Environmental Technology Verification Report Spectrex Inc. Safeye 420 Ultraviolet Open-path Monitor

March 2001

Environmental Technology
Verification Report

Spectrex Inc. SafEye 420
Ultraviolet
Open-Path Monitor

Prepared by
^llS?

BaneJie

. . . Putting Technology Ta Waiic

Battel le

Under a cooperative agreement with

SEPA U.S. Environmental Protection Agency

ETvf ETt? ET\/

-------
March 2001

Environmental Technology Verification

Report

ETV Advanced Monitoring Systems Center

Spectrex Inc. SafEye 420
Ultraviolet
Open-Path Monitor

By

Jeffrey Myers
Thomas Kelly
Charles Lawrie
Karen Riggs

Battelle
Columbus, Ohio 43201

-------
Notice

The U.S. Environmental Protection Agency (EPA), through its Office of Research and
Development, has financially supported and collaborated in the extramural program described
here. This document has been peer reviewed by the Agency and recommended for public release.
Mention of trade names or commercial products does not constitute endorsement or
recommendation by the EPA for use.

ii

-------
Foreword

The U.S. 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. At present, six environmental technology areas are covered
by ETV. Information about each of the environmental technology areas covered by ETV 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 assess-
ment. In 1997, through a competitive cooperative agreement, Battelle was awarded EPA funding
and support 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/07/07_main.htm.

iii

-------
Acknowledgments

The authors wish to acknowledge the support of all those who helped plan and conduct the
verification test, analyze the data, and prepare this report. In particular we would like to thank
Paul Webb and Andy Montgomery of Battelle. We also acknowledge the participation of
Jay Cooley and Eric Zinn of Spectrex Inc. in this verification test.

iv

-------
Contents

Notice	 ii

Foreword 	iii

Acknowledgments	 iv

List of Abbreviations	ix

1.	Background	1

2.	Technology Description 	2

3.	Test Design and Procedures 	3

3.1	Introduction	3

3.2	Test Design 	5

3.3	Experimental Apparatus and Materials	7

3.3.1	Standard Gases 	7

3.3.2	Dilution Gas 	7

3.3.3	Gas Dilution System	7

3.3.4	Gas Cell	7

3.3.5	Temperature Sensor 	7

3.3.6	Ozone Sensor 	7

3.3.7	Nitrogen Oxides/Ammonia Monitor	8

3.3.8	Benzene/Carbon Disulfide Measurement	8

3.4	Test Parameters 	8

3.4.1	Minimum Detection Limit 	8

3.4.2	Linearity 	8

3.4.3	Accuracy	9

3.4.4	Precision	9

3.4.5	Interferences 	9

4.	Quality Assurance/Quality Control	11

4.1	Data Review and Validation 	11

4.2	Changes from the Test/QA Plan 	11

4.3	Calibration	13

4.3.1	Gas Dilution System 	13

4.3.2	Temperature Sensor	13

v

-------
4.3.3	Ozone Sensor	13

4.3.4	Nitrogen Oxides/Ammonia Monitor	13

4.3.5	Benzene/Carbon Disulfide Measurement 	13

4.4	Data Collection 	14

4.5	Audits	14

4.5.1	Technical Systems Audit	14

4.5.2	Performance Evaluation Audit	15

4.5.3	Data Quality Audit	17

5.	Statistical Methods	18

5.1	Minimum Detection Limit	18

5.2	Linearity	18

5.3	Accuracy 	18

5.4	Precision	19

5.5	Interferences	19

6.	Test Results	20

6.1	Minimum Detection Limit	20

6.2	Linearity	20

6.2.1	Source Strength Linearity 	20

6.2.2	Concentration Linearity	23

6.3	Accuracy 	23

6.4	Precision	26

6.5	Interferences	27

6.6	Other Factors	29

6.6.1	Costs	29

6.6.2	Data Completeness	29

7.	Performance Summary	30

8.	References	32

Appendix A: Data Recording Sheet	 A-l

vi

-------
Figures

Figure 2-1. Spectrex SafEye 420 UV Open-Path Monitor	2

Figure 3-1. Test Site at West Jefferson Facility	5

Figure 3-2. Optical Open-Path Monitor Setup	6

Figure 6-1. Source Strength Linearity Plot of the SafEye 420 	 23

Figure 6-2. Concentration Linearity Plot of the SafEye 420 Challenged with

Carbon Disulfide	24

Figure 6-3. Concentration Linearity Plot of the SafEye 420 Challenged with Benzene 	25

Figure 6-4. Concentration Linearity Plot of the SafEye 420 Challenged with Ammonia .... 25

Tables

Table 3-1. Target Gases and Concentrations for Testing the SafEye 420 	3

Table 3-2. Optical Open-Path Monitor Verification: Measurement

Order for Each Target Gas 	4

Table 4-1. Summary of Data Recording Process for the SafEye 420

Verification Test	14

Table 4-2. Summary of Performance Evaluation Audit Procedures 	16

Table 6-1. Minimum Detection Limits Data for the SafEye 420 	 21

Table 6-2. Minimum Detection Limits of the SafEye 420 	 22

Table 6-3. Source Strength Linearity of the SafEye 420 	 22

Table 6-4. Concentration Linearity Data for the SafEye 420 	 24

Table 6-5. Results of Accuracy Tests for the SafEye 420 	 26

Table 6-6. Data from Precision Tests on the SafEye 420 	 27

Table 6-7. Results of Precision Tests on the SafEye 420 	 27

vii

-------
Table 6-8. Concentration Data from Interference Tests on the SafEye 420 	 28

Table 6-9. MDL Data from Interference Tests on the SafEye 420 	 28

viii

-------
List of Abbreviations

AMS

Advanced Monitoring Systems

API

Advanced Pollution Instrumentation

CEM

continuous emission monitor

co2

carbon dioxide

EPA

U.S. Environmental Protection Agency

ETV

Environmental Technology Verification

FIO

flame ionization detector

GC

gas chromatograph

GC/FID

gas chromatography/flame ionization detection

Hg

mercury

m

meters

MDL

minimum detection limit

n2

nitrogen

ND

neutral density

nh3

ammonia

NIST

National Institute of Standards and Technology

NO

nitrogen oxide

ppb*m

parts per billion meters

ppm

parts per million

ppm*m

parts per million meters

QA/QC

quality assurance/quality control

QMP

Quality Management Plan

RSD

relative standard deviation

TSA

technical systems audit

UV

ultraviolet

ix

-------
Chapter 1
Background

The U.S. Environmental Protection Agency (EPA) has created the Environmental Technology
Verification (ETV) Program to facilitate the deployment of innovative environmental tech-
nologies through performance verification and dissemination of information. The goal of the
ETV Program is to further environmental protection by substantially accelerating the acceptance
and use of improved and cost-effective technologies. ETV seeks to achieve this goal by provid-
ing high quality, peer-reviewed data on technology performance to those involved in the design,
distribution, permitting, purchase, and use of environmental technologies.

ETV works in partnership with recognized standards and testing organizations; with stakeholder
groups that consist of buyers, vendor organizations, and permitters; and with the full participation
of individual technology developers. The program evaluates the performance of innovative tech-
nologies 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
protocols to ensure that data of known and adequate quality are generated and that the results are
defensible.

The EPA's National Exposure Research Laboratory and its verification organization partner,
Battelle, operate the Advanced Monitoring Systems (AMS) Center under ETV. The AMS Center
recently evaluated the performance of optical open-path monitors for use in ambient air or fence
line measurements. This verification report presents the procedures and results of the verification
test for the Spectrex Inc. SafEye 420 ultraviolet (UV) open-path monitor.

-------
Chapter 2
Technology Description

The objective of the ETV AMS Center is to verify the performance characteristics of environ-
mental monitoring technologies for air, water, and soil. This verification report provides results
for the verification testing of the SafEye 420. The following description of the SafEye 420 is
based on information provided by the vendor.

The SafEye 420 is an alarm system that detects ammonia, aromatics, and, hydrogen sulfide, using
a high-intensity UV flash source. The detector's three-sensor design includes two absorbed and
one reference band sensor. Depending on the gas to be monitored, the band range of the sensors
can be tailored with a dip switch to meet specific absorption zones.

The SafEye 420 is made up of two components: a flash source and a detector. These components
can be separated to measure ambient gas concentrations over a path length from 1 to 140 meters.
The flash source projects a wavelength (specific for the type of gas to be measured) to the
detector over an unobstructed line of sight. The beam is attenuated when a hazardous gas
traverses it at any point along its path. The detector measures the amount of attenuation by means
of two narrow-band sensors and compares this information to a third reference sensor input that
is not affected by the subject gas or environmental factors.

The detector's microprocessor software interprets the data and provides output signals in terms of
parts per million meters (ppm*m). The detector transmits the data via a 4 to 20 mA signal or an

RS485 port or, if a pre-set gas
concentration is exceeded, closes one of
three contacts.

All the SafEye models (ultraviolet and
infrared) are approved for industrial
applications by international standards:
CENELEC explosion-proof enclosures
(per EN 50014, 50018, and 50019),
Underwriter's Laboratory, and Factoiy
Method (Class I Division 1, Groups B,
C, and D and Class II Division 1,

Groups E, F, and G).

Flash Source

SafEye

Detector

Figure 2-1. Spectrex SafEye 420 UV Open-Path
Monitor

-------
Chapter 3
Test Design and Procedures

3.1 Introduction

This verification test was conducted according to procedures specified in the Test/QA Plan for
Verification of Optical Open-Path Monitors 1(1) The test was designed to challenge the SafEye 420
in a manner simulating field operations and was modeled after Compendium Method TO-16.(2)
The monitor was challenged in a controlled and uniform manner, using an optically transparent
gas cell filled with known concentrations of a target gas. The gas cell was inserted into the
optical path of the monitor during operation under field conditions, simulating the presence of
the target gas in the ambient air.

The monitor was challenged with the three target gases commonly measured by this monitor at
known concentrations, and the measurement results were compared to the known concentration
of the target gas. The gases and concentrations used for testing the SafEye 420 are shown in
Table 3-1. The verification was conducted by measuring the three gases in a fixed sequence over
three days. The one-day sequence of activities for testing the monitor for a single gas is shown in
Table 3-2.

Table 3-1. Target Gases and Concentrations for Testing the SafEye 420

Gas

Concentration
Level

Target Gas Concentration
(ppm*m)a

Equivalent Gas Cell
Concentrationb (ppm)

50.3

335

Carbon

100

665

Disulfide

201

1296

50.3

335

Benzene

100

665

201

1342

50.3

335

Ammonia

100

665

201

1342

appm*m=parts per million meters.
bLength of gas cell = 15.0 cm.

-------
ex
&

- =

fin

C/2 O

cs o

C .-y .-y .-y

c3 c3 c3

c 0) 0) -• fa fa fa c

o 3 3 3 o in in

15 ^ !0 ^

Q Q

z; z;

=a &=a & a a a =«

0) 0) 0) S

a a a gfc

CO CO CO w

^ ^ ^
-------
3.2 Test Design

The verification test was performed near West Jefferson, Ohio, at an outdoor testing area
belonging to Battelle, between October 23 and October 27, 2000. Testing began between 7 and
8 a.m. and ended between 5 and 7 p.m. during these five days. During each of the test days, there
was consistently heavy fog (visibility was less than 100 meters) and precipitation ranging from a
light drizzle to a moderate rain. This location provided sufficient length and a direct line of sight
for each of the path lengths used during the test and provided an area that was away from any
chemical sources that might affect the testing. The same sampling location was used during a
previous period of testing of open-path optical monitors in April and May of 2000. The open
space in the foreground of Figure 3-1 shows the test site at Battelle's West Jefferson facility.

Figure 3-1. Test Site at West Jefferson Facility

The SafEye 420 was challenged with the target gases at the concentrations shown in Table 3-1,
and the SafEye 420 measurement was compared to the known concentration of the target gas. For
each target gas, the monitor was set up as if it were operating in the field, except that an optically
transparent gas cell was placed in the light beam's path (see Figure 3-2). National Institute of
Standards and Technology (NIST-)-traceable or commercially certified standard gases, a
calibrated gas diluter, and a supply of certified high-purity dilution gas were used to supply the
target gases to the gas cell.

-------
GAS CELL

OPTICAL
PATH

GAS
DILUTION
SYSTEM

TO VENT

TARGET GAS
OR GASES

fOf

DILUTION GAS

Figure 3-2. Optical Open-Path Monitor Setup

Target gases were measured at different path lengths, integration times, source intensities, and
numbers of replicate measurements to assess

¦ Minimum detection limit (MDL)

¦ Source strength linearity

¦ Concentration linearity

¦ Accuracy

¦ Precision

¦ Sensitivity to atmospheric interferences.

The test procedures shown in Table 3-2 were nested, in that each measurement was used to
evaluate more than one of the above parameters. In Table 3-2, N2 in the gas cell concentration
column denotes a period of cell flushing with high-purity nitrogen. The denotations cl, c2, and
c3 refer to the concentrations shown in Table 3-1. The last column shows the parameters to be
calculated with the data from that measurement.

-------
3.3 Experimental Apparatus and Materials

3.3.1 Standard Gases

The standard gases used to produce target gas levels for the verification testing were NIST-
traceable gases provided by Scott Specialty Gases Inc. Gravimetrically blended cylinders of
carbon disulfide, benzene, and ammonia were used and were specified to have an accuracy of 2%
of the certified concentration.

3.3.2 Dilution Gas

The dilution gas was acid rain continuous emission monitor (CEM) zero grade nitrogen obtained
from Scott Specialty Gas.

3.3.3 Gas Dilution System

The dilution system used to generate known concentrations of the target gases was an Environics
2020 (Serial No. 2428). This system had mass flow capabilities with an accuracy of approxi-
mately ±1%. The dilution system accepted a flow of compressed gas standard for dilution with
high-purity nitrogen. It was capable of performing dilution ratios from 1:1 to at least 100:1.

3.3.4 Gas Cell

A vendor-provided gas cell 15 centimeters in length was integrated into the end of the receiver.
This cell had two 1/4-inch tube fittings that allowed the target gas to flow through.

3.3.5 Temperature Sensor

An Omega CT485B temperature monitor (Serial No. 704012206W1) with a thermocouple and a
digital temperature readout was used to monitor ambient air and gas cell temperatures. This
sensor was operated in accordance with the manufacturer's instructions and was calibrated
against a certified temperature measurement standard within the 12 months preceding the
verification test.

3.3.6 Ozone Sensor

The sensor used to determine ozone in ambient air was a commercial UV absorption monitor
(ThermoEnvironmental Model 49) designated by EPA as an equivalent method for this
measurement. The UV absorption method is preferred for this application over the reference
method (which is based on ethylene chemiluminescence) because the UV method is inherently
calibrated and requires no reagent gases or calibration standards. The sensor was operated in
accordance with the manufacturer's instructions.

-------
3.3.7 Nitrogen Oxides/Ammonia Monitor

A chemiluminescent nitrogen oxides monitor [Advanced Pollution Instrumentation (API) Model
200, Serial No. 142] was used with a high-temperature ammonia converter (API Model 1000,
Serial No. 100-233-120F-120H) to monitor the nitrogen oxide and ammonia concentrations
supplied to the gas cell for verification testing. This monitor sampled gas immediately down-
stream of the gas cell to confirm the nitrogen oxide or ammonia concentrations prepared by
dilution of high-concentration nitrogen oxide or ammonia standards. The API monitor was
calibrated with a NIST-traceable commercial standard cylinder of nitrogen oxide in nitrogen. The
conversion efficiency for ammonia was checked by comparing the calibration slope for nitrogen
oxide with that found in calibrations with ammonia. All ammonia measurements were corrected
for the ammonia conversion efficiency, which was generally greater than 95%.

3.3.8 Benzene/Carbon Disulfide Measurement

Benzene and carbon disulfide concentrations provided to the gas cell were checked by collecting
a sample at the exit of the cell using pre-cleaned Summa® stainless steel air sampling canisters. A
Hewlett Packard 5880 gas chromatograph (GC) was used to analyze the canister samples for
benzene and carbon disulfide at the ppm concentration levels. A flame ionization detector (FID)
was used to measure the signal response. The compounds were resolved using a fused silica
capillary column (HP-1, 30 m by 0.3 mm with 1.05-//m film thickness). After an initial hold of
2 minutes, the column was temperature programmed from -50 to 220 °C at a rate of 8°C/minute.
Helium was the carrier gas (~3 cc/minute). A syringe filled with a 1-cc sample was used for
benzene analyses. A syringe filled with a 10-cc sample was used for carbon disulfide analyses.
The syringe samples were directed through a heated sampling line and onto a cold trap (-150) for
preconcentration. The trap was then heated to 150°C, and a six-port valve was used to inject the
contents of the trap onto the column. Data acquisition and peak integration were accomplished
with a PC equipped with Chrom Perfect software.

3.4 Test Parameters

3.4.1 Minimum Detection Limit

The MDL was calculated for each target gas by supplying pure nitrogen to the gas cell in the
optical path of the monitor and taking a series of 25 measurements using integration times of
1 and 5 minutes. The MDL was defined as two times the standard deviation of the calculated
concentrations from the 25 absorption spectra. The sequence of measurements was conducted at
both integration times: twice at a 30-meter path length and once at a 90-meter path length for
carbon disulfide and once at a 30-meter path length and once at a 90-meter path length for
benzene and ammonia.

3.4.2 Linearity

Two types of linearity were investigated during this verification: source strength and con-
centration. Source strength linearity was investigated by measuring the effects on the monitor's

-------
performance by changing the source intensity. In the field, light signal levels can be attenuated by
mist, rain, snow, or dirty optical components. As a constant concentration of target gas was intro-
duced into the gas cell, the light intensity of the source was reduced by placing a series of
aluminum wire mesh screens in the path of the light to determine how the monitor's measure-
ments were affected by an attenuated light source. Three aluminum wire screens of various
meshes were placed in the beam path. These screens were approximately 1 foot square and had a
mesh spacing of approximately lA, V2, and 1 inch. At each of these attenuation levels, a measure-
ment was made and the monitor analyzed for the target gas. The test was performed at two
concentrations (50 ppm*m and 200 ppm*m) using benzene.

Concentration linearity was investigated by challenging the SafEye 420 with each target gas at
the concentrations shown in Table 3-1, while the path length and integration time were kept
constant. At each concentration, the monitor response was recorded and its linearity evaluated by
comparing the recorded response with the input target gas concentration.

3.4.3 Accuracy

Accuracy of the monitor relative to the gas standards was verified by introducing known
concentrations of the target gas into the cell. The gas cell was first flushed with at least five cell
volumes of nitrogen, and five zero measurements were recorded. The target gas was then intro-
duced into the cell and, after flushing with at least five cell volumes, five measurements of the
target gas were obtained. The cell was again flushed with at least five cell volumes of nitrogen,
and five more zero measurements were recorded. The concentration of the target gas was the
average value with the target gas in the cell, minus the average of the zero measurements.

The accuracy was evaluated at concentrations denoted as cl through c3, using an integration time
of 1 minute. The accuracy was then evaluated at concentration c2 using a longer integration time,
and then again at concentration c2 using a 1-minute integration time during the interference
measurements (Table 3-2). The percent relative accuracy for an experimental condition is the
absolute value of the difference between the average monitor response and the reference monitor
response, divided by the reference monitor response, times 100 (see Section 5.3).

3.4.4 Precision

The procedure for determining precision was very similar to the procedure for determining
accuracy. The gas cell was flushed with at least five cell volumes of nitrogen. The target gas was
then introduced into the cell and, after flushing with at least five cell volumes, 25 measurements
of the target gas were obtained. The relative standard deviation (RSD) of this set of measure-
ments was the precision at the target gas concentration. Precision was evaluated by this
procedure at one concentration of the target gas (see Table 3-2).

3.4.5 Interferences

The effects of interfering gases were established by supplying the gas cell with a target gas and
varying the distance (i.e., the path length) between the source and detector of the monitor. For the
UV measurement of the target gases, the main interferences in ambient air are oxygen and ozone,

-------
and changing the path length effectively changed the amount of interferants in the light path for
the measurement. The purpose of the interference measurements was to determine the effects that
the ambient atmospheric gases have on the accuracy and MDL of the SafEye 420. Using two
different integration times, these tests were conducted to determine the effect of integration time
on the monitor's ability to perform measurements with interfering gases in the light path.

To determine the effect of the interferences, the path length was first set to 30 meters. The gas
cell was supplied with nitrogen and, after flushing with at least five cell volumes, five measure-
ments were recorded. Next, the target gas was introduced into the cell; and, after similarly
flushing the cell, five measurements were recorded. Finally, the cell was flushed again, and five
more measurements were recorded. Atmospheric concentrations of oxygen and ozone were
recorded at the beginning and the end of these measurements.

The path length was then set to 100 and 90 meters, and the entire measurement procedure was
repeated. The sensitivity of the monitor to the interferant was calculated by comparing the results
at different path lengths (i.e., different ppm*m levels of oxygen and ozone).

-------
Chapter 4
Quality Assurance/Quality Control

Quality assurance/quality control (QA/QC) procedures were performed in accordance with the
quality management plan (QMP) for the AMS Center(3) and the test/QA plan(1) for this
verification test.

4.1 Data Review and Validation

Test data were reviewed by the Verification Testing Coordinator and disclosed to the
Verification Testing Leader. The Verification Testing Coordinator reviewed the raw data and the
data sheets that were generated each day. Laboratory record notebook entries also were signed
and dated.

4.2 Changes from the Test/QA Plan

Two types of changes from the test/QA plan could occur: planned changes to improve the test
procedures for a specific vendor (amendments) and changes that occurred unexpectedly
(deviations).

Before the verification test began, several planned amendments were made to the original
test/QA plan to improve the quality or efficiency of the test. These procedural changes were
implemented and, in each case, either increased the quality of the collected data set or removed
inefficiencies in the test, ultimately resulting in a reduced test duration. A brief summary of these
amendments is provided below:

¦ MDL was determined using twice the standard deviation, as described in section 3.4.1 of the
test/QA plan. The test/QA plan inadvertently called for the MDL to be determined by two
different methods. The correct method was chosen and used during the verification test.

¦ The benzene analysis procedure was changed from that specified in the test/QA plan. The
test/QA plan specified using Method 18 to determine the hydrocarbon emissions from
combustion or other source facilities. This method broadly describes an analysis procedure,
but does not specify how the analysis is to be done, and calls for the use of Tedlar® bags
rather than Summa® canisters. Instead of as described in the test/QA plan, the analysis was
done according to Battelle's GC/FID analysis procedure for canister samples.

-------
¦ The short and the long path lengths in the test/QA plan, which were specified as 100 and
400 meters, were changed to meet the specific technology requirements of the SafEye 420.
In this verification test, path lengths of 30, 90, and 100 meters were used.

¦ Gases for this UV technology were selected based upon the monitor's capability. In addition,
the operating range only permitted using three concentrations. Because of this change in the
specific concentration, measurement #9 rather than measurement #14 was used to calculate
precision.

¦ The order of testing in the test/QA plan was changed. The test order was originally
developed to maximize the efficiency of the test procedure. Several improvements were
made to the test matrix to further improve its efficiency. For example, instead of conducting
all of the measurements for one gas and then changing to the next gas, all of the short path
measurements were conducted before moving to the long path. This was done because
changing the path length was more time consuming than changing the target gas.

¦ The test/QA plan specified that source strength linearity would be tested for each of the
gases. The original intent was to conduct this test for one gas only. The source strength
linearity test, therefore, was conducted only for a single gas.

¦ The original test/QA plan specified that the ambient oxygen concentration be monitored by
an oxygen analyzer. Instead, the ambient oxygen concentrations were assumed to be 20.9%.

¦ Although monitoring ambient carbon monoxide was part of the test/QA plan, it was decided
that carbon monoxide measurements would not add any useful information to the verifica-
tion. Therefore, no carbon monoxide monitoring was performed.

¦ The test/QA plan called for determining ammonia converter efficiency by placing two
converters in series with the nitrogen oxide monitor. Instead, conversion efficiency was
calculated by comparing nitrogen oxide and ammonia calibration curves.

Amendments required the approval of Battelle's Verification Testing Leader and Center

Manager. An amendment form was used for documentation and approval of all amendments.

Deviations from the test/QA plan were as follows:

¦ No independent performance evaluation was conducted for temperature during the
verification test.

¦ Measurement #15 was not conducted for benzene and ammonia.

Deviation reports have been filed for each deviation.

Neither the amendments nor the deviations had a significant impact on the test results used to

verify the performance of the SafEye 420.

-------
4.3 Calibration

4.3.1 Gas Dilution System

Mass flow controllers in the Environics gas dilution system were calibrated by the manufacturer
prior to the start of the verification test by means of a soap bubble flow meter. Corrections were
applied to the bubble meter data for pressure, temperature, and water vapor content.

4.3.2 Temperature Sensor

The thermocouple was calibrated by Battelle's instrument calibration facility on September 21,
2000. This instrument has a one-year calibration period, and so was still within its calibration
interval.

4.3.3 Ozone Sensor

The UV absorption method of ozone measurement is inherently calibrated, relying as it does on
the accurately determined absorption coefficient of ozone. As a result, routine calibration of the
ozone monitor is not needed. However, the monitor was operated according to the manufac-
turer's directions, with careful attention to the diagnostic indicators that assure proper operation.

4.3.4 Nitrogen Oxides/Ammonia Monitor

The nitrogen oxides/ammonia monitor was calibrated with both nitrogen oxide and ammonia
standards. The nitrogen oxide standard was a Certified Master Class Calibration Standard of
6,960 ppm nitrogen oxide in nitrogen, of ± 1% analytical uncertainty (Scott Specialty Gases,
Cylinder No. K026227). The ammonia standard was also a Certified Master Class Calibration
Standard, of 494 ppm ammonia in air, of ± 2% analytical uncertainty (Scott, Cylinder No. ALM
005256). The ratio of the slopes of the ammonia and nitrogen oxide calibration curves
established the ammonia conversion efficiency.

A performance evaluation audit was also conducted once during the test, in which the API
monitor's response was tested with a different nitrogen oxide standard. For that audit, the
comparison standard used was a NIST-traceable EPA Protocol Gas of 3,925 ppm nitrogen oxide
in nitrogen, with ± 1% analytical uncertainty (Scott Specialty Gases, Cylinder No. ALM 057210).

4.3.5 Benzene/Carbon Disulfide Measurement

The GC/MI ) instrumentation was calibrated using a cylinder of benzene in nitrogen, with an
analytical uncertainty of ± 2% (Cylinder No. AAL 18549). Calibration for carbon disulfide was
conducted using a cylinder of carbon disulfide in nitrogen, with ± 2% analytical uncertainty
(Cylinder No. ALM 003452, Scott Specialty Gases).

-------
4.4 Data Collection

Data acquisition was performed primarily by Battelle and the vendor. Table 4-1 summarizes the
type of data recorded (see also the example data recording form in Appendix A); where, how
often, and by whom the recording was made; and the disposition or subsequent processing of the
data. Test records were then converted to Excel spreadsheet files.

Table 4-1. Summary of Data Recording Process for the SafEye 420 Verification Test

Data Recorded

Recorded
By

Where Recorded

When Recorded

Disposition of Data

Dates, Times, Test
Events

Battelle

Data Sheet

Start of each test,
whenever testing
conditions changed

Used to compile results,
manually entered into
spreadsheet as necessary

Test Parameters
(temp., etc.)

Battelle

Data Sheet

Every hour during
testing

Transferred to spreadsheet

Interference Gas
Concentrations

Battelle

Data Sheet

Before and after
each measurement of
target gas

Transferred to spreadsheet

Target Gas
Concentrations

Battelle

Data Sheet

At specified time
during each test

Transferred to spreadsheet

GC Concentrations

Battelle

PC Stored
Chromatograms

After GC analysis

Stored on PC and
on printouts

Optical Open-Path
Monitor Readings

Battelle

Data Sheet

At specified time
during each test

Transferred to spreadsheet

4.5 Audits

4.5.1 Technical Systems Audit

No technical systems audit (TSA) was performed during this verification test. A TSA was
performed on another open-path verification test during the initial testing of this type of
technology. The TSA of similar test procedures was conducted on April 13 and 14 during the
period of open-path monitor verification testing in early 2000. The TSA was performed by
Battelle's Quality Manager as specified in the AMS Center QMP. The TSA ensures that the
verification test is conducted according to the test/QA plan and that all activities associated with
the test are in compliance with the AMS QMP. Specifically, the calibration sources and methods
used were reviewed and compared with test procedures in the test/QA plan. Equipment
calibration records and gas certificates of analysis were reviewed. The conduct of the testing was
observed, and the results were assessed.

All findings noted during the TSA on the above dates were documented and submitted to the
Verification Testing Coordinator for correction. The corrections were documented by the

-------
Verification Testing Coordinator and reviewed by Battelle's Quality Manager, Verification
Testing Leader, and Center Manager. None of the findings adversely affected the quality or
outcome of this verification test, and all were resolved to the satisfaction of the Battelle Quality
Manager. The records concerning the TSA are permanently stored with the Battelle Quality
Manager.

In addition to the internal TSA performed by Battelle's Quality Manager, an external TSA was
conducted by EPA on April 14, 2000, during a previous set of open-path monitor verifications.
The TSA conducted by EPA included all the components listed in the first paragraph of this
section. A single finding was noted in that external TSA, which was documented in a report to
the Battelle Center Manager for review. A response and corrective action were prepared and
returned to EPA. The finding did not adversely affect the quality or outcome of this verification
test. The results of both the Battelle and EPA TSAs were accounted for in preparing for testing
the SafEye 420.

4.5.2 Performance Evaluation Audit

A performance evaluation audit was conducted during the testing period to assess the quality of
the measurements made in the verification test. This audit addressed only those measurements
made by Battelle in conducting the verification test. The performance audit procedures
(Table 4-2) were performed by the technical staff responsible for the measurements. Battelle's
Quality Manager assessed the results. The performance evaluation audit was conducted by
comparing test measurements to independent measurements or standards.

Each of the required procedures for the performance evaluation audit was conducted during the
testing period in accordance with the direction specified in the test/QA plan, except for the
deviation concerning the temperature performance evaluations listed in Section 4.2 of this report.
The results from the performance evaluation are shown in Table 4-2. The temperature measure-
ment agreed to within 0.4°C and the ozone to within 0.4 ppm. The monitor used for nitrogen
oxides/ammonia determination agreed with the performance evaluation standard within 2% at a
concentration of 75 ppm*m.

The benzene and carbon disulfide concentrations were audited by independent analysis of the test
gas mixture supplied to the gas cell during verification testing.

The ( iC/MI ) analysis of both the benzene and carbon disulfide measurements showed that the
performance evaluation failed, resulting in an investigation into the reason for failure. It was
determined that a leaky orifice caused an additional amount of ambient air to flow into the
Summa® canister during sample collection.

The GC/FID results for both benzene and carbon disulfide were lower than the expected value,
based upon the controlled concentration being delivered by the Environics 2020 diluter
(SN 2428) gas dilution system. The Environics, which was calibrated and passed the

-------
Table 4-2. Summary of Performance Evaluation Audit Procedures

Measurement
Audited

Audit Procedure

Reference
Reading

Monitor
Reading

Difference

Acceptance
Criteria

Temperature"

Compare to independent
temperature measurement
(Hg thermometer)

19°C

18.6°C

-0.4°C

< 3°C

Ozone

Compare to independent
ozone measurement

16 ppm

16.4 ppm

2.5%

< 10%

NO/NH3

Compare using another
NO standard from the
same supplier

100 ppm

98 ppm

-2.0%

<5%

Benzene

Compare to results of GC
analysis of canister
sample

335 ppm
665 ppm
1342 ppm

214 ppm
352 ppm
910 ppm

-36.1%
-47.1%
-39.6%

< 10%

Carbon
disulfide

Compare to results of GC
analysis of canister
sample

335 ppm
665 ppm
1296 ppm

280 ppm
392 ppm
1010 ppm

-16.4%
-41.1%
-22.1%

< 10%

Performed on January 27, 2001.

performance evaluation, delivered ammonia, benzene, and carbon disulfide to the target gas cell
at a controlled concentration. The benzene and carbon disulfide performance evaluations were
performed using Summa® canisters. The ammonia performance evaluation used an ammonia
converter and an nitrogen oxide monitor. Ammonia was converted to nitrogen oxide and then
monitored using the API (SN 142). The ammonia performance evaluation passed, as did the
dilution system performance evaluation, confirming that the Environics was functioning
properly. Next, the GC/FID analysis was checked by analyzing an independent cylinder of both
benzene and carbon disulfide. The results from these two analyses agreed with the expected
concentration; and, therefore, it was concluded that the GC analysis procedure was correct.

At the same time that the carbon disulfide and benzene samples were being collected, methane,
propane, and a mixed hydrocarbon were being sampled for verification of a different SafEye
monitor. The same operator conducted both sampling efforts. The only difference in the Summa®
canister collection technique was that two different flow orifices were used on the inlets of the
canisters to control the flow rate into the canister. It was desired that, over a 5-minute period,
approximately 3 liters of sample be collected. This was, in fact, what occurred. The performance
evaluation results from the methane, propane, and mixed gas were satisfactory. Since the only
difference between the two sampling efforts (benzene/carbon disulfide vs. methane/propane/
mixed gas) was the use of different critical flow orifices, it was concluded that the orifice used
during the benzene and carbon disulfide sampling effort allowed ambient air to flow into the
Summa® canister, effectively diluting the canister samples and causing the audit to show lower

-------
than expected results. Therefore, the concentration delivered to the target gas cell is reported as
the nominal value displayed by the Environics 2020 dilution system.

4.5.3 Data Quality Audit

Battelle's Quality Manager audited at least 10% of the verification data acquired in the verifi-
cation test. The Quality Manager traced the data from initial acquisition, through reduction and
statistical comparisons, to final reporting. All calculations performed on the data undergoing
audit were checked.

-------
Chapter 5
Statistical Methods

The following statistical methods were used to reduce and generate results for the performance
factors.

5.1 Minimum Detection Limit

The MDL is defined as the smallest concentration at which the monitor's expected response
exceeds the calibration curve at the background reading by two times the standard deviation (j J
of the monitor's background reading, i.e.,

MDL = 2i.

5.2 Linearity

Both concentration and source strength linearity were assessed by linear regression with the
certified gas concentration as independent variable and the monitor's response as dependent
variable. Linearity was assessed in terms of the slope, intercept, and correlation coefficient of the
linear regression.

y = mx + b

where _y is the response of the monitor to a target gas, x is the concentration of the target gas in
the gas cell, m is the slope of the linear regression curve, and b is the zero offset.

5.3 Accuracy

The relative accuracy (A) of the monitor with respect to the target gas was assessed by

T-R

xlOO

where the bars indicate the mean of the reference (R) values and monitor (7) results.

-------
5.4 Precision

Precision was reported in terms of the percent RSD of a group of similar measurements. For a set
of measurements given by T\, T2,Tn, the standard deviation (j ) of these measurements is

O =

n -1

'L(Tt-fy

k= 1

1/2

where T is the average of the monitor's readings. The RSD is calculated from

RSD =

x 100

and is a measure of the measurement uncertainty relative to the absolute value of the
measurement. This parameter was determined at one concentration per gas.

5.5 Interferences

The extent to which interferences affected MDL and accuracy was calculated in terms of
sensitivity of the monitor to the interferant species, relative to its sensitivity to the target gas, at a
fixed path length and integration time. The relative sensitivity is calculated as the ratio of the
observed response of the monitor to the actual concentration of the interferant. For example, a
monitor that indicates 26 ppm*m of cyclohexane in air with an interference concentration of
100 ppm*m of carbon dioxide indicates 30 ppm*m of cyclohexane when the carbon dioxide
concentration is changed to 200 ppm*m. This would result in an interference effect of
(30 ppm*m - 26 ppm*m)cyclohexane/(200 ppm*m - 100 ppm*m) C02 = 0.04, or 4% relative
sensitivity.

-------
Chapter 6
Test Results

The results of the verification test of the SafEye 420 are presented in this section, based upon the
statistical methods described in Chapter 5. The monitor was challenged with carbon disulfide,
benzene, and ammonia over path lengths of 30 to 100 meters. These gases were chosen because
they are representative of gases monitored by this monitor. Test parameters included MDL,
linearity, accuracy, precision, and the effects of atmospheric interferants on concentration
measurements. The SafEye 420 was programmed to respond using theoretical and limited
empirical calibration data. The vendor indicated that the performance results from this
verification test will be used to make calibration adjustments that improve performance as part of
the SafEye 420's development program.

6.1 Minimum Detection Limit

The MDL was calculated from measurements in which there were no target gases in the gas cell,
but the monitor analyzed the absorption spectra for the presence of a target gas. The data used to
determine the MDL were obtained under several experimental conditions, including different
path lengths and integration times, as shown in Table 6-1. Table 6-2 shows the results of the
MDL calculations.

The results in Table 6-2 show that the SafEye 420 has an MDL of between 0.096 and
0.515 ppm*m for carbon disulfide, 0.111 and 0.340 ppm*m for benzene, and 0.081 and
3.53 ppm*m for ammonia, at the path lengths and integration times tested. Changing the
integration times from 1 to 5 minutes increased the MDL for carbon disulfide. Changing the path
lengths between 30 and 90 meters substantially reduced the MDLs for carbon disulfide and
benzene. The opposite path length effect was seen for ammonia.

6.2 Linearity

6.2.1 Source Strength Linearity

Table 6-3 shows the results from this evaluation of source strength linearity, and Figure 6-1
shows a plot of the effect that the light signal level has on the monitor's measurements. In
Table 6-3, the relative signal power is the measure of light attenuation during that measurement.

-------
Table 6-1. Minimum Detection Limits Data for the SafEye 420

Carbon Disulfide Benzene Ammonia

Path Length (m) Path Length (m) Path Length (m)

30 30 65 30 90 30 90

Measure-

Integration Time (min)

ment

Number

Concentration (ppm*m)

-0.542

0.613

-0.241

0.362

-0.241

-0.542

4.28

0.060

-0.492

-0.241

-0.291

-0.241

-0.492

3.73

-0.492

-0.542

-0.241

-0.291

-0.542

3.73

-0.492

-0.593

-0.241

-0.342

-0.492

5.63

-0.492

-0.593

-0.291

-0.191

-0.342

-0.492

3.63

-0.492

-0.593

-0.291

-0.191

-0.291

-0.492

4.33

-0.442

-0.593

-0.291

-0.191

-0.291

-0.492

2.97

-0.492

-0.593

-0.291

-0.141

-0.342

-0.492

3.83

-0.442

-0.643

-0.291

-0.090

-0.342

-0.492

2.92

-0.492

-0.593

-0.291

-0.040

-0.342

-0.492

5.13

-0.492

-0.643

-0.342

0.060

-0.392

-0.492

3.68

-0.442

-0.643

-0.342

0.161

-0.342

-0.492

3.07

-0.442

-0.643

-0.342

0.060

-0.342

-0.492

2.07

-0.442

-0.643

-0.342

0.110

-0.392

-0.492

2.82

-0.492

-0.643

-0.342

0.161

-0.342

-0.492

2.32

-0.492

-0.643

-0.342

0.161

-0.392

-0.492

2.62

-0.442

-0.693

-0.342

0.060

-0.392

-0.492

2.67

-0.492

-0.643

-0.342

0.010

-0.392

-0.442

4.38

-0.492

-0.693

-0.342

0.161

-0.442

3.12

-0.492

-0.693

-0.342

0.010

-0.442

3.22

-0.442

-0.693

-0.392

-0.141

-0.392

-0.492

2.37

-0.442

-0.693

-0.392

-0.090

-0.392

-0.442

3.32

-0.492

-0.693

-0.392

-0.090

-0.392

3.48

-0.492

-0.743

-0.392

-0.241

-0.392

5.03

-0.492

-0.743

-0.342

-0.141

-0.442

-0.392

11.1

-------
Table 6-2. Minimum Detection Limits of the SafEye 420

Path Length Integration

MDL

Target Gas (m)

Time (min)

(ppm*m)

Carbon disulfide 30

0.222

Carbon disulfide 30

0.515

Carbon disulfide 90

0.096

Benzene 30

0.340

Benzene 90

0.111

Ammonia 30

0.081

Ammonia 90

3.53

Table 6-3. Source Strength Linearity of the SafEye 420

Relative Benzene Concentration

Monitor

Signal Power

(ppm*m)

Response (ppm*m)

1.00

50.3

84.1

0.79

50.3

84.2

0.57

50.3

85.7

0.38

50.3

86.3

1.00

201

213

0.79

201

213

0.57

201

213

0.38

201

214

For example, a relative signal power of 0.79 means that the light level for that test is 79% of
what the light level is during normal operating conditions. The benzene concentration is the con-
centration of gas being delivered to the gas cell during the measurement, and the monitor
response is the resulting reading from the SafEye 420. The source strength results show that there
is little degradation in monitor performance during conditions of declining source strength. The
data indicate a slight effect of source strength on benzene measurement, with source reductions
of up to 62%. The slopes of the linear regression lines of -0.49 and -4.0 , shown in Figure 6-1,
indicate that reducing the source strength had a slightly positive effect on the monitor's response
over the range tested.

-------
y=-0.5C" +2:::
r=; 59

~ Benzene Concentration (ppnfm)
s MonitorResporise (ppnfm)
— i ine;¦ (Mo-1o:R^enonef isn•-jj

y = -4.0x - 88

0.20 0 30 0 40 0 50 0 6C C.7C C:.8C 0.9:! VO! 1;

Relative Signal Strength (Arb. Units)

Figure 6-1. Source Strength Linearity Plot of the SafEye 420
6.2.2 Concentration Linearity

Table 6-4 and Figures 6-2 through 6-4 show the path-average results of the evaluation of
concentration linearity. The regression analysis results are shown on the individual figures.

The target gas concentration values used for this calculation are based on the concentration of gas
delivered by the Environics 2020 dilution system and not the concentrations as determined by
GC (as explained in Section 4.5.2 of this report).

The concentration linearity results show that the SafEye 420 has a linear response over the
concentration ranges tested. The monitor response as given by the slope of the linear regression
line is 0.56 for carbon disulfide, with an r2 value of 0.47; a slope of 0.73 for benzene, with an r2
value of 0.59; and a slope of 1.2 for ammonia, with an r2 value of 0.95.

The best results were found when the monitor was challenged with ammonia. The results from
challenges of benzene and carbon disulfide show that the instrument generally responds to both
compounds; however, the low r2 values show that the instrument's response is variable and not
linear over the tested range.

6.3 Accuracy

The accuracy of the SafEye 420 was evaluated at each target gas concentration introduced into
the cell. These concentrations were introduced at the path lengths and integration times shown in
Table 6-5. The accuracy results compare the monitor response with the target gas concentration
as delivered by the Environics 2020 diluter. The SafEye 420's relative accuracy ranged from

-------
Table 6-4. Concentration Linearity Data for the SafEye 420

Target Gas

Target Gas Concentration
(ppm*m)

Monitor Response
(ppm*m)

Carbon disulfide

50.3

113

Carbon disulfide

99.8

213

Carbon disulfide

194

214

Carbon disulfide

99.8

190

Carbon disulfide

99.8

208

Benzene

50.3

86.7

Benzene

99.8

200

Benzene

201

209

Benzene

99.8

118

Benzene

99.8

156

Ammonia

50.3

29.3

Ammonia

99.8

66.7

Ammonia

201

210

Ammonia

99.8

108

Ammonia

99.8

101

y = 0.56x + 130
r2 = 0.47

100 150

Target Gas Concentration (ppm*m)

Figure 6-2. Concentration Linearity Plot of the SafEye 420
Challenged with Carbon Disulfide

-------
r2 = 0.59

100 150

Target Gas Concentration (ppm*m)

Figure 6-3. Concentration Linearity Plot of the SafEye 420
Challenged with Benzene

100.0 -

y = 1.2X-28

r2 = 0.95

100 150

Target Gas Concentration (ppm*m)

Figure 6-4. Concentration Linearity Plot of the SafEye 420
Challenged with Ammonia

-------
Table 6-5. Results of Accuracy Tests for the SafEye 420

Target Gas

Target Gas
Concentration
(ppm*m)

Path
Length (m)

Integration
Time
(min)

Monitor

Response

(ppm*m)

Relative
Accuracy

(%)

Carbon disulfide

50.3

113

126

Carbon disulfide

99.8

213

113

Carbon disulfide

194

214

Carbon disulfide

99.8

100

190

Carbon disulfide

99.8

208

108

Benzene

50.3

86.7

Benzene

99.8

200

100

Benzene

201

209

Benzene

99.8

100

118

Benzene

99.8

156

Ammonia

50.3

29.3

-41

Ammonia

99.8

66.7

-33

Ammonia

201

210

Ammonia

99.8

100

108

Ammonia

99.8

101

10 to 126% for carbon disulfide, from 4 to 100% for benzene, and from -41 to 8% for ammonia.
Integration time had little effect on the accuracy of the SafEye 420.

The results from the accuracy tests show that the monitor is most accurate when challenged with
ammonia. Both carbon disulfide and benzene have widely varying accuracy results, with the best
relative accuracy at the shortest tested path length (30 meters) and concentrations at the high end
of the instrument's operating range (200 ppm*m in the cell); however, the same 30-meter path
length also resulted in the poorest accuracy when challenged at lower concentrations. Increasing
the integration time from 1 to 5 minutes had no consistent effect on the relative accuracy results.

6.4 Precision

Precision data were collected during measurement #9 (see Table 3-2) using an integration time of
1 minute and a path length of 100 meters. The target gas was introduced into the gas cell at a
fixed concentration, and 25 successive analyses were made for the target gas. The data from these
measurements are found in Table 6-6, and the results are shown in Table 6-7. Table 6-7 shows
precision of 0.00% RSD for carbon disulfide, 3.52% RSD for benzene, and 2.45% RSD for
ammonia. The variability for benzene and ammonia occurs in the form of sporadic individual
readings that differ sharply from the other values, which are highly consistent.

The 0.00% RSD found for the carbon disulfide is probably the result of a saturated signal. At the
concentration used for this test, the monitor produced a constant, saturated signal of 213 ppm*m,
resulting in a zero reading. This may also be the case with benzene, although there were several

-------
Table 6-6. Data from Precision Tests on the SafEye 420

Target Gas

Carbon Disulfide

Benzene

Ammonia

Analysis #

(ppm*m)

213

208

213

211

213

192

210

213

214

213

191

213

214

213

214

213

214

213

192

214

213

215

213

215

213

214

213

215

213

215

213

215

213

215

213

215

213

192

215

213

215

213

215

213

215

213

215

213

215

Table 6-7. Results of Precision Tests on the SafEye 420a

Gas Cell

Average Monitor

Standard

Concentration

Response

Deviation

Relative Standard

Target Gas

(ppm*m)

Deviation (%)

Carbon disulfide

194

213

0.000

Benzene

201

211

7.09

3.52

Ammonia

201

213

4.92

2.45

a Integration time = 1 minute, path length = 30 meters.

-------
occasions where lower than saturation level readings (three instances of 192 ppm*m) were
recorded.

6.5 Interferences

Interference tests of the SafEye 420 evaluated the effects that the common atmospheric inter-
ferants ozone and oxygen have on the monitor's ability to determine the concentration of the
target gases and on the MDL for the target gases. Because of the large relative accuracies that
were seen during the accuracy tests, it is difficult to determine whether changes in the monitor's
ability to perform properly when challenged with the target gas is a result of interfering com-
pounds in the atmosphere or of other effects. The accuracy results were best for ammonia.
Examining the results from the ammonia challenge, it can be seen that the longer path lengths of
90 and 100 meters are more accurate than the 30-meter path length, indicating that the increasing
presence of interfering compounds did not adversely affect the monitor's ability to measure
ammonia. Tables 6-8 and 6-9 show the data used to determine the interference effects of ozone
and oxygen on the concentration and MDL.

Table 6-8. Concentration Data from Interference Tests on the SafEye 420

Path

Concentration

Target Gas

Monitor

Relative

Target

Length

of Oxygen

of Ozone

Concentration

Response

Accuracy

Gas

(m)

(%*m)

(ppb*m)

(ppm*m)

(%)

Carbon disulfide

627

1740

99.8

213

114

Carbon disulfide

1881

3330

99.8

208

109

Carbon disulfide

100

2090

2700

99.8

190

90.5

Benzene

627

480

99.8

200

101

Benzene

1881

4860

99.8

156

56.3

Benzene

100

2090

1300

99.8

118

18.4

Ammonia

627

1380

99.8

66.7

-33.2

Ammonia

1881

1620

99.8

101

1.71

Ammonia

100

2090

3500

99.8

108

8.55

Table 6-9. MDL Data from Interference Tests on the SafEye 420

Concentration

Target

Path Length

of Oxygen

of Ozone

MDL

Gas

(m)

(%*m)

(ppb*m)

(ppm*m)

Carbon disulfide

627

1650

0.222

Carbon disulfide

1881

3960

0.096

Benzene

627

420

0.340

Benzene

1881

4680

0.111

Ammonia

627

1170

0.081

Ammonia

1881

1890

3.53

-------
Both ozone and oxygen have absorption features in the same spectral region that the SafEye 420
uses to analyze for the target compounds. Because the concentration of these two potential
interferants is usually much greater than the concentration of the compounds of interest, the
presence of these compounds can make analyzing for the target compounds difficult. The
SafEye 420 uses various methods to deal with these interferants, and this test evaluated the
effectiveness of these methods.

These results did not permit calculation of relative sensitivity, as described in Section 5.5.
Instead, a comparison of the measured concentrations was made to the input concentrations.

Changing the total number of ozone and oxygen molecules in the path length had little effect on
the monitor's MDL for the target gas. For both carbon disulfide and benzene, lower MDLs were
found with the longer path length, despite the increased amounts of ozone and oxygen in the
optical path.

6.6 Other Factors

6.6.1 Costs

The cost of the SafEye 420, as tested, ranges from $7,000 to $12,000, according to Spectrex.

6.6.2 Data Completeness

All of the expected data were collected except for measurement #15 for benzene and ammonia.
Data from measurement #15 for benzene and ammonia were not collected because the vendor
declined to conduct these two measurements.

-------
Chapter 7
Performance Summary

The SafEye 420 minimum detection limits ranged between 0.096 and 0.515 ppm*m for carbon
disulfide, 0.111 and 0.340 ppm*m for benzene, and 0.081 and 3.53 ppm*m for ammonia, at the
path lengths and integration times tested. Changing the integration times from 1 to 5 minutes
increased the MDL for carbon disulfide. Changing the path lengths between 30 and 90 meters
substantially reduced the MDLs for carbon disulfide and benzene. The opposite path length effect
was seen for ammonia.

The tests of the effects of source strength on the measurement capability of the monitor showed
that there was little to no degradation of monitor performance, with reductions in source strength
of up to 62%. The slopes at two different test concentrations were slightly negative, suggesting
that reducing the source strength may have a slight positive effect on the monitor's response over
the range tested.

The concentration linearity results showed that the SafEye 420 had a slope of 0.56 for carbon
disulfide, with an r2 value of 0.47 over a range of 50.3 to 194 ppm*m; a slope of 0.73 for
benzene, with an r2 value of 0.59 over a range of 50.3 to 201 ppm*m; and a slope of 1.2 for
ammonia, with an r2 value of 0.95 over a range of 50.3 to 201 ppm*m.

Percent relative accuracy was evaluated over the same ranges of concentration noted above for
concentration linearity testing. Relative accuracy over these ranges was 10 to 126% for carbon
disulfide, from 4 to 100% for benzene, and from -41 to 8% for ammonia. The accuracy tests
show that the monitor is most accurate when challenged with ammonia. Both carbon disulfide
and benzene showed the best relative accuracy at the shortest tested path length (30 meters) and a
concentration at the high end of of the instrument's operating range (200 ppm*m in the cell);
however, the same 30-meter path length also resulted in the poorest accuracy when challenged at
lower concentrations.

Precision results showed that the SafEye 420 had an RSD of about 0.00% for carbon disulfide
at a gas cell concentration of 194 ppm*m, a 3.52% RSD for benzene at a concentration of
201 ppm*m, and a 2.45% RSD for ammonia at a concentration of 201 ppm*m at a path length
of 30 meters.

Analysis of the effects of interferences of oxygen and ozone on the measuring ability of the
SafEye 420 showed that the MDLs were not affected. However, when examining only the
accuracy results from the ammonia challenge, it can be seen that the longer path lengths of

-------
90 meters and 100 meters are more accurate than the 30-meter path length, indicating that the
increasing presence of interfering compounds did not adversely affect the monitor's ability to
measure ammonia. The results from the benzene and carbon disulfide challenges showed no
consistent effects, especially in light of the large relative accuracy values found for these two
gases during the accuracy test.

-------
Chapter 8
References

1. Test/QA Plan for Verification of Optical Open-Path Monitors, Battelle, Columbus, Ohio,
October 28, 1999.

2. Compendium Method TO-16 Long-Path Open-Path Fourier Transform Infrared Monitoring
of Atmospheric Gases, EPA-625/R-96/010b, U.S. Environmental Protection Agency,
Cincinnati, Ohio, January 1999.

3. Quality Management Plan (QMP) for the ETV Advanced Monitoring Systems Pilot, Version
2.0, U.S. EPA Environmental Technology Verification Program, Battelle, Columbus, Ohio,
October 2000.

-------
Appendix A
Data Recording Sheet

A-l

-------
ETV - Spectrex, Inc

Etv Advanced Monitoring Systems Pilot Verification of Optical Open Path Monitor Round Two

Meas. # (From tesWA plan Table 'i'f1:

Cell Temp

Int. Time

Vendor; Spectrex, Inc.

(F):

(min):

Iristru merit

C02 Cone.

Pathlength

Mode!:

(ppm):

(meters):

Battelle, West Jefferson,

Ambient RH

Cell Length

Location: Ohio

(%)¦

(cm):

Vendor

Ambient

Operator: Jay Cooley

Temp (F):

Sample Gas:

ozone uoric.

Time:

(ppb):

Sample Gas

Cone, in Ceil

Date:

(ppm):

Data Point #

Meas. Result
(voits)

Note: Measurement #s (3,4,5,10,11&12
only)

Neutral

Mon itor

Density Filter

Desired

response
(volts)

Attenuation

none

20%

40%

60%

none

Data taken by:

Date:

Data reviewed by:

Date:

data sheet.xls

A-2

-------