tari
United States GRi-94 / 0257.22
Environmental Protection EPA-600/R-96-080e
Agencv Juneigg6
<&EPA Research and
Development
METHANE EMISSIONS FROM THE
NATURAL GAS INDUSTRY
Volume 5: Activity Factors
Prepared for
Energy Information Administration (U. S. DOE)
Prepared by
National Risk Management
Research Laboratory
Research Triangle Park, NC 27711
-------
TECHNICAL REPORT DATA ... .. ¦¦¦¦n iimn mi i ii iii
(Please read Instructions on the reverse before compi || | |||| || |||||| 111| III 1III I
1. REPORT NO. 2.
EPA-600/R-96-080 e
PB97-142 962
4. TITLE AND SUBTITLE
Methane Emissions from the Natural Gas Industry,
Volumes 1-15 (Volume 5: Activity Factors)
5. REPORT DATE
June 1996
6. PERFORMING ORGANIZATION CODE
7. author(s) 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 AND ADDRESS
Radian International LLC
P. O. Box 201088
Austin, Texas 78720-1088
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
5091-251-2171 (GRI)
68-D1-0031 (EPA)
12. SPONSORING AGENCY NAME AND ADORESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 3/91-4/96
14. SPONSORING AGENCY CODE
EPA/600/13
15.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).
is- abstract The 15-volume report summarizes the results of a comprehensive program
to quantify methane (CH4) emissions from the U. S. natural gas industry for the base
year. The objective was to determine CH4 emissions from the wellhead and ending
downstream at the customer's meter. The accuracy goal was to determine these
amissions 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
13 B
14 G
04A
21D
15E
07 C
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
137
20. SECURITY CLASS (Thispage)
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 Laboratory1 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.
PROTECTED UNDER INTERNATIONAL COPYRIGHT
ALL RIGHTS RESERVED.
NATIONAL TECHNICAL INFORMATION SERVICE
U.S. DEPARTMENT OF COMMERCE
-------
EPA-600/R-96-080e
June 1996
METHANE EMISSIONS FROM
THE NATURAL GAS INDUSTRY,
VOLUME 5: ACTIVITY FACTORS
FINAL REPORT
Prepared by:
Blake E. Stapper
Radian International LLC
8501 N. Mopac Blvd.
P.O. Box 201088
Austin, TX 78720-1088
DCN: 95-263-081-14
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 5: Activity Factors
Final Report
Contractor Radian International LLC
GRI Contract Number 5091-251-2171
EPA Contract Number 68-D1-0031
Principal
Investigator Blake E. Stapper
Report Period March 1991 - June 1996
Final Report
Objective This report describes a study to quantify the national activity factors for
the methane emission source types in the gas industry.
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 on global warming of natural gas versus coal and oil.
Results Activity factors are documented for use in determining the total
emissions from the natural gas industry. Since there are very few
published activity factors for the gas industry, many activity factors were
developed specifically for this study. Confidence intervals were
calculated for the activity factors so that the overall accuracy could be
determined. Precautions were also taken to ensure that the activity
factors are statistically representative of the industry.
iii
-------
The program reached its accuracy goal and provides an accurate estimate
of methane emissions that can be used to construct U.S. methane
inventories and analyze fuel switching strategies.
Technical Since it is impractical to sample emissions from every source in the
Approach natural gas industry, it is necessary to develop a method of scaling up the
sampled emissions from a representative set of sources within the
industry. The activity factor extrapolation method was developed for this
purpose.
This method is used to scale-up the average annual emissions from a
source (determined by a limited sampling effort) to represent the entire
emissions from the national population of similar sources in the gas
industry. The typical activity factor extrapolation approach uses
emission factors (EF) and activity factors (AF) to do this. An emission
factor for a source category is defined as the average annual emissions
per source. Activity factors are the number of sources in the entire target
population or source category. Thus, the product of the EF and the AF
equals the annual national emissions from a specific source in the natural
gas industry. Since there are very few published activity factors for the
gas industry, many were developed specifically for this study.
The accuracy of the activity factors is dependent upon precision and bias.
The precision of the activity factors was determined by calculating the
90% confidence limits. Potential bias in the activity factors were
eliminated by three methods: peer review by experienced sources,
analysis of the data, and extrapolation by different parameters.
For the 1992 base year the annual methane emissions for the U.S. natural
gas industry are 314 Bscf ±106 Bscf (± 34%). This is equivalent to
1.4% ± 0.5% of 1992 gross natural gas production. Results from this
program were used to compare greenhouse gas emissions from the fuel
cycle for natural gas, oil, and coal using the global warming potentials
(GWPs) recently published by the Intergovernmental Panel on Climate
Change (IPCC). The analysis showed that natural gas contributes less to
potential global warming than coal or oil, which supports the fuel
switching strategy suggested by 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
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
Project
Implications
iv
-------
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 GAS INDUSTRY CHARACTERIZATION 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 DEVELOPMENT OF ACTIVITY FACTORS 15
4.1 Definition of Activity and Emission Factors 15
4.2 Effect of Sampling Approach on Activity Factors 15
4.3 Activity Factor Sources 16
4.4 Extrapolation of Site Activity Factor Data 17
4.4.1 Selection of Site-Weighted (Ratio) Extrapolation Technique 19
4.4.2 Extrapolation Parameter Selection 20
4.5 Activity Factor Accuracy 21
4.5.1 Precision 23
4.5.2 Propagation of Error 24
4.5.3 Screening for Bias 26
5.0 PRODUCTION SEGMENT 28
5.1 Development of a Regional Approach 28
5.2 Gas Well Count and Production (Onshore and Offshore) 34
5.3 Oil Well Count and Production (Onshore and Offshore) 34
5.4 Oil Wells Marketing Gas 35
5.5 Offshore Platforms 36
5.6 Specific Equipment 39
5.7 Vessels and Pressure Relief Valves 40
5.8 Gathering Pipeline Miles 46
5.9 Maintenance Activities 48
5.10 Compressors and Dehydrators 50
vi
-------
TABLE OF CONTENTS
(Continued)
Page
6.0 GAS PROCESSING SEGMENT 51
6.1 Gas Plant Count 51
6.2 Acid Gas Recovery (AGR) Units 52
6.3 Compressors and Dehydrators 52
7.0 TRANSMISSION SEGMENT 53
7.1 Compressor Station Count 53
7.2 Storage Station Count 56
7.3 Storage Well Count 57
7.4 M&R Station Count 57
7.5 Miles of Pipeline 60
7.6 Compressors and Dehydrators 60
7.7 Pneumatics 60
8.0 CROSS-SEGMENT ACTIVITY FACTORS 61
8.1 Compressors 61
8.1.1 Compressor Horsepower-Hours 61
8.1.2 Compressor Counts 67
8.2 Dehydrators 70
8.2.1 Dehydrator Count 70
8.2.2 Dehydrator Throughput 72
8.2.3 Dehydrator Flash Drums 75
8.2.4 Dehydrator Gas Pumps 75
8.2.5 Dehydrator Stripping Gas Use 75
8.2.6 Dehydrator Vapor Recovery Use 76
9.0 DISTRIBUTION SEGMENT 78
9.1 Distribution Annual Equivalent Leaks 78
9.2 Distribution Metering and Pressure Regulating Stations 85
9.3 Customer Meters 87
9.4 Distribution Mains and Services 89
10.0 RESULTS 90
11.0 REFERENCES 93
APPENDIX A - Site Data A-l
APPENDIX B - Methods for Estimating Activity Factors B-l
APPENDIX C - Conversion Table C-l
vii
-------
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 Major Producing States 29
5-2 Selected Production Regions 30
B-l Conceptual Comparison of Normal and Lognormal Distributions B-4
B-2 Comparison of 90% Confidence Limits B-7
B-3 Distribution of Separators per Well B-8
B-4 Ratio of Separators per Unit of Production B-9
viii
-------
LIST OF TABLES
Page
4-1 Nationally-Tracked Activity Factors 17
4-2 Production Extrapolation Parameter Selection 22
4-3 Error Propagation for a Sum 25
4-4 Error Propagation for a Product 25
5-1 Regional Differences in Production Rates and Well Counts 31
5-2 Sample Set Comparison to Actual Regional Data 33
5-3 National Gas Industry Wells 37
5-4 Platform Counts 39
5-5 Specific Equipment Activity Factors 40
5-6 Separators in Production 41
5-7 In-Line Heaters in Production 42
5-8 Pneumatic Devices in Production 43
5-9 Chemical Injection Pumps in Production 44
5-10 Compressors (Non-Gas Lift) in Production 45
5-11 Pressure Relief Valves 46
5-12 Production Gathering Pipeline Mileage - Gas Wells 48
6-1 Divisions of Gas Processing Plants by Type 51
7-1 Transmission Compressor Station Counts 54
7-2 Transmission M&R Station Populations 59
8-1 Compressor Activity Factors for Each Industry Segment 62
8-2 Compressor Counts in the Gas Industry 68
ix
-------
LIST OF TABLES
(Continued)
Page
8-3 U.S. Gas Industry Dehydrator Counts 71
8-4 Industry Segment Glycol Dehydrator Throughput 73
8-5 U.S. Gas Industry Dehydrator Flash Drums 76
8-6 Site Survey Gas Pump Data 76
8-7 Site Survey Stripping Gas and Vapor Recovery Data 77
9-1 National Leak Repairs Allocated by Pipe Material Category 80
9-2 Summary of Leak Record Data from Participating Companies 83
9-3 Summary of Activity Factors for Distribution Underground
Pipelines 84
9-4 Activity Data Provided by Individual Companies 86
9-5 Summary of Activity Factors for Distribution Metering/
Pressure Regulating Stations 88
10-1 Well-Defined Activity Factors 90
10-2 Examples of Developed Activity Factors 91
A-l Production Sites (Summary) A-2
A-2 Production Sites (Offshore Data) A-3
A-3 Production Sites (Gulf Coast Onshore Data) A-4
A-4 Production Sites (Central Plains Data) A-5
A-5 Production Sites (Pacific/Mountain Onshore Data) A-6
A-6 Production Sites (Atlantic & Great Lakes Data) A-7
A-7 Gas Processing Plants A-9
x
-------
LIST OF TABLES
(Continued)
Page
A-8 Transmission Compressor Stations A-10
A-9 Storage Compressor Stations A-12
xi
-------
1.0 SUMMARY
This document is one of several volumes that provide background
information supporting the Gas Research Institute and U.S. Environmental Protection
Agency Office of Research and Development (GRI-EPA/ORD) methane emissions project.
The objective of this comprehensive program is to quantify the methane emissions from the
gas industry for the 1992 base year to within ± 0.5% of natural gas production starting at
the wellhead and ending immediately downstream of the customer's meter. Activity factors
are needed to estimate the national emissions from the natural gas industry and are based on
estimated populations of equipment and equipment characteristics. This report presents a
definition for activity factors, methods of calculation, and this project's calculated values
for industry activity factors.
1
-------
2.0 INTRODUCTION
If emissions could be sampled from every source in the natural gas industry,
then the total national emissions for the industry could be determined by summing the
emissions from each source. Unfortunately, because of the size of the industry, measuring
emissions from each source is impractical. Therefore, a method of scaling up (i.e.,
extrapolating) the sampled emissions from a representative set of sources within the
industry is necessary. The activity factor extrapolation method was developed for this
purpose.
This method is used to scale-up the average annual emissions from a source
(determined by a limited sampling effort) to represent the entire emissions from the national
population of similar sources in the gas industry. The typical activity factor extrapolation
approach uses emission factors (EF) and activity factors (AF) to do this. An emission
factor for a source category is defined as the average annual emission per source. Activity
factors are the number of sources in the entire target population or source category. An
activity factor is usually presented as an equipment count, but a few exceptions exist, such
as: 1) horsepower-hours for compressors, 2) average equivalent leaks for underground
pipe, and 3) events/year for maintenance activities. They can be used to estimate the
nationwide emissions from the natural gas industry since emission rates are often measured
on a per equipment or per activity basis. Thus, the emission factor and activity factor are
defined such that their product equals the annual national emissions from a specific source
in the natural gas industry.
EF x AF = National Emission Rate (1)
Since there are very few published activity factors for the gas industry, many
were developed specifically for this study, especially in the production segment of the
industry. This report discusses precautions taken to ensure that the data used to determine
activity factors were statistically representative of the industry.
2
-------
This report explains in detail how the activity factors for the GRI/EPA
methane emissions project were determined for the 1992 base year. The following sections
summarize the methods used to calculate gas industry activity factors. Section 3 provides
the general industry characterization used to identify the equipment populations. Section 4
defines the techniques used to develop the activity factors. Sections 5 through 9 provide
the specific activity factor development for equipment within each segment of the gas
industry. Summary results are presented in Section 10. This report is one of several
volumes under the GRI/EPA methane emissions project.
3
-------
3.0
GAS INDUSTRY CHARACTERIZATION
Most activity factors define equipment populations. Therefore, to determine
activity factors, it is important to first define and describe the major divisions in the
industry that affect the selection of specific equipment. 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.
While this section draws a general picture of the industry developed by the
GRI/EPA methane emissions project, it is not intended to represent a definitive picture of
the industry regarding all typical operational parameters. Many details of the operation are
unnecessary for development of an accurate methane emissions inventory. Other details are
given in specific reports where the details were needed to estimate specific emission factors.
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 by local distribution
pipeline networks to the customer.
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.
This project has set specific boundaries on each segment of the industry that
specify what equipment is included or excluded from the study. These boundaries were set
4
-------
PRODUCTION
Surface Facilities
Ul
PROCESSING
TRANSMISSION/
STORAGE
Direct
Sales
M&PR Stations
Compressor
Stations
Pipelines
Gas P ant
Pipelines
Liquids
Storage
Liquids
Underground
Storage
Reservoir
DISTRIBUTION
Main and
Service Pipelines
(M)
"(§)
-<9>
Customer Meters
m
Compressor
©
Meter
£
Pressure
Regulator
Figure 3-1.
Gas Industry Flow Chart
-------
using input from industry experts. 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 certain equipment mainly 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 this study's activity factor estimates.
As shown in Figure 3-2, the gas industry production segment includes all
equipment at gas well sites, except emissions from liquid hydrocarbon condensate or oil
tanks at the site. These emissions are considered to be a result of marketing the liquid
products from the well. The gas industry production segment excludes equipment that
6
-------
Petroleum Industry
Oil Wei
Wei head
Compressor
Dehydrator
Rc-inieclion/
Compressor
Gas Lift Gas
Meter
Oil Well
Wellhead
(producing
gas)
Field Use
Gas
Separator
Chemical
Injection
Pump
Pnuematic
Control
Valve
Heater/
Treater
Salt Water and
Oil Stock Tanks
q | Vapor Recovery
Compressor
Coal Bed
Methane Well
3_j:
^ Separator ^
Compressor
H h
Fresh Water
Dehydrator
Gas Well
Wellhead
Chemical
Injection
Pump
Dehydrator
Compressor
Meter
Field Use
Gas
^ Separator J
Vapor Recovery
Compressor
Hydrocarbon
Condensate
or Oil Tank
Pnuematic
Control Valve
Petroleum Industry
Salt Water
Tank
Figure 3-2. Industry Boundaries
7
-------
exists primarily for the production of oil, since it would exist even if the gas were not
marketed. Therefore, the definition excludes all oil tanks, and excludes 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 production, 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 compression and dehydration 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 separator(s), a dehydrator, or a compressor. The use of the equipment depends upon
the free liquid production, the absorbed moisture content, and the 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.
8
-------
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
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 also 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 those few 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
9
-------
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.
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. 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 pipeline, compressor stations, and metering
facilities. All of these facilities and all of the equipment they contain are considered part of
the natural gas industry.
Transmission compressor stations usually consist of piping manifolds,
reciprocating 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.
10
-------
c ~
Absorption or
Compression
& Cooling Process
Wet
Gas
Dehydration
Backend of the Plant
-> C2S
Liquids
-> C„
-> C„
Natural
Gasolines
Figure 3-3. Gas Processing Plant
11
-------
to
Gas
from
Pipeline
TRANSMISSION
COMPRESSOR STATION
ABOVE GROUND STORAGE
Cooling
Compressors
Heaters ^
Transmission Pipeline
from
Pipeline
BELOW GROUND STORAGE
Compressor
Station
Wells
Dehydrator storage Field"
Figure 3-4. 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 local
distribution companies (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. Storage facilities are often located close to consumption centers
so that the cross-country transmission pipeline does not have to be sized for peak winter
demand. Storage facilities can be below ground or above ground. Above-ground facilities
are liquified natural gas (LNG) facilities that liquify 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 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
regulating 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)
®
Cm)
Customer Meters
(Residential,
Commercial,
Industrial,
Electric Utility)
Figure 3-5. Distribution Segment Equipment
-------
4.0
DEVELOPMENT OF ACTIVITY FACTORS
This section discusses the sampling approach, sources of equipment counts,
and methods used to develop activity factors.
4.1 Definition of Activity and Emission Factors
The activity factor is typically defined as a count of the total population of
sources in a particular category, and the emission factor is defined as the average annual
emissions per source. Where a source category is emissions from an equipment type, the
activity factor is the national population of that equipment type. Exceptions to the general
definitions for the factors are those sources that have an emission rate that can be more
accurately represented by a parameter that directly influences the rate of operation. (For
example, the emission factor for internal combustion (IC) engines is given in terms of
annual emissions per horsepower.) For such exceptions, the activity factor is defined as the
parameter that influences the emissions from the source (e.g., the activity factor is annual
horsepower-hours from the IC engine).
4.2 Effect of Sampling Approach on Activity Factors
As discussed in Volume 3 on general methodology,1 the methods used to
measure and collect samples directly impact the generation of activity factors. Improper
sampling can easily bias the results. For this reason, particular attention was paid to the
sampling methods used for this project. However, an absolutely random sample of the
natural gas industry could not be ensured because participation in this program was
voluntary.
In addition to ensuring as random of a sample as possible, the project team
attempted to identify and compensate for any biases in the data from the sampling. For
example, if regional differences in equipment configurations were known to exist, then data
15
-------
were gathered and evaluated regionally to generate regional activity factors. These regional
activity factors were then combined to determine a national total activity factor for a
particular source type.
The project team attempted to compensate for each of the biases identified by
adjusting the sampling method used. For the example of regional bias, the United States
was divided into six regions to account for the regional differences in most of the activity
factors, and samples were taken in each of those regions (see Section 5 for more details).
No significant regional differences were observed in the major activity factors in the gas
processing, transmission, or distribution segments; in these categories all samples were
treated equally and averaged to produce a national figure.
4.3 Activity Factor Sources
Activity factors were determined from publications such as the American Gas
Association (A.G.A.) Gas Facts,2 company data, or through site visits.
Organizations such as A.G.A. track national statistics such as gas wells, miles
of transmission and distribution pipelines, and total national production in the natural gas
industry. Table 4-1 shows some of the published and well-defined activity factors.
However, in many cases, the total population of a source type within the gas industry is
unknown.
For sources that are not tracked nationally, data from individual companies
(for the entire company) or regional surveys (surveys by state agencies or trade
organizations) are sometimes available. Metering/pressure regulating stations, glycol
dehydrators, and compressor engines/gas turbines are tracked on a company-wide basis or
by regional organizations. 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.
16
-------
TABLE 4-1. NATIONALLY-TRACKED ACTIVITY FACTORS
Segment
Activity Factor Name
Number
Total Industry
1992 Gross Gas Production (Tscf)
22.13
1992 Marketed Gas Production (Tscf)
18.71
Production
No. of Gas Wells
276,000
No. of Oil Wells
602,200
Processing
No. of Gas Plants
726
Transmission and Storage
Miles of Trans Pipeline
284,500
No. of Storage Facilities
475
No. of Storage Wells
18,000
Distribution
Miles of Mains
836,760
No. of Services
51,520,000
For sources that have an unknown population, a number of site visits were
conducted to determine the number of sources at each location. The average values from
multiple sites were then scaled-up to represent the total population. Site visits to collect
activity factor data were typically done in conjunction with the data collection efforts for the
emission factors. For the activity factors determined from site visit data, a method was
needed for extrapolating the site count to a national value. The following section explains
how the activity factors were determined from site visit data.
4.4 Extrapolation of Site Activity Factor Data
In the GRI/EPA study, some activity factors were determined based on
extrapolative site data. For example, no data were available on the nationwide population of
17
-------
production separators. The number of separators at a site, gathered as part of site visits, was
divided by the number of wells at each site. Then, the average ratio of separators to wells
from all site visit data was used to extrapolate nationally by multiplying by the national well
count. This method of scaling-up the site visit data uses extrapolation parameters (EP), a
well-defined, published factor that can relate the activity between the site and the national or
regional level. The EPs are usually one of the nationally-tracked activity factors presented in
extrapolated to regional or national activity factors (AFR or AFN, respectively) by using EP.
Examples of extrapolation parameters are the number of wells for production and storage, the
number of plants for processing, and gas throughput for all segments of the industry.
To scale-up a site activity factor to the regional or national level, the site
activity factor must be divided by the extrapolation parameter for that site and multiplied by
the regional or national value for the extrapolation parameter. In effect, the site activity
factor (AFS) is being multiplied by the ratio of two known factors: one on the site basis
(EPS) and the other at the regional (EPR) level. This is illustrated in the following equation:
parameter for multiple sites within a region can be done using one of two equations: a site-
weighted method or an approach using an average count per site (not site-weighted). The
first uses a ratio of the summation of activity factors for all sites to the summation of
extrapolation parameters for all sites:
Table 4-1.
Equipment activity factors from one individual site's data (AFS) can be
(2)
Determining the ratio of the site activity factor to the site extrapolation
18
-------
£ AF,
(AF/EP), -
£ EPi
i = 1
where n = number of sites. The second is an average of individual site ratios. This
equation determines the ratio to be used for scale-up, (AF/EP)S, by summing the ratio of the
site activity factor to the extrapolation factor, (AF/EP);, calculated independently for each
site.
52 (AF/EP). Method 2: (4)
(AF/EP) = — — (Not site-weighted)
s n
4.4.1 Selection of Site-Weighted (Ratio) Extrapolation Technique
The site-weighted method (Equation 3) was selected for the extrapolation
technique for all production activity factors. In project review meetings, this technique was
accepted by production industry representatives and their company statisticians as the best
approach. The site-weighted approach is based on the assumption that the sites are randomly
sampled, so that a large site is proportionately representative of a large section of the
population. The site-weighted method, which will hereafter be referred to as the ratio
method is discussed in greater detail in Appendix B.
The non-weighted approach was initially used in earlier drafts of this report as
the extrapolation method. It was based upon an assumption that the "representativeness" of
any site could not be determined. It assumed that a small sampled site (15 wells) may
actually be part of a heterogeneous large field (10,000 wells), while a large site (100 wells)
might be the full sample of a small field. However, industry reviewers found that this
assumption was not the best representation for the production segment, and therefore the
ratio method was used. The ratio method was also used in the gas processing, transmission,
and storage segments.
Method 1: (3)
(Site-weighted)
19
-------
4.4.2
Extrapolation Parameter Selection
Some equipment activity factors can be scaled up by any of several available
extrapolation parameters (EP). These parameters must be known for each site as well as for
the national or regional total. For example, in the production segment there are only three
extrapolation parameters: 1) gas well count, 2) gas production rate, and 3) total "gas
industry" well count. Gas wells are simply the total count of gas wells at the site, and gas
production rate is the rate of gas marketed from the site. Both of these values are well
known for the nation and for the region. "Gas industry" wells are the sum of gas wells plus
the oil wells that market gas (oil-only wells are not counted). This last extrapolation
parameter is not known exactly, since the regional or national number of oil wells marketing
gas is not known. Nevertheless, the number of oil wells marketing gas was determined to a
reasonable degree of accuracy (see Section 5.5).
If there was a physical or technical relation between the equipment and one of
the three candidates for the extrapolation parameters (EP), then that parameter was selected.
This was the case, for example, for in-line heaters and chemical injection pumps, which are
related to well count based on engineering design reasons. However, where the relation
between the source population and various extrapolation parameters is not definitive from a
technical perspective, a single EP cannot be selected. Therefore some alternate approaches
for combining the various EP's can be used, such as: 1) averaging the results from both
methods of EP extrapolation, or 2) statistically analyzing the data to determine the best
approach.
For example, statistical analysis was used in selecting the activity factor
determination for meter stations in distribution. 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 by system throughput, which were the only known population statistics. The
station counts from individual companies were examined both on a per mile of main and per
20
-------
system throughput basis. A linear regression analysis showed that extrapolating the data
using a per mile main basis produced less variability in the national extrapolation.
In the production segment, Radian examined the technical relationship between
equipment and the three extrapolation parameters (gas well counts, total well counts, or
production rate). The EP selected was based upon Radian's and the industry reviewer's
knowledge of production operations. If the equipment was most strongly related to one
parameter, and not related or very weakly related to the other parameters, then the first
parameter was selected. If there was a rationale for a strong technical relation between both,
then an average of the approaches was used.
In several cases, the total "gas industry well" count could be excluded as an
EP if the boundaries selected (see Figure 3-2) excluded that equipment on associated oil
wells. In many cases, there was a reason to expect that equipment counts were related to
both well counts and production rate. This was true for pneumatics, separators, and
compressors.
The rationale for relation to both parameters on this equipment is as follows.
Surface facilities may be organized as central facilities or individual well site facilities. If
surface facilities are sited on individual well sites, which occurs frequently, then a relation
between equipment count and wells would be expected. If the equipment was sited at central
facilities, then it might be specified and sized based purely upon production rate. Since both
relations clearly existed, the average was selected. The production activity factors,
extrapolation parameters, and the basis for the selection are listed in Table 4-2.
4.5 Activity Factor Accuracy
Determining the accuracy of the activity factor is a key part of estimating the
accuracy of the national emission rate. Accuracy is dependent on precision and bias, as
discussed in Volume 3 on general methodology1 and Volume 4 on statistical methodology.3
21
-------
TABLE 4-2. PRODUCTION EXTRAPOLATION PARAMETER SELECTION
Activity Factor
Extrapolation Parameter(s) Selected
Basis
In-Line Heaters
Gas wells
In-line heaters exist only on gas wells and have a strong relation to well count.
While some heaters (such as heater-treaters) do exist on oil wells, they are not
within the boundaries of the gas industry (see Figure 3-2).
Separators
Mean of: 1) gas well and 2) production rate.
Exception: offshore, where the relation is to
production rate only.
Separators can be related to both production rate and well count. Separators can
be at individual wells, or can be at central facilities, where production rate may
play a factor. In the offshore area, the relation is stronger to production rate.
Compressors
Mean of: 1) total gas industry wells and 2)
production rate
Compressors exist within the gas industry on the overhead lines from gas well
separators and associated-oil well separators. Therefore there is a relation to total
well count. There is also a relation to production rate since there are size limits
for compressors.
Pneumatics
Mean of: 1) total gas industry wells and 2)
production rate
Pneumatics exist on the overhead gas lines from associated oil wells and from
gas wells. Gas lines from any gas producing well may have dehydrators and
compressors with pneumatics. Therefore pneumatics are related to total gas
industry well count. Pneumatics can also be related closely with production rate
where central separation facilities exist. (The siting of equipment and pneumatics
at central facilities can be largely dependent on production rates.)
Chemical
Injection Pumps
(CIPs)
Gas wells
CIPs were found mostly at individual well sites, even where central separation
facilities existed. CIPs therefore have a strong relation to well count. While
CIPs do exist on oil wells, oil well CIPs are not within the boundaries of the gas
industry (see Figure 3-2). Therefore gas well count was used.
-------
Precision, the random invariability in the measurement, is calculated by propagating error
from each individual group of measurements. However, bias, a systematic error in the
estimate, must be discovered and eliminated.
4.5.1 Precision
If the activity factor estimate is assumed to be approximately normally
distributed, then the 90% confidence limits for the activity factors determined by the site-
weighting method can be estimated using Cochran's4 equation 6.14. The equation for the
90% confidence interval (symmetric) for the site-weighted activity factor is:
± n-1) ^
The equation for the variance is:
„ . (6)
n(n-l) i=i ' '
where
yt = the number of equipment at site i in the sample set;
X; = the number of wells or amount of production at site i in the sample
set;
n = number of sites sampled in the given region;
N = the total number of sites in the region;
f = sampling fraction = n/N;
R = activity factor ratio = (AF/EP)sample; and
t(i-a/2, n-i) = the l-a/2 probability of the Student's t Distribution with n-1
degrees of freedom.
23
-------
The total number of sites (N) is not known for each region or nationally,
estimated by the following equation:
N= (Production or Wells)totalregion
(Production or Wells)tota] sample /n
Either production rate or wells can be used in the equation, depending on which extrapolation
parameter is used.
4.5.2 Propagation of Error
This section discusses the general techniques used to propagate the error
bounds (for precision) that are calculated in Section 4.5.1. The error bounds of two numbers
can be propagated to determine the error bound of their sum and/or of their product. These
techniques are covered in more detail in Volume 4 on statistical methodology,3 but are
summarized here for the reader's convenience. Products are often used in this study since
the basic extrapolation technique was to take the product of AF X EF to obtain the source's
emission rate (see Section 2). Sums are also used frequently since all of the individual
source emission rates are summed to obtain the national annual emissions from the natural
gas industry.
Section 4.5.1 discussed the calculation of 90% confidence half widths for a
single term, such as an emission factor or activity factor. These confidence half widths can
be plugged into the following equations to determine the confidence bounds for sums and
products.
For uncorrelated values, the error bound (90% confidence half width) of a sum
is the square root of the sum of the squares of the absolute errors of the values being
summed. An illustrative example follows. Suppose the following values, A and B, are to be
summed, and that the confidence bound of value "A" is expressed as "a" (in absolute terms).
The bottom cell of Table 4-3 shows the resulting error calculation for the sum.
Thus, it must be
(7)
24
-------
TABLE 4-3. ERROR PROPAGATION FOR A SUM
Values to be
Summed
90% Confidence Half Widths
Absolute Value
Percent Value
A
a
a% = 100 x a/A
B
b
b%= 100 x b/B
Sum = (A + B)
absolute error of sum = square root of (a2 + b2)
For correlated values, the equation for error becomes (a2 + b2 + 2rab)1/2, where r =
correlation coefficient between A and B. However, r was assumed to be zero for most
categories since they were derived from different data and were unrelated.
The error bound (90% confidence half width) on a product is also calculated
with the absolute errors of the terms being multiplied. Suppose that A X B = C, and that
the errors for A and B are expressed as a and b, respectively. The errors expresed as a
fractional value would be fa and fb, respectivley. The bottom cell of Table 4-4 shows the
resulting error calculation for the product.
TABLE 4-4. ERROR PROPAGATION FOR A PRODUCT
Values to be
Multiplied
90% Confidence Half Widths
Absolute Value
Fractional Value
A
a
fa = a/A
B
b
fb= b/B
Product = A x B
relative error of product = square root of [fa2 + fb2 +(fa2 x fb2)]
25
-------
4.5.3
Screening for Bias
It is impossible to prove that there is no bias in any data set. While tests can
be designed that are capable of revealing some bias, there are no tests nor group of tests that
will reveal all possible biases. Assuming that a data set has no bias, even after extensive
testing, is only a hypothesis. Such hypotheses can be disproved, but cannot be proven.
However, the data collected during this project were carefully checked and rechecked to
eliminate all identifiable biases. Three basic methods were used to screen for bias: peer
review of the activity factors by experienced sources, analysis of the data, and extrapolation
by different parameters (EPs).
Methods for gathering the data were tested repeatedly through extensive
technical and industrial review. Numerous individual reviews and project advisor's meetings
over a number of years were used to examine the project data with industry representatives
and other experts so that systematic errors could be identified 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. One example of this review
process is the identification of regional differences in production practices. These differences
were identified during the advisor meeting review process. The regional bias was then
overcome by subdividing the production data into two offshore and four onshore regions,
sampling within each region, and extrapolating by region.
Some of the production data that were collected and analyzed were also
checked for bias by extrapolating the activity factors with multiple parameters (i.e., EPs).
For a subset of data that is perfectly representative of the gas industry, equipment counts
from the data set could be extrapolated to national totals by any variable in the data set. Any
extrapolation from the perfect subset of data would deliver the correct answer, regardless of
the parameter used. For an imperfect data set, which all data sets are, extrapolation by
multiple variables allows for a cross check for bias. For example, in production, the
equipment counts can be extrapolated by production rate or well count. These two methods
26
-------
often produced different answers that were averaged to minimize the potential bias from a
single method.
27
-------
5.0 PRODUCTION SEGMENT
Some of the production segment activity factors were well-known, published
figures, but most production activity factors had to be developed. Production activity factors
were extrapolated from observed equipment counts at sampled sites based on the only three
nationally known extrapolation parameters: total gas well count, gas industry well count,
and gas production rate. Any scaled-up site data were extrapolated using the ratio of site
wells to national wells or using the ratio of site production to national production.
5.1 Development of a Regional Approach
The extrapolations were performed on a regional basis since regional
differences were known to exist and because each of the well-known activity factors (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 upon known differences in practices in various regions. The regions
are: Gulf Coast Onshore, Gulf Coast Offshore, Central Plains (onshore), Atlantic and Great
Lakes (onshore), Pacific and Mountain (onshore), and Pacific Offshore. Figure 5-1 shows
the major producing states and Figure 5-2 shows the regions selected and which states are
included.
The differences in the regions justify their selection and can be seen in Table
5-1. Specifically, Table 5-1 shows the regional differences that exist in production versus
well count. Each region has a unique oil well versus gas well split and has 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 and 1.2% of the oil wells) but produce 27.7% of the U.S.
marketed gas production.
28
-------
States with Production
>50,000 MMscf/yr
Figure 5-1. Major Producing States
-------
ntfial Plai
l&iSn
P.M. Offshore
G.C. Offshore
Figure 5-2. Selected Production Regions
-------
TABLE 5-1. REGIONAL DIFFERENCES IN PRODUCTION RATES AND WELL COUNTS
1992 Producing Gas
1992 Producing Oil
1992 Marketed
1992 Gross
Wells*
Wells9
Productionh
Productionb
States in Region
Percent of
Percent
that are
Percent of
Percent of
Marketed
of Gross
Regional Groupings
> SO Bscfy
Count
Total
Count
Total
Bscfy
Total
Bscfy
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,
80924
29.3
199103
33.1
5424
29.0
5672
25.6
WY,KS
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,
129157
46.8
138805
23.0
790
4.2
796
3.6
(onshore)
WV
TOTAL U.S.
276014
100.0%
602197
100.0
18712
100.0%
22132
100.0%
a Table 3-17, Gas Facts2
b Natural Gas Annual5
-------
In regard to the other regions, as shown in Table 5-1, the majority of natural
gas production in the United States (61.5% of marketed production) is in the Gulf Coast
Region. All other regions only account for 38.5% of the production, and the majority of that
fraction of production occurs in the Central Plains region. However, the split on well count
is completely different. The Atlantic & Great Lakes region, which accounts for only 4.2%
of the marketed national gas production, has the largest portion of gas wells (46.8% of the
national total) as well as a large fraction of oil wells (23%).
If wells or equipment associated only with wells (such as heaters and chemical
injection pumps) are the sources being evaluated, then bias would potentially be introduced if
sampling was performed only on wells in the Gulf Coast region, where most of the gas is
produced. The bias is avoided if the sources are combined regionally, and the regional
averages should then be added in the same proportion that they are distributed in the actual
population.
Table 5-2 shows the relative representativeness of the sample data set for the
region that it represents. The percentage of site samples (site visits and surveys) in the
region can be compared with the actual percent of national well count and national
production that the region represents. The sample set does not exactly match the actual
regional averages for production per well. The sample sites have a higher production per
well than the actual region, probably due to lower participation of sites with very low
production rates. (Lower production rates may be tied to fewer resources and less ability to
participate in surveys.)
However, this bias has not caused significant error in the data set since the
extrapolation technique for activity factors does not use the production rate per well.
Instead, the extrapolation technique uses both well count and production rate independently.
Therefore, the high production rate per well should not introduce any error into the final
estimates.
32
-------
TABLE 5-2. SAMPLE SET COMPARISON TO ACTUAL REGIONAL DATA
Regional Groupings
Sample Data Set*
1992 Actual Regional Datab'c
Number
of Site
Samples
Percent of Total
Site Samples
from this Region
Production Rate
per Gas Industry
Well (MMcfd)
Percent of
Total Oil
Percent of
Total Gas
Wells
Percent of
Marketed
Gas Total
ProductionRate
per Gas Industry
Well (MMcfd)
Marketed
Gross
Marketed
Gross
Gulf Coast Region Total
GC Offshore
GC Onshore
13
4
9
26.0
8.0
18.0
1.66
0.54
1.66
0.54
36.1
0.9
35.3
23.1
1.5
21.6
61.5
26.7
34.8
2.36
0.13
2.38
0.15
Central Plains (onshore)
7
14.0
0.22
0.22
33.1
29.3
29.0
0.10
0.10
Pacific and Mountain Total
PM Offshore
PM Onshore
11
2
9
22.0
4.0
18.0
5.22
0.12
5.22
0.32
7.8
0.3
7.4
pop
bo o bo
5.3
1.0
4.3
0.66
0.12
0.99
0.48
Atlantic and Great Lakes (onshore)
19
38.0
0.018
0.019
23.0
46.8
4.2
0.012
0.012
TOTAL U.S.
50
100.0
0.14
0.16
100.0
100.0
100.0
0.106
0.125
a Includes Radian site visits (24), Radian phone surveys (15), and Star site visits (12)
b Table 3-10, Gas Facts2
c Natural Gas Annual5
-------
5.2
Gas Well Count and Production (Onshore and Offshore)
Gas well counts for 1992 were available from several sources (Gas Facts,
Table 3-17,2 or from World Oil,6 or from DOE EIA's Natural Gas Annual5). These sources
give total U.S. gas wells for each year. The data are also sorted by state, which allowed
regional groupings to be made, as was done in Table 5-1.
The split of counts into offshore and onshore wells has to be made for the
states that have offshore production in state waters; these states are Alabama, California,
Louisiana, Texas, and Alaska. World Oil (Feb 93) gives figures for offshore gas well counts
for Alabama, California, and Louisiana. Texas and Alaska data were not split. Texas data
were obtained from contacts with the Texas Railroad Commission (RRC).7 The Alaska
offshore count was provided by the Alaska Oil and Gas Conservation Commission.8
Gas well production data for 1992 were also available from several sources
(Gas Facts, Table 3-17,2 or from the U.S. Department of Energy (DOE) EIA's Natural Gas
Annual5). These sources give total U.S. gas wells for each year. The data are also sorted by
state which allowed regional groupings to be made, as was done in Table 5-1. Production
was sorted into onshore and offshore production using DOE EIA's Natural Gas Annual,
Table 4, "Offshore Gross Withdrawals of Natural Gas By State, 1992."
5.3 Oil Well Count and Production (Onshore and Offshore)
Oil well counts for 1992 were available from the Oil and Gas Journal and
from World Oil (Feb 1993).6 Similar to gas wells, the onshore and offshore split of oil wells
was made by World Oil for Alabama, Alaska, California, and Louisiana. Texas data were
not split. Again, Texas offshore oil well data were obtained from contacts with the Texas
Railroad Commission.7
34
-------
Gas production from oil wells for 1992 was also available from DOE EIA's
Natural Gas Annual5. These sources give total U.S. gas wells for each year. The data are
also sorted by state, which allowed regional groupings to be made, as was done in Table 5-1.
Production was sorted into onshore and offshore production using DOE EIA's Natural Gas
Annual, Table 4, "Offshore Gross Withdrawals of Natural Gas By State, 1992."
5.4 Oil Wells Marketing Gas
Oil wells that produce natural gas and sell the gas into the natural gas system
have been considered a part of the natural gas industry for the purposes of this report. The
wellhead is not considered part of the gas industry, but downstream gas production
equipment associated with these wells is included. However, oil wells that do not produce
gas, or that produce gas that is vented or consumed on site are not considered part of the gas
industry. Methane emissions from this latter type of oil well are not included in our
estimates.
There are no national statistics on the fraction of oil wells that market gas.
Only the total oil well count is known. Even differentiation of gas wells from oil wells are
based on a gas-to-oil ratio (GOR) and may vary state-to-state. The definition in Texas for an
oil well is any well that has a GOR lower than 100,000 cubic feet (of gas) per barrel (of oil).
A GOR greater than 100,000 qualifies the well as a gas well.
Since there are no national statistics on oil wells that market gas, the national
statistics on total oil wells had to be manipulated to produce this number. To determine the
number of oil wells that produce gas which is not marketed, detailed state data from
Louisiana, Colorado, California, and Texas were examined for disposition designations that
would indicate whether gas from the wells was marketed or internally consumed. Only
Texas data allowed such a differentiation.
35
-------
Texas RRC magnetic tapes for 1989's P-l and P-2 disposition reports were
purchased, downloaded, and analyzed with a database management program.10 Oil wells are
not tracked on an individual well basis but are tracked on a lease basis, so the data will be
discussed on a lease basis. The analysis showed that oil well leases with some marketed gas
account for only 34.7% of all oil well leases. This figure was obtained by taking the mean
of the percentages for each month's worth of data. The minimum observed monthly value
was 28.0% (March 1989), and the maximum was 46.7% (October 1989). The standard
error, used in the calculation of statistical confidence limits, was 2.8 percentage points, or
7.8% of the mean value. The reader should note that this error demonstrates the variability
of the Texas data and has been assumed to be equal to the variability of the national data.
Dwight's Energy Data, Inc. was asked to query their database to determine the
fraction of oil leases that were producing gas for sale in Oklahoma. 11 They found that 52%
of the oil leases in Oklahoma had sold gas. However, Dwight's was unable to confirm what
GOR was used to differentiate between an oil lease and a gas lease. Thus, the 52% was not
used in the calculation of associated gas wells, but was thought to confirm the magnitude of
the 34.7% value that was obtained from the RRC tapes.
Total gas industry well counts were determined by multiplying known regional
oil well counts by 0.347, and adding these oil wells to all of the gas wells. These
calculations were performed on a regional basis, as shown in Table 5-3. The total gas
industry well counts were used to extrapolate populations of chemical injection pumps, in-
line heaters, separators, and compressors as detailed in Section 5.7.
5.5 Offshore Platforms
A separate method for determining offshore platform count was necessary
because offshore production fugitive studies produced emission rates per platform, rather
than per equipment. While published data are available on the number of active drilling rigs,
no published data were found on the number of producing platforms. Two separate sources
were contacted to obtain the required information.
36
-------
TABLE 5-3. NATIONAL GAS INDUSTRY WELLS
Region
Gas Wellsb
(1992 data)
Oil Wells"
(1992 data)
Oil Wells
Marketing Gas"
(1992 data)
± 90%
Confidence
Limits
(absolute)
Total Gas
Industry Wells
(1992 data)
±90%
Confidence
Limits
(% of Mean)
Gulf Coast Offshore
4021
5140
1780
133
5800
2.3
Gulf Coast Onshore
59646
212427
73700
5479
133000
4.1
Central Plains (onshore)
80924
199103
69100
5135
15000
3.4
Pacific Offshore
65
2040
708
53
773
6.5
Pacific & Mountain Onshore
2201
44682
15500
1152
17700
6.5
Atlantic & Great Lakes (onshore)
129157
138805
48200
3580
177000
2.0
TOTAL U.S. (1992 data)
276014
602197
209000
15533
485000
3.2
a Based upon analysis of Texas RRC data that shows 34.7% of all oil wells market gas.7'10
b Source on well counts: Gas Facts.1
-------
The first source for platform count information was Offshore Data Services,
Inc.,12 which tracks statistical data on offshore production. They estimated that in 1993 there
were 1607 producing platforms in the Gulf of Mexico, and 45 platforms in the rest of the
United States. This includes platforms in state waters.
The second source for platform data was the Minerals Management Service
MMS Outer Continental Shelf Activity Database (MOAD).13 This database provided a count
of 1,857 producing platforms in the Gulf of Mexico (excluding platforms located in state
waters). However, since MOAD only has an 85% response rate from industry, the number
was extrapolated to 2,185. MOAD does not account for platforms located in state waters in
the Gulf.
The MOAD number could be overestimated since it counts some platforms with no
production, and could be underestimated since platforms in state waters were not included.
However, since the Offshore Data Services number was lower, the MOAD count was likely
not underestimated. The MOAD data were used for the Gulf of Mexico platform count
because they were conservatively high and could be examined more readily. The sum of the
extrapolated MOAD Gulf of Mexico count and the Offshore Data Services count for the rest
of the United States was used in the study.
The total platform count was split 50/50 between "oil" industry and "gas"
industry platforms. No data were available to define the actual split, but industry reviewers
approved this approach. The total counts for the two regions and the split of gas industry
platforms are shown in Table 5-4.
The total offshore oil and gas well count is 11,300 (see Table 5-3). The
result, when divided by the total platform count, is 5.1 wells per platform. Site visit data on
platforms produced a ratio of 5 to 10 wells per platform. This ratio validates the relative
magnitude of the platform counts provided by Offshore Data Services and the MOAD
38
-------
TABLE 5-4. PLATFORM COUNTS
Total
Gas Industry
Region
Platform Count
Platform Count
Rest of U.S.
45
22
Gulf of Mexico
2,185
1,092
Total U.S. Offshore
2,230
1,114
database. Therefore, the 90% confidence interval was assigned to be + 10% based on
engineering judgement.
5.6 Specific Equipment
Nationally tracked data are not available for most equipment in the production
segment. Therefore, this study estimated total equipment in the production segment from
specific counts taken during the study's site visits. A total equipment count for a facility was
estimated while on site. Then the site equipment counts were extrapolated to regional counts
with an extrapolation parameter (something that is known both for the site and for the
region). Section 4.4, Equation 3 discussed the method used for extrapolation.
For major equipment activity factors in the production segment, equipment
counts could be extrapolated by either active wells or marketed production for each site. As
explained in Section 4.4.2, the active wells extrapolation parameter can be either gas only
wells or total "gas industry" wells. "Gas industry" wells include gas wells and oil wells that
also market gas. These three extrapolation methods (gas wells, total wells, and production
rate) can yield very different results. Therefore, technical relations were used to select the
appropriate extrapolation parameter(s), as was explained in Section 4.4.2, Table 4-3.
The extrapolations were also performed on a regional basis in order to account
for regional differences in the key parameters such as operating practices, gas quality, well
39
-------
production rates, equipment design, and field ages. The site data were organized into
regions, and regional ratios were then determined. The regional ratios were then multiplied
by the regional count of active wells (gas and total wells) or by regional marketed production
to obtain a total count of equipment in the region. Regions were added together to determine
the national number. The final national equipment counts are summarized in Table 5-5.
TABLE 5-5. SPECIFIC EQUIPMENT ACTIVITY FACTORS
Activity Factor
Count and Confidence Limits
Separators
167,000 + 28%
Heaters
51,000 + 95%
Pneumatics
249,000 ± 48%
Chemical Pumps
17,000 ± 143%
Compressors
17,100 + 52%
The results of the extrapolations for separators, in-line heaters, pneumatics,
chemical injection pumps, and compressors are presented in Tables 5-6 through 5-10. The
last column in Tables 5-6 through 5-10 presents the selected activity factor value. The site
equipment count data used in developing the regional counts in Tables 5-6 through 5-10 are
presented in Appendix A.
5.7 Vessels and Pressure Relief Valves
Specific equipment activity factors were combined and used to extrapolate
emission rate estimates for vessel blowdowns and pressure relief valves (PRVs). The
number of production "vessels" (used in estimating vessel blowdown emissions) was based
on the sum of separators and in-line heater counts from Tables 5-6 and 5-7 and the number
of dehydrators (discussed in Section 8), resulting in 256,000 vessels. Compressors were not
included in the vessel activity factor since emission factors for compressor start-ups and
blowdowns were treated separately. A confidence interval of + 26% was determined by
40
-------
TABLE 5-6. SEPARATORS IN PRODUCTION
Gas Well Basis :
( Marketed Gas Basis
Arith. Mean
(Gas and Mkt.)
Pneumatic Devices
Activity Factor
Equip/Gas Well
Extrapolated
Count
90%
Conf
Activity Factor
Equip/MktGas
Extrapolated
Count
90%
Conf
Wt. Mean
Equipment
Count
90%
Conf
Final
Extrapolated
Count
NATIONWIDE Basis (non-regional)
NATIONWIDE Sum by Region
Not
Applicable
0.727
200,566
200,312
26%
4.182
214,399
134,558
36%
167,172
Gulf Coast Offshore Region
0.335
1,349
129%
0.060
825
42%
+/-
Gulf Coast Onshore Region
0.985
58,763
23%
0.931
16,614
48%
-0.310
37,689
17%
28%
Central Plains Region
0.486
39,335
119%
2.144
31,865 .
123%
0.868
35,600
117%
Pacific/Mountain Offshore Region
0.083
5
300%
0.006
3
101%
Pacific/Mountain Onshore Region
0.849
1,869
300%
0.412
902
268%
0.700
1,385
271%
Atlantic/Great Lakes Region
0.766
98,991
17%
38.972
84,349
32%
0.859
91,670
23%
-------
TABLE 5-7. IN-LINE HEATERS IN PRODUCTION
Gas Well Basis
Marketed Gas Basis
In-Line Heaters
Total Well
Basis
Activity Factor
Equip/Gas Well
Extrapolated *
Count
90%
Conf
Activity Factor
Equip/Mkt Gas
Extrapolated
Count
Conf
i,
Final
Extrapolated
Count
NATIONWIDE Basis (non-regional)
Not
0.282
77,865
Not Applicable
Not
NATIONWIDE Sum by Region
Applicable
51,000
95%
Applicable
51,000
Gulf Coast Offshore Region
0.000
0
0%
+/-
Gulf Coast Onshore Region
0.224
13,342
121%
95%
Central Plains Region
0.435
35,197
128%
Pacific/Mountain Offshore Region
0.000
0
0%
Pacific/Mountain Onshore Region
1.000
2,201
300%
Atlantic/Great Lakes Region
0.002
260
196%
-------
TABLE 5-8. PNEUMATIC DEVICES IN PRODUCTION
Pneumatic Devices
Total Well Basis
Gas Well
: ^ Basis
Marketed Gas Basis
Arith. Mean
(Gas and Mkt.)
Activity Factor
Equip/Gas Well
Extrapolated:
90% Conf
Activity Factor
Equip/Mkt Gas
Extrapolated
Count
Conf
Wt. Mean
Equipment
Count
90%
Conf
Final
Extrapolated
Count ::
NATIONWIDE Basis (non-regional)
0.517
250,958
3.458
177,253
NATIONWIDE Sum by Region
353,356
56%
144,867
35%
249,111
48%
249,111
Gulf Coast Offshore Region
0.127
739
224%
Not
0.077
1,049
26%
-0.102
894
92%
+/-
Gulf Coast Onshore Region
0.575
76,649
64%
Applicable
1.060
18,917
69%
0.930
47,783
64%
48%
Central Plains Region
1.325
198,890
94%
4.947
73,509
42%
0.950
136,199
80%
Pacific/Mountain Offshore Region
0.000
0
0%
0.000
0
0%
0.000
0
0%
Pacific/Mountain Onshore Region
0.080
1,423
231%
0.666
1,455
249%
0.703
1,439
222%
Atlantic/Great Lakes Region
0.427
75,654
54%
23.072
49,937
77%
0.950
62,796
62%
-------
TABLE 5-9. CHEMICAL INJECTION PUMPS IN PRODUCTION
Gas Well Basis
Marketed Gas Basis
Arith. Mean
(Gas arid Mlit.)
Chemical Injection Pumps
Well Basis
Activity Factor
Equip/Gas Well
Extrapolated
¦Count
90%
Conf
Activity Factor
Equip/Mkt Gas
Extrapolated
Count
90%
Conf
r
Wt. Mean
Equipment
Count
90%
Conf
Filial
Extrapolated
Count
NATIONWIDE Basis (non-regional)
NATIONWIDE Sum by Region
Not
Applicable
0.036
9,903
16,971
143%
Not Applicable
Not Applicable
16,971
Gulf Coast Offshore Region
0.000
0
0%
+/-
Gulf Coast Onshore Region
0.033
1,962
142%
143%
Central Plains Region
0.159
12,857
184%
Pacific/Mountain Offshore Region
0.000
0
0%
Pacific/Mountain Onshore Region
0.679
1,495
300%
Atlantic/Great Lakes Region
0.005
657
48%
-------
TABLE 5-10. COMPRESSORS (NON-GAS LIFT) IN PRODUCTION
Compressors
(excluding gas lift)
Total Well Basis
Gas Well
Basis
Marketed Gas Basis
Arith. Mean :
(Gas and Mkt.)
Final
Extrapolated
Count
Activity Factor
Equip/Gas Well
Extrapolated
Count
90% Conf
Activity Factor
Equip/Mkt Gas
Extrapolated
Count
90%
Conf
Wt. Mean
Equipment
Count :
90%
Conf
NATIONWIDE Basis (non-regional)
0.067
32,317
0.195
9,977
NATIONWIDE Sum by Region
26,000
59%
8,223
41%
17,112
52%
17,112
Gulf Coast Offshore Region
0.000
0
0%
Not
0.000
0
0%
0.000
0
0%
+ /-
Gulf Coast Onshore Region
0.067
8,991
129%
Applicable
0.124
2,219
119%
0.950
5,605
126%
52%
Central Plains Region
0.095
14,189
70%
0.347
5,157
38%
0.157
9,673
54%
Pacific/Mountain Offshore Region
0.167
129
300%
0.013
6
300%
0.700
68
296%
Pacific/Mountain Onshore Region
0.142
2,512
55%
0.348
761
121%
0.903
1,637
69%
Atlantic/Great Lakes Region
0.001
179
37%
0.037
80
23%
0.950
129
33%
-------
combining the confidence intervals for each equipment type using the square root of the sum
of the squares approach (discussed in Volume 4 on statistical methodology3).
An activity factor for the number of PRVs was based on an observed PRV
count per equipment type as shown in Table 5-11.
TABLE 5-11. PRESSURE RELIEF VALVES
Equipment Type
PRV Count
Per Equipment
Separators
2 + 68%
Heaters
1 ± 89%
Compressors
4 + 84%
Dehydrators
2 + 53%
Details of these observations are provided in Volume 8 on equipment leaks.14
Using the PRV ratios shown above and the equipment counts reported in Sections 5 and 8,
results in an estimated 529,400 + 53% PRVs in production.
5.8 Gathering Pipeline Miles
Total gathering pipeline mileage is not reported or tracked nationally and must
be estimated. The gathering pipeline mileage was divided into three segments. The first of
these was gas well gathering pipeline miles, and the second was miles for oil wells that
market gas. The third segment consisted of gathering pipeline miles owned by transmission
companies.
Total miles of gathering pipeline were estimated using site visit data from the
thirteen production sites shown in Table 5-12. 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.
46
-------
Gas well gathering pipeline miles were determined using data from the thirteen
sites. The gathering pipeline miles and the gas wells from the sites were each summed. The
sum of the gathering pipeline miles (1,359) was divided by the sum of the gas wells (2,033)
to determine an average number of pipeline miles per gas well (0.668). This value was
extrapolated by the nationwide activity factor for gas wells (276,000) to give the nationwide
activity factor of 184,000 gathering pipeline miles for gas wells.
Gathering pipeline mileage for oil wells that market gas was based on the gas
well mileage data. It was assumed that one half of the gathering pipeline mileage at an oil
well that markets gas was in the oil industry, and half was in the gas industry (see Figure
3-2). Therefore, the average number of pipeline miles per oil well marketing gas was
divided by two to determine the average number of pipeline miles per associated gas well
(0.334). This average was extrapolated by the nationwide activity factor for these wells
(209,000) to give the nationwide activity factor of 70,000 gathering pipeline miles for oil
wells that market gas.
The third segment, gathering pipeline miles owned by transmission companies,
is reported by A.G.A. to be 86,200 miles (Table 5-12). Utility-owned pipelines are assumed
to be part of the total production gathering pipeline miles, and were not included in the
mileage counts in the production site visit data.
The total production gathering pipeline miles from gas wells, from oil wells
that market gas, and from transmission companies was summed. This sum gives a
nationwide activity factor of 340,200 for production gathering pipeline miles. A rigorous
determination of the 90% confidence interval gave an error less than + 4%. This value was
considered to be too low based on the quality of the data that were used to generate the
activity factor. Thus, a 90% confidence interval of + 10% was assumed based on
engineering judgement.
47
-------
TABLE 5-12. PRODUCTION GATHERING PIPELINE MILEAGE - GAS WELLS
Site
Gathering
Number of 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
a Includes 55 oil wells.
5.9 Maintenance Activities
Emission rate estimates for well workovers and completion flaring were based
on annual counts of a specific activity, as opposed to an equipment count. Completion
flaring occurs immediately following the drilling process for a new well, where the gas is
flared to determine the available pressure and flow rate. Most completion flaring is
associated with exploratory wells to allow proper sizing of meters and surface equipment.
(The production rates for new wells in existing fields can usually be determined before the
well is completed.) The Energy Information Agency (EIA)15 estimated 844 wells are
48
-------
completed each year, based on 1987 data. A confidence interval of 10% was assigned to this
value based on engineering judgement.
Well workovers are another type of maintenance activity. During a well
workover, the tubing is pulled from the well to repair tubing corrosion/erosion or other
downhole equipment problems. A study performed by Pipeline Systems Incorporated (PSI)
reported the number of gas well workovers for two sites.16 PSI data showed 1 workover/yr
per 21 gas wells at Site 1, and 1 workover/yr per 50 gas wells at Site 2. The annual average
for these two sites is 0.0338 + 258% workovers/gas well. (Workovers for oil wells are
outside the natural gas industry boundaries defined for this report.) The confidence interval
for this value was calculated using Equation 5. Multiplying this ratio by the total number of
gas wells (276,014 + 5% from Table 5-1) results in the number of well workovers per year:
276,014 gas wells x 0.0338 workovers/yr-well = 9,329 workovers/yr
A confidence interval of 258% was calculated based on propagating the confidence intervals
for the terms in the equation above.
A third type of maintenance, venting or gas well unloading, requires an
activity factor for the number of gas wells that require unloading. Gas wells operated under
low pressures can accumulate water in the wellbore due to their low flow rate. This water
chokes the flow of the well, reducing the gas production. To clear the water, the well is
blown to a tank at atmospheric pressure, where the gas is vented. Data from 22 production
site visits indicated that 41% (+ 45%) of gas wells require this unloading practice to
maintain production. Additional details on this particular activity factor can be found in the
Volume 7 on blow and purge activities.17 Multiplying the total number of gas wells
nationally (276,014 + 5%) by this value results in 114,139 + 45% gas wells requiring
unloading each year.
49
-------
5.10 Compressors and Dehvdrators
Activity factors for compressors and dehydrators were calculated in a single
extrapolation procedure for all segments of the natural gas industry. The details of this
procedure are discussed in Section 8.
50
-------
6.0 GAS PROCESSING SEGMENT
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 number of gas plants (of each type), dehydrators, acid gas
recovery units (AGRs) and compressors. All of these activity factors were either published
and well-defined or were developed through other studies such as the Wright Killen report18
or Volume 11 on compressor driver exhaust.19
6.1 Gas Plant Count
Gas processing plants recover high-value hydrocarbon liquids from natural gas.
The plants can be divided by process type based upon data from the Oil & Gas Journal
annual plant survey9 as shown in Table 6-1:
TABLE 6-1. DIVISIONS OF GAS PROCESSING PLANTS BY TYPE
Type
Count
% of Total
Cryogenic
302
42
Refrigerated absorption
117
16
Refrigeration
192
26
Other
115
16
TOTAL PLANTS
726
100
Since the sites visited sampled all of the major types, and since no emission
factor data were correlated to plant type, only the total plant count was used for the plant
activity factor. Based upon 1992 data from the Oil and Gas Journal, the total gas plant
count is 726. A confidence limit for the gas plant count was set at + 2% based upon
engineering judgement.
51
-------
6.2
Acid Gas Recovery (AGR) Units
Acid gas recovery (AGR) units are used to remove carbon dioxide (C02) and
hydrogen sulfide (H2S) from natural gas. Absorption of C02 and H2S in aqueous amine
solutions (e.g., monoethanolamine, diethanolamine) is the most widely used process for
AGRs. Some methane is absorbed along with the acid gas in the amine solution and may be
emitted when the solution is regenerated by heating and decrease in pressure. Purvin &
Gertz, Inc., estimated the total count of AGRs using an amine process to be 371, with a total
capacity of 30,433 MMscfd.20 Most AGRs have flares or sulfur recovery processes that
prevent the release of methane. However, it is estimated that only 18% of AGRs vent
regenerator offgas (including acid gas and methane) to the atmosphere.21 Other AGR
processes may result in methane emissions but are much more difficult to quantify because of
the wide variety of processes available.
6.3 Compressors and Dehvdrators
Activity factors for compressors and dehydrators were calculated in a single
extrapolation procedure for all segments of the natural gas industry. The details of this
procedure are outlined in Section 8.
52
-------
7.0 TRANSMISSION SEGMENT
Activity factors for the transmission segment are easier to determine than
production segment factors because the transmission segment is more homogeneous than the
production segment. The transmission segment has only four basic components: pipeline,
compressor stations, storage facilities, and meter and regulating (M&R) stations. Most
transmission pipelines are one of two types: cross-country or regional network.
7.1 Compressor Station Count
The total number of compressor stations was calculated to be 1700 + 10%.
The number was extrapolated from data submitted by major transmission companies to the
Federal Energy Regulatory Commission (FERC).22 Pages 508 and 514 of FERC Form 2 list
the number of compressor stations and miles of pipeline for 46 transmission companies. The
pipelines owned by these companies comprise 70% of the total transmission pipeline mileage,
making the sample large but incomplete. The total number of transmission compression
stations reported in the forms was extrapolated by pipeline mileage to determine the total
number of compression stations in the transmission segment. A summary of the compression
station data obtained from FERC Form 2 is presented in Table 7-1. The confidence interval
of ± 10% was based on engineering judgement.
In a 1990 survey titled "Natural Gas Industry Environmental Organization
Structure Survey," A.G.A. surveyed 14 gas transmission companies and found an average of
138 miles per station.23 The number of miles per station corresponds to a higher count of
2062 compressor stations. However, the 1990 A.G.A. survey was less formal, and the
stations used in the survey also included some gas processing plants; therefore, the 2062
value determined by the A.G.A. is an overestimate of transmission compressor stations. Its
relative size does, however, lend credence to the selected activity factor of 1700 stations.
53
-------
TABLE 7-1. TRANSMISSION COMPRESSOR STATION COUNTS
Company
Number of Stations
Miles of Pipeline
1
5
1,023
2
47
9,915
3
29
6,556
4
28
3,253
5
64
4,517
6
119
11,188
7
13
4,288
8
78
10,244
9
6
495
10
17
1,149
11
20
4,460
12
14
1,960
13
1
203
14
0
375
15
5
925
16
51
9,725
17
2
548
18
7
354
19
15
1,972
20
1
362
21
2
552
22
12
1,457
23
56
10,973
24
7
971
25
48
24,182
26
42
3,551
27
0
89
28
13
799
29
58
6,532
30
13
1,718
31
2
462
32
48
7,896
33
2
275
Continued
54
-------
TABLE 7-1. (Continued)
Company
Number of Stations
Miles of Pipeline
34
69
14,666
35
69
9,386
36
28
5,819
37
0
436
38
43
10,670
39
20
4,194
40
19
4,218
41
0
30
42
29
7,272
43
8
549
44
45
6,093
45
30
3,224
46
1
269
Total
1186
199,795
Extrapolated No. of Stations
1,700
55
-------
7.2
Storage Station Count
There are two types of gas storage facilities: liquefied natural gas (LNG)
storage and underground storage. LNG facilities condense the natural gas at high pressures
for storage and transportation in refrigerated vessels. Underground gas storage areas
typically operate to accept excess gas from major gas pipelines during surplus production
flow periods. The stored surplus gas is reinjected into the pipeline when demand is higher.
Typically, the net amount of gas to storage for any year is zero; if there is an actual long-
term over-supply, the producing wells will simply be cut back.
Underground storage areas are unique. Unlike liquid storage facilities,
expensive tanks are not used. Underground gas storage facilities are typically one of the
following: old, spent gas fields; salt domes; or aquifers. The storage operator simply injects
the surplus gas into the underground formation, which effectively "stores" the gas.
Compression is sometimes necessary to inject the gas into the formation or to return the gas
to the pipeline.
Underground (UG) station count data were available from Gas Facts,
Table 4-5 "Amount of Pools, Wells, and Horsepower in Underground Storage Fields, By
State, 1990-1992.1,2 The number of LNG facilities was also available from Gas Facts Table
4-3, "LNG Storage Operations in the U.S. as of Dec. 31, 1987.1,24 The total number of
storage stations is the sum of the underground stations and LNG facilities as shown:
UG: 386 (storage stations)
LNG: + 89 (54 complete plants, 32 satellite plants, 3 import
terminals)
TOTAL: 475 (storage stations)
This project assigned a +5% accuracy based upon engineering judgement.
56
-------
7.3 Storage Well Count
Underground storage facilities have wellheads at the surface similar to
wellheads at producing gas wells. These wellheads sit above the wellbore's connection to the
underground reservoir. Each wellhead has components that can emit gas.
The number of active underground storage wells for 1992 is given in Gas
Facts, Table 4-5, as 17,999.2 This project assigned a + 5% accuracy since this number is
tracked annually by A.G.A.
7.4 M&R Station Count
Meter and regulating (M&R) stations are used to measure flow of gas at a
custody transfer point, and/or to reduce and regulate the pressure and flow into a
downstream pipeline system. Some metering facilities do exist within compressor stations,
but these have already been counted in the compressor station emissions. In the transmission
segment, separate M&R facilities are usually fenced, above ground facilities that contain
valves, piping, and meter runs.
In the U.S. gas industry, most M&R stations exist in the distribution and
production segments and have already been counted there. M&R facilities counted by the
transmission companies are generally of the following type:
1. Receipt from production (meter only),
2. Delivery to distribution (M&R),
3. Pressure reduction to inter-company transmission lines with low
maximum operating working pressure (MOWP),
4. Interconnects/custody transfer (bi-directional) to another transmission
company, or
57
-------
5.
Direct sales (farm taps, direct industrial sales from transmission lines).
The first three types are already counted in other activity factors as follows:
• In the production segment, meter runs that transfer the gas to
transmission companies were included in the well site fugitive
component count data.
• Delivery to distribution has been counted in the distribution segment
M&R stations (see Section 9 for more detail).
• Pressure reduction stations that reduce the pressure to a line that has a
lower maximum operable working pressure (MOWP) also are already
counted in the distribution M&R count (Section 9).
The two types of M&R stations that have not been included elsewhere are
transmission company interconnects and direct sales. Most large transmission companies
have interconnects with other transmission companies to allow for flexibility of supply.
These shared stations can flow in either direction. The last category (direct sales) is
comprised of two types: farm taps and direct industrial sales. Direct industrial sales from
transmission lines are owned by local distribution companies (LDCs), 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.
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 Table 7-2.
58
-------
TABLE 7-2. 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%
938 ± 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.
The activity factor for direct industrial sales was developed from FERC Form
No. 2, page 306.22 Industrial sales greater than 50,000 Mcf are listed individually, while
sales less than 50,000 Mcf are combined into a single item. In the latter case, the total
amount of gas sold was divided by 25,000 to provide an estimate of the number of sales.
The national activity factor was extrapolated from the FERC data to give a total count of 938
direct industrial sales. Due to the uncertainty that this approach introduced to the activity
59
-------
factor and to the complexity of retrieving data from FERC, a confidence bound of + 100%
was assigned based on engineering judgement.
7.5 Miles of Pipeline
The total miles of transmission pipeline equals 284,500 miles for 19922 and is
available from Gas Facts, Table 5-1. A confidence interval of + 5% was assigned based on
engineering judgement.
7.6 Compressors and Dehvdrators
Activity factors for compressors and dehydrators were calculated in a single
extrapolation procedure for all segments of the natural gas industry. The details of this
procedure are outlined in Section 8.
7.7 Pneumatics
The number of natural gas operated pneumatic devices in the transmission and
storage segment was calculated based on the average number of devices per station multiplied
by the total number of transmission and storage stations nationally. Data for determining the
average number of pneumatic devices per station are presented in Volume 12 on pneumatic
devices.25 The total number of transmission and storage stations is the sum of 1,700
transmission compressor stations, 386 underground storage facilities, and 89 liquid natural
gas storage facilities (2,175).
60
-------
8.0 CROSS-SEGMENT ACTIVITY FACTORS
Many equipment types are unique to a particular segment of the natural gas
industry. The activity factors from these types of equipment, such as the count of
distribution customer meters, were determined for each segment and are discussed in detail
in the Sections 5, 6, 7, and 9. However, a few equipment types are common to all segments
of the gas industry. Activity factors for these equipment types were determined in one step
for the entire industry. This section of the report discusses two equipment types,
compressors and dehydrators, that were determined on an industry-wide basis.
8.1 Compressors
Activity factors were developed for compressor horsepower-hours and for
compressor counts on an industry-wide basis. Compressor horsepower was used as the basis
for the compressor exhaust calculations, and compressor counts were used as the basis for
the fugitive and unsteady emission calculations. Compressors exist in all segments of the
industry except distribution.
8.1.1 Compressor Horsepower-Hours
The two main types of compressor drivers in the industry are reciprocating gas
engines and gas-fired turbines. Two pieces of information are needed to calculate the
activity factors, expressed as horsepower-hours (hp-hr), for the two types of drivers in each
segment of the industry except production. These data are the installed horsepower and the
average annual operating hours. For the production segment, horsepower-hour data were
available for the activity factor calculation. Table 8-1 presents the necessary parameters and
the resulting activity factors for both engines and turbines by industry segment. This table
also includes the 90% confidence limits for each factor.
61
-------
TABLE 8-1. COMPRESSOR ACTIVITY FACTORS FOR EACH INDUSTRY SEGMENT
Industry Segment
Installed
Engine
MMhp *
Installed
Turbine
||l|pii||l||||l
Annual
Hours
Engine
Annual
Hours
Turbine
Engine
MMHphr
Turbine
MMHp-hr
Production
NA
NA
NA
NA
27,460 ± 200%
NA
Processing
4.19 ± 132% b
5.19 ± 99% b
6626 ± 12%
6345 + 48%
27,760 ± 133%
32,910 ± 121%
Transmission:
Compressor Drivers
Generator Drivers
10.2 ± 10%
1.45 + 23%
4.55 ± 10%
0.045 ± 166%
3964 ± 14%
1352 ± 38%
2118 ± 31%
474 ± 620%
40,380 ± 17%
1962 ± 45%
9635 + 33%
21.2 ± 1215%
Storage
Compressor Drivers
Generator Drivers
1.33 ± 14%
0.085 ± 126%
0.59 + 14%
0.057 ± 184%
3707 ± 23%
191 ± 377%
2917 ± 620%
36 ± 620%
4922 ± 27%
16.3 ± 621%
1729 ± 626%
2.05 ± 1312%
" Does not include horsepower associated with gas lift for oil recovery or with electric drivers.
b Average of two estimation methods.
-------
To determine the horsepower and operating hours for engines and turbines,
four major sources of data are available: 1) the methane project site visits, 2) the A.G.A. /
GRI/Southwest Research Institute (SwRI) compressor databases, 3) the Federal Energy
Regulatory Commission (FERC) Form No. 2 Database, and 4) individual company
databases. The site visits and company databases were used to determine horsepower and
operating hours for the processing industry segment and for transmission and storage
generators. The A.G.A./GRI/SwRI databases were used to determine horsepower for the
transmission industry segment. Operating hours for the storage industry segment were based
on data from site visits and company databases. Operating hours for the transmission
industry segment were based on data from FERC.22
The horsepower-hours for the production industry segment were determined
using data provided by a major natural gas production company. A brief discussion of the
development of each of the activity factors follows. (For further details, refer to the report
Methane Emissions from the Natural Gas Industry, Volume 11: Compressor Driver
Exhaust.19)
Production—The production segment horsepower is based on horsepower-hour
data from one company for 516 engine drivers. The database included drivers in the Gulf
Coast (Onshore and Offshore) and Central Plains regions. The total horsepower-hours were
divided by the company's total gas production before scaling to a national estimate. National
horsepower-hours were calculated to be 27,460 MMhp*hr using the 1992 marketed gas
production for the entire United States (Natural Gas Annual, 1992).5 Confidence limits for
the hp-hr/throughput ratio for the database were calculated to be + 576%. However, this
value was shown by engineering analysis to be unreasonably high.19 Therefore, the
confidence limits on the national estimate were set at + 200 % based on engineering
judgement.
A few turbine compressor drivers do exist in the production segment;
however, the total population is insignificant and was therefore not included in the summary
63
-------
calculations. This emission source would be an insignificant source due to the small number
of these drivers in this industry segment and the low methane emission factor for turbines.
Processing—The processing segment horsepower was determined by
extrapolating site visit data for ten gas plants. The average of two extrapolation methods was
used. The first method extrapolates the site data to a national number by multiplying the
total U.S. gas plant throughput as of January 1, 1993 (46,510.7 MMcfd, Oil & Gas
Journal9) by the site visit horsepower per throughput (47.8 hp/MMcfd for engines and 59.2
hp/MMcfd for turbines). This gives a total of 2.22 MMhp for engines and 2.75 MMhp for
turbines.
The second method scales up the data by plant count, assuming the visited
plants are perfectly representative. A national estimate is produced by multiplying the
horsepower from the site visits by a ratio of 726/10 [the 726 gas plants in the United States
(Oil & Gas Journal) to the 10 sites visited]. This second method, therefore, uses a scaling
factor of approximately 73 for processing horsepower and assumes that all gas plants have
approximately the same throughput, despite the variation found in the site data (40 MMcfd to
750 MMcfd). The activity factors based on the number of gas plants are 6.16 MMhp for
engines and 7.62 MMhp for turbines.
To provide a more conservative estimate, the two methods were averaged
when calculating the national estimate for processing horsepower. The resulting activity
factors are 4.19 MMhp for engines and 5.18 MMhp for turbines.
The annual operating hours were based on the ten site visits and data provided
from two companies (18 additional sites). For eight of the sites visited, typical operating
hours for each of the compressors were not available. Therefore, all compressors that were
running during these site visits were assigned annual operating hours of 8760 and all
compressors that were idle were assigned annual operating hours of 0. An average of the
average operating hours per site was calculated to get the processing segment operating hours
64
-------
(203 engines and 9 turbines). Confidence limits were calculated for the horsepower and
operating hours estimates from the variation of the data.
Transmission—Transmission losses from compressor exhaust have been
divided into two subsections: compressor stations and storage fields. For each of these
areas, activity factors were calculated for compressor drivers and for generator drivers.
Transmission Compressor Stations-The transmission segment horsepower for
each compressor type was determined using the GRI TRANSDAT database.26 GRI
TRANSDAT has not been revised since the 1989 data were collected, and it is assumed that
the data are valid for the 1992 base year. Installed horsepower was taken from the Industry
Database module of GRI TRANSDAT (11.2 MMhp for engines and 5.0 MMhp for turbines).
This horsepower number accounts for about 97% of the gas utility industry installed 16.7
MMhp reported by A.G.A. for 1989.23
The horsepower values in GRI TRANSDAT are a combination of transmission
and storage compressor drivers. GRI TRANSDAT also reports that there were 1.5 MMhp in
the storage segment in 1989. Since GRI TRANSDAT did not give separate values for
engines and turbines in storage, the ratio of engines to turbines in the overall database
(11.2:5.0) was used giving 1.0 MMhp for engines and 0.5 MMhp for turbines in storage.
The overall GRI TRANSDAT horsepower numbers were then adjusted for the 1989 storage
horsepower numbers to give 10.2 MMhp for engines and 4.5 MMhp for turbines in the
transmission segments.
Confidence limits for horsepower could not be rigorously calculated from the
GRI TRANSDAT database because the installed horsepower was given by installation, or
site, and then summed to calculate a national estimate. Therefore, confidence limits were set
based on engineering judgement.
65
-------
The annual operating hours are based on data reported on FERC Form No. 2
for the year 1992.22 The FERC database did not identify the type of driver (i.e.,
reciprocating engine or turbine). As a result, additional data from GRI TRANSDAT and one
transmission company were used to split the FERC hours between the two driver types.
Confidence limits were calculated for the operating hours estimate from the variation of the
FERC data.
Transmission Storage Fields-The storage segment horsepower (1,920,441 hp)
came from Gas Facts Table 4-5 for 1992.2 The split between engines and turbines was
assumed to be the same as the engine and turbine split found in GRI TRANSDAT (69.1%
for engines and 30.9% for turbines). The horsepower data were gathered by A.G.A.
through a survey of their member companies, whose production makes up approximately
96% of total gas industry sales. Confidence limits for horsepower data were calculated
based on an assumed + 5% uncertainty for the A.G.A. horsepower and an assumed ± 10%
uncertainty for the GRI TRANSDAT horsepower splits.
The annual operating hours are based on site visit data and a company
database for a total of eleven storage stations (50 engines and 6 turbines). Confidence limits
were calculated for the operating hours estimates from the variation of the data.
Transmission Generators—Generators were found at both compressor stations
and storage fields. The generator horsepower for compressor stations is based on the total
installed horsepower at seven sites and data provided by one company for 34 compressor
stations (35,006 hp for engines and 1080 hp for turbines). To extrapolate to a national
estimate, the horsepower per station was multiplied by the total number of transmission
compressor stations in the United States (1700, FERC Form No. 2).22
The generator horsepower for storage fields is based on the total installed
horsepower for one company with nine storage fields (3 engines and 1 turbine). To
66
-------
extrapolate to a national estimate, the horsepower per field was multiplied by the total
number of storage fields in the U.S. (475; A.G.A. Gas Facts, Section 7.22).
Generator annual operating hours were also provided by the same sources
providing horsepower data. An average of the average generator operating hours per site
was calculated to determine generator operating hours for compressor stations (87 engines
and 1 turbine) and storage fields (2 engines and 1 turbine). Confidence limits were
calculated for the horsepower and operating hour estimates from the variation of the
site/company data except for the case of generator turbines. Since there was only one data
point for both compressor station generator turbines and storage field generator turbines,
confidence limits were assumed to be equal to those calculated for turbine compressor drivers
at storage fields.
The final activity factors (Table 8-1), in horsepower-hours, were calculated
using the national estimates for compressor horsepower and the average operating hours for
each industry segment except for the production segment. Production horsepower-hours
were estimated as described above. Activity factor confidence limits were propagated from
the confidence limits for the individual terms using a standard statistical approach.
8.1.2 Compressor Counts
Compressor counts were developed using a variety of sources. National data
from DOE and site visit data were used to make the estimates. Table 8-2 provides a
summary of the results and the 90% confidence bounds.
Production—The compressor counts in the production segment were developed
from the site data collection and a regional extrapolation that is explained in detail in Section
5.7. Confidence bounds were calculated for these estimates from the variation of the site
data.
67
-------
TABLE 8-2. COMPRESSOR COUNTS IN THE GAS INDUSTRY
Segment
Reciprocating
Engines
Turbines
Total
Compressors
Production
17,112 ± 52%b
0a
32,327 ± 64%
Gas Plants
4,092 + 47.7%
726 + 76.6%
4,818 + 44%
Transmission
6,785 + 16.6%
681 + 26.2%
7,466 ± 15%
Storage
930 ± 58.3%
136 ± 119%
1,066 ± 54%
Combined Transmission & Storage
7,715 + 10%
817 ± 10%
8,532 + 9.1%
Total Industry
44,134
1,543
45,677
a Only one production site (out of 27 with data on compressors) had any turbines. However, those
"gathering compressors" actually were physically part of an attached gas plant, and were therefore
added into the gas plant counts.
b Non-gas lift compressors
Along with the count of total compressors used in production, it was also
necessary to determine an activity factor for compressors at large gathering compressor
stations. Compressors in transmission and storage stations were found to have blowdown vent
lines routed to the atmosphere through a separate stack. Because of the lack of a good seal
on the blowdown valves, these blowdown lines were found to have extremely large
continuous fugitive emission rates. These emission rates had not been considered or
measured until the GRI/EPA campaign to measure fugitive emissions from transmission
compressors that occurred in late 1994.
Originally, it was assumed that compressors used in production did not have
these separate vent lines, or that the vent lines were small and located proximate to the
compressors so that the API/EPA fugitive measurement efforts had included them.
However, several companies reported that some large gathering compressor stations (that had
a large number of engines) were similar to transmission stations, and did have these separate
vent lines to a separate stack.
Therefore, for production, the FERC22 data were analyzed for gathering
stations. (Gathering was included in the production segment.) In FERC Form No. 2, the
number of stages of compression operating on the peak demand day is reported. It was
assumed that there are an average of 3.3 stages of compression per compressor, and that a
68
-------
large station would contain at least five compressors. Therefore, a station with 16 or more
stages of compression was counted as a large gathering compressor station. Based on this
criteria, three large gathering compressor stations were identified in the FERC data. This
number was then extrapolated from the 86,200 miles of gathering pipeline mileage owned by
transmission companies, to the national activity factor of 340,000 gathering pipeline miles,
for a total of twelve large compressor stations in production. The FERC data also showed
that the large compressor stations averaged 26 stages of compression, or eight compressors
per large station. This value was extrapolated in the same manner to provide a national
activity factor of 96 compressors at large gathering compressor stations. A confidence
interval of + 100% was assigned to the activity factor for large stations and for compressors
at large stations based on engineering judgement.
Processing--The compressor counts for the processing segment were
determined from eleven gas plant site visits. The average ratio of compressors per plant was
multiplied by the 742 plants (see section 6.1) to obtain the values in the table. Confidence
bounds were calculated for these estimates from the variation of the site data.
Transmission & Storage—The GRI TRANSDAT industry database was used
to estimate the total compressor counts.26 The industry database of 16.2 MMHp had 7489
engines and 793 turbines. However, the database is not quite a complete sample of the
transmission/storage sector since its horsepower is slightly lower than the total
transmission/storage horsepower of 16.7 MMHp for the year 1992 (as shown in Gas Facts,
Table 5-5, "Gas Utility Industry Installed Compressor Horsepower, 1968-1993"2).
Therefore, the database totals were multiplied by the ratio of the two values (16.7/16.2) to
determine the adjusted totals of 7715 engines and 817 turbines. Confidence bounds were not
calculated but were estimated to be ± 10% based upon the large size of the TRANSDAT
database and the tight confidence of the Gas Facts data.
The combined transmission/storage count of storage compressors was divided
into separate categories (i.e., transmission and storage) based upon site visit data and Gas
69
-------
Facts data. Site visits to eight storage sites produced ratios of 0.750 turbines per site (±
119%), 5.13 reciprocating engines per site (± 58.3%), and 13,033 HP/site. The site data
are assumed to have come from larger than average stations since Gas Facts shows that the
actual horsepower per underground storage site for 1992 was only 4975 HP/site (Table 4-5,
"Amount of Pools, Wells, Stored Gas, and Horsepower in Underground Storage Fields, By
State, 1990-1992"2). Therefore, the site ratios were adjusted to account for the higher than
average horsepower by multiplying by 4975/13033. The adjusted ratios were then multiplied
by the 475 storage compressor stations (see Section 7.2). Transmission compressor counts
were then determined by subtracting the count of compressors in storage (1066) from those
determined by the TRANSDAT database analysis.
8.2 Dehvdrators
Multiple activity factors associated with dehydrators were determined for the
entire gas industry. These factors are based on industry segment gas throughput, and the
percentage of units with flash drums, gas pumps, and stripping gas. This study is only
concerned with glycol-based dehydrators (as opposed to dry-bed) since only the glycol
dehydrators emit methane.
8.2,1 Dehydrator Count
A national count of 41,700 dehydrators in the gas industry was recently
determined by a 1994 GRI study performed by Wright Killen & Company.18 The study,
"Natural Gas Dehydration, Status and Trends," used several states' large survey databases
[Wyoming, Texas Mid-continent Oil & Gas Association (TMOGA), and the Louisiana
Department of Environmental Quality (LDEQ)] to determine the count of dehydrators in the
natural gas industry. The Wright Killen study divided the dehydrators into segments of the
natural gas industry based on data from the TMOGA and Gas Processors Association (GPA)
survey.27 The information in this survey allowed Wright Killen to estimate the counts of
70
-------
dehydrators in the different industry segments. These values can be found in the first
column of Table 8-3.
TABLE 8-3. U.S. GAS INDUSTRY DEHYDRATOR COUNTS
Segment
Wright-Killen Total Dehydrator
Counts
(All types)
Glycol Dehydrator Count
Based on Adjusted Totals
Production
25270
37824
Gas Processing Plants
7923
498
Transmission
8507
201
Storage
NA
1092
Total Gas Industry
41700
39615
This GRI/EPA study used the Wright Killen report as the basis for the overall
count of dehydrators because the report provides a reasonably accurate estimate of the
industry. The Wright Killen report estimates that there are 7923 dehydrators at 742 plants,
or 10.6 dehydrators per gas plant. However, in the Wright Killen study, the dehydrators
were split into industry segments using data from a single regional survey; as a result, the
gas plant dehydrator count is unrealistically high. This study adjusted the Wright Killen split
based on GRI/EPA's information from 11 gas plants that shows only 1.41 dehydrators per
plant. Starting with the Oil & Gas Journal annual gas processing plant list, an estimate of
the count of glycol-based dehydrators (as opposed to dry-bed) was made. All refrigerated
processes were assumed to use glycol dehydration, and all cryogenic processes were assumed
to use dry-bed dehydration. Based on these assumptions, the percentage of glycol-based
units in the industry is estimated as 48.6% (a total of 498 glycol dehydrators).
Similarly, the Wright Killen estimate produces 8507 dehydrators for 2190
transmission/storage compressor stations, or 3.9 dehydrators per compressor station.
GRI/EPA's data show 2 dehydrators for 17 mainline transmission stations and 17 dehydrators
for 6 storage stations. The Wright Killen split is probably in error since the TMOGA/GPA
production segment survey has too few small dehydrators and too many large dehydrators to
71
-------
represent the true national production totals. This assessment of the data is validated by the
fact that Texas, where the study was performed, has a higher production rate per well than
the national average. Therefore, the splits between industry segments may be biased if the
national percentage of production dehydrators is larger than in Texas.
This GRI/EPA study also corrected the production total to 37,824 units to
maintain the total national count estimated by Wright Killen. These corrected industry
segment counts are listed in the second data column of Table 8-3.
By using the site visit ratios for 1700 mainline transmission stations, an
estimate of 201 dehydrators in transmission service was determined. For the 386
underground storage stations, the site visit ratio produces an estimate of 1092 in storage
service.
The Wright Killen study also found that on an industry-wide basis, 95% of the
dehydrators were glycol-using (as opposed to molecular sieve adsorbers or other types of
water removal). Based on the 95% value, the total number of glycol dehydrators in the
production, transmission, and storage segments was calculated to be 39,615. These data
have confidence limits of + 20% based on Wright Killen's 90% confidence limits. All
dehydrators in this survey are assumed to be active.
8.2.2 Dehydrator Throughput
Emissions from dehydrators are proportional to the gas throughput of the
dehydrator. Therefore, an activity factor for throughput for each industry segment was
determined. There is a direct relationship between the estimated count of dehydrators and
the average throughput per dehydrator. In fact, the Wright Killen study determined the
production count by dividing known gas production rates by an estimated dehydrator
throughput. Therefore, the GRI/EPA study used the Wright Killen report as the basis for the
72
-------
dehydrator gas throughput used in this project,
project are shown in Table 8-4.
The throughput numbers selected for this
TABLE 8-4. INDUSTRY SEGMENT GLYCOL
DEHYDRATOR THROUGHPUT
Segment
Average Throughput with 90%
Confidence Intervals
Production
12.4 Tscf/yr ± 62%
WK&C (Adjusted Wyoming data)18
Gas Processing Plants
8.63 Tscf/yr + 22%
Oil and Gas JoumaP
Transmission
1.09 Tscf/yr ± 144%
TMOGA26
Storage
2.00 Tscf/yr ± 25%
A.G.A. Gas Facts2
Production—Wright Killen estimated that the average throughput per
dehydrator in production is 2 MMscfd. This value has been validated by analysis of three
data sources, as follows:
The Wyoming dehydrator survey (as adjusted by Wright Killen), which
is believed to be the most comprehensive survey, reports a value of 2
MMscfd per production segment dehydrator.
The TMOGA survey produced an average throughput value in the 8.1
to 8.5 MMscfd range, but the study is believed to be biased toward the
higher values in the production segment. This bias is caused by the
higher production rate per well in the Texas region. Therefore, the
median value (as opposed to the average value) of the Texas survey is
likely to be more representative of a true national average, since the
median value is not heavily weighted by a few large dehydrators. The
median value of the Texas (TMOGA) survey is 2 MMscfd.
The site visits produced an average of 12.5 MMscfd/dehydrator.
However, many of the sites visited have higher production rate-to-
number of wells ratios than the national average. If the two largest
dehydrators, which are in the Gulf Coast, are excluded from the total,
the resulting average is only 1.14 MMscfd/dehydrator.
73
-------
The total production throughput was calculated by multiplying the average
throughput per dehydrator by the number of dehydrators in the production segment and by a
capacity utilization factor. Operators of several production sites studied for another GRI
project indicated that the typical production dehydrator was using less than half its original
capacity. The capacity utilization factor was determined to be 0.45 ± 33%.
Gas Processing—Throughput for the gas processing industry segment was
determined from the annual Oil and Gas Journal survey of processing plants.9 It was
assumed that plants using a refrigerated process have glycol dehydration and plants using a
cryogenic process have some type of dry-bed dehydration with no methane emissions. For a
total gas processing throughput of 17.44 Tscf/year, these assumptions lead to a flow of 8.63
Tscf/year that is dehydrated by glycol. No distinction was made between triethylene glycol
(TEG) and ethylene glycol (EG), since the methane emissions for the two glycol processes
are similar. [Note that benzene, toluene, ethyl benzene, and xylene (BTEX) emissions for
EG may be several orders of magnitude lower than for TEG.]
Transmission-Transmission segment throughput was determined in much the
same way as was production. The Wright Killen survey indicated an average transmission
dehydrator throughput of 14.8 MMscfd and an adjusted count of 201 transmission
dehydrators, for a total throughput of 1.09 Tscf/year.
Storage—Storage segment throughput was determined from data in A. G.A. Gas
Facts, Table 4-3.2 Withdrawals from storage in 1992 were 2.41 Tscf. It was assumed that
most gas taken from storage is dehydrated, so storage dehydrator throughput was estimated
to be 2.00 Tscf/year ±25%.
74
-------
8.2.3
Dehydrator Flash Drums
One of the dehydrator equipment accessories that can have a large impact on
emissions is the three-phase flash drum. Data on flash drum existence were available from
the TMOGA survey27 and from the site visits. The results are shown in Table 8-5.
As discussed previously, the production segment of the TMOGA database is
believed to be biased toward larger dehydrators, since the Texas production per well ratio is
higher than the national average. Since analysis of the TMOGA database does show a
relation between dehydrator size and the existence of a flash drum (especially for sizes
below 2 MMscfd), the number of large dehydrators in the TMOGA database may bias the
average flash drum percentage high. The site visits may be biased toward smaller
dehydrators but are not believed to have the same degree of bias as the TMOGA data set.
The result was a 3.31% number for production dehydrators.
8.2.4 Dehydrator Gas Pumps
Another important piece of glycol dehydrator equipment is the type of pump
that circulates glycol. The activity factor for gas-driven glycol circulation pumps is
calculated on the same basis as the dehydrator vented emissions (i.e., Tscf/year gas
throughput for each industry segment). The total gas throughputs listed in Table 8-4 were
multiplied by the fraction of dehydrators using gas-driven pumps to determine the gas-
driven pump activity factor. This value is based on the average from Radian site surveys of
the number of pumps per dehydrator. The results are displayed in the Table 8-6.
8.2.5 Dehydrator Stripping Gas Use
The percent of dehydrators using stripping gas was determined from the site
surveys. The site surveys showed that less than 0.5% of all production dehydrators use
stripping gas. The use of stripping gas is more common in the other three industry
segments, with 11.1% for gas processing, 7.4% for transmission, and 8.0% for storage.
75
-------
TABLE 8-5. U.S. GAS INDUSTRY DEHYDRATOR FLASH DRUMS
Percent of Dehvdrators with Flash Drums3
(with 90% confidence bounds)
Industry Segment
Unadjusted
Survey Results
Site
Visits
Selected
Numbers
Basis
Production
42.4 ± 8.5%
3.31 ± 43.2%
26.5 + 8.35%
TMOGA and site visits
Gas Processing Plants
66.7 + 8.5%
66.7 + 46.5%
66.7 ± 10.1%
TMOGA and site visits
Transmission
66.9 ± 9.7%
NA
66.9 ± 9.70%
TMOGA
Storage
NA
52.0 ± 33.6%
52.0 + 33.6
Site visits
a The number of dehydrators were divided by survey: Unadjusted TMOGA results (618 in
production, 207 in processing, and 192 in transmission/storage) and site visits (408 in
production, 11 in processing, 2 in transmission, and 25 in storage).
TABLE 8-6. SITE SURVEY GAS PUMP DATA
Segment
Active Gas Pumps per
Dehydrator
(with 90% confidence limits)
Total Industry
Throughput
Tscf/year
Gas-driven Pump
Throughput
i :ij|i
Production
0.891 ± 2.79%
12.4 ± 48.2%
11.1 ± 62.0%
Gas Processing Plants
0.111 ± 186%
8.63 ± 22.4%
0.958 + 192%
Transmission
0
1.09 ± 130%
0
Storage
0
2.00 + 25.0%
0
Total
24.1 ± 26.8%
12.1 + 58.6%
8.2.6 Dehydrator Vapor Recovery Use
Due to regulations restricting benzene, toluene, ethylbenzene, and xylene
(BTEX) emissions, many sites have begun to install vapor recovery systems on glycol
dehydrator reboiler/regenerator vents. However, not all of these systems prevent the release
of the methane since methane emissions are not regulated. Because of the lack of prevention
against methane release, the only vapor recovery systems counted for this study were those
that captured and burned the methane from the vent. The results are displayed in Table 8-7.
The site visits did not identify any gas processing plants with vapor recovery systems that
consume methane. However, there are likely to be some plants with methane vapor
76
-------
recovery. For calculation purposes, the fraction of plants with methane emission controls
was assumed to be 0.10 + 90%.
TABLE 8-7. SITE SURVEY STRIPPING GAS AND VAPOR RECOVERY DATA
Segment
Fraction of Dehydrators with
Active Stripping Gas
(with 90% confidence limits)
Fraction of Dehydrators With
Vapor Recovery that Consumes
Methane
(with 90% confidence limits)
Production
0.0047 ± 116%
0.012 ± 73.1%
Gas Processing Plants
0.111 ± 186%
0.0 + 0% a
Transmission
0.074 + 118%
0.148 ± 80.3%
Storage
0.080 ± 118%
0.160 ± 80.0%
a No vapor recovery was found during the site visits but it is likely that some processing plants have
flares. A value of 0.10 + 90% was used in the calculations.
77
-------
9.0 DISTRIBUTION SEGMENT
Activity factors were developed for each of the sources of methane emissions
from the distribution segment of the gas industry. The activity factors that were used in the
distribution segment are the following: total leaks in underground distribution mains and
services; metering and pressure regulating stations; customer meters; and total miles of mains
and services used to determine losses from pipeline third-party damage (i.e., dig-ins) as well
as pipeline blow and purge for maintenance/installation activities. A discussion of the
methodology used to derive these activity factors is presented in this section.
9.1 Distribution Annual Equivalent Leaks
Since the emission factor for quantifying emissions from underground
distribution mains and services was stratified by pipe use (mains versus services) and by pipe
material (i.e., cast iron, cathodically protected steel, unprotected steel, plastic, and copper),
the activity factor was also stratified to extrapolate emissions.
With the exception of cast iron main pipeline, the activity factor used to
extrapolate the leakage estimate for underground distribution mains and services was the
number of annual equivalent leaks. (For cast iron pipeline, the activity factor was the total
mileage of cast iron mains in the United States, which is a nationally tracked statistic.28)
Annual equivalent leaks are defined as the number of leaks that leak continuously year round.
For example, if leaks that are repaired during the year are leaking for half the year, on
average, then each repaired leak would be counted as half an annual equivalent leak.
The number of annual equivalent leaks was derived from the national database
of leak repair records broken down by mains and services (U.S. Department of
Transportation, Research and Special Programs Administration).28 To allocate leak repairs
into pipe material categories, data were collected from ten local distribution companies
78
-------
representing different regions within North America. An estimate of the national leak repairs
allocated by pipe material type is shown in Table 9-1.
To derive annual equivalent leaks from the national leak repair records,
additional information was needed including the number of leaks found during the year (leak
indications) and the unrepaired leaks at the beginning of the year (outstanding leaks). Since
leak indications and outstanding leaks are not tracked nationally, this information was
requested from individual companies.
Data were collected from the companies participating in the cooperative leak
measurement program on the annual number of leak repairs, number of leak indications, and
outstanding leaks at the beginning of the year (reference year in most cases was 1991). The
data were requested to be disaggregated by mains versus services and by pipe material.
Complete data were provided by only four companies, coupled with a breakdown of the total
mileage of mains and number of services by pipe material. Two additional companies
provided data without a breakdown by pipe material or use. (The leak data disaggregated by
pipe use and material type are difficult to obtain from many companies, since leak records
are often not maintained in this manner.)
An estimate of the total annual equivalent leaks for each of the six companies
was developed for each pipe material category except cast iron, based on the following
formula:
TEL = OL + LI + UDL + URL - (0.5 X RL)
where
TEL = Total annual equivalent leaks;
OL = Outstanding leaks at the beginning of the year;
79
-------
TABLE 9-1. NATIONAL LEAK REPAIRS ALLOCATED BY PIPE MATERIAL CATEGORY
Pipe Use
Pipe Material
Average Leak
Repairs/Mile"
National
Miles/Servicesb
Extrapolated
Leak Repairs
Percent Leak
Repairs
Estimated
Total Leak
Repairs
Precision
(Leak
Repairs)
Mains
Cast Iron
1.38
55,288
76,353
34
69,750
39,957
Protected Steel
0.08
451,466
34,943
16
31,922
14,006
Unprotected
Steel
1.09
82,109
89,417
40
81,685
50,645
Plastic
0.08
299,421
25,151
11
22,976
22,862
Subtotal
255,864
100
206,333b
Services
Protected Steel
0.006
20,352,983
126,188
42
181,752
206,405
Unprotected
Steel
0.027
5,446,393
148,686
50
214,156
190,569
Plastic
0.001
17,681,238
22,986
8
33,107
26,238
Copper
0.011
233,246
2,519
1
3,628
3,165
Subtotal
300,379
100
432,643b
a Based on data provided by ten companies.
b Based on nationally tracked database, U.S. Department of Transportation, Research and Special Programs Administration.27
-------
LI = Leak indications recorded during the year, including call-ins;
UDL = Undetected leaks which cannot be found using an industry
standard survey procedure;
URL = Unreported leaks that have developed in parts of the network not
survey during the current year; and
RL = Repaired leaks - estimated to be leaking half the year, on
average.
Undetected leaks that cannot be found using an industry standard survey procedure were
quantified based on information provided by Southern Cross.29 According to their experience
in performing leak surveys and survey audits, Southern Cross predicts that a standard
industry survey procedure using a flame ionization detector (FID) instrument finds 85 % of
the leaks. (Note: The standard industry survey procedure involves using either a walking or
mobile survey, as appropriate for the area being surveyed, using an FID instrument. Any
potential leak that is found with the FID instrument, registering any concentration above
background, is investigated using bar holing procedures.) Therefore, the number of
undetected leaks is estimated by:
UDL = [(1/0.85) - 1] x LI
The total annual equivalent leaks are derived using the estimated leak duration for each type
of leak, based on the following:
• Repaired leaks are assumed to be leaking half the year, on average.
• Outstanding leaks, leak indications, and undetected leaks are estimated
to be leaking the entire year (i.e., 8760 hours per year).
The leak duration of unreported leaks is factored into the estimation methodology for these
leaks. Unreported leaks are those leaks that occur in parts of the network not surveyed
during the year (i.e., multi-year survey cycle). The number of unreported leaks is based on
81
-------
the annual leak indications and the undetected leaks as well as the frequency of the leak
survey. The number of unreported leaks in the system that is surveyed every n years is
calculated based on the following:
• For the first year in the cycle — 1/n X (LI + UDL) leaks are leaking
half the year; (n-l)/n x (LI + UDL) leaks are not yet leaking.
• For the second year in the cycle — 1/n x (LI + UDL) leaks are
leaking the entire year; 1/n x (LI + UDL) leaks are leaking half the
year; and (n-2)/n x (LI + UDL) leaks are not yet leaking.
• For the third year in the cycle — 2/n x (LI + UDL) leaks are leaking
the entire year; 1/n X (LI + UDL) leaks are leaking half the year; and
(n-3)/n X (LI + UDL) leaks are not yet leaking.
• For the fourth year in the cycle — 3/n x (LI + UDL) leaks are leaking
the entire year; 1/n x (LI + UDL) leaks are leaking half the year; and
(n-4)/n x (LI + UDL) leaks are not yet leaking.
Based on the methodology described above, the number of equivalent leaks was determined
for each of the six companies providing detailed data. The ratio of equivalent leaks to leak
repairs was then calculated for each of the companies. The average ratio of equivalent leaks
to leak repairs was used to extrapolate the national database of leak repairs. Table 9-2
presents a summary of the leak record data provided by the six companies, the estimated
equivalent leaks, and the corresponding ratio of equivalent leaks to leak repairs. As shown,
the average ratio of equivalent leaks to leak repairs is 2.14.
The national number of annual equivalent leaks, broken down by pipe use and
material type, is shown in Table 9-3. As shown, the activity factor for cast iron mains is
miles of pipeline, to correspond to the emission factor in units of scf/mile-year. The
estimate of annual equivalent leaks is highest for unprotected steel services, followed by
protected steel services. For mains, unprotected steel is the category with the highest
estimated annual equivalent leaks. The precision of the number is based on the variability in
leak repair data allocated by material type from ten companies and the variability in the ratio
of equivalent leaks per leak repair from six companies.
82
-------
TABLE 9-2. SUMMARY OF LEAK RECORD DATA FROM PARTICIPATING COMPANIES
Company
Annual Leak
Indications
Annual I.cak Repairs
Annual Outstanding
Leaks
Ksti mated Total
Equivalent Leaks
Ratio of Equivalent Leaks
to Leak Repairs
A
3,747a
2,061a
0
3,378
1.64
B
9,249
17,003
11,701
18,796
1.11
C
2,115
2,443
0
2,832
1.16
D
b
14,681
b
41,286
2.81
E
1,999
2,287
2,396
6,250
2.73
F
5,992
3,421
1,558
11,597
3.39
Average
2.14
Precision
0.79
a Mains only.
b Data not available.
-------
TABLE 9-3. SUMMARY OF ACTIVITY FACTORS FOR DISTRIBUTION
UNDERGROUND PIPELINES
Pipe Use
Material Category
Estimated Total
Leak Repairs2
Average Activity
Factor
(equivalent leaks)1'
Precision of Activity
Factor (equivalent
leaks)
Mains
Cast Iron
69,750
55,288°
2,764°
Protected Steel
31,922
68,308
42,545
Unprotected Steel
81,685
174,657
101,685
Plastic
22,976
49,226
58,018
Subtotal
206,333
Services
Protected Steel
181,752
390,628
526,354
Unprotected Steel
214,156
458,476
499,850
Plastic
33,107
68,903
66,840
Copper
3,628
7,720
8,521
Subtotal
432,643
a Based on national leak repair database27 and data provided by six companies (see Tables 9-1 and 9-2).
b Based on ratio of annual equivalent leaks to leak repairs of 2.14 (see Table 9-2).
c Miles.
84
-------
9.2
Distribution Metering and Pressure Regulating Stations
Since the emission factor for predicting emissions from metering and pressure
regulating stations was stratified into station type, inlet pressure range, and location
categories, the activity factor was also stratified to produce an overall estimate of emissions
from stations in each segment of the gas industry.
For distribution stations, a questionnaire was sent to distribution companies
participating in the underground leak measurement program and the companies participating
in the metering/pressure regulating station measurement program. A total of ten companies
provided demographic information on the stations in their distribution network. The station
counts provided by the companies were disaggregated by station type and inlet pressure
range. Table 9-4 presents the demographic information provided by the respective
companies.
To extrapolate the individual company data to a national number of stations, a
known industry statistic was used. The individual company demographic data were
combined with the miles of main and total gas throughput for each company to determine the
number of stations in each category on a per mile of main and per unit of throughput basis.
Linear regression analyses were performed on the number of stations versus both miles of
mains and total gas throughput, respectively. It was concluded that the number of stations
should be extrapolated by miles of mains since the variability within the sample set was
lower for number of stations per mile than for number of stations per throughput.
The average number of stations per mile was calculated from the individual
company data and multiplied by the total mileage of main pipeline in the United States.28
The standard deviation of each number was calculated based on the variability in the
individual company data.
85
-------
TABLE 9-4. ACTIVITY DATA PROVIDED BY INDIVIDUAL COMPANIES
Station Type
Inlet
Pressure
Range (psig)
Comp
A
Comp
B
Comp.
C
Comp.
D
Comp.
E
Comp.
F
Comp.
G
Comp.
H
Comp.
I
Comp.
J
Comp.
K
Comp
L"
M&R Stations
>300
14
6
18
19
-
25
1
0
-
29
128
15
100-300
8
15
0
0
-
600
0
0
-
286
431
2
40-100
1
2
0
0
-
0
0
0
-
170
252
0
<40
0
0
0
0
-
0
0
0
-
48
97
0
Regulating
Stations
>300
720
0
0
0
0
30
2
1
34
29
87
0
100-300
1,187
0
258
94
25
210
0
0
44
286
835
273
40-100
207
1,257
0
325
44
400
11
8
136
170
2,122
120
<40
12
0
130
28
203
1,000
0
0
53
48
935
0
Miles Main
40,930
30,934
7,594
4,000
3,924
14,900
64
78
4,109
3,396
29,073
1,263
" Not considered representative of national average.
-------
Data were collected from three of the ten companies on the number of
regulating stations in vaults versus above ground. Based on the data collected, about half of
the regulating stations with an inlet pressure greater than 300 psig were observed to be
located in vaults. For regulating stations with an inlet pressure between 40 and 300 psig, the
majority of stations in urban areas were in vaults and the majority in rural areas were above
ground. On average, it was estimated that 55% 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, about
70% of the stations are located in vaults. Based on the data collected, essentially all low-
pressure (<40 psig) stations are located in vaults. Table 9-5 shows the number of stations
disaggregated by station type, inlet pressure range, and location in a vault versus above
ground.
9.3 Customer Meters
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, Table
8-1.2 The number of residential customer meters located indoors versus outdoors was
estimated based on a regional breakdown of total customers presented in Gas Facts. Based
on data collected from 22 companies located in six regions within the United States, the
majority of customer meters in urban, northern areas of the country are located indoors,
while the majority of customer meters located in non-urban areas and in the southern areas
are located outdoors. On average, approximately 22% of residential customer meters are
located indoors. 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.
87
-------
TABLE 9-5. SUMMARY OF ACTIVITY FACTORS FOR DISTRIBUTION METERING/PRESSURE
REGULATING STATIONS
Station Type
Inlet Pressure
(psig)
Location
(vault or abme
ground)
Stations per Mile
Average Activity
Factor
(stations)
Precision of Activity
factor
(stations)
M&R
>300
A-G
0.004
3,460
2,458
M&R
100-300
A-G
0.016
13,335
14,091
M&R
<100
A-G
0.009
7,127
10,374
Reg.
>300
A-G
0.005
3,995
2,702
Reg.
>300
Vault
0.003
2,346
1,587
Reg.
100-300
A-G
0.015
12,273
7,461
Reg.
100-300
Vault
0.007
5,514
3,352
Reg.
40-100
A-G
0.043
36,328
23,375
Reg.
40-100
Vault
0.039
32,215
20,729
Reg.
<40
a
0.018
15,377
9,922
a The above-ground and in-vault categories were combined for the low pressure regulating station category.
-------
Therefore, the number of outdoor residential customer meters, 40,049,300,
was 78% of the total population of residential customer meters. The confidence interval
around the mean activity factor, + 10%, was estimated based on the variability in company
supplied data.
The precision of the total commercial/industrial customer meters is assumed to
be + 5% of the estimated 4,608,000 meters.
9.4 Distribution Mains and Services
The total mileage of distribution mains and services was used as the activity
factor to estimate the methane emissions from distribution third-party damage (e.g. dig-ins)
and from blow and purge due to pipeline abandonment, installation, and repair. The total
miles of main pipeline in the gas industry, 836,700 miles, was based on U.S. Department of
Transportation (DOT) records.28 The total miles of service, 460,809 miles, was reported by
A.G.A.2 Therefore, the total combined main and service mileage in the United States is
1,297,569 miles. The 90% confidence limit around the activity factor is assumed to be 5%,
or 64,879 miles.
89
-------
10.0 RESULTS
This section summarizes the major activity factors collected during this
project. Some activity factors were already known on a national basis and were taken from
published sources. Others were generated from company data or site visit data. Most
equipment counts taken at site visits excluded equipment that was not in use. The following
tables (10-1 and 10-2) summarize the activity factor data. Previous sections of this report
explain the detailed basis for all of the activity factors.
TABLE 10-1. WELL-DEFINED ACTIVITY FACTORS
Segment
Activity Factor Name
Number
Total Industry
1992 Gross Gas Production (Tscf)
22.13
1992 Marketed Gas Production (Tscf)
18.71
Production
No. of Gas Wells
276,000
No. of Oil Wells
602,200
Processing
No. of Gas Plants
726
Transmission and Storage
Miles of Trans Pipeline
284,500
No. of Storage Facilities
475
No. of Storage Wells
18,000
Distribution
Miles of Service
460,800
Miles of Mains
836,760
Number of Residential Customer Meters
51,520,000
90
-------
TABLE 10-2. EXAMPLES OF DEVELOPED ACTIVITY FACTORS
Segment
Activity Factor Name
Number ¦¦
Total Industry
Recip 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
Engine MMHp-hr
27,460
Offshore Platforms
1,115
Dehydrators (% of industry dehydrators)
37,820 (95%)
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
340,200
Well Workovers
9,329
Pneumatic Devices
249,100
Processing
Compressor Drivers
4,092 recips
726 turbines
Annual Compressor Operation
27,780 MMhp-hr
32,910 MMhp-hr
Dehydrators (% of industry dehydrators)
498 (1.3%)
Acid Gas Recovery Units
371
Continued
91
-------
TABLE 10-2. (Continued)
Segment
Activity Factor Name
Number
Transmission and
Storage
Compressor Drivers
7,715 recips
817 turbines
Annual Driver Operating Hours (average)
- Transmission Compressor Drivers
- Storage Compressor Drivers
40,380 MMhp-hr
9,635 MMhp-hr
4,922 MMhp-hr
1,729 MMhp-hr
Transmission Compressor Stations
1,700
Dehydrators (% of industry dehydrators)
1,293 (3.3%)
Pneumatic Devices
2,175
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
92
-------
11.0 REFERENCES
1. Harrison, M.R., H.J. Williamson, and L.M. Campbell. Methane Emissions
from the Natural Gas Industry, Volume 3: General Methodology, Final
Report, GRI-94/0257.20 and EPA-600/R-96-080c. Gas Research Institute and
U.S. Environmental Protection Agency, June 1996.
2. American Gas Association. Gas Facts: 1993 Data, Arlington, VA, 1994.
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. Cochran, William G. Sampling Techniques Third Edition, John Wiley &
Sons, New York, 1977.
5. U.S. Department of Energy/Energy Information Administration. Natural Gas
Annual, 1992. DOE/EIA-0131 (89), Washington, DC, September 1993.
6. Gulf Coast Publishing Company, World Oil, Annual Forecast/Review, Vol.
214, No. 2, February 1993.
7. Texas Railroad Commission Information Services Dept., Paula Middleton,
Personal Contact with K. Hummel, Radian Corporation, June 1994.
8. Alaska Oil & Gas Conservation Commission, Joan Miller, Phone Contact
with Radian Corp., February 9, 1995.
9. Oil & Gas Journal. 1992 Worldwide Gas Processing Survey Database, 1993.
10. Texas Railroad Commission 1989, P-l, P-2 Tapes, Radian files.
11. Dwight's Energy Data, Inc., Mike Morgan, FAX contact with Radian Corp.,
December 29, 1995.
12. Offshore Data Services, David Sutherland, Phone Contact with Radian Corp.,
June 27, 1994.
13. Minerals Management Service. User's Guide: MMS Outer Continental Shelf
Activity Database (MOAD). OCS Study MMS 94-0018. U.S. Department of
Interior, New Orleans, LA, June 1994.
14. 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.
93
-------
15. U.S. Department of Energy/Energy Information Agency. Natural Gas
Production Responses to a Changing Market Environment, 1978-1988. Table
6-3, Washington, D.C.
16. Pipeline Systems Incorporated. Annual Methane Emission Estimate of the
Natural Gas Systems in the United States, Phase 2. Prepared for Radian
Corporation, September 1990.
17. 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.
18. Wright Killen & Co. Natural Gas Dehydration: Status and Trends, Final
Report, GRI-94/0099, Gas Research Institute, Chicago, IL, January 1994.
19. Stapper, C.J. Methane Emissions from the Natural Gas Industry, Volume 11:
Compressor Driver Exhaust, Final Report, GRI-94/0257.28 and EPA-600/R-
96-080k. Gas Research Institute and U.S. Environmental Protection Agency,
June 1996.
20. Tannehill, C.C. and C. Galvin. Business Characteristics of the Natural Gas
Conditioning Industry, Topical Report, GRI-93/0342, prepared by Purvin &
Gertz, Inc, Gas Research Institute, May 1993.
21. Dalrymple, D.A., F.D. Skinner, and N.P. Meserole. Investigation of U.S.
Natural Gas Reserve Demographics and Gas Treatment Processes, Topical
Report, GRI-91/0019. Gas Research Institute, Chicago, IL, prepared by
Radian Corporation, January 1991.
22. Federal Energy Regulatory Commission (FERC) Form No. 2: Annual Report
of Major Natural Gas Companies, 1992 database.
23. American Gas Association. Natural Gas Industry Environmental
Organization Structure Survey, Arlington, VA, 1990.
24. American Gas Association. Gas Facts-. 1991 Data, Arlington, VA, 1992.
25. 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.
26. Biederman, N. GRI TRANSDAT Database: Compressor Module, (prepared
for Gas Research Institute) npb Associates with Tom Joyce and Associates,
Chicago, IL, August 1991.
94
-------
27.
Texas Mid-Continent Oil and Gas Association (TMOGA) and Gas Processors
Association (GPA), Dehydrator Survey Database, 1991.
28. U.S. Department of Transportation, Research and Special Programs
Administration, 1991.
29. Southern Cross Corporation. "Comments on Docket PS-123 Notice 1, Leak-
age Surveys," 49 CFR Part 192, Department of Transportation, Research and
Special Programs Administration, Materials Transportation Bureau, Office of
Pipeline Safety Regulations, December 19, 1991.
95
-------
APPENDIX A
Site Data
A-l
-------
TABLE A-l. PRODUCTION SITES (Summary)
Region"
Offshore
(if
CP
PM
ac;l
Total liS
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 surveys
- Star Site Visit**
0
0
0
0
12
12 star sites**
~Region Key: GC = Gulf Coast; CP = Central Plains; PM = Pacific Mountain; AGL = Atlantic & Great Lakes.
**This does not include all sites visited by Star or other fugitive emission contractors. Only the sites used for activity factor data collection
included.
-------
TABLE A-2. PRODUCTION SITES (Offshore Data)
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 Wells
Gas Wells
(1/MMcfd)
(1/MMcfd)
Site
1
2
3
4
6
Company
1
2
2
2
3
4
6 Sites
Survey Type
V
P
P
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.
-------
TABLE A-3. PRODUCTION SITES (Gulf Coast Onshore Data)
Region
GC
GC
GC
GC
GC
GC
GC
GC
GC
Total GC
Site
7
8
9
10
11
12
13
14
15
Equip./
Equip./
Company
5
6
7
8
9
10
11
11
11
9 Sites
Equip./
Equip./
Mkt. Gas
Prod. Gas
Survey Type
V
V
V
V
V
V
V
V
V
7 Companies
Total
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.
-------
TABLE A-4. PRODUCTION SITES (Central Plains Data)
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
Prodi Gas
Survey Type
V
V
iii!
P
P
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
-t-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 = 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: CP = Central Plains.
-------
TABLE A-5. PRODUCTION SITES (Pacific/Mountain Onshore Data)
Region
PM
PM
PM
PM
PM
PM
PM
Total PM
Site
iiii
24
Wmm
26
27
28
29
30
illill
Equip./
Equip./
Company
19
20
21
21
21
21
22
21
MziW,
Equip./
Equip./
Mkt. Gas
Prod. Gas
Survey Type
fill:-
V
V
V
V
ill:::
V
V
'MMIi
4 Companies
Total
Total Wells
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: PM = Pacific Mountain.
-------
TABLE A-6. PRODUCTION SITES (Atlantic & Great Lakes Data)
Region
AGL
AGL
AGL
AGL
AGL
AGL
AGL
AGL
AGL
Site
32
33
34
35
36
37
38
39
40
Company
23
24
25
26
27
28
29
30
30
Survey Type
V
P
P
V
P
P
P
S
S
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). (Continued)
2) * = Gas lift compressors not included.
3) Y = Yes, N = No; = No Data.
4) Region Key: AGL = Atlantic & Great Lakes.
-------
TABLE A-6. (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./
Mkti Gas
Prod. Gas
Survey Type
S
S
S
S
S
S
S
S
S
mm
10 Companies
(Sites 32-50)
Total Wells
Gas Wells
(1/MMcfd)
(1/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
Fupitive 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: AGL = Atlantic & Great Lakes.
-------
TABLE A-7. GAS PROCESSING PLANTS
Site
1
2
3
4
5
6
7
8
9
in
II
Total
It Sitfcs
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
44**
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 blowdown open-ended lines, site open-ended lines, and compressor pressure relief valves.
** Gas lift compressors not counted in the totals for this site with gas lift for oil recovery.
*** 1 turbine drives 2 propane compressors. No natural gas compressors.
"Y" = Yes
-------
TABLE A-8. TRANSMISSION COMPRESSOR STATIONS
Site Number
1
2
3
4
5
6
7
8
9
10
11
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 A-8. (Continued)
Site Number
12
13
14
15
16
17
IS
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 A-9. STORAGE COMPRESSOR STATIONS
Site
1
2
3
4
5
6
7
8
Total
8 Sites
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
*** _ Rec}p Engine w/Centrifugal Compressor
means no data available, "Y" = Yes
UG = Underground Storage Station
LNG = Above ground, Liquefied Natural Gas Station
-------
APPENDIX B
Methods for Estimating Activity Factors
B-l
-------
B.l RATIO (SITE-WEIGHTED) METHOD FOR ESTIMATION OF AN
ACTIVITY FACTOR
As discussed in Section 3.4, the ratio (site-weighted) method has been used to
estimate activity factors on the basis of well counts or production. In that section, a
numerical example is given which illustrates the ratio method for that purpose. This
appendix provides a further description of the ratio method, including calculation of a
confidence interval for an estimate obtained by this method. In Section B.l.l, estimation
using the ratio method is described. In Section B. 1.2, methods for computing a confidence
interval for the estimate produced by the ratio method are discussed. Further discussion of
the ratio method and methods for computing confidence intervals is provided by Cochran4
(1977).
B.l.l Estimation Using the Ratio Method
Suppose
y; = device count (e.g., number of separators) at the i4 sampled site;
x, = value of the extrapolation parameter (number of wells or gas
production) at the ith site;
n = number of sites sampled;
X = the regional value of the extrapolation parameter, e.g. the total
number of wells in the region; and
N = the total number of sites in the region.
For the purposes of illustration, this section discusses the estimation of the
regional number of separators by using the well method. Then, by the ratio method, the
following is the estimate of the number of separators per well:
r = i
x
or
B-2
-------
n
R =
n
E*.
i=l
This estimated number of separators per well and the regional number of wells is then used
to estimate the number of separators in the region:
Yr = RX
B.1.2 Confidence Interval for Estimate Produced by the Ratio Method
Cochran presents two approaches for estimation on the basis of confidence
intervals. One method for calculation of the confidence interval is based on the assumption
that the ratio estimate, ft, is approximately normally distributed. In many applications, the
normality assumption is satisfied only if the sample size (the number of sites visited in our
application) is sufficiently large (at least 30) and the relative uncertainties (coefficients of
variation) in both the average number of separators per site and the average number of
wells per site are both sufficiently small (less than 10%). The suggested rules of thumb are
given by Cochran. If the ratio is normally distributed, its confidence interval will be
symmetric.
If the ratio itself is not approximately normally distributed, but the numerator
and denominator are both normally distributed, the ratio will tend to have an asymmetric
confidence interval in which the upper confidence limit is more separated from the mean
than is the lower confidence limit (see Figure B-l). A second method handles this case.
As is discussed below, the cause of the asymmetry in some applications is a fundamental
consideration in the selection of a method.
B-3
-------
90% mean 90%
Confidence x Confidence
Limit Limit
Emissions
a) Normal Distribution
90% 90%
Confidence mean Confidence
limit x Limit
Emissions
b) Lognormal Distribution
Figure B-l. Conceptual Comparison of Normal and Lognormal Distributions
B-4
-------
For the ratio a/b, "a" and "b" are both subject to random variability but both
are non-negative. Given that "b" is subject to random variability and bounded below only
by zero, a value very close to zero could occur. The ratio has no upper bound as "b"
approaches zero; thus the error in the ratio is unbounded above. But the ratio has an
absolute lower bound of zero. The possibility of values extremely larger than the true
value, without a corresponding possibility of values extremely lower than the true value,
tends to cause the uncertainty in the ratio to be asymmetric.
The method based on the assumption that the ratio is approximately normally
distributed will be called method 1. The method that produces asymmetric confidence
intervals will be called method 2. Radian has performed calculations to compare these two
methods. Tests revealed that method 2 is capable of producing an upper confidence level
that is unreasonably large from an engineering point of view (see the discussion below
pertaining to separators for the Central Plains Region). The confidence limits produced by
method 1 under these circumstances are much more reasonable from this perspective.
Both engineering judgement and further statistical calculations have indicated
that method 2 is preferable for this application. First, the asymmetric confidence interval is
based on the general mathematical situation described above, in which the denominator can
become arbitrarily close to zero. But in our application, the denominator is the sum of the
production levels or of the numbers of wells for the sites visited in a region. From a
practical perspective, it is not reasonable to expect that either of these sums can become
arbitrarily close to zero, causing an extremely large ratio of separators per well or
separators per unit of production.
The number of wells at a site of interest must be at least one. Thus, the sum
of the numbers of wells has a lower bound equal to the number of sites visited. The
production does not have a definable lower bound of this nature. The argument above still
applies; however, it is not reasonable to expect an arbitrarily small production rate at all
visited sites in a region, allowing an unbounded ratio of devices per unit of production on
the basis of the data for all sites.
B-5
-------
The argument above pertains to the possibility of an arbitrarily small
denominator, which could cause extreme skewness; Figure B-2 depicts a hypothetical
distribution that is skewed, or asymmetric. The relationship between the number of devices
and the number of wells or amount of production is also relevant. Theoretically, positive
skewness in a ratio could result from positive skewness in the numerator; this would be a
special concern if the numerator could increase without bound, independently of the value
of the denominator. In this application, however, it is not reasonable to expect that the
number of separators attached to a given well is unbounded; similar comments apply for
other device types. Further, it is not reasonable to expect that the number of separators at
the visited sites in a region is independent of the total production at those sites and can
become arbitrarily large, independently of the production level.
The intuitive arguments above indicate that certain mathematical causes of
marked asymmetry do not exist in this application. However, these arguments do not prove
that asymmetry cannot exist at all. A further investigation was performed on the basis of
statistical calculations. For each of a selected set of regions and device types, the number
of devices was divided by the extrapolation parameter (wells or production) for each site.
This produced a ratio for each site visited for a given region and device type. In most
cases, the number of sites is too small to allow a detailed characterization of the
distribution. For separators for the Atlantic/Great Lakes region, however, there were 19
sites. The distribution of separators per well is displayed for this case in Figure B-3. The
histogram is somewhat ragged, because of the sample size; even 19 is a small sample size
to characterize a distribution. Nevertheless, there is no evidence of positive skewness.
Despite the raggedness of this empirical distribution, a hypothesis test indicated that this
distribution does not differ to a significant extent from a normal distribution.
Figure B-4 presents the histogram for the ratio of separators per unit of
production for the same region. In this case, there is evidence of asymmetry in the
distribution of the site-by-site ratios, and the hypothesis test indicated that this distribution
differed significantly from a normal distribution. The primary reason for the visual
B-6
-------
0 10 20 30 40 50 60 70 80 90 100
Uncertainty (%, Normal Assumption)
—A— Emissions * Normal Limit a Lognormal Limit
Figure B-2. Comparison of 90% Confidence Limits (Normal and Lognormal Assumptions)
-------
0.00 0.25 0.50 0.75 1.00
Separators/Well
Figure B-3. Distribution of Separators per Well
B-8
-------
10 30 50 70 90
Separators/Production (MMcfd)
Figure B-4. Ratio of Separators per Unit of Production
B-9
-------
impression of asymmetry is a single site with a ratio of 99.9 separators per MMcfd of
production. Asymmetry in the site ratios, however, does not necessarily imply that the
error in the ratio for the region is asymmetrically distributed. For the site with the large
ratio, there are 1,582 separators and 16 MMcfd of production. Another site with a more
moderate ratio has a much larger impact on the ratio for the region. This site has
3,227 separators and 81 MMcfd of production, so the ratio is 39.8 separators per MMcfd.
The site values of the number of separators per well for the Central Plains
region revealed evidence of negative skewness. That is, instead of a long tail to the right,
as in Figure B-4, there was some evidence of a long tail to the left. Since there were only
seven sites, a histogram of this data set would not be meaningful and is not shown. In this
case, method 2 produced an upper confidence limit for the ratio of separators per well that
was unreasonably large in this case from an engineering or a statistical point of view. This
upper limit was several times the largest site ratio of separators per well for the Central
Plains region and exceeded almost all the separator-per-well site values for several regions.
In this example, method 1 produced results that were considered to be much more
reasonable. While negative skewness was the exception, this example provides another
illustration of why method 1 was preferred over method 2.
Moreover, asymmetric uncertainties of individual parameters exist for other
reasons. Confidence intervals with greater than 100% uncertainty exist for activity factors,
emission factors, or emission rates for some source categories. One possible explanation is
that the error in the estimated parameter is not normally distributed. The ultimate objective
of the study, however, is to estimate the national emission rate. The sum of the emission
rates for 82 source categories will tend toward normality, even if some of the individual
values summed are nonnormal. Thus, even if some category parameters were not normal,
this would not necessarily invalidate the confidence interval for the national emission rate.
Moreover, an assessment has been made of the effect of a lognormal error in the industry
emission rate. The upper confidence limits based on the normal and lognormal assumptions
differ by a small amount, and the target precision is met on the basis of either' assumption
(Section A.4).
B-10
-------
Based on a finite sample of size n (i.e., n sampled sites), the following is an
approximation of the variance of the error in YR:
n\n i=l
The quantity N, the total number of sites in the region, is not known and,
therefore, must be estimated. The total number of separators, X, in the region is known.
The quantity X divided by the average number of separators per site is an approximation of
the number of sites in the region. This method of estimating N was suggested to Radian by
Jonathan Cohen of ICF Kaiser in a private communication.
Thus, N is an estimate rather than a known constant. The value N is used
only in quantifying the uncertainty of YR, however, and not in estimating YR. The quantity f
is the sampling fraction, n/N.
The equation given by Cochran for a symmetric confidence interval for YR is
as follows:
where z is a tabulated value of the standard normal distribution selected according to the
confidence level; for a 90% confidence interval, the z value is 1.645. The z value is
appropriate when the quantity estimated (^r) has an uncertainty, but the uncertainty of the
variance [v(Yr)] can be neglected. The use of the z statistic is generally accepted if the
sample size is greater than 30.
According to Cochran's rules of thumb, the sample size would be at least 30
when this expression for the confidence interval was used. In our case, however, the
decision that the symmetric confidence interval was preferable to the asymmetric
confidence interval even if the sample size was less than 30 was based on engineering
B-ll
-------
considerations and data analysis, as is discussed above. To account for the uncertainty in
the variance as well as in the estimate, we have replaced z in the expression above by the
appropriate t value. Even though the t-distribution does not apply exactly in this context,
replacing z by t provides a degree of conservatism; that is, somewhat wider confidence
intervals are produced, which tends to account for the uncertainty in v(YR). The resulting
confidence interval is as follows:
rB
B.2 COMBINATION OF ESTIMATES OF AN ACTIVITY FACTOR
The methods discussed in the preceding section were used to estimate the
activity factor and its uncertainty on the basis of both well counts and production. The
arithmetic average of the two estimates was computed to obtain the final estimate.
The two estimates are based on different extrapolation factors (values of the
Xj) but common device counts (values of the y,). The device counts vary by site and are
subject to sampling error. Thus, this source of sampling error was common to the two
estimates of the activity factor. It has been discussed elsewhere that separate measured
quantities (e.g., emission rates from different types of devices) may have correlated
sampling errors. The evidence here for correlation is much stronger, however, since
common data are used in the two estimates. Thus, steps were taken to account explicitly
for the correlation. To address this issue, we introduce the following notation:
xw i = number of wells at the ith site;
xpi = production at this site;
= estimate of R on the basis of wells;
B-12
-------
Rp = estimate of R on the basis of production;
^r,w = estimate of YR on the basis of wells; and
Yr = estimate of YR on the basis of production.
By substituting xw i for Xj in the appropriate equations in the preceding
section, for example, one obtains the estimate of the number of devices in the region on the
basis of wells and the confidence interval for this estimate. The following is a sample
estimate of the covariance between the errors in the two estimates:
UrW (yrW
n\n i) i=1
This expression satisfies important required properties of the covariance, such as the
symmetry property:
cov(tR w, tRp) = cov(tR p, tR w)
Additionally, the covariance between a quantity and itself equals the variance of that
quantity. This can be confirmed simply by replacing all "w" subscripts with a "p"
subscript, to obtain the variance of Yp.
In a textbook application of the ratio method, the quantity N would be
known. As discussed earlier, N must be estimated in this application. In estimating the
covariance above, the average of the two estimates of N was used, both where N appears
explicitly in the covariance equation and in calculating f.
Now, the expression for the final estimate of the activity based on the
arithmetic average approach is as follows:
The variance of this expression is:
B-13
-------
Y +Y
f = R,w R,p
R,avg r>
^ ^ ™*V+2cov(V V>+v^V>
Var\lRaVg) 4
The confidence interval for the final estimate is as follows:
YR,avg ± *
In some instances, the number of sites for which data existed for both wells
and production did not coincide exactly. In these cases, the covariance was computed on
the basis of the sites for which common data did exist. This provided a somewhat
conservative (large) estimate of the covariance. This calculation of the covariance
represents the case in which the sites in common for the two extrapolation parameters are
the only sites. But the fact that sites exist with data for wells but not production (or vice
versa) introduces an element of independence between the estimates of YR based on the two
extrapolation parameters. The somewhat conservative covariance estimate produces a
somewhat conservative confidence interval for the final estimate of YR.
To account for this case, the correlation between the errors in the two
estimates was computed as follows:
r_
Jvar(YR,Jvar^Rj
Then the half-width of the confidence interval for YR was computed as
follows:
where tp and tw are the t-values appropriate for the sample sizes for the two extrapolation
parameters. The expression involving the correlation coefficient was written in the manner
B-14
-------
shown to emphasize that this is approximately an error propagation using half-widths of
confidence intervals. Each t-value is grouped with its respective standard error (the square
root of an error variance is a standard error). The expression above can be simplified
algebraically to the following:
\tiVarWR*>+2W0VWx,~<
This expression involving different sample sizes for the two estimates reduces
to the simpler expression for the half-width of the confidence interval given earlier if the
sites for which data exist for wells and production are the same.
B-15
-------
APPENDIX C
Conversion Table
C-l
-------
Unit Conversion Table
English to Metric Conversions
1 scf methane = 19.23 g methane
1 Bscf methane = 0.01923 Tg methane
1 Bscf methane = 19,230 metric tonnes methane
1 Bscf = 28.32 million standard cubic meters
1 short ton (ton) = 907.2 kg
1 lb = 0.4536 kg
1 ft3 - 0.02832 m3
1 ft3 = 28.32 liters
1 gallon = 3.785 liters
1 barrel (bbl) = 158.97 liters
1 inch = 2.540 cm
1 ft = 0.3048 m
1 mile = 1.609 km
1 hp = 0.7457 kW
1 hp-hr = 0.7457 kW-hr
1 Btu = 1055 joules
1 MMBtu = 293 kW-hr
1 Ib/MMBtu = 430 g/GJ
T (T) = 1.8 T (°C) + 32
1 psi = 51.71 mm Hg
Global Warming Conversions
Calculating carbon equivalents of any gas:
MMTCE = (MMT of gas) x
MW, carbon
MW, gas ,
x (GWP)
C-2
-------
Calculating C02 equivalents for methane:
MMT of CO, equiv. = (MMT CH.) x
MW, C09\
MW, CH,
x (GWP)
where MW (molecular weight) of C02 = 44, MW carbon = 12, and MW CH4= 16.
Notes
scf = Standard cubic feet. Standard conditions are at 14.73 psia and 60°F.
Bscf = Billion standard cubic feet (109 scf).
MMscf = Million standard cubic feet.
Mscf = Thousand standard cubic feet.
Tg = Teragram (1012 g).
Giga (G) = Same as billion (109).
Metric tonnes = 1000 kg.
psig = Gauge pressure.
psia = Absolute pressure (note psia = psig + atmospheric pressure).
GWP = Global Warming Potential of a particular greenhouse gas for a given
time period.
MMT = Million metric tonnes of a gas.
MMTCE = Million metric tonnes, carbon equivalent.
MMT of C02 eq. = Million metric tonnes, carbon dioxide equivalent.
C-3
------- |