GRI-94 / 0257.20
EPA - 600/R-96-080c
June 1996
«>EPA Research and
Development
METHANE EMISSIONS FROM THE
NATURAL GAS INDUSTRY
Volume 3: General Methodology
United States
® Environmental Protection
Agency
Prepared for
Energy Information Administration (U. S. DOE)
Prepared by
National Risk Management
Research Laboratory
Research Triangle Park, NC 27711
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before complel
1. REPORT NO. 2
EPA-600/R-96-080c
3.
4. TITLE AND SUBTITLE
Methane Emissions from the Natural Gas Industry,
Volumes 1-15 (Volume 3: General Methodology)
5. REPORT OATE
June 1996
6. PERFORMING ORGANIZATION CODE
7. authoris) 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 ORGANIZATION NAME ANO 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 ANO ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVEREO
Final; 3/91-4/96
14. SPONSORING AGENCY CODE
EPA/60.0/13
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).
i6. ABSTRACTXhe l5_volume report summarizes the results of a comprehensive program
to quantify methane (CH4) emissions from the U. S. natural gas industry for the base
pear. 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
|as 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.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Emission
Greenhouse Effect
Natural Gas
Gas Pipelines
Methane
Pollution Prevention
Stationary Sources
Global Warming
13B
14G
04A
21D
15E
07C
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
188
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
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.
-------
EPA-600 /R-96-080c
June 1896
METHANE EMISSIONS FROM
THE NATURAL GAS INDUSTRY,
VOLUME 3: GENERAL METHODOLOGY
FINAL REPORT
Prepared by:
Matthew R. Harrison
Hugh J. Williamson
Lisa M. Campbell
Radian Corporation
8501 N. Mopac Blvd.
P.O. Box 201088
Austin, TX 78720-1088
DCN: 95-263-081-13
For
GRI Project Manager: Robert A. Lott
GAS RESEARCH INSTITUTE
Contract No. 5091-251-2171
8600 West Bryn Mawr Ave.
Chicago, IL 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
-------
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 Methane Emissions from the Natural Gas Industry,
Volume 3: General Methodology
Final Report
Contractor Radian International LLC
GRI Contract Number 5091-251-2171
EPA Contract Number 68-D1-0031
Principal
Investigator Matthew R. Harrison
Report Period March 1991 - June 1996
Final Report
Objective This report describes the methods used to quantify the annual methane
emissions from the natural gas industry. The methods include the
general methods used for emission factor measurement, activity factor
quantification, and extrapolation.
Technical The increased use of natural gas has been suggested as a strategy for
Perspective 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 results of this
study can be used to construct global methane budgets and to determine
the relative impact of natural gas on global warming versus coal and oil.
This report is Volume 3 of a multi-volume set of reports that fully
describe the project. While general methodology is covered in this
report, specific statistical methodology is covered in Volume 4, and
detailed activity factor methodology is covered in Volume 5.
iii
-------
This report provides a brief summary of the methods used by the
GRI/EPA project to estimate emissions from the natural gas industry.
The methods have been extensively reviewed by industry experts and by
project advisors throughout the duration of this multi-year project. The
following is included: methods used to characterize the industry and to
identify each emission source, the measurement techniques used to
measure emissions, the procedures used for calculating emissions from
sources that couldn't be measured, methods used to extrapolate per-
device emissions to national totals, and the sampling techniques and
statistical methods used to determine the accuracy of the emissions
estimate.
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
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.
The techniques used to determine methane emissions were developed to
be 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 emissions estimate is the product of
the emission factor and activity factor. This report documents this
overall technical approach.
The development of specific emission factors and activity factors for
each source category are presented in separate reports (Volumes 6
through 15).
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 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 global warming than coal or oil, which supports the
fuel switching strategy suggested by IPCC and others.
In addition, results from this study are being used by the natural gas
industry to reduce operating costs while reducing emissions. Some
companies are also participating in the Natural Gas-Star program, a
iv
-------
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 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
v
-------
TABLE OF CONTENTS
Page
1.0 SUMMARY 1
2.0 INTRODUCTION 2
3.0 METHOD USED TO CHARACTERIZE THE NATURAL GAS INDUSTRY ... 4
3.1 Natural Gas Industry Definition 4
3.2 Production Segment Definition 6
3.3 Gas Processing Segment Definition 9
3.4 Transmission and Storage Segment Definition 10
3.5 Distribution Segment Definition 13
4.0 EMISSION ESTIMATION TECHNIQUES 15
4.1 Measurement Techniques for Steady Emissions 15
4.2 Calculation Approach for Unsteady Emissions 17
5.0 GENERAL EXTRAPOLATION METHODOLOGY 19
5.1 Emission Factors 20
5.2 Activity Factors 21
6.0 SAMPLING AND STATISTICAL ACCURACY 45
6.1 Sampling Approach 47
6.2 Screening for Bias in Activity Factors 53
6.3 Screening for Bias in Emission Factors 54
6.4 Accuracy Target 57
6.5 Overall Statistical Accuracy 61
7.0 REFERENCES 63
APPENDIX A - Production Source Sheets A-l
APPENDIX B - Gas Processing Source Sheets B-l
APPENDIX C - Transmission/Storage Source Sheets C-l
APPENDIX D - Distribution Source Sheets D-l
APPENDIX E - Conversion Table E-l
vi
-------
LIST OF FIGURES
Page
3-1 Gas Industry Flow Chart 5
3-2 Industry Boundaries 7
3-3 Gas Processing Plant 11
3-4 Transmission and Storage Stations 12
3-5 Distribution Segment Equipment 14
5-1 Selected Production Regions 29
6-1 Illustration of Random and Bias Errors 46
vii
-------
LIST OF TABLES
Page
5-1 Well-Defined Activity Factors 23
5-2 Examples of Developed Activity Factors 24
5-3 Example Data Compilation of Sites in Region X 27
5-4 Regional Differences in Production Rates and Well Counts 30
5-5 Production Sites Summary 32
5-6 Data for Offshore Production Sites 33
5-7 Data for Gulf Coast Onshore Production Sites 34
5-8 Data for Central Plains Production Sites 35
5-9 Data for Pacific/Mountain Production Sites 36
5-10 Data for Atlantic and Great Lakes Production Sites 37
5-11 Gas Processing Plants 40
5-12 Transmission Compressor Stations 41
5-13 Storage Compressor Stations 43
6-1 Estimated Methane Emissions from Distribution Metering and Pressure
Regulating Stations 56
6-2 Percentage of Error in Total Emissions 59
viii
-------
1.0 SUMMARY
Fuel switching has been suggested by the International Panel on Climate
Change (IPCC), U.S. Environmental Protection Agency (EPA), and others as a strategy for
reducing global warming. During the combustion process, natural gas generates less carbon
dioxide (C02) per unit of energy generated than either coal or oil. On the basis of the
amount of C02 generated, 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 the production, transmission, and distribution of natural gas could reduce
or even eliminate the inherent advantage of its lower carbon dioxide emissions during
combustion. For this reason, Gas Research Institute (GRI) and the U.S. Environmental
Protection Agency Office of Research and Development (EPA-ORD) developed a jointly
funded and managed program to better define methane emissions from the U.S. natural gas
industry. The objective of this comprehensive program is to quantify methane emissions
from the gas industry, beginning at the wellhead and ending immediately downstream of the
customer's meter, with an overall accuracy goal of 0.5% of natural gas production (±111
Bscf) based on a 90% confidence level.
This report provides a brief summary of the methods used by the GRI/EPA
project to estimate emissions from the natural gas industry. The methods have been
extensively reviewed by industry experts and by project advisors throughout the duration of
this multi-year project. The methods include the following: the industry characterization
that was used to identify each emission source, the measurement techniques used to directly
measure emissions, the calculation approach used to estimate unmeasured emissions, the
activity factor approach used to extrapolate per-device emissions to national totals, and the
sampling techniques and statistical methods used to determine the accuracy of the emission
estimate. These methods are more completely described in other volumes prepared as part
of this project.
1
-------
2.0 INTRODUCTION
The GRI/EPA project was conducted in three phases: scoping, methods
development, and implementation. During the scoping study, methane emissions from each
source in the gas industry were estimated on the basis of available data and engineering
judgement. These initial estimates were used to set priorities for data collection according
to the relative contribution to emissions or the uncertainty in emissions.
In the second phase of the program, methods were developed to measure
and/or calculate emissions from all sources of methane emissions in the gas industry.
These methods were validated through tests designed to quantify the accuracy in the
measurement approach (i.e., proof of concept tests), and through industry review for the
calculation approaches. However, emissions could not be measured or calculated from each
individual source (e.g., glycol dehydrator, compressor engine) in the natural gas industry
because of the vast number of sources. Therefore, part of the second phase was to develop
defensible techniques for extrapolating limited data collected for sources in a specific
source category to similar sources nationwide.
The third phase of the program focused on collecting data needed to define
emissions from all sources and extrapolating these data to develop a national estimate.
Data collection in this phase concentrated on high priority sources. An Advisory
Committee consisting of industry representatives, project sponsors, and other interested
parties from both the government and private sectors provided guidance and peer review for
all phases of the program. In addition, an Industry Review Panel provided more detailed
technical review of the project.
This report briefly describes the methods used in all phases of the GRI/EPA
methane emissions study. Section 3 presents the methods for characterizing the industry
and identifying sources of emissions (part of Phase 1 of the program). Section 4 presents
the measurement and calculation methods used to determine emissions from each source
2
-------
type or category. Section 5 presents the methods used to extrapolate emissions from
individual sources to a national total. Section 6 provides a brief summary of the sampling
and statistical methods used to determine the accuracy of each emissions estimate. The
appendices (A, B, C, D) present some of the data collected as part of the program and the
worksheets for determining emissions and extrapolating national annual emissions for each
source of methane emissions in the natural gas industry.
3
-------
3.0
METHOD USED TO CHARACTERIZE THE NATURAL GAS
INDUSTRY
The first step for estimating methane emissions from the U.S. natural gas
industry is to identify and characterize each emission source so that all significant sources
are included. This section of the report characterizes the industry by outlining the segments
of the industry as well as the types of equipment found within each segment. (This section
is identical to Section 3 of Volume 5 on activity factors.1)
While this section draws a general picture of the industry developed by the
GRI/EPA methane emissions project, it is not intended to present a definitive picture of the
industry regarding all typical operational parameters but only those that are necessary to
identify all sources and causes of methane emissions.
3.1 Natural Gas Industry Definition
The natural gas industry produces and delivers natural gas to various
residential, commercial, and industrial customers. The industry uses wells to produce
natural gas existing in underground formations, then processes and compresses the gas and
transports it to the customer. Transportation to the customer involves intra- and interstate
pipeline transportation, storage, and finally distribution of the gas to the customer through
local distribution pipeline networks.
The generally accepted segments of the natural gas industry are 1)
production, 2) gas processing, 3) transportation, 4) storage, and 5) distribution. Each of
these segments is shown in the overall flow chart for the industry in Figure 3-1. Each
segment is described in more detail in the following subsections.
Boundaries for the study were defined to specify what equipment is included
or excluded from the study. These boundaries were set using input from industry experts.
4
-------
PRODUCTION
Surface Facilities
PROCESSING
TRANSMISSION/
STORAGE
Direct
Sales
M&PR Stations
Compressor
Stations
Pipelines
Gas Plant
Pipelines
Liquids
Storage
Liquids
Underground
Storage
Reservoir
DISTRIBUTION
Main and
Service Pipelines
<@>
-®>
-®)
-<0)
-®)
-®)
Customer Meters
0
Compressor
©
Meter
£
Pressure
Regulator
Figure 3-1. Gas Industry Flow Chart
-------
The guideline used for setting the boundaries was to exclude equipment in each segment
not required for the marketing of natural gas. For example, certain oil production
equipment is excluded from the production segment since it exists to produce oil and is not
needed for gas production (see Figure 3-2). Similarly, in gas processing, equipment
associated with the fractionation of propane, butane, and natural gas liquids is excluded. In
distribution, all equipment up to and including the customer's meter are included. End user
piping, combustion, and vented emissions are not included.
3.2 Production Segment Definition
Emissions of methane that result from oil production, or that occur naturally
(non-anthropogenic) from formations, are excluded. Unmarketed natural gas, such as that
produced by oil wells that vent some gas, are not considered part of the gas industry.
The production segment is composed of gas wells, oil wells, and surface
equipment. The well includes the holes drilled through subsurface rock that 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.
However, the segment definition for gas industry production equipment
excludes equipment associated with oil production. Figure 3-2 shows the general
equipment found in the oil and gas production segment as well as the selected boundaries
for gas industry equipment used by this study. Equipment outside the boundaries were not
included in the activity factor estimates developed in this study.
The rationale for defining the boundaries is that all equipment at a gas well
site, except equipment used to collect and handle liquids that are marketed (oil or
condensate), is part of the gas industry, but that all equipment at an oil well site is excluded
unless it is used to collect, process or transport marketed natural gas (Figure 3-2).
6
-------
Oil Well
Wellhead
Dehydrator
Gas Industry
Compressor
Meter
Oil Well
Wellhead
(producing
gas)
Field Use
Gas
^ Separator ^
Vapor Recovery
Compressor
Chemical
Injection
Pump
Pnuematic
Control
Valve
Heater/
Treater
Salt Water and
Oil Stock Tanks
Gas Industry
Compressor
Dehydrator
Coal Bed
Methane Well
Meter
^ Separator j
Fresh Water
Dehydrator | *
Chemical
Injection
Pump f
Compressor
Meter
Gas Well
Wellhead
Field Use
Gas
^ Separator j
Vapor Recovery
Compressor
Hydrocarbon
Condensate
or Oil Tank
Pouematic
Control Valve
Salt Water
Tank
Figure 3-2. Industry Boundaries
7
-------
Therefore, the definition excludes all oil tanks and equipment at all oil wells that do not
market gas. In addition, it excludes much of the equipment at oil wells that do market gas.
At oil wells that market gas, the gas production is secondary and usually generates lower
revenue; the well exists primarily because it produces oil. Therefore, the wellhead, the
separator, the pneumatic control valves, the well's chemical injection pumps, any field use
gas lines, and all of the liquid piping are considered part of the oil industry and are
excluded from the GRI/EPA gas industry study. The gas industry equipment begins only
on the gas line downstream of the separator, at the first piece of gas line equipment, such as
the sales meter, compressor, or dehydrator.
In general, an oil or oil and gas field may have centralized surface treatment
facilities or each well site may have its own independent surface facilities. In centralized
facilities, all of the separators, dehydrators, and compressors may be in one location, with
gas flowing in from gathering pipelines connected to many wellheads. Decentralized
facilities have all the necessary surface equipment (separators, compressor, dehydrator, etc.)
at each individual well site. Centralized facilities can have lower equipment counts per
well than decentralized facilities. Sometimes the facilities may be primarily decentralized
but have a few centralized components. For example, separators may be at each well
(decentralized), while compressors and dehydrators may be centralized.
Whatever the field configuration, all gas wells have a wellhead and most
have a gas meter. Also independent of the field configuration, gas wells may or may not
have a separator(s), dehydrator, or compressor. The use of the equipment depends upon
the free liquid production, absorbed moisture content, and well pressure. For example,
some sweet, dry gas wells can produce directly to a pipeline. However, most wells require
separation for free-liquid products (salt water, hydrocarbon condensate, and oil) and some
dehydration.
Oil wells that market gas (the only oil wells included in this study) may also
have centralized or individual well site facilities. They will always have a separator and a
8
-------
meter. As with gas wells, they may or may not have a dehydrator or a compressor,
depending upon the absorbed moisture content and the pressure.
Oil wells that market gas may be either free-flowing or artificial-lift wells.
Free-flowing wells often have absorbed or co-produced gas that is marketed. Therefore,
some of the equipment at these free-flowing oil wells is considered part of the gas industry
if it exists to market the gas. Artificial-lift oil wells are most often not part of the gas
industry, but a few do produce gas and therefore are included in the gas industry definition.
Artificial-lift oil wells that have downhole pumps or surface pump jacks
usually do not produce or market any gas and therefore are not part of the gas industry.
Artificial-gas lift oil wells push compressed gas downhole and inject the gas into the tubing,
thus using the gas to aerate the oil in the tubing string. This brings the oil back to the
surface. Only the gas-lift wells that produce and market gas in excess of the amount
injected are considered part of the natural gas industry. For gas-lift oil wells that market
gas, the compressors associated with the gas-lift circulation are not considered to be part of
the gas industry.
3.3 Gas Processing Segment Definition
Natural gas processing plants exist primarily to recover high value liquid
products from the gas stream and to maintain the quality (content and heating value) of the
gas stream. The liquid products include natural gasoline, butane, and propane. (Ethane is
sometimes recovered as well.) The products are removed by compression and cooling or
by absorption. Absorption processes use a fluid, such as lean oil, to absorb the liquid
components from the gas stream in a tower; the rich oil is then heated to release the
recovered products. A compression and cooling process uses a turboexpander or a
refrigeration process to supercool the natural gas so that the products will condense.
9
-------
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 it exists to
purify and market liquid products. The back end of the gas plant has negligible methane
emissions since the liquids handled contain only trace amounts of methane.
The front end of the gas plant often contains dehydration facilities, wet gas
compression, and the absorption or compression and cooling process. All gas plants are
considered part of the natural gas industry. Therefore, all methane emissions from natural
gas processing plants are included in this study. Figure 3-3 shows a schematic diagram of
a gas plant.
3.4 Transmission and Storage Segment Definition
The transmission segment moves the natural gas from the gas plant or
directly from the field production to the local distribution companies (LDCs). Gas is often
moved across many states, such as from the Gulf Coast to the eastern seaboard of the
United States. The segment consists of large diameter pipelines, 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 or gas turbine (centrifugal) compressors, and generators. Dehydrators may be
included but are not usually present because of upstream drying. The station may also
include metering facilities. Figure 3-4 shows a schematic diagram of transmission and
storage stations.
Transmission companies also have metering and regulating stations where
they exchange gas with other transmission companies or where they deliver gas to the
LDCs. Storage facilities exist to store natural gas produced during off-peak times (summer)
so that the gas can be produced and delivered during periods of peak demand in the winter.
10
-------
Residue Gas
to Sales
C ~
Wet
Gas
Dehydration
Absorption or
Compression
& Cooling Process
Backend of the Plant
-*c,
Liquids
"~ C,
-* C.
Natural
Gasolines
Figure 3-3. Gas Processing Plant
11
-------
Ivj
Gas
from
Pipeline
TRANSMISSION
COMPRESSOR STATION
ABOVE GROUND STORAGE
Compressors
Cooling
Process
->
LNG
Tank
Heaters
V Gas
Transmission Pipeline
from
Pipeline
BELOW GROUND STORAGE
f
Compressor
Station
Wells
_L _L ±
Dehydrator storage Field
Figure 3-4. Transmission and Storage Stations
-------
Storage facilities are often located close to consumption centers so that a cross-country
transmission pipeline does not have to be sized for peak winter 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 store 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.
Most storage stations consist of a compressor station that is very similar to a
transmission compression station (see Figure 3-4). Underground storage facilities also have
storage field wells and often have dehydrators to remove the water absorbed by the gas
while underground. Except for emissions from underground leaks in storage formations, all
storage equipment is included in the boundaries for the gas industry as defined by this
project.
3.5 Distribution Segment Definition
The distribution segment receives high-pressure gas from the transmission
pipelines, reduces the pressure, and delivers the gas to all of the residential, commercial,
and industrial consumers. The segment includes pipeline (mains and services), meter and
regulation stations (city gates), and customer meters. All of these facilities are considered
to be an integral part of the gas industry. Figure 3-5 shows a schematic of the distribution
segment and associated equipment.
13
-------
Meter and Pressure
Regulating (M&PR) Station
Pressure-Reducing
Station
Services (Small Pipe)
Transmission
Pipeline
Meter
Pressure-Reducing
Regulators
Mains (Pipeline)
[M)
©
(M)
(M)
©
(m)
©
Cm)
Customer Meters
(Residential,
Commercial,
Industrial,
Electric Utility)
Figure 3-5. Distribution Segment Equipment
-------
4.0 EMISSION ESTIMATION TECHNIQUES
After all potential sources of methane emissions in the industry were identified
and characterized, the emissions were quantified. The method selected to quantify the
emissions from a source depended on the variability of emissions with time. The sources
were divided into two categories: steady emitters and unsteady emitters. Emission sources
with continuous bleed rates, or with reasonably steady bleed rates over a typical
measurement time, are considered "steady" sources and can be more easily measured.
Fugitive emissions are generally considered steady. Unsteady emitters are defined as sources
with highly variable emissions, such as a pneumatic device on an isolation valve or a
maintenance activity that requires blowdown. These emissions sources vary from company
to company and site to site because of different maintenance practices and operating
conditions. "Steady" is a relative term and is defined by the time period of data needed for
the study. For this study, the annual value of methane emissions is needed. Because it
would be impractical to measure emissions all year for every source, it is important that a
single measurement is representative of the annual emissions.
This section describes the measurement techniques used for steady emissions
and the calculation approach used to estimate unsteady emissions.
4.1 Measurement Techniques for Steady Emissions
The techniques used in the study for measuring steady emissions is briefly
described below.
Fugitives Measurement Methods
Emission factors for estimating fugitive emissions were determined based upon
measurements of emissions from individual sealed surfaces (components) associated with the
equipment, such as valve packing, flange gaskets, screwed fittings, and compressor/pump
15
-------
seals. Emissions from a large number of components were measured and an average
emission rate per component determined for each component type. Emissions from an
equipment source, such as a compressor, or a facility, such as a compressor station, were
then calculated by multiplying the number of components associated with that equipment or
facility with the average emission rate per component.2
Emissions from individual components were measured using one of several methods:
• A high flow organic vapor analyzer captures the entire leak and
measures the methane concentration and flow rate. The emissions rate
is determined from the product of the concentration and flow rate.
• A total enclosure technique called bagging. Uncontaminated air is
blown through a plastic bag enclosing the component. The flow rate
and outlet concentration is measured by an organic vapor analyzer, and
the leak rate is determined from the product of the concentration and
flow rate.
• A screening technique where the methane concentration is measured at
the point of the leak using a standard organic vapor analyzer. The
concentration is related to an emission rate by a correlation equation
that was developed in other studies and that related bagged emissions to
the concentration measured during the screening test.
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 the tracer gas and methane. Assuming complete mixing of the methane and
tracer gas and 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 meter and pressure
regulating stations.3
16
-------
Leak Statistics Method
The leak statistics method is used to quantify methane emissions from
underground mains and services in distribution systems. 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 leak statistics program
was conducted as a cooperative program between EPA/GRI and industry. The industry
participants used specially designed equipment to measure leakage rates from underground
distribution mains and services. To perform the measurement, 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 measured. The leak statistics
method combines the measured leakage rate per leak with the historical leak records of the
companies and national leak repair data to determine the number of leaks per mile for
different pipe material. The emissions are determined by multiplying the leak rate per leak
by the leaks per mile and number of miles of pipe.4
Direct Flow Measurement
The direct flow method was used to measure emissions from some pneumatic
devices and leaks in some open ended lines. Since the gas consumption of certain pneumatic
devices is emitted to the atmosphere, a flow measurement of the supply gas to the pneumatic
device can be used to characterize the device emissions. Measurements used by this study
involved a direct flow turbine meter that was installed in the supply gas lines.5 Leaks from
some open ended lines were measured by directly connecting a gas meter to the line.
4.2 Calculation Approach for Unsteady Emissions
Emissions that are intermittent or unsteady have highly 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 often calculated rather than measured.
17
-------
For some sources, such as maintenance emissions, detailed company
calculations are available for multiple years.6 However, most sources of 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:
• A detailed technical characterization of the source, identifying the
important parameters affecting emissions. (This information was
documented for individual source types in other Tier 3 reports on each
major emission type.1'7,8,9)
• Data gathered (from multiple sites) that would be needed to calculate
emissions per event.
• Sufficient data to define the frequency of events.
An example of emissions calculated for an unsteady source is the estimate of
emissions from a vessel blowdown for routine maintenance. In this case, the volume,
pressure, and temperature of gas contained in the vessel before blowdown is "calculated" to
quantify the losses from the blowdown event. Additionally, an average frequency of these
vessel blowdown events is necessary to determine the annual losses from this source of
methane emissions.
In some cases, measurement from other studies were used to establish the
emissions per event. Therefore, these emissions data were combined with site data collected
in this study to quantify the number of events per year in order to calculate the annual
emissions from these sources. Examples of sources where data from other studies were used
are: emissions from compressor exhaust, gas-operated pneumatic devices, glycol dehydrator
regenerator overhead vents, and gas-operated chemical injection pumps by measuring the
emissions per event.
18
-------
5.0 GENERAL EXTRAPOLATION METHODOLOGY
By necessity, data were collected on a relatively small percentage of sources
within each source category. These data were extrapolated to obtain nationwide estimates for
similar sources throughout the industry. The extrapolation techniques for creating nationwide
emission estimates were developed so that the emissions for each source category could be
estimated to meet a specified level of precision and negligible bias.
The extrapolation approach is a method to scale-up the average emissions from
a source, determined by a limited sampling effort, 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 the limited number of samples. These
factors are defined in such a way that their product equals the total emissions from a source
category.
EF x AF = National Emissions (1)
Typically, the emission factor is the average measured or calculated emissions
from a large number of randomly selected sources in a source category and the activity factor
is the total number of sources in the entire target population or source category. However,
in applying this simplified approach to developing emission and activity factors it is
important to ensure that there is no bias in the data.
The extrapolation methodology involves more than just the scale-up of
emissions data; it also includes the sampling approach which is fundamental to the accuracy
of the emissions data.
This section describes the two components for the extrapolation equation-
emission factors and activity factors.
19
-------
5.1
Emission Factors
The emission factor is generally defined as the annual emissions per source.
In many cases, the emission factor was calculated by simply summing the emissions data
from each source and dividing by the number of sources sampled. The emissions data would
be measurements or estimates for each source. In some cases, the variability of the
emissions data from source to source is very large. For source types of this nature, it is
often possible to reduce variability by redefining the emission factor or by stratification.
Reducing the variability is important because it reduces the number of data points needed to
achieve the accuracy target.
Redefining the Emission Factor
For a few types of sources, emissions can be more accurately estimated with
fewer data points when the emission factor is defined not as a simple average for the source
but in relation to key parameters that influence the emissions from the source. This is
essentially the same as subdividing the source category into subsets. Since the variability is
significantly reduced, fewer total data points are required to achieve an acceptable level of
accuracy.
For example, the internal combustion engines that drive compressors in the gas
industry vary in size (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.
The number of data points required may also be reduced by stratifying on the
basis of parameters that affect emissions. An example is quantification of the methane
20
-------
emissions from underground distribution mains and services. Based on limited data, the
variability in emission measurements for underground distribution lines was determined to be
very large. By defining parameters that influence the emission rate from distribution lines
and stratifying the emission factor and activity factor for this source by these parameters, the
variability of emissions from source to source may be reduced. Data collection resources
can be allocated to the parameters that contribute the most to the overall uncertainty of the
estimate. Therefore, by subdividing a source into categories with differing emission
characteristics and allocating data collection resources to the parameters that influence
emissions the most, the overall number of samples required to meet the accuracy target can
be reduced.
5.2 Activity Factors
This section on activity factors is an abbreviated version of the text presented
in Volume 5.1 The reader is referred to the activity factors report for specific details on
particular activity factors that cannot be found in Section 5.2 or in the appendices to this
report.
In general, the activity factor is the total population of the source when the
emission factor is defined as the annual emissions per source. Exceptions to this general
definition of an activity factor would only include sources that have an emission factor that
can be more accurately represented by a parameter(s) that influences emissions (e.g., the
emission factor for internal combustion engines is in terms of emissions per horsepower-
hour). For these exceptions, the activity factor would be the parameter that influences
emissions (e.g., horsepower-hours/year).
In some cases, existing programs track the total nationwide population of a
source type, such as gas wells, miles of transmission and distribution pipelines, and total
national production, within the natural gas industry, as shown in Table 5-1. However, in
many cases, the total population of a source type within the gas industry is unknown.
21
-------
Table 5-2 presents some of the activity factors that are not tracked nationally, but that were
generated by this project.
For sources that have an unknown population, site visits were conducted to
determine the number of sources at each site and to scale-up the site data to represent the
total population. These site visits to collect activity factor data are typically conducted in
conjunction with the data collection efforts for the emission factor. The number of sites
visited, gas produced or marketed, and equipment counts are presented in Section 6.1,
Sampling Approach. These site count data are scaled by using population data that is known
and is related to the source. For example, in the GRI/EPA study, no data were available on
the nationwide population of production separators. To calculate this value, the number of
production separators at a site, gathered as part of site visits, were divided by the number of
wells at each site. Then, the average ratio of separators to wells from all site visit data were
used to extrapolate nationally by multiplying by the national well count. However, when
scaling the site visit data to represent the entire population, a check for bias is made. (Refer
to Section 6.2, Screening for Bias in Activity Factors.)
For sources that are not tracked nationally, individual company data or
regional surveys (surveys by state agencies or trade organizations) were sometimes available.
Metering/pressure regulating stations, glycol dehydrators, and compressor engines/gas
turbines are tracked on a company wide basis or through regional surveys. For regional or
company tracked activity factors, sufficient company/regional data had to be gathered to
comprise a representative sample to extrapolate to a national population. In most cases,
entire companies or regions could be represented by the data collected from one sample;
therefore, few samples were required, in general, to represent the national population
accurately.
The extrapolation of equipment activity factors from individual site data within
a stratum is usually handled by selecting an "extrapolation parameter" (EP) that is known for
the site as well as regionally or nationally. Examples of extrapolation parameters are the
22
-------
TABLE 5-1. WELL-DEFINED ACTIVITY FACTORS
Segment
Activity Factor Name
Number
Total Industry
Gross Gas Production (Tscf)
22.13
Production
No. of gas wells
276,000
No. of oil wells
602,000
Processing
No. of gas plants
726
No. of AGR units
371
Transmission and Storage
Miles of transmission pipeline
284,500
No. of storage facilities
475
No. of wells
18,000
Distribution
Miles of mains
888,000
No. of services
43,714,000
(Continued)
23
-------
TABLE 5-2. EXAMPLES OF DEVELOPED ACTIVITY FACTORS
Segment Activity Factor Name
Number
Total Industry
Reciprocating compressor drivers
44,130
Turbine compressor drivers
1,543
Number of glycol dehydrators
39,620
Production
No. of oil wells marketing gas
209,000
No. of gas wells requiring unloading
114,100
Compressor drivers
17,100 recips
Engineer MMHp-hr
27,460
Offshore platforms
1,114
Glycol dehydrators
37,820
Glycol dehydrator throughput (MMscfy)
12,400,000
Separators
167,200
In-line heaters
51,000
Total production vessels
256,000
Chemical injection pumps
16,970
Pressure relief valves
529,400
Gathering pipeline miles
340,000
Pneumatic devices
249,100
Processing
Compressor drivers and installed HP
4,092 recips (4.19 MMHP)
726 turbines (5.19 MMHP)
Annual compressor operating hours (average)
6,626 (recips)
6,345 (turbines)
Glycol dehydrator throughput (MMscfy)
8,630,000
Acid gas recovery units
371
(Continued)
24
-------
TABLE 5-2. EXAMPLES OF DEVELOPED ACTIVITY FACTORS (Continued)
Segment
Activity Factor Name
Number "
Transmission and Storage
Compressor drivers and installed HP
7,715 recips (13.4 MMHP)
817 turbines (5.1 MMHP)
Annual driver operating hours (average)
- Transmission compressor drivers
- Transmission generating drivers
- Storage compressor drivers
- Storage generating drivers
3,964 (recips)
2,118 (turbines)
1,352 (recips)
474 (turbines)
3,707 (recips)
2,917 (turbines)
191 (recips)
36 (turbines)
Transmission compressor stations
1,700
Glycol dehydrator (throughput (MMscfy)
3,086,000
M&R stations
- Farm taps
- Interconnects
- Direct industrial sales
71,690
2,533
938
Distribution
M&R stations
132,000
Outdoor customer meters
40,049,000
Leak frequency
Various
25
-------
number of wells for production, number of plants for processing, and number of compressor
stations for transmission. Populations of other equipment, such as the count of separators at
the site, are then divided by that term, allowing the resulting ratio to be easily extrapolated
to a regional or national total.
However, the regional ratio of
AF,_ (2)
EP
where: AF! = activity factor (population) of equipment type 1
EP = extrapolation parameter,
can be determined from 1) regional sums, or 2) by averaging the ratio from each site. The
extrapolation plan must select one of these two methods based upon technical merit. These
two methods can be described as: 1) weighting the site counts by the extrapolation activity
factor, or 2) using an average count per site (not weighting).
For example, to determine the number of separators in a region, the
production site count of separators and wells at a site could be extrapolated to the regional
total by two methods: 1) summing the separators and dividing by the total well count (each
site data is weighted by the total well counts), or 2) by averaging all of the site ratios of
separators/well (thus treating each site as an equally representative sample). The decision on
which method to use depended upon a technical analysis of whether that method would
introduce bias. The method selected might vary from segment to segment, but was generally
constant across most calculations within a segment. The first method, summing equipment
from all sites and dividing by the sum of the extrapolation parameter, was used almost
exclusively by this project. This is discussed in detail in Volume 4 on statistical
methodology.10
26
-------
The following hypothetical example illustrates the two options for extrapolating
activity factors. The following table (Table 5-3) and calculations give an example of the two
methods for determining the number of separators in a region in the natural gas production
segment.
TABLE 5-3. EXAMPLE DATA COMPILATION OF SITES IN REGION X
Site Count of Site Count of Gas Site Ratio
Site Separators Wells (separators/well)
1 140 138 1.01
2 324 321 1.01
3 100 100 1.00
4 5 15 0.33
5 10 1000 0.01
TOTAL 579 1574
On a basis weighted by the total wells at a site, the regional ratio is 579/1574,
which equals 0.37 separators/well. This number is heavily weighted by one of five sites that
had a low separator per well count but a high number of wells (about ten times as many as
any other site). If the second method is used, each site is treated as an equally representative
sample and the average of the site ratios is used; the result is 0.67 separators/well.
The first method was selected for all production activity factor extrapolations
since there is a reason to believe that a randomly selected site that has many wells is
representative of a larger portion of the population than a randomly selected site with only a
few wells. Weighting by well count assumes that a larger number of wells at a site means
that the site is representative of a larger population than a site with a smaller well count.
Volume 5 on activity factors1 provides additional details on this method.
In addition, some equipment activity factors sources could be scaled up by
several possible EPs. If a known physical/technical relationship existed between the source
27
-------
population and one EP, then that parameter was selected. However, where the relationship
between the source population and the other parameters is not obvious from a technical
perspective, many approaches having technical merit were used, and either a) the average of
the methods was used, or b) the resulting data from individual companies statistically
analyzed to determine the appropriate extrapolation approach.
For example, it was not clear from a technical perspective whether to scale-up
the number of metering/pressure regulating stations by miles of main pipeline or system
throughput, which were the only known population statistics. The station counts from
individual companies were examined both from a per mile main and per system throughput
basis. A linear regression analysis showed that if the data were preferentially extrapolated
using a per mile main basis, the resulting national extrapolation would have lower variability.
In production, the number of separators appears to be technically related to both well count
and throughput. Therefore, separator count was extrapolated by both methods, and the
average of the two national estimates was used.
Activity Factors Developed for Production Sources
Most production activity factors were extrapolated by ratio to known EPs.
The extrapolations were done on a regional basis, since regional biases were known to exist
and each of the well-known EPs (i.e., well count and production throughput) were also
known on a regional basis. Six regions were selected based upon an analysis of the
production and well population centers in the United States., as well as based upon known
differences in practices in various regions. The regions are: 1) Gulf Coast Onshore, 2) Gulf
Coast Offshore, 3) Central Plains (onshore), 4) Atlantic & Great Lakes (onshore), 5) Pacific
and Mountain (onshore), and 6) Pacific Offshore. Figure 5-1 shows the regions selected and
which states are included.
The differences in the regions justify their selection and can be seen in Table
5-4. Specifically, Table 5-4 shows the regional biases that exist in production versus well
28
-------
P.M. Offshore
'vv**-"
G.C. Offshore
Selected Production Regions
-------
TABLE 5-4. REGIONAL DIFFERENCES IN PRODUCTION RATES AND WELL COUNTS
Regional Groupings
States in Region that are
> 50 Bscl/yr
1992 Prodocing Gas
Wells*
1992 Producing Oil Well**
1992 Marketed
Production'
1992 Gross
Production'
Coont
Percent of
Total
Count
Percent of
Total
Bsefy
Percent of
Marketed/v;
Total
Bacfy
Percent
of Gross Total
Gulf Coast Region Total
TX,LA,FL
63667
23.1
217567
36.1
11514
61.5
12272
55.4
GC Offshore'
4021
1.5
5140
0.9
5000
26.7
5045
22.8
GC Onshore
59646
21.6
212427
35.3
6514
34.8
7227
32.6
Central Plains (onshore)
OK,AR,CO,MO,NM,WY,KS
80924
29.3
199103
33.1
5424
29.0
5672
25.6
Pacific and Mountain Total
UT,CA,AK
2266
0.8
46722
7.8
984
5.3
3392
15.3
PM Offshore'
65
0.0
2040
0.3
186
1.0
279
1.3
PM Onshore
2201
0.8
44682
7.4
798
4.3
3113
14.1
Atlantic and Great Lakes
PA,MI,OH,WV
129157
46.8
138805
23.0
790
4.2
796
3.6
(onshore)
TOTAL U.S.
276014
100.0%
602197
100.0
18712
100.0%
622132 100.0%
• Table 3-17, Gas Facts 11
b Natural Gas Annual 12
c Table 3-10, Gas Facts "
-------
count. Each region has a unique oil well versus gas well split and a unique production rate
per well. Two offshore regions exist to account for the known differences in practices
between onshore and offshore production operations. The well and production demographics
also support this split, since the offshore regions account for a small portion of the wells
(1.5% of the gas wells, 1.2 % of the oil wells), but produce 26.1 % of the U.S. marketed
gas production.
As shown in Table 5-4, the majority of natural gas produced in the United
States (more than 64% of total production) occurs is in the Gulf Coast region. Other regions
only account for 36% of the production, and the majority of that occurs in the Central Plains
region. However, the split on well count is completely different. The Atlantic and Great
Lakes region, which accounts for only 3.8% of the gross national gas production, has the
largest portion of gas wells (46% of the national total) as well as a large fraction of oil wells
(24%).
If the source being evaluated is wells or equipment associated with wells (such
as separators and chemical injection pumps), then bias would potentially be introduced if
only wells in the Gulf Coast region were sampled, where most of the gas is produced. The
sources should be combined regionally, and the regional averages then added in the same
proportion that they are distributed in the actual population.
The site data used to develop production activity factors are presented in
Tables 5-5 through 5-10.
Activity Factors Developed for Processing Sources
Activity factors in gas processing are significantly simpler than in gas
production, since the segment consists of one type of facility: gas processing plants. Major
activity factors were limited to the count of gas plants (and gas plant type), the count of
dehydrators, the count of acid gas recovery units (AGRs), and compressor data. All of these
31
-------
TABLE 5-5. PRODUCTION SITE SUMMARY
Offsho
Region
re
(iC
CP
PM
AGL
Total US
Site
6
9
7
9
19
50 sites*
Companies
4
7
7
4
10
32 companies
Survey Type
- Site Visit
1
9
3
9
2
24 site visits
- Phone Survey
5
0
4
0
5
14 phone
- Star Site Visit*
0
0
0
0
12
surveys
12 star sites*
This does not include all sites visited by Star or other fugitive emissions contractors. Only the sites used for activity
factor data collection are included.
-------
TABLE 5-6. DATA FOR OFFSHORE DATA PRODUCTION SITES
Total
Equipment/
Equipment/
Equip./Mkt. Gas
Equip.Prod. Gas
Region
GC-Off GC-Off GC-Off GC-Off PM-Off PM-Off
Offshore
Totals
Total Wills
Gas Wells
(1/MMcfd)
(1/MMcfd)
Site
2
3
4
5
6
Company
2 ¦¦¦¦¦
2
2
3
4
6 Sites
Survey Type
V:' •:
.-.J*-.
P
P
P
4 Companies
GC
PM
GC
PM
GC
PM
GC
PM
GC
PM
Gas Marketed (MMcfd)
0.365
12.5
440
4
17.5
160
456.9
177.5
1.66
5.22
Gas Produced (MMcfd)
0.5
12.5
440
4
17.5
160
457.0
177.5
1.66
5.22
Equipment Counts:
Gas Wells
2
0
80
0
0
12
82
12
Oil Wells
3
150
0
40
22
0
193
22
+Oil wells that market gas
3
150
0
40
22
0
193
22
Separators
4
0
24
0
0
1
28
1
0.10
0.03
0.34
0.08
0.06
0.01
0.06
0.01
In-line Heaters
0
0
0
0
2
0
2
0.00
0.06
0.00
0.00
0.00
0.01
0.00
0.01
Pneumatic Devices
3
0
32
0
0
0
35
0
0.13
0.00
0.43
0.00
0.08
0.00
0.08
0.00
Chem Inj Pumps
0
0
0
0
0
0
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Compressors*
0
-
-
-
-
2
0
2
0.00
0.17
0.00
0.17
0.00
0.01
0.00
0.01
Dehydrators
1
0
8
1
2
3
10
5
0.04
0.15
0.11
0.25
0.02
0.03
0.02
0.03
+ Dehy with 3 pH Flash
1
1
0
+ Dehy with Vent Control
0
0
0
+ Dehy w/Kimray Pumps
1
1
0
+ Dehy w/Stripping Gas
0
0
0
Miles of Gathering Pipeline
-
-
-
-
-
-
Fugitive Component count
Y
N
N
N
N
N
Vented (Site Blow & Purge Data)
Y
Y
Y
Y
Y
Y
Notes: 1) Survey Type V = Site Visit (Radian); P = Phone Survey; S = Site Visit (Star)
2) * = Gas lift compressors not included.
3) Y = Yes, N = No; = No Data;
4) Region Key: GC = Gulf Coast, PM = Pacific Mountain; CP = Central Plains; A = Atlantic & Great Lakes
-------
TABLE 5-7. DATA FOR GULF COAST ONSHORE PRODUCTION SITES
Region
GC
GC
GC
GC
GC
GC
GC
GC
GC
Total GC
Site :
7
8
9
10
ii
12
13
14
15
Equip./
Equip./
Company
5
6
7
8
9
10
U
11
11
9 Sites
Equip./
Equip,/
Mkt, Gas
Prod. Gas
Survey Type
V
V
V
V
¦ V
V
V
¦ y ;
7 Companies
TotBl
Total Wells
Gas Wells
(1/MMcfd)
(1/MMcfd)
Gas Marketed (MMcfd)
23.1
25.5
124
54
28
250
1.9
7
130
643.4
0.54
Gas Produced (MMcfd)
23.1
25.5
124
54
28
250
1.9
7
130
643.4
0.54
Equipment Counts:
Gas Wells
13
80
18
130
26
300
0
10
31
608
Oil Wells
50
0
3
3
0
300
155
127
0
638
+ Oil wells that market gas
50
0
3
3
0
300
155
68
0
579
Separators
38
80
42
71
26
300
0
11
31
599
0.50
0.99
0.93
0.93
In-line Heaters
2
56
17
23
26
0
0
12
0
136
0.11
0.22
0.21
0.21
Pneumatic Devices
68
170
0
68
109
225
0
11
31
682
0.57
1.12
1.06
1.06
Chem Inj Pumps
10
5
0
5
0
0
0
0
0
20
0.02
0.03
0.03
0.03
Compressors*
12
4
2
37
0
0
0
15
10
80
0.07
0.13
0.12
0.12
Dehydrators
7
2
2
12
26
2
0
4
26
81
0.07
0.13
0.13
0.13
+Dehy with 3 pH Flash
0
0
2
2
0
0
0
4
0
8
+Dehy with Vent Control
4
0
2
0
0
0
0
0
0
6
+ Dehy w/Kimray Pumps
7
1
1
6
26
2
0
0
26
69
+ Dehy w/Stripping Gas
0
0
0
0
0
0
0
0
0
0
Miles of Gathering Pipeline
-
46.3
26.4
40
8
-
-
-
-
Fugitive Component count
Y
Y
Y
Y
Y
Y
-
-
-
Vented (Site Blow & Purge Data)
Y
Y
Y
Y
Y
Y
Y
Y
Y
Notes: 1) Survey Type V = Site Visit (Radian); P = Phone Survey; S = Site Visit (Star)
2) * = Gas lift compressors not included.
3) Y = Yes, N = No; = No Data;
4) Region Key: GC = Gulf Coast, PM = Pacific Mountain; CP = Central Plains; A = Atlantic & Great Lakes
-------
TABLE 5-8. DATA FOR CENTRAL PLAINS PRODUCTION SITES
Region
CP
CP
CP
CP
CP
CP
CP
Total CP
Site
16
17
18
19
20
21
22
Equip./
Equip,/
Company
12
13
14
15
16
17
18
7 Sites
Equip;/
Equip,/
Mkt.Gas
Prod. Gas
Survey Type
V
V
V :
!::.:PP::.
:::': P : :
P ::
7 Companies
Total
Total Wells
Gas Wells
(1/MMcfd)
(1/MMcfd)
Gas Marketed (MMcfd)
42.7
180
196
7
0.2
19.8
2
447.7
0.22
Gas Produced (MMcfd)
42.7
180
196
7
0.2
20
2.1
448.0
0.22
Equipment Counts:
Gas Wells
138
321
1000
400
1
100
15
1975
Oil Wells
55
11
0
0
0
0
4
70
+Oil wells that market gas
55
11
0
0
0
0
4
70
Separators
130
321
7
400
1
100
1
960
0.47
0.49
2.14
2.14
In-line Heaters
138
321
0
400
0
0
0
859
0.42
0.43
1.92
1.92
Pneumatic Devices
449
963
667
100
0
2179
1.33
1.38
4.95
4.94
Chem Inj Pumps
28
273
0
13
0
0
0
314
0.15
0.16
0.70
0.70
Compressors*
31
50
64
1
146
0.09
0.10
0.35
0.35
Dehydrators
16
220
0
400
0
25
1
662
0.32
0.34
1.48
1.48
+Dehy with 3 pH Flash
0
0
0
0
-
0
+Dehy with Vent Control
0
0
0
-
-
0
+Dehy w/Kimray Pumps
16
220
0
25
0
261
+Dehy w/Stripping Gas
0
0
0
-
0
Miles of Gathering Pipeline
5.2
-
600
-
-
-
-
Fugitive Component count
Y
Y
Y
-
-
-
-
Vented (Site Blow & Purge Data)
Y
Y
Y
Notes: 1) Survey Type V = Sile Visit (Radian); P = Phone Survey; S = Site Visit (Star)
2) * = Gas lift compressors not included.
3) Y = Yes, N = No; = No Data;
4) Region Key: GC = Gulf Coast, PM = Pacific Mountain; CP = Central Plains; A = Atlantic & Great Lakes
-------
TABLE 5-9. DATA FOR PACIFIC/MOUNTAIN PRODUCTION SITES
Region. -
PM
PM
PM
PM
PM
PM
PM
PM
PM
Total PM
Site
23
24
25
26
27
28
29
30
¦ 31
Equip./
Equip./
Company
19
20
2!
21
21
21
¦: 22
21
21
9 Sites
Equip./
Equip./
Mkt. Gas
Prod. Gas
Survey Type
V
V
V
V
V.
:' V :
V
V
4 Companies
: Total
Total WeUs
Gas Wells
(1/MMcfd)
(1/MMcfd)
Gas Marketed (MMcfd)
4
104
0.138
0.03
0.02
0.035
0.8
0.1
11.082
120.2
0.12
Gas Produced (MMcfd)
4
307
0.138
0.03
0.02
0.035
0.8
0.1
11.082
323.2
0.32
Equipment Counts:
Gas Wells
53
0
0
0
0
0
0
0
0
53
Oil Wells
0
913
18
8
10
15
20
7
728
1719
+Oil wells that market gas
0
137
18
8
10
15
20
7
728
943
Separators
45
0
0
0
0
0
0
0
45
0.17
0.85
0.41
0.14
In-line Heaters
53
5
3
2
0
0
0
0
63
0.24
1.00
0.58
0.20
Pneumatic Devices
80
0
0
0
0
0
0
0
0
80
0.08
1.51
0.67
0.25
Chem Inj Pumps
36
0
0
0
0
0
0
0
0
36
0.04
0.68
0.30
0.11
Compressors*
17
19
0
0
0
0
1
1
38
0.14
0.32
0.35
0.12
Dehydrators
5
0
0
0
0
0
1
0
6
0.02
0.09
0.05
0.02
+ Dehy with 3 pH Flash
0
0
0
0
+Dehy with Vent Control
0
0
0
0
+Dehy w/Kimray Pumps
5
0
1
6
+Dehy w/Stripping Gas
0
0
0
0
Miles of Gathering Pipeline
-
-
-
-
-
-
-
-
-
Fugitive Component count
Y
Y
N
N
N
N
N
N
N
Vented (Site Blow & Purge Data)
Y
Y
N
N
N
N
N
N
N
Notes: 1) Survey Type V = Site Visit (Radian); P = Phone Survey; S = Site Visit (Star)
2) * = Gas lift compressors not included.
3) Y = Yes, N = No; = No Data;
4) Region Key: GC = Gulf Coast, PM = Pacific Mountain; CP = Central Plains; A = Atlantic & Great Lakes
-------
TABLE 5-10. DATA FOR ATLANTIC & GREAT LAKES PRODUCTION SITES
Region
AGL
AGL
AGL
AGL
AGL
AGL
AGL
AGL
AGL ,¦
Site
Company
Survey Type
: 32
33
24 '
p :
34
¦¦¦2SV;V: ":
P
1... 35
V
36
27
P
37
28
p
38
29
' P
39
30
¦¦:¦¦¦¦¦ ' S:: .
40
30
: ¦ 23
V
Gas Marketed (MMcfd)
24
6
15
17
12
16
81
0.18
0.18
Gas Throughput (MMcfd)
24
6
15
17
12
20
81
0.19
0.19
Oil Throughput (1000 B/D)
0
Equipment Counts:
Gas Wells
800
250
1000
520
450
1582
4034
11
11
Oil Wells
0
0
0
0
163
418
0
0
0
+ Oil wells that market gas
0
0
0
163
418
0
0
0
Separators
151
250
500
520
450
1582
3227
-
2
In-line Heaters
0
-
0
Pneumatic Devices
76
0
10
520
450
1582
1294
-
Chem Inj Pumps
0
0
0
12
0
8
25
Y
0
Compressors*
1
-
-
-
-
-
-
-
0
Dehydrators
0
2
1
30
0
0
41
-
+Dehy with 3 pH Flash
0
0
0
0
5
+Dehy with Vent Control
0
3
2
+Dehy w/Kimray Pumps
0
2
1
30
8
0
+Dehy w/Stripping Gas
0
21
0
Miles of Gathering Pipeline
-
-
-
-
-
-
-
-
-
Fugitive Component count
Y
N
N
N
Y
Y
Y
Y
Vented (Site Blow & Purge Data)
Y
Y
Notes: 1) Survey Type V = Site Visit (Radian); P = Phone Survey; S = Site Visit (Star)
2) * = Gas lift compressors not included.
3) Y = Yes, N = No; = No Data;
4) Region Key: GC = Gulf Coast, PM = Pacific Mountain; CP = Central Plains; A = Atlantic & Great Lakes ( Continu ed)
-------
TABLE 5-10. (Continued)
Region
AGL
AGL
AGL AGL
AGL
AGL
AGL AGL AGL
AGL
Total AGL
Site
41
42
43
44
45
46
47
48
49
50
Equip./
Equip./
Company
30
31
31
31
31
31
31
31
31
' 31
19 Sites
Total
Equip./
Equip,/
Mkt. Gas
Prod. Gas
Survey Type
S
S
S
S
S
S '¦
S
S
::
10 Companies
(Sites 32-50)
Total Wells
Gas WellS
(1/MMcfd)
(i/MMcfd)
Gas Marketed (MMcfd)
0.17
0.39
0.37
0.37
0.18
0.23
0.30
0.35
0.13
0.35
173.6
Gas Throughput (MMcfd)
0.17
0.39
0.37
0.37
0.19
0.23
0.30
0.35
0.13
0.35
178.0
Oil Throughput (1000 B/D)
Equipment Counts:
Gas Wells
10
23
22
22
11
14
18
21
8
21
8828
Oil Wells
0
0
0
0
0
0
0
0
0
0
581
+Oil wells that market gas
0
0
0
0
0
0
0
0
0
0
581
Separators
0
10
7
8
5
3
15
17
5
7
6766
0.72
0.77
38.97
38.01
In-line Heaters
0
0
0
0
1
1
0
0
0
0
2
0.00
0.00
0.07
0.07
Pneumatic Devices
3932
0.43
0.46
23.07
22.50
Chem Inj Pumps
0
0
0
0
0
0
0
0
0
0
45
0.00
0.01
0.26
0.25
Compressors*
0
0
0
0
0
0
0
0
0
0
1
0.00
0.00
0.04
0.04
Dehydrators
74
0.01
0.01
0.43
0.42
+Dehy with 3 pH Flash
5
+Dehy with Vent Control
5
+ Dehy w/Kimray Pumps
41
+ Dehy w/Stripping Gas
21
Miles of Gathering Pipeline
Fugitive Component count
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Vented (Site Blow & Purge Data)
Notes: 1) Survey Type V = Site Visit (Radian); P = Phone Survey; S = Site Visit (Star)
2) * = Gas lift compressors not included.
3) Y = Yes, N = No; = No Data;
4) Region Key; GC = Gulf Coast, PM = Pacific Mountain; CP = Central Plains; A = Atlantic & Great Lakes
-------
activity factors were either published and well-defined or were developed through other
studies such as the report by Wright Killen & Co.13 or Volume 11 on compressor driver
exhaust.14 The site data used to developed processing activity factors are presented in Table
5-11.
Activity Factors Developed for Transmission Sources
Activity factors for the transmission segment were simpler than production
segment factors. The transmission segment is definitely more homogeneous than the
production segment. Transmission facilities are either surface compressor stations, surface
metering and regulating stations, buried pipelines, or underground storage. Most
transmission pipelines are one of two types: interstate (cross-country) or intrastate (strictly
regional). Therefore, transmission company data can be extrapolated by using pipeline
miles, station count, or storage facility count.
The total number of compressor stations was extrapolated from data on major
transmission companies listed in Gas Facts.11 The total miles of transmission pipeline and
the number of storage facilities and storage wells is also published. Counts of transmission
metering stations by type were produced by extrapolating data from Radian's company
surveys of several transmission companies. Other transmission activity factors were
developed from Radian site visits to transmission facilities. The site data used to develop the
transmission activity factors is presented in Tables 5-12 and 5-13.
Activity Factors Developed for Distribution Sources
In the distribution segment, activity factors were developed for total number of
leaks in underground mains and services. These activity factors were desegregated by pipe
service (i.e., mains versus services) and pipe material (i.e., cast iron, cathodically protected
steel, unprotected steel, plastic, and copper). The estimates of total leaks for each company
39
-------
TABLE 5-11. GAS PROCESSING PLANTS
Site
1
2
3
4
5
6
7
8
9
10
tl
Total
llllllllllll
Companies
1
1
2
3
3
4
5
5
5
6
7
7
companies
Type
Cryo
Cryo
Cryo
Lean Oil
Abs.,
Cryo
Cryo
Cryo
Lean Oil
Abs.
Refrig.
Refrig.
Refrig./
Lean Oil
Abs.
Capacity (MMscfd)
100
75
70
850
900
40
130
130
140
Current Throughput
(MMscfd)
49
60
56
350
750
140
40
130
130
70
Compr. Units
7
4
6**
9
1
o***
4.4**
1.4**
1.4**
20
19
72
- Turb. Eng
0
0
0
2
1
1
0
0
0
5
2
10
- Recip. Eng
7
4
6**
7
0
0
4.4**
1.4**
1.4**
15
17
62
- Total HP
11000
3700
6740**
43300
27000
20000
5925**
6267**
6267**
59600
189799
Dehys
0
1
2
3
0
0
1
1
1
1
10
Dehys w/Kimray
Pumps
1
0
0
0
0
0
1
Pneum Ongas
2
3
0
25
25
17
0
0
0
0
72
Vented Data
- Site
- Company
Y
Y
Y
Y
-
Y
Y
Y
Y
Fugitive CC
639
357
799
1458
-
-
6831
5902
5902
-
Y*
* Count only of compressor BD OELs, site OELs, and Compressor PRVs.
** Gas lift compressors not counted in the totals for this site with gas lift for oil recovery.
*** 1 turbine drives 2 propane compressors. No NG compressors.
"Y" = Yes
-------
TABLE 5-12. TRANSMISSION COMPRESSOR STATIONS
Site Number
1
2
3
4
5
6
7
8
9
10
iiiiiiiiiiii
Company Number
1
1
1
2
2
2
3
3
4
4
4
Compr. Units
13
2
2
6
7
13
12
13
2
10
6
- Turb. Eng
0
2
2
0
0
0
2
1
2****
3
2
- Recip. Eng
13
0
0
6
7
13
10
12
0
7
4***
- Total HP
32650
6900
6900
16900
10400
24800
14560
17570
40000
-
-
Dehydrators
0
0
0
0
1
-
0
0
6
0
0
- Flash Tanks
1
6
- Kimray Pumps
0
6
- Stripping Gas
0
0
- Vapor Recovery
0
0
- Vent Flash Gas
Pneum
48
12
-
8
-
20
75
40
68
83
50
Wells
Not Applicable
Fugitive CC
741**
223**
165**
-
-
-
3038
3949
1730
1467
956
Site B.D Practices
Y
Y
-
Y
Y
-
-
-
Y
Y
Y
Co. B/D No's
-
-
-
Y
Y
Y
-
-
-
-
-
(Continued)
-------
TABLE 5-12. (Continued)
Site Number
12
13
14
15
16
17
18
19
20
21
21 Sites
Company Number
5
6
6
7
8
8
9
9
10
10
10 Co's
Compr. Units
18
2
2
26
5
3
7
13
7
2
171
- Turb. Eng
0
2
2
1
1
0
0
3
0
2
25
- Recip. Eng
18
0
0
25
4
3
6
10
7
0
145
- Total HP
21000
-
-
-
-
-
-
-
-
-
191,680
Dehydrators
1
0
0
0
0
0
0
0
-
-
8
- Flash Tanks
1
- Kimray Pumps
0
- Stripping Gas
0
- Vapor Recovery
0
- Vent Flash Gas
Pneum
3
38
-
0
Wells
Not Applicable
Fugitive CC
1123
134
284
1706
345
12
508
792
-
-
Site B.D Practices
Y
-
-
Y
-
-
Y
-
Y
Y
Co. B/D No's
-
-
-
-
-
-
-
-
-
-
Fug cc does not include connections or tubing
* = Elec driven compressors
** = Not including hydraulic valves
*** _ Recip Engine w/Centrifugal Compressor
**** = Does not include third turbine that was permanently out-of-service
means no data available, "Y" = Yes
-------
TABLE 5-13. STORAGE COMPRESSOR STATIONS
Site
1
2
3
4
5
6
7
$
iiiwiiiiiii
Companies
1
2
3
4
5
6
7
7
7 Co's
Type
UG
UG
UG
UG
LNG
UG
UG
UG
Compr. Units
3
2
2
4
5*
18
9
9
52
- Turb. Eng
0
0
0
2*
0
4
0
0
6
- Recip. Eng
3
2
2
2
0
14
9
9
41
- Total HP
6250
2200
9400
7000
10300*
48510
9000
11600
104260
Dehydrators
4
1
1
8
0
1
-
-
- Flash Tanks
1
8
1
- Kimray Pumps
0
0
0
- Stripping Gas
0
0
1
- Vapor Recovery
0
0
- Vent Flash Gas
1
0
0
Pneumatics
18
-
68
127
4
-
-
-
217
Wells
?
50
22
83
0
64
-
-
219
Fugitive CC
1750
-
1113
8326
1679
887
-
-
13700
Vented Data:
- Site
- Company
_
Y
Y
Y
Y
Y
Y
Y
Y
Y
Fug cc does not include connections or tubing
* = Elec driven compressors
** = Not including hydraulic valves
*** = Recip Engine w/Centrifugal Compressor
means no data available, "Y" = Yes
UG = Underground Storage Station
LNG = Above ground, Liquefied Natural Gas Station
-------
were based on historical leak records and the average leak per mile (or per service)
extrapolated by the total national mileage (or number of services).
For metering/pressure regulating stations, the number of stations was
collected from each company and extrapolated by the total miles of distribution mains.
Other activity factors used in the distribution segment were based on well-defined activity
data, such as total mileage of mains/services and throughput.
44
-------
6.0
SAMPLING AND STATISTICAL ACCURACY
A key part of this project was the estimation of the accuracy of the overall
emission rate. This section explains the techniques used during sampling to maximize
precision and eliminate bias. The general approach used for statistical accuracy calculations
is also discussed.
Accuracy is made up of precision and bias. Precision can be calculated from
a set of replicate measurements. Bias cannot be calculated, and must be discovered and
eliminated where possible. Figure 6-1 illustrates the role of random and bias errors in the
estimation process. In each of the four illustrations in this figure, the center of the
concentric circles represents the correct answer. In the upper left, there is a significant
amount of random scatter in the points. The term "precision" refers to random variability
alone; in this case, the precision is poor. Additionally, the points are predominantly below
and to the right of the target. The systematic difference between the points and correct
answer is a bias. The term "accuracy" refers to the total error, including random and bias
errors. Because of the large bias and the poor precision, the accuracy is also poor.
In the upper right of Figure 6-1, the points are randomly scattered about the
correct answer; there is little or no bias in this case, but the precision and accuracy are both
poor. In the lower left, there is good precision, but there is again a large bias; thus, the
accuracy is poor. In the lower right, the bias is small and the precision is also good. Thus,
the accuracy is good in this case.
The following subsections discuss the approaches used in sampling to handle
precision, bias, accuracy targets, and overall accuracy calculations.
45
-------
High Bias + Low Precision
= Low Accuracy
Low Bias + Low Precision
= Low Accuracy
High Bias + High Precision
— Low Accuracy
Low Bias + High Precision
= High Accuracy
Figure 6-1. Illustration of Random and Bias Errors
46
-------
6.1
Sampling Approach
In general, a sampling program should gather enough replicate samples to
meet the precision target. If the sample-to-sample variability is high, then more samples
may need to be taken to reach the desired precision target. 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.
Sampling bias occurs if the methodology is flawed in a manner that leads to
a systematic under-representation of parts of the population and a systematic over-
representation of other parts. Bias, in a statistical sense, can be explained as follows.
Suppose it was possible to repeat the sampling and measurement process infinite times, and
that each time the process was repeated, an independent estimate of a given emission factor
was obtained. If the average of the entire infinite set of emission factor estimates equalled
the true value, then bias would not exist. If the average of these estimates differed from the
true value, then the process would be wrong in a systematic sense, and bias would be said
to exist. The point here is that averaging an infinite set of independent estimates of the
emission factor would remove random error altogether, leaving only bias error, if any.
While it is clearly impossible to obtain an infinite set of estimates of an emission factor, the
example given serves to illustrate the meaning of bias.
Even if there was no bias, the actual estimate of an emission factor would be
expected to differ from the true value. First, the estimate is based on less than the total
number of sources. Random differences between the set of sampled sources and the
population of sources introduce a sampling error. Second, physical measurements have
uncertainties. As is indicated above, the term "accuracy" refers to the closeness of an
estimate of a quantity to the true value. Accuracy is a measure of random error plus bias
error. The term "precision" refers to random error alone.
47
-------
An estimate is precise if it has a small random error, regardless of the bias.
Suppose, for example, that sources had been selected only from the Gulf Coast, but that a
very large number of sources had been sampled. The averaging of a large number of
emission measurements would lead to an emission factor estimate that had a small random
error. Unless Gulf Coast sources were representative of the source type for the entire
nation, however, the estimate could have a large bias because the sample of sources did not
represent the general population. Bias in this example is avoided by sampling in a variety
of regions of the country. More subtle potential sources of sampling bias and methods for
avoiding them are discussed in this subsection.
Several sampling approaches can be applied to avoid bias.
Random Sampling
In random sampling, each source in the population has an equal probability
of being selected. A random sample is expected to "match" the industry population because
no biases are introduced in selecting the sample sites. The number of data points required
in a random sample depends on the target accuracy of the final emissions estimate, the
confidence with which this accuracy is to be met, and the underlying variability among the
emissions rates of the complete set of sources.
Random sampling is not a guarantee of accurate results. It is possible, for
example, that by pure chance random sampling would produce a disproportionately large
number of sources from the Gulf Coast and an under-representation of sources from the
West Coast. While such an outcome is unlikely if the sample size is sufficiently large, this
particular problem can be avoided altogether by selecting an acceptable number of sources
from each of a set of regions. (See the discussion of stratified random sampling below.)
There are two major reasons why truly random sampling was not possible in
the GRI/EPA program. First, a complete list of sources did not exist and still does not
48
-------
exist. It was possible, for example, to list all compressor stations whose owners were GRI
members. While this might account for 90% of the compressor stations, the list was not
complete. Another example is the production segment, where it was not possible to
produce a list of all the individual well owners for random selection. The second point is
that the owners of the randomly selected sources could not be required to participate in the
study. For this reason, there is no guarantee that a truly random sample of the available list
could be tested.
Stratified Random Sampling
In stratified random sampling, the population of interest is divided into
subsets, or strata. Then random samples are drawn from each stratum. For example, the
sources of interest in this program could be stratified by geographical region, and random
sampling could be applied within each region.
Stratified random sampling can be performed proportionately or
disproportionately. In proportionate stratified random sampling, the number of sources
sampled in a stratum is in proportion to the total number of sources in that stratum. For
example, if Region A had twice as many sources as Region B, then the sample would
include twice as many sources from Region A as from Region B. From an intuitive point
of view, then, a proportionate stratified random sample "matches" the population, at least
with respect to the criteria used to specify the stratification.
Proportionate stratified random sampling can be used to address the issue of
regional differences, but only if applied properly. In the paragraph above, it is suggested
that sources could be sampled in proportion to the total number of sources by region.
Alternatively, proportionality could be achieved on the basis of gas production, rather than
on the basis of the number of sources. The variable or variables used to achieve
proportionality must be closely related to emissions or proportionate random sampling could
lead to biased results.
49
-------
It is common in practice, however, to sample in such a way that the sample
size for a stratum is not in proportion to the total number of sources in the stratum (and the
throughput of the sampled sources is not proportional to total throughput in the stratum).
This type of sample is called a "disproportionate stratified random sample." This type of
sample does not "match" the population in the sense described above. As long as the
disproportionality is accounted for in computing the final statistics (e.g., mean emission rate
by source or total emissions), disproportionate sampling will not cause bias in the final
results.
Stratified sampling can lead to increased accuracy for the total sample size if
there is less variability within any given stratum than there is in the total population.
Similarly, a smaller sample size might suffice to meet the target accuracy if stratified
random sampling were used rather than random sampling.
Neither type of stratified random sampling was feasible in this study. The
obstacles to random sampling, discussed earlier in this section, were also obstacles to
random sampling within strata.
Further, at the outset of the program, it was not known which variables were
related to emissions; thus, it was not known which variables should be used as a basis for
stratification. If stratification had been performed on the basis of all variables that could
possibly have an influence on emissions, the number of strata (determined by the number of
variables and the number of categories for each variable) could have become unreasonably
large. For example, for leakage from underground distribution mains and services, a
number of parameters were identified that potentially influence emissions: pipe material,
age, operating pressure, diameter, soil type, and parameters characterizing the leak detection
and repair practices of the company. The required sample size can become large because
of the total number of strata, especially if proportional stratified random sampling is used.
One company has embarked upon an independent program to quantify leakage from
underground mains and services using a proportional sampling approach. Even within this
50
-------
single company, hundreds of samples were required to produce a proportionate stratified
random sample for underground pipelines.
Additionally, stratified sampling is of no use unless there are activity factors
that can be used to estimate the emission rate for the population. Complete information for
all variables of potential interest does not exist. For example, the age of a dehydrator, in
most cases, is not even known by the owner of the equipment. It would be pointless to
stratify dehydrator emission factors with respect to age if the necessary activity factors
cannot be obtained.
Sampling Approach Selected for This Program
Thus, because of various practical limitations, neither random sampling nor
stratified random sampling was perfectly 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 very similar to
disproportionate stratified random sampling, with certain differences.
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 sources to be sampled was calculated based on an estimate of the accuracy target
and the estimated standard deviation for the source category. The accuracy targets are
based on the need ultimately to estimate the national emission rate to within 0.5% of the
national production rate based on a 90% confidence limit. Sites were selected in a random
fashion from known lists of facilities, such as GRI or American Gas Association (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.
51
-------
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.
The topic of screening for bias is discussed further in Sections 6.2 and 6.3. If a
relationship between emissions and a parameter was found, then the population, or the
number of sources in the industry, was stratified by that parameter. For example, station
type was found to influence the emission rates from metering and pressure regulating
stations, so the number of stations under each station type in the nation was determined.
To stratify the population of sources by a parameter, data were collected from companies
on the distribution of sources in each stratum and an average over all companies sampled
was determined.
It is important to realize that just because a parameter or set of strata is
identified that has a large effect on the emissions from a given source category, it does not
mean that there is bias in the data. A second condition is necessary. The condition is 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
number of sources in a given stratum to the total number of sources for both the data set
and at the national level. If this known national ratio is different from the 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 were identified, the precision of the emission rate extrapolated
to a national basis was evaluated and compared to the accuracy target. (Note: The
accuracy target is a function of the magnitude of the emissions from the source.) 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 is computed based on the standard
deviation and a 90% confidence interval.
52
-------
6.2
Screening For Bias in Activity Factors
It is impossible to technically prove that there is no bias in any dataset.
While tests can be designed that are capable of revealing some bias, there are no tests nor
group of tests that would reveal all possible biases. Assuming that any dataset has no bias,
even after extensive testing, is only a theory. Such theories can be disproved, but not
proved. The following examples in this section show some of the many bias tests used in
this project.
The sample sets were tested for bias by continuous technical and industry
review. Numerous individual reviews and project advisor's meetings were used to review
the project data with knowledgeable industry experts, so that systematic errors could be
discovered and eliminated. When possible biases in the activity factor sampling plan or
extrapolation method were theorized, the project was altered to test for that bias and
eliminate it if it existed. All provable biases were corrected.
One example of the success of this bias review process includes the
identification of regional differences in production practices. These differences were
brought up by the advisor's meeting review process. The differences were then accounted
for by stratifying the production data into two offshore and four onshore regions, sampling
within each region, and extrapolating by region.
Another example of activity factor screening bias includes extrapolation by
two methods that validate each other. If both methods are technically sound and
independent, and if they deliver the same result for national totals, then this indicates that
there is no bias in the data set related to either variable. If a data set existed that was the
perfect microcosm of the gas industry, one could extrapolate equipment counts from the
data set to national totals by any variable in the data set. Any extrapolation from the
perfect microcosm would deliver the right answer, even technically unrelated data such as
extrapolating separator count by number of employees at the site.
53
-------
Therefore, for an imperfect data set, which all data sets are, extrapolation by
two variables allows for a cross check. For example, production activity factors could be
extrapolated by two nationally known extrapolation parameters: well count and production
rate. It was possible that the extrapolated variable is actually technically related to one
variable more than the other, or that the sample set had some bias related to one or both
variables. Nevertheless, if the extrapolation by well count produces the same value as the
extrapolation by production rate, then the methods tend to validate one another. This also
indicates that there is no bias made in selecting the extrapolation technique, nor between the
relations of the two extrapolation variables in the data set.
If the two methods produced results which differed widely, that might
indicate that one or the other method has a bias. In fact, as mentioned in Volume 5 on
activity factors,1 there was a tendency for the well method to be high-biased and for the
production method to be low-biased. Therefore, the average of the two techniques should
minimize the bias.
6.3 Screening for Bias in Emissions Factors
Screening for bias can be accomplished by identifying design, operational,
and regional parameters that may cause differences in emissions across a source type and
then analyzing the data to determine whether there is an established relationship between
those parameters and the emission rate. Usually, these parameters are chosen on the basis
of industry expertise and/or engineering judgement. If these parameters are determined to
exhibit statistically different emission characteristics, then the population of sources is
stratified into distinct categories by these design, operational, or regional parameters.
Emission factors and activity factors are determined for each category within the source
type to uniquely characterize emissions.
Metering/pressure regulating stations provide an example where the process
of screening for bias was beneficial. Table 6-1 shows the average measured emission factor
54
-------
for metering/pressure regulating stations, in units of scf/station-hour, based on 86
measurements which were performed in 19 cities in the United States. The activity factor
was also derived from data provided by the distribution companies participating in the
study, which was scaled to a national estimate of metering/pressure regulating stations.
Assuming that the sample selection was random or representative, the extrapolated
emissions are 104.1 Bscf, based on the average of all measurements to estimate the
emission factor. However, if the data is subdivided, or stratified, by station type (i.e.,
metering/pressure regulating versus pressure regulating), then the estimated emissions from
this source type decrease to 73.7 Bscf. Furthermore, if this source type is further
subdivided into discrete operating pressure ranges and by enclosure status, the emissions
decrease to 27.3 Bscf. As illustrated, the bias, which was caused by testing a
disproportionate number of high pressure stations, can be minimized by stratifying the
emission and activity factors.
The previous example also illustrates that it is equally as important to
accurately stratify the activity factor samples as the emission factor samples. In some
cases, the activity factor can only be stratified to the necessary level of disaggregation by
gathering data from industry.
Even if a process does not produce a bias in the statistical sense described
above, it is possible for a given segment of the population to be seriously under represented
and another segment to be over represented by random chance (i.e., by an anomaly in the
random selection of sources). The error that results is a larger than expected random error;
an error from a correct sampling and measurement process is not a bias.
55
-------
TABLE 6-1. ESTIMATED METHANE EMISSIONS FROM DISTRIBUTION
METERING AND PRESSURE REGULATING STATIONS
Category
Location
(vault or
above-ground)
Emission Factor
(scf/station-hr)
Activity Factor
(number of
stations)
Emissions
(Bscf)
All Stations
--
90.2
131,799
104.1
M&R Stations
--
154.1
23,922
32.3
Reg. Stations
--
43.7
108,048
41.4
Total
--
131,970
73.7
M&R Stations
>300 psig
A-G
179.8
3,460
5.45
100-300 psig
A-G
95.6
13,335
11.2
40-100 psig
A-G
4.31
7,127
0.269
<40 psig
A-G
—
0
0
Res. Stations
>300 psig
A-G
161.9
3,995
5.67
>300 psig
Vault
1.30
2,346
0.0266
100-300 psig
A-G
40.5
12,273
4.35
100-300 psig
Vault
0.180
5,514
0.0087
40-100 psig
A-G
1.04
36,328
0.332
40-100 psig
Vault
0.0865
32,215
0.0244
<40 psig
Vault
0.133
15,377
0.0179
Total
131,970
27.3
The screening process serves to identify variables that are related to emission
characteristics. Then it is possible to determine whether sources are disproportionately
sampled in the different strata of these variables. Such a disproportionality need not lead to
a bias in the final estimate of emissions, if this condition is identified and accounted for
56
-------
properly. Moreover, the screening process has been carried out during the course of the
study. Thus, additional sampling to correct a disproportionality, if present, is possible.
Note that the screening process would identify unrepresentativeness in the
sample, whether the problem resulted from an inadvertent bias in the sampling process or a
purely random effect. The protection against both bias and anomalies in the random
selection of sources is considered to be a significant benefit of the method used in this
study.
6.4 Accuracy Target
The target uncertainty in the emission estimate is 0.5% of the national
methane emissions, on the basis of a 90% confidence limit for the emissions estimate.
Practical considerations allow sampling only a small percentage of the large number (tens
of thousands) of sources that exist nationwide. Moreover, there is typically a large amount
of variability among the sources in a given category. In view of these considerations,
meeting the accuracy target may seem insurmountable. Despite these facts, the target
precision for the industry emission rate was achieved. The purpose of this section is to
illustrate, through hypothetical calculations, how large errors in emissions estimates for
individual source strata can combine to allow this to occur.
As is discussed in the preceding sections, bias is minimized by randomly
selecting sites (although from a limited list), analyzing the data, and creating strata in a
systematic way. The estimate of total emissions is the sum of the emissions for all the
strata. An essential point is that the uncertainties are not additive; the uncertainty of a sum
is related to the sum of squares of the individual uncertainties. If the errors in a sum vary
independently, then they tend to "average out"; as a result, the relative uncertainty in a sum
of terms (with equal means and variances) is smaller than the relative uncertainty in the
individual terms. (Several statistical points made in this paragraph are discussed in further
detail later in this subsection.)
57
-------
The steady emission sources have been split into five major segments. Each
segment has two to seven significant source categories, and each source category is divided
into 10 to 40 strata. In total, steady sources have been divided into nearly 100 strata.
Unsteady or vented sources have been divided into approximately 40 strata. Thus, in all
there are approximately 140 strata.
Hypothetical calculations are presented that illustrate the effect of summing
the errors in the different strata. For the purposes of the hypothetical calculations, it has
been assumed that there are "n" strata with equal emissions and equal uncertainties based
on random errors. While it is recognized that both the emission rate and the variability
change from stratum to stratum, in actuality, the simplifying assumptions facilitate a
calculation that illustrates the effect of summing the emission estimates from a large
number of strata.
Also, it has been assumed that undiscovered bias, if any, varies
"independently" from stratum to stratum. This type of error would exist if the sources
within a stratum were sampled in an unrepresentative manner, resulting in a bias error.
Clearly, a systematic bias that was common to a large number of strata would have a more
serious effect on the final result. The processes described earlier for screening for bias
provide a protection against this (or any type of) bias error. Additionally, given the large
number and diversity of strata, it is reasonable to believe that any undetected bias will
exhibit a high degree of "independence" among the strata.
Table 6-2 presents the results of the calculations. The random error was
chosen to be as large as plus or minus 100% of the emissions for each stratum, based on a
90% confidence interval. This is equivalent to assuming a coefficient of variation of
approximately 0.6.
58
-------
TABLE 6-2. PERCENTAGE OF ERROR IN TOTAL EMISSIONS
V: Bias
(Percent of
Emissions)
Number of Strata
20 Strata
40 Strata
100 Strata
0%
0.40
0.29
0.18
15%
0.41
0.29
0.18
30%
0.43
0.31
0.19
(Percent random error in a given stratum based upon a 90% confidence interval = 130%)
The bias error is represented as the stratum-to-stratum standard deviation of
the biases in the emission estimates; this quantity is presented as a percent of the emissions
for a stratum. In the calculations, three values have been considered for the bias: 0%,
15%, and 30%. In view of the methods used for screening for bias, 30% is considered to
be a high estimate.
As previously mentioned, the total number of strata is approximately 140. It
has been assumed that there are approximately 100 strata with nearly equal emissions that
represent the major part of the industry emissions. Some of the strata (such as distribution
pipe type) have been aggregated in the final summary table that is presented in Volume 4
on statistical methodology.10 The summary table includes 86 source categories.
Further calculations were performed assuming 40 and 20 strata, in addition to
the case with 100 strata. Given that the parameters discussed above of the random and bias
errors are fixed, the relative uncertainty in the final result decreases as the number of strata
increases. This is because the "error averaging effect" is greater if a larger number of
independent estimated quantities are summed. This does not mean that artificially
increasing the number of strata would improve the accuracy. There would be fewer data
points per stratum, and the uncertainty of the emission estimate for each stratum would
increase.
59
-------
Table 6-2 presents the percentage of uncertainty in the fugitive emissions as a
fraction of the national production rate. The uncertainty is expressed in terms of a 90%
confidence interval. Since bias errors were considered as well as random errors, the
numbers in Table 6-2 represent accuracy, not just precision.
The accuracy target is met if the percentage error is no greater than 0.5%.
Under all scenarios modelled, the uncertainty is less than 0.5%. This is true even in the
case in which there are only 20 strata with approximately equal emissions, and the bias is
30%. These calculations, while hypothetical, illustrate the way in which errors combine in
a sum and show that meeting the accuracy target is feasible, even in the presence of large
percentage random errors in the individual strata and an assumed large undetectable bias
error.
It must be remembered that the random and bias errors were expressed as a
fraction of the emissions in the strata. For these test calculations, the national emissions
were assumed to be approximately 307 Bscf. The accuracy target is expressed as a
percentage (0.5%) of the national gas production per year, which is 22,132 Bscf.
It is stated earlier in this subsection that the uncertainty of a sum is related to
the sum of squares of the individual uncertainties. The uncertainty of the industry annual
emissions have been calculated as the square root of the sum of the squares of the
uncertainties of the emission rates by category. This method is strictly valid if the errors in
the different terms are uncorrelated. Two terms would have uncorrelated errors if there
were no common source of error. The method would still be valid if the correlations were
negligible.
A method recommended in the Quality Control Handbook15 by Juran, Gryna,
and Bingham was used for quantifying the uncertainty of a sum (on the basis of assumed
uncorrelated errors). The different terms being summed do not have to have the same
statistical distribution. In fact, Juran, et al., illustrate the method with an example in which
60
-------
three terms are summed, no two of which have estimated values that are the same within
uncertainty. Also, no two of the terms have the same uncertainty. This method is based on
a theorem that is proved by Mood, Graybill, and Boes in Introduction to the Theory of
Statistics16 (p. 178). The methods used for analyzing error propagation, as well as
alternative schemes, are discussed in much further detail in a separate project report
documenting the statistical methods10 used in this study.
It has been pointed out that there may be correlations between the errors in
the emission rates for different source categories. For example, in some instances, the same
activity factor applies to more than one category. Also, there are instances in which data
for more than one source category were collected from the same field. If inspection and
maintenance practices at that field were better than the industry average, for example, this
fact could have a common effect on the data for all categories sampled at that field. An
assessment of the effect of correlated errors among categories has been performed. The
results are discussed in Volume 4 on statistical methodology.10
6.5 Overall Statistical Accuracy
The precision is computed by rigorously calculating errors for average values
produced from replicate measurements, and then by propagating error from each individual
group of measurements into the national numbers. This section provides a brief discussion
of the statistical methods used for the overall methane emissions project. Volume 5 on
activity factors' summarizes the general statistical propagation techniques used, and
Volume 410 provides the full details of the various statistical techniques, tests, and
considerations involved with this project.
During 1992, this project used a statistical analysis to apportion resources to
refine the categories with the highest emission rates and/or the poorest accuracies. This
method allowed for additional measurements or calculations that quickly refined and
tightened the precision of the estimate. The 1992 plan also set an absolute accuracy target
61
-------
of ± 0.5% of gross production for the 90% upper confidence bound of the annual national
emission rate. This is equivalent to an absolute accuracy target of approximately ±111
Bscf. For the estimate of 307 Bscf, this means that the annual national estimate must have
an upper confidence interval (at 90% confidence) that all possible answers will fall within ±
35% (106/307) of the actual estimate. Any one source category within the sum of all
sources may have a much larger confidence interval than ± 35%, since error bands are not
additive directly, but are added as the square-root of a sum of squares. Consequently, it is
possible for a sum to have a smaller relative uncertainty than do the individual terms
(although the absolute uncertainty of the sum is larger than that of any individual term).
The project progress was tested in 1993 after significant data had been
collected. The test showed that the precision target had already been reached, so additional
tests were not needed to improve the precision. However, accuracy is made of precision
and bias, and data collection continued throughout 1993 and early 1994 in order to reduce
and eliminate bias. Precision can be calculated and improved by additional measurements,
but bias must be eliminated through proper testing design, proper extrapolation, bias tests,
and detailed data analysis and review. Testing design is a key factor in eliminating bias
since sources cannot really be randomly selected for a number of practical reasons.
Another potential source of bias can be eliminated by subdividing the data on an
appropriate basis (such as grouping metering and regulating stations into various pressure
categories), and reanalyzing the data. Stratifying, or subdividing, the data is beneficial if
there are differences in the means and variances within these strata.
62
-------
7.0 REFERENCES
1. 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.
2. 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.
3. Campbell, L.M. and B.E. Stapper. Methane Emissions 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.
4. Campbell, L.M., M.Y. 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.
5. 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.
6. 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.
7. 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.
8. 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.
63
-------
Myers, D.B. 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.
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.
American Gas Association. Gas Facts, 1992 Data, Arlington, VA, 1993.
U.S. Department of Energy/Energy Information Administration. Natural Gas
Annual, 1992. DOE/EIA-O131(92), Washington, DC, Sept. 1992.
Wright Killen & Co. Natural Gas Dehydration: Status and Trends, Final
Report. GRI-94/0099, Gas Research Institute, Chicago, IL, January 1994.
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.
Juran, J.M., F.M. Gryna, Jr., and R.S. Bingham, Jr. Quality Control
Handbook, Third edition, McGraw-Hill Book Company, New York, NY,
1974, p. 178.
Mood, A.M., F.A. Graybill, and D.C. Boes. Introduction to the Theory of
Statistics, Third edition, McGraw-Hill Book Company, New York, NY, 1974.
-------
APPENDIX A
Production Source Sheets
A-l
-------
SOURCE SPECIFIC EMISSION ESTIMATES
This appendix presents the source specific emission estimates derived from
emission factors and activity factors for the sources of emissions within the natural gas
industry. Each significant source of emissions has a "source sheet" that gives a synopsis of
the basis of the estimate. These estimates are presented in a format which documents the
approach for extrapolation of data to a national estimate. The emission factor and activity
factor presented represent the final estimates from the program.
Each source sheet is divided into two sections, each one describing the basis
for the emission factor and activity factor, respectively. The approach used to determine the
accuracy of the emission and activity factor was discussed in Section 6.0, unless otherwise
stated.
A-2
-------
APPENDIX A
Production Source Sheets
Page
P-l - Compressor Exhaust A-5
P-2 - Fugitive Emissions A-10
P-3 - Gathering Pipeline Leaks A-17
P-4 - Pneumatic Devices A-22
P-5 - Chemical Injection Pumps A-25
P-6 - Glycol Dehydrator Vents A-28
P-7 - Glycol Pumps A-32
P-8 - Maintenance Venting A-35
P-9 - Upsets A-38
P-10 - Pipeline Dig-ins A-41
P-l 1 - Well Completion and Workovers A-42
A-3
-------
PRODUCTION SOURCE SHEETS
This section contains the specific source sheets for the production segment of the natural gas
industry. The following table serves as a guide for finding sheets in this section. The cells
in the table give the sheet number (P-l, P-2, etc.) of the source sheet. The rows define the
equipment covered, while the columns define the operating mode and emission type. A
category with no sheet number means that the emissions from that area were determined to
be negligibly small. The label for each source sheet is shown at the top of the cover page
for that sheet.
TABLE OF
CONTENTS
OPERATING MODE,
EMISSION TYPE (Fugitive, Vented, or Combusted)
EQUIPMENT:
Start Up
Normal Operations
Maintenance
Upsets
Mishaps
V
C
F
V
C
V
C
V
C
V
Wellheads
P-2
P-4,
P-5
P-8,
P-ll
P-9
Heaters
P-2
P-4
P-8
P-9
Separators
P-2
P-4
P-8
P-9
Dehydrators
P-2
P-4,
P-6,
P-7
P-8
P-9
Compressors
P-2
P-4
P-l
P-8
P-9
Metering
P-2
P-8
P-9
Pipelines
P-3
P-8
P-9
P-10
A-4
-------
p-1
ALL-SEGMENT SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
Compressors, Generators
Normal Operation
Unsteady, Combusted (Compressor Driver Exhaust)
24.4 Bscf ±64%
BACKGROUND:
Compressors are used to move gas through the system. They are located in production fields, processing
plants, gas storage fields, and along transmission lines. Methane emissions are found in compressor driver
exhaust (reciprocating engines and gas turbines) because of the incomplete combustion of the natural gas burned
as fuel.
EMISSION FACTOR: (0.240 ± 5% scf/hp-hr, engines and 0.0057 ± 30% scf/hp hr, turbines)
An average emission rate was calculated for each model of compressor engine and turbine in the GRI
TRANSDAT Emissions Database (1), which is based on compressor tests conducted by Southwest Research
Institute (SwRI). The emission rates were calculated from the reported methane emissions per unit of fuel and
the reported fuel use rate (FUR) for each compressor model, as follows:
The following equation was used to determine the total emissions for the 86 turbines and 775 reciprocating
engines in the Emissions Database.
(1)
where: ER(m) = average emission rate for model, m (scf/hr)
EP(m) = average emission parameter for model, m (scf CRJscf fuel)
FUR(m) = average fuel use rate for model, m (scf fuel/hr)
K M
TE = £ £ (ER(m) X HR, )
(2)
m = 1 i = 1
m
where: TE = total emissions for database, (scf)
HRj = annual operating hours for compressor i, (hr/yr)
K = number of unique compressor models
M = number of compressors of model, m
A-5
-------
The emission factors, for engines and turbines, were then calculated using the following equation.
Emission Factor = TE /
N N
£ HP; X ( £ HR; / N)
i = 1 i =1
(3)
where: HP = average operating horsepower during HR, (hp)
HR = annual operating hours, (hr/yr)
N = number of compressors
This equation considers that some models could be operated at a higher percentage of the time because they are
base loaded compressors. The average emission factors for the compressor drivers in the Emissions Database
are 0.240 scf/hp-hr for reciprocating engines and 0.0057 scf/hp-hr for turbines.
EF DATA SOURCES:
1. "National Estimate of Methane Emissions from Compressors in the U.S. Natural Gas
Industry" (2).
EF ACCURACY: +5%, engines and + 30%, turbines
Basis:
The accuracy for the EF is estimated based on propagation of error from the spread of samples in the
database. However, engineering judgement was used to assign accuracy for two of the individual terms
in the equation, as follows:
1. Hydrocarbon analysis was estimated to be + 10%, based on the generally accepted accuracy
of gas chromatographs (flame ionization detector).
2. Likewise, fuel flow measurements were estimated to be ± 2.5%.
ACTIVITY FACTORS: (horsepower-hour)
Horsepower-hour data were available for the production industry segment activity factor calculation. Two
pieces of information are needed to calculate the activity factor, which is expressed as horsepower-hours (hp-hr)
for each type of driver in each of the remaining industry segments. These are the installed horsepower and the
average operating hours. The following table presents these parameters and the resulting activity factors for
both engines and turbines in each segment of the industry. The sources and methods for calculating all the
values presented in the table below are given in the next section: AF Data Sources.
It is estimated that about 94% of the emissions in compressor and generator driver exhaust are from
reciprocating engines used in production, processing, and transmission, with about 5% attributable to
reciprocating engines used in storage. All other categories are negligible in comparison. Therefore, it is more
important to accurately determine the activity factors for reciprocating engines in production, processing, and
transmission.
A-6
-------
COMPRESSOR DRIVER ACTIVITY FACTORS FOR EACH INDUSTRY SEGMENT
Installed
Installed
Annual
Annual
Industry
Engine
Turbine
Honrs
Hours
Engine
Turbine
Segment
MMhp"
MMhp"
Engine
Turbine
MMHp • jhr
MMHpir
Production
NA
NA
NA
NA
27,460 ± 200%
0
Processing
4.19 ± 132%"
5.19 ± 99.4%"
6626 ± 11.5%
6345 ± 48.4%
27,760 ±133%
32,910 ± 121%
Transmission
Compressor Drivers
10.2 ± 10.0%
4.55 ± 10.0%
3964 ± 13.8%
2118 ± 31.3%
40,380 ± 17.1%
9635 ± 33.0%
Generator Drivers
1.45 ± 23.3%
0.045 ± 166%
1352 ± 38.0%
474 ± 620%
1962 ± 45.4%
21.2 ± 1215%
Storage
Compressor Drivers
1.33 ± 13.5%
0.59 ±13.5%
3707 ±23.1%
2917 ± 620%
4922 ± 26.9%
1729 ± 626%
Generator Drivers
0.085 ± 126%
0.057 ±184%
191 ± 377%
36 ± 620%
16.3 ± 621%
2.05 ± 1312%
1 Does not include horsepower associated with gas lift for oil recovery or with electric drivers.
h Average of two estimation methods.
AF DATA SOURCES:
1. The production segment horsepower is based on the total installed horsepower-hours for data
provided by one company for 516 compressor drivers (all reciprocating engines). The
horsepower-hours for the company was divided by their production before scaling to a national
estimate. National horsepower-hour was calculated using the 1992 marketed production for the
U.S. [Natural Gas Annual 1992, (3)].
2. The processing segment horsepower was determined by taking the average of two methods.
Each of the methods uses site data for the 10 gas plants visited. The first method scales to a
national estimate by multiplying the total U.S. gas plant throughput as of January 1, 1993
[46,510.7 MMcfd, Oil & Gas Journal (4)] by the total site visit horsepower per throughput
(47.8 hp/MMcfd, engines and 59.2 hp/MMcfd, turbines). The second method scales to a
national estimate by multiplying the total number of gas plants in the U.S. [726, Oil & Gas
Journal (4)] by the total site visit horsepower per number of gas plants visited (10), which is a
scale-up ratio of about 73. The annual operating hours are based on the 10 sites plus data
from two companies for an additional 18 gas plants. An average of the average operating
hours per site was calculated to get the processing segment operating hours (203 engines and 9
turbines).
3. The transmission segment compressor station horsepower for each compressor driver type is
based on the GRI TRANSDAT database. Installed horsepower was taken from the Industry
Database module of GRI TRANSDAT. The annual operating hours are based on information
reported on FERC Form No. 2. FERC data does not distinguish between driver type. The
FERC data were split between engines and turbines based on data in GRI TRANSDAT and
data provided by one transmission company (524 engines and 89 turbines).
4. The storage segment horsepower came from Gas Facts (5) data for 1992 (1,920,441 hp). The
split between engines and turbines was assumed to be the same as the engine and turbine splits
found in GRI TRANSDAT (69.1%, engines and 30.9%, turbines). The annual operating
hours are based on 11 storage stations (50 engines and 6 turbines). An average of the average
operating hours per station was calculated to get the storage segment operating hours.
A-7
-------
5. The generator driver horsepower (compressor stations) is based on the total installed
horsepower for 7 of the transmission sites visited and company data for 34 transmission
compressor stations. To scale to a national estimate, the total horsepower per station was
multiplied by the total number of transmission compressor stations [1700, FERC Form No. 2
(6)] in the U.S. The annual operating hours are also based on data from the site visits and
company data. An average of the average generator operating hours per station was calculated
to get generator operating hours (87 engines and 1 turbine).
6. The generator driver horsepower (storage fields) is based on the total installed horsepower for
9 storage fields (one company). To scale to a national estimate, the total horsepower per field
was multiplied by the total number of storage fields [475, Gas Facts (5)] in the United States.
The annual operating hours are also based on the company data. An average of the average
generator operating hours per field was calculated to get generator operating hours (3 engines
and 1 turbine).
AF ACCURACY:
Basis:
Errors were propagated from each of the following terms:
1. Production Hp-hr: The production Hp-hr accuracy is based upon an engineering analysis and
set at ± 200%.
2. Transmission Hp-hr: The transmission Hp-hr accuracy is based upon an assigned estimated
error of + 10% for the horsepower data in the GRI TRANSDAT database and error
propagation from the FERC operating hours.
3. Other segment Hp-hr: The accuracy of the site visit data for horsepower and operating hours
was also propagated using the spread of the data, but from much smaller data sets. The
accuracy of the horsepower-hour activity factors for each industry segment are calculated
statistically using the individual terms for horsepower and operating hours.
ANNUAL METHANE EMISSIONS: (24.57 ± 65.1% Bscf, engines + 0.256 ± 97.8% Bscf, turbines)
The annual emissions were determined by multiplying an emission factor by the horsepower-hour activity factor
for reciprocating engines and turbines and summing these values for each segment. The following table shows
the resulting emissions for each industry segment and the overall national estimate.
ANNUAL COMPRESSOR EMISSIONS FOR THE NATURAL GAS INDUSTRY BY SEGMENT
Compressor
Production
Processing
Transmission
Storage
Generators
TOTAL
Engines, Bscf
6.58± 200%
6.65 ±133%
9.68±17.9%
1.18 ±26.9%
0.474+45.6%
24.57±65.1%
Turbines, Bscf
0.00
0.186±129%
0.0546 ±45.7%
0.00979 ±654%
0.000132±1163%
0.256±97.8%
REFERENCES
1. Biederman, N. GRI TRANSDAT Database: Compressor Module, (prepared for Gas Research
Institute) npb Associates with Tom Joyce and Associates, Chicago, IL, August 1991.
2. Jones, D.L., L.M. Campbell, C.E. Burklin, M. Gundappa, and R.A. Lott, "National Estimate of
Methane Emissions from Compressors in the U.S. Natural Gas Industry." Radian Corporation and
Gas Research Institute, Air & Waste Management Association Conference Proceedings, Paper # 92-
142.02, Kansas City, MO, 1992.
A-8
-------
3. U.S. Department of Energy/Energy Information Administration. Natural Gas Annual, 1992.
DOE/EIA-0131 (92), Washington, DC, September 1992.
4. Oil & Gas Journal. 1992 Worldwide Gas Processing Survey Database, 1993.
5. American Gas Association, Gas Facts: 1992 Data, Arlington, VA, 1993.
6. Federal Energy Regulatory Commission (FERC) Form No. 2: Annual Report of Major Natural Gas
Companies, 1992 database.
A-9
-------
P-2
PRODUCTION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
All Production Equipment (See Below)
Normal Operation
Steady, Fugitive
17.4 Bscf ± 41%
BACKGROUND:
Equipment leaks are typically low-level, unintentional losses of process fluid (gas or liquid) from the sealed
surfaces of above-ground process equipment. Equipment components that tend to leak include valves, flanges
and other connectors, pump seals, compressor seals, pressure relief valves, open-ended lines, and sampling
connections. These components represent mechanical joints, seals, and rotating surfaces, which in time tend to
wear and develop leaks.
EMISSION FACTOR: (scf/equipment-yr, see below)
In the component method for estimating emissions from equipment leaks, an average emission factor is
determined for each of the basic components, such as valves, flanges, seals, and other connectors that comprise
a facility. The average emission factor 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. An average estimate of the emissions per equipment or facility are determined as the product of the
average emission factor per component type (i.e., the component emission factor) and the average number of
components associated with the major equipment or facility:
where:
Nx = average count of components of type x per plant, and
EFX = average methane emission rate per component of type x.
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 Eastern U.S. (i.e., Atlantic and Great Lakes region) and the
Western U.S. (i.e., rest of the country, excluding the Atlantic and Great Lakes region) and between offshore
production in the Gulf of Mexico and the Pacific Outer Continental Shelf (OCS). Separate measurement
programs were conducted to account for these regional differences.
Onshore Production in the Eastern U.S. Region. Gas production in the Eastern U.S. accounts for only 4.2%
of gross national gas production, but includes 47% of the total gas wells in the country. Component emission
factors for onshore production in the Eastern U.S. were based on a measurement program conducted by
GRI/Star Environmental (1) of 192 individual well sites at 12 eastern gas production facilities. Component
counts for gas wellheads, separators, meters and the associated above-ground piping, and gathering compressors
were based on information collected as part of the Eastern U.S. production measurement program. Site visits
and phone surveys of 7 additional sites provided data used for determining the number of heaters and
dehydrators in the Eastern U.S. region. Component counts for heaters and dehydrators were assumed to be
identical to those derived from data collected in the Western U.S. The following table presents the component
emission factors, average component counts, and average equipment emissions for onshore gas production in the
Eastern U.S. region.
EF = [(Nvlv x EFV|V) + (Ncn x EFJ + (NMl x EF^,) + (N^ X EF^)]
A-10
-------
Average Equipment Emissions for Onshore Production in the Eastern U.S.
Equipment Type
Component Type
Component
Emission Factor,
Mscf/component-yr
Average
Component
Count
Average Equipment
Emissions,"
scf/equipment-yr
Gas Wellheads
Valve
0.184
8
2,595 (27%)
Connection
0.024
38
Open-Ended Line
0.42
0.5
Separators
Valve
0.184
1
328 (27%)
Connection
0.024
6
Heaters
Valve
0.184
14
5,188 (43%)
Connection
0.024
65
Open-Ended Line
0.42
2
Pressure Relief
Valve
0.279
1
Glycol
Valve
0.184
24
7,938 (35%)
Dehydrators
Connection
0.024
90
Open-Ended Line
0.42
2
Pressure Relief
Valve
0.279
2
Meters/Piping
Valve
0.184
12
3,289 (30%)
Connection
0.024
45
Gathering
Valve
0.184
12
4,417 (27%)
Compressors
Connection
0.024
57
Open-Ended Line
0.42
2
1 Values in parentheses represent the 90% confidence interval.
Onshore Production in the Western U.S. Region. Component emission factors for onshore production in the
Western U.S. were based on a comprehensive fugitive emissions measurement program conducted by API/GRI
(2) at 12 oil and gas production sites. In this program, measurement data were collected from 83 gas wells at 4
gas production sites in the Pacific, Mountain, Central, and Gulf regions. The average component counts for
each piece of major process equipment associated with gas production in the Western U.S. were based on data
collected during the API/GRI study and additional data collected for GRI during 13 site visits to gas production
fields. The following table presents the component emission factors, average component counts, and average
equipment emissions for onshore gas production in the Western U.S. region.
A-ll
-------
Average Equipment Emissions for Onshore Production in the Western U.S.
Equipment Type
Component
Type
Component
Emission Factor,
Mscf/component-yr
Average
Component
Count
Average Equipment
Emissions,"
scf/equipment-yr
Gas Wellheads
Valve
0.835
11
13,302 (24%)
Connection
0.114
36
OEL
0.215
1
Separators
Valve
0.835
34
44,536 (33%)
Connection
0.114
106
OEL
0.215
6
PRV
1.332
2
Heaters
Valve
0.835
14
21,066 (40%)
Connection
0.114
65
OEL
0.215
2
PRV
1.332
1
Glycol Dehydrators
Valve
0.835
24
33,262 (25%)
Connection
0.114
90
OEL
0.215
2
PRV
1.332
2
Meters/Piping
Valve
0.835
14
19,310 (30%)
Connection
0.114
51
OEL
0.215
1
PRV
1.332
1
Gathering
Valve
0.835
73
97,729 (68%)
Compressors
Connection
0.114
179
OEL
0.215
3
PRV
1.332
4
Compressor
Seal
2.37
4
Large Compressor
Stations
Station Components
Compressor-
Related Comnonents
b
b
b
3.01 x 106 (102%)
b
b
b
5.55 x 106 (65%)
a Values in parentheses represent the 90% confidence interval.
b Refer to T-l source sheet for a discussion of the basis for estimated emissions from large compressor
stations.
A-12
-------
Offshore Gas Production. Emissions from equipment leaks from offshore production sites in the U.S. were
based on two separate measurement programs:
• The API/GRI oil and natural gas production operations study, which included 4 offshore
production sites in the Gulf of Mexico; and
• The Minerals Management Service study of 7 offshore production sites in the Pacific Outer
Continental Shelf.
The component emission factors and component counts were taken directly from the field test reports from these
studies. The following table presents the component emission factors, component counts, and average facility
emissions for offshore production in the Gulf of Mexico and Pacific OCS.
Average Facility Emissions for Offshore Production
Equipment Type
Component Type
Component
Emission Factor,
Mscf/component-yr
Average
Component Count
Average Facility
Emissions,8
Mscf/yr
Gulf of Mexico
Platform
Valve
0.187
2,207
1,064 (27%)
Connection
0.046
8,822
Open-Ended Line
0.368
326
Other
2.517
67
Pacific OCS
Platform
Valve
0.048
1,833
430 (36%)
Connection
0.021
13,612
Open-Ended Line
0.092
313
Other
0.091
307
a Values in parentheses represent the 90% confidence interval.
EF DATA SOURCES:
1. Emission Factors for Eastern Gas Production based upon data from the GRI/Star program (1)
for the component EF's at 12 gas production sites.
2. Fraction of methane (78.8 mol%) based on data from Methane Emissions from the Natural Gas
Industry, Volume 6: Vented and Combusted Source Summary (3). Conversion of emission
factors from (pounds THC per day) to (methane Mscf/yr) also required estimation of gas
average molecular weight. Based on data from Perry's Chemical Engineer Handbook (5th
Edition) (4), Table 9-15, selected most similar gas composition speciation from C, through C6+
and performed linear extrapolation from average of 3 lowest data (87 mol% methane) to 78.8
mol% methane. Resultant weight percent of 69.6 wt% methane used to speciate methane
emissions.
3. Component counts in Eastern gas production were based on average counts per equipment
from the GRI/Star program at 12 gas production sites. Component counts for heaters and
dehydrator in the Eastern region were based on data collected in the Western region.
Component counts for onshore production in the Western U.S. were based on the averages
A-13
-------
from the GRI/Star program at 4 gas production sites and GRI/Radian data from 13 site visits
to gas production fields.
4. Offshore data from API/GRI/Star 20-site program for Gulf of Mexico platforms (4 platforms,
site numbers 17 through 20), and Minerals Management Service/ABB Pacific OCS fugitive
study (7 platforms). See respective test reports (Gulf of Mexico Offshore: API/Star 20-site
study (5); Pacific OCS Offshore: MMS report 92-0043 November 30, 1992) (6).
5. Large gathering compressors and large gathering compressor station emission factors are taken
from Transmission segment (see Sheet T-l).
EF PRECISION:
Gas Wells - Eastern
±
27%
Separators - Eastern
+
27%
Heaters - Eastern
+
43%
Dehydrators - Eastern
+
35%
Meters/piping - Eastern
±
30%
Gathering compressors - Eastern
+
27%
Gas Wells - Western
+
24%
Separators - Western
±
33%
Heaters - Western
+
40%
Dehydrators - Western
±
25%
Meters/piping - Western
+
30%
Gathering compressors - Western
+
68%
Large Gathering Compressors
+
65%
Large Gathering Stations
+
102%
Offshore (Gulf)
+
27%
Offshore (Pacific)
±
36%
Basis:
The accuracy is rigorously propagated through the EF calculation from the range of individual
measurements. Ninety percent confidence intervals were calculated for the sites using the t-statistic
method. Computed 90% confidence intervals for site average component counts were combined with
90% confidence intervals for component emission factors to obtain pooled uncertainty in aggregate
emission factor.
ACTIVITY FACTOR: (129157 Gas Wells - Eastern)
±
5%
(91670 Separators - Eastern)
+
23%
(260 Heaters - Eastern)
±
196%
(1047 Dehydrators - Eastern)
±
20%
(76262 Meters - Eastern)
+
100%
(129 Gathering Compressors - Eastern)
±
33%
(142771 Gas Wells - Western
+
5%
(74674 Separators - Western
+
57%
(50740 Heaters - Western
±
95%
(36777 Dehydrators - Western
+
20%
(301180 Meters - Western
+
100%
(16915 Gathering Compressors - Western)
+
52%
(96 Large Gathering Compressors)
±
100%
(12 Large Gathering Stations)
+
100%
(1092 Gulf of Mexico platforms)
+
10%
(22 Western offshore)
+
10%
A-14
-------
AF DATA SOURCES:
1. The gas well count is from A.G.A.'s Gas Facts 1992 data (7).
2. Eastern gas wells and equipment AFs were regionalized using site visit data. Eastern meter
AF based on 0.43 meter per gas industry well (per Star Environmental). Western U.S. meter
AF based on industry advisor information of 1:1 meter per gas industry well.
3. Dehydrator counts are based on 37,824 glycol dehydrators in production (see Sheet P-6 for
details). Adjustment to activity factor for Eastern gas production: subtract 1,047 dehydrators
(included in Eastern gas production component counts).
4. Offshore platform counts provided by Offshore Data Services, Inc., Houston, Texas, and
Minerals Management Service MO AD database for producing platforms. Assumed 50/50 split
between "oil" industry and "gas" industry.
5. Large gathering compressors and compressor station counts were estimated from FERC Form
2 database. Large gathering compressor stations were those with at least 16 stages of
compression (5 compressors per station and an average of 3.3 stages per compressor). The
result was extrapolated to the national total by ratioing on gathering miles covered in FERC to
total gathering mileage.
6. The other equipment counts were produced from equipment count data taken during the site
visits by Radian and Star. As explained in the activity factor section of the text of this report,
extrapolation to national counts was done on a regional basis to account for regional equipment
configuration differences.
AF PRECISION:
Basis:
1. The precision for the active wells is assigned by engineering judgement, based upon the fact
that the number of active wells is tracked nationally and known accurately by A.G.A./DOE,
etc.
2. The accuracy for the other equipment types is based upon rigorous propagation of error from
the range in averages from the 9 production sites visited.
REFERENCES:
1. Star Environmental. Fugitive Hydrocarbon Emissions: Eastern Gas Wells, Final Report, GRI-
95/0117, Gas Research Institute, Chicago, IL, July 1995.
2. Star Environmental. Fugitive Hydrocarbon Emissions from Oil and Gas Production Operations, (API
Publication No. 4589). American Petroleum Institute, December 1993.
3. 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.
4. Perry, R.H. et al. (ed.). Perry's Chemical Engineers' Handbook. Sixth Edition, McGraw -Hill Book
Co., New York, NY, 1984.
5. Star Environmental. Emission Factors for Oil and Gas Production Operations. (API Publication No.
4615). American Petroleum Institute, January 1995.
6. R. Countess, et al. Fugitive Hydrocarbon Emissions from Pacific OCS Facilities. Volume 1, Final
Report. MMS 92-0043, Minerals Management Service, Washington, DC, 1992.
7. American Gas Association. Gas Facts, 1992 Data, Arlington, VA, 1993.
A-15
-------
ANNUAL EMISSIONS: (17.4 Bscf/yr ± 7.1 Bscf/yr)
The annual emissions were determined by multiplying the average equipment emissions by the population of
equipment in the segment.
Category
Emission Factor
Activity Factor
Emission Rate
Uncertainty
Gas Wellhead (Eastern U.S.)
2595 scf/yr methane
129157 gas wells (Eastern U.S.)
0.34 Bsctfyr methane
27%
Separators (Eastern U.S.)
328 scf/yr methane
91670 separators (Eastern U.S.)
0.03 Bscf/yr methane
36%
Heaters (Eastern U.S.)
5187 scf/yr methane
260 heaters (Eastern U.S.)
0.001 Bscf/yr methane
218%
Dehydrators (Eastern U.S.)
7939 scf/yr methane
1047 dehydrators (Eastern U.S.)
0.008 Bsctfyr methane
41%
Meters/Piping (Eastern U.S.)
3289 scf/yr methane
76262 meters (Eastern U.S.)
0.25 Bscf/yr methane
109%
Gathering Compressors (Eastern U.S.)
4417 scf/yr methane
129 gathering compressors (Eastern U.S.)
0.0006 BscCyr methane
44%
Gas Wellheads (Western U.S.)
13302 scf/yr methane
142771 gas wells (Western U.S.)
1.9 Bscf/yr methane
25%
Separators (Western U.S.)
44536 scffyr methane
74674 separators (Western U.S.)
3.33 Bsctfyr methane
69%
Heaters (Western U.S.)
21066 scf/yr methane
50740 in-line heaters (Western U.S.)
1.07 Bscf/yr methane
110%
Dehydrators (Western U.S.)
33262 scf/yr methane
36777 dehydrators (Western U.S.)
1.22 Bsctfyr methane
32%
Meters (Western U.S.)
19310 scf/yr methane
301180 meters (Western U.S.)
5.82 Bscf/yr methane
109%
Small Gathering Compressors (Western
U.S.)
97729 scf/yr methane
16915 compressors (Western U.S.)
1.65 Bscf/yr methane
93%
Large Gathering Compressors (Western
U.S.)
5.55 MMscf/yr methane
96 large compressors
0.53 BscCyr methane
136%
Large Gathering Compressor Stations
(Western U.S.)
3.01 MMscf/yr methane
12 large gathering compressor stations
0.04 Bscf/yr methane
176%
Offshore Oil/Gas (Gulf)
1064 Mscf/yr methane
1092 Gulf of Mexico platforms
1.16 Bsctfyr methane
29%
Offshore Oil/Gas (Pacific)
430.0 MscCyr methane
22 platforms (Pacific)
0.01 Bscf/yr methane
38%
TOTAL
17.4 Bscf/yr methane
41%
A-16
-------
P-3
PRODUCTION SEGMENT SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
Gathering Pipelines
Normal Operations
Steady, Fugitives (Pipeline Leaks)
6.6 Bscf ± 108%
BACKGROUND:
Gathering field pipelines transport the gas from the production well to gas conditioning or processing
facilities. Leakage from gathering pipelines occurs from corrosion, joint and fitting failures, pipe wall
fractures, and external damage.
EMISSION FACTOR: (scf/Ieak-year)
The emission factors for leakage from gathering pipelines are based on the arithmetic average leakage rates
for main pipelines from the cooperative underground distribution leakage measurement program. A mean
value of the estimated leak rate per leak was calculated using the test data for all pipe materials except cast
iron. For east iron mains, a segment test approach was used which quantifies the leakage rate for a long
isolated segment of pipe; therefore, the mean leakage rate for cast iron is in terms of leakage per unit length
of pipe. The natural gas leak rate is adjusted for methane by multiplying by the volume percent of methane
for production (78.8 vol. %), and is adjusted for the soil oxidation of methane. The value of the emission
factor and standard deviation for each pipe material category is given below:
Pipe Material
Number of
Samples
Average
Emission
Factor
Units of
Emission
Factor
90%
Confidence
Interval of
Emission
Factor
Protected Steel
17
17,102
scf/leak-yr
14,548
Unprotected
Steel
20
43,705
scf71eak-yr
40,675
Plastic
6
84,237
scf/leak-yr
139,729
Cast Iron
21
201,418
sef'mile-yr
128,290
EMISSION FACTOR DATA SOURCES:
1. Leakage rate data on a rate per leak basis for cathodically protected steel mains, unprotected
steel mains, and plastic mains from the cooperative leak measurement program.
2. Leakage rate data on a rate per unit length basis for cast iron mains from the cooperative
leak measurement program for distribution mains.
3. Assumes that the leak rates from gathering lines are identical to leak rates from distribution
mains.
A-17
-------
ACTIVITY FACTOR:
The estimated number of leaks in field gathering pipelines is based on a leak repair frequency for gathering
lines owned and operated by transmission companies reported in the 1991 DOT RSPA database (1). This
database reports an estimated 8,153 repaired leaks and 270 outstanding leaks in 31,918 miles of gathering
pipeline. The leak frequency is derived by compensating for leaks that are repaired during the year and,
therefore, not contributing to leakage year round. On average, the repaired leaks are assumed to be leaking
for half the year, and each leak repair is counted as half an equivalent leak. Outstanding and unreported leaks
are assumed to be leaking the entire year.
Most production lines owned and operated by production companies are not regulated by DOT and many are
not monitored for leaks in the rigorous fashion employed by distribution and transmission companies.
Therefore, unreported leaks are accounted for based on the effectiveness of the survey method performed,
which is estimated to find 35% and 85% of the total leaks for a vegetation and walking survey, respectively,
based on one contract company specializing in distribution surveys. It is estimated that production company
owned gathering lines are only surveyed using a vegetation method. However, transmission company owned
gathering lines are estimated to be surveyed annually using a walking method, based on conversations with
several transmission companies.
Based on this analysis of equivalent leaks, the leak frequency is 0.18 leaks per mile for a walking survey and
0.63 leaks per mile for a vegetation survey. This leak frequency was used to ratio the number of leaks to the
total estimated population of gathering pipeline.
Total gathering pipeline mileage is not reported or tracked nationally and must be estimated. The "gathering
pipeline" designation includes three categories of pipeline: 1) production company owned gathering pipeline
for gas wells not associated with oil production (i.e., non-associated gas wells); 2) production company owned
gathering pipeline for oil wells that produce marketed gas (i.e., associated gas wells); and 3) transmission
company owned gathering pipeline. The third category of utility-owned pipelines are assumed to be in
addition to the production pipeline miles associated with wells. This is consistent with the site visit data
since gathering lines owned by transmission companies were intentionally excluded from the site mileage
totals. (The production companies did not report pipeline miles beyond their custody transfer meters.)
Total miles of gathering pipeline for non-associated gas wells were estimated using site visit data from the
thirteen production sites shown in the following table. Seven of the thirteen sites provided estimates of their
total miles of pipeline. The fifth site's mileage was estimated from a map of its pipelines.
A-18
-------
Site
Gathering Miles
Number of Wells
Miles per Total
Wells
Site 1
46.3
80
Site 2
8
26
Site 3
40
130
Site 4
15.4
12
Site 5
11
6
Site 6
5.2
193a
Site 7
600
1000
Site 8
441.3
425
Site 9
0.7
1
Site 10
27.7
24
Site 11
2.1
3
Site 12
7.1
7
Site 13
154.2
126
TOTAL
1359.0
2033
0.67 +/- 28%
"Includes 55 oil wells.
The estimate of total gathering miles per non-associated gas well was derived as the weighted average total
miles divided by total wells (0.67 ± 28%). The average mile per well ratio was extrapolated by the nationally
tracked number of non-associated gas wells (276,000). The resulting estimate of national gathering pipeline
miles associated with gas wells is 184,000.
For the gathering pipeline mileage associated with oil wells that market gas, the same ratio of gathering miles
per well was applied. However, it was assumed that only half of the gathering pipeline mileage was
attributed to the gas industry; the other half was attributed to the oil industry. Therefore, the average ratio of
pipeline miles to oil wells marketing gas was estimated to be 0.33. This average ratio was extrapolated by
the estimated number of oil wells marketing gas in the U.S. (209,000). The resulting estimate of gathering
pipeline mileage associated with oil wells that market gas is 70,000.
The third category of gathering pipeline owned by transmission companies is reported by the American Gas
Association (A.G.A.) (2) to be 86,200 miles. Utility-owned pipelines were assumed to be included in the
total production owned gathering pipeline miles and are not included in the transmission company owned
gathering line mileage.
The resulting total national gathering pipeline mileage from gas wells, oil wells marketing gas, and
transmission companies was estimated to be 340,200 miles. A rigorous determination of the 90% confidence
interval gave an error less that 4%. which was considered low based on the quality of the data used to
generate the activity factor. Thus, a 90% confidence interval of ± 10% was assumed based on engineering
judgement.
A-19
-------
Based on the analysis resulting in a leak frequency of 0.18 leaks per mile for transmission-owned gathering
lines employing a walking survey, and 0.63 leaks per mile for production-owned gathering lines employing a
vegetation survey, the activity factor can be calculated as follows:
[(86,200 x 0.18)] + [(340,200 - 86,200) x 0.63] = 174,779 equivalent leaks/year
The breakdown of total equivalent leaks by pipe material category is based on the breakdown of pipe mileage
reported in the 1991 DOT RSPA database (1) for transmission-owned gathering lines. It was estimated that
production-owned gathering line mileage is equivalent to the transmission-owned pipelines, with the exception
of cast iron. It was assumed that no additional cast iron gathering lines are in service. (That is, the cast iron
gathering line mileage reported in the RSPA database accounts for the total in the United States.)
The total number of estimated gathering line leaks was allocated on a pipeline material category basis in the
same proportion (adjusted for the fraction of mileage in each material category) as in the distribution sector.
The precision of the estimated total leaks was calculated based on the estimated 90% confidence interval
associated with each parameter in the activity factor equation:
repaired leaks; outstanding leaks: ± 100%
leak duration: ± 25%
leak survey effectiveness: ± 15%
A statistical software program [(@ RISK (3)] was used to determine the overall 90% confidence interval of
the activity factor: ± 76%.
For cast iron gathering lines, the mileage is based on the 1991 DOT RSPA database for transmission and
gathering lines. The precision of the cast iron mileage estimate is assumed to be ± 10%. The following table
summarizes the estimated average activity factor and the precision:
Pipe Material
Total Miles
Average Activity
Factor
Units of Activity
Factor
90% Confidence
Interval of
Activity Factor
Protected Steel
268,082
53,657
equivalent leaks
40,779
Unprotected Steel
41,400
114,655
equivalent leaks
87,138
Plastic
29,862
6,467
equivalent leaks
4,915
Cast Iron
856
856
miles
86
ACTIVITY FACTOR DATA SOURCES:
1. Leak repair frequency from [(DOT RSPA (1)] gathering line data.
2. Leak survey effectiveness provided by Southern Cross Company (4).
3. The gathering miles for gas and oil wells marketing gas was estimated using Phase 3 site
visit data for seven production companies. The number of gas and oil wells for these
companies was also used to extrapolate out to the national estimate.
4. The number of producing gas wells in the United States was taken from A.G.A. Gas Facts
(2) for 1992.
5. The number of oil wells producing marketed gas in the United States was estimated by
Radian (5). See activity factor section and sheet P-2.
A-20
-------
6. The field and gathering miles owned by transmission companies was taken from A.G.A. Gas
Facts (2) for 1992.
ANNUAL EMISSIONS: (6.6 Bscf ± 108%)
The activity factor was multiplied by the emission factor to derive this total leakage rate. The 90%
confidence intervals were propagated through this multiplication.
Pipe Material
Average Emission
Factor
(scCleak-yr)
Average Activity
Factor
(equivalent leaks)
Annual Emissions
Estimate,
(Bscf)
90% Confidence
Interval of
Leakage Estimate
(Bscf)
Protected Steel
17,102
53,657
0.9
1.2
Unprotected Steel
43,705
114,655
5.0
7.0
Plastic
84,237
6,467
0.6
1.2
Cast Iron
201,418"
856b
0.2
0.1
Total
6.6
7.2
ascf/mile-yr.
bmiles.
REFERENCES
1. U.S. Department of Transportation, Research and Special Programs Administration. 1991.
2. American Gas Association, Gas Facts, Arlington, VA, 1992.
3. Palisade Corporation. @ Risk, Risk Analysis and Simulation Add-In for Lotus 1-2-3, Version 1.5,
March 1989.
4. Southern Cross Corporation. Comments on Docket PS-123, Notice 1, Leakage Surveys, 49 CFR Part
192, Department of Transportation, Research and Special Programs Administration, Materials
Transportation Bureau, Office of Pipeline Safety Regulations, December 19, 1991.
5. 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.
A-21
-------
P-4
PRODUCTION SOURCE SHEET
SOURCES: Various Equipment
(wells, heaters, separators, dehydrators, compressors)
COMPONENTS: Pneumatic Devices
OPERATING MODE: Normal Operation
EMISSION TYPE: Unsteady, Vented
ANNUAL EMISSIONS: 31.4 Bscf ± 65 %
BACKGROUND:
Most of the pneumatic devices in the industry are valve actuators and controllers that use natural gas pressure as
the force for valve movement. There is a large population of pneumatic devices throughout the gas industry.
Gas from the valve actuator is vented to the atmosphere during every valve stroke, and gas may be continuously
bled from the valve controller pilot as well.
EMISSION FACTOR: 125,925 scf per average device ± 40%
(This was adjusted for the production methane fraction of natural gas at 78.8 mol%.)
Pneumatic devices (valve controllers) linked to control valves are the largest source of pneumatic emissions in
the production segment. There are two types of devices with distinct bleed modes: intermittent and continuous.
Intermittent bleed devices emit methane to the atmosphere only when the control valve actuates; when the device
is not moving the bleed rate is zero. Continuous bleed devices emit methane both when the valve actuates and
when the device is not moving. An emission rate for a generic pneumatic device combines the bleed rates of the
two types of devices, weighted by the population of the device types as follows:
EFavg pneum. device (Fraction intermittent ^ EF intermittent Fraction continuous ^ ^^"continuous)
x % methane
where:
Fraction in^^tten, = 0.65 ± 43%
Fraction= 0.35 ± 43%
% Methane = 78.8 mol % + 5%
Emissions for intermittent and continuous bleed devices were based on measured data provided by a Canadian
study and U.S. field measurements from a separate contractor's program. The average measured emissions for
intermittent and continuous bleed devices are 323 ± 34% and 654 ±31% scfd/device, respectively. The
fraction of each type of device was determined from site visits.
Therefore the average annual emission factor for a generic pneumatic device is:
EF avg. pneuma[ic device = 125,925 ± 40% scf/device
EF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 12: Pneumatic Devices (1)
establishes the important emission-affecting characteristics.
2. Site visit device counts establish the fraction of continuous bleed versus intermittent bleed
A-22
-------
devices for multiple sites.
3. The Canadian Producers Association (CPA) determined an average emission factor per device
based on 19 measurements.
4. An independent contractor provided 18 measurements of pneumatic devices in onshore and
offshore production services.
EF PRECISION:
Basis:
EF accuracy is based on rigorous error propagation from the spread of site device counts and
measured emission rates.
ACTIVITY FACTOR: 249,111 pneumatic controllers ± 48 %
The average count of devices per equipment type was determined from multiple site visits. The ratios for the
number of devices per gas well and the number of devices per marketed gas production were compiled by
region. The regional values were summed to give national device counts based on well counts and marketed gas
production. These values were averaged to give the final national device count of 249,111.
AF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 5: Activity Factors (2) establishes
the methodology for extrapolating the site data to a national count.
2. Site visit device counts, well counts, and production rates establish the number of devices per
well and the number of devices per gas production.
3. Total regional gas well counts and 1992 marketed gas production rates are from A.G.A. Gas
Facts (3).
4. The oil wells that market gas were calculated by this report and World Oil (4). Total oil wells
for 1992 are reported as 602,197 by the Oil & Gas Journal (5). The active oil wells that
market gas are determined by multiplying the total national active wells by the fraction that
market gas. The fraction is determined from a Texas Railroad Commission database (6) on oil
leases and gas disposition from those leases; an analysis that shows the percent of oil leases
that market the associated gas in Texas is 34.7%.
AF PRECISION:
Basis:
1. The accuracy for the devices per well and devices per gas production rate are calculated from
the spread of site data collected for each region (a total of 36 sites).
2. The accuracy for wells that market gas are based on the spread of data from the Texas
Railroad Commission database.
ANNUAL METHANE EMISSIONS: 31.4 Bscf ± 65 %
The national annual emissions were determined by multiplying an emission factor for an average pneumatic
device by the population of devices in the production segment.
125,925 scf x 249,111 devices = 31 Bscf
A-23
-------
REFERENCES
1. 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.
2. 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.
3. American Gas Association. Gas Facts: 1993 Data, Arlington, VA, 1994.
4. Gulf Coast Publishing Company, World Oil, Annual Forecast/Review, Vol. 214, No. 2, February 1993.
5. Oil and Gas Journal. 1992 Worldwide Gas Processing Survey Database, 1993.
6. Texas Railroad Commission, P-l, P-2 Tapes, Radian files, Austin, TX, 1989.
A-24
-------
P-5
PRODUCTION SOURCE SHEET
SOURCES:
COMPONENTS:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Gas-driven chemical injection pumps use gas pressure acting on a piston to pump a chemical on the opposite
side of the piston. The gas is then vented directly to the atmosphere. The pumps are used to add chemicals
such as corrosion inhibitors, scale inhibitors, biocide, demulsifier, clarifier, and hydrate inhibitors to operating
equipment. Two types of pumps were observed: 1) piston pumps, and 2) diaphragm pumps. Some of the
pumps observed were inactive at the time or had seasonal operation.
Wells, Gathering Facilities
Chemical Injection Pumps
Normal Operation
Unsteady, Vented
1.5 Bscf + 203%
EMISSION FACTOR: 248 scfd/average pump ± 83 %
(This was adjusted for the production methane content in natural gas at 78.8 mol%.)
This average emission factor is based upon the following equation:
PP _ p w pp
E'avg. pump •'piston ^-^piston
+
F X FF
*¦ diaphragm ^ diaphragm
where:
F .
x piston
EFpislon
P
^diaphragm
^^diaphragm
fraction of the pump population that is the piston type = 49.8% ± 38%
emission factor of an average piston pump = 48.9 scfd/pump ± 106%
fraction of the pump population that is the diaphragm type = 50.2% + 38%
emission factor of an average diaphragm pump = 446 scfd/pump ±11%
The average device emission factor was determined by an aggregation of device emissions calculated for
multiple U.S. sites. For piston pumps, the emission factor was determined by the following equation:
PP
^-''piston
Gas usage (acf/stroke) x Density (scf/acf) x Frequency (strokes/day) x
Operating time x % methane
where:
Gas usage
Density
Frequency
Operating time
% methane
calculated gas usage based on piston diameter and stroke length (in actual ft3);
scf/acf at supply gas pressure (average 30 psig) (combined average value of
volume and density is 0.0037 ±65% scf/stroke);
strokes per day of the average pump (37,901 ± 29% strokes/day);
portion of time that the pump is operating (0.446 + 62%); and
78.8 mol% ± 5% for the production segment.
Based on site and manufacturer data, the resulting national piston pump emission factor is 48.9 scfd/pump ±
106%.
A-25
-------
For diaphragm chemical injection pumps, the emission factor was determined by the following equation:
where:
FF
diaphragm
Gas usage
Volume
Frequency
Operating time
% methane
Gas usage (scf/gal) X Volume (gal/stroke) x Frequency (strokes/day) x
Operating time x % methane
volume of gas (in standard ft3) required to pump one gallon of liquid chemical
(provided by the manufacturer);
liquid displaced per stroke based on the plunger diameter and stroke length
(combined average value of gas consumption and volume is 0.0719 + 10%
scf/stroke);
strokes per day of the average pump (19,642 ±49% strokes/day);
portion of time that the pump is operating (0.40 + 52%); and
78.8 mol% ± 5% for the production segment.
Using the site, manufacturer, and measured data to calculate the emission factor equation terms, the total
diaphragm pump emission factor was determined to be 446 scfd/pump ±11%.
Stroke volume was calculated from pump manufacturers' data and site observations of manufacturer and model
number. Density was calculated based upon observed site supply gas pressure, and frequency was based upon
timed stroke intervals observed while on site. Operating time was estimated by site personnel (if seasonal), or
was based upon the percent of pumps at the site that were operating during the visit. The emission factors
shown above (in scfd/pump) have been corrected for the natural gas composition in the production segment of
78.8 mol % methane.
EF DATA SOURCES:
1.
2.
3.
4.
5.
6.
The report entitled Methane Emissions from the Natural Gas Industry, Volume 13:
Chemical Injection Pumps (1) establishes the important emission-affecting
characteristics.
Site visit data and reference material established the density from supply gas pressure
at 30 psig.
For the piston pumps, the stroke volume was estimated from manufacturers' data of
pumps found at each site.
Manufacturers' data for the diaphragm pumps provided scf of gas required to pump
one gallon of chemical. This information was used with the calculated liquid
displaced for a range of pumps to give an average gas volume.
The frequency of actuations per day was determined from 40 timing measurements
taken at 12 sites. The operating time was determined from data at 13 sites.
Measurements of 5 diaphragm chemical injection pumps were provided from an
emissions estimate program by the Canadian Petroleum Association.
EF ACCURACY:
Basis:
1.
2.
3.
4.
Operating time confidence bounds (at 90% confidence) were calculated by analysis of
the spread of 7 sites for piston pumps and 10 sites for diaphragm pumps.
Actuation confidence bounds (at 90% confidence) were based on measurements from 7
sites for the piston pumps and 5 sites for the diaphragm pumps.
It was assumed that the manufacturers' data are completely accurate. Data for the
piston pumps were based on information from 4 manufacturers. Diaphragm pump
data were provided by 2 manufacturers.
90% confidence bounds for each value were carried through error propagation to
result in the final 90% confidence bound.
A-26
-------
ACTIVITY FACTOR: 16,971 pumps in the production segment + 143 %
The number of gas actuated pumps used in the production segment was determined by establishing the ratio of
the number of pumps to active wells (oil or gas) that market gas. Site data were organized into regions and
regional values were determined. The regional ratios were then multiplied by the regional count of active wells
that market gas in that region to produce the total count of chemical injection pumps in the region. Finally,
regions were added together to determine the national number. The activity factor is then:
n
(1) National AF = £ (Regional AF) where n = total number of regions
i = l
(2) Regional AF = (Rj's) x (W)
where Rj = ratio of total pumps to total wells in Region j
where W = number of wells in the region
AF DATA SOURCES:
1. The active oil and gas wells are from A.G.A. Gas Facts (2). The active oil wells that
market gas are determined by multiplying the total national active oil wells times the
fraction that market gas. The fraction is determined from a Texas Railroad
Commission lease study that shows the percent of oil leases that market the associated
gas in Texas (3).
2. The pump counts were obtained during the site visits. Inactive, electrically driven, or
air driven pumps were not counted.
3. Regional extrapolation by gas well count was used.
AF ACCURACY:
Basis:
1. The accuracy for the active gas wells is assigned by engineering judgement, based
upon the fact that the number of active wells is tracked nationally and known
accurately by A.G.A./DOE, etc.
2. The accuracy for the national AF is based upon error propagation from the production
sites visited.
ANNUAL METHANE EMISSIONS: 1.5 Bscf ± 203 %
The national annual emissions were determined by multiplying an emission factor for a typical pump by the
population of chemical injection pumps in the production segment.
REFERENCES
1. 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.
2. American Gas Association. Gas Facts: 1993 Data, Arlington, VA, 1994.
3. Texas Railroad Commission, P-l, P-2 Tapes, Radian files, Austin, TX, 1989.
A-27
-------
P-6
PRODUCTION SOURCE SHEET
SOURCES: Glycol Dehydrators
COMPONENTS: N/A
OPERATING MODE: Normal Operation
EMISSION TYPE: Vented
ANNUAL EMISSIONS: 3.42 Bscf ± 192%
BACKGROUND:
Glycol dehydrators remove water from a gas stream by contacting the gas with glycol and then driving the
water from the glycol by heating in the glycol reboiler and into the atmosphere. The glycol also absorbs a
small amount of methane, and some methane can be driven off to the atmosphere through the reboiler vent.
EMISSION FACTOR: (275.57 scf/MMscf gas processed ± 154.48%)
A thermodynamic computer simulation was used to determine the most important emission-affecting variables
for dehydrators. The variables are: gas throughput, existence of a flash tank, existence of stripping gas,
existence of a gas driven pump, and existence of vent controls routed to a burner. Throughput, since its
effect is linear, is handled by establishing an emission rate per unit of gas throughput. Emission rates per unit
of throughput are then established for the other important emission affecting characteristics. Gas driven
pumps are ignored here and handled in a separate source analysis (see Methane Emissions from the Natural
Gas Industry, Volume 15: Gas-Assisted Glycol Pumps) (1). The emission factor is then:
EF = [ ( Fpjr x EFjr,- ) + ( FOT x EF^ ) + ( FSG x EFSG ) ] x F^j- x OC
= [ (0.265 x 3.57) + (0.735 x 175.10) + (0.00473 x 670) ] x 0.9882 x 2.1
Fpp = Fraction of the population WITH flash tanks
0.265 ± 8.35%
Fot = Fraction of the population WITHOUT flash tanks
0.735 ± 2.99%
Fsg = Fraction of the population WITH stripping gas
0.00473 ± 115.78%
Fnvc = Fraction of the population WITHOUT combustion vent
controls
0.9882 ± 0.87%
EFp-,. = Total methane emission rate scf per 1 MMscf throughput
with a flash tank
3.57 -+-102%/-58%
EFnt = Total methane emission rate scf per 1 MMscf throughput
WITHOUT a flash tank
= 175.10 +101%/-50%
EFsg = Incremental methane emission rate per 1 MMscf
throughput per dehydrator that has stripping gas
= 670 +40%/-60%
OC = Overcirculation factor for glycol—number of times the
industry rule-of-thumb of 3 gallons glycol/lb water
= 2.1 ±41%
A-28
-------
EF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators
(2) establishes emission affecting characteristics of dehydrators.
2. GRI/EPA site visit data establishes the FSG and for multiple sites (19 PROD
sites).
3. An analysis of a combined database including TMOGA's 1019 dehydrators and
GRI/EPA site visits 444 dehydrators established Fpu and F^ for production dehydr-
ators.
4. ASPEN computer simulations were used in combination with measured data to
determine EF^, and EF^ from the dehydrator vent.
5. Sampling data from the GRI Glycol Sampling and Analytical Program for one
dehydrator was used to determine EFSG (Glycol Dehydrator Emissions: Sampling
and Analytical Methods and Estimation Techniques) (3). The upper bound was
calculated by assuming that all of the measured noncondensable vent gas was due to
stripping gas that was 100% methane. The lower bound was calculated as the rule-
of-thumb stripping gas rate recommended by a glycol dehydrator manufacturer.
6. Overcirculation factor determined using data from the GRI Glycol Sampling and
Analytical Program data for ten dehydrators.
EF PRECISION: 275.57 sctfMMscf gas processed ± 154.48%
Basis:
The accuracy is propagated through the EF calculation from each term's accuracy:
1. ASPEN has been demonstrated to match actual dehydrators within ±20% within the
calculated confidence intervals obtained from site data.
2. Individual EF confidence intervals were calculated from the data used in the
calculation.
3. Data from site visits has been assigned confidence intervals based upon the spread
of the 444 dehydrators from GRI/EPA site data.
ACTIVITY FACTOR: (12.4 Tscf/year gas throughput in the production segment)
The amount of gas processed by glycol dehydrators in the production segment was calculated from the
estimated number of glycol dehydrators in production and the average throughput capacity for production
dehydrators (Wright Killen and Co., 1994). A capacity utilization factor was estimated based on observations
at several sites in the GRI Glycol Dehydrator Sampling and Analytical Program.
AF DATA SOURCES:
The report: Natural Gas Dehydration: Status and Trends (4) by Wright Killen for the Gas Research Institute,
provides data and describes the methodology used to develop an estimate of the gas dehydrator count for the
U.S. The count also estimated the number in several industry segments: production, transmission, and gas
processing.
A-29
-------
Basis:
1. A GRI study by Wright Killen & Co. found 41700 dehydrators in the U.S. gas
industry for 1993. Wright Killen & Co. also used a TMOGA/GPA database on
dehydrators to split the population into the following industry segments:
Production: 25270
Processing: 7923
Transmission: 8507
TOTAL: 41700
The study also found that 95.0 % of the dehydrators were glycol for a total of
39,615 (versus molecular sieve or other types).
2. Site visit data on 24 transmission compressor stations shows: 2/17 = 0.118 per
transmission compressor station, and 17/6 = 2.83 per storage compressor station.
The site visit numbers would lead to an estimate of 1293 total transmission and
storage dehydrators. Site visit data on 11 gas plants show 1.41 dehydrators per
plant, or 1,024 in gas plants.
Subtracting processing, transmission, and storage glycol dehydrators from the total
of 39,615 yields 37824 glycol dehydrators in production.
3. Average capacity of production dehydrators was reported to be 2 MMscfd by Wright
Killen & Co.
Information on actual dehydrator throughput as compared to design capacity is, in general, difficult to obtain
especially for production field units. Data from several sites in the GRI Glycol Dehydrator Sampling and
Analytical Program and other anecdotal information from various site visits indicate that capacity utilization
may be less than 50%, so a value of 45% was chosen for the AF calculations.
AF PRECISION: 12.4 Tscf/year ± 61.87%
Basis:
The 90% confidence limits for total glycol dehydrators were established in the Wright Killen & Co.
report. The confidence limits for the segments other than production were based on site visit data.
Confidence limits for the capacity utilization was based on engineering judgement.
ANNUAL METHANE EMISSIONS: (3.4171 BscCyr ± 191.90%)
The annual methane emissions were determined by multiplying the dehydrator emission factor by the activity
factor.
REFERENCES
1. Myers, D.B. and M.R. Harrison. Methane Emissions from the Natural Gas Industry, Volume 15:
Gas-Assisted Glycol Pumps, Final Report, GRI-94/0257.43 and EPA-600/R-96-080o. Gas Research
Institute and U.S. Environmental Protection Agency, June 1996.
2. 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.
A-30
-------
Radian Corporation. Glycol Dehydrator Emissions: Sampling and Analytical Methods and
Estimation Techniques. GRI-94/0324, Gas Research Institute, Chicago, IL, March 1995.
Wright Killen & Company. Natural Gas Dehydration: Status and Trends, Final Report. GRI-
94/0099, Gas Research Institute, Chicago, IL, October 1993.
A-31
-------
P-7
PRODUCTION SOURCE SHEET
SOURCES:
COMPONENTS:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
Dehydrators
Gas Driven Kimray Pumps
Normal Operation
Unsteady, Vented
10.96 Bscf ± 110.0%
BACKGROUND:
Gas driven Kimray glycol circulation pumps use a mixed phase of wet glycol liquid and absorber gas to drive
pistons that pump dry (lean) glycol circulation. Unlike chemical injection pumps which vent the driving gas
directly to the atmosphere, Kimray pumps pass the driving gas along with the wet glycol to the reboiler. In
the reboiler the methane is driven off into the vent line. Depending on dehydrator vent gas dispositions, the
methane may be vented to the atmosphere or controlled and burned.
EMISSION FACTOR: (992.0 scf CH4/MMscf gas processed)
The average glycol pump gas emission factor was determined by an equation describing the gas generation
and disposition of gas from the pump. The disposition of gas generated by the pump depends upon the
existence of a flash tank and vent controls. Measured and estimated parameters were input into the equation.
In general, the emission factor for a gas-assisted pump was determined by the following equation:
EFp„mp PGU x CR x WR x OC
EF DATA SOURCES:
1. Equation 1, i.e. the effects of operating variables on emissions, was defined by the report on Methane
Emissions from the Natural Gas Industry, Volume 15, Gas-Assisted Glycol Pumps.(I)
2. CR = glycol circulation ratio = 3.0 gal glycol/lb water ± 33.3%.
3. WR = water removed from gas
= 53 Ib/MMscf ± 20% for high pressure
= 127 lb/MMscf ± 20% for low pressure
4. OC = factor to account for overcirculation of glycol = 2.1 ± 71.4%.
5. Fnt = fraction of dehydrators without flash tanks = 0.735 ± 2.99%.
6- Fjjvc = fraction of the dehydrators without combustion vent controls = 0.9882 ± 0.87%.
7. PGU = pump gas usage (assume 83% methane)
= 3.73 scf CH4/gal glycol ± 30% for high pressure
A-32
-------
= 2.31 scf CH4/gaI glycol ± 30% for low pressure
CALCULATION METHOD:
It is estimated that 80% of the production dehydrators would be high pressure (R. Garrett memo). The
overall production emission factor is then calculated as a weighted average of the high and low pressure
emission factors.
EF (high pressure) = (3.73 scf/gal) x (3.0 gal/lb H-,0) x (53 lb H20/MMscf)
x (2.1) x (0.735) x (0.9882)
= 904.45 scf/MMscf ± 95.04%
EF (low pressure) = (2.31 scf/gal) x (3.0 gal/lb H20) x (127 lb H20/MMscf)
x (2.1) x (0.735) x (0.9882)
= 1342.18 scf/MMscf ± 95.04%
EF (Production) = (0.80 ± 12.5%) (904.45 scf/MMscf ± 95.04%) +
(0.20 ± 50%) (1342.18 scf/MMscf ± 95.04%)
992.00 scf CH„/MMscf ± 77.29%
EF ACCURACY: (± 77.29%)
Basis:
1. Assumption: The manufacturer's data and ranges are relatively accurate (±30%).
2. Dehydrator characteristics based on site visit observations and TMOGA survey.
ACTIVITY FACTOR: (11.05 Tscf/year in the production segment with gas-assisted pumps)
The volume of gas processed through dehydrators using gas-assisted pumps was calculated from the total
throughput for production dehydrators and the fraction of dehydrators using gas-assisted pumps determined
from site visits. The activity factor is then:
AF = (fraction of dehydrators with gas-assisted pumps) x (throughput for production dehydrators)
= (0.8913 ± 2.79%) x (12.4 Tsctfyear ± 48.21%)
= 11.05 Tscf/year ± 61.96%
AF DATA SOURCES:
1. See Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators (2) for
an explanation of production dehydrator throughput. See the Methane Emissions from the
Natural Gas Industry, Volume 5: Activity Factors (3) for more details.
2. Fraction of dehydrators using gas-assisted pumps came from data from site visits.
AF ACCURACY: (± 61.96%)
Basis:
Calculated from confidence limits of gas throughput and fraction of dehydrators by standard error
propagation analysis.
A-33
-------
ANNUAL METHANE EMISSIONS: (10.962 Bscf ± 110.03%)
The annual methane emissions were determined by multiplying an emission factor (scf CH4/MMscf) by the
total throughput for production dehydrators using gas-assisted pumps.
(992.00 scfTMMscf) x (11.05 Tscf) = 10.962 Bscf (± 110.03%)
REFERENCES
1. 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.
2. 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.
3. 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.
A-34
-------
P-8
PRODUCTION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Maintenance activities can emit gas to the atmosphere through blowdown or through purge. Blowdown is the
direct, intentional venting to the atmosphere of gas contained inside operating equipment. The gas is released
to provide a safer working environment for maintenance activities around or inside the equipment. After the
equipment is serviced, the oxygen inside the equipment is often cleared to the atmosphere by purging natural
gas through the equipment.
Another type of maintenance venting is associated with low pressure gas wells that sometimes accumulate
water in the wellbore due to their low flow rate. This water chokes the flow of the well, reducing gas
production. To clear the water, the well is blown to a tank at atmospheric pressure where the gas is vented.
EMISSION FACTORS: Gas Well Unloading 49,570 ± 344% scf/unloading gas well
Compressor Blowdown 3,774 ± 147% scf/compressor
Compressor Starts 8,443 ± 157% scf/compressor
Pipeline Blowdown 309 + 32% scf/mile
Vessel Blowdown 78 ± 266% scf/vessel
(Emission factors were adjusted for the production methane fraction of
natural gas of 78.8 mol%)
Blowdown volumes and frequencies were averaged from calculations for each GRI/EPA site visit. The
volume times the frequency results in the annual emissions. The volumes were calculated at each site using
equations of state, observed vessel dimensions, and pre-blowdown pressures. Frequencies were gathered at
each site from operator interview. The annual emission factor (scf/unit) for each category was calculated as
follows:
EF = Volume x Frequency x % Methane
where:
Volume = Gas released to the atmosphere during an event (scf/event/unit);
Frequency = Number of events annually;
% Methane = 78.8 mol % ± 5% for the production segment.
More details are available in the Methane Emissions from the Natural Gas Industry, Volume 7: Blow and
Purge Activities (1).
EF DATA SOURCES:
1. The blow and purge report establishes emission affecting characteristics of blowdown practices.
Various Production Equipment
(wells, vessels, compressors, pipelines)
Maintenance
Unsteady, Vented
6.0 Bscf ± 359%
A-35
-------
2. Volume and frequency data were available from the following number of sites:
LP Gas Well Unloading (12 sites)
Compressor Blowdown (17 sites)
Compressor Starts (12 sites)
Vessel BD (12 sites)
Pipeline BD (18 sites)
3. The count of equipment at each site was gathered during the site visits by observation, record
search, or interview.
EF PRECISION: ± 32% to 344%
Basis:
The accuracy was calculated from the variance of the site data. A 90% confidence interval is
calculated for the sites using the method outlined in the Methane Emissions from the Natural Gas
Industry, Volume 4: Statistical Methodology (2).
ACTIVITY FACTORS:
114,139 ± 45% gas wells requiring unloading
17,112 + 52% compressors
340,000 ± 10% miles of pipeline
255,996 + 26% production vessels
The activity factors for equipment in the production segment were compiled from GRI/EPA site visit averages
as well as published statistics on the gas industry (see activity factor sections in previous sheets). The
number of production vessels was assumed to be the sum of separators, heaters, and dehydrators.
AF DATA SOURCES:
1. The well, compressor, and vessel counts came from the activity factor extrapolation based on
GRI/EPA site visits or surveys (previously discussed in the production fugitives sheet). The
count of "vessels" is from the addition of dehydrator, separator, and in-line heater counts.
2. The miles of production gathering pipelines were determined from a site extrapolation of seven
sites and data from Gas Facts Table 5-3 (3). This extrapolation was previously discussed in the
production gathering pipeline fugitive leaks sheet,
P-3.
3. The number of gas wells requiring unloading is based on the ratio of gas wells requiring
unloading to all active gas wells from 25 GRI/EPA sites (41.4% ± 162%).
AF PRECISION: Range ± 10% to 52%
Basis:
The accuracy for all equipment types is based on error propagation from the spread of available
production site data.
ANNUAL METHANE EMISSIONS: 6.0 Bscf ± 359%
The annual methane emissions were determined by multiplying an emission factor (rate per average unit) for
each category by the activity factor (population) of the category.
A-36
-------
REFERENCES
1. 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.
2. 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.
3. American Gas Association. Gas Facts, 1992 Data (Table 5-3), Arlington, VA, 1993.
A-37
-------
P-9
PRODUCTION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Upsets in process conditions can cause pressure rises that exceed the maximum design pressure for
equipment. To prevent equipment overpressure and damage, pressure relief valves (PRVs) open and vent the
excess gas to the atmosphere. These PRVs are spring loaded or pilot actuated valves that are designed to
handle the upset conditions. A few offshore production facilities (but no onshore facilities) have Emergency
Shutdown Systems (ESDs) that depressure the entire facility to a vent or a flare.
EMISSION FACTORS: PRV Discharge Blowdown 34 ± 252% scf/PRV
ESD Blowdown 257 ± 200% Mscf/platform
(Corrected for the production methane composition of 78.8 mol%)
Emergency blowdown volumes and frequencies were estimated at each site visited. The average volume of
gas released at lift pressure was calculated for a typical PRV size and duration, and corrected for the fraction
of PRVs that release gas to the atmosphere. ESD blowdown volumes were based on the platform volume and
corrected for the fraction of platforms with ESDs and the fraction that vent gas to the atmosphere. The
annual emission factor (scf/unit) for each category was calculated as follows:
EF = Volume x Frequency x % Methane
where:
Volume = Gas released to the atmosphere during an event (scf/event/unit);
Frequency = Number of events per year;
% Methane = 78.8 mol % ± 5% for the production segment.
EF DATA SOURCES:
1. The GRI/EPA Methane Emissions from the Natural Gas Industry, Volume 7: Blow and Purge
Activities (1) establishes emission affecting characteristics of blowdown practices.
2. Volumes (duration, release rate, % to atmosphere) and frequencies were calculated from each site
visit based on data collection, observation, and interview. Data were available from the
following number of sites:
PRV discharge (11 sites)
ESD activation (5 platforms)
3. The count of equipment at each site was gathered during the site visits by observation, record
search, or interview.
EF PRECISION:
Basis:
The accuracy was propagated from the spread of the site data. A 90% confidence interval is calculated
using the method presented in the Methane Emissions from the Natural Gas Industry, Volume 4:
Statistical Methodology (2).
Various Production Equipment (vessels)
Upsets
Unsteady, Vented
0.3 Bscf ± 190%
A-38
-------
ACTIVITY FACTORS: 529,440 ± 53% Production PRVs
1,115 + 10% Platforms
The activity factors for equipment types in the segment were compiled from GRI/EPA site visit data as well
as published statistics on the gas industry.
AF DATA SOURCES:
1. The count of platforms is from Offshore Data Services and the Minerals Management System
Outer Continental Activity Database as reported in Methane Emissions from the Natural Gas
Industry, Volume 5: Activity Factors (3).
2. The number of production PRVs is based on counts of PRVs per equipment type from site visit
data:
Equipment PRV Number
Type Count of Sites
Separators 2 ± 68% 20
Heaters 1 ± 89% 11
Dehydrators 2 ± 53% 10
Compressors 4 ± 84% 13
Details are provided in the Methane Emissions from the Natural Gas Industry, Volume 8: Equipment Leaks
report (4).
AF PRECISION: Range ± 10% to 53%
Basis:
1. Confidence intervals for the platform count were assumed and assigned based upon an excellent
recorded source of data [see Methane Emissions from the Natural Gas Industry, Volume 5:
Activity Factors (3)].
2. Ninety percent confidence limits for production vessels with PRVs were calculated from the
confidence intervals of each type of equipment. See Methane Emissions from the Natural Gas
Industry, Volume 5: Activity Factors (3) for details of equipment count determination.
ANNUAL METHANE EMISSIONS: 0.30 Bscf ± 190%
The annual methane emissions were determined by multiplying an emission factor (rate per avg unit) by the
activity factor (population) of the category. Each emission factor was adjusted for the average methane
content in the production segment of 78.8 mol%.
REFERENCES
1. 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.
2. 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.
A-39
-------
3. Stopper, 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.
4. 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.
A-40
-------
P-10
PRODUCTION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Dig-ins are gathering pipeline ruptures caused by unintentional (sometimes third-party) damage. Production
companies do NOT estimate and record the quantity of gas lost during a dig-in event; therefore, distribution
data has been used.
EMISSION FACTOR: 669 ± 1,925% scCmile
(Corrected for the production methane composition of 78.8 mol%)
The emission factor was derived from four distribution company estimates of the losses from dig-ins: the
Pacific Gas and Electric unaccounted-for (UAF) gas study (1) results showed that losses from dig-ins were
estimated at 91,178 Mscf for 58,024 miles of distribution mains and services; the Southern California Gas
Company estimate (2) of losses from dig-ins was 170,457 Mscf for 82,337 miles of distribution mains and
services; a third company estimate of losses from dig-ins was 19,581 Mscf for 24,916 miles of distribution
mains and services; and a fourth company reported dig-in losses of 10,453 Mscf for 18,713 miles of
distribution mains. The ratio of the total dig-in emissions to the total pipeline miles from these companies
was used to estimate the annual national methane emission factor, resulting in 2.06 Mscf/mile.
This value was halved (and adjusted for the different methane compositions of the two industry segments)
based upon an engineering assumption that production dig-ins occur much less frequently than distribution
dig-ins, and so account for approximately one-half of the distribution emission rate per mile. This is
supported by the fact that most production sites are remotely located, while distribution sites are by definition
located in population centers where third-party dig-ins are more likely.
ACTIVITY FACTOR: 340,000 + 10% miles of production gathering pipeline
The annual number of miles of gathering pipeline in the U.S. gas industry was derived from site data. See P-
3 and Methane Emissions from the Natural Gas Industry, Volume 5: Activity Factors (3) for more details.
ANNUAL METHANE EMISSIONS: 0.23 Bscf ± 1,934%
REFERENCES
1. Pacific Gas & Electric Company and Gas Research Institute. Unaccounted-For Gas Project. Volume
1, Final Report, San Ramon, CA, June 1990.
2. Southern California Gas Company and Gas Research Institute. A Study of the 1991 Unaccounted-For
Gas Volume at the Southern California Gas Company, Final Report, Los Angeles, CA, April 1993.
3. 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.
Pipeline
Mishaps (Dig-ins)
Unsteady, Fugitive
0.2 Bscf ± 1,934%
A-41
-------
p-11
PRODUCTION SOURCE SHEET
SOURCES: Gas Wells
OPERATING MODE: Maintenance
EMISSION TYPE: Venting and Flaring
ANNUAL EMISSIONS: 0.02 Bscf ± 1,263%
BACKGROUND:
Two minor sources of maintenance releases are completion flaring and well workover. Completion flaring
occurs at a new well's open ended pipe flare immediately following the drilling process. During completion
testing, the gas is flared to determine the available pressure and flow rates at the surface. This allows proper
sizing of meters and surface equipment. Most completion flaring occurs at exploratory wells, since the
production rates and needed facilities for in-fill wells (also called development wells) are often available or
can be determined before the well is completed.
Well workovers are another type of maintenance venting. During a well workover, the tubing is pulled from
the well to repair tubing corrosion/erosion or other downhole equipment problems. The well is "killed" by
replacing the gas in the column with water or mud, thus stopping all production flow. The well can then be
opened to the atmosphere.
EMISSION FACTORS: Completion Flaring 733 ± 200% scf7completion well
Well Workovers 2,454 ± 459% scf/well workover
(Emission factors were adjusted for the production methane fraction of natural gas of 78.8 moI%.)
The flow rate of gas at completion is the highest that the well will produce. For emission estimate purposes,
the maximum gas flow rate was not available. Instead, the completion flaring emission factor was calculated
based on the average annual natural gas production per well and an assumed flaring efficiency as shown:
^^"completion flaring = Average Annual Volume x Duration x % Methane x Flaring Efficiency
where:
Average Annual Volume = 16.97 MMscf for natural gas
Duration = Flaring duration is one day/completion well
% Methane = 78.8 mol% for production
Flaring Efficiency = 98% efficient (2% methane not burned)
This results in an emission factor of 733 ± 200% scf/completion well for completion flaring.
The emission factor for well workovers was determined from two gas production fields. Data from these
fields are shown in the following table:
Site 1 Site 2
Total number of wells 21 400
Number of workovers/year 1 8
Methane emissions/workover, scf/workover 670 4,238
Average scf methane/workover 2,454 ± 459%
A-42
-------
EF DATA SOURCES:
1. One operator provided data on the typical duration of completion flaring and which types of
completions were flared. Average is one day/exploratory completion well.
2. Average gas production per well from Gas Facts (1).
3. Multiple reports on methane flare combustion efficiency support 98% combustion.
4. Pipeline Systems Incorporated (PSI) reported gas well workover emissions from two sites (2).
EF PRECISION: ± 200% to 459%
1. Engineering judgement was used to establish the upper confidence limit for the completion
flaring emission factor.
2. Confidence bound for well workover emission factor is based on the average of data from two
sites.
ANNUAL ACTIVITY FACTORS:
844 ± 10% completed gas wells
9,329 + 258% well workovers
AF DATA SOURCES:
1. Number of exploratory wells completed per year based on data from the Energy Information
Administration (EIA) Drilling and Production under Title I of the Natural Gas Policy Act (3).
This data excludes Alaska.
2. PSI data showed 1 workover/yr per 21 wells at Site 1 and 1 workover/yr per 50 wells at Site 2.
3. The Activity Factors Report (4) provides details on the total number of gas producing wells
(276,014 ± 5%).
AF PRECISION: Range ± 10% to 258%
1. 10% upper confidence bound for completion wells is assigned based on good precision from
national statistics of 1987 data.
2. Well workover confidence interval is based on the average of data from two sites combined with
the confidence bound for the total number of gas producing wells.
ANNUAL METHANE EMISSIONS: Completion Wells: 619 ± 201% Mscf
Well Workovers: 22.9 ± 1296% MMscf
The annual methane emissions were determined by multiplying an emission factor (methane emissions per
event) for each category by the activity factor (events/year) of the category.
REFERENCES
1. American Gas Association. Gas Facts: 1992 Data (Table 3-3), Arlington, VA, 1993.
2. Pipeline Systems Incorporated. Annual Methane Emission Estimate of the Natural Gas Systems in the
United States, Phase 2. For Radian Corporation, September 1990.
3. Energy Information Administration. Annual Energy Review 1994, Table 4.5 "Oil and Gas Exploratory
Wells, 1949-1994." EIA, Office of Oil and Gas, U.S. Department of Energy, DOE/EIA-0384(94),
Washington, DC, July 1995.
4. 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.
A-43
-------
APPENDIX B
Gas Processing Source Sheets
B-l
-------
APPENDIX B
Gas Processing Source Sheets
Page
GP-1 - Fugitive Emissions B-4
GP-2 - Glycol Dehydrator Vents B-9
GP-3 - Acid Gas Removal (AGR) Vents B-ll
GP-4 - Maintenance Venting B-13
GP-5 - Glycol Pumps B-15
GP-6 - Pneumatic Devices B-18
B-2
-------
PROCESSING SOURCE SHEETS
This section contains the specific source sheets for the processing (gas plant) segment of
the natural gas industry. The following table serves as a guide for finding sheets in this
section. The cells in the table give the sheet number (GP-1, GP-2, etc.) of the source sheet.
The rows define the equipment covered, while the columns define the emission type. A
category with no sheet number means that the emissions from that area were determined to
be zero or negligibly small. The label for each source sheet is shown at the top of the
cover page for that sheet.
TABLE OF
CONTENTS
OPERATING MODE,
EMISSION TYPE (Fugitive, Vented, or Combusted)
EQUIPMENT:
Start Up
Normal Operations
Maintenance
Upsets
Mishaps
V
C
F
V
C
V
C
V
C
V
Entire Plants
GP-1
GP-6
GP-4
GP-4
Vessels
GP-1
GP-6
GP-4
GP-4
Acid Gas
Removal (AGR)
Units
GP-1
GP-3
GP-6
GP-4
GP-4
Dehydrators
GP-1
GP-2
GP-5
GP-6
GP-4
GP-4
Compressors
GP-1
GP-6
P-l
GP-4
GP-4
Metering
GP-1
GP-4
GP-4
B-3
-------
GP-1
PROCESSING SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Equipment leaks are typically low-level, unintentional losses of process fluid (gas or liquid) from the sealed
surfaces of above-ground process equipment. Equipment components that tend to leak include valves, flanges
and other connectors, pump seals, compressor seals, pressure relief valves, open-ended lines, and sampling
connections. These components represent mechanical joints, seals, and rotating surfaces, which in time tend
to wear and develop leaks.
EMISSION FACTOR:
a. Plant = 2.89 MMscCyr methane per plant
b. Reciprocating Compressor = 4.09 MMscl/yr methane per recip
c. Centrifugal Compressor = 7.75 MMscf/yr methane per turbine
The average fugitive emission rate for gas processing plants was determined to be composed of two parts:
a) plant component counts (excluding compressor components), and b) compressor-related components.
Fugitives from the compressor-related components have much higher emission factors than components in the
rest of the facility. Part of this is due to the high vibration that compressors generate, but most of the larger
emissions are due to unique compressor components, as explained below.
a. The contribution from non-compressor components was determined by multiplying the average component
count by the component emission factor. The number of components was subdivided into valves,
connections/flanges, small open-ended lines, site blowdown (B/D) OELs, control valves, and other
components (such as pressure relief valves). (Tubing components were determined to be insignificant.) All
of these components are typical fugitive components [as described in the EPA Fugitive Emissions Protocol
(1)] with the exception of control valves and site B/D OELs. Control valves emit at a higher rate than
manual isolation valves since their packing is stressed more often as they are activated much more frequently.
Site B/D OELs are the large diameter emergency station blowdown valves that are designed to depressure the
entire site to the atmosphere when the valve is opened.
The component emission factors for gas plant components (i.e., non-compressor related) were based on an
API/GRI measurement program conducted at 8 gas plants. The average facility emissions are then calculated
as follows:
EF = [(Nvlv x EFvlv) + (Ncn x EFJ + (Noe, x EFoeI) + (Nprv x EFprv) + (NsitcB/D x EFsitcB/D)]
where:
Nx = average count of components of type x per plant, and
EFX = average methane emission rate per component of type x.
b. The contribution from compressor-related components was obtained by multiplying the average number of
fugitive components per compressor engine by component emission factors. The component emission factors
were based on the GRI/lndaco measurement program conducted at 15 compressor stations. Some compressor
All Equipment at Gas Processing Plants
Normal Operation
Steady and Unsteady, Fugitive
24.45 Bscf ± 68%
B-4
-------
components are unique, while others have higher leak rates than identical components elsewhere in the plant
due to vibration. Compressors have the following types of components:
1) Comp. B/D OEL A blowdown (B/D) valve to the atmosphere that can depressure the compressor
when idle. The B/D valve or the large unit block valves (depending on the
operating status of the compressor) can act as an open-ended line that leaks at an
extraordinarily high rate through the valve seat. The leak rate is dependent upon
whether the compressor is pressurized (in operation or idle, pressurized) or
depressurized (idle, depressurized).
2) Comp. PRV The pressure relief valve (PRV) is usually installed on a compressor discharge line,
and leaks at a higher than average rate due to vibration.
3) Comp. Starter OEL Most compressors have a gas starter motor that turns the compressor shaft to start
the engine. Some use natural gas as the motive force to spin the starter's turbine
blades, and vent the discharge gas to the atmosphere. The inlet valve to the starter
can leak and is therefore an OEL unique to compressors.
4) Comp. Seal All compressors have a mechanical or fluid seal to minimize the flow of pressurized
natural gas that leaks from the location where the shaft penetrates the compression
chamber. These seals are vented to the atmosphere. Reciprocating compressors
have sliding shaft seals while centrifugal compressors have rotating shaft seals.
5) Miscellaneous There are many components on each compressor, such as valve covers on
reciprocating compressor cylinders and fiiel valves.
Each compressor has one B/D OEL, one PRV, and one starter OEL. Reciprocating compressors have one
compressor seal per compression cylinder (which averaged 2.5 per engine), while centrifugal compressors
have 1.5 seals per gas turbine. For the miscellaneous component category, there are many components per
compressor engine, but the emission rates were minor and so were added into one lump emission factor per
compressor for miscellaneous components.
All of the compressor emission factors take several correction factors into account. First, the various phases
of compressor operations [such as the amount of time that compressors are a) idle and depressured, b) idle
and pressured up, or c) running]. This is actually a complex adjustment that takes into account valve position
practices. [See Methane Emissions from the Natural Gas Industry, Volume 8: Equipment Leaks (2) for
details.] Correction factors were also added for fraction of starter gas turbines using air instead of gas (75%
for recip, 33% for turbines in gas processing), and for sites with flares handling PRV or compressor B/D
discharge (approximately 11% of the compressor blowdown OELs were routed to a plant flare).
EF DATA SOURCES:
1. Component emission factors based on screening results from API/GRI/Star program for the
component EF's for eight gas processing plants and EPA's current default zero factors,
correlation equations, and pegged source factors. Confidence limits derived from analysis of
screening data by Radian in April 1995.
2. OEL (site B/D) emission factor based on results from GRI/Indaco program for compressor
stations (June 1994).
3. Plant component counts were based on average of 8 API/Star sites, 6 EPA/Radian sites in 1982,
and 7 sites visited under this project in 1992.
4. Compressor emission factors based on results from GRI/Indaco program for 15 compressor
stations (June 1994). Compressor operating hours (% running) based on data from 3 gas
processing company databases.
B-5
-------
Average Facility Emissions for Gas Processing
Equipment Type
Component Type
Component
Emission Factor,
Mscf/component-yr
Average Component
Count
Average Equipment
Emissions,"
MMscf/yr
Gas Plant (non-
Valve
1.305
1,392
2.89 (48%)
compressor related
components)
Connection
0.117
4,392
Open-Ended Line
0.346
134
Pressure Relief
Valve
0.859
29
Site Blowdown
Open-Ended Line
230
2
Reciprocating
Compressor
Compressor
Blowdown Open-
Ended Line
2,036bc
1
4.09 (74%)
Pressure Relief
Valve
349bc
1
Miscellaneous
189c
1
Starter Open-Ended
Line
1,341
0.25®
Compressor Seal
450c
2.5
Centrifugal
Compressor
Compressor
Blowdown Open-
Ended Line
6,447bd
1
7.75 (39%)
Miscellaneous
31d
1
Starter Open-Ended
Line
1,341
0.667f
Compressor Seal
228d
1.5
a Values in parentheses represent 90% confidence interval.
b Adjusted for 11.1% of compressors which have sources routed to flare.
c Adjusted for 89.7% of time reciprocating compressors in processing are pressurized.
d Adjusted for 43.6% of time centrifugal compressors in processing are pressurized.
e Only 25% of starters for reciprocating compressors in processing use natural gas.
f Only 66.7% of starters for centrifugal compressors in processing use natural gas.
B-6
-------
EF ACCURACY: a. Plant Emission Factor = + 48%
b. Recip. Compressor = + 74%
c. Turbine Compressor = + 39%
Basis:
1. The accuracy was propagated through the EF calculation from each terms accuracy. 90%
confidence intervals were calculated for the sites using the t-statistic method. The 90%
confidence intervals accounted for variability in component count from the range in site averages
and estimates were also provided for the component emission factors from the API/Star and
GRI/Indaco program.
ACTIVITY FACTOR a. Plant Activity Factor = 726 plants
b. Compressor Activity Factor = 4092 recip engines, 726 turbines
The number of gas processing plants was determined from the Oil and Gas Journal (3) (July 1993).
The number and type of gas processing compressor engines were determined from eleven gas plant site visits.
The average ratio of compressors per plant was multiplied by the total number of plants, 726, to obtain these
estimates. [See Methane Emissions from the Natural Gas Industry, Volume 5: Activity Factors (4) for
details.]
AF DATA SOURCES: Oil and Gas Journal (July 1993) (3)
AF ACCURACY: a. Plant Activity Factor: + 2%
b. Compressor Activity Factor: Recip engines = ± 48%; Turbines = ± 77%
Basis:
1. An accurate count of gas plants by the Oil and Gas Journal (3) is very likely since counting
such large, discreet facilities should be straightforward. The ± 2% was assigned by engineering
judgement.
2. The compressor count accuracy was determined by statistical analysis of the "compressor per
site" averages for 11 gas plant sites.
3. A check was performed to estimate whether gas plant sites visited for compressor counts were
representative of industry average. Based on Oil and Gas Journal (3), the average plant capacity
was 88.3 MMscfd and throughput was 51.2 MMscf/d. Site visit data averaged 271 MMscfd and
throughput was 182 MMscf/d, suggesting that plants visited were larger than average. However,
further investigation revealed that there is no correlation between plant capacity/throughput and
number of compressors (The plant visited with the most compressors had 20 engines with 20,000
HP and a low throughput of 56 MMscfd, while the plant with the highest current operating rate
of 750 MMscfd had only one compressor at 17,500 HP.)
B-7
-------
ANNUAL EMISSIONS: (24.45 Bscf/yr ± 16.7 Bscf/yr)
The annual emissions were determined by multiplying the average equipment/facility emissions by the
population of equipment in the segment.
Category
Emission
Factor
Activity
Factor
Emission
Rate
Uncertainty
Gas processing
plants
2.89 MMscf/yr
methane
726 plants
2.1 Bscf/yr
methane
48%
Recip Comp
4.09 MMscf/yr
methane
4092 recip
16.7 Bscf/yr
methane
95%
Turbine Comp
7.75 MMscf/yr
methane
726 turbine
5.6 Bscf/yr
methane
91%
TOTAL
24.4 Bscf/yr
methane
68%
REFERENCES
1. Hausle, K.J., Protocol for Fugitive Leak Emission Estimates, Final Report, EPA-453/R-93-026, U.S.
Environmental Protection Agency, June 1993.*
2. 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.
3. Oil and Gas Journal. 1992 Worldwide Gas Processing Survey Database, 1993.
4. 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.
(*) NTIS PBS5-22S21S.
B-8
-------
GP-2
PROCESSING SOURCE SHEET
SOURCES:
COMPONENTS:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Glycol dehydrators remove water from a gas stream by contacting the gas with glycol and then driving the
water from the glycol and into the atmosphere. The glycol also absorbs a small amount of methane, and
some methane can be driven off to the atmosphere through the reboiler vent.
EMISSION FACTOR: (121.55 scf/MMscf ± 201.96%)
A thermodynamic computer simulation was used to determine the most important emission-affecting variables
for dehydrators. The variables are: (gas throughput, existence of a flash tank, existence of stripping gas,
existence of a gas-assisted pump, existence of vent controls routed to a burner). Throughput, since its effect
is linear, is handled by establishing an emission rate per gas throughput. Emission rates per throughput are
then established for the other important emission affecting characteristics. Gas driven pumps are ignored here
and handled in a separate source analysis. The emission factor is then:
EF — [ ( Fpj x EFpj ) + ( Fnt x EFnt ) + ( FSG x EFSG ) ] x F^^ x OC
EF = [ (0.667 x 3.57) + (0.333 x 175.10) + (0.111 x 670) ] x 0.900 x 1.0
Fpp = Fraction of the population WITH flash tanks
0.667 ± 10.13%
Fjn- = Fraction of the population WITHOUT flash tanks = l-F^
0.333 ± 20.12%
Fsg = Fraction of the population WITH stripping gas
0.111 ± 186%
Fnvc= Fraction of the population WITHOUT combusted vent controls
0.90 ± 10%
EFp^ Total CH4 emission rate per 1 MMscf throughput for dehydrator that has a flash
tank
3.57 (+102% / -58%)
EFfj^ Total CH4 emission rate per 1 MMscf throughput for dehydrator that does NOT
have a flash tank
175.10 (+101% / -50%)
EFSG= Incremental emission rate per 1 MMscfd throughput for dehydrator that has stripping
gas
670 (+40% / -60%)
OC = Overcirculation factor for glycol—number of times the industry rule-of-thumb of 3
gallons glycol/lb water
1.0 ± 0%
Glycol Dehydrators
Reboiler Vent
Normal Operation
Unsteady, Vented
1.05 Bscf ± 208%
B-9
-------
EF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators (1)
establishes emission affecting characteristics of dehydrators.
2. Site visit data establish the FSG and Fwc for multiple sites (7 PROC sites with dehydrators).
3. TMOGA/GPA survey of 207 gas plant dehydrators established F3P and FNT) and TP for
dehydrators for the processing segment.
4. ASPEN computer simulations were used to determine EF3P, and EF^ from the dehydrator vent.
5. Sampling data from the GRI Glycol Dehydrator Sampling and Analytical Program for one
dehydrator was used to determine EFSG (Glycol Dehydrator Emissions: Sampling and Analytical
Methods and Estimation Techniques') (2). The upper bound was calculated by assuming that all
of the measured noncondensable vent gas was due to stripping gas that was 100% methane. The
lower bound was calculated as the rule-of-thumb stripping gas rate recommended by a glycol
dehydrator manufacturer.
EF ACCURACY 121.55 sctfMMscf ± 201.96%
Basis:
The accuracy is rigorously propagated through the EF calculation from each term's accuracy:
1. ASPEN has been demonstrated to match actual dehydrators within ± 20% within the calculated
confidence intervals obtained from site data.
2. Individual EF confidence intervals were calculated from the other data based upon the spread of
the 11 site averages.
ACTIVITY FACTOR: (8.63 TscCyear gas throughput in the gas pocessing segment)
The glycol dehydrator throughput is estimated from the fraction of gas processed by refrigerated processes (as
opposed to dry bed dehydration for cryogenic processes). The estimate was obtained from the Oil & Gas
Journal (3) annual Gas Processing Survey. Of a total of 17.44 Tscf, 8.63 Tscf were determined to be
dehydrated by glycol.
AF ACCURACY: 8.63 Tscf/year ± 22.45%
Basis:
1. Uncertainty based on estimate of confidence limits for Oil and Gas Journal survey.
AF DATA SOURCES:
1. Oil & Gas Journal (3) annual Gas Processing Survey.
ANNUAL METHANE EMISSIONS: (1.0490 Bscf ± 208.20%)
The annual methane emissions were determined by multiplying the dehydrator emission factor by the activity
factor.
REFERENCES
1. Myers, D.B. 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.
2. Radian Corporation. Glycol Dehydrator Emissions: Sampling and Analytical Methods and Estimation
Techniques. GRI-94/0324, Gas Research Institute, Chicago, IL, March 1995.
3. Oil & Gas Journal. 1992 Worldwide Gas Processing Survey Database, 1993.
B-10
-------
GP-3
PROCESSING SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
Acid Gas Removal (AGR) Units
Normal Operation
Unsteady, Vented
0.82 Bscf ± 109%
BACKGROUND:
AGR units remove acid gas (H2S and C02) from a natural gas stream by contacting the gas with material
(usually amines) and then driving the absorbed components from the solvent. The amines can also absorb a
small amount of methane, and some methane can be driven off to the atmosphere through the reboiler vent to
the atmosphere.
EMISSION FACTOR: (6083 scfd/avg AGR)
AGRs were assumed to have an absorption of methane similar to water, since the typical AGR solution
contains over 50% water. The methane emissions were calculated using an ASPEN PLUS process simulation
based on an actual DEA unit (1). AGRs were assumed to have no three-phase flash tanks nor stripping gas.
The average AGR throughput (MMscfd) was determined from a 1982 API study, and multiplied times the
emission rate (CH„/MMscfd). The emission factor is then:
EF = EFOT x FnvC x TP
Fmvp = Fraction of the AGRs that do vent the waste stream
0.18 ± 10%
TP = Average throughput for AGRs (MMscfd)
35.02 ± 20%
EFfjr = Total "CH4 scfd emission rate per 1 MMscfd throughput" for an AGR
965 ± 100%
EF DATA SOURCES:
1. ASPEN PLUS process simulations based on an actual DEA unit were used to determine EFOT
from the reboiler vent. It was assumed that AGRs have an absorption of methane similar to
water.
2. 1982 API Survey, quoted in Investigation of US Natural Gas Reserve Demographics and Gas
Treatment Processes, shows 287 AGR units, with a cumulative throughput of 10052 MMscfd (2).
The survey also shows split of AGR vent dispositions: 50% burned, 32% to sulfur recover, and
18% vented.
EF ACCURACY: 6083 ± 104.92%
Basis:
1. The accuracy is based upon engineering judgement that the methane solubility in AGR solutions
is similar to the solubility in water.
B-ll
-------
ACTIVITY FACTOR: (371 active AGR units in the U.S.)
The number of AGR units in the U.S. have all been assumed to be in the processing segment. The activity
factor was extracted from the Purvin & Gertz survey.
AF DATA SOURCES:
1. Purvin & Gertz, Business Characteristics of the Natural Gas Conditioning Industry (3), 1993.
AF ACCURACY: 371 ± 20%
Basis:
1. The accuracy is based upon engineering judgement. The survey should have excellent accuracy
(± 5%), but the upper bound at 90% confidence was revised upward to 20% to be conservative.
ANNUAL METHANE EMISSIONS: (0.8237 Bscf ± 108.85%)
The annual methane emissions were determined by multiplying an emission factor for an average dehydrator
by the population of AGRs in the segment.
REFERENCES
1. 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.
2. Radian Corporation. Investigation of U.S. Natural Gas Reserve Demographics and Gas Treatment
Processes, Topical Report, Gas Research Institute, January 1991.
3. Tannehill, C.C. and C. Galvin. Business Characteristics of the Natural Gas Conditioning Industry,
Topical Report. GRI Contract 5088-221-1753, Gas Research Institute, May 1993.
B-12
-------
GP-4
PROCESSING SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Blowdown is the direct, intentional venting to the atmosphere of gas contained inside operating equipment.
The gas is released to provide a safer working environment for maintenance activities around or inside the
equipment.
EMISSION FACTOR: 4,060 ± 262% Mscf/gas plant
(Corrected for the gas processing methane composition of 87 mol%)
Blowdowns at gas plants consist primarily of the following types of events: compressor blowdown,
compressor starts, pipeline pig receiver blowdown, and miscellaneous vessel blowdown. Due to the
similarities in station blowdown practices between the gas processing and transmission segments, transmission
station company tracked data were applied to gas plants. Blowdown volumes per station were provided based
on company tracked data from 9 transmission companies.
EF DATA SOURCES:
1. The Methane Emissions from the Natural Gas Industry, Volume 7: Blow and Purge Activities (1)
establishes emission affecting characteristics of blowdown practices.
2. Company tracked data were provided from 9 transmission companies representing a total of 328
stations.
EF ACCURACY: ± 262%
Basis:
The accuracy was calculated from the spread of the company tracked data. A 90% confidence interval
is calculated for the data using the method presented in Methane Emissions from the Natural Gas
Industry, Volume 4: Statistical Methodology (2).
ACTIVITY FACTOR 726 ± 2% gas plants
AF DATA SOURCES:
1. The number of gas processing plants for 1992 is repeated in the Oil and Gas Journal (3).
AF ACCURACY:
Basis:
An accurate count of gas plants by the Oil and Gas Journal is very likely since counting such large,
discrete facilities should be straightforward. The ± 2% was assigned by engineering judgement.
All Equipment (vessels, compressors, pig traps, manifolds)
Maintenance
Unsteady, Vented
3.0 Bscf ± 262%
B-13
-------
ANNUAL METHANE EMISSIONS: 2.95 ± 262 Bscf
The annual methane emissions were determined by multiplying an emission factor by the activity factor
(population). Each emission factor was adjusted for the average methane content in the gas processing
segment of 87 mol%.
REFERENCES
1. 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.
2. 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.
3. Bell, L. "Worldwide Gas Processing," Oil and Gas Journal, July 12, 1993, p. 55.
B-14
-------
GP-5
PROCESSING SOURCE SHEET
SOURCES:
COMPONENTS:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
Glycol Dehydrators
Gas Assisted Kimray Pumps
Normal Operation
Unsteady, Vented
0.170 scf ± 228%
BACKGROUND:
Most glycol circulation pumps in gas plants are electric. However, some gas driven pumps do exist. Gas-
assisted Kimray glycol circulation pumps use a mixed phase of wet glycol liquid and absorber gas to drive
pistons that pump dry (lean) glycol circulation. Unlike chemical injection pumps which vent the driving gas
directly to the atmosphere, Kimray pumps pass the driving gas along with the wet glycol to the reboiler. In
the reboiler the methane is driven off into the vent line. Depending on dehydrator vent gas dispositions, the
methane may be vented to the atmosphere or controlled and burned.
EMISSION FACTOR: (177.75 scf CH4/MMscf gas processed)
The average glycol pump gas emission factor was determined by an equation describing the gas generation
and disposition of gas from the pump. The disposition of gas generated by the pump depends upon the
existence of a flash tank and vent controls. Measured and estimated parameters were input into the equation.
In general, the emission factor for a gas-assisted pump was determined by the following equation:
EFpump = PGU x CR x WR x OC x x FNTyC
= (3.73 scf/gal) x (3.0 gal/lb H20) x (53 lb H20/MMscf) x (1.0) x (0.333) x (0.900)
= 177.75 scf CH4/MMscf gas ± 56.85%
EF DATA SOURCES:
1. Equation 1, i.e. the effects of operating variables on emissions, was defined in Methane Emissions from
the Natural Gas Industry, Volume 15: Gas-Assisted Glycol Pumps (1).
2. CR = glycol circulation ratio = 3.0 gal glycol/lb water ± 33.3%.
3. WR = water removed from wet gas = 53 lb water/MMscf gas + 20%. For inlet gas stream of 95°F and
800 psig dried to 7 lb water/MMscf gas.
4. OC = factor to account for overcirculation of glycol = 1.0 ± 0%.
5. F^ = fraction of dehydrators without flash tanks = 0.333 ± 20.12%.
6- Fnvc = fraction of the dehydrators without combustion vent controls = 0.900 ± 10%.
7. PGU = pump gas usage = 3.73 scf CH4/gal glycol ± 30%. Determined by multiplying the volume of
gas used by high-pressure pump models by a typical fraction of methane in the natural gas (83 mole%).
B-15
-------
PGU
= 4.49 scf/gallon x 83%
= 3.73 sctfgallon ± 30%
EF ACCURACY: (± 56.85%)
Basis:
1. Assumption: The manufacturer's data and ranges are relatively accurate (±30%).
2. Dehydrator characteristics based on site visit observations and TMOGA survey.8
ACTIVITY FACTOR: (0.9579 Tscf/year in the processing segment with gas-assisted pumps)
The volume of gas processed through dehydrators using gas-assisted pumps was calculated from the total
throughput for gas processing dehydrators and the fraction of dehydrators using gas-assisted pumps
determined from site visits. The activity factor is then:
AF = (fraction of dehydrators with gas-assisted pumps) x (throughput for gas processing
dehydrators)
= (0.111 ± 186%) x (8.63 Tscf/year ± 22.4%)
= 0.9579 Tscf/year ± 191.95%
AF DATA SOURCES:
1. See Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators (2) for
an explanation of processing dehydrator throughput (8.63 Tscf/year). See the Methane Emissions
from the Natural Gas Industry, Volume 5: Activity Factors (3) for more details.
2. Fraction of dehydrators using gas-assisted pumps came from data from site visits.
AF ACCURACY: (± 192%)
Basis:
Calculated from confidence limits of gas throughput and fraction of dehydrators by standard error
propagation analysis.
ANNUAL METHANE EMISSIONS: (0.1703 Bscf ± 228%)
The annual methane emissions were determined by multiplying an emission factor (scf CH4/MMscf) by the
total throughput for processing dehydrators using gas-assisted pumps.
(177.75 scfTMMscf) x (0.9579 Tscf) = 0.1703 Bscf (± 228.00%)
REFERENCES
1. 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.
2. Myers, D.B. Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators.
Final Report, GRI-94/0257.32 and EPA-600/R-96-080n. Gas Research Institute and U.S.
Environmental Protection Agency, June 1996.
B-16
-------
3. Stopper, 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.
B-17
-------
GP-6
GAS PROCESSING SOURCE SHEET
SOURCES:
COMPONENTS:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
The gas processing segment uses compressed air to power the majority of the pneumatic devices within the
plant, although some devices may be powered by natural gas. Many plants use gas driven pneumatic
controllers on isolation valves for emergency shut-down or maintenance work.
The same type of devices used in the transmission segment are also commonly used in the gas processing
segment — continuous bleed throttling/regulating valves, displacement operators, and turbine operators.
EMISSION FACTOR: 165 Mscf per average plant ± 133%
(This was adjusted for the gas processing methane fraction of natural gas at 87 mol%.)
The average device gas emission factor was determined from a combination of vendor information on device
emission rates and device counts from several sites. The average emission factor was calculated using the
following equation:
Various Equipment (vessels, compressors, piping)
Pneumatic Devices
Normal Operation
Unsteady, Vented
0.1 Bscf ± 133%
n
^2 (Annual Site Emissions, scf Natural Gas)
EF , . = K x ili x % Methane
avg.pneum.device ^
K = fraction of sites that use natural gas rather than air (0.56 ± 59%)
n = number of sites operating with natural gas
Each term in this equation was determined from site specific information. The summation of the site specific
data was then adjusted based on the number of sites with gas operated devices versus the total number of
sites surveyed. The site results are shown in the following table.
B-18
-------
Site
Device Type
Number
of Devices
Operations/
Year
Annual
Displacement/
Device, scf
Displacement/
Site, scf
1
Throttling (Fisher)
2
Continuous
497,584
995,168 ± 29%
2
Isolation (Fisher)
3
12
214,675
644,025 ± 29%
3
Air
-
--
~
--
4
Isolation (Turbine)
25
1
780
19,500 ± 112%
5
Isolation (Rotary Vane)
7
18
12
1
48
1,206 ± 49%
6
Isolation
(Turbine & Rotary Vane)
1
16
1
12
3,376
44,115 ± 68 %
7
Air
-
-
-
--
8
Air
--
-
-
--
9
Air
-
-
~
--
TOTAL
1,704 Mscf ± 21%
Average (for gas sites)
341 Mscf ± 103%
EF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 12: Pneumatic Devices (1)
establishes the important emission-affecting characteristics.
2. Site visit device counts establish the number of continuous bleed devices, turbine operators, and
displacement operators for each site.
3. The emission factor for continuous bleed devices was estimated using data provided by one site
and measurements for transmission pneumatic devices.
4. Gas usages for the displacement operators were provided by Pantex Valve Actuators and Systems
and Shafer Valve Operating Systems. The number of devices, supply gas pressure, and operating
frequency were based on site information.
5. Gas usages for the turbine operators were provided by Limitorque Corp. Operating duration,
frequency, and supply gas pressure were based on site information.
EF ACCURACY:
Basis:
1. EF accuracy is based on error propagation from the spread of data for the nine sites visited.
2. It was assumed that the manufacturers' data are completely accurate.
ACTIVITY FACTOR: 726 gas processing plants + 2%
The activity factor for the gas processing segment was taken from Oil and Gas Journal (2) published
information from the year 1992.
B-19
-------
AF DATA SOURCES:
1. The number of gas processing plants was taken from the Oil and Gas Journal (2).
AF PRECISION:
Basis:
1. AF accuracy is based on engineering judgement.
ANNUAL METHANE EMISSIONS: 0.12 Bscf ± 133%
The annual emissions were determined by multiplying an average site emission factor (adjusted for the
methane composition) by the total number of gas processing sites.
165 Mscf/site x 726 sites = 0.12 Bscf
REFERENCES
1. 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.
2. Bell, L. "Worldwide Gas Processing," Oil and Gas Journal, July 12, 1993, p. 55.
B-20
-------
APPENDIX C
Transmission/Storage Source Sheets
C-l
-------
APPENDIX C
Transmission/Storage Source Sheets
Page
T-l - Fugitive Emissions (Transmission) C-4
T-2 - Meter and Regulating Station Emissions C-9
T-3 - Pipeline Leaks C-13
T-4 - Pneumatic Devices C-16
T-5 - Maintenance Venting C-19
T-6 - Glycol Dehydrator Vents C-22
S-l - Fugitive Emissions (Storage Facilities) C-24
S-2 - Glycol Dehydrators (Storage Facilities) C-29
C-2
-------
TRANSMISSION/STORAGE SOURCE SHEETS
This section contains the specific source sheets for the transmission and storage segment of
the natural gas industry. The following table serves as a guide for finding sheets in this
section. The cells in the table give the sheet number (T-l, T-2, etc.) of the source sheet.
The rows define the equipment covered, while the columns define the emission type. A
category with no sheet number means that the emissions from that area were determined to
be zero or negligibly small. The label for each source sheet is shown at the top of the
cover page for that sheet.
TABLE OF
CONTENTS : rr.
OPERi8
EMISSION T
iTING MODE,
YPE (Fugitive, Vented, or Combi
isted)
EQUIPMENT:
Start Up
Normal Operations
Maintenance
Upsets
Mis-
haps
V
C
F
V
C
V
C
V
C
V
Entire Transmission
Compressor Stations
T-l
T-4
T-5
T-5
Trans. Comp. Station
Dehydrators
T-l
T-4,
T-6
T-5
T-5
Trans. Comp. Station
Vessels
T-l
T-4
T-5
T-5
Trans. Comp. Station
Compressors
T-l
T-4
P-l
T-5
T-5
Transmission Pipelines
T-3
T-4
T-5
T-5
Trans. M&R Stations
T-2
T-2
T-5
T-5
Entire Storage Stations
S-l
T-4
T-5
T-5
Stor. Sta. Wells
S-l
T-4
T-5
T-5
Stor. Sta. Compressors
S-l
T-4
P-l
T-5
T-5
Stor. Sta. Vessels
S-l
T-4
T-5
T-5
Stor. Sta. Dehydrators
S-l
T-4,
S-2
T-5
T-5
C-3
-------
T-l
TRANSMISSION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
Compressor Stations
Normal Operation
Steady and Unsteady, Fugitive
50.7 Bscf ± 52%
BACKGROUND:
Equipment leaks are typically low-level, unintentional losses of process fluid (gas or liquid) from the sealed
surfaces of above-ground process equipment. Equipment components that tend to leak include valves, flanges
and other connectors, pump seals, compressor seals, pressure relief valves, open-ended lines, and sampling
connections. These components represent mechanical joints, seals, and rotating surfaces, which in time tend
to wear and develop leaks.
EMISSION FACTOR: a. Station = 3.2 MMscf/yr methane per plant
b. Recip. Compressor = 5.55 MMscf/yr methane per recip
c. Turbine Compressor = 11.1 MMscf/yr methane per turbine
The average fugitive emission rate for transmission compressor stations was determined to be composed of
two parts: a) station components (excluding compressor-related components); and b) compressor-related
components. Fugitives from the compressor-related components have much higher emission factors than
components in the rest of the facility. This is due in part to the high vibration that compressors generate, but
most of the larger emissions are due to unique compressor components, as explained below.
a. The contribution from non-compressor components was determined by multiplying the average component
count by the component emission factor. The number of components was subdivided into valves,
connections/flanges, small open-ended lines, site blowdown (B/D) OELs, control valves, and other
components (such as pressure relief valves). (Tubing components were determined to be insignificant.) All
of these components are typical fugitive components (as described in the EPA Fugitive Emissions Protocol)
with the exception of control valves and site B/D OELs. Control valves emit at a higher rate than manual
isolation valves since their packing is stressed more often as they are activated much more frequently. Site
B/D OELs are the large diameter emergency station blowdown valves that are designed to depressure the
entire site to the atmosphere when the valve is opened.
The component emission factors for station components were based on a GRI/Indaco measurement program
conducted at 6 compressor stations. The average facility emissions are then calculated as follows:
EF [(Nv)v x EFv!v) + (NC.V|V x EFc_vlv) +(Ncn x EF^) + (NmI x EF^,) + (Nprv x EFprv) + (NsiteB/D x EFsileB/D)]
where:
Nx = average count of components of type x per plant, and
EFX = average methane emission rate per component of type x.
b. The contribution from compressor-related components was obtained by multiplying the average number of
fugitive components per compressor engine by the component emission factors. The component emission factors
were based on the GRI/Indaco measurement program conducted at 15 compressor stations. Some compressor
components are unique, while others have higher leak rates than identical components elsewhere in the plant due
to vibration. Compressors have the following types of components:
C-4
-------
2) Comp. PRY
1) Comp. B/D OEL A blowdown (B/D) valve to the atmosphere that can depressure the compressor when
idle. The B/D valve or the large unit block valves (depending on the operating status
of the compressor) can act as an open-ended line that leaks at an extraordinarily high
rate through the valve seat. The leak rate is dependent upon whether the compressor
is pressurized (in operation or idle, pressurized) or depressurized (idle, depressurized).
The pressure relief valve (PRV) is usually installed on a compressor discharge line and
leaks at a higher than average rate due to vibration.
3) Comp. Starter OEL Most compressors have a gas starter motor that turns the compressor shaft to start the
engine. Some use natural gas as the motive force to spin the starter's turbine blades
and vent the discharge gas to the atmosphere. The inlet valve to the starter can leak
and is therefore an OEL unique to compressors.
All compressors have a mechanical or fluid seal to minimize the flow of pressurized
natural gas that leaks from the location where the shaft penetrates the compression
chamber. These seals are vented to the atmosphere. Reciprocating compressors have
sliding shaft seals while centrifugal compressors have rotating shaft seals.
There are many components on each compressor, such as valve covers on
reciprocating compressor cylinders and fuel valves.
4) Comp. Seal
5) Miscellaneous
Each compressor has one B/D OEL, one PRV, and one starter OEL. Reciprocating compressors have one
compressor seal per compression cylinder (which averaged 3.3 per engine), while centrifugal compressors have
1.5 seals per gas turbine. For the miscellaneous component category, there are many components per
compressor engine, but the emission rates were minor and so were added into one lump emission factor per
compressor for miscellaneous components.
All of the compressor emission factors take several correction factors into account. First, the various phases of
compressor operations (such as the amount of time that compressors are a) idle and depressured, b) idle and
pressured up, or c) running). This is actually a complex adjustment that takes into account valve position
practices. [See Methane Emissions from the Natural Gas Industry, Volume 8: Equipment Leaks (1) for more
details.] Correction factors were also added for fraction of starter gas turbines using air instead of gas (100%
for recip, 0% for turbines in Transmission).
EF DATA SOURCES:
1. Component emission factors based on results from GRI/Indaco program for the component EF's for
6 transmission compressor stations (June 1994). Adjustment of station EF is to account for data
obtained from one interstate transmission pipeline company that was found to have higher emissions
than average.
2. Plant component counts were based on an average of 8 Indaco sites in 1994 and 9 sites visited
under this project in 1993, plus 7 industry sites.
3. Compressor emission factors based on results from GRI/Indaco program for 15 compressor stations
(June 1994). Compressor operating hours (% running) based on data from FERC database, GRI
TRANSDAT (2) database, and data supplied by one large interstate transmission pipeline company.
4. Fraction of methane (93.4 mol%) based on data from GRI TRANSDAT database.
C-5
-------
Average Facility Emissions for Gas Transmission
Equipment Type
Component Type
Component
Emission Factor,
Mscf/component-yr
Average
Component
Count
Average Equipment
Emissions,"
MMscf/yr
Compressor Station
Valve
0.867
673
3.01 (102%)
(non-compressor
related components)
Control Valve
8.0
31
(Note: 3.2 MMscf/yr
used in national
Connection
0.147
3,068
emission estimate)b
OEL
11.2
51
PRV
6.2
14
Site B/D OEL
264
4
Reciprocating
Compressor
Compressor B/D
OEL
3,683
1
5.55 (65%)
PRV
372c
1
Miscellaneous
180°
1
Compressor Starter
OEL
d
d
Compressor Seal
396c
3.3
Centrifugal
Compressor
Compressor B/D
OEL
9,352
1
11.1 (34%)
Miscellaneous
18c
1
Compressor Starter
OEL
1,440
1
Compressor Seal
165°
1.5
1 Values in parentheses represent 90% confidence interval.
b Adjusted for data received from one company that was not considered representative of national average.
c Adjusted for the fraction of time the compressor is pressurized (79.1% and 24.2% for reciprocating and
centrifugal
compressors, respectively).
d Reciprocating compressor starters were assumed to use compressed air or electricity instead of natural gas.
C-6
-------
EF ACCURACY: a. Station = 102%
b. Recip. Compressor = 65%
c. Turbine Compressor = 34%
Basis:
Rigorous propagation of error from the spread of thousands of individual measurements taken by Indaco.
ACTIVITY FACTOR: a. Station Activity Factor = 1700 stations
b. Compressor Activity Factor = 6799 recip engines, 681 turbines
AF DATA SOURCES:
1. 1992 FERC Form 2 responses accounted for 70% of national transmission pipeline mileage. Total
station count extrapolated using national total transmission mileage of 276,900 miles from A.G.A.
Gas Facts (3).
2. Compressor engine count based on GRI TRANSDAT "industry database" with adjustments for total
industry horsepower. Transmission compressor station counts were split from storage based upon
storage station site visit data and Gas Facts (3) data on storage stations. Added 0.2% to recip
count account for electric motor drivers.
AF ACCURACY: a. Station Activity Factor: ± 10%
b. Compressor Activity Factor: Recip engines = ± 17 %; Turbines = + 26 %
Basis:
1. FERC Form 2 data have a high percentage (70%) of all transmission companies. Therefore a
national extrapolation should not add much error. This 10% figure was assigned based on
engineering judgement.
2. The compressor count accuracy was assigned based upon the propagation from: a) rigorous error
propagation for the 8 storage station "compressor/station" averages; and b) engineering judgement
assignment of + 10% error to the large GRI TRANSDAT database.
ANNUAL METHANE EMISSIONS: (50.73 Bscf/yr ± 26.3 Bscf/yr)
The annual emissions were determined by multiplying the average facility /equipment emissions by the population
of equipment in the segment.
Category
Emission Factor
Activity Factor
Emission Rate
Uncertainty
Station
3.2 MMscf/yr
CH4
1700 stations
5.4 Bscf/yr
CH4
103%
Recip Comp
5.55 MMscf/yr
CH4
6799 recip
37.8 Bscf/yr
CH4
68%
Turbine Comp
11.1 MMscf/yr
CH4
681 turbine
7.5 Bscf/yr
CH4
44%
TOTAL
50.7 Bscf/yr
CH4
52%
C-7
-------
REFERENCES
1. 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.
2. Biederman, N. GRI TRANSDAT Database: Compressor Module, (prepared for Gas Research Institute)
npb Associates with Tom Joycxe and Associates, Chicago, IL, August 1991.
3. American Gas Association, Gas Facts. Arlington, VA, 1992.
C-8
-------
T-2
TRANSMISSION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Metering/pressure regulating (M&PR) stations are located throughout the transmission network to meter gas where
a custody transfer occurs. Emissions from M&PR include continuous fugitive losses and also may include
intermittent emissions from pneumatic devices such as pressure regulators, if they exist at the station. Fugitive
emissions are relatively low-level emissions of process fluid (gas or liquid) from process equipment. Specific source
types include various fittings such as valves, flanges, or compressor seals. These components represent mechanical
joints, seals, and rotating surfaces, which in time tend to wear and develop leaks.
The transmission segment contains many "metering and regulation stations" (M&R stations) where flow is measured
for custody transfer or system control. The table below shows the types of M&R stations that transmission
companies count in their system. Most of the meter station types associated with the transmission system have
already been counted in other segment calculations (receipt stations in production and delivery stations in
distribution).
Only three types remain to be accounted for under the transmission system M&PR stations: 1) farm taps, 2) direct
industrial sales from the transmission pipeline, and 3) transmission company-to-transmission company transfer
stations.
Transmission Meter and Regulation Station Types
GENERAL
STATION
SERVICE
SPECIFIC TYPE
STATION TECHNICAL
DESCRIPTION
ACCOUNTED FOR IN
OTHER SEGMENT
SOURCE SHEETS?
RECEIPT TO
THE
SYSTEM:
1. Gathering meters at produc-
tion sites
Meter Only
Yes, P-2
INTER-
SYSTEM:
2. Meters at compressor
stations
Meter Only
Yes, T-l
DELIVERY
TO CUS-
TOMERS:
3. City Gate M&R stations
Meter and Regulation
(Pressure regulation)
Yes, D-l
4. Industrial sales directly off
of transmission pipelines
Meter and Regulation
(Pressure regulation)
Some in D-l, but those owned
by transmission companies in
this sheet (T-2)
5. Farm sales off gathering
and transmission pipelines
Meter and Regulation
(Pressure regulation)
No, so accounted for in this
sheet (T-2)
6. Sales to Other Transmission
Companies (Inter-connects)
Most often Meter only, but
can have some flow
regulation
No, so accounted for in this
sheet (T-2)
Meter and Regulating Stations
Normal Operation
Steady, Fugitive, and Vented
4.5 Bscf + 835%
C-9
-------
Although direct customer connections (sales) on the transmission pipeline are rare, where they exist they are often
owned by distribution companies, even if they only own a few feet of line. Many farm taps are still owned by
transmission companies, even though there is a trend to let LDCs handle the farm taps or to remove them entirely.
Therefore, many direct sales from the transmission pipeline are already accounted for in the distribution M&PR
calculations. Only the direct sales from the transmission pipeline that are owned completely by the transmission
companies are counted under this source sheet.
Most large transmission companies have interconnects with other transmission companies to allow for flexibility
of supply. These shared stations can flow in either direction.
EMISSION FACTOR: (see below)
The average fugitive emission rate for transmission M&R stations was determined by analysis of the GRI tracer
measurement tests for gas industry M&R stations. Transmission farm taps and industrial meters are both direct-
connects to high pressure pipelines, and will have one pressure regulator (and not 3 to 22 regulators, as some city
gates had) in addition to a meter. The pressure regulator is a self contained device, and so does not have significant
pneumatic emissions. Therefore the tracer data set was sorted and adjusted as follows:
1) include only stations with one regulator,
2) include only stations in vaults (which were known to have no-bleed regulators similar to farm taps),
3) delete regulator only stations in the low pressure range (0 to 100 psig inlet pressure), and
4) delete meter only stations.
The average of the 14 samples in the new transmission direct sales (farm tap & industrial meters) data set is used
for the emission factor.
The transmission company inter-connect meter stations were taken by sorting the tracer data set for M&R stations
with inlet pressures above 100 psig. Thirty-seven samples met this criterion.
Summary of Component Counts and Overall Emission Factors (scf/day)
METER STATION TYPE
SAMPLES
(Number of Tracer Measure-
ments Fitting this Type)
EMISSION FACTOR
(Methane SCFD)
Trans-Trans Co.
Interconnect points
37
3984 ± 80%
Farm Taps and direct industrial
sales
14
31.2 ± 80%
EF DATA SOURCES:
1. Tracers result based on downwind tracer measurements performed by AerodyneAVSU (1) at over
100 gas industry meter/regulation stations.
2. Analysis of tracer results was based upon technical descriptions of meter station types given by
several transmission company measurement experts.
3. Definition of transmission segment boundaries and other measurement programs shows that several
meter types have already been accounted for. See sheet D-l for sales to distribution M&R
stations, see sheet T-l and S-l for compressor station meter fugitive emissions and see sheet P-2
for production receipt meters which have already been accounted for at gas production sites.
C-10
-------
EF PRECISION:
Basis:
The transmission meter/pressure regulation station (M&PR) upper bounds are based upon rigorous
propagation of error from the standard deviation of the multiple tracer measurements.
ACTIVITY FACTOR:
Trans-to-trans company interconnects 2533 ± 776%
Farm taps and Direct Industrial Sales 72630 ± 780%
As discussed above, types 1,2, and 3 of transmission M&R stations are actually already accounted for in other
activity factors. In the production segment meter runs were counted in the well site data. Delivery to distribution
has been counted in the distribution segment M&R stations (i.e. city gates). There is also a trend to let LDCs
handle the farm taps, or to remove them entirely; however, many farm taps are still owned by transmission
companies.
Transfers to other transmission companies and farm taps were calculated from survey data provided by the metering
departments of three large (over 10,000 miles of pipeline) transmission companies, and from three companies with
fewer than 10,000 miles of pipeline, as shown in the following table.
Transmission M&R Station Populations
Company
Transfer to
another
Transmission Co.
Farm Taps
Direct Industrial
Sales
Miles of Pipeline
1
323
23
Confidential
2
5
0
Confidential
3
60
0
Confidential
4
62
48
Confidential
5
40
3,800
Confidential
6
0
10,000
Confidential
Total
490
13,871
658
55,045 (19.3% of
U.S. total)
Total U.S.
Activity Factor
Extrapolated by
Miles
2,533 ± 776%
71,690 ± 787%
937 ± 100%
284,500
Only five of the six companies that responded to the survey reported having interconnects with other transmission
companies. The activity factor was extrapolated based on pipeline miles and was calculated to be 2533 interconnects
(transfers). The 90% confidence bound was determined to be + 776%.
The count of farm taps appears to be extremely regional. Based on interviews, it seems that most companies have
no farm taps, while others have thousands. The activity factor for farm taps was calculated to be 71,690 ± 787%.
The calculated activity factor is believed to be conservatively high, since only a small percentage of all transmission
companies have these M&R stations, yet two of the six companies in our data set reported a large number of farm
taps.
C-ll
-------
The activity factor for direct industrial sales was developed from FERC Form No. 2, page 306 (2). Industrial
sales greater than 50,000 Mcf are listed individually, while sales less than 25,000 Mcf are combined into a
single item. In the latter case, the total amount of gas sold was divided by 50,000 to provide an estimate of the
number of sales. Due to the uncertainty that this approach introduced to the activity factor and to the
complexity of retrieving data from FERC, a confidence bound of + 100% was assigned based on engineering
judgement.
The activity factor for the direct industrial sales was combined with that for farm taps based upon similar
construction of the two station types.
AF DATA SOURCES:
1. For interconnects and farm taps, six transmission companies responded to the GRI/EPA survey to
determine average ratios of meter types per mile of transmission line. Averages from the survey
were extrapolated to national interconnect M&R number by multiplying the ratio by the known
miles of U.S. transmission line.
2. Miles of transmission line were from Gas Facts (3).
3. Direct industrial sales were determined from gas sales reported to FERC (2).
AF PRECISION:
Basis:
1. For interconnects and farm taps, rigorous propagation of error based upon the standard deviation of
the ratio data from individual transmission companies.
2. For direct industrial sales: An engineering estimate based upon interview data.
ANNUAL METHANE EMISSIONS: (4.5 Bscf ± 835%)
The annual emissions were determined by multiplying an emission factor for an each equipment type by the
population of equipment in the segment.
REFERENCES
1. 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, Chicago, IL, March 1995.
2. Federal Energy Regulatory Commission (FERC) Form No. 2, page 306: Annual Report of Major
Natural Gas Companies, 1992 database.
3. American Gas Association. Gas Facts: 1992 Data, Arlington, VA, 1993.
C-12
-------
T-3
TRANSMISSION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Transmission pipelines are the inter- and intrastate high pressure underground pipelines that transport natural gas
from the production/processing operations to the end user or distribution network. Leakage from underground
transmission lines occurs from corrosion pits, joint and fitting failures, pipe wall fractures, and external damage.
EMISSION FACTOR: (scf/leak-year)
Leak survey practices for transmission lines are generally more stringent than for distribution mains. Transmission
lines are required to be surveyed annually, and more frequently in populated areas. In addition, many transmission
companies perform additional routine aerial surveys to monitor the transmission lines for leakage. Based on
conversations with several transmission companies, any leaks found in the pipewall are extremely small and are
repaired immediately for safety reasons. Based on the rigorous leak survey and repair practices of transmission
companies (i.e., leaks are discovered and repaired earlier in transmission lines), the average leak rate from a
transmission leak is believed to be of the same order of magnitude as a leak found in a distribution main, even
though there may be a substantial difference in the operating pressure of the pipelines.
Therefore, the emission factors for leakage from transmission pipelines are based on the arithmetic average leakage
rates for main pipelines from the cooperative underground distribution leakage measurement program. A mean
value of the estimated leak rate per leak was calculated using the test data, for all pipe materials except cast iron.
For cast iron mains, a segment test approach was used which quantifies the leakage rate for a long isolated segment
of pipe; therefore, the mean leakage rate for cast iron is in terms of leakage per unit length of pipe. The natural
gas leak rate is adjusted for methane by multiplying by the volume percent of methane for transmission (93.4 vol.
%), and is adjusted for the soil oxidation of methane. The value of the emission factor and standard deviation for
each pipe material category is given below:
Pipe Material
Number of
Samples
Average
Emission
Factor
Units of
Emission
Factor
90%
Confidence
Interval of
Emission
Factor
Protected Steel
17
20,270
scf/leak-yr
17,243
Unprotected Steel
19
51,802
scf/leak-yr
48,212
Plastic
6
99,845
scf/leak-yr
165,617
Cast Iron
21
238,736
scf/mile-yr
152,059
Preliminary data from the underground distribution program indicate that the leakage rate is not a function of the
pipeline pressure. Therefore, the leakage rates for transmission pipelines have not been adjusted based on the
difference in average operating pressure of the transmission lines versus distribution lines.
Transmission Pipelines
Normal Operations
Unsteady, Fugitives (Pipeline Leaks)
0.16 Bscf +/- 89%
C-13
-------
EMISSION FACTOR DATA SOURCES:
1. Leakage rate data on a rate per leak basis for cathodically protected steel mains, unprotected steel
mains, and plastic mains from the cooperative leak measurement program.
2. Leakage rate data on a rate per unit length basis for cast iron mains from the cooperative leak
measurement program for distribution mains.
3. Assumes that the leak rates from transmission pipelines are identical to leak rates from distribution
mains, based on the more rigorous leak survey and repair practices of transmission companies.
4. Assumes that the leak rates from underground pipelines are independent of pressure and pipe
diameter, based on preliminary results from the underground distribution leak measurement
program.
ACTIVITY FACTOR: (equivalent leaks)
The mean activity factor and precision for each pipe material category is given below:
Pipe Material
Total Miles
Average Activity
Factor
Units of Activity
Factor
90% Confidence
Interval of
Activity Factor
Protected Steel
287,155
5,077
equivalent leaks
3,859
Unprotected Steel
5,233
659
equivalent leaks
501
Plastic
2,621
14
equivalent leaks
11
Cast Iron
96
96
miles
10
The number of total leaks (excluding pipeline incidents) in transmission pipelines is based on the 1991 DOT RSPA
database (1) for transmission pipelines, including both repaired leaks (6,120 leaks) and outstanding leaks (1,369
leaks). Because transmission lines are surveyed at least once per year using a walking survey method, the number
of unreported leaks is estimated based on the effectiveness of the walking survey. According to one contract
company specializing in distribution surveys, roughly 85 percent of the leaks are found using a walking survey.
This estimated survey efficiency was applied to transmission surveys, resulting in roughly 1,320 unreported leaks.
The leak duration for outstanding leaks and unreported leaks is estimated to be 8,760 hours per year, and the leak
duration for repaired leaks is half a year (4,380 hours/year), on average. The resulting estimate of equivalent leaks
represents the number of leaks with a year round leak duration. (That is, each leak repair is counted as half an
equivalent leak to compensate for the leak duration.) Therefore, the equation used to estimate equivalent leaks is:
0.5 x (repaired leaks) + {[(repaired leaks + outstanding leaks)/0.85] - repaired leaks}
The total number of estimated transmission pipeline leaks, 5,750, was allocated on a pipeline material category basis
in the same proportion (adjusted for the fraction of mileage in each material category) as in the distribution sector.
(That is, the ratio of percent leaks to percent miles in the transmission segment is the same as the ratio in the
distribution segment.) The precision of the estimated total leaks was calculated based on the estimated 90%
confidence interval associated with each parameter in the activity factor equation:
repaired leaks; outstanding leaks: ± 100%
leak duration: +25%
leak survey effectiveness: ± 15%
C-14
-------
A statistical software program [(@RISK (2)] was used to determine the overall 90% confidence interval of the
activity factor: ±76%.
For cast iron transmission lines, the mileage is based on the 1991 DOT RSPA database for transmission and
gathering lines. The precision of the estimate is assumed to be + 10%.
ACTIVITY FACTOR DATA SOURCES:
1. 1991 DOT RSPA database (1) for transmission and gathering pipelines.
2. Total number of leaks is assumed equal to the total number of leak repairs plus the outstanding
(unrepaired leaks) and unreported leaks.
3. Leak survey effectiveness estimation provided by Southern Cross Company (3).
4. The allocation of estimated leaks per pipe material category is based on the leak frequency for
underground distribution main pipelines, adjusted for the fraction of total mileage per pipe material
category.
5. @RISK statistical software program (2) used to estimate the 90% confidence interval.
ANNUAL METHANE EMISSIONS: (0.16 Bscf ± 89%)
Pipe Material
Average Emission
Factor
(scf/Ieak-yr)
Average Activity
Factor
(equivalent leaks)
Annual Emissions
Estimate,
(Bscf)
90% Confidence
Interval of
Emissions
Estimate,
(Bscf)
Protected Steel
20,270
5,077
0.10
0.14
Unprotected Steel
51,802
659
0.03
0.05
Plastic
99,845
14
0.001
0.003
Cast Iron
238,736a
96"
0.02
0.02
Total
0.16
0.14
ascf/mile-yr
bmiles
The total leakage was determined by multiplying an emission factor for each type of pipeline material by the
estimated number of leaks in each respective pipe material category.
REFERENCES
1. U.S. Department of Transportation. Research and Special Programs Administration. 1991.
2. Palisade Corporation. @ Risk, Risk Analysis and Simulation Add-in for Lotus 1-2-3, Version 1.5, March
1989.
3. Southern Cross Corporation. Comments on Docket PS-123 Notice 1, Leakage Surveys, 49 CFR Part 192,
Department of Transportation, Research and Special Programs Administration, Materials Transportation
Bureau, Office of Pipeline Safety Regulations, December 19, 1991.
C-15
-------
T-4
TRANSMISSION AND STORAGE SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
COMPONENTS:
ANNUAL EMISSIONS:
BACKGROUND:
The transmission segment is comprised of compressor stations, pipelines, and storage stations. There are
essentially no pneumatic devices associated with the pipelines. Within the storage and compressor stations, most
of the pneumatics are gas-actuated isolation valves, and there are a few continuous bleed controllers.
Meter-only stations do not have venting pneumatics. Meter and regulation (M&R) stations do have regulating
pneumatic controllers (the pressure regulator valves), but all of the M&R station pneumatic emissions are counted
in the fugitive calculation for M&R stations and so are not included in this sheet.
The continuous bleed controllers in transmission compressor stations are used for liquid level control in filter-
separators and pressure reduction. The higher pressures and large pipe diameters associated with transmission
operations require larger actuators and valves than typically found in production, resulting in larger emissions than
similar devices in production.
Within the storage and mainline compressor stations, most of the pneumatic devices are gas-actuated isolation
valves. These valves block the flow to or from a pipeline and can isolate the facility for maintenance work or
in the case of an emergency. Therefore, the isolation valves are actuated infrequently and their emissions are
intermittent.
EMISSION FACTOR: 162,197 scf/device ± 44%
(This was adjusted for the transmission methane fraction of natural gas at 93.4 mol%.)
The average pneumatic device emission factor was determined from a compilation of information from several
sites. Counts of devices per site were taken during Radian site visits. The devices were classified into three
categories: continuous bleed valves, isolation valves with turbine operators, and isolation valves with displacement
operators. The emission factor was determined based on the following equation:
pneumatic devices — ( cont. bleed valves X Fraction bl«d valves ~*~
turbine operators * Fraction t.jrt)]ne operators
displacement operators * Fraction djSp|accmeru operators ) * methane
Listed below are the average fraction of devices for each of the three valve categories:
Fraction blecd valves = 0.32 ± 69%
Fraction ^ine operators = 0.16 ±94%
Fraction disp,acOTen, operatore = 0.52 ± 48%
Emissions from continuous bleed pneumatics in the transmission segment were measured by an independent
contractor. The average emission factor, based on 23 measurements, is 1,363 scfd/device ± 29% (497,584
scf/device).
Various Equipment (vessels, compressors, piping)
Normal Operation
Unsteady, Vented
Pneumatic Devices
14.1 Bscf ± 60%
C-16
-------
For the isolation valves with turbine operators, the emission factor depends on the gas usage for a given supply
gas pressure, the time required to complete one movement of the valve, and the number of operations per year.
The annual emission factor is then:
EF 0pera,0I3 = Gas Usage (scf/min) x Operating Duration (min/operation) x 2
(operations/cycle) x Frequency (cycles/year)
EF wAmc optraIore = 67,599 ± 276% scf/device
The equation for isolation valves with displacement operators is similar:
EF displacement opera,ore = Gas UsaSe (sctfpsia) x Supply Pressure (psia) X 2
(operations/cycle) x Frequency (cycles/year)
EF displacCTiem operatoIJ = 5,627 ± 112% scf/device
EF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 12: Pneumatic Devices (1)
establishes the important emission-affecting characteristics of transmission pneumatic devices.
2. Device counts from 16 compressor and storage stations establish the fraction of turbine valve
operators, and displacement valve operators. Counts from a total of 54 stations were used to
establish the fraction of continuous bleed devices.
3. The emission factor for the continuous bleed valves was based on 23 field measurements.
4. Gas usages for the turbine valve operators were provided by Limitorque. Operating duration
and frequency were estimated based on information from two transmission stations.
5. Gas usages for the displacement valve operators were provided by Shafer Valve Operating
Systems. Supply pressure and frequency of operation were estimated based on information from
four transmission stations.
EF ACCURACY:
Basis:
1. EF accuracy is based on error propagation from the combination of site information and
measured data.
2. It was assumed that the manufacturers' data are completely accurate.
ACTIVITY FACTORS: 87,206 pneumatic devices ± 38%
The number of gas operated pneumatic devices in the transmission and storage segment was calculated based on
the average number of devices per station and multiplied by the total number of transmission and storage stations
nationally. The average number of devices per site was determined to be 40 ± 37%. The total count of
transmission compression facilities is 2,175, based on 1,700 compressor stations, 386 UG storage stations, and
89 LNG storage stations.
AF DATA SOURCES:
1. The number of transmission compressor stations was compiled from 1992 Fossil Energy
Commission Form No. 2: Annual Report of Major Natural Gas Companies (2).
2. The number of underground storage facilities is taken directly from A.G.A. Gas Facts: "Number
of Pools, Wells, Compressor Stations, and Horsepower in Underground Storage Fields." Data
from base year 1992 were used (3).
3. The number of liquefied natural gas storage facilities was summed from A.G.A. Gas Facts,
C-17
-------
"Liquefied Natural Gas Storage Operations in the U.S. as of December 31, 1987 (4)." The table
lists 54 complete plants, 32 satellite plants, and 3 import terminals for a total of 89 facilities.
4. The number of devices per site is based on the total number of devices observed during site
visits.
AF ACCURACY: 38%
Basis:
1. Extremely tight confidence limits are expected due to the well documented and reviewed
numbers published in A.G.A. Gas Facts and FERC forms. A 10% confidence bound was
assigned to the number of compressor stations and a 5% confidence bound was assigned to the
number of storage stations.
2. The confidence bound on the number of devices per station was determined based on the spread
of site data.
ANNUAL METHANE EMISSIONS: 14.1 Bscf ± 60 %
The annual emissions were determined by multiplying an emission factor per device (corrected for the methane
composition) by the population of pneumatic devices in the transmission segment.
162,197 scf/device x 87,206 devices = 14.1 Bscf
REFERENCES
1. 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.
2. Department of Energy. FERC Form No. 2: Annual Report of Major Natural Gas Companies. OMB
No. 1902-0028, Department of Energy Federal Energy Regulatory Commission, Washington, DC,
December 1994.
3. American Gas Association. Gas Facts:. 1993 Data, Arlington, VA, 1994.
4. American Gas Association. Gas Facts:. 1991 Data, Arlington, VA, 1992.
C-18
-------
T-5
TRANSMISSION AND STORAGE SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Maintenance activities can release gas to the atmosphere through blowdown or through purge. Blowdown is the
direct, intentional venting to the atmosphere of gas contained inside operating equipment. The gas is released
to provide a safer working environment for maintenance activities around or inside the equipment. After the
equipment is serviced, the oxygen inside the equipment is often cleared to the atmosphere by purging natural gas
through the equipment.
Upsets can also emit gas directly to the atmosphere. Upsets in process conditions can cause pressure rises that
exceed the maximum design pressure for equipment. To prevent equipment overpressure and damage, pressure
relief valves (PRVs) or remotely actuated valves open and vent the excess gas to the atmosphere. PRVs are
spring loaded or pilot actuated valves that are designed to handle the upset conditions. Remotely actuated valves
are usually designed to vent entire compressor stations or areas (such as compressor piping) in the event of a
station emergency such as a fire or a large gas release.
EMISSION FACTORS: Station Blowdowns 4,359 ± 262% Mscf/station
Pipeline Blowdowns 31.6 ± 236% Mscf/mile
(Corrected for the transmission methane composition of 93.4 mol%)
Company tracked data were available from either company gas use estimates reported to accounting departments
from each site (accounted-for), or from special "unaccounted-for" studies that searched for unmetered company
gas use. Most of the company data could be separated into two event types: station blowdowns (includes
compressor blowdowns, compressor starts, PRV lifts, ESD activation, and other venting sources) and pipeline
blowdowns. These data are summarized in the following table.
EF DATA SOURCES:
1. GRI/EPA Methane Emissions from the Natural Gas Industry, Volume 7: Blow and Purge
Activities (1) establishes emission affecting characteristics of blowdown practices.
2. Company tracked data were available from 8 companies.
EF ACCURACY: Range ± 236% to 262%
Basis:
The accuracy was calculated from the spread of the company data. A 90% confidence interval is
calculated for the 8 companies using the method presented in the Methane Emissions from the Natural
Gas Industry, Volume 4: Statistical Methodology (2).
Various Equipment
Maintenance/Upsets
Unsteady, Vented
18.5 Bscf ± 177%
C-19
-------
Company
Annua! Station
Blowdown
Emissions,
Mscf
Annua!
Pipeline
Blowdowns,
Mscf :
Total Annual
Blowdowns,
: Mscf
Total
Number of
Stations
Total Number
of
Pipeline Miles
1
120,757
189,044
309,801
11
3,857
2
272,589
11,358
283,947
15
4,000
3
33,731
138,988
172,719
27
5,886
4
-
-
172,776
(19)°
(5,450)
5
325,418
Unknown
Unknown
47
(4,725)
6
Unknown
161,628
Unknown
(48)
7,896
7
60,956
750,000
810,956
69
14,666
8
194,541
315,058
509,599
47
9,915
TOTALS
1,007,992
1,566,076
216
46,220
ANNUAL AVERAGE, Mscf natural gas/station 4,667 ± 262%
ANNUAL AVERAGE, Mscf natural gas/mile 33.9 ± 236%
'Parentheses indicate that the value was not included in the total because a station or pipeline emission rate
was not available.
ACTIVITY FACTORS: 2,175 ± 8% compression facilities
284,500 ± 5% transmission pipeline miles
The activity factors for the segment were compiled from published statistics on the gas industry. The total
count for transmission compressor stations was 1700; the total underground and liquefied natural gas storage
station count was 475. The number of transmission pipeline miles comes from A.G.A. Gas Facts (3) which
shows 284,500 miles of pipeline in the United States for 1992.
AF DATA SOURCES:
1. The number of transmission compressor stations was compiled from FERC Form No. 2:
Annual Report of Major Natural Gas Companies (4).
2. The number of underground storage facilities is taken directly from A.G.A. Gas Facts, Table
4-5, "Number of Pools, Wells, Compressor Stations, and Horsepower in Underground
Storage Fields" (3).
3. The number of liquefied natural gas storage facilities was summed from A.G.A. Gas Facts,
Table 4-3, "Liquefied Natural Gas Storage Operations in the U.S. as of December 31, 1987"
(3). The table lists 54 complete plants, 32 satellite plants, and 3 import terminals for a total
of 89 facilities.
4. The number of transmission pipeline miles comes from A.G.A. Gas Facts which shows
284,500 miles of pipeline in the U.S. for 1992 (3).
AF ACCURACY: Range ± 5% to 8%
Basis:
Extremely tight confidence limits are expected due to the well documented and reviewed DOE
numbers published in A.G.A. Gas Facts (3).
C-20
-------
ANNUAL METHANE EMISSIONS: 18.5 Bscf ± 177%
The annual methane emissions were determined by multiplying an emission factor (rate per avg unit) for each
category by the activity factor (population) of the category. Each emission factor was adjusted for the average
methane content in the transmission segment of 93.4 mol%.
REFERENCES
1. 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.
2. 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.
3. American Gas Association. Gas Facts: 1992 Data, Arlington, VA, 1993.
4. Department of Energy. FERC Form No. 2: Annual Report of Mayor Natural Gas Companies. OMB
No. 1902-0028, Department of Energy, Federal Energy Regulatory Commission, Washington, DC,
December 1994.
C-21
-------
T-6
TRANSMISSION SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
COMPONENTS:
ANNUAL EMISSIONS:
Glycol Dehydrators
Normal Operation
Unsteady, Vented
Reboiler Vents
0.10 Bscf ± 392%
BACKGROUND:
Glycol dehydrators remove water from a gas stream by contacting the gas with glycol and then driving the water
from the glycol and into the atmosphere. The glycol also absorbs a small amount of methane, and some methane
can be driven off to the atmosphere through the reboiler vent.
EMISSION FACTOR: (93.72 scf/MMscf gas processed ± 207.99%)
A thermodynamic computer simulation was used to determine the most important emission-affecting variables
for dehydrators. The variables are: gas throughput, existence of a flash tank, existence of stripping gas, existence
of a gas driven pump, existence of vent controls routed to a burner. Throughput, since its effect is linear, is
handled by establishing an emission rate per gas throughput. Emission rates per throughput are then established
for the other important emission affecting characteristics. The emission factor is then:
EF = [ ( Fft x EFp,. ) + ( Fnt x EF^ ) + ( FSG x EFSG ) ] x FNIVC x OC
EF = [ (0.669 x 3.57) + (0.331 x 175.10) + (0.0741 x 670) ] x 0.852 x 1.0
Fn =
Fraction of the population WITH flash tanks
0.669 ± 9.70%
Fnt =
Fraction of the population WITHOUT flash tanks
0.331 ± 19.6%
Fsg =
Fraction of the population WITH stripping gas
0.0741 ± 118.26%
FnVC=
Fraction of the population WITHOUT combusted vent controls
0.852 ± 14.0%
efft=
Total CH„ emission rate per 1 MMscf throughput for dehydrator that has a flash tank
3.57 scf/MMscf (+102% / -58%)
EF^
Total CH4 emission rate per 1 MMscf throughput for dehydrator that does NOT have
a flash tank
175.1 scf/MMscf (+101% / -50%)
EFsg= Incremental emission rate per 1 MMscf throughput for dehydrator that has stripping gas
670 scf/MMscf (+40% / -60%)
OC = Overcirculation factor for glycol-number of times the industry rule-of-thumb of 3
gallons glycol/lb water
1.0 ± 0%
C-22
-------
EF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators (1)
establishes emission affecting characteristics of dehydrators.
2. Site visit data establishes the FSG and for multiple sites. Wyoming ADQ data also
verifies F^, though it implies a higher F, and thus a higher overall EF.
3. TMOGA/GPA survey of 1019 dehydrators established Ffd and F^ and TP for dehydrators.
4. ASPEN computer simulations were used to determine EF^, and EF^ from the dehydrator
vent.
5. Sampling data from the GRI Glycol Dehydrator Sampling and Analytical Program for one
dehydrator was used to determine EFSG (1). The upper bound was calculated by assuming
that all of the measured noncondensable vent gas was due to stripping gas that was 100%
methane. The lower bound was calculated as the rule-of-thumb stripping gas rate
recommended by a glycol dehydrator manufacturer.
EF ACCURACY: 93.72 scf/MMscf ± 207.99%
Basis:
The accuracy is propagated through the EF calculation from each term's accuracy:
1. ASPEN has been demonstrated to match actual dehydrators within ±20% within the
calculated confidence intervals obtained from site data.
2. Individual EF confidence intervals were calculated based upon the spread of the site
averages.
ACTIVITY FACTOR: (1.086 Tscf/year gas throughput in the transmission segment)
The amount of gas processed by glycol dehydrators in the transmission segment was calculated from the
estimated number of glycol dehydrators in transmission service and the average throughput capacity for
transmission dehydrators (Wright Killen & Co., 1994). See Source Sheet P-6 for a detailed discussion of the
breakdown of glycol dehydrators into industry segments. The capacity utilization factor for transmission was
assumed to be 1.
AF ACCURACY: 1.086 Tscf/year ± 143.85%
Basis:
1. Uncertainty based on confidence limits from the site visit data.
ANNUAL METHANE EMISSIONS: (0.1018 Bscf/yr ± 391.75%)
The annual methane emissions were determined by multiplying the dehydrator emission factor by the activity
factor.
REFERENCES
1. Myers, D. Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators, Final
Report, GRJ-94/0257.31 and EPA-600/R-96-080n. Gas Research Institute and U.S. Environmental
Protection Agency, June 1996.
2. Wright Killen & Co. Natural Gas Dehydration: Status and Trends, Final Report, Gas Research
Institute, GRI-94/0099, January 1994.
C-23
-------
S-l
STORAGE SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Equipment leaks are typically low-level, unintentional losses of process fluid (gas or liquid) from the sealed
surfaces of above-ground process equipment. Equipment components that tend to leak include valves, flanges
and other connectors, pump seals, compressor seals, pressure relief valves, open-ended lines, and sampling
connections. These components represent mechanical joints, seals, and rotating surfaces, which in time tend
to wear and develop leaks.
EMISSION FACTOR: a. Station = 7.85 MMscf/yr methane per station
b. Wellhead = 41.8 Mscf/yr methane per wellhead
c. Recip. Compressor = 7.71 MMscf/yr methane per recip
d. Turbine Compressor = 11.16 MMscf/yr methane per turbine
The average fugitive emission rate for storage facilities was determined to be composed of three parts: a)
storage compressor station components (excluding compressor-related components), b) injection/withdrawal
wellhead components, and c) compressor-related components. Fugitives from the compressor-related
components have much higher emission factors than components in the rest of the facility. This is due in part
to the high vibration that compressors generate, but most of the larger emissions are due to unique compressor
components as explained below.
a) The contribution from non-compressor components was determined by multiplying the average number of
fugitive components by the component emission factor. The number of components was subdivided into
valves, connections/flanges, small open-ended lines, and other components (such as pressure relief valves);
tubing components were determined to be insignificant. All of these components are typical fugitive
components (as described in the EPA Fugitive Emissions Protocol) with the exception of site blowdown
(B/D) open-ended lines (OELs). Site B/D OELs are the large diameter emergency station blowdown valves
that are designed to depressure the entire site to the atmosphere when the valve is opened. Emission factors
for storage station components were based on the GRI/Indaco program at 6 transmission compressor station
sites.
b) The contribution from storage injection/withdrawal wells was determined in the same manner as storage
compressor stations (see below). Emission factors for storage injection/withdrawal wells were based on the
updated API/GRI/Star 20-site study (4 gas production sites). Physical and operational characteristics of
injection/withdrawal wells were compared to gas production wells, and were found to be similar but typically
larger (more components). This was taken into account in the component count data.
The number of components was subdivided into types, such as valves, connections/flanges, open-ended lines,
and other components (such as pressure relief valves). The average facility/equipment emissions are
calculated as follows:
EF = [(NvIv x EFV|V) + (Ncn x EFJ + (Noel x EFoel) + (N0[h x EFoth) + (Npiv x EFprv) + (NsiteB/D x EFsiKB/D)]
where:
Nx = average count of components of type x per plant, and
EF„ = average methane emission rate per component of type x.
Storage Facilities (Compressor Stations and Wells)
Normal Operation
Steady and Unsteady, Fugitive
16.76 Bscf ± 57%
C-24
-------
c) The contribution from compressor-related components was obtained by multiplying the average number of
fugitive components per compressor engine by the component emission factors. The component emission
factors were based on the GRI/Indaco measurement program conducted at 15 compressor stations. Some
compressor components are unique, while others have higher leak rates than identical components elsewhere
in the plant due to vibration. Compressors have the following types of components:
1) Comp. B/D OEL A blowdown (B/D) valve to the atmosphere that can depressure the compressor
when idle. The B/D valve or the large unit block valves (depending on the
operating status of the compressor) can act as an open-ended line that leaks at an
extraordinarily high rate through the valve seat. The leak rate is dependent upon
whether the compressor is pressurized (in operation or idle, pressurized) or
depressurized (idle, depressurized).
2) Comp. PRV The pressure relief valve (PRV) is usually installed on a compressor discharge line
and leaks at a higher than average rate due to vibration.
3) Comp. Starter OEL Most compressors have a gas starter motor that turns the compressor shaft to start
the engine. Some use natural gas as the motive force to spin the starter's turbine
blades and vent the discharge gas to the atmosphere. The inlet valve to the starter
can leak and is therefore an OEL unique to compressors.
4) Comp. Seal All compressors have a mechanical or fluid seal to minimize the flow of pressurized
natural gas that leaks from the location where the shaft penetrates the compression
chamber. These seals are vented to the atmosphere. Reciprocating compressors
have sliding shaft seals while centrifugal compressors have rotating shaft seals.
5) Miscellaneous There are many components on each compressor, such as valve covers on
reciprocating compressor cylinders and fuel valves.
Each compressor has one B/D OEL, one PRV, and one starter OEL. Reciprocating compressors have one
compressor seal per compression cylinder (which averaged 4.5 per engine), while centrifugal compressors
have 1.5 seals per gas turbine. For the miscellaneous component category, there are many components per
compressor engine, but the emission rates were minor and so were added into one lump emission factor per
compressor for miscellaneous components.
All of the compressor emission factors take several correction factors into account. First, the various phases
of compressor operations (such as the amount of time that compressors are a) idle and depressured, b) idle
and pressured up, or c) running). This is actually a complex adjustment that takes into account valve position
practices. [See Methane Emissions from Natural Gas Industry, Volume 8: Equipment Leaks (1) for more
details.] Correction factors were also added for fraction of starter gas turbines using air instead of gas (40%
for recip, 50% for turbines in storage).
EF DATA SOURCES:
1. Emission Factors for storage compressor stations are based upon GRI/Indaco transmission
compressor station fugitive leak measurement surveys at 6 compressor stations. Compressor
operating hours (% running) based on data from 5 national gas storage companies.
2. Component counts for storage compressor stations and injection/withdrawal wellheads are
based on Radian site visits to 5 storage facilities.
3. Component emission factors for compressor-related components based on GRI/Indaco
transmission compressor station fugitive leak measurement program at 15 compressor
stations.
4. Wellhead emission factors based on simple average of GRI/Star data for gas production
wellheads (Atlantic/Eastern region and Rest of U.S.).
5. Fraction of methane (93.4 mol%) based on data from GRI TRANSDAT database.
C-25
-------
Average Facility Emissions for Gas Storage
Equipment Type
Component Type
Component
Emission Factor,
Mscf/component-yr
Average Component
Count
Average Equipment
Emissions,2
MMscf/yr
Storage Facility
Valve
0.867
1,868
7.85 (100%)
(non-compressor
related components)
Connection
0.147
5,571
OEL
11.2
353
PRV
6.2
66
Site B/D OEL
264
4
Injection/Withdrawa
Valve
0.918
30
0.042 (76%)
1 Wellhead
Connection
0.125
89
OEL
0.237
7
PRV
1.464
1
Reciprocating
Compressors
Compressor B/D
OEL
5,024b
1
7.71 (48%)
PRV
317b
1
Miscellaneous
153b
1
Compressor Starter
OEL
1,440
0.6C
Compressor Seal
300"
4.5
Centrifugal
Compressors
Compressor B/D
OEL
10,233"
1
11.16 (34%)
Miscellaneous
17b
1
Compressor Starter
OEL
1,440
0.5C
Compressor Seal
126b
1.5
a Values in parentheses represent 90% confidence interval.
b Adjusted for the fraction of time the compressor is pressurized (67.5% and 22.4% for reciprocating and
centrifugal compressors, respectively).
c Adjusted for the fraction of compressor starters using natural gas (60% and 50% for reciprocating and
centrifugal compressors, respectively).
C-26
-------
EF ACCURACY: a.
b.
b.
c.
Station = + 100%
Wellhead = ± 76%
Recip. Compressor = ± 48%
Turbine Compressor = + 34%
Basis:
Rigorously propagation of error from the spread of thousands of individual measurements taken by
Indaco and Star.
ACTIVITY FACTOR a. Station Activity Factor = 475 stations
b. Wellhead Activity Factor = 17999 wellheads
b. Compressor Activity Factor = 1396 recip compressors, 136 turbines
The activity factors for the segment were compiled from published statistics in Gas Facts (2). The total count
for Underground storage stations was 386, and the total LNG storage count was 89.
AF DATA SOURCES:
1. The number of underground storage facilities was taken directly from A.G. A. Gas Facts, (2),
Table 4-5: Number of Pools, Wells, Compressor Stations, and Horsepower in Underground
Storage Fields. Data from base year 1992 were used.
2. The number of Liquefied Natural Gas Storage Facilities was summed from A.G.A. Gas
Facts (2), Table 4-3, "Liquefied Natural Gas Storage Operations in the U.S. as of December
31, 1987." The table lists 54 complete plants, 32 satellite plants, and 3 import terminals for
a total of 89 facilities.
3. Compressor engine count based on GRI TRANSDAT "industry database" with adjustments
for total industry horsepower. Storage site visits to 8 storage sites provided number of
reciprocating engines and turbines per site [see Activity Factor Report (3)]. Also, the number
of reciprocating compressors in storage was increased by 31% to account for electric motor
drivers.
AF ACCURACY: a. Station Activity Factor: + 5%
b. Wellhead Activity Factor: ± 5%
b. Compressor Activity Factor: Recip engines = ± 58 %; Turbines = ± 119 %
Basis:
1. A.G.A. Gas Facts (2) has a high percentage of all storage facilities represented in Tables 4-5
and 4-3. Therefore a national extrapolation should not add much error. This 5% figure was
assigned based on engineering judgement.
2. The compressor count accuracy was assigned based upon the propagation from: a. Rigorous
error propagation for the 8 storage station "compressor/station" averages; and b. Engineering
judgement assignment of ± 10% error to the large GRI TRANSDAT database.
ANNUAL METHANE EMISSIONS: (16.76 Bscl/yr ± 9.6 Bscf/yr)
The annual emissions were determined by multiplying an emission factor for an average equipment type by
the population of equipment in the segment.
C-27
-------
Category
Emission Factor
Activity Factor
Emission Rate
Uncertainty
Station
7.85 MMscf/yr CH4
475 stations
3.73 Bscf/yr CH4
100%
Inj/With Wellheads
41.8 Mscf/yr CH4
17999 wellheads
0.752 BscCyr CH4
76%
Recip Comp
7.71 MMscCyr CH4
1396 recip
10.76 Bscf/yr CH4
80%
Turbine Comp
11.16 MMscf/yr CH4
136 turbine
1.52 Bscf/yr CH4
129%
TOTAL
16.76 Bscf/yr CH4
57%
REFERENCES
1. 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.
2. American Gas Association. Gas Facts, Arlington, VA. 1992.
3. 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.
C-28
-------
S-2
STORAGE SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
COMPONENTS:
ANNUAL EMISSIONS:
Glycol Dehydrators
Normal Operation
Unsteady, Vented
Reboiler Vents
0.23 Bscf ± 167%
BACKGROUND:
Glycol dehydrators remove water from a gas stream by contacting the gas with glycol and then driving the
water from the glycol and into the atmosphere. The glycol also absorbs a small amount of methane, and
some methane can be driven off to the atmosphere through the reboiler vent.
EMISSION FACTOR: (117.18 scf/MMscf ± 159.76%)
A thermodynamic computer simulation was used to determine the most important emission-affecting variables
for dehydrators. The variables are: gas throughput, existence of a flash tank, existence of stripping gas,
existence of a gas-assisted pump, existence of vent controls routed to a burner. Throughput, since its effect is
linear, is handled by establishing an emission rate per gas throughput. Emission rates per throughput are then
established for the other important emission affecting characteristics. The emission factor is then:
EF = [ ( Fft x EF^ ) + ( Fnt x EF) + ( FSG x EFSG ) ] x F,^. x OC
EF = [ (0.520 x 3.57) + (0.480 x 175.10) + (0.080 x 670) ] x 0.840 x 1.0
Fpr = Fraction of the population WITH flash tanks
0.520 ± 33.56%
F^ = Fraction of the population WITHOUT flash tanks
0.480 ± 36.25%
Fso = Fraction of the population WITH stripping gas
0.080 ± 118.44%
Fnvc= Fraction of the population WITHOUT combusted vent controls
0.840 ± 15.24%
EFft= Total CH4 emission rate per 1 MMscf throughput for dehydrator that has a flash
tank
3.57 (+102% / -58%)
EF^ Total CH4 emission rate per 1 MMscf throughput for dehydrator that does NOT
have a flash tank
175.10 (+101% / -50%)
EFSC= Incremental emission rate per 1 MMscfd throughput for dehydrator that has stripping
cr^s
670 (+40% / -60%)
OC = Overcirculation factor for glycol—number of times the industry rule-of-thumb of 3
gallons glycol/lb water
1.0 ± 0%
C-29
-------
EF DATA SOURCES:
1. Methane Emissions from the Natural Gas Industry, Volume 14: Glycol Dehydrators (1)
establishes emission affecting characteristics of dehydrators.
2. Site visit data establishes the FSG and F^ for multiple sites. Wyoming ADQ data also
verifies F^, though it implies a higher F, and thus a higher overall EF.
3. TMOGA/GPA survey of 1019 dehydrators established F3P and F^ and TP for dehydrators.
4. ASPEN computer simulations were used to determine EF3P, and EFNT) from the dehydrator
vent.
5. Sampling data from the GRI Glycol Dehydrator Sampling and Analytical Program for one
dehydrator was used to determine EFSG (1). The upper bound was calculated by assuming
that all of the measured noncondensable vent gas was due to stripping gas that was 100%
methane. The lower bound was calculated as the rule-of-thumb stripping gas rate
recommended by a glycol dehydrator manufacturer.
EF ACCURACY: 117.18 ± 159.76%
Basis:
The accuracy is propagated through the EF calculation from each term's accuracy:
1. ASPEN has been demonstrated to match actual dehydrators within ±20% within the
calculated confidence intervals obtained from site data.
2. Individual EF confidence intervals were calculated based upon the spread of the site
averages.
ACTIVITY FACTOR: (2.00 Tscf/year gas throughput in the storage segment)
The amount of gas processed by glycol dehydrators in the storage segment was calculated from the estimated
amount of gas withdrawn from underground storage. A total of 2.4 Tscf was withdrawn in 1992, and it is
assumed that most stored gas is dehydrated.
AF ACCURACY: 2.00 Tscf/year ± 25%
Basis:
1. Uncertainty based on estimate of confidence limits.
ANNUAL METHANE EMISSIONS: (0.2344 Bscf ± 166.56%)
The annual methane emissions were determined by multiplying the dehydrator emission factor by the activity
factor.
REFERENCES
1. 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.
C-30
-------
APPENDIX D
Distribution Source Sheets
D-l
-------
APPENDIX D
Distribution Source Sheets
Page
D-l - Meter and Pressure Regulating Stations D-4
D-2 - Pipeline Leaks D-7
D-3 - Maintenance/Upsets (PRV) D-ll
D-4 - Pipeline Dig-ins D-l2
D-5 - Customer Meters D-l4
D-6 - Pipeline Blowdown D-l7
D-2
-------
DISTRIBUTION SOURCE SHEETS
This section contains the specific source sheets for the distribution segment of the natural
gas industry. The following table serves as a guide for finding sheets in this section. The
cells in the table give the sheet number (D-l, D-2, etc.) of the source sheet. The rows
define the equipment covered, while the columns define the emission type. A category
with no sheet number means that the emissions from that area were determined to be zero
or negligibly small. The label for each source sheet is shown at the top of the cover page
for that sheet.
TABLE OF
CONTENTS
OPERATING MODE,
EMISSION TYPE (Fugitive, Vented, or Combusted)
EQUIPMENT:
Start Up
Normal Operations
Maintenance
Upsets
Mishaps
V
C
F
V
C
V
C
V
C
V
Meter/Pressure
Regulating Stations
D-l
D-l
D-3
D-3
Main and Service
Pipelines
D-2
D-6
D-4
Customer Meters
D-5
D-3
-------
D-l
DISTRIBUTION SEGMENT SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Metering/pressure regulating stations are located throughout the distribution network to meter gas where a
custody transfer occurs and/or to reduce and regulate the pressure in the downstream main pipeline.
Emissions from fugitive losses and normal operations at meter and pressure regulating stations include both
continuous and intermittent emissions from equipment components, such as pneumatic devices, valves,
flanges, flow meters, and pressure regulators.
EMISSION FACTOR: (scf/station-hour)
The emission factor and standard deviation are given below for facilities located in vaults and above ground
for different inlet pressure ranges:
Station
Type
Inlet
Pressure
(psig)
Location in
Vault?
Number of
Samples
Average
Emission
Factor
(scf/sta.-hr)
Standard
Deviation
of Emission
Factor
(scf/sta.-hr)
Precision
of
Emission
Factor
(scf/sta.-hr)
M&R
>300
N
31
179.8
236.1
69.8
M&R
100-300
N
6
95.6
130.6
107.4
M&R
<100
N
3
4.3
5.8
9.8
Regulating
>300
N
13
161.9
188.8
93.3
Regulating
>300
Y
4
1.3
2.0
2.4
Regulating
100-300
N
7
40.5
36.4
26.7
Regulating
100-300
Y
10
0.2
0.3
0.2
Regulating
40-100
N
7
1.0
1.1
0.8
Regulating
40-100
Y
8
0.1
0.1
0.1
Regulating
<40
Y
6
0.1
0.2
0.2
The emission factors were derived from data collected using a tracer gas measurement method. Downwind
tracer measurements were performed by Aerodyne/Washington State University at 2 West Coast companies, 3
northeastern companies, 4 midwestern towns, and 3 southern plains towns. In total, 95 measurements were
performed on metering/regulating stations in distribution and transmission systems.
Meter/Pressure Regulating Stations
Normal Operations
Steady, Fugitive
27.3 +/- 23~3 Bscf
D-4
-------
The test data were analyzed to evaluate the differences in emissions from stations with different
configurations (i.e., metering/regulating versus regulating only), inlet pressure ranges, and locations (i.e., in
vaults versus above-ground). The test data were disaggregated into four distinct inlet pressure categories
(>300 psig, 100-300 psig, 40-100 psig, and <40 psig), two station types (meter/pressure regulating facilities
and pressure regulating facilities), and into stations in vaults versus above-ground, resulting in a total of 10
categories. These categories were selected for disaggregation of the data based on knowledge of the gas
industry, and were confirmed to be statistically significant based on the data analyses.
ACTIVITY FACTOR: (number of stations)
The mean activity factor and standard deviation for each station type/inlet pressure/location category is given
below:
Station
Type
Inlet
Pressure
(psig)
Location in
Vault?
Stations per
Mile
Average
Activity
Factor
(stations)
Standard
Deviation
of Activity
Factor
(stations)
Precision
of Activity
Factor
(stations)
M&R
>300
N
0.004
3,460
3,965
2,458
M&R
100-300
N
0.016
13,335
22,728
14,091
M&R
<100
N
0.009
7,127
13,550
8,401
Regulating
>300
N
0.005
3,995
4,946
2,702
Regulating
>300
Y
0.003
2,346
2,905
1,587
Regulating
100-300
N
0.015
12,273
13,656
7,461
Regulating
100-300
Y
0.007
5,514
6,136
3,352
Regulating
40-100
N
0.043
36,328
42,785
23,375
Regulating
40-100
Y
0.039
32,215
37,942
20,729
Regulating
<40
Y
0.018
15,377
18,161
9,922
The number of stations in each inlet pressure/station type category were provided by twelve distribution
companies. The data were extrapolated based on the total mileage of distribution main pipeline in the
respective companies. The mean number of stations in each category per mile of main was estimated as the
average of the values from eleven of the twelve companies supplying data. Based on conversations with one
of the companies supplying data, the average number of stations per mile for the one company were not
considered representative of typical industry practices. Therefore, this company was not included in the
overall average, but rather was treated separately. The standard deviation represents the variation in the
estimated number of stations per mile of main pipeline for each company. The precision represents the 90%
confidence interval around the estimated mean activity factor.
The extrapolation from stations per mile to total stations in the U.S. was implemented by multiplying the
stations per mile for each category by the total U.S. mileage of main pipeline: 836,760 miles.
Data were collected from five companies representing urban, rural, and suburban areas on the number of
regulating stations in vaults versus above-ground in the U.S. On average, 37% of the regulating stations with
D-5
-------
an inlet pressure greater than 300 psig are located in vaults. For regulating stations with an inlet pressure
between 40 and 300 psig, it was found that the majority of stations in urban areas were in vaults and in rural
areas were above-ground. On average, it was estimated that 31% of the stations are located in vaults with an
inlet pressure between 100 and 300 psig. For regulating stations with an inlet pressure between 40 and 100
psig, 47% of the stations are located in vaults. Based on the data collected, the majority of the low pressure
(<40 psig inlet pressure) stations are located in vaults.
ANNUAL EMISSIONS ESTIMATE: (27.3 +/- 23.3 Bscf)
Station
Type
Inlet
Pressure
(psig)
Location
in Vault?
Average
Activity
Factor
(stations)
Average
Emission
Factor
(scf/sta.-hr)
Annual
Emissions
Estimate
(Bscf)
90% Confidence
Interval of
Emissions
Estimate
(BscO
M&R
>300
N
3,460
179.8
5.5
4.7
M&R
100-300
N
13,335
95.6
11.2
21.7
M&R
<100
N
7,127
4.3
0.3
1.2
Regulating
>300
N
3,995
161.9
5.7
5.5
Regulating
>300
Y
2,346
1.3
0.03
0.06
Regulating
100-300
N
12,273
40.5
4.4
4.3
Regulating
100-300
Y
5,514
0.2
0.01
0.01
Regulating
40-100
N
36,328
1.0
0.3
0.4
Regulating
40-100
Y
32,215
0.1
0.02
0.02
Regulating
<40
Y
15,377
0.1
0.02
0.03
Total
131,970
27.3
23.3
The emissions estimate for each category of station was derived by multiplying the respective emission factor
(scf/station-hr) by the activity factor (number of stations), and converted to an annualized estimate by
assuming continuous fugitive leakage (i.e., 8760 hour per year leakage). The precision represents the 90%
confidence interval around the estimated mean emissions for each category.
D-6
-------
D-2
DISTRIBUTION SEGMENT SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Distribution mains are the pipelines that serve as a common source of natural gas supply for more than one
customer. Services are the branch connection lines from the mains to the customer meters. Leakage from the
underground distribution network occurs from corrosion pits, joint and fitting failures, and pipe wall fractures.
Gas distribution operators use leak detection procedures to locate and classify leaks. The leak is classified
and prioritized for repair based on the concentration of gas detected and the proximity of the leak to existing
structures.
EMISSION FACTOR: (scf/leak-year)
The value of the emission factor and standard deviation for each pipe material category is given below:
Material
Category
Pipe Use
Number
of
Samples
Average
Emission
Factor"
Units of
Emission
Factor
90%
Confidence
Interval of
Emission
Factor
Cast Iron
Main
21
238,736
scf/mi-yr
152,059
Unprotected
Steel
Main
20
51,802
scf/lk-yr
48,212
Protected Steel
Main
17
20,270
scf/lk-yr
17,243
Plastic
Main
6
99,845
scf/lk-yr
165,617
Unprotected
Steel
Service
13
20,204
scflk-yr
21,129
Protected Steel
Service
24
9,196
scf/lk-yr
5,581
Plastic
Service
4
2,386
scf/lk-yr
3,412
Cooper
Service
5
7,684
scf/lk-yr
5,559
a Adjusted for the soil oxidation of methane.
A cooperative leak measurement program has been developed to measure a representative sample of
underground leaks to estimate the average leak intensity, which is combined with company leak records to
estimate leak frequency. Leak measurements were performed at five U.S. companies and two Canadian
distribution companies in accordance with the testing protocol developed as part of the program. The test
data were disaggregated by material type and mains versus services, based on a combination of statistical
analyses and engineering judgement. A mean value of the estimated leak rate per leak was calculated using
Main and Service Pipeline
Normal Operations
Steady, Fugitives (Leakage)
41.6 Bscf +/- 65%
D-7
-------
the test data, for all pipe materials except cast iron. In these tests, an individual leak was randomly selected
for testing based on criteria outlined in the program plan. For cast iron, long segments of pipe were tested to
measure the leak rate per mile rather than the leak rate per leak. Cast iron was tested in long segments since
it tends to have a very high frequency of leaks (due to the joint spacing of 10 to 16 feet) and the relatively
high occurrence of undetectable leaks in cast iron. The measured natural gas leak rates were adjusted for the
average volume percent of methane in pipeline-quality gas (93.4 vol. %), and the soil oxidation rates of
methane.
ACTIVITY FACTOR:
The mean activity factor and standard deviation for each pipe material category is given below:
Material
Category
Pipe Use
Estimated
Total Leak
Repairs
Average
Activity
Factor
(Equivalent
Leaks)
Units of Activity
Factor
90%
Confidence
Interval of
Activity Factor
Cast Iron
Main
69,776
55,288
miles
2,764
Unprotected
Steel
Main
81,627
174,657
equivalent leaks
101,685
Protected Steel
Main
31,924
68,308
equivalent leaks
42,545
Plastic
Main
23,006
49,226
equivalent leaks
58,018
Unprotected
Steel
Service
214,271
458,476
equivalent leaks
499,850
Protected Steel
Service
182,562
390,628
equivalent leaks
526,354
Plastic
Service
32,202
68,903
equivalent leaks
66,840
Copper
Service
3,608
7,720
equivalent leaks
8,521
The national database of leak repairs was used to extrapolate data provided by individual companies. Data
were requested from each company participating in the underground leak test program, based on their
historical leak records. To allocate leak repairs into pipe material categories, data were collected from ten
local distribution companies representing different regions within North America.
Data on the total number of annual leak repairs, leak indications, and outstanding leaks within the distribution
system were provided by six companies. An estimate of the number of annual equivalent leaks for each of
the six companies was developed based on the following methodology:
Total Equivalent Leaks = Outstanding Leaks + New Leaks - Leak Repairs
The total number of annual equivalent leaks represents the equivalent leaks which are leaking all year. (That
is, for leaks with a leak duration of half year, these leaks are counted as half an equivalent annual leak.)
The total number of leaks in the system are quantified by incorporating the leak duration into the estimated
equivalent leaks. For example, if a leak is only leaking half the year, it is counted as 0.5 equivalent leaks.
The assumptions made in deriving the estimated number of equivalent leaks for each company include:
D-8
-------
Approximately 85 percent of leaks are found during a leak survey when an organic vapor
analyzer (OVA) instrument is used along with bar holing.
Leaks that are repaired during the year are leaking half of the year, on average.
Outstanding leaks are leaking at the beginning of the year.
The number of new leaks in the system is estimated based on the annual leak indications and
the frequency of the leak survey.
The number of new leaks in a system that is surveyed every n years is calculated based on the following:
For the first year in the cycle ~ 1/n leaks are leaking half the year; (n-l)/n leaks are not yet
leaking.
For the second year in the cycle — 1/n leaks are leaking the entire year; 1/n leaks are leaking
half the year; and (n-2)/n leaks are not yet leaking.
For the third year in the cycle — 2/n leaks are leaking the entire year; 1/n leaks are leaking
half the year; and (n-3)/n leaks are not yet leaking.
For the fourth year in the cycle - 3/n leaks are leaking the entire year; 1/n leaks are leaking
half the year; and (n-4)/n leaks are not yet leaking.
Based on the data provided by each of the six companies, a ratio of the annual equivalent leaks to leak repairs
was calculated. The average ratio (2.14) was multiplied by the estimated number of leak repairs in each pipe
material category to extrapolate the national database of leak repairs to represent annual equivalent leaks. The
precision of the estimate is based on the variability in the leak repair disaggregation provided by ten
companies and the variability in the calculated ratio of annual equivalent leaks to leak repairs provided by six
companies.
The activity factor for cast iron mains is the total estimated mileage of cast iron mains in the U.S., as reported
by the U.S.DOT RSPA (1). The standard deviation was assumed to be 5% of the estimated mileage, based
on engineering judgement.
D-9
-------
EMISSIONS ESTIMATE: (41.6 +/- 65 %)
The emissions estimate for each category of pipe material/use was dervied multiplying the respective emission
factor (scf/leak-yr or scf/mile-yr) by the activity factor (total number of leaks or miles).
Material
Category
Pipe
Use
Average
Emission
Factor
(scf/lk-yr)
Average
Activity
Factor
(equivalent
leaks)
Annual
Emissions
Estimate
(Bscf)
90% Confidence
Interval of
Emission Estimate
(Bscf)
Cast Iron
Main
238,736*
55,288"
13.2
8.4
Unprotected Steel
Main
51,802
174,657
9.1
11.1
Protected Steel
Main
20,270
68,308
1.4
1.6
Plastic
Main
99,845
49,226
4.9
13.9
Unprotected Steel
Service
20,204
458,476
9.3
17.5
Protected Steel
Service
9,196
390,628
3.6
6.1
Plastic
Service
2,386
68,903
0.2
0.4
Copper
Service
7,684
7,720
0.1
0.1
Total
41.6
27.1
ascf/mile-yr
bmiles
REFERENCES
1. U.S. Department of Transportation. Research and Special Programs Administration. 1991.
D-10
-------
D-3
DISTRIBUTION SEGMENT SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Pressure relief valves (PRVs) are often used in the distribution network to prevent the over-pressure of
distribution main pipelines. Typically, PRVs are used in conjunction with pressure regulators as a secondary
protection mechanism in the event of regulator failure. Gas is released during any emergency actuation of the
PRVs.
EMISSION FACTOR: 0.050 ± 3,914% Mscf/mile
(Adjusted for the distribution methane fraction of natural gas of 93.4 mol%)
The estimated emission factor was based on two separate distribution company studies which quantified losses
from PRVs as part of unaccounted-for (UAF) gas studies. The studies calculated PRV releases per mile of
pipeline mains. The GRI/EPA emission factor was estimated as the ratio of emissions per mile of main from
the two companies, and corrected for the methane composition in distribution.
EF PRECISION: ± 3,914%
Basis:
The precision was calculated using the method outlined in the Statistics Report (1).
ACTIVITY FACTOR: 836,760 ± 5% miles of main
The activity factor is based on the total miles of distribution main pipeline in the U.S.
AF PRECISION: ± 5%
Basis:
The accuracy was assigned based on engineering judgement.
ANNUAL METHANE EMISSIONS: 0.042 ± 3,919% Bscf
The annual methane emissions were determined by multiplying an emission factor (annual methane emissions
per mile of main) by the activity factor (miles of distribution main pipeline nationally).
REFERENCES
1. 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.
Pressure Relief Valves
Maintenance/Upsets
Unsteady, Vented
0.04 Bscf ± 3,919%
D-ll
-------
D-4
DISTRIBUTION SEGMENT SOURCE SHEET
SOURCES: Pipeline
OPERATING MODE: Mishaps (Dig-ins)
EMISSION TYPE: Unsteady, Fugitive
ANNUAL EMISSIONS: 2.1 Bscf+ 1,925%
BACKGROUND:
Dig-ins are distribution main or service pipeline ruptures caused by unintentional third-party damage. Some
distribution companies estimate and record the quantity of gas lost during a dig-in event; therefore, they keep
records of estimated annual losses due to dig-ins. From these annual records, a national emission rate for dig-
ins was determined.
ANNUAL EMISSION FACTOR: 1.59 ± 1,922% Mscf/mile
(Adjusted for the distribution methane fraction of natural gas of 93.4 mol%)
The emission factor was derived from four distribution company estimates of the losses from dig-ins: the
Pacific Gas and Electric unaccounted-for (UAF) gas study (1) results showed that losses from dig-ins were
estimated at 91,178 Mscf for 58,024 miles of distribution mains and services; the Southern California Gas
Company estimate (2) of losses from dig-ins was 170,457 Mscf for 82,337 miles of distribution mains and
services; a third company estimate of losses from dig-ins was 19,581 Mscf for 24,916 miles of distribution
mains and services; and a fourth company reported dig-in losses of 10,453 Mscf for 18,713 miles of
distribution mains. The ratio of the total dig-in emissions to the total pipeline miles from these companies was
used to estimate the national methane emission factor, resulting in 2.06 Mscf/mile.
EF PRECISION: ± 1,922%
Basis:
The precision was calculated from the spread of the company data using the method presented in the
Methane Emissions from the Natural Gas Industry, Volume 4: Statistical Methodology (3).
ACTIVITY FACTOR: 1,297,569 ± 5% miles of mains and services
The total number of miles of main pipeline in the U.S. gas industry was based on U.S. Department of
Transportation, Research and Special Projects Administration (4). The total miles of service pipeline was
reported in A.G.A.'s Gas Facts, 1990 (5).
AF PRECISION: ± 5%
Basis:
A 5% confidence bound was assigned on the basis of good precision from national statistics of 1990
data.
ANNUAL METHANE EMISSIONS: 2.06 ± 1,925% Bscf
The annual methane emissions were determined by multiplying an emission factor (annual methane emissions
per mile of pipeline) by the activity factor (number of miles).
D-12
-------
REFERENCES
1. Pacific Gas & Electric Company and Gas Research Institute. Unaccounted-For Gas Project. Volume
1, Final Report, San Ramon, CA, June 7, 1990.
2. Southern California Gas Company and Gas Research Institute. A Study of the 1991 Unaccounted-For
Gas Volume at the Southern California Gas Company, Final Report, Los Angeles, CA, April 1993.
3. 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.
4. U.S. Department of Transportation, Research and Special Projects Administration, Washington, DC,
1991.
5. American Gas Association. Gas Facts, 1992 Data, Arlington, VA, 1993.
D-13
-------
D-5
DISTRIBUTION SEGMENT SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Losses from customer meters are caused by fugitive leakage from the connections and other fittings
surrounding the meter set.
EMISSION FACTOR: (outdoor residential meters: 138.5 + 23.1 scf/meter-yr
commercial/industrial meters: 47.9 ± 16.7 scf/meter-yr)
The estimate of leakage from customer meters is based on screening and bagging studies conducted at ten
sites throughout the United States. The initial study was conducted by Indaco to measure customer meters in
the west coast [Indaco Air Quality Services, Inc., Methane Emissions from Natural Gas Customer Meters:
Screening and Enclosure Studies, draft report, August 15, 1992 (1)]. Data were also collected at nine
additional sites across the United States, including three east coast sites, a mid-western site, a rocky mountain
site, and five western U.S. sites. A summary of the average emissions from residential customer meters from
each of the ten sites is shown in the following table:
Site
Number of Meters
Screened
Number of Meters
Leaking
Average Leak Rate a
(lb methane/day)
Standard
Deviation3
(lb methane/day)
Site 1 — West Coast
134
37
0.0098
0.0239
Site 2 — East Coast
40
29
0.0002
0.0004
Site 3 — East Coast
158
37
0.0789
0.1753
Site 4 - Mid-West
156
8
0.0057
0.0061
Site 5 - Rocky
Mountain
188
28
0.0035
0.0082
Site 6 — West Coast
194
5
0.0002
0.0001
Site 7 — South East
201
56
0.0146
0.0328
Site 8 — North West
101
31
0.0101
0.0199
Site 9 — South West
150
50
0.0222
0.0404
Site 10 — North West
150
40
0.0125
0.0230
"Average value for all meters (i.e., leaking and non-leaking) screened at the site.
Customer Meters
Normal Operations
Steady, Fugitive
5.8 ± 1.1 bscfy
D-14
-------
The average emission factor for residential customer meters was derived by averaging the emission rates for
the ten sites. The emission factor was converted to units of scf/meter-yr by assuming that the losses from the
leaking meters were continuous throughout the year.
The precision represents the 90 % confidence interval and was calculated by averaging the standard deviations
for the ten sites.
The emission factor for commercial/industrial customer meters was derived from screening data collected at a
total of four sites. A summary of the average emissions from each of the four sites is shown in the following
table:
Site
Number of Meters
Screened
Number of Leaking
Meters
Average Leak Ratea
(lb methane/day)
Standard Deviation2
(lb methane/day)
Site 3 -- East Coast
45
12
0.0112
0.0251
Site 4 — Mid-West
61
0
—
—
Site 5 — Rocky
Mountain
21
6
0.0088
0.0076
Site 6 — West Coast
22
1
0.0018
-
"Average value for all meters (i.e., leaking and non-leaking) screened at the site.
The average emission factor for commercial/industrial customer meters was derived by averaging the emission
rates for the four sites. The emission factor was converted to units of scf/meter-yr by assuming that the
losses from the leaking meters was continuous throughout the year.
ACTIVITY FACTOR: (outdoor residential meters: 40,049,306 ± 4,200,135
commercial/industrial meters: 4,608,000 ± 230,400)
The total number of customer meters in the U.S. gas industry, 56,132,300, and the number of residential
customer meters, 51,524,600, were based on Gas Facts, American Gas Association, 1992 (2). The number of
residential customer meters located indoors versus outdoors was estimated based on a regional breakdown of
total customers presented in Gas Facts (2) combined with data obtained from 22 individual gas companies
within different regions of the country. (Note: The number of customers in each region was used to estimate
the number of indoor meters because data on number of customer meters segregated by region were not
available.)
Following is the average percentage of customer meters located indoors in each region:
D-15
-------
Region
Total
Residential
Customers
Average Percent
Indoor Meters
Sample
Size
Estimated
Indoor
Meters
Precision
New England
1,886,500
52
1
980,980
471,625"
Middle Atlantic
8,403,400
61
7
5,126,074
1,905,371
East North
Central
11,633,500
17
7
1,977,695
1,461,663
West North
Central
4,684,100
40
1
1,873,640
1,873,640*
South Atlantic
4,987,700
21
4
l,030,680b
1,030,680'
East South
Central
2,465,200
0
--
0
123,260°
West South
Central
5,666,600
0
~
0
283,330°
Mountain
3,318,700
0
—
0
331,870°
Pacific
9,724,500
5
2
486,225
486,225"
TOTAL
52,770,200
22
11,475,294
3,317,254
'Estimated based on engineering judgement.
bEstimated for each state separately in region.
'Estimated based on industry comments suggesting that customer meters in southern regions are essentially all
located outdoors.
The estimated number of indoor meters, 11,475,294, was subtracted from the total number of reported meters,
51,524,600, to derive an estimated 40,049,306 outdoor residential customer meters in the United States. The
precision was estimated from the data provided by the companies, engineering judgement for some regions,
and an estimated 5% error in the nationally reported number of residential customer meters.
The leakage rates from customer meters located indoors was assumed to be negligible based on the increased
probability that leaks on indoor meter sets are detected and repaired promptly. This assumption of negligible
leakage from indoor meters is consistent with the findings from pressure regulating stations located in vaults.
The precision of the total estimated commercial/industrial customer meters is assumed to be ± 5% of the
estimated 4,608,000 meters.
ANNUAL METHANE EMISSIONS: (5.8 ± 1.1 Bscf/yr)
REFERENCES
1. Indaco Air Quality Services, Inc. Methane Emissions from Natural Gas Customer Meters: Screening
and Enclosure Studies, Draft Report, August 15, 1992.
2. American Gas Association. Gas Facts. Arlington, VA. 1992.
D-16
-------
D-6
DISTRIBUTION SEGMENT SOURCE SHEET
SOURCES:
OPERATING MODE:
EMISSION TYPE:
ANNUAL EMISSIONS:
BACKGROUND:
Gas is blown to the atmosphere as a result of pipeline abandonment, installation, and repair.
ANNUAL EMISSION FACTOR: 0.102 + 2,521% Mscf/mile
(Adjusted for the distribution methane fraction of natural gas of 93.4 mol%)
The emission factors for pipeline blowdown are based on estimates from four companies: the Pacific Gas &
Electric Unaccounted-for Gas (UAF) Project, 1987 (1); the Southern California Gas Company (SoCal) project
(2); and two additional company estimates. The estimated total gas losses were adjusted for 93.4 volume
percent methane. The annual methane emissions per mile of mains and services for each of the four
companies was calculated based on the ratio of emissions to miles of distribution mains and services. The
following table summarizes the individual company estimates and the national emission factor. The precision
of the estimate is based on the 90 percent confidence level for the four companies providing data.
Company
Annua!
Slowdown
Methane
Emissions, Mcsf
Pipeline
Miles
Annual
Blowdown Methane
Emission Factor,
scf/mile
1
8,972
58,024
0.155
2
5,688
82,337
0.069
3
2,360
24,916
0.095
4
1,695
18,713
0.091
TOTALS
18,715
183,990
ANNUAL BLOWDOWN EF, Mscf methane/mile
0.102 ± 2,521%
ACTIVITY FACTOR: (1,297,569 ± 5% miles mains and services)
The total number of miles main pipeline in the U.S. gas industry was based on U.S. Department of
Transportation, Research and Special Projects Administration (3). The total miles of service was reported in
Gas Facts (4). The precision, or 90 percent confidence level, was estimated to be ± 5%, based on
engineering judgement.
ANNUAL METHANE EMISSIONS: 0.13 ± 2,524% Bscf
The annual methane emissions were determined by multiplying an emission factor (annual methane emissions
per mile of pipeline) by the activity factor (number of miles).
Pipeline
Maintenance
Unsteady, Vented
0.13 Bscf ± 2,524%
D-17
-------
REFERENCES
1. Pacific Gas & Electric Company and Gas Research Institute. Unaccounted-For Gas Project.
Volume 1, Final Report, San Ramon, CA, June 7, 1990.
2. Southern California Gas Company and Gas Research Institute. A Study of the 1991 Unaccounted-For
Gas Volume at the Southern California Gas Company, Final Report, Los Angeles, CA, April 1993.
3. U.S. Department of Transportation, Research and Special Projects Administration, Washington, DC,
1991.
4. American Gas Association. Gas Facts, 1992 Data, Arlington, VA, 1993.
D-18
-------
APPENDIX E
Conversion Table
E-l
-------
Unit Conversion Table
English to Metric Conversions
1 scf methane
1 Bscf methane
1 Bscf methane
1 Bscf
1 short ton (ton)
1 lb
1 ft3
1 ft3
1 gallon
1 barrel (bbl)
1 inch
1 ft
1 mile
1 hp
1 hp-hr
1 Btu
1 MMBtu
1 lb/MMBtu
T (°F)
1 psi
19.23 g methane
0.01923 Tg methane
19,230 metric tonnes methane
28.32 million standard cubic meters
907.2 kg
0.4536 kg
0.02832 m3
28.32 liters
3.785 liters
158.97 liters
2.540 cm
0.3048 m
1.609 km
0.7457 kW
0.7457 kW-hr
1055 joules
293 kW-hr
430 g/GJ
1.8 T CO + 32
51.71 mm Hg
Global Warming Conversions
Calculating carbon equivalents of any gas:
MMTCE = (MMT of gas) x carbon] x (GWP)
{ MW, gas )
E-2
-------
Calculating C02 equivalents for methane:
MMT of CO, equiv. = (MMT CH.) x
MW, CO^
MW, CH,
' 4 /
x (GWP)
where MW (molecular weight) of C02 = 44, MW carbon = 12, and MW CH4= 16.
Notes
scf = Standard cubic feet. This is the cubic feet that the gas would
occupy if it were at the standard conditions of 14.73 psi, absolute,
and 60°F. For a fixed gas composition, a scf defines a mass.
Bscf = Billion standard cubic feet (109 scf).
Tscf = Trillion standard cubic feet (1012 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).
scfd.scfy = Standard cubic feet per day, standard cubic feet per year.
mol% = Percent of molecules in a stream that are of one type.
vol% = Percent of the volume of a stream that is of one species. For an
ideal gas, vol% is equal to mol%.
E-3
-------
wt%
Percent of the mass (weight) of a stream that is of one species.
Prod =
Production
Proc =
Gas Processing
Trans
Transmission
Dist
Distribution
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.
E-4
------- |