November 2010
Environmental Technology
Verification Report
FLIR SYSTEMS
GASFiNDlR™ MIDWAVE (MW) CAMERA
Prepared by
Battelle
Batfeile
The, Business of I tin ovation
Under a cooperative agreement with
U.S. Environmental Protection Agency
ET v ET v ET w
-------
November 2010
Environmental Technology Verification
Report
ETV Advanced Monitoring Systems Center
FLIR SYSTEMS
GAsFiNDlR™ MIDWAVE (MW) CAMERA
by
Brian Boczek and Amy Dindal, Battelle
John McKernan, U.S. EPA
-------
Notice
The U.S. Environmental Protection Agency, through its Office of Research and Development,
funded and managed, or partially funded and collaborated in, the research described herein. It
has been subjected to the Agency's peer and administrative review. Any opinions expressed in
this report are those of the author(s) and do not necessarily reflect the views of the Agency,
therefore, no official endorsement should be inferred. Any mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
11
-------
Foreword
The EPA is charged by Congress with protecting the nation's air, water, and land resources.
Under a mandate of national environmental laws, the Agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, the EPA's Office of Research and
Development provides data and science support that can be used to solve environmental
problems and to build the scientific knowledge base needed to manage our ecological resources
wisely, to understand how pollutants affect our health, and to prevent or reduce environmental
risks.
The Environmental Technology Verification (ETV) Program has been established by the EPA to
verify the performance characteristics of innovative environmental technology across all media
and to report this objective information to permitters, buyers, and users of the technology, thus
substantially accelerating the entrance of new environmental technologies into the marketplace.
Verification organizations oversee and report verification activities based on testing and quality
assurance protocols developed with input from major stakeholders and customer groups
associated with the technology area. ETV consists of six environmental technology centers.
Information about each of these centers can be found on the Internet at http://www.epa.gov/etv/.
Effective verifications of monitoring technologies are needed to assess environmental quality
and to supply cost and performance data to select the most appropriate technology for that
assessment. Under a cooperative agreement, Battelle has received EPA funding to plan,
coordinate, and conduct such verification tests for "Advanced Monitoring Systems for Air,
Water, and Soil" and report the results to the community at large. Information concerning this
specific environmental technology area can be found on the Internet at
http ://www. epa.gov/etv/centers/centerl .html.
in
-------
Acknowledgments
The authors wish to acknowledge the contribution of the many individuals, without whom, this
verification testing would not have been possible. Quality assurance (QA) oversight was
provided by Michelle Henderson, U.S. EPA, and Zachary Willenberg, Battelle. We thank Mr.
David Fashimpaur of BP, for hosting the laboratory testing phase of this verification test at the
BP, Naperville, IL research complex. Also, we acknowledge the support of Mr. Jeffrey Panek
and Dr. Paul Drayton of Innovative Environmental Solutions, Inc. for operating the leak
generation equipment and performing data collection during the laboratory testing phase. We
thank Ms. Julie Woodard, Ms. Fran Quinlan Falcon, and Mr. Barry Kelley of the Dow Chemical
Company for providing the field test site and for supporting the verification test team during the
field testing phase. We gratefully acknowledge the support of the American Chemistry Council
(ACC) and the Texas Chemical Council (TCC) as collaborators to this verification test and
would like to specifically thank Mr. Jim Griffin (ACC) and Ms. Christina Wisdom (TCC) for
their personal dedication to this verification test. Finally, we want thank Mr. David Williams
and Mr. Eben Thoma of the U.S. EPA for their review of the test/QA plan and/or this verification
report.
IV
-------
Contents
Page
Foreword iii
Acknowledgments iv
List of Abbreviations ix
Chapter 1 Background 10
Chapter 2 Technology Description 11
Chapters Test Design and Procedures 12
3.1 Test Overview 12
3.2 Experimental Design 14
3.2.1 Detect!on of a Chemical Gas Leak Using FLIR GasFindIR™ 14
3.2.2 Method Detection Limit 15
3.2.3 Confounding Factors 15
3.2.4 Detection of a Chemical Gas Species Relative to a Portable Monitoring Device.. 18
3.2.5 Field Testing Procedures 18
3.3 Qualitative Evaluation Parameters 19
Chapter 4 Quality Assurance/Quality Control 20
4.1 Reference Method Quality Control 20
4.1.1 Bias and Accuracy of Sample Screening Measurements Using Portable Monitoring
Device 21
4.1.2 Confirmation of Detected Leaks 22
4.1.3 Bias and Accuracy of Enclosure Equilibration Gas 22
4.1.4 Bias and Accuracy of Bagging Procedure 24
4.1.5 Bias and Accuracy of Gas Chromatography Analytical Method 24
4.2 Audits 25
4.2.1 Technical Systems Audit 25
4.2.2 Data Quality Audit 27
Chapters Statistical Methods 28
5.1 Method Detection Limit 28
5.2 Percent Agreement 28
Chapter 6 Test Results 30
6.1 Method Detection Limit 30
6.2 Detection Agreement to a Portable Monitoring Device 35
6.2.1 Laboratory Testing 35
6.2.2 Field Testing 36
6.3 Confounding Factors 40
6.4 Operational Factors 40
Chapter 7 Performance Summary 42
Chapter 8 References 45
Appendix A FLIR GasFindIR™ LW Camera Results 46
A.I Method Detection Limit 46
A.2 Detection Agreement to a Portable Monitoring Device 48
A.2.1 Laboratory Testing 48
v
-------
A.2.2 Field Testing 48
A.3 Confounding Factors 49
A.4 Operational Factors 50
VI
-------
Table 1.
Table 2.
TableS.
Table 4.
Table 5.
Table 6.
Table 7.
Table 8.
Table 9.
Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Table 15.
Tables
Chemical Leaks Evaluated with the FLIR GasFindIR™ MW Camera During
Laboratory Testing 15
Test Conditions Evaluated During Laboratory Testing 17
TVA Calibration Responses 22
TVA Calibration Check Samples 23
Confirmation of Detected Leaks by TVA 23
Known Leak Rate Test Results 24
Summary of Positive Control Check Responses 26
FLIR GasFindIR™ MW Method Detection Limits at 10 Feet Stand-off
Distance with a Cement Board Background 31
FLIR GasFindIR™ MW Method Detection Limits at 30 Feet Stand-off
Distance with a Cement Board Background 32
. FLIR GasFindIR™ MW Method Detection Limits at 10 Feet Stand-off
with a Curved Metal Gas Cylinder Background 33
. FLIR GasFindIR™ MW Method Detection Limits at 30 Feet Stand-off
Distance with a Curved Metal Gas Cylinder Background 34
. FLIR GasFindIR™ MW Range of Method Detection Limits and Overall Method
Detection Limit Variation (g/hr) 35
TA /(
. Summary of Detection Agreement Between FLIR GasFindIR MW Camera
and a Method 21 Portable Monitoring Device 36
TM
Summary of Field Testing Results Using the FLIR GasFindIR MW Camera 37
Summary of FLIR GasFindIR™ MW Camera Method Detection Limits and Percent
Agreement with a Method 21 Monitoring Device During Laboratory Testing 43
TM
Table 16. Summary of Field Testing Results Using the FLIR GasFindIR MW Camera 44
vn
-------
FIGURES
Figure 1. FLIR GasFindIR™ MW Camera 11
Vlll
-------
List of Abbreviations
ACC American Chemistry Council
AMS Advanced Monitoring Systems
CH4 Methane
DQI Data Quality Indicator
EPA Environmental Protection Agency
ETV Environmental Technology Verification
FLIR FLIR Systems, Inc.
ft Foot, feet
GC Gas Chromatography
g/hr Grams per hour
Hz Hertz
IR Infrared
kg/hr Kilogram per hour
LOD Limit of Detection
LW Longwave
mm Millimeter
mph Miles per hour
MW Midwave
NRMRL National Risk Management Research Laboratory
PID photoionization
ppmv Parts per million by volume
QA Quality assurance
QC Quality control
QMP Quality Management Plan
SF6 Sulfur hexafluoride
TCC Texas Chemical Council
TQAP Test Quality Assurance Plan
TVA Toxic Vapor Analyzer
U.S. United States
VOC Volatile organic compounds
°F Degrees Fahrenheit
IX
-------
Chapter 1
Background
The U.S. Environmental Protection Agency (EPA) supports the Environmental Technology
Verification (ETV) Program to facilitate the deployment of innovative environmental
technologies through performance verification and dissemination of information. The goal of the
ETV Program is to further environmental protection by accelerating the acceptance and use of
improved and cost-effective technologies. ETV seeks to achieve this goal by providing high-
quality, peer-reviewed data on technology performance to those involved in the design,
distribution, financing, permitting, purchase, and use of environmental technologies.
ETV works in partnership with recognized testing organizations; with stakeholder groups
consisting of buyers, vendor organizations, and permitters; and with the full participation of
individual technology developers. The program evaluates the performance of innovative
technologies by developing test plans that are responsive to the needs of stakeholders,
conducting field or laboratory tests (as appropriate), collecting and analyzing data, and preparing
peer-reviewed reports. All evaluations are conducted in accordance with rigorous quality
assurance (QA) protocols to ensure that data of known and adequate quality are generated and
that the results are defensible. The definition of ETV verification is to establish the performance
of a technology under specific, pre-determined criteria or protocols and a strong quality
management system. High quality data are assured through implementation of the ETV Quality
Management Plan. ETV does not endorse, certify, or approve technologies.
The EPA's National Risk Management Research Laboratory (NRMRL) and its verification
organization partner, Battelle, operate the Advanced Monitoring Systems (AMS) Center under
ETV. The AMS Center recently evaluated the performance of the GasFindIR1 Midwave (MW)
camera by FLIR Systems, Inc. (FLIR), a portable, passive infrared (IR) camera operating in the
spectral range of 3 to 5 micrometers.
10
-------
Chapter 2
Technology Description
TM
This verification report provides results for the verification testing of FLIR's GasFindIR MW
Following is a description of the FLIR GasFindIR™ MW camera technology (hereafter referred
to as FLIR GasFindIR™ MW), based on information provided by the vendor. The information
provided below was not verified in this test. Figure 1 shows the FLIR GasFindIR™ MW
camera.
The GasFindIR™ MW camera takes focal plane arrays and
optical systems that are tuned to very narrow spectral
infrared ranges to enable the camera to detect the energy
emitted from certain gases. Images are processed and
enhanced by the GasFindIR High Sensitivity Mode™
feature to show the presence of gases against stationary
backgrounds. Gases that are detectable by the GasFindIR™
camera appear on screen as smoke.
GasFindIR™ MW camera is designed for use in harsh
industrial environments and operates in wide temperature GasFindIR11V1 MW Camera
ranges. The GasFindIR™ MW camera is a real-time
infrared camera that scans at 30 hertz (Hz) or 30 images per second. The camera includes a 25-
millimeter (mm) wide-angle lens for scanning of a variety of components and operations. For
longer-range needs, 50-mm and 100-mm lenses are available from FLIR Systems.
Figure 1. FLIR
,TM
11
-------
Chapter 3
Test Design and Procedures
3.1 Test Overview
This verification test was conducted according to procedures specified in the Test/QA Plan for
Verification of Leak Detection and Repair Technologies^\TQAP) and adhered to the quality
system defined in the ETV AMS Center Quality Management Plan (QMP).(2) Battelle conducted
this verification test with support from British Petroleum (BP), Innovative Environmental
Solutions, Inc., The Dow Chemical Company, Sage Environmental Consulting, and Enthalpy
Analytical, Inc.
This verification test simulated gas leaks of various chemicals in a controlled laboratory
environment. The ability of the FLIR GasFindIR™ MW camera to qualitatively detect gas leaks
of select chemicals species by visual images under controlled environmental conditions -
including varied stand-off distances, wind speeds, and background materials - was verified and
the method detection limits under each test condition were determined. This passive IR camera
has not been evaluated under the ETV Program for other compounds or species other than those
tested under this verification test. The potential exists for the identification of other species that
have an IR absorbance feature(s) in this spectral range under ideal test conditions.
Additionally during laboratory testing, the ability of the FLIR GasFindIR™ MW camera to
qualitatively detect the gas leak by visual images relative to a quantitative concentration
measurement made by a portable monitoring device acceptable under U.S. EPA Method 21 -
Determination of Volatile Organic Compound (VOC) Leaks^ for the determination of VOC
leaks from process equipment was verified for each chemical at each test condition during
laboratory testing. During laboratory testing, acceptable under U.S. EPA Method 21 meant that
the portable monitoring device met all of the performance requirements of Section 6 in U.S. EPA
Method 21 with the exception of those requirements related to a specific leak definition
concentration specified in any applicable regulation. A specific leak definition concentration
was not used to qualify leaks during laboratory testing in a regulatory sense.
This verification test also verified the ability FLIR GasFindIR™ MW camera to detect gas leaks
of various chemicals relative to a portable monitoring device acceptable under U.S. EPA Method
21 under "real world" conditions at a chemical plant in Freeport, TX. During field testing,
acceptable under U.S. EPA Method 21 meant that the portable monitoring device met all of the
performance requirements of Section 6 in U.S. EPA Method 21; a specific leak definition
concentration of 500 parts per million by volume (ppmv) was utilized. Reference sampling was
conducted to determine the mass rate of specific chemical species emitted from each leaking
component observed with the FLIR GasFindIR™ MW camera and with the portable monitoring
device acceptable under U.S. EPA Method 21.
12
-------
This verification test of the GasFindIR™ MW camera was conducted October 20 through
October 24, 2008 at the BP research complex in Naperville, Illinois (laboratory testing) and
December 1 through December 5, 2008 at the Dow Chemical Company plants (field testing) in
Freeport, TX in compliance with the data quality requirements in the AMS Center Quality
Management Plan (QMP). The TQAP for this verification test indicated that field testing would
be conducted at two field sites. Due to production scheduling issues, a second field site could
not be obtained in a timely manner and this verification test was completed using only one field
test location. Confirmation from a second field site was obtained during the writing of these
reports and field testing occurred outside of this verification test in March 2010. The reader is
encouraged to contact either FLIR Systems or the Texas Chemical Council (TCC) to obtain the
results of testing completed at the second field site. As indicated in the test/QA plan, the testing
conducted satisfied EPA QA Category III requirements. The test/QA plan, the verification
statement, and this verification report were reviewed by the following experts.
• Dave Fashimpaur, BP
• Julie Woodward, Dow Chemical
• Jim Griffin, American Chemistry Council
• Christina Wisdom, Texas Chemical Council
• EbenThoma, U.S. EPA.
One technical expert came to the laboratory testing, and one technical expert came to the field
site to observe testing. Verification testing was conducted by appropriately trained personnel
following the safety and health guidelines for BP and Dow's facilities.
The GasFindIR™ MW camera was verified by evaluating the following four parameters.
• Method detection limit - The minimum mass leak rate that three separate individuals can
observe using the GasFindIR™ MW camera under controlled laboratory conditions. Thi
parameter was not evaluated during the field testing phase.
• Detection of chemical gas species relative to a portable monitoring device - The ability
of the GasFindIR™ MW camera to qualitatively detect a gas leak by visual images
relative to a quantitative concentration measurement made by a portable monitoring
device acceptable under U.S. EPA Method 21. This parameter was evaluated in both the
laboratory and field testing phases.
• Confounding factors effect - Background materials, wind speed, and stand-off distance
were carefully controlled during laboratory testing to observe their effects on the method
detection limit. During field testing, these variables as well as meteorological conditions
were recorded.
• Operational factors - Factors such as ease of use, technology cost, user-friendliness of
vendor software, and troubleshooting/downtime were evaluated.
Due to unavailability of a second FLIR GasFindIR™ MW camera during the laboratory and field
testing portions of this verification test, inter-unit comparability could not be completed during
laboratory and field testing.
13
-------
A FLIR GasFindIR™ LW camera was used during a portion of both the laboratory and field
testing. This camera was not evaluated against the entire suite of chemicals used in the
laboratory portion of this verification testing; rather the vendor used the FLIR GasFindIR™ LW
camera for 1,3 -butadiene, acetic acid, and acrylic acid because these compounds have an
absorption peak within the 10 to 11 micrometer operating wavelength of the FLIR
GasFinderlR™ LW camera. The camera was evaluated in the field for all chemical gas leaks
identified, regardless of whether the gas leak contained compounds with an absorption peak
within the 10 to 11 micrometer operating wavelength of the FLIR GasFinderlR1 LW camera on
the days that the camera was available to the verification test team. Because the FLIR
GasFindIR™ LW camera was not used during the entire portion of the laboratory and field
testing phases of this verification test, test results obtained with the FLIR GasFindIR™ LW
camera are not included in the body of this verification report. Rather, the results obtained with
TA/f
the FLIR GasFindIR LW camera are included as an appendix to this report for reference by
the reader.
Prior to the start of the verification test, FLIR setup the FLIR GasFinderlR™ MW camera
according to their recommended configuration for optimal performance.
3.2 Experimental Design
3.2.1 Detection of a Chemical Gas Leak Using FLIR GasFindIRrM
During both the laboratory testing and field testing, the FLIR GasFindIR™ MW camera was
operated by a representative of FLIR. This verification test used two additional confirming
individuals beyond the camera operator to confirm the observation of a leak in an effort to
eliminate potential operator bias. The two additional confirming individuals were the Battelle
verification test coordinator and an additional verification test team member. The use of three
individuals to observe a chemical leak with the FLIR GasFindIR™ MW camera is not standard
practice when using the FLIR GasFindIR™ MW camera; typical operation relies on a single
camera operator to observe the presence of a chemical gas leak.
The detection of a chemical gas leak in either the laboratory or field setting was determined by
the camera operator, as well as two confirming individuals who reported the results qualitatively
as either "detect" or "non-detect" observation. All three individuals must have agreed on the
results for the observation to be considered a "detect." When all three individuals did not agree
on a detection, the observation was reported as a "non-detect." A non-detect was also recorded if
the camera operator did not observe a detection (i.e., no confirmation of a non-detect was
performed). Each observation was conducted using the eye piece of the FLIR GasFindIR™ MW
camera.
The TQAP for this verification test required that camera observers have five seconds to identify
the origin of the leak or be able to track the plume back to the leaking component when
observing chemical gas leaks (i.e., identify the source of the leak). However, during laboratory
and field testing, the observers were allowed two minutes. This change was made during
laboratory testing to account for system hysteresis and upon discovering that several liquid
compounds at very low flow rates did not generate a continuous plume. Rather, the leaks were
observable as intermittent "puffs" of chemicals emanating from the valve at a frequency on the
order of 10 seconds to two minutes. This time lag resulted from lower syringe pump feed rate
settings, and the reduced hot nitrogen carrier gas volume flow rates.
14
-------
3.2.2 Method Detection Limit
Method detection limits were determined only in the laboratory portion of this verification test.
To determine the method detection limit, a known mass leak rate from the packing of a 1-inch
valve attached to certified gas cylinders and calibrated flow meters was set at a nominally
detectable level either specified by the vendor's limit of detection (LOD) for a particular test
condition, or based on previous literature by Panek et al.(4) When all three observers identified
the leak, the leak rate was reduced by the testing staff using calibrated flow meters. Once a leak
rate that was not identifiable by all three people was reached, the mass emission rate was again
increased using the calibrated flow meters to the level where all three could again identify the
leak using the FLIR technology (i.e. passive infrared imager). This rate was then established as
the method detection limit for the passive infrared imager under the tested conditions. This
process was completed for every testing trial identified in Section 3.2.3. Table 1 identifies the
type of chemical leaks evaluated with the FLIR technology during laboratory testing.
Table 1. Chemical Leaks Evaluated with the FLIR GasFindIR™ MW Camera During
Laboratory Testing
Chemical Chemical Group
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Methyl ene chloride
(di chl or om ethane)
Ethyl ene
Methanol
Pentane
Propane
Styrene
Olefm
Acetate
Acid
Aromatic
Chlorinated
Olefm
Alcohol
Alkane
Alkane
Aromatic
The TQAP for this verification test stated that propylene di chloride (1,2-dichloropropane) and
hydrochloric acid would also be used during laboratory testing. The stock solution of propylene
di chloride was suspected by laboratory personnel of having been cross-contaminated by a
different chemical compound. A second stock solution of propylene di chloride could not be
obtained from a chemical vendor before the conclusion of laboratory testing. Thus, propylene
dichloride was not used during laboratory testing. The laboratory staff also expressed concerns
of causing damage to the delivery syringe in the chemical delivery system with the use of
hydrochloric acid. Because hydrochloric acid could not be delivered through the chemical
delivery system without causing damage to the system, a known leak rate could not be generated
during laboratory analysis, therefore hydrochloric acid was not evaluated.
3.2.3 Confounding Factors
Because passive IR imagers such as the FLIR technology rely on the physical characteristics of
the environment and the molecules being imaged to create an image viewed by the operator (via
temperature/emissivity differences between naturally occurring ambient IR radiation and the
thermal emission or absorption of the leaking gas), environmental characteristics may confound
the measurement. For example, if there is not sufficient thermal emission or absorption by the
15
-------
leaking gas, the passive IR imager may not be able to detect a leak against the ambient thermal
background.
During laboratory testing, experimental factors of background materials, wind speed, and stand-
off distance were altered for each chemical tested. These experimental factors were chosen,
because the performance of passive imagers is dependent on physical characteristics of the leak,
atmospheric conditions, and background materials. The change of background material
demonstrates the ability of the FLIR GasFindIR™ MW camera to detect the leak with a
background scene similar to petrochemical process piping and vessels (curved metal gas
cylinders) and with a background that is different than the leaking component and more uniform
in nature (cement board - representing control buildings, sidewalks, and other uniform flat
background surfaces). The wind speed variations and the stand-off distances inform on the
atmospheric and optical pathway effects on the method detection limit, and in turn on real-world
limitations. Table 2 presents the specific test conditions evaluated during laboratory testing.
It was originally intended that all test conditions would be completed for all chemicals; however,
it was not possible for 1,3-butadiene, acrylic acid, methylene chloride, methane, and styrene for
the following reasons.
Previous testing of the FLIR GasFindIR™ MW camera using methane had been completed by
the laboratory facility outside of the verification test. Consequently, methane was used during
test equipment setup to confirm that the equipment produced method detection limits for
methane that were consistent with those produced during previous testing by the laboratory.
16
-------
Table 2. Test Conditions Evaluated During Laboratory Testing
Chemical Species
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Methylene chloride
Ethylene
Methanol
Pentane
Propane
Styrene
Laboratory Test Conditions
1
&
T3
g
^r-j
1
o ^
„ fi
b fi
S 2
3 4?
t stand-off dis
ent board bac
'l~l S
o <3
S
i C3
fS *-!
S ^
+3 o"
t stand-off dis
ed metal gas
^H ^
o s
^H O
^
^
-------
all personnel in respirators. However, documentation of respirator fit testing was not available
for test team members. Respirators could not be used without this documentation.
To address this problem, the leaking valve was placed next to the side hood during wind speed
testing and testing of those chemical compounds which are liquids at standard conditions
commenced in order of increasing boiling point. Upon completion of wind testing for acetic
acid, the laboratory had a slight odor of acetic acid. This indicated to laboratory personnel that
locating the leaking valve next to the side hood during wind speed testing did not adequately
capture all of the chemical compounds exhausting from the test system. Rebalancing of the hood
was attempted, but the problem was caused by an increase in general building exhaust, rather
than at the local hoods. At this point, wind speed testing of 1,3-butadiene and styrene was
abandoned because these compounds have higher chemical toxicity and exposure by the
verification test team, vendor, and laboratory staff to these compounds would have occurred
during wind speed testing.
During methylene chloride testing, several of the wind speed tests and background tests were not
conducted because the method detection limit for lower wind speed (or background) conditions
exceeded the highest reliable flow rate capable of being provided by the chemical leak delivery
system at test conditions which were expected to produce a lower method detection limit (refer
to Section 6.3 for discussion of the observed influence of confounding factors). For example, a
5-mph wind speed test was not conducted at a 10 ft stand-off distance with a cement board
background because the method detection limit exceeded the highest reliable flow rate of the
chemical delivery system for the 10 ft stand-off distance, cement board background, and 2.5-
mph.
3.2.4 Detection of a Chemical Gas Species Relative to a Portable Monitoring Device
The detection of a single chemical gas leak in either the laboratory or field environments was
determined by the operator as well as two confirming individuals as previously described in
Section 3.2.1 and reported qualitatively as either "detect" or "non-detect."
During laboratory testing a portable monitoring device, a factory-calibrated Industrial Scientific
IB RID MX6 with photoionization (PID) sensor and SP6 motorized sampling pump, acceptable
under U.S. EPA Method 21, sampled the leak after the method detection limit was determined
for the specified test conditions. During laboratory testing, "acceptable under U.S. EPA Method
21" meant that the PID met all of the performance requirements of Section 6 in U.S. EPA
Method 21 with the exception of those requirements related to a specific leak definition
concentration specified in any applicable regulation. A specific leak definition concentration
was not used to qualify leaks during laboratory testing in a regulatory sense.
During field testing a portable monitoring device, a Thermo-Environmental Toxic Vapor
Analyzer (TVA), acceptable under U.S. EPA Method 21 was used to screen each leaking
component as part of the bagging reference method used. During field testing, "acceptable under
U.S. EPA Method 21" meant that the TVA met all of the performance requirements of Section 6
in U.S. EPA Method 21; a specific leak definition of 500 ppmv was utilized.
3.2.5 Field Testing Procedures
Field testing was conducted to allow for performance evaluation under "real world" conditions.
Chemicals that were tested in the laboratory were targeted for evaluation at the field sites. The
mass flow rates of field leaks were quantitatively determined by a reference method called EPA
18
-------
Protocol for Equipment Leak Emission Estimates,^ referred to as the "bagging method."
Specific details and procedures for this reference method are provided in the TQAP for this
verification test. This method involves completely enclosing the leak with non-permeable
material, collecting the leak with ambient air entering the bag, and performing mass
measurement of the bagged leak by an analytical method. Only those leaks above the field test-
assigned 500 ppmv leak definition concentration, as measured by the Thermo-Environmental
TV A, were observed with the passive infrared imagers and collected as reference samples under
this verification test.
The verification test team moved through the plant screening for possible leaking components
using the Thermo-Environmental TVA. Once a leak was detected with the portable monitoring
device, leak characteristics and environmental factors such as type of component, background
material, temperature, and time were recorded qualitatively. Meteorological data were retrieved
from the nearest meteorological station, which was on Dow Chemical's site. As space permitted,
the camera operator took readings at three stand-off distances (10, 30, and greater than 30 ft if
possible). Every reading was verified by an additional two confirming individuals and recorded
as either "detect" or "non-detect" as specified in Section 3.2.1. Once the camera had scanned the
leak, the bagging team members (Sage Environmental Consulting) commenced collecting
duplicate reference samples of the leak into evacuated SUMMA canisters. Reference sampling
concluded with a final screening by the TVA to verify that the leak concentration had not
changed from the beginning to the end of testing the component. Only those leaks which
showed less than a 20% difference between the pre- and post-screening with the TVA were
considered consistent enough to report in the results without a data qualifier. The concentration
of the collected reference samples was determined according to the analytical method in U.S.
EPA Method 18 -Measurement of Gaseous Organic Compound Emissions by Gas
Chromatography.^ Upon conclusion of the five days of field testing, all reference samples were
shipped to Enthalpy Analytical, Inc. for U.S. EPA Method 18 analysis.
3.3 Qualitative Evaluation Parameters
Operational factors such as maintenance needs, ease of use, data output, and software
requirements were documented based on observations by Battelle.
19
-------
Chapter 4
Quality Assurance/Quality Control
QA/quality control (QC) procedures were performed in accordance with the QMP for the AMS
Center and the TQAP for this verification test. As noted throughout Chapter 3, there were
deviations from the TQAP, but the work was performed as described in the previous sections.
None of the deviations from the test/QA plan resulted in any adverse impacts on the quality of
the data produced by this verification test. QA/QC procedures and results are described in the
following subchapters.
4.1 Reference Method Quality Control
Laboratory testing did not use a specified reference method for determining the leak rate of the
test conditions. Rather, certified gas cylinders and laboratory grade liquid compounds were used
with calibrated flow meters and a calibrated syringe pump to generate a known leak rate in terms
of mass per unit time from the leaking valve. As a laboratory QC measure, laboratory personnel,
randomly and without the knowledge of the camera operator or the additional confirming
individuals, increased or decreased the mass leak rate to reduce the opportunity to predetermine
an outcome. In addition, laboratory blanks (i.e., pure nitrogen gas) and replicate tests were used
to reduce uncertainties and verify method detection limits established in prior tests.
The field testing portion of this verification test used accepted methods to generate reference
samples. Reference samples were collected using EPA Protocol for Equipment Leak Emission
Estimates and the concentrations of compounds in the collected reference samples were
determined according to the analytical method in U.S. EPA Method 18 Measurement of Gaseous
Organic Compound Emissions by Gas Chromatography
The quality of the reference measurements collected during field testing was assured by
adherence to the requirements of the data quality indicators (DQIs) and criteria for the reference
collection and analytical method critical measurements, including requirements to perform initial
calibrations and calibration checks of the portable monitoring device acceptable under U.S. EPA
Method 21, confirming the leak rates changed less than 20% before and after bagging, assessing
the bias and accuracy of the bagging procedure, and assessing the bias and accuracy of the gas
chromatography (GC) laboratory analysis by developing calibration curves traceable to certified
gas standards, and performing positive and negative control checks. The following sections
present key data quality results from these methods.
20
-------
4.1.1 Bias and Accuracy of Sample Screening Measurements Using Portable Monitoring
Device
A DQI is established in the TQAP for this verification test for the bias and accuracy of sample
screening measurements using a portable monitoring device. This DQI is assessed by
performing calibrations of the Thermo-Environmental TVA used to screen leaking components
during the field portion of the verification test and analyzing calibration check samples. During
laboratory testing the portable monitoring device was an Industrial Scientific IBRID MX6 with
PID sensor and SP6 motorized sampling pump which was supplied calibrated from the
instrument supplier; per the TQAP for this verification test, no additional calibrations were
performed during laboratory testing.
Calibration of the TVA was conducted using various levels of certified methane (CH/^-in-air gas
standards. The TQAP for this verification test required the use of five calibration points (an un-
spiked gas standard plus four additional concentrations); however, only three additional gas
standard concentrations were obtained. Because component leaks were only bagged as reference
samples if their concentration was greater than 500 ppmv and because the calibration response of
the TVA was evaluated using an un-spiked gas standard (0 ppmv) and three additional
concentrations of gas standards (500, 1000, and 9600 ppmv) thereby bounding the 500 ppmv
reference sample bagging threshold, there was no effect on data quality.
The calibration response of the TVA was analyzed at the start and end of each verification test
day or if the overall TVA sensitivity changed by greater than 10% (based on the calibration
check data, which are presented in Table 5). The minimum acceptance criterion for this
reference method DQI was that the TVA calibration response must agree within 10% of the
concentration of each gas standard. Table 3 presents the results of all TVA calibration responses
collected during this verification test. Inspection of the data present in Table 3 shows that all
calibration response measurements were confirmed to be within 10% of the calibration gas
standard concentration.
The TQAP for this verification test required that a calibration check sample be analyzed using
one concentration of the calibration gas standards at a minimum frequency of 5% of all bagged
reference samples collected. Sixteen calibration check samples were analyzed with the TVA
during the course of field testing and nine duplicate reference samples were collected resulting in
a calibration check sample frequency of 178% of all bagged reference samples collected (i.e., 16
calibration check samples completed during the collection of nine duplicate reference samples).
These checks were performed more frequently to ensure no drifting of the instrument occurred
during downtimes to ensure optimum performance. The minimum acceptance criterion specified
in the TQAP for this verification test is that the check standard must be within less than or equal
to a 10% change in response from the previous calibration of the TVA. If the calibration check
sample showed a change in response greater than 10%, then recalibration of the TVA was
performed and any affected reference sample components collected would be rescreened.
During this verification test, calibration check samples were performed using a certified 500
ppmv CH/j-in-air gas standard. Table 4 presents the results of all calibration check standards
performed during verification testing. Inspection of the data presented in Table 4 indicate that
reference samples 08 A and 08B should have been rescreened after recalibration of the TVA and,
therefore, are considered suspect data and reported with a data qualifier.
21
-------
Table 3. TVA Calibration Responses
Date [Time]
12/1/2008 [13:33](a)
12/2/2008 [09:01]
12/2/2008 [14:08]
12/2/2008 [16:05]
12/3/2008 [08:41]
12/3/2008 [09:30]
12/3/2008 [10: 12]
12/3/2008 [17:06]
12/4/2008 [10:04]
12/4/2008 [13:20]
12/4/2008 [16: 12]
12/4/2008 [17:23]
12/5/2008 [08:59]
12/5/2008 [11:20]
12/5/2008 [14:01]
Calibration
0
TVA Output
Concentration
(ppmv CH4)(b)
0.70
0.40
1.0
1.0
0.80
0.70
0.80
0.60
0.60
ND
0.60
0.20
0.60
1.2
0.20
Gas Standard Concentration (ppmv
500
TVA
0.40
-0.80
1.2
5.6
-1.4
-0.60
-1.2
-7.2
-0.60
ND
-0.80
-1.4
ND
4.0
3.4
1000
Calibration Response (as %
-1.3
-0.10
1.0
4.2
ND
-4.4
-0.60
-8.2
-0.30
-0.10
-1.5
-1.7
-0.70
3.0
3.3
CH4)
9600
Error)(c)
-0.80
-0.60
2.1
4.2
-0.70
-4.9
0.10
-8.0
-1.0
-0.30
-1.0
-1.1
-0.70
-8.3
-3.1
(a) An end-of-day TVA response was not collected on 12/1/2008. Data for leak location 1 is included but flagged
because there are acceptable reference and bagging measurements.
(b) Concentration data presented for un-spiked gas standard, since % error calculation is not possible. This point is
used in calibrating the Thermo-Environmental TVA.
(c) Percent (%) error is calculated as [(TVA calibration response, ppmv CH4 - Calibration Gas Standard
Concentration, ppmv CH4)/ Calibration Gas Standard Concentration, ppmv CH4] x 100%.
ND - Not detected
4.1.2 Confirmation of Detected Leaks
A DQI is established in the TQAP for this verification test for the confirmation of detected leaks.
This DQI is assessed by analyzing the concentration of a leaking component before and after
bagging the component. These measurements were completed for all leaking components which
were bagged and collected as reference samples. The acceptance criterion for this DQI is that
the pre and post screening measurements collected with the TVA agree within 20%. Table 5
presents the results of all pre- and post-bagging measurements completed during the collection of
reference samples.
4.1.3 Bias and Accuracy of Enclosure Equilibration Gas
A DQI is established in the TQAP for this verification test for bias and accuracy of the enclosure
equilibration gas. This DQI requires that if the blow-through bagging procedure is used to
collect reference samples, then the equilibration gas in the bag is collected and analyzed for
contamination prior to collection of reference samples. During the verification testing, reference
samples were collected using the vacuum-method which does not require the use of an
equilibration gas; therefore, this DQI was not applicable.
22
-------
Table 4. TVA Calibration Check Samples
Date [Time]
Calibration Check
Response
(as % Error)(a)
Comments
12/2/2008 [11:17]
12/2/2008 [12:15]
12/2/2008 [14:05]
12/2/2008 [14:08]
12/2/2008 [15:10]
12/2/2008 [15:43]
12/3/2008 [9:23]
12/3/2008
12/3/2008
12/3/2008
12/3/2008
12/4/2008
12/4/2008
[10:30]
[11:32]
[13:57]
[15:45]
[11:43]
[13:23]
0.40
-5.2
-16
1.2
1.4
2.0
64
0.80
-0.60
0.60
0.60
1.6
-17
Recalibration only. No rescreening necessary because no
reference samples had been collected between this
calibration check sample and TVA calibration.
Found leak; recalibrated only. No rescreening necessary
because reference samples had yet to be collected this
day.
Recalibration only. No rescreening necessary because no
reference samples had been collected between this
calibration check sample and the previous check.
12/4/2008 [15:30]
12/4/2008 [17:25]
12/5/2008 [10:38]
24
-1.4
-3.0
Recalibration only. Reference samples 08A and 08B
were inadvertently not rescreened and are therefore
considered suspect and results reported with qualifier.
(a) Percent (%) error is calculated as [(TVA calibration check response, ppmv CH4 - Calibration Gas Standard
Concentration, 500 ppmv CH4)/ Calibration Gas Standard Concentration, 500 ppmv CH4] x 100%.
Table 5. Confirmation of Detected Leaks by TVA
Reference
Sample
Numbers
01C, 01D
02A, 02B
03A, 03B
05A, 05B
06A, 06B
07A, 07B
08A, 08B
09A, 09B
10A, 10B
Concentration
Pre-bagging
>100,000(a)
20,500
>100,000(a)
>100,000(a)
18,000
18,000
8,000
800
>100,000(a)
Measured by TVA
Post-bagging
>100,000(a)
20,500
>100,000(a)
>100,000(a)
23,000
17,000
8,000
870
>100,000(a)
(ppmv CH4)
Relative %
Difference^
0%
0%
0%
0%
24%
5.7%
0%
8.4%
0%
Comments
Data is considered suspect and
results reported with qualifier.
(a) The concentration of the leak at the component was high enough to cause the TVA to flameout. Concentration
estimated as greater than 100,000 ppmv CH4.
(b) Relative percent (%) difference calculated using the following calculation:
23
-------
2 x ?re — bagging concentration — Post bagging concentration
Relative % difference = ^-^ ^^ X 100%
?re — bagging concentration + Post bagging concentration
4.1.4 Bias and Accuracy of Bagging Procedure
A DQI is established in the TQAP for this verification test for the bias and accuracy of the
bagging procedure. This DQI is assessed by bagging an artificial leak at a known rate in the
middle of the analytical calibration curve. The procedure followed is that specified in U.S. EPA
Protocol for Equipment Leak Emission Estimates using certified CH/i-in-air gas standards and
calibrated flow meters. This DQI indicator was assessed at the beginning and end of the week of
field sampling. An acceptance criterion of 80 to 120% recovery is required for the bagging
equipment to pass the known leak rate test. Table 6 presents the results of the known leak rate
test. As shown in Table 6, this DQI was met before and after reference sampling.
Table 6. Known Leak Rate Test Results
Date [Time]
Leak Rate
Level
Emission Rate
(kilogram per hour [kg/hr] CH4)
Theoretical
Measured
% Recovery(a)
P re- Test
11/28/2008 [12:45]
11/28/2008 [12:20]
Low
High
4.31xlO'4
1.75X10'3
4.23x 10'4
1.60X10'3
98%
91%
Post-Test
12/5/2008 [14:35]
12/5/2008 [14:43]
Low
High
1.25X10'3
2.43 x 10'3
1.32 xlO'3
2.50 xlO'3
106%
103%
(a) Percent (%) Recovery is calculated as (measured emission rate, kg/hr CH4) / (theoretical emission rate, kg/hr
CH4) x 100%
4.1.5 Bias and Accuracy of Gas Chromatography Analytical Method
A DQI is established in the TQAP for this verification test for the bias and accuracy of the GC
analytical method used to quantify the concentration of leaks collected during reference
sampling. This DQI was assessed through initial calibration, and by performing positive and
negative control samples. These assessments are discussed in the following paragraphs.
Initial Calibration. Initial calibration of the GC was conducted by using various levels of
certified calibration gases starting with an un-spiked gas standard and then a minimum of four
additional concentrations of gas standards. The TQAP for this verification test required that the
initial calibration be performed at the start and end of every analytical sequence or if overall
instrument sensitivity changed by greater than 10%. To ensure accuracy of the initial calibration,
the instrument must be calibrated using certified gas standards. The minimum acceptance
criteria specified for this assessment is that all gas standards must be within 2% of their certified
value.
The analytical laboratory that performed the GC analytical method (Enthalpy Analytical, Inc.)
purchased gas standards with certification accuracies of ± 2%, as specified by the gas supplier.
In addition, the GC analytical laboratory produced diluted gas standards from these purchased
standards using a gas dilution system compliant with U.S. EPA Method 205(7) which specifies
gas dilution systems must produce calibration gases whose measured values are within ± 2% of
the predicted levels from a certified gas standard.
24
-------
Positive Control Checks. The TQAP for this verification test required that positive control
checks be performed at a minimum frequency of 10% of all samples tested using one
concentration of calibration gas standard. The minimum acceptance criteria for positive control
checks is that the positive control check response is less than or equal to a 10% change in
response from the initial calibration after adjustment of the overall instrument sensitivity. Forty
sample measurements were conducted by the GC analytical laboratory using triplicate injections
and 19 positive control checks were performed exceeding the minimum frequency of 10% of
samples tested. The results of the positive control checks are provided in Table 7. As
demonstrated by Table 7, all positive control checks met this acceptance criterion.
Negative Control Checks. The TQAP for this verification test required that negative control
checks be performed at a minimum frequency of one out of every 10 samples tested. The
minimum acceptance criterion for this assessment is that all negative control responses must
remain lower than the lowest calibration standard for the chemical analyzed. Forty sample
measurements were conducted by the GC analytical laboratory using triplicate injections and
four negative control checks were performed meeting the minimum frequency of one negative
control check per 10 samples analyzed. All negative control checks performed were non-detect
for the compounds analyzed indicating an analytical result below the method detection limit for
the compound. The method detection limit for methane, ethylene, styrene, benzene,
1,3-butadiene, methylene chloride, and propylene dichloride was 1.00 ppmv for each compound.
4.2 Audits
Two types of audits were performed during the verification test, a technical systems audit (TSA)
of the verification test procedures, and a data quality audit. Because of the nature of bagging
reference method, a performance evaluation audit, as is usually performed to confirm the
accuracy of the reference method, was not applicable for this verification test. Audit procedures
for the TSA and the data quality audit are described further below.
4.2.1 Technical Systems Audit
The Battelle AMS Center Quality Manager performed a TSA during both the laboratory and
field testing portions of this verification test to ensure that the verification test was performed in
accordance with the QMP for the AMS Center and the test/QA plan.
The TSA of the laboratory portion of the verification test was performed on October 22, 2008.
During this TSA, the Battelle AMS Center Quality Manager observed the test procedures used to
determine method detection limits and the response of the Industrial Scientific IBRID MX6 with
PID sensor and SP6 motorized sampling pump at the each method detection limit. These
procedures were observed during some of the testing conducted with acrylic acid, benzene,
dichloromethane (methylene chloride), and styrene. The TSA of the field testing portion of the
verification test was performed on December 3, 2008. During this TSA, the Battelle AMS
Center Quality Manager observed the procedures of the bagging reference method, including the
confirmation of the detected leaks by means of pre- and post-bagging screening of the leaking
component with the Thermo-Environmental TVA, construction of the bagging enclosure, and
duplicate reference sample collection, as well as audited the observations of the leak component
with camera. In addition, the Battelle AMS Center Quality Manager observed both the
performance of a calibration drift check and recalibration as well as an end-of-day calibration
response check of the Thermo-Environmental TVA.
25
-------
Table 7. Summary of Positive Control Check Responses
Positive Control
Check Sample ID
GC100pgl67#2
GC100pgl67#2
GC100pfl69F#4
GC100pfl69F#4
GC100pfl69F#4
GC100pfl69F#4
GC100pfl69F#4
GC100pfl69F#4
GC102pg44 #3
GC102pg44 #3
GC100pgl69 #2
GC100pgl69 #2
GC100pgl69 #3
GC100pgl69#4R
GC100pgl69#4R
GC102pg52 #4
GC102pg52 #4
GC102pg52 #4
GC102pg52 #4
Compounds
Measured by GC
Method
Benzene
Benzene
Ethylene
1,3 -butadiene
Ethylene
1,3 -butadiene
Ethylene
1,3 -butadiene
Ethylene
1,3 -butadiene
Ethylene
1,3 -butadiene
Ethylene
1,3 -butadiene
Methane
Methane
Methane
Methane
Methane
Methane
Methane
Pentane
Methylene chloride
Benzene
Propylene dichloride
Styrene
Pentane
Methylene chloride
Benzene
Propylene dichloride
Styrene
Pentane
Methylene chloride
Benzene
Propylene dichloride
Styrene
Pentane
Methylene chloride
Benzene
propylene dichloride
Styrene
Expected
Response
(Picoampere
Second)
39.8
39.8
13.7
27.3
13.7
27.3
13.7
27.3
13.7
27.3
13.7
27.3
13.7
27.3
22.4
22.4
7.10
7.10
15.9
15.9
15.9
122
17.6
148
36.1
31.9
122
17.6
148
36.1
31.9
67.7
10.2
82.0
21.0
17.8
67.7
10.2
82.0
21.1
17.8
Actual Response
(Picoampere
Second)
39.3
39.0
13.8
26.9
13.7
26.7
13.5
26.3
13.4
25.7
13.7
26.9
13.8
27.2
22.8
22.7
6.95
6.73
15.3
15.5
15.8
127
17.7
150
35.4
34.0
125
17.3
147
34.4
32.7
67.5
9.86
79.5
20.5
18.4
70.3
10.2
82.3
21.2
18.6
Percent Error(a)
-1.1%
-1.9%
+0.39%
-1.6%
-0.61%
-2.4%
-1.6%
-3.9%
-2.4%
-5.8%
-0.44%
-1.5%
+0.39%
-0.43%
+1.6%
+1.3%
-2.1%
-5.3%
-3.4%
-2.5%
-0.39%
+4.2%
+0.60%
+1.1%
-2.1%
+6.7%
+2.7%
-1.9%
-0.75%
-4.6%
+2.4%
-0.35%
-3.4%
-3.1%
-2.9%
+3.8%
+3.7%
+0.16%
+0.35%
+0.49%
+4.5%
(a) Percent error is calculated as [(Actual Peak Response, peak area - Expected Response, peak area)/ Expected
Response, peak area] x 100%.
26
-------
The TSA of both the laboratory and field testing portions resulted in one finding and one
observation. The finding identified that only one field test (at a chemical plant) has been
conducted as part of this verification test as opposed to the two field sites (one a chemical plant
and the other a petrochemical plant) identified in the TQAP for this verification test. The
observation noted documentation errors and improvements to the manner in which data were
recorded were discussed on-site with the Verification Test Coordinator; immediate changes
based on the discussed improvements were implemented.
A TSA report was prepared, and a copy was distributed to the EPA AMS Center Quality
Manager.
4.2.2 Data Quality Audit
Records generated in the verification test received a one-over-one review before these records
were used to calculate, evaluate, or report verification results. Data were reviewed by a Battelle
technical staff member involved in the verification test. The person performing the review added
his/her initials and the date to a hard copy of the record being reviewed.
100% of the verification test data were reviewed for quality by the Verification Test Coordinator,
and at least 10% of the data acquired during the verification test were audited. The data were
traced from the initial acquisition, through reduction and statistical analysis, to final reporting to
ensure the integrity of the reported results. All calculations performed on the data undergoing
the audit were checked.
The data quality audit resulted in four findings (on three separate topics) that were addressed
related to the documentation of the number of confirming individuals at the method detection
limits in the laboratory phase raw data, exclusion from the verification report of concentration
measurements made by the PID sensor for dichloromethane (methylene chloride), methanol, and
propane during the laboratory phase of this verification test, and data transcription errors.
A data audit report was prepared, and a copy was distributed to the EPA AMS Center Quality
Manager.
27
-------
Chapter 5
Statistical Methods
The statistical methods used to evaluate the quantitative performance factors listed in Section 3.2
are presented in this chapter. Qualitative observations were also used to evaluate verification test
data.
5.1 Method Detection Limit
The method detection limit was assessed using the procedures described in Section 3.2.2 and the
TQAP for this verification test. The overall detection limit variation was calculated as the
standard deviation of the method detection limits determined under all the conditions tested for
each chemical of interest. The equation for standard deviation is as follows:
5,, =
n
k-l
(1)
where Sx is the standard deviation of all method detection limits determined for chemical x, n is
the number of replicate samples, Ck is the leak rate measured for the Mi sample, and is the
average leak rate of the replicate samples. If the sample sizes were small (n < 10), standard
deviations provide a biased estimate of variability. Therefore the range is provided when there
were fewer than 10 samples collected.
5.2 Percent Agreement
Percent agreement was used to assess the agreement between the FLIR GasFindlR™ cameras and
the monitoring device acceptable under U.S. EPA Method 21 in the laboratory for each compound
tested. The inverse of the percent agreement is the percentage of the results that the technology
would detect a leak when U.S. EPA Method 21 would not. The equation for percent agreement is as
follows:
.4
Percent Agreement — — K 100%
* T
where A the number of tests that both units agree and Tis the total number of tests. To determine if
both the monitoring device acceptable under U.S. EPA Method 21 and the FLIR GasFindlR™
camera agreed, the method detection limits at each test condition were first reviewed. If the
method detection limit of the FLIR GasFindlR™ camera was below the highest reliable flow rate
of the chemical delivery system (reported as <), then the FLIR GasFindlR camera was noted
28
-------
as being able to detect the chemical gas leak under those specified test conditions. Similarly, if
the method detection limit of the FLIR GasFindIR™ camera was equal to or above the highest
reliable flow rate of the chemical delivery system (reported as >), then the FLIR GasFindIR™
camera was noted as not being able to detect the chemical gas leak under those specified test
conditions.
Next, the response of the monitoring device acceptable under U.S. EPA Method 21 was
reviewed for the same test conditions. If the monitoring device acceptable under U.S. EPA
Method 21 produced a response greater than zero, the monitoring device was considered capable
of detecting the chemical gas leak. Similarly, if the monitoring device acceptable under U.S.
EPA Method 21 produced a response equal to zero, the monitoring device was considered
incapable of detecting the chemical gas leak.
The responses of the FLIR GasFindIR™ MW camera and the monitoring device acceptable
under U.S. EPA Method 21 under the same test conditions were compared. If both the FLIR
GasFindIR™ MW camera and the monitoring device acceptable under U.S. EPA Method 21
proved capable of detecting the chemical gas leak, then both units were considered to have
agreed under the specific test condition. Likewise, if either the FLIR GasFindIR™ MW camera
or the monitoring device acceptable under U.S. EPA Method 21 proved incapable of detecting
the chemical gas leak under the specified test conditions, then the units were considered to have
TA if
disagreed. Test conditions, under which a response from the either the FLIR GasFindIR MW
camera or the monitoring device acceptable under U.S. EPA Method 21 were not obtained, were
excluded from the comparison.
29
-------
Chapter 6
Test Results
As mentioned previously, this verification test included both quantitative and qualitative
evaluations. The quantitative evaluation was conducted to assess the method detection limits of
the FLIR GasFindIR™ MW camera, the detection of chemical gas species relative to a portable
monitoring device acceptable under U.S. EPA Method 21, as well as, by testing the influence of
confounding factors. The qualitative evaluation was performed to document the operational
aspects of FLIR GasFindIR M MW camera used during verification testing. The following
sections provide the results of the quantitative and qualitative evaluations.
6.1 Method Detection Limit
The method detection limit of each chemical compound was determined according to the
procedures discussed in Section 3.2.2. Table 8 through Table 11 present the method detection
limits of each chemical compound determined during laboratory testing. Table 8 through Table
11 identify each test condition evaluated (i.e., stand-off distance, background material, and wind
speed), the temperatures of the laboratory and of the chemical leak, the response of the portable
monitoring device acceptable under U.S. EPA Method 21, and the method detection limits for
each test condition. Table 12 summarizes the range of method detection limits in units of grams
per hour (g/hr) found during the laboratory testing as well as presents the overall detection limit
variation for each compound. The overall detection limit variation presented in Table 12 was
calculated using Equation 1 in Chapter 5.
30
-------
>TM
Table 8. FLIR GasFindIR MW Method Detection Limits at 10 Feet Stand-off Distance
with a Cement Board Background
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Methylene chloride
Ethylene
Methanol
Pentane
Propane
Styrene
Wind
Speed
(mph)(a)
0
0
2.5
5
0
0
2.5
5
0
2.5
0
Q(e)
0®
2.5
5.0
0
2.5
5.0
0
Q(e)
0®
2.5
2.5(e)
2.5®
5.0
5.0(e)
5.0®
0
2.5
5.0
0
Ambient
Temp.
(°F)
70.3
72.7
75.1
75.0
71.2
72.7
74.3
74.4
70.9
72.3
71.4
70.9
71.1
71.4
71.3
71.3
70.1
70.1
72.1
71.7
71.9
71.3
71.3
71.4
71.1
71.0
70.8
71.0
71.8
71.3
71.8
Leak Temp.
(°F)
70.9
82.1
85.5
80.4
84.8
89.3
81.7
77.5
79.2
78.4
71.9
71.2
71.5
72.2
72.1
77.0
88.8
82.0
79.0
77.6
80.1
83.4
82.2
81.9
78.6
77.3
76.8
70.6
71.8
71.6
82.4
M21 Device
Cone.
(ppmv)
843
4.0
526
32
4.9
220
737
684
N.A.(8)
N.A.(8)
No data(d)
No data(d)
No data(d)
253
554
N.A.(8)
N.A.(8)
N.A.(8)
1.7
No data(d)
No data(d)
45
18
0.20
77
26
12
N.A.(8)
N.A.(8)
N.A.(8)
212
Method
Detection
Limit (g/hr)
1.3
<0.02(b)
<4.6(b)
<4.6(b)
0.92
0.70
11
28
18
> 70(c)
1.4
0.70
0.35
68
83
0.35
2.8
14
< 0.2800
<0.28(b)
<0.28(b)
8.3
2.2
0.28
28
9.4
4.1
< 0.44 (b)
4.4
8.2
0.70
(a) The leak was viewed using the camera's standard lens (25-mm) at these conditions unless otherwise noted.
(b) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
(c) The leak could not be detected below the highest reliable flow rate supplied by the delivery system.
(d) No data - the leak concentration was inadvertently not collected by laboratory personnel using the M21 device.
(e) The leak was viewed using the optional 50-mm lens at these conditions.
(f) The leak was viewed using the optional 100-mm lens at these conditions.
(g) N. A. - not applicable. The ionization potential of this compound is higher than is capable of detection by the
device used. Therefore, any raw data measured with this device is not reported in this table.
-------
>TM
Table 9. FLIR GasFindIR MW Method Detection Limits at 30 Feet Stand-off Distance
with a Cement Board Background
(a)
(b)
(c)
(d)
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Methylene chloride
Ethylene
Methanol
Pentane
Propane
Styrene
The leak was viewed
Wind
Speed
(mph)(a)
0
0
2.5
5
0«0
0«0
25(c)
5(0
0«0
0
0®
0«0
2.5
5.0
0(0
25(c)
5(0
0(0
25(c)
5(0
0(0
25(c)
5(0
0(0
Ambient
Temp. (°F)
71.0
70.8
74.8
74.8
71.7
71.4
74.5
74.8
69.9
71.3
70.5
70.1
71.3
71.2
71.8
72.4
70.2
72.0
71.3
69.9
70.5
71.8
71.9
71.4
using the camera's standard lens
Leak
Temp
71.8
88.7
85.5
79.9
92.0
76.2
82.8
78.7
87.7
71.8
71.1
70.4
72.2
72.0
77.9
90.4
81.4
77.1
85.6
80.0
70.6
71.7
71.7
77.1
(25-mm)
M21 Device
Cone, (ppmv)
876
1.8
7.8
7.8
0.80
203
323
1042
N.A.(d)
No data(e)
No data(e)
No data(e)
287
241
N.A.(d)
N.A.(d)
N.A.(d)
17
84
46
N.A.(d)
N.A.(d)
N.A.(d)
85
Method
Detection Limit
(g/hr)
1.6
<0.02(b)
<4.6(b)
<4.6(b)
0.92
0.35
15
31
4.9
3.8
2.1
1.1
83
243
0.28
2.1
19
<0.28(b)
8.3
17
< 0.44 (b)
3.3
6.3
0.35
at these conditions unless otherwise noted.
Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
The leak was viewed
N.A. - not applicable
using the optional
100-mm lens at these conditions.
. The ionization potential of this
compound is higher than is capable of detection by the
device used. Therefore, any raw data measured with this device
(e)
(f)
No data - the leak concentration was inadvertently not
The leak was viewed
using the optional
collected
is not reported in this table.
by laboratory personnel
using the M21 device.
50-mm lens at these conditions.
32
-------
,TM
Table 10. FLIR GasFindIR MW Method Detection Limits at 10 Feet Stand-off with a
Curved Metal Gas Cylinder Background
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Methylene chloride
Ethylene
Methanol
Pentane
Propane
Styrene
Wind
Speed
(mph)(a)
0
0
2.5(c)
5(0
0
0
2.5
5
0
2.5
0
2.5
5
0
2.5
5
0
2.5
5
0
2.5
5
0
Ambient
Temp. (°F)
70.0
72.8
74.8
74.8
71.7
72.6
74.4
74.2
70.7
74.2
71.4
71.1
71.4
71.3
70.5
70.4
71.6
71.9
72.1
70.7
71.9
70.9
72.1
Leak
Temp.
(°F)
70.8
80.6
85.7
78.7
93.9
86.2
82.0
77.9
81.0
82.1
71.4
72.1
72.1
95.0
91.8
81.6
87.1
85.8
80.5
71.4
71.9
71.5
82.8
M21 Device
Cone, (ppmv)
> 2,000
2.9
1.3
29
20
364
33
227
N.A.
N.A.
No data(e)
225
600
N.A.(f)
N.A.(f)
N.A.(f)
8.0
58
142
N.A.
N.A.
N.A.
104
Method
Detection Limit
(g/hr)
2.7
<0.02(b)
<4.6(b)
<4.6(b)
1.2
0.70
11
35
18
> 70(d)
1.7
68
122
0.35
3.1
17
0.44
8.3
19
< 0.44 (8)
7.1
13
0.70
(a) The leak was viewed using the camera's standard lens (25-mm) at these conditions unless otherwise noted.
(b) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
(c) The leak was viewed using the optional 100-mm lens at these conditions.
(d) The leak could not be detected below the highest reliable flow rate supplied by the delivery system.
(e) No data - the leak concentration was inadvertently not collected by laboratory personnel using the M21 device.
(f) N. A. - not applicable. The ionization potential of this compound is higher than is capable of detection by the
device used. Therefore, any raw data measured with this device is not reported in this table.
(g) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
33
-------
,TM
Table 11. FLIR GasFindIR MW Method Detection Limits at 30 Feet Stand-off Distance
with a Curved Metal Gas Cylinder Background
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Methylene chloride
Ethylene
Methanol
Pentane
Propane
Styrene
Wind
Speed
(mph)(a)
0
0
2.5(c)
5(0
Q(c)
0«0
2.5(c)
5(0
0(c)
0
0®
2.5
5
0
2.5
5
0«0
2.5(c)
5(0
0(c)
2.5(c)
5(0
0(c)
Ambient
Temp. (°F)
71.1
71.0
74.7
74.7
70.7
71.9
74.9
75.0
69.6
71.3
71.4
71.3
71.3
71.7
71.2
70.3
71.9
74.0
71.6
70.3
70.9
70.7
72.8
Leak
Temp.
(°F)
71.9
83.6
88.0
78.3
80.2
86.1
82.0
80.6
80.8
72.1
72.0
72.2
72.1
81.4
88.7
82.6
78.2
85.3
81.3
69.9
71.4
71.7
88.3
M21 Device
Cone, (ppmv)
468
2.2
161
No data(d)
1.2
337
526
521
N.A.(e)
No data(d)
No data(d)
571
473
N.A.(e)
N.A.(e)
N.A.(e)
18
19
61
N.A.(e)
N.A.(e)
N.A.(e)
No data(d)
Method
Detection Limit
(g/hr)
1.6
<0.02(b)
<4.6(b)
<4.6(b)
0.92
0.77
16
35
11
7.0
5.2
156
278
0.35
2.8
22
<0.28(b)
2.8
17
< 0.44 ^
3.3
6.6
0.70
(a) The leak was viewed using the camera's standard lens (25-mm) at these conditions unless otherwise noted.
(b) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
(c) The leak was viewed using the optional 100-mm lens at these conditions.
(d) No data - the leak concentration was inadvertently not collected by laboratory personnel using the M21 device.
(e) N. A. - not applicable. The ionization potential of this compound is higher than is capable of detection by the
device used. Therefore, any raw data measured with this device is not reported in this table.
(f) The leak was viewed using the optional 50-mm lens at these conditions.
34
-------
Table 12. FLIR GasFindIR™ MW Range of Method Detection Limits and Overall
Method Detection Limit Variation
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Dichloromethane
(methylene chloride)
Ethylene
Methanol
Pentane
Propane
Styrene
Minimum
1.3
<0.02
0.92
0.35
4.9
0.35
0.28
<0.28
<0.44
0.35
Maximum
2.7
<4.6(c)'(d)
1.2
35(d)
> 70(c)
278(d)
22(d)
28(d)
13(d)
0.70
Overall Variation^
2.3
14
88
8.5
8.2
3.8
(a) Minimum and maximum values shown were measured at a 0-mph wind speed unless otherwise noted.
(b) When sample sizes are small (N < 10), standard deviations provide a biased estimate of the variability, therefore
only the range is provided when there were fewer than 10 method detection limits determined.
(c) Measured at a 2.5-mph wind speed condition.
(d) Measured at a 5-mph wind speed condition.
TM
6.2 Detection Agreement to a Portable Monitoring Device
The detection of a single chemical gas leak in either the laboratory or field environments was
determined by the operator as well as two confirming individuals as discussed in Section 3.2.1.
The leak rate was know from certified gas cylinders and calibrated flow meters in the laboratory,
or was determined through the bagging method during field testing. During both the laboratory
and field tests, a portable monitoring device acceptable under U.S. EPA Method 21 was used to
sample the leaks. The following sections present results on the ability of the FLIR GasFindIR
MW camera to detect a chemical gas species relative to a portable monitoring device acceptable
under U.S. EPA Method 21.
6.2.1 Laboratory Testing
Table 13 presents the percent agreement between the ability of the FLIR GasFindIR™ MW
camera and of a portable monitoring device acceptable under U.S. EPA Method 21 to detect a
chemical gas leak under the conditions tested. Percent agreement was calculated according to
Equation 2 in Chapter 5. The calculation of percent agreement excludes those laboratory test
conditions for which a response was not collected using a portable monitoring device acceptable
under U.S. EPA Method 21. In addition, percent agreement was not evaluated for methylene
chloride, methane, methanol, and propane because these compounds have an ionization potential
greater than that which could be supplied by the Industrial Scientific IBRID MX6 with PID
sensor.
35
-------
Table 13. Summary of Detection Agreement Between FLIR GasFindIR
and a Method 21 Portable Monitoring Device
™
MW Camera
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Ethylene
Pentane
Styrene
No. of Tests in which
Agreed
4
11
4
12
8
16
3
Total No. of Tests
Completed
4
11
4
12
8
16
3
Percent Agreement
100%
100%
100%
100%
100%
100%
100%
6.2.2 Field Testing
™
During field testing, nine leaking components were viewed using the FLIR GasFindIR MW
camera using the procedures described in Section 3.2.1. Table 14 identifies whether each
chemical species gas leak was observed by the FLIR GasFindIR™ MW camera and the
concentration of the leak as determined by a portable monitoring device acceptable under U.S.
EPA Method 21 . In addition, these tables identify the type of component that was leaking, the
average chemical-specific mass leak rate from the component as determined by reference
sampling, the distance the leak was observed, and the wind speed. Daily meteorological
conditions were obtained from Dow Chemical's on-site weather station. Although the wind
speed and daily maximum and minimum temperatures were obtained from this weather station,
the actual wind speed and ambient and background temperatures at each leak location at the time
of observation are unknown. Additional discussions describing each leak location are provided
in the following sections.
Leak Location 1. A leak was identified originating from a 3 -inch plug in service with a process
stream containing ethane, ethyl ene, methane, and propane. Screening of the component with the
TVA caused an over range reading (estimated as > 100,000 ppmv). The leak was viewed and
detected with the FLIR GasFindIR™ MW camera at stand-off distance of 12 ft with the sun at
the observers back. The leak was bagged and a duplicate reference sample was collected into
two evacuated SUMMA canisters. The SUMMA canisters were shipped to the off-site GC
laboratory and analyzed for ethylene and methane concentrations. Daily weather conditions, as
reported by the on-site weather station, were clear conditions, a daily minimum and maximum
temperature of 41 and 61 degrees Fahrenheit (°F), respectively, with wind out of the east at up to
8 mph.
36
-------
TM
Table 14. Summary of Field Testing Results Using the FLIR GasFindIR MW Camera
Leaking
Leak Component
Location Type
1 3 -inch Plug
2 %-inch Tube
_ !/2-inch
Connector
6-inch Block
Vnl VP
V CU V V»-
8-inch Block
T/^1 T^
Valve
7 Control
Valve Flange
2-inch Block
T/^1 T^
Valve
„ 1 -inch Valve
Plug
6-inch
10 Pressure
Relief Valve
Wind
Speed
(mph)
8
21
21
21
21
18
18
18
5
Stand-
off
Distance
(ft)
12
10
30
10
30
10
10
10
10
10
10
M21 Device
Screening
Cone, (ppmv)
>100,000
20,500
>100,000
>100,000
20,500
17,500
8,000^
835
>100,000
Leak
Detected
by
Camera?
Yes
No
No
Yes
Yes
Yes
No
No
No
No
No
No
Bagging Results:
Average Leak Rate
(g/hr)
8.79 (methane)
4.31 (ethylene)
0.951 (ethylene)
2.32 xlO'3 (ethylene)
7.78 (methane)
5.24x 10"2 (ethylene)
8.68 x 10"3 (styrene)
0.077 (benzene)
3. 44(a) (benzene)
1.95x 10"3 (ethylene)
0.282 (benzene)
1.92(b)(l,3-butadiene)
0.350 (methylene
chloride)
6.78
(propylene dichloride)
(a) As reported in Table 5, the pre- and post-bagging leak concentrations, as measured by the TVA, differed by
24.4%. This exceeds the minimum acceptance criterion of 20% for the DQI for the confirmation of detected
leaks. Thus, this data is considered suspect and reported with this data qualifier.
(b) As reported in Table 4, the calibration check response for the TVA, conducted after screening this component,
resulted in a 24% difference. This exceeded the minimum acceptance criterion of 10% for the DQI for the bias
and accuracy of sample screening measurements using a portable monitoring device. After recalibration of the
TVA, the leak concentration from this component was not reconfirmed with the TVA. Thus, this data is
considered suspect and reported with this data qualifier.
Leak Location 2. A leak was identified originating from a H-inch tube in service with a process
stream containing ethane and ethylene. Screening of the component with the TVA resulted in a
concentration reading of 20,500 ppmv. The leak was viewed with the FLIR GasFindIR™ MW
camera at stand-off distances of 10 and 30 ft with the sun to the left of the observer. The camera
did not detect the leak at either stand-off distance. Wind direction at the location was noted as
originating from behind the observer and the site was shaded by piping and other equipment.
The leak was bagged and a duplicate reference sample was collected into two evacuated
SUMMA canisters. The SUMMA canisters were shipped to the off-site GC laboratory and
analyzed for ethylene concentration. Daily weather conditions, as reported by the on-site
weather station, were clear conditions, a daily minimum and maximum temperature of 42 and 70
°F with wind out of the south southeast at 21 mph.
37
-------
Leak Location 3. A leak was identified originating from a %-inch connector in service with a
process stream containing acetylene, ethane, ethylene, methane, propane, and propylene.
Screening of the component with the TVA caused an over range reading (estimated as > 100,000
ppmv). The leak was viewed with the FLIR GasFindIR™ MW camera at stand-off distances of
10, 30, and 45 ft, with the sun to the right of the observer. The FLIR GasFindIR™ MW camera
detected the leak at each of the three stand-off distances. Wind direction at the location was
noted as originating from the right of the observer and the site was shaded by piping and other
equipment. The leak was bagged and a duplicate reference sample was collected into two
evacuated SUMMA canisters. The SUMMA canisters were shipped to the off-site GC
laboratory and analyzed for ethylene and methane concentrations. Daily weather conditions, as
reported by the on-site weather station, were clear conditions, a daily minimum and maximum
temperature of 42 and 70 °F with wind out of the south southeast at 21 mph.
The average mass leak rate of ethylene measured at this leak location was 2.23 x 10"3 g/hr. This
value is below the lowest ethylene method detection limit measured with the FLIR GasFindIR™
MW camera during the laboratory phase of this verification test.
Leak Location 4. Leak location 4 contained a leaking component that was misidentified as
being in service with styrene. This sample location was confirmed to be in ethylbenzene service
and thus no analytical results are reported for this leak location. The FLIR GasFindIR™ MW
camera was able to detect this leak.
Leak Location 5. A leak was identified originating from a 6-inch block valve in service with a
process stream containing benzene, ethane, ethylene, ethylbenzene, styrene, and toluene.
Screening of the component with the TVA caused an over range reading (estimated as >
100,000 ppmv). The leak was viewed with the FLIR GasFindIR™ MW camera at a stand-off
distance of 10 ft; the leak could not be detected at this distance. The site was shaded and the
viewing background was concrete. The leak was bagged and a duplicate reference sample was
collected into two evacuated SUMMA canisters. The SUMMA canisters were shipped to the
off-site GC laboratory and analyzed for benzene, ethylene, and styrene concentrations. Daily
weather conditions, as reported by the on-site weather station were clear conditions, a daily
minimum and maximum temperature of 48 and 79 °F with wind out of the north at 21 mph.
The average mass leak rates of ethylene, styrene, and benzene measured at this leak location
were 5.24 x 10"2, 8.68 x 10"3, and 0.077 g/hr, respectively. These values are all below the lowest
method detection limits measured with the FLIR GasFindIR™ MW cameras for these
compounds during the laboratory phase of this verification test.
Leak Location 6. A leak was identified originating from an 8-inch block valve in service with a
process stream containing benzene, toluene, hexane, and other aromatic hydrocarbons.
Screening of the component with the TVA resulted in a concentration reading of 20,500 ppmv.
The leak was viewed with the FLIR GasFindIR™ MW camera at a stand-off distance of 10 ft
with the sun to the right of the camera observer; the leak could not be detected at this distance.
The site was an exterior location and weather conditions were noted as slightly overcast with
moderate wind originating from the right of the observer. The leak was bagged and a duplicate
reference sample was collected into two evacuated SUMMA canisters. The SUMMA canisters
were shipped to the off-site GC laboratory and analyzed for benzene concentration. Daily
weather conditions, as reported by the on-site weather station were clear conditions, a daily
minimum and maximum temperature of 48 and 79 °F with wind out of the north at 21 mph.
38
-------
Leak Location 7. A leak was identified originating from a control valve flange in service with a
process stream containing benzene, butane, butylbenzene, all isomers of diethylbenzene, ethane,
ethylbenzene, ethylene, hexane, toluene, and other aromatic hydrocarbons. Screening of the
component with the TVA resulted in a concentration reading of 17,500 ppmv. The leak was
viewed with the FLIR GasFindIR™ MW camera at a stand-off distance of 10 ft with the sun
behind the camera observer; the leak could not be detected at this distance. The site was located
on the second deck of the chemical plant and weather conditions were qualitatively noted as very
windy. The viewing background was other plant piping and equipment. The leak was bagged
and a duplicate reference sample was collected into two evacuated SUMMA canisters. The
SUMMA canisters were shipped to the off-site GC laboratory and analyzed for benzene and
ethylene concentrations. Daily weather conditions, as reported by the on-site weather station,
were partly cloudy conditions, a daily minimum and maximum temperature of 43 and 65 °F with
wind out of the north at 18 mph.
The average mass leak rates of ethylene and benzene measured at this leak location were 1.95 x
10"3 and 0.282 g/hr, respectively. These values are all below the lowest method detection limits
measured with the FLIR GasFindIR™ MW camera for these compounds during the laboratory
phase of this verification test.
Leak Location 8. A leak was identified originating from a 2-inch block valve in service with a
process stream containing 1,3-butadiene. Screening of the component with the TVA resulted in
a concentration reading of 8,000 ppmv. The leak was viewed with the FLIR GasFindIR™ MW
camera at a stand-off distance of 10 ft; the leak could not be detected at this distance. The site
was an exterior location on a marine vapor recovery line at a marine vapor recovery system and
weather conditions were qualitatively noted to be very windy and overcast. The leak was bagged
and a duplicate reference sample was collected into two evacuated SUMMA canisters. The
SUMMA canisters were shipped to the off-site GC laboratory and analyzed for 1,3-butadiene
concentration. Daily weather conditions, as reported by the on-site weather station, were partly
cloudy conditions, a daily minimum and maximum temperature of 43 and 65 °F with wind out of
the north at 18 mph.
Leak Location 9. A leak was identified originating from a 1-inch valve plug in service with a
process stream containing methyl ene chloride. Screening of the component with the TVA
resulted in a concentration reading of 835 ppmv. The leak was viewed with the FLIR
GasFindIR™ MW camera at a stand-off distance of 10 ft; the leak could not be detected at this
distance. The site was an exterior location and weather conditions were qualitatively noted as
overcast with calm winds. The viewing background was concrete ground and a few metal pipe
supports. The leak was bagged and a duplicate reference sample was collected into two
evacuated SUMMA canisters. The SUMMA canisters were shipped to the off-site GC
laboratory and analyzed for methylene chloride concentration. Daily weather conditions, as
reported by the on-site weather station, were partly cloudy conditions, a daily minimum and
maximum temperature of 43 and 65 °F with wind out of the north at 18 mph.
The average mass leak rate of methylene chloride measured at this leak location was 0.350 g/hr.
This value is below the lowest ethylene method detection limit measured with the FLIR
GasFindIR™ MW camera during the laboratory phase of this verification test.
Leak Location 10. A leak was identified originating from a 6-inch pressure relief valve in
service with a process stream containing 1,2,3-trichloropropane, 2,3-dichloropropanol, 2-methyl-
2-pentenal, l-chloro-2,3-epoxypropane, and propylene di chloride. Screening of the component
39
-------
with the TVA caused an over range reading (estimated as > 100,000 ppmv). The leak was
viewed with the FLIR GasFindIR M MW camera at a stand-off distance of 10 ft; the leak could
not be detected at this distance. The site was an exterior location (on top of a storage tank
platform) and weather conditions were qualitatively noted as overcast, breezy, and cold. The
leak was bagged and a duplicate reference sample was collected into two evacuated SUMMA
canisters. The SUMMA canisters were shipped to the off-site GC laboratory and analyzed for
propylene dichloride concentration. Daily weather conditions, as reported by the on-site weather
station, were partly cloudy conditions, a daily minimum and maximum temperature of 41 and 50
°F with wind out of the north at 5 mph.
6.3 Confounding Factors
The method detection limits generated during laboratory testing presented in Table 8 through
Table 11 were inspected to identify general trends that the confounding factors of stand-off
distance, wind speed, and background materials impart on the method detection limits for the
gaseous chemical species leaks observed using the FLIR GasFindIR™ MW camera. In addition,
the effect of lens size was also inspected. The following general trends were noted when using
the FLIR GasFindIR™ MW camera.
• Stand-off Distance - Method detection limits generally increased as the viewing distance
increased
• Wind Speed - Method detection limits generally increased with increased wind speed
• Background Materials- Method detection limits were generally lower when viewed
against the cement board background. Two exceptions to this observation were noted
when viewing ethylene. The first occurred when viewing the leak at a 10 ft distance at a
5-mph wind speed with the standard 25-mm lens. The second occurred when viewing the
leak at a 30 ft distance at a 2.5-mph wind speed with the optional 100-mm lens.
• Camera Lens - Method detection limits generally decreased with an increase in camera
lens size
During field testing, confounding factors were recorded either quantitatively or qualitatively and
are reported in Table 14 and Table 15. A rigid analysis of the influence of confounding factors
was not undertaken using field testing data, however, it is generally noted that because the
cameras detected only a few of the chemical leaks in the field, the confounding factors of wind
speed, stand-off distance, and background materials affected the detection capability of the
cameras.
6.4 Operational Factors
The FLIR GasFindIR™ MW camera was found to be easily set up in a small, two ft by two ft
area and deployed within approximately 10 minutes for portable gas leak observations. In terms
of field portability, the camera was light in weight (approximately 4.6 pounds with battery and
camera), easily carried by one person and was provided with a rugged shipping case for
transportation.
The FLIR GasFindIR™ MW camera may be powered with either an AC adaptor for stationary
applications or with a six volt, 4200 milliampere-hour nickel-metal hydride battery for mobile
field observations. The battery for the camera was used and held its charge when performing
40
-------
visual screening of leaking components. The FLIR GasFindIR™ MW camera comes equipped
with a standard 25-mm camera lens; optional 50-mm and 100-mm lenses may be purchased
separately for use with the camera. The camera observer sees the infrared image through a
standard eyepiece when using both the FLIR GasFindIR™ MW camera; these images are also
recordable to any off-the-shelf video recorder for image storage.
Ease of use was not investigated with a newly trained operator, as the vendor operated the FLIR
GasFindIR™ MW camera during both laboratory and field testing. Verification test team
members, however, did observe that both cameras were operated by the camera operator with
relative ease. The FLIR GasFindIR™ MW camera is not intrinsically safe, and cannot be used
in explosive atmospheres or environments.
During this verification test, all chemical leaks were required to be observed by the camera
operator and two additional confirming individuals to be considered as "detected" by the camera.
During verification testing, there were instances where either one or two of the three observers
(not the required three) were able to observe the chemical leak. This indicates that the ability of
the operator using the camera to positively identify the chemical leak may have an influence on
the operation of the camera.
The cost of the FLIR GasFindIR™ MW camera is $64,950. The base price of the camera
includes an intelligent battery charger and three lithium ion batteries, an alternating current
power supply, a video cable, a personal video recorder and battery, audio/video cable for the
personal video recorder, camera neck strap, shipping/carrying case, and operating manual.
The cost of optional 50 and 100-millimeter lenses for the FLIR GasFindIR™ MW camera are
$7,500 and $9,950, respectively.
41
-------
Chapter 7
Performance Summary
Method Detection Limits. Method detection limits were determined during the laboratory
testing. Table 15 summarizes the minimum and maximum method detection limit obtained
during laboratory testing using the FLIR GasFindIR™ MW camera. Specific details, including
the test conditions at which these method detection limits were obtained and the lens size used,
are provided in Table 8 through Table 11 in Chapter 6. The overall detection limit variations for
each chemical obtained using each camera are presented in Table 12 in Chapter 6.
Detection of Chemical Gas Species Relative to a Portable Monitoring Device. The ability of
the FLIR GasFindIR™ MW camera to detect a gaseous leak of a chemical relative to a portable
monitoring device acceptable under U.S. EPA Method 21 was assessed during both laboratory
and field testing. During laboratory testing, after the method detection limit had been reached
for a particular chemical under the specified test conditions, the leak was sampled by the portable
monitoring device. Table 15 presents the percent agreement between the ability of the FLIR
GasFindIR™ MW camera and of a portable monitoring device acceptable under U.S. EPA
Method 21 to detect a chemical gas leak under the conditions tested in the laboratory.
During field testing a portable monitoring device acceptable under U.S. EPA Method 21 was
used to screen each leaking component as part of the bagging reference method used. Table 16
reports the responses of the portable screening device when screening leaking components,
identifies whether the FLIR GasFindIR™ MW camera was able to detect the chemical leak from
the leaking component, and reports the chemical-specific mass rate of emissions from the leaking
component as obtained through the bagging method.
Confounding Factors. Stand-off distance, wind speed, and background materials generally
impacted the performance of the FLIR GasFindIR™ MW camera (e.g., increasing the stand-off
distance from the leak increased the method detection limits). Changing to an optional
magnifying camera lens that can be purchased separately lowered the method detection limit.
Details of the effects of confounding factors may be found in Section 6.3.
42
-------
Table 15. Summary of FLIR GasFindIR™ MW Camera Method Detection Limits(a) and
Percent Agreement with a Method 21 Monitoring Device During Laboratory Testing
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Benzene
Methylene chloride
Ethylene
Methanol
Pentane
Propane
Styrene
Method Detection
Minimum
1.3
<0.02
0.92
0.35
4.9
0.35
0.28
<0.28
<0.44
0.35
Limit (g/hr)
Maximum
2.7
<4.6(b)'(c)
1.2
35(o)
> 70(c)
278(c)
22(c)
28(c)
13(o)
0.70
Agreement with Method 21 Monitoring
Device
Total No. of Tests
Performed Percent Agreement
4 100%
11 100%
4 100%
12 100%
No data(d)
8 100%
No data(d)
16 100%
No data(d)
3 100%
(a)
Minimum and maximum method detection limits shown were measured at a 0-mph wind speed unless
otherwise noted.
(b) Measured at a 2.5-mph wind speed.
(c) Measured at a 5-mph wind speed.
(d) Percent agreement was not evaluated for methylene chloride, methanol, and propane because these compounds
have an ionization potential greater than the energy which could be supplied by the Industrial Scientific IBRID
MX6 with PID sensor.
Operational Factors. The FLIR GasFindIR™ MW camera was found to be easily setup and
ready to deploy in 10 minutes. The camera is light (4.6 pounds or less) and operated on batteries
when performing visual screening of leaking components. The camera may also utilize optional
lenses that can be used to further magnify the images. Because the camera was operated by
FLIR and there were some disagreements on detections with the two other confirming
individuals, the ability of the operator may influence the operation of the camera. The FLIR
GasFindIR™ MW camera is not intrinsically safe, and cannot be used in explosive atmospheres
or environments.
The cost of the FLIR GasFindIR™ MW camera is $64,950 and includes an intelligent battery
charger and three lithium ion batteries, an alternating current power supply, a video cable, a
personal video recorder and battery, audio/video cable for the personal video recorder, camera
neck strap, shipping/carrying case, and operating manual.
The cost of optional 50 and 100-millimeter lenses for the FLIR GasFindIR
$7,500 and $9,950, respectively.
TM
MW camera are
43
-------
TM
Table 16. Summary of Field Testing Results Using the FLIR GasFindIR MW Camera
Leak
Location
1
2
3
5
6
7
8
9
10
Leaking
Component
Type
3 -inch Plug
'/4-inch Tube
'/2-inch
Connector
6-inch Block
Valve
8-inch Block
T Tn\,~n
Valve
Control
Valve Flange
2-inch Block
Valve
1 -inch Valve
Plug
6-inch
Pressure
Relief Valve
Wind
Speed
(mph)
8
21
21
21
21
18
18
18
5
Stand-off
Distance
(ft)
12
10
30
10
30
45
10
10
10
10
10
10
M21 Device
Screening
Cone, (ppmv)
>100,000
20,500
>100,000
>100,000
20,500
17,500
8,000^
835
>100,000
Leak
Detected by
Camera?
Yes
No
No
Yes
Yes
Yes
No
No
No
No
No
No
Bagging Results:
Average Leak Rate
(g/hr)
8.79 (methane)
4.31 (ethylene)
0.951 (ethylene)
2.32 x 10"3 (ethylene)
7.78 (methane)
5.24 x 10"2 (ethylene)
8.68 x 10"3 (styrene)
0.077 (benzene)
3.44(a) (benzene)
1.95 x 10"3 (ethylene)
0.282 (benzene)
1.92(b) (1,3-butadiene)
0.350
(methylene chloride)
6.78
(propylene dichloride)
(a) As reported in Table 5, the pre- and post-bagging leak concentrations, as measured by the TVA, differed by
24.4%. This exceeds the minimum acceptance criterion of 20% for the DQI for the confirmation of detected
leaks. Thus, this data is considered suspect and reported with this data qualifier.
(b) As reported in Table 4, the calibration check response for the TVA, conducted after screening this component,
resulted in a 24% difference. This exceeded the minimum acceptance criterion of 10% for the DQI for the bias
and accuracy of sample screening measurements using a portable monitoring device. After recalibration of the
TVA, the leak concentration from this component was not reconfirmed with the TVA. Thus, this data is
considered suspect and reported with this data qualifier.
44
-------
Chapter 8
References
1. Test/QA Plan for Verification of Leak Detection and Repair Technologies, Battelle,
Columbus, Ohio, September 18, 2008.
2. Quality Management Plan for the ETV Advanced Monitoring Systems Center, Version 7.0,
U.S. EPA Environmental Technology Verification Program, Battelle, Columbus, Ohio,
November, 2008
3. EPA Method 21- Detection of Volatile Organic Compound Leaks, EP A-600/2-18-110; U. S.
EPA, September 1981.
4. Panek, J., P. Drayton, and D. Fashimpaur. Controlled Laboratory Sensitivity and
Performance Evaluation of Optical Leak Imaging Infrared Cameras for Identifying Alkane,
Alkene, and Aromatic Compounds, Proceedings of the 99* Annual Conference and
Exposition of the Air and Waste Management Association, New Orleans, June 20-23, 2006,
Manuscript number 06-A-159-AWMA, Curran Associates, Inc., Red Hook, New York,
March 2007.
5. EPA Protocol for Equipment Leak Emissions Estimates, EPA-453/R-95-017; U.S. EPA:
Research Triangle Park, NC, November 1995.
6. EPA Method 18 - Measurement of Gaseous Organic Compound Emissions by Gas
Chromatography, 40 CFR, Part 60, Appendix A; April, 1994.
7. EPA Method 205 - Verification of Gas Dilution Systems for Field Instrument Calibrations,
40 CFR, Part 51, Appendix M, September, 1996.
45
-------
Appendix A
FLIR GasFindIR™ LW Camera Results
A FLIR GasFindIR™ LW camera underwent a limited amount of testing during both the
TA/f
laboratory and field testing phases of this verification test. The FLIR GasFindIR LW camera
was not evaluated against the entire suite of chemicals used in the laboratory portion of this
verification testing; rather the vendor used the FLIR GasFindIR™ LW camera for 1,3-butadiene,
acetic acid, and acrylic acid because these compounds have an absorption peak within the 10 -
11 micrometer operating wavelength of the FLIR GasFindIR™ LW camera. The camera was
evaluated in the field for all chemical gas leaks identified, regardless of whether the gas leak
contained compounds with an absorption peak within the 10 - 11 micrometer operating
wavelength of the FLIR GasFindIR M LW camera, on only those days that the camera was
available to the verification test team during field testing.
A.I Method Detection Limit
The method detection limit for 1,3-butadiene, acetic acid, and acrylic acid was determined
according to the procedures discussed in Section 3.2.2. Tables Al through A4 present the
method detection limits of each these compounds determined during laboratory testing. Tables
Al through A4 identify each test condition evaluated (i.e., stand-off distance, background
material, and wind speed), the temperatures of the laboratory and of the chemical leak, the
response of the portable monitoring device acceptable under U.S. EPA Method 21, and the
method detection limits for each test condition. Table A5 summarizes the range of method
detection limits in units of gram per hour (g/hr) found during the laboratory testing as well as
presents the overall detection limit variation for each compound. The overall detection limit
variation presented in Table A5 was calculated using Equation 1 in Chapter 5.
Table Al. FLIR GasFindIR™ LW Method Detection Limits at 10 Feet Stand-off Distance
with a Cement Board Background
Compound
1,3-butadiene
Acetic acid
Acrylic acid
Wind
Speed
(mph)(a)
0
0
2.5
5
0
Ambient
Temp.
70.1
72.7
75.1
75.0
71.2
Leak Temp.
CD
71.2
82.1
85.5
80.4
84.8
M21 Device
Cone.
(ppmv)
> 2,000
4.0
526
32
4.9
Method
Detection Limit
(g/hr)
2.7
0.02
<4.6(b)
<4.6(b)
0.92
(a) The leak was viewed using the camera's standard lens (50-mm) at these conditions unless otherwise noted.
(b) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
46
-------
Table A2. FLIR GasFindIR™ LW Method Detection Limits at 30 Feet Stand-off Distance
with a Cement Board Background
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Wind
Speed
(mph)(a)
0
0
2.5
5
0
Ambient
Temp. (°F)
71.7
70.8
74.8
74.9
71.7
Leak
Temp.
72.1
88.7
85.5
80.5
92.0
M21 Device
Cone, (ppmv)
> 2,000
1.8
7.8
17
0.8
Method Detection
Limit (g/hr)
13
0.02
<4.6(b)
14
0.92
(a) The leak was viewed using the camera's standard lens (50-mm) at these conditions unless otherwise noted.
(b) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
Table A3. FLIR GasFindIR™ LW Method Detection Limits at 10 Feet Stand-off with a
Curved Metal Gas Cylinder Background
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Wind
Speed
(mph)(a)
0
0
2.5
5
0
Ambient
Temp. (°F)
70.2
72.8
74.8
74.8
71.4
Leak
Temp.
(°F)
71.0
80.6
85.7
78.7
97.7
M21 Device
Cone, (ppmv)
> 2,000
2.9
1.3
29
1.2
Method
Detection Limit
(g/hr)
3.4
0.02
<4.6(b)
<4.6(b)
^0_46^
(a) The leak was viewed using the camera's standard lens (50-mm) at these conditions unless otherwise noted.
(b) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
Table A4. FLIR GasFindIR™ LW Method Detection Limits at 30 Feet Stand-off Distance
with a Curved Metal Gas Cylinder Background
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Wind
Speed
(mph)(a)
0
0
2.5
5
0
Ambient
Temp. (°F)
71.0
71.0
74.7
74.7
70.7
Leak
Temp.
(°F)
71.9
83.6
88.0
77.9
80.2
M21 Device
Cone, (ppmv)
> 2,000
2.2
161
28
1.2
Method
Detection Limit
(g/hr)
13
0.02
<4.6(b)
18
0.92
(a) The leak was viewed using the camera's standard lens (50-mm) at these conditions unless otherwise noted.
(b) Leak observable at the lowest reliable flow rate capable of being supplied by the chemical delivery system.
Table A5. FLIR GasFindIR™ LW Range of Method Detection Limits and Overall Method
Detection Limit Variation (g/hr)(a)
Compound
1,3 -butadiene
Acetic acid
Acrylic acid
Minimum
2.7
0.02
<0.46
Maximum
13
18(D)
0.92
Overall Variation^
5.7
(a) Minimum and maximum values shown were measured at a 0-mph wind speed unless otherwise noted.
(b) When sample sizes are small (N < 10), standard deviations provide a biased estimate of the variability, therefore
only the range is provided when there were fewer than 10 method detection limits were determined.
47
-------
A.2 Detection Agreement to a Portable Monitoring Device
The detection of a single chemical gas leak in either the laboratory or field environments was
determined by the operator as well as two confirming individuals as discussed in Section 3.2.1.
The leak rate was known from certified gas cylinders and calibrated flow meters in the
laboratory, or was determined through the bagging method during field testing. During both the
laboratory and field tests, a portable monitoring device acceptable under U.S. EPA Method 21
was used to sample the leaks. The following sections present results on the ability of the FLIR
GasFindIR™ LW camera to detect a chemical gas species relative to a portable monitoring
device acceptable under U.S. EPA Method 21.
A. 2.1 Laboratory Testing
Table A6 presents the percent agreement between the ability of the FLIR GasFindIR™ LW
camera and of a portable monitoring device acceptable under U.S. EPA Method 21 to detect a
chemical gas leak under the conditions tested. Percent agreement was calculated according to
Equation 2 in Chapter 5. The calculation of percent agreement excludes those laboratory test
conditions for which a response was not collected using a portable monitoring device acceptable
under U.S. EPA Method 21.
Table A6. Summary of Detection Agreement Between FLIR GasFindIR™ LW Camera
and a Method 21 Portable Monitoring Device
Compound
1,3 -Butadiene
Acetic acid
Acrylic acid
No. of Tests in which
Agreed
4
12
4
Total No. of Tests
Completed
4
12
4
Percent Agreement
100%
100%
100%
A.2.2 Field Testing
During field testing, three leaking components were viewed using the FLIR GasFindIR™ LW
camera using the procedures described in Section 3.2.1. Table A7 identifies whether each
chemical species gas leak was observed by the FLIR GasFindIR™ LW camera and the
concentration of the leak as determined by a portable monitoring device acceptable under U.S.
EPA Method 21. In addition, these tables identify the type of component that was leaking, the
average chemical-specific mass leak rate from the component as determined by reference
sampling, the distance the leak was observed and the wind speed. Daily meteorological
conditions were obtained from Dow's on-site weather station. Although the wind speed and
daily maximum and minimum temperatures were obtained from this meteorological tower, the
actual wind speed and ambient and background temperatures at each leak location at the time of
observation are unknown. Additional discussions describing each leak location are provided in
the following sections.
Leak Location 2. A leak was identified originating from a H-inch tube in service with a process
stream containing ethane and ethylene. Screening of the component with the TVA resulted in a
concentration reading of 20,500 ppmv. The leak was viewed with the FLIR GasFindIR™ LW
camera at stand-off distances of 10 and 30 ft with the sun to the left of the observer. The camera
did not detect the leak at either stand-off distance. Wind direction at the location was noted as
originating from behind the observer and the site was shaded by piping and other equipment.
The leak was bagged and a duplicate reference sample was collected into two
48
-------
TM
Table A7. Summary of Field Testing Results Using the FLIR GasFindIR LW Camera
Leak
Location
2
3
5
Leaking
Component
Type
!/4-inch Tube
'/2-inch
Connector
6-inch Block
Valve
Wind
Speed
(mph)
21
21
21
Stand-off
Distance
(ft)
10
30
10
30
45
10
M21 Device
Screening
Cone, (ppmv)
20,500
>100,000
>100,000
Leak
Detected by
Camera?
No
No
Yes
Yes
Yes
No
Bagging Results:
Average Leak Rate
(g/hr)
0.951 (ethylene)
2.32 x 10"3 (ethylene)
7.78 (methane)
5.24 x 10"2 (ethylene)
8.68 x 10"3 (styrene)
0.077 (benzene)
evacuated SUMMA canisters. The SUMMA canisters were shipped to the off-site GC
laboratory and analyzed for ethylene concentration. Daily weather conditions, as reported by the
on-site weather station, were clear conditions, a daily minimum and maximum temperature of 42
and 70 °F with wind out of the south southeast at 21 mph.
Leak Location 3. A leak was identified originating from a %-inch connector in service with a
process stream containing acetylene, ethane, ethylene, methane, propane, and propylene.
Screening of the component with the TVA caused an over range reading (estimated as > 100,000
ppmv). The leak was viewed with the FLIR GasFindIR™ LW camera at stand-off distances of
10, 30, and 45 ft with the sun to the right of the observer. The FLIR GasFindIR™ LW camera
detected the leak at each of the three stand-off distances. Wind direction at the location was
noted as originating from the right of the observer and the site was shaded by piping and other
equipment. The leak was bagged and a duplicate reference sample was collected into two
evacuated SUMMA canisters. The SUMMA canisters were shipped to the off-site GC
laboratory and analyzed for ethylene and methane concentrations. Daily weather conditions, as
reported by the on-site weather station, were clear conditions, a daily minimum and maximum
temperature of 42 and 70 °F with wind out of the south southeast at 21 mph.
Leak Location 5. A leak was identified originating from a 6-inch block valve in service with a
process stream containing benzene, ethane, ethylene, ethylbenzene, styrene, and toluene.
Screening of the component with the TVA caused an over range reading (estimated as >
100,000 ppmv). The leak was viewed with the FLIR GasFindIR™ LW camera at a stand-off
distance of 10 ft; the leak could not be detected at this distance. The site was shaded and the
viewing background was concrete. The leak was bagged and a duplicate reference sample was
collected into two evacuated SUMMA canisters. The SUMMA canisters were shipped to the
off-site GC laboratory and analyzed for benzene, ethylene, and styrene concentrations. Daily
weather conditions, as reported by the on-site weather station, were clear conditions, a daily
minimum and maximum temperature of 48 and 79 °F with wind out of the north at 21 mph.
A.3 Confounding Factors
The method detection limits generated during laboratory testing presented in Table Al through
Table A4 were inspected to identify general trends that the confounding factors of stand-off
distance, wind speed, and background materials impart on the method detection limits for the
49
-------
gaseous chemical species leaks observed using the FLIR GasFindIR™ LW camera. The
following general trends were noted when using the FLIR GasFindIR™ LW camera.
• Stand-off Distance - Method detection limits generally increased as the viewing distance
increased;
• Wind Speed - Method detection limits generally increased with increased wind speed;
• Background Materials - Method detection limits were generally lower when viewed
against the cement board background. An exception to this observation was noted when
viewing acrylic acid at a 10 ft distance at a 0-mph wind speed with the standard 50-mm
lens.
A.4 Operational Factors
The FLIR GasFindIR™ LW camera was found to be easily setup in a small, two ft by two ft area
and deployed within approximately 10 minutes for portable gas leak observations. In terms of
field portability, the camera was light in weight (approximately six pounds with battery and
camera), easily carried by one person and was provided with a rugged shipping case for
transportation.
The FLIR GasFindIR™ LW cameras may be powered with either an AC adaptor for stationary
applications or with a six volt, 4200 milliampere-hour nickel-metal hydride battery for mobile
field observations. The battery for each camera was used and held its charge through the whole
of each testing day. The FLIR GasFindIR™ LW camera comes equipped with a standard 50-mm
camera lens. The camera observer sees the infrared image through a standard eyepiece when
using both the FLIR GasFindIR™ LW cameras; these images are also recordable to any off-the-
shelf video recorder for image storage.
Ease of use was not investigated with a newly trained operator, as the vendor operated both the
FLIR GasFindIR™ LW cameras during the both laboratory and field testing. Verification test
team members, however, did observe that the camera was operated by the camera operator with
relative ease. The FLIR GasFindIR™ LW camera is not intrinsically safe, and cannot be used in
explosive atmospheres or environments.
During this verification test, all chemical leaks were required to be observed by the camera
operator and two additional confirming individuals to be considered as "detected" by the camera.
During verification testing, there were instances where either one or two of the three observers
(not the required three) were able to observe the chemical leak. This indicates that the ability of
the operator using the camera to positively identify the chemical leak may have an influence on
the operation of the camera.
The cost of the FLIR GasFindIR™ LW camera is $80,000. The base price of the camera
includes an intelligent battery charger and three lithium ion batteries, an alternating current
power supply, a video cable, a personal video recorder and battery, audio/video cable for the
personal video recorder, camera neck strap, shipping/carrying case, and operating manual.
50
------- |