93-WP-104.03
EPA/600/A-93/153
Methane Emissions from the Natural Gas Industry
Matthew R. Harrison
R. M. CowgUl
L. M. Campbell
Radian Corporation
P. O. Box 201088
Austin, Texas 78720-1088
Robert A, Lott
Gas Research Institute
8600 W. Bryn Mawr Avenue
Chicago, Illinois 60631
-------
Page Intentionally Blank
-------
AEERL-P-1037
TECHNICAL REPORT DATA nc
(Please read Instructions on the reverse before con r k
J3212421.
1. REPORT NO.
EPA/600/A-93/153
2.
4. TITLE AND SUBTITLE
Methane Emissions from the Natural Gas Industry
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
M.R.Harrison, R. M. Cowgill, L.M.Campbell
(Radian); R. A. Lott (GRI)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
P.O. Box 201088
Austin, Texas 78720-1088
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-Dl-0031 (Radian)
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Published paper; 5/92-12''92
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES AEERL project officer is David A. Kirchgessner, Mail Drop 63,
919/541-4021. Presented at AWMA National Conference, Denver, CO, 6/13-18/93.
/The paper discusses a project to quantify methane (CH4) emissions from
the U. S. natural gas industry. This study will measure or calculate all gas industry
CH4 emissions--from production at the wellhead, through the system, to the custo-
mer's meter. Emissions downstream of the consumer's meter, such as the end-
user's burner emissions, are not included. When these data are combined with data
from other studies, a definite comparison of the relative environmental impact of
using natural gas versus other fuels will be possible. The study will also provide
data that can be used by the industry to identify cost-effective mitigation techniques
to reduce losses.c^A target of + or - 100 billion standard cubic feet (BCF) (0. 5% of
the total U. S. annual production) has been established as the overall accuracy goal
of the current system. (NOTE: 1 std cu ft = 0.0283 std cu m.) While individual cate-
gories may be less accurate, the cummation of the accuracy figures for all categor-
ies will be within this target. The CH4 emissions project is being conducted in three
phases: Phases 1 and 2 have identified and ranked all known potential CH4 emitting
sources and established methods for measuring, calculating, and extrapolating emis-
sions from those sources. Phase 3, currently in progress, will gather sufficient data
to achieve the accuracy goal.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Pollution
Natural Gas
Petroleum Industry
Methane
Emission
Pollution Control
Stationary Sources
REPRODUCED BY
U.S. DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL
INFORMATION SERVICE-
SPRINGFIELD, VA 22161
13 B
21D
05C
07C
14G
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report/
Unclassified
21. NO. OF PAGES
11
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
Page Intentionally Blank
-------
93-WP-104.03
INTRODUCTION
The Gas Research Institute (GRI) and the U.S. Environmental Protection Agency (EPA) have co-
funded a project to quantify methane emissions from the natural gas industry. This study will measure or
estimate all gas industry methane emissions in the 'fuel chain"-£rom production at the wellhead, through the
system, to the customer's meter. Emissions downstream of the consumer's meter, such as the end-user's
burner emissions, are not included. When these data are combined with data from other studies, a definite
comparison of the relative environmental impact of using natural gas versus other fuels will be possible. The
study will also provide data that can be used by the industry to identify cost-effective mitigation techniques to
reduce losses.
A target of ± 100 billion standard cubit feet (BCF)r(0.5 percent of total U.S. annual production) has
been established as the overall accuracy goal of the current program. While individual categories may be
less accurate, the summation of the accuracy figures for all categories will be within this target.
The methane emissions project is being conducted in three phases: the first two phases have identified
and ranked all known potential methane emitting sources and established methods for measuring, calculating,
and extrapolating emissions from those sources. Data collected during the first two phases of the project
indicate that the amount of methane emitted by the gas industry is approximately 1 percent of production or
approximately 200 BCF. The third phase, which is currently in progress, will gather sufficient data to achieve
the accuracy goal. This paper briefly summarizes the methodology being used for the completion of the
third phase.
BACKGROUND
As a result of human activities, trace gases such as carbon dioxide (COfe), methane (CH*), nitrous oxide
(NjO), ozone (Qj), and chlorofluorocarbons (CFC's) have been accumulating in the atmosphere. Unlike
oxygen and nitrogen, these gases absorb a portion of the energy radiated by the earth that would otherwise
be radiated back to space. Described as the greenhouse effect, this process eventually leads to an increase in
the Earth's temperature, and may create environmental, economic, and social problems.
Approximately half of the warming calculated to have occurred during the 1980' s was caused
Methane contributed approximately 20 percent to global warming; the CFC's, 15 percent; NjO, 5 percent;
and Oj and other gases, 10 percent1 . A substantial amount of COj, CH,, and other greenhouse gases are
emitted in conjunction with the production, transportation, and combustion of fossil fuels. Nearly 75 percent
of all COj emissions are caused by the combustion of fossil fuel. However, some fuels, such as natural gas,
emit far less COj per unit of energy generated2. For this reason, EPA and the United Nations' Inter-
governmental Panel on Climate Change (IPCC) have suggested that global wanning could be reduced if
more energy was generated using gas rather than fuels such as coal.
Methane, which is the major constituent of natural gas, is a more potent greenhouse gas than CQ, and
leakage from the natural gas system could reduce or even eliminate the advantage that natural gas has
because of its lower COj emissions. To determine whether CH's emissions counteract its benefits, GRI and
EPA are co-funding a study to quantify CH,, emissions from the natural gas industry.
PROJECT HISTORY
The study is being conducted in three phases: Phase 1 (scoping), Phase 2 (initial emission estimates,
characterizations, and method selection), and Phase 3 (method execution through measurement, data
gathering, and extrapolation). Phases 1 and 2 are complete, and Phase 3 should be finished in early 1994.
The conclusions from the first two phases will be described briefly in this section.
1 std ft3 = 0.0283 std m3
-------
Page Intentionally Blank
-------
93-WP-104.03
Phase 1
The Phase 1 scoping study revealed that:
There are sources that could not be easily measured because their emission rates vary with time
and, therefore, must be calculated. Techniques needed to be developed to calculate the "unsteady
emissions" from the many different source types that could not be measured.
There were no proven techniques for measuring emissions from the different types of steady,
measurable sources, so new measurement techniques needed to be developed and validated.
The number of emission sources is enormous. It would be impossible to physically visit every
source in the industry, therefore, scientifically defensible techniques needed to be developed that
would allow data obtained for a source type or category to be extrapolated to similar sources
throughout the industry,
Phase 2
During Phase 2 of the project it was recognized that, in addition to developing methods for (1)
measuring the steady emissions, (2) calculating the unsteady emissions, and (3) extrapolating emissions data,
accuracy targets needed to be defined and a framework for accounting for all emissions needed to be
established.
'The overall accuracy goal for the program was established to be ± 100 BCF. Then, accuracy targets
were established for each source in such a way that, if the target is met for each source, the overall accuracy
of the program (± 100 BCF) would automatically be achieved. In addition, the accuracy target was
established as a function of the size of the source. A higher degree of accuracy is required for the larger
sources. This approach provides an automatic mechanism for apportioning project resources to the most
important emission sources.
Phase 2 also developed a framework that characterized all emissions to account for every possible
source. With every known emission source categorized, the entire listing of emissions can be more easily
checked for completeness. The important characteristics that define emissions were determined to be:
Emission source. Where the emission occurs in the industry.
a) Industry segment,
b) Source type category (equipment types), and
c) Component type category (equipment components).
Operational mode. What the cause of the emission is (the facility operating mode that produces
the emission).
Emission time dependency. How constant the emission rate is (steady vs. unsteady).
Emission path type. What path the emission takes to reach the atmosphere (vented, combusted, or
leaked).
The emission source was defined according to a three-part industry source definition: (1) industry
segment, (2) major equipment type, and (3) component type. The five industry segments are: production,
processing, transmission, storage, and distribution. Subdivisions under each segment are the major
equipment types such as wells, separators, compressors, pipelines, and metering stations. Components of the
equipment types are items such as valves, flanges, and pneumatic devices. Table I shows a grouping of
emissions into the emission source categories.
The "emission time dependency" characterization groups all emissions by the steadiness of the emission
rate. Some emissions are defined as continuous and "steady," and as a result are easily measured. Others
-------
93-WP-104.03
are defined as intermittent and "unsteady" and are, therefore, difficult to measure. "Steady" is a relative term
and is defined by the time period considered for measurement. An emission source whose rate, when
measured over a short time period, is representative of its yearly average emission rate is considered a steady
source.
The "emission path* characterization groups emissions by common paths or means for reaching the
atmosphere. There are three categories under this grouping: vented, combusted, and leaked. They are
defined as:
* Vented Emissions. Vented emissions are intentional releases that reach the atmosphere by design
or operational practice. Vented emissions include vessel blowdowns (where the operator
intentionally depressures the vessel to the atmosphere) and pneumatic device emissions, PRV lifts,
and dehydrator vent emissions (where the basic engineering design intends that there is methane
released). Most of the vented sources are intermittent and are classified as unsteady emissions.
Combusted Emissions. Combusted emissions include compressor exhaust, burner, and flare
emissions. These emissions are uncombusted methane that escapes due to the inefficiency of a
combustion process.
* Leaked Emissions. Leaked emissions refer to unintentional emissions from sealed surfaces such as
packings and gaskets, as well as from pipes (resulting from corrosion, etc.). These emissions are
unintentional releases, and most are steady. Leaks are also called fugitive emissions.
There are five possible operating modes associated with any equipment or emission source: (1) startup
activities; (2) normal operations; (3) maintenance; (4) upsets; and (5) mishaps. Since emission rates and
types can be completely different under each mode, these categories were defined as specific causes/modes
for an emission.
Table II shows emission operating modes, emission path type, and the emission time dependency
associated with specific emission source examples. Once the emissions are characterized using the terms
defined above, an approach to estimating each emission source is developed. The total emissions for each
source type are then determined by summing the emissions created from that source in each operating mode.
For some source types it may be necessary to subdivide the task, adding the emissions from all components
that comprise the source type to get the total. Then, all source types emissions in a segment are added
together, and all segment totals are added to reach the national total.
The primary choices for estimating emissions are: (1) direct measurement, and (2) calculation. Since
measurements are only performed on steady sources, the emission type was the primary means for selecting
the estimation method. An array of methods were needed for measuring the steady emissions and
calculating the unsteady emissions from all sources in the industry. Some methods were applicable to entire
segments of the industry, while others were only applicable to some source types. In a number of cases
some sources could only be evaluated by summing the emissions from their basic components.
Phase 3
Phase 3 of the program consists of four major activities: (1) measuring emissions from steady sources;
(2) collecting the information needed to calculate emissions from unsteady sources; (3) gathering data for
extrapolating a source estimation to national totals for that source; and (4) calculating national totals and
evaluating the uncertainty of the national emissions estimate.
Emissions determined for a set of similar sources are extrapolated to national totals by using "emission
factors* and "activity factors." These factors are defined in such a way that their product will be equal to the
total emissions from a given source type or category.
-------
93-WP-104.03
In the simplest case, the emission factor would be defined as the average emission rate calculated by
adding the steady and unsteady emissions for a large number of randomly selected sources of a given type.
The activity factor would then be the number of sources within the United States. For example, total U.S.
emissions from pressure regulating and metering stations could be calculated by multiplying the emission
factor, which would be the average emission rate (cubic feet/station-year), by the activity factor, which would
be the total number of stations.
The methods for determining these two factors may be completely independent. The emission factor
may be determined by direct measurement, while the activity factor is calculated from site visit data. The
options for emission estimation and activity factor estimation will be discussed in the following sections.
Phase 3 Emission Measurements. The first task was to apportion available resources to focus on
sources with the largest emissions and highest uncertainty in the emissions estimate. Each source defined in
Phase 2 was examined to determine if the accuracy target has been achieved. If not, the number of
additional data points needed to achieve the accuracy target was calculated from the standard deviation for a
90 percent confidence level. If there is more than one method of obtaining emissions data, a cost
optimization program was used to select the most cost effective approach.
The most cost effective approach typically depends on the accuracy of the method, the cost per data
point, and the number of data points needed. Because the standard deviation of the emissions data will
change as additional data are collected, optimizing the Phase 3 effort is an iterative procedure.
Due to the diversity of source types and emission types, it was necessary to develop a menu of methods
for the many different types of sources. Some of the methods were applicable to entire facilities (such as gas
plants), some to source types (such as wellheads), and some to components (such as valves and flanges).
The project established the following approach for measurement of steady, fugitive emissions in
Phase 3:
Packed or sealed surfaces (on valves, flanges, screwed fittings, pump and compressor seals). A
component count approach will be used for all industry segments. Multiple site visits will produce
average component counts, and existing measured component emission rate data [from recent
American Petroleum Institute (API) emission factor data] will be used to establish the total
emission rates for each site. The national fugitive component count for a segment will be
determined by establishing average component counts per facility type, and multiplying the average
by the population of faculties in that segment.
Underground pipeline leakage in the distribution segment. A leak measurement technique will be
used for the distribution segment only. (Transmission pipeline leakage is considered unsteady.)
Pipeline Leak Measurement Programs use a flow measurement device to quantify leakage from an
isolated segment of pipe containing single or multiple leaks. The isolated pipe segment is
maintained at its normal operating pressure using the gas supply routed through the flow
measurement device. The leakage rate from the isolated segment is equivalent to the flow rate
required to maintain normal operating pressure. Leakage for an entire distribution network can
be determined using the measured leakage rates combined with the leak frequency. This program
has been developed as a cooperative effort with gas distribution company participation.
Meter and pressure regulator station leakage. A tracer measurement program has been used to
quantify the steady emissions from meter and pressure regulating stations in the distribution and
transmission segments. Tracers are downwind measurement techniques that release a known rate
of tracer gas to the atmosphere next to the emission source. Assuming identical dispersion, the
ratio of the downwind concentrations is equal to the ratio of release rates. The downwind
concentrations of methane and tracer gas are measured, and the emission rate of CfL, is then
determined.
-------
93-WP-104.03
The material balance approach. This method is being tested for the transmission segment. It will
determine all emissions (primarily fugitive emissions). A material balance method determines the
amount lost to the air by accounting for all other dispositions of gas. A detailed balance around
the system is completed based on all known incoming and outgoing streams. This approach can
be applied to systems that are relatively constant in their operation, have few changes in state, and
have an extensive flow measuring network. It can be applied to some sections of gas plants, and
to transmission and distribution systems. The current test is based on one transmission company's
system and meter data.
The planned tests for Phase 3 are shown in Table III. Table III presents the number of tests conducted
to date for each source category and the number of additional data points required to reach the accuracy
target as a function of the test method.
Phase 3 Emissions Calculations. Emissions are calculated for unsteady sources that cannot be easily
measured. The calculation must be based upon a detailed understanding of what parameters contribute to
the emission and upon the ability to quantify these important parameters. Determination of these important
characteristics is the first major goal of the calculation approach; this goal has been accomplished by studying
the source and producing a detailed characterization for each major source.
Once the important characteristics are defined, a plan is set to gather data to establish national
averages for those characteristics. The data are usually gathered by one of the following methods:
Gathering aggregate data from published statistics, reports, or surveys;
Gathering data already compiled by individual companies; or
Visiting multiple company sites and generating the data from observation.
A few examples of techniques for large unsteady sources follow.
Pneumatic Emissions. A separate characterization study has been completed for pneumatic
devices. This study examined technical and manufacturer data to determine the important
emission-affecting characteristics of pneumatic devices. Statistical data on the characteristics will
be gathered from various companies during Phase 3 site visits.
Glvcol Dehvdrator Vents. A separate characterization study using a computer model identified the
important emission-affecting parameters for dehydrators. Data on these parameters will be
gathered during the Phase 3 site visits and fed back into the computer model. The model's output
will estimate average vent emissions from dehydrators.
Maintenance Emissions. A separate characterization study is being conducted for maintenance
emissions. This study outlines all of the important maintenance activities in each industry
segment, so that data on these activities can be collected during the Phase 3 site visits. For
example, on compressor blowdown emissions, data will be gathered at each site on:
the frequency of compressor blowdowns;
the volume blowdown per event; and
the disposition of the blowdown gas.
The data from multiple visits will then be compiled, allowing a final national emission estimate to
be produced.
Compressor Exhaust Emissions. A separate study was performed on compressor exhaust to
characterize national emissions. The study made use of several existing databases that included
some direct measurements of stack emissions. The study extrapolated average compressor
emissions to overall industry emissions by using the total industry horsepower and total fuel use.
-------
93-WP-104.03
Phase 3 Activity Factors. Activity factors (national populations, counts, or activity frequencies) for each
source are generally not available in industry literature sources. Therefore, activity factors will be generated
using the same general approach described in the Phase 3 Emissions Calculations section. For the common
equipment populations, data will be gathered from reports and surveys, from individual companies, and
aggregated from multiple site visits to estimate national populations.
For some common facility and equipment counts (such as the number of gas plants, the number of
storage facilities, the miles of transmission pipeline, and the total number of active wells), definitive data are
available on national populations. Most categories, such as the count of dehydrators, have to be extrapolated
from existing (but not comprehensive) industry surveys and reports. Finally, some equipment has no
aggregate data available, and must be established from counts at multiple sites visited during Phase 3.
Phase 3 Validation. The basic strategy for validating the estimate and method has been;
« Conduct proof-of-concept (POO and Field demonstration tests. These tests will confirm the
concept and validate the accuracy of the selected steady emission measurement techniques.
* Establish a comprehensive industry and emissions characterization. This has been discussed in
detail in the previous sections.
« Seek continuous peer review. Through presentations like this one, the project has sought the gas
industry's response to the selected methods. The Gas Research Institute (GRI) also reviews
project plans with a project advisory committee composed of sponsors and industry
representatives. In addition, some interim project reports are circulated for a wider industry
review.
Perform comprehensive site measurement tracer tests. The tests will verify some of the various
steady and unsteady emission estimation methods. Tracer measurements will be made for a few
entire sites, such as gas plants, transmission compressor stations, and separation facilities. This
site tracer measurement naturally will include all steady and unsteady emissions at the time of the
test. The other methods for estimating steady and unsteady emissions will be used to calculate
total emissions at the site, and the summary of these methods will be compared with the tracer
results.
CONCLUSIONS
This report has presented a synopsis of the methods being used to estimate atmospheric CH^ emissions
that result from operations in the natural gas industry. The basic method is to recognize, delineate,
characterize, and understand each emission source so that no source is overlooked, and so that the major
sources can be appropriately estimated. The study uses many techniques, employing many measurement and
diverse-calculation-based techniques.
The basic strategy for validating the approach has been to establish a comprehensive characterization,
to seek industry review, and to perform some comprehensive tests. The results of the final estimate should
be available in late 1993 or early 1994.
As a follow-up to this study, mitigation techniques can be identified that will reduce emissions from
each major source. The economic impact of the reduction (cost versus benefit to the gas company) can then
be used as a criterion for selecting appropriate areas for application.
REFERENCES
1. Climate Change. Intergovernmental Panel on Climate Change (IPCC) WMO/UNEP, New York, 1990.
2. The Impact of Synthetic Fuels on Carbon Dioxide Emissions, Clark, W.C. ed., Carbon Dioxide Review
1982, New York, pp 406-410.
-------
93-WP-104.03
Table I, Emission Source Categories (within each segment).
Segment
Emission Source Types
(Facilities / Equipment)
Equipment Components and Practices (that
cause emissions)
Production Wells, Surface Equipment,
Gathering Facilities
Processing Gas Plants
Transmission Transmission Pipelines, Meter
and Pressure Regulating Stations,
and Compressor Stations
Storage Underground Injection /
Withdrawal Facilities, Liquefied
Natural Gas (LNG) Facilities
Distribution Mains (pipelines), Services
(pipelines), Meter and Pressure
Regulating Stations
Valve Stems, Flanges, Compressor Seals,
Pneumatic Devices, Glycol Dehydrator Vents,
Chemical Injection Pumps, Maintenance
Slowdowns, Flaring, Compressor Exhaust, Line
Leaks, Pressure Relief Valve (PRV) Lifts, Etc.
Valve Stems, Flanges, Compressor Seals, Glycol
Dehydrator Vents, Maintenance Slowdowns,
Flaring, Compressor Exhaust, Line Leaks, PRV
Lifts, Etc.
Valve Stems, Flanges, Compressor Seals,
Pneumatic Devices, Maintenance and
Emergency Slowdowns, Flaring, Compressor
Exhaust, Line Leaks, Etc.
Valve Stems, Flanges, Compressor Seals,
Pneumatic Devices, Glycol Dehydrator Vents,
Chemical Injection Pumps, Maintenance
Slowdowns, Flaring, Compressor Exhaust, Line
Leaks, PRV Lifts, Etc.
Valve Stems, Flanges, Pneumatic Devices,
Maintenance Slowdowns, Line Leaks, Etc.
-------
93-WP-104.03
Table II. Emission Characterizations.
Emission
Path Type
Fugitive
Combusted
Vented
Specific Emission Source
Examples
Packed or Sealed Surfaces
Leaks (holes in pipes)
Compressor Exhaust
Flaring
Flaring
Burners
Dehydrator Vents
Pipeline Purge
Pneumatic Devices
Compressor Starts
Equipment Slowdown
Equipment Slowdown
Chemical Inj. Pump Vents
PRVLift
Pipeline Dig Ins
Operating Mode
Category
Leaks
Leaks
Normal Operations
Upsets
Normal Operations
Normal Operations
Normal Operations
Maintenance
Normal Operations
Normal Operations
Maintenance
Upsets
Normal Operations
Upsets
Mishaps
Emission Time Dependency
Steady
Steady (for distribution)
Unsteady (for transmission)
Steady
Unsteady
Unsteady
Steady
Steady & Unsteady
Unsteady
Steady & Unsteady
Unsteady
Unsteady
Unsteady
Steady & Unsteady
Unsteady
Unsteady
Table III. Data and Tests.
Current Estimates Total Sites to be Visited for Sites to be Directly
Segment (BCF/yr) Characterization Measured During the Visit
Production
Processing
Transmission
Storage
Distribution
75
21
86
1
16
23
6
18
L5
1
1
70
200
Fields
Plants
Compressor Stations
Meter & Pres. Reg. Stations
Injection/Withdrawal
LNG
Meter & Pres. Reg. Stations
Underground Mains & Services
6
6
4
15
0
0
70
200
-------
Page Intentionally Blank
------- |