SRI/USEPA-GHG-VR-16
June 2002
Environmental Technology
Verification Report
MIRATECH Corporation
GECO™ 3001 Air/Fuel Ratio Controller
(Manufactured by Woodward Governor Company)
Phase II Report
Prepared By:
Greenhouse Gas Technology Center
Southern Research Institute
Under a Cooperative Agreement With
U.S. Environmental Protection Agency
-------
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.
-------
SRI/USEP A-GHG-VR-16
June 2002
Greenhouse Gas Technology Center
A U.S. EPA Sponsored Environmental Technology Verification ( £IY ) Organization
Environmental Technology Verification Report
MIRATECH Corporation
GECO™ 3001 Air/Fuel Ratio Controller
(Manufactured by Woodward Governor Company)
Phase II
Prepared By:
Greenhouse Gas Technology Center
Southern Research Institute
PO Box 13825
Research Triangle Park, NC 27709 USA
Telephone: 919/806-3456
Under EPA Cooperative Agreement CR 826311-01-0
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711 USA
EPA Project Officer: David A. Kirchgessner
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES iii
LIST OF TABLES iii
ACKNOWLEDGMENTS iv
1.0 INTRODUCTION 1-1
1.1 BACKGROUND 1-1
1.2 GECO™ 3001 AIR/FUEL RATIO CONTROLLER DESCRIPTION 1-2
1.3 TEST FACILITY DESCRIPTION 1-6
1.4 OVERVIEW OF VERIFICATION PARAMETERS AND EVALUATION
STRATEGIES 1-6
1.4.1 Changes in Lubrication Oil Degradation Rate 1-7
2.0 VERIFICATION RESULTS 2-1
2.1 OVERVIEW 2-1
2.2 RESULTS OF OIL PROPERTIES MEASUREMENTS 2-2
2.2.1 Degradation Profiles: Lubrication Oil Nitration and Oxidation 2-2
2.2.2 Degradation Profile: Lubrication Oil Total Acid Number 2-3
2.2.3 Degradation Profile: Lubrication Oil Viscosity 2-4
2.2.4 Numerical Data Analysis 2-5
3.0 DATA QUALITY ASSESSMENT 3-1
3.1 DATA QUALITY INDICATORS 3-1
3.1.1 Introduction 3-1
3.1.2 Completeness, Accuracy, and Repeatability Results 3-1
3.2 AUDITS 3-4
4.0 REFERENCES 4-1
-------
LIST OF FIGURES
Page
Figure 1-1 Relationship of Excess Air and Exhaust Gas Characteristics 1-3
Figure 1-2 Schematic of the GECO 3001 Controller 1-4
Figure 2-1 Oil Nitration During Verification Period 2-2
Figure 2-2 Oil Oxidation During Verification Period 2-3
Figure 2-3 Oil Total Acid Number Throughout the Verification Period 2-4
Figure 2-4 Oil Viscosity Throughout the Verification Period 2-5
LIST OF TABLES
Page
Table 1-1 GECO Air/Fuel Controller System Components 1-5
Table 1-2 Lubrication Oil Analyses 1-8
Table 2-1 Schedule of Lubrication Oil Sampling 2-1
Table 2-2 Results of Analysis of Oil Properties 2-6
Table 2-3 Summary of Changes in Engine Emission Rates from Phase 1 2-7
Table 3-1 Summary of Data Quality Indicators for Lubrication Oil Analyses 3-3
Table 3-2 Results of Duplicate Oil Analyses 3-4
111
-------
ACKNOWLEDGMENTS
The Greenhouse Gas Technology Center wishes to thank personnel at the natural gas transmission station
where testing was conducted for hosting this verification. The name and location of the host site are
omitted at their request. Thanks are also extended to the members of the GHG Center's Oil and Gas
Industry Stakeholder Group for reviewing and providing input on the testing strategy and this Verification
Report.
IV
-------
1.0 INTRODUCTION
1.1 BACKGROUND
The U.S. Environmental Protection Agency's Office of Research and Development (EPA-ORD) operates
the Environmental Technology Verification (ETV) program to facilitate the deployment of innovative
technologies through performance verification and information dissemination. The goal of ETV is to
further environmental protection by substantially accelerating the acceptance and use of improved and
innovative environmental technologies. Congress funds ETV in response to the belief that there are many
viable environmental technologies that are not being used for the lack of credible third-party performance
data. With performance data developed under ETV, technology buyers, financiers, and permitters in the
United States and abroad will be better equipped to make informed decisions regarding environmental
technology purchase and use.
The Greenhouse Gas Technology Center (GHG Center) is one of six verification organizations operating
under ETV. The GHG Center is managed by the U.S. EPA's partner verification organization, Southern
Research Institute (SRI), which conducts verification testing of promising GHG mitigation and
monitoring technologies. The GHG Center's verification process consists of developing verification
protocols, conducting field tests, collecting and interpreting field and other test data, obtaining
independent peer-review input, and reporting findings. Performance evaluations are conducted according
to externally reviewed Test and Quality Assurance Plans (Test Plans) and established protocols for quality
assurance.
The GHG Center is guided by volunteer groups of stakeholders. These stakeholders offer advice on
specific technologies most appropriate for testing, help disseminate results, and review Test Plans and
Verification Reports. The GHG Center's stakeholder groups consist of national and international experts
in the areas of climate science and environmental policy, technology, and regulation. Members include
industry trade organizations, technology purchasers, environmental technology finance groups,
governmental organizations, and other interested groups. In certain cases, industry specific stakeholder
groups and technical panels are assembled for technology areas where specific expertise is needed. The
GHG Center's Oil and Gas Industry Stakeholder Group offers advice on technologies that have the
potential to improve operation and efficiency of natural gas transmission activities. They also assist in
selecting verification factors and provide guidance to ensure that the performance evaluation is based on
recognized and reliable field measurement and data analysis procedures.
In the natural gas industry, transmission pipeline operators use internal combustion (1C) gas-fired engines
to provide the mechanical energy needed to drive pipeline gas compressors. As such, owners and
operators of compressor stations are interested in the performance of these engines with regard to engine
fuel consumption, reliability, availability, and emissions. MIRATECH Corporation has developed a
technology that has the potential to improve engine performance and committed to participate in a
verification of this technology. MIRATECH's GECO™ 3001 Air/Fuel Ratio Controller (Controller) is
designed to balance lean-burn engine fuel mixtures and improve fuel economy, maintenance
requirements, and emissions performance.
Performance testing of the Controller was carried out at a natural gas processing station in the southern
U.S. This station employs several reciprocating engines in support of its gas processing and transmission
activities. The design of the Controller is applicable to two of the lean-burn engines at this facility, and
these units were used for evaluation of the technology. The verification was executed in two phases of
1-1
-------
testing. Phase I evaluated Controller installation requirements, and changes in fuel consumption and
engine emission rates realized through use of the Controller. Phase II evaluated changes in lubrication oil
degradation rates realized by using the Controller. Details on the verification test design, measurement
test procedures, and Quality Assurance/Quality Control (QA/QC) procedures for both phases of testing
can be found in the Test Plan titled Test and Quality Assurance Plan, MIRATECH Corporation GECO™
3001 Air/Fuel Ratio Controller (SRI 200 la). Results of the Phase I testing are documented in a separate
report titled Environmental Technology Verification Report, MIRATECH Corporation GECO™
3001Air/Fuel Ratio Controller (Manufactured by Woodward Governor Company) (SRI 2001b). Both can
be downloaded from the GHG Center's Web site (www.sri-rtp.com). This report presents results of the
Phase II testing for evaluation of lubrication oil degradation rates.
The remaining discussion in this section describes the Controller technology, presents the operating
schedule of the test facility, and lists the performance verification parameters that were quantified.
Section 2 presents the Phase II verification test results, and Section 3 assesses the quality of the data
obtained. Section 4, provided by MIRATECH, provides additional information regarding the Controller.
The GHG Center has not independently verified information provided in Section 4.
1.2 GECO™ 3001 AIR/FUEL RATIO CONTROLLER DESCRIPTION
As engine operations and conditions change over time, engine performance and emissions can be affected
by these changes. Variables such as engine speed and load, fuel gas quality, and ambient air conditions
can have significant effects on engine operation and the air/fuel ratio in the cylinders. The Controller is
an air/fuel ratio controller designed to improve performance of natural-gas-fired, four-cycle, lean-burn
reciprocating engines by optimizing and stabilizing the air/fuel ratio over a range of engine operations and
conditions.
This device was first introduced in 1997, and currently there are about 25 units in operation in the gas
transmission industry. The technology uses a closed-loop feedback system to continuously optimize the
air/fuel mixture introduced to the engine. This system provides the potential to improve engine fuel
consumption and reduce engine emissions, particularly when changes in engine load, fuel quality, or
ambient conditions occur. Optimized and stabilized air/fuel ratios may also improve engine performance,
reduce lubrication oil degradation, and help minimize wear to major engine components, thereby reducing
engine maintenance. The Controller can be configured to operate based on engine exhaust oxygen (O2)
feedback, or generator output (kW) feedback (for engines used to drive electrical generators). Using
either approach, the Controller monitors the O2 or kW sensor inputs and controls the air/fuel ratio
generated by the carburetor. This verification test addressed only the exhaust O2 feedback system
because the Test Engine was not used to drive a generator.
The Controller uses relationships between excess air in the combustion chamber, measured exhaust gas
O2 concentrations, and engine emissions to calculate optimum air/fuel ratios at various engine loads.
Typical relationships between excess air and emissions in lean-burn, gas-fired engines are illustrated in
Figure 1-1. Using exhaust gas O2, intake manifold air pressure (MAP), intake manifold air temperature
(MAT), and magnetic pickup engine speed (MAG) as primary indicators of engine operation, the
Controller continuously adjusts air/fuel ratios in the engine by adjusting and controlling fuel flow to the
carburetor. Fuel flow is adjusted using a full-authority fuel valve that is supplied by the vendor and
installed directly into the engine fuel line upstream of the carburetor/mixer. Figure 1-2 presents a
schematic of the Controller. Table 1-1 summarizes the components that are included in a typical
Controller installation and their function.
1-2
-------
Figure 1-1. Relationship of Excess Air and Exhaust Gas Characteristics
Typical Uncontrolled Engine Emissions
2000 M1RATECH Corporation
1.031 1.020 1.010 1.000 0.910 0.833 0.769 0.714 0.667 0.625 0.588 0.555 0.526 0.500 0.476
Excess Air Ratio (Phi *)
• Phi = StOKhiometric APR 1 Operating APR
1-3
-------
Figure 1-2. Schematic of the GECO 3001 Controller
(provided by MIRATECH Corporation)
CARBURETOR/MIXER --1
OOOOOO
E^B TH GF:CIJ« D
h E'"Htf.1 CCt.lFL ";
REMOTE 5 TA1US
fCQMM MODULE
(OPTOKALI
— DIAGHOSTC R4ILILT
ALARM BELAV
B • 3) '.C.C. UC I HO
i^,.- ircy-.i j:..••.. «irj. ^.i. .I.*;;
1-4
-------
Table 1-1. GECO Air/Fuel Controller System Components
Component
Engine Control Unit
(ECU) Control Board
Keyterm
User Interface Module
Full-Authority Fuel Valve
Manifold Temperature
Sensor
Manifold Pressure Sensor
Engine Speed Sensor
Exhaust Oxygen Sensor
GECO Diagnostic
Software
Function
Includes the microprocessor controller and all electronics associated with power
regulation, signal inputs and filtering, controlled outputs, and communications. Also
includes the closed-loop enable switch.
A terminal useful for communication with the Controller in applications where a
computer is not available.
Allows the user to view the Controller status using three LED displays including
Controller power, shutdown relay, and fault relay.
An electronically actuated, full-authority valve used to control fuel flow to the air/fuel
carburetor/mixer.
A thermal resistor used to monitor intake MAT to determine M-dot air and
calculations (M-dot air is a default air temperature setpoint used during engine start-
up).
A 5-volt reference pressure sensor used to monitor intake MAP from 0 to 43 psia, used
as an indicator of engine load.
A MAG-pickup sensor used to determine engine speed (RPM) by counting pins on the
flywheel.
A universal exhaust gas oxygen (UEGO) sensor used to continuously monitor the
oxygen concentration in the exhaust gas.
Provides advanced troubleshooting capabilities using diagnostic fault codes,
oscilloscope plotting, and data logging.
Figure 1-2 and Table 1-1 show that the four input variables to the Controller during operation are exhaust
gas O2 content, MAP, MAT, and MAG-pickup. The O2 signal indicates the excess air level, the MAP
signal is used by the Controller to estimate engine load, the MAT signal is used to calculate the intake air
flow breakpoint (a preprogrammed exhaust gas O2 threshold level that disables the Controller during
engine start-up), and the MAG-pickup sensor monitors engine speed. After all system components are
installed on an engine and confirmed to be functional, the Controller must then be programmed to control
air/fuel ratios to levels most desirable for a specific engine and application. During programming, the
engine air/fuel ratios are varied while monitoring emissions to determine the optimum ratios with respect
to engine NOX emissions or fuel consumption. The optimum air/fuel ratio value is identified as Phi-
desired. The engine is then operated at a range of loads and, while monitoring the input variables (O2,
MAT, MAP, and engine speed or rpm) to the Controller, the fuel valve is adjusted to achieve the Phi-
desired ratio at each load. The valve positions and input variables at each operating point are stored by
the Controller as the Phi-target table. When in operation, the Controller produces a continuous valve
command that controls valve position, and subsequently, the air/fuel ratio.
The Controller can be used in three different modes of operation: open-loop, closed-loop, and manual.
When the engine is started, the Controller sets the fuel valve to a crank default position that can be preset
as desired. The valve remains in this position until the engine reaches 400 rpm, at which point the
Controller goes into open-loop mode and sets valve positions according to a preprogrammed valve learn
table. The Controller will operate in open-loop mode until the preprogrammed target air/fuel ratio is
surpassed, at which point the Controller will go into closed-loop mode of operation. Once in closed-loop
mode, the Controller uses input signals for engine speed and air pressure (the MAG-pickup and MAP
sensors) to look up the Phi-target valve positions from the preprogrammed valve table, and set the valve
at that position to optimize the air/fuel ratio. Manual mode is primarily a troubleshooting tool that allows
the user to disable the Controller and manually control the fuel valve to observe the sensor and emissions
responses and to program the Controller during system installation and setup.
1-5
-------
1.3 TEST FACILITY DESCRIPTION
The facility that hosted this verification is a natural gas plant where gas is extracted and processed for
subsequent transport and sale. The plant recovers hydrocarbons of C2 and heavier from the natural gas,
then compresses the residual gas for transport and subsequent sale. The plant has a capacity of greater
than 20 million cubic feet of gas per day, and is equipped with five internal combustion (1C) engines
including two Caterpillar Model 3516-LE recompressors that were used to conduct this verification.
One engine was equipped with the Controller and designated as the Test Engine. The twin engine without
the Controller was designated as the Control Engine and used for comparison of engine oil conditions.
Both units were exchanged during a scheduled overhaul with zero-hour units (engines with no run time)
during the first week of August 2000. Shortly before field testing was conducted, both engines received
scheduled maintenance including fresh lubrication oil, a tune-up, and other routine maintenance.
Both engines have a rated power output of 1,085 brake horsepower (bhp) and each consumes
approximately 7,200 cubic feet per hour (cfh) natural gas from a common fuel header during normal
operation. During normal plant operations, plant residual gas used to fuel the engines is very uniform in
composition with methane concentrations of approximately 91 percent and fuel lower heating value
(LHV) between 980 and 990 Btu/scf. Engine fuel composition can change in response to plant upsets or
changes in station operations, but these occurrences are rare. The engines are lean-burn design and no
additional emission controls are employed. Both engines drive reciprocating gas compressors that elevate
pipeline gas pressure from approximately 250 to 850 psig. The compressors are Ariel Model JGK two-
stage units. The two engine/compressor sets operate on the same schedule and load during normal station
operation. Engine speed may vary somewhat between the engines depending on inlet gas volumes.
Under normal operations, the engines run at or near full capacity with an average annual utilization of
approximately 96 percent. The engines were operated at reduced operating loads for short periods in
order to facilitate the testing planned for this verification. The station monitors engine operations
continuously, but has limited data acquisition capabilities. Therefore, engine operating parameters that
were key to this verification were monitored by the GHG Center using procedures described in Section
1.4.1 of this report and detailed in Section 2.2.1 of the Test Plan (SRI 2001 a).
For this verification, one Controller was installed on the designated Test Engine by MIRATECH on May
8, 2001. GHG Center personnel were on-site to observe and document installation activities and
requirements. On the day immediately preceding the verification testing, MIRATECH personnel
programmed the Controller using the procedures outlined in Section 1.2 to operating conditions specific
to the Test Engine.
1.4 OVERVIEW OF VERIFICATION PARAMETERS AND EVALUATION STRATEGIES
This verification was designed to quantify changes in engine fuel consumption rates, criteria pollutant and
greenhouse gas (GHG) emissions, and oil degradation rates that occur with the use of the Controller. The
evaluation was designed to characterize, via measurements and other means, the following verification
parameters:
• Changes in fuel consumption rates (Btu/bhp-hr)
• Changes in emissions of NOX, CO, THC, CO2, and CH4 emissions (g/bhp-hr)
• Controller installation requirements (labor and capital)
• Lubrication oil degradation rates (extended Phase II evaluation)
1-6
-------
Changes in fuel consumption rate and engine emissions were evaluated during Phase I testing over a 4-
day period after completion of Controller installation, shakedown, and start-up activities on the Test
Engine. The performance evaluation approach and test procedures for those evaluations are detailed in
the Phase I report and are not repeated here.
1.4.1 Changes in Lubrication Oil Degradation Rate
Evaluation of oil degradation rates continued over an additional 8-month period from June 2001 through
February 2002. This extended evaluation was conducted by comparing the oil characteristics of the
engine equipped with the Controller (Test Engine) to the oil in an identical engine (Control Engine) that
was not equipped with a Controller and operated under similar conditions.
Users of 1C engines typically collect oil samples from the engines at routine intervals and analyze the
samples for compounds that can corrode and degrade combustion equipment. These analyses are a useful
preventive maintenance tool for operators and can help to evaluate the performance and condition of the
engines. The test host site typically performs these oil analyses every 45 days. Poor fuel quality,
excessive fuel blow-by (unburned fuel passing the piston rings and entering the crankcase), unstable
air/fuel ratios, and fuel mixtures that are too rich or lean can all accelerate the rate of oil degradation. In
support of this verification, oil samples were collected and analyzed for both the Test and Control
Engines each month to evaluate if use of the Controller on the Test Engine slowed the oil degradation
process.
Both engines were equipped with fresh oil in June 2001, approximately 2 weeks prior to the verification
test period. The first set of samples was collected on June 23 immediately after completion of the Phase I
testing. For the remainder of the Phase II verification period through February 4, 2002, the Test Engine
was operated with the Controller operating in closed-loop mode (air/fuel ratios were continuously
controlled). Samples were then collected on a monthly basis for the duration of the 8-month verification
period to enable the development of oil degradation profiles for each engine. Engine operators collected
the samples from a sampling port in each engine oil system located at a point between the oil filters and
the oil cooler. The Test Plan specified that duplicate samples would be collected each month in order to
increase the size of the dataset. However, instead of collecting duplicate samples, the GHG Center
extended the verification period to 8 months in order to generate additional data and improve the oil
quality profiles.
Samples were shipped to CTC Analytical Laboratories in Phoenix, Arizona, each month after collection
to quantify each of parameters listed in Table 1-2. Station operating logs were used to document the
operating hours of both engines during the verification period. In order to make a meaningful comparison
of oil degradation rates on the two engines, operating hours needed to be similar. Typically, the engines
operate on the same schedule so long as equipment malfunctions do not occur.
1-7
-------
Table 1-2. Lubrication Oil Analyses
Analyte
Oxidation
Nitration
Viscosity @ 40°C
Total Acid Number
Reference Method
Not Specified
Not Specified
ASTM-D445
ASTM-D974
Principle of Analysis
Fourier- Transform Infra-red
Spectroscopy
Fourier- Transform Infra-red
Spectroscopy
Kinematic
Potentiometric Titration
Reporting
Units
absorbance per
centimeter (cm)
absorbance per
centimeter (cm)
centistokes (cSt)
mg KOH/g
The analytes listed in the table are indicators of oil condition and often times related. Oil nitration,
quantified in units of absorbance per centimeter (cm), is a result of piston blow-by and fuel and/or
combustion products mixing with the engine oil. The products of nitration are highly acidic and therefore
have an obvious impact on total acid numbers, but also can increase or accelerate the effects of oxidation,
and increase the oil viscosity. Oxidation, also quantified as absorbance per cm, is a chemical change in
oil composition caused by nitration and high-temperature operation. Oxidation can also increase oil
viscosity and reduce the oil's ability to lubricate engine components.
Viscosity, quantified as centistokes (cSt), is a measure of the thinness of the oil and is used as a primary
indicator of the oil's lubricating abilities. Abnormally high or low oil viscosity can be caused by dilution,
contamination, or oxidation and can be damaging to engine components. Total base and total acid
numbers are also indicators of oil condition and contamination. Most oils contain alkaline additives to
help neutralize the effects of acidic products that accumulate in the oil over time. In an engine
experiencing excessive blow-by, improper air/fuel ratios, or poor fuel quality, the total acid number can
increase dramatically over time, thereby reducing ability of the oil to maintain neutral pH.
The trends observed in the viscosity, oxidation, nitration, and total acid levels between the oil in the two
engines were used to develop degradation profiles, and identify differences that developed between the
Test and Control Engines. The operator that hosted this test has a strict policy of changing engine oil
whenever abnormal analytical parameters are reported, or every 4 months, whichever occurs first. During
planning of this test, the GHG Center recognized that, in a 4-month period, oil degradation may not have
been severe enough to observe conclusive trends regarding how use of the Controller impacts the
condition of the oil, or reduced oil degradation. During this verification, the laboratory's recommendation
to change oil occurred after the fourth monthly samples were collected, and therefore coincided with the
facility's policy of changing after a 4-month period. Both engines were charged with fresh oil after two
consecutive 4-month periods.
QA/QC procedures specified in the above referenced analytical methods were followed by the laboratory,
including instrument calibrations and performance checks. In addition, duplicate analyses were
conducted on two samples from each engine to demonstrate analytical repeatability. A detailed
discussion of the data quality of the oil sampling is provided in Section 3.0 of this report.
1-8
-------
2.0 VERIFICATION RESULTS
2.1
OVERVIEW
The Test and Control Engines were charged with fresh lubricating oil on June 2, 2001. Installation and
programming of the Controller was completed on June 19, 2001. The Phase I field testing for fuel
consumption and engine emissions was conducted from June 20 to 23, 2001. The first set of oil samples
was collected on June 23 at the conclusion of the Phase I test period. Samples were then collected each
month through February 2002, except during January 2002 when site operators did not collect samples.
Station operating logs were used to document the operating hours of both engines during the verification
period. Engine operating hours and sample collection dates are summarized in Table 2-1.
Table 2-1. Schedule of Lubrication Oil Sampling
Date
06/02/01*
06/23/01
07/12/01
08/08/01
09/05/01
10/10/01
10/11/01*
11/08/01
12/12/01
02/04/02
Activity (Both Engines)
1st Oil Change
Samples Collected
2nd Oil Change
Samples Collected
Operating Hours on Current Oil Charge
Test Engine
na
496
945
1590
2252
3054
na
722
1526
2736**
Control Engine
na
497
946
1594
2251
3051**
na
722
1549
2808**
* An oil change with fresh lubricant occurred on both engines.
** Based on the analytical results from these samples, the laboratory recommended that the engine oil be changed.
na = Not applicable
For samples collected on October 10, analytical results indicated that only the oil in the Control Engine
had degraded to abnormal levels. The laboratory flags oil analysis results as abnormal (or critical in some
cases), when values exceed industry-accepted standards. For the oil parameters examined here, abnormal
values include total acid numbers greater than 3.0 mg KOH/g oil, nitration greater than 35 absorbance
units in cm, and oxidation greater than 25 cm.
The engine oil was changed with fresh lubricant on both engines on October 11, 2001, even though
results of the oil analysis on the Test Engine were within normal tolerances. This is because the host site
changes oil if abnormal values are indicated, or after 4 months have elapsed since the last oil change.
Results of the samples collected on February 4 indicated abnormal oil conditions (i.e., elevated total acid,
nitration, and oxidation levels) on both engines. At this time, the oil in both engines was changed again
and the verification testing was concluded.
The operational hours presented in Table 2-1 show that both engines operated on nearly identical
schedules during the period following the first oil change (June to October 2001). The Control Engine
operated 72 hours longer than the Test Engine during the second period, about 3 percent longer. Site
2-1
-------
operators confirmed that the Test Engine's Controller remained in closed-loop mode for the entire test
period, continuously trying to optimize the air/fuel mixture introduced to the engine.
2.2
RESULTS OF OIL PROPERTIES MEASUREMENTS
In the Test Plan, the GHG Center proposed to prepare degradation "profiles" or graphical plots of
viscosity, oxidation, nitration, and total acid. These profiles would be used to identify significant
differences in the rate of oil degradation between the Test and Control Engines (e.g., accelerated
degradation in one engine compared to the other). It was envisioned that, if significant differences in the
oil properties for both engines occurred, they would be revealed using graphical profiles, but it was
recognized that these differences might be small and/or obscured by the variability that commonly occurs
in oil analyses.
Graphical profiles for each oil parameter are shown in the sections that follow. With the possible
exception of nitration, these graphs suggest that clear and significant differences were not evident in the
rate of oil degradation between the Test and Control Engines. However, some trends observed in the
profiles are identified and analyzed in the following section. Following this graphical assessment is a
more traditional numerical analysis of the data.
2.2.1 Degradation Profiles: Lubrication Oil Nitration and Oxidation
Each of the samples was analyzed for nitration and oxidation to develop degradation profiles for the Test
and Control Engines. Results of the nitration and oxidation analyses are presented in Figures 2-1 and 2-2.
Figure 2-1. Oil Nitration During Verification Period
oc _
on _
Si
Don -
is
z
1 n -
c .
n .
u
I '
' .
, ' •
I Test Engine
• Control Engine
*
! hirst un unange j [ second uil Change i
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500
Hours of Operation Since
First Oil Change
2-2
-------
Figure 2-2. Oil Oxidation During Verification Period
30
25
E 20
o
IS
'><
O
10
•i
! f
I
!
I I I Test Engine
7~" | • • Control
j First Oil Change j j j Second Oil Change j
=ngine
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500
Hours of Operation Since
First Oil Change
Both the Test and Control Engines operated nearly continuously as expected, and nitration and oxidation
in both engines increased over time. It is generally accepted in the industry that fuel combustion products
cause increases in nitration, and that nitration can accelerate the oil oxidation process. One cause of
nitration is fuel mixtures that are either too rich or too lean, and exposure of lubrication oil to combustion
products that contain acid gases like NOX.
With the Controller maintaining optimum air/fuel ratios and lower NOX formation on the Test Engine, it
is possible that nitration rates, and subsequently oxidation rates, could be reduced. The data in Figure 2-
1 suggest that oil degradation, as expressed by increases in nitration (an oxidation precursor), is lower on
the Test Engine. The Test Engine often had lower nitration values than the Control Engine. It should be
noted that both the nitration and oxidation figures are skewed somewhat by the presence of data outliers,
which are identified and examined more closely in Section 2.2.4.
2.2.2 Degradation Profile: Lubrication Oil Total Acid Number
Most nitration products formed in the oil are highly acidic and, therefore, whenever nitration levels
increase, as was the case in these samples, the oil's total acid number is also likely to increase. Increasing
oil acidity is evident for both engines during this verification, as illustrated in Figure 2-3.
2-3
-------
Figure 2-3. Oil Total Acid Number Throughout the Verification Period
•^ Q n _
9. or
2£ Z.O "
£ 9 n -
!5 1 c
— -i n -
ns I-U
0.0 -
|
i
!
• i
|
1 Test Engine
• Control Engine
: : : • :
I Fiibl Oil Chanye • • I Secund Oil Change •
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500
Hours of Operation Since
First Oil Change
Acidity in the Control Engine oil after about 3,000 hours of operation was identified by the laboratory as
abnormal prior to the second oil change, while the acidity in the Test Engine oil was at a narrowly
acceptable level. Near the end of the second oil charge, the acid number for the oil in both engines was
identified as abnormal after approximately 2,700 hours. The profile for total acid in the graph above does
not present a clear indication that one engine's oil was acidifying more than the other.
2.2.3 Degradation Profile: Lubrication Oil Viscosity
Many variables can impact oil viscosity, including nitration, oxidation, and acidity of the lubrication oil.
Therefore, whenever levels of these three parameters increase, the oil's viscosity is likely to degrade. On
both engines, viscosity increased over time at similar rates as illustrated in Figure 2-4. The viscosity data
do not present a clear indication that one engine's oil was degrading more significantly than the other.
2-4
-------
Figure 2-4. Oil Viscosity Throughout the Verification Period
4-1
w
o
o
'in
o
o
(0
>
14 n -
19 n -
1 n n -
A n -
2.0 -
0.0 -
T t * f * 1 *
L
1 Test Engine
• Control Engine
• ,_. ± ^., ^, • • i ^ , ^., ^,
s I list DM unange \ i -: oecond DM on
ange i
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500
Hours of Operation Since
First Oil Change
2.2.4
Numerical Data Analysis
In addition to the trends described above, an analysis of the profiles revealed the presence of several
apparent outliers. To assess the validity of these apparent outliers, the data's reasonableness was assessed
based on input from experts at the oil analysis laboratory, and an assessment of data trends and oil
properties for both engines. The outliers identified and actions taken are outlined below.
• Test Engine: In Figure 2-1 (nitration data), the fourth data point for the Test Engine appears
unreasonably low, especially considering the previous two samples had significantly higher nitration
values, and nitration levels should not be reduced with added exposure to engine operations. Dilution
with large quantities of fresh oil could cause such a large reduction in nitration, but evidence of a
large dilution is not apparent in the other results, and if additions did occur, site operators indicate
they would be small (under 4 percent of total volume per month). Later in this section, the average
difference in nitration values between the Test and Control Engines is calculated. When this is done,
the questionable fourth data point for the Test Engine (October 10, 2001) is considered invalid and
not used in the calculations. To avoid biasing average results for the Control Engine, the
corresponding October 10, 2001, sample for this engine was also invalidated and not used.
• Control Engine: In Figure 2-2 (oxidation data), the fifth oxidation data point in the chart for the
Control Engine appears unreasonably low (2.0 cm). This oxidation sample was the first one collected
after the second engine oil change occurred and, in comparison with other samples collected soon
after an oil change, the 2.0 cm value is much lower. Specifically, the initial oxidation values for the
Test Engine are 6 to 8 times higher, and the initial oxidation value for the Control Engine in June
2001 is 6 times higher. As such, the questionable fifth oxidation data point for the Control Engine
(November 8, 2001) is considered invalid, and is not used in determining the average difference in
oxidation values between the Test and Control Engines. To avoid biasing average results for the Test
Engine, the corresponding November 8, 2001, sample collected for this engine was also invalidated
and not used.
After removing the outliers as described above, differences between the Test and Control Engine oil
properties were examined further. For each sample pair collected on the same day, the GHG Center
subtracted oil properties measured for the Test Engine from the same oil properties measured on the
2-5
-------
Control Engine. The average of these differences is calculated and reported in Table 2-2 for each oil
parameter (Average Difference). Separate Average Difference values are reported, one including all valid
samples collected following the first oil change, and one including all valid samples collected following
the second oil change.
Table 2-2 presents the results described above. In addition, the Overall Average Difference is reported
near the end of the table for each oil parameter. It is calculated by averaging all of the differences
reported in Table 2-2 for that parameter. Finally, the Overall Average Percent difference is calculated and
presented at the end of the table. This value provides an indication of the significance of any differences
found, and was calculated by dividing the Overall Average Difference, calculated as described above, by
the overall average value of that oil parameter for the Control Engine. The example below illustrates this
calculation.
Overall Average Percent Difference For Viscosity =
100*[average of all viscosity differences/average of all Control Engine viscosity values]
= 100*[0.21 cSt/14.52 cSt]
= 1.4%
Table 2-2. Results of Analysis of Oil Properties
Date of
Sample
R/9/O1
6/93/01
7/19/01
R/R/CI1
Q/R/m
10/10/01
m/11/m
1 1 /8/01
19/19/01
2/4/02
Viscosity @ 40°C
(cSt)
Control
Enaine
Test
Enaine
Difference
Oxidation
(cm)
Control Test
Enaine Enaine
Difference
Nitration
(cm)
Control
Enaine
First Oil Change
14.23
14.26
14.59
15.18
14.75
14.11
14.26
14.72
15.01
15.35
-0.12
0.00
0.13
-0.17
0.60
12 12
14 15
22 21
0
1
-1
17
18
29
Test
Enaine
Difference
Total Acid Number
(mg KOH/g)
Control
Enaine
10
22
24
-7
4
-5
2.37
2.80
2.96
Test
Enaine
Difference
2.09
2.88
2.51
-0.28
0.08
-0.45
see footnote a
26 20
-6
36b
18b
na
3.50
2.90
-0.60
Second Oil Change
13.94
14.57
14.64
14.60
14.55
15.20
Average Difference
Met ni rhannp'^
Average Difference
(9nri oil changpl
Overall Average Difference
Overall Average Percent
Diffprpnnp
0.66
-0.02
0.56
0.09
0.40
0.21
1.4
2b 17b
13 18
25 24
na
5
-1
-2
2
-0.3
-2
16
22
38
13
11
28
-3
-11
-10
-3
-8
-5
-21
1.50
2.74
3.23
1.86
2.81
3.79
0.36
0.07
0.56
-0.31
0.33
-0.04
-1.4
a Sampling data unavailable: analyses of these parameters were not specified on the chain of custody forms submitted by site
operators.
b Data considered outlying values are included in the table for information, but not included in the analysis.
na: Not aDDlicable
2-6
-------
There appear to be no consistent and significant Average Difference between the oxidation, viscosity, and
total acid numbers for the Control Engine, and the values of these parameters for the Test Engine. Small,
and often inconsistent, differences can be seen in these parameters and, for several samples, differences
reported are within the repeatability of the measurements (repeatability results of 1.11 to 2.18 percent are
reported in Chapter 3).
Consistent with the graphical analysis discussed earlier, the data in Table 2-2 suggest that the amount of
lubrication oil nitration is significantly less in the Test Engine: 21 percent less than the Overall Average
nitration value associated with the Control Engine. This is consistent with the Phase I finding that NOX
emissions were reduced by about 30 percent, and that exposure of lubrication oil to acid gases like NOX
can increase nitration as combustion gases or "blow-by" mix with lubrication oil in the crank case. Table
2-3 shows the NOX emission reductions verified as part of the Phase I evaluation. The emission changes
measured for other air pollutants and GHGs are also shown. More details on the data from Phase I can be
obtained from the Phase I report Environmental Technology Verification Report, MIRATECH
Corporation, GECO™ 3001 Air/Fuel Ratio Controller (Manufactured by Woodward Governor Company)
(SRI 2001b). This report can be obtained online at www.sri-rtp.com or www.epa.gov/etv.
Table 2-3. Summary of Changes in Engine Emission Rates from Phase I
Test
ID
1
10
11
Engine
Operation
Full Load
Average Change
2
12
13
Reduced Load
Average Change
Reduction3 in Engine Emissions (%)
NOX
18.1
38.6
34.9
30.5
19.1
38.2
32.7
30.0
CO
9.9
3.1
2.3
5.1
3.6
2.2
1.5
2.4
THC
22.3
(11.6)
(4.3)
(2.1)
(9.2)
(14.0)
(9.9)
(11.0)
CH,
23.0
(11.7)
(4.0)
(2.4)
(8.8)
(13.8)
(9.9)
(10.8)
C02
(0.9)
(1.1)
(0.2)
(0.7)
0.6
(1.9)
(1.2)
(0.8)
a Values in parentheses indicate percent increases.
2-7
-------
-------
3.0 DATA QUALITY ASSESSMENT
3.1 DATA QUALITY INDICATORS
3.1.1 Introduction
In verifications conducted by the GHG Center and EPA-ORD, measurement methodologies and
instruments are selected to ensure that a desired level of data quality occurs in the final results. The
desired levels of data quality are normally defined as data quality objectives (DQOs). The process of
establishing DQOs starts with determining the desired level of confidence in the primary verification
parameters (e.g., during Phase I, these were changes in fuel consumption and engine emission rates). The
next step is to identify all measured values that impact the primary verification parameters and estimate
the level of error that can be tolerated. Prior to testing, error propagation is often used to estimate the
cumulative effect of all measured variables on the quality of the verification parameters measured. This
allows individual measurement methods and instruments to be chosen which perform well enough to
satisfy the DQO for each verification parameter. The technique used to determine if DQOs are met is to
identify data quality indicators (DQIs) for each of the required critical measurements. The DQIs usually
define the accuracy, precision, and completeness goals for each measured variable.
In the Test Plan, the GHG Center intentionally did not include a DQO for changes in oil degradation rates
because degradation rates can vary widely and contain large variation. There are many engine design and
operational variables that can contribute to oil degredation. The GHG Center did identify DQIs for each
of the oil analysis parameters, to ensure that the measurements were as reliable as possible and consistent
with generally accepted industry standards.
3.1.2 Completeness, Accuracy, and Repeatability Results
This section presents the results of all QA/QC checks for the oil analyses methods, and reconciles the
DQIs. Table 3-1 summarizes the range of measurements observed in the field, the DQI goals, the
achieved DQIs, and completeness goals. In all cases, collected samples were analyzed within 23 days of
collection. This deviates from the interval of 1 day specified in the Test Plan but, given the stability of
sealed oil samples during storage, this deviation is not expected to affect results.
The completeness goal for the oil analyses was to obtain a valid sample each month during 90 percent of
the verification period. Achieved completeness of 78 percent was short of that goal. This was because
the September 5, 2001, samples were mistakenly analyzed for viscosity only because the other analyses
were not specified on the chain-of-custody form prepared by the host site operators. A second reason for
the low completeness was that the host site did not submit samples for January 2002. Both of these
omissions contributed to difficulties encountered in interpreting the oil degredation profiles presented in
Section 2. However, valid samples were generally available for the preceding and following months,
allowing the GHG Center to bound the period where sampling was not conducted. This significantly
reduces the impact of losing data, allowing the GHG Center to form reliable conclusions with a
reasonable degree of confidence. One exception was the nitration variable where, as a result of the outlier
analysis discussed in Section 2.2.4, one sample was removed for the Test Engine. In this case, nitration
trends and performance results are reasonably clear in spite of the data loss, the trends observed are
consistent with emissions findings in Phase I and, as such, the conclusions reached for nitration are
considered reliable.
3-1
-------
Instrument calibration and performance data supplied by the laboratory were reviewed to document
proper instrument operation, and to qualitatively verify achievement of instrument accuracy DQI goals for
viscosity, nitration, and oxidation. These records were obtained and reviewed by the GHG Center, and
they revealed that the accuracy-related DQI goals were achieved for the three parameters listed above.
Instrument performance data for total acid number were not obtained and reviewed. The CTC Analytical
Laboratory maintains continuous QA/QC procedures as standard operation for all of the instrumentation
used in these analyses. Oxidation and nitration levels are determined using a Perkin Elmer Model 1600
Fourier Transformation Infrared Spectrometer (FTIR). The instrument has background checks run every
half-hour during use and is internally standardized to an accuracy of ± 1-cm absorbance or lower.
Viscosity is determined in accordance with the American Society for Testing and Materials (ASTM)
D445 using a Houllion Viscometer. This instrument measures sample viscosity under controlled
conditions (100 °C). A Statistical Process Control (SPC) standard sample is checked daily to verify that
the viscometer has an accuracy of ± 0.05 cSt. For total acid number (ASTM D974), the method uses a
color indicator to indicate the titration endpoint, and the titrant is certified for its normality. Triplicate
titrations were conducted on each sample. In addition, daily calibration is conducted using an SPC
sample which yields a method accuracy of ± 0.5 percent of reading.
Duplicate analyses for all the oil variables examined were conducted to confirm the achievement of the
repeatability DQI goal for the total acid number method, and to provide important repeatability
characterizations for the other three methods used. Many Phase II performance assessments involve the
calculation of a difference between two measured parameters (e.g., the Test Engine nitration minus the
Control Engine nitration), so repeatability is perhaps more important than accuracy or bias as an indicator
of data reliability (measurement bias would tend to cancel out when measured values are subtracted).
Duplicate analyses were carried out on four of the samples submitted, and these results are presented in
Table 3-2. For samples with duplicate analyses, the average value of the two results was used to report
results in Section 2.2. The DQI goal for total acid number was to demonstrate repeatability of ± 5 percent
using duplicate analyses and, as Tables 3-1 and 3-2 show, this goal was achieved (1.7 percent was
achieved). Repeatability results for the other three methods ranged from about 1 percent for the viscosity
and oxidation methods, to about 2 percent for the nitration method. The Test Plan called for duplicate
analyses on every sample collected. This was not conducted on every sample by the laboratory but, given
the results of the duplicates conducted, demonstrated repeatability is acceptable.
3-2
-------
Table 3-1. Summary of Data Quality Indicators for Lubrication Oil Analyses
Measurement Variable
Lubrication
Oil Analyses
Viscosity
Nitration
Oxidation
Total Acid
Number
Instrument Type /
Manufacturer
Kinematic Capillary
Viscometer
NicoletFTIR
Spectrometer
Automatic Karl
Fischer (KF1 Titrator
Instrument
Rated
Accuracy
+ 0.05 cSt
+ 1 cm
+ 0.5% of
readina
Measurement
Range
Observed
13.95to 15.35
cSt
10 to 38 cm
2 to 26 cm
1.50 to 3.79 mg
KOH/g oil
Frequency of
Measurements
Once every
month for
duration of
verification
period
DQI Goal
+ 0.05 cSt
± 1 cm
±5%
repeatability
DQI Achieved
+ 0.05cSt
+ 1 cm
+ 1.6%
repeatability
How Verified /
Determined
Reviewed
laboratory
calibration
records
Duplicate analyses
Completeness
Achieved
78 percent
(this excludes
outliers removed
as discussed in
Section 2.2.4)
3-3
-------
Table 3-2. Results of Duplicate Oil Analyses
Parameter
Viscosity (cSt)
Nitration (cm)
Oxidation (cm)
Total Acid (mg
KOH)
Analytical Results (initial/duplicate)
6/23/01, Test
Engine
14.12/14.09
10/10
12/12
2.07/2.10
6/23/01,
Control
Engine
14.18/14.27
17/17
12/12
2.36/2.37
8/8/01, Test
Engine
14.56/14.87
23/25
21/21
2.56/2.45
8/8/01,
Control
Engine
14.48/14.69
29/29
21/22
2.95/2.96
Average
Repeat-
ability (%)
1.1
2.1
1.2
1.7
3.2
AUDITS
The GHG Center's QA Manager conducted an audit of data quality (ADQ). The ADQ confirmed that all
data handling and calculations were adequate and correct. This was done by selecting a random sample of
verification results, duplicating all the calculations performed to determine those results, confirming the
proper extraction and tabulation of measurements data by examining a section of raw data reports
supplied by the laboratory, and confirming the proper use and interpretation of laboratory-supplied
duplicate analysis results. A field activities technical systems audit (TSA) was not conducted on this
portion of the verification because samples were collected only at 1-month intervals.
3-4
-------
4.0 REFERENCES
ASTM D445. American Society for Testing and Materials, Standard Test Method for Kinematic
Viscosity of Transparent and Opaque Liquids (the Calculation of Dynamic Viscosity (D445-01), West
Conshohocken, PA, 2001.
ASTM D974. American Society for Testing and Materials, Standard Test Method for Acid and Base
Number by Color-Indicator Titration (D974-01), West Conshohocken, PA, 2001.
SRI 2001 a. Southern Research Institute. Test and Quality Assurance Plan, MIRATECH Corporation,
GECO™ 3001 Air/Fuel Ratio Controller, Research Triangle Park, NC, February 2001.
SRI 2001b. Southern Research Institute. Environmental Technology Verification Report, MIRATECH
Corporation GECO™ 3001 Air/Fuel Ratio Controller (Manufactured by Woodward Governor Company),
SRI/USEPA-GHG-VR-11, Research Triangle Park, NC, September 2001.
4-1
------- |