P                    United Stales                            '  GRI-94 / 0257,28
P       J^t^S®       Environmental Protection-                        EPA-600/R-96-080k
                     Agency                                 June 1996
       HEPA     Research  arid
                     Development
P       ^^'•••i: ff-l     • ;11^€9^OI %*l I OIIO             PB97-143028
P
P
P
P
P
P
P

P                    METHANE EMISSIONS FROM THE

                     NATURAL GAS INDUSTRY

                     Volume 11: Compressor Driver Exhaust
                     Prepared lor
                     Energy Information Administration (U. S. DOE)
                     Prepared by

                     National Risk Management
                     Research Laboratory
                     Research Triangle Park, MC 27711
                                 REPRODUCED BY:
                               US. DSpstmtiiKifCoinmorce
                               Btlonal T«eh£t^al Infanntriion Servka
                                Splir,f,i>«ii VSjInla JJ161

-------
4, TITLE AWQ SUBTjTiE
Methane Emissions from the Natural Gaa Industry,
 Volumes 1-15 (Volume 11: Compressor Driver Exhaust
                                TECHNICAL REPORT DAT*
                         {Pltme read hutai&tam on the revase before eoatph
 RCPOB-T NO.
EPA-100/R-96-080IC
            n, REPORT OATE
             June 1996
            a. PERFORMING ORGANIZATION CODE
                   PB.97-143028
 . AUTMOBISI
           , Campbell, M. Campbell, M. Cowgill, D. Ep-
3erson» M. Hall, M. Harrison, K. Hummel, D. Myers,
T. Shires, B. Stapper, C. Stapper, J. Wessels,  and *
            B. PGBFOnMINO ORGANIZATION flEPQBT NO.

             DCN 96-263-081-17
B, PERFORMING ORGANIZATION NAME AND ADDRESS
Radian International L/LC
P. O.  Box 201088
Austin, Texas 78720-1088
                                                      10. PROGRAM ELEMENT NO.
            11. CONTHACWBHANT NO,
            5091-251-2171 (GR!)
            68-Dl-0031 (EPA)
12. SPONSORING AGENCV NAME AND AOOBESS
EPA, Office of Research and Development
.Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
            13.TVP6 OF HfPOHT AND PERIOD COWEBED
                   3/01-4/96
            14. SPOMSOBIS6 A61NCY COOS
             EPA/6Q0/13
«.SUPPLEMENTARY MOTES fipA project        is D. A. KIretigessixer, MTM83,919/ 541-4011.
 oaponeor GRI project officer is E, A. Lott,  Gas Hesearch Institute, 8600 West Bryn
Mawr Jve., Chicago, IL 60631.  (*)H, Williamson {Block 7), 	   	
              IS'Volume report summarizes the results of a comprehensive program
to quantify methane (CH4) emissions from the TJ. 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
 missions within +/-0, 5% of natural gas production for a 90%  confidence interval. For
the 19S2 base year, total CH4 emissions for the t). S* natural gas industry was 314
+/- 105 Bscf (6.04 +/- 2.01 Tg). TMs is equivalent  to L4 +/- 0.5% of gross natural
    production, and reflects neither emissions reductions (per the voluntary Ameri-
Gas  Association/EP^ 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  ooal 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.
                             KEY WORDS AND DOCUMENT ANALYSIS

                                          PUDENWIERS/OPEN ENDED TERM*.
                QESGBSPTORS
                           COSATI ReU/6iaup
Pollution
Emission
Greenhouse Effect
Natural Gas
Gas Pipelines
Methane
Pollution Prevention
 Stationary Sources
 Global Warming
13B
14G
04A
21D
15E
07C
t8. DiSTWBUTION STATEMENT

 Release to Public
                                          18. SECURITY CLASS (This Repatif
                                                                   31. NO. OP PAGES-
                               81
TO, SECURItV CLASS fTMs page)
 Unclassified
                         22. PBIC6
EPA Form 2220-1 (9-73)

-------
                            FOREWORD
The 0. 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 oi the Laboratory's
research program Is on methods for the prevention and control of pollution to air,
land,  water,  and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwaterj 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 sad engineering Information needed by EPA to
support regulatory and  policy decisions; and provide technical support and infdr-
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  ia published and made available by EPA1 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 tie 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 puttie through the National Technical Information
     Service, Springfield, Virginia 22161
                                                    PROTECTED UNDEH Kl

-------
                                    EPA - 600 / R- 66- 0 80 k
                                    June 1996
        METHANE EMISSIONS FROM:
       THE NATURAL GAS INDUSTRY,
VOLUME 11:  COMPRESSOR DRIVER EXHAUST
              FINAL REPORT
                 Prepared by:

               Carole I. Stopper

            Radian International LLC
             8501 N. Mopac Blvd.
               P.O. Box 201088
            Austin, TX  78720-1088


             DCN:  95-263-081-07


                    For

       GRI Project Manager:  Robert A. Loft
         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.

-------
                            RESEARCH  SUMMARY
Tille             Methane Emissions from the Natural Gas Industry^
                  Volume 11:  Compressor Driver Exhaust
                  Finai Report

Contractor        Radian International LLC

                  GRI Contract Number 5091-251-2171
                  EPA Contract Number 68-D1-0051
Principal
Investigator

Report Period
Objective
Teclinical
Perspective
Results
Carole J, Stapper
March 1991 -June  1996
Final Report

This report describes a study to quantify the annual methane emissions
from compressor driver exhaust, which is a significant source of methane
emissions within the gas industry.

The increased use of natural gas has been suggested  as a strategy far
reducing the potential for global warming. During combustion, natural
gas generates less carbon dioxide (CO2) per unit of energy produced than
either coal or oil. On the basis of the amount of CO2 emitted, the
potential for global  wanning 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 CO2 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.

The national annual emissions  for compressor drivers in each industry
segment are as follows; production, 6.58 ± 200% Bscf; gas processing,
6.84 ± 130% Bscf; transmission, 10.2 ± 17.1%  Bscf;  and storage, 1.19 ±
27.2% Bscf.
                                          ill

-------
                  Based  on data from the entire program, methane emissions from natural
                  gas operations are estimated to be 314 + 105 Bscf for the 1992  base
                  year.  This is about 1.4 ± 0.5% of gross natural gas production.   The
                  overall program also showed that  the percentage of methane  emitted for
                  an incremental  increase in natural  gas sales  would be significantly lower
                  than the  baseline case.

                  The program reached  its accuracy  goal  and provides an accurate  estimate
                  of methane emissions that can be  used to construct  U.S. methane
                  inventories and analyze fuel switching strategies.

Technical         The industry has two  primary types of compressor drivers that fire
Approach         natural gas:  1) reciprocating engines and 2) turbines.  Methane
                  emissions result from  the incomplete combustion of natural gas  in the
                  driver.,  which allows methane to exit the driver in the exhaust stream.

                  The techniques used to determine  methane  emissions  were developed to
                  be representative  of annual emissions from the natural gas industry.
                  However, it is impractical to measure every source continuously  for a
                  year.  Therefore,  annual emissions for compressor drivers were
                  determined by extrapolating  measured emissions using activity factors
                  where the national emissions estimate is the product  of the emission
                  factor and the  activity factor.

                  Emissions test data for each driver type were collected by Southwest
                  Research Institute (SwRI) for compressors in natural gas industry service.
                  SwRI data for emissions,  fuel use  rates, and compressor  model  numbers
                  were used with data in OKI's TRANSDAT compressor database  to
                  develop the emission factors. Equations relating the SwRI data  and  the
                  distribution of compressor models  and operating hours found in
                  TRANSDAT were developed to calculate emission factors for each type
                  of compressor  driver.

                  Activity factors for each Industry segment were developed using  site visit
                  data, company surveys and databases, and data published in the
                  American Gas Associatten's Gas Facts  and  the Oil &  Gas Journal.  The
                  national annual emissions for each industry segment  were then calculated
                  as the product  of the emission factor and activity factor for each
                  compressor driver.

Project            For the 1992 base year the annual  methane emissions estimate for the
Implications       U.S. natural  gas industry is 314 Bscf ±  105 Bscf (± 33%).  This is
                  equivalent to 1.4% ± 0.5% of gross natural gas production.  Results from
                  this program were used to compare greenhouse gas emissions from the
                  fuel cycle for natural gas, oil, and  coal using the global wanning
                                           IV

-------
potentials (GWPs) recent'y 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 the EPA. Since
this program was begun after the 1992 baseline year, any reductions in
methane emissions from this program are not reflected in this study's
total emissions.
Robert A, Lott
Senior Project Manager,  Environment and Safety

-------
                          TABLE OF CONTENTS

                                                                     Page

 1.0   SUMMARY  	,	, , . . ,	1

2.0	2

3.0   DATA SOURCES	   3

      3,1    Previous Studies	.,..,.,.	  3
      3.2                      	,		..........  4
      3.3    Current Study	  5

4.0                                                     .....	.....  7

5.0   COMPRESSOR DRIVER ACTIVITY FACTORS	  11

      5.1    Production Industry	....,.,..,.... 12
      5.2	  14
      5.3    Transmission Industry Segment	 17
            5.3.1 Compressor Station Horsepower and Hours	 19
            5.3.2 Underground                 and Hours	 22
            5.3.3 Generator Driver Horsepower and Hours	, 22
      5.4    Industry Activity Factors	27

6.0                                                        ,.,.,,,... 29

7,0   CONCLUSIONS	30

8.0                . .	.		.32

      APPENDIX A - National Estimate of Methane Emissions from Compressors
                   in the U.S. Natural Gas Industry	 A-l

               B -           P-l	 B-l

      APPENDIX C - Production Segment Site Visit Results	 C-l

      APPENDIX D - Transmission Compressor           Visit       ........ D-l

      APPENDIX E - Conversion Table			 E-l
                                   VI

-------
                                LIST OF FIGURES
A-l    Effect of Horsepower and Speed on Total Hydrocarbon Emissions for
       a Cooper Bessemer 2 Cycle Engine ...............................  A-6

A-2    Relative  Horsepower Contribution of Each Engine Model in the
       Emissions and Industry Databases ...................... , .......... A-14

A-3    Relative  Horsepower Contribution of Each Turbine Model  in the
       Emissions and Industry Databases .......... .. ............ .... ........  A- 15
                                        Vil

-------
                               LIST OF TABLES




                                                                          Page




4-1    SwRI Field Test Emission Factor Statistics . . . ,	.,,,.,,,,.,,  9




5-1    Compressor Driver Data for Production Segment	, . , ,	  12




5-2    Regional Data for Production Segment Extrapolation	,...,.,,..,...  13




5-3    Compressor Driver Data from Processing Segment Site Visits  ,,,,,....,.,, 15




5-4    Annual Operating Hours for Processing Segment	 ,	 IS




5-S    Annual Operating Hours for Storage Segment	 23




5-6    Generator Driver Data from Transmission Compressor Stations  , , ,	24




5-7    Generator Driver Data from Storage  Fields ...,,...,.,....,,..	26




5-8    Compressor Driver Activity  Factors for Each Industry Segment  . .	28




6-1    Compressor Driver Emissions for the Natural Gas Industry by Segment	29




7-1    Ranking of Compressor Driver Exhaust Emissions by Industry Segment  ...... 30




A-4    Description of the Databases  .,.,.........,.	A-8




C-l   Gathering Compressor Data  from Production Segment Site Visits .......... C-4




C-2   Regional Data for Production Segment Horsepower ExtrapoL"ion ,,.,,,,,.. C-5




C-3   Annual Operating Hours for Production Segment	 C-6




D-l   Compressor Driver Data from Transmission  Segment Site  Visits	D-4




D-2   Annual Operating Hours for Transmission Segment	.,...,,	 . D-6
                                      Vlli

-------
1.0           SUMMARY

              Thif report 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  (GR1-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,

              This report quantifies the amount of unburned methane released in
compressor driver exhaust in natural  gas production, gas processing, and transmission,
Emissions from generator driver exhaust, while minor, are also included in this report.
Emission  estimates for each industry segment were based on data from one or more of the
following sources: 1) site visits, 2) company databases;  and 3) published data.  The factors
that affect the quantity of methane emissions from compressor drivers are: type of driver,
horsepower, and operating hours.

              Compressor driver exhaust is a significant source of methane emissions.  It
accounts for 24.8 Bscf of methane emissions, which is about 7.9% of methane emissions
from the natural gas industry.

-------
2.0           INTRODUCTION

              In the natural gas industry there are two primary types of compressor drivers
that fire natural  gas:  ]) reciprocating engines, and 2) gas-fired turbines.  Methane
emissions result from the incomplete combustion of the natural gas, which  allows some of
the methane  in the fuel to exit in the exhaust stream.  Compressor driver exhaust emissions
represent a significant  source of methane emissions is all of the industry segments where
these sources are present. Emissions from generator driver exhaust, while minor, are also
included in this  report.

              Annual emissions were calculated as the product of the emission  factor and
the activity factor.  To develop the ;;pmpressor driver emission factors, test data from
Southwest Research Institute (SwRI) and CRTs  TRANSDAT compressor database1 were
used.  The activity  factors were developed  to characterize the compressor drivers in each
industry segment.  Data were gathered from site visits, company databases, and published;
data from the American Gas Association (A.G.A.) and: the Oil and Gas Journal.

              This report describes how the emissions from compressor driver exhaust  were
determined.   Section 3 discusses the data used to make the emission estimates.  Section 4
presents the development  of the emission factors for engines and turbines.  Section 5
describes the development of the activity factors for each industry segment (production,
processing, and  transmission, including  storage and generators).  The annual emissions  for
each segment and the overall national emissions estimate are provided in Section 6.
Conclusions are given  in Section 7, This report is one of several documents prepared for
the GRI/EPA methane  emissions project.

-------
 3,0           DATA SOURCES

              Compressor driver exhaust emissions had been identified in earlier studies as
 a significant source  of methane emissions.  However, these studies were based on limited
 data and did not calculate error bounds for the estimated emissions.  This study collected
 and evaluated a large number of data for compressor  drivers and performed an error
 analysis of the emissions  estimate.  This section provides  background information on
 previous studies and an introduction to the data and the approach used for the current stud\

 3.1           Previous Studies

              Previous estimates of methane emissions from compressor driver exhaust
 have ranged  from  about  114 Bscf (2.2 Tg/yr) to 11.5  Bscf (0.22 Tg/yr)."  The 114 Bscf
 estimate, from Pipeline Systems, Inc. in 1990,  was based on an estimated emission factor
 established by the South Coast Air Quality Management District (1120 Ib/MMscf fuel).3
 This estimate used a "model" installation  to describe typical compressor  facilities for field
 production, transmission, storage, and  gas processing; then used these "mode!"  installations
 to extrapolate to a national estimate.  The "model" installations were based on site visits to
 three production facilities, three pipeline systems,  five injection/withdrawal plants, and two
 gas plants.   Eight  of the site visits were in California, tv/o  were in Texas, and the remaining
 site visits were in  the Central Plains Region. The estimate from Pipeline Systems, Inc. was
 based on a limited number of data and are probably biased due to the disproportionate use
 of California sites.

              The 11,5 Bscf estimate was a preliminary estimate developed in  1992 as part
 of the GRJ/EPA methane emissions project.3 (The preliminary  emissions estimate is
described in detail in the paper titled "National Estimate of Methane Emissions from
Compressors in  the U.S.  Natural Gas Industry," which can be found in Appendix A.) This
estimate was based on data  published by A.G.A. and data contained in the  GRI
TRANSDAT compressor database.  The GRI TRANSDAT database includes data from

-------
A.G.A.  and SwRI.  A.G.A. data are         from government agencies, such as U.S.
Department of Energy (DOE) and Federal Energy Regulatory Commission  (FERC), and
from surveys of its member companies in transmission and distribution.  SwRI data were
          during a field testing program on natural gas compressor driver emissions and


             This preliminary               that the GRI TRANSDAT         included
information on all compressors used in the natural  gas industry when it only contained data
collected from transmission companies.  Therefore, this study underestimated  methane
emissions from compressor driver exhaust for the natural gas industry,

3,2

             The GRI TRANSDAT          is actually composed of three data subsets:
Industry Database, Operating Database, and Test Database,  The Industry Database lists
8282 compressor drivers (engines and turbines) used in the gas industry in  1989,
Horsepower in this database is given by compressor model number for each gas company
and accounts  for a     of 16.2 million horsepower (MMhp), installed,  TRANSDAT
were collected                                It     include a limited        of
on                  for production and            The
                     and                    for 1515             (3.2 MMhp, total).
The Test Database contains emissions data from field      collected by SwRI  and includes
data on  methane emissions, fuel use, fuel use rate,  and horsepower  for 241  compressors,

             The Operating     Test                combined to         a fourth
         the                                      to  develop                   factors,
This                                   number,                   fuel use rate, and
                     for 775                    and 86                   Therefore,
the emissions can be calculated for each of the 775 engines and 86 turbines as they were
operated during the year using SwRI emissions data.

-------
 3.3           Current Study

              Data were gathered during this study to supplement the GRI TRANSDAT
 database to account for emissions from all industry segments.  Data were coL^ved during
 site  visits and company surveys to determine the number of compressors and their
 horsepower and operating hours.  Because it is not practical to visit every site or test every
 compressor driver, a method was developed using emission factors and  activity factors to
 calculate total emissions from compressor drivers based on a limited data set.  The emission
 and  activity factors wer  defined such that their product would equal the total emissions
 from compressor driver exhaust for each segment of the industry.  Emissions from
 reciprocating engines and turbines were evaluated separately.

              For most source  categories in the gas industry, the emission factor is defined
 as the average emission rate determined from a large  number of randomly selected  sources.
 The  activity factor is then the total number of sources within the source category, such as
 the total count of compressor drivers.  However, for compressor  driver exhaust, the
 emission factor was evaluated in terms of emissions per horsepower-hour (scf/hp-hr) since
 this  has less variability than an emission factor per compressor unit.  By calculating
 emissions on a per horsepower basis the variability  in emissions  due to compressor size can
 be eliminated. As a result, the activity factor is horsepower 'hours per year.  This means
 that:

                   Total Emissions (scf) = EF (scf/hp-hr) x AF (hp-hr)                 (1)

             The site visit  and company survey data provided information on activity
factors for the various  segments of the industry, and SwRI test data were  used to develop
the emission factors.  However, the emission factor was not based on a  simple average of
the SwRI test data because the emission factor  needed to be weighted to reflect the
compressor population of the industry  and because some compressor models are used as
baseloaded compressors s»nd operate  a higher percentage of the time. For these reasons,

-------
A.G.A.'s Operating Database, which; contains horsepower, operating  hoyrs, and model
numbers for over  1500 compressors,  was used to develop weighted industry emission
factors for reciprocating engines and  .turbines.  The emission factors for engines and
turbines are 0.240 sef/hp-hr ± 5% and 0.0057 scf/hp-hr ± 30%, respectively. The method
for calculating  these weighted emission factors is described in the following section.

-------
4.0

              The emission factors for reciprocating engines and gas-fired turbines were
determined from compressor test data collected by SwRI and weighted to reflect both the
population of compressor        and their operating hours in the natural  gas industry.  The
                   to develop the'          procedure'      available in 'GUI's
TRANSDAT

              SwRI followed several procedures to       the quality of their compressor
data,  First, the emissions analyzers  were calibrated using gases certified by the National
Institute of Standards and Testing.  Second, an oxygen molar balance was performed
         the          and the products of combustion to verify the        Third,      of
the                          graphically  in an effort to                  outliers.  There
     generally five            for          and any of the points     did not follow the
trend              as outliers. Data         collected by                and provided to
SwRI for inclusion in their study were also scrutinized graphically to validate the results.
Where technically justified,  statistical outliers were discarded in the development of the
emission factors for the GRI/EPA methane emissions project.

              An                                    for      model of
                  driver in the                         the test     collected by SwRI
for     model.  Using the feel use    and                           for
compressor model, the emission rate was  calculated as follows:

              ER(m)    =EPwxFURW)                                            (2)

where:        ER tmj    =                      for model, m (sef/hr)
              EP (m)    =                            for model, m (scf CH4/scf
              FUR 
-------
              The methane emissions for each compressor driver in the Emissions Database
 were calculated by multiplying the emission rate (Equation 2) by the annual operating hours
 for each engine or turbine.  The total emissions for each model of compressor driver in  the

 database were  calculated  as follows:
                              M
                                        xHR,)                                      (3)
where:        E (m)    = total emissions for model, m(scf)
              HR |    = annual operating lours for compressor i, (lir/yr)
              M      = number of compressors of model, m
              The total emissions for all compressor drivers in the database were the sum

of the total emissions calculated for each model.
                  TE-E   E(m)                                                     (4)
                        m=l

where:        TE     = total emissions for database, (scf)
              K      — number of unique compressor models
              These emissions were calculated based on data from 775 engines and 86

turbines and weighted to reflect the distribution of compressor models found in the gas

industry as well as the percentage of time these models are operated. Emissions  can also

be affected by the compressor load.  However, the emissions are generally linear over a

fairly large range of operating horsepower.  To ensure that significant differences did not

exist between  the horsepower at which the compressor drivers are typically operated and

the horsepower at which they were tested, the following comparison was made;


                   N         ?      N
                  ^ "P (test) i   "      L WP (operating)  i

-------
The two              to within
             The emission factors, for engines and turbines, were then calculated using the
following
                               N	     N
                      = TE /  [I HP , x (£ HR , / N)]
                               i=l      j=f
                                                         (5)
where:       HP     = average operating horsepower during HR, (hp)
             HR     =                      (hr/yr)
             N      = number of compressors
The ¥alues obtained are:
               Engines;
              Turbines:
                                1.165 x  109 scf
                         (1.3 x  106 hp) x
      0.005 x  109 scf
(0.4  x 10* hp) x
                             = 0.240
                = 0,0057 sef/hr-hr
      Confidence limits were calculated for the emission factors based on the variation in
SwRFs      test       Table 4-1       the              to           the 90%
limits for each of the emission factors.  The resulting uncertainty is ± 5% for reciprocating
       and ± 30% for gas-fired               a 90%           limit).
        TABLE 4-1, SwRI FIELD TEST EMISSION FACTOR STATISTICS
        Engines
        TorMnes
 902
 105
229
 12
 0.0307
0.000107
 0.0279
0,00018?

-------
       Appendix B contains the source sheet for compressor driver exhaust emissions, which
samtmrlzes the approach taken to calculate the emission factors.  The source sheet also
summarizes the activity factors used to determine the overall national emissions estimate
from compressor drivers.  The next section describes the development of these activity
factors.
                                          10

-------
 S.O           COMPRESSOR DRIVER ACTIVITY FACTORS

              Information from site visits, company Surveys, A.G.A.'s-'Gas Fii-~ts,  and the
 Oil & Gas Journal was used to develop the activity factors.  Data on horsepower and
 operating hours were obtained from site visits and company databases far compressor drivers
located at 297 field gathering stations, 18 gas processing plants, 51 transmission compressor
 stations, and 11 underground storage stations. Data for electric generator drivers at 41
 transmission compressor stations and 3 underground storage fields were also obtained from
 the site visits and company databases.

              Compressors used fat gas lift (oil production) were excluded from the
 following analysis, since this gas is wot marketed and is- not considered part of the natural gas
industry,  Some sites used electrically driven  compressors, These electrically driven
compressors were also excluded in the development of the activity factors, since they have
no exhaust and therefore have no methane emissions.

              For the production industry  segment, horsepower-hour data were provided by
one company for  516 engines located at 244 gathering stations.  Production data for 53
gathering stations (from site visits and company surveys) were analyzed separately and are
presented in Appendix C for comparison.  For each of die remaining industry segmesrts,
separate horsepower and operating hour estimates were made, which were multiplied together
to get the horsepower-hour activity factors.

              Data were gathered fer engines and turbines, allowing separate activity factors
to be developed for the two driver types.  Date were also gathered for installed veisus active
horsepower in each industry segment (based on operating hours). Finally, data were
gathered that coultt be used to estimate nations! tmissions from each of the natural gas
industry segments (e.g, site data for gas throughput).
                                          11

-------
              The national estimates were extrapolated using 1992 data as the base year,
except for the transmHion segment estimate of compressor station horsepower that uses
1989 data as the base year,  A more detailed explanation of this variation appears in Section
5,3,1. The following sections describe the development of the horsepower-hour estimate for
production and the horsepower and annual operating hour estimates for processing and
transmission (including storage and generators).  Section 5,4 summarizes the resulting
activity factors for all of the industry segments.

5.1           Production Industry Segment

              Table 5-1 shows the production segment horsepower-hour data provided by
one company for 516 reciprocating engine compressor drivers.  These data included
compressor units with operating hours  listed as 8760 hr/yr.  It was assumed that these were
maximum operating hours, not actual operating hours. Since no other data were available
for these units, extensive quality checks were not possible, so the data were used as
provided. Turbine drivers were excluded from this analysis due to the small number of these
drivers in production (only one reported in the company database used for this analysis) and
the low methane emission factor for turbines,

    TABLE 5-1.  COMPRESSOR DRIVER DATA FOR PRODUCTION SEGMENT
Business Unit
Gulf Coast
BUI
BU2
BU3
Central Plains
BU4
BUS
TOTAL
• No. •Engines
15
\'J.
177
68
244
516
MMbpte
121.0
28.8
755.9
233.1
104,4
1243.2

478.6
89.8
396.6
731.6
624.0
2320.6
                                         12

-------
             The company data were provided by business unit and then grouped into two
geographic areas corresponding to the Gulf Coast (onshore and offshore) and Central Plains
regions described in Volume 5 on activity factors,4 There were no production data available
for fbe Pacific Mountain and Atlantic & Great Lakes  regions.  However, the two regions
without data account for only 9,5% of tile total U.S.  marketed natural gas production.  Wfaiie
this company database was not a perfect microcosm of the production segment, it was the
most comprehensive database available and should be a reasonable representation of the
production segment.

             The total company horsepower-hours per production (hp-hr/MMcfd) and the
total U.S. marketed natural gas production for 1992 were used to extrapolate to a national
estimate of compressor engine driver horsepower-hours.  Table 5-2 shows fee company
production data  and the U.S. production data usid for this.-calculation.3 The resulting
activity factor is 27,4€Q MMhp-hr.

           Engine hp- hr  = Compalty MMfr>te x  U.S. Marketed  Gas MMcfd       (6)
                            Company MMcfd
                            1243,2 MMhpfcr   «,«« *«, «
                            	:	-— x 51265 MMcfd
                             2320.6 MMcfd
                         = 27,460 MMhpir
TABLE 5-2. REGIONAL DATA FOR PRODUCTION SEGMENT EXTRAPOLATION
                               Company Production .  '•„  National Marketed Production
        Region   '  :lyy              ' -MMca      "- .              '
   .Pacific Mountain                      NA                         2696
   Gulf Coast                         965,0                        3.1.544
   Central Plains                       1355.6                       14860
   Atlantic & Great Lakes _ NA _ 2166
   TOTAL                            2320.6                       5J265
                                         13

-------
              Confidence limits were assigned to be ± 200% based upon an engineering
analysis.  Rigorous propagation of the error using the ratio method produced an estimate for
the upper bound of ± 576% (185,630 MMhp-hr),  The ratio method assumes that sites are
randomly sampled, so that a large site is proportionately representative of a large section of
the population (see Volume 4 on statistical methodology).6 However, an analysis of this
bound showed it to be unreasonably high.  A.G.A.'s Gas Facts reports that lease and plant
Mel usage for 1992 was 1.2 trillion sef.7 Assuming all of this fuel was fired in compressor
drivers, this would produce 150,080 MMhp-hr.  In reality, much of this fuel is actually used
in heater burners;  therefore, the actual upper bound must be well below 150,080 MMhp-hr
(± 447%).  A separate analysis of the company database produced an upper bound  of only
66,490 MMhp-hr (± 119%). This estimate was based on average horsepower-hours by
business unit and applied a confidence limit analysis to the average of these data.  Each
alternative approach resulted in lower confidence limits ton the calculated ± 576%;
therefore, the confidence limits were set at ± 200%.

5.2           Processing Industry Segment

              Horsepower

             Horsepower in gas processing could be extrapolated to estimate a national total
by using either plant throughput or plant count.  Horsepower is technically related to plant
pressure (AP) and  throughput.   However, there are no national figures for average plant AP»;
Without the AP factor, plant horsepower is not related to throughput alone.  Therefore,
national gas plant horsepower was determined by averaging the values obtained when
extrapolating by throughput and by plant count.

             Table 5-3 shows the installed horsepower and gas throughput for each of the
10 sites visited in the processing segment (gas plants).  The engine/turbine horsepower split
for this segment is 44.7% engines and55.3% turbines.  The first extrapolation method uses
the average site horsepower per plant throughput and the national gas pltnt throughput for

                                          14

-------
1992 to calculate the processing segment horsepower.8  This gives a total compressor driver
horsepower of about 2,2 MMhp for engines and 2,8 MMhp for turbines in the processing
segment.
               TABLE 5-3. COMPRESSOR DRWBR DATA PROM
                    PROCESSING SEGMENT SITE VISITS
Installed Engine Ms-tailed Turbine
Site ' . • hp • ' . " ' ' hp •
GP1
GP2b
GPS
GP4
GP5
GP6
GP7
GP8
GP9
GP10
TOTAL1
AVERAGE
* Total site horsepower (including gas
b Estimated,
' Excludes gas lift.
THROUGHPUT METHOD:
Engine hp-hp
8300
0;
0
3700
11000
6740(20000)
5925 {17490}
6267(18500)
6267(18500)
36600
84799
8480
lift for oil recovery) is shown


Total Site hp
Total Site MMcfd
35000
27000
20000
0
0
0
0
0
0
23000
105000
10500
in parentheses.


U.S. Throughput
WMrfd
MMcfd
350
750
140
60
49
56
40
130
130
70
1775
177.5



(7)
                         1775 MMcfd

                       = 2,22 MMhp
                                     15

-------
                     Turbine fap =               x 46510.7 MMcfd
                                   1775 MMcfd

                                = 2,75 MMhp
              The second extrapolation method uses the average installed horsepower per site
and the total number of gas plants in: the United States in 1992 to calculate the processing
segment horsepower.7  This gives a total compressor driver horsepower of about 6,2 MMhp
for engines and 7.6 MMhp for turbines in the processing segment.
SITE METHOD:
                Engine hp-hp = 	Total Site hp	 x o g Qas
                               Total Number of Sites
                                   .            x 726 Plants
                                      10 Plants

                                   = 6.16 MMhp

                        Turbtae hp =             * 726 Pknte
                                  -     10 Plants

                                   = 7.62 MMhp

             An average of the results of the two extrapolation methods is used to
determine the national estimate for compressor driver horsepower in the processing segment.
The average engine horsepower is 4,19 MMhp, and the average turbine tiorsepower is 5.19
MMhp.  Both estimates are based  on the total installed horsepower from the sites visited,
excluding horsepower used  for gas lift.

             Confidence limits were calculated for the average site vistt ftp/MMcfd and
average site visit Sip/plant using the ratio method. Confidence limits for the U.S. gas plant
                                          16

-------
throughput and the total U.S. number of gas plants were estimated to be: ± 5% and ±2%,
respectively.4 The uncertainty for the first extrapolation method (based on throughput) is
+ 114% for engines and ±120% for turbines.  The uncertainty for the second extrapolation
method (based on number of plants) is ± 71.4% for engines and ± 7.6% for turbines.   For
fiie average of the two methods, the uncertainty is ± 132% for engines and ± 99.4% for
turbines (90% confidence limits).

              Operating Hours

              Table 5-4 shows the "operating hours" for the compressor drivers at the 10
sites visited plus operating hour data from two company databases for an additional 18 gas
plants.  For the site visit data, actual operating hours were only available for two of the  sites.
For the remaining eight sites, compressors that were running during the site visit were
assigned annual operating hours of 8760.  Compressors that were idle were assigned annual
operating hours of 0.  One company database (17 plants) included operating hours of 8760
for some units. These data may be maximum hours, not actual hours.   However,  no other
data were available for these units, so the data were used as provided.   The operating hours
listed for each site in Table 5-4 are the average hours for all of the drivers at the site.
Confidence limits were calculated for the operating hoars estimate from the variation of the
site data.  The resulting uncertainty is ± 11.5% for engines and x 48.4%  for turbines (90%
confidence limits),

S,3           Transmission Industry Segment

              The activity factor estimates for the transmission segment include horsepower
and operating hours for compressor stations and storage fields. These estimates have been
calculated for compressor drivers and generator drivers at both types of facilities.  Section
5.3,1 discusses the compressor driver activity factor estimates for compressor stations.  (Site
visit data for transmission compressor stations are included in Appendix D  for comparison.)
                                          17

-------
   TABLE 5-4.  ANNUAL OPERATING HOURS FOR PROCESSING SEGMENT
Site
GPjb
GP2
GP3b
QW
GP5b
GP6b
GP7b
GP8b
GP9b
GP10
GP11
GP12
GP13
GP14
GP15
GPI6
GP17
GPIE
GH9
GP20
GP21
GP22
GP23
GP24
GP25
GP26
GP27
GP28
"TOTAL
AVERAGE*
90% LIMIT*
No, of Engines*
7
0
0
4
7
6(16)
4(12)
1 (4)
1(4)
15
2
5
4
5
9
J
14
10
3
11
7
46
8
10
8
5
1
9
203
•Average
Operating .Hours
6257
mm.
—
4380
5006
2920
6570
8760
8760
5724
0
7376
4617
8570
7949
8760
8760
4043
584Q 	
8760
5325
8760
8760
5011
6487
8176
8760
7949
6626
± 11,5%
No. of
Turbines 4
2
1
1
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
Average
Operating Hours
4380
8423
8760
__
__
mm
_„"
„
—
3816

__
__
mm
—
_-
,-
mm
„,
_
„„

„_
_,„
„_
—
__
__
6345
±48.4%
" Total number of drivers (including gas lift for oil
6 Based on waive (8760 fas) versus idle (0 his),
     1 on site averages.
recovery) are shown in parentheses.
                                      18

-------
Section 5.3.2 discusses the compressor driver activity factor estimates for underground
storage fields. Activity factor estimates for generator drivers are discussed in Section 5.3,3.

5,3.1        Compressor Station Horsepower and Hours

             Horsepower

             The transmission compressor station horsepower estimate is  based on the GRI
TRANSDAT database, which includes transmission, storage,  and associated field
compression for 1989, TRANSDAT has not been revised since the 1989 data were
collected; therefore, for these activity factors, 1989 was "~r.d as the base year. It is not
possible to separate the compressor data in GRI TRANSDAT by the specific application;
however, an adjustment for storage compressor horsepower can be r.nde based on data from
A.G.A. for 1989. While the GRI TRANSDAT data  could be adjusted to a 1992 basis using
A.G.A.'s transmission horsepower data, this approach was rejected because the annual data
showed no trend and differences were attributed to changes in reporting protocols by
A.G.A.'s member companies.

             GRI TRANSDAT's Industry Database  contains information on 7489
reciprocating engines and 793 gas turbines.  The total installed horsepower for these
compressors is 11.2 MMhp for engines and 5.0 MMhp for turbines, which accounts for
about 97% of the gas utility industry  installed horsepower reported by A...G.A. for 1989
£16.7 MMhp).7  A.G.A.'s data were gathered through a survey of their member companies
and includes approximately 96% of total gas industry sales.  The GRI TRANSDAT
horsepower can be adjusted for the 1989 storage horsepower reported  by A.G.A, (1,462,971
hp) to give 10.2 MMhp for engines and 4.5 MMhp for turbines.  GRI TRANSDAT'S
engine/turbine horsepower split of 69.1% engines and 30.9% turbines was applied to the
storage horsepower before adjusting the totals.
                                         19

-------
             A.G.A. also separately reports that 13,9 MMhp is for transmission in 1992,7
This is based upon the annual A.G.A. survey issued to member companies. GRI
TRANSDAT horsepower, less the storage segment, is about 6%  higher than the A.G.A.
number.  This discrepancy probably results from differences in the field gathering
horsepower reported by GRI TRANSDAT and A.G,A,,  Currently, there is no way to
separate this type of horsepower from that used for mainline transmission.  No other
adjustment to the GRI TRANSDAT horsepower data were made during this study.

             Horsepower in the GRI TRANSDAT Industry Database is reported as site
totals.  The national estimate for transmission compressor stations is simply a sum  of all of
these site totals.  As a result, the uncertainty could not he calculated from the data as was
done for the other industry segments.  Therefore, confidence limits for compressor station
horsepower were conservatively estimated by engineering judgement to be ±  10%  for each
type of compressor driver.

             Operating Hours

             The average annual compressor operating hours for the transmission segment
was determined from information reported on FBRC Form No, 2 for ffie year  1992,  In the
FERC database, transmission companies report the total number of operating hours per stage
of compression for each compressor station.  The total number of stages of compression
operating on the  date of peak demand is also reported by station. The vatoe for operating
hours was divided by the total number of stages to give the average annual operating hours
for the station. The average annual operating hours at every station for all the companies in
the FERC database were then summed and divided by the total number of stations to produce
the nationwide activity factor. Transmission compressor station annual operating hours were
3787 hours per year. For a few stations the information in the FERC database was not
considered to be  accurate (less than 1% of the total), so the average operating hours for the
rest of fee company's stations was applied to the station with questionable data.  Confidence
limits for the FERC operating hours were calculated to fee ± 4,0%.
                                         20

-------
              Because the FERC database does not identify the type of driver, reciprocating
engine or turbine, a procedure was developed to split the total average operating hours (3787
hr/yrj between the two types.  The FERC data were weighted by the total number of each
type of driver using data from GRI TRANSDAT as shown in the following equation.

                                  7489,^ (RJ =
where:       Thr    ~ Turbine operating hours, hr/yr
             Rhr    — Reciprocating engine operating hours, hr/yr
             The following relationship between Thr and R^ was based on data provided by
a company for 107 transmission compressor stations, which included 89 turbines and 524
reciprocating engines.

                               T±  = 2718 hr/yr =
                               R^    5086 hr/yr
Using these two equations to solve for Thc and Rhr resulted in annual operating hours of 2118
hr/yr for turbines and 3964 lir/yr for reciprocating engines.

             Confidence limits for the transmission compressor station operating hours  were
calculated to be + 31,3% for turbines and ± 13.9% for reciprocating engines. The
confidence limits were based on assumed uncertainties of ± 10% for the GRI  TRANSDAT
data and calculated confidence limits for both the FERC and company data.
                                         21
                                                                             92-142.05

-------
5.3.2        Underground Storage Horsepower and Hours

             Horsepower

             Horsepower data from A.G.A,'s Gas Fasts were used for (his industry
segment.  A.G.A. reports that for 1992 the total installed compressor horsepower associated
with underground storage was 1,920,441 horsepower (entire United States).1 The
horsepower data were gathered by A.6.A. through a surrey of their member companies,
which includes approximately 96% of total gas industry sales.  Since GRI TRANSDAT
includes storage  compressor horsepower, the horsepower split between engines and  turbines
was assumed to be the same as the split found in this database (69.1%, engines and 30,9%,
turbines).   As a result, the horsepower attributed to engines is  1,33 MMhp-and to turbines is
0,59 MMhp,  Confidence limits for storage horsepower (engines and turbines) were
calculated based  on an assumed ± 5%  uncertainty for tte A.G.A. horsepower and an
assumed ± 10% uncertainty for the GM TRANSDAT horsepower splits.

             Operating Hours

             Table 5-5  shows the operating hours of compressors at four sites surveyed and
the operating hours provided by one company for their seven storage fields.  The operating
hoars listed are the average hours for all  of the drivers at the site.  Confidence limits were
calculated for the operating hour estimate from the variation of the site data. The resulting
uncertainty is ±  23.1% for engines and ± 620% for turbines (90% confidence limits).

5.3.3        Generator Driver  Horsepower and Hours

             The generator driver estimates are based oa site visit data and data provided
by one company  for 34 transmission compressor stations and three storage fields. Site vMt
data were  collected for generator  drivers  (for electricity generators) at seven, of the
transmission compressor stations surveyed.  For compressor stations, an average generator
                                         22

-------
      TABLE S-S, ANNUAL OPERATING HOURS FOR STORAGE SEGMENT
Site
US1
US2
US3
US4
US5-1
US5-2
US5-3
US5-4
US5-5
US5-6
US5-7
TOTAL
AVERAGE"
90% LIMIT"
No. of
Engines
9
14
2
3
5
3
3
4
3
4
0
50


Average
Operating
Hours
2400
5122:
5607
1155
2700
3813
5832
3727
3390
3327
-

3707
± 23.1%
No. of
Turbines
0
4
0
0
0
0
0
0
0
0
2
6


Average
. Operating
Hours
..
52,5
_
,_
..
-,
..
—
—
—
5781

2917
±620%
 * Based on site averages,

horsepower per station was established to determine total generator horsepower,  Similarly,
for storage fields, an average generator horsepower per field was established to determine
total generator horsepower.

             Data for generators used in transmission compressor stations are shown in
Table 5-6.  Only one turbine driver was installed at the 41 facilities surveyed; therefore, the
data  shown in Table f-6 represent an engine/turbine horsepower split of 97%/3%.  Using the
average installed horsepower per station, the total number of transmission compressor
stations in the United States for 1992 was used to extrapolate to a national estimate of 1.45
MMfap for reciprocating engine generator drivers and 0,045 MMhp for turbine generator
drivers.9 The number of compressor stations is an estimate based on data from FERC Form
No. 2 for 46  transmission companies and is described  in Volume 5 on activity  factors,4
                                         23

-------
        TABLE 5-6.  GENERATOR DRIVER DATA FROM TRANSMISSION
                          COMPRESSOR STATIONS
Site
GSl
GS2
GS3
OS4
GS5
GS6
G57
GSS-1
GS8-2
GS8-3
GS8-4
GS8-5
OSS-6
GS8-7
. OS8-8
QSS-9
GS8-10
GS8-11
OSS-12
GS8-I3
GS8-14
OS8-I5
USB-16
OS8-17
QS8-18
GS8-19
GS8-20
GS8-2I
GS842
QSB-23
OS8-24
GS8-15
GS8-26
GS8-27
GS8-28
GSS-29
C-SS-30
OSS-3J
QSB-32
GSS-33
GS8-34
TOTAL
AVERAGE*
90% LIMIT'
No.
Engines
3
1
2
0
2
I
1
3
5
1
i
4
2
1
4





3
5
I
2
3
7
2
2
2
4
t
2
J
I
i
2
1
7
i
2
1
87
installed
Engine
Up
1542
514
1028
0
3500:
500
JOO
1110
1501
30
435
1190
590
435
net-
as
225
6Q
227
640
1148
1974
435
940
tI77
2135
BK
816
290
1556
435
370
435
100
195
784
436
2365
625
816
150
35006
854
Average
Operating
Hours
3232
0
1095
,.
88
88
88
9
1688
0
2
4455
4379
0
2279
895
0
0
12
5217
47J6
5290
0
4392
2060
3660
33
0
0
27
0
129
1
0
398
4902
6
4370
254
269
0
1352
± 38.0®
*to.
Turbines
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
Instance! Average •
Turbine Operating
hp Hours
0
0
0
0
0
tt
0
0
0
0
0
1080 474
0:
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
fl
0
0
0
0
0
0
0
0
0
108»
26 474
± 620S"
" Based on site averages,
*-Assumed equal to compressor turbine drivers at storage--fields.
                                     24

-------
                    Engine hp-hr  =    Total hp -  x No, U.S. Stations                (11)
                                    No. Stations
                                      35006 hp     ,_rtA „„ .
                                   =:._—,	ti_ x  noo Stations
                                     41  Statious

                                   = 1,45 MMhp

                       Turbine hp =  1Q8° hp x 1700 Stations
                                     41 Stations

                                   = 0.045 MMhp
              Confidence limits were calculated for die horsepower estimates from the
variation of the data.  The uncertainty calculated for the compressor station estimate is
± 10%. The resulting uncertainties of the  national estimates of generator driver horsepower
are ± 23.3% for reciprocating engines and ± 166% for turbines (90% confidence limits).

              Table 5-6 also shows the annual operating hoars for the generators at
compressor stations. The operating hours listed are the average hours for all of the drivers
at fee station.  The wide variation reflects the sites that use generators for continuous
electricity generation versus the sites that use generators oitly for emergency backup.  Since
there was only one data point for turbine drivers,  the uncertainty was assumed to be the same
as that for turbine compressor drivers in underground storage at ± 620%. Confidence  limits
were calculated for the reciprocating engine operating hours estimate from the variation of
the data. The resulting uncertainty is ± 38.0% (90% confidence limits).

              Data for generators used to storage fields ape shown in Table 4-7.  The data
were provided by one company for 9 storage fields, but only 3 fields had generators.  The
date shown represent an engine/turbine horsepower split of 60%/40%.  Using the average
installed horsepower per field, the total number of storage fields in the United States for
1992 was used to extrapolate to a national estimate of 0.085 MMhp for reciprocating engine
                                          25

-------
generator drivers and 0.057 MMhp for turbine generator drivers.  The number of storage
fields is from A.O,A.'s Gas Facts (475 storage fields}.7 Confidence limits were calculated
for the horsepower estimates from the variation of the data.  The resulting uncertainties of
the national estimates of generator horsepower are ± 126% for reciprocating engines and
± 184% for turbines (90%  confidence limits).
                      Engine hp-hr =  Total hp   x  No. U.S. Fields                 (12)
                                      No. Fields
                                         9Fields

                                      - 0.085 MMhp

                          Turbine hp  = 1Q8Q hpx 475Fields
                                        9 Fields

                                      = 0.057 MMhp
              Table 5-7 also shows the annual operating hours for the generators at storage
fields. The operating hours listed are the average hours for all of the drivers at the field.
Since there was only one data point for turbine drivers , the uncertainty MS assumed to be
the same as that for turbine compressor drivers in underground storage at ± 620%.
Confidence limits were calculated from the variation in  the operating hoars data. The
resulting uncertainty is ± 377% (90% confidence limits).

       T4BLR 5-7.                 iBtfMCIt DATA        STORAGE
Hk
nr-L
GF2
GF3
GF4
GF5
QF6
GF7
GF8
OF9
TOTAL
AYEHAGE'
f«% LIMIT"
*«£*
S
0
0
0
0
0
0
2
0
3


	 I* 	
ii=K
0
0
0
0
0
f;
611
0
1611
179

•S "
?r
-,
_
-,
_
-
_
306
_

I9t
±377*
*JSw
-3
0
0
0
0
1
0
0
0
1


TsirtiP?
fl
Q
0
0
Q
1060
a
0
ft
1080
K«

flisssra,
--
_
—
,.
_
36
_
—
--

36
±6»r»*
1 Based on site averages,
* Assumed equal to compressor turbine drivers at storage fields.
                                          26

-------
              Generator drivers in production and processing were not Included in this
analysis since these sources are negligible.  Production generators  are rare.  Processing
generators are more coRunon but rarely operate.  The generator date reported by one
production and processing company accounted for less than 3% of their total drivers and
corresponded to less than 0.01 Bsef methane emissions.

5.4           Indsistry Activity Factors

              The two main types of compressor and generator drivers in the natural gas
industry are reciprocating gas engines and gas-fired turbines.  Horsepower-hour data were
available for the production industry segment activity factor calculation.  Two pieces of
information were needed to calculate the activity factors (hp-hr) for each type of driver in
each of the remaining industry segments.  These were the installed horsepower and the
average annual operating hours.  The final hp-hr activity factors were calculated using the
national estimates for compressor horsepower and average operating hours for these industry
segments, as described in the preceding sections.  Table 5-8 summarizes  each of these
estimates and stows the resulting activity factors for beth engines and turbines by industry
segment.  The table also includes the 90% confidence limits calculated for each factor.
Activity factor confidence limits were propagated from the  confidence limits for the
individual terms using a standard  statistical approach.
                                           27
                                                                              Q9-14?

-------
             TABLE 5-8.  COMPRESSOR DRIVER ACTIVITY FACTORS FOR EACH INDUSTRY SEGMENT
KJ
OQ
industry Segment
Production
Processing
Transmission
Compressor Drivers
Generator Drivers
Storage
Compressor Drivers
Generator Drivers
Instilled
Engine
MMhp"
NA
4.19 .£ 132%"

10.2 ±
1.45 ±

1.33 ±

10,0%
23,3%

13.5%
0.085 ± 126%
- Installed
_ TOri>itte
NA
5.19 ± 99.4%b

4.55 ± 10.0%
0.045 ± 166%

0.59 ± 13.5%
0.057 ± 184%
Annual
Hbiirs
Battue .
6626

3964
1352

3707
191
NA
±:11.5%

±13.8%
± 38.0%

± 23.1%
±377%
Annual
' Hours
Turbine .
6345

2118
474

2917
36
NA
±:48.4%

±31.3%
±620%

±620%
±: 620%
Engine
MMHp -br
27,460 ± 200%
27,760 ± 133%

40,380
1962

4922
16.3

± 17.1%
± 45.4%

± 26.9%
±621%
Turbine
MMHp -fer
0
32,910- ± 121%

9635 ±
21.2 ±

1729 ±
2,05 ±

33.0%
1215%

626%
1312%
       Does not include horsqjower associated with gas "ft for oil recovery or with el«:tric drivers.
       Average of two estimation methods.

-------
6.0          NATIONAL COMPRESSOR DRI¥ER EMISSION ESTIMATES

             The annual emissions for compressor drivers were determined by multiplying
the emission factors (Section 4) by the horsepower'hour activity factors (Section 5) for
reciprocating engines and gas-fired turbines and adding these values for each industry
segment.  Table 6-1 shows the resulting emissions for each industry segment and the overall
national emissions estimate for eaeh type of compressor driver.  Using the same approach for
error propagation, the total annual emissions from compressor drivers and generator drivers
for the entire industry are estimated to be 24,8 Bscf ± 64%,
    TABLE 6-1. COMPRESSOR DRIVER EMISSIONS FOR THE NATURAL GAS
                            INDUSTRY BY SEGMENT
.--•-" Driver

Production
Processing
Transmission
Storage
Generators
TOTAL
Engines, Bscf
6.58 ± 260%
6,65 ± 133%
9.68 ± 17,9%
1.18 ± 26 J%
0,474 ± 45,6%
24,57 ± 65.1%
Turbines, Bscf
0,00
0,186 ± 129%
0,0546 ± 45.7%
0.00979 ± 654%
0.000132 ± 1163%
0,256 ± 97,8%
             Approximately 92% of the emissions from compressor driver exhaust are
estimated to be from reciprocating engines used in Ine production, processing, and
transmission segments; another 5 % of the emissions are from engines in the storage segment.
If generator drivers are included, the percentages for all drivers are 94% and 5%,
respectively, AM other categories are negligible in comparison; therefore, it is more
important to accurately determine the activity factors for reciprocating engines in production,
processing, and transmission.
                                        29
                                                                            i jo rkc

-------
 7.0

              As a result of the              in this study, the       gas industry
                          driver                      to be 24,8 Bscf ±64%.  Nearly
 99% of the methane emissions are from reciprocating engines, and approximately 40% of the
                  are from        in the                               The
          from            driver exhaust               7.9% of the
 from the entire natural gas  industry.10 Table 7-1 shows how these emissions rank with
       to the                      in
           7-1.            OP                                                   BY

Production
Processing
Storage
Transmission
Distribution
TOTAL"
6.58
6.84
1.19
10.2
0
24.8
84.4
36.4
18,2
98.3
77.0
314.3
7.8
18,9
6.5
10.4
0
7.9
 a  Compressor driver totals are for reciprocating engines and gas-fired turbines combined.
 *                    drivers.

             Compressor driver exhaust emissions represent a significant source of methane
          in all of the                                    are                     all
of the           are due to                     any                          on
types of drivers,  The confidence limits on the engine emission factor were ±5% while the
•srror bound on the activity                  ± 17.1% to  ± 621%. Therefore,
the       activity       would      the               on the          of the
estimate.
                                          30

-------
              As part of the requirements for Title V operating permits, many facilities will
implement recordkeeping practices that may benefit this type of data collection,  If companies
were to collect horsepower-how data for each compressor driver, the emission factors
developed as part of this project would allow a direct way  to determine their methane
emissions.  Access to these databases would also provide the resource to improve the natural
gas industry activity  factors for compressor drivers.   Using this kind of data would further
improve the precision of the overall compressor driver exhaust emissions estimate,
                                          31
                                                                             92-142,05

-------
8.0          REFERENCES
 1.           Biederman, N.  GRI TRANSDAT Database:  Compressor Module, (prepared
             for Gas Research Institute),  npb Associates with Tom Joyce and Associates,
             Chicago,  IL, August 1991.

2.           Pipeline Systems, Incorporated. Annual Methane Emission Estimate of the
             Natural Gas Systems in the United States, Phase 2, Prepared for Radian
             Corporation, September 1990.

3,           Jones, D.L., L.M. Campbell, C.E, Burklin, M. Gundappa, and R.A. Lott,
             "National Estimate of Methane Emissions from Compressors In the U.S.
             Natural Gas Industry," Radian Corporation and Gas Research Institute, Air &
             Waste Management Association Conference Proceedings, Paper # 92-142,02,
             Kansas City, MO, 1992,

4.           Stapper, B.E. Methane Emissions from the Natural Gas Industry, Volume 5:
             Activity Factors,  Final Report, GRI-94/0257.22 and EPA-600/R-96-080e.
             Gas Research Institute and U.S.  Environmental Protection Agency, June 1996.

5.           U.S. Department of Energy/Energy Information Administration.  Natural Gas
             Annual, 1992. DOE/EIA-0131 (92), Washington, DC, September 1992.

6.           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-Q80d, Gas Research Institute and U.S,
             Environmental Protection Agency, June 1996,

7,           Gas Facts;  1992 Data, American Gas Association, Arlington,  VA, 1993.

8.           Oil & Gas Journal. 1992 Worldwide Gas Processing Survey Database, 1993.

9.           Federal Energy Regulatory Commission (FERC) Form Mo, 2:  Annual Report
             of Major Natural Gas  Companies, 1992 database.

10.          Harrison,  M.R., L.M. Campbell, T.M, Shires, and R.ML Cowgill. Methane
             Emissions from the Natural Gas Industry, Volume 2: Technical Report, Final
             Ileport, GRI-94/0257.1 and EPA-600/R-96-08Qb, Gas Research Institute and
             U.S. Environmental Protection Agency, June 1996.
                                        32
                                                                          92-142.05

-------
              APPENDIX A

National Estimate of Methane Emissions from
Compressors in the U.S. Natural Gas Industry
                  A-l

-------
                                                                      92-142,05

National Estimate of Methane Emissions froai CuiapEessors in the U.S. Natural
                                  Gas  Industry

   Donna  Lee  Janes,  Lisa  M.  Campbell,  Clint  E,  Bwrklln,  and  Mahesh Gundappa
       Radian Corporation, Research Triangle Park, North Carolina  27709

                                Robert A. Lot:^:
                Gas Research Institute, Chicago, Illinois 60631
INTRODUCTION

      The  combustion of natural  gas  emits  much  less  carbon dioxide  per unit  of
energy  generated  than other  fossil fuels.   For  this  reason,  one  strategy that
has been suggested for reducing  global warming  is  to encourage switching from
other fossil  fuels to natural  gas.   However, methane is  currently thought to
be a more  potent  greenhouse  gas  than carbon dioxide;  If  so,  leakage of natural
gas (which is approximately  90 percent methane)  could reduce or  eliminate the
advantage  of  using natural gas because of  its lower  carbon dioxide  emissions.

      Two  major "issues must  be addressed before: the  consequences: of the fuel
switching  strategy can be evaluated.  First, there is a  need to  better define
the impact of mefiliane relative to carbon dioxide on  global warming.   Second,
it is important to better define methane emissions from  the gas  industry,
Because of the latter issue, the §as Research Institute  (GRI) and the U.S.
Environmental Projection Agency  (EPA) have developed a cooperative  program to
quantify methane  andssions from  U.S.  gas operations.   Currently,  estimates of
methane emissions from the gas Industry range from 0,5 to  4,0 percent of
production.  The  GR1/EPA program Is  a comprehensive  program to determine
methane emissions from the wellhead  to the customer's meter.  "Ehe goal is to
determine  emissions to within  approximately 3 billion cubic meters  (a3), or
100 billion standard cubic feet  (scf),, which is  approximately -0.5' percent of
U.S. production.   To achieve this overall  accuracy,  an accuracy  Cargat has
been established  for each source category  in the natural gas industry.

      One  source  category is the exhaust from compressor engines  that are used
to move the gas through the  system.   Compressors are  located in  production
fields, processing plants, gas storage facilities, and along transmission
lines.  A  preliminary study  indicated that ths exhaust from compressor engines
might account for more than  50 percent of  the industry's methane  emissions,1
Because the uncertainty in this  early estimate was quite large,  Radian
Corporation conducted a study  to determine methane emissions from both
reciprocating and turbine compressor engines,  Ttie accuracy target  established
by Che  GRI/EPA program was to  determine the emissions for  this source category
within  30  billion acf.

      In this study,  methane emissions were estimated by multiplying the
methane emission  rate  of the unit by the activity  of the unit.  The  emission
rate expresses the  amount of emissions per operating  characteristic,
independent of the  other features of  the unit,  Ths activity expresses the


                                       A-2
                                                                        92-W2.05

-------
                                                                      92-L42. OS-

operating  characteristic  of  the  unit  which  was,  in this  ease,  either  a
reciprocating engine  or gas  turbine.   For example,  when  the  operating
characteristic  is  fuel usage,  thfe-activity  is  annual  fuel  used fay  compressors.
When the operating characteristic is  operating time,  the activity  is  hours  of
operation  (par  year),

      The  emission rates  used  in this study were in the  form of methane
emissions  per hour.   The  activity of  each unit was  the annual  operating hours
for that unit.  A  national estimate of annual  methane emissions was determined
from emissions  estimates  using available  data  far compressors  in the  natural
gas industry,

APPROACHES

The Ideal  Emission Estimate

      Ideally,  the annual methane emissionn from compressors in the natural
gas industry would be determined from the sum  of the  annual  emissions from
each unit  in the Industry, where the  emission  rate  may vary  over time and from
unit to unit.   This relationship is expressed  by the  following equation;

                                  N
                      Emissions- S  flER, (t) x A, (t)  1 dt                 CD
-  S l[ERi (t)  xAi. Ml At
  i-lj
where ER^t) is the instantaneous emission rate at time (t) for
AI-{C) is the instantaneous activity at time (t) for compressor^,  and N is the
total number of compressor units  in the industry,  To reflect annual
emissions,  the integral in Equation (1) ,is evaluated over  the time  interval  tj
to t2, where t2 -  tt is equal  to one year.  This  approach produces no
uncertainty in annual emissions.

Emission Estimate Using Test Data

      The ideal approach was considered impractical because it would  require a
massive data-gathering effort by  the industry.  Fortunately, a substantial
amount of data was available for both the emission rates and activity factors
of compressor engines.  For the emission rates, data for reciprocating engines
and gas turbines were available for a number of models that ware tested during
relatively short time periods,  Values for the annual activity of compressors,
in terms of operating hours, were also available for some models.   If data for
all compressors in the industry were available, the following equation could
be used to estimate methane emissions:

                               N
                   Emissions - S  |ER, |At,t8rt x A^t) L_lyaaJ               <2>
                                       A-3
                                                                        ,149

-------
                                                                      92-142.9-5

where ERj Is the emission rate of compressor (in grams of methane per hour)
evaluated over  the  time period of the test; where the  activity, Alt of
compressor! corresponds to the annual operating hours of eompEBssori!  and
where N Is  tha  total number of compressor units  in the  industry.   This
equation assumes that  the emission rate  of the compressors  during  ths test
period represents the  emission rate of the compressors  during  the  year,  on  the
average.

      Methane emissions from compressors will vary from manufacturer  to
manufacturer and model to model.  Assuming that  the compressors are properly
maintained, differences in methane emissions, even for  the  same model, also
will be caused  by operating the compressors at different horsepower levels  or
speeds, or  for  longer  or shorter periods of time in order  to satisfy  the
operational needs of the system,

      Figure 1  Illustrates the relationship between operating  load and
emissions for a Cooper Bessemer two-cycle engine.2  The figure shows  the
total hydrocarbon emission rate at varying loads, where speed:  Is used to lower
horsepower.  Total  hydrocarbon emissions can be  considered  a surrogate for
methane emissions,  since methane is expected ta  comprise a  large portion (over
90 percent) of  total hydrocarbon emissions from  compressors.   The  data show
Chat the emission rate for a Cooper Bessemer emgine increased  at lower loads,
with a 40 percent increase in emissions at 50 percent of rated horsepower
(full load), and a  possible 100 percent  increase if loads close to 30 percent:
of full load are used.  A similar increase in wathane emissions can be
expected for other  engine and twrbine models, although  the  magnitude  of  the
Increase for other  models is not known at this cine.

      The emission  rate of a\iy compressor, then, is a function of  the rated
horsepower  and  operating horsepewer.  The equation below summarizes this
relationship:

                             ERj <= f (ER[, HpJ, Hp4)                         (3)

where ERi is the operating emission rate of cosapressori; ERj0 is the emission
rate of compressor^ at the rated horsepower,  Hp^0; and Hp£ is the operating
horsepower.  The data  from the Cooper Bessemer Illustrate that a linear
relationship can be used to represent emissions  for this engine as a  function
of horsepower,  within  3 percent accuracy, over the range of horsepower from
50 to 100 percent of rated horsepower.  In this  situation,  the average
horsepower  at which the engine operates over the year  (as in Equation (2))
could be used to evaluate the integral in Equation (1)  without any loss of
accuracy.

      If the information needed feo develop Equation (3) was known  for all
engines and Che relationship between horsepower and emissions  for  each engine
was linear, the following equation could be used to more accurately estimate
annual methane  eratssions from compressor engines from test  data;
                                      A-4

-------

11 -

19 -

&S 7-
i*
Ik _
Hfi. 8 1
I 5
6. s

2M
<












» S
-4-



N
^






n 4
196 HPW




^
^^





II S
fe£





%
* . "




i * f f "
» e
P
231 RPW











0 7
-a






e».




0 i
264 RPI







'"•-%.



ID S
« -o







"X7t



10 1
29/RPH







'* * .^ . A



00 1'





,






0 1
330RP











20 ia
M











10

Figure 1.  Effect of horsepower and speed on total hydrocarbon emissions for a
         Cooper Bessemer 2 cycle engine2
e-
IN3

O

-------
                                                                      92-142.05


              Annual Emissions    _  . „_, ,,, ,,    ,     , ,,     ,,,  .  .          f&\
                  ,     .      —  S  [ ER (Up) i x Annual Hours (Ht>), ]          t*>
                  (grams)        i—l


where ER  (Hp)4 is the emission rate of compressor as a function of horsepower

-------
                                                                      92-142,05

 contains  Information  concerning  8,282  compressors  in  the natural  gas  Industry,
 However,  individual compressor horsepower  la not reported  In the  database;
 only  the  total horsepower  for all  compressors of a specific  type  is reported
 for each  gas  company,

      A second smaller  database  in the compressor  module (the "Operating
 Database")  is a subset  of  the Industry Database (with 1,515  units)
 corresponding to  3,2  million total horsepower.  The Operating Database  is the
 only  database in  fHANSDAT  that contains information on annual operating hours
 fe£ each  unit,  TBe operating horsepower of eaeS compressor  la also recorded
 In the database.  The data In the  Operating Database  were  obtained frcm an AG&.
 survey of 112 companies In ozone nonattaintnent areas,  where  NO^ emissions were
 minimized.  Because emissions of NO* and hydrocarbons are  inversely related,
 the operation of  compressors to  minimize ROX will likely Increase emissions of
 hydrocarbons and, hence, of metHane,   Consequently, the data In the Operating
 Database, such as the operating  horsepower, may represent  maximum methane
 emission  conditions.  The  use of these data, then,  presents  a, conservative
 estimate  of industry  emissions,  if the data is representative of  the  industry
 as a  whole.

      Data  from .emissions  tests  performed by Southwest Research Institute are
 contained in a tMrd  database In TRANSDAT  (the "Test  Database"),  During the
 emission  tests, die compressors  were operated a£ close to  full load (rated
 horsepower).  Methane emissions, fuel  use, fuel use rate,  and horsepower were
 recorded  for eacK emissions test for 241 models of  engines and turbines.
 Since there was seme  variation ta  horsepower in the multiple emissions  tests
 for the same compressor, it may  be possible to define  the  relationship  between
 horsepower  and meChane  emission  fate for the 241 models of compressors  In this
 database,  With this  information,  it would be possible to  extrapolate from the
 test  conditions to actual  operating conditions at  lower horsepower levels.
 Because of  the limitations  of this study, an analysis  of this type was  not
 performed,

      Because the Test  Database  contains only amission data  and the other
 databases contain only  operating data, a fourth database was  developed  that
 contains  data for compressors for  which information was found in both, the
 Operating Database and  Test Database,  This fourth  database was called  the
 "Emissions Database."   A total of  775  reciprocating engines  and 86 gas
 turbines  (out of 1,515  units in  Che Operating Database) fit  the criteria and
 weire  included.

      Table 1 describes the contents of each of the four databases.

Model-Matching Hierarchy

      A model-matching  schema was  designed Co maximise the amount of
 correlation between the Test Database  data and the  Industry  and Operations
 Databases. Originally, when the  compressors were matched according to exact
model names, the Test Database accounted for only 38 percent  of the units


                                     A-7

-------
                                                     TABLE 1.  DESCRIPTION OF THE DATABASES
Industry Database
Data
Total Units
Hanuf/Hodel'
Total Hp (10s)
Total Hours (10«)
Engines
7,489
922
11.2
NAb
Turbines
793
144
5.0
NA
Total
8,282
1,066
16.2
: KA
Operating Database
Engines
1 , 385
318
2.A
4 . 6
Turbines
130
31
0.8
0.3
Total
1,515
34 <
3.2
4.9
test Database
Engines
229
229
NA
NA
Turbines
12
12
NA
NA
Total
241
24t
NA
NA
Emissions Database
Engines
775
120
1.3
2.9
Turbines
86
7
0,4
0,2
Total
861
127
1.7
3.1
'Number of unique compressor models in the database.


bNot applicable.
                                                                                                                                            >£>
                                                                                                                                            to
                                                                                                                                            ro

                                                                                                                                            o

-------
                                                                     92-142.05

In  the Operating Database and only 7 percent of the Industry Database,
Through consultations with experts in the field of compressor emissions, a
three-step hierarchy was designed to match more data in the Industry and
Operating Databases to the Test Database.

      The first step, therefore, in the hierarchy was based on an exact match,
FOE instance, an Ajax OPC-360 engine in the Industry or Operating Database was
identified with Test Database data for an Ajax DPC-360 in Step (1) of the
hierarchy.

      The second step in the matching procedure matched compressors by
substring name, where the horsepower of both compressors was within
± 20 percent.  An example of Step (2) in the hierarchy was: a. match of a Clark
BAST engine with a Clark HBA8-T engine (common substring o£ EAST), where the
horsepower for the Clark HBAfiT engine in the Emissions Database  was 1,911,. and
was 2,050 for the Clark BA8T engine In the Industry Database,  The third step
in  the hierarchy consisted of matching the rated horsepower per  cylinder
(±  20 percent) in compressors with similar initial names (manufacturer) but
with varying substrings.  An example of Step (3) is a matcli of a Clark BA5
engine, with 248 horsepower per cylinder (total horsepower — 1,242) and a
Clark BA8 engine, with 211 horsepower per cylinder (total horsepower — 1,000).

      Following the execution of the three steps In the model-matching
hierarchy, 37 percent of the models and 57 percent of the units  In the
Operating Database (and 23 and 55 percent, respectively, in the  Industry
Database) were matched to emissions data.  Trio execution of the  three-step
hierarchy also increased the amount of data in the Emissions Database to a
total of 1,7 million horsepower, from the previous total of 1.3  million
horsepower after Step (1) only, and accounting for over one-half of the
horsepower in the Operating Database and Industry Database.

METHOD TO ESTIMATE METHANE EMISSIONS

      The Emissions Database was used to estimate methane emissions from
compressors in the natural gas industry.  The method used to calculate
emissions is described below,

Equations

      An average emission rate was obtained for each model of compressor
engine and turbine In the Emissions Database from the average of all methane
emission tests in the Test Database for that model.  Since the time period in
which each test was conducted was not given, the emission rates, in units of
grains of methane per hour, were calculated from the reported methane emissions
per unic of fuel (in grams per m3)  and the reported fuel USB race (in m3 per
hour) for each compressor, as In the equation below;
                                     A-9

-------
                                                                      92-142,05
                Emission Rate,   Average Emission Rate,    FOR,
                             *  »m                       x
                    (g/hr)              (g/m3)           (
where FDRj was the average fuel use rate for

      The methane emissions  for each  compressor  were  then calculated as in
Equation (8) below;

                    Emissions,   __                                         /c*
                             * - ER, x Annual Operating Hours,                W
                     (grams)      *                         *

where ER4 was the methane emission rate of eompressori calculated with
Equation (5):, and Annual Operating Hours t were obtained froa the data for
compressor^ in the Emissions Database,

      To determine the  total methane  emissions for  the compressors  in the
IBIS s ions Database, the following equation was used;


             Annual Emissions .  *  [ ER, x Annual Operating Hours. ]          (?)
                  (grams)        1_1 l   »•           r       B      *J


where the emissions Iron each unit are calculated as  in Equation (6), and  H
was the total number of compressors in the database.

Estimate of National Emissions

      If data for all compressors in  the industry were available, a national
estimate could be calculated using Equation (7) ,  Since the compressors in the
Emissions Database were only a subset of the compressors  in the natural gas
industry, a procedure was necessary to relate the methane emissions calculated
with Equation (7) to a  national estimate,

      The ratio of the  total horsepower from compressors  in the industry
(16,7 million horsepower) to the  total horsepower of  the  compressors in the
Emissions Database (1.7 Billion horsepower) was used  to scale up the methane
emissions calculated in Equation  (7)  by a factor of (16.7/1.7),  or  9.8,  to
estimate national emissions.  This relationship  is shown  in the equation
below:

                                                      Scaling Factor
           National Methane _ Methane  Emissions for         up,             (§)
               Emissions       Emissions Database          — — -
where Hpj is the total horsepower in the industry* and Hpa is the total
horsepower in the Enissions Database, which produce m scaling factor of  9.8,
                                      A-10

-------
                                                                     92-142.05

RESULTS

National Emission Estimate

      An estimate: of national methane emissions from compressors using the
approach diseussad above was 0,22 teragrams (TgJ of methane.  Over 99 percent
of the emissions ware estimated to be from reciprocating engines and less than
1 percent from gas turbines.  This estimate was based on the assumption that
the compressors in the Emission Database represent compressors used in the
natural gas industry, on the average.  If the industry compressors are
operated at horsepower levels much less than the rated horsepower, the methane
emissions estimated here could underestimate the industry emissions, since
methane emissions are thought to vary inversely with horsepower.

Field, Plant, and Pipeline Compressor Emissions

      Compressors are used in field and plant operations as well as in
transmission activities.  The emission estimate above does not apportion
methane emissions among the segments of the natural gas industry that use
compressors.  The U.S. Department of Energy (DOE) provides estimates of the
amount of fuel used in field (lease), plant, and pipeline applications, and
these estimates were used to apportion methane emission estimates among the
sources,3  The methane emission estimates  were based on the assumption that
all the fuel reported for field and plant purposes was used by compressors.
This assumption is likely to be an overestimatet because fuel is known to be
used for field and plant purposes by equipment other than compressors.
Although the portion of lease and plant fuel used for other purposes was not
known, estimates of the fuel used by these sources served as rsugh estimates
of fuel used by compressors in field, plant, and pipeline operations.

      The result of using DOE estimates of fuel use was that 39 percent of
total compressor fuel was attributed to field compressors, 26 percent to plant
compressors, and 35 percent to pipeline.  It is likely, however, that a higher
percentage of total compressor fuel is used for pipeline, ana that lower
portions are used for field and plant.  The AGA estimated that 84 percent of
total compressor horsepower was used for the pipeline4; if fuel use can be
assumed to be proportional to horsepower,  the percent of fuel used by pipeline
compressors could be over- twice as high as the estimate based on the DOE
Information.

      Since methane emissions were assumed to be proportional Co fuel use, the
PDE breakdown in fuel among the three sources of compressor emissions in the
natural gas industry was used to apportion the national estimate of 0.22 Tg of
methane emissions between the three sources.  The results were that, of the
total 220 megagraas (Mg) of annual methane emissions estimated for compressors
in the industry, 86 Mg of methane was estimated to be emitted from field
compressors, 57 Mg from plant compressors, and 77 Mg from pipeline
compressors,  As discussed above for apportioning fuel use between these three
sources, the emissions from pipeline compressors were probably underestimated,
end the portions for the other two sources were overestimated.  Another


                                      A-ll

-------
                                                                       92-142.05

  possible error In the breakdown of emissions was the assumption that the
  emission rate and operating charactaristlcs o£ compressors Iti field, plant,
  and pipeline applications were similar,

  UNCERTAINTY IN EMISSION ESTIMATES

        Since the true value of methane emissions from compressors In the
  natural gas industry is not known, the accuracy of the estimate of 0,22 Tg of
  methane emissions cannot be assessed.  However, the data can be used to
  produce an approximation of the uncertainty In the emission estimate.  Like
  confidence intervals, the uncertainty estimated for a value can describe, with
  a high degre* of confidence,  the range in which the true value lies.

        The target uncertainty for the total estimate of methane emissions from
  all sources in the natural gas industry was close to 3 billion m3 (100 billion
  scf),   Because there are many segments in the natural gas industry that emit
  methane that are part of the overall methane emission estimate in the GRI/EFA
  project,  the goal for each segment was to produce a methane estimate with as
  little uncertainty as possible.  The target uncertainty for compressors was
  approximately 8QO million m3  (30 billion scf).

        The following explains the procedure used to estimate the uncertainty in
  Che annual metisane emission estimate from compressors and the results of this
  estimate,

  Sasis  of Uncertainty in the Compressor Emission Estimate

        The only sources of uncertainty in the approach described above that
  could  be estimated were the analytical error associated with the test data
  used to calculate the methane emission fates and the scale-up to national
  amlssions based on the ratio of the horsepower in the Emissions Database to
  the estimate of compressor horsepower In the industry.

        The uncertainty in the hydrocarbon analysis was estimated to be
  ± 10 percent,  based on the expected gas chrojnatograph (with a flame ionizatlon
  detector)  capabilities.   Likewise, the uncer .ainty associated with fuel flow
  measurements was estimated to be ±2,5 percent.  The uncertainty associated
  with the  scale-up between the Emissions Database and total industry horsepower
  was more  difficult to quantify.  Two indirect assessments of this uncertainty
  are discussed below,  using comparisons between the Emissions Database and the
  Industry Database,  which was  taken to be a fairly good representation of the
  industry  as a whole.   The uncertainty for the scale-up  to nationwide emissions
  was estimated to be approximately 1 percent,  based on estimated significant
  figures in total horsepower.   For an estimated industry horsepower of
  16.7 million,  the uncertainty was estimated to be 0.17  million horsepower; for
  the Emissions  Database,  with  1,7 million horsepower,  the uncertainty was
  estimated to be 0.017 million horsepower.
                                       A-12
3.     U.S. Deaartment of Energv/EnerEV Information Administration. Natural Gas Annual, 1992,

-------
                                                                      92-142:, 05

 Procedure  to Calculate  Uncertainty

       The  software  "@R.isk"e was used to estimate the uncertainty In the
 estimate of methane emissions  from compressors.  The variables  of  uncertainty
 were  entered into the program  along with  the  values discussed above.   The
 "@Risk" uncertainty estimating procedure  recalculated  the  total methane
 emission estimate using Latin  Hypercube sampling from  normally  distributed
 intervals  along  the range  of each  variable, defined by the expected value  snct
 the (estimated)  standard deviation of  that value.  In  the  case  where  the
 standard deviation  was  estimated from  significant  figures,  a  uniform  interval
 (as opposed to a normally  distributed  interval) was used for  the Latin
 Hyperoube  sampling.

       In the "jaRisk"  program,  the  recalculation was performed 500  times, each:
 Cine  using a different  value itt the interval  o:f each variable in the  emission
 equations.   The  results of the 500 calculations, excluding values  below zero,
 were  analyzed for variation about  the  mean.   The uncertainty  was reported  as
 the coefficient  of  variation,  or the standard deviation divided by the mean,
 as a  percent.

 Estimates  of Uncertainty

      The  uncertainty for  the  estimate of  methane  emissions from compressor
 engines and turbines, at 0.22  Tg (11.5 billion scf), was estimated to  be
 approximately 1  percent, or 4.3 million m3 (0.15 billion scf).  This value is
 well  below the 800  million m3  target for compressor engines and turbines.

 Comparison of the Database

      One  component of  uncertainty in  the  methane  emission  estimate that could
 not be quantified for the  uncertainty  analysis was the  degree to which the
 compressors  in the  Emissions Database  represented  the compressors  in  the
 IttdiisCry.  The Emissions Database  was compared with Che  Industry Database  in
•two procedures that were designed  to assess the: representativeness of  the
 Emissions  Database  for "compressors  in  the  industry and,  indirectly, the
 uncertainty  associated  with the scaling factor.

      Figures 2  and 3 are  plots for anginas and turbines, respectively,
 showing the  total horsepower for each model in the Industry Database versus
the total  horsepower  of that model in  the  Emissions Database.  The line formed
 from  the relationship between  the  total horsepower for  each database was
plotted on each  graph.   For engines, a line through 11,2 million and
 1.3 million,  the total  horsepower  for engines in the Industry and  Emissions
Batabases, respectively, constituted the ideal relationship between the total
horsepower for each type of engine  in the  two databases.  Similarly,  for
turbines,  a  line through 5 million and 0.4 million produced the ideal
relationship for the  total horsepower for each type of turbine  in  Che  two
databases,  The  figures  show that  the horsepower contribution of the
compressors  in the Emissions Database was higher than the horsepower
                                     A-13

-------
                                200             300

                                 (Thousands)

                     Horeepower Coiitiibufcri in Mw Industry Database
4QQ
Figure 2.   Relative horsepower contribution of each engine modal in the

           Emissions and Industry Databases
                                                                                       o
                                                                                       un

-------
I
CO
4
220
210
200
190

170
160
150
140
'<30

110
 80
 70
 60
 SO
 40
 30
 20
 10
  0
                                                                                Ideal
                                     200
                                              (Thousands)
                                  HP Contribution - Industry Database
                                                              400
                                                                                       600
             Figure 3.   Relative horsepower contribution of each turbine model in the
                         Emissions and Industry Databases
                                                                                                       o
                                                                                                       Ul

-------
                                                                     92-142,05

contribution in the Industry Database.  However,; the trend in general is in a
positive direction for both databases, showing that compressors that
contribute a large amount of hoirsepower to the total in one database also
contribute significantly to the other.  Conversely, compressors that are small
(Hintributors in one database are of similar ranking
in the other.  This comparison presents a qualitative assessment of the
representativeness of Che Emissions Database for the industry's compressors,

      A more quantitative assessment of how well the Emissioije Database
represents industry compressors was performed using a comparison between the
emission factors derived from the Emissions Database and the Industry
Database, with the Industry Database used as a close approximation of
compressors in the natural gas industry.  To determine the emission factors
for the Emissions Database, an overall average omission factor was calculated
for engines and turbines, separately, after Step 1 in the hierarchy, where
only the data that were exact matches with the compressors its the Test
Database were included.  Overall average emission factors were calculated for
engines and turbines, weighting the model-specific emission factors for each
engine and turbine, respectively, by the horsepower contribution of that model
in the Emissions Database.

      To determine the emission factors for thfe Industry Database, model-
specific emission factors were developed after Step III in the hierarchy, and
overall average emission factors were calculated for engines and turbines,
weighted by the horsepower contribution of the engine and turbine models,
respectively, in the Industry Database,  The results were that the overall
average emission factors calculated for the Emissions Database were 18 grains
of methane emitted per m3 of fuel for engines,  and 0.16 grains of methane
emitted per m3 of fuel for turbines.   The overall average emission factors
calculated for the Industry Database were 20 grams of methane emitted per vf"
of fuel for engines, and 0.17 grams of methane emitted per m3 of fuel for
turbines.  This analysis showed: quantitatively that the Emissions Database was
a fairly good representation of the Industry Database and, therefore, the
industry,

CONCLUSIONS ANF RECOMMENDATIONS'

      The annual methane emissions estimate of 0,22 Tg for compressors in the
natural gas industry was less than the previous estimate of 1,2 Tg by a power
of 10,  Although compressors are still a significant source o£ methane
emissions, this estimate reduces the importance of compressors in the
assessment of the natural gas industry's contribution to global warming,

      The uncertainty for the estimate, at 4.3 million m3 of Methane
(0.IS billion scf), was much less than the portion of total industry methane
emissions contributed by compressors, and was 0,15 percent of the target
uncertainty (100 billion scf) for the industry.

      Future worlK in this area should include an assessment of- the effect of
compressor operation, in terms cf horsepower, on the methane emission rate.

                                      A-16

-------
                                                                     92-142.05

On the basis of the data on this subject, the annual methane emissions could
be higher than fche current estimate by a factor of 2, If compressor operatidirv
is at horsepower levels significantly lower than compressors * rated
horsepower,

REFERENCES

1,    Annual Methane Emission Estimates of the National Gas Systems in the
      United States, Phase 2 by B, Tilkieioglu.  Pipeline Systems, Inc. under
      subcontract to Radian Corporation, Research Triangle Pack, North
      Carolina,  U.S. Environmental Protection Agency Contract No, 68-02-4288.
      Research Triangle Park, North Carolina,  September 1990,

2,    Two Cycle Glean Burn (Integral) Gas Engines.  D, Blizzard, and A. fi,
      Gillette.  Presented at the Small Prime Movers NO,, Control Workshop, S&&
      Diego, CA,  June 1991.

3,    TRANSDAT:  Compressor Module.  Tom Joyce and Associates, Washington,
      D.C,, for the Gas Research Institute, Washington, D.C,  August 1991.

4.    1990 Gas Facts.  The American Gas Association, Arlington, Virginia.
      1990,  Page 65.

S,    Natural Gas Annual 1989,  Department of Energy, Energy Information
      Administration, Washington, D, C.  Volume I,  1989,

€-.    @Riak, Risk: Analysis and Simulation Add-la for Lotus 1-2-3,
      Version 1.5,  Palisade Corporation, Newfield, NY.  March 1989.

ACKNOWLEDGMENTS

      This projeet was cofunded by Bob Lett, of the Gas Research Institute,
Chicago, Illinois, and David Kirehgessner, of the Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina.  Radian gratefully acknowledges the contribution made by the
sponsors.
                                     A-17

-------
 APPENDIX B



Source Sheet P-i
     B-l

-------
                                               p-1
                                 ALL-SEGMENT SOURCE; SHEET
SOURCES;                          Compressors, Generators
OPERATING MODE;                 Normal Operation
EMISSION TYPE:                    Uiisteady, Combusted (Compressor Driver Exhaust)
-ANNUAL EMISSIONS:               24,4 Bscf ±64%

BACKGROUND:
Compressors are used to move gas through (fie system.  They are located in productlos fields, processing
plants, gas storage fields, and along transmission lines.  Methane emissions are found in compressor driver
exhaust (reciprocating engines and gas turbtn.es) because of the incomplete combustion of the natural gas burned
as fuel.

EMISSION FACTOR: (0.240 ± 5% scf/hp-hr, engines and 0.0057 ± 30% scf/hp-hr, turbines)
An average emission rate was calculated for each model of compressor engine and turbine in the GRI
TRANSDAT Emissions Database (1), which is based on compressor tests conducted by Southwest Research
Institute (SwRl).  The emission rates were calculated from the reported methane emissions per unit of fuel and
the reported fuel use rate (PUR) for each compressor model, as follows:

                                     M(ra) = EP(m) x FUl(m)                                  (1)


               where;  ER(hl,  •= average emission rate for model, m (scf/hr)


                       EP(tn)  = average emission parameter for model, m (scf CH/sef fnel)


                            = average fuel use rate fof model, in (scf fuel/hr)
The following equation was used to determine the total emissions far the 86 turbines ani "775 reciprocating
engines in the Emissions Database.
                           TE
                                   K
                                 m .= 1
 M
i = 1
                                               (2)
                    where:  TE = total emissions for database, (scf)

                           HRj - annual operating hours for compressor i, (hr/yr)


                             K = number of unique compressor models

                             M = number of compressors of model, m
                                              B-2

-------
 The emission factors, for engines and turbines, were Ihen calculated using the following equation,

                                                 N            N
                        Emission Factor = TE /
                                               ! = 1         i  =1
HR,/N)
(3)
                      where:  HP = average operating horsepower during HR, (hp)

                              HR = annual operating hours, flr/yr)

                               N = number of compressors

This equation considers thai some models could be operated at a higher percentage of the time because they are
base loaded compressors.  The average emission factors for the compressor drivers in the Emissions Database
are 0,240 scf/hp-hr for reciprocating engines and 0,0057 scf/hp-hr for turbines,

EF DATA SOURCES:

        I.      "National Estimate of Methane Emissions from Compressors in the US. Natural Gas
                Industry" (2).

EF ACCURACY;  ±5%, engines  and ± 30%, turbines

        Basis;
        The accuracy for the EP is estimated based on propagation of error from the spread of samples in the
        database.  However, engineering judgement was used ID assign accuracy for two of the individual terms
        in the equation, as follows:

        i.      Hydrocarbon analysis was estimated to be -f ID%, based on  the generally accepted accuracy
                of gat dtromatograpfas (flame ionkatknt detector).

        2,      Likewise, firel flow measurements were estimated to be ± 2,5%,

ACTIVITY FACTORS: (horsepower tour)

Horsepowerhour data were available forihe production industry segment activity factor calculation. Two
pieces of information are needed to  calculate the activity factor, which is expressed as horsepower-hours (hp-hr)
for each type of driver in each of the remaining industry segments. These are the  installed horsepower and the
average operating hours. The following table presents these parameters and the resulting activity factors for
both engines and turbines in each segment of the industry.  The sources and methods for calculating all the
values presented in She table below are given in the nest section: AF Data Sources.

ft is estimated that about 94% of the emissions in compressor and generator driver exhaust are from
reciprocating engines used in production, processing, and transmission, with about 5% attributable to
reciprocating engines used in storage. AU other categories arc negligible in comparison.  Therefore, it is  more
important  to accurately determine the activity factors for reciprocating engines in production, processing, awl
transmission.
                                                 B-3

-------
                  COMPRESSOR DRIVER ACTIVITY FACTORS FOR EACH INDUSTRY SEGMENT
Industry
Segment
Production
Processing
Transmission
Compressor Drivers
Generator Drivers
Storage
Compressor Drivers
Generator Drivers
< Installed
•Engine
•MMhp"
NA
4,19 ± 1329?

10,2 ± 10,0%
1,43 ± 23.3%

U33 i 13,5%
0,085 ± 126%
'"• Insjalfed
Turbines
'• MMhp1 ,
NA
5,19 ± 9$,4%*

4,55 ± KM
0.045 ± 166%

0.59 ± 13.3K
0.057 ± 1S458
Annunt
Haul's
Engtae
NA
6626 ± 11.5JS

3964 ± I3,8«
13SZ ± 3B.05J

3707 ±23. IS
191 ± 377*
Annual
. Hours
Turbine
NA
634S ± 48,4%

2118 ± 313%
474 ± 620%

2917 ± 620%
36 ± 620K
Engine
POTp-hr
27,460 ± 200%
27,760 ±133 %

40,380 ±17, 1 K
1962 ±45,4%

4922 ±26.9%
16,3 ± 62! %
Turbine
MMHp*r ,
0
32,910 ± 121%

96?5 ± 33.0 %
11,2 ± 3215®

1729 ± 626%
2.05 ± 1312%
 'Does not inqlude horsepower associated with gas lift for oil recovery or with electric-drivers.
 k Average of two estimation melhods.

AP DATA SOURCES:

        I.     The production segment horsepower is based on fee total installed horsepower-tours for data
               provided by one company for 516 compressor drivers (all reciprocating engines). The
               horsepower-hours for the company was divided by their production before scaling to a national
               estimate.  National horeepowerhour was calculated using tbe 1992 marketed production for the
               U.S. {Natural Gas Annual 1992, <3)J.

        2.     The processing  segment horsepower was determined by taking the average of two methods.
               Each of the methods uses  site data for the 10 gas plants  visited.  The first method scales to a
               national estimate by multiplying the total U.S. gas  plant throughput as of January 1, 1993
               [46,510,7 MMcfd, Oil &  Gas Journal (4)] by the total site visit horsepower per throughput
               (47,8 %/MMcfd, engines and 59.2 hp/MMcfd, turbines). The second method scales to a
               national estimate by multiplying the total number of gas  plants in the U.S. [726, Oil & Gas
               Journal (4)3 by  the total site Visit horsepower per uninber of gas plants visited (10), which is a
               scale-up ratio of about 73, The annual operating hours are based on Hie 10 sites plus data
               from two companies for an additional 18 gas plants.  An average of she average operating
               hours per site was calculated to get the processing segment operating hours (203 engines and 9
               turbines),

        3.      The transmission segment compressor station horsepower for each compressor driver type is
               based on the GRI TRANSDAT database.  Installed horsepower was taken from die Industry
               Database module of GRI TRANSDAT. The annual operating hours are based on information
               reported on PBRC Form No, 2, FERC data does not distinguish between driver type.  The
               FERC data were split between engines and turbines based on data in GRI TRANSDAT and
               data provided by one transmission company (524 engines and 89 turbines),

        4.      The storage segment horsepower came from Gas Facts (5) data for 1992  (1,920,441 hp). The
               split between engines and  turbines was assumed to  be the same as the engine and  turbine splits
               found in GRI TRANSDAT{€9.1%, engines and 30J%, turbines). The annual operating
               hours we based on 11 storage stations (50 engines sad 6 turbines). An average of the  average
               operating hours per station was calculated to get the storage segment operating hours.
                                                B-4

-------
                The generator driver horsepower (compressor stations) is based on the total installed
                horsepower for 7 of the transmission sites visited and company data for 34 transmission
                com^-essor stations. To scale to a national estimate, the total horsepower per station was
                multiplied by the total number of transmission compressor stations [1700, PERC Form No, 2
                (6)] in the U.S.  The annual operating hours  are also based on data from the site visits and
                company data.  An  average of the average generator operating hours per station was calculated
                lo get generator operating hours (87 engines  and 1 turbine).

                The generator driver horsepower (storage fields) is based on the total installed horsepower for
                9 storage fields (one company),  To scale to  a national estimate, the total horsepower per field
                was multiplied by the total number of storage fields [475, Gas Fads (5)] in the United States.
                The anniKi'i operating hours are also based on the company data,  An average of the average
                generator operating  hours per field was calculated to get generator operating hours (3 engines
                and 1 turbine).
AF ACCURACY;

        Basis:
        Errors were propagated from each of the following terms;
        1.       Production Hp-hr: The production Hp-br accuracy is based upon an engineering analysis and
                set at ± 300%.
        2.       Transmission Hp-hr: The transmission Hp-hr accuracy is based upon an assigned estimated
                error of ± 10% for the horsepower data in the GUI TRANSDAT database and error
                propagation from  the FERC operating hours,
        3.       Other segment Hp-hr:  The accuracy of the site visit data for horsepower and operating hours
                was also propagated using the spread of the data, but from much smaller data sets.  The
                accuracy of the horsepower Mm activity factors for each industry segment are calculated
                statistically using the individual terms  for horsepower and operating hours,

ANNUAL EMISSIONS:  (24.57 ± 65.1% Bscf, engines -f 0.256  £.37.8% Bsef, turbines)
The annual  emissions were determined by multiplying an emission factor by the horsepower'hour activity factor
for reciprocating engines and turbines and summing these values for each segment.  The following table shows
the resulting emissions for each industry segment and Hie overall national estimate,

    ANNUAL                 EMISSIONS FOR THE NATOEAL GAS INDUSTRY BY SEGMENT
Corapresior
Engines, Bscf
Turbines, Bscf
Production
6.38*200%
o.oe
Processing
fi.65 ±13356
O.J«6±129«
Transmission
MS±i7.9SS
0,0546*45,7 ?&
Storage
i,!8±26,9%
0,00»9±654%
Generators
0.474 ±4S«6*
0,00013241163%
TOTAL
24,57±65.1«:
0.236*97.8* !
REFERENCES

1,       Biederman, N. ORJ TRANSDAT Database:  Compressor Module,  (prepared for Gas Research
        Institute),  npb Associates with Tom Joyce and Associates, Chicago, IL, August  1991.

2.       Jones, D.L., L,M Campbell, C.E. BurfcHn, M. Gundappa, and R..A, Lett, "National Estimate of
        Methane Emissions from Compressors in the U.S. Natural Gas Industry,"  Radian Corporation and
        Gas Research Institute, Air & Waste Management Association Conference Proceedings, Paper # 92-
        142.02, Kansas City, MO, 1992.
                                                B-5

-------
:3.      U.S. Department of Energy/Energy Information Administration. Natural Gfe Annual, 1992,
        DOE/EIA-OO1 (92), Washington, DC, September 1992,

4,      Oil & Gas Journal.   1992 Worldwide Gas Proeesfing Survey Database, 1993,

5,      Gas Facts;  1992 Data, American Gas Association, Arlington, VA. 1993.

5,      Federal Eneigy Regulatory Commission (FERC) Fomi No. 2: Annual Report of Major Natural Gas
        Companies,  1992 database.
                                              B-6

-------
          APPENDIX C   ,



Production Segment Site Visit Results
              C-l

-------
       Table C-l  shows the production segment data collected for each of the 28 sites visited
 and the data provided by one company for 25 gathering stations,  The data have been
 grouped into four geographic areas so that a regional extrapolation can be conducted for this
 industry segment.  The following four regions were selected based on differences in
 production rates and equipment populations: Pacific Mountain, Gulf Coast, Central Plains,
 and Atlantic & Great Lakes. Differences in onshore versus offshore production were not
 considered because of the lack  of available site data,  (This approach  is described further in
 the Methane Emissions from the Natural Gas Industry, Volume 5;  Activity Factors),1

       No turbines were installed at any of the production facilities surveyed; therefore, the
 data shown in Table C-l represent art engine/turbine horsepower split of 100%/0% for this
 segment.   Excluding compressors used for gas lift and electrically driven compressors, the
 average horsepower per MMcfd was calculated for each region.  The United States marketed
 oatural gas production,  by region, for 1992 was used to extrapolate from the regional ratios
 (hp/MMcfd) to a national estimate of compressor engine driver horsepower in the production
 segment of 3.61 MMhp (0 MMhp for turbine drivers)..* The following equation describes the
 national extrapolation using the regional approach.

                              region 4 ,
              Engine hp =     £    Av»
                              region 1 'V

       This estimate for the production segment is based on the total installed horsepower
 from the sites surveyed. Table C-2  shows the regional data used to extrapolate to the
 national horsepower estimate.   The table also shows the 90% confidence intervals for each of
 the values  used in. the calculation. Confidence limits were calculated using the ratio method
 f&r each region's average hp/MMcfd and the total horsepower from the variation of the site
data. The resulting uncertainty is ±- 63.2%,

       Table C-3 shows Hie "operating hours" for the compressors at the sites surveyed. For
the production segment, annual operating hears were not available for most of the sites
Visited; therefore,  toe compressors that were running during the site visit were assigned
annual operating hours of 8760,  Compressors that were idle were assigned annual operating
hours of 0. This approach was used for 9 of the sites  visited.  Actual operating hours were
available for two of the 2 sites  visited and for the 25 gathering stations surveyed (one
company). Operating hours for the remaining sites were not included in this analysis since
Compressors  were either used for gas lift, had electric drivers, or were not found at ihe site.
The operating howrs listed for these sites are the average hours for all of the drivers at the
site.  This  table also reflects the faet that no turbines were surveyed in fc production
industry segment.  Confidence  limits were calculated for the operating hours estimate from
the variation of the data.  The resulting uncertainty is -fc 9.7% (90% confidence limits).

       Activity factors based on the  site data presented here are 23,820 MMhp-hr for
engines and 0 MMhp-hr for turbines. Confidence limits were calculated for the engine
driver activity factor to  be ± 64.3%,  This activity factor is within the error bounds of the

                                          C-2

-------
activity factor calculated using the company database for horsepower-hours {Section 4,1).
This database provided one data point for turbine drivers in the production segment, thus
affirming their presence in production.  However, the number of these drivers is assumed to
be very small and not significant for tills estimate,

REFERENCE

1,     Stapper, B.E, Methane Emissions from the Natural Gas Industry, Volume 5: Activity
       Factors, Final Report, GRI~94/0257,22 and EPA~600/R-96-080e, Gas Research
       Institute and U.S. Environmental Protection Agency, June 1996.

2,     U.S. Department of Energy/Energy Information Administration,  Namral Gas
       Annual, 1992. DOE/EIA-0131 02), Washington, DC, September K92.
                                        C-3

-------
TABLE C-l. GATHERING COMPRESSOR DATA FROM
      PRODUCTION SEGMENT SITE VISITS
:SM£ -. , '
Pacific Mountain
PM1
PM2
PM3
Total f
Gulf Coast
GC1
GC2
GC3
GC4
GC5
GC6
GC7
GC8
GC9
Total f
Central Plains
CP1
CP2
CP3
CP4*
Total '
Atlantic & Great Lakes
AGi
AG2
AG3
AG4
AG5
AG6
installed
hp*
3794 J (112000)
1650
-0 (200 b)
39591
OCNA')
7467"
4950
700*
0"
S87S
0
625
0(NAC)
22620
6000
4650
4480
72192
87322
0
0
0
0
0
0
MMcfd
104
4.5
12
120.5
4,5
23
25,5
112
12
54
IS
250
0.50
499.5
180
42.7
240
542
1004.7
24
0.18
0.18
0.17
0.39
0.37
-'•".:. AVg
hp/MMcfd



328.6









45J




86.9






Continued
                   C-4

-------
                            TABLE C-l.
Site
AO7
ACS
AO9
AGIO
AG11
AC 12
A013
Total '
Installed
hp '
50"
0
0
0
0
0
0
50
• MMcfd
0,37
OJ8
0.23
0,30
0.35
0.13
0,35
27,2
Avg
hp/MMcfd







1.84
* Total site horsepower (including gas lift for oil recovery) is shown in parentheses.
* Horsepower associated: with electric drivers.
c Not available.
d Estimated.
' Total for 25 gathering stations; peak production rate.
f Excludes gas lift and electric drivers.
                TABLE C-2.  REGIONAL DATA FOR PRODUCTION
                   SEGMENT HORSEPOWER EXTRAPOLATION

Region
Site t
. hp/MMcfd '
Pacific Mountain 328.6
Gulf Coast 45,3
Central Plains 86.9
Atlantic & Great Lakes 1.84
Total U.S.
90% Limit
ivs
90% Limit .
. -' National Marketed
Production
' -{ MMcfd

Engine
MMhp
± 48.2% 26% OJ86
± 132% 31544 1,43
±94.6% 14860 1.29
± 244% 2166 0.0039S
51265 3.61
± 63.2%
                                        C-5

-------
TABLE C-3. ANNUAL OPERATING HOURS FOR PRODUCTION SEGMENT
Site
PMI h
PM2
PM3
GCi
GC2"
GC3 "
GC4 fc
OC5 b
GC6"
GC7
GC8
GC9
CP1 "
CP2b
CP3 b
CP4-1
CP4-2
CP4-3
CP5-4
CP4-5
CP4-6
CP4-7
CP4-8
CP4-9
CP4-10
CP4-11
CP4-12
CP4-13
CP4-14
No, of Engines '
19 (56)
17
0 (9)"*
0(14)
12
4
1
0 (If
37
0
25 "
0(1)
50 d
31
64
I
1
2
1
1
IS
I
1
3
2
5
1
1
!
Average Operating Hours
8760
78S4
;,,
...
8030
87fifl
87SO
„»
8523
—
4380
--..
8760
8195
8760
5924
83:64
4753
ft
6884
5193
7631
7309
6953
5833
«Sfi
5505
7235
7333
                                                     Continued
                           C-6

-------
                                TABLE; c-3.  {CONTINUED)
Site
CP4-15
CP4-16
CP4-I7
CP4-18
CP4-19
CP4-20
CP4-21
CP4-22
CP4-23
CP4-24
CP4-25
AG1
AG2
AG3
AG4
AG5
AG6
AG?
AQS
AG9
A" 10
AG11
A012
AG13
TOTAL AVERAGE '
90% LIMIT '
No. of Engines *
2
I
I
2
I
2
t
t
2
2
1
S
0
0
0
0
0
2
0
0
0
0
0
0
315
Average Operating Hours
8031
6841
7607
6303
6946
6340
748
7384
6908
7176
_. sn
,,
..
..
..
_
„
NAB
..
—
„
„
..
-
6599
± 9.1%
' Total number of compressors (including gas lift for oil recovery) is shown in parentheses,
* Based on active (8760 hrs) versus idle (0 hrs).
* Electric drivers.
" Estimated,
" Nat available,
f Based on site averages.
E Includes vapor recov«y.
                                             C-7

-------
                APPENDIX D
Transmission Compressor Station Site Visit Results
                     D-l

-------
       Data were collected during site visits and company surveys for 51 transmission
 compressor stations.  The compressor horsepower is related to the gas throughput and the
 pressure. Throughput is available for most of the sites surveyed, but a national figure is not
 available.  Pressure data are not available for any of the sites  surveyed of nationally.  As a
 Jesuit, the only extrapolation method currently available  for this industry segment is based on
 the number of compressor stations.

       Table D-l shows the number of compressors, installed horsepower, and gas
 throughput for each of the sites surveyed in the traesmission segment.  The engine/turbine
 Horsepower split for this segment is 63,8% engines and 36,2% turbines (compared to 69.1%
 for engines and 30,9% for turbines in TRANSDAT),  Using the average site horsepower per
 station and the tola! number of transmission compressor stations in the United States for 1992
 from the FERC Form No, 2, the site data were scaled up to give a national estimate of
 compressor horsepower in transmission compressor stations of about 11,1 MMhp for engines
 and 6,3 MMhp for turbines.

                      Engine hp  »  Total hp  x No. U.S, Stations
                                   No. Stations

                                 _ 334240 hp
                                   51 Stations

                                =  11.1 MMhp


                     Turbine hp = 19Q03° hp x  1700 Stations
                                  SI Stations

                                = 6.33 MMhp

       Confidence limits were calculated for the horsepower from the variation of the data,
The uncertainty calculated for the compressor station estimate used to scale up the site visit
horsepower is ± 10%.  The resulting uncertainties  foi the national horsepower estimates for
this industry segment are ± 22.0% for engines and ± 44,9% for turbines.

       Table D-2 shows the operating hours for the compressors at the 9 sites surveyed aM
for one company that provided data for 39 transmission compressor stations.  The operating
hours listed are the average hours for all of the drivers at each of the sites. Confidence
limits were calculated for the operating hours  estimate from the variation of the average site
data.  The resulting uncertainty is ± 25.0% for engines and ± 48.1% for turbines (90%
confidence limits).

       Activity factors based  on site visit data alone are 26,100 MMhp-hr for engines and
13,460 MMhp-hr for turbines.  The -uncertainties for these activity factors are ± 33.7% for

                                         D-2

-------
engines and ± 69,2% for turbines. These activity factors are within the error bounds of the
activity factors calculated using TRANSDAT (horsepower) and FERC  (operating hours),
However, the TRANSDAT database and the FERC database are much more representative of
the transmission industry segment and were used in the overall emissions estimates.
                                        D-3

-------
 TABLE D-l. COMPRESSOR DRIVER DATA
FROM TRANSMISSION SEGMENT SITE VISITS
Site
TS1
TS2
TS3
TS4
TS5
TS6
TS7
TS8
TS9
TS10-1
TS10-2
TS10-3
TS104
TSIO-5
TS1G-6
TS10-7
TS10-8
TSiO-9
TS10-10
TS10-11
TS10-12
TS10-13
TS10-14
TS10-15
TS10-16
TS10-17
TS10-18
TS10-19
TS 10-20
TS 10-21
No. of
Engines
13
0
0
2
10
12
0
4
7
4
4
5
3
4
2
1
2
2
3
3
0
10
7
2
8
2
1
19
0
5
Installed
Engine hp
32650
0
0
800
14560
17570
Q
14000
14000
6900
8000
1340
815
8400
1.200
2400
4090
1330
2475
4050
.,
13000
11200
4800
10600
1200
2400
32800
—
850
No. of
Turbines
0
2
2
0
2
1
1
2
3
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
Installed
Turbine hp
—
6900
6900
—
20090
12090
40000
26500
37500
—
—
—
—
...
—
—
—
—
—
—
1000

—
—
—
—
—
—
9300
—
MMcfd
843
200
37.5
50
NA
NA
NA
4.9
4.9
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
                                             Continued
                  D-4

-------
                                 TABLE D-l,
                                 (CONTINUED)
Site
TS10-22
TS10-23
TS 10-24
TS 10-25
TS10-26
TS10-27
TS 10-28
TS10-29
TS 10-30
TS10-31
TS10-32

TS 10-34
TS 10-35
TS10-36
TS 10-37
TS10-38
TSIO-39
TSII
TSI2
TS13
TOTAL
AVERAGE0
No, of
Engines
1
1
3
1
I
7
2
0
1
1
1
3
7
2
1
1
11
1
6
7
13
208
Installed
Engine hp
2400
2400
SOOO
2400
2400
7000
2TOO
,.,
4800
3350
2:400
6000
7510
4000
2400
270
12380


10400
24800
334240
6554
No, of
Turbines
0
0
0-.
0
0
2
0
1
0
1
0
0
0
0
0
0
3
e;
0
o:.
0:
22:
Installed
Turbine hp

...
—
	
—
8500-
—
4250
—
4250
_
—
—
...
—
—
12750
—
—
0
0
190030
3726
MMcfd
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
140
NA
1409.5
HA = Not Available
'Based on sile averages
                                      D-5

-------
TABLE D-2. ANNUAL OPERATING HOURS FOR
         TRANSMISSION SEGMENT
Site
TS1
TS2
TS3
TS4
TS5
TS6
TS7
TS8
TS9
TS10-1
TS10-2
TS10-3
TS10-4
TS10-5
TS10-6
TS10-7
TSiO-8
TSIO-9
TS10-10
TS10-11
TS10-12
TSIO-I3
TS10-14
TS10-15
TS10-16
TS10-17
TS10-18
TS10-19
TS 10-20
TS10-21
TS10-22
No. of
Engines
1.3.
0
0
2
10
12
0
4
7
4
4
5
3
4
2
1
2
2
3
3
0
10
7
2
8
2
1
19
0
5
1
Average
Operating Hours
3303
—
—
7884
4066
2082
—
6366
6813
5019
3020
2343
0
4310
0
5914
3150
35
1574
2307
—
3653
52
158
2179
1351
4576
6117
—
470
5695
No. of
Turbines
0
2
2
0
2
1
1
2
3
0
0
0
0
0
0
0
0
o :
0
0
i
0 ;
0
0
0
0
0
0
1
0
0
Average
Operating Hours
•
! 3066
: 1139
:
3420
6314
0
4517
4698
	
_.
..
..
—
—
—
	
„_
—
—
0
„
	
„
„
—
—
..
579
__
,
                                                Continued)
                   D-6

-------
D
p
p
p
p
•p
p
p
p
p
 TABLE D-2,
(CONTINUED)
Site
TS10-23
TS1G-24

TS10-2S
TS10-27
TS10-28:
TS10-29
TSIO-30
TS10-31
TS10-32
TS10-33
TS 10-34
TS10-3S
TS10-36
TS10-37
TS10-38
TS10-39
TOTAL
AYERAGI1
90% LIMIT -a
No, Of
Engines
]
3
1
1
7
2:
i
2:
I
1
3
7
2:
1
1
11
1
182-
Average
" Operating |f 0%rs
2747
570.5
502,9
688J:
557
95
_
615
1015
497
476?
509
736
474.5
360
1240
586,9
2343
±
No. of '
' TuritinW
0
0
0
0
2
0
1
0
1
0
0
0
0
0
0
3
0
22
<. Average
Operating HOIOT
__
„„
**.
«
557
„«,
1823
„
1015
„
—
_
-.
..
	
497
„
2125
±48,1%
               "Based on site averages.
                                                      D-7

-------
 APPENDIX E
Conversion Table
      E-l

-------
                               Unit Conversion Table
    1 scf methane
    1 Bsef methane
    1 Bsef methane
    1 Bsef
    1 short ton (ton)
    1 Ib
    1 ft3
    1 ft3
    1 gallon
    1 barrel (bbl)
    1 inch
    1 ft
    1 mile
    1 hp
    1 hp-kr
    1 Btu
    1 MMBtu
    1 Ib/MMBtu
    T(°F)
    1 psi
English (o Metric Conversions

  19.23 g methane
  0,01923 Tg methane
  19,230 metric tonnes methane
  28.32 million standard cubic meters
  907.2 kg
  0.4536 kg
  0.02832 m3
  28.32 liters
  3.785 liters
  158.97 liters
  2.540 cm
  0.3048 m
  1.609 km
  0.7457 kW
  0.7457 kW-hr
  1055 joules
  293  kW-hr
  430  g/GJ
  1.8 T (°C)  + 32
  51.71 mm Hg
                            Global Warming Conversions
Calculating carbon equivalents of any gas:
                MMTCE ~ (MMT of gas) x
               /'MW, carbon\
               \  MW, gas  J
x (GWP)

-------
Calculating CD.2 equivalents for methane;
                                                  /MW, CO,\
             MMT of CO2 equiv.  = (MMT CH4) x	  x (GWP)
                                                  I MW,
where MW (molecular weight) of CO2 = 44, MW carbon = 12,  and MW CH4= 16.
Notes

scf


Bscf


MMscf


Mscf


Tg


Giga (G)


Metric tonnes


psig


psia


GWP



MMT


MMTCE


MMT of CO2 eq.
Standard cubic feet.  Standard conditions are at 14.73 psia and 60°F.


Billion standard cubic feet (10* scf).


Million standard cubic feet,


Thousand standard cubic feet


Teragrara (1012 g).


Same as. "billion (109),


1000 kg,


Gauge pressure.


Absolute pressure (note psia = psig + atmospheric pressure).


Global Warming Potential of a particular greeiihouse gas for a given
time period.


Million metric tonnes of a gas.


Million metric tonnes, carbon equivalent.


Million metric tonnes, carbon dioxide equivalent.
                                          E-3

-------

J
      "S
•       -
8   -
 aj >ȣ
+J CLOJQ
 0£ = o
                  Reproduced by NTIS

                  National Technical Information Service
                  Springfield, VA  22161
        This report was printed specifically for your order
      from nearly 3 million titles available in our collection.


F0r economy  and  efficiency, NTJS does not maintain stock of its vast
collection of technical reports.  Rather, most documents are printed for
each order. Documents that are not In electronic format are reproduced
from master archival copies and  &ts the best possible  reproductions
available.  If you have any questions concerning this document or any
order you have placed with NTIS,  please call our Customer Service
Department at (703) 487-4660,

About NTIS
NTIS  collects  scientific,  technical, engineering, and business  re'ated
information — then  organizes,  maintains,  and  disseminates  that
Information in a variety of formats —from microfiche fe online services,
The NTIS collection of nearly 3 million titles includes reports describing
research  conducted  or sponsored  by federal agencies  and their
contractors;  statistical  and  business  information;   U.S.  military
publications; audiovisual products; computer software and electronic
databases developed by federal agencies; training tools;  and technical
reports prepared by research organizations worldwide.  Approximately
100,000 new  titles are added and indexed  into the  NTIS collection
annually.
                      For more information about NTIS products and services, call NTIS
                      at (703) 487-4650 and request the free NTIS Catalog of Products
                           and Services, PR-827LPQ, or visit the NTIS Web site
                                         http://www.rife.gov
                                               NT®
                        Your indispensable resource for government-sponsored
                                   information—U.S, and worfefwrade

-------