practices.
The final analysis of the data and the methodologies used in the program have
been documented in a series of 31 reports. Table 2-1 shows the report name, report volume
number, report reference number, and the author of each report. The first 15 reports present final
data and analysis, and these reports have been assigned volume numbers. The first 15 reports are
available through the National Technical Information Service (NTIS) or from GRI. The
remaining reports represent field data, proof of concept tests, and efforts cofunded by others, and
have not been assigned volume numbers. These reports are listed here only as references, and
must be ordered from the listed author by the reference number.
The first five volumes present the executive summary, the technical report, the
general methodology, the statistical methodology, and the activity factors. These first five
volumes are the most important source of overall information on the project. Volumes 6 through
15 present the details of the test program and calculation procedures for determining specific
emission and activity factors.
4
-------
TABLE 2-1. U. S. GAS INDUSTRY METHANE EMISSIONS STUDY REPORT LIST
Report Name
Report Number or
Reference
Authors/
Contractor
Methane Emissions from the Natural Gas Industry, Volume 1: Executive Summary4
GRI-94/0257
EPA-600/R-96-080a
M.R. Harrison
et al.
Methane Emissions from the Natural Gas Industry, Volume 2: Technical Report
GRI-94/0257.1
EPA-600/R-96-080b
M.R. Harrison
et al.
Methane Emissions from the Natural Gas Industry, Volume 3: General Methodology 5
GRI-94/0257.20
EPA-600/R-96-080c
M.R. Harrison
et al.
Methane Emissions from the Natural Gas Industry, Volume 4: Statistical Methodology6
GRI-94/0257.21
EPA-600/R-96-080d
H.J. Williamson
et al.
Methane Emissions from the Natural Gas Industry, Volume 5: Activity Factors7
GRI-94/0257.22
EPA-600/R-96-080e
B.E. Stapper
Methane Emissions from the Natural Gas Industry, Volume 6: Vented and Combustion
Source Summary8
GRI-94/0257.23
EPA-600/R-96-080f
T.M. Shires and
M.R. Harrison
T.M. Shires and
M.R. Harrison
Methane Emissions from the Natural Gas Industry, Volume 7: Blow and Purge
Activities9
GRI-94/0257.24
EPA-600/R-96-080g
Methane Emissions from the Natural Gas Industry, Volume 8: Equipment Leaks 10
GRI-94/0257.25
EPA-600/R-96-080h
K.E. Hummel
et al.
Methane Emissions from the Natural Gas Industry, Volume 9: Underground
Pipelines 11
GRI-94/0257.26
EPA-600/R-96-080i
L.M. Campbell
et al.
Methane Emissions from the Natural Gas Industry, Volume 10: Metering and Pressure
Regulating Stations in Natural Gas Transmission and Distribution 12
GRI-94/0257.27
EPA-600/R-96-080j
L.M. Campbell
and B.E. Stapper
Methane Emissions from the Natural Gas Industry, Volume 11: Compressor Driver
Exhaust13
GRI-94/0257.28
EPA-600/R-96-080k
C.J. Stapper
(Continued)
-------
TABLE 2-1. U. S. GAS INDUSTRY METHANE EMISSIONS STUDY REPORT LIST
(Continued)
Report Name
Report Number or
Reference
Authors/
Contractor
Methane Emissions from the Natural Gas Industry, Volume 12: Pneumatic Devices 14
GRI-94/0257.29
EPA-600/R-96-0801
T.M. Shires and
M.R. Harrison
Methane Emissions from the Natural Gas Industry, Volume 13: Chemical Injection
Pumps 15
GRI-94/0257.30
EPA-600/R-96-080m
T.M. Shires
Methane Emissions from the Natural Gas Industry, Volume 14: Glycol
Dehydrators 16
GRI-94/0257.31
EPA-600/R-96-080n
D. Myers
Methane Emissions from the Natural Gas Industry, Volume 15: Gas-Assisted Glycol
Pumps 17
GRI-94/0257.33
EPA-600/R-96-080o
D. Myers and
M.R. Harrison
An Engineering Estimate of the Incremental Change in Methane Emissions with Increasing
Throughput in a Model Natural Gas System 18
GRI-94/0257.32
Columbia Gas
Results of Tracer Measurements of Methane Emissions from Natural Gas System
Facilities, Final Report19
Evaluation of Methane Emissions from Natural Gas Production Operations Using Tracer
Methodologies20
GRI-94/0257.43
GRI-92/0102
Aerodyne,
Washington State
University,
University of New
Hampshire
SRI International
Fugitive Hydrocarbon Emissions from Oil and Gas Production Operations2'
API 4589
STAR
Environmental
Fugitive Hydrocarbon Emission: Eastern Gas Wells (Final Report)22
GRI-95/0117
STAR
Environmental
(Continued)
-------
TABLE 2-1. U. S. GAS INDUSTRY METHANE EMISSIONS STUDY REPORT LIST
(Continued)
Report Name
Report Number or
Reference
Authors/
Contractor
Leak Rate Measurements for Natural Gas Customer Meters (Draft)23
GRI-94/0257.36
Indaco Air Quality
Services
Leak Rate Measurements at U.S. Natural Gas Transmission Compressor Stations (Draft)24
GRI-94/0257.37
Indaco Air Quality
Services
Emission Factors for Oil and Gas Production Operations25
API 4615
STAR
Environmental
Fugitive Methane Emissions: Customer Meter Sets (Final Report)26
GRI-95/0204
STAR
Environmental
Phase 3 Program Plan - Implementation Plan27
Radian DCN
92-263-081-02
Radian
Corporation
Southwest
Research Institute
Mass Balance of a Natural Gas Transmission System for Improved Estimates of Methane
Emissions28
SwRI 04-4447
Soil Consumption of Methane from Natural Gas Pipeline Leaks29
GRI-94/0257.35
American Oil and
Gas Reporter,
March 1995
Aerodyne,
Washington State
University,
University of New
Hampshire
Sampler Enables Measurement of Leaks on Site-Specific Basis30
Lott, Webb,
Howard
-------
This technical report is organized as follows: Section 2 presents background
information; Section 3 provides an overview of the methodologies developed and followed in the
study which includes methods for characterizing the industry, measuring and calculating
emission factors, collecting activity factor data (equipment and component counts), and
extrapolating the data to derive an annual methane emissions estimate for the U.S. natural gas
industry. Section 4 provides summaries of the largest methane emission sources. Section 5
provides an overview of the major conclusions drawn from the study.
8
-------
3.0 METHODS
This section characterizes the natural gas industry and describes in general terms
the methods used to define and extrapolate emissions for all source types or categories that
comprise the industry.
3.1 Emission Source Characterization
The first step for estimating methane emissions from the U.S. natural gas industry
is to identify and characterize each emission source within the industry, so that all significant
sources are included. To fully characterize the industry, sources were defined by equipment
type, mode of operation, and type of emissions.
While this section draws a general picture of the industry, it is not intended to be a
definitive picture of any company or of the industry regarding specific operational practices and
procedures. Rather, it is intended to define the general industry equipment practices and
procedures used in 1992, the base year of the program, that could lead to measurable emissions
of methane. Details that were useful for determining methane emissions are contained in specific
reports (see Table 2-1).
3.1.1 General Industry Description
The natural gas industry uses wells to produce natural gas existing in underground
formations, then processes, compresses, and transports the gas to the customer. Transportation
and distribution of natural gas involve interstate and intrastate pipeline transportation, storage,
and finally distribution of the gas by local distribution pipeline networks.
9
-------
The generally accepted segments of the natural gas industry are:
1) Production
2) Processing
3) Transmission/storage
4) Distribution
Each of these segments is shown in the flow chart for the industry in Figure 3-1. Some of the
major equipment in each segment is shown in Table 3-1. Each segment is described in more
detail in the following subsections.
This project set specific boundaries for each segment of the industry that specify
what equipment is included in the study. The guideline used for setting the boundaries was to
include only the equipment in each segment that is required for marketing natural gas. For
example, oil production equipment is excluded if it is used to produce oil and not natural gas.
Similarly, gas processing equipment associated with the fractionation of propane, butane, and
natural gas liquids are excluded from consideration. In distribution, all equipment up to and
including the customer's meter are included. End-user emissions are not included in this
estimate.
Each industry segment is described in more detail in the following subsections:
Production Segment Description
The production segment is comprised of gas and oil wells and the surface
equipment required to produce gas. The well includes the holes drilled through subsurface rock
to reach the producing formation and the subsurface equipment such as casing and tubing pipe.
Gas and oil surface equipment can include separators, heaters, heater-treaters, tanks, dehydrators,
compressors, pumps, and pipelines.
10
-------
PRODUCTION
Surface Facilities
PROCESSING
TRANSMISSION/
STORAGE
Direct
Sales
Compressor
Stations
Pipelines
Gas Plant
Pipelines
Liquids
Storage
Liquids
Underground
Storage
Reservoir
DISTRIBUTION
Main and
Service Pipelines
M&PR Stations (Qj
-<§)
-O)
-<@>
-<9>
-(0)
"(0)
-®>
Customer Meters
~
Compressor
©
Meter
£
Pressure
Regulator
Figure 3-1. Gas Industry Flow Chart
-------
TABLE 3-1. INDUSTRY CHARACTERIZATION
Segment j Facilities
Production | Well Sites,
Central Gathering Facilities
1
/
Equipment at the Facility
Wellheads, Separators,
Pneumatic Devices, Chemical
Injection Pumps, Dehydrators,
Compressors, Heaters, Meters,
Pipelines
Processing
Gas Plants
Vessels, Dehydrators,
Compressors, Acid Gas
Removal (AGR) Units, Heaters,
Pneumatic Devices
Transmission
Transmission Pipeline Networks,
Compressor Stations,
Meter and Pressure Regulating Stations
Vessels, Compressors, Pipelines,
Meters/Pressure Regulators,
Pneumatic Devices
Storage
Underground Injection/Withdrawal
Facilities, and Liquefied Natural Gas
(LNG) Facilities
Wellheads, Vessels,
Compressors, Dehydrators,
Heaters, Pneumatic Devices
Distribution
Main and Service Pipeline Networks,
Meter and Pressure Regulating Stations
Pipelines, Meters and Pressure
Regulators, Pneumatic Devices,
Customer Meters
12
-------
The definition for gas industry production equipment excludes equipment
associated with oil production. Also, unmarketed natural gas, such as that produced by oil wells
that vent gas or that reinject gas for gas lift circulation only, are not considered part of the natural
gas industry. Figure 3-2 shows the general equipment found in the oil and gas production
segment, as well as the boundaries between gas and oil production equipment used by this study.
The boundary between oil and gas equipment shown in Figure 3-2 affects the gas
industry emissions estimate since it excludes some high emission rate production equipment
associated with oil production. An accounting of total production segment emissions, or just oil
industry emissions, will have to include the oil industry equipment excluded from this study
(such as some pneumatics, some chemical injection pumps, and oil tanks).
Gas Processing Segment Description
Natural gas processing plants recover high value liquid products from the gas
stream and maintain the quality (i.e., content and heating value) of the gas stream. The liquid
products include natural gasoline, butane, propane, and in some cases, ethane. The products are
removed by compression and cooling or by absorption.
A gas plant may have fractionation towers and stabilization towers to further
purify the individual components of the product stream. The back end of the gas plant, such as
the fractionation train, is excluded from the gas industry definition since its function is to purify
and market liquid products. Also, the back end of the gas plant has negligible methane emissions
since the liquids handled have little methane content.
The front end of the gas plant often contains dehydration facilities, wet gas
compression, and the absorption or compression and refrigeration process. All natural gas
processing plants are considered part of the natural gas industry, and methane emissions from
these facilities are included in this study.
13
-------
Petroleum Industry
Oil Wei!
Wellhead
Compressor
Compressor
Gas Lift Gas
Meter
Oil Well
Wellhead
(producing
gas)
Field Use
Gas
^ Separator ^
Vapor Recovery
Compressor
Pnuemaiic
Control
Valve
Heater/
Treater
Salt Water and
Oil Stock Tanks
Coal Bed
Methane Well
V/////Z
u:
^ Separator ^
Compressor
Meier
Fresh Water
Dehydrator
Y/S////
Chemical
Injection
Pump
Dehydrator 1
/^S
Compressor
Meier
Gas Well
Wellhead
Field Use
^ Separator j
Vapor Recovery
Compressor
' ^ Hydrocarbon
Condensate
or Oil Tank
Pmiemauc
Petroleum Industry
Control Valve
Salt Water
Tank
Figure 3-2. Gas Industry Boundaries
14
-------
Transmission and Storage Segment Description
The transmission segment moves the natural gas from the gas plant or directly
from field production to local distribution companies (LDCs). Gas is often transported across
many states, such as from the Gulf Coast to the Eastern seaboard of the United States. The
transmission segment consists of large diameter pipeline, compressor stations, and metering
facilities. All of these facilities and all of the equipment they contain are considered part of the
natural gas industry.
Transmission compressor stations usually consist of piping manifolds,
reciprocating engines or gas turbines, reciprocating or centrifugal compressors, and generators,
as shown in Figure 3-3. Dehydrators may be included but are not typically present because of
upstream gas drying. Some transmission compressor stations may also include metering
facilities.
Transmission companies also have metering and regulating stations (M&PR)
where they exchange gas with other transmission companies, or where they deliver gas to LDCs
or industrial customers. These stations may contain heaters, small dehydrators, and odorant
addition equipment.
Most storage facilities exist to store natural gas produced during off-peak times
(usually summer) so that gas can be produced and delivered during peak demand. Storage
facilities are often located close to consumption centers so that cross-country transmission
pipelines do not have to be sized for peak demand. Storage facilities can be below or above
ground. Above-ground facilities are liquefied natural gas (LNG) facilities that liquefy the gas by
supercooling and then storing the liquid phase methane in above ground, heavily insulated
storage tanks. Below-ground facilities compress and store the gas (in vapor phase) in one of
several formations: 1) spent gas production fields, 2) aquifers, or 3) salt caverns. Below-ground
storage is the predominant means of gas storage.
15
-------
Gas
from
Pipeline
On
TRANSMISSION
COMPRESSOR STATION
ABOVE GROUND STORAGE
Compressors
Cooling
Process
->
LNG
Tank
Heaters
\f Gas
Transmission Pipeline
from
Pipeline
BELOW GROUND STORAGE
Compressor
Station
t
Wells
_ 4= 4=
Dehydrator storage Field
Figure 3-3. Transmission and Storage Stations
-------
Most storage stations consist of a compressor station that is very similar to a
transmission compression station (see Figure 3-3). Underground storage facilities also have
storage field wells, and usually have dehydrators to remove water absorbed by the gas while
underground. All storage equipment is included in boundaries of the natural gas industry defined
by this project.
Distribution Segment Definition
The distribution segment receives high pressure gas from transmission pipelines,
reduces the pressure, and delivers the gas to residential, commercial, and industrial consumers.
This segment includes pipelines (mains and services), M&PR stations and customer meters. All
of these facilities are considered to be an integral part of the gas industry. Figure 3-4 shows a
schematic of the distribution segment and the equipment that it includes.
3.1.2 Operating Mode
After identifying the major equipment (source types) in each industry segment,
emissions from each source were identified by examining the operating modes of the equipment
that may lead to emissions, and by associating one of three possible types of emissions from the
source: fugitive emissions, vented emissions, or combustion emissions.
The cause of emissions is directly related to the operating mode of the equipment.
Since more than one cause of emissions can be associated with a particular piece of equipment, it
is important to identify the various operating modes in order to identify all emissions. In general,
the operating modes are:
• Start-up;
• Normal operations;
• Maintenance;
• Upsets; and
• Mishaps.
17
-------
Meter and Pressure
Regulating (M&PR) Station
Pressure-Reducing
Station
Services (Small Pipe)
00
Transmission
Pipeline
<=>
Meter
Pressure-Reducing
Regulators
Mains (Pipeline)
(M)
©
(m)
©
(M)
(M)
©
Cm)
Customer Meters
(Residential,
Commercial,
Industrial,
Electric Utility)
Figure 3-4.
Distribution Segment Equipment
-------
Start-up operations, such as purging a newly constructed plant or pipeline, can
involve purging natural gas directly to the atmosphere. Emissions associated with normal
operations include emissions from process vents, fugitive emissions from packed or sealed
surfaces or underground pipeline leaks, and emissions from gas-operated pneumatic devices.
Maintenance operations involve blowing down equipment, such as compressors, pipelines, or
vessels, before equipment maintenance. Process upsets usually involve releasing natural gas to
the atmosphere or to a combustion device, such as a flare, as the result of overpressure or
emergency shutdown conditions. Mishaps are intended to include accidental occurrences that
result in emissions, such as third-party damage to pipelines (dig-ins).
3.1.3 Emission Types
Emissions from each piece of equipment in the natural gas industry can be
classified in one of three general emission types: 1) fugitive emissions; 2) vented emissions; and
3) combustion emissions. Fugitive emissions are unintentional leaks emitted from sealed
surfaces, such as packings and gaskets, or leaks from underground pipelines (resulting from
corrosion, faulty connections, etc.). Vented emissions are releases to the atmosphere by design
or operational practice. Examples of vented emissions include emissions from continuous
process vents, such as dehydrator reboiler vents; maintenance practices, such as blowdowns; and
small individual sources, such as gas-operated pneumatic device vents. Combustion emissions
are exhaust emissions from combustion sources such as compressor engines, burners, and flares.
In summary, the facilities and equipment comprising each segment of the industry
were identified. Each source (i.e., piece of equipment) was then examined for different
emissions during different operating modes. Emissions from each source were also categorized
as either combustion, vented, or fugitive. Equipment, such as compressors, might emit gas under
all three categories (fugitive emissions when pressurized, vented emissions when blown down
for maintenance, and combustion emissions during normal operations).
19
-------
3.2
Emission Estimation Technique
After all potential sources of methane emissions in the industry were identified
and characterized, the annual emissions were estimated. Because it would be impractical to
measure emissions all year for every source, it is important that a measurement be representative
of the annual emissions. Some emissions from natural gas industry sources are continuous and
nearly "steady" and a single measurement is representative of annual emissions. ("Steady" is a
relative term and to some extent is dependent on the time period of data needed for the study.
For this study, the annual value of methane emissions is needed.) The measurement techniques
used in this study depended on the variability of the emission rate with time.
Emissions that are intermittent are considered "unsteady" and have variable
emission rates during a year. Because it would not be practical to collect data continuously for a
year for each source, emissions from these sources were calculated rather than measured. Table
3-2 shows examples of emission sources characterized by operating mode emission type and
whether the emissions are steady or unsteady.
3.2.1 Measurement Techniques for Steady Emissions
Steady emissions result from unintentional leaks from sealed surfaces such as pipe
connectors, valve packing, flange gaskets at surface facilities, and from components and small
holes in below-ground equipment (i.e., pipelines). One method for measuring these steady
fugitive emissions from above-ground facilities (surface production equipment, gas plants,
compressor stations, etc.) is to measure emissions from individual components, and then sum all
the component emissions for the facility. Other surface facility methods include the tracer gas
method. Measuring emissions from buried pipelines is done through a leak statistics method.
Each of these methods is described in the following subsections.
20
-------
TABLE 3-2. EMISSION CHARACTERIZATIONS
Emission
Type
Specific Source Examples
Operating Mode
Steady or
Unsteady
Fugitive
Packed or Sealed Surfaces
Normal Operations
Steady
Leaks (holes in gathering & distribution
pipes)
Normal Operations
Steady
Leaks (holes in transmission pipes)
Normal Operations
Steady
Vented
Dehydrator Vents
Normal Operations
Steady
Pipeline Purge/Blowdown
Maintenance
Unsteady
Pneumatic Devices
Normal Operations
Unsteady
Compressor Starts
Normal Operations
Unsteady
Equipment Blowdown
Maintenance
Unsteady
Chemical Injection Pump Vents
Normal Operations
Unsteady
Pressure Relief Valve Lift
Upsets
Unsteady
Combusted
Compressor Driver Exhaust
Normal Operations,
Unsteady
Flaring
Upsets/Maintenance
Unsteady
Burners
Normal Operations
Unsteady
Component Measurement Methods
One method for determining fugitive emissions from above-ground facilities is to
determine emissions from basic components such as valves, flanges, seals, and other connectors
and then sum these for a given facility to determine total emissions. As part of this program,
GRI cofunded studies with API and others to update emission factors for pipe fittings and other
components used in oil and gas production.21,22'23'24'25'26 Nearly 200,000 components were
screened at 33 facilities throughout the country. The approach was to measure emissions from a
large number of randomly selected components and to determine the average emission rate (i.e.,
21
-------
emission factor) for each type of component. After the components were screened to determine
if they were leaking, the average emission rate was measured using one of several test methods:
• A high flow organic vapor analyzer that captures the entire leak and
measures the methane concentration and flow rate. The emission rate is
determined from the product of the concentration and flow rate. This
method was developed as part of this natural gas industry program to
provide a more accurate and cost-effective technique for measuring a
methane emission rate directly.
• A total enclosure technique called bagging. Uncontaminated air is blown
through an enclosure surrounding the component; the flow rate and outlet
concentration are then measured. The leak rate is determined from the
product of the concentration and flow rate.
• A screening technique in which the methane concentration is measured by
passing a standard organic vapor analyzer around the sealed surface. The
concentration is related to an emission rate by a correlation equation that
relates bagged emissions to measured screening values.
Tracer Gas Method
The tracer gas method of measuring methane emissions consists of releasing
tracer gas (at a known constant rate) near the emission source and measuring the downwind
concentrations of tracer and methane. Assuming complete mixing of the methane and tracer gas,
and assuming identical dispersion, the ratio of the downwind concentrations is equal to the ratio
of the release rates. Based upon the downwind concentrations of methane and tracer gas and the
known release rate of the tracer, the emission rate of methane can then be determined. This
method was used primarily to measure emissions from M&PR stations.12,19
22
-------
Leak Statistics Method
The leak statistics method is used to quantify methane emissions from
underground main and service pipelines." Emission rates are measured for a large number of
leaks to accurately determine the average emission rate per leak as a function of pipe material,
age, pressure, and soil characteristics. The measurement program was conducted as a
cooperative effort between EPA/GRI and industry. The industry participants used specially
designed equipment to measure leak rates from underground distribution mains and services. In
the procedure, a pipe segment containing the leak is isolated, the isolated segment repressurized,
and the volumetric flow required to maintain normal operating pressure in the isolated segment is
equal to the leak rate. Historical leak records are analyzed to determine the number of leaks per
mile for different pipe materials. Total emissions are determined by multiplying the average leak
rate per leak by the estimated total number of leaks in the distribution segment.
3.2.2 Calculation Approach for Unsteady Emissions
For some methane emission sources, such as releases during maintenance,
detailed company records are available for multiple years. However, many other sources of
unsteady emissions are not tracked by companies and, therefore, must be calculated.
Each unsteady source of emissions requires data gathering and a unique set of
equations to quantify the average annual emissions. In general, all unsteady sources of emissions
require the following information to quantify annual emissions:
• Detailed technical characterization of the source and identification of the
important parameters affecting emissions. (This information is
documented for individual source types in the reports for each major
source category.)
• Data from multiple sites that allow the methane emitted per emission event
to be calculated from the governing equations.
23
-------
• Data on the frequency of releases.
The estimate of emissions from a vessel blowdown for routine maintenance is an
example of emissions calculated for an unsteady source . In this case, the volume, pressure, and
temperature of gas contained in the vessel before blowdown is used to calculate losses from a
blowdown event. Additionally, an average frequency of these vessel blowdown events is
necessary to determine the annual loss.
In some cases, emissions per event from some unsteady sources were measured.
These emissions data were combined with site data collected in this study to quantify the annual
emissions from these sources. Examples of sources where emission measurements per event
were used include emissions from compressor driver exhaust, gas-operated pneumatic devices,
glycol dehydrator regenerator overhead vents, and gas-operated chemical injection pumps.
3.3 General Extrapolation Methodology
By necessity, data in this project were collected for a relatively small percentage
of sources in each source category. Therefore, these data had to be extrapolated to develop
national estimates for each source category. The extrapolation techniques for creating national
emission estimates were developed so that the emissions from each source could be estimated
with a relatively high level of precision (given the nature of this study) and negligible bias. (See
Section 3.4 for definitions of precision and bias.)
The extrapolation approach is a method to scale-up the average emissions from a
limited number of sources to represent the entire population of similar sources in the gas
industry. The extrapolation approach uses the concept of emission and activity factors to
estimate emissions based on a limited number of samples. These factors are defined in such a
way that the product of emission and activity factors equals the annual national emissions from
the source category.
24
-------
Emission Factor x Activity Factor = National Emissions
Typically, the emission factor (EF) for a source represents the average emissions
rate per source and the activity factor (AF) represents the total industry population of the source
category.
3.3.1 Sampling Approach
Even if the overall precision of an estimate is acceptable because the variability in
the data is relatively low, the overall accuracy may still be poor if the data are biased. Several
approaches can be applied to avoid bias.
Because of various practical limitations, neither random sampling nor stratified
random sampling was feasible in this study. For this reason, an alternate approach was used.
While this approach is not a textbook sampling method, it is believed to be very effective for the
specific needs of this project. This approach is similar to disproportionate stratified random
sampling, with certain differences. These conventional sampling techniques and the reason why
they were not applicable in this project are discussed in Volume 4 on statistical methodology.6
Initially, some data were collected to determine if a given source was a major
contributor to methane emissions. For each source category, an initial estimate of the number of
data points needed was calculated based on an estimate of the target precision and the estimated
standard deviation for the source category. The accuracy targets for precision are based on the
need to estimate the 1992 national emissions to within 0.5% of U.S. natural gas production with
a 90% confidence limit. Sites were selected in a random fashion from known lists of facilities,
such as GRI or A.G.A. member companies. However, the companies contacted were not
required to participate, and a complete list of all sources in the United States was generally not
available. Therefore, the final set of companies selected for sampling was not truly random.
Each company that agreed to participate in the program was asked to select representative sites
for sampling, rather than one-of-a-kind facilities.
25
-------
After a limited set of data was collected, the data were screened for bias by
evaluating the relationship between emission rate and parameters that may affect emissions. It is
important to realize that just because a parameter or set of strata is identified that has a large
effect on emissions from a given source category, it does not mean that there is bias in the data.
A second condition is necessary, namely, that the sampling procedure would have to produce a
disproportionate number of samples in the strata. To determine whether this has occurred,
information is needed on the ratio of the total number of sources in a given stratum to the total
number of sources throughout the country. If this ratio is different from the corresponding ratio
for the sample data set, then there may be bias. But this bias can be eliminated by applying the
correct emission factors and activity factors for the different strata.
Once the strata are identified, the precision of the emission rate extrapolated to a
national basis was evaluated and compared to the accuracy target. Where necessary, additional
data were collected in various strata to improve the precision of the national estimate of
emissions from the source. The number of additional data points needed to meet the newly
calculated accuracy target was computed based on the standard deviation and a 90% confidence
interval.
In some cases, variability of the emissions data from source to source is very
large. For source types of this nature, it is normally possible to reduce variability by redefining
the emission factor or by stratification. This is important because reducing variability reduces
the number of data points needed to achieve the accuracy target.
3.3.2 Redefining the Emission Factor
For a few types of sources, emissions can be more accurately estimated when the
emission factor is defined not as a simple average of the data but is expressed in terms of a key
parameter that influences the emissions from the source. Since this would significantly reduce
the variability, fewer data points are required to achieve the desired level of accuracy. For
example, the internal combustion engines that drive compressors in the gas industry vary in size
26
-------
(i.e., horsepower rating). If data were collected on individual engines in the industry, and an
average emission rate per engine was established, the variability from engine to engine would be
very large because of size differences. However, if the emission factor for the engines is defined
by horsepower of the engine (i.e., annual emissions per horsepower), then the variability from
engine to engine and therefore the number of samples required to reach an acceptable accuracy
are both significantly reduced.
As discussed previously, the number of data points required also may be reduced
by stratifying on the basis of parameters that affect emissions. A source type can be stratified
into categories with different emission characteristics; the objective is to produce strata with
much less variability than the total data set. The sampling is performed within the strata and
because the variability within the strata is smaller, fewer total data points are required to achieve
target precision.
3.4 Accuracy
A key part of this project is the estimation of the accuracy of the annual national
emissions. Accuracy is dependent on precision and bias, as discussed in Volume 4 on statistical
methodology.6 Precision, the random variability in the measurement, is calculated rigorously by
propagating error from each individual group of measurements into the final numbers. However,
bias, a systematic error in the measurements must be prevented or discovered and eliminated,
rather than identified and calculated.
3.4.1 Precision
Most source activity factors and emission factors are made up of an average of
multiple measurements or calculations. Therefore, assuming a normal distribution around a
mean and error independence, standard deviations and 90% confidence limits can be calculated
directly for each group of measurements in an activity or emission factor.
27
-------
The confidence intervals or error bounds can be propagated through the addition
of multiple emission source estimates to arrive at a confidence bound for the national emission
estimate. These generally accepted statistical techniques are described in detail in the statistical
methods report cited previously.
3.4.2 Bias
It is impossible to prove that there is no bias in any data set. While tests can be
designed that are capable of revealing some bias, there are no tests nor group of tests that will
reveal all possible biases. Assuming that a data set has no bias is only a hypothesis, even after
extensive testing. Such hypotheses can be disproved, but not absolutely proven. However, the
data collected during this project were extensively checked and rechecked to identify and then
eliminate biases. Three basic methods were used to screen for bias: peer review by experts,
subdivision of the data into strata, and extrapolation by different parameters. Some of these
techniques were discussed previously in Section 3.3.
Data sets were tested repeatedly through extensive technical and industrial review.
Numerous project advisor's meetings were held during the course of the study to examine the
data with industry representatives and other experts so that systematic errors could be identified
and eliminated. When biases in the sampling plan or extrapolation method were postulated, the
project was altered to test for that bias and eliminate it if it existed. One example of the success
of this review process is the identification of regional differences in production practices. These
differences were identified during the advisor meeting review process. The regional bias was
then eliminated by subdividing the production data into two offshore and four onshore regions,
collecting random samples within each region, and extrapolating by region.
3.5 Quality Assurance and Quality Control Approach
As defined during the 1980s quality initiatives, quality is conformance to
requirements.31 The programmatic quality assurance requirement of this project was to develop a
28
-------
national emission estimate of defined uncertainty and no known bias. Accordingly, the
GRI/EPA-ORD program included quality assurance and quality control (QA/QC) activities
designed to control and assess the quality of the data collected and the resultant conclusions.
Other QA/QC activities associated with the various data sources, data handling,
project review, and statistical analysis are outlined in subsequent reports associated with this
project. The report on general methodology explains the industry characterization used to
identify each emission source, the measurement techniques, and calculation approaches.5 The
statistical approach for this project is presented in the statistical methodology report.6 In
addition, the individual reports for each emission source provide detailed statements regarding
data quality efforts and uncertainties associated with the specific components that make up
each emission estimate.
3.5.1 Overview
The first step in this project's QA/QC efforts was the establishment of project
phases that had clear QC goals and that outlined QA review steps. This allowed the nature and
breadth of data collection to be modified to ensure consistent data collection with minimal bias.
The three phases of this study, and their inherent QA/QC goals were:
Scoping phase—The scoping phase included defining the boundaries of the
natural gas industry and a comprehensive characterization of all equipment in the
natural gas industry that could be a source of methane emissions. This process
minimized the potential bias of missing sources or double counting sources in
other industries, such as the oil production industry. Steps taken during this
planning process ensured that all sources of emissions were examined and that the
accuracy and bias goals of the project could be met.
Methods development phase—Based on the factors that contributed to each
emission source, methods and protocols were developed to measure and/or
29
-------
calculate each emission factor. Measurement methods were validated through
controlled experiments (laboratory), tests in the field, and proof of concept tests
designed to quantify the accuracy in the measurement approach. Methods were
also developed to extrapolate limited emission estimates to a national emission
rate, accounting for regional differences in equipment and operational practices.
The methods were peer reviewed before they were implemented.
Implementation phase—The implementation phase focused on collecting the
final field data required for emission factors and activity factors based on the
developed methodologies. QC steps were used for data collection, and QA was
performed on the data collected. Data were screened for bias and further stratified
if a relationship between the emission rate and a parameter affecting emissions
was identified. Uncertainty bounds were calculated to quantify precision and
results were compared to the target precision. Where necessary, additional data
were collected to improve the precision of the national emission estimate for a
particular source.
The following sections outline the specific QA and QC goals and methods used throughout the
project.
3.5.2 Definitions
In general, QC activities include those designed to control the data collection and
data handling efforts to ensure consistency and reliability throughout the process. The QC
activities incorporated throughout the project included:
• Proof of concept tests;
• Protocols for test methods and data collection;
• Methodology for data handling, and extrapolation; and
30
-------
• Established documentation, reporting, and filing systems.
Quality assurance activities are generally considered those that are independent of the data
gathering effort, per se. The QA activities incorporated throughout the project included:
• Quality audits;
• Industry peer review;
• Comparison to other studies; and
• Statistical analysis.
Both QC and QA steps were aimed to minimize any potential bias in the estimate. The following
subsections describe the QC and QA efforts in more detail.
3.5.3 Quality Control
The GRI/EPA study was designed from the beginning to implement standard QC
procedures, such as defined methods and protocols for data collection and handling. The most
significant QC step was the development and use of general methodologies that ensured
consistent results. During the methods development phase, a sampling plan and data gathering
protocol were developed. Most of the plans and protocols are outlined in the Volume 3
Methodology Report,5 or in the Phase 3 Program Plan.27
Emission factor measurement programs had a QC plan for measurement data
gathering that included:
• Adherence to formal protocols for data collection; and
• Sampling and analysis QC checks, including
Sample collection during representative operations,
Instrument calibration,
Analysis of blank samples,
31
-------
Analysis of known standards, and
Analysis of replicate samples.
In addition, where new measurement technologies were being applied, proof of
concept tests were performed and documented. For example, for the new distribution tracer
measurements of meter and regulation stations, QC efforts associated with emission factor
measurements are outlined in the Phase 3 Program Plan27 and in the tracer measurement field
report.19 Since the measurement technology is new to this distribution application, proof of
concept tests were performed and are reported in a separate volume.20 Similar QC efforts and
proof of concept tests exist for underground pipeline leak measurements: QC plans were
documented in the Phase 3 Program Plan and in the detailed field planning protocol32, and QC
results are documented in the Underground Pipeline Leaks report.11 Other QC efforts for
emission measurements, such as other fugitive emission efforts, are outlined in the specific field
reports cited by this project (see Table 2-1).
For activity factors, a general data collection methodology was developed that is
described in the Activity Factor Report.7 The collection of activity factor data included the
following QC efforts:
• Establishing a site visit protocol and data gathering form for each type of
site;
• Establishing a data entry protocol (for spreadsheet data entry from the site
visit forms and files);
• Validation of data entry;
• Comparison among site entries to identify unusual data; and
• Verification of unusual data.
In most cases, activity factor data were gathered directly through site visits or from published
sources. In a few instances, data were collected from efforts outside of this program and for
which no published field reports exist. For example, one production company provided their
compressor database which was used in the production activity factor estimate of
32
-------
horsepower-hours.13 In these cases, QC efforts performed by this project were limited, and QA
efforts were therefore intensified, as is described in the following section on QA.
Data on emission and activity measurements were collected and condensed
electronically, so that auditable electronic files contain all of the major data points, calculations,
and extrapolations. Many of these data are also printed in the field reports, and in the reports
comprising the 15 volumes of this set.
In addition to methodologies and QC efforts directed at activity and emission data
gathering, methodologies and QC efforts were developed for data handling and extrapolation
techniques. These are outlined in the Activity Factor report7 and the Statistical Methods report.6
3.5.4 Quality Assurance
The main goal of the QA program was to ensure the validity of the estimate
through data audits, result reviews, and statistical analysis. As with the QC steps, one of the main
goals of QA was to identify and eliminate bias. The main QA steps were audits, statistical
analysis, technical review, and comparison to other studies. Each of these are described in the
following paragraphs.
Quality audits of the databases and calculations were conducted to verify accuracy
of the mathematical applications. Audits included the following:
• Checks of the calculations made by spreadsheets. These were provided by
hand checking the results using the equations and data published in the
various reports. Also, independent calculation was performed by the
summary spreadsheet (in Appendix A). This validated the individual
emission rate and confidence bound calculations made in each report.
• Checks of conformance to known technical relations and first principles.
For example, in the QC checks of activity factor data provided on annual
operating hours for compressors, data were rejected if operating hours
exceeded 8760, the maximum number of hours in a year.
33
-------
QA audits were also performed on industry databases provided by participant
companies. In a few cases, specific company data were provided to a particular emission
estimation process, such as compressor HP-hrs for an entire company's production division, or
vented quantities from an entire transmission company's system. The data requirements were
listed in a letter to the particular company. These data were checked for completeness by the
project team, using follow-up questions to the supplier of the data, and some specific QA
requirements for the data supplied. Some supplied data that did not meet the QA/QC validation
criteria were rejected or not used. Data that were gathered in violation of typical QC controls
such as consistently following a generally accepted measurement method. For example,
pneumatic device emission rate data that did not follow the QC protocol of a single measurement
for a single device were rejected from the dataset. (Some measurements were emissions from
multiple devices; this rejected only 2 data points from a set of 43.)14 In some cases, the project
team visited the company to discuss the data. Specific data discussions are provided in the
detailed emission source reports (Volumes 6 through 158"17 in Table 2-1).
Another QA step was the use of statistical analysis, using error propagation to
define the precision and confidence in the final estimate. Uncertainty in the emission factors and
activity factors was calculated for each emission source based on the variability in the data. The
few exceptions relate to well documented data or emission sources with a very small contribution
to the overall emission estimate. Narrow confidence bounds were assigned to well-known, often
published values, where the confidence bounds were not published and the supporting data were
not available to calculate a confidence bound (e.g., the natural gas production rate published in
Gas Facts33). For source categories with a very small emission rate but unknown uncertainty,
wide confidence bounds were assigned rather than expending resources to collect additional data
for a source that had an insignificant contribution to the end result. The method for the
confidence bound is carefully documented for each value in the applicable emission
characterization reports.
34
-------
In all cases, the resulting confidence limits on the emission rate (the product of the
emission factor and the activity factor) were rigorously propagated from the confidence limits of
the activity and emission factor values. The result is that statistical analysis was very robust. The
analysis was made even more robust through analysis of potential correlation between the data,
and potential bias effects. In addition, separate tests of the input data sets were performed. For
example, outlier tests were performed on input datasets. (See the Statistical Methods6 report for
further details.) Any anomalies were verified and documented or corrected.
Another very important and unique QA step was the extensive technical review
process. All stages of this project received detailed review by an advisory panel comprised of
gas industry experts and representatives from other related industries, such as coal and oil. The
panel approved the goals and scope of the project and verified that the general results of the
project were acceptable. The advisory panel met six times during the 5-year duration of the
project to review and approve the methods and protocols. In addition, the advisory panel
reviewed the draft and final versions of the project reports.
Other industry reviewers were involved in the final stages of the project (spanning
approximately two years). These individuals, who had industry experience relating to one or
more specific project areas, reviewed emission estimates and the supporting data and
methodology to verify that the results were not biased. In addition, the reviewers provided
comments on individual reports in their areas of expertise. The involvement of these reviewers
served as a QA measure by ensuring that all emission sources were accounted for and that all
data handling methods were representative of the natural gas industry. A list of the advisors and
reviewers is included in Appendix D of this report. The reviewers met four times to examine the
detailed results and review the project team's own QA efforts that checked for:
• Representativeness—Data were analyzed to determined if the sample set
was properly stratified with respect to pertinent emission affecting
parameters and representative of the U.S. natural gas industry, including
regional differences in equipment and operating practices.
35
-------
• Technical and scientific validity—Data were reviewed for conflicting
results, for data that was inconsistent with physical possibilities, and for
results that contradicted common industry experience.
In addition to the review provided by industry experts, production activity factors
developed by this project were compared to a separate source of national equipment counts.34
EPA's Office of Air and Radiation (OAR) worked with an independent team of industry experts
to estimate production activity factors using a consensus approach. Although the EPA-OAR
results were not based on measured data, they provided an alternate method for estimating
equipment counts and provided another check for potential bias. The EPA-OAR results
compared well to the results of this GRI/EPA-ORD project.
36
-------
4.0 DETAILED RESULTS
The natural gas industry's total methane emissions are 314 Bscf for the 1992
baseline year with a 90% confidence bound of ± 105 Bscf. (See Section 4.7 for further
explanation of the confidence bound.) The total emissions can be expressed as a percent of
production: 314 Bscf is 1.4% of gross 1992 production, which is 22,130 Bscf, or 1.7% of
marketed gas production, which is 18,710 Bscf.
This section presents the detailed methane emission estimates produced by this
project. The results are presented by emission type in Section 4.1, and the methods used for
estimating emissions are briefly discussed in Sections 4.2 through 4.4. The largest sources
within each segment are discussed in Section 4.5. The emissions are also presented for different
types of equipment in Section 4.6.
4.1 Emission Type Summary
This section presents a summary of annual methane emissions by emission type.
The emission types are fugitive, vented, and combusted, as described earlier in Section 3.1.3.
Table 4-1 lists the largest sources of methane emissions in the U.S. gas industry by emission
type. Fugitive emissions are the largest (195 Bscf), followed by vented emissions (94 Bscf), then
combusted emissions (25 Bscf). Figure 4-1 shows the percentage of emissions by type for the gas
industry. The major contributors to each emission type are discussed in more detail in the
following subsections.
37
-------
TABLE 4-1. UNITED STATES NATURAL GAS INDUSTRY LARGEST METHANE
EMISSIONS SOURCES
Source
Annual Methane
Emissions (Bscf)
% of Total
Fugitive Emissions (Sec 4.2) SUBTOTAL
195.2
62.1
Equipment Leaks
Compressor Stations (transmission and storage)3
67.5
21.5
Production Facilities
17.4
5.5
Gas Plants
24.4
7.8
Metering and Pressure Regulating Stations'5
31.8
10.1
Customer Meter Sets
5.8
1.8
Underground Pipeline Leaks (all segments)
48.4
15.4
Vented Emissions (Sec 4.3) SUBTOTAL
94.2
30.0
Pneumatics2 (4.3.1)
45.7
14.6
Blow and Purge (4.3.2)
30.2
9.6
Dehydrator Glycol Pumps (4.3.3)
11.1
3.5
Dehydrator V ents (4.3.4)
4.8
1.5
Chemical Injection Pumps (4.3.5)
1.5
0.5
Other (AGR)
0.9
0.3
Combusted Emissions (Sec 4.4) SUBTOTAL
24.9
7.9
Compressor Exhaust (4.4)
24.9
7.9
TOTAL
314
100
"Includes wells at storage facilities.
bEmissions from meter and pressure regulating (M&PR) stations result from both pneumatic and fugitive
emissions. Since these components cannot be separated, M&PR emissions are shown as fugitive by default.
Combusted
8%
Vented
30%
Fugitive
62%
Figure 4-1. Emissions by Type
38
-------
4.2
Fugitive Emissions
Fugitive emissions are defined as unintentional releases that include methane
emissions from equipment leaks at sealed surfaces (component fugitive emissions), as well as
from underground pipeline leaks. Figures 4-2 and 4-3 show the major contributors to fugitive
emissions. Total fugitive emissions for the natural gas industry are 195.2 Bscf. Underground
pipeline leaks account for 48.4 Bscf of emissions, and include leaks from production gathering
lines, transmission pipelines, and distribution pipe systems. Equipment leaks account for 146.9
Bscf, and are typically low-level emissions of process fluid (gas or liquid) from the sealed
surfaces on above-ground process equipment. Specific fugitive emission source types include
various fittings such as valves, flanges, pump seals, compressor seals, or sampling connections.
These components represent mechanical joints, seals, and rotating surfaces, which in time tend to
wear and develop leaks.
Facilities and equipment that are significant contributors to equipment leak
emissions include: production facilities, gas processing plants, compressor stations/facilities in
transmission and storage, and meter and pressure regulating stations in transmission and
distribution. The following subsections describe each of the major fugitive emission sources in
more detail.
Customer Meter Seto UNDERGROUND
Gas Piants 3% PIPELINES
13%
25%
Production Facilities
9%
M&PR Stations
16%
Compressor Stations
(T&S)
EQUIPMENT LEAKS 34o/o
75%
Figure 4-2. Major Contributors to Fugitive Emissions From the Natural Gas Industry
39
-------
Total
195
Equipment
Leaks
147
Underground
Pipeline
Leaks
48.4
M&PR
31.8
Production
17.4
Gas
Plants
24.4
Compressor
Stations
(Transmission
& Storage)
67.5
Customer Meters 5.8
15.2
22.4
Compressors
57.5
9.9
Compressors 2.2
Facility
Compressors
Facility 2.1
Facilities (wells, stations)
Figure 4-3. Major Contributors to Fugitive Emissions - By Segment Facilities
-------
4.2.1
Equipment Leaks
Fugitive emissions from equipment leaks in the natural gas industry were
estimated to be 146.9 Bscf. Of this total, 82.1 Bscf was attributed to compressors, 31.8 Bscf to
meter and pressure regulating stations, 5.8 Bscf from customer meter sets, and 27.2 Bscf from
other surface facilities. Other surface facilities are the non-compressor portion of production
facilities, gas plants, and transmission and storage stations.
There are two general approaches for estimating fugitive methane emissions from
equipment leaks: the tracer gas method and the component method. Tracer tests are conducted
by releasing a tracer gas such as SF6 at a known constant rate near the methane emissions source.
The concentration of methane and tracer are then measured downwind. The methane emissions
are calculated based on the relationship that the ratio of emissions is equal to the ratio of
concentrations. The tracer method measures total emissions from the facility, and was used to
measure emissions from metering and pressure regulating (M&PR) stations. The tracer method
for M&PR stations is described in more detail in Volume 10.12 The component method is
described in more detail in Volume 8 on equipment leaks.10 Both techniques are described in the
following subsections.
In the component method for estimating emissions from equipment leaks, an
average emission rate is determined for each of the basic components, such as valves, flanges,
seals, and other connectors that comprise a facility. The average emission rate for each type
of component is determined by measuring the emission rate from a large number of randomly
selected components from similar types of facilities throughout the country. By knowing the
average emission rate per component type (i.e., the component emission factor) and the
average number of components associated with the major equipment or facility, an estimate of
the average emissions per equipment/facility can be determined. Extrapolation to a national
emission estimate can then be made by determining the total count of that specific
equipment/facility in the United States.
41
-------
The component approach was used to estimate fugitive emissions from gas
production facilities, processing plants, transmission/storage facilities, and customer
meters.21,24'26'35 Separate component emission factors were developed for each industry segment
because of differences in design and operating practices that could lead to differences in
emissions characteristics. Some regional differences were also determined to have an impact
on fugitive emissions; therefore, regional component emission factors were developed. (That
is, regional component emission factors were developed for onshore and offshore production.)
For gas processing, transmission, and storage, separate emission factors were
developed for components physically connected to, or directly adjacent to, compressors.10,35
These compressor-related components were found to have significantly higher emission rates
than components associated with other equipment. The higher emission rate from compressor-
related components is due to the unique design, size, and operation, as well as from the
vibrational wear associated with compressors. For gas processing, transmission, and storage
facilities, emissions were calculated as a sum of compressor-related components and station
(non-compressor related) components. Table 4-2 presents an example of the calculational
approach used to calculate fugitive emissions using the component method.
Two approaches were used to quantify the component emission factors for
valves, flanges, seals, and other connectors. The first approach is based on the EPA protocol
document using EPA Reference Method 21.36 The EPA protocol approach involves screening
components using a portable instrument to detect total hydrocarbon (THC) leaks. The
corresponding screening value for a component, which is a concentration measurement, is then
converted to an emission rate by using a correlation equation developed from data collected
using an enclosure measurement method. The enclosure method allows the actual leakage rate
to be measured as the product of the flow rate of inert gas through the enclosure and the THC
concentration. The correlation equation is developed by correlating the screening or
concentration data with the emission rate data measured using the enclosure method. The
correlation equation can then be applied to the same component type in similar service within
the gas industry to estimate emissions using only screening data. The EPA protocol approach
42
-------
TABLE 4-2. EXAMPLE OF NATIONAL EMISSIONS ESTIMATION FOR GAS STORAGE FACILITIES
Equipment Type
Component Type
Component Emission
Factor,
Mscf/component-yr
Average
Component
Count
Average
Equipment
Emissions,
MMscf/yr
Activity Factor,
Number of
Plants/Compressors
National
Methane
Emissions,
Bscf
Storage Facility (non-
compressor related
components)
Valve
0.867
1868
7.85
475
3.7
Connection
0.147
5571
Open-Ended Line
11.2
353
Pressure Relief Valve
6.2
66
Site Blowdown Open-Ended
Line
264
4
30
Injection/W ithdrawal
Wellhead
Valve
0.918
0.042
17,999
0.75
Connection
0.125
89
Open-Ended Line
0.237
7
1
Pressure Relief Valve
1.464
Reciprocating Compressors
Compressor Blowdown
Open-Ended Line
5024
1
7.71
1,396
10.8
Pressure Relief Valve
317
1
Miscellaneous
153
1
Compressor Starter Open-
Ended Line
1440
0.6
Compressor Seal
300
4.5
Centrifugal Compressors
Compressor Blowdown
Open-Ended Line
10233
1
11.16
136
1.5
Miscellaneous
17
1
Compressor Starter Open-
Ended Line
1440
126
0.5
Compressor Seal
1.5
-------
was used to quantify emissions from equipment leaks in onshore production (except for
production facilities in the Atlantic and Great Lakes region), offshore production, and gas
processing.
The second approach used to quantify component emission factors modifies the
EPA protocol approach by using the GRI Hi-Flow™ sampler and direct measurements to
replace the data collected using an enclosure approach. The GRI Hi-Flow™ sampler is a
newly developed device which allows the leak rate of a component to be measured directly.
The sampler creates a flow field around the component in order to capture the entire leak. As
the stream passes through the instrument, the flow rate and concentration are measured. The
GRI Hi-Flow™ sampling approach was used to quantify emissions from equipment leaks in
onshore production in the Atlantic and Great Lakes region, gas transmission and storage, and
customer meters. Direct measurements, such as rotameter readings, were also used on very
high leak rates from open-ended lines at transmission and storage compressor stations.
The following subsections explain how fugitive emissions were calculated for
each of the facility types that were significant contributors to total national emissions.
Compressor Stations (Transmission and Storage)
Compressor stations in transmission and storage are one of the largest sources
of fugitive emissions. Equipment leaks from transmission compressor stations were separated
into two distinct categories because of differences in leakage characteristics:
• Station components including all sources associated with the station inlet
and outlet pipelines, meter runs, dehydrators, and other piping located
outside of the compressor building; and
• Compressor-related components including all sources physically
connected to or immediately adjacent to the compressors. The types of
components associated with compressors include compressor blowdown
44
-------
open-ended lines, starter open-ended lines, compressor seals, pressure
relief valves, and other components such as cylinder valve covers and
fuel valves.
Fugitive emissions from compressor stations are dominated by emissions from
components related to compressors, which emit 57.5 Bscf, while emissions from all of the
remaining components not associated with compressors contribute only 9.9 Bscf.
Fugitive emissions were estimated from measurement data collected at 15
compressor stations using the GRI Hi-Flow™ approach.24 Leaking components were
identified using soaping tests and all leaking components were directly measured using the GRI
Hi-Flow™ sampler or a direct flow measurement, such as a rotameter. Based on the
measurement data, fugitive emissions from the compressor blowdown open-ended line were
found to be the largest source. Compressor blowdown open-ended lines allow a compressor to
be depressurized when idle, and typically leak when the compressor is operating or idle.
There are two primary modes of operation leading to different emission rates for compressor
blowdown open-ended lines:
• Blowdown valve is closed and the compressor is pressurized, either
during normal operation or when idle.
• Compressor blowdown valve is open. This occurs when the compressor
is idle, isolated from the compressor suction and discharge manifolds,
and the blowdown valve is opened to depressurize the compressor.
The fugitive emission rate is higher for the second operating mode when the
blowdown valve is open, since leakage occurs from the valve seats of the much larger suction
and discharge valves. Separate component emission factors were developed for the two
operating modes of the compressor blowdown open-ended line. An overall average
component emission factor was derived for compressor blowdown open-ended lines by
45
-------
determining the fraction of time transmission compressors operate in each mode (i.e.,
pressurized and depressurized).
The majority of compressor fugitive emissions result from the transmission and
storage segments, where a high number of very large compressors exist. Since compressors
are also a part of production facilities and gas plants, the compressor component emission
factors developed for the transmission and storage segments were also used for compressor
components in those segments.
Production Facilities
Annual fugitive emissions from gas production facilities in the United States
were estimated to be 17.4 Bscf. Component emission factors for fugitive equipment leaks in
gas production were estimated separately for onshore and offshore production due to
differences in operational characteristics. Regional differences were found to exist between
onshore production in the Atlantic and Great Lakes region (i.e., Eastern U.S.) and the rest of
the country (i.e., Western U.S.), and between offshore production in the Gulf of Mexico and
the Pacific Outer Continental Shelf (OCS). In general, these regional differences were due to
differences in the number, type, age, and leak detection and repair characteristics of
equipment. Therefore, separate measurement programs were conducted to account for these
regional differences.
For onshore production in the Eastern U.S., component emission factors and
average component counts were based on a measurement program using the GRI Hi-Flow™
sampler to quantitate emission rates from leaking components.22 A total of 192 individual well
sites were screened at 12 eastern gas production facilities.
Fugitive emissions from onshore production in the rest of the U.S. (excluding
the Eastern U.S.) were estimated using the EPA protocol approach. Component emission
46
-------
factors were based on screening and enclosure data collected from 83 gas wells at 4 gas
production sites in the Western U.S.21 The average component counts were based on data
from the onshore production measurement program and additional data collected during 13 site
visits to gas production fields.10
Emissions from equipment leaks from offshore production sites in the U.S.
were quantified based on two separate screening and enclosure studies using the EPA protocol
approach:
• The oil and natural gas production operations measurement program,21
which included 4 offshore production sites in the Gulf of Mexico; and
• The offshore production measurement program,37 which included 7
offshore production sites in the Pacific OCS.
Gas Processing Plants
Fugitive emissions from gas processing plants contribute 24.4 Bscf to national
annual methane emissions. The majority of fugitive emissions from gas processing plants are
attributed to compressor-related components, which account for 22.4 Bscf. The component
emission factors for compressor-related components in gas processing plants were based on the
fugitives measurement program at 15 compressor stations.10 Fugitive emissions from the
remaining gas plant components, not associated with compressors, were estimated based on the
oil and gas production measurement program.21 In the oil and gas production measurement
program, equipment leaks from a total of 8 gas processing plants were measured using EPA
protocol approach.
47
-------
Meter and Pressure Regulating Stations
Fugitive emissions from meter and pressure regulating stations (M&PR stations)
contribute 31.8 Bscf to total annual methane emissions. Emissions from this category of surface
equipment were measured using the tracer measurement approach, and therefore were reported
separately from other categories of surface equipment fugitives. A total of 95 M&PR facilities
were measured using the tracer technique.12
The primary losses from M&PR stations include both fugitive emissions and, in
some cases, emissions from pneumatic devices. Since the tracer measurement technique used
does not differentiate between fugitive and vented emissions, the vented pneumatic emissions are
therefore included in the fugitive category by default. Some pressure regulating stations use gas-
operated pneumatic devices to position the pressure regulators. These gas-operated pneumatic
devices bleed to the atmosphere continuously and/or when the regulator is activated for some
system designs. Other designs bleed the gas downstream into the lower pressure pipeline and,
therefore, have no losses associated with the pneumatic devices.
Tracer measurements were used to derive the emission factors for estimating
emissions from M&PR stations in both the transmission and distribution segments of the gas
industry. The total emissions are a product of the emission factor and activity factor, which were
stratified into inlet pressure and location (above ground versus in a vault) categories to improve
the precision of the emissions estimate.
Metering/pressure regulating stations in the distribution segment include both
transmission-to-distribution custody transfer points and the downstream pressure reduction
stations. The emission factors for distribution are based on the average measured emissions for
each station category, and the activity factors are based on the average data supplied by 12
distribution companies. The annual methane emissions for the M&PR stations in the distribution
segment of the gas industry are 27.3 Bscf.
48
-------
For the transmission segment, the stations include transmission to transmission
custody transfer points and transmission-to-customer transfer. Emission factors for the
transmission segment are derived from the tracer measurement database for M&PR stations, and
the activity factors are based on survey data from six transmission companies. The annual
estimated methane emissions for the transmission segment are 4.5 Bscf.
Customer Meter Sets
Fugitive emissions from commercial/industrial and residential customer meter
sets contribute 5.8 Bscf to total national emissions. The average leak rate per residential meter
set is only 0.01 scf/hr, but there are approximately 40 million customer meters located
outdoors. The meter sets include the meter itself and the related pipe and fittings. Methane
emissions from commercial and residential customer meter sets are caused by fugitive losses
from the connections and other fittings surrounding the meter set. No losses have been found
from the meter itself; only the pipe fittings surrounding the meter have been found to be
leaking.
Methane emissions from customer meter sets were estimated based on fugitives
screening data collected from 10 cities across the United States.10,24 26 Although a total of
around 1600 meter sets were screened as part of the GRI/EPA study, only about 20% of the
meter sets screened were found to be leaking at low levels. For the majority of customer
meter sets screened, the GRI Hi-Flow device was used to develop emission factors. For the
other meter sets screened, the EPA protocol approach was used to convert the screening data
into emission rates.
Emission factors for residential customer meter sets were defined as the average
methane leakage rate per meter set for outdoor meters. Emissions from indoor meters are
much lower than for outdoor meters because gas leaks within the confined space of a residence
are readily identified and repaired. This is consistent with the findings that pressure regulating
49
-------
stations located in vaults have substantially lower emissions than stations located above
ground. Emission factors for commercial/industrial meter sets were estimated separately as
the average emission rate per meter set.
The activity factors for residential customer meter sets were defined as the
number of outdoor customer meters in the United States. The activity factor was based on
published statistics including a breakdown of residential customer meters by region in order to
estimate the number of meter sets located indoors. Data were obtained from 22 individual gas
companies within different regions of the United States to estimate the number of indoor
residential customer meters.
4.2.2 Underground Pipeline Leaks
Fugitive leakage from underground piping systems contributes 48.4 Bscf to total
methane emissions. Pipeline leaks are caused by corrosion, material defects, and joint and fitting
defects/failures. Based on limited leak measurement data from two distribution companies,
leakage from underground distribution mains and services was targeted as a potentially large
source of methane emissions from the gas industry.
A leak measurement technique was developed (Section 3.2.1) and was
implemented as a method to quantify methane emissions from underground pipelines in the
natural gas industry.11 A total of 146 leak measurements were collected from the participating
companies. These data were used to derive the emission factors for estimating methane leakage
from distribution, transmission, and production underground pipelines.
The total emissions are a product of the emission factor and activity factor, and
are stratified by pipe use (mains versus services) and pipe material categories to improve the
precision of the estimate. The total annual methane emissions from underground pipeline leaks
in all segments are 48.4 Bscf.
50
-------
The soil oxidation rates of methane were experimentally determined to be a
function of the methane emissions rate, pipe depth, and soil temperature. The methane leakage
rate for underground pipelines was determined to be a function of the pipe service (main versus
services) and the pipe material type. In general, the larger the leakage rate per leak, the lower the
soil oxidation rate. Because of the type of pipelines in service in the distribution segment, the
overall leakage rate per peak is lower. Therefore, the overall oxidation rates for distribution
pipelines is higher than for transmission or gathering lines.
In the distribution segment, activity factors were based on the national database of
leak repairs broken down by pipe material using information from ten companies, and then
combined with historical leak records provided by six companies. The activity factors represent
the number of equivalent leaks that are continuously leaking year round. (Repaired leaks are
counted as fractional leaks.)
The activity factor combined with the emission factors derived from the leak
measurement data produced an overall methane emissions estimate of 41.6 Bscf, which includes
an adjustment for soil oxidation. The largest contributor to the overall annual emissions was cast
iron mains, followed by unprotected steel services and mains. The average soil oxidation rate
applicable to distribution piping was 18%, which primarily affects the emissions from cast iron
mains, which have low leak rates per leak.
In the transmission and production segments, the estimated methane leakage was
based on the emission factors derived from the leak rates measured on distribution mains and on
activity factors derived from a nationally tracked database of pipe mileage/leak repairs. For
transmission pipeline leakage, the estimated annual methane emissions were 0.2 Bscf, which
includes an adjustment for soil oxidation.
51
-------
For gathering pipeline in the production segment, the estimated annual methane
emissions were 6.6 Bscf. The estimated methane emissions to the atmosphere from gathering
lines includes an adjustment of 5% average methane oxidation in the soil.
4.3
Vented Emissions
Vented emissions primarily result from three categories: 1) pneumatic devices,
2) blow and purge emissions, and 3) dehydrator emissions. Emissions from chemical injection
pumps is a minor category. Figure 4-4 shows each of the contributions to vented emissions.
Each of these are described in more detail in the following sections.
Dehy. Glycol Pumps
12%
Dehy. Vents
5%
Blow and Purge
32%
Other
3%
Pneumatics
48%
Figure 4-4. Contributions to Vented Emissions
4.3.1 Pneumatic Devices
Pneumatic devices in the natural gas industry are valve actuators and controllers
that use natural gas pressure as the force for valve movement. Gas from the valve actuator is
vented during every valve stroke, and gas may bleed continuously from the valve controller pilot
52
-------
as well. Pneumatic devices are a major source of unsteady emissions and account for 45.7 Bscf
of methane emissions.14 Methane emissions from pneumatic devices were calculated based on
field measurements, site data, and manufacturers' data.
There are two primary types of these devices: 1) control valves that regulate flow,
and 2) isolation valves that block or isolate equipment and pipelines. Of the two main types,
isolation valves typically have lower annual emission rates, although the emission rate per
actuation can be large. This is because isolation valves are moved infrequently for emergency or
maintenance activities that require isolating a piece of equipment or section of pipeline.
Alternatively, control valves typically move frequently to make adjustments for changes in
process conditions, and some types of control valves bleed gas continuously.
Emission factor estimates for pneumatic devices were based on a combination of
site information, manufacturers' data, and measured emissions from devices in the field. Each
segment of the industry has very different practices regarding the use of pneumatic devices.
These differences and a summary of the data collected to characterize the different pneumatic
devices are described below.
Production
The production segment accounts for the majority of the pneumatic emissions:
31.4 Bscf, or 69% of all pneumatic emissions. High pressure natural gas is used to operate most
of these devices, since production facilities are usually located at remote sites. Natural gas is
readily available and less expensive than compressed air or electricity at the remote sites. The
majority of devices are used to regulate flow and can emit methane either on a continuous basis
or only when the device actuates. Data were collected from 22 sites to determine the fraction of
continuous bleed devices versus intermittent bleed devices. A total of 44 measurements of
various device types in field operation were used to estimate the emission factor. In addition, the
four most common manufacturers of these devices were contacted for information regarding the
53
-------
characteristics of the devices that affect emissions. The total number of pneumatic devices in the
production segment were determined based on data from more than 35 sites.
Gas Processing
Pneumatic device emissions from the gas processing segment are very small: 0.1
Bscf, or less than 1% of all pneumatic emissions. Emissions were based on data collected from
nine gas processing plants and from the four manufacturers of the devices observed. Of the gas
processing plants surveyed, only one-half (56%) use natural gas to operate pneumatic controllers
and isolation valves. (Other sites use compressed air or electric motors.) The natural gas
powered isolation valves in this industry segment are operated infrequently (once per month or
once per year), so the emissions per site are relatively small.
Transmission/Storage
Emissions from pneumatic devices at transmission compression stations and
storage stations account for 14.1 Bscf, or 31% of pneumatic emissions. In this industry segment,
most of the pneumatics are gas-actuated isolation valves. Data for these types of devices were
provided by 16 sites and two manufacturers. There are a few pneumatic control valves used to
reduce pressure or to control liquid flow from a separator or scrubber. Emissions for these
devices were based on information collected from 54 sites and 23 measurements of operating
devices. Site data from 54 stations were also used to determine the number of devices per
station, which was extrapolated to a national number of pneumatic devices in the transmission
segment.
54
-------
Distribution
Pneumatic emissions for the distribution segment are included in the "fugitive"
emission factor for M&PR stations. The M&PR pneumatics cannot be separated from fugitives,
since M&PR total emissions were measured using the downwind tracer technique.
4.3.2 Blow and Purge
Blow and purge is a major source of unsteady emissions and accounts for
approximately 30.2 Bscf of methane emissions.9 Blow (or blowdown) gas refers to intentional
and unintentional venting of gas for maintenance, routine operations, or emergency conditions.
A piece of process equipment or an entire site is isolated from other gas containing equipment
and depressured to the atmosphere. The gas is discharged to the atmosphere for one of the
following reasons:
1) Maintenance Blowdown - The gas is vented from equipment to eliminate
the flammable material inside the equipment, thus providing a safer
working environment for personnel that service the equipment or enter the
equipment.
2) Emergency Blowdown - The gas is vented from a site to eliminate a
potential fuel source. For example, if an equipment fire begins at a
compressor station, the station emergency shutdown and emergency
blowdown system blocks the station away from the pipelines and
discharges the gas inside the station, thus reducing the fuel that could feed
the fire.
The factors that affect the volume of methane blowdown released to the atmosphere are:
frequency, volume of gas blowdown per event, and the disposition of the blowdown gas.
Blowdown from maintenance releases were determined by equipment category:
compressor blowdown, compressor starts, pipeline blowdown, vessel blowdown, gas wellbore
55
-------
blowdown, and miscellaneous equipment blowdowns. Emergency blowdowns refer to the
unexpected release of gas by a safety device, such as a pressure relief valve (PRV), on a vessel or
the automatic shutdown/emergency blowdown of a transmission compressor station. Dig-ins,
pipeline ruptures caused by unintentional damage, were also classified under emergency release
of gas and included in the blow and purge estimates.
Emission estimates for each industry segment were based on data from site visits
or company tracked data. Blow and purge emissions from the production segment, accounting
for approximately 6.5 Bscf of the total blow and purge emissions, were based on data from 25
sites. Emissions for transmission and gas processing plants, which have similar station
blowdown practices, were based on data from eight companies. These industry segments
account for 18.5 Bscf and 2.9 Bscf of the total blow and purge emissions, respectively. The
distribution segment makes up about 2.2 Bscf of the total blow and purge emissions, and the
emission estimate for this segment was based on detailed unaccounted-for gas studies from two
distribution companies.
4.3.3 Dehydrator Glycol Pumps
Glycol dehydrator circulation pumps are a major source of unsteady emissions
and account for 11.1 Bscf of methane emissions.17 These pumps use the high pressure of the rich
glycol from the absorber to power pistons that pump the low pressure, lean glycol from the
regenerator. The pump configuration pulls additional gas from the absorber along with the rich
glycol (more gas than would flow with the rich glycol if conventional electrical pumps and level
control were used). This gas is emitted through the dehydrator vent stack along with the methane
absorbed in the rich glycol stream (see Section 4.3.4).
Gas-powered glycol circulation pumps are common throughout the industry, even
at sites where electrical pumps are the standard for other equipment. The dehydrator equipment
is often specified as a separate bid package, and the vendors most often use the Kimray gas pump
56
-------
as their standard pumping unit. The pumps are an integral part of the glycol dehydrator unit and
their emissions occur through the same point. However, the pumps are the cause for nearly half
of the methane emissions from dehydrators, so they are considered separately.
Unlike chemical injection pumps which vent the driving gas directly to the
atmosphere, dehydrator pumps pass the driving gas along with the rich (wet) glycol to the
reboiler. Therefore, methane emissions from the pump depend on the design of the dehydrator,
since gas recovery on the dehydrator will also recover gas from the pump. The demographics
generated for the glycol dehydrator control system (flash drum recovery and vent vapor
recovery) were also used to determine the net emission rate for glycol pumps. Design data from
Kimray were used to establish the amount of gas used by these pumps. Gas-assisted glycol
pumps were found almost exclusively in production dehydrators, with a few in gas processing.
No active gas-assisted pumps were found during the site visits to transmission or storage
facilities, which is consistent with the fact that larger facilities tend to have electricity available.
4.3.4 Dehydrator Vents
Glycol dehydrator vents are a major source of methane emissions and account for
4.8 Bscf of methane emissions.17 The majority of the glycol dehydrators are located in
production, but dehydrators are also used in gas processing, transmission, and storage. Methane
emissions are highest in the production segment; 71% of the total dehydrator vent emissions are
attributed to dehydrators in the production segment. This is due to the high activity and emission
factors for this segment. The absence of flash tanks in most production dehydrators leads to an
emission rate per volume of gas dehydrated that is higher in production than in the other
segments.
Glycol dehydrators remove water from the natural gas through continuous glycol
absorption. The water-rich glycol is regenerated, or heated, which drives the water back out of
the glycol. The glycol also absorbs some other compounds from the gas, including a small
57
-------
amount of methane. The methane is driven off with the water in the regenerator and vented to
the atmosphere.
The important emission-affecting variables for dehydrators are: gas throughput,
use of a flash tank, use of stripping gas, and use of vent controls where the gas is routed to a
burner. An emission factor per unit of gas throughput was established for glycol dehydrator
regenerator vents using three sources of data: 1) computer simulations of dehydrator operations
using first principles; 2) data from actual samples taken from regenerator vents; and 3) multiple
site visits. The emission factor was combined with an activity factor to generate the emission
rate. The activity factors are the volumes of gas dehydrated in each industry segment. The total
glycol dehydrator throughput compares well with a separate study conducted by API.38
4.3.5 Chemical Injection Pumps
Chemical injection pumps are a source of unsteady emissions and account for
1.5 Bscf of methane emissions solely in the production segment.15 Emission estimates for this
source were based on data from 17 sites, 6 manufacturers, and emission measurements from a
Canadian study.39 The total number of chemical injection pumps nationally was extrapolated
from data relating the number of chemical injection pumps to the number of gas wells at 38 sites.
Gas-driven chemical injection pumps use gas pressure to move a piston which
pumps the chemical on the opposite end of the piston shaft; the power gas is then vented to the
atmosphere at the end of the stroke. The power gas may be natural gas or compressed air. Two
types of chemical injection pumps were observed: 1) piston pumps, and 2) diaphragm pumps.
The larger diaphragm pumps emit more gas per stroke, and they are used to pump a higher flow
rate of chemical or to pump the chemical into high pressure equipment.
Chemical injection pumps are used to add chemicals such as corrosion inhibitors,
scale inhibitors, biocides, demulsifiers, clarifiers, and hydrate inhibitors to operating equipment.
58
-------
These additives protect the equipment or help maintain the flow of gas. The vast majority of
these pumps exist in the production segment where the gas is wet and has a high non-methane
content. The pumps are most often located at the well sites, so that the chemical can protect all
of the downstream and downhole equipment. Most of the chemical injection pumps in oil and
gas production are associated with oil production and were not included in this study. As with
pneumatic control valves, the chemical injection pumps in production are primarily powered by
natural gas.15
In the production segment, significant regional differences exist. Depending on
the gas composition and conditions, some regions use very few pumps, while other regions use
the pumps frequently. Many pumps also have seasonal operation since they protect against
hydrate formation, which winter temperatures exacerbate.
Only a few pumps exist in the gas processing and transmission segments. The
pumps that do exist are powered by compressed air at these stations, and as a result, have no
methane emissions.
4.4 Combusted Emissions
Combusted emissions result from incomplete combustion of methane in burners,
flares, and engines. Incomplete combustion of methane in compressor engine exhaust is the only
significant source of methane in this category.
Methane emitted to the atmosphere in compressor driver exhaust is a major source
of unsteady emissions and accounts for 24.9 Bscf of methane emissions.13 Methane emissions
result from the incomplete combustion of the natural gas fuel, which allows some of the methane
in the fuel to exit in the exhaust stream. There are two primary types of compressor drivers: 1)
reciprocating gas engines, and 2) gas turbine drivers. A few compressors in the industry are
driven by other means such as electrical motors, but the majority are natural gas fueled. In
59
-------
addition to compressors, there are some natural gas drivers that run electrical generators at gas
plants and compressor stations.
Reciprocating engines emit approximately 40 times more methane per horsepower
or per unit of fuel consumed than gas turbine drivers. Reciprocating engines account for over
two-thirds of all installed horsepower in the gas industry. Therefore, reciprocating engine
compressor drivers account for over 98% of the methane emissions for this category.
Emissions were determined by analyzing and combining several databases to
generate emission factors and activity factors. A GRI database, the TRANSDAT compressor
module,40 contains data from A.G. A. on types and models of compressors in use, as well as data
on compressor driver exhaust from the Southwest Research Institute (SwRI). A.G.A. gathers its
data from government agencies, such as the U.S. Department of Energy (DOE) and the Federal
Energy Regulatory Commission (FERC), and from surveys of its member companies in
transmission and distribution. SwRI data were generated through actual field testing. These data
were combined to generate emission factors for this project by correlating compressor driver
type, methane emissions, fuel use rate, and annual operating hours for 775 reciprocating engines
and 86 gas turbines.
Horsepower*hour activity factors were developed for each industry segment using
data from GRI TRANSDAT, FERC, A.G.A., company databases, and site visits. GRI
TRANSDAT includes horsepower data for 7489 reciprocating engines and 793 gas turbines in
transmission. Transmission operating hours were based on FERC data for 1992 and one
company's data for 524 reciprocating engines and 89 gas turbines. Storage horsepower and
operating hours were based on A.G.A. data and data from 11 storage stations, respectively.
Since national totals for transmission and storage horsepower were available, no industry
extrapolation was necessary for these activity factors. Production horsepower*hours were based
on one company's data for 513 reciprocating engines and 6 gas turbines. Processing horsepower
and operating hours were based on 10 site visits and company data for 11 gas processing plants.
60
-------
Activity factors for production and processing were extrapolated to the industry using published
data for national marketed gas production and gas processing, respectively.
4.5 Largest Sources by Industry Segment
This section summarizes the segment emissions and presents the data by largest
emission categories within each segment. Table 4-3 presents a summary of emissions by gas
industry sector. Figure 4-5 shows the same data in a chart format.
TABLE 4-3. SUMMARY OF METHANE EMISSIONS
Segment
Emissions (Bscf)
Percent of Total
Emissions (%)
Emissions as a Percent of
Gas Produced (Gross
National Product)
Production
84.4
26.8
0.38
Processing
36.4
11.6
0.16
Transmission/Storage
116.5
37.1
0.53
Distribution
77.0
24.5
0.35
TOTAL
314 ±105
100.0
1.42
*Gross national production of natural gas = 22,132 Bscf (22.13 Tscf)41
(Accuracy Goal is ± 110.7 Bscf or ± 0.5% of production)
Distribution
24%
Production
27%
Processing
12%
Transmission/
Storage
37%
Figure 4-5. Summary of Methane Emissions
61
-------
The total segment emissions presented in Table 4-3 and Figure 4-5 are split into
emission type in Table 4-4. The largest emission type for the entire U.S. natural gas industry is
fugitive emissions; however, the largest emission category in each segment varies. Vented
emissions are the largest emission category in production because of the contribution from
pneumatic devices. In the other segments of the industry, fugitive emissions are the largest
source.
Segment emissions also can be broken down into the largest categories that were
presented in Table 4-1, U.S. Natural Gas Industry Largest Methane Emission Sources. These
categories are actually a mixture of emission types and equipment types, since some
measurement programs were specific to a type of equipment (such as the buried pipeline leak
statistics method), while others were not.
Since the characteristics of each segment of the natural gas industry are quite
unique, and since companies within each segment will want to know their segment's emissions,
the data have been recast by segment. Tables 4-5 through 4-8 show the largest sources within
each segment. Figures 4-6 through 4-9 show the same data in chart format.
Table 4-5 shows that the largest sources in production were pneumatic devices
and fugitive emissions. Table 4-6 shows that the largest sources in gas plants are fugitive
emissions and compressor driver exhaust. Table 4-7 shows that the largest sources in
transmission and storage are fugitives, pneumatic devices, blow and purge, and compressor
driver exhaust. Table 4-8 shows that the largest sources in distribution are M&PR stations and
underground pipeline leaks. There are nine categories (rows) on Tables 4-7 through 4-8 that
exceed 10 Bscf, and four of these are in the transmission segment.
62
-------
TABLE 4-4. EMISSIONS BY TYPE
Emission Type
Production
Processing
Segment
Transmission
and Storage
Segment (Bscf)
Distribution
Segment
(Bscf)
Natural Gas
Industry
Emissions
Emission Type as
Percent of Total
(%)
Segment (Bscf)
(Bscf)
(Bscf)
Fugitive
24.0
24.4
72.1
74.7
195.2
62.1
Vented
53.8
5.1
33.0
2.2
94.2
30.0
Combusted
6.6
6.9
11.4
N/A
24.9
7.9
TOTAL*
84.4
36.4
116.5
77.0
314
100%
* Individual categories may not sum exactly to totals shown due to roundoff errors.
o\
u>
-------
TABLE 4-5. PRODUCTION SEGMENT LARGEST SOURCES
Source
Annual Methane
Emissions (Bscf)
% of Segment Total
Pneumatic Devices
31.4
37.2
Fugitive Emissions2
17.4
20.6
Underground Pipeline Leaks
6.6
7.8
Blow and Purge
6.5
7.8
Compressor Driver Exhaust
6.6
7.8
Glycol Dehydrator Pumps
11.0
13.0
Glycol Dehydrator Vent
3.4
4.0
Chemical Injection Pumps
1.5
1.8
Other
<0.1
<0.1
TOTAL
84.4
100
"Excludes underground pipeline leaks.
Dehydrator Vents and Other
Pumps < 2% Fugitive Emissions
21%
17%
Underground P/L
Leaks
8%
Blow and Purge
8%
Compressor
Driver Exhaust
Pneumatic Devices
37%
Figure 4-6. Production Segment Largest Sources
64
-------
TABLE 4-6. GAS PROCESSING SEGMENT LARGEST SOURCES
Source
•• •¦ • :
Annual Methane
Emissions (Bscf)
% of Segment Total
Fugitive Emissions
24.4
67.1
Compressor Driver Exhaust
6.9
18.8
Blow and Purge
2.9
8.1
Other
0.9
2.6
Glycol Dehydrator Vent
1.0
2.9
Glycol Dehydrator Pumps
0.2
0.5
TOTAL*
36.4
100
*Individual categories may not sum exactly to total shown due to roundoff errors.
Blow and Purge
8%
Compressor Driver
Exhaust
19%
Fugitive Emissions
67%
Figure 4-7. Gas Processing Segment Largest Sources
65
-------
TABLE 4-7. TRANSMISSION AND STORAGE SEGMENT LARGEST SOURCES
Source
Annual Methane
Emissions (Bscf)
% of Segment Total
Fugitive Emissions"
67.5
57.9
Blow and Purge
18.5
15.9
Pneumatic Devices
14.1
12.1
Compressor Driver Exhaust
11.4
9.8
M&PR Stations
4.5
3.9
Glycol Dehydrator Vent
0.3
0.3
Underground Pipeline Leaks
0.2
0.1
Glycol Dehydrator Pumps
0.0
0.0
Other
0,
0.0
TOTAL
116.5
100
aExcludes underground pipeline leaks and M&PR leaks.
Compressor Driver
Exhaust
10%
Other
4%
Blow and Purge
16%
Pneumatic Devices
12%
Fugitive Emissions
58%
Figure 4-8. Transmission and Storage Largest Sources
66
-------
TABLE 4-8. DISTRIBUTION SEGMENT LARGEST SOURCES
/Source |
Annual Methane
Emissions (Bscf)
% of Segment Total
Underground Pipeline Leaks
41.6
54.1
Meter and Pressure Regulating Stations !
(includes fugitive and pneumatic device emissions) j
27.3
35.5
Customer Meters
5.8
7.5
Other
2.2
2.9
TOTAL*
77.0
100
* Individual sources may not sum exactly to total shown due to roundoff errors.
Customer Meters
7%
Other
3%
Underground P/L
Leaks
54%
M/R Stations
36%
Figure 4-9. Distribution Largest Sources
67
-------
4.6
Equipment Emissions
The data presented in Sections 4.1 through 4.4 and in the Summary Table in
Appendix A are grouped by emission source or emission category. An alternate method for
grouping the emissions is by equipment type. Since some companies may wish to use the
methane emissions data to make decisions on equipment choices, it is important to know all of
the methane emissions associated with each equipment type.
For example, this grouping would allow a company to make a better choice
between turbine and reciprocating compressors, if methane emissions from the compressors were
important to the company. Instead of using only the difference in compressor exhaust emissions
between the two types, all of the compressor emissions should be used in the comparison. For
example, all turbine compressor emissions would include: turbine compressor exhaust, turbine
compressor blow and purge, turbine compressor fugitives, and turbine compressor pneumatics.
Unfortunately, recasting the data in this form cannot be done with precision since
many emission categories cannot be accurately split into equipment types. The methods used to
estimate the emissions simply do not provide this breakdown. Blow and purge emissions from
compressors, for example, were calculated from total volumes for all events provided by a
company. Since the companies did not provide the data by engine type, the data cannot be
accurately split into compressor start gas for turbines, compressor start gas for reciprocating
engines, blowdown gas for turbines, and blowdown gas for engines.
The assumptions used to split emissions into equipment types are listed in Table
4-9. Table 4-9 shows that reciprocating compressors contribute the most emissions among the
categories (100 Bscf). This is due to the large number of reciprocating compressors, combined
with large emission rates from the following: fugitive emissions associated with compressor
components, the large compressor exhaust emissions from reciprocating compressors, and
relatively large blowdown emissions associated with reciprocating compressors. The next
highest equipment category is pipelines (60 Bscf), which have high emissions due to the
68
-------
TABLE 4-9. EMISSIONS BY EQUIPMENT
!
Estimated Annual Emissions (Bscf)
Equipment Type
Emissions Included
Fugitives
Vented
I Combusted
Total
Reciprocating
Compressors
Exhausts, blow and purge (starts and
blowdowns), fugitives, pneumatics,
production stations
67.4
6.4
24.6
98.4
Pipelines (Gathering
Transmission, Dist.)
Fugitives, dig-ins, blow and purge
48.4
11.5
59.9
Separators
Fugitives, pneumatics, chemical
injection pumps, production vessel
blowdowns, production PRV's
3.4
29.8
33.2
M&R Stations
Fugitives, distribution PRV's
31.8
<0.1
31.9
Transmission Station
Vessels/Piping
Fugitives, pneumatics, station venting
9.2
22.2
31.4
Centrifugal Compressors
Exhausts, blow and purge (starts and
blowdowns), fugitives
14.7
0.4
0.3
15.3
Glycol Dehydrators
Fugitives, pneumatics, dehydrator vents,
AGR vents, dehydrator pumps
1.2
17.4
18.6
Wellheads
Fugitives, well workovers, well clean
ups, completion flaring
3.0
5.7
<0.1
8.7
Production Meters/Piping
Fugitives
6.1
6.1
Customer Meters
Residential, commercial/industry
5.8
5.8
Gas Plant Vessels/Piping
Fugitives, pneumatics, blow and purge
2.1
0.4
2.5
Offshore Platforms
Fugitive, ESD
1.2
0.3
1.5
Heaters
Fugitives
1.1
Negl.
1.1
TOTAL
195
94.2
24.9
314
Assumptions:
- Production pneumatics are broken down as: 90% separators, 2% dehydrators, 8% reciprocating compressors.
- Gas processing pneumatics are broken down as: 90% vessel/pipes, 10% reciprocating compressors.
- Transmission and storage pneumatics are broken down as: 90% vessel/pipes, 10% reciprocating compressors.
- Gas processing blowdowns are broken down as: 76% reciprocating compressors, 14% turbine, and 10% vessel/pipes.
69
-------
tremendous mileage of pipe in the United States combined with relatively large dig-in and blow
and purge emissions rates. The third highest equipment category is production separators (3
Bscf). Separators have a high emission rate due to the large population, combined with high
emission rates from associated pneumatics, fugitives, and chemical injection pumps. There are
four other equipment categories that each exceed 10 Bscf: M&R stations, transmission station
vessels and piping (i.e. everything but the compressors), turbine compressors, and glycol
dehydrators.
Many of the categories in Table 4-9 have high emissions due to a high population
of equipment, rather than due to a high emission rate per equipment. Table 4-10 recasts the total
data in Table 4-9 into equipment emission factors by using aggregate activity factors. Many of
these aggregate factors are groupings of multiple categories, such as all types of pipeline miles.
They are therefore not as specific as the individual activity factors presented in Appendix A, and
should be used only for the purposes of comparison in this table.
Table 4-10 shows that the highest single sources on the list are gas plants and
transmission and storage stations. These are large facilities with large equipment counts that
result in relatively high fugitive and blow and purge emissions. The highest emission factors for
individual equipment types are : 1) compressors, 2) glycol dehydrators, 3) separators, and 4)
M&PR stations. Each of these are explained in more detail below.
While turbine compressors have the highest emission rates per compressor unit
(due to fugitives and blowdowns), reciprocating engine-driven compressors have higher methane
emissions per million horsepower hour. This makes sense because turbine driven compressors
have specific maintenance practices that result in higher blowdown and fugitive emissions on a
per compressor basis, yet have far lower driver exhaust emissions on a per HP-hr basis.
70
-------
TABLE 4-10. ESTIMATED EQUIPMENT EMISSION FACTORS
| Estimated Activity j Estimated E mission Factor (Mscflequipment)
Equipment Types
Factor
Fugitives
Vented
Combusted
Total
Reciprocating
Compressors
29,000
compressors
2,327
222
832
3,381
102,500
MMHp-hr
658*
62.3*
240*
960*
Total Pipelines (Gathering
Transmission, Distribution)
1,620,000
miles
29.9
7.1
37.0
Separators
166,000
separators
20.2
180
200
M&R Stations
207,000
stations
154
0.202
154
Transmission & Storage Station
Vessels/Piping
2,175
stations
4,219
10,212
14,430
Turbine/Centrifugal
Compressors
1,540
compressors
9,530
268
164
9,962
44,000
MMHp-hr
334*
9.4*
5.7*
349*
Glycol Dehydrators
38,000
dehydrators
32.4
458 |
490
Wellheads
272,000
wellheads
11.0
20.9
0.0
31.9
Production Meters/Piping
377,000
meters
16.1
16.1
Customer Meters
45,000,000
meters
0.128
0.128
Gas Plant Vessels/Piping
726
plants
2,886
554
3,440
Offshore Platforms
1,110
platforms
1,055
258
1,313
Heaters
51,000
heaters
21.0
Negl.
21.0
Assumptions: See Table 4-8
Note: * Mscf/MMHp-hr, not per equipment
71
-------
Glycol dehydrators have high emission factors due to contributions from multiple
sources on the dehydrator: the glycol vent, the glycol pump, fugitives, and pneumatics. The
separator also has high emissions, mostly due to the high number of pneumatic devices
associated with separators. Similarly, M&PR stations also have high emissions mostly due to the
pneumatic devices associated with the stations.
4.7 Accuracy Results
The accuracy goal was to determine emissions from the natural gas industry to
within ± 0.5% of natural gas production. This goal was established based on the accuracy needed
for constructing emission inventories for use in global climate change models and for assessing
the validity of the fuel switching strategy. Accuracy, which is made up of precision and bias, has
been rigorously propagated through the calculations using techniques described in Volume 4 on
statistical methodology.6 The propagation of error resulted in a calculated uncertainty of ± 89.6
Bscf (0.4% of gross production). However, this assumes that the errors are normally distributed
and that there is no correlation between source categories.
Since there are some correlated errors among categories, and since some
categories might have lognormal distributions, the uncertainty estimate for the total emissions
was modified. The effect of inter-category correlations was calculated, and the additional
uncertainty was added to the uncertainty total. In addition, the effect of lognormal distribution
assumptions was also calculated. A point midway between the result for normal and lognormal
errors was used as a more reasonable conservative case than is the result based on the normal
assumption. The midway point represents the possibility that there is asymmetry in the
distribution of the error in the industry emission rate. While the selection of the midway point is
arbitrary, it is considered a reasonable postulated conservative case, given the various issues
discussed in the Volume 4 on statistical methodology.6
72
-------
Therefore, with assumptions of inter-category correlations and some lognormality,
the uncertainty is calculated to be ± 104.6 Bscf, which is slightly under 0.5% of national
production. The conclusion is that, under assumptions that are not unrealistically conservative,
the target precision was achieved.
The project has reached its accuracy goal for the annual emissions. The objective
of the project was to determine the overall national methane emissions, not to accurately
determine methane emissions for individual equipment or processes. The emission estimates for
source categories represent industry average values and are not meant to be representative of any
company's individual emissions or operations. Also, although the project has reached its
accuracy goal for the total emissions, the percent accuracy of an emissions estimate for a specific
category will likely have a much wider confidence bound than the national estimate.
73
-------
5.0 ANALYSIS AND CONCLUSIONS
As presented in Section 4, the methane emissions estimate from the U.S. natural
gas industry for the 1992 base year is 314 Bscf, which is 1.4% of gross natural gas production
(i.e., 1992 gross production was 22,130 Bscf).
As part of this program, a rigorous calculation of the uncertainty in emissions
from the significant sources was made to help plan the program. An overall accuracy target of
0.5% of natural gas production (±111 Bscf) was set as a benchmark to address the fuel switching
issue. The overall accuracy of the total methane emissions estimate generated from this program
is ± 106 Bscf, or 0.5% of natural gas production. Therefore, the accuracy goal originally set
forth for the program has been met (see Section 4.7).
Methane emissions from all U.S. anthropogenic sources are reported in the U.S.
EPA Report To Congress (RTC).3 Excluding the gas industry, the report states that total U.S.
anthropogenic methane emissions are estimated to be between 1190 to 1336 Bscf. Therefore, the
gas industry (based upon the new GRI/EPA estimate) accounts for 19% to 21% of total U.S.
methane emissions. According to the RTC, landfills (421 to 614 Bscf) and livestock (328 to 546
Bscf), each has higher emissions of methane (Figure 5-1).
Other
6%
Natural Gas Systems
20%
Landfills
31%
Coal Mining
15%
Domesticated
Livestock
Livestock Manure
Figure 5-1. Contribution of Major Methane Sources to Total
U.S. Anthropogenic Emissions
74
-------
The following sections analyze the results of the GRI/EPA study in various
contexts. Section 5.1 uses the results to examine the validity of the fuel switching strategy.
Section 5.2 compares the results to previous estimates. Section 5.3 discusses trends in the
natural gas industry that have changed total emissions since the base year of 1992. Finally,
Section 5.4 summarizes some of the key lessons learned during this study.
5.1 Impact of Natural Gas Use on Global Warming
The primary purpose of the GRI/EPA methane emissions study was to help
answer the question of whether the strategy of switching from other fossil fuels to natural gas
would be successful in reducing global warming. To address this question, the amount of
greenhouse gas released during the fuel cycle for each fossil fuel and the impact of these gases on
the atmosphere are needed. For fossil fuels, only emissions of carbon dioxide (C02) and
methane play a significant role. For methane emissions, it is important to account for emissions
from the production of gas, oil, and coal and also from the transmission and distribution of
natural gas. Methane emissions from the transportation and distribution of coal and oil are
negligible, as are methane emissions from end-use combustion. Nearly all the C02 emitted
results from end-use combustion of the fossil fuels.38 Only 7 to 9% of the C02 emitted from
natural gas is associated with upstream production, processing, and transportation, while 11% of
the C02 emissions associated with oil are from production through product transport.
Approximately 1% of the C02 emitted from coal is associated with production, processing and
transportation.42
After determining the emissions of C02 and methane over the fuel cycle for each
fossil fuel, the second step is to determine the impact of those emissions on global wanning.
This is a difficult problem because C02 and methane behave very differently when released into
the atmosphere; they have different lifetimes and absorb substantially different amounts of
infrared energy. As discussed in Appendix B, an index referred to as the Global Warming
Potential (GWP) can be calculated that describes the impact of a given greenhouse gas on global
warming compared to C02. The GWP can then be used to convert emissions of one greenhouse
75
-------
gas, such as methane, into equivalent quantities of C02. For example, if the GWP of methane
was seven, then one pound of methane would have the same impact on global warming as seven
pounds of C02.
The impact of greenhouse gases such as C02 and methane is dependent on the
amount of infrared energy they absorb (referred to as their radiative forcing) and their
concentration. Since the concentration is a function of time, the GWP is calculated by
integrating the ratio of the impact of methane to the impact of C02 as the concentration of the
gases decreases with time.
The value of the GWP is highly dependent on the time period over which the
integral is evaluated because the lifetime of methane is significantly shorter than the lifetime of
C02. Some studies select a period long enough for concentrations of both gases to decrease to
the original value (approximately 500 years), while others have chosen a shorter time period of
50 to 100 years. The GWP for methane is approximately 6.5 for an integration interval of 500
years, while the value of the GWP using a 50-year period is 34. Faced with such a large
difference, two approaches were taken to examine the validity of the fuel switching strategy.
The first approach is to determine the breakeven percentage. The breakeven
percentage is the amount of methane that would have to be emitted to the atmosphere from
natural gas operations in order for natural gas to have the same impact on global warming as
coal or oil (i.e., the amount of methane that would have to be leaked to eliminate the inherent
advantages that gas has because of its lower C02 emissions). Comparing the breakeven
percentage to the 1992 emission estimate provides an indication of the advantage that natural gas
has over coal or oil. Likewise, the breakeven percentage can be compared to the percentage of
natural gas emissions resulting from an incremental increase in gas use to determine the validity
of the fuel switching strategy.
The analysis presented in Appendix B indicates that between 8 and 34% of the
natural gas produced would have to be lost to the atmosphere for natural gas to have the same
76
-------
impact on global warming that coal has, depending on whether the GWP was evaluated over a
50- or 500-year time period. A similar comparison for oil indicates that methane emissions from
natural gas operations would have to be between 5 and 23% of production to have the same
impact that oil has on global warming.
As discussed in Section 5.3.2, the GRI/EPA study not only evaluated emissions
for the 1992 baseline system, but also estimated emissions from incremental increases in natural
gas use ranging from 5 to 30%. The study found that incremental emission increases were
proportionally less than the increases in gas usage for the scenarios examined.
Since the breakeven percentages for coal (8 to 34%) and oil (5 to 23%) are much
larger than even the upper limit of the percent of gas lost per gas produced from an incremental
increase in gas use (i.e., 1.38% compared to 1.42 for the 1992 baseline) the breakeven analysis
shows that switching from other fossil fuels to natural gas is a valid strategy for reducing global
warming.
In the second approach, the amount of "equivalent" C02 emissions was evaluated
for each fossil fuel over the fuel cycle by converting methane emissions to "equivalent" C02
emissions. Since the GWP is a factor that relates the impact of releasing a pound of methane on
global warming to that of releasing a pound of C02, the GWP can be used to convert methane
emissions into equivalent amounts of C02. For the fuel switching analysis, emissions are
expressed as the mass of equivalent C02 emissions per unit of energy (based on the higher
heating value of the fuel). Thus for an energy requirement of one million Btu, the equivalent C02
emission contribution of each fuel can be compared. Table 5-1 presents the results of this
comparison for GWPs of 6.5 and 34. (A more detailed discussion is presented in Appendix B.)
Table 5-1 also shows the ratio of equivalent C02 emissions per MMBtu for coal and oil divided
by the value for natural gas. This "equivalent C02 ratio" shows that oil has 1.2 to 1.4 times the
impact on global warming compared to natural gas, and coal contributes 50 to 60% more
equivalent C02 emissions than natural gas.
77
-------
TABLE 5-1. EQUIVALENT COz EMISSIONS
Fuel Source
lbs C02/MMBtu
Equivalent C02 Ratio
GWP = 6.5
GWP = 34
GWP = 6.5
GWP = 34
Gas
132
152
1.0
1.0
Oil
184
186
1.4
1.2
Coal
212
228
1.6
1.5
An analysis of the fuel switching strategy based on examining the equivalent C02
emissions from each fuel supports the conclusion reached by evaluating the fuel switching
strategy using the breakeven percentage. Based on the results of both approaches, fuel switching
is a valid strategy for reducing global warming. This conclusion is consistent with the
Intergovernmental Panel on Climate Change (IPCC) report43 on climate change.
5.2 Comparison to Previous Estimates
This project began in 1989 by posing the following questions: 1) "What are the
methane emissions from the U.S. natural gas industry from the wellhead to the customer meter?"
and 2) "Based on this emission estimate, is it reasonable to recommend switching from oil or
coal to natural gas as a strategy for reducing the U.S. contribution to global climate change?"
The project sponsors agreed that it would not be prudent to attempt to answer the second
question unless an accuracy goal for the emission estimate of ± 0.5% of gas production could be
achieved with 90% confidence. An emission estimate with this degree of accuracy, therefore,
became the project objective.
A literature survey conducted at the outset of this project verified that previous
studies contained insufficient data, individually or collectively, to meet the accuracy goals of this
project. The majority of studies that were found during the literature survey employed a method
common at that time in which "unaccounted-for gas" was assumed to be equivalent to losses to
78
-------
the atmosphere.44'45,46-47 "Unaccounted-for gas" is simply an accounting term which includes
numerous categories in addition to losses to the atmosphere and therefore could greatly overstate
gas industry losses.
These studies were followed by a report written by Pipeline System Incorporated
(PSI) and funded by EPA-OAR and GRI in 1990.48 The purpose of the early GRI/EPA study
was to initially guide the more comprehensive GRI/EPA efforts that are presented in this report.
The early PSI study produced an estimate showing that methane emissions were 1% of gross gas
production. However, this study was only an attempt to identify major sources, and no emission
measurements were made.
A Report to Congress (RTC) by EPA estimated that methane emissions from the
natural gas industry were between 0.55 and 1.07% of gross production for 1990.3 This study
provided a reasonable synthesis of existing data at that time but did not expand the database. The
need remained for an extended field sampling program and a statistical framework within which
the data could be analyzed and accuracy targets could be calculated.
The 140 Bscf difference between the emission estimates from the RTC (110 to
220 Bscf) and the GRI/EPA study (307 Bscf) result from recent data that were not available at
the time the RTC was written. The GRI/EPA study used new data to refine many source
categories. The single category with the most significant difference was fugitive emissions,
which accounts for almost 90% of the difference between the RTC and the GRI/EPA reports.
The fugitive differences result from two major sources of new data: 1) compressor components
(82 Bscf difference), and 2) distribution sources (60 Bscf difference), such as pipelines, meter
and regulation (M&R) stations, and customer meters.
Compressor components, which are very large sources of fugitive emissions, were
measured as part of the GRI/EPA study, but no measurements were available at the time of the
RTC. Compressor components in processing, transmission, and storage facilities resulted in
GRI/EPA estimates of 82 Bscf; these components had not been accounted for by the RTC.
79
-------
The GRI/EPA report also refined the estimates for leakage from distribution
pipelines and M&R stations through additional data gathering efforts. When the RTC was
prepared, the only data available on pipeline leakage were for two very tight distribution systems
that had very little cast iron pipe. In addition, no data were available on the number and type of
M&R stations used in the gas industry. These data were gathered during the GRI/EPA study.
Also, emissions from customer meters, which were not included in the RTC, were included in
the GRI/EPA study and measured to be 6 Bscf. The new GRI/EPA data show that the total
distribution segment emissions are approximately 60 Bscf higher than estimated by the RTC.
5.3 Current and Future Industry Emissions
Since the 1992 base year, emissions from the natural gas industry have changed
because the amount of gas produced has increased and because gas industry practices have
changed. In 1993 a joint industry-government program was started to reduce emissions. The
impact of increased production and changes in practices is discussed in more detail in the
following subsections.
5.3.1 Industry Practices to Reduce Methane Emissions
The natural gas industry has always been concerned with reducing natural gas
losses. Every year the industry's practices continue to evolve and many companies have policies
to recover gas or reduce losses. Examples are company programs to reduce losses through
fugitive leak detection and repair programs (LDAR) for underground piping and above-ground
facilities. Also as a result of this study, a number of companies became aware of ways to reduce
operating costs while reducing emissions, and many of these companies are implementing
cost/emission reduction programs.
In 1993, a joint industry-government effort began. The Environmental Protection
Agency (EPA), in conjunction with the natural gas industry, created the Natural Gas STAR
Program to help reduce methane emissions from its major sources.2 The Natural Gas STAR
Program was established as a flexible, voluntary partnership to reduce methane emissions using
80
-------
cost-effective practices. The EPA49 and industry identified several best management practices
(BMPs) in each sector of the natural gas industry, including:
Distribution Sector
• Implement directed inspection and maintenance programs at surface
facilities
• Identify and rehabilitate leaky distribution pipe
Transmission Sector
• Implement directed inspection and maintenance programs at compressor
stations
• Consider use of turbines at compressor stations in lieu of reciprocating
engines
• Identify and replace high-bleed pneumatic devices
Production Sector
• Identify and replace high-bleed pneumatic devices
• Install flash tank separators on dehydrators
The program also facilitates technology transfer among partners on other practices
that cost-effectively reduce methane emissions. As of April 1996 the program included 54
partners, representing over 60% of all transmission pipeline, 30% of all distribution pipeline and
25% of all U.S. natural gas production. As the new Producers Program (launched in March,
1995) gets under way and as new distribution and transmission companies join, the program is
expected to continue to reduce emissions of methane by 35 Bcf through the year 2000.
81
-------
The gas industry may also decrease methane emissions in the future as it complies
with maximum achievable control technology (MACT) standards for hazardous air pollutants
(HAPs). The MACT rule will result in a reduction of certain hydrocarbon emissions and may
also reduce methane emissions. However, the actual impact of the MACT is unclear at this point
in time, due to questions on the language of the final rule, the compliance schedule, source
applicability, and the required control technologies.
Since the Oil and Gas MACT Rule is scheduled for promulgation in 1997, the
only information publicly available is from the preliminary Background Information Document
(BID).50 The effect of the Oil and Gas MACT on methane emissions cannot be easily determined
because the language of the MACT does not specifically address these emissions. The BID
suggests that glycol dehydrators and some sources of fugitive emissions will require controls for
HAP emissions. Depending on the kind of controls implemented by the industry for these
sources, methane emissions may be reduced as well.
The preliminary draft MACT proposes that equipment leaks at major sources,
including gas processing plants and offshore platforms, must be controlled by a LDAR program.
If this requirement becomes part of the final MACT rule, methane emissions from fugitive
sources in the gas production and processing segments will decrease.
53.2 Incremental Increases in System Throughput
As part of this program, a study was conducted to determine the percent increase
in emissions caused by an incremental increase in natural gas production and sales.18 The study
found that increases in throughput did, in many cases, produce increases in emissions. However,
the average increase in emissions was proportionally smaller than the increase in system
throughput.
82
-------
This study examined the consequences of increasing gas sales by 5,15, and 30
percent under three scenarios: uniform, winter peak, and summer peak load profiles. All
segments of the gas industry were examined to determine the percent increase in equipment that
would be needed in order to meet the increased demand. The percent increase in emissions was
then estimated based on changes in the current system that would be required to accommodate
the increase in gas sales. The GRI/EPA's emission estimates were used to calculate the percent
increase in emissions that would result from an incremental increase in natural gas sales for
several scenarios examined in the study. The most realistic scenario assumes that the system will
be expanded using current technology, whereas the most conservative scenario assumes that the
expanded system mirrors the existing system. Generally, emissions would only increase 2% to
21% for corresponding load increases of 5% to 30%. The incremental methane emission
increases, when divided by the incremental production rate increases, result in emissions per
production percentages of 0.3 to 1%, which are only one-third to two-thirds of the base emission
rate (1.42% for 1992). Thus, the incremental emission increases are proportionally less than the
load increase for all scenarios examined. (Results are explained further in Appendix B.)
5.4 Lessons Learned for Future Studies
The project team learned some key lessons during this multi-year project that may
benefit other similar studies. The key lessons learned are grouped below in two categories:
sampling/statistical methods and measurement methods.
5.4.1 Sampling/Statistical Methods
Because of the complexity and diversity of the natural gas industry, a detailed
plan was implemented to meet the goals of the program.27 Some of the procedures used in
sample selection and statistical methodology were developed/implemented specifically for this
program but would have potential utility in other similar studies. These sampling/statistical
methods include:
83
-------
• Sampling technique that is dependent upon the source population;
• Sampling techniques and statistical methods to minimize bias in a dataset;
• Use of accuracy targets to plan the program and allocate resources; and
• Statistical tests to handle small datasets that are highly variable.
Sampling Technique/Bias Minimization
Because of the complexity and often unknown equipment populations for a given
source within the gas industry, the selection of a proper sampling approach was not
straightforward. For some sources, such as production separators, even the population size was
not known at the onset of the program. These factors made the selection of representative
samples for measurement or observation difficult, and traditional sampling methods, such as
random or stratified random sampling were not directly applicable in most cases. Therefore, an
alternative approach, which is similar to disproportionate stratified random sampling, was used.
The sampling approach included selecting sites from known lists of facilities in as
random a fashion as possible. However, the companies contacted were not required to
participate and a complete list of all sources in the United States was generally not available;
therefore, site selection was not truly random. Companies that elected to participate were asked
to identify potential sites that were considered representative of company-wide operations.
The limited data set collected was screened for bias by evaluating the relationship
between the emission rate and parameters that may affect emissions. The data set was then
stratified by the parameter(s) found to significantly influence emissions. Because the sample set
collected was not necessarily representative of the nationwide proportions of sites in each strata,
an emission factor per strata was produced along with an activity factor per strata to eliminate
bias in the disproportionate sample set.
84
-------
Other techniques employed to minimize bias included evaluating regional
differences in operating practices or gas composition. In many cases, regional differences were
found and had to be accounted for in the emissions estimation approach. A group of industry
experts was used to review the data and approach for estimating emissions, so that any additional
biases could be identified and eliminated. Industry experts from each segment and other
reviewers were called upon to regularly review the project sampling approach, extrapolation
techniques, and preliminary estimates. These reviewers identified potential biases that were
eliminated through changes to techniques or through additional data collection.
Use of Accuracy Targets
To effectively allocate resources within the budget constraints of the program,
accuracy targets were established for each emissions source such that resource could be assigned
to emission sources based on the impact of each source on accuracy. An overall target accuracy
was set for the industry-wide methane emissions estimate, and individual source target
accuracies were calculated based upon overall accuracy goal. Target accuracies were set so that if
individual source accuracies were met, the overall accuracy for the project would be met. The
individual source accuracy targets were calculated based on precision estimates of the activity
and emission factors. After the individual source target accuracies were calculated, the required
number of additional samples needed to meet the target was calculated. By setting accuracy
targets for individual sources, small, highly uncertain sources of emissions could then be
appropriately handled. This process was used continuously throughout the data collection phase
of the program to help direct the most efficient use of resources required to meet the overall
program goal.
85
-------
5.4.2
Measurement Methods
A number of unique measurement methods were developed and tested as a result
of this program. Many of these methods are applicable to sources outside of the natural gas
industry. The most noteworthy of these are listed and described below:
• High flow device for fugitive emissions measurements;
• Tracer gas measurement method for estimating emissions from meter and
pressure regulating stations;
• A cooperative effort between industry and GRI/EPA in measuring
emissions from underground pipeline leaks; and
• A detailed mass balance approach for system-wide emissions from a
sample transmission network.
High Flow Fugitives Measurement Device
At the beginning of the GRI/EPA methane emissions study, it was clear that new
component emission factors would be needed to evaluate fugitive emissions from gas industry
equipment. The factors developed by EPA in the 1970s for natural gas production facilities were
no longer applicable because of the changes that took place in the industry over the past 15 to 20
years. In addition, emission factors for gas processing, transmission and distribution equipment
were needed since these had not been developed previously.
The standard EPA approach for determining emission factors uses a combination
of screening and enclosure methods. First, all components are screened using an organic vapor
analyzer (OVA) to determine which pipefittings are leaking and to measure the maximum
concentration at the point of the leak. This is done for thousands of components at sites
throughout the country. The leak rate is measured using the enclosure method for hundreds of
leaking fittings of each type. A correlation equation is developed that correlates the
86
-------
concentration value measured with the OVA with the emission rate measured using the enclosure
method. The correlation equation is then used to calculate the emission rate for all components
based on OVA readings, and an average emission rate (i.e., emission factor) is calculated for
each type of component.
The problem is that the scatter in the concentration versus emission rate data is 3
to 4 orders of magnitude. Because the correlation is poor, thousands of measurements are
needed, and this is time consuming and expensive. Therefore, GRI funded a study to develop a
new instrument that could accurately measure the emission rate directly in about the time
required to measure the concentration. This method was used not only for developing emission
factors for production equipment, but also for processing, transmission, and distribution.
The new instrument, called the GRI Hi-Flow sampler, can also be used to reduce
operating cost. Since it provides a quick accurate measurement of the leak rate, the operator can
determine if it is cost effective to fix the leak. It also can be used to accurately measure the
fugitive emissions from a facility and determine whether the facility is subject to regulations and
costly control and reporting requirements.
Tracer Gas Measurement Method
Tracer techniques were developed to measure methane emissions from sources of
widely varying sizes and types. These sources included single regulator installations (above
ground and below ground), city distribution M&PR stations, transmission tie-in points,
transmission and production facilities, industrial gas users, municipal wastewater treatment
facilities, landfills, and total city emissions.
The principle for each of these emission measurements was the same, but the
application varied depending on the scale of the measurement. In each case, the tracer was used
to measure the dilution of the methane from the source as it was transported to the receptor where
87
-------
concentration was measured. For underground vaults and enclosures of single above-ground
regulators, air was flushed through the enclosed volume. The methane emissions were measured
by measuring the resulting dilution with a tracer released at a known rate while measuring both
the tracer and methane concentrations. For larger sources such as M&PR stations, gas plants,
and landfills, tracer was released at a known rate from an area inside the source boundaries, and
the tracer and methane were measured at a downwind distance where the tracer and methane
were well mixed. This again provided a measurement of the dilution of methane as it was
transported from the source to receptors and allowed the calculation of the source strength from
the ambient methane concentration.
Several lessons were learned concerning the application of tracers during this
work. Real time instruments were used to track both the methane and tracer plumes and helped
to identify interfering sources, to determine appropriate sampling points, and to integrate the
plumes from very large scale sources such as landfills and cities. Measurements were validated
using techniques developed in past studies which included comparing results from samplers at
different crosswind locations in the methane plume, comparing plume traverses at different
downwind distances, and conducting replicate measurements with different tracer source
configurations or under different meteorological conditions.
Tracer emission measurement techniques have both advantages and disadvantages
compared to techniques that measure emissions from individual components or flux chamber
measurements made at landfills or treatment plants. The accuracy of the tracer technique is
susceptible to some meteorological conditions and interferences from other sources. However,
the tracer technique can provide the total site emission rate in a fraction of the time (a few hours
under the appropriate conditions) that is required using individual component techniques or flux
chambers. At a natural gas facility, this total emission rate will include non-fugitive sources such
as compressor engine exhaust. Measuring total emissions proved to be an advantage in this
study because it was used to determine if any sources were missed by comparing the sum of all
known fugitive vented and combusted emissions to the total value measured using the tracer
88
-------
technique. It was found that leakage from the blowdown valve was overlooked by the component
measurement method because it was directed to a roof vent system.
Due to the inherent uncertainty of the component screening techniques originally
used for the individual component measurements, the tracer method was the most accurate
method available for determining total emissions from a facility. The development of the high
flow sampler during this project now provides a component measurement method of the same or
better accuracy than the tracer method. Consequently, the best measurement method will often
depend on the goals of the measurement work. Tracer techniques do not provide any data on the
location or magnitude of sources within a site. For a natural gas facility, effective emissions
reductions cannot be accomplished without knowing which components are leaking and how
much each is leaking. However, these individual component methods are more time consuming,
have the potential to miss significant sources, and are not applicable to many sources. When
trying to obtain as much data on total facility emissions as possible in the shortest amount of
time, tracer techniques may provide the best method of emissions measurement.
Cooperative Industry Measurement Effort
Early in the program, leakage from underground pipelines in the distribution
segment was targeted as a potentially large source. A measurement technique identified as being
very accurate was proposed for the GRI/EPA program. However, this technique was extremely
costly to implement on a per test basis, and due to the population size and uncertainty in
emissions, the estimated sample size to reach the target accuracy was very large. Therefore,
GRI/EPA solicited participation in a cooperative program between industry and the program
sponsors to share the cost of collecting data. The GRI/EPA program provided a detailed test
protocol, specifications for the measurement device, and training/auditing/support to the industry
participants. The actual measurements were performed and funded by the companies agreeing to
participate in the program. This cooperative effort proved to be a successful means to meet the
objectives within the budget constraints of the GRI/EPA program.
89
-------
Mass Balance Measurement Approach
An extremely detailed mass balance was performed on a sample transmission
system to determine if emissions to the atmosphere could be determined by examining the
differences in upstream/downstream meter readings. This effort did not prove successful due to
the many uncertainties in mass balance measurements that could not be completely resolved and
emissions could not be determined to meet the accuracy target.
5.4.3 Significant Sources
Several significant sources of methane emissions were identified or found to be
much larger than anticipated. Compressor blowdown valve fugitives, M&PR stations, pneumatic
devices, dehydrators, and maintenance emissions all were determined to be larger than estimated
by previous studies.
90
-------
REFERENCES
United Nations Environmental Programs, Intergovernmental Panel on Climate
Change (IPCC), Greenhouse Gas Inventories, 1995.
U.S. Environmental Protection Agency. Natural Gas STAR: The Second Year.
EPA 430-F-96-010, Office of Air and Radiation, Winter 1995/1996.
U.S. Environmental Protection Agency. Anthropogenic Methane
Emissions in the United States: Estimates for 1990. Report to Congress,
EPA 430-R-93-003, Environmental Protection Agency, Office of Air and
Radiation, Washington, DC, April 1993.
Harrrison, M.R., T.M. Shires, J.K. Wessels, and R.M. Cowgill. Methane Emission
from the Natural Gas Industry, Volume 1: Executive Summary, Final Report,
GRI-94/0257 and EPA-600/R-96-080a. Gas Research Institute and U.S.
Environmental Protection Agency, June 1996.
Harrison, M.R., H.J. Williamson, and L.M. Campbell. Methane Emissions
from the Natural Gas Industry, Volume 3: General Methodology, Final Report,
GRI-94/0257.20 and EPA-600/R-96-080c. Gas Research Institute and U.S.
Environmental Protection Agency, June 1996.
Williamson, H.J., M.B. Hall, and M.R. Harrison. Methane Emissions from the
Natural Gas Industry, Volume 4: Statistical Methodology, Final Report, GRI-
94/0257.21 and EPA-600/R-96-080d. Gas Research Institute and U.S.
Environmental Protection Agency, June 1996.
Stapper, B.E. Methane Emissions from the Natural Gas Industry, Volume 5:
Activity Factors, Final Report, GRI-94/0257.22 and EPA-600/R-96-080e. Gas
Research Institute and U.S. Environmental Protection Agency, June 1996.
Shires, T.M. and M.R. Harrison. Methane Emissions from the Natural Gas
Industry, Volume 6: Vented and Combustion Source Summary, Final Report,
GRI-94/0257.23 and EPA-600/R-96-080g. Gas Research Institute and U.S.
Environmental Protection Agency, June 1996.
Shires, T.M. and M.R. Harrison. Methane Emissions from the Natural Gas
Industry, Volume 7: Blow and Purge Activities, Final Report, GRI-94/0257.24
and EPA-600/R-96-080g. Gas Research Institute and U.S. Environmental
Protection Agency, June 1996.
91
-------
10. Hummel, K.E., L.M. Campbell, and M.R. Harrison. Methane Emissions from
the Natural Gas Industry, Volume 8: Equipment Leaks, Final Report, GRI-
94/0257.25 and EPA-600/R-96-080h. Gas Research Institute and U.S.
Environmental Protection Agency, June 1996.
11. Campbell, L.M., M.V. Campbell, and D.L. Epperson. Methane
Emissions from the Natural Gas Industry, Volume 9: Underground
Pipelines, Final Report, GRI-94/0257. 26 and EPA-600/R-96-080i. Gas
Research Institute and U.S. Environmental Protection Agency, June
1996.
12. Campbell, L.M. and B.E. Stapper. Methane Emission from the Natural Gas
Industry, Volume 10: Metering and Pressure Regulating Stations in Natural Gas
Transmission and Distribution, Final Report, GRI-94/0257. 27 and EPA-600/R-
96-080j. Gas Research Institute and U.S. Environmental Protection Agency, June
1996.
13. Stapper, C.J. Methane Emissions from the Natural Gas Industry, Volume 11:
Compressor Driver Exhaust, Final Report, GRI-94/0257. 28 and EPA-600/R-96-
080k. Gas Research Institute and U.S. Environmental Protection Agency, June
1996.
14. Shires, T.M. and M.R. Harrison. Methane Emissions from the Natural Gas
Industry, Volume 12: Pneumatic Devices, Final Report, GRI-94/0257. 29 and
EPA-600/R-96-0801. Gas Research Institute and U.S. Environmental Protection
Agency, June 1996.
15. Shires, T.M. Methane Emissions from the Natural Gas Industry, Volume 13:
Chemical Injection Pumps, Final Report, GRI-94/0257. 30 and EPA-600/R-96-
080m. Gas Research Institute and U.S. Environmental Protection Agency, June
1996.
16. Myers, D. Methane Emissions from the Natural Gas Industry, Volume 14:
Glycol Dehydrators, Final Report, GRI-94/0257.31 and EPA-600/R-96-080n.
Gas Research Institute and U.S. Environmental Protection Agency, June 1996.
17. Myers, D.B. and M.R. Harrison. Methane Emissions from the Natural Gas
Industry, Volume 15: Gas-Assisted Glycol Pumps, Final Report, GRI-94/0257.
33 and EPA-600/R-96-080o. Gas Research Institute and U.S. Environmental
Protection Agency, June 1996.
18. Columbia Gas. An Engineering Estimate of the Incremental Change in Methane
Emissions with Increasing Throughput in a Model Natural Gas System, GRI-
94/0257.32, Columbus, OH, March 1993.
92
-------
19. Aerodyne Research Inc., Washington State University, and University of New
Hampshire. Results of Tracer Measurements of Methane Emissions from Natural
Gas System Facilities, Final Report, GRI-94/0257.43, Gas Research Institute,
March 1995.
20. SRI International, Evaluation of Methane Emissions from Natural Gas Production
Operations Using Tracer Methodologies, Final Report, Gas Research Institute,
GRI-92/0102, September 1992.
21. Star Environmental. Fugitive Hydrocarbon Emissions from Oil and Gas
Production Operations (API Publication No. 4589). American Petroleum
Institute, December 1993.
22. Star Environmental. Fugitive Hydrocarbon Emissions: Eastern Gas Wells, Final
Report, GRI-95/0117, Gas Research Institute, 1995.
23. Indaco Air Quality Services. Leak Rate Measurements for Natural Gas
Customer Meters, Draft Report, GRI-94/0257.36. Gas Research Institute, 1994.
24. Indaco Air Quality Services. Leak Rate Measurements at U.S. Natural Gas
Transmission Compressor Stations, Draft Report. Gas Research Institute, GRI-
94/0257.37, 1996.
25. Star Environmental. Emission Factors for Oil and Gas Production Operations.
(API Publication No. 4615). American Petroleum Institute, January 1995.
26. Star Environmental. Fugitive Methane Emissions, Customer Meter Sets, Report,
GRI-95/0204, Gas Research Institute, 1995.
27. Radian Corporation. Phase 3 Program Plan - Implementation Plan, prepared for
Gas Research Institute, Radian DCN 92-263-081-02, 1992.
28. Southwest Research Institute. Mass Balance of a Natural Gas Transmission
System for Improved Estimates of Methane Emissions. SwRI 04-4447, February
1993.
29. Washington State University, University of New Hampshire, and Aerodyne
Research, Inc., "Soil Consumption of Methane from Natural Gas Pipeline Leaks,"
Environmental Science & Technology, December 1994.
30. Lott, R., M. Webb, and T. Howard. "Sampler Enables Measurement of Leaks on
Site-Specific Basis."American Oil and Gas Reporter, March 1995.
93
-------
31. Philip Crosby, Quality Improvement through Defect Prevention, PCA, Inc.,
Winter Park, FL, 1985.
32. Radian Corporation, Program Plan for the Cooperative Leak Test Program, Gas
Research Institute, March 1993.
33. American Gas Association. Gas Facts: 1993 Data, Arlington, VA, 1994.
34. ICF Inc. Estimation of Activity Factors for Gas E&P Facilities, Final Report, U.S.
Environmental Protection Agency, Office of Air and Radiation, July 12, 1995.
35. Indaco Air Quality Services, Inc. Methane Emissions from Natural Gas
Customer Meters: Screening and Enclosure Studies, Draft Report, August 15,
1992.
36. Radian Corporation. Protocol for Equipment Leak Emission Estimation, EPA-
453/R-93-026 (NTIS PB93-229219). U.S. Environmental Protection Agency,
Emission Standards Division, June 1993.
37. ABB Environmental Services. Fugitive Hydrocarbon Emissions from Pacific
OCS Facilities (MMS Report 92-0043). U.S. Department of the Interior,
Minerals Management Service, November 1992.
38. Radian Corporation. Global Emissions of Carbon Dioxide from Petroleum
Sources, American Petroleum Institute, July 1991.
39. Canadian Petroleum Association, A Detailed Inventory of CH4 and VOC
Emissions from Upstream Oil and Gas Operations in Alberta, 1992.
40. Biederman, N. GRI TRANSDAT Database: Compressor Module, (prepared for
Gas Research Institute) npb Associates with Tom Joyce and Associates,
Chicago, IL, August 1991.
41. American Gas Association, Gas Facts. Arlington, VA, 1992.
42. Energy International. Energy Utilization and Greenhouse Gas Emissions: End-
Use Analysis, GRI-9310335. Gas Research Institute, June 1994.
43. World Meteorological Organization. Climate Change 1995. Intergovernmental
Panel on Climate Change, United Nations Environment Programme, 1995.
44. Ehhalt, D.H. The Atmospheric Cycle of Methane, Tellus, 26 55-70, 1974.
94
-------
45. Seiler, W.R. Conrad and D. Scharffe. "Field Studies of Methane Emission
from Termite Nests into the Atmosphere and Measurements of Methane Uptake
by Tropical Soils," J. Atmos. Chem, 1, 171-186, 1984.
46. Crutzen, P.J. "Role of the Tropics in Atmospheric Chemistry," Geophysiology
of Amazonia, edited by R.E. Dickinson, pp. 107-130, John Wiley, New York,
NY, 1987.
47. Sheppard, J.C., H. Westberg, J.F. Hopper, K. Ganessan, and P.
Zimmerman. "Inventory of Global Methane Sources and Their Production
Rates,"/. Geophys. Res., 87, 1305-1312, 1982.
48. Pipeline Systems Incorporated. Annual Methane Emission Estimate of the
Natural Gas Systems in the United States Phase 2, Chicago, IL, September 1990.
49. Environmental Protection Agency. Options for Reducing Methane Emissions
Internationally, Volume 1: Technological Options for Reducing Methane
Emissions, Report to Congress, EPA-430-R-93-006, Environmental Protection
Agency, Office of Air and Radiation, Washington DC, July 1993.
50. Preliminary Draft, Background Information Document for Proposed Oil and Gas
Production, Maximum Achievable Control Technology (MACT), U.S. EPA
Docket No. A-94-04, 1994.
95
-------
APPENDIX A
Summary Table of Emission Sources
A-l
-------
METHANE EMISSION AND ACCURACY ESTIMATES
Percent
Percent
Activity
Emission
Precision
Conservative
Target
PROCESS SEGMENT
1992
1992
of Total
of Total
Upper
Upper
of Annual
Precision
Precision
Emission Type
Emissions
Emissions
Emissions
Production
Value
Units
Bound
Value
Units
Bound
Emissions
of Annual
(%)
Source
flg)
(Bscf)
(%)
%
(a)
(b)
(b)
Emissions
fd)
(c)
PRODUCTION
Normal Fugitives
Oas Wells (Eastern on shore)
0.0064
0.3352
0.11
0.002
129,157
wells
5%
7.11
Scfd/well
27%
27.49%
31.39%
1077 82
Field Separation Equipment
(Eastern on shore)
Heaters
0.0000
0.0013
0.00
0.000
260
heaters
196%
14.21
scfd/heater
43%
217.64%
423.13%
1500.00
Separators
0.0006
0.0301
0.01
0.000
91,670
separators
23%
0.90
scfd/sep
27%
36.01%
42.74%
1500.00
Gathering Compressors
Small Recip. Compr.
0 0000
0.0006
000
0.000
129
compressors
33%
12.1
scfd/comp
27%
43.56%
53.45%
1500.00
Meters/Piping
0.0048
02508
0.08
0.001
76,262
meters
100%
9.01
scfd/meter
30%
108.63%
169.02%
1246.02
Dehydrators
00002
0.0083
000
0000
1,047
dehydrators
20%
21.75
scfd/dehy
35%
40.91%
49.63%
1500.00
Gas Wells (Rest of US on shore)
0.0365
1.8969
0.60
0.009
142,771
wells
5%
36.40
Scfd/well
24%
24.54%
27.65%
453.08
Guff of Mexico (offshore pltfrms)
0.0223
1.1615
0.37
0.005
1,092
platforms
10%
2914
Scfd/plat
27%
28.92%
33.24%
579.01
Rest of US (offshore platforms)
0.0002
0.0095
0.00
0.000
22
platforms
10%
1178
Scfd/plat
36%
37.54%
44.86%
1500 00
Field Separation Equipment
(Rest of US on shore)
Heaters
0.0206
1 0686
034
0.005
50.740
heaters
95%
57.7
scfd/heater
40%
109.86%
171.56%
603.64
Separators
0.0639
3.3252
1.06
0.015
74,674
separators
57%
122.0
scfd/sep
33%
68.50%
93.05%
342.20
Gathering Compressors
Small Recip. Compr.
0.0318
1.6534
0.53
0.007
16,915
compressors
52%
267.8
scfd/comp
68%
92.62%
137.08%
485.29
Large Recip. Compr.
0.0102
0 5328
0.17
0 002
96
compressors
100%
15205.0
scfd/comp
65%
135.83%
227.42%
854.90
Large Recip. Stations
0.0007
0.0361
0.01
0.000
12
stations
100%
8247.0
scfd/station
102%
175.52%
319.62%
1500 00
Meters/Piping
0.1118
5.8153
1.85
0.026
301,180
meters
100%
52.9
scfd/meters
30%
108.63%
169.02%
25876
Dehydrators
0.0235
1.2229
0.39
0.006
36,777
dehydrators
20%
91.1
scfd/dehy
25%
32.40%
37.84%
564.28
Pipeline Leaks
0.1269
6.6000
2.10
0.030
340,200
miles
10%
532
scfd/mile
107%
108.00%
167.72%
242.89
Vented and Combusted
Drilling and Well Completion
Completion Flaring
0.0000
0.0006
000
0 000
844
compl/yr
10%
733
scf/compl
200%
201.25%
382.35%
1500.00
Normal Operations
Pneumatic Device Vents
0.6037
31.3948
9.99
0.142
249,111
controllers
48%
345
Scfd/device
40%
64.99%
87.10%
111.37
Chemical Inj Pumps
0.0295
1.5365
0.49
0 007
16,971
active pumps
143%
248.05
Scfd/pump
83%
203.53%
388.00%
503.41
Kimray Pumps
0.2108
10.9616
3.49
0.050
1.105E+07
MMscf/yr
62%
992 00
scf/MMscf
77%
110.03%
171.90%
188.47
Dehydrator Vents
0.0657
3.4171
1.09
0015
1.240E+07
MMscf/yr
62%
275.57
scf/MMscf
154%
191.90%
359.36%
337 57
Compressor Exhaust Vented
Gas Engines
0.1267
65904
2.10
0.030
27,460
MMHPhr
200%
0240
scf/HPhr
5%
200.31%
380 04%
243.07
Routine Maintenance
Well Workovers
Gas Wells
0.0004
0.0230
0.01
0.000
9,392
w.o./yr
258%
2,454
scfy/w.o.
459%
1296.00%
2746.84%
1500.00
Well Clean Ups (LP Gas Wells)
0.1088
5.6579
1.80
0.026
114,139
LP gas wells
45%
49570
scfy/LP well
344%
379.90%
834.56%
262.34
Slowdowns
Vessel BD
0.0004
0.0200
0.01
0.000
255,998
vessels
26%
78
Scfy/vsl
266%
276.07%
571.10%
1500.00
Pipeline BD
0.0020
0.1051
0.03
0.000
340,000
miles(gath)
10%
309
Scfy/mile
32%
33.68%
39.56%
1500.00
Compressor BD
0.0012
0.0646
. 0.02
0000
17,112
compressors
52%
3774
Scfy/comp
147%
173.66%
315.14%
1500.00
Compressor Starts
0.0028
0.1445
0.05
0.001
17,112
compressors
52%
8443
Scfy/comp
157%
184.44%
341.16%
1500.00
Upsets
Pressure Relief Valves
0.0003
0.0180
0.01
0.000
529,440
PRV
53%
34
Scfy/PRV
252%
290.09%
606.86%
1500.00
ESD
0.0055
0.2864
0.09
0.001
1,115
platforms
10%
256688
Scfy/plat
200%
201.25%
382.35%
116595
Mishaps (Dia-ins)
0.0044
02275
0.07
0.001
340.000
miles
10%
669
scf/mile/vr
1925%
1934.63%
3766.68%
1308.38
(a) Based on a total gross national production of 22132 Bscf for 1992.
(b) Precision based on a 90% confidence interval.
(c) Target Precision = 100*(6.24/SQRT(ER)), where ER = emissions in Bscf. Overall TP is +/-110.66 Bscf.
Maximum Relative Category TP Is +/-1500%, Minimun Relative Category TP is +/• 75%, where TP = target precision.
(d) Conservative precision based on upper limit of a 90% confidence interval. This confidence interval is based on a lognormal assumption.
BSCF EMISSION REPORT 10/16/96
-------
METHANE EMISSION AND ACCURACY ESTIMATES
Percent
Percent
Activity
Emission
Precision
Conservative
Target
PROCESS SEGMENT
1992
1992
of Total
of Total
Upper
Upper
of Annual
Precision
Precision
Emission Type
Emissions
Emissions
Emissions
Production
Value
Units
Bound
Value
Units
Bound
Emissions
of Annual
(%)
Source
(Tg)
(Bscf)
(%)
%
(a)
(b)
(b)
Emissions
(d)
(c)
Gas Processing Plants
Normal Fugitives
Plants
0.0403
2.0950
0.67
0.009
726
plants
2%
7906
scfd/plant
46%
48.05%
60.11%
431.12
Recip. Compressors
0.3216
16.7251
5.32
0.076
4,092
compressors
46%
11198
scfd/comp
74%
95.09%
141.67%
15258
Centrifugal Compressors
0.1082
5.6257
1.79
0.025
726
compressors
77%
21230
scfd/comp
39%
91.39%
134.71%
263.09
Vented and Combusted
Normal Operations
Compressor Exhaust
Gas Engines
0.1281
66824
2.12
0030
27,760
MMHPhr
133%
0.240
scf/HPhr
5%
133.26%
221.71%
241 75
Gas Turbines
0.0038
0.1876
006
0001
32.910
MMHPhr
121%
0.0057
scf/HPhr
30%
129.84%
214.17%
1440.74
AGR Vents
0.0158
0.8237
0.26
0.004
371
AGR units
20%
6083
scfd/AGR
105%
108.85%
169.48%
687.54
Kimray Pumps
0.0033
0.1703
0.05
0.001
957900
MMscflyr
192%
177.75
scf/MMscf
57%
228 00%
449.12%
1500.00
Oehydrator Vents
0.0202
1.0490
0.33
0.005
8,630,000
MMscf/yr
22%
121.55
scf/MMscf
202%
208.20%
399.58%
609.26
Pneumatic Devices
0.0023
0.1196
0.04
0.001
726
gas plants
2%
164721
scfy/plant
133%
133.04%
221.23%
1500.00
Routine Maintenance
Blowdowns/Ventinq
0.0567
2.9475
0.94
0.013
726
gas plants
2%
4060
Mscfy/plant
262%
262.16%
535.66%
363.46
(a) Based on a total gross national production of 22132 Bscf for 1992.
(b) Precision based on a 90% confidence interval.
(c) Target Precision = 100*(6.24/SQRT(ER)), where ER = emissions in Bscf. Overall TP is ~/• 110.68 Bscf.
Maximum Relative Category TP is +/• 1500%, Minimun Relative Category TP is +1- 75%, where TP = target precision.
(d) Conservative precision based on upper limit of a 90% confidence interval. This confidence interval is based on a lognormal assumption.
>
U>
BSCF EMISSION REPORT 10/16/96
-------
METHANE EMISSION AND ACCURACY ESTIMATES
>
Percent
Percent
Activity
Emission
Precision
Conservative
Target
PROCESS SEGMENT
1992
1992
of Total
of Total
Upper
Upper
of Annual
Precision
Precision
Emission Type
Emissions
Emissions
Emissions
Production
Value
Units
Bound
Value
Units
Bound
Emissions
of Annual
(%)
Source
(Tg)
(Bscf)
<*)
%
(a)
(b)
(b)
Emissions
(d)
(c)
TRANSMISSION/STORAGE
Fugitives
Pipeline Leaks
0.0031
0.1600
0.05
0.001
284,500
miles
5%
1.541
scfd/mile
89%
89.00%
130.14%
1500.00
Compressor Stations (TRANS)
Station
0.1047
5.4467
1.73
0.025
1,700
stations
10%
8778
scfd/station
102%
103.00%
157.55%
267.37
Recip. Compressor
0.7256
37.7333
12.01
0.170
6,799
comp.
17%
15205
scfd/comp.
65%
68.09%
92.35%
101.58
Centrifugal Compressor
0.1449
7.5328
2.40
0.034
681
comp.
26%
30305
scfd/comp.
34%
43.71%
53.67%
227.36
Compressor Stations (STOR)
Station
0.0717
3.7288
1.19
0.017
475
stations
5%
21507
scfd/station
100%
100.25%
152.05%
323.15
Recip. Compressor
0.2069
10.7594
3.42
0.049
1,396
comp.
58%
21116
scfd/comp
48%
8027%
113.88%
190.24
Centrifugal Compressor
0.0292
1.5176
0.48
0 007
136
comp.
119%
30573
scfd/comp.
34%
13021%
214 97%
506.53
Wells (STOR)
0.0145
07522
024
0 003
17,999
wells
5%
1145
scfd/well
76%
76.26%
106.64%
719.47
M&R (Trans. Co. Interconnect)
0.0708
36834
1.17
0017
2,533
stations
776%
3984
scfd/station
80%
996.98%
2197.40%
325.14
M&R (Farm Taps + Direct Sales)
0.0159
0.8271
0.26
0.004
72,630
stations
780%
31.2
scfd/station
80%
1002 09%
2207.28%
686.13
Vented and Combusted
Normal Operations
Dehydrator Vents (TRANS)
00020
0.1018
0.03
0.000
1,086,000
MMscf/yr
144%
93.72
scf/MMscf
208%
391.75%
864.25%
1500.00
Dehydrator Vents (STOR)
0.0045
0.2344
0.07
0.001
2,000,000
MMscf/yr
25%
117.18
scf/MMscf
160%
166.56%
298.24%
1288 98
Compressor Exhaust
Engines (TRANS)
0.1864
9.6912
3.08
0.044
40,380
MMHPhr
17%
0.240
scf/HPhr
5%
17.74%
19.35%
200.45
Turbines (TRANS)
0.0011
0.0549
0.02
0.000
9,635
MMHPhr
33%
0.0057
scf/HPhr
30%
45.68%
56.58%
1500.00
Engines (STOR)
00227
1.1813
038
0.005
4,922
MMHPhr
27%
0.240
scf/HPhr
5%
27.49%
31.39%
574.13
Turbines (STOR)
0 0002
0.0099
0.00
0.000
1729
MMHPhr
626%
0.0057
scf/HPhr
30%
654.25%
1485.73%
1500.00
Generators (Engines)
0.0091
04748
0.15
0.002
1,978
MMHPhr
45%
0.240
scf/HPhr
5%
45.25%
55.94%
905.60
Generators (Turbines)
0.0000
0.0001
0.00
0.000
23.3
MMHPhr
1114%
0.0057
scf/HPhr
30%
1163.33%
2510.01%
1500.00
Pneumatic Devices
0.2720
14.1446
4.50
0.064
87,206
devices
38%
162197
scfy/device
44%
60.49%
79.65%
165.92
Routine Maintenance/Upsets
Pipeline Venting
0.1732
9.0044
2.87
0.041
284,500
miles
5%
31.65
Mscfy/mile
236%
236.25%
469.92%
207.95
Station Venting
0.1823
9.4800
3.02
0.043
2.175
cmp stations
8%
4359
Mscfv/station
262%
26266%
536.93%
202.67
(a) Based on a total gross national production of 22132 Bscf for 1992.
(b) Precision based on a 90% confidence interval.
(c) Target Precision = 100*(6.24/SQRT(ER)), where ER « emissions in Bscf. Overall TP is ~/-110.66 Bscf.
Maximum Relative Category TP is ~/-1500%, Minimun Relative Category TP is 75%, where TP = target precision.
(d) Conservative precision based on upper limit of a 90% confidence interval. This confidence interval is based on a tognormat assumption.
BSCF EMISSION REPORT 10/16/96
-------
METHANE EMISSION AND ACCURACY ESTIMATES
Percent
Percent
Activity
Emission
Precision
Conservative
Target
PROCESS SEGMENT
1992
1992
of Total
of Total
Upper
Upper
of Annual
Precision
Precision
Emission Type
Emissions
Emissions
Emissions
Production
Value
Units
Bound
Value
Units
Bound
Emissions
of Annual
(%)
Source
300
0.1048
5.4510
1.73
0.025
3,460
stations
71%
179.8
scfh/station
39%
85.46%
123.47%
267 27
M&R 100-300
0.2149
11.1731
3.56
0.050
13,335
stations
106%
95.0
scfh/station
112%
19497%
366.89%
18068
M & R < 100
0.0052
0.2693
009
0.001
7,127
stations
118%
4.31
scfh/station
227%
370.94%
812.08%
1202.35
Reg > 300
0.1090
56655
1.80
0.026
3,995
stations
68%
161.9
scfh/station
58%
97 37%
146.35%
262.16
R-Vault > 300
0.0005
0.0266
0.01
0000
2,346
stations
68%
1.30
scfh/station
182%
230.44%
455.26%
1500.00
Reg 100-300
0.0837
4.3520
1.38
0.020
12,273
stations
61%
40.5
scfh/station
66%
98.47%
148.52%
299.12
R-Vault 100-300
0.0002
0.0087
0.00
0.000
5,514
stations
61%
0.180
scfh/station
94%
126.14%
206.09%
1500.00
Reg 40-100
0.0084
0.3317
0.11
0.001
36,328
stations
64%
1 04
scfh/station
74%
109.09%
169.96%
1083.42
R-Vault 40-100
0.0005
0.0244
0.01
0.000
32,215
stations
64%
0.0865
scfh/station
64%
98.97%
149.51%
1500.00
Reg < 40
0.0003
0 0179
0.01
0.000
15,377
stations
65%
0.133
scfh/station
135%
173.87%
315.67%
1500.00
Customer Meters
Residential
0.1007
5.5468
1.78
0.025
40,049,306
outdr meters
10%
1385
scfy/meter
17%
19.80%
21.80%
264.95
Commercial/Industry
0.0042
0.2207
0.07
0.001
4,608,000
meters
5%
479
scfy/meter
35%
35.40%
41.91%
1328.20
Vented
Routine Maintenance
Pressure Relief Valve Releases
0.0008
0.0418
0.01
0000
830,700
mile main
5%
0.050
Mscf/mile
3914%
391889%
6199.19%
1500.00
Pipeline Slowdown
0.0025
01324
0.04
0.001
1,297,569
miles
5%
0.102
Mscfy/mile
2521%
2524.15%
4579.76%
1500.00
Upsets
Mishaps (Dia-ins)
0.0397
2.0031
060
0.009
1.297.509
miles
5%
1.59
Mscfy/mile
1922%
1924.41%
3751.65%
434.43
INDUSTRY TOTAL EMISSIONS
0.0437
314.2714
100.0000
1.4200
28.51%
32.71%
35.21
UNCERTAINTY (~/-)
0.0172
89.6029
(a) Based on a total gross national production of 22132 Bscf for 1992.
(b) Precision based on a 90% confidence interval.
(c) Target Precision = 100*(6.24/SQRT(ER)), where ER = emissions in Bscf. Overall TP is ~/-110.06 Bscf.
Maximum Relative Category TP is +/• 1600%, Minimun Relative Category TP is ~/- 75%, where TP = target precision.
(d) Conservative precision based on upper limit of a 90% confidence interval. This confidence interval Is based on a lognormal assumption.
BSCF EMISSION REPORT 10/16/96
-------
APPENDIX B
Effect of Methane Emissions on Global Warming
B-l
-------
B.O EFFECT OF METHANE EMISSIONS ON GLOBAL WARMING
Based on the recent climate change reports by the Intergovernmental Panel on
Climate Change (IPCC), "switching from coal to oil or natural gas, and from oil to natural gas,
can reduce (greenhouse gas) emissions.'" The GRI/EPA study to estimate methane emissions
from natural gas operations was undertaken primarily because this information was needed to
determine if it makes sense to promote the increased use of natural gas as a strategy for reducing
greenhouse gas emissions. This appendix attempts to put the results into perspective by
examining whether the current estimate of methane emissions from natural gas operations is
likely to affect this fuel switching strategy.
Carbon dioxide contributes as much to global warming as all other greenhouse
gases combined. Natural gas emits substantially less C02 per unit of energy generated than
either coal or oil.1 However, methane, a more potent greenhouse gas than C02, is also emitted in
the production, transmission and distribution of natural gas. The question raised was whether the
fuel switching strategy is valid when emissions of all greenhouse gases are considered over the
complete fuel cycle (production through end use combustion). To address this question, it is
necessary to account for emissions of all greenhouse gases throughout the fuel cycle and to
determine the impact of these gases on global warming.
Fortunately, in evaluating fossil fuel emissions, emissions of greenhouse gases
other than methane and C02 are negligible and do not need to be considered. In addition, most
of the C02 emissions result from fuel combustion and are accurately known. The uncertainty in
estimating C02 emissions from production and transportation of the fuel is higher, but this is a
relatively small value and does not have a large effect on the overall accuracy of the analysis.
Estimates of methane emissions from natural gas operations prior to the GRI/EPA study
generally ranged from 2 to 5 % of production.2-3,4,5 The uncertainties in methane emissions from
coal and oil production were equally as large. Although the uncertainty in emissions is still
relatively large, the largest uncertainty in addressing the validity of the fuel switching strategy is
B-2
-------
in determining the relative impacts of C02 and methane emissions on global warming. In order
to simplify a comparison of the impact of one greenhouse gas with another, a global warming
potential (GWP) has been defined.6 The GWP is an index that relates the impact of a given
greenhouse gas to an equal amount (by mass) of C02. The projected effect on global warming of
a greenhouse gas over a chosen time horizon can be estimated by multiplying the appropriate
GWP by the amount of gas emitted. Considerable work has been done in this area by the
Intergovernmental Panel on Climate Change (IPCC). However, as discussed below, the
uncertainty in the GWP is still very large.
B.l Global Warming Potential
greenhouse heating potential depends on its lifetime and the rate at which it is injected into the
atmosphere. The GWP for a trace gas addresses the net effect of the radiative forcing and the
lifetime of the gas by calculating the time integrated radiative forcing of a unit mass impulse to
the atmosphere.6 The GWP is defined as the impact on global warming caused by an incremental
amount of a given greenhouse gas divided by the impact of releasing an equivalent amount of
C02. The GWP for methane can be approximated by the following equation:
Although a trace gas can have a strong radiative forcing per molecule, its
GWP
(Bl)
where
C02
to
(ti-0
A
C
ch4
radiative forcing per unit mass
concentration
subscript designating methane
subscript designating C02
time of release
time period over which the GWP is evaluated
B-3
-------
The concentration is often approximated by the expression:
C - e* (B2)
where t = time; and
A, = 1/lifetime
In addition to the direct effect of methane on global warming (the radiative
forcing due to methane itself, as given above), methane can also contribute to the formation of
other greenhouse gases such as tropospheric ozone and water vapor in the stratosphere. These
indirect effects of methane must be added to the direct effect to determine the total contribution
of methane.
IPCC has published the results of studies to evaluate the GWP for the various
greenhouse gases. Their findings in 1990,1992, and 1994 are shown in Table B-l for the direct
effects of methane for different integration time periods (t,-t0).7,8'9 The GWPs for methane,
including both the direct and indirect effects, are also presented in Table B-l for 1990, 1994, and
1995. Because the IPCC believed that the uncertainties in the indirect effects were very large,
they decided not to publish a total GWP for methane in 1992. The change in values from 1994
to 1995 reflects a change in the lifetimes of gases that react primarily with tropospheric hydroxyl
radical (OH) concentration, based on a revised estimate in the mean global OH concentration.6
B-4
-------
TABLE B-l. GLOBAL WARMING POTENTIAL OF METHANE
Integration Interval (years)
Direct Effects
Direct and Indirect Effects
Year
20
100
500
L 20
100
500
1990
35*
11
4
63
21
9
1992
35
11
4
—
--
—
1994
43-52
12-21
5-6
62 ±20
24.5 ± 7.5
7.5 ±2.5
1995
56 ±20
21 ± 7.4
6.5 ±2.3
* A value of 35 indicates that one pound of methane has the same effect as 35 pounds of C02
—Data not available
As implied by the variations in the values shown in Table B-l, there are
significant uncertainties in the GWP for methane. Some of these uncertainties result from
differences in the models and model limitations. In 1990 and 1992 the lifetime of methane was
determined by calculating the decay rate while the composition of the atmosphere was held
constant. In the 1994 and 1995 calculations, the atmosphere was allowed to respond to the
change in methane by coupling the methane chemistry to the calculation of the radiative forcing.
This resulted in a reduction in the OH concentration. Since OH is primarily responsible for the
oxidation of methane, the lifetime of methane in the atmosphere increased. The effect of
including the chemistry was initially thought to be small, but as shown in Table B-l, the 1994
GWPs increased by approximately 35 to 50%. The change in GWP from 1994 to 1995 includes
a decrease of about 10% based on an improved estimate in the concentration of methyl
chloroform which is used as a reference compound in determining the mean global OH
concentrations.
Other issues could also have significant effects on the calculated GWP when they
are eventually addressed. Some examples are:
B-5
-------
• The size of the incremental increase in methane. A relatively large pulse of
methane is used in the models to evaluate the GWP. Because the chemistry
is highly nonlinear, a large pulse can generate a nonlinear change in the
lifetime of methane that would produce a much larger GWP than using a
small pulse.
• Grid size. Emissions of NOx and other gases are smeared over large grid
cells and are artificially diluted. Because NOx/methane/ozone chemistry is
highly nonlinear, this could have a significant effect on the tropospheric
ozone calculation and the evaluation of indirect effects of methane on GWP.
• Nonmethane hydrocarbons fNMHCsl NMHCs are not included in the
current models. Since most NMHCs are more reactive than methane, the
impact of increased methane emissions on tropospheric ozone could be
overstated. This would cause an erroneously high value for the indirect
contribution to GWP for methane.
Of all the parameters discussed, however, the parameter that has the largest effect
on GWP is the time interval (t,-t0) used in evaluating GWP. There is not a consensus on the
proper value, particularly between policy analysts and scientists. Some policy analysts use a
time interval as short as 50 years. For methane, the time interval is an important question
because there is a large difference in the lifetime of methane (approximately 12.2 years ± 25%)
and the effective lifetime of C02 (200 to 250 years). If a time period of 50 years is selected, the
GWP calculated by the IPCC would be approximately 34. The implication is that one pound of
methane released into the atmosphere would have the same impact on global warming as
34 pounds of C02. The problem is that this is only true for the first 50 years. The amount and
percentage of methane and C02 in the atmosphere based on releasing 34 pounds of C02 for each
pound of methane (i.e., a GWP of 34) is presented in Table B-2 as a function of time after
release. At the end of 50 years, the methane concentration would have decreased to a negligible
level, but approximately 80% of the C02 would remain in the atmosphere and still contribute to
global warming.
B-6
-------
TABLE B-2. AMOUNT OF C02 AND METHANE REMAINING IN
ATMOSPHERE WITH TIME
Greenhouse Gas
Time (Y rs.)a
20 >'1fH
10°
500
% C02
90
78
61
8
% ch4
19
1.6
0.03
—
lbs. C02b
31
26
21
3
lbs. CH4
0.19
0.02
—
--
a Lifetime of methane and C02 used were 12.2 and 200 years, respectively.
b Assumed GWPM = 34.
If a 50-year time interval is used to develop emission trading policies, then
1 pound of methane emissions could be traded for 34 pounds of C02. The problem is that after
50 years the 1 pound of methane would have decayed to less than 0.02 pounds, but there would
still be 26 pounds of C02 remaining in the atmosphere. A century later, 21 pounds of the
original 34 pounds of C02 released would still be contributing to global warming. These
contributions of C02 are neglected by choosing a time period of 50 years.
In considering the impact of using different types of fuels, the time interval should
be chosen so that both gases (in this case methane and C02) would have time to decay to
negligible values. This suggests that the time interval for evaluating the GWP for methane
should be in the range of 500 to 1000 years.
There currently is not a consensus on the integration interval. Because the GWP
could be as low as 6.5 for a 500-year integration interval and as high as 34 for a 50-year interval
(over five times larger), two approaches were taken in this analysis to examine the validity of the
fuel switching strategy. In the first approach, a breakeven percentage is calculated. The
breakeven percentage is the amount of methane that would have to be released during the natural
gas fuel cycle to eliminate the advantage that natural gas has over coal and oil because of its
lower C02 emissions. The breakeven percentage can be compared to the 1992 emission
inventory for the natural gas industry to evaluate the relative advantage that natural gas has over
B-7
-------
coal and oil. This approach is presented in Section B.2. In addition, to determine the validity of
the fuel switching strategy, the breakeven percentage can be compared to the percentage of gas
leaked due to the incremental increase in gas use that results from fuel switching from coal or oil
to natural gas. This is presented in Section B.3. The second approach, presented in Section B.4,
is to evaluate the amount of equivalent C02 emissions for each fossil fuel over the fuel cycle by
converting total greenhouse gas emissions to "equivalent C02."
B.2 Breakeven Percentage
The first approach for evaluating the fuel switching analysis requires comparing
the breakeven percentages of the various fuels. The breakeven percentage (BP) is the amount of
methane that would have to be released in the production, distribution, and end use of natural gas
for it to have the same impact on global warming that the fuel cycle of coal or oil would have.
The breakeven percentage can be calculated knowing the GWP for the different greenhouse
gases and the amount of each greenhouse gas released per unit of energy from the fuel cycle of
natural gas, coal, and oil. The equation used is given below, along with the parameters used in
the calculation.
8? = 200 E, ~ E
G GWP
— + EM;
(Bll)
M
where
E
pounds of C02 emitted from the fuel cycle for 106 Btu of fuel
G
pounds of methane in 106 Btu of natural gas
EMj
pounds of methane emitted from the fuel cycle for 106 Btu's of fuel
GWP
global warming potential calculated on a mass basis
subscript denoting type of fossil fuel (c for coal, o for oil)
NG
subscript denoting natural gas
B-8
-------
M = subscript denoting methane
The results for coal are presented in Figure B-l as a function of the GWP. Table
B-3 presents the breakeven percentage for oil and coal based on IPCC's GWPs calculated for
various time intervals, using the following values.
G
38 pounds per MMBtu, based on an HHV of 1,031 for natural gas
and a methane composition of 93.4%
Ec =
208 pounds per MMBtu
Eng
127 pounds per MMBtu
Eo =
184 pounds per MMBtu
EMC =
0.6 pounds per MMBtu
EM0 =
0.06 pounds per MMBtu
TABLE B-3. BREAKEVEN PERCENTAGE (BP) FOR COAL AND OIL FOR
VARIOUS GWP INTEGRATION INTERVALS
Integration Interval
GWP
BP
BP
(for coal)
(for oil)
50
34
8
5
100
21
12
7
500
6.5
34
23
As shown for a GWP of 6.5, approximately 34% of the natural gas produced
would have to be leaked for natural gas to have the same impact on global warming as coal, or
for oil, 23% of natural gas would have to be leaked to have the same impact. For a GWP of 34,
the percentage is approximately 8% for coal and 5% for oil.
All the breakeven percentages are substantially larger than the percent of methane
emitted from natural gas operations (1.42% of production for 1992). This indicates that natural
gas has an inherent advantage over the other fuels for the 1992 base case.
B-9
-------
30
25
20
15
10
5
0
100
0
GWP
Figure B-l. Breakeven Percentage - Natural Gas Compared with Coal
B-10
-------
For the fuel switching strategy, where natural gas consumption could replace
some of the energy supplied by coal or oil, the breakeven percentage needs to be compared with
the emissions that would result from an incremental increase in gas use. As will be shown in the
next section, the incremental increase in emissions (above the 1992 baseline) are between 0.3%
and 1.0% of the incremental increase in gas production, which is approximately one-third to two-
thirds of the base year methane emissions (1.42% of the total gas production rate).
B.3 Emissions from Increased Gas Sales
As part of the GRI/EPA project, a separate study was conducted to determine the
percent increase in emissions caused by an incremental increase in natural gas production and
sales.10 This study examined the consequences of increasing gas sales by 5, 15 and 30% under
three scenarios: uniform, winter peak, and summer peak load profiles.
All segments of the gas industry were examined to determine the percent increase
in equipment that would be needed to meet the increased demand. The percent increase in
emissions was then estimated based on changes in the current system that would be required to
accommodate the increase in gas sales. GRI/EPA's emission estimates were used to calculate
the percent increase in emissions that would result from an incremental increase in natural gas
sales for seven scenarios. The results are presented in Tables B-4 and B-5 for two cases:
expected and upper limit, respectively. The assumption for the values listed under "expected"
was that the system will be expanded using the latest technologies. The assumption for the
values listed under "upper limit" was that the expanded system mirrors the existing system (i.e.,
new equipment or technologies for reducing emissions are not utilized).
For most components, facility and operating changes are not linearly related to
increased gas throughput due to excess capacity or practices such as pipeline looping. Therefore,
the study showed that an increase in gas use for either a system mirroring current technology or a
system utilizing the latest technology would increase emissions by an amount less than the
B-ll
-------
TABLE B-4. INCREMENTAL CHANGES IN EMISSIONS
RESULTING FROM INCREASED GAS SALES - EXPECTED CASE
Base Case
Increased System Throughput (%)
Uniform Load
Winter Peak
Summer
Peak
5,^;
30
, ,1 S^i-
Total Emissions, Bscf
314
319
328
343
320
333
352
324
% Increase over Base Case
-
1.37
4.43
9.20
1.90
5.84
12.0
2.98
Total Emissions/
1.42%
1.37%
1.29%
1.19%
1.38%
1.31%
1.22%
1.27%
Total Gas Production Rate
A Emissions/A Production Rate
--
0.39%
0.42%
0.44%
0.54%
0.55%
0.57%
0.28%
TABLE B-5. INCREMENTAL CHANGES IN EMISSIONS
RESULTING FROM INCREASED GAS SALES - UPPER LIMIT CASE
Base Case
Increased System Throughput (%)
Uniform Loa<
Winter Peak
Summer
Peak
30
15
Total Emissions, Bscf
314
319
336
361
321
346
380
331
% Increase over Base Case
--
1.37
6.98
15.0
2.12
9.98
21.0
5.42
Total Emissions/
1.42%
1.37%
1.32%
1.26%
1.38%
1.36%
1.32%
1.30%
Total Gas Production Rate
A Emissions/A Production Rate
~
0.39%
0.66%
0.71%
0.60%
0.95%
0.99%
0.51%
-------
percent load increase. For the expected system, total emissions (Bscfy of methane) increase by
1.4% to 12% over the load scenarios examined for corresponding increases in gas sales of 5% to
30%. The incremental methane emission increase (5 to 38 Bscfy), when divided by the
incremental production rate increase (1,110 to 6,640 Bscf natural gas), results in emissions per
production percentages of 0.3 to 0.6%. For the upper limit case, total emissions increase by
1.4% to 21% for the same scenarios. The incremental methane emission increase for these
scenarios (5 to 66 Bscfy), when divided by the incremental production rate increase, results in
emissions per production percentages of 0.4 to 1.0%. Compared to the base year emissions per
production percentage of 1.42 %, the incremental emission rates (A methane emissions per A
production volume) are only one-third to two-thirds of the base emission rate. The incremental
emission percentages are much lower than the breakeven percentages of either coal (8 to 34%) or
oil (5 to 23%) based on a GWP time interval of 50 or 500 years, respectively. Therefore, this
analysis supports the validity of the fuel switching strategy.
B.4 Equivalent CO, Emissions
The second approach used to examine the validity of the fuel switching strategy is
based on quantifying the emissions of methane and C02 for each fuel over the fuel cycle and then
converting the methane emissions to equivalent C02 by multiplying by the GWP. The GWP
relates the radiative forcing of other greenhouse gases, such as methane, to the radiative forcing
of C02 over a period of time, accounting for the changing concentration of the greenhouse gases
over time. For a given fuel, the equivalent C02 emissions are calculated using the following
equation:
n GHG/
Equivalent C02 = £ Ec0 + £ (GWPjxEj) (®4)
i=l[ 2 j=l ' JJ
where:
B-13
-------
i = subscript that denotes the various fuel cycle operations
(production, transportation, processing, and combustion)
j = subscript to denote the various greenhouse gases (GHGs)
ECo2 = mass of C02 emissions (lb) per energy input
GWPj = global warming potential used to convert emissions of GHG, "j,"
to equivalent C02 emissions
Ej = mass of other greenhouse gas emissions (lb) per energy input
This equation results in the total fuel cycle emissions by accounting for:
1) End use C02 emissions from fuel combustion;
2) C02 emissions by the industry resulting from the production, processing,
and transportation of the fuel; and
3) C02 equivalent emissions that result from industry methane emissions.
For the purpose of the fuel switching analysis, equivalent C02 emission factors of
the various energy sources are reported as the mass of equivalent C02 emissions per unit of
energy. Therefore, for an energy requirement of one million Btu, the relative contribution of
equivalent C02 emissions of the various fuels can be compared.
The energy content of the fuel can be expressed in terms of either the lower
heating value (LHV) or the higher heating value (HHV). The difference between lower (or net)
and higher (or gross) heating value is the heat of vaporization (AfJ) from the moisture produced
during combustion, where the higher heating value includes this amount:
HHV = LHV + (H20) (B5)
where:
n = moles of water produced
B-14
-------
A^(H20) = heat of vaporization of water at 25 °C
The difference between the higher and lower heating values can be significant
when comparing the combustion efficiency from various end use equipment, since the latent heat
is recovered by some end-use equipment, but not by all end-use equipment. For this study, the
efficiencies of end use equipment are not considered. The fuels are compared on a higher
heating value basis, which is the convention commonly used in the U.S. and is also the
convention used by IPCC.111 The general methodology is shown in the following equation:
Emissions (mass/yr) v (mass) — Emissions (massV_r.
1 fuel (ov)
Marketed Fuel Production (mass/yr) (MMBtu) MMBtu
C02 and methane are the only greenhouse gases, related to fuel use, that make a
substantial contribution to global warming.12 (Fuel combustion also contributes to N20
emissions, but these emissions result primarily from mobile source combustion which is not
considered in this analysis.12) Methane emissions resulting from the production of gas, oil, and
coal must be considered, as well as emissions from the transportation and distribution of natural
gas. Methane emissions from the transportation of crude/refined product and coal are small, and
methane emissions from the end use combustion of natural gas, oil, and coal are negligible.
The results of the equivalent C02 emissions analysis are presented in Table B-6,
which summarizes the data sources used to develop each equivalent C02 emission estimate and
the values that resulted for a GWP of 34. As stated earlier, the equivalent C02 emissions for
each type of fuel were developed from three basic parts: 1) combustion end use emissions of
C02,2) C02 emissions from production through transport, and 3) methane emissions converted
to equivalent C02.
As the table shows, combustion end use emissions are the largest contributor for
all fuel types. Approximately 76% (for GWP of 34) of the C02 equivalent emissions per
MMBtu from natural gas are from end use combustion. For fuel oil and coal, nearly 90% of total
B-15
-------
TABLE B-6. SOURCES OF C02 EQUIVALENTS FOR EACH FUEL TYPE
Emission Type/Source
Natural Gas
Fuel Oil
Coal
lb Eq.
C02 per
MMBtu
Data Source
lb Eq.
C02 per
MMBtu
Data Source
lb Eq.
C02 per
MMBtu
Data Source
Combustion of fuel in end
use (direct C02)
115.6
Based on fuel
content and HHV
(this study)
164.4
ABB/Combustion
Engineering13
205.8
EPA Greenhouse Gas
Report10
Methane emissions from
production, processing,
refining, transportation
(methane converted to C02
based on a GWP of 34)
24.5
GRI/EPA Methane
Emissions (this
study)
2.0
API Study14
20.4
EPA Reports12'15
C02 emissions during
production, processing,
refining, transportation
(direct C02)
11.7
AP-42 Emission
Factors, Activity
Factors from this
study
19.6
API Study'4
2.1
Energy International13
TOTAL
lb Equivalent C02/MMBtu
152
186
228
-------
C02 emissions per MMBtu are attributable to end-use combustion. The C02 emissions from end
use combustion are well defined, since they depend primarily upon the carbon content of the
fuels. Therefore, the uncertainty associated with the largest portion of the equivalent C02
estimate is relatively small.
The C02 emissions from production through transport and the methane emissions
from coal and oil operations play a much smaller role in the overall comparison.
Methane emissions comprise only 16% of the equivalent C02 emission estimate
for natural gas, 1.1% for oil, and 9% for coal. Therefore the impact of the methane emission
estimate is far less than that of the end use component. Methane emissions for natural gas are
well known (± 33 %), while the estimates for oil and coal may have much wider confidence
bounds (possibly with an upper bound larger than 100%). Therefore the equivalent C02 emission
comparison for natural gas is conservative, since the emissions from coal and oil may be much
higher.
The following sections on natural gas, coal, and oil describe the methods and
assumptions used to determine the equivalent C02 emissions for each fuel type that were
presented in Table B-6.
Natural Gas
Approximately 116 lbs of C02/MMBtu is emitted from the combustion of natural
gas. This emission rate was calculated assuming the complete combustion of marketed natural
gas (17.84 x 106 scf for 1990)16 with a gross energy content of 1031 Btu/scf12 and the
corresponding composition of 93.4% methane, 4.0% ethane, 0.5% propane, and 2.1% inerts.
(Note: the mole percents of ethane, propane, and inerts were determined by weighting the
respective higher heating values to achieve the desired methane composition and energy content
of the gas mixture.)12'17
B-17
-------
Carbon dioxide is also emitted through combustion in compressors, burners, and
flares. Emissions from gas-fired compressor engines used to transport natural gas from
production to market were determined using a C02 emission factor of 0.89 lb C02/hp hr18 and the
total hp-hr for production, transmission, and processing of 145 x 109 hp-hr.19 This results in 7.0
lbs C02/MMBtu emissions. Similarly, carbon dioxide emissions result from burning natural gas
for other plant, lease, or pipeline fuel requirements. The amount of natural gas used for fuel
purposes other than compressors was estimated in the Vented and Combustion Source Summary
to be approximately 558 Bscfy.17 This results in an additional 3.6 lb C02/MMBtu, based on the
C02 emission factor from natural gas combustion of 120 lb C02/Mscf.2° It should be noted that
this estimate is conservatively high since a portion of the fuel gas is used by the petroleum
industry to operate equipment such as gas-lift compressors and heater-treaters. Finally, a small
amount of C02 is generated from flaring natural gas. The Vented and Combustion Source
Summary estimates 15.1 Bscfy of methane is flared from production through distribution.17
Based on a 98% to 99% combustion efficiency, where all of the methane combusted is assumed
to form C02, 0.1 lb C02/MMBtu result from flaring.
Methane emissions from the production, transmission, gas processing, and
distribution of natural gas are approximately 314 Bscf or 6.04 Tg/yr (methane emissions from
end uses are negligible). Based on the marketed gas volume of 17.84 x 106 scf and the natural
gas HHV of 1,031 Btu/scf,12,14 the methane emissions equate to 0.72 lbs CH4/MMBtu.
Converting the methane emissions to equivalent C02 emissions requires a GWP, which for
methane is 34 for an integration interval of 50 years and is 6.5 for an integration interval of 500
years.6-7'8'9 Applying these conversion factors, the equivalent C02 emissions for methane are then
24.5 lbs C02/MMBtu for a GWP of 34, and 4.7 lbs C02/MMBtu for a GWP of 6.5.
Natural gas results in a total of 132 lbs C02/MMBtu for a GWP of 6.5, and 152
lbs C02/MMBtu for a GWP of 34. Table B-7 summarizes the various components that
contribute equivalent C02 emissions from the natural gas fuel cycle.
B-18
-------
TABLE B-7. CP2 EQUIVALENT EMISSIONS FROM NATURAL GAS
lb C02 Equivalent/MMBtu
Emission Source
GWP = 6.5
GWP = 34
End-use
Combustion C02 Emissions
115.6
115.6
Industry
Compressor C02 Emissions
Burner C02 Emissions
Flare C02 Emissions
Methane Emissions
7.0
4.6
0.1
4.7
7.0
4.6
0.1
24.5
TOTAL
132.0
151.8
Coal
Methane emissions result primarily from coal mining; emissions of methane from
the transport or end uses of coal are negligible. EPA's Inventory of Greenhouse Gas Emissions
and Sinks reports methane emissions from coal mining activities ranging between 3.2 and 5.0
million metric tonnes of methane for the year 1992.12 Based on the 1992 coal production of
997.5 million short tons,21 and the coal higher heating value of 10,395 Btu/lb,15 methane
emissions from this source equate to 0.44 lbs methane/MMBtu. In comparison, another EPA
report shows 30 to 50 million metric tonnes of methane emitted globally corresponding to coal
production of 5 billion tonnes.13 For the same heating value, these values result in 0.77 lbs
methane/MMBtu. An average of the two sources (0.60 lbs methane/MMBtu) was used to
estimate equivalent C02 emissions of 3.9 lbs C02/MMBtu for a GWP of 6.5 and approximately
20.4 lbs C02/MMBtu for a GWP of 34.
The primary source of C02 emissions from coal results from combustion. EPA
reported that C02 emissions from energy production were 430.4 million metric tonnes of carbon
equivalent (MMTCE) for 1992.12 The energy generated from coal consumption for that year was
16,910 trillion Btu.12 Based on these values, C02 emissions are approximately 206 lbs
C02/MMBtu. In addition, C02 emissions from production and transportation equipment and the
B-19
-------
loss of coal during transport are estimated to result in an additional increase in C02 emissions of
approximately 1%, or 2.1 lbs C02/MMBtu.14
Table B-8 summarizes greenhouse emissions from the coal fuel cycle. The result
is an equivalent C02 emission rate of 212 to 228 lbs C02/MMBtu, depending on the GWP for
methane.
TABLE B-8. COz EQUIVALENT EMISSIONS FROM COAL
Emission Source
lbC02 Equivalent/MMBtu
GWP = 6.5
GWP = 34 .t
, T T CO, Combustion Emissions
EndUse (Electric Utilities)
205.8
205.8
t j Other CO, Emissions
Industry MethaneEmissions
2.06
3.9
2.06
20.4
TOTAL
212.3
228.3
Oil
A study was conducted by the American Petroleum Institute (API) for the
petroleum industry to quantify methane and C02 emissions resulting from petroleum operations
(production through transportation of refinery products) for the base year 1990.22 End use
emissions were not included in the API study. Emission estimates for production, crude
transportation, refining and product transportation are presented in Table B-9.
B-20
-------
TABLE B-9. 1990 METHANE AND C02 EMISSIONS FROM CRUDE
PRODUCTION THROUGH REFINED PRODUCT TRANSPORTATION
Industry Segment
Methane Emissions
tons Methane
COz Emissions,
Million tons C02/yr
Production
823,609
95.16
Crude Transport
11,192
8.87
Refining
13,845
171.24
Product Transport
0
8.77
TOTAL
848,646
284.04
For the fuel switching analysis, estimating equivalent C02 emissions from the oil
industry on a basis comparable to emissions from natural gas and coal is complicated by the
many different products generated from crude oil. Starting in refining, emissions from individual
fuel products are directly related to the emission sources associated with those products; before
the refining segment, a direct relation is not possible. Some portion of emissions generated from
crude production and transport must be assumed to be associated with individual fuel products.
For the purpose of this study, the fuels of interest (i.e., those comparable with the primary uses of
natural gas and coal—residential heating and generating electricity) are distillate and residual
fuel oils. Emissions associated with these fuel oils are assumed to be proportional to the ratio of
the mass of distillate and residual fuel produced in refining to the mass of the total refinery crude
charge:
Fuel oil produced (mass) ^ Total fuel oil emissions (mass)
Refinery crude charge (mass) Total crude emissions (mass)
Therefore, the total emissions reported in Table B-7 will be scaled according to the following
equation to estimate emissions associated with distillate and residual fuel oil only:
. . Fuel oil produced (mass)
E _ /t (mass/yr) * - -— (B8)
p ~ p Refinery crude charge (mass)
B-21
-------
where:
Ep-»p/t = Emissions of C02 or methane from production through product
transportation
The mass of fuel oil produced for market in 1992 is 2.215xl08 tons. This is based
on 1.09xl09 bbls of distillate23 with a specific gravity of 0.8654,24 resulting in 1.65xl08 tons
distillate produced in 1992. In addition, residual production of 5.64xl07 tons is based on
3.26xl08 bbls of residual23 with a specific gravity of 0.982 (which is estimated from the specific
gravities of fuel oils Nos. 2,4, 5, and 6 weighted by the relative volumes of each produced).24
The mass of refinery crude charge for 1992 is 7.53 x 108 tons (based on 4.91 xlO9 bbls of crude
and an average crude specific gravity of 0.876).23,25 The resulting ratio of fuel oil to crude charge
is approximately 0.294.
The general methodology for estimating emissions from the fuel cycle of distillate
and residual oils is then:
EP - p* (mass/yr) * 0.29 x rhv (mass) + E.„d (mass> = E (mass)
Marketed Fuel Oil Production (mass/yr) FueI 0,1 (MMBtu) MMBTUend use MMBtu
(B9)
which includes the addition of end use emissions from distillate and residual fuels (not included
in the API study). The combined higher heating value for distillate and residual fuel oil is 19,194
Btu/lb based on the individual HHVs (19,524 Btu/lb for distillate and 18,228 Btu/lb for
residual)25 weighted by the relative production rate of each fuel (presented above).
The resulting C02 emissions from production through product transport are 19.6
lb C02/MMBtu. Methane emissions equate to 0.059 lb methane/MMBtu. When converted to
equivalent C02 emissions, this results in 0.38 lb equivalent C02/MMBtu for a GWP of 6.5 and
2.0 lb equivalent C02/MMBtu for a GWP of 34.
B-22
-------
End use emissions of methane, from the combustion of petroleum products in
turbines or boilers, are negligible. However, C02 end use emissions from these sources are
significant. A Combustion Engineering report provided fuel oil properties,24 which were used to
calculate combustion emissions based on the following equation:
lb fuel lb C 44 lb COz gaj fuei lb C02
x x x — (jJlUJ
gal fuel lb fuel 12 lb C MMBtu MMBtu
This assumes all of the carbon present in the fuel oil is combusted to form C02. The properties
of distillate and residual fuel oils and the corresponding C02 combustion emissions are shown in
Table B-10.
TABLE B-10. PROPERTIES OF FUEL OILS
Property
Distillate Fuel Oil
Residual Fuel Oil
Density, lb/gal
7.206
8.212
% Carbon
86.4
85.7
HHV, MMBtu/gal
0.141
0.150
lb C02/MMBtu
161.9
172.0
These values were combined to generate one end use emission estimate for fuel
oils based on a weighted average with respect to the production rate of each fuel oil type
(1.65xl08 tons distillate and 5.64x107 tons residual, as discussed previously). The resulting C02
end use emissions from fuel oils used in residential heating and electricity generation are 164.4 lb
C02/MMBtu.
Table B-l 1 summarizes the emission estimates for the fuel cycle of residual and
distillate fuel oils.
B-23
-------
TABLE B-ll. C02 EQUIVALENT EMISSIONS FROM FUEL OIL
Emission Source
lb C02 Equivalent/MMBtu
GWP = 6.5
GWP = 34
End Use Fuel Combustion C02 Emissions
164.4
164.4
t , Prod, through Product Transport C02 Emissions
Prod, through Product Transport CH4 Emissions
19.6
0.38
19.6
2.0
TOTAL
184.4
186.0
B.5 Global Warming Conclusions
Table B-12 lists the greenhouse gas emissions, expressed as pounds of equivalent
C02 per MMBtu for natural gas, coal, and oil. To quantify the relative impacts on global
warming of coal and oil compared to natural gas, the equivalent C02 emissions per unit of energy
for coal and oil are divided by the value for natural gas. This "equivalent C02 ratio" is listed in
Table B-12 for GWPs of 6.5 and 34.
TABLE B-12. EQUIVALENT CO, EMISSIONS FOR NATURAL GAS,
OIL, AND COAL
Fuel Source
lbs C02/MMBtu
Equivalent
C02 Ratio
GWP = 6.5
GWP = 34
GWP = 6.5
- GWP = 34
Gas
132
152
1.0
1.0
Oil
184
186
1.4
1.2
Coal
212
228
1.6
1.5
Using oil has between 1.2 and 1.4 times the impact on global warming emissions
than the use of natural gas. Similarly, coal contributes 50 to 60% more equivalent C02 emissions
than natural gas, resulting in 1.5 to 1.6 times the impact on global warming. These results are in
basic agreement with IPCC's conclusions on fuel switching to reduce greenhouse gas emissions:
B-24
-------
Switching from coal to oil or natural gas would reduce carbon emissions in
proportion to the carbon intensity of the fuel. For example, switching from coal
to natural gas would reduce emissions by 40%. In addition, the higher energy
efficiency available with natural gas would reduce emissions further—for
example, a shift from coal to natural gas in power generation by 20%.1
The net result is that switching from other fuels to natural gas can help the United States reach its
goals on limiting greenhouse gas emissions and their potential impact on global warming.
B-25
-------
B.6
REFERENCES
1. World Meteorological Organization. Climate Change 1995 Impacts, Adaptations
and Mitigation of Climate Change: Scientific-Technical Analyses.
Intergovernmental Panel on Climate Change, United Nations Environment
Programme, 1996.
2. Ehhalt, D.H. The Atmospheric Cycle of Methane, Tellus, 26 55-70, 1974.
3. Seiler, W.R. Conrad and D. Scharffe. "Field Studies of Methane Emission
from Termite Nests into the Atmosphere and Measurements of Methane Uptake
by Tropical Soils, J. Atmos. Chem, 1, 171-186, 1984.
4. Crutzen, PJ. "Role of the Tropics in Atmospheric Chemistry," Geophysiology
of Amazonia, edited by R.E. Dickinson, pp. 107-130, John Wiley, New York,
NY, 1987.
5. Sheppard, J.C., H. Westberg, J.F. Hopper, K. Ganessan, and P.
Zimmerman. "Inventory of Global Methane Sources and Their Production
Rates," J. Geophys. Res., 87, 1305-1312, 1982.
6. World Meteorological Organization. Climate Change 1995 The Science of
Climate Change. Intergovernmental Panel on Climate Change, United Nations
Environment Programme, 1996.
7. World Meteorological Organization. Climate Change, 1990. Intergovernmental
Panel on Climate Change, United Nations Environment Programme, 1990.
8. World Meteorological Organization. Climate Change, 1992. Intergovernmental
Panel on Climate Change, United Nations Environment Programme, 1992.
9. World Meteorological Organization. Climate Change, 1994. Intergovernmental
Panel on Climate Change, United Nations Environment Programme, 1995.
10. Columbia Gas. An Engineering Estimate of the Incremental Change in Methane
Emissions with Increasing Throughput in a Model Natural Gas System, Final
Report, GRI-94/0257.32, Gas Research Institute, 1993.
11. Energy Information Administration, Emissions of Greenhouse Gases in the
United States 1987-1994, Department of Energy, DOE/EIA-0573(87-
94),Washington, DC, October 1995.
B-26
-------
12. U.S. Environmental Protection Agency. Inventory of U.S. Greenhouse Gas
Emissions and Sinks: 1990-1993, EPA-230/R-014 (NTIS PB95-138079), Office
of Policy, Planning and Evaluation, Washington, DC, September 1994.
13. International Workshop on Methane Emissions from Natural Gas Systems, Coal
Mining, and Waste Management, Sponsored by the Environment Agency of
Japan, U.S. Agency for International Development, and U.S. EPA, April 9-11,
1990.
14. Energy International. Energy Utilization and Greenhouse Gas Emissions: End-
Use Analysis, Final Report, GRI-93/0335, Gas Research Institute, June 1994.
15. Energy Information Administration. Coal Industry Annual 1993, Department of
Energy, DOE/EIA-0584(93), Washington, DC, December 1994.
16. American Gas Association, Gas Facts: 1992 Data, "Supply and Disposition of
Gas in the United States, 1969-1992," Arlington, VA, 1993.
17. Shires, T.M. and M.R. Harrison. Methane Emissions from the Natural Gas
Industry, Volume 6: Vented and Combustion Source Summary, Final Report, GRI-
94/0257.23 and EPA-600/R-96-080f. Gas Research Institute and U.S.
Environmental Protection Agency, June 1996.
18. U.S. Environmental Protection Agency, Compilation of Air Pollution Emission
Factors: Volume I Stationary Point and Area Sources, AP-42 (GPO 055-000-005-
001), Table 3.2-2, U.S. EPA Office of Air Quality Planning and Standards, 5th
Edition, January 1995.
19. Stapper, C.J. Methane Emissions from the Natural Gas Industry, Volume 11:
Compressor Driver Exhaust, Final Report, GRI-94/0257.28 and EPA-600/R-96-
080k, Gas Research Institute and U.S. Environmental Protection Agency, June
1996.
20. U.S. Environmental Protection Agency, Compilation of Air Pollution Emission
Factors: Volume I Stationary Point and Area Sources, AP-42 (GPO 055-000-005-
001), Table 1.4-3, U.S. EPA Office of Air Quality Planning and Standards, 5th
Edition, January 1995.
21. Energy Information Administration. Cost and Quality of Fuels for Electric Utility
Plants 1993, Department of Energy, DOE/EIA-O191(93), Washington, DC, July
1994.
B-27
-------
22. Radian International LLC. Methane and Carbon Dioxide Emission Estimates from
U.S. Petroleum Sources, Draft Report, American Petroleum Institute, January
1996.
23. Energy Information Administration. Petroleum Supply Annual1993, Volume 1.
Department of Energy, DOE/EIA-0340(93)/1, Washington, DC, June 1994.
24. Combustion Engineering. Combustion Fossil Power, ABB/Combustion
Engineering, Windsor, CT, 1991.
25. Baumeister, T., E.A. Avallone, and T. Baumeister III. Mark's Standard
Handbook for Mechanical Engineers, Eighth Edition, McGraw-Hill Book
Company, New York, NY, 1978, p. 7-14.
B-28
-------
APPENDIX C
Conversion Table
C-l
-------
Unit Conversion Table
English to Metric Conversions
lscf methane = 19.23 g methane
lBscf methane = 0.01923 Tg methane
lBscf methane = 19,230 metric tonnes methane
1 Bscf = 28.32 million standard cubic meters
1 short ton (ton) = 907.2 kg
1 lb = 0.4536 kg
1 ft3 = 0.02832 m3
1 ft3 = 28.32 liters
1 gallon = 3.785 liters
1 barrel (bbl) = 158.97 liters
1 inch = 2.540 cm
1 ft = 0.3048 m
1 mile = 1.609 km
1 hp - 0.7457 kW
1 hp-hr - 0.7457 kW-hr
1 Btu = 1055 joules
1 MMBtu = 293 kW-hr
1 lb/MMBtu = 430 g/GJ
T(°F) = 1.8 T (°C) + 32
1 psi = 51.71 mm Hg
Global Warming Conversions
Calculating carbon equivalents of any gas:
MMTCE = (MMT of gas) x
/ \
MW, carbon
MW, gas
X (GWP)
C-2
-------
Calculating C02 equivalents for methane:
MMT of CO, equiv. = (MMT CH.) x
/ MW, C02^
MW, CH,
4/
X (GWP)
where MW (molecular weight) of C02 = 44, MW carbon = 12, and MW CH4= 16.
Notes
scf = Standard cubic feet. Standard conditions are at 14.73 psia and 60°F.
Bscf = Billion standard cubic feet (109 scf).
MMscf = Million standard cubic feet.
Mscf = Thousand standard cubic feet.
Tg = Teragram (1012 g).
Giga(G) = Same as billion (109).
Metric tonnes = 1000 kg.
psig = Gauge pressure.
psia = Absolute pressure (note psia = psig + atmospheric pressure).
GWP = Global Warming Potential of a particular greenhouse gas for a given
time period.
MMT = Million metric tonnes of a gas.
MMTCE = Million metric tonnes, carbon equivalent.
MMT of C02 eq. = Million metric tonnes, carbon dioxide equivalent.
C-3
-------
APPENDIX D
Project Reviewers
D-l
-------
PROJECT REVIEWERS*
Name ' ¦
Company
Ackell, John
Oil Heat Task Force
Ammirato, Vincent
Columbia Gas
Ball, Richard H.
U.S. Department of Energy
Benjey, Bill
EPA
Bjerklie, John
Consolidated Natural Gas
Boss, Terry
INGAA
Bradford, Ray
Phillips 66
Brasseur, Guy
National Center for Atmospheric Research
Busch, William
NOAA
Carter, Doug
DOE
Chai, Eric
Shell Development
Childress, P.D.
Colorado Interstate Gas
Ching, Jason
EPA
Cohen, Jonathan
ICF Kaiser
Cook, Tracy
SoCal
Cormier, Michael
Amoco Production Co.
Craig, Bruce
Natural Gas Supply Assoc.
Derkowski, Carrie
Coastal
Doyle, Terry
Enron Corp
Doyle, William J.
Marathon Oil Company
Eberle, Art
Columbia Gas
Enright, Jeffrey
ICF Kaiser
Erickson, John
American Gas Association
Farrand, David
Williams Natural Gas
Fisher, Diane
Environmental Defense Fund
Fitzgibbon, Timothy
ICF
Fritz, Eric
Natural Gas Pipeline
Fung, Inez
NASA Goddard Institute for Space Studies
Gibbs, Michael
ICF
Godec, Michael
ICF
Goens, Dick
Union Gas Ltd.
Haines, Deanna
SoCal Gas
Hansen, Anne
NGL
Hansford, James E.
Enron Gas Pipeline
Hare, Marika
Consumers Gas
Hay, Nelson
American Gas Association (A.G.A.)
Hogan, Kathleen
EPA-OAR
Innerarity, Mike
Tenneco Energy
Isaacson, Ron
GRI
Johnson, Donald
Argonne National Laboratory
Kalkstein, Larry
EPA
Kirchgessner, Dave
EPA
Klein, Gary
API, Oil Heat Task Force
Knight, Bruce
Marathon Oil
D-2
(Continued)
-------
PROJECT REVIEWERS* (Cont'd)
Name
Company
Konecki, Mark
AMOCO
Kothari, Kiran
GRI
Lajiness, Vincent D.
ANR Pipeline
Lashof, Dan
Natural Resources Defense Council
Lawrence, Steve
PG&E
Lott, Bob
GRI
Magid, Hillel
Allied-Signal
Malarkey, Patrick
Phillips 66
Martino, Paul
API
Matthews, Neil
Southern Natural Gas
Magid, Hillel
Allied Signal Corp.
Mathis, Michael J.
Niagara Mohawk Power Corporation
Mercado, Donna
American Gas Association
Meshkati, Shahed
SoCal
Minotti, Marcello
Tenneco
Mize, Ed
Williams Natural Gas
Mobley, David J.
EPA
Morse, William
Columbia Gas
Mroz, Gene
Los Alamos National Lab
Mussio, Peter
Union Gas
Newsom, Vick
Amoco
Nunn, William
Texas Gas Transmission
Ollison, W.W.
American Petroleum Institute
Orfeo, Robert
Allied Signal Corp.
Osborne, Andrea
EPA
Parrotta, Daniel
Con Edison
Perhac, Ralph
EPRI
Philips, Marc
Enron
Prather, Michael
NASA Goddard Institute for Space Studies
Preusser
Brooklyn Union Gas Company
Primus, Frank
Chevron
Reilly, Mike
CNG
Reiquam, Howard
GRI
Resch, Rhone
EPA OAR
Riordan, Mike
Brooklyn Union Gas Company
Roose, Tom
GRI
Schievelbein, Vernon
Texaco
Shah, Andy
Conoco
Shervill, Paul
Union Gas
Stern, Richard
EPA
Swenson, Paul
Consolidated Natural Gas Service Corp.
Tate, Jack
Texaco
Tie, Xuexi
Los Alamos National Lab
D-3
(Continued)
-------
PROJECT REVIEWERS* (Cont'd)
' Name
Company
Tixier, Charles
Shell
Traweek, Lori
American Gas Association
Van Wyck, Robert
Con Edison
VanDerZanden, Daniel
Chevron
Warner, John
Amoco
Weynand, Gordon
EPA
Woodbury, Jonathan
ICF
Yundt, Paul
Gas Technology Canada
Zinkham, Jeffrey
Amoco
^Project reviewers have participated in one or more review meetings for this project over the
time period 1990-1996.
D-4
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before comple
1. REPORT NO. 2.
EPA-600/R-96-080b
3.
4. TITLE AND SUBTITLE
Methane Emissions from the Natural Gas Industry,
Volumes 1-15 (Volume 2: Technical Report)
5. REPORT DATE
June 1996
6. PERFORMING ORGANIZATION CODE
7. author(s) l. Campbell, M. Campbell, M. Cowgill, D. Ep-
person, M.Hall, M. Harrison, K. Hummel, D.Myers,
T. Shires, B. Stapper, C. Stapper, J. Wessels, and *
8. PERFORMING ORGANIZATION REPORT NO.
DCN 96-263-081-17
9. PERFORMING OROANIZATION NAME AND ADDRESS
Radian International LLC
P.O. Box 201088
Austin, Texas 78720-1088
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
5091-251-2171 (GRI)
68-D1-0031 (EPA)
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 3/91-4/96
14. SPONSORING AGENCY CODE
EPA/600/13
16.supplementary NOTES EPA project officer is D. A. Kirchgessner, MD-63, 919/541-4021.
Cosponsor GRI project officer is R. A. Lott, Gas Research Institute, 8600 West Bryn
Mawr Ave., Chicago, IL 60631. (*)H. Williamson (Block 7).
16. ABSTRACT-phe 15-volume report summarizes the results of a comprehensive program
to quantify methane (CH4) emissions from the U. S. natural gas industry for the base
year. The objective was to determine CH4 emissions from the wellhead and ending
downstream at the customer's meter. The accuracy goal was to determine these
emissions within +/-0. 5% of natural gas production for a 90% confidence interval. For
the 1992 base year, total CH4 emissions for the U. S. natural gas industry was 314
+/- 105 Bscf (6.04 +/- 2.01 Tg). This is equivalent to 1.4 +/- 0.5% of gross natural
gas production, and reflects neither emissions reductions (per the voluntary Ameri-
Gas Association/EPA Star Program) nor incremental increases (due to increased
gas usage) since 1992. Results from this program were used to compare greenhouse
gas emissions from the fuel cycle for natural gas, oil, and coal using the global war-
ming potentials (GWPs) recently published by the Intergovernmental Panel on Climate
Change (IPCC). The analysis showed that natural gas contributes less to potential
global warming than coal or oil, which supports the fuel switching strategy suggested
by the IPCC and others. In addition, study results are being used by the natural gas
industry to reduce operating costs while reducing emissions.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. 1DENTIF1ERS/OPEN ENDED TERMS
c. cosati Field/Group
Pollution
Emission
Greenhouse Effect
Natural Gas
Gas Pipelines
Methane
Pollution Prevention
Stationary Sources
Global Warming
13 B
14 G
04A
21D
15E
07C
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
146
20. SECURITY CLASS (This page J
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |