EPA/6QO/A-94/086
92-142.06
RCN #275-026-62-04
Methane Emissions from the Natural Gas Industry:
Production and Transmission Emissions
Prepared for:
Air and Waste Management Association
Prepared by:
Matthew R. Harrison
Radian Corporation
8501 Mopac Blvd.
P. O. Box 201088
Austin, TX 78720-1088
1 May 1992
-------
92-142.06
INTRODUCTION
The Gas Research Institute (GRI) and the U.S. Environmental Protection Agency (EPA) have co-funded a
project to quantify methane emissions to the atmosphere resulting from operations in the natural gas
industry. The results of the study will measure or calculate all methane emissions, from production at the
well and up to, but not including, the point of final use. When these data are combined with those of other
studies that quantify greenhouse gas emissions from methane consumers (i.e., various combustion sources), a
definitive comparison of the relative environmental effects of using methane versus other fuels will be
possible.
The methane emissions project is being executed in three phases: Phases 1 and 2 identified all potential
emitting sources and then established methods for measuring or calculating emissions from those sources.
Phase 3 will gather statistical samples of data to complete the extrapolation to national estimates. An
accuracy target of ± 100 billion standard cubic feet (Bscf) per year of calculated emissions has been
established.
Currently, Phase 2 of the project is complete, and Phase 3 is beginning This report presents the methods
and preliminary conclusions from Phases 1 and 2 as well as the plans for completing Phase 3. Some
calculations have been completed in Phase 2 and, although the calculations are not yet based on statistically
significant data, some preliminary conclusions can be drawn. The purpose of this paper is to describe the
methodology and sources of emissions and present the current estimate of methane emissions from the
production and transmission phases of the natural gas industry.
BACKGROUND
Global warming or "the greenhouse effect" is an anticipated global climate change caused by the
accumulation of energy-absorbing gases in the atmosphere. As trace gases (such as COj, Cr^, N>O, and
CFCs) that absorb energy at various wavelengths accumulate, it is anticipated that the radiative properties of
the Earth's atmosphere will change, and more heat will be trapped in the atmosphere. The relative impacts
of several greenhouse gases can be seen by comparing the Radiative Forcing Potential of the gases, as shown
in Table I. It is predicted that the greenhouse effect will have a significant impact on the global environment
and economy, and it is a worldwide concern.
Many of the anthropogenic sources of greenhouse gases are directly related to energy production and
consumption. The combustion of all types of fuels generates greenhouse gases (COX, l^O, hydrocarbons),
but some fuels produce more greenhouse gas products than others. Table II shows the relative amount of
one combustion product, COi, produced by various fuels. Some fuels, like methane, produce fewer
greenhouse gas combustion products but may emit unburned fuel directly to the atmosphere through leaks
wherever the fuel is handled.
The types of greenhouse gases and the sources of greenhouse gas emissions are diverse. Some are natural,
some anthropogenic. Methane, for example, can be emitted from many sources: agriculture (rice paddies),
ruminant animals, animal and municipal waste management facilities, coal mining, biomass burning, and
natural ecosystems (tundra, swamps). The sources of concern in this report are natural gas production,
transmission, and distribution systems.
RESULTS OF PHASE 1 AND PHASE 2 INVESTIGATIONS
Overview
In the natural gas system, methane is produced from oil or gas wells and routed via pipe to treatment
systems. The treated methane is then compressed and sent via pipeline to the end users. The entire system
is designed to contain the methane, since it is a valuable product and is a high-pressure gas. Nevertheless, a
small percentage of methane is emitted during the process.
-------
92-142.06
TABLE I. Radiative forcing potentials for principal greenhouse gases.
Greenhouse Gases
Carbon dioxide
Methane (including indirect)
Nitrous oxide
CFC-12
20 yr.
1
63
210
7100
Time Horizon
100 yr.
1
21
290
7300
500 yr.
1
9
190
4500
Source: Ref. 1.
TABLE n. Carbon dioxide emission rates for conventional fuel types.
Fuel
CQj Emission Rate (g C/109J)
Natural gas
Crude oil-based fuck
Bituminous coal
135-142
18.2-20.d*
23.7-23.9
Ranges are attributable to product mix (i.e., gasoline vs. fuel oil and gasoline).
Source: Ref. 2.
-------
92-142.06
Methane emissions from the system can generally be grouped into two major categories: Fugitive Emissions
and Vented Emissions. Fugitive emissions are slow, steady, and continuous losses of process fluid through
the sealing mechanism separating the fluid from the atmosphere. These losses are particularly common
around the moving parts of valve stems and pump shafts. Although most individual equipment leaks are
small emission points compared to typical process or storage vents, the large number of components makes
the aggregate emission rate a significant source of emissions.
Vented emissions are all the remaining emissions that occur from the intentional or unintentional purging,
flaring, or releasing of methane during all phases of operations. Venting is a broad category that covers
numerous emission sources from normal operations, routine maintenance, and upsets. Vented emissions that
result from flaring count only the unhurried methane released because of incomplete combustion.
The initial results of the Phase 1 and Phase 2 calculations can be seen in Table III, which shows the current
emissions estimates for the fugilive and the venting categories under each natural gas system segment. The
methane emission rate for production, processing,and transmission systems is 167 Bscf per year, or
approximately 81% of total natural gas industry emissions (205 Bscf). Distribution segment emissions are
discussed in a companion paper. All of the subcategories used to calculate the summary numbers b Table
III are broken out in Table IV. Although emission rates are not shown for each subcategory, they have been
calculated and will be presented when completed during Phase 3.
Detailed Categorizations
Calculation of the current estimates for fugitive and vented emissions followed various approaches. The
following sections will explain the Phase 1 and Phase 2 approaches in detail.
Fuotives.
Overview - Fugitive methane emissions are estimated from equipment leak emission factors and
component counts for model plants. The national estimate of methane emission from the natural gas
industry then extends these estimates by factoring in population estimates (i.e., number of gas wellheads, or
number of gas plants) and average emission rates. The fugilive emission rate is summed for all component
types (valves, connections, compressor seals, relief valves, open-ended lines) and may involve multiple
streams.
Phase 1 Efforts - During the Phase 1 program, existing data on fugitive emissions were reviewed
from various phases of natural gas production and use. A total of eight studies were reviewed, and it was
found that most of the studies covered fugitive organic emissions and were not concerned with methane
emissions. Table V summarizes reports from the three major studies available during this review.
API/Rockwetf: This study yielded probably the most comprehensive data currently available describing
fugitive hydrocarbon emissions from the oil and gas production and processing segment. The study was
published in 1980 and included soap score screening of a total of 173,236 components. A total of 8,466 leaks
were quantified for volumetric emission rate, and 1,914 chromatograms were obtained for speciation.
Drawbacks of the API/Rockwell data include the questionable accuracy of soap score screening, the
questionable validity of a large number of emission factors, and the age of the data.
EPA/Radian (Gas Plants)4: This study combined data from two gas plants tested by API/Rockwell plus
four gas plants tested by EPA/Radian. The study provided correlation equations for total hydrocarbon
(THC) emissions from valves, compressor seals, and pump seals. It was published in 1982 and included
Organic Vapor Analyzer screening of 6,585 components. A total of 212 components were bagged, and leaks
were quantified using GC. Drawbacks of the EPA/Radian data include the age of the data.
-------
TABLE III. Current Phase 2 methane emission estimates.
92-142.06
Category
Percent of Total
Industry
Emissions
5.2
5.2
PRODUCTION (field production, gathering)
Fugitives I2L6
Venting 26.6
PROCESSING PLANTS
Fugitives
Venting
TRANSMISSION (transmission pipelines and compressors, and gas storage facilities)
Fugitives
Venting
DISTRIBUTION (mains and service pipeline)
Fugitives
Venting
12.9
18.9
16.6
2.0
TOTAL
-------
92-142.06
TABLE IV. Production and transmission subcategories.
Segment
Process Area Emission Type
Process Subcategory (Emitting
Mode Process or Component)
PRODUCTION Field Production Fugitives
Venting
Normal Gas Wells
Operation Oil Wells
Field Separation Equipment
Normal Drilling and Completion
Operation Flaring
Compressor Engine Exhaust
Pneumatic Device Vents
Gas Letdown Pump Vents
Dehydrator Vents
AGR Vents
Tank Vents
Routine
Maintenance
System
Upsets
PRODUCTION Gathering Lines Fugitives Normal
Operation
Venting Normal
Operation
Routine
Maintenance
Upsets
PRODUCTION Gas Plants Fugitives Normal
Operation
Well Maintenance
Surface Equip. Maintenance
Pressure Relief Discharge
Overflow Tank Vents
Pipeline Leaks
Compressor Stations
Metering and Pressure
Regulator
Compressor Engine Exhaust
Pneumatic Device Vents
Pipeline Blow and Purge
Drip Blow and Purge
Relief Valve Operation
Dig-Ins
Cryogenic
Refrigerated Adsorption
Refrigeration
(Continued)
-------
TABLE IV. (Continued)
92-142.06
Process
Segment Process Area Emission Type Mode
PRODUCTION,
Coot'd
Venting Normal
Operation
Routine
Maintenance
Upsets
TRANSMISSION Pipelines and Fugitives Normal
Compressors Operation
Venting Normal
Operation
Routine
Maintenance
Upsets
TRANSMISSION Gas Storage Fugitives Normal
Operation
Venting Normal
Operation
Routine
Maintenance
Subcategory (Emitting
Process or Component)
Others
Compressor Engine Exhaust
Pneumatic Devices
Dehydrator Vents
Vessel Slowdown
Compressor Start/Stop
Pressure Relief Discharge
Pipeline Leaks
Compressor Stations
Metering and Pressure
Regulator
Compressor Engine Exhaust
Pneumatic Device Vents
Pipeline Blow aad Purge
Drip Blow and Purge
Dehydrator Blow and Purge
Relief Valve Operation
Dig-Ins
-
Pneumatic Device Vents
Compressor Engine Exhaust
Vessel Purge and Slowdown
Well Workover
-------
TABLE V. Summary of fugitive emission reports reviewed.
Sponsor/Contractor
Publication Date
Number of
Sites Tested
Sources
Screened
Sources
Bagged
Test Methodology
API/Rockwell Intl.
March 1980
21'
173,236
EPA/Radian Corp.
July 1982
4C
6,585
co MMS/ABB Environmental
1991'
r
89,466
8,466b Soap score technique. All heavy1
leakers were bagged. Statistical
selection of low to moderate
leakers also bagged. Non leaking
(or zero soap score) sources were
not bagged.
212 OVA screening. Did not include
small number of streams
containing
-------
92-142.06
MMS/ABB Offshore Study5: This study was conducted on Pacific DCS (offshore) oil and gas facilities. A
Draft Final report was issued in 1991, but it has been subject to the review and revision of various regulatory
agencies and a revised Final Report is expected in May or June 1992. This study included results from the
screening of 89,466 components on seven platforms. A total of 294 components were bagged and leaks were
quantified using GC. Drawbacks of the MMS/ABB data include limited applicability to sources outside the
Pacific OCS region, questions about the emission rates of inaccessible components, and questions about the
effect of "pre-notification."
The conclusion of the review was that these reports could be used as the starting point for an investigation of
the current program.
Also during the Phase 1 program, Radian conducted a short-term field study in 1990 of fugitive emissions at
Amoco's Ft. Lupton gas production/gas plant facility. This study involved screening 20 gas wellheads, a
bagging test on a single gas wellhead, and screening 565 components in the gas plant.
Other work included the development of model plants to direct future data collection activities. For
example, the number of gas plants using various processing techniques were obtained from industry surveys,
as shown in Figure I.
Phase 2 Efforts - During Phase 2, Star Environmental conducted a fugitive emission study funded
by the American Petroleum Institute (API) to validate existing correlation equations and emission factors for
petroleum production fugitive emissions, and to develop a profile of air toxics emissions from petroleum
production sources. GRI co-funded the API/Star program to obtain updated methane fugitive emission
factors. The API/Star program consisted of fugitive screening and bagging at twelve sites: four gas
production facilities, four oil production facilities, and four gas processing plants. In addition, four of the
twelve facilities were the same as written up in the original 1980 API/Rockwell report. The preliminary
results for gas production facilities are shown in Table VI.
As seen in Table III, fugitive emissions appear to contribute approximately 47% of total industry emissions.
The gas production and transmission segments are both major sources of fugitive methane emissions.
Because the API/Star program will provide updated emission factors for gas production and processing,
these segments will have less uncertainty than the transmission segment.
Vented Emissions.
Overview - Radian has completed interim work on vented emissions for Phase 2 of the project
which builds on previous work performed by Pipeline Systems Incorporated (PSI) in Phase 1 and in the early-
stages of Phase 2. The Radian Phase 2 work added new categories for venting and calculated preliminary
numbers for each category. In addition, many numbers from the original PSI reports were revised with new
data. The major sources were determined via interviews and site visits to gas and oil fields, gas processing
plants, and gas storage facilities.
Options for evaluating the emissions from venting in the production segment were to: 1) use the nationally
tracked "vented and flared" numbers reported by each operator, 2) calculate the total vented and flared
emissions by calculating an emissions range for each known emitter, 3) calculate the total vented and flared
emissions by measuring the emissions from each known emitter or group of emitters, or 4) use some
combination of 2) and 3).
Reported National Numbers - A detailed evaluation of the nationally reported vented and flared
emissions statistics revealed that the numbers were not valuable as true emissions measurements. Vented
and flared emissions numbers reported to state agencies come from two sources. Gas and oil field operators
report all dispositions of their produced gas, including a "vented and flared" disposition, and gas processing
-------
92-142.06
Cryogenic-
Expander
(36.4%)
Cryogenic-
Joule-Thomson
(6.0%)
Absorption (5.8%)
Adsorption (1,7%)
Compression (2.2%)
Refrigeration
Absorption
(21.5%)
Refrigeration
(26.4%)
Source: Ref. 6.
FIGURE I. Gas processing plants by process type.
10
-------
92-142.06
TABLE VI. Preliminary fugitive methane emissions factors for gas production fields.
From West Coast Dry Gas Field
Site contains 27 wells and 7,700 components
Each well has a separator (27 total)
Each well produces less than one million SCF/day
Site contains 7 compressors and 8
Wells
Heaters
Separators
Compressors
TOTALS
Count
27
8
27
7
Connections
1000
800
3000
1300
6100
From Rocky
Valves
300
150
625
200
1275
Mountain
fired heaters
Open-
ends
50
35
120
20
225
Gas Field
Relief V.
7
10
65
30
112
(some oil also)
Leaks
60
70
100
30
260
CH4
(lb/day)
14
12
19
6
51
Site contains 15 wells and 8,300 components
Each well
Each well
has a separator (15 total)
produces less
than one million SCF/day
Site contains 7 dehydrators and 15
Wells
Separators
Sales Areas
Dehydrators
TOTALS
Count
L5
15
15
7
Connections
700
2600
2100
1300
6700
Valves
150
450
350
200
1150
sales areas
Open-
ends
15
75
100
50
240
Relief V.
0
90
80
55
225
Leaks
33
240
200
215
688
CH4
(lb/day)
25
125
100
100
350
11
-------
92-142.06
plant operators also produce a similar disposition report. These monthly numbers are then routed from the
state agencies to the Department of Energy's (DOE) Energy Information Administration (EIA), where they
are published nationally.
Unfortunately, the vented and flared data generated by the gas and oil field operators have little scientific
basis, primarily because there is no regulatory driving force for a consistent basis. The state agencies are
only concerned with "venting and flaring" from a toxic gas (H^S) release standpoint (they want to make sure
the gas is burned), or from a tracking of production "allowables" standpoint (they want to ensure equity
among the producers in a field). States have very few regulations covering the methods of calculating vented
and flared emissions, and tracking venting and flaring numbers vary from state to state.
Since wellhead flare lines have no meters, any material flared or vented at the well is estimated by the
operator and is based on how many events are remembered and recorded and on an undefined method for
estimating the gas release. The operator may use the field's gas to oil ratio, the most recent well test, or
his/her best guess as to the amount released. The basis is far from scientific.
Gas processing plants use a more scientific basis, performing a total plant material balance using all the plant
disposition meters. However, the imbalance, or Unaccounted For (UAF), is automatically assigned to the
"vented and flared" disposition in the monthly report. Thus, the vented and flared category for gas plants is
not really measured, but simply the result of a material balance. If all other dispositions were perfectly
measured, the UAF would represent the amount vented and flared plus the amount emitted from fugitives.
However, since meters often have uncorrected biases, since the total gas plant balance may vary by an
amount larger than the UAF, and since fugitive emissions can be large, the UAF is probably not a realistic
measure of venting and flaring. Some proof of this is that the plant's "vented and flared" amounts may vary
and may be relatively high despite a flare meter reading of zero and despite routine maintenance practices
within the plant (no venting),
The national numbers for venting and flaring are therefore useful only as yardsticks against which to
compare the more accurate detailed calculations and measurements methods that follow.
Measurement Methods - Although calculations are the basis for measuring most of the vented
categories, there are categories where measurements can be taken to populate the database. Some methods
considered were: tracer tests, direct on-line gas chromatograph (GC) measurement of vent stacks, and
bagging and screening for specific devices. Phase 2 made little use of these measurements because of the
vast number of categories that first had to be identified. There were, however, a few applications of on-line
GC tests to glycol dehydrator vents. There were also some tracer tests for total emissions (fugitive and
vented) from oil and gas fields and gas plant facilities. Future use of the measurement methods will be
defined during the Phase 3 planning.
Calculation Methods - Venting sources often require a calculation.*! approach. Unlike fugitive
emissions, which is a rather narrow category, venting covers all kinds of emission sources. Whereas fugitive
components can be broadly classified and then measured by the same method for all component types,
venting components are too diverse to measure by the same method and often too numerous to measure at
all. Venting emissions vary from unburned methane at the flare tip, to pneumatic device emissions, to
continuous glycol regenerator stack emissions, to vessel purging, to intermittent pressure relief valve lifting.
To quantify methane emissions from venting, each possible venting source was defined by examining all the
equipment used by the industry, as well as the practices for operating and maintaining the equipment. A
data search was then performed for each source to find applicable studies and emission measurements. If
measurements were unavailable, a method was devised to calculate unit emissions based on other engineer-
ing data. For example, data on the average size of separators, the average pressure of separators, the
12
-------
92-142.06
population of separators, and the frequency of separator blowdown might be used to calculate maintenance
emissions from separator blowdowns when direct measurements are unavailable. Thus, an individual plan
was developed for each source.
PSI initially populated the database for some calculations by visiting a few facilities. PSI made an estimate of
emissions for each field and then extrapolated to total U.S. emissions, assuming the fields were "models" of
all fields in the U.S. PSI extrapolated by multiplying the field emissions by a ratio of U.S. wells to model
field wells (or U.S. production to model field production). Since a few fields of any type are not statistically
significant, Radian visited and interviewed a number of other operators of multiple fields. Radian has added
many new emission categories to the database in the production area and has produced initial calculations in
each new category, as well as updated many of the existing categories.
Asseen in Table III, venting emissions appear to contribute approximately 53% of total industry emissions.
The gas production segment is estimated as the largest source of venting emissions, with the transmission
segment as the second largest source.
CONCLUSIONS AND PHASE 3 PLANS
In Phase 3, the plan for populating of the database will be finalized. The calculations will then be made and
presented in a final report to GRI and EPA. The plan will include selections of measurement techniques (if
any) for particular categories, and selection of the number of site visits and surveys necessary to statistically
populate the database. Data gathering will then begin.
Fugitives
Future work during Phase 3 will include reviewing new reports as they become available. In particular, the
API/Star Fugitive Emission Project and the MMS/ABB Pacific OCS Fugitives Report should be available
soon. Data from these reports will be incorporated into revised emission factors and component counts.
Also, Radian has proposed to validate the emission factors with a field test program consisting of screening
and bagging at possibly two additional gas production sites (with adjoining gas processing plants). At the
same time, component inventories and stream composition data will be collected. This validation work could
focus on either "typical" faculties with minimal leak detection and repair (LDAR) programs, or it could
attempt to target facilities with more intensive fugitive emissions control. This might allow an estimate to be
made of the degree of reduction in emissions actually achievable through improved inspection and
maintenance (I&M).
The goal of the Phase 3 work will be to obtain a national emissions estimate within the specified target
accuracy.
Vented Emissions
The four largest venting emission subcategories (see Table IV for all subcategories) were compressor
exhaust, pneumatic devices, dehydrator vents, and vessel blowdowns. These subcategories made up a large
percentage of total vented emissions and will receive the most attention for refinement in Phase 3.
For venting categories in production, gas processing, gathering lines, and gas storage, the calculational
approach will be the primary method of completing the emissions estimates. Data gathered from various
operators of field equipment through interviews, site visits, and questionnaires will populate the database for
equipment types and emission rates. Some measurement techniques (tracer tests) will also be used to
supplement the calculational approach in high emission rate categories. As with fugitives, the goal will be to
obtain a national emissions estimate within the specified target accuracy.
13
-------
92-142.06
REFERENCES
1 Intergovernmental Panel on Climate Change (IPCC). (1990) Climate Change: The IPCC Scientific
Assessment. J.T. Houghton, GJ. Jenkins, and JJ. Ephrams, eds. Cambridge University Press.
Cambridge, UK. 365 pp.
2 Marland, G. (1982) The Impact of Synthetic Fuels on Carbon Dioxide Emissions." In: Clark, W.C.,
ed. Carbon Dioxide Review 1982. Oxford University Press, New York. pp. 406-410.
3 Eaton, W.S., F.G. Bush III, J. Coster, J.C. Delwiche, and H.O. Hartley. Fugitive Hydrocarbon
Emissions from Petroleum Production Operations (2 volumes). Prepared for the American Petroleum
Institute by Rockwell International, Los Angeles, CA. API Publication No. 4322. March 1980.
4 DuBose, DA., J.I. Steinmetz, and G.E. Harris. Data Analysis Report Frequency of Leak Occurrence
and Emission Factors for Natural Gas Liquid Plants. USEPA, Office of AQPS, EMB Report No.
80-FOL-l. July 1982.
5 Countess, RJ., and D. Herkhof. "Fugitive Hydrocarbon Emissions from Pacific OCS Facilities."
Paper No. 91-91.2 presented at 84th Annual Meeting of the Air and Waste Management Association,
Vancouver, B.C. June 16-21, 1991.
6 Cepica, W J. "Gas-processing Industry Shows Signs of Maturity." Oil and Gas Journal, Penn Well
Publishing Company, New York, NY July 10, 1989.
14
-------
AEERL-P-1127
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before compl
1. REPORT NO.
EPA/600/A-94/086
2.
4. TITLE ANDSUBTITLE
Methane Emissions from the Natural Gas Industry:
Production and Transmission Emissions
5, REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Matthew R. Harrison
8. PERFORMING ORGANIZATION REPORT NO.
RCN 275-026-62-04
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
P. C. Box 201088
Austin, Texas 78720-1088
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-Dl-0031
3, SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Published paper; 4/92-9793
14, SPONSORING AGENCY CODE
EPA/600/13
75. SUPPLEMENTARY NOTES AEERL project officer is David A. Kirchgessner, Mail Drop 63,
919/541-4021. Presented at AWMA Conference, Kansas City, MC, 6/22-26/92.
. ABSTRACT
paper discusses a eofunded, Gas Research Institute/EPA project to
quantify methane emissions to the atmosphere resulting from operations in the natu-
ral gas industry. Study results will measure or calculate all methane emissions,
from production at the well and up to, but not including, the point of final use. When
these data are combined with those of other studies than quantify greenhouse gas
emissions from methane consumers (i. e. , various combustion sources), a definitive
comparison of the relative environmental effects of using methane versus other fuels
will be possible. The methane emissions project is being executed in three phases:
Phases 1 and 2 identified all potential emitting sources and then established methods
for measuring or calculating emissions from those sources, and Phase 3 will gather
statistical samples of data to complete the extrapolation to national estimates. An
accuracy target of +/-100 billion standard cubic feet per year of calculated emissions
has been established. Currently Phase 2 is complete, and Phase 3 is beginning. The
paper presents the methods and preliminary conclusions from Phases 1 and 2, as
well as the plans for completing Phase 3. Some calculations have been completed in
Phase 2 and, although the calculations are not yet based on statistically significant
data, some preliminary conclusions can be drawn.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Natural Gas
Methane
Emission
Production
Transmission
Pollution Control
Stationary Sources
13 B
21D
07C
14G
13. DISTRIBUTION STATEMENT
Release to Public
19, SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page/
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |