EPA-600/R-96-020a
GRI-95/0270.1
February 1996
Criteria Pollutant Emissions from
Internal Combustion Engines
in the
Natural Gas Industry
Volume I. Technical Report
by:
Gerald S. Workman Jr., Rachel G. Adams, and
Gunseli Sagun Shareef
Radian Corporation
P.O. Box 13000
Research Triangle Park, NC 27709
EPA Contract No. 68-D2-0160 GRI Contract No. 5091-254-2293
Work Assignment No. 33
EPA Project Officer: GRI Project Manager
Charles C. Masser James M. McCarthy
Air Pollution Prevention and Control Division Environment and Safety Research
U.S. Environmental Protection Agency Gas Research Institute
Research Triangle Park, NC 27711 Chicago, JJL 60631
Prepared fan
Office of Air Quality Planning and Standards Office of Research and Development
U.S. Environmental Protection Agency U.S. Environmental Protection Agency
Research Triangle Park, NC 27711 Washington, DC 20460
-------
EPA Review Notice
This report has been peer and administratively reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information Service,
Springfield, Virginia 22161.
GRI Disclaimer
LEGAL NOTICE: This report was prepared by Radian Corporation as an
account of work sponsored by the Gas Research Institute (GRI). Neither 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.
-------
Abstract
This report contains emissions data for nitrogen oxides (NOJ, carbon monoxide
(CO), methane (CH<), ethane (C^H^), nonmethane hydrocarbons (NMHC), and nonmeth.'Jie*
ethane hydrocarbons (NMEHC) from stationary internal combustion (1C) engines and gas
turbines used in the natural gas industry. The emission factors were calculated from test
results based on five test campaigns conducted as part of the Gas Research Institute's air
toxics study, three of which were cofunded by the EPA. Test results for individual engines
tested are presented, along with full load engine family-specific factors, and the calculated
emissions factors are evaluated relative to the emission factors published in EPA report
AP-42. Units tested included eleven 2-stroke engines and five 4-stroke engines, with and
without controls, and two gas turbines. The data will enhance the current database in AP-42
for stationary 1C engines. It will not only enlarge the population of engine types covered,
but will enhance the emission factor quality of several engine categories which have a limited
data set.
ii
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate. EPA's research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
iii
-------
Executive Summary
Background
One function of the Air Pollution Prevention and Control Division (APPCD) of the
U.S. Environmental Protection Agency's (EPA's) Office of Research and Development is
improving current air pollutant emission inventory methodologies, especially for those
pollutants associated with tropospheric ozone formation. As part of the improvement to
emission inventory methodologies, APPCD supports field emission measurement efforts.
These data are used by EPA's Office of Air Quality Planning and Standards (OAQPS) to
enhance their reference document "Compilation of Air Pollutant Emission Factors" (AP-42),
which contains emission factors for oxides of nitrogen (NOJ, carbon monoxide (CO),
methane (CH,), ethane (CjH*), nonmethane hydrocarbon (NMHC), and nonmethane-ethane
hydrocarbon (NMEHQ emissions from the large, stationary internal combustion (1C)
reciprocating engines and turbines used in the natural gas industry. In AP-42, emission
factors for some types of engines, especially those with air pollution controls, are based on
an inadequate amount of emissions test data. To improve the understanding of emissions
from these sources, additional testing is needed to enhance the emissions database, giving
OAQPS the ability to revise AP-42.
Emissions characterization of 1C engines in the natural gas industry is currently
underway through a program sponsored by the Gas Research Institute (GRI), with the
primary focus on determining the potential for air toxics emissions. Since information on
NO,, CO, CH*, CJli, NMHC, and NMEHC emissions is needed to completely characterize
the 1C engine emissions, EPA/APPCD provided cofunding to the GRI program to support
gathering such data for enhancement of the emissions database currently used in AP-42 for
the development of emission factors. The work described in this document was conducted as
part of this joint effort between GRI and EPA and involved the following:
• Field measurements of NOt, CO, CH<, CjH,, and total hydrocarbon (THC)
emissions at three test sites (GRI Campaigns 4, 5, and 6);
IV
-------
• Incorporation of field data collected at two earlier test sites (Campaigns 2
and 3) by GRI into the data set for evaluation; and
• Evaluation of all test data for use in enhancing the emissions database
currently in AP-42.
Results
Table S-l presents a summary of full load emission factors for NOX, CO, CH«,
THC, NMHC, and NMEHC expressed in grams per horsepower-hour (g/hp-hr) and pounds
per million British thermal units (Ib/MMBtu). The emission factors were averaged by engine
family, and are presented for 2-stroke, lean-bum; 2-stroke, clean-burn; 4-stroke, lean-bum;
4-stroke, clean-bum; and 4-stroke, rich-bum engines; and gas turbines. Separate emission
factors were calculated for engines using emission control equipment, e.g., nonselective
catalytic reduction (NSCR), or selective catalytic reduction (SCR), CO oxidation catalyst, or
pre-combustion chamber (PCQ. Only data from test periods during which the engines were
operated within 90 percent of rated load and 95 percent of rated speed were used to calculate
the average emission factors, except when the engine tested was the only one of a particular
classification included in the test program, and the engine did not meet the minimum load
and speed criteria during any of the test periods.
Oxides of nitrogen, CO, and THC emission factors are based on continuous emissions
monitoring system (CEMS) measurements while the methane and ethane emission factors are
based on gas chromatography (GQ. Emission factors expressed as NMHC and NMEHC are
calculated by subtracting the methane and methane/ethane concentrations, respectively, from
the THC concentrations. In some cases, the difference between the measured THC and
methane/ethane concentrations was less than the analytical precision of the instruments. In
these cases, NMHC/NMEHC emissions were not quantified.
Except for the 2-stroke, lean-burn engine family, the information presented in
Table S-l is considered limited since the emission factors are based on tests conducted on
only one to three engines/turbines. As expected, there arc differences between the emission
factors calculated in this study and those in AP-42. The differences between the data from
this study and AP-42 can be attributed to the variability associated with the population of
engines tested, and difference* in the type of instrumentation used during the two studies.
-------
Table S-1. Full Load Average Emission Factors
Engine Family
2-strcke; lean-burn
2-stroke; clean-bum
4-stroke; lean-burn
4-*troke; clean-burn
4-stroke; rich-burn
Gas turbine
Emission
Control
—
—
CO
catalyst
—
SCR
catalyst
PCC
~
NSCR
catalyst
•™*
No. of Engines/
Runs*
7/16
1/3
1/1
3/6
1/2
1/1
le/l
lf/2
2/4
Units
(g/hp-hr)b
Ob/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
Ob/MMBtu)
(g/hp-hr)
Ob/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
Ob/MMBtu)
(g/hp-hr)
Ob/MMBtu)
NO.
14
3.4
0.48
0.14
0.54
0.17
14
3.7
5.0
1.3
0.56
0.14
18
5.2
0.050
0.015
1.4
0.31
CO
0.63
0.15
1.4
0.41
0.11
0.030
0.83
0.21
0.43
0.11
2.0
0.51
15
4.2
0.26
0.075
0.162
0.0382
CH<
4.6
I.I
NA
NA
NA
NA
5.5
1.5
NA
NA
NA
NA
NA
NA
NA
NA
ND
ND
CA
0.31
0.077
0.38C
0.1 lc
NA
NA
0.16
0.044
0.15
0.036
NA
NA
NA
NA
NA
NA
ND
ND
THC
5.7
1.4
6.8
2.0
6.3
1.9
4.1
1.1
2.7
0.69
8.0
2.0
3.0
0.85
1.7
0.49
ND
ND
NMHC
1.1
0.28
_c
_c
_c
_c
-c.d
_c.d
_c,d
_c,d
_c
_c
_c
_c
_c
_c
ND
ND
NMEHC
0.80
0.19
_c
_c
_c
_c
-c.d
_c,d
-c.d
_c,d
_c
_c
_c
_c
_c
_c
ND
ND
NA •* Not available. ND - Not detected. NSCR >• nonselective catalytic reduction. SCR =• selective catalytic reduction.
PCC - Pre-combu*Uon chamber.
*For some poUutants, the number of engines/runs used in the average is less than the total number tested.
There is uncertainty in the horsepower measuremrnU made by the engine analyst for 4 of the 16 runs.
CGC hardware malfunction during Campaign 4 prevented collection of data for methane and/or ethane.
Difference between recorded methane and THC measurements was less than the precision of either instrument.
"Based on one engine tested at 91 percent speed and below 90 percent load.
Based on one engine tested at 90 percent speed.
*Te*t results below the detection limits were averaged as zero.
-------
Conclusions
Based on examination of the test results from this study, the following conclusions are
offered to enhance the emissions database currently in AP-42:
• Incorporate emissions data used to develop the emission factors for
uncontrolled 2-stroke, lean-burn; 4-stroke, lean-burn; and 4-stroke, rich-burn
engines; and gas turbines into the current AP-42 emissions database. Although
the current factors are "A" quality, incorporation of these data will broaden
the population of the engines covered.
• Incorporate the data used to develop the emission factors for 2-stroke,
clean-bum engines into the current AP-42 emissions database. The current
AP-42 factors are "C" quality. The additional data may upgrade the emission
factor quality rating for this category.
• Use data for the NSCR-controlled 4-stroke, rich-burn engine, PCC-controUed
4-stroke, lean-bum engine, and the 2-stroke, clean-burn engine with a CO
oxidation catalyst to build and/or improve an emissions database for these
categories.
• The current version of AP-42 has separate emission factors for "clean-burn"
and "PCC" controlled engines. "Gean-burn" is a trade name used by one
manufacturer to describe modifications to a lean-burn engine to lower
emissions. A PCC is a primary component of the "clean-burn" modification to
these engines. An engine equipped with PCC may also have all of the other
clean-burn modifications, as did the one engine with PCC tested under this
program. Consideration should be given to combining the emissions database
for these control scenarios under a single generic description.
vii
-------
Contents
Volume I
Executive Summary iv
Figure xi
Tables xi
Acronyms and Abbreviations xiv
Section 1.0 Introduction 1-1
1.1 Background 1-1
1.2 Objectives and Approach 1-2
1.3 Report Contents 1-3
Section 2.0 Summary of Emission Factors 2-1
2.1 Engine Families 2-1
2.1.1 2-Stroke Engines 2-1
2.1.2 4-Stroke Engines 2-4
2.1.3 Gas Turbines 2-6
2.2 Full Load Emission Factors 2-6
2.2.1 Emission Factors 2-6
2.2.2 Test Engines/Turbines 2-9
2.2.3 Operating Data 2-10
2.3 Engine Family-Specific Emission Factors 2-10
Section 3.0 Test Results 3-1
3.1 Overview 3-1
3.2 Campaign 2 3-5
3.2.1 Site Description 3-5
3.2.2 Operating Conditions and Measurement of Oj, CO,, CO,
THC, and NO, 3-5
3.3 Campaign 3 3-6
3.3.1 Site Description 3-6
3.3.2 Operating Conditions and Measurement of Oj, COj, CO,
THC, and NO, 3-6
3.3.3 Measurement of Methane and Ethane Emissions 3-10
3.4 Campaign 4 3-14
3.4.1 Site Description 3-14
3.4.2 Operating Conditions and Measurement of Oj, COj, CO,
THC, and NO, 3-14
viu
-------
Contents (continued)
3.4.3 Measurement of Methane and Ethane Emissions 3-23
3.5 Campaign 5 3-24
3.5.1 Site Description 3-24
3.5.2 Test Results 3-24
3.6 Campaign 6 3-26
3.6.1 Site Description 3-26
3.6.2 Engine Operating Conditions and Measurement of Oj,
CO2) CO, THC, and NO, 3-26
3.6,3 Measurement of Methane and Ethane Emissions 3-33
Section 4.0 Sampling and Analytical Methods 4-1
4.1 Sampling Traverse Point Determination 4-3
4.2 Exhaust Gas Flow Rate 4-3
4.3 Exhaust Gas Molecular Weight 4-3
4.4 Exhaust Gas Moisture Content 4-4
4.5 Fuel Gas Composition and Heating Value 4-4
4.6 Exhaust Gas Composition 4-4
4.6.1 Sample Gas Extraction and Transfer 4-5
4.6.2 Calibration and Quality Control Standard Delivery 4-5
4.6.3 Measurement of O^, COj, CO, THC, and NO,
Concentrations 4-5
4.6.4 On-Site GC Analysis 4-6
Section 5.0 Quality Assurance/Quality Control and Documentation 5-1
5.1 Process Data Quality 5-1
5.2 Continuous Emission Monitors Data Quality 5-2
5.2.1 CEMS Calibration 5-2
5.2.2 CEMS Drift Checks 5-2
5.2.3 CEMS Bias Checks 5-3
5.2.4 CEMS Precision 5-4
5.2.5 Other CEMS QC 5-4
5.3 Manual Sampling Methods Data Quality 5-5
5.4 Method 18 Data Quality 5-5
Section 6.0 Evaluation and Comparison of Emission Factors 6-1
6.1 Uncontrolled Engines/Gas Turbines 6-1
6.2 Controlled Engines 6-3
6.3 Other 6-5
6.4 Conclusions 6-6
Section 7.0 References 7-1
ix
-------
Contents (continued)
Volume n
Appendices
A Sample Calculations A-l
B CEMS and Operation Data B-l
C CEMS Summary Data C-l
D GC Data D-l
E GC Summary Data E-l
F Raw Operating Data F-l
G Manual Measurement Data G-l
H Fuel Analysis ResulU H-l
I Quality Assurance and Quality Control (QA/QQ M
-------
Figure
4-1 Measurement System Schematic 4-2
Tables
S-l Full Load Average Emission Factors vi
2-1 Engine* Tested in Each Family 2-2
2-2 Summary of Emission Factors 2-7
2-3 Average Emission Factors 2-11
2-4 Emission Factor Range (g/hp-hr) 2-13
3-1 Engines Tested 3-2
3-2 Station 3A: Engine Operating Conditions and CEMS Results, Engine 102,
Cooper-Bessemer LSV-16 3-8
3-3 Station 3A: Engine Operating Conditions and CEMS Results, Engine 101,
Cooper-Bessemer LSV-16 3-9
3-4 Sweet Gas Plant 3B: Engine Operating Conditions and CEMS Results,
Engine 4, Cooper-Bessemer GMWA-8 3-11
3-5 Sweet Gas Plant 3B: Engine Operating Conditions and CEMS Results,
Engine 2, Cooper-Bessemer GMV-10TF 3-12
3-6 Station 3A: GC Results 3-13
3-7 Sweet Gas Plant 3B: GC Results 3-13
3-8 Campaign 4: Operating Conditions and CEMS Results, Waukesha L7042GU
(Tests 1 and 2) 3-15
xi
-------
Tables (continued)
3-9 Campaign 4: Operating Conditions and CEMS Results, Waukesha L7042GU
(Test 3) 3-16
3-10 Campaign 4: Operating Conditions and CEMS Results, Engine 4,
Cooper-Bessemer GMVC-10C (Tests 7, 9, and H) 3-17
3-11 Campaign 4: Operating Conditions and CEMS Results, Engine 4,
Cooper-Bessemer GMVC-10C (Test 8) 3-18
3-12 Campaign 4: Operating Conditions and CEMS Results, Engine 8,
I-R KVS-412 (PCQ 3-19
3-13 Campaign 4: Operating Conditions and CEMS Results, Engine 9,
I-R KVS-412 (Tests 14 and 17) 3-20
3-14 Campaign 4: Operating Conditions and CEMS Results, Engine 9,
I-R KVS-412 (Tests 15 and 16) 3-21
3-15 Campaign 4: GC Results (Ethane) 3-23
3-16 Campaign 5: Operating Conditions and CEMS Results, Westinghouse 191 ... 3-25
3-17 Station 6A: Operating Conditions and CEMS Results, Engine 10,
Clark BA-5 3-27
3-18 Station 6B: Operating Co' ''lions and CEMS Results, Solar Taurus T-6502 . . 3-29
3-19 Station 6C: Operating Conditions and CEMS Results, Engine 15,
Cooper-Bessemer GMVC-10 3-30
3-20 Station 6C: Operating Conditions and CEMS Results, Engine 11,
Cooper-Bessemer GMVC-10 3-31
3-21 Station 6C: Operating Conditions and CEMS Results, Engine 9,
Cooper-Bessemer GMVA-10 3-32
3-22 Station 6C: Operating Conditions and CEMS Results, Engine 13,
Cooper-Bessemer GMWC-10 3-34
3-23 Campaign 6: GC Results 3-35
4-1 Target Parameters and Measurement Methods 4-1
xii
-------
Tables (continued)
6-1 Emission Factors for Uncontrolled Natural Gas Prime Movers:
Gas Turbines and 2-Stroke Engines 6-2
6-2 Emission Factors for Uncontrolled Natural Gas Prime Movers:
4-Stroke Engines 6-2
6-3 Emission Factors for Controlled Natural Gas Prime Movers:
NSCR On 4-Stroke Rich-Burn Engines 6-4
6-4 Emission Factors for Controlled Natural Gas Prime Movers:
SCR On 4-Stroke Lean-Burn Engines 6-4
6-5 Emission Factors for Controlled Natural Gas Prime Movers:
"Clean Burn" On 2-Stroke Lean-Burn Engines 6-5
6-6 Emission Factors for Controlled Natural Gas Prime Movers:
"Pre-combustion Chamber (PCQ" On 4-Stroke Lean-Burn Engines 6-6
6-7 Emission Factors for Controlled Natural Gas Prime Movers:
"Clean Bum" and CO Catalyst On 2-Stroke Lean-Burn Engines 6-6
xiii
-------
Acronyms and Abbreviations
APPCD Air Pollution Prevention and Control Division
ASSET Air Sciences and Engine Technology, Inc.
BMEP Brake mean effective pressure
C-B Cooper-Bessemer
CEMS Continuous emissions monitoring system
CO Carbon monoxide
CH< Methane
CjHa Ethane
EPA Environmental Protection Agency
Frf F factor (dry basis)
FID Flame ionization detector
g/hp-hr Gram per horsepower-hour
GC Gas chromatography
GC/MD Gas chromatography with multiple detectors
GPA Gas Processors Association
Gkl Gas Research Institute
HHV Higher heating value
hp Horsepower
1C Internal combustion
I-R Ingersoll-Rand
LHV Lower heating value
Ib/MMBtu Pounds per million British thermal units
NMEHC Non methane-ethane hydrocarbon
NMHC Non methane hydrocarbon
NO, Oxides of nitrogen
NSCR Nonselective catalytic reduction
PCC Pre-combustion chamber
QA Quality assurance
QAPP Quality assurance project plan
QC Quality control
SCR Selective catalytic reduction
THC Total hydrocarbon
xiv
-------
Section 1.0
Introduction
1.1 Background
One function of the Air Pollution Prevention and Control Division (APPCD) of the
U.S. Environmental Protection Agency's (EPA's) Office of Research and Development is
improving current air pollutant emission inventory methodologies, especially for those
pollutants associated with tropospheric ozone formation. As part of the improvement to
emission inventory methodologies, APPCD supports field emission measurement efforts.
These data are used by EPA's Office'of Air Quality Planning and Standards (OAQPS) to
enhance their reference document "Compilation of Air Pollutant Emission Factors" (AP-42),
which contains emission factors for oxides of nitrogen (NOJ, carbon monoxide (CO),
methane (CH4), ethane (C^Hj), nonmethane hydrocarbon (NMHC), and nonmethane-ethane
hydrocarbon (NMEHC) emissions from the large, stationary internal combustion (1C)
reciprocating engines and turbines used in the natural gas industry.' In AP-42, emission
factors for some types of engines, especially those with air pollution controls, are based on
an inadequate amount of emissions test data. To improve the understanding of emissions
from these sources, additional testing is needed to enhance the emissions database, giving
OAQPS the ability to revise AP-42.
Emissions characterization of 1C engines in the natural gas industry is currently
underway through a program sponsored by the Gas Research Institute (GRI), with the
primary focus on determining the potential for air toxics emissions. Since information on
NOX, CO, CH4, CjHj, NMHC, and NMEHC emissions is needed to completely characterize
the 1C engine emissions, EPA/APPCD provided cofunding to the GRI program to support
gathering such data for enhancement of the emissions database currently used in AP-42 for
the development of emission factors. The work described in this document was conducted as
part of this joint effort between GRI and EPA.
1-1
-------
1.2 Objectives and Approach
The primary objectives of this study were to:
• Characterize emissions of NO,, CO, CH4, CjH*. and total hydrocarbons
(THCs) from 1C engines including turbines;
• Evaluate the emissions data for use in enhancing the emissions database
currently in AP-42.
The scope of this joint effort covered measurements conducted as part of the following field
campaigns:
• Campaign 4--Compressor station (four engines);
• Campaign 5-Sweet gas plant (one turbine);
• Campaign 6A--Compressor station (two engines);
• Campaign 6B--Compressor station (one turbine); and
• Campaign 6C--Compressor station (four engines).
Additionally, field data collected as part of previous GRI efforts are also included in this
document:
• Campaign 2--Sour gas plant (two engines);
• Campaign 3A--Compressor station (two engines); and
• Campaign 3B--Sweet gas plant (two engines).
The host sites for the field measurements were selected according to criteria
developed for the GRI program. These criteria included engine make/model, family (type),
size, age, presence of controls, and operating load flexibility to ensure the data collected
would be applicable to a broad population of units in the industry. Emissions data collection
and reduction for Campaigns 4,5, and 6 were conducted according to procedures
documented in the Quality Assurance Project Plan (QAPP) prepared for EPA and the test
1-2
-------
plans prepared for GRI and EPA. During the engine testing, the GRI prcgiam engine
consultant, Jon Tice of Air Sciences and Engine Technology, Inc., and engine analyst(s)
were on-sile to ensure the operation of the engines being tested was satisfactory and to
measure and confirm engine operating data (e.g., horsepower, fuel flow).
For Campaigns 2 and 3, emissions data collection and reduction were performed
according to test plans prepared for GRI, similar to those prepared for both GRI and EPA in
Campaigns 4, 5, and 6. Engine horsepower measurements were performed by host site
engine analysts for Campaigns 2 and 3B, with the GRI program's engine consultant (Jon Tice
of ASSET) providing initial engine operation assessment for Campaign 3.
1.3 Report Contents
Section 2.0 presents an overview of broad engine categories, followed by a summary
of emission factors calculated from the test data tabulated according to engine classification.
Section 3.0 gives detailed test results for each engine characterized as part of this effort,
including descriptions of the test sites. Descriptions of the test methods used during the
measurement campaigns are included in Section 4.0, with the summary of the quality
assurance (QA) and quality control (QC) procedures used, and documentation of the data
quality indicators presented in Section 5.0. Section 6.0 presents the average emission factors
calculated from the test data by engine classification including a comparison of these factors
with current AP-42 emission factors. Evaluation of the data for enhancing the emissions
database used in AP-42 to improve the emission factors for large, stationary internal
combustion engines is also included in Section 6.0. Finally, Section 7.0 lists the references,
and supporting data are presented in Volume II of this document in Appendices A through I.
1-3
-------
Section 2.0
Summary of Emission Factors
2.1 Engine Families7-1
Natural gas-fired reciprocating engines can be classified into five broad categories or
"families" according to design differences which may lead to differences in emission
characteristics. These families include:
• 2-stroke; lean-bum;
• 2-stroke; clean-bum;
• 4-stroke; lean-bum;
• 4-stroke; clean-bum; and
• 4-stroke; rich-bum.
Following is a brief description of the engine families, with each family composed of units
that share typical engine power cycles, air-to-fuel (A/F) ratios, and combustion and exhaust
temperatures. Table 2-1 presents the engines tested in this study by engine family. In
addition, two gas turbines, a Westinghouse 191 and a Solar Taurus T-6502, were also tested
in this study.
2.7.7 2-Stroke Engines
A 2-stroke engine completes the power cycle in one revolution of the crankshaft. In
the first stroke, air or an air and fuel mixture is drawn or forced into the cylinder as the
piston begins the compression stroke. Near the end of the compression stroke, the mixture is
ignited, which forces the piston downward through the cylinder and begins the second stroke.
2-1
-------
Table 2-1. Engines Tested in Each Family
Air Scavenging
Turbochargcd
Blower scavenged
Piston scavenged/
Naturally aspirated
2-Stroke
Lean-burn
Cooper GMVC-10 (2)
Cooper CMWC- 10
Cooper GMVA-10 (2)
Cooper GMWA-8
Clark BA-5
Clark BA-6
Clark HBA-5
Cooper GMV-10TF
2-Stroke
Clean-burn
Cooper GMVC-10C*
_b
_b
4-Stroke
Lean-burn
Cooper LSV-16 (2)
Ingcrsoll Rand KVS-412*
_d
_d
4-Stroke
Clean-hurn
Ingersoll Rand KVS^I2
_b
_b
4-Stroke
Rich-burn
—c
_f
Waukesha L7042GUK
K>
K>
'Equipped with CO oxidation catalyst.
All clean-bum engines are turbocharged.
cEquipped with selective catalytic reduction control.
No engines of this design were identified.
*No eoginea of this design were tested.
4-stroke, rich-burn engines do not utilize scavenging air.
'Equipped with nonselective catalytic reduction control.
-------
During the second stroke, power is transferred to the crankshaft. As the piston continues to
move downward, the piston passes and uncovers exhaust ports (or exhaust valves open), and
the combustion gases exit. Intake ports then open, and the fresh fuel and air mixture is
forced into the cylinder, displacing the remainder of the combustion gases. Finally, the
exhaust ports are closed, and the cycle begins again.
Because scavenging air is used to sweep the cylinder of exhaust gases, 2-stroke
engines operate with an overall A/F ratio that is greater than the stoichiometric ratio. This is
also referred to as a fuel-lean condition. As such, 2-stroke engines are classified as having
lean-burn combustion. Newer model 2-stroke engines arc designed to utilize turbochargers
and high-energy ignition systems to achieve stable combustion at even higher A/F ratios.
The high A/F ratio lowers bulk combustion temperatures and, thereby, reduces NO,
formation. Due to the reduced NO, levels, these models are commonly called "clean-burn"
engines.
2-Stroke, Lean-Burn Engines
A lean-burn engine is classified as one with an A/F ratio operating range that is
greater than stoichiometric, and cannot be adjusted to operate with an exhaust O2
concentration of less than 1 percent. A/F mass ratios for lean-burn engines range from 20:1
to 60:1, with stack temperatures normally ranging from 550 to 850°F. All 2-stroke engines
and 4-stroke, scavenged, turbocharged engines operate under lean-bum conditions due to
scavenging air; however, some engines may have fuel-rich combustion zones.
The higher air content in lean-burn combustion increases the heat capacity of the
mixture in the combustion chamber, which lowers combustion temperatures and generally
results in increased THC emissions due to the high quench volume in the cylinder.
All 2-stroke, lean-bum engines are direct-injected (i.e., fuel is injected directly into
the cylinder) and experience nonuniform mixing of the air and fuel prior to combustion.
Therefore, thermal and concentration gradients are more prominent in the combustion
chamber for 2-stroke engines than 4-stroke engines which have carbureted (prc-mixed) fuel
delivery systems. Because of the potential for nonuniform mixing of the air and fuel,
2-stroke engines tend to have higher THC levels than 4-stroke carbureted engines.
2-3
-------
2-Stroke, Clean-burn Engines
Clean-burn engines use turbochargers to force more air into the combustion chamber,
with the increased A/F ratio reducing bulk gas temperatures and combustion temperatures,
resulting in lower NOX formation. However, the reduced temperatures can also increase
THC and CO emissions.
Engines with large cylinder bores and conventional ignition systems cannot reliably
ignite and sustain combustion at the higher A/F ratios used in clean-bum designs. In these
cases, a pre-combustion chamber (PCC) design is utilized. Although PCC engine designs
vary among manufacturers, the PCC is typically a small volume antechamber in which a
fuel-rich mixture is ignited. The ignited mixture from the PCC propagates into the main
cylinder and ignites a very lean combustion charge. The exit velocity of the combustion
products from the PCC has a torch-like effect that creates multiple ignition fronts and
promotes mixing in the main chamber. Both of these factors create a more stable and cooler
temperature profile in the main combustion chamber with a PCC design than with an
open-chamber design. Although the lower temperatures and leaner A/F ratios reduce NOX
emissions, they may result in higher levels of THC in the exhaust stream of a PCC engine.
2.1.2 4-Stroke Engines
A 4-stroke engine completes the power cycle in two full revolutions of the crankshaft.
During the intake stroke, the downward motion of the piston draws air into the cylinder.
The second stroke compresses the air, or air and fuel mixture, and begins to increase
cylinder temperatures. The third-stroke begins with ignition of the gases, which causes the
gases to expand, driving the piston downward and delivering power to the crankshaft.
Finally, the piston moves upward and forces the exhaust gases out of the cylinder. Four-
stroke engines are available in three basic configurations:
• 4-stroke, rich-burn;
• 4-stroke, lean-bum; and
• 4-stroke, clean-bum.
2-4
-------
4-Stroke, Rich-Burn Engines
Rich-bum engines operate with an A/F ratio that is near stoichiometric (approximately
16:1 to 20:1), or fuel-rich, and have an exhaust O2 concentration ranging from nearly zero to
about five percent. Rich-bum engines include all naturally aspirated and non-scavenged,
turbocharged 4-stroke engine models. Because of the low levels of O2 present, combustion
temperatures and consequently exhaust temperatures are higher than for lean-burn engines.
Exhaust temperatures for rich-burn engines typically range from 1,000 to 1,250°F.
4-Stroke, Lean-Burn Engines
Four-stroke, lean-bum engines are available in two basic designs: direct injected and
pre-mixed (carbureted or port injected). The conditions in the combustion zone for these two
designs can be very distinct. The direct injected 4-stroke, lean-bum engines have a hot
combustion zone before the flame front mixes with the remainder of the combustion air.
This hot zone is similar to conditions present in a rich-bum engine. The pre-mixed 4-stroke,
lean-bum engines combust a homogeneous air/fuel mixture which leads to a cooler
combustion zone, similar to a 2-stroke engine.
The additional mixing in 4-stroke engines reduces the presence of high concentration
and temperature gradients in the cylinder during combustion when compared to 2-stroke
engines. In addition, the -esidence time of combustion products in the cylinder of a 4-stroke
engine is up to twice that of a 2-stroke engine operating at the same speed. The longer
residence time at elevated temperatures typically results in lower THC emissions compared to
2-stroke, lean-burn engines.
4-Stroke, Clean-burn Engines
As with 2-stroke engines, newer model 4-stroke engines are frequently designed with
very high A/F ratios to minimize NO, formation. Four-stroke, clean-bum engines can be
classified into two subcategories: injected and carbureted.
Four-stroke, clean-burn injected engines are characterized by either a direct-injected
or port-injected fuel delivery system. Compared to the carbureted air/fuel design, the
injected clean-burn design is expected to have higher fuel concentration gradients, leading to
nonuniform temperature distribution in the combustion chamber.
2-5
-------
Four-stroke, carbureted engines are characterized by pre-mixing the air and fuel prior
to charging the combustion cylinder. Because of the homogeneity of the A/F mixture during
the combustion process, the pre-mix design provides a relatively uniform combustion
temperature profile. Compared to uncontrolled rich- or lean-bum, 4-stroke engines, the
carbureted clean-bum design exhibits lower combustion and exhaust temperatures due to the
higher A/F ratio.
2.1.3 Gas Turbines*
A gas turbine is an internal combustion engine which uses rotary rather than
reciprocating motion to generate shaft horsepower. Three primary sections are present in gas
turbines: the compressor, the combustor, and the turbine. The compressor draws in ambient
air, compresses it with a compression ratio of up to 30:1, and directs the compressed air into
the combustion zone. Fuel is injected and combusted in the combustor. Flame temperatures
can reach 3,6008F; however, additional ambient air is quickly added to reduce temperatures
to around 2,000 to 2,300'F before the gases enter the turbine section. The turbine recovers
the energy released during combustion in the form of shaft horsepower.
Combustion in a gas turbine takes place under fuel lean conditions; however, due to
imperfect mixing, fuel rich zones frequently occur in the combustor. By maintaining overall
fuel lean conditions, NO, formation is minimized.
2.2 Full Load Emission Factors
2.2.1 EmissJon Factors
Table 2-2 presents a summary of emission factors for NOZ, CO, CH4, CjH^, THC,
NMHC, and NMEHC expressed in grams per horsepower-hour (g/hp-hr) and pounds per
million British thermal units (lb/MMBtu), based on the higher heating value of fuel. In cases
where individual engines were tested during more than one test period, an average emission
factor for the engine is reported in the table. Only data from test periods during which the
engines were operated within 90 percent of rated load and 95 percent of rated speed were
used to determine average emission factors to represent engines operated at or near full load
conditions. There are a few cases where data from engines tested at slightly lower loads or
speeds are included. In these cases, the engines tested were the only ones of their particular
2-6
-------
Table 2-2. Summary of Emission Factors
Campaign
Engine Make/Model
2-stroke; lean-burn
6
3
6
6
6
3
6
CUrk BA-5. Unit 10
Cooper-Bessemer GMV-10TF,
Unit 2
Cooper-Bessenier GMVA-10.
Unit 9
Cooper-Bessemer GMVC-10,
Unit 11
Cooper-Bessemer GMVC-10,
Unit 15
Cooper-Bessemer GMWA-8,
Unit 4
Cooper-Bessemer GMWC-10,
Unit 13
2-stroke; clean burn
4
4
Cooper-Bessemer GMVC-10C
(before CO catalyst), Unit 4
Cooper-Bessemer GMVC-10C
(after CO catalyst), Unit 4
4-stroke; lean-burn
3
3
Cooper-Bessemer LSV-16,
Unit 101
Cooper-Bessemer LSV-16,
Unit 102
No. of
Test
Periods
Load
(*)
Units
NO.
CO
2
3
2
3
2
1
3
94-100
104
100
95-101
102
98
91-96
(g/bp-br)
(Ib/MMBtu)
(S/hp-hr^
(Ib/MMBtu)
(g/hp-br)
(Ib/MMBtu)
(g/hp-br)
(Ib/MMBtu)
(g/hp-br)
(Ib/MMBtu)
(g/hp-hr^
(Ib/MMBtu)
(g/hp-hr)
Gb/MMBtu)
19
3.6
13
3.3
5.0
1.3
13
3.5
8.4
2.1
17
4.3
19
5.1
0.90
0.17
0.53
0.10
0.52
0.14
0.47
0.13
0.60
0.15
0.40
0.10
0.87
0.23
3
1
93-97
99
(g/hp-hr)
(Ib/MMBtu)
(g/hp-br)
Ob/MMBtu)
0.48
0.14
0.54
0.17
1.4
0.41
0.11
0.030
2
2
98-99n
98-101h
(g/hp-hr)
(Ib/MMBtu)
(g/bp-hr)
(Ib/MMBtu)
9.5
2.6
12
3.2
1.1
0.30
0.90
0.20
CH.
CA
THC
NMHC"
NMEHCb
3.8
0.73
2.4°
0.57e
5.0
1.3
3.8*
1.0»
3.6»
0.89r
NA
NA
7.6
2.0
0.13
0.025
0.86°
0.21°
0.24
0.060
0.19*
0.050f
0.1 8f
0.043f
NA
NA
0.34
0.087
6.1C
1.2C
3.5
0.87
6.4
1.7
4.4
1.2
5.0
1.3
6.0
1.5
8.8
2.3
1.8C
0.35C
1.1
0.27
1.4
0.35
0.56
0.14
1.5
0.37
-
-
1.1
0.29
1.7C
0.32C
0.23
0.054
1.1
0.29
0.35
0.091
1.3
0.32
-
—
0.76
0.20
NA
NA
NA
NA
0.38*
0.118
NA
NA
6.8
2.0
6.3
1.9
-
-
-
-
-
—
-
-
5.6
1.5
5.3
1.5
0.17
0.048
0.15
0.041
5.3
1.5
4.7
1.3
_i
_i
_i
_i
_i
_i
_i
_>
to
-------
Table 2-2. (Continued)
Campaign
4
4
Engine Make/Modd
Ingersoll-Rand KVS-412
(before SCR catalyst). Unit 9
Ingersoll-Rand KVS-412
(after SCR catalyst). Unit 9
4-stroke; clean burn
4
Ingersoll-Rand KVS-412. Unit 8
4-stroke; rich-bum
4
4
Waukesha L7042GU
(before NSCR catalyst)
Waukesha L7042GU
(after NSCR catalyst)
Gas Turbine
6
5
Solar Taurus T-65021
Westinghouse 19 lm
No. of
Test
Periods
2
2
Load
(*)
91
90
Units
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
NO.
22
5.4
5.0
1.3
CO
0.55
0.14
0.43
0.11
1
92
(g/hp-hr)
(Ib/MMBtu)
0.56
0.14
2.0
0.51
1
2
88*
92-95k
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
18
5.2
0.050
0.015
15
4.2
0.26
0.075
2
2
95
Full
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
1.3
0.30
1.5
0.32
ND
ND
0.33
0.075
CH,
NA
NA
NA
NA
C.H,
NA
NA
0.15J
0.036J
THC
2.5
0.64
2.7
0.69
NMHC"
-
-
-
-
NMEHCb
-
-
-
-
NA
NA
NA
NA
8.0
2.0
-
-
-
-
NA
NA
NA
NA
NA
NA
NA
NA
3.0
0.85
1.7
0.49
-
-
-
-
-
-
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
CO
NA = Not available. ND = Not detected. NSCR = nonselective catalytic reduction. SCR = selective catalytic reduction.
'Calculated as THC minus methane.
Calculated as THC minus methane and ethane.
'Emission factors based on Test Period 2 only.
Uncertainty in the horsepower measurements by the engine analyst for these runs.
cEmission factors based on Test Periods 2 and 3 (not 4).
Instrument drift exceeded specified limit.
gBased on GC data from Test Period 7.
. Rating based on operation at maximum brake mean effective pressure (BMEP).
'Difference between THC and methane measurement is less than precision of the instruments.
JBased on GC data from Test Period 15.
^
Engine spercl is 81 % of the rated speed.
Detection limits are as follows: CO: 1 ppm; CH4: 2 ppm; C:H4: 2 ppm; and THC: I ppm.
"Detection limits are as follows: CH4: 1 ppm; C,H6: 1 ppm; and THC: 10 ppm.
2-8
-------
family included in the test program. [Note: See Section 3.0 for details on the test data
presented in Table 2-2. Data for low load and low speed test conditions are included in the
appendices.]
Oxides of nitrogen, CO, and THC emission factors are based on continuous emissions
monitoring system (CEMS) data, whereas methane and ethane emission factors are based on
gas chromatography (GC). Emission factors for NMHC and NMEHC were calculated by
subtracting the methane and methane plus ethane concentrations from the THC
concentrations, respectively. In some cases, the difference between the measured THC and
methane/ethane concentrations was less than the analytical precision of one or both of the
instruments. This is not unusual for combustion sources where THC emissions are composed
of high fractions of methane and ethane. In these cases, NMHC and NMEHC emissions
were not quantified.
2.2.2 Test Engines/Turbines
The engines/turbines tested under Campaigns 2, 3B, and 5 were located at gas
processing plants, while the others were located at pipeline transmission/storage stations.
Discussions with industry representatives have indicated differences in operating and
maintenance practices at gas processing and transmission/storage stations which may impact
engine emission rates. Engines at gas processing plants tend to be run continuously
year-round, and are rarely shut down for maintenance unless engine problems are affecting
production rates. Engines at transmission/storage stations have more operating flexibility,
because the stations do not run at full capacity all year long, and most engines used at
transmission stations are subject to regular shutdowns which allow the opportunity for repairs
and preventive maintenance. Therefore, on average, engines used at transmission/storage
stations tend to be in better physical condition than those used at gas processing plants.
Neither of the two engines tested at a sour gas processing plant under Campaign 2
was included in the engine family emission factor averages. Test data from these two
engines were excluded from the averages because they were both running at less than
90 percent load, and one of the engines appeared to be operating especially poorly, as
described in Section 3.0.
Except for the 2-stroke lean-burn family, the information presented in Table 2-2 for
each category is limited, as it is based on test(s) conducted on only one to three engines.
2-9
-------
The, data for the 2-stroke lean-burn engine family are from seven engines, representing two
manufacturers and six models, encompassing a broader population of engines than
represented by the data for the other engine families. Two manufacturers and two models
(three engines) are represented with the data on 4-stroke lean-burn engine category. Of the
two 4-stroke lean-burn engine models tested, the Cooper-Bessemer LSV-16 engines use a
port injection system, while the Ingersoll-Rand KVS-412 engines use direct injection. This
difference in injection techniques accounts for some of the difference in emission levels
between these two engine models.
As shown in Table 2-2, three of the engines tested were equipped with.catalytic
emission controls: selective catalytic reduction (SCR) for NO, control, nonselective catalytic
reduction (NSCR) for THC, NO,, and CO control, and CO oxidation catalyst. The CO and
NSCR catalysts had recently been installed on the Cooper GMVC-10C and
Waukesha L7042GU engines, respectively.
2.2.3 Operating Data
For all but four engines in Table 2-2, the emission factors were calculated using
horsepower measurements performed by an engine analyst and exhaust flow rates derived
from fuel flow measurement data. The horsepower data were based on measurements of
actual pressure changes in each cylinder of the compressor. For the Cooper LSV-16 engines
in the 4-stroke lean-bum family, the horsepower data were calculated from site-specific
performance curves. Although the horsepower during the Cooper GMV-10TF and GMWA-8
engine testing was measured by an engine analyst, calculations based on fuel flow rate data
indicate the horsepower data may be about 10 to 15 percent high.
Gas turbine test data are based on one model from each of two turbine manufacturers.
The process data for the Solar turbine were provided by the host site, while the load during
the tests on the Westinghouse 191 turbine was estimated based on exhaust flow rate data and
discussions with the manufacturer.
2.3 Engine Family-Specific Emission Factors
The emission factors calculated from engine tests have been averaged by engine
family, as shown in Table 2-3. For the 2-stroke, clean-burn; 4-stroke, clean-burn; and
4-stroke, rich-burn categories, the emission factors are the same as those presented in
2-10
-------
Table 2-3. Average Emission Factors
Engine Family
2-stroke; lean-burn
2-stroke; clean-burn
4-stroke; lean-bum
4-stroke; clean-burn
4-stroke; rich-bum
Gas turbine
Emission
Control
—
—
CO
catalyst
—
SCR
catalyst
PCC
—
NSCR
catalyst
~~
No. of Engines/
Runs*
7/16
1/3
1/1
3/6
1/2
1/1
le/l
lf/2
2/4
Units
(g/hp-hr)b
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp>-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
(g/hp-hr)
(Ib/MMBtu)
NO,
14
3.4
0.48
0.14
0.54
0.17
14
3.7
5.0
1.3
0.56
0.14
18
5.2
0.050
0.015
1.4
0.31
CO
0.63
0.15
1.4
0.41
0.11
0.030
0.83
0.21
0.43
0.11
2.0
0.51
15
4.2
0.26
0.075
0.168
0.0388
CH,
4.6
1.1
NA
NA
NA
NA
5.5
1.5
NA
NA
NA
NA
NA
NA
NA
NA
ND
ND
CM.
0.31
0.077
0.38
0.11
NA
NA
0.16
0.044
0.15
0.036
NA
NA
NA
NA
NA
NA
ND
ND
THC
5.7
1.4
6.8
2.0
6.3
1.9
4.1
1.1
2.7
0.69
8.0
2.0
3.0
0.85
1.7
0.49
ND
ND
NMHC
1.1
0.28
_c
_c
_c
_c
_c,d
_c,d
_c,d
_c,d
_c
_c
_c
_c
_c
_c
ND
ND
NMEHC
0.80
0.19
_c
_c
_c
_c
_c.d
_c,d
_c,d
_c.d
_c
_c
_c
_c
_c
_c
ND
ND
NA = Not available. ND = Not detected. NSCR = nonselective catalytic reduction. SCR = selective catalytic reduction.
PCC = Pre-combustion chamber.
For some pollutants, the number of engines/runs used in the average is less than the total iiumber tested.
'There is uncertainty in the horsepower measurements made by the engine analyst for 4 of the 16 runs.
CGC hardware malfunction during Campaign 4 prevented collection of data for methane and/or ethane.
Difference between recorded methane and THC measurements was less than the precision of either instrument.
eBased on one engine tested at 91 percent speed and below 90 percent load.
Based on one engine tested at 90 percent speed.
*Test results below the detection limits were averaged as zero.
-------
Table 2-2, since only one engine was tested in each of these categories. For the 2-stroke,
lean-burn; 4-stroke, lean-burn; and gas turbine categories, the emission factors are based on
seven, three, and two units, respectively. In calculating the engine/turbine family average
emission factors, data from all test periods collected on units that fall in the particular family
were included. Table 2-4 provides an indication of the range of the emission factors for
individual test periods in these three categories. The largest range was observed for NO,
data in the 2-stroke, lean-burn category, where the most models were tested.
2-12
-------
Table 2-4. Emission Factor Range (g/hp-hr)
Pollutant
NO,
CO
CH4
QH.
THC
NMHC
NMEHC
2 -Stroke Lean-burn
(7 engines/ 16 runs)
4.9 -22 (14)
0.40 -0.94 (0.63)
2.3 - 8.0 (4.6)
0.090-0.88 (0.31)
3.5 - 9.2 (5.7)
0.29 - 1.8 (I.I)
0.087 - 1.8 (0.80)
4-Stroke Lean-burn
(3 engines/6 runs)
8.7 -22 (14)
0.51-1.1 (0.83)
4.7 -6.0 (5.5)
0.13-0.17 (0.16)
2.4 -5.3 (4.1)
- -
_ _
Gas Turbines
(2 turbines/4 runs)
1.2- 1.6(1.4)
ND- 0.35 (0.16)
ND
ND
ND
ND
ND
Note: The number in parenthesis is the average (mean) emission factor.
to
C*
-------
Section 3.0
Test Results
3.1 Overview
This section presents results from the individual engine/turbine tests conducted during
Campaigns 2 through 6 and provides a brief description of the host sites for these campaigns.
Campaigns 2 and 3 were conducted before EPA Work Assignment No. 33 was initiated.
Data from these two test campaigns, therefore, were collected using the procedures
established in the GRI program. No methane or ethane emissions data were collected during
Campaign 2, because such measurements were beyond the scope of the GRI program at that
time.
Table 3-1 describes the engines/turbines tested in this effort (e.g., make/model, size,
speed, family, controls) including information on the test runs conducted. All
engines/turbines listed in Table 3-1 were tested at 90 percent of full load (or higher) and
95 percent of full speed (or higher), except as noted in the table and explained in the text.
Only data that meet these criteria are included in this section, with low load and low speed
data presented in the appendices.
Full load for reciprocating engines was determined on the basis of torque. Torque at
full load was taken to be the manufacturer's rated power divided by the manufacturer's rated
speed except at Station 6A, where torque was based on the site-rated power and
manufacturer's rated speed. For the turbines, load was taken as a percentage of maximum
available power calculated as the manufacturer's rating, adjusted for site elevation and
temperature. For most test periods, stack gas flow rate was monitored via both EPA
Method 2 and EPA Method 19. EPA Method 2 may be unreliable when used on
reciprocating engines because of possible pulsations in the exhaust gas flow. EPA
Method 19 calculations, based on fuel flow to the engine/turbine and exhaust oxygen content,
were used to calculate emission rates except as noted in the tables and text.
3-1
-------
Table 3-1. Engines Tested
Campaign
2
2
3A
3A
3B
3B
Engine
Make/Model
Cooper-Bessemer*
GMVA-10
Clark BA-6«
Cooper- Bessemer
LSV-16
Cooper-Bessemer
LSV-16
Cooper-Bessemer
GMWA-8
Cooper-Bessemer
GMV-10TF
Unit
No.
2
1
102
101
4
2
Rated
Power
(M
1.350
1,140
4.200b
4.200b
2,000
1,100
Rated
Speed
(rpm)
300
300
327
327
250
300
Year
Installed
1966
1982
1957
1955
1958
1945
Engine Family
2-stroke, lean-burn; BS
2-stn>kc, lean-burn; PS
4-stroke, lean-bum; TC
4-stroke, lean-bum; TC
2-stroke, lean-bum; BS
2-stroke, lean-burn; PS
Emission
Control
None
None
None
None
None
None
Test
Dale
3/12/94
3/13/94
6/14/94
6/15/94
6/16/94
6/16/94
6/18/94
6/19/94
6/19/94
6/20/94
Sample
Location
stack
slack
stack
stack
stack
stack
stack
stack
stack
stack
Test
Period
No.
7
8
1
2
2
3
1
2
3
4
Test Time
Period
1042-1550
1240-1810
1600-2100
0832-1315
0915-1415
1415-1806
0925-1 1 15
1105-1428
1630-2018
1240-1747
OJ
K>
-------
Table 3-1. (Continued)
Campaign
4
4
4
4
Engine
Make/Model
Waukesha
L7042GU
Cooper-Bessemer
GMVC-10C
Ingersoll-Rand
KVS-4I2
Ingersoll-Rand
KVS-4I2
Unit
No.
4
8
9
Rated
Power
(hp)
896
1.300
2.000
2.000
Rated
Speed
(rpm)
1.000
300
330
330
Year
Installed
1982
1956
1956
1956
Engine Family
4-strokc, rich-burn; NA
2-stroke, clean-burn; TC
4-stroke, clean-burn; TC
4-stroke, lean-burn; TC
Emission
Control
NSCR
Clean-burn
and CO
catalyst
PCC
SCR
Test
Date
8/23/94
8/23/94
8/23/94
8/25/94
8/26/94
8/26/94
8/30/94
8/30/94
8/29/94
8/29/94
8/29/94
8/29/94
Sample
Location
after
catalyst
after
catalyst
before
catalyst
before
catalyst
after
catalyst
before
catalyst
before
catalyst
stack
before
catalyst
ifter
catalyst
after
catalyst
before
catalyst
Test
Period
No.
1
2
3
7
8
9
H
A
14
15
16
17
Test Time
Period
1146-1345
1424-1722
1805-2028
1705-1906
1051-1226
1310-1537
1605-1630
0745-0815
0858-0932
0959-1330
1413-1549
1613-1700
U)
-------
Table 3-1. (Continued)
Campaign
5
6A
6A
6B
6C
6C
6C
6C
Engine
Make/Model
Wcstinghouse
191 (Turbine)
Clark BA-5
Clark HBA-5
Solar Taurus
T-6502 (Turbine)
Cooper-Bessemer
GMVC-10
Cooper-Bessemer
GMVA-10
Cooper-Bessemer
GMWC-10
Cooper-Bessemer
GMVC-10
Unit
No.
10
12
15
9
13
11
Rated
Power
(hp)
20.000
911
1,000
5.419
1,800
1,235
3.500
1.800
Rated
Speed
(rpm)
300
300
14,300
300
300
250
300
Year
Installed
1948
1951
1993
1963
1954
1960
1957
Engine Family
2-stroke, lean-bum; PS
2-stn>ke, lean-bum; PS
2-stroke. lean-bum; TC
2-stroke, lean-bum; BS
2-stroke. lean-bum; TC
2-stroke. lean-bum; TC
Emission
Control
None
None
None
None
None
None
None
None
Test
Date
10/5/94
10/5/94
11/5/94
11/5/94
11/7/94
11/8/94
11/10/94
11/10/94
1 1/12/94
11/12/94
11/13/94
11/14/94
11/14/94
11/15/94
li/15/94
11/15/94
Sample
Location
stack
slack
stack
stack
stack
stack
slack
stack
stack
stack
stack
stack
stack
stack
stack
stack
Test
Period
No.
3
4
2
5
6
7
9
10
13
14
15
18
23
19
20
24
Text Time
Period
1230-1755
1843-2133
936^1209
1345-1802
1709-1823
915-1344
1030-1500
1546-1801
1000-1405
1515-1645
1030-1130
1855-1925
2025-2100
1020-1120
1622-1655
1730-1813
BS =» Blower scavenged. PS = Piston scavenged. TC = Turbocharged.
NSCR = Nonselective catalytic reduction. SCR = Selective catalytic reduction. NA = Naturally aspirated. PCC = Pre-combustion chamber.
Loads greater than 90 percent could not be achieved for these engines.
Manufacturer's rated hp is 3,500; however, site operates unit at maximum brake mean effective pressure (BMEP) equivalent to 4,200 hp.
-------
Continuous emissions monitoring data for NO,, CO, and THC were collected for all
units. Gas chromatography measurements for methane and ethane were performed on all
units except in following cases:
• Campaign 2 (Cooper-Bessemer GMVA-10, Clark BA-6);
• Campaign 3 (Cooper-Bessemer GMWA-8); and
• Campaign 4 [Waukesha L7042GU, Cooper-Bessemer GMVA-10, and
Ingersoll-Rand KVS-412 (Unit 8)].
3.2 Campaign 2
3.2. 1 Site Description
The site tested during GRI Campaign 2 was a sour gas processing plant. Raw natural
gas containing carbon dioxide (COj) and hydrogen sulfide (HjS) is treated with a
monoethanolamine absorption unit to remove acid gases. The sweetened gas is dehydrated
using triethylene glycol absorption and a molecular sieve unit prior to treatment in a
cryogenic extraction unit for removal of nonmethane hydrocarbons. The site utilizes 1C
engines for recompression of refrigerants for the cryogenic plant and for recompressing the
treated natural gas prior to transfer to a pipeline. Two 1C engines were tested in this
campaign. All of the engines on-site burn treated natural gas from the processing plant;
however, due to an equipment malfunction, a small amount of raw, high Btu natural gas may
have bypassed treatment and mixed with the treated gas fired during the testing.
3.2.2 Operating Conditions and Measurement of Op CO^ CO, THC, and NO
Cooper-Bessemer GMVA-10
The Cooper- Bessemer GMVA-10 engine is used to power a refrigeration compressor
used in the cryogenic extraction unit. This 2-stroke, lean-burn engine is rated at
1,350 horsepower (hp) and 300 rpm, and was operating at 77 percent load during the tests.
Horsepower measurements were performed by the host site engine analyst. The NO,
measurements failed to meet the QA requirement for daily calibration drift. The data from
this engine are not included in the average emission factor calculations because of the
operating load level during the tests (see Appendix B for test data).
3-5
-------
Clark BA-6
The Clark BA-6 engine is used to compress the natural gas leaving the cryogenic
extraction unit. This 2-stroke, lean-burn engine is rated at 1,140 hp and 300 rpm, and was
operating at 89 percent load during the tests. As for the Cooper-Bessemer GMVA-10
engine, the horsepower measurements were performed by the host site engine analyst. A
stack gas temperature of 1,092°F and visible exhaust suggest that the engine's operation may
not be representative (see Section 2.1.1 typical stack temperatures). Due to the low load and
stack gas temperature, the data from this engine are excluded from the average emission
factor calculations (see Appendix B for test data).
3.3 Campaign 3
3.3.1 Site Description
Testing was performed at a natural gas transmission station and a sweet gas
processing plant. The engines tested at Station 3A include two Cooper-Bessemer LSV-16
4-stroke, lean-burn, turbocharged engines which are used to drive compressors for natural
gas transmission in a pipeline. All equipment on-site fire pipeline-quality natural gas.
The engines tested at Sweet Gas Plant 3B are used to compress the raw natural gas at
the inlet to the extraction plant. These engines include a Cooper-Bessemer GMV-10TF,
2-stroke, lean-bum, piston-scavenged model and a Cooper-Bessemer GMWA-8, 2-stroke,
lean-burn, piston-scavenged model. Both of these engines fire raw gas from a sweet gas
field which has a high Btu natural gas.
3.3.2 Operating Conditions and Measurement of O^ CO^ CO, THC, and NOX
During this campaign, CEMS and GC data were collected, except on Engine 4
(GMWA-8) at Sweet Gas Plant 3B, which did not include GC measurements. For all test
periods, the accuracy of the Method 19 calculations for determination of stack gas flow rate
was confirmed with host site staff and the GRI program engine consultant. Therefore, the
emission rate calculations are based on Method 19 results. During 8 of the 12 test periods,
at least one manual sampling test requiring a full duct traverse was performed, thereby
allowing an independent calculation of the stack gas volumetric flow rate by EPA Method 2.
Volumetric flow rate of the stack gas during the other four test periods was estimated using
3-6
-------
differential pressure measurements made at a single point in the duct. Both Method 2 and
Method 19 results are shown in the summary tables.
Cooper-Bessemer LSV-16
The Cooper-Bessemer LSV-16 engines tested at Station 3A are turbocharged 4-stroke
engines, rated at 4,200 hp and 327 rpm, based on operation at maximum brake mean
effective pressure (BMEP). Two test runs were performed on each engine at full load.
Operating parameters and results of the CEMS measurements for NO,, CO, THC, O2, and
COj are presented in Tables 3-2 and 3-3.
Engine fuel flow rate and horsepower data were obtained from the host site's
computer control system. The engine power was calculated from site-specific performance
curves for the engines. Heat content of the natural gas fuel was based on two canister
samples taken from the fuel supply header-one taken on June 14, and one taken on June 16.
The fuel composition data from the canister analyses were used to determine the higher
heating value (HHV) for engine heat input calculations.
Engine 102 at Station 3A was tested by the host site approximately one week prior to
the Campaign 3 testing. The Campaign 3 full-load results show good agreement with this
data set. The NO, and THC values measured in this program were somewhat higher than
the earlier results, approximately 20 percent and 5 percent, respectively, although the
remaining pollutant measurements agree within 1 percent. One possible explanation for the
higher NOX values may be the higher ambient temperatures during this testing (85 °F versus
67°F).
The Method 2 volumetric flow rates measured at Station 3A were lower than expected
based on the earlier testing and Method 19 calculations; however, the Method 19 values from
the two sets of test data agree well. Follow-up discussions with the site personnel and the
GRI program engine consultant, and subsequent calculations have confirmed the accuracy of
the fuel flow measurements. As noted in Tables 3-2 and 3-3, the Method 19 values were
used to calculate all mass emission rates.
3-7
-------
Table 3-2. Station 3A: Engine Operating Conditions and CEMS Results, Engine 102,
Cooper-Bessemer LSV-16
Test Period
Load Condition
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (*F)
Relative Humidity (%)
Absolute Humidity Ob H2O/1000 Ib dry air)
Engine Operation Conditions
Horsepower (hp)*
Load (%)b
Engine Speed (rpm)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)c
NO HHV (Btu/scf)
NO LHV (Btu/scf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow(d«cfm)-MI9, F,d
Stack Gas Temperature (*F)
Moisture (*V)
02(*V)
C02(%V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO. (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (£/hp-hr)
THC (Ib/MMBtu)
1
98%
6/14/94
1600-2100
29.2
87
62
17.9
4,055
93
321
487
33.3
1.159
1,049
7,381
8,735
976
13.0
9.6
7.0
1,749
203
1,666
109
12.2
3.3
7.7
0.9
0.2
41.7
4.7
1.2
2
101%
6/15/94
0832 - 1314
29.2
83
73
18.4
4,202
101
325
497
33.7
1,148
1,039
6,947
8,678
976
14.1
9.4
7.3
1,657C
217
1,677
103
11.1
3.1
8.2
0.9
0.2
42.2
4.6
1.3
*Based on lite-specific load performance curves.
'Rating baaed on operation at maximum brake mean effective pressure (BMEP).
'Based on HHV.
Used in emission rate calculations.
"Analyzer calibration drift exceeded quality criteria for this test by 1 percent.
3-8
-------
Table 3-3. Station 3A: Engine Operating Conditions and CEMS ResuKs, Engine 101,
Cooper-Bessemer LSV-16
Test Period
Load Condition
Date
T
-------
Cooper-Bessemer GMWA-8
One Cooper-Bessemer GMWA-8 engine, Engine 4, was tested at Sweet Gas Plant 3B.
This engine is a 2-stroke, blower-scavenged engine, rated at 2,000 hp and 250 rpm. One
test run was performed on this engine at full load on June 18 as shown in Table 3-4. Fuel
flow rate data were obtained from the plant's computer control system, with the horsepower
measurements collected by the host site engine analyst.
An on-line gas analysis system was used by the host site to measure the composition
and calculate the heat content of the natural gas fuel. The results from this system were
averaged over each test period. Two gas samples were also taken in sample bombs for
analysis by Southern Petroleum Laboratories (SPL) to confirm the accuracy of the station's
analysis system. The results from SPL agreed within 4 percent of the station analyses.
The exhaust flow measurements for this engine agreed well between the two methods
(Method 2 and Method 19), with differences on the order of 1 percent. Since the accuracy
of the fuel flow rate measurements was confirmed by follow-up discussions with host site
personnel and the GRI program engine consultant, Method 19 results were used to calculate
mass emission rates.
Cooper-Bessemer GMV-10TF
One Cooper-Bessemer GMV-10TF engine, Engine 2, was tested at Sweet Gas
Plant 3B. This is a 2-stroke, piston-scavenged engine, rated at 1,100 hp and 300 rpm.
Three test runs were performed on this engine at full load—two on June 19 and one on June
20-with the test results shown in Table 3-5. Engine operating parameters, including
horsepower, speed, fuel flow, and fuel heat content were measured as described above for
the Cooper-Bessemer GMWA-8 engine tested at the same station.
3.3.3 Measurement of Methane and Ethane Emissions
On-site analysis for methane and ethane was performed for Engines 101, 102, and 2
using a GC with a flame ionization detector (FID). Results are shown in Tables 3-6 and 3-7
for the full (or highest) load conditions. Note that the differences between the measured
THC and methane/ethane concentrations were less than the analytical precision of the
instruments for the results presented in Table 3-6 for the Cooper-Bessemer LSV-16 engines.
3-10
-------
Table 3-4. Sweet Gas Plant 3B: Engine.Operating Conditions and CEMS Results, Engine 4,
Cooper-Bessemer GMWA-8
Test Period
Load Condition
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Abs. Humidity (Ib H3O/lb dry air)
Engine Operation Conditions
Horsepower (hp)'
Load(%)
Engine Speed (rpm)
Fuel Flow (»cf/min)
Heat Input (MMBtu/hr)b
NG HHV (Btu/scf)
NG LHV (Btu/icf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dscfm) -M19, Ff
Stack Gas Temperature (*F)
Moisture (%V)
O2 (%V)
C02(%V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBtu)
1
98%
6/18/94
0925-1115
29.3
85
66.0
17.72
1,958
98
249
244
17.4
1,206
1,098
8,665
8,290
645
9.3
14.7
4.2
1,246
53
1,129
74.0
17.1
4.3
1.9
0.4
0.1
25.7
6.0
1.5
*There is some uncertainty in the horsepower measurement by the engine analyst
"Based on HHV.
cUsed in emission rate calculations.
3-11
-------
Table 3-5. Sweet Gat Plant 3B: Engine Operating Conditions and CEMS Results, Engine 2,
Cooper-Bessemer GMV-10TF
Test Period
Load Condition
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Abs. Humidity (lb H2O/lb dry air)
Engine Operation Conditions
Horsepower (hp)*
Load (%)
Engine Speed (rpm)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)b
NO HHV (Btu/*cf)
NO LHV (Btu/»cf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dscfm) -M19, F,c
Suck Gas Temperature (°F)
Moisture (%V)
02(*V)
CO2 (%V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBtu)
2
104%
6/19/94
1105- 142S
29.4
91
51.0
16.74
1,145
104
300
146
10.4
1,204
1,095
3,377
3,885
661
10.3
13.0
d
1,287
81.0
828
35.8
14.2
3.5
1.4
0.5
0.1
8.9
3.5
0.9
3
104%
6/19/94
1630 - 2018
29.4
91
54.0
17.72
1,142
104
300
146
10.4
1,204
1,095
3,391
3.890
660
9.3
13.0
5.0
1,366
80.4
818
38.1
15.1
3.7
1.4
0.5
0.1
8.7
3.5
0.8
4
104%
6/20/94
1240 - 1747
29.4
91
51.0
16.71
1146
104
300
141
10.1
1205
1096
3217
3868
651
7.8
13.2
5.1
994
83.7
865
27.5
10.9
2.7
1.4
0.6
0.1
9.0
3.6
0.9
*There is uncertainty in horsepower measurements by the engine analyst.
bBased on HHV.
cUsed in emission rate calculations.
dCOj concentration not reported because analyzer calibration drift exceeded quality criteria for this test
penod.
3-12
-------
Table 3-6. Station 3A: GC Results
Engine
Make/Model
C-B LSV-16
C-B LSV-16
C-B LSV-16
C-B LSV-16
Unit
No.
102
102
101
101
Test
Period
1
2
2
3
Load
(%)
98
101
98
99
Pollutant
Methane
Ethane
Methane
Ethane
Methane
Ethane
Methane
Ethane
Stack
Cone
(ppmrd)
1,924
28
2,566
39
2,339
37
2,444
38
Emission Rale/Factor
(Ib/hr)
42
1.1
55
1.6
49
1.5
51
1.5
(g/hp-hr)
4.7
0.13
6.0
0.17
5.5
0.16
5.6
0.17
(Ib/MMBtu)
1.3
0.034
1.7
0.047
1.5
0.045
1.6
0.050
Table 3-7. Sweet Gas Plant 3B: GC Results
Engine
Malce/Modd
C-B GMV-10TF
C-B GMV-10TF
Unit
No.
2
2
Test
Period
2
3
Load
(%)
104
104
Pollutant
Methane
Ethane
Methane
Ethane
Slack
Cone
(ppmvd)
604
117
626
121
Emission Rate/Factor
(Ib/hr)
5.9
2.1
6.1
2.2
(g/hp-hr)
2.3
0.84
2.4
0.88
(Ib/MMBtu)
0.56
0.20
0.58
0.21
3-13
-------
This is not unusual for combustion sources where a large fraction of THC emissions is
composed of methane/ethane. No methane/ethane data were collected for Engine 4.
3.4 Campaign 4
3.4.1 Site Description
The host site for Campaign 4 was a storage station where engines are used to pump
natural gas to and from storage fields. The engines characterized during this campaign
included a Cooper-Bessemer GMVC-10C 2-stroke, clean-burn engine, two
Ingersoll-Rand KVS-412 4-stroke, lean-burn engines, and a Waukesha L7042GU 4-stroke,
rich-burn engine. All of the engines burn pipeline-quality natural gas. As shown in
Table-3-1, three of the engines were equipped with catalytic controls on the exhaust gas
streams.
3.4.2 Operating Conditions and Measurement of O~ CO^ CO, THC, and NO
Tables 3-8 through 3-14 present the engine operating conditions and results from the
CEMS testing performed at this site. During each test period, CEMS and GC data were
collected at the inlet and outlet of the control devices for engines equipped with catalytic
controls. Fuel flow rates were measured by the station control system and confirmed by the
GRI program engine consultant. The Method 19 volumetric flow rate estimates were used in
the emission rate calculations except where noted, under the recommendation of the GRI
program engine consultant. Both Method 2 and Method 19 flow rate estimates are shown in
the result summary tables. Heat content of the fuel was based on samples taken daily from
the fuel supply header and analyzed by the host site's laboratory. The heat content of
individual fuel samples varied less than 0.3 percent from the average during the testing
period.
Engine horsepower measurements were conducted by an engine analyst under
subcontract to Radian. Prior to testing, minimal maintenance was performed where needed
to balance the engine cylinders at maximum load conditions.
3-14
-------
T«bl« 3-«, Campaign 4: Operating Conditions and CEMS Results, Waukesha L7042GU
(Ttsts 1 and 2)
Test Period
Load Condition
Sampling Location
Date
Test Tim*
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Abs. Humidity Ob H.O/1000 Ib dry air)
Engine Operating Conditions
Horsepower 0>p)
Lo*d(*)
Engine Speed (rpm)
Fuel Flow (scf/min)»
Heat Input (MMBtu/hr)b
NO HHV (Btu/icf)
NO LHV (Btu/icf)
Exhaust Gas Conditions
Vol. Flow (dacfm) - M2C
Vol. Flow (dscfm) - Ml 9, Ft
Suck Gai Temperature (*F)
Moisture (%V)
02(*V)
CO, (*V)
NO, (ppravd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO. Ob/hr)
NO, (g/hp-hr)
NO. Ob/MMBtu)
CO Ob/hr)
CO (g/bp-hr)
CO Ob/MMBtu)
THC Ob/hr)
THC (g/hp-hr)
THC Ob/MMBru)
*Fuel flow rate is suspect.
bB*sed on HHV.
eUsed in emission rate calculations.
1
95%
After NSCR
8/23/94
114« - 1345
29.5
87
35.2
10.1
692
95
809
88.0
5.3
1,011
911
1,274
746
796
20.7
0.04
11.7
8.03
68.5
607
0.07
0.05
0.01
0.38
0.25
0.07
2.4
1.6
0.46
2
92%
After NSCR
8/23/94
1424 - 1722
29.5
87
40.1
11.3
671
92
811
84.6
5.1
1.011
911
1.274
718
796
20.7
0.06
11.7
8.38
69.5
640
0.08
0.05
0.02
0.39
0.26
0.08
2.6
1.7
0.51
3-15
-------
Table 3-9 Campaign 4: Operating Conditions and CEMS ResuKs, Wauketha L7042GU (Test 3)
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Abs. Humidity Ob H.O/1000 Ib dry air)
Engine Operating Conditions
Horsepower flip)
Load(*)
Engine Speed (rpm)
Fuel Flow (scf/min )•
Heat Input (MMBtu/hr)b
NG HHV (Btu/fcf)
NG LHV (Btu/icO
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2C
Vol. Flow (dscfm)- Ml 9. F/
Slack CM Temperature (*F)
Moisture (%V)
Oj (*V)
C02(%V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, Ob/hr)
NO. (g/hp-hr)
NO, Ob/MMBtu)
CO Ob/hr)
CO (g/bp-hr)
CO Ob/MMBtu)
THC Ob/hr)
THC (g/hp-hr)
THC Ob/MMBtu)
*Fuel flow rate i* suspect.
bBased on higher beating value O^HV).
cUsed in emission rate calculations.
3
88%
Before NSCR
8/23/94
1805 - 202*
29.5
74
65.4
12.0
644
88
813
82.8
5.0
1,011
911
1,274
716
796
20.7
0.42
11.2
2,843
3,714
1,048
25.9
18.3
5.2
20.6
14.5
4.2
4.2
3.0
0.85
3-16
-------
Table 3-10. Campaign 4: Operating Conditions and CEMS Results, Engine 4,
Cooper-Bessemer GMVC-10C (Tests 7, 9, and H)
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
At*. Humidity (lb H,O/1000 Ib dry air)
Engine Operating Conditions
Brake horsepower (hp)
Load (*)
Engine Speed (rpm)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)*
NG HHV (Btu/scf)
NG LHV (BtWjcO
Exhaust Gas Conditions
Vol. Flow (dscfm) • M2
Vol. Flow (dscfm) - Ml 9. F,b
Stack Gas Temperature (°F)
Moisture (*V)
0,<*V)
CO, (*V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (lb/hr)
NO, (g/hp-hr)
NO, Gb/MMBtu)
CO Gb/hr)
CO (g/hp-hr)
CO Gb/MMBtu)
THC Gb/hr)
THC (g/hp-hr)
THC Gb/MMBtu)
'Based on HHV.
bUsed in emission rate calculations.
7
97%
Before Catalyst
8/25/94
1705 - 190«
29.4
85
37.8
10.1
1.732
97
299
211
12.6
1,009
910
8.486
6.920
574
7.26
15.5
2.86
39.2
162
1,218
1.94
0.51
0.15
4.89
1.28
0.39
22.6
5.93
1.80
9
93%
Before Catalyst
8/2*/94
1310 - 1537
29.5
89
40.5
12.4
1.674
93
300
213
12.7
1,009
910
NA
6.866
517
7.30
15.4
2.96
34.1
181
1,559
1.68
0.45
0.13
5.41
1.47
0.43
28.8
7.79
2.26
H
95%
Before Catalyst
8/30/94
1605 - 1630
29.4
85
29.4
8.0
1.705
95
299
212
12.7
1.011
911
8.338
6.633
506
8.20
15.3
2.92
36.9
178
1.426
1.75
0.47
0.14
5.15
1.37
0.41
25.6
6.82
2.02
3-17
-------
Tabl« 3-11. Campaign 4: Operating Conditions and CEMS Results, Engine 4,
Cooper-Bessemer GMYC-10C (Test 8)
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Abs. Humidity (lb H:O/1000 Ib dry air)
Engine Operating Conditions
Horsepower (hp)
Load (*)
Engine Speed (rpm)
Fuel Flow (*cf/min)
Heat Input (MMBtu/hr)*
NG HHV (Btu/scf)
NG LHV (Btu/scf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dscfm) -M19, F,b
Slack Gas Temperature (*F)
Moisture (%V)
O, (%V)
COj (%V)
NO, (ppmvd)
CO (pprnvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, (lb/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (lb/MMBtu)
'Based on HHV.
Used in emission rate calculations.
8
99%
After Catalyst
8/26/94
1051 - 1226
29.5
84
47.8
12.5
1,773
99
299
211
12.6
1.009
910
8,751
6,696
517
7.3
15.4
3.0
44.0
14.2
1,358
2.11
0.54
0.17
0.41
0.11
0.03
24.4
6.25
1.94
3-18
-------
Table 3-12. Campaign 4: Operating Conditions and CEMS Results, Engine 8,I-R KVS-412 (PCC)
Test Period A
Load Condition 92%
Sampling Location Stack
Date 8/30/94
Test Time 0745 - 0815
Ambient Conditions
Barometric Pressure (In Hg) 29.4
Ambient Temperature (*F) 73
Relative Humidity (%) 66.4
Ab«. Humidity (Ib H.O/1000 Ib dry air) 11.7
Engine Operating Conditions
Horsepower (hp) 1,846
Lo«d (*) 92
Engine Speed (rpm) 330
Fuel Flow (»cf/min) 267
Heat Input (MMBtu/hr)« 15.9
NG HHV (Btu/scO 1.009
NG LHV (Btuy*cf) 910
Exhaust Gas Conditions
Vol. Flow (dacfm) - M2 4,531
Vol. F1ow(d*cfm) - M19, F,b 4.927
Suck Ga« Temperature (*F) 746
Moisture (%V) 12.0
02(%V) 11.3
C02(*V) 5.19
NO, (ppmvd) 64.6
CO (ppmvd) 378
THC (ppmvw) 2,333
Exhaust Emissions
NO, (lb/hr) 2-28
NO, (j/hp-hr) 0.56
NO, (lb/MMBtu) 0.14
CO (lb/hr) 8-10
CO (g/hp-hr) 1.99
CO Ob/MMBtu) 0.51
THC (lb/hr) 32.5
THC (g/hp-hr) 8.0
THC (Ib/MMBtu) 2.04
*B«ed on HHV.
Utcd in emission rate calculations.
3-19
-------
Table 3-13. Campaign 4: Operating Conditions and CEMS Results, Engine 9, I-R KVS-412
(Tests 14 and 17)
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (96)
Aba. Humidity (lb H2O/1000 Ib dry air)
Engine Operating Conditions
Horsepower (hp)
Load (96)
Engine Speed (rprn)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)1
NG HHV (Btu/scf)
NG LHV (Btu/scf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (d*cfm) - Ml 9, F,b
Slack Gas Temperature (*F)
Moisture (%V)
O2 (%V)
C02(96V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO. (Ib/hr)
NO, (i/hp-hr)
NO, (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBtu)
14
91%
Before SCR
8/29/94
0858-0932
29.5
77
52.1
10.6
1,840
91
332
272
16.3
1,011
911
4,259
3,914
779
14.4
8.56
6.93
3,042
141
938
85.2
21.0
5.24
2.41
0.59
0.15
10.7
2.63
0.66
17
91%
Before SCR
8/29/94
1613 - 1700
29.4
91
31.7
10.3
1,830
91
331
266
15.9
1,011
911
4,259
3,754
779
14.4
8.33
7.10
3,339
126
900
89.7
22.3
5.65
2.06
0.51
0.13
9.8
2.44
0.62
'Based on HHV.
bUsed in emission rate calculations.
3-20
-------
Table 3-14. Campaign 4: Operating Conditions and CEMS Results, Engine 8,1-R KVS-412
(Tests 15 and 16)
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Tetnpermftire (*F)
Relative Humidity (%)
Abs. Humidity Ob H2O/1000 Ib dry air)
Engine Operating Conditions
Horsepower (hp)
Lo*d(*)
Engine Speed (rpm)
Fuel Flow (jcf/mia)
Heat Input (MMBtu/hr)*
NG HHV (Btu/scO
NG LHV (Btu/scf)
Exhaust GAS Conditions
Vol. Flow (dscfm) - M2
Vol. How (dicfm) - M19, F
-------
Waukesha L7042GU
One Waukesha L7042GU engine equipped with NSCR control was tested. This
model is a naturally-aspirated, 4-stroke, rich-burn engine, rated at 896 hp and 1,000 rpm.
The test results are included in this section because it is the only 4-stroke, rich-burn engine
tested in this effort. Two runs were performed downstream of the catalyst bed, and one run
was performed upstream of the catalyst bed. Operating parameters and results of the CEMS
measurements for NO,, CO, CO2, THC, and O2 are presented in Tables 3-8 and 3-9. As
noted in these two tables, the fuel flow rate data for this engine are suspect. Therefore, the
emission rate calculations are based on Method 2 results.
Cooper-Bessemer GMVC-10C
One Cooper-Bessemer GMVC-10C engine (Engine 4) retrofitted with clean-burn
modifications for NO, control and an oxidation catalyst for CO control was tested. This is a
2-stroke, turbocharged engine rated at 1,800 hp and 300 rpm. Engine 4 was tested at full
load, upstream and downstream of the CO catalyst. Operating parameters and CEMS
measurements are summarized in Tables 3-10 and 3-11.
Ingersoll-Rand KVS-412
One Ingersoll-Rand KVS-412 engine (Engine 8) equipped with a PCC for NOX control
was tested. This is a 2-stroke, turbocharged engine rated at 2,000 hp and 330 rpm. In
addition, measurements were conducted on a sister unit, Engine 9, which has been retrofitted
with SCR for NO, control. Engine 9 was tested upstream and downstream of the SCR
catalyst at full load. During the tests on Engine 9, the SCR system was operating under a
condition of excess ammonia injection, creating ammonia slip through the SCR catalyst.
This condition is not typical for SCR operation, and the ammonia slip nvy have caused NO,
measurements downstream of the catalyst for this engine to be biased high. Operating
parameters and CEMS measurements for these two units are summarized in Tables 3-12
through 3-14.
3-22
-------
3.4.3 Measurement of Methane and Ethane Emissions
On-site analysis for methane and ethane was performed using a GC/FID system which
was calibrated specifically for methane and ethane, among other straight chain hydrocarbons.
Due to instrument failure, no GC data were collected on August 23, 1994. Therefore no
methane/ethane data are available for the Waukesha engine. For the remaining engines,
methane measurements were not collected due to an instrument malfunction. This
malfunction did not affect readings for ethane, as summarized in Table 3-15. Since no
methane data were available, it was not possible to calculate NMHC and NMEHC emission
factors for this campaign.
Table 3-15. Campaign 4: GC Result! (Ethane)
Engine
Make/Modd
C-BGMVC-10C
C-B GMVC-10C
I-R KVS-412
I-R KVS^12
I-R KVS-4I2
Unil
No.
4
4
8
9
9
Test
Period
7
9
11
13
15
Load
(%)
97
93
99
88
90
Sample
Location
Before
CO
catalyst
Before
CO
catalyst
Stack
Before
SCR
After
SCR
Slack
Cone.
(ppmvd)
44
45
60
29
32
Emission Rate/Factor
(lb/hr)
1.4
1.5
1.4
0.50
0.58
(x/hp-hr)
0.37
0.39
0.36
0.13
0.15
(Ib/MMBlu)
0.11
0.11
0.086
0.032
0.037
3-23
-------
3.5 Campaign 5
3.5.1 Site Description
The sweet gas plant tested in Campaign 5 is a cryogenic expansion plant which is
designed to remove 90 percent of the ethane and 100 percent of the propane and higher
molecular weight hydrocarbons from raw natural gas. The plant fractionates the
hydrocarbons extracted from the raw natural gas along with a mix of hydrocarbons purchased
from outside the facility. The facility's products include: ethane, propane, butane,
isobutane, and natural gasoline.
A natural gas-fired Westinghouse 191 (20,000 hp) combustion turbine is used to
power two refrigeration compressors at the facility. The first of the compressors is a
propane compressor rated at 10,800 hp, and the second is an ethylene compressor rated at
4,865 hp.
3.5.2 Test Results
Two full load test runs were conducted on the turbine. The CEMS was used to
gather NO,, CO, CO2, O2, and THC concentration data during both test periods, while
methane and ethane data were collected only during the first test period.
Table 3-16 summarizes the operating parameters and the CEMS measurements
collected during the two test periods. No actual measurement of brake horsepower of the
turbine was available because of instrumentation limitations at the facility. Based on
discussions with the turbine manufacturer, the measured fuel flow and the exhaust flow rates
(Method 2 and Method 19) were considered representative of full load operation for the
turbine.
Measurements of the methane and ethane concentrations in the gas turbine exhaust
were conducted using a GC/FID system. As consistent with the THC measurements, neither
methane nor ethane were detected in the exhaust gas, resulting in NMHC and NMEHC
-concentrations below detection limits as well.
3-24
-------
Table 3-16. Campaign 5: Operating Conditions and CEMS Results, Westlnghouse 191
Test Period
Load Condition
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Absolut* Humidity (Ib HaO/1000 Ib dry air)
Operating Conditions
Horsepower (hp)*
Fuel Flow (scf/min)
Heat Input (MMBru/hr)b
NG HHV (Btu/scf)
NG LHV (Btu/scf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2C
Vol. Flow (dscfm) - M19
Slack Gas Temperature (*F)
Moisture (%V)
02(%V,dry)
CO,(%V, dry)
NO, (ppmvd)
CO (ppmvd) •
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO,-,(£/hp-hr)
NO; (lb/MMBtu)
CO Ob/hr)
CO (g/hp-hr)
CO (lb/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBru)
Estimated to be tt full load.
3
Full
10/05/94
1130 - 1755
29.3
57.24
51.26
5.11
20,000
3,298
203
1,026
924
232,503
185,603
239d
4.03
17.7
2.00
41.4
13.2
ND°
68.9
1.56
0.34
13.4
0.30
0.07
ND
ND
ND
4
Full
10/05/94
1843-2133
29.4
51.48
75.12
6.04
20,000
3,298
203
1,026
924
214,147
184,460
430
4.46
17.6
2.00
39.6
16.6
NDe
60.7
1.38
0.30
15.5
0.35
0.08
ND
ND
ND
Used in emission rate calculations.
"Temperature reading suspect.
'Detection limit: 10 ppm.
3-25
-------
3.6 Campaign 6
3.6.1 Site Description
Testing was performed at three natural gas transmission stations (6A, 6B, and 6C)
where the engines and the turbine characterized in these tests are used to drive compressors
for gas transmission. Station 6A operates several Clark BA-5 engines and one Clark HBA-5
engine. As shown in Table 3-1, testing was conducted on one of the Clark BA-5 units
(Engine 10) and the single Clark HBA-5 unit (Engine 12). At Station 6B, there is one Solar
Taurus T-6502 gas turbine, which was tested during this campaign. Station 6C operates
several engines, including Cooper-Bessemer GMVA-10, Cooper-Bessemer GMVC-10, and
GMWC-10 engines. Four engines were tested at Station 6C, including one
Cooper-Bessemer GMVA-10 (Engine 9), two Cooper-Bessemer GMVC-10 (Engines 11 and
15), and one Cooper-Bessemer GMWC-10 (Engine 13).
3.6.2 Engine Operating Conditions and Measurement of O^ CO^ CO, THC, and NOX
During each test period, concentrations of O2, CO2, CO, THC, and NO, in the stack
gas were measured using the CEMS. Methane and ethane concentrations were measured by
GC/FID.
For all engines, horsepower was measured by the host site's engine analyst under the
direction of GRI program's engine consultant. For the turbine at Station 6B, horsepower and
other process data were obtained from the host site control system. Pressure readings from
an orifice meter on the fuel header were used to calculate fuel flow rates at Stations 6A and
6B. At Station 6C, the orifice meters on the fuel headers were monitored by the station's
data acquisition system, and the calculated fuel flow rates were available from the station
computer. Heat content of the natural gas fuel was based on samples taken daily from the
fuel supply header and analyzed by the host site.
Clark BA-5
A Clark BA-5 engine, Engine 10, was tested at 93 percent load on November 4, and
at 100 percent load on November 5. This model is a naturally-aspirated, 2-stroke, lean-burn
engine, rated at 911 hp at 300 rpm at site conditions. Operating parameters and results of
the CEMS measurements for O2, C02 CO, THC, and NO, are presented in Table 3-17.
3-26
-------
Table 3-17. Station 6A: Operating Conditions and CEMS Results, Engine 10,
Clark BA-5
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature ("F)
Relative Humidity (%)
Ab«. Humidity Ob HjO/1000 Ib dry tir)
Engine Operating Conditions
Brake horsepower (hp)
Load (%)
Engine Speed (rpm)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)*
NO HHV (Btu/scf)
NO LHV (Btu/scf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dscfm) - M19, F,b
Stack Gu Temperature (°F)
Moisture (%V)
02 (BV)
CO, (%V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBtu)
*Based on higher heating value (HHV).
bUsed in emission rate calculations.
°Not available due to instrument failure.
1
94%
Exhaust
11/04/94
1000-1214
26.4
62
62.7
8.2
851
94
299
161
9.8
1,025
928
7,163
4,272
458
10.4
14.2
4.3
914
86.3
NAC
28.0
14.9
2.87
1.61
0.86
0.16
—
—
—
2
100%
Exhaust
11/05/94
093* - 1209
29.3
62
29.3
3.5
910
100
299
174
10.5
1,021
924
7,493
4,679
455
8.0
14.3
4.4
1,337
92.6
970
44.8
22.3
4.26
1.89
0.94
0.18
12.3
6.12
1.17
3-27
-------
Clark HBA-5
One Clark HBA-5 engine, Engine 12, was tested at 84 percent load on November 5.
This was the highest achievable load under the pipeline pressure conditions during the testing
at Station 6A. This model is a piston-scavenged, 2-stroke, lean-burn engine, rated at
1,000 hp at 300 rpm. Because the engine was tested at a load below 90 percent, operating
parameters and CEMS results are summarized in Appendix B.
Solar Taurus T-6502
Two test runs were performed on a Solar Taurus T-6502 gas turbine at 95 percent
load, on November 7 and 8. The turbine has a site power rating of 5,419 hp. Although the
loads are similar for the two full load tests, ambient conditions and pipeline suction and
discharge pressures were slightly different. Operating parameters and CEMS results are
summarized in Table 3-18.
Cooper-Bessemer GMVC-10
A Cooper-Bessemer GMVC-10 engine, Engine 15, was tested twice at 102 percent
load on November 10. This model is a turbocharged, 2-stroke, lean-burn engine, rated at
1,800 hp at 300 rpm. Operating parameters and results of the CEMS measurements for O2,
C02, CO, THC, and NO, are presented in Table 3-19. A second Cooper-Bessemer
GMVC-10 engine, Engine 11, was tested three times at 95-101 percent load on
November 15. Operating parameters and results of the CEMS measurements for O2, CO2,
CO, THC, and NO, are presented in Table 3-20 for Engine 11. Note that only one of the
three tests was conducted at 100 percent speed (Test Period 19). As indicated in the tables,
the emission rates for the three tests were affected by the different engine speeds.
Cooper-Bessemer GMVA-10
A Cooper-Bessemer GMVA-10 engine, Engine 9, was tested twice at 100 percent
load on November 12. This model is a blower scavenged, 2-stroke, lean-burn engine, rated
at 1,235 hp at 300 rpm. Operating parameters and results of the CEMS measurements for
02, CO2, CO, THC, and NO, are presented in Table 3-21.
3-28
-------
Tabte 3-18. Station 68: Operating Conditions and CEMS Results, Solar Taurus T-6502
Test Period
Load Condition
Sampling Location
Date
Test Tim*
Ambient Conditions
BATomeirk Pressure (in Hg)
Ambient Temperature (*F)
ReUlive Humidity (%)
Ab*. Humidity (Ib H:O/1000 Ib dry air)
Engine Operating Conditions
Horsepower (hp)
Percent Load*
Turbine Speed (rpm)
Fuel Flow (scf/min)
He*t Input (MMBru/hr)b
NO HHV fBtuyjcQ
NO LHV (Btu/*cf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dscfm) -M 19, F,c
Slack Cu Temperature (°F)
Moisture (%V)
02(%Vdry)
COj (%V dry)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, Ob/MMBtu)
CO (Ib/hr)
CO (^/hp-hr)
CO (lb/MMBtu)
THC (Ib/hr)
THC (i/hp-hr)
THC (lb/MMBru)
6
95%
Exhaust
11/07/94
1709 - 1823
26.5
69
44.2
7.5
4,803
95
12,458
709
42.8
1,020
924
29,192
25,757
954
6.40
16.0
3.2
71.5
NT*1
NDe
13.2
1.2
0.3
ND
ND
ND
ND
ND
ND
7
95%
Exhaust
11/08/94
0915 - 1344
26.5
76
47.2
10.4
4,709
95
12,250
745
45.0
1.021
925
30,964
26,581
941
6.50
15.9
3.2
70.8
NDd
NDe
13.5
1.3
0.3
ND
ND
ND
ND
ND
ND
*Turbinc lo»d is bwed on tvailible power tt ambient conditions during testing.
bB*«ed on higher hearting value (HHV).
eUsed in emission rate calculations.
dDetection limit: 1 ppm.
'Detection limit: 1 ppm.
3-29
-------
Table a-19. Station 6C: Operating Conditions and CEMS Resuttt, Engine 15,
Cooper-BetMmer GMVC-10
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Ab«. Humidity (Ib H:O/1000 Ib dry tir)
Engine Operating Conditions
Horsepower (hp)
Lo«d (%)
Engine Speed (rpm)
Fuel Flow (scf/min)
Htat Input (MMBtu/hr)*
NG HHV (Btu/«cf)
NO LHV (Btu/*cf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dacfm) - Ml 9, F,b
Suck Gu Temperature (°F)
Moisture(%V)
0,(*V)
C02(%V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBtu)
*Based on higher heating value (HHV).
Used in emission rate calculations.
9
102%
Exhaust
11/10/94
1030 - 1500
26.7
53
46.5
4.4
1,827
102
300
266
16.0
1.020
924
7,570
7,972
581
7.0
15.0
3.68
560
71.2
971
31.9
7.93
1.99
2.47
0.61
0.15
20.7
5.15
1.29
10
102%
Exhaust
11/10/94
1544- 1801
26.6
58
36.2
4.2
1,830
102
300
266
16.0
1,020
924
7,588
7,749
576
7.0
14.8
3.69
638
71.1
957
35.4
8.77
2.21
2.40
0.59
0.15
19.9
4.92
1.24
3-30
-------
Table 3-20. Station 6C: Operating Conditions and CEMS Remit*, Engine 11,
Cooper-Bet»«mef GMVC-10
Test Period
Load Condition
Sampling Location
Date
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Abs. Humidity (lb H-O/1000 ]b dry air)
Engine Operating Conditions
Brake horsepower (hp)
Load (*)
Engine Speed (rpm)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)1
NG HHV (Btu/scf)
NO LHV (Btu/scf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dscfm) - M19, F,b
Slack Gas Temperature (°F)
Moisture (%V)
0, (*V)
C02(*V)
NO, (ppmvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBtu)
"Based on higher heating value (HHV).
bUsed in emission rate calculations.
19
100%
Exhaust
11/15/94
1020- 1120
27.0
50
33.5
2.8
1.799
100
300
236
14.2
1.021
925
7,501
6,983
592
6.7
14.9
3.8
870
63.6
958
43.5
11.0
3.05
1.94
0.49
0.14
17.9
4.51
1.25
20
101%
Exhaust
11/15/94
1622 - 1655
26.9
55
29.3
2.9
1,724
101
285
224
13.5
1,021
925
7,019
6,312
583
7.0
14.6
3.9
1,426
57.0
895
64.5
17.0
4.77
1.57
0.41
0.12
15.1
3.98
1.12
24
95%
Exhaust
11/15/94
1730 - 1813
26.9
51
36.2
3.1
1,669
95
292
221
13.3
1,021
925
7.172
6.528
576
6.8
14.9
3.7
757
65.3
973
35.4
9.6
2.65
1.86
0.50
0.14
17.0
4.61
1.27
3-31
-------
Table 3-21. Station 6C: Operating Conditions and CEMS Result*, Engine 9,
Cooper-Bessemer G MY A-10
Test Ptriod
Load Condition
Sampling Location
Date
Test Tiro*
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Abs. Humidity Ob H:O/1000 Ib dry »ir)
Engine Operating Conditions
Horsepower (hp)
Load (*)
Engine Speed (rpm)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)*
NO HHV (Btu;*cf)
NO LHV (BtuJscf)
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (dsc fin) -M19. F,b
Stack CM Temperature (*F)
Moisture (%V)
02(*V)
CO,(*V)
NO, (pprnvd)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (lb/hr)
NO, (g/hp-hr)
NO, (Ib/MMBtu)
CO (lb/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THCflb/MMBtu)
*Basod on higher heating value (HHV).
Used in emission rate calculations.
13
100%
Exhaust
11/LZ/W
1000 - 1405
26.5
57
S5.6
9.3
1,232
100
300
174
10.5
1,020
928
4,715
4,845
586
8.3
14.6
3.91
396
67.6
1,342
13.7
5.06
1.31
1.43
0.53
0.14
17.7
6.50
1.69
14
100%
Exhaust
11/12/94
1515- 1645
26.5
59
82.9
10.0
1.234
ICO
300
173
10.4
1.020
928
4,690
4.756
590
8.4
14.5
3.95
391
67.1
1.309
13.3
4.89
1.28
1.39
0.51
0.13
16.9
6.22
1.62
3-32
-------
Cooper-Bessemer GMWC-10
A Cooper-Bessemer GMWC-10 engine, Engine 13, was tested three times on
November 14. This model is a turbocharged, 2-stroke, lean-burn engine, rated at 3,500 hp
at 250 rpm. Tests were conducted at 91, 92, and 96 percent load. Operating parameters and
results of the CEMS measurements for O2, CO2, CO, THC, and NOZ are presented in
Table 3-22.
3.6.3 Measurement of Methane and Ethane Emissions
On-site analysis for methane and ethane was performed using a GC/FID system,
which was calibrated daily using a certified mixture of methane and ethane. Methane and
ethane concentrations as measured on the GC/FID system and the emission rate/factor for
each test run are presented in Table 3-23. As noted in the table, the instrument drift
exceeded the specified limit for runs conducted on the two Cooper-Bessemer GMVC-10 units
(Engines 11 and 15). For the two runs on the Solar gas turbine, methane and ethane
concentrations were below the detection limits, as consistent with the THC concentrations
measured at non-detect levels.
3-33
-------
Table 3-22. Station 6C: Operating Conditions and CEMS Results, Engine 13,
Cooper-Bessemer GMWC-10
Test Period
Load Condition
Sampling Location
Dale
Test Time
Ambient Conditions
Barometric Pressure (in Hg)
Ambient Temperature (°F)
Relative Humidity (%)
Ab«. Humidity (Ib H,O/1000 Ib dry air)
Engine Operating Conditions
Horsepower (hp)
Load (%)
Engine Speed (rpm)
Fuel Flow (scf/min)
Heat Input (MMBtu/hr)1
NO HHV (Btu/fcf)
NO LHV (Btu/«cO
Exhaust Gas Conditions
Vol. Flow (dscfm) - M2
Vol. Flow (djtcfm) -M19, F,b
Suck Gas Temperature (°F)
Moisture (%V)
02 (%V)
C02(56V)
NO, (ppmv!)
CO (ppmvd)
THC (ppmvw)
Exhaust Emissions
NO, (Ib/hr)
NO, (g/hp-hr)
NO, (Ib/MMBtu)
CO (Ib/hr)
CO (g/hp-hr)
CO (Ib/MMBtu)
THC (Ib/hr)
THC (g/hp-hr)
THC (Ib/MMBtu)
15
96%
. Exhaust
11/14/94
1030-1130
26.8
62
24.0
3.2
3,352
96
250
459
27.8
1,025
928
14,164
13,104
653
6.91
14.7
3.84
1,695
114
1,710
159
21.5
5.72
6.53
0.88
0.23
60.0
8.11
2.16
18
91%
Exhaust
11/14/94
1855 - 1925
26.8
55
42.7
4.3
3,172
91
249
440
26.6
1,025
928
13,942
13,514
621
6.77
15.1
3.68
1,204
105
1,785
116
16.7
4.37
6.16
0.88
0.23
64.4
9.21
2.42
23
92%
Exhaust
11/14/94
2025-2100
26.9
51
52.5
4.5
3,092
92
240
426
25.8
1.025
928
13.360
12,978
617
6.85
15.1
3.72
1,449
103
1.754
135
19.8
5.22
5.82
0.85
0.23
60.9
8.93
2.36
*Based on higher heating value (HHV)
bUsed in emission rate calculations.
3-34
-------
Table 3-23. Campaign 6: GC Results
Engine Make/Model
Clark BA-5
Clark BA-5
Solar Tiunis T-6502
Sol«r Tiuru* T-6502
C-B GMVA-10
C-B GMVA-10
C-B GMVC-10
C-B GMVC-10
C-B GMVC-10
C-B GMVC-10
C-B GMVC-10
C-B GMWC-10
C-B GMWC-10
C-B GMWC-10
Unit
No.
10
10
•
•
9
9
11
11
11
15
15
13
13
13
Test
Period
2
1
6
7
13
14
19
20
24
9
10
15
18
23
Load
(*)
100
94
95
95
100
100
100
101
95
102
102
96
91
92
Pollutant
Methane
Ethane
Methane
Ethane
Methane
Ethane
Methane
Ethane
Methane
Ethane
.Methane
Ethane
Methaneb
Ethaneb
Methaneb
Ethaneb
Methaneb
Ethaneb
Methaneb
Ethaneb
Methaneb
Ethaneb
Methane
Ethane
Methane
Ethane
Methane
Ethane
Stack
Cone.
(ppmrd)
739
14.7
593
8.7
NO*
ND*
ND*
ND*
1185
29
1067
28.3
908
23.0
895
24.2
852
23.7
772
19.9
695
18.6
1597
35.8
1651
40.0
1636
38.8
Emission Rate/Factor
(Ib/hr)
8.6
0.32
6.3
0.17
-
-
-
-
14
0.66
13
0.63
16
0.75
14
0.71
14
0.72
15
0.74
14
0.68
52
2.2
56
2.5
53.0
2.4
(g/hp-hr)
4.3
0.16
3.4
0.093
-
-
-
-
5.3
0.24
4.7
0.23
4.0
0.19
3.7
0.19
3.8
0.20
3.8
0.18
3.3
0.17
7.1
0.30
8.0 •
0.36
7.8
0.35
(Ib/MMBtu)
0.82
0.031
0.64
0.018
-
-
-
-
1.4
0.063
1.2
0.060
1.1
0.053
1.04
0.053
1.04
0.054
0.95
0.046
0.84
0.042
1.9
0.079
2.1
0.095
2.1
0.091
Detection limit: 2 ppm.
Instrument drift exceeded specified limit.
3-35
-------
Section 4.0
Sampling and Analytical Methods
The sampling and analytical methods used in the test campaigns are described in this
section. The list of target parameters and measurement methods used is presented in
Table 4-1, with a schematic of the measurement system shown in Figure 4-1.
Table 4-1. Target Parameters and Measurement Methods
Location
Suck
Fuel Header
Ambient Air
Process Data
Parameter
GM Flow Rale
Gas Molecular Weight
Gas Moisture Content
Methane
Ethane
Oxygen
Ctrbon Dioxide
Total Hydrocarbons
Oxides of Nitrogen
Carbon Monoxide
Gas Composition and
Heating Value
Barometric Pressure
Temperature
Relative Humidity
Fuel Flow Rate
Brake Horsepower
Engine Speed
Collection Method
Manual, Traverse
Extractive Probe (Dry)
Manual, Single Point
Extractive Probe (Wet)
Extractive Probe (Wet)
Extractive Probe (Dry)
Extractive Probe Pry)
Extractive Probe (Wet)
Extractive Probe (Dry)
Extractive Probe (Dry)
Sample Bombs
Barometer
Thermometer
Wei Bulb/Dry Bulb
Orifice Meter
Engine Analyst^
Engine Analyst"
Sampling and
Analytical Method
EPA Method 2; EPA Method 19
EPA Method 3A
EPA Method 4
EPA Method 18
EPA Method 18
EPA Method 3A
EPA Method 3A
EPA Method 25A
EPA Method 7E
EPA Method 10
GPA Method 22611
*GPA •= Gas Processors Association. This method was used in all campaigns except in Campaign 3B where
EPA Method TO-14 was used.
bAn engine analyst measured engine horsepower except in the following cases: Campaign 3A: site-specific
performance curves were used to determine brake horsepower; Campaign 5: turbine broke horsepower was
not measured; and Campaign 6B: turbine brake horsepower data were obtained from the station's control
system.
4-1
-------
Exhaust
Stack
Sample
Prob«
Insulated. Mufti-Tub*
Heat-Traced Sample Une
To Atmosphere
Solenoid
CaHbraUon
•nd/or
QC Standard*
Figure 4-1. Measurement system schematic
-------
4.1 Sampling Traverse Point Determination
The procedures specified in EPA Method 1 were used to determine the number and
location of sampling points required for estimating an average gas velocity. EPA Method 1
parameters are based on the length of duct separating the sampling ports from the closest
downstream and upstream flow disturbances. The minimum number of total traverse points
for a duct is specified in Method 1 based on the duct size and distance from the nearest flow
disturbances.
4.2 Exhaust Gas Flow Rate
The volumetric flow rate of the exhaust streams was estimated using EPA Methods 2
and 19. In EPA Method 2, a Type S pilot tube is used to measure the velocity of the
exhaust gas, and a Type K thermocouple is used to measure exhaust gas temperature.
During the measurements, the thermocouple and pitot tube were incorporated into a single
stainless steel sheathed probe. An oil manometer or calibrated electronic pressure transducer
was used to measure the pressure drop across the pitot tube, while a calibrated barometer
was used to obtain barometric pressure readings. During some runs, the pitot tube and
thermocouple readings were made in conjunction with other manual tests conducted as part of
the GRI test program.
In EPA Method 19 exhaust flow rate calculations, the theoretical volume of
combustion products is estimated from the fuel flow rate and the resulting volumetric flow
rate is corrected for excess air based on the O2 concentration in the exhaust gas. In this
study, the fuel flow rate data were used in conjunction with higher heating value of the fuel,
exhaust 02 concentration, and the Fd factor for natural gas to estimate exhaust gas flow rates.
The higher heating value and the Fd factor were based on fuel analysis results, while the O2
concentrations were obtained from the CEMS measurements. The Fd factor is the ratio of
the gas volume of the products of combustion to the heat content of the fuel. The "d"
subscript indicates that it is calculated on a dry basis and includes all components of
combustion less water.
4.3 Exhaust Gas Molecular Weight
The molecular weight of the exhaust gas was determined based on analysis of the
major components of the gas stream. The concentrations of 02 and CO2 in the exhaust gas
4-3
-------
were determined using continuous analyzers as specified in EPA Method 3A. The remainder
of the gas composition was assumed to be moisture and nitrogen, for the purposes of
molecular weight determination. Moisture content of the gas stream was determined by EPA
Method 4 (see below).
4.4 Exhaust Gas Moisture Content
The moisture content of the exhaust gases was determined using the procedures
specified in EPA Method 4. In this method, moisture is condensed from a metered volume
of gas in a series of chilled impingers, and the remainder is absorbed in silica gel. The total
mass of the water collected and the total sample gas volume are used to determine the
moisture content of the gas.
In most cases, the EPA Method 4 tests were completed in conjunction with other
manual tests conducted as part of the GRI program, using chilled impingers and silica gel as
specified in Method 4. In some cases, the approximation method using midget impingers, as
described in Method 4, was used for moisture content determination.
4.5 Fuel Gas Composition and Heating Value
Natural gas samples for composition analysis and heating value calculations were
collected in evacuated stainless steel containers during Campaigns 2, 4, 5, and 6; and at
Compressor Station 3A of Campaign 3. Mechanical flow controllers were used to fill the
containers at line pressure over a five-minute period. The sample containers were analyzed
using Gas Processors Association (GPA) Method 2261.
In Campaign 3B, a modified version of EPA Method TO-14 was used to collect and
analyze canister samples of the natural gas for composition analysis. Samples were collected
in evacuated SUMMA* polished stainless steel canisters, using mechanical flow controllers
for time-integrated samples over a one- to two-hour period. Analysis was performed on a
gas chromatograph with multiple detectors (GC/MD).
4.6 Exhaust Gas Composition
In determining the emissions of the pollutants of interest, a CEMS including O2,
CO2, CO, THC, and NO, analyzers, and a GC/FID were used.
4-4
-------
4.6.1 Sample Gas Extraction and Transfer
Samples were extracted from the engine/turbine exhaust stack using a stainless steel
filter and probe assembly. Sample gas was pulled through a heat-traced sample line to the
mobile sampling laboratory using a heated head pump. The sample gas was then delivered to
an insulated sample manifold for further distribution. The total sample extraction flow rate
was controlled by a flow control valve located at the exit of the heated head pump. Sample
gas was then conditioned, if required, by passing it through a series of chilled cyclones.
Conditioned gas was delivered to the O2, CO2, CO, and NO, analyzers, while unconditioned
gas was delivered to the GC and THC analyzers.
A rotameter located at the exit of the sample manifold was used to set the overall
sample extraction flow rate prior to monitoring, and also as a visual flow indicator to ensure
an adequate flow of sample gas was maintained.
4.6.2 Calibration and Quality Control Standard Delivery
The calibration and QC gas standards consisting of target analytes contained in
high-pressure gas cylinders were introduced into the sampling system through a dedicated
tube in the multi-tube heat-traced sample line. A solenoid valve located at the exit of the
sampling probe was actuated to allow the gas standards to flow through the heat-traced
sample line, the heated head pump, the insulated sample manifold, and the gas conditioner as
appropriate. Gas standards were delivered at a flow rate in excess of the total sample
extraction flow rate.
4.6.3 Measurement of O^ CO^ CO, 77-/C, and NOX Concentrations
Measurement of O2 and C02 concentrations in the sample gas was conducted
according to EPA Method 3A. A Servomex Model 1400 analyzer with a range of
0-25 percent O2 was used for O2 measurements, except during part of Campaign 2. The
Servomex analyzer uses a paramagnetic cell to produce a linearized voltage signal
proportional to the ratio of the oxygen concentration of a reference gas to the oxygen
concentration of the sample. An Ametek Model WDG-III O2 analyzer was used during part
of Campaign 2. The Ametek O2 analyzer uses an electrochemical cell to produce a
linearized voltage signal proportional to the ratio of oxygen concentrations of the sample and
a reference gas. The instrument range was 0 to 25 percent O2.
4-5
-------
Three different analyzers were used for measurement of CO2 concentrations during
this study. A Beckman Model 865-23 analyzer was used during Campaign 2. A Horiba CO2
analyzer was used during Campaigns 3 and 4 while a Servomex Model 1400 analyzer was
used during Campaigns 5 and 6. All three analyzers use a nondispersive infrared test cell to
determine CO2 concentrations based on the infrared light absorption of the gas sample. The
analyzer range for each instrument was 0 to 20 percent CO2.
Measurement of NO, was conducted according to EPA Method 7E. A TECO
Model 10AR analyzer was used during all test campaigns to measure total concentrations of
NO, in the gas streams. The analyzer operation is based on the chemiluminescence
principle, where all nitrogen oxides in the sample are converted to nitric oxide (NO),
followed by reaction of the NO with ozone in a photomultiplier tube, providing a signal
proportional to the NO, concentration in the sample. The instrument was calibrated over a
range suitable for concentration of NO, at the particular source being tested.
Measurement of CO was conducted according to EPA Method 10. A TECO
Model 48H analyzer was used during all test campaigns to measure CO in the exhaust gas
stream. The instrument was calibrated over a range suitable for concentrations of CO at the
particular source being tested. The analyzer uses a non-dispersive infrared test cell to
determine CO concentration from the infrared light absorption of the gas sample, based on a
gas filter correlation technique to eliminate interferences from other gaseous compounds
present in the sample gas.
Measurement of THC was conducted according to EPA Method 25A. A Ratfisch
Model RS-55 analyzer was used during all test campaigns except for Campaign 6, where a
J.U.M. Model VE7 analyzer was used. The instruments were calibrated over a range
suitable for concentrations of THC at the particular source being tested. Both instruments
use an FID to detect THC as it is combusted in a hydrogen flame. The FID for each
instrument was calibrated using certified concentrations of methane in air.
4.6.4 On-Sfte GC Analysis
Analysis of the exhaust gas by GC was implemented at four of the five campaigns
included in this report. Sampling and analysis of the methane and ethane concentrations in
the exhaust gases was conducted according to EPA Method 18, using a Hewlett-Packard
Model 5890 GC with an FID detector at Campaigns 3, 4, 5, and 6. The instrument was
4-6
-------
calibrated with a mixture of straight-chain hydrocarbons, including methane and ethane, in
air prior to testing. Unconditioned sample gas was delivered directly to the GC from the
sample gas distribution manifold through a heated sample line. The sample gas was passed
continuously through a heated six-port sampling valve. The valve was used for injecting
sample into the GC on a sequential basis during each testing period. Sampling frequency
was determined by the total cycle time for sample analysis, which includes the retention time
of the sample in the GC column and the time for the GC oven temperature to return to the
initial value. Typical cycle time was 15-20 minutes.
4-7
-------
Section 5.0
Quality Assurance/Quality Control and Documentation
The QA/QC procedures used during the test campaigns were largely based on the
procedures described in the QAPP. This section describes the QA/QC procedures used in
the field and laboratory, with the supporting information presented in the appendices.
5.1 Process Data Quality
Where available, process data were recorded at approximately 30-minute intervals
during each test period. At the end of each day, the process data collected for each engine
tested was reviewed by the Radian field engineer. In addition, an engine analyst was
available at all test sites except at Station 3A during Campaign 3, Campaign 5, and
Station 6B during Campaign 6. The GRI program engine consultant was on-site during
Campaigns 4 and 6 (except at Station 6B), and during startup testing at Station 3A. [Note:
Campaigns 5 and 6B involved turbine testing.]
The engine horsepower data for Station 3A were based on site-specific performance
curves, while the engine horsepower measurements at Sweet Gas Plant 3B were conducted by
the host site engine analyst. Horsepower estimates for the turbine tested during Campaign 5
were based on the manufacturer's rating since direct horsepower measurements were not
possible because of instrument limitations. The horsepower data for the turbine characterized
during Campaign 6B were based on the station's data acquisition system.
Power outages were experienced during some of the testing conducted under
Campaign 4. Time periods during which facilities experienced power outages or other
operational problems were excluded from data analysis. The results from one
Cooper-Bessemer GMVA-10 engine tested at Campaign 4 have been excluded from all
analysis in this report. This engine was not part of the original Campaign 4 test matrix, and
upon review of the fuel flow measurements, the data was considered highly suspect.
5-1
-------
5.2 Continuous Emission Monitors Data Quality
Quality assurance procedures for the CEMS were implemented according to the
reference methods. These specifications include requirements to determine calibration drift
and error of the instruments. The primary method of measuring these parameters was daily
analysis of control standards.
5.2.7 CEMS Calibration
All continuous analyzers were calibrated at the beginning and end of each test day.
Certified calibration gases were introduced at the probe and transferred through the entire
CEMS. Each analyzer was calibrated at two points, using pure nitrogen as a zero gas and a
certified span gas chosen at a concentration appropriate to the instrument range and expected
exhaust gas composition. Instrument voltage responses to the calibration gases were
recorded for each instrument during the calibration routine, along with the slope and
intercept coefficients from the linear calibration equation. The equations established during
the calibration routine were used by the data acquisition computer to calculate pollutant
concentrations based on instrument voltage responses. Instruments were recalibrated
whenever the range was changed, or when adjustments were made to the instrument
linearizer potentiometers. All ranpe changes and potentiometer adjustments were recorded in
the field log.
5.2.2 CEMS Drift Checks
At the end of each test day, or at intermediate instrument calibrations, the calibration
drift of each instrument was determined by calculating the difference between the pre-
sampling and post-sampling responses to the zero and span calibration gases, as percentages
of the full-scale reading of the analyzer. This calculation is simplified by the fact that the
response to the calibration gases is mathematically set to the actual concentration \alues in
the pretest calibration routine. The equation is thus:
D - X ^ " C) (5-D
FS
5-2
-------
where
D = calibration drift, in percent;
PO* = post-test response, in ppm or percent;
C = certified gas concentration, in ppm or percent; and
FS = full scale value (range) of instrument, in ppm or percent.
The calibration drift for each instrument was tabulated for each day or partial day.
Drifts during Campaigns 4, 5, and 6 were all below the ±5 percent limit specified in the
QAPP. In Campaigns 3 and 2, drift in excess of specified limits was experienced. The NO,
analyzer experienced drift in excess of 6 percent during testing of the two Cooper-Bessemer
LSV-16 engines on June 15 and 16, 1994, during Campaign 3A. The drift for the CO2
analyzer during the June 19 (Campaign 3B) test on the Cooper-Besemer GMV-10TF engine
was higher than the specified limit. Data collected during these test days are flagged
accordingly in the results tables.
Additionally, the NO, analyzer experienced drift in excess of 10 percent during
testing of the Cooper-Bessemer GMVA-10 engine on March 12, 1994, during Campaign 2.
The data were not used in the calculation of average emission factors because the tests were
conducted at less than 90 percent load.
5.2.3 OEMS Bias Checks
After completion of the calibration routine for each instrument, at least one QC gas
standard at the approximate concentration expected in the stack gas was introduced to the
instrument through the entire sampling system. For each QC gas (O2, CO, CO2, and NOJ,
the difference between the measured concentration of the QC standard and the known value
was calculated as a percentage of the full-scale value of the instrument range to estimate the
measurement bias (accuracy). For THC, the bias was calculated relative to the QC standard
gas concentration. This procedure was repeated periodically throughout each test day. The
equation used to calculate bias is similar to Equation 5-1:
m 100 x (R^ - C)
FS
5-3
-------
where
B = bias, in percent; and
OC = response to the QC gas, in ppm or percent.
The measurement bias data for each test are tabulated in Appendix I. Bias measurements
during Campaigns 4, 5, and 6 were all below the ±10 percent limit. One bias test of the
C02 analyzer used to monitor emissions from the Cooper-Bessemer GMV-10TF tested during
Campaign 3B showed a bias of approximately 10 percent. The affected CO2 measurements
are flagged in the data tables and are not used in any subsequent calculations.
5.2.4 OEMS Precision
The relative standard deviation (RSD) was calculated for the multiple QA/QC checks
to provide an indication of the precision (repeatability) of the species measurements. The
RSD is calculated by taking the ratio of response standard deviation to the average of the
response values times 100. The RSD values for all the tests were less than 5 percent except
on June 19, 1994 during Campaign 3B where the RSD for CO2 measurements was
approximately 9 percent, and the RSD for NO, measurements was approximately 6 percent.
Additionally, on August 23, 1994 during Campaign 4, the RSD on THC measurements was
5.1 percent.
5.2.5 Other CEMSQC
Each time the CEMS was set up at a new location, a leak check was performed on
the sampling system by sending pure nitrogen through the system at the probe, and checking
the oxygen monitor for signs of inleaking air. No tests were performed until a satisfactory
leak check with an overall O2 concentration at the analyzer less than 0.5 percent was
achieved. Additionally, instrument response lime was checked each time the system was set
up at a new location. The response time was less than 2 minutes for each test.
The efficiency of the NO2 converter on the NO, analyzer was checked periodically to
ensure that it was operating correctly. The converter was replaced if the NO2-to-NO
conversion efficiency dropped below 90 percent.
5-4
-------
5.3 Manual Sampling Methods Data Quality
Quality control procedures for the manual sampling and analysis methods consisted of
the following:
• All sampling equipment passed a thorough visual and functional check prior to
and after shipment to ensure clean and operable parts. Equipment which failed
the checks was not used in the field.
• Manometers were leveled and zeroed before measuring the differential
pressure across the Type S pitot tubes.
• The temperature measurement system was capable of measuring the ambient
temperature prior to each traverse to within ± 2°C of the average measured
ambient temperature.
• Type S pitot tubes were measured and passed the inspection criteria specified
in EPA Method 2.
• The pitch and yaw angles of the Type S pitot tube were maintained within
10 degrees of perpendicular to the flow velocity traverses.
• Each leg of the Type S pitot tube achieved the leak criteria specified in EPA
Method 2.
• The field personnel reviewed sampling and traverse data forms daily on-site
during testing.
• Any unusual occurrences during testing were noted on the field data sheets or
log book.
5.4 Method 18 Data Quality
The quality control procedures specified in Method 18 were followed. The retention
times of the analytes of interest were determined using certified calibration gases, and
multipoint calibration curves were developed for use in quantitation.
No GC data were collected during Campaign 2 since collection of methane/ethane
data was outside the scope of the GRI program at that time. During Campaign 4, hardware
problems on the gas chromatograph caused all the methane readings to be off-scale, with no
5-5
-------
methane data available for the engines tested during Campaign 4. Most of the ethane data
for this test campaign was of good quality based on the QA criteria.
During Campaign 6, on November 10, response to the high-level calibration standard
(1,000 pprnv) was unusually low. Data generated using this standard were not used in
preparing the calibration curve for this day, and the GC was calibrated at only two points.
On November 12, 14, and 15, weighted (1/x) least squares curves were used to generate the
calibration curves. This was required because a higher level standard of 2,000 ppmv was
used, and because response to the 1,000 ppmv standard was variable. The weighing routine
minimized the influence of the two high standards and improved the accuracy of
measurements at the lower end of the calibration range. Use of the higher level standard and
weighing routine are not expected to affect the final data quality.
5-6
-------
Section 6.0
Evaluation and Comparison of Emission Factors
Section 3.2 of the fifth edition of AP-42 lists emission factors for heavy-duty, natural
gas-fired pipeline compressor engines. The emission factors in AP-42 are grouped in tables
according to engine classification and emission control equipment. This section presents the
emission factors calculated for ihe eufclncs/Unbiaes vested uudw tfus, joM e.((o*\v \u tafoks.
>*rrth -41 covmterpam, \o aftcw a direct comparison between ine data collected during
this project and the emission factors listed in AP-42. [Note: The emission factors presented
in this section are from Table 2-3, except where footnoted.] Additionally, recommendations
are presented on how the data from this study can be used to enhance the emissions database
used in AP-42, thereby improving the emission factors for large internal combustion engines.
6.1 Uncontrolled Engines/Gas Turbines
Tables 6-1 and 6-2 show emission factors for uncontrolled units where Table 6-1
contains emission factors for turbines and 2-scroke engines, and Table 6-2 contains emission
factors for 4-stroke engines. Emissions listed as "THC" in the tables correspond to total
organic carbon (TOC) emission factors in AP-42. Emissions listed as "NMHC" in the tables
correspond to total nonmethane organic compounds (TNMOC) emission factors in AP-42.
There are no emission factors currently listed in AP-42 for NMEHC. The NMEHC factors
are listed in the tables in this report because the EPA excludes methane and ethane from its
definition of volatile organic compounds (VOCs) (40 CFR 51). The AP-42 emission factors
are derived from measurements conducted by Southwest Research Institute in the late '70s
and early '80s and cover a wide population of engines.
As expected, there are differences between the emission factors calculated in this
study and those in AP-42. For 2-stroke, lean-burn engines, the NO, emission factor is
higher than the AP-42 factor by about 25 to 30 percent (14 vs. 11 g/hp-hr) while the CO
6-1
-------
Table 6-1. Emission Factors for Uncontrolled Natural Gas Prime Movers: Gas Turbines and 2-Stroke Engines
Pollutant
NO,
CO
CH4
C,H6
THC
NMHC
NMEHC
Gas Turbines
GRI/EPA Tests*
(£/hp-hr)
1.4
0.16
ND
ND
ND
ND
ND
(Ib/MMBtu)
0.31
0.038
ND
ND
ND
ND
ND
AP-42
(j!/hp-ru-)c
1.3
0.83
0.17
NA
0.18
0.01
NA
(Ib/hp-hr)
2.87E-03
1.83E-03
3.75E-04
NA
3.97E-04
2.20E-05
NA
(Ib/MMBtu)
0.34
0.17
0.051
NA
0.053
0.002
NA
2-Stroke Lean-Burn
GRI/EPA Testsb
(jt/hp-hr)
14
0.63
4.6
0.31
5.7
1.1
0.80
(Ih/MMBtu)
3.4
0.15
1.1
0.077
1.4
0.28
0.19
AP-42
(*/hp-hr)C
11
1.5
5.6
NA
6.1
0.43
NA
(Ib/bp-hr)
0.024
3.31E-03
•0.012
NA
0.013
9.48E-O4
NA
(Ib/MMBtu)
2.7
0.38
1.4
NA
1.5
0.1 1
NA
Based on two turbines.
Based on seven engines.
g/hp-hr emission factors are calculated from the AP-42 Ib/hp-hr factors.
to
Table 6-2. Emission Factors lor Uncontrolled Natural Gas Prime Movers: 4-Stroke Engines
Pollutant
NO,
CO
CH4
C,H6
THC
NMHC
NMEHC
4-Stroke Lean-Burn
GRI/EPA Tests*
(S/hp-hr)
14
0.83
5.5
0.16
4.1
NA
NA
(Ib/MMBtu)
3.7
0.21
1.5
0.044
I.I
NA
NA
AP-42
(£/hp-hr)c
12
1.6
4.1
NA
4.9
0.72
NA
(Ib/hp-hr)
0.026
3.53E-03
9.04E-03
NA
0.011
1.59E-03
NA
(Ib/MMBtu)
3.2
.42
1.1
NA
1.2
0.18
NA
4-Stroke Rich-Burn
GRI/EPA Testsb
ds/hp-hr)
18
15
NA
NA
3.0
NA
NA
(Ib/MMBtu)
5.2
4.2
NA
NA
0.85
NA
NA
AP-42
(£/hp-hr)c
10
8.6
1.1
NA
1.2
0.14
NA
(Ib/hp-hr)
0.022
0.019
2.43E-03
NA
2.65E-03
3.09E-04
NA
(Ib/MMBtu)
2.3
1.6
0.24
NA
0.27
0.03
NA
"Based on three engines
b
Based on one engine tested at 81 percent speed.
cg/hp-hr emission factors are calculated from the Ib/hp-hr factors.
-------
emission factor is less than half of the AP-42 factor. Although the THC emission factors are
similar (5.7 vs. 6.1 g/hp-hr), the methane values from this study are lower by about
20 percent (4.6 vs. 5.6 g/hp-hr), resulting in NMHC factors being different (1.1 vs.
0.43 g/hp-hr). In the 4-stroke, lean-burn category, the NO, and CO emission factors show
the same trend described for the 2-stroke, lean-burn category. However, the methane
emission factor is higher (5.5 vs. 4.1 g/hp-hr) and the THC factor is lower (4.1 vs.
4.9 g/hp-hr) than the AP-42 factor. The largest differences are observed for 4-stroke,
rich-burn engines, where the emission factors from this study are based on data from one
engine only. The 4-stroke, rich-burn engine tested was an intermediate speed engine rated at
1,000 rpm, but was tested at approximately 810 rpm.
The differences between the limited data from this study and AP-42 may largely be
attributed to the variability associated with the smaller population of engines tested. Another
likely contributing factor is the type of sampling/analysis instrumentation employed during
ihe two studies.
For turbines, the NO, emission factors are similar, however, the CO emission factor
from this study is 20 percent of the AP-42 factor. Methane and THC emissions were found
at non-detect levels while the respective AP-42 values are 0.17 and 0.18 g/hp-hr.
6.2 Controlled Engines
The data for controlled engines in this report are limited because the information is
based on tests of single engines in each category. These engines were all tested during
Campaign 4, where problems with the GC hardware prevented collection of data for methane
emissions. Therefore, only NO,, CO, and THC emission factors are included in this
comparison.
Tables 6-3 and 6-4 list the emission factors based on single engine testing for
NSCR-controlled 4-strokc, rich-burn engines and SCR-controlled 4-stroke, lean-burn engines,
respectively, including the AP-42 factors. The NSCR-controlled 4-stroke, rich-bum emission
factors for NO, and CO are 50 and 40 times smaller (0.050 vs. 2.5 g/hp-hr; 0.26 vs.
10 g/hp-hr) than the AP-42 factors, respectively, while the THC emission factor is larger
than the AP-42 factor (1.7 vs 0.2 g/hp-hr). Note that the data in this study arc based on a
recently installed NSCR catalyst.
6-3
-------
Table 6-3. Emission Factors tor Controlled Natural Gas Prime Movers: NSCR On 4-Stroke Rich-Bum Engines
Pollutant
NO,
CO
THC
Inlet
GRI/EPA Tests*
(g/hp-hr)
18
15
3.0
(Ib/MMBtu)
5.2
4.2
0.85
AP-42
(j5/hp-hi-)b
7.8
12
0.33
(Ih/hp-hr)
0.017
0.026
7.28E-04
(Ib/MMBtu)
1.8
2.8
0.079
Outlet
GRI/EPA Tests*
(2/hp-hr)
0.050
0.26
1.7
(Ib/MMBtu)
0.015
0.075
0.49
AP-42
(*/hp-hr)l>
2.5
10
0.2
(Ib/hp-hr)
5.5IE-03
0.022
4.41E-O4
(Ib/MMBtu)
0.58
2.4
0.047
Based on one engine tested at 81 percent rated speed.
g/hp-hr emission factors are calculated from the AP-42 Ib/hp-hr factors.
Table 6-4. Emission Factors for Controlled Natural Gas Prime Movers: SCR On 4-Stroke Lean-Bum Engines
Pollutant
NO,
CO
THC
Inlet
GRI/EPA Tests*
(S/hp-hr)
22
0.55.
2.5
(Ib/MMBtu)
5.4
0.14
0.64
AP-42
0«/hp-hr)b
19
1.2
NA
(Ib/hp-hr)
0.042
2.65E-03
NA
(Ib/MMBtu)
6.4
0.38
NA
Outlet
GRI/EPA Tests*.c
te/hp-hr)
5.0
0.43
2.7
(Ih/MMBtu)
1.3
0.1 1
0.69
AP-42
(j;/hp-hr)b
3.6
1.1
NA
(Ib/hp-hr)
7.94E-03
2.43E-03
NA
(Ib/MMBtu)
1.2
0.37
NA
o\
-u
Based on one engine.
g/hp-hr emission factors are calculated from the AP-42 Ib/hp-hr factors.
Data suspect due to excess NH3 injection.
-------
For the SCR-controllcd engine, the outlet NO, emission factor is higher (5.0 vs.
3.6 g/hp-hr) and the CO emission factor is lower (0.43 vs. 1.1 g/hp-hr) than the AP-42
factor. The SCR system used on this engine was operating under a condition of excess
ammonia injection, creating ammonia slip through the catalyst. The ammonia slip may have
interfered with the NO, measurements, causing them to be biased high, hence making the
data suspect.
Table 6-5 presents the emission factors generated from testing a single 2-stroke,
lean-burn engine recently retrofitted with clean-burn. When the factors for the 2-stroke
clean-burn category arc compared, the NO, emission factor is much smaller than the AP-42
factor (0.48 vs. 2.3 g/hp-hr), while the CO emission factor at approximately the same level
as the AP-42 factor. The THC emission factor is significantly larger than the AP-42 factor
(6.8 vs. 2.5 g/hp-hr).
Table 6-5. Emission Factors for Controlled Natural Gas Prime Movers: 'Clean Burn' On
2-Stroke Lean-Burn Engines
Pollutant
NO,
CO
THC
GR I/EPA Tests8
(2/hp-hr)
0.48
1.4
6.8
(Ih/MMBtu)
0.14
0.41
2.0
AP-42
(j!/hp-hr)h
2.3
1.1
2.5
(Ih/hp-hr)
5.07E-03
2.43E-03
5.51E-03
(Ih/MMBtu)
0.83
0.30
0.77
Based on one engine.
g/hp-hr emission factors are calculated from the AP-42 Ib/hp-hr factors.
6.3 Other
AP-42 currently has no listing of emission factors for 4-stroke engines equipped with
PCC. Emission factors generated from the testing on such an engine are shown in
Table 6-6. Another emission control scenario not addressed in AP-42 is the presence of a
CO catalyst in conjunction with clean-burn on a 2-stroke, lean-burn engine. Table 6-^sjjpws
the emission factors based on testing on a single engine.
6-5
-------
Table 6-6. Emission Fact on (or Controlled Natural
Gat Prime Movers: 'Pre-combuitlon Chamber (PCC)'
On 4-Stroke Lean-Burn Engines
Pollutant
NO,
CO
THC
GRI/EPA Tests"
(jj/hp-hr)
0.56
2.0
8.0
(Ib/MMBtu)
0.14
0.51
2.0
aBased on one engine.
Table 6-7. Emission Factors for Controlled Natural
Gas Prime Movers: 'Clean Burn' and CO Catalyst
On 2-Stroke Lean-Burn Engines
GRI/EPA Tests8
Pollutant
NO,
CO
THC
(jj/hp-hr)
0.54
0. 11
6.3
(Ib/MMBtu)
0.17
0.030
1.9
aBased on one engine.
6.4 Conclusions
Based on examination of the test results from this study, the following conclusions are
presented to enhance the emissions database currently in AP-42:
• Incorporate emissions data used to develop the emission factors presented in
Tables 6-1 and 6-2 for uncontrolled 2-stroke, lean-burn; 4-stroke, lean-burn;
and 4-stroke, rich-burn engines; and gas turbines into the current AP-42
emissions database. Although the current factors are "A" quality,
incorporation of these data will broaden the population of the engines covered.
• Incorporate the emissions data used to develop the emission factors in
Table 6-5 for 2-stroke, clean-burn engines into the current AP-42 emissions
database. The current AP-42 factors arc "C" quality. The additional data may
upgrade the emission factor quality rating for this category.
6-6
-------
Use data shown in Tables 6-3, 6-6, and 6-7 for an NSCR-conlrolled 4-strokc,
rich-burn engine, PCC-controlled 4-stroke, lean-burn engine, and a 2-stroke,
clean-bum engine with a CO oxidation catalyst, respectively, to build and/or
improve an emissions database for these categories.
The current version of AP-42 has separate emission factors for "clean-burn"
and "PCC" controlled engines. "Clean-burn" is a trade name used by one
manufacturer to describe modifications to a lean-burn engine to lower
emissions. A PCC is a primary component of the "clean-burn" modification to
these engines. An engine equipped with PCC may also have all of the other
clean-burn modifications, as did the one engine with PCC tested under this
program. Consideration should be given to combining the emissions databases
for these control scenarios under a single generic description. The emission
factor resulting from this incorporation may be more representative of this
class of engines.
In summary, full load engine emissions test data following the QA/QC approved
procedures were obtained during 36 test runs for eight of the 11 tested 2-stroke engines,
five 4-stroke engines, and two gas turbines. These data will enhance the current database in
AP-42 for stationary 1C engines. These emissions data will not only enhance the population
of engine types covered, but will also upgrade the emission factor quality of several engine
categories which have a limited data set.
6-7
-------
Section 7.0
References
1. U.S. Environmental Protection Agency, "Compilation of Air Pollutant Emission
Factors: Stationary Point and Area Sources", Fifth Edition. AP-42
(GPO-005-000-005-001), January 1995.
2. U.S. Environmental Protection Agency, "Alternative Control Techniques
Document—NO, Emissions from Stationary Reciprocating Internal Combustion
Engines", EPA-453/R-93-032 (NTIS PB94-104494), July 1993.
3. Wilke, C., et al., "Exhaust Emissions Reduction Retrofits Available for Existing
Dresser-Rand Gas Engines", Dresser-Rand Company, Painted Post, New York,
October 1993.
4. Kim, C. and D. E. Foster, "Aldehyde and Unburned Fuel Emissions Measurements
from a Methanol-Fueled Texaco Stratified Charge Engine", Society of Automotive
Engineers, Inc., Paper No. 852120, 1986.
5. Snyder, R.B., U.S. Environmental Protection Agency, "Alternative Control
Techniques Document-NO, Emissions from Stationary Gas Turbines",
EPA-453/R-93-007 (NTIS PB93-156586), January 1993.
7-1
------- |