EPA-650/2-74-019
DEVELOPMENT AND TESTING
OF AN AIR MONITORING
SYSTEM
by
C. E. Decker, T. M. Royal, and J. B. Tommerdahl
Research Triangle Institute
Research Triangle Park, North Carolina 27709
Contract No. 68-02-1011
Program Element No . 1A1003
EPA Project Officer: Robert K. Stevens
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
December 1973
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
ABSTTMCT
The purpose of this contract was to test and evaluate instrumentation
based on specified measurement principles selected for the Regional Air
Pollution Study (RAPS), where an extensive network of air monitoring
stations equipped with state-of-the-art monitoring instrumentation and
sophisticated data acquisition and computer processing systems will be
required. The primary objective of the program was to equip an experi-
mental trailer with selected instrumentation and to evaluate these
instruments at a non-urban site to: determine if the monitors selected for
the study can meet the required performance specifications or need to be
modified; determine the operating environment needed to obtain optimum
performance from these monitors; evaluate the latest calibration techniques
and select calibration procedures to provide the most reliable measurements;
and recommend, based on the results of the evaluation program, instrumenta-
tion for use in the RAPS program.
Instrumentation for the measurement of ozone, sulfur compounds (sulfur
dioxide, hydrogen sulfide, total sulfur), nitric oxide, nitrogen dioxide,
hydrocarbons (total hydrocarbon, methane, non-methane hydrocarbon) and
carbon monoxide in ambient air were included in the program. The evaluation
of each instrument included in the program was a systematic comparison of
its ability to obtain reliable data with emphasis placed on implementation,
if possible, of design changes to instrumentation that could not meet the
operational requirements specified for this program.
iii
-------
ACKNOWLEDGE
The work on this project was performed by the Instrumentation,
Measurements and Device Research Department, Engineering Division of
Research Triangle Institute for the Field Methods Development Section,
Chemistry and Physics Laboratory, Environmental Protection Agency.
Mr. J. B. Tommerdahl, Manager, Instrumentation, Measurements and Device
Research Department, served as Laboratory Supervisor; and Mr. C. E. Decker
served as Project Leader. Mr. T. M. Royal, Mr. R. W. Murdoch, and
Mr. J. H. White participated in the instrument evaluation program.
Appreciation is expressed to Messrs. Stevens, Clark, Baumgardner and
others of the Field Methods Development Section for their cooperation and
assistance during this program.
iv
-------
TABLE OF CONTENTS
ABSTRACT
ACKNOWLEDGMENT
SECTION PAGE
1.0 INTRODUCTION 1
1.1 BACKGROUND AND OBJECTIVES 1
1.2 EXPERIMENTAL APPROACH 2
2.0 DESCRIPTION OF FACILITIES 5
2.1 SHELTER/ENVIRONMENTAL CONTROL 5
2.2 AMBIENT AIR SAMPLING SYSTEM 5
2.3 DATA ACQUISITION AND PROCESSING OF DATA 8
2.3.1 Data Acquisition System 8
2.3.2 On-Line Computer 12
3.0 SELECTION OF INSTRUMENTS 21
3.1 SELECTION CRITERIA 21
3.2 TEST PLAN 24
4.0 GAS ANALYZERS 26
4.1 BENDIX MODEL 8002 GAS PHASE CHEMILUMINESCENT
OZONE ANALYZER 26
4.1.1 Instrument Description 26
4.1.2 Operational Summary 30
4.2 DASIBI MODEL 1003-AH ULTRAVIOLET OZONE ANALYZER 32
4.2.1 Instrument Description 32
4.2.2 Operational Summary 35
V
-------
TABLE OF CONTENTS (Cont'd)
SECTION PA6E
4.3 BENDIX MODEL 8101-B CHEWLUMINESCENT
NO-NOX-N02 ANALYZER 36
4.3.1 Instrument Description 36
4.3.2 Operational Summary 43
4.4 MELOY MODEL 520 CHEMILUMINESCENT
NO-NOX-N02 ANALYZER 44
4.4.1 Instrument Description 44
4.4.2 Operational Summary 48
4.5 THERMO ELECTRON MODEL 14 CHEMILUMINESCENT
NO-NO -N0? ANALYZER 49
rt fc
4.5.1 Instrument Description 49
4.5.2 Operational Summary 53
4.6 REM MODEL 642 CHEMILUMINESCENT NO-NOX-N02 ANALYZER 53
4.6.1 Instrument Description 53
4.6.2 Operational Summary 58
4.7 MELOY MODEL SA-185R FLAME PHOTOMETRIC TOTAL
SULFUR ANALYZER 59
4.7.1 Instrument Description 59
4.7.2 Operational Summary 64
4.8 BENDIX MODEL 8300 FLAME PHOTOMETRIC TOTAL
SULFUR ANALYZER 64
4.8.1 Instrument Description 64
4.8.2 Operational Summary 67
4.9 BECKMAN MODEL 6800 AIR QUALITY CHROMATOGRAPH
THC, CH4, CO ANALYZER 69
4.9.1 Instrument Description 69
4.9.2 Operational Summary 72
vi
-------
TABLE OF CONTENTS (Cont'd)
SECTION PAGE
4.10 BENDIX MODEL 8201 AMBIENT HYDROCARBON ANALYZER 73
4.10.1 Instrument Description 73
4.10.2 Operational Summary 76
4.11 BENDIX MODEL 8501-5FA CARBON MONOXIDE ANALYZER 76
4.11.1 Instrument Description 76
4.11.2 Operational Summary 80
4.12 ANDROS MODEL 7000 DUAL-ISOTOPE-FLUORESCENCE
CARBON MONOXIDE ANALYZER 81
4.12.1 Instrument Description 81
4.12.2 Operational Summary 83
4.13 TRACOR MODEL 270-HA GAS CHROMATOGRAPHIC-FLAME
PHOTOMETRIC ANALYZER 84
4.13.1 Instrument Description 84
4.13.2 Operational Summary 88
4.14 BENDIX MODEL 8700 GAS CHROMATOGRAPH-FLAME
PHOTOMETRIC ANALYZER 88
4.14.1 Instrument Description 88
4.14.2 Operational Summary 91
5.0 INSTRUMENT EVALUATION PROCEDURES 92
5.1 DEFINITIONS 92
5.2 TEST PROCEDURES 94
5.2.1 Stability - Zero and Span Drift 94
5.2.2 Linearity 95
5.2.3 Precision 95
5.2.4 Minimum Detectable Concentration 95
vii
-------
TABLE OF CONTENTS (Cont'd)
SECTION PAGE
5.2.5 Response Time 98
5.2.6 Noise 99
5.2.7 Line Voltage Variation 99
5.2.8 Flow Rate/Pressure Variation 101
6.0 CALIBRATION SYSTEMS/PROCEDURES 102
6.1 CALIBRATION SYSTEMS 102
6.1.1 Ozone 102
6.1.2 Nitric Oxide/Nitrogen Dioxide 102
6.1.3 Sulfur Dioxide/Hydrogen Sulfide 107
6.1.4 Hydrocarbons/Carbon Monoxide 109
6.2 CALIBRATION PROCEDURE 110
7.0 SUMMARY OF INSTRUMENT PERFORMANCE 112
7.1 EVALUATION TESTS 112
7.1.1 Stability - Zero and Span Drift 113
7.1.2 Linearity 114
7.1.3 Precision 115
7.1.4 Minimum Detectable Limit 116
7.1.5 Response (Lag, Rise, Fall time) 118
7.1.6 Noise 120
7.1.7 Line Voltage Variation 120
7.1.8 Sample Flow Rate Test 122
7.2 OPERATIONAL SUMMARY 124
viii
-------
TABLE OF CONTENTS (Cont'd)
SECTION PAGE
7.3 SUBSYSTEM COMPONENT EVALUATION/RECOMMENDED
MODIFICATIONS TO IMPROVE PERFORMANCE 124
7.3.1 Ozone Analyzers 127
7.3.2 Oxides of Nitrogen Analyzers 128
7.3.3 Sulfur Analyzers 131
7.3.4 Hydrocarbon Analyzers 132
7.3.5 Carbon Monoxide Analyzers 137
7.3.6 General List of Design Deficiencies/
Considerations 140
8.0 RECOMMENDATIONS FOR SELECTION OF INSTRUMENTS 144
9.0 REFERENCES 150
APPENDIX A: RECOMMENDED CALIBRATION PROCEDURES 153
APPENDIX B: RECOMMENDED OPERATING PROCEDURES 163
APPENDIX C: RECOMMENDED MAINTENANCE PROCEDURES 171
APPENDIX D: GRAPHICAL PRESENTATION OF ZERO AND SPAN DRIFT 179
APPENDIX E: SUMMARY OF DATA FORMATS 203
ix
-------
1.0 INTRODUCTION
1.1 BACKGROUND AND OBJECTIVES
During the past five (5) years the Environmental Protection Agency
has been participating in a demanding instrument development program.
Instrumentation based on newly developed measurement principles, such as
chemiluminescence, flame photometry, gc-flame photometry, and gc-flame
ionization, have been developed and marketed by various manufacturers.
Improvements and modifications to instrumentation based on these measure-
123
merit principles are constantly occurring. Previous field studies ' '
conducted to evaluate some of these analyzers and to compare them to
classical measurement principles, such as colorimetry, coulometry, and
conductivity, have shown the need for intensive evaluation of new instru-
mentation for monitoring atmospheric pollutants. In most cases where
prototype or first production models of an ambient air analyzer were
evaluated, excessive failures occurred which seriously impaired the
analyzer's performance. In every case these failures resulted from
electronic and/or flow problems and interferences, such as water vapor,
interferent gases, vibrations, etc., that could be solved by engineering
modifications. This observation indicates that important output
requirements of an evaluation program are to point out deficiencies, to
document required engineering modifications, and to feed back this
information to manufacturers and other interested parties sothat
future procurements of similar equipment can incorporate these
modifications and/or changes.
During the Regional Air Pollution Study (RAPS) to be conducted in
St. Louis, Missouri, an extensive network of air monitoring stations
equipped with state-of-the-art monitoring instruments and automatic data
acquisition systems will be established. Prior to this, it was deemed
necessary to equip an experimental trailer with the latest instrumentation
and to evaluate these instruments at a non-urban site to:
-------
determine whether these new monitors, can adequately measure
concentrations of H2S, SO.,, CO, NO, NO,, NO , 0., THC, CH.,
and non-methane hydrocarbons in a typical non-urban location;
determine if the monitors selected for this study can meet
the required performance specifications or need to be
modified;
determine the operating environment needed to obtain optimum
performance from the monitors;
evaluate the latest calibration techniques and select
calibration procedures to provide the most reliable
measurements; and
recommend, based on the results of the evaluation program,
instrumentation for use in the RAPS program.
The objective of this program was to test and evaluate instrumentation based
on the specified measurement principles selected for the RAPS Program with
emphasis being placed on implementation, if possible, of design changes to
instrumentation that cannot meet the operational requirements specified for
the objectives of the program*
1.2 EXPERIMENTAL APPROACH
An environmentally controlled mobile laboratory was used to house the
instrumentation, data acquisition system, on-line computer, and supporting
laboratory equipment required to complete this program. Figure 1.1 shows
two internal views of the mobile laboratory, including instrumentation and
the data acquisition and processing equipment. These facilities are
described in detail in Section 2.0 and includes a description of the mobile
laboratory, the ambient air sampling system, the data acquisition system,
an on-line computer for processing of data in real-time, and a general
discussion of the automatic calibration technique, mode switches, and
computer programs utilized to process the data in real-time.
-------
Figure 1.1: Two Internal Views of the Mobile Laboratory
3
-------
Criteria for selection of instruments and the test plan are presented
in Section 3.0. The gas analyzers under evaluation are described in
Section 4.0. To expedite the presentation, a standard format was followed
for each analyzer and is presented in two parts. Part I includes the
following: (1) principle of operation, (2) physical characteristics,
(3) gas flow system, (4) detector system, and (5) signal processing.
Part II includes the operational summary for the evaluation period,
failures, operation, calibration, maintenance requirements, and other
pertinent data.
Instrument evaluation procedures, including definitions and test
procedures, are presented in Section 5.0. The calibration schedule varied
depending on the test plan and evaluation procedure being performed.
Calibration procedures used during this study are given in Section 6.0.
A summary of instrument performance is presented in Section 7.0 and
includes results of the evaluation tests, operational summary, and
subsystem component evaluation. Recommendations for selection of instru-
mentation for the RAPS Program are given in Section 8.0, and recommended
calibration, operational, and maintenance procedures are presented in
Appendices A, B, and C, respectively. Graphical presentations of zero and
span drift for each analyzer evaluated are presented in Appendix D.
Computer output formats are presented in Appendix E.
-------
2,0 EESCRIPTION OF FACILITIES
2.1 SHELTER/ENVIRONMENTAL CONTROL
The semi-mobile environmental monitoring laboratory owned by Research
Triangle Institute and previously described in the final report for
Contract CPA 70-101, "Field Evaluation of New Air Pollution Monitoring
Systems" , was used to provide the housing and environment required for
this evaluation program. This facility contained the necessary structural
and electrical facilities and sufficient interior space for housing all
the sensors. The environmental control system was capable of maintaining
the interior temperature to within ± 4°F within the range 60-90°F. For
this evaluation program the semi-mobile environmental monitoring laboratory
was located on the Research Triangle Institute's campus in Research
Triangle Park, North Carolina. Figure 2.1 shows the mobile laboratory on
location at Research Triangle Institute.
2.2 AMBIENT AIR SAMPLING SYSTEM
The ambient air sampling system utilized in the program was constructed
of 1-inch (0.4 cm) O.D. pyrex glass and consisted of a cane to prevent
moisture and particulates from setting into the inlet; a particulate trap
to remove large suspended particulate; six 1.5-meter sections of glass
manifoldeach section having four sampling ports; and a blower. Ambient
air was aspirated through the manifold at a rate of approximately 3 CFM.
Sampling ports made of 12/5 ball-and-socket joints were used for each
hookup of instrument sample inlet lines which were all Teflon. A diagram
of the manifold system is shown in Figure 2.2.
An experiment was conducted to determine possible wall losses of
pollutants on the manifold surfaces of glass and teflon. Ozone was chosen
as the pollutant for investigation due to its reactivity. Two chemilu-
minescent gas phase ozone meters were installed in the mobile laboratory
at each end and calibrated simultaneously. Immediately after calibration
one analyzer was connected to the first sampling port on the manifold and
-------
cd
1-1
o
8
(J
JO
±
ctf
s
Q)
4J
|W
CN
0)
M
a,
TK
I*
-------
-------
the other analyzer connected to the last sampling port on the manifold
some 9 meters away. Ozone concentration measurements were made at both
sampling locations on ambient air and on ambient air spiked with ozone
generated using an ultraviolet lamp for a period of four hours. Examina-
tion of the data indicated that the two measurements were identical down
to the part-per-billion level and that ozone was not being destroyed in
the manifold under the conditions the air sampling system was being
operated. Under these conditions it is very unlikely that other pollutants
of interest, such as S02» NO, N02, THC, NMHC, and CO, are being destroyed
due to wall effects in the air sampling system.
2.3 DATA ACQUISITION AND PROCESSING OF DATA
2.3.1 Data Acquisition System
The basic purpose of the data acquisition system was to automatically
acquire and record in digital form the output signals derived from the gas
analyzers under evaluation and their associated instrument status mode
switches. The data acquisition system used in the RTI mobile laboratory
basically consists of a Hewlett-Packard 2015H data system, modified and
expanded to meet the requirement of a flexible instrument evaluation
program. A block diagram of the system as it now stands is shown in
Figure 2.3.
Analog signals from the gas analyzers and instrument status switches
enter the system by means of an external Junction box affixed to the side
of the rack panel cabinets. From the input junction box, signals are
carried to the sensor coupler for signal conditioning. This unit houses
the necessary bridge circuits, scaling networks, bias voltages, etc., which
are utilized in converting or modifying the sensor output signal to forms
or levels more suitable for recording. Up or down scaling of the signals
is sometimes required in order to match the input signal requirements of
the analog recorders. In addition, filter networks are incorporated where
it is necessary to smooth the signal in order to obtain sampled data that
are representative of the preceding sampling interval.
-------
a:
CD UJ
OQ
_l DC.
.
CO
O
H
4-1
H
W
H
3
to
n
6
(fl
i-l
M
n!
H
n
c
o
o
c
CN
3
60
H
-------
A plug-in card is provided for each signal or data channel in which
the scaling and filtering networks are incorporated. There are two signal
outputs for each channel. One is tied into the digital recording system,
and the other into the analog-monitoring or recording system through a
front-panel-located patch panel. Isolation is provided between each of the
outputs so that on-line tests may be made without disturbing the digital
recording system.
From the signal conditioning unit, the analog signals are passed to
the digital recording system for recording on magnetic tape. This system,
a Hewlett Packard Model 2015H, consists of a scanner capable of accepting
up to 200 floating inputs. For this evaluation program, only 65 active
channels were needed to monitor sufficiently all the necessary parameters.
After passing through the scanner the analog signals are converted to
digital numbers by means of a digital voltmeter. The particular model
used is a conventional 5-1/2 digit, autoranging model, capable of measuring
dc voltage levels from 1.0 volt full scale to 100 volts full scale with an
accuracy of 0.01%. Resolution on the lowest scale is 6 digits, yielding a
resolution of 10 uV on the most sensitive scale. System control is accom-
plished by a clock controller unit, the purpose of which is to supply
timing pulses and a BCD representation of time to the remaining system.
This time code is entered as the first word in each scan, which is days,
hours and minutes.
The BCD output of the digital voltmeter along with the time from the
clock and a channel indicator from the scanner are input to a tape coupler
which feeds the data to the tape deck with the proper accompanying control
signals for interrecord gapping and end of files.
Manual data such as instrument mode or status are introduced into the
system via manual data entry channels. By utilizing codes, the status or
operational mode information was placed in the respective channels and
used in the data processing phase to indicate the operational status of the
respective sensors being evaluated at any point in time. The manual data
entry modes shown in Table 2.1 were those used to describe instrument
operational status.
10
-------
TABLE 2.1 INSTRUMENT OPERATIONAL MODES
Output
Symbol
C
A
B
T
X
Q
R
L
M
99999 or $$$$
Mode
Switch
Setting
0
1
2
3
4
5
6
7
8
9
Operating Condition
Measure - - - Valid Ambient Data
Calibration - Stabilizing
Calibration - Zero Averaging
Calibration - Multi-point Averaging
Routine Test Procedures
Offline - Not Setup or Available
Awaiting Repair
Repair
Awaiting Maintenance
Maintenance
Data Not Available
The auxiliary output of the tape coupler of the data acquisition
system is capable of feeding voltage data directly to a column printer or
to an on-line computer. This type of system is a satisfactory arrangement,
although not ideal, in that it allows data to be fed to the computer; yet
the computer does not control the system. For the instrument evaluation
program this feature was not a handicap in that real-time data could be
obtained from the on-line computer and also processed from magnetic tapes
on a large computer.
The manual data inputs presented in Table 2.1 were also used to
facilitate an automatic calibration procedure in which calibration
2
concentrations in ug/m were entered directly on magnetic tape or input to
11
-------
the computer via a teletype keyboard. The calibration concentration data,
when combined with the instrument voltage output and the appropriate mode
switch status information, can be used to automatically compute a transfer
function which relates voltage output to pollutant concentration. This
equation then is used for processing the data. A stripchart recording and
a computer printout demonstrating the automatic calibration procedure are
presented in Figures 2.4 and 2.5, respectively.
2.3.2 On-Line Computer
One of the major requirements for this instrument evaluation program
was the need for immediate readout or processing of data in quasi-real
time to assist in early identification of weaknesses and/or degradation of
analyzer performance. To do this required the use of an on-line computer.
This feature vjas accomplished by installing a Hewlett Packard 2100A
computer in the mobile van and coupling it to the output of the data
acquisition system described in the preceding section.
The software program written for the HP-2100A computer can effectively
be used to process data for up to 20 instruments at one time, the limiting
factor here being the size of the memory core (8 k). Instrument names,
along with the channel number (on the data acquisition system) of the data
and status for that instrument are input quantities, which are read into
the computer at the beginning of program execution. This allows ease of
reconfiguring the system, adding or deleting instruments, changing cali-
bration equations coefficients, etc. The program monitors the instrument
outputs, as they are scanned every 5 minutes, and converts the voltage
reading obtained to a concentration in ppm or yg/m by use of an equation
of the form
c = A v + w
C A VOTJT + B
where
o
C = concentration in ppm or yg/m ,
A = slope of calibration curve,
VQITT = analyzer voltage output, and
B = intercept of calibration curve.
12
-------
Recorder Scale
0 12345678 9 10
J^ V
I
3
*
1
«
3
mm
0
MODE SWITCH
SETTING
| TIME
CHART SPEED 2 IN/HR
0.097 PPM
0.055 PPM
ZERO AIR
AMBIENT AIR
t
Figure 2.4: Automatic Calibration Procedure Showing Mode Switch Settings
13
-------
->-*
* 03
* «IO
* 0
*
0
*
* K
-»o
* O-l
* * s
^
:
(L
L C3
** 5
z
a
X
3
«4
-1
3
* O
U
*
*
*
*
*
* o
* -J
o
* U
1
^
ij
X
, ' .
«»«»»»*'««»»«»«»»»»»»»»»»
«««»»»»»«*»»o-»»»»»»e>««a>»
»«»»»«»»»^»»c*»»»»»»»»»»»
»«»«O>«»«*»C»»0>O>O>»O>»»«»»»»
»»»o>««»»»»»«o>»»»o>»«»»«»*
o-»«»»»»»»»*»e>»o>-««m«««n«
**************0.00000,0,0,°,°
UU«<4OOOU(BCBflOtJO
oeooomrmr«h>»««iM>>nNri«o>4HiMm
o o o
tJO-«<
z
« o
*
*
*
^
ui x <**imcMiafOcwNNMC4csi{>Jo>o>9>»o>0tetot»o>»
o x
ft O O>99^O*999O*O'9
* c
a.
k. «.
o
s x «s;;s;;;ssssSooo5oSo"^sa
- z
X 1 1 1 1 I 1 1
« r- « o o o o o «*««« o> » o. » o. » o- o- » » »
Ul
0
X
z a.
o
' _, «^«0000.0.«mK>»«iJ^5iU53o^^S^
* O ooooooooooooooaooooooooo
» i
4J
3
O
4J
H
^i
PH
C
o
cd
5
T-l
(0
o
(J
H
*J
4-1
J;
*
CM
a)
H
V\ tfvo o «4 *4M CMIO
o o »4 *4 M M IO **> ^ ^ Ifs M\ O o «- r4 CM M H> »*> » V tf> tf\
14
-------
The computed concentrations are retained and printed out at the end of each
hour along with the hourly average of the data for that instrument.
In addition to the output at the end of each hour, data may also be
output after each five-minute scan if requested by the operator. This
includes a list of:
(1) the voltages read by the data acquisition system,
(2) the concentrations computed, or
(3) the equations used to convert the voltage reading
to a concentration.
The program was also designed to assist the operator in performing
calibration procedures. It senses when an instrument is being calibrated
by means of the analyzer mode switch status channel. The program then
stores fqr each instrument in calibration (up to 10) all the valid cali-
bration data points (i.e., the measured instrument output voltage and the
corresponding calibration gas). Upon completion of the multipoint
qalibration as indicated by the appearance of the proper operational mode
switch code in the instrument status channel, the calibration data points
are fed into a regression subroutine; and a least-squares, best-fit
equation is automatically computed. This equation then replaces the one
already in storage and is used to process data from that point on. Both
the new and old regression equations can be printed out for comparison.
A flow chart of the program used in the evaluation program is
presented in Figure 2.6 and consists basically of two loops. The outer
loop is completed every five minutes (once for every scan of the data
acquisition system). The inner loop is scanned once for each instrument
every five minutes. Since it is a real-time program, it has no end but is
intended to run until manually halted.
Program operation starts with an initialization phase in which elements
of arrays are set to zero and data regarding system configuration are read
in. Then the program enters the loop. Here the first step is to print out
the system configuration data for checking and editing, if any exist. Then
the voltages are read from the data acquisition system and stored along
with the time of the reading. This reading can occur only during a scan,
15
-------
I
1
iimum 1
AMMTS J
|
KAD IV 1
CMFiaKATIOH 1
MTA I
J
1
1 MI NT OUT
OWFICMEB MTA
I
1 ,
KAO WCTA6ES 1
AWTIK |
M
I
PUT ««T LAKE
MME* I« AMM*
CAUSES I SIEM TO
»E OUTPUT.
1
1 COMPUTE 1
AVERAGES |
1
IMS1PTTOUT 1
NOIW.V DATA J
r |
Figure 2.6: Program Flow Chart
16
-------
and since the hardware is set only to scan once every five minutes, the
voltages are read every five minutes also. The program then enters the
inner loop where the individual instrument voltages are picked from the
voltage list for each instrument. First, from the array which contains
the data on how the system is configured, two numbers are obtained:
(1) the channel number of the data for that instrument, and
(2) the channel number of the status line for that instrument.
The program then enters a decision phase to determine what to do with
the data. This is accomplished by examining the individual instrument
status channel. The voltage on the status channel, which normally ranges
from 0-1 volt, is multiplied by 10 and rounded off to the nearest integer.
The value of this integer is tested to see what to do with the data point.
If the instrument was off line for any of the reasons coded on the status
switch, an excessively large number is entered in the slot in a 20 x 12
3
array corresponding to 20 instrument readings (to be in pg/m ) by
12 scans/hour. This will cause $$$ to be printed out when this array is
listed at the end of the hour. Also, the appropriate letter code is
inserted in the status array to indicate the reason why the instrument was
off line.
If the instrument was being calibrated, the data must be treated
differently. First, the actual concentration is computed and entered in
the DATA array and the code for either calibration, stabilization or
calibration data point is entered in the STATUS array. If the calibration
point was a valid point, another step must be taken as well. The voltage
measured and the standard value obtained from the operator are entered in
the appropriate array. Only two possibilities remain for the status.
The first is a special test for which no action is taken other than putting
computed data points in the DATA array and the appropriate Literal in the
STATUS array. The last possible status choice, measure, is the most often
used. This means a valid data point resulting from a measurement of
ambient air. For this data point the computed value is entered in the data
array, a blank entered in the STATUS array, and the value is added to a
summation location for computing an hourly average at the end of the hour.
17
-------
Also a test ±s made to see if this is the first data point after a
calibration; if so, the set of calibration data is fed to a regression
subroutine for the computation of the calibration equation. The new
equation then replaces the old and both are output.
This procedure is repeated for all instruments after each data scan
until the end of the hour. At this time a summary of all data points is
printed out along with the computed hourly averages. The DATA and STATUS
arrays are then reinitialized, and the program starts over. A typical
example of a five-minute, long-form data printout with regression analysis
and updated transfer equation is shown in Figure 2.7, and an hourly average
summary printout is shown in Figure 2.8.
18
-------
o o o oo r-- co
O O O
O O O O 0. rH
in r i r r
i O O rH 00 O VO vo
O O rH CO O O I--.
O O O O O rH rH
rH 4J
CO >W
a iH
P. B H
3 C O
O -H 4J
1
rH
o
C
CO
U 4J
td G
o eu
t-l
C >* C
030
H U iH
4-1 4J
CO B CO
3 O r4
CT* J-4 *£)
Q) >4-l -H
rH
M T3 CO
CU 01 CJ
m G
CO -H )-l
G to o
CO 4-1 -H
r4 *Q M
H o a.
i
CM
a
T3 CO
G
«0 T)
G
CO CO
rH
i) TJ
C CU
G 4-1
0) 3
CJ P*
o
'O CJ
cu
u ca
u cu
cu 3
rH rH
01 CO
en >
1
CO
O
o
es?s
o\
CO
m
CM
1 1
OOrHCOOrHOOrH
rHOvOCMHOOOO
rH O O O O VO H
II I I CO
i-l
I
a\ o o\ co rH CM m
c^ o co CM o ii i
rH O CM O O in rH
I I co
1
rO
r
o\ i
rHvOCT>vOOOOOvO
O rH 00 CO O O r^
rHOrHOOvOrH
CM
HI I I CO
O
O
m
&
S;
_i
FH vo
h
00
V)
M ^~
co m
^ SO
o
0
U
I M
tti
+ g
xg
* Pd
CO OH
CM g
O O
0\
« CO
VO M
in co
VO >H
CD r^
in CM
O\ CM
rH (^
o o
I CM
++aaaaaaaaaaaa
BBB
H
>-i
Pi
wmcoo^coooovo
§00 rH iI 00 O O O
o> o o in o vo I-H
g
XXcJCJCJOCJOOCJO
* * t3E3&»O!3£3C)C3
00 00
O O
1-xOO XJXX a HHH
in m
cMvooocM CM
\O VO "CO* -CO- -CO- VO VO H>H CO--CO--CO-I-HCM-CO- ^,_||
CO- -CO- -CO- -CO-
| M PM f!4 H
o
&
I
1 CO
rH
1
00
c/> O
W CJ
2
O Q
W nJ
P4 O
W
O C/3
O O
U O O O
CM cj
-------
00
S £i
TO 1-1
js n
w a
o
CO
CO CO
CO O
0) Px
4J
« to
o *j
H cu
T) O
s-a
H
"3
<0->
co-
--<
O -^co- co- -co- -co- -co- -co- -co- -co- -co- -co- <«-
<«-«/>«/>£/>/>«>«>-
CO-
CO-
CO-
CO-
CO-
CO-
cj
pq
.
OrHcMCMCSHeNCNr-l
CM r^ sr CN cs
«o ^ o «^ o
i-t CN CM tH CM
~
gS
^^
tj
oo oo o\
oo oo oo
CJ
B
o vo
OO
oo
M
O
oo
CN
(U
5
00
CO
*
M
*
H
Pd
OS
CO
CM
G»
^
HHHHHHHHUOCJCJ
a«»cMi-4r4cMi-tcNco
**** CO CO CO CO CO CO CO CO
Oi-li-ti-tl-lF-»i-ti-lr-l
o
omomoinomoiooiri
u i-(HCMCMcoco-*-*mm
t£ CO CO CO CO CO CO CO CO CO CO CO CO
-co-
co-
co-
CO-
CO-
20
-------
3,0 SELECTION OF INSTRUCTS
3.1 SELECTION CRITERIA
Operational requirements, definitions describing sensor specifications,
and sensor specifications applicable to the special needs of the RAPS
Program were prescribed in the section of RFP-DU-73-A016 describing instru-
mentation and are presented in Table 3.1. These specifications and
requirements provided the basic criteria for selection of analyzers and
are as follows: (1) accuracy, (2) maximum reliability, and (3) long-term
unattended operation. Other factors considered important in the selection
process were commercial availability, prior evaluation, proven field
reliability, and accessibility. It was readily apparent from the onset of
the program that all available brand name models of chemiluminescent ozone
monitors, for instance, could not be included in the evaluation program due
to monetary and time constraints. Therefore, subjective judgment was
utilized to choose a particular model for evaluation to determine if the
measurement principle, as represented by that particular model, could meet
the specifications required for the RAPS Program. Much consideration was
given to all available manufacturers of instrumentation within each cate-
gory prior to selection of an instrument or instruments for that category.
All available results obtained from previous instrument evaluation programs
were utilized in the selection process. In the case of ozone instrumenta-
tion one instrument was considered adequate to conduct the evaluation;
while for the other pollutants numerous analyzers available at the time of
the evaluation were considered and tested. For example, several chemi-
luminescent NO- analyzers were tested. This judgment was based on prior
knowledge of proven field reliability of the measurement principle and, to
some extent on the reliability of the instrument chosen for the evaluation.
This should not be interpreted to mean, however, that other instruments
not included in the evaluation program within each category would not or
could not meet the required specifications. The instrumentation and
measurement methods evaluated in this program are summarized in Table 3.2.
The principle of operation of these instruments and measurement methods
are described in detail in Section 4.0.
21
-------
CO
jgj
Q
M
H
3
£
M
O
H
PH
w
04
O
CO
55
W
CO
*
i-H
*
co
0)
rH
n)
H
H CO
d «)
O T)
H >-'
U
18 TJ
r4 O
> o
O T-I
nJ co
ri -H
3 0
U V
U U
^^ OH
^
4-1
H*
> c?
v4 S3
4J (^
H PH
d
(U
CA
0) ,-N
MX
d CM
^
4-1
d , a
H W M
O 4-1 0)
4-i a) u
o a co
J3 0
co
+ 1
X
CNJ
*4" 1
&*?
fsj
+ 1
0
o
rH
d
o
in o
CM rH
44
OfN
**
-i m
O 0
d
o
H
4J
cd
(4
T-I
g
M
a
cd
fa
1
rj
0
8
0
^
CS
4*1
gs^
CO
+l
»*
CN|
^ 1
gs}»
cs
+l
o
o
d
in
CM
o
i-H
4
d
o
H
4-1
cd
N
H
d
o
H
d
o
0
o
H
«
rH
ON
CO
rH
1
U)
CM
M
00
vO
CM
1
CM
O
co
CO
vD
cn
rH
1
CO
O
r^
CM
CM
rH
1
O
o
oo
00
rH
1
CM
g
m
^
rH
rH
1
O
-a-
in
vO
i
-!
B
22
-------
TABLE 3.2. INSTRUMENTATION AND MEASUREMENT METHODS EVALUATED
Pollutant
Instrument
Principle of Operation
Ozone
Nitric Oxide/
Nitrogen Dioxide
Sulfur Dioxide
Hydrocarbons
Carbon Monoxide
Total Sulfur/Hydrogen
Sulfide/Sulfur Dioxide
Bendix Model 8002
Dasibi Model 1003-AH
Bendix Model 8101-B
Meloy Model 520
Thermo Electron Model 14
Rem Model 642
Bendix Model 8300
Meloy Model SA-185R
Beckman Model 6800
Bendix Model 8201
Beckman Model 6800
Andros Model 7000
Bendix Model 8501-5FA
Tracer Model 270-HA
Bendix Model 8700
Chemiluminescence
UV-Absorption
Chemiluminescence
Chemiluminescence
Chemiluminescence
Chemiluminescence
Flame Photometry
Flame Photometry
GC-Flame lonization
GC-Flame lonization
GC-Flame lonization
DI-Fluorescence
Non-Dispersive Infrared
GC-Flame Photometry
GC-Flame Photometry
23
-------
3.2 TEST PLAN
Operational requirements expressed in the specifications for each
sensor refer to the characteristics of the sensor as a unit from the
inlet sample port of the instrument to the actual output of the sensor.
The original test plan for the evaluation program is presented in
Figure 3.1 and was divided into two parts. Briefly, Part I of the test
plan was concerned with identification of critical points in the measure-
ment process and the sensor's ability to meet the specifications presented
in Table 3.1, Section 3.1. Test procedures described in detail in
Section 5.0 and computer programming techniques developed during the
performance of Contract CPA 70-101, "Field Evaluation of New Air Pollution
Monitoring Systems," were utilized to determine the following performance
characteristics for each sensor: lag, rise and fall time, sensitivity,
precision, zero and span drift, linearity, operational period, etc.
Part II of the evaluation program involved recommendations of engineering
modifications to correct deficiencies if a given sensor failed to meet any
specification.
Outputs of the test plan considered important were documentation of
performance characteristics; modification, if necessary, of sensor to
improve performance; retesting of sensor after modification to determine
compliance with specifications; and rejection of sensor and/or documen-
tation of performance characteristics if sensor met all requirements.
Early identification of weaknesses or problems was considered imperative
to accomplish the program objectives. To do this required the use of
on-line data processing in quasi-real time. This feature of the evaluation
program was accomplished via the use of an on-line computer. Another
important facet of this program was to document in detail all modifications
and/or recommendations to EPA or the manufacturer of a particular instrument
to improve performance.
24
-------
Analysis
sor and
g System
1
Detailed Analysis of
Calibration System
and Procedures
Selection of
Sensors
Detailed ;
of Manuf;
Specif ical
Mamie
1
Identify Critical
Points in
Measurement Process
Identify Points
Requiring Periodic
Checks and
Maintenance
I
Submit Sensor to Standard
Test Procedures to Determine
Compliance with Specifications
Document Performance
Characteristics, if
Sensor Meets All
Specifications
Modify Sensor or
Component, if
Sensor Fails to
Meet Specifications
Document Engineering
Modification
Retest Sensor to
Determine
Compliance
JL
Documen^ Performance
Characteristics, if
Sensor Meets All
Specifications
Reject Sensor
and Notify
Project Officer
Documentation
Draft of Final Report
Documentation
Final Report
Figure 3.1: Test Plan Flowchart
25
-------
4,0 GAS WLYZERS
4.1 BENDIX MODEL 8002 GAS PHASE CHEMILUMINESCENT OZONE ANALYZER
4.1.1 Instrument Description
The theory of operation of the Bendix Model 8002 ozone analyzer is
based on the gas phase chemiluminescent reaction of ozone with ethylene .
Ozonized or sample air (1 £/min) and ethylene (25 cc/min) are mixed in a
shallow cavity closely coupled to the cathode face of a photomultiplier
tube. The photomultiplier tube current resulting from the detection of
the chemiluminescence of the flameless reaction of ethylene gas with ozone
is directly proportional to the concentration of ozone in the air sample.
A block diagram of the chemiluminescent ozone analyzer is shown in
Figure 4.1. The ethylene reaction has been reported in the literature to
be specific for ozone, and no known components of the lower troposphere
other than ozone have been observed to give chemiluminescence with the
reactive gas .
The physical characteristics and other descriptors of the Bendix ozone
monitor are as follows:
(1) overall dimensions: 16.5 in (42 cm) width,
8.5 in (2.15 cm) height,
18.5 in (47 cm) depth;
(2) weight: 45 pounds (20.4 Kg)
(3) power requirements: 350 watts at 105 to 125 volts,
60 Hz;
(4) measurement ranges: 0.01, 0.02, 0.05, 0.1, 0.2, 0.5,
and 1.0 ppm full scale
(19.60-1960 yg/m3);
(5) mode selector: ambient, zero, internal calibrate
(fixed ozone concentration)
(6) adjustments: zero and span adjust potentiometers;
(7) outputs: 0-10 mV recorder; 0-1 VDC
26
-------
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: 0.001 ppm
(1.96 yg/m3);
(b) linearity: + 0.5%;
(c) noise: + 1% on 0 to 0.1 ppm (0-196 yg/m ) range;
(d) measurement repeatability: + 2% from mean value
on the 0-0.1 ppm
(0-196 ug/m3) range;
(e) accuracy: not specified
(f) zero drift: + 1% per day, + 2% per three days
(g) span drift: + 1% per day, + 2% per three days
(h) operational period: 7 days or more unattended.
(9) required accessories: cylinder of ethylene gas, two-stage
regulator, stripchart recorder
The gas flow system of the Bendix analyzer is composed of two parts,
ethylene flow and sample flow. A flow diagram of the system is shown in
Figure 4.2. Ethylene is provided by an external cylinder and two-stage
regulator, and the ethylene flow rate is maintained by a pressure regulator
and capillary designed to give a particular flow at a specified input
pressure. A solenoid valve is provided for safety and is designed to shut
off ethylene flow whenever AC-line voltage to the instrument is interrupted.
Sample flow rate is maintained by a vacuum pump, needle valve, and flow
meter arrangement. The needle valve is used to adjust the flow rate from
the reaction chamber, and the flow meter is utilized to monitor this rate.
Particulates greater than 5 microns in diameter are removed from the ambient
air sample via a Teflon filter and prevented from entering the reaction
chamber.
The detector cell of the analyzer consists of a reaction chamber
located within the end plate block of the detector cell and a photo-
multiplier assembly. The assembly consists of a photomultiplier tube,
27
-------
CLEAN
AIR IN
AIR
SAMPLE
IN
SELECTOR SOLENOID
ETHYLENE
IN
1
OZONE
GENERATOR
AC CURRENT
REGULATOR
|
REACTION
CHAMBER
PHOTO
MULTIPLIER
ASSEMBLY
I
DETECTOR
CELL
PHOTO
MULTIPLIER
TEMPERATURE
CONTROL
ALARM AND
VALVE CONTROL
SIGNAL TO
RECORDER
ZERO
ADJUST
CALIBRATE
ADJUST
±110V
POWER
SUPPLY
03 CONCENTRATION
Figure 4.1:
Ozone Monitor Simplified Block Diagram
28
-------
£ °
°9
as
ij O
in w
u
z'
<
ui tr
z ui
o z
N ui
i
u u.
M
cd
H
Q
t-l
O
o
s
ai
c
o
N
O
CM
3
60
Z
ui
0.
to
z
IT
5
UI
Z
ui
I
H
ui
<(
$
UI
29
-------
a high voltage power supply, thermoelectric coolers which maintain the
cathode of the tube at 5°C, and a housing to enclose the three items.
The basic function of the photomultiplier assembly is to convert the
low-level photons resulting from the flameless reaction between ethylene
and ozone to electrical energy and to amplify the electrical signal for
further processing through the electrometer amplifier and recorder output/
meter display circuitry. A schematic diagram of the detector cell is
shown in Figure 4.3.
4.1.2 Operational Summary
The gas phase chemiluminescent analyzer was installed in the mobile
van on December 18, 1972 and operated continuously until August 2, 1973.
During this period of time both short-term and long-term tests were
performed on the analyzer in accordance with the test plan requirements.
The frequency of performance monitoring during these tests was dependent
on the particular requirements of the evaluation procedure being applied.
Calibration frequency for short-term drift was on a daily basis, and that
for longer term drift studies was on a weekly or biweekly basis.
During the seven-month period in which the analyzer was operated and
subjected to the performance evaluation tests, two failures occurred. The
first failure involved the detector assembly and resulted from a power
failure that lasted approximately two hours. Upon resumption of power, the
analyzer failed to respond. The failure was attributed to failure of the
detector power supply and required replacement of the detector cell. The
other problem involved the inability of the analyzer to maintain a constant
sample air-ethylene flow rate. This problem was traced finally to a baffle
in the metal bellows vacuum pump and corrected. With the exception of
these two problems, the analyzer required a minimum amount of maintenance.
Routine maintenance involved replacement of the Teflon prefilter element
on the sample inlet line on a monthly basis and replacement of the ethylene
cylinder at six-month intervals.
30
-------
OLAM WINDOW FOB LIGHT EXPOCUHC
TO PHOTOMULTIPLIER TUM
AMPLI
INO PLATE BLOCK
PHOTOMULTIH.lt« AMEMILY
PC IOAHD
fXHAUtT
OLAM
WINDOW
PHOTOMNtlTIVE
COATING
Figure 4.3: Detector Cell
31
-------
4.2 DASIBI MODEL 1003-AH ULTRAVIOLET OZONE ANALYZER
4.2.1 Instrument Description
The operation of the Dasibi Model 1003-AH analyzer is based upon the
measurement of the absorption of ultraviolet (UV) light by ozone within a
sample volume of air. Ultraviolet light is generated by a low-pressure
mercury vapor lamp, which emits its strongest line at 2536.52 angstroms.
This wavelength coincides with the region of maximum light absorption of
ozone. By measurement of the amount of light absorbed by the sampled gas,
the ozone concentration can be determined. The ozone concentration can be
calculated from the following expression:
, i, -act
lfno = e
where I, = photocurrent of absorbed light,
I = photocurrent for zero concentration of ozone,
-4 -1
a = absorption coefficient = 2.74 x 10 cm ,
c = concentration of ozone in ppm by volume, and
t = length of absorption chamber in centimeters.
The manner by which the above equation is mechanized in the Dasibi requires
a variation of the expression as follows:
-r i -act
I - I, = 1 - e
o r
Basic components of the Dasibi analyzer are a gas flow system, UV
source, sample chamber, absorption and reference detectors, digital elec-
trometers, integration control logic and computation logic systems.
Figure 4.4 is a simplified block diagram of the Dasibi analyzer.
Sample gas enters the inlet and is divided into two gas streams. A. scrubber
is utilized to remove ozone from the air sample in one line. Both sample and
air minus ozone alternately enter the sample chamber through an electrically
operated three-way solenoid valve. The absorption chamber has two windows
32
-------
CO
rt
W
OJ »-l
rH Q)
11
0)
O
t-l
3
O
CO
00
H
Q
cfl 3
4-1 p
CO
Q 3
O
O
H
M
O
§
rl
4-1
0] CO
I-
O
O
t-l
4-1
§
u
§
H (X
4J O
cfl
IH
00
0)
4-J
n
c
a)
4-1
CO
c
0)
Pi
O
N
O
c
O
O
CO
1-1
00
cd
H
Q
01
s
oo
-a-
~*
00
H
H C
< H
33
-------
at opposite ends through which UV light passes, with the UV source being
located at one end and the absorption detector at the other. Measurement
begins when a signal from the electrically operated valve indicates that
sample air minus ozone is flowing through the absorption chamber. At this
time, the absorption detector starts to measure photons and integrates the
result. At the same time the reference detectors starts counting to a
preset value. The sequence is repeated again for the unfiltered gas. The
difference between the integrated results for the filtered and unfiltered
gas is computed and displayed digitally as the output concentration.
The physical characteristics and other descriptors of the Dasibi
UV ozone analyzer are as follows:
(1) overall dimensions: 15 in (38.1 cm) width,
5.25 in (13.3 cm) height,
22 in (55.9 cm) depth;
(2) weight: 31 pounds (14.1 Kg);
(3) power requirements: 75 watts at 110 V, 60 Hz;
(4) measurement ranges: 0.003 to 20 ppm
(5.9 to 39200 yg/m3);
(5) mode selectors: ambient, zero, calibrate;
(6) outputs:
(a) digital display - 0.003 to 20.0 ppm
(5.9 to 39,200 yg/m3)
(b) analog - 0 - 0.999 V (Std)
(c) BCD (optional) - 8-4-2-1, Standard TTL;
(7) Stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: 0.003 ppm
(5.9 yg/m3);
(b) linearity: 1% (0.003 to 1.0 ppm)
(5.9 to 1960 yg/m3);
34
-------
(c) noise: not specified;
(d) precision + 1%;
(e) accuracy: + 3% (based on Beer's Law);
(f) zero drift: Digital and BCD output - (zero)
analog output - 0.5% per day;
1% per three days;
(g) span drift: Digital and BCD output - (zero)
analog output - 0.5% per day;
1% per three days;
(h) operational period: not specified;
(8) required accessories: none; stripchart recorder.
The gas flow system of the Dasibi analyzer is composed of two flow
paths. Sample air flows through one flow path directly to the sample
chamber and is alternately diverted through a scrubber which removes ozone
from the sample stream. The time sequence is controlled by a timer, with
switching occurring at 10 second intervals. Sample flow rate is maintained
at approximately 2 liters per minute by a blower.
The absorption chamber consists of a folded tube approximately
0.4 inch (1.0 cm) in diameter and 30 inches (77 cm) in length with inlet
and outlet gas ports. The chamber has two windows which are optically
o
transparent at the 2537 A emission. A solid-state photodiode detector is
positioned adjacent to the outlet port and senses the UV radiation through
the absorption chamber. At the light source end of the chamber, a second
photodiode detector, which serves as a reference, senses the portion of
light from the source. The electronic signals from the photodiode
detectors are processed through the digital electrometers and computation
logic systems.
4.2.2 Operational Summary
The Dasibi ozone analyzer was installed in the mobile van on July 12,
1973 and operated for a week's period of time. Due to time constraints and
the unavailability of the analyzer from EPA for longer periods of time,
35
-------
a limited number of performance evaluation tests were performed. The major
efforts were concerned with interference tests, minimum detectable concen-
tration, and comparison of ambient data with that from a chemiluminescent
ozone analyzer. No failures occurred during this period of time and no
maintenance was required.
4.3 BENDIX MODEL 8101-B CHEMILUMINESCENT NO-NOX-N02 ANALYZER
4.3.1 Instrument Description
The principle of operation of the Bendix Model 8101-B analyzer is
based on the chemiluminescent gas phase reaction of nitric oxide (NO) and
7 8
ozone (0_) ' :
NO + 03 f NO* + 02
N02 > N02 + hv.
Basic components of the analyzer are a reaction chamber, photomultiplier
tube with a thermoelectric cooler, filter, electric discharge ozone
generator, carbon-coated and heated converter, flow system, vacuum pump,
electronic circuitry, and amplifier systems. Sample air (= 130 cc/min) and
ozone (a 50 cc/min) are mixed prior to entering the reaction chamber which
is closely coupled to the cathode face of a photomultiplier tube. Detection
of the light emission from the reaction by the photomultiplier tube yields
a signal with an amplitude proportional to the concentration of NO in the
sample stream. The concentration of nitrogen dioxide (N0_) in ambient air
can be indirectly determined after quantitative conversion to nitric oxide
by this chemiluminescent reaction. The NO concentration in the ambient air
sample is first determined; then the sample air is diverted through a carbon
converter (heated to 285°C) where NO- is reduced quantitatively to NO. The
detector alternately measures NO and total oxides of nitrogen (NO + N09
reduced to NO) 30 seconds in each mode. By electronic subtraction of the
amplitude of the NO signal from that of the NO signal, the concentration
X
of N0» in the air sample is determined. Total time for the measurement
36
-------
cycle for the Bendix analyzer is one minute. Thus, 60 NO- measurement
values are available for computing an hourly average. A block diagram of
the chemiluminescent NO-NO -NO,, analyzer is shown in Figure 4.5. Measure-
X 2.
ment of N0_ via the chemiluminescent reaction of NO with 0~ is not subject
to interference from any of the common air pollutants found in the ambient
a
air, such as C02» CO, C H^, NH , S02, and H~0 ; but recent evidence
indicates a possible response to PAN (peroxy acetyl nitrate).
The physical characteristics and other descriptors of the Bendix
NO-NO -N07 analyzer are as follows:
X £,
(1) overall dimensions: 16.5 in (42.0 cm) width;
8.5 in (21.5 cm) height; and
17 in (43.0 cm) depth;
(2) weight: 60 pounds (27.2 Kg);
(3) power requirements: 350 watts at 105 to 125 volts, 60 Hz;
(4) measurement ranges: 0.5, 1.0, 2.0, 5.0 ppm full scale
(940-9400 yg/m3);
(5) mode selectors: ambient, zero, span, NO mode only,
NO mode only, N0-N00-N0 mode;
X £. X
(6) adjustment: zero and span adjustment potentiometers for
NO, NO and N0_ outputs;
X £,
(7) outputs: 0-10 mV recorder; 0-1 VDC;
(8) stated performance specifications (manufacturer)
3
(a) minimum detectable sensitivity: 0.005 ppm (9.4 yg/m );
(b) linearity: + 0.5% full scale;
(c) noise: 0.5% full scale;
3
(d) precision: + 0.01 ppm (18.8 yg/m ) from
0.005 to 2.0 ppm (9.4-3760 yg/ni )
measured at integrator output;
(e) accuracy: + 0.01 ppm (18.8 yg/m ) or + 2%
whichever is greater on the 0 to 2.0 ppm
(0-3760 yg/m3) scale;
37
-------
SMUT-OFF
SOLENOID
VALVE
MIOXVBEN)
PHOTOMULTIPUER
TIMPERATURE
CONTROL
SIGNAL
INTEGRATOR
t
4 RESET
^
1
COMPARATOR
R£3tlk
4
^
TIMER
*I5 -15
t t
*«-*>«
GATE
NO SPAN
ADJUSTMENT
CONTROL
COMMON
No ZENO ADJUSTMENT CONTROL
NO. SPAN
AtMUSTMENT
CONTHOL
No. ZERO ADJUSTMENT CONTROL
Nox SPAN
ADJUSTMENT
CONTROL
*is
NO. ZERO
ADJUSTMENT
CONTROL
SI9NALS
tO-IV)
TO RECORDERS
Figure 4.5: Oxides of Nitrogen Analyzer
Simplified Block Diagram
38
-------
(f) zero drift: less than + 1% per 24 hours;
less than + 2% per 3 days;
(g) span drift: less than + 2% per 24 hours;
less than + 2% per 3 days;
(h) operational period: 7 days or more unattended;
(9) required accessories: cylinder of oxygen, two-stage
regulator, stripchart recorder.
The gas flow system of the Bendix analyzer is composed of two parts
(i.e., oxygen/ozone flow system and sample flow system). A flow diagram
of the system is shown in Figure 4.6. The ozone generator used in this
analyzer is a gas discharge type and produces approximately 1% of ozone
at a flow rate of 50 cc/min of oxygen. Oxygen is provided by an external
cylinder thru a two-stage regulator. The oxygen flow rate is maintained
by use of a pressure regulator and capillary designed to give approxi-
mately 50 cc/min flow at a specified input pressure. A solenoid valve is
provided for safety and designed to shut off oxygen flow anytime AC-line
voltage to the instrument is interrupted. Sample flow rate is maintained
at approximately 130-150 cc/min via use of a glass capillary which acts
as a flow restrictor (critical orifice) and a two-stage diaphragm vacuum
pump. Particulates greater than 5 microns in diameter are removed from
the ambient air sample via a Teflon filter and prevented from entering the
inlet lines and plugging the glass capillary.
The detector cell of the analyzer consists of a reaction chamber,
optical filter, photomultiplier tube, high voltage power supply, thermo-
electric coolers which maintain the cathode of the tube at 5°C, and a
detector cell housing. The basic function of the photomultiplier assembly
is to convert the low level photons resulting from the flameless reaction
to electrical energy and amplify the electrical signal for further
processing through the electrometer amplifier, integrator and memory
circuitry. A schematic diagram of the detector cell is shown in
Figures 4.7 and 4.8.
39
-------
HE AW
PANEL
OXYGEN
IN
TO
EXTERNAL
VENT
EVACUATION PUMP
Figure 4.6: Oxides of Nitrogen Analyzer Flow Diagram
40
-------
6
n)
!-i
00
n)
r-t
Q
0)
u
t-l
o
4J
o
QJ
4-1
0)
Q
r-^
vf
60
OU.
-------
c
H
CO
3
o
BC
u
0)
01
Q
00
H
-------
The catalytic converter used in the Bendix analyzer to reduce NO. to
NO contains two tubes filled with ultra pure carbon. The catalyst block
is heated to a temperature of 285°C to control the efficiency of the
conversion process. Conversion efficiency of 98% has been reported.
4.3.2 Operational Summary
The chemiluminescent analyzer was installed in the mobile van on
January 19, 1973 and operated continuously until June 13, 1973. During
this period of time both short-term and long-term tests were performed
on the analyzer in accordance with the test plan requirements. The
frequency of performance monitoring during these tests was dependent on
the particular requirements of the evaluation procedure being applied.
Calibration frequency for short-term drift was on a daily basis, while for
longer term drift studies the calibration frequency was on a weekly or
biweekly basis.
During the six month period in which the analyzer was operated and
subjected to the performance evaluation tests, two failures occurred.
Both failures involved the two stage vacuum pumps used to maintain cell
pressure within certain limits (i.e., p < 7 inches (17.8 cm) absolute
pressure) and the proper flow rate at approximately 130-150 cc/min through
the instrument. The response of the Bendix analyzer has been shown to be
highly dependent on cell pressure with a minimum vacuum of 58.5 cm of Eg
required for proper operation. Minor changes in the cell pressure will
affect both the sensitivity of the measurement and calibration. Documen-
tation of these results are presented in Section 7.0. A minimum amount of
maintenance was required during this period of time. Routine maintenance
involved replacement of the Teflon prefilter element on the sample inlet
line and cleaning the residue from the ozonator electrode on a monthly
basis. Accumulation or buildup of residue on the ozonator electrode
contributed to propagation of radio frequency interference which adversely
affects the operation of the memory and signal processing circuitry.
Adequate shielding would have eliminated this effect.
-------
4.4 MELOY MODEL 520 CHEMILUMINESCENT NO-NOX-N02 ANALYZER
4.4.1 Instrument Description
Operation of the Meloy NO-NO -NO- analyzer is based on the flameless
X £,
chemiluminescent reaction between NO and 0_:
NO + 0- -> N0_ + 02 -> N02 + hv .
The intensity of the radiation of the electronically excited N0~ molecules
as they revert to lower energy states is directly proportional to the
concentration of NO molecules involved in the reaction.
Basic components of the analyzer are a pneumatic sampling system,
catalytic converter, high-voltage ozone generator, detector-chopper unit,
flow controller, and associated electronics. Sample air and ozone
(a 1000-3500 ppm) are mixed and chemically react within the detector
chamber, which is coupled to the cathode face of a photomultiplier tube.
Detection of the light emission from the reaction yields a signal whose
amplitude is proportional to the concentration of NO in the sample stream.
The concentration of N02 in the ambient air sample is then determined after
quantitative conversion to NO, when the sample air is diverted through a
catalytic converter. The detector alternately measures NO and total oxides
of nitrogen (NO 4- N02 reduced to NO), 30 seconds in each mode. By elec-
tronic subtraction of the NO signal from the NO signal, the concentration
X
of N02 in the air sample is determined. Total time for the measurement
cycle is one minute. A block diagram of the chemiluminescent NO-NO -N09
X £.
analyzer is shown in Figure 4.9. Measurement of NO- via the reaction of NO
with 0_ is not subject to interference from any of the common air pollutants
J 9
in ambient air, such as C02, CO, C2H , NH , S02, and H20 ; but recent
evidence indicates a possible response to PAN (peroxy acetyl nitrate).
The physical characteristics and other descriptors of the Meloy
NO-NO -NO, analyzer are as follows:
X £
(a) overall dimensions: 17 in (43 cm) width,
12.5 in (31.8 cm) height,
20.75 in (52.7 cm) depth;
44
-------
(U
N
cfl
-S
tO
M
00
n)
O
o
l-<
PQ
CTi
-------
(2) weight: 60 pounds (27.2 Kg);
(3) power requirements: 100 watts at 105 to 125 volts,
60 Hz;
(4) measurement ranges: 0.5, 1.0, 2.0, 5.0 ppm full scale
(940 - 9400 yg/m3);
(5) Mode selectors: ambient, zero, span;
(6) adjustments: zero and span potentiometers for NO, NO ,
and NO- outputs;
(7) outputs: 0-100 mV recorder; 0-1 VDC;
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: 0.015 ppm
(28.2 yg/m3);
(b) linearity: + 2%;
(940 yg/m ) range;
(c) noise: 0.8% full scale of 0.5 ppm
(940 yg/m-
(d) precision: + 2%;
(e) accuracy: 0.01 ppm (18.8 yg/m );
(f) zero drift: + 3% per 7 days;
(g) span drift: + 2% per 7 days;
(h) operational period: 7 or more days unattended;
(9) required accessories: stripchart recorder.
The gas flow system of the Meloy analyzer is composed of three parts:
ambient air/ozone flow system and NO and NO sample air flow systems. A
X
flow diagram is shown in Figure 4.10. Ozone concentrations (* 1000-3500 ppm)
are generated via a high voltage ozone generator using ambient air scrubbed
through a chemical filter. Separate NO and NO sample streams are provided
X
through identical flow systems, except that in the case of NO the air
X
stream passes through a catalytic converter. Constant flow rates are
46
-------
PS.
o PS
H w
U PL, H
W PM M
H O Z
W ffi 5
PS
W
M
M
K
O
NI
O
I
W
////I
W W
2: H
O OH
6C
£0
H
O
0)
s
o
(U
M
S
60
pp
47
-------
maintained through the three inlets (ozone, NO sample air, and NO sample
.A
air) by use of a flow controller module and a vacuum regulator. The flow
controller is comprised of a temperature-controlled cartridge which
presents a constant pneumatic impedance (stainless steel capillaries) to
the NO, NO , and ozone carrier streams over a 5 to 40°C ambient temperature
change. A vacuum regulator and pump are utilized to maintain the pneu-
matic system at a constant pressure. Particulates are prevented from
entering the inlet lines by a 7y particle filter.
The detector/chopper unit of the analyzer consists of a detector
chamber, light chopper, and a photomultiplier tube. The functions of this
unit are to provide a chamber in which the reaction can take place and to
modulate and amplify the chemiluminescent signal. The detector arrangement
has gas entry ports on one end and an optical window/filter arrangement and
photomultiplier tube coupled to the other end of the detector chamber. The
detector chamber and photomultiplier tube are separated by a light chopper,
which modulates the dc chemiluminescent signal and transforms it into a
pulsed dc signal.
The catalytic converter used in the Meloy analyzer to reduce NO- to
NO contained copper sulfate and was maintained at a temperature of 300°C.
Conversion efficiency of 95 to 98% was reported by Meloy.
4.4.2 Operational Summary
The chemiluminescent analyzer was installed in the mobile van on
January 19, 1973 by the service representatives of Meloy and was operated
continuously through July 16, 1973. During this period of time both
short-term and long-term tests were performed on the analyzer in accordance
with the test plan requirements. The frequency of performance monitoring
during these tests was dependent on the particular requirements of the
evaluation procedure being applied. Calibration frequency for short-term
drift was on a daily basis, and that for longer term drift studies was on
a weekly or biweekly basis.
During the six-month period in which the analyzer was operated and
subjected to the performance evaluation tests, several failures occurred.
48
-------
One failure involved conversion of N0» to NO by the heated stainless steel
capillary used to regulate sample air flow rate and caused N02 interference
with the NO measurement. This was considered a design deficiency and
reported to Meloy. In addition, catalyst material was found deposited in
the cell capillary and thereby reduced flow. The other failures, four in
all, resulted from an underdesigned ozonator transformer. Insufficient
insulation on the secondary winding resulted in a voltage breakdown and
shorted out the transformer. The recommendation to upgrade the transformer
was not heeded by Meloy and the defective part was simply replaced with the
same model transformer. Routine maintenance involved replacement of the
Teflon prefilter element on the sample inlet line, charcoal in the zero air
scrubber, and silica gel and molecular sieve in the ozonator air cleanup
scrubber on a monthly basis.
4.5 THERMO ELECTRON MODEL 14 CHEMILUMINESCENT NO-NO -NO- ANALYZER
A C-
4.5.1 Instrument Description
Operation of the Thermo Electron NO-NO -N09 analyzer is based on the
X ^
flameless chemiluminescent reaction between NO and 0,:
NO + 03 -* N02 + 02 -> N02 + hv .
Light emission resulting from the reversion of electronically excited NO-
molecules to their ground state is directly proportional to the concen-
tration of NO molecules involved in the reaction.
Basic components of the analyzer are a gas flow system, catalytic
converter, reaction chamber, high-voltage ozone generator, thermoelectric-
cooled photomultiplier tube, and associated electronic circuitry. To
measure NO concentrations, the gas sample is mixed with ozone in the
reaction chamber, and the resulting chemiluminescence is monitored through
an optical filter by a high sensitivity photomultiplier tube. To measure
NO concentrations (NO + N02 reduced to NO), the sample gas flow is diverted
49
-------
through a catalytic converter, with the resulting chemiluminescent response
being linearly proportional to the NO concentration entering the converter.
Nitric oxide and NO concentrations are alternately measured and stored in
X
electronic circuits which automatically subtract the NO concentration from
the NO concentration and display the N00 value continuously. Measurement
X 2*
of N02 via the reaction of NO with 0- is not subject to interference from
any of the common air pollutants found in ambient air, such as C09, CO,
9 z
C»H;, NH_, S0_, and H^O ; but recent evidence indicates a possible response
to PAN (peroxy acetyl nitrate).
The physical characteristics and other descriptors of the Thermo
Electron NO-NO -NO- analyzer are as follows:
Ji £,
(1) overall dimensions: 19 in (48.3 cm) width,
17 in (43.2 cm) height,
18 in (45.7 cm) depth;
(2) weight: 80 pounds (36.4 Kg);
(3) power requirements: 500 watts at 115 V, 60 Hz;
(4) measurement ranges: 0.05, 0.1, 0.25, 0.5, 1, 2.5. 5
and 10 ppm (94 to 18,800 yg/m );
(5) mode selectors: automatic mode (NO-NO -N09),
A £,
manual mode (NO or NO );
X.
(6) adjustments: zero (background) and calibrate potentiometers;
(7) outputs: 0-10 V (NO, NO , NO, outputs);
X. £.
(8) stated performance specifications (manufacturer);
(a) minimum detectable sensitivity: not specified;
(b) linearity: + 1% full scale;
(c) noise: not specified;
(d) precision: precision of span setting better than 1%;
(e) accuracy: not specified;
(f) zero drift: < 1% in 24 hours;
50
-------
(g) span stability: + 1% in 24 hours,
+ 3% in 7 days;
(h) operational period: not specified;
(9) required accessories: stripchart recorder.
The gas flow system of the Thermo Electron analyzer is composed of two
separate systems, sample air and NO/NO flow systems. A flow diagram is
X
shown in Figure 4.11. Ozone concentrations are generated by an electric
discharge ozone generator using ambient air scrubbed through an air dryer.
Sample air containing NO and NO enter the reaction chamber through either
X
of two loops with the direction of flow controlled by a solenoid valve.
In the NO mode of operation, sample air is diverted through the catalytic
X
converter. Sample air and ozone flow rates are maintained constant by use
of capillary restrictors, a bypass flow system, and two metal-bellows
vacuum pumps. Constant flow to the reaction chamber is maintained by
providing a differential pressure across a 20-mil capillary of approximately
22 in Hg (55.9 cm Hg). Stability of the instrument calibration is dependent
on maintaining constant flow rates of sample and ozone and maintaining an
absolute pressure of 8 in Hg (20.3 cm Hg) in the reaction chamber.
Particulates are prevented from entering the instrument by a particulate
filter.
The detector unit consists of a reaction chamber, optical filter,
thermoelectric-cooled photomultiplier tube, electrometer, amplifier, memory,
and comparator circuitry.
Three catalytic converters are available for use in the Thermo Electron
analyzer. The recommended operating temperatures for the various converter
options available are as follows:
Model 300 - Molybdenum Converter - 400°C
Model 325 - Carbon Converter - 300°C
Model 350 - Stainless Steel Converter - 400°C.
The standard converter supplied with the Model 14, unless otherwise speci-
fied, is the Model 350, Stainless Steel. Conversion efficiency of 98%
has been reported.
51
-------
M
0)
N
x
CM
i
(U
XI
I
o
§
0
H
52
-------
4.5.2 Operational Summary
The chemiluminescent analyzer was installed in the mobile van on
May 8, 1973 and operated continuously through July 15, 1973. During this
period of time both short-term and long-term tests were performed on the
analyzer in accordance with the test plan requirements. The frequency
of performance monitoring during these tests was dependent on the
particular requirements of the evaluation procedure being applied. Cali-
bration frequency for short-term drift was on a daily basis, and that for
longer term drift studies was on a weekly or biweekly basis.
During the two-month period in which the analyzer was operated and
subjected to the performance evaluation tests, one failure occurred which
resulted from failure of the converter heater. The converter heater was
promptly replaced and the evaluation continued. Routine maintenance
consisted of replacement of the drierite in the air dryer for the ozone
generator flow system.
4.6 REM MODEL 642 CHEMILUMINESCENT NO-NOX-N02 ANALYZER
4.6.1 Instrument Description
Operation of the Rem NO-NO -NO analyzer is based on the flameless
X £,
chemilunrfnescent reaction between NO and 0~:
NO + 0- -> N02 + 0- -» N02 + hv.
*
The intensity of the radiation of the electronically excited N0~ molecules
as they revert to lower energy states is directly proportional to the
concentration of NO molecules involved in the reaction.
Basic components of the analyzer are a reaction chamber, photo-
multiplier tube with thermoelectric cooler, filter, catalytic converter,
gas flow system, electric discharge ozone generator, and electronic
circuitry and amplifier systems as shown in Figure 4.12. Sample air and
ozone are mixed and chemically react within the detector chamber, which is
53
-------
w
-L.
1
I
- - 8
8
§
i
if
i
ii
I
I
I
I
i-i
oo
(0
o
0
0)
4-1
01
>s
to
I
vt
0)
00
H
-------
coupled to the cathode face of a photomultiplier tube. Detection of the
light emission from the reaction yields a signal whose amplitude is
proportional to the concentration of NO in the sample stream. The
concentration of NO. in the ambient air sample is determined after quanti-
tative conversion to NO, when the sample air stream is diverted through a
catalytic converter. The detector alternately measures NO and total
oxides of nitrogen (NO + NO,, reduced to NO), 30 seconds in each mode. By
electronic subtraction of the NO signal from the NO signal, the concen-
X
tration of NO in the air sample is determined. Total time for the
measurement cycle is one minute. Measurement of N0? by the reaction of NO
with 0 is not subject to interference from any of the common air pollutants
9
found in ambient air, such as C02> CO, C-H,, NH«, SO, and H~0 ; but recent
evidence indicates a possible response to PAN (peroxy acetyl nitrate).
The physical characteristics and other descriptors of the Rem
NO-NO -N0_ analyzer are as follows:
X 2.
(1) overall dimensions: 19.6 in (49.8 cm) width,
16.5 in (41.9 cm) height,
18 in (46 cm) depth;
(2) weight: 125 pounds (56.8 Kg);
(3) power requirements: 350 watts at 115 + 110 Volts, 60 Hz;
(5) mode selectors: ambient, zero, span;
(6) adjustments: separate zero and span adjustments for NO,
N0x, and NO channels not provided;
(7) outputs: adjustable from 10 MV to 5 Volts full scale;
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: 0.005 ppm
(9.4 yg/m3);
(b) linearity: 1% for given range;
(c) noise: not specified
(d) precision: + 1% of full scale;
55
-------
(e) accuracy: not specified;
(f) zero drift: + 1% per day,
+ 2% per 3 days,
+ 3% per month;
(g) span drift: + 1% per day,
+ 2% per 3 days,
+ 3% per month;
(h) operational period: not specified;
(9) required accessories: stripchart recorder.
The gas flow system of the Rem analyzer is composed of three separate
systems: (1) sample-span-zero gas system, (2) air ozonator system, and
(3) exhaust system. The Rem analyzer operates at atmospheric pressure,
with sample air being provided via a metal-bellows pressure/vacuum pump.
Components of the gas handling system are the sample pump, pressure
regulator, and 600 cc/min orifice. Constant sample flow rate is achieved
by a pressure drop across the 600 cc/min critical flow type orifice of at
least 2 to 1. Constant pressure is maintained by the sample pump and
pressure regulator which is set at 20 psig. Under these conditions the
reaction chamber is essentially at atmospheric pressure. Ozonated air
is provided through a separate system which consists of a pump, snubber,
air dryer (drierite column), and an electric discharge ozonator. The
exhaust gas system consists of a de-ozonizer and vent to the atmosphere.
A gas flow diagram is shown in Figure 4.13. Particulates are prevented
from entering the inlet lines by a 5y Teflon particulate filter.
The detector cell of the analyzer consists of a reaction chamber,
optical filter, thermoelectrically cooled photomultiplier tube (temperature
controlled to 0°C), and detector cell housing. The optical filter prevents
the photomultiplier tube from responding to light produced by interfering
reactions, while the function of the photomultiplier tube is to convert the
low-level photons resulting from the flameless reaction to electrical
energy and to amplify the electrical signal for further processing the
electrometer amplifier, memory and comparator circuitry.
56
-------
ttf
H
O
TO
«
O
-3"
0)
3
H
57
-------
The Rem catalytic converter uses molybdenum to convert N02 to NO, and
the temperature of the converter is maintained at 700°F. Rejuvenation of
the catalytic converter with dry hydrogen is required when the converter
efficiency drops below 95%.
4.6.2 Operational Summary
The chemiluminescent analyzer was installed in the mobile van on
March 12, 1973 and operated continuously until May 7, 1973 at which time
the analyzer was removed by Rem for modifications and replacement. During
this period of time both short-term and long-term tests were performed on
the analyzer in accordance with the test plan requirements. The frequency
of performance monitoring during these tests was dependent on the particular
requirements of the evaluation procedure being applied. Calibration
frequency for short-term drift was on a daily basis, and that for longer
term drift sudies was on a weekly or biweekly basis.
During the two-month period in which the analyzer was operated and
subjected to the performance evaluation tests, two failures occurred. Both
failures involved the electronic circuitry and amplifier systems and
required service calls by Rem personnel. On May 7, 1973, after the second
failure, Rem personnel removed the analyzer from the evaluation and were
to replace it with an updated model within three days. The replacement
analyzer was never submitted by Rem for reevaluation. As originally
designed, the Rem analyzer required an excessive amount of maintenance.
Routine maintenance involved replacing the drierite in the drying column
for the ozonator air supply at two-day intervals and regenerating the
catalytic converter with hydrogen at 30-day intervals. Replacement of the
particulate filter element of the air sample filter was required at periodic
intervals.
58
-------
4.7 MELOY MODEL SA-185R FLAME PHOTOMETRIC TOTAL SULFUR ANALYZER
4.7.1 Instrument Description
Operation of the Meloy Model SA-185R analyzer is based on the flame
photometric principle, previously described in the literature ' . The
flame photometric principle utilizes detection of the 394-nm centered band
emitted by sulfur-containing compounds when burned in a hydrogen-rich flame,
thus creating an excited state species of sulfur. The resultant release
of light energy when the species returns to ground state sulfur is measured
by a photomultiplier tube that is optically shielded from the flame by the
geometric arrangement of the detector and a narrow band-pass interference
filter. The response of the detector is directly related to the concen-
tration of sulfur entering the detector per unit time.
Basic components of the analyzer are a gas flow system, burner
chamber, photomultiplier tube and hydrogen restrictor line enclosed in a
thermoelectric cooler oven (temperature maintained at 20°C), linearizer
network (converts log output of sulfur concentration versus output
voltage to linear output voltage), hydrogen shut-off valve, and additional
electronic circuitry and amplifier systems. A block diagram of the flame
photometric analyzer is shown in Figure 4.14.
The physical characteristics and other descriptors of the Meloy SA-185R
analyzer are as follows:
(1) overall dimensions: 19.0 in (48.2 cm) width,
10.0 in (25.4 cm) height,
17.0 in (43.2 cm) depth;
(2) weight: 60 pounds (27.2 Kg);
(3) power requirements: 350 watts at 115 + 10% Volts, 60 Hz;
2
(4) measurement ranges: 0.01 to 1.0 ppm (26.1 to 2612 pg/m );
(5) mode selectors: ambient, zero, calibrate;
(6) adjustments: zero and span potentiometers;
(7) outputs: 0-1 Volt;
59
-------
^
> . .
i-4 \<
O /~^Y
M
Nw/
r
erf erf
0 H
H M
H J3
Z U
G Crf
M l-l
0
AMPLIFIER
P
I t
o >-
H p.
O t-
M F-
5
erf
H
g
PQ
c
t
t
1-
(
[
t-
*"
^
^
MODULE
i>
1
q >-t
' ^
H S
^ W
H CO
2 <3
Prf
H
§
1
x;
i]
N]
i
d
3
-I
J
i-
H
b
e
o
c
t-
li
tl
^
4
3
J
J
4
5
-t
J
4
4
4
M
-------
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: < 0.01 ppm
(26.1 yg/m3);
(b) linearity: not specified;
(c) noise: not specified;
(d) precision: + 1% or better at full scale;
(e) accuracy: not specified;
(f) zero/span drift: not specified;
(g) operational period: 14 days or morelimited
by hydrogen supply;
(9) required accessories: supply of hydrogen gas;
stripchart recorder.
The gas flow system of the Meloy analyzer is composed of two parts,
hydrogen and sample air flow. Hydrogen passes from an external source
(cylinder or H_ generator) to the H» flow control valve or through the
optional pressure regulator which maintains 125 cc/min H_ flow. In either
case, hydrogen flow is indicated on the rotameter. If the optional pressure
regulator is used, a capillary restrictor is located in the thermoelectric
cooler oven and maintained at 20°C. A schematic diagram of the gas flow
system with and without options is shown in Figure 4.15 and Figure 4.16,
respectively. The ambient air/calibration gas sample can be brought in
through either the direct inlet line or the metered inlet line. During
normal operation sample air is drawn into the burner chamber through the
direct sample line. Exhaust gases from the burner chamber are diluted
with ambient air, and the exhaust line is heated to prevent condensation
of moisture in the pump. A hydrogen shut-off valve is available, which
discontinues !! flow upon flame-out or power failure.
The detector cell of the analyzer consists of a burner chamber
thermoelectrically-cooled photomultiplier tube optically shielded from the
flame by a narrow band-pass interference filter, electrometer amplifier,
61
-------
DILUTION ^
KXHAITST ^
HYDROGEN |_
: >n
r
i"
^ THERMOELECTRIC
' COOLER (OPTION)
1
J OVEN
i
'l tft
CHAMBER
PM TUBE ASSY
^^n
' SV3 |
(OPTION)
b-
PRESS. RE
I
L
SVl
(OPTION)
G
^ SV2
^t (OPTION")
''ZERO / 1
AIR -*V 1
CHARCOAL FI1
(OPTION)
-O
-\
.
A
,
, BURNER
1 '
^_J H
\\ <*.
}
<,
f
;5
JU
H2
ROTAMETER
AIR
ROTAMETER
Figure 4.15: Gas Flow Diagram (with options)
62
-------
DILUTION
AIR
EXHAUST
DIRECT
INLET
METERED
INLET
HYDROGEN \.
iJo>
PM TUBE ASSY
BURNER
BLOCK
H2
ROTAMETER
AIR
ROTAMETER
Figure 4.16: Gas Flow Diagram (without options)
63
-------
and linearizer network. The purpose of the linearizer network is to
convert the log output (sulfur concentration versus output voltage) to a
linear output voltage in the range (f to 1 V for sulfur concentrations of
0.01 to 1.0 ppm (28.1 to 2612 yg/m ). The linearizer is used in conjunc-
tion with electrometer's log scale only and operates over the range
0 to 1 V.
4.7.2 Operational Summary
The flame photometric analyzer was installed in the mobile van on
December 15, 1972 and operated continuously until May 14, 1973. During
this period of time both short-term and long-term tests were performed on
the analyzer in accordance with the test plan requirements. The frequency
of performance monitoring during these tests was dependent on the particular
requirements of the evaluation procedure being applied. Calibration
frequency for short-term drift was on a daily basis, and that for longer
term drift studies was on a weekly or biweekly basis.
During the five-month period in which the analyzer was operated and
subjected to the performance evaluation tests, no failures occurred and a
minimum amount of maintenance was required. Routine maintenance involved
replacement of the Teflon prefilter element on the sample inlet line on a
monthly basis and replacement of hydrogen cylinders (if used) when the
pressure decreased to 100 psig.
4.8 BENDIX MODEL 8300 FLAME PHOTOMETRIC TOTAL SULFUR ANALYZER
4.8.1 Instrument Description
Operation of the Bendix Model 8300 analyzer is based on the flame
photmetric detection principle. The flame photometric principle utilizes
detection of the 394-nm centered band emitted by sulfur-containing
compounds when burned in a hydrogen rich flame, thus creating an excited
state species of sulfur. The resultant release of light energy when the
species returns to ground state sulfur is measured by a photomultiplier
tube that is optically shielded from the flame by the geometric
64
-------
arrangement of the detector and a narrow band-pass interference filter.
The response of the detector is directly related to the concentration of
sulfur entering the detector per unit time.
Basic components of the analyzer are a gas flow system, burner
chamber, photomultiplier tube with thermoelectric cooling (temperature
maintained at 5°C), linearizer network (converts log output of sulfur
concentration versus output voltage) to linear output voltage, hydrogen
shut-off valve, and additional electronic circuitry and amplifier systems.
A block diagram of the flame photometric analyzer is shown in Figure 4.17.
* The physical characteristics and other descriptors of the Bendix
total sulfur analyzer are as follows:
(1) overall dimensions: 17.6 in (44.8 cm) width,
8.5 in (21.5 cm) height,
17.0 in (43.0 cm) depth;
(2) weight: 60 pounds (27.2 Kg);
(3) power requirements: 300 watts at 115 + 10% Volts, 60 Hz;
3
(4) measurement ranges: 0.01 to 1.0 ppm (26.1 to 2612 yg/m );
(5) mode selectors: ambient;
(6) adjustments: zero and span adjustment potentiometers;
(7) outputs: 0-10 MV (front panel); 0-1 V (rear panel);
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: 0.005 ppm
(13.1 yg/m3);
(b) linearity: + 1%;
(c) noise: 0.5% of full scale;
(d) precision: + 1% of full scale;
(e) accuracy: + 2% of full scale;
(f) zero/span drift: less than + 1% per 24 hours,
less than + 2% per 3 days;
(g) operational period: 7 or more days unattended;
65
-------
HYDROGEN
IN
SOLV. 1
-------
(9) required accessories: supply of hydrogen gas; dryer
and pressure regulator for
hydrogen supply; stripchart
recorder.
The gas flow system of the Bendix analyzer is composed of two parts,
hydrogen and sample air flow. Hydrogen passes from an external source
(cylinder or H« generator) to the H~ flow control valve, through a
capillary restrictor and rotameter, and into the burner chamber. The
ambient air and/or calibration gas sample can be brought in through either
the direct inlet line or the metered inlet line. During normal operation
sample air is drawn into the burner chamber through the direct sample line.
A schematic diagram of the gas flow system is shown in Figure 4.18. Exhaust
gases from the burner chamber are diluted with ambient air, and the exhaust
line is heated to prevent condensation of moisture in the pump. A hydrogen
shut-off valve is available, which discontinues H_ flow upon flame-out or
power failure. Particulates are prevented from entering the inlet line by
a 5y particulate filter (Teflon).
The detector cell of the analyzer consists of a burner chamber,
thermoelectrically cooled photomultiplier tube optically shielded from the
flame by a narrow band-pass interference filter, electrometer amplifier
and linearizer network. The purpose of the linearizer network is to convert
the log output (sulfur concentration versus output voltage) to a linear
output voltage in the range 0 to 1 V for sulfur concentrations of 0.01 to
1.0 ppm (28.1 to 2612 pg/m ). The linearizer is used in conjunction with
electrometer's log scale only and operates over the range 0 to 1 V.
4.8.2 Operational Summary
The flame photometric analyzer was installed in the mobile van on
December 27, 1972 and operated continuously until May 14, 1973. During
this period of time both short-term and long-term tests were performed on
the analyzer in accordance with the test plan requirements. The frequency
of performance monitoring during these tests was dependent on the particular
requirements of the evaluation procedure being applied. Calibration
frequency for short-term drift was on a daily basis, and that for longer
term drift studies was on a weekly or biweekly basis.
67
-------
w «! o
u w erf
O H U
QMS
^j C^
53 N O
W IX H
C C3 <
O co ,J
Prf to £3
P fa1 O
>H 2 w
M a. B
w
r|j
»
H (*S
CO W
iJ W H W
fJ H P4 >
E3 W nJ
PH O > <
H P
O W tt3
J H U
W M §
H fa
CM W
O
CO
CO
n)
H
P
0)
N
a
$
3
CO
o
H
oo
o>
M
3
00
H
fa
68
-------
During the five-month period in which the analyzer was operated and
subjected to the performance evaluation tests, one failure occurred and
required readjustment of the linearizing network. Routine maintenance
involved replacement of the Teflon prefilter element on the sample inlet
line on a monthly basis, replacement of hydrogen cylinders (if used) when
the pressure decreased to 100 psig, and replacement of the molecular sieve
drying column on the hydrogen supply line at periodic intervals.
4.9 BECKMAN MODEL 6800 AIR QUALITY CHROMATOGRAPH
THC, CH4, CO ANALYZER
4.9.1 Instrument Description
The Beckman Model 6800 Air Quality Chromatograph utilizes an
automatic gas chromatographic flame ionization detector (GC-FID) to measure
total hydrocarbons, methane, and carbon monoxide in ambient air. The
analyzer is comprised of a gas sampling valve, a back flush valve, a
pre-column, a molecular sieve column, a catalytic reactor, and a flame
ionization detector (FID). Measured volumes of air are delivered at
five-minute intervals to the FID to determine its total hydrocarbon content,
An aliquot of the same air sample is introduced into the pre-column
(stripper column) which removes carbon dioxide, water, and heavy hydro-
carbons other than methane and prevents them from reaching the molecular
sieve column. Methane and carbon monoxide are passed to a gas chromato-
graphic column where they are separated. Methane is eluted first and
passes unchanged through the catalytic reduction tube. Carbon monoxide is
eluted into the catalytic reduction tube where it is reduced to methane
prior to passing through the FID. Between analyses, the stripper column
is backflushed to prepare it for subsequent analysis. A nickel catalyst
is used in the reduction tube with hydrogen as the carrier gas to satisfy
the requirements for the converter and also as a fuel for the FID.
The measuring electronics is comprised of all solid state devices
which control all the timing functions such as valve switching, gating the
amplifiers, and automatic attenuation for each pollutant. The analyzer is
69
-------
designed for semicontinuous operation and is capable of performing one
analysis every five minutes. An instrument diagram is shown in Figure 4.19.
High specificity for the pollutants of interest is inherent in the
hydrocarbon analyzer design. It is important that the hydrogen carrier gas
be relatively free from CH , CO, C^H,, and C^H- or exist in concentrations
much less than the ambient levels to be measured.
The physical characteristics and other descriptors of the Beckman
chromatograph are as follows:
(1) overall dimensions: 17 in (43.2 cm) width,
40 in (102 cm) height,
20 in (50.8 cm) depth;
(2) weight: -175 Ibs (79.4 Kg)
(3) power requirements: 500 watts, 107 to 127 Volts A.C.
50/60 Hz;
(4) measurement ranges: individually selectable, automatically
actuated, ranges for each component;
maximum sensitivity 1.0 ppm full scale;
(5) mode switch: sample or calibrate
operation: automatic repetitive analysis at selectable
rate of 4, 6, or 12 analyses per hour for
three component system;
(6) adjustment: independent span potentiometers for chromato-
graph and memory for each component;
(7) output: 0-10 MV recorder, 0-5 Vdc output from memory for
trend record;
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: not specified;
(b) linearity: 1% full scale;
(c) noise: not specified;
(d) precision: + 0.5% of full scale or 0.05 ppm,
whichever is greater
(e) accuracy: not specified;
70
-------
BURNER A«
REGULATOR
AIR
HYDROGEN
HYDROCARBON
OXIDIZER
SAMPL^OOP
HYDROCARBON
OXIDIZER
AIR CARRIER
REGULATOR
SAMPLE IN*
SAMPLE
LOOP
SAMPLE-
INJECTION
VALVC B
(TOTAL
HYORO-
CAIWON8)
FID
VENT
1
SAMPLE
VENT
H2 CARRIER
PRESSURE REGULATOR
SAMPLE
INJECTOR
STRIPPER
VALVE A
(CH4a CO)
SAMPLE
AAr
LuuJ
ANALYSIS
COLUMN C3
-fc SAMPLE
VENT
CO
METHANATOR
STRIPPER COLUMN Cl
H2 FUEL PRESSURE
REGULATOR
Figure 4.19: Schematic Diagram of Beckman 6800 Hydrocarbon Analyzer
71
-------
(f) zero drift: automatic zero adjustment during
each analysis cycle compensates
for zero drift;
(g) span drift: not specified;
(h) operational period: not specified;
(i) operating temperature: 35°F to 120°F, 95% relative
humidity;
(9) required accessories: hydrogen cylinder or generator,
two-stage regulator, air supply
or cylinder of compressed air.
The atmospheric sample is drawn into the analyzer by an internal
metal-bellows pump. The pump provides a sample flow rate of 5 liters per
minute; however, most of the ambient sample is released to the atmosphere
to prevent pressurizing the sample loops. The Model 6800 consists of three
sections: (1) gas control section, (2) chromatographic oven, and (3) elec-
tronic control section. Pressure regulators and associated gauges for air
and hydrogen carrier gases, hydrogen fuel, air support gases for the FID,
and service air to operate the slider valves are located on the front
panel of the gas control section. The oven section contains the columns,
slider valves, fan to equalize the temperature, and the flame ionization
detector. The electronic control section includes circuitry for flame
Ignition, amplifiers for FID output signal, automatic control of all time-
related functions, memory and analog presentation of data and power supplies.
4.9.2 Operational Summary
The Beckman chromatograph was installed in the mobile van on
December 14, 1972 and operated continuously until June 29, 1973. During
this period of time both short-term and long-term tests were performed on
the analyzer in accordance with the test plan requirements. The frequency
of performance monitoring during these tests was dependent on the particular
requirements of the evaluation procedure being applied. Calibration fre-
quency for short-term drift was on a daily basis, and that for long-term
drift studies was on a weekly or biweekly basis.
72
-------
During the six-month period in which the analyzer was operated and
subjected to the performance evaluation tests, two failures occurred.
The first failure involved a sticking valve slider, and the second
involved the temperature controller for the oven. A minimum amount of
maintenance was required during this period of time. Routine maintenance
involved replacing the sample inlet filter element, maintaining the
hydrogen generator, and replacing compressed air cylinders periodically.
4.10 BENDIX MODEL 8201 AMBIENT HYDROCARBON ANALYZER
4.10.1 Instrument Description
The Bendix ambient hydrocarbon analyzer utilizes a hydrogen flame
ionization detector (FID) for the measurement of reactive hydrocarbons.
The analyzer is capable of measuring total hydrocarbons, methane, and
non-methane hydrocarbons by difference method (i.e., THC minus CH ).
During the operation, a metered quantity of ambient air is routed from
the sample inlet through the sample valves and a GC-column where methane is
separated from the THC. Methane is then routed from the column into the
FID to yield a measurement. A second quantity of ambient air is then
routed from the sample inlet through the sampling valve directly to the
FID, bypassing the analytical column. The value of the methane analysis
is then electronically subtracted from the total hydrocarbon value,
providing a measurement of non-methane hydrocarbons. Analytical functions
and signal processing are programmed by a solid state electronic timer.
The output is updated after each cycle which is approximately three minutes.
The signals are stored in permanent memories that are updated after each
cycle.
The physical characteristics and other descriptors of the Bendix
hydrocarbon analyzer are as follows:
(1) overall dimensions: 17.6 in (44.8 cm) width,
9.2 in (23.4 cm) height,
17.2 in (43.8 cm) depth;
73
-------
(2) weight: 50 Ibs (22.7 Kg)
(3) power requirements: 700 watts maximum,
500 watts normal at
105-125 volts,
50 or 60 Hz;
(4) measurement ranges: 1, 2, 5, 10 ppm
(5) Mode switch: Sample or calibrate
(6) adjustment: zero and span adjustment potentiometers
for THC, CH^, THC-CH^;
(7) outputs: 0-10 MV recorder, 0-1 Vdc;
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: 0.005 ppm;
(b) linearity: +0.5%;
(c) noise: + 1%;
(d) precision: + 1% of full scale;
(e) accuracy: + 1% of full scale;
(f) zero drift: + 1% per 24 hours
(automatically zeroed each cycle);
(g) span drift: + 1% per 24 hours or
+ 2% per 3 days;
(h) operational period: 7 or more days unattended
(i) operating temperature fluctuations: + 5°C
(j) operating temperature extremes: 5°C to 40°C
(9) required accessories: cylinder of hydrogen or hydrogen
generator, two-stage regulator,
air supply unit with hydrocarbon
oxidizer.
The instrument requires a source of clean air and a source of pure
hydrogen. The hydrogen consumption is approximately 50 cc per minute.
74
-------
60
TO
H
Q
u
H
r-H
<1
c
-------
Sample flow rate is approximately 200 cc per minute. The analyzer flow
diagram is shown in Figure 4.20.
Automatic zeroing of the detector output is accomplished during each
cycle to ensure a stable baseline for long-term, unattended operation.
4.10.2 Operational Sinnmary
The hydrocarbon analyzer was installed in the mobile van on
February 16, 1973 and operated continuously until July 19, 1973. During
this period of time both short-term and long-term tests were performed
on the analyzer in accordance with the test plan requirements. The
frequency of performance monitoring during these tests was dependent on
the particular requirements of the evaluation procedure being applied.
Calibration frequency for short-term drift was on a daily basis, and that
for long-term drift studies was on a weekly or biweekly basis.
During the five-month period in which the analyzer was operated and
subjected to the performance evaluation tests, two failures occurred. One
failure involved valve actuation, and the other failure resulted in loss
of output signal. A minimum amount of maintenance was required during
this period of time. Routine maintenance involved replacement of the Teflon
filter on the sample inlet and maintaining hydrogen generator and/or cylinder
replacement. The air supply unit operated for the five-month period with no
failures.
4.11 BENDIX MODEL 8501-5FA CARBON MONOXIDE ANALYZER
4.11.1 Instrument Description
The Bendix carbon monoxide analyzer operates on the principle that
CO has a sufficiently characteristic infrared absorption spectrum so that
the absorption of infrared radiation by the CO molecule can be used to
measure the CO concentration in the presence of other gases. A schematic
diagram is shown in Figure 4.21. Optical beams from the infrared source
travel through adjacent parallel cells, one beam for the sample cell and
one for the reference cell. The infrared radiation from the radiation lamp
76
-------
cfl
H
Q
O
O
PQ
(1)
N
CO
cfl
O
T)
(U
M
ctj
M
M-l
C
CN
-j
01
00
H
Pm
77
-------
is modulated in phase reversal by the rotating chopper into two equally
intense portions. After passing through a diffuser, both radiation portions
enter the detection chamber which contains the gas constituent to be
measured. The reference cell is filled with nitrogen which does not absorb
the infrared radiation. The gas mixture to be analyzed flows through the
sample cell. The absorbed energies heat the gas in these chambers, causing
the pressure of the gas to rise. The gas concentration in the two measuring
chambers and the geometry of these chambers are such that the pressures are
equal on both sides of the diaphragm capacitor when none of the constituent
to be measured is in the analysis chamber. If, however, the radiation in
the analysis chamber passes through a gas mixture containing CO, some of
the specific radiation energy is absorbed, and the resulting reduction of
radiation intensity takes place primarily in the center of the infrared
absorption band. The condition disturbs the pressure equilibrium existing
between the two chambers, and the pressure differential is proportional to
the deflection of the diaphragm. This results in a change of measuring
capacitor which is excited to vibration at a frequency corresponding to the
radiation modulating chopper. The amplitude of this vibration is a measure
of the gas concentration to be determined. The signal is further amplified,
filtered, and demodulated to provide stable output. A diagram of the gas
flow system is shown in Figure 4.22.
The physical characteristics and other descriptors of the Bendix CO
analyzer are as follows:
(1) overall dimensions: 30 in (76 cm) height,
19 in (48 cm) width,
11 in (27 cm) depth;
(2) weight: 60 pounds (27.22 Kg);
(3) power requirements: 300 watts at 105-125 volts, 60 Hz;
(4) measurement ranges: 0-20, 50, 250, 500, 1000 ppm;
(5) mode switch: zero, ambient, calibrate;
(6) adjustments: zero and span potentiometers;
78
-------
w o
HJ IS
CL H O M
S 5 Pi H
<2 o w iz:
CO
1
/
H W
CO Pi
ex CD W
co O
0 co S
f"> W <3
i 3 o
O P-i
Ofn fti
W Q
M H
eJ <
g g
o
w w
pi Pi
n
CO W
CO >
W J
Pi **S
PJ >
n \
CM
H
M
1 (
. v
:
pi
o
W H W
Q U >
O W >J
S J <
W >
CO
i r^
i
n
W I ^X.
ij fel \
ill 9)
^ m v *
CO \y____ ^
a
0 Pi
pi w ._JL_
0 HU- ]
M eJ 1 , 1
s y\^
(-^
c»
/^ /
(> ^
^
Si
1 4 I
' /V
Pi
IV)
N
>-<
i-l
3
^
iH
0)
C
3
en
ca
o
T3
Q)
Vj
& m -
W LU ^
Pi M
w
CO H
-------
(7) output: 0-10, 0-1000 MV, 0-1.0 volt;
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: 0.5 ppm;
(b) linearity: + 0.5% of full scale on 20, 50 ppm range,
+ 1.0% of full scale on 25, 500,
1000 ppm range;
(c) noise: not specified;
(d) precision: 1% full scale;
(e) accuracy: not specified;
(f) zero drift: 0.05% per hour or + 1% per 24 hours,
whichever is lower,
+ 2% of full scale for 3 days;
(g) span drift: 0.05% per hour or +_ 1% per 24 hours,
whichever is lower,
+ 2% of full scale for 3 days;
(h) operational period: 7 days
(9) required accessories: stripchart recorder.
4.11.2 Operational Summary
The carbon monoxide analyzer was installed in the mobile van on
April 26, 1973 and operated continuously until July 31, 1973. During this
period of time both short-term and long-term tests were performed on the
analyzer in accordance with the test plan requirements. The frequency of
performance monitoring during these tests was dependent on the particular
requirements of the evaluation procedure being applied. Calibration
frequency for short-term drift was on a daily basis, and that for longer
term drift studies was on a weekly or biweekly basis.
During the three-month period in which the analyzer was operated and
subjected to the performance evaluation tests, two failures occurred. One
failure was attributed to a faulty diode resulting in blown fuses and the
flexible coupling on chopper motor shaft dragging causing an erratic
80
-------
modulation frequency. A minimum amount of maintenance was required during
this period of time. Routine maintenance involved replacement of the
filter element on the sample inlet line.
4.12 ANDROS MODEL 7000 DUAL-ISOTOPE-FLUORESCENCE
CARBON MONOXIDE ANALYZER
4.12.1 Instrument Description
The principle, of operation of the Andros 7000 Dual-Isotope-
Fluorescence NDIR analyzer is based on the ratio comparison of alternate
-i / 10
infrared absorption of CO and CO radiation in a sample chamber. The
IR radiation spectra that are an exact match of the vibrational-rotational
Ifi 18
absorption bands of CO and CO are produced and alternately allowed to
enter the single sample chamber. The CO present in the sample chamber
will absorb CO radiation but not CO radiation. The alternating CO
18
and CO radiation levels are then converted to electrical signals by a
solid-state IR detector. Ratio comparison of the two signals yields a
voltage level corresponding to the CO-concentration in the sample air. A
broad-band "blackbody" IR source is used to stimulate fluorescence of
1 (\ 1 ft "" ' '
CO and CO isotopes contained in two sealed cells mounted on a rotating
chopper wheel. Each cell, when appearing in the optical path, will filter,
I / -I Q
by absorption, either CO or CO IR radiation. The net result-is a .
1 f\ 18
single beam of IR energy which consists of alternating CO and CO IR
radiation. Particulate matter and moisture present in the gas under test
or accumulated on the "optical windows attenuates both IR pulses equally.
Since 99.8% of all naturally occurring CO is CO , no ratio difference will
be detected, thus eliminating drift problems associated with conventional
NDIR analyzers using a separate reference chamber. A diagram of the D.I.F.
analyzer is shown in Figure 4.23.
The physical characteristics and other descriptors of the Andros CO
analyzer are as follows:
(1) overall dimensions: 19 in (48.3 cm) width,
5.25 in (13.3 cm) height,
16.7o in (42.5 cm) depth;
81
-------
(I)
N
H
X
O
i
c
O
§
M
a
0)
1-1
3
bO
H
82
-------
The detector of the Andros 7000 uses a solid-state IR detector
(photo-diode) which is mechanically rugged and non-microphonic. The
detector is operated at peak CO-sensitivity by being thermally stabilized
at -30°C by a Peltier Cooler. Since the detector is a solid-state device,
high mechanical stability is achieved. The Andros analyzer is much less
sensitive to shock and vibration than conventional NDIR analyzers.
4.12.2 Operational Summary
The dual-isotope-fluorescence NDIR analyzer was received from
Andros and installed in the mobile van on March 22, 1973, but was found to
be inoperative due to improper mounting of the pump prior to shipment. The
pump was repaired, and the analyzer was set up'on April 3, 1973 and oper-
ated continuously until May 25, 1973. During this period of time both
short-term and long-term tests were performed on the analyzer in accordance
with the test plan requirements. The frequency of performance monitoring
during these tests was dependent on the particular requirements of the
evaluation procedure being applied. Calibration frequency for short-term
drift was on a daily basis, and that for longer term drift studies was on
a weekly or biweekly basis.
During the approximately two-month period in which the analyzer was
operated and subjected to the performance evaluation tests, two failures
occurred, both as a result of power failures. "Immediately upon resumption
of AC power, the analyzer output was pegged off-scale on the 200-ppm scale
when recording zero CO concentration for approximately 36 hours. Eventually
the output signal decreased to 'a zero air value approximately 50% higher
than the original zero level. After stabilization at that level, the
proper zero level was reset using the zero adjustment control. This
occurrence repeated -itself on May 25, 1973, at which time the analyzer was
returned to Andros for repairs.
83
-------
4.13 TRACOR MODEL 270-HA GAS CHROMATOGRAPHIC-FLAME
PHOTOMETRIC ANALYZER
4,13.1 Instrument Description
The principle of operation of the Tracer 270-HA analyzer is based
on gas chromatographic separation of sulfur compounds (i.e., total sulfur,
H?S, and S0~) and quantitative determination using a flame photometric
detector equipped with a 394-nm optical filter. Low concentration levels
(i.e., parts per billion) of total sulfur, HjS, and S02 in ambient air can
be quantitatively separated and measured. A typical chromatogram of a
mixture of these sulfur compounds requires five minutes. A precolumn and
backflushing are employed to prevent mercaptans from entering the analytical
column. The system utilizes two sample loops, one for total sulfur and one
for SO and H2S,
A schematic diagram is shown in Figure 4.24. Since a chromatographic
column is used to quantitatively separate sulfur-containing compounds and
since a 30,000-to-l specificity ratio of sulfur to non-sulfur compounds is
achieved by the detector, the GC-FPD measurement for H9S and SO- is very
specific. No interferences were observed from gases under study.
The basic components of the analyzer are a flame photometric detector,
thermally stabilized oven housing for valves and columns, two-range signal
linearizer, timing and control circuitry. The linearizer has two signal
outputs for each pollutant. The low range is 0 to 0.1 ppm (0-261.8 pg/m )
3
and the high range is 0.1 to 1.0 ppm (261.8-2618 Mg/m ).
The physical characteristics and other descriptors of the Tracer TS,
H2S, SO analyzer are as follows:
(1) overall dimensions: 19 in (48.2 cm) width,
9 in <22.8 cm) height,
24 in (61.0 cm) depth;
(2) weight: 55 pounds (25.2 Kg);
(3) power requirements: 250 watts;
(4) measurement ranges: 0.1, 1.0 ppm full scale;
84
-------
a
H
J3
ex
R)
f-l
M
o
U M
nj (U
B N
O >,
U H
J3 CO
CO
td
O
cs
M
O -
O CM
cO O
tH C/J
H
O H
S o
CO -H
)-i VJ
bO 4-i
CO 0)
H 0
P O
4J
O O
H JS
W Oi
CN
-------
(5) mode switch: inject/load sample (manual operation);
(6) cycle selector: automatic, manual and remote;
(7) output: trend and chromatograph, 0-1 volts DC
(8) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: not specified;
(b) linearity: not specified;
(c) noise: not specified;
(d) precision: not specified;
(e) accuracy: not specified;
(f) zero drift: not specified;
(g) span drift: not specified;
(h) operational period: not specified;
(i) operating temperature: not specified;
(9) required accessories: cylinder H_, air supply, two-stage
regulator, stripchart recorder.
The gas flow system of the Tracer analyzer is composed of two 10-port
valves, two sample loops, rotameter, flow control valve and pump. Sample
flow rate is maintained at approximately 50 cc/minute by a variable
restrictor and vacuum pump. Particulates greater than 5 microns in
diameter are removed from the ambient air sample by a Teflon filter. The
detector assembly of the analyzer consists of a flame chamber, optical
filter, photomultiplier tube and housing. The basic functions of the
photomultiplier assembly are to convert the low-level photons resulting from
the excited sulfur atoms to electrical energy and to amplify the electrical
signal for further processing through the high- or low-range linearizer,
analog-to-digital converter, digital storage, and digital-to-analog converter
output module circuitry. A block diagram of the detector and signal
processing is shown in Figure 4.25.
86
-------
00
C
r-l
W
W
0)
O
O
n)
a
M
r-t
CO
O
0)
4-1
0)
Q
M
O
O
a)
M
H
14-4
O
60
n)
a
o
PP
-------
4.13.2 Operational Summary
The sulfur chromatograph was installed in the mobile van on
December 18, 1972 and operated continuously until July 2, 1973. During
this period of time both short-term and long-term tests were performed on
the analyzer in accordance with the test plan requirements. The frequency
of performance monitoring during these tests was dependent on the particular
requirements of the evaluation procedure being applied. Calibration
frequency for short-term drift was on a daily basis, and that for long-term
drift studies was on a weekly or biweekly basis.
During the six-month period in which the analyzer was operated and
subjected to the performance evaluation tests, five failures occurred, two
involving vacuum pumps, and one each involving valve actuator, high range
linearizer and total sulfur memory. A considerable amount of maintenance
was required during this period of time. Routine maintenance involved
replacing the Teflon filter element on the sample inlet line and changing
compressed gas cylinders on a monthly basis.
4.14 BENDIX MODEL 8700 GAS CHROMATOGRAPH-FLAME
PHOTOMETRIC ANALYZER
4.14.1 Instrument Description
The Bendix Environmental Chromatograph Model 8700 principle of
operation is based on gas chromatographic separation of sulfur compounds
(sulfur dioxide, hydrogen sulfide, and total sulfur) and quantitative
determination using a flame photometric detector equipped with a 394-nm
optical filter. The instrument consists of three sections: (1) control
and signal processing electronics, (2) oven and analysis, and (3) sample
system. The control electronics, timer, and electrometer are solid state
and are mounted in a slideout drawer.
The electrometer output is linearized to provide an output signal
directly proportional to the sulfur concentration. After the respective
signals for each component have been linearized, they are stored in
respective memories which are updated each cycle.
88
-------
The analytical components are mounted in an insulated, temperature-
controlled oven with associated valves and detector. The gas-handling
system of the Bendix analyzer is composed of two 10-port valves, two sample
loops, rotameter pressure regulators, and pump. The sample system includes
a pump and associated flow control valves. The detector assembly of the
analyzer consists of a flame chamber, optical filter, photomultiplier tube
and housing. A schematic diagram is shown in Figure 4.26.
The physical characteristics and other descriptors of the Bendix
chromatograph are as follows:
(1) overall dimensions: 22 in (56 cm) width,
41 in (104 cm) height,
23 in (58.5 cm) depth;
(2) weight: 250 pounds (113.6 Kg);
(3) power requirements: 800 watts, 115 Volt, 60 Hz;
(4) measurement range: 0.02 to 20 ppm on continuously
adjustable attenuator for individual
components
(5) outputs: 0-10 Volts trend-type for each component
10 MV chromatographic output
(6) stated performance specifications (manufacturer)
(a) minimum detectable sensitivity: H2S - 0.005 ppm
S02 - 0.010 ppm
TS - 0.010 ppm
(b) linearity: less than 2% of full scale
(c) noise: 0.5% of full scale
(d) precision: + 4% of full scale
(e) accuracy: + 2% of full scale
(f) zero drift: less than +_ 1% per day,
2% per three days;
automatic zero of baseline
before each component;
89
-------
8-
1-1
60
O
§
M
43
O
O
O
r^
oo
X
H
o
-------
(g) span drift: less than +_ 1% per day,
2% per three days;
(h) operational period: more than three days
unattended;
4.14.2 Operational Summary
The sulfur chromatograph was installed in the mobile van on
March 27, 1973 and operated continuously until July 3, 1973. During this
period of time both short-term and long-term tests were performed on the
analyzer in accordance with the test plan requirements. The frequency
of performance monitoring during these tests was dependent on the particular
requirements of the evaluation procedure being applied. Calibration fre-
quency for short-term drift was on a daily basis, and that for long-term
drift studies was on a weekly or biweekly basis.
During the three-month period in which the analyzer was available for
evaluation, a minimum number of performance evaluation tests were performed
due to time constraints; and no failure occurred. A minimum amount of
maintenance was required during this period of time. Routine maintenance
involved replacement of sample inlet filter element, maintaining hydrogen
generator, and replacing compressed gases.
91
-------
5,0 INSTRUMENT EVALUATION PROCEDURES
A general set of performance criteria was established in earlier
reports describing field evaluation of advanced sensors for monitoring air
123
pollutants ' ' . These criteria were based on instrument characteristics
that are independent of the type of instrument being evaluated or its
application. Instrument performance characteristics fall into four major
groups: physical characteristics, measured responses to standard test
procedures, field data quality, and functional capacity. Physical
characteristics are usually obvious and require minimal analysis. Response
of instruments to standard test procedures poses no significant problems
in determination of values for comparison with performance or desired
operational specifications. Field data quality is concerned with cali-
bration requirements, stability, accuracy, and limits of detection.
Performance in the area of functional capacity is concerned primarily with
instrument failure. The most obvious negative functional characteristics
are instrument downtime and maintenance costs.
5.1 DEFINITIONS
The technical definitions used to describe sensor specifications for
this program are as follows:
Range- - - - The minimum and maximum measurement limits which
the instrument is capable of measuring.
Noise - - - Spontaneous, random, short-duration deviations in
the instrument output about the mean output,
which are not caused by input pollutant concen-
tration changes.
Zero drift - The change in instrument output over a stated time
period of unadjusted, continuous operation when
the input concentration of pollutant is zero.
Span drift - The change in instrument output over a stated time
period of unadjusted, continuous operation when the
pollutant concentration is a stated value.
92
-------
Linearity- - The standard deviation between the actual
instrument reading and the reading predicted by
the line of best fit as determined by the method
of least squares.
Precision- - Variation about the mean of repeated measurements
of the same concentration.
Minimum detectable concentration or limit - The smallest
amount of input concentration which can be
detected with a specified degree of confidence.
Lag time - - The time interval from a step change in the input
concentration at the instrument inlet to the first
corresponding change in the instrument output. This
can be most reproducibly determined by extrapolating
the slope at the 5%.point.
Rise time- - The interval between the initial response time and
the time to 95% response after a step change in the
input concentration.
Fall time- - The interval between the lag time, t., and the time
to 95% response after a step decrease in the inlet
concentration. This time is not necessarily equal
to the rise time in the instrument.
Line voltage - The RMS voltage supplied to the instrument.
Operational period - The period of time over which the instrument
can be expected to give meaningful results without
benefit of maintenance, service, or adjustment.
Accuracy - - The standard deviation between the indicated and the
true values of the parameter being measured, unless
otherwise stated. The true value is determined by
analytical procedures consistent with the accepted
reference methods.
93
-------
5.2 TEST PROCEDURES
Test procedures were developed to determine the following parameters:
stability, linearity, precision, minimum detectable concentration (sensi-
tivity), response times (lag, rise, fall), noise, line voltage variation,
and flow rate/pressure variation. The test procedures devised and utilized
to determine each of these parameters are described in stepwisa fqrm in
the following subcategories.
5.2.1 Stability - Zero and Span Drift
(1) Allow sufficient time for instrument to warm up and
stabilize.
(2) Sample zero air until a stable reading is obtained and
adjust baseline to 10% of chart. Instrument is operated
at normal line voltage controlled at 117 volts. Normal
room temperature is maintained at 25 + 2°C.
(3) Sample test concentration of pollutant equal to 80 + 5%
of operating range selected. Adjust span control for
appropriate reading.
(4) Generate and measure a test concentration equal to
20 + 5% of the selected range. Do not adjust span
control.
(5) At this point no adjustments can be made to the instrument
except in the case of failure or discontinuation of test.
Instrument can sample ambient air or undergo other tests
during idle periods.
(6) Generate and measure test pollutant concentrations of
0, 20, and 80% of the range limit each day, 24-hours after
the previous day's readings. Test concentrations must be
consistent from day to day. Allow sufficient time for
stabilization of zero and span output and record data in
concentration units. Subtract each day's readings from the
previous day's respective readings to obtain zero and span
drift for that day.
94
-------
5.2.2 Linearity
(1) Allow sufficient time for instrument warm-up and
stabilization.
(2) Sample zero air and adjust baseline to 10% of chart
after stabilization period.
(3) Sample and measure at least eight different test
concentrations evenly distributed on the operating range.
(4) Plot these test points, including zero, on a linear plot
with test concentrations on the horizontal axis and instru-
ment reading on the veritical axis. (See Figure 5.1.)
(5) Draw line of best fit determined by linear regression
analysis.
(6) Calculate the standard deviations of the test points
about the line of best fit.
5.2.3 Precision
(1) Allow sufficient time for instrument warm-up and
stabilization.
(2) Sample and measure zero until a stable reading is
obtained.
(3) Sample test concentration of 80% of range and record
steady-state reading.
(4) Rapidly reduce pollutant concentration to zero air.
(5) Repeat Steps 3 and 4 six times and record each value.
(6) Calculate and report mean and standard deviations for
this data.
5.2.4 Minimum Detectable Concentration (Limit)
(1) Allow sufficient time for instrument warm-up and
stabilization.
(2) Run calibrations using a minimum of eight calibration
standard, recording concentrations and instrument output.
95
-------
INCREMENT (&) = INSTRUMENT READING DATA POINT VALUE
MINUS REGRESSION LINE VALUE
PL,
O
Y = y + M (x - x)
H
C/3
where: M =
N
- v =
.2 y N
Zx
i N
x =
N
S.D
4
N - 1
I I I I I
I I i I i l
TEST CONCENTRATION (PPM)
Figure 5.1: Method of Determining Linearity
96
-------
(3) Range of calibration standards should be as follows:
03 0.2 ppm
NO - - 0.5 ppm
N02 0.5 ppm
S02 0.2 ppm
H2S 0.2 ppm
CO 10.0 ppm
THC 10.0 ppm
CH, 10.0 ppm
(4) Plot signal output and calibration standards on ordinate
and abscissa, respectively.
(5) Using linear regression, calculate and plot best fit
line Y = y + b (x - x) as shown in Figure 5.2.
(6) Determine from a confidence level to 90% the students'
t value corresponding to N-2 degrees of freedom.
(7) Calculate the signal level, (y ), from the following
equation for x = o.
y = y + b(x - x)
Z(xi - x)
where
s = JE(VJ - Yi)2 .
II N-2
(8) Calculate the minimum detectable level (JLj from the
following equation:
= y + b(X - X) -
97
-------
Y = y + b (x - x)
Concentration
Figure 5.2. The Linear Calibration Line, With Its
Upper and Lower Confidence Limits
5.2.5 Response Time
(Lag Time)
(1) Sample zero air until stabilized.
(2) Switch sample inlet to test concentration of 80% of
operating range and simultaneously mark stripchart
record.
(3) After instrument has stabilized, determine the 5%
and 95% points of the stabilized final value.
(4) Draw a tangent line at the 5% point and note the
zero intercept. Determine the time between the
initial mark and the intercept to yield lag time.
(Rise Time)
(5) The rise is determined from the intercept to the
95% point.
98
-------
(Fall Time)
(6) Sample at concentration of 80% of operating range and
obtain a stable reading.
(7) Switch sample inlet to zero air and simultaneously
mark stripchart recorder.
(8) Determine intercept using 5% (as in Step (4)) and
95% point. (See Figure 5.2.)
5.2.6 Noise
(1) Allow sufficient time for instrument warm-up using
zero air.
(2) Connect instrument output to recording digital
voltmeter sampling rate set at one-minute intervals.
(3) Sample zero air for 60 minutes.
(4) Locate the largest positive and negative excursions
during the period and determine the maximum "peak-to-
peak" value. Convert this value to concentration units.
(5) Repeat Steps (3) and (4) using a concentration of 80%
of operating range.
(6) Record the maximum value in concentration units.
5.2.7 Line Voltage Variation
(1) Connect a variable transformer between the instrument
and supply voltage. Allow the instrument to warm up
with the variable transformer set at 117 volts.
(2) Calibrate the instrument using at least five concen-
tration levels and determine the best fit line by
regression analysis.
(3) Decrease line voltage to 105 volts and allow instrument
to stabilize electrically and thermally.
(4) Repeat Step (2).
99
-------
CO
o
D,
CO
0)
CO
IT)
100
-------
(5) Increase line voltage to 125 volts and allow
Instrument to stabilize.
(6) Repeat Step (2).
(7) Calculate slope and intercept changes with respect
to calibration data obtained at 117 volts.
5.2.8 Flow Sate/Pressure Variation
(1) Allow instrument to warm up and stabilize using a
concentration level of 80% of the operating range.
(2) Adjust the flow rate or pressure up and down by a
small increment (1-20% range) simulating a typical
change preceding a failure for the instrument.
(3) Record instrument output for several pressure and/or
flow rate increments to establish instrument response
dependence.
(4) Express instrument response dependence as concentration
units per pressure unit or concentration units per
volumetric flow rate.
101
-------
6,0 CALIBRATION SYS1H-1S/PROCEDURES
6.1 CALIBRATION SYSTEMS
6.1.1 Ozone
12
A dynamic calibration system as described by Hodgeson et al. and
published with the National Primary and Secondary Ambient Air Quality
13
Standards was used to calibrate the ozone instrumentation evaluated
during this study. Briefly, the ozone source consists of an 8-inch
ultraviolet mercury lamp which irradiates a 5/8-inch quartz tube through
which clean (compressed) air flows at 5 liters/minute. Ozone concentrations
from 0 to approximately 1 ppm (1960 yg/m ) can be generated by moving the
shield and exposing various lengths of the lamp. Although the UV 0,. gener-
tr£
13
1 12
ator has been shown to be quite stable and reproducible ' , the neutral-
buffered potassium iodide technique was used as the reference method
A permanent calibration setup consisting of a zero air source, calibrated
rotameter, UV generator, and a glass manifold was installed in the labor-
atory facility and calibrated by the manual neutral-buffered potassium
iodide procedure periodically during the study. A diagram of the calibra-
tion system is shown in Figure 6.1.
6.1.2 Nitric Oxide/Nitrogen Dioxide
Due to problems associated with long-term use of NO. permeation
tubes and the need to routinely determine the efficiency of the carbon,
stainless steel, or molybdenum converters (which reduce NO,, to NO), an
alternate procedure (gas phase titration) was used for routine dynamic
calibration of the chemiluminescent NO-NO -N00 analyzers. The gas phase
X £.
titration technique is based upon application of the rapid gas phase
7 8
reaction between NO and 0. to produce a stoichiometric quantity of NO- ' ,
NO + 0
hv
102
-------
ex.
LU
co
M
*-!
e
(U
o
H
c
a)
C
O
tsl
O
(U
Vi
103
-------
Nitric oxide from a cylinder of NO in N~ (100 ppm) is diluted with a
constant flow of clean air to provide 1.0 ppm and used to calibrate the NO
and NO cycles of the chemiluminescent NO-NO -N00 analyzer. By incor-
X A ^
poration of a calibrated ozone generator in the calibration apparatus
upstream from the point of NO addition, precise N0» concentrations can be
generated by oxidation of NO to NO- with 0». A schematic diagram showing
the component parts of the calibration system is presented in Figure 6.2.
As long as a slight excess of NO is present, the concentration of 0., added
is equivalent to the concentration of NO consumed and is equivalent to the
concentration of NO- generated.
A general description of this calibration scheme is presented in the
following paragraphs. Primary calibration of the NO concentration in the
pressurized cylinder containing nitrogen as a diluent is accomplished by
titration of an NO concentration of 1.0 ppm produced by dilution with
successive concentrations of ozone (0-0.8 ppm) generated by an ozone
13
generator which has been referenced to the neutral-buffered KI procedure
The resultant NO detector outputs after stabilization at each titration
point (i.e., 0.0, 0.1, 0.2, 0.8 ppm ozone added) are plotted in ppm on
coordinate graph paper (y-axis) versus 0- concentrations added, ppm (x-axis)
A straight line is drawn through the linear portion of the titration curve
and extrapolated to the x-axis. The concentration at the x-axis intercept,
C', is the 0» concentration equivalent to the initial diluted NO concen-
tration. An example of a typical gas phase titration curve is presented in
Figure 6.3. The concentration of NO in the cylinder can then be determined
as follows:
°NO= FNO
where
(l^n = cylinder NO concentration, ppm,
F..,, = measured NO flow, cc/min,
NO
Cl = equivalence point 0- concentration, ppm,
F- = total clean air flow, cc/min.
104
-------
CALI BRAT ION SYSTEM
ADJUSTABLE SLEEVE
PEN-RAY
LAMP
CYLINDER
AIR
Figure 6.2: Gas Phase Titration System for
Calibrating NO-NO -NO Analyzers
X £.
105
-------
EQUIVALENCE
POINT
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0- Concentration (ppra) (0. Generator)
Figure 6,3: Gas-phase Titration of NO with 0,
106
-------
Once the NO concentration in the cylinder has been determined, this cylinder
can be used over its lifetime to provide a working standard for routine
14
calibration ; however, to assure validity of data, the NO concentration
should be verified at one-month intervals.
During routine calibration, the NO and NO channels of the chemilumi-
X
nescent NO-NO -N0~ analyzers were calibrated by dynamic flow dilution of
the NO in nitrogen cylinder gas. To calibrate the NO- output channel and
to determine the converter efficiency (i.e., efficiency of reduction of
N0« to NO), a constant concentration of 0.5 or 1.0 ppm of NO is produced
in the flow system. Ozone is then added in increments from the variable
0- source. The incremental decrease of the NO measurement is then
equivalent to the concentration of NO- produced by the gas phase titration
reaction. In this scheme the calibrated 0_ source becomes a calibrated
N0_ source when NO is present in excess.
6.1.3 Sulfur Dioxide/Hydrogen Sulfide
A dynamic calibration system using a gravimetrically calibrated
sulfur dioxide (SO-) permeation tube as a primary standard and zero air
as diluent was used to provide known concentrations of SO- for calibration
of total sulfur and sulfur dioxide analyzers. Gravimetrically calibrated
SO. permeation tubes were obtained from the National Bureau of Standards.
Hydrogen sulfide (H_S) permeation tubes were purchased from Metronics, Inc.,
gravimetrically calibrated at EPA on a Cahn balance, and used to provide
known concentrations of H-S for the gas chromatographic-flame photometric
analyzers evaluated during this study. Figure 6.4 shows the permeation
tube calibration system used for this investigation. The permeation tube
was housed in a pyrex glass holder and immersed in a constant temperature
bath which maintained the tube at a temperature of 20.3°C + 0.1°C. Dry,
compressed air, conditioned to the temperature of the bath and metered
through a rotameter, was passed over the permeation tube and into a 1-inch
O.D. glass manifold from which the analyzers were allowed to sample. This
allowed for simultaneous calibration of S0~ and H2S analyzers. Two
permeation tube holders were utilized in this studyone each for the S0
107
-------
PERMEATION TUBE
^CONSTANT-TEMPERATURE
WATER BATH
V.
COMPRESSED AIR
THERMISTOR TEMPERATURE MONITOR
Figure 6.4: Permeation Tube Calibration System
108
-------
and H_S tube. By varying diluent air flow rate, SO- or H S concentrations
3 3
from 0 to 0.2 ppm (524 yg/m S02. 279 yg/m H2S) could be generated and
supplied to the respective analyzers during calibration.
6.1.4 Hydrocarbons/Carbon Monoxide
Calibration of hydrocarbon and carbon monoxide instruments was
accomplished utilizing standard calibration gases certified by the
supplier. For this study cylinders of methane and carbon monoxide in air
were obtained from Scott Research Laboratories. Several concentrations
ranging from 0 to 10 ppm of CH, and CO were used for calibration purposes.
Cylinders of zero air (hydrocarbons _^ 0.1 ppm) were supplied by several
manufacturers for comparison purposes. The total hydrocarbon, methane,
and carbon monoxide concentrations of each zero air cylinder were
determined by analysis on a Beckman Model 6800 air quality chromatograph.
The results of the analyses are as follows:
Supplier
Matheson
Scott Research
Laboratories
Linde
Scientific Gas
Products
Certification by
Manufacturer
THC as CH^
<0.01 ppm
THC < 0.1 ppm
CO < 0.1 ppm
THC as CH^
<0.1 ppm
THC as CH^
<0.1 ppm
CO < 1.0 ppm
THC
(ppm
0.08
Analysis Results
CH
4
m)
0.06
CO
0.22
0.20 0.10 0.13
0.18 0.08 0.63
0.20 0.08 0.67
The Matheson cylinder was utilized for zeroing the two hydrocarbon units
evaluated during this investigation
109
-------
6.2 CALIBRATION PROCEDURE
A general procedure applicable to the dynamic calibration of any
analyzer was utilized in this investigation. Procedures and features
developed during the previous Los Angeles Study and in the St. Louis
2 3
Study, Phase I and Phase II, ' such as mode switch inputs describing
instrument operational status and automatic entry of calibration data on
magnetic tape were combined with on-line computer computation of transfer
equations by linear regression analysis of calibration data and processing
of air quality data in quasi real-time.
The basic step^wise procedure employed for dynamic calibration of all
air quality analyzers was as follows:
(1) Verify operational status of each analyzer prior to
beginning calibration.
(2) Connect instrument inlet line or instrument calibration
inlet line, as the case may be, to the manifold of the
calibration apparatus or for hydrocarbon and carbon
monoxide instruments, directly to cylinders containing
calibration gas.
(3) Allow instrument to sample zero air (i.e., air minus the
pollutant of concern) for a period of time sufficient to
establish a valid zero output. Indicate the proper
manual entry status code zero and average the instrument
output for zero input concentration for at least
15 minutes.
(4) Introduce a pollutant calibration concentration equal to
approximately 80% of the operating range and adjust the
span of instrument as required upon initial setup of the
instrument. This adjustment is normally required only
upon initial setup of an instrument or if excessive span
drift occurred during the evaluation period. Minor
adjustments in the span of each instrument can be performed
by the on-line computer easier than by manual adjustment of the
110
-------
span control knob, except for cases where drastic
changes in span occurred. Omit Step (4) except on
initial setup of analyzer.
(5) Introduce successive pollutant calibration concentrations
of 10, 20, 40, 60, and 80% of the operating range of the
instrument being calibrated. Allow sufficient time to
establish a valid instrument output for each calibration
concentration, and average the instrument output for that
input calibration concentration for at least 15 minutes.
Indicate the proper manual entry status codes for multi-
point calibration and proceed to the next higher calibration
concentration and repeat the sequence of events for multi-
point calibration.
(6) Return the instrument inlet line to the ambient air sampling
manifold and compute the transfer equation, which relates
pollutant concentration input to instrument voltage output,
for each instrument. This function was automatically
computed at the end of each calibration by the on-line
mini-computer employed during this instrument evaluation
program.
The frequency of calibration performed during this investigation varied
from daily to weekly to biweekly to monthly depending on the type of
instrument being evaluated and on the type of test procedures being run.
The frequency of calibration varied from daily to every other day when
short-term zero and span drift was being determined, while the frequency
of calibration might have been weekly for long-term drift (monthly basis).
All calibration concentrations utilized in this investigation were
generated using the apparatus and procedures described in the preceding
section.
Ill
-------
7,0 SUMW OF IICTWENT PERFORMRNCE
7.1 EVALUATION TESTS
In determining performance specifications, it was of prime importance
to recognize a properly operating instrument as opposed to a malfunctioning
one. With new instrumentation, especially gas chromatographs, it was
considered essential to have an experienced factory representative assist
in the start-up and familiarization phase. This procedure was generally
followed for most analyzers and aided in the assurance of a properly
operating analyzer. During this evaluation several analyzers were returned
to the manufacturers for repair and checkout prior to performance testing
due to initial problems of instability, drifting, and erratic signal output.
Standard tests or those tests most widely accepted were used insofar
as possible; however, for some performance characteristics (i.e., minimum
detectable limit), existing test procedures are inadequate. Some
test procedures, for example, recommend using test atmospheres equivalent
to concentrations of two or three times the noise level. The accuracy with
which test atmospheres of extremely low pollutant concentration can be
generated at the present state-of-the-art is insufficient to evaluate minimum
detectable limits. In this test procedure it was possible to statistically
predict the minimum detectable limit more consistently than to determine
the minimum detectable concentration by generating low concentration test
atmospheres.
Most manufacturers' performance specifications tend to converge to the
same value, typically 1 or 2% of full scale, for many performance factors.
This indicates that performance specifications are more easily printed than
demonstrated. Test procedures and statistical methods can influence
results and cause resulting conclusions to deviate significantly from the
manufacturer's specifications. Industry would do well to state and describe
fully the test procedures used in performance testing.
112
-------
7.1.1 Stability - Zero and Span Drift
The performance tests for stability were conducted as outlined in
Section 5.2.1. The tests were conducted during a minimum time period of
12 days; however, long-term drift data were recorded intermittently during
idle periods between tests or in conjunction with other tests that allowed
unadjusted status of zero and span controls. The following data in
Table 7.1 show zero and span drift, standard deviation and mean as percent
of full scale from consecutive calibrations obtained on a daily basis
except for week-ends and holidays.
Table 7.1. DRIFT TEST RESULTS
Zero Drift/Week (%) Span Drift/Week (%)
Evaluation Test Result Evaluation Test Result
Instrument Pollutant S.D. X S.D. X
Bendix 8002
Dasibi 1003-AH
Bendix 8101-B
Bendix 8101-B
Meloy 520
Meloy 520
Rem 642
Rem 642
Thermo Electron 14
Thermo Electron 14
Bendix 8300
Meloy SA-185R
Tracer GC 270-HA
Tracer GC 270-HA
Tracer GC 270-HA
Andros 7000
Bendix 8501-5FA
Beckman GC 6800
Beckman GC 6800
Bendix GC 8201
Beckman GC 6800
Bendix GC 8201
°3
°3
NO
N02
NO
N°2
NO
N02
NO
NO 2
TS
TS
TS
so2
CO
CO
CO
THC
THC
CH4
CH,
0.1
*
0.4
0,4
1.5
.5
*
*
0.3
0.2
5.0
0.8
0.3
0.2
0.2
0.1
0.4
1.8
1.9
0.9
0.3
0.7
0.03
*
0.05
-0.16
0.68
0.21
*
-0.01
0.06
1.82
-0.04
0.00
-0.05
0.06
-0.16
1.56
-0.42
0.14
-0.27
-0.04
-0.35
3.0
1.1
0.9
6.9
6.4
*
*
2.2
2.5
10.7
2.5
9.2
8.0
3.1
1.4
1.0
3.1
1.6
2.4
1.9
1.7
-0.50
*
-0.68
-0.62
2.80
3.35
*
*
0.00
-0.80
4.18
-0.38
-2.4
-2.3
0.6
-0.56
0.14
0.78
-0.24
1.27
0.96
0.88
S.D. - Standard Deviation
X - Mean 113
* - Not Available for Testing
-------
7.1.2 Linearity
The linearity performance test was performed as outlined in
Section 5.2.2. The data summary in Table 7.2 lists the standard deviation
and mean value in concentration units of the vertical increments from data
test points to the line of best-fit determined by linear regression
analysis.
Table 7.2. LINEARITY TEST RESULTS
Instrument
Bendix 8002
Dasibi 1003-AH
Bendix 8101-B
Bendix 8101-B
Meloy 520
Meloy 520
Rem 642
Rem 642
Thermo Electron 14
Thermo Electron 14
Bendix 8300
Meloy SA-185R
Tracor GC 270-HA
Tracer GC 270-HA
Tracor GC 270-HA
Andros 7000
Beckman GC 6800
Bendix 8501-5FA
Beckman GC 6800
Bendix GC 8201
Beckman GC 6800
Bendix GC 8201
Pollutant
°3
°3
NO
N02
NO
N02
NO
NO 2
NO
N02
TS
TS
TS
so2
CO
CO
CO
THC
THC
CH4
CH,
Standard
Deviation
(ppm)
0.00051
0.00075
0.00890
0.00100
0.02700
0.00620
0.02800
0.01100
0.00390
0.00330
0.00430
0.00360
0.00428
0.00448
0.00099
0.32000
0.01600
0.32900
0.01800
0.02100
0.00585
0.01620
Mean
(ppm)
0.0000220
0.0000310
0.0000510
-0.0001100
-0.0056000
0.0000052
-0.0000068
0.0000025
-0.0002000
0.0000077
-0.0003000
-0.0003000
0.0007500
0.0002000
0.0000000
0.0038700
-0.0010700
0.1300000
0.0023000
0.0041000
-0.0002200
0.0032000
114
-------
7.1.3 Precision
Precision performance tests were performed as outlined in
Section 5.2.3. Precision is defined as the variation about the mean of
repeated measurements of the same concentrations. The variation in
Table 7.3 is expressed as one standard deviation about the mean which
includes both instrument and calibration system errors.
Table 7.3. PRECISION TEST RESULTS
Instrument
Bendix 8002
Dasibi 1003-AH
Bendix 8101-B
Bendix 8101-B
Meloy 520
Meloy 520
Rem 642
Rem 642
Thermo Electron 14
Thermo Electron 14
Bendix 8300
Meloy 8A-185R
Tracor GC 270-HA
Tracer GC 270-HA
Tracor GC 270-HA
Andros 7000
Bendix 8501-5FA
Beckman GC 6800
Beckman GC 6800
Bendix GC 8201
Beckman GC 6800
Bendix GC 8201
Pollutant
°3
°3
NO
N02
NO
N02
NO
N02
NO
N02
TS
TS
TS
so2
H2S
CO
CO
CO
THC
THC
CH,
CH,
Standard
Deviation
(ppm)
0.0030
0.0040
0.0030
0.0021
0.0069
0.0073
0.0150
0.0040
0.002,9
0.0056
0.0110
0.0050
0.0020
0.0030
0.0080
0.0460
0.2040
0.0680
0.0120
0.0330
0.0100
0.0360
Mean
0.395
0.357
0.403
0.401
0.397
0.405
0.388
0.400
0.404
0.408
0.135
0.126
0.136
0.138
0.238
4.190
9.710
9.830
2.340
4.220
2.160
4.270
115
-------
7.1.4 Minimum Detectable Limit
The minimum detectable limit tests were performed as outlined in
Section 5.2.4. The minimum detectable limit is dependent upon many of the
other performance parameters. Both the test procedure and the method of
evaluating test data will often reflect a dependency upon other performance
factors in varying degrees. The method used in this evaluation is related
to precision of the method, linearity, number of calibration points (8 to
10 points), range of operation, mode of partitioning the calibration points,
mean value of the calibration concentrations (i.e., should be as low as
practical), and replicate measurements performed on ambient samples.
The evaluation approach used in assessing the minimum detectable limit
was conducted with the following guidelines:
(1) All analyzers would be tested for each pollutant on
identical ranges, or as nearly so as feasible, being
aware that a few instruments would undergo tests at
ranges almost a decade higher than the most sensitive
range;
(2) All analyzer signal outputs would behave linearly with
pollutant concentration, and calibration data yield
readily to linear regression analysis for a best-fit
line;
(3) Output data would be converted to physical units
2
(i.e., ppm or yg/m ) using a transfer equation from the
regression analysis;
(4) The appropriate operation and test range would be selected
for measuring non-urban pollutant concentration levels for
each pollutant; and
(5) The data acquisition system signal handling capability
could function reliably with signal voltages less than
0.1 millivolt.
116
-------
Analyzers measuring each pollutant underwent identical tests insofar as
possible. Confidence levels of 90% and 60% were applied to all analyzers.
This was satisfactory for 0_, NO, N02, SO^, H-S analyzers and respective
calibration systems; however, hydrocarbon and carbon monoxide calibration
standards tended to be the limiting factor in the evaluation. To alleviate
this problem, standards were selected on a relative basis from an original
lot of approximately 20 cylinders using data from all analyzers under test.
The minimum detectable limit for each analyzer in Table 7.4 should not
be interpreted to mean that a particular analyzer cannot produce an
Table 7.4. MINIMUM DETECTABLE LIMIT TEST RESULTS
Instrument
Bendix 8002
Dasibi 1003-AH
Bendix 8101-B
Bendix 8101-B
Meloy 520
Meloy 520
Rem 642
Rem 642
Thermo Electron 14
Thermo Electron 14
Bendix 8300
Meloy SA-185R
Tracer GC 270-HA
Tracor GC 270-HA
Tracer GC 270-HA
Andros 7000
Bendix 8501-5FA
Beckman GC 6800
Beckman GC 6800
Bendix GC 8201
Beckman GC 6800
Bendix GC 8201
Pollutant
°3
°3
NO
N02
NO
N02
NO
N02
NO
N02
TS
TS
TS
so2
H2S
CO
CO
CO
THC
THC
CH4
CH4
Concentration
90%
0.005
0.008
0.031
0.005
0.070
0.025
0.118
0.027
0.016
0.009
0.008
0.007
0.016
0.016
0.006
0.740
0.820
0.120
0.052
0.082
0.040
0.140
117
(ppm)
60%
0.003
0.004
0.016
0.003
0.034
0.012
0.059
0.013
0.008
0.005
0.004
0.004
0.008
0.008
0.003
0.370
0.410
0.060
0.026
0.041
0.020
0.070
Output
Signal Level (90%)
(Millivolts)
0.59
5.67
17.10
1.50
87.50
18.90
94.40
24.80
1.25
0.69
0.41
21.90
56.80
52.30
13.40
2.56
6.51
72.90
26.20
38.30
63.75
63.79
-------
observable response to a concentration level less than the minimum detectable
limit as determined by these tests. It should be noted that some analyzers
produced an observable response different from the zero reading for estimated
concentrations levels in the range of 1/2 to 1/10 of the minimum detectable
limit. These concentration levels were estimated by linear extrapolation
of calibration data. Generation and transport of test atmospheres containing
extremely low concentration levels are difficult to maintain with a signif-
icant degree of confidence. Tests were conducted using the Bendix NO , TS,
X
and Tracor analyzers to determine the minimum concentration level that would
yield a response different from the zero air measurement level. The
analyzers were flushed with zero air and with low pollutant concentrations
for one hour prior to attaining a stable reading. The test results for the
Bendix NO , TS and Tracor S09 analyzers are shown as follows:
X ^L
Test Point //I Test Point #2 Test Point #3
Bendix
Bendix
Tracor
0.
N0x
TS
so2
005
*
Cone.
0
0
0
ppm (NO)
S.D.
1.0
0.6
0.02
= 6.12
Mean
0.4
1.6
1.5
yg/m
Cone.
3.1
13.5
13.5
*
S.
1.
1.
3.
0.010
D. Mean
2
1
7
ppm
4.6
4.7
2.6
(SO.)
*
Cone.
6.1
15.8
39.5
= 26.2
S.D.
1.3
0.9
1.0
Mg/m
Mean
5.1
6.7
8.6
o
* - Estimated Concentration (yg/m )
The mean values of the concentration levels are indicative of the non-zero
pollutant levels; however, the large standard deviations exhibited show
considerable overlapping of the analyzer readings.
7.1.5 Response (Lag, Rise, Fall time)
The response times were determined using the procedure outlined in
Section 5.2.5. This procedure is applicable only for continuous analyzers.
For semi-continuous analyzers, such as NO,, instruments and gas chromato-
graphs, the cycle time is reported as lag time. Response time test
results are shown in Table 7.5.
118
-------
Table 7.5. RESPONSE TIME TEST RESULTS
Instrument
Bendix 8002
Dasibi 1003-AH
Bendix 8101-B
Bendix 8101-B
Meloy 520
Meloy 520
Rem 642
Rem 642
Thermo Electron 14
Thermo Electron 14
Bendix 8300
Meloy SA-185R
Bendix GC 8700
Bendix GC 8700
Bendix GC 8700
Tracer GC 270-HA
Tracor GC 270-HA
Tracer GC 270-HA
Andres 7000
Bendix 8501-5FA
Beckman GC 6800
Beckman GC 6800
Beckman GC 6800
Bendix GC 8201
Bendix GC 8201
Pollutant
°3
°3
NO
N02
NO
N02
NO
N02
NO
TS
TS
TS
SO 2
H2S
TS
so2
CO
CO
CO
THC
CH4
THC
CH,
Lag
(sec)
4.8
20.0
5.0
60.0
10.0
25.0
6.0
5.0
110.0
4.5
3.0
300.0
300.0
300.0
225.0
225.0
225.0
5.0
39.8
300.0
300.0
300.0
215.0
215.0
Rise
(sec)
10.2
15.0
15.0
9.0
10.0
30.0
20.0
__
__
22.0
85.0
-~
Fall
(sec)
10.2
15.0
9.0
10.2
6.0
20.0
«._
_ _
22.0
85.0
__
Remarks
Cycle Period
NO Only
Cycle Period
NO Only
Cycle Period
NO Only
Cycle Period
NO, NO Only
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
Cycle Period
119
-------
7.1.6 Noise
The performance tests for noise were conducted as outlined in
Section 5.2.6. The noise tests were performed while the analyzers operated
in their normal mode of operation. The oxides of nitrogen analyzers were
operated in cyclic mode (i.e., periodically yielding NO- data). In this
mode of operation the signals receive considerable integration before
being stored in the memory circuits. The output from the memory circuit
may be free from noise; however, the analyzer may be very noisy when
operated in continuous mode with the signal bypassing the integrator and
memory circuits. With gas chromatographs, the chromatographic signal may
be noisey but may not be readily seen after the peak is stored in memory.
For this reason, noise tests were not performed on gas chromatographic
analyzers, both hydrocarbon and sulfur units. Noise test results are
shown in Table 7.6.
7.1.7 Line Voltage Variation
Line voltage variation tests were performed only on total sulfur
and oxides of nitrogen analyzers as outlined in Section 5.2.7. Each
analyzer subjected to the test was initially operated and calibrated at
a regulated line voltage of 117 volts. Each analyzer was then operated
at 105 and 125 volts for a 16-hour stabilization period at each level and
recalibrated. The calibration curve of the analyzer was then observed
for offset and slope changes.
Variation in line voltage from 105 to 125 volts had no measurable
effect on zero baseline for either the Bendix or Meloy total sulfur
analyzers. Span sensitivity for both analyzers appeared to increase
approximately 1% between the low voltage level (105 volts) and the two
higher levels (117 and 125 volts).
The Bendix and Thermo Electron N0-N0_-N0 analyzers were operated on
£. X
105, 117, and 125 ac line voltages on consecutive days with a 16-hour
equilibration period overnight. The test employed a three-point dynamic
calibration at each of the line voltage settings. The calibration curve
of the analyzer was observed for offset and slope changes. The maximum
120
-------
Table 7.6. NOISE TEST RESULTS
Instrument
Bendix 8002
Dasibi 1003-AH
Bendix 8101-B
Bendix 8101-B
Meloy 520
Meloy 520
Rem 642
Rem 642
Thermo Electron 14
Thermo Electron 14
Bendix 8300
Meloy SA-185R
Pollutant
°3
°3
NO
NO
NO
NO
TS
TS
Mean
Concentration
(ppm)
0.000
0.400
0.000
0.400
0.000
0.400
0.000
0.400
0.000
0.400
0.000
0.400
0.000
0.400
0.000
0.400
0.000
0.406
0.000
0.400
0.00025
0.2044
0.00025
0.2044
Noise
(ppm)
0.0002
0.0003
0.0005
0.003
0.0005
0.0005
0.0005
0.0020
0.0020
0.0035
0.0050
0.0090
0.0060
0.0100
0.0020
0.0070
0.0005
0.0005
0.0005
0.0005
0.0003
0.0016
0.0006
0.0022
121
-------
slope change was less than 0.4% for each of the outputs (NO, N0«, NO )
^ X
but did appear to change proportionately with line voltage changes. The
Meloy NO-N00-NOV analyzer was tested under the same conditions as
£. A.
mentioned in the foregoing paragraph. Operation at 117 and 125 volts
appeared to be typical performance; however, at 105 volts the zero
immediately went negative, equivalent to 0.06 ppm, and no response was
observed for NO or N0~ concentrations of 0.1 and 0.4 ppm. The probable
cause of the malfunction was failure of the ozonator to produce ozone
required for the reaction. When operating at a normal line voltage of
117 volts, the analyzer responded properly. No voltage variation tests
were performed on the Rem analyzer, since it was removed from the
evaluation prior to this performance test.
Minor excursions or variations in line voltage (105 to 125 volts)
appear to have an insigificant effect on the performance of most of the
analyzers evaluated; however, for optimum performance the use of a voltage
regulation system capable of maintaining line voltage at 117 + 2 volts
is recommended.
7.1.8 Sample Flow Rate Test
Sample flow rate tests were performed according to the test
procedure described in Section 5.2.7 on two analyzers (i.e., Bendix and
Meloy total sulfur analyzers). These analyzers were selected for this
test, since dilution air is added to the exhaust stream from the detector
and maintaining proper flow rates has presented problems in the past.
Each analyzer was calibrated at the manufacturer's recommended flow rate
of 200 cc/min, and the flow rate was adjusted + 10% while sampling a test
concentration of sulfur dioxide. Results of this test indicate a signif-
icant decrease in response with a 10% increase in flow rate for both
analyzers. This condition is usually not observed with field use of
analyzers, in that flow rate normally decreases with time. Decreasing
the flow rate by 10% had a much greater effect on the Bendix analyzer
than it did on the Meloy. The minor difference in the indicated value of
122
-------
the calibration concentration for both analyzers at a flow rate of
200 cc/min can be attributed to analyzer drift between calibration and
performance of the flow rate test.
Sample flow rate tests were not performed on several instruments due
to the nature of their principle of operation or the flow system employed.
Analyzers utilizing critical orifices to regulate or maintain a constant
flow rate normally do not experience flow problems. The exception to this
occurs when the orifice is plugged with particulate or the vacuum pump
fails to maintain the proper pressure drop across the orifice. All
chemiluminescent NO-NO -N02 analyzers, with the exception of the Rem unit,
operate with critical orifices and at high vacuum within the reaction
chamber. Variations in cell pressure or vacuum of 10 to 20 cm of Hg
would not affect the flow rate through the instrument but can and do affect
significantly the sensitivity of the Bendix, Meloy, and Thermo Electron
analyzers. Tests indicated that the response of these analyzers is
directly proportional to cell pressure (absolute pressure) and that the
response decreases approximately 102 per 2.54 cm of Hg decrease from normal
operating pressure (i.e., absolute pressure). Sample flow rate test results
are shown in Table 7.7.
Table 7.7. SAMPLE FLOW RATE TEST
Meloy (TS) Bendix (TS)
Flow Rate (cc/min) Concentration Reading (ppm) Concentration Reading (ppm)
190 0.138 0.169
200 0.135 0.140
220 0.103 *
*
Instrument has very low response at flow rates exceeding 200 cc/min.
123
-------
7.2 OPERATIONAL SUMMARY
Operational data are summarized in Table 7.8 for each analyzer
evaluated during this program. Since individual analyzers were available
for varying periods of time during the evaluation period, the significance
of the data are somewhat questionable for comparison purposes. The Dasibi
ozone analyzer, for instance, was available for one week of evaluation,
whereas the Bendix ozone analyzer was operational for six months. Therefore,
it would be unfair to try to compare percent operational time and failure
rates for instruments with unequal evaluation periods. The data presented
in Table 7.8 should be interpreted individually on the basis of percent
operational time, failure rate, and type of failure. For this evaluation,
operational time is divided into the following categories:
(1) percent operational time, and
(2) percent downtime.
Percent operational time includes all categories or operational status
other than downtime, which includes routine maintenance, awaiting repair,
or repair. This includes calibration, performance testing, special tests,
and"availability for ambient air monitoring when tests were not being
performed.
It should also be pointed out that the data presented in Table 7.8
are results of an evaluation of one unit assumed to be a fair representative
of that particular model. Also, consideration should be given to the fact
that most of the mechanical and electrical failures experienced are of the
type that can be corrected by redesign or replacement of the component part
and are not basic flaws in the measurement principle itself.
7.3 SUBSYSTEM COMPONENT EVALUATION/RECOMMENDED MODIFICATIONS
TO IMPROVE PERFORMANCE
To meet the objectives of this'program and adequately evaluate the
performance of the analyzers under investigation, it was necessary to
separate the analyzers into three subsystems, to look at the contribution
of each subsystem to the total performance of the analyzer, to identify
124
-------
JH
pi
3
D
§
5
W
,j
^3
a
0
l2
w
CM
o
oo
I*"*
CU
rH
XI
cd
H
<4H
O * M M
Cd T) XI 00 X!
cd
CMOCMin rH CO rH O CM CN CN
oomo o o o ~a- o o o
^omm - vo c^ r*^ o\ o> cr\ vo c^
rH
or->min mvoooo-*oo,I| 55-H >,giH B
o-HTJoCo ti o 3 t> 3
COOCrHCUSSBCrHCJCCJ
a) cd cu cu xi eu cu cu Q) cu cu
OQ Q ffl S H PCiPQSPQPQpC)
r-t
00
00 00
M
o
CJ
CM CM in O
o o o o
o o o -»
rH rH CM
o o o o
O O O vo
^^ O^ 00 O^
O f-» CM O
in t^ o o\
CM
/^^ y-x
CM CM
0 O
CO CO
CO CO
CM CM
pFJ JlJ
1 1
CO CO
8 8 £ £
co X K X
O iH O TH
M -O O TJ
TJ C Cd C
Cj CU M fl)
< m H eq
cu
3
rH
H
cd
<4H
a)
3 0)
rH rH
CU -H rH
M Cd O
3 UH IH a)
rH 4J M
H ^ a c 3
cd M l-i O rH
l*H 4-1 3 CJ -H
H rH Cd
rJ 3 iH -l I4H 3 IH
cd -H 4-i O
eu cj In cd 4J
jc cu n cd
cj T) cu 3
U -H -H O, 4J
4J O CO 8 Cd
1 1 ti lj
cu 4-i cu a)
> o > c >
C (U rH (U rH
O rH Cd > Cd
0 CU > O >
1 1 1 1 1
14H 00X1 -H M-i
^
a
a.
3
co
cu
cu 3
> rH
O -H
P. cd
>4H
a)
XI rH
3 cd CU
4J CJ IH
H 3
S C rH
CM cd -H
*^ xi cd
O *4H
cu >> cu
M 4-i a M
3 -H CU "-^ CU
rH rH IH C! S
iH -H 3 OO C
cd XI rH iH O
>4H Cd iH CO <4H
4-1 cd eu co
i-t co IM -a c
rH C Cd
01 iH D. 00 M
CU 3 iH
M 4J (X M IH
o id cu o
4-1 K 9 CU 4->
U 9 C cd
CU S 3 -H C
4J O O 00 O
CU rH Cd Pi N
T) <4H >
-------
weaknesses and general deficiencies in each subsystem, to recommend
modifications and/or design changes to correct deficiencies whenever
possible, and to document and report these findings. From the onset of
this program, it was obvious that a thorough subsystem component evaluation
was not within the scope of this work, due to time and money constraints.
The optimum approach would be to divide each analyzer into three basic
component subsystems (i.e., gas flow system, detector system, and signal
processing system); to examine subjectively the components of each sub-
system with special attention being given to anticipated effects on the
total performance of the analyzer; to perform testing to validate
subjective judgment on identifiable weaknesses and/or deficiencies in the
design of a particular component and report test results to the Project
Monitor and the respective manufacturer; to recommend or suggest engineering
modifications or alternate approaches to upgrade the subsystem components;
to document these findings; and to retest and evaluate the effect of the
recommended design changes on instrument performance. Unfortunately, it
was not possible to accomplish all these tasks within the specified time
limits. In most cases all the objectives were completed with the exception
of retesting each analyzer to evaluate the effect of modifications on
instrument performance. Some recommended modifications were implemented
immediately, and their effect on improvement of instrument performance is
easily identifiable; others were completely ignored and never implemented
or are being implemented at the present time.
To report in detail the results of the subsystem component evaluation
would require several volumes. Therefore, a general description of the
major component systems for each set of analyzers will be presented in the
following sections, with attention being given to design consideration
rather than simple component failure. Section 7.6 will present what is
considered to be a composite list of general deficiencies encountered with
the analyzers evaluated in this investigation. This list is segmented by
subsystems (i.e., gas flow, detector, and signal processing), applies to
some but not all the analyzers evaluated, and is not all inclusive.
126
-------
7.3.1 Ozone Analyzers
Flow systems for ozone analyzers evaluated in this program can be
classified into two groups: a chemiluminescent analyzer with both ethylene
and sample flow rates and a UV-adsorption instrument with only sample flow.
The sample flow rate of the Dasibi ozone analyzer may be set over a rela-
tively wide range, typically 1 to 5 liters per minute, without introducing
any measurement errors. The sample flow rate of the Bendix ozone analyzer
must be accurately controlled to ensure calibration stability. An increase
in sample flow rate increases the available ozone per unit time to react
with the excess ethylene, yielding more chemiluminescence and, therefore,
increased signal output. The ethylene flow rate was varied + 10% of the
nominal flow rate (25 cc per minute) without significantly affecting a
given signal level of the analyzer. The ethylene and sample flow patterns
within the detector cell are designed to minimize deposits on the optical
window. This is accomplished with the ethylene and sample flows converging
at right angles in a plane parallel to the optical window and exhausting
opposite the PM tube cathode.
The Bendix analyzer has a chemical filter mounted on the rear of the
cabinet for providing zero air. This filter is connected to the sample inlet
line by a three-way, Teflon solenoid valve. The chemical filter contains
molecular sieves, indicating silica gel, and charcoal, and with proper
maintenance provides adequate zero air for the analyzer for routine checks.
The Dasibi analyzer flow path alternates the flow between the ozone
catalyst (to provide zero air) and the ambient air during the measurement
cycle. The analyzer is referenced to zero air in each cycle and automat-
cally compensates for lamp intensity output irregularities throughout the
cycle. The ozone catalyst is adequate for providing a reference sample for
several months.
The detector systems utilized in both UV-adsorption and chemilumi-
nescence analyzers appear to have adequate response and sensitivity. The
high voltage power supply and the photomultiplier tube share a common
thermostated housing. Power outages during the evaluation resulted in
127
-------
failure of the high voltage power supply. Although field repair of the
detector cell is not recommended by the manufacturer, an accessible test
point from the high voltage circuitry would facilitate fault isolation.
No major deficiencies were observed with the UV-detector system.
The signal processing in the chemiluminescence analyzer uses an
electrometer with a selectable time constant for filtering or smoothing
the signal. The time constants available are 1, 10 and 40 seconds. The
longer time constants are not essential to decrease noise of the analyzer
to an acceptable level, but provide smoothing of the signal to aid in
output sampling of instantaneous points by the data acquisition system.
The UV-adsorption analyzer measures the ambient sample and reference
sample for two very short periods with respect to the cycle period. The
ratio of the actual measurement period to cycle period is very small and
may be undesirable where finely detailed data are required (i.e., response
times less than 20 seconds).
7.3.2 Oxides of Nitrogen Analyzers
Flow systems for oxides of nitrogen analyzers evaluated during this
study can be divided into two types with respect to vacuum pump location
or reactor cell pressure. One type of analyzer utilizes the pump upstream
from the detector cell and operates near ambient pressure; the other type
employs the evacuation pump downstream from the reactor cell and, hence,
has cell pressures of approximately 23 to 30 in (58.4 to 76.2 cm) of Hg.
Each of the design considerations has its advantages and disadvantages,
some of which are listed as follows:
Ambient Pressure Reactor (REM Analyzer)
Advantages Disadvantages
(1) flow rate independent of ambient (1) larger sample required for
pressure changes or altitude existing detector design
(i.e., 600 cc/min)
(2) no oxygen cylinder required for (2) possible variable loss rate
ozonator of sample in sample pump
(3) desiccant change every 2 days
due to moisture in the air
flowing to ozonator
128
-------
Low Pressure Reactor (Bendix, Meloy, Thermo Electron)
Advantages Disadvantages
(1) small sample flow rate (1) response dependent upon cell
(100-200 cc/mln) pressure (approximately 10%
decrease per 2.54 cm of Hg
decrease)
(2) small air or oxygen flow (2) desiccant change every ten
rate (50-100 cc/min) for days (high R.H.)
ozonator
The use of air as a source of oxygen for the ozonator is satisfactory
and stable, providing the desiccant is changed properly and timely. The
replacement period is variable often by a factor of two (i.e., dry ambient,
20 day period, versus high relative humidity, 10 day period) and provides
additional logistics for a monitoring site. The use of oxygen cylinders is
also a logistical problem, but usually very infrequent, requiring replacement
twice a year.
The pressure along sample flow paths employed by low reactor pressure
analyzers varies significantly. In the Bendix analyzer low pressure is
maintained from the pump, ozone scrubber, detector cell and to each of two
capillaries used for flow control for ozone and sample. Therefore, a
minimum number of fittings or connections are exposed under vacuum
conditions. In the Thermo Electron analyzer low pressure is maintained
from the vacuum pump, accumulator, reaction cell, catalytic converter,
cycling solenoid valve, ozone capillary and sample capillary. The poten-
tial for development of leaks is about twice as great for the Thermo
Electron as for the Bendix analyzer.
Converters may use many materials, such as molybdenum, stainless
steel, carbon, and carbon-coated gold, for converting N0« to NO. Conversion
efficiency is typically 93 to 100%. The molybdenum converter used with the
Thermo Electron analyzer yielded the highest conversion efficiency lifetime,
with the Bendix carbon unit, Meloy converter, and Rem molybdenum converter
placing second, third, and fourth, respectively. The molybdenum converters
of Thermo Electron and Rem are easily rejuvenated to 100% using hydrogen
for a short period of time.
129
-------
The detectors and reaction chambers used by the various manufacturers
have some very subtle differencesfor example, background level vari-
ability for ozonator on and off conditions. The Thermo Electron and Rem
possess the lowest ozone contribution to background level, while the
Bendix has the highest. In addition to background level contributions
from ozone, the units were tested using zero air, very dry (dew point -30°F),
and slight relative humidity (1-4%). The Bendix unit yielded a greater
response for dry air than for air containing a slight amount of moisture.
Light leaks entering the detector cell presented problems in the
Thermo Electron analyzer. All analyzers except Thermo Electron utilized
stainless steel tubing in the near vicinity of detector parts. Teflon
tubing without proper shielding is notorious for conducting light great
distances. The Thermo Electron unit utilized two sizes of heat-shrinkable,
black tubing, using the larger size over tubing connectors and the smaller
size over Teflon tubing to shield it from ambient light. Poor quality
control/design allowed it to slip away from the tubing and fittings. This
deficiency was relatively easy to correct.
Signal processing can be subdivided into three distinct groups with
respect to range circuitry location and type of amplification used and
are as follows:
Analyzer
Bendix
Amplification
electrometer (0.5 - 5.0 ppm)
Individual ranges for
NO-NO-NO
Range
Meloy
Thermo Electron
Rem
chopper amplifier
demodulator (0.5 - 5.0 ppm)
range attenuator network
input to electrometer
(0.5 - 10.0 ppm)
range attenuator network
input to electrometer
(0.5 - 10.0 ppm)
control associated with
memory circuits.
Individual ranges for
NO-NO, -NO . Range
£. X
control associated with
memory circuits.
One range for all
NO-NO-NO
One range for all
NO-NO,-NO
£m X
130
-------
One significant difference in the analyzers is the feature of
individual ranges for NO-NO -N0_ under ambient monitoring conditions. NO
A. £*
concentration levels are typically 2 to 5 times higher than NO- concen-
tration levels. All analyzers utilize some averaging or integrating step
prior to sampling and hold operation of the memories while functioning in
the cyclic mode. When operating in the manual mode (i.e., NO or NO only),
X
the Bendix and Meloy exhibited a greater noise bandwidth than when operated
in the cyclic mode. Lack of adequate radio frequency (RF) shielding and/or
design of the ozonator in the Bendix analyzer allowed the RF to propagate
into the signal processing electronics and through rectification in the
transistor junctions, added and sometimes subtracted values from the actual
dc signal level.
Major problems encountered during this evaluation and other field
programs with chemiluminescent NO-NO -N02 analyzers involved the gas flow
system (i.e., maintaining the required cell pressure), electronic component
failures, generation and destruction of high concentrations of ozone, and
the cyclic mode of operation where rapid changes in the ratio of NO/NO- in
ambient air can yield negative N0_ measurements. These problems were
pointed out to the various manufacturers of chemiluminescent NO-NO -NO-
analyzers as they occurred, and appropriate recommendations to alleviate
the problem were made.
7.3.3 Sulfur Analyzers
Two measurement methods, which utilized the same detection principle,
were evaluated.. Both a gas chromatographic flame photometric detector
(GC-FPD) and a flame photometric detector (FPD) were tested for total
sulfur, sulfur dioxide, hydrogen sulfide and total sulfur.
The flow system of GC-FPD analyzers usually consists of sample loop,
pump, valves, stripper column, analytical column and detector. Two sample
loops may be employed: one for total sulfur and one for the analytical column.
The stripper column is used in automated gas chromatographs designed for
continuous monitoring. The stripper column removes from the sample aliquot
interferences (i.e., H-O, methyl mercaptans, etc.). After the components
131
-------
of interest elude the stripper column, it is subsequently back-flushed and
vents to the atmosphere most of the trapped interferences. For some
interferences, the process is nonreversible, giving the column a finite
lifetime. The lifetime of the column is dependent on contamination level
of the environment sampled, the frequency of analysis, packing material,
size of column, and sample size.
The analytical column used in a gas chromatograph will undergo a
change in its characteristics with time. Column deterioration is usually
recognized by a closing of the space separating the chromatographic peaks,
concurrent change in retention times, peak overlap, base broadening, or
some combination of these symptoms.
It is imperative to maintain proper operational documentation as
follows:
(1) periodically run chromatograms comparing retention times
with prior run;
(2) place in record book for future reference;
(3) record data such as flow rate, chart speed, pressure,
oven temperature, pollutant concentration, and any
additional pertinent parameters; and
(4) record all modifications made by the manufacturer.
The FPD requires proper preventive maintenance for reliable, trouble-
free performance. The most important precautions include:
(1) provision for continuous removal of all combustion
products including water vapor, and
(2) utilization of sulfur-free carrier gas.
7.3.4 Hydrocarbon Analyzers
The flow system of gc-flame ionization hydrocarbon analyzers usually
consists of a pump, sample loop, valves, stripper column, analytical column
and detector. Most manuals stress the importance of cleanliness which will
be reiterated here. A dirty cylinder outlet, contaminated regulator, or
service tubing can foul an analyzer and contribute to poor performance.
Each gas cylinder used should be equipped with a clean, hydrocarbon-free,
132
-------
two-stage regulator with appropriate indicating gauges. Each regulator
should have a metallic, not elastomeric, diaphragm and have a cleanliness
equivalent to materials used in handling liquid oxygen. The use of all
new tubing is strongly recommended for all connecting gas lines. The
tubing material type should be compatible with the pollutant of interest
and the accompanying gases. The material type should not be a source of
the contaminant of interest nor an adsorber. Polyethylene tubing is
frequently used because of its flexibility, strength, and inertness. In
some applications, the preferred type is new "refrigeration grade" copper
tubing sealed at the ends. Generally, stainless steel tubing is less
desirable, as it contains hydrocarbon contaminants, necessitating scrupulous
cleaning before installation. Precleaned "chromatographic grade" stainless
steel tubing, available at premium price, is recommended if stainless is to
be used. Stainless steel tubing of other than "chromatographic grade" may
be used if properly cleaned. The following cleaning procedure is
recommended:
(1) Wash with hydrocarbon solvent such as acetone or
trichloroethylene.
(2) Rinse with distilled or deionized water.
(3) Connect tubing to the regulator on a cylinder
containing nitrogen using a flow rate of 1 to
10 liters per minute.
(4) Heat the tubing with a propane or natural gas torch
working the heat source slowly from the regulator end
to the open end of the tubing. This will remove
contaminants from the inner walls of the tubing and
expel them at the open end.
Caution: Never heat lines when connected to a gas
chromatograph.
Tubing of questionable cleanliness and new tubing should be cleaned
with the exception of "refrigeration" and "chromatographic grade" tubing.
It is customary to employ release agents such as oil for metal tubing or
talcum for plastic tubing during the manufacturing process.
133
-------
Tubing such as copper and plastic may be cleaned as follows:
(1) Flush tubing with methylene chloride.
(2) Rinse with acetone.
(3) Purge with an inert gas such as helium or nitrogen.
(4) If cleanliness is questionable, zero-air can be passed
through it to the analyzer for a reading and then
bypassed. A difference in analyzer reading indicates
further cleaning is necessary.
A stripper column is used in automated gas chromatographs designed for
continuous monitoring. The stripper column removes from the sample aliquot
interferences (i.e., H-O, CCL, and hydrocarbons other than CH,). After the
components of interest elude the stripper column, it is subsequently back-
flushed and vents to the atmosphere most of the trapped interferences. For
some interferences, the process is nonreversible, giving the column a finite
lifetime. The lifetime of column is dependent on the contamination level
of the environment sampled, the frequency of analysis, packing material,
size of column and sample size. Failure to replace an expended stripper
column will result in the eventual contamination of the analytical column.
A degradation of the stripper column and analytical column can yield
similar failing symptoms. It is generally recommended to replace the
stripper column first if there is doubt about the analytical column. If
difficulty is encountered in the analytical column, it is advisable to
replace both columns at the same time.
The analytical column used in a gas chromatograph will undergo a change
in its characteristics with time. The rate of change may be dramatic,
occurring within a few days, or insignificant, over a year's time. Column
deterioration is usually recognized by a closing of the space separating the
chromatographic peaks, concurrent change in retention times, peak overlap,
base broadening, peak flattening and tailing of the top of the chromatogram,
or some combination of the symptoms.
It is imperative to maintain proper operational documentation as
follows:
(1) periodically run chromatograms comparing retention times
with previous run;
134
-------
(2) place in record book for future reference;
(3) record data such as chart speed, flow rates, pressures,
oven temperature, pollutant concentration and any
additional pertinent parameters.
The flame ionization detector requires the proper preventive
maintenance for reliable trouble-free performance. The most important
precautions include:
(1) provide for continuous removal of all combustion
products including water vapor;
(2) prevent the contamination of any component with
hydrocarbons even in trace amounts;
(3) utilize contaminant-free hydrogen and burner air.
Combustion products or other contaminants allowed to accumulate inside the
flame ionization detector form electrical leakage paths between the
collector and burner contact. Slight traces of such contaminants can
degrade the performance sensitivity of an analyzer almost an order of
magnitude. For best performance, it is absolutely necessary that the flame
ionization detector be kept free of any kind of contamination.
Proper venting of the flame ionization detector is essential,
especially in small trailers or monitoring facilities to prevent pressure
disturbances from upsetting the flame. The opening and closing of outside
3
doors in relatively air-tight buildings (1000-4000 ft ) can produce
significant effects in the flame background signal level or can yield false
peaks. Detector venting to a protected area outside the building is
recommended for two reasons: the internal building static pressure upsets
are avoided, and should hydrogen "fail-safe" devices fail, the hydrogen will
be safely vented to the outside.
The electronics section in hydrocarbon gas chromatograph analyzers
includes the following subsystems: timer, amplifier, peak detectors,
memory storage, programming capability, fail-safe circuitry, temperature
controllers, automatic baseline zero capability, and power supplies. Timers
are generally divided into two categories: (1) line frequency divider
circuitry and (2) resistance-capacitance, time-constant-controlled voltage
135
-------
ramp. Both techniques are sufficiently reliable and accurate and easily
facilitate programming the various functions to an accuracy of one second.
Automated gas chroma tographs usually incorporate electronic zeroing
circuits to connect for baseline drift. The programming of the automatic
zero is critical and requires periodic checking to ensure it is occurring
at the appropriate time. Initiation of the automatic zero too early or
too late may result in an inaccurate measurement. Automatic zeroing too
early may not have allowed sufficient time for baseline recovery from
flow/pressure upsets subsequent to valve switching. Automatic zeroing
too late may result in zeroing on the leading edge of peak elution.
The programming of the peak detectors and associated memory circuitry
is somewhat less ciritical than automatic zeroing procedures but should
receive careful scrutiny.
The analyzer must be correctly calibrated if meaningful data are to
be obtained. Errors in labeled concentrations will be directly related to
ambient air measurements. Vendors of pressurized calibration mixtures tend
to emphasize preparation and certification capability of mixtures far
exceeding the specifications of the gas mixture evaluated when received.
In approximately a one-year comparison of about twenty cylinders on a
relative basis, three major discrepancies occurred: two cylinders contained
concentrations that deviated by exactly a factor of 2, and one cylinder
mixture contained a diluent of non-air (30% 0~, 70% N_).
Blended air mixtures that deviate from ambient air (21% O^, 79% N~)
are unacceptable and will yield incorrect calibration of total hydrocarbon
analysis. Oxygen concentration exceeding ambient air causes the analyzer
to read a higher total hydrocarbon value than is actually present in the
calibration mixture. Verification procedures ultimately rest with the user
and are not within the scope of this report.
No major problems were encountered with the operation of the Beckman
6800 after replacement of the analytical column with the updated version
prior to the evaluation period. Minor problems were encountered with
failure of electronic components and circuitry which required the attention
136
-------
of a component operator. Daily attention by a competent operator is
required for optimum performance.
A major problem, requiring an engineering modification, was
encountered with the Bendix ambient hydrocarbon analyzer and involved
installation of precolumn to remove moisture from the sample prior to
injection of the sample directly into the flame for the total hydrocarbon
measurement. Abnormally high non-methane hydrocarbon measurements were
being obtained and were attributed to moisture by Bendix personnel.
Installation of a scrubber column prior to the detection chamber appears
to have eliminated the problem. Additional tests are being run to verify
that the modification has indeed accomplished its purpose.
7.3.5 Carbon Monoxide Analyzers
Analyzers employing three different measurement principles were
evaluated during this study. The measurement principles are infrared
fluorescence photodiode detection, non-dispersive infrared capacitance-
microphone type detection, and gas chromatograph methanation with flame
ionization detection. These principles are described in detail in
Section 4.0.
All three measurement principles are tolerant to flow rate changes of
+ 20%; however, precautions should be observed to prevent pressurizing the
flow system above the normal pressure mode of operation.
The optical-flow systems of the analyzers vary significantly with
respect to losses in the transmission optical paths. The Andros infrared
fluorescence analyzer can tolerate an approximate 50% decrease in optical
transmission efficiency due to condensation, particulates in the sample,
infrared source intensity or detector response drift. The signal output
1 f\ 1 ft
is directly related to the ratio of CO to CO spectral lines.
The Andros carbon monoxide analyzer has an internal zero air system
which was compared periodically with an external zero air source. During
the evaluation ambient air containing an average of 1.0 ppm CO was passed
through the internal zero module used, and the module produced zero air
with 0.05 ppm or less of carbon monoxide. No degradation of the zero
module was observed during the evaluation period.
137
-------
Analyzers employing non-dispersive infrared capacitance-microphone
detection schemes are susceptible to condensation of particulates and
water vapor. Two sample handling deficiencies were noted for the Bendix
analyzer and are as follows:
(1) The input sample filter (8 micron) mounted within the
bowl failed to sufficiently remove the water during high
humidity periods and allowed condensation in the gas
filter mounted within an aluminum housing, external to
the analyzer.
(2) Replacement of the gas filter further downstream was
difficult and attributed to seizing of the threaded
portions of the aluminum housing. This made filter
element replacement nearly impossible.
Both problems were reported to Bendix for appropriate modifications.
The flow system for the gas chromatograph was outlined in detail for
the total hydrocarbon and methane analysis in the previous section and will
not be reiterated here except for pertinent facets of the carbon monoxide
measurement. The Beckman 6800 air quality chromatograph utilizes a
converter to reduce the carbon monoxide to methane in the presence of
hydrogen on a heated nickel catalyst. The conversion efficiency is
typically 70 to 100%. Current data indicate a decrease of approximately
5% in one year of operation. Converter efficiency appears to be constant
for CO concentrations up to 100 ppra. Converter efficiency was checked by
using known concentrations of methane and carbon monoxide and determining
the areas under the respective peaks. For equal concentrations it is
simply the ratio of the areas. If the converter efficiency is constant
between calibration period and not 100%, satisfactory and meaningful data
can be obtained by simply increasing the electronic gain of the amplifier.
Subsequent CO peak decreases from successive calibrations would warrant
checking the converter efficiency. Converter efficiencies appear to be
very reliable, and periodic checks are not recommended. With careful
maintenance and use of proper calibration and operating procedures, a
minimum detectable limit of 0.1 ppm was achieved, using the test procedure
specified in Section 5.2.4.
138
-------
The capacitance-microphone detector utilized in the Bendix analyzer is
marginally capable of making CO measurements with a minimum detectable limit
in the neighborhood of 1.0 ppm after a few weeks of operation. With careful
maintenance, sample conditioning, and use of proper calibration and
operating procedures, the minimum detectable limit may be reduced to below
1.0 ppm. This type of detector is very vibration and temperature sensitive,
and precautions should be observed for proper location of the analyzer and
insulation from vibrations and temperature extremes. Pumps and other
monitoring equipment can drastically interfere with optimum performance
of the analyzer and should be located accordingly.
The detector employed in the Andros isotope fluorescence analyzer is
very rugged and will easily tolerate hostile vibration environments such as
helicopters and mobile vans. The solid-state photodiode detector is
thermally stabilized at -30°C to achieve maximum sensitivity by a Peltier
cooler. With careful maintenance and use of proper calibration and
operating procedures, a minimum detectable limit of 0.7 ppm can be achieved,
using the test procedures specified in Section 5.2.4.
For convenience in troubleshooting, the Andros 7000 contains a
continuous self-check function which will generate a panel-indicated NO-GO
if any of the following occur:
(1) Excessive postamplifier signal;
(2) Insufficient postamplifier signal that results in
closed loop span's AGC drive voltage being too far
negative;
(3) Closed.loop span fault, which results in AGC drive
voltage being too far positive.
The self-check circuits are all part of the comparator assembly. In
addition to generating a panel NO-GO, the comparator has five diagnostic
indicators which include a GO as well as a NO-GO situation. There are
three other indicators which, if a NO-GO is displayed, indicate which of
the above abnormal conditions exist. The indicator lights on the panel
and diagnostic lights on the printed circuit board are intended to help
identify the probable cause of improper operation, as well as indicate
operating modes.
139
-------
In addition to the above, the automatic frequency control circuit
board contains a self-test switch, which sends a control pulse train to
the automatic gain control board. This causes an upscale span increment
of approximately 50% of full scale. This test procedure will only work if
Ifi 18
the optical system and the CO and CO signals coming from it are in
good order.
The GO/NO-GO indicator is the most advanced concept for subsystem
component check for air monitoring instrumentation on the market. Each
printed circuit board has an individual card guide and retainer. The elec-
tronic subsystem is adequately separated and shielded from the pump,
plumbing and internal zero. The Andros 7000 evaluated during this study
definitely does not have the production line appearance. The development
of subsystem component test points and checks is recommended for inclusion
on all future analyzers.
7.3.6 General List of Design Deficiencies/Considerations
Results of the subsystem component evaluation indicate certain
deficiencies or design considerations that are applicable or common to all
instrumentation with respect to the gas flow, detector, and signal
processing subsystems. This list of considerations is segmented by sub-
system, is presented in outline form, and is by no means all inclusive.
The reader is also advised that many of the design considerations presented
for one subsystem may also be applicable to other subsystems.
7.3.6.1 Gas Flow Subsystem
(1) Plumbing of sample handling system should be minimized in
length and diameter to enhance instrument response time and
reduce possible sample deterioration.
(2) Tubing, valves, and filters should be compatible with the
pollutant being sampled and measured.
(3) Tubing should have sufficient wall thickness to prevent
collapse or pinch-off at tubing bends in the instruments.
Operating temperature and the gas being handled should be
considered in selecting tubing used in the analyzer.
140
-------
(4) Hot and cold regions should be avoided in the sample
handling system.
(5) Sample handling scrubbers or conditioners that require
constant or frequent attention should be avoided. If
preconditioning is essential for proper operation, the
conditioning element should have a visible or useful
life status indicator.
(6) Sufficient monitoring gauges or indicators should be
used to warn the user of degradation of performance or
upcoming failure of vacuum pumps or other sample
handling devices.
(7) Cooling air for components internal to the instrument
should be filtered to prevent accumulation or buildup
of dirt and dust on electronic component parts.
(8) Avoid using incompatible materials where vibration,
temperature, voltage, stress levels, etc., would induce
failure (i.e., stainless steel housing for glass
capillary where vibration causes glass to fracture).
(9) Particulate filters, preferably Teflon filter elements,
should be included on all instrument inlet lines,
regardless of the pollutant being measured.
(10) Safety devices, such as automatic shut-off valves which
close upon loss of vacuum or AC-line voltage to the
instrument should be included for handling hydrogen,
oxygen, ethylene, etc.
7.3.6.2 Detector Subsystem
Detectors employing photomultiplier tubes as sensors should be
designed with provisions to:
(1) Check the HV with at least one accessible test point.
(2) Check the dark current by use of a shutter in front of
the sensor or a positive method for reducing
141
-------
luminescence to an insignificant amount; the latter
might be accomplished by passing a quenching gas into
the detection chamber.
(3) Check the gain of the detector assembly by means of a
built-in calibrated light source.
A desirable way to check the operation of a detector utilizing a
photomultiplier (PM) tube or other type of photo-sensor would involve the
use of a calibrated (or known) intensity light source with an output in or
near the wavelength(s) of interest. If it is located in the detection cell,
then the performance of the entire electro-optical system may be checked
out. This would indicate (indirectly) problems with buildup on the detec-
tion window, i.e., light transmission, and in the performance of the PM
tube. Assuming no change in the light transmission through the detection
window, the gain and signal-to-noise (S/N) ratio of the PM tube may be
ascertained.
The dark current is a direct function of temperature and any ambient
light. A means for recording dark current, under normal operating
conditions, is essential to the determination of malfunction in the PM tube,
a light leak in the detection, or some unwanted chemiluminescence reaction
or prior exposure to light. For example, exposure to ambient light
(without voltage applied) may require up to 48 hours before the dark current
has returned to its normal and stable level.
The implications of such testing procedures cannot be overemphasized,
provided standard techniques are used and the data are recorded for future
reference. Such data will be invaluable in the troubleshooting of indi-
vidual instruments and collecting data which will help indicate trends or
degradation rates of the instruments. Compilation of data from several
instruments over time will produce the data necessary to statistically
describe the distribution of the various performance parameters and allow,
to some degree, prediction of individual unit performance as well as
indicated individual component weakness or out-of-tolerance situations.
142
-------
7.3.6.3 Signal Processing Subsystem
(1) Card guides and retainers should be used for all printed
circuit cards.
(2) Circuit board contacts should be compatible with connectors.
Mechanical vibration of tinned contacts on printed circuit
boards mated with gold-plated connectors often yield
undesirable resistance, especially in series with control
potentiometers.
(3) Electronic subsystems should be adequately isolated or
shielded from devices such as valves, ozonators, ovens,
and other mechanical subsystems which may induce undue
stress from electromagnetic, vibrational, or thermal
sources.
(4) Heavy components should not be mounted on printed circuit
cards causing undue torque or bending moments.
(5) GO/NO-GO indicators are essential for subsystem checkout
and failure isolation. Future development of instrumentation
should demand incorporation of these indicators for all
feasible subsystem checkpoints.
(6) Instrument controls, such as potentiometers, should be
turn-counting with locking-type dials.
(7) Commercially available packaged operational amplifiers and
signal processing devices with demonstrated reliability are
preferable to home-built devices.
143
-------
8,0 RECOffENDATIONS FOR SELECTION OF INSTRUCTS
Operational requirements, definitions describing sensor specifications,
sensor specifications applicable to the special needs of the RAPS Program,
criteria for selection of instrumentation, and the test plan were presented
in Section 3.0 of this report. Standardized performance evaluation tests
were developed to evaluate selected instrumentation on an equivalent basis
(Section 5.0). Results of the performance evaluation tests are presented
in Section 7.0 along with pertinent comments, observations, and recommen-
dations concerning general deficiencies or considerations observed in the
design of selected instrumentation.
Results obtained from the performance evaluation tests were compared
to the required specifications and are presented in Table 8.1 for each
analyzer. Qualifications that should be considered prior to the inter-
pretation of these results are as follows:
(1) Evaluation results reflect the performance of a single
brand-name analyzer assumed to be representative of
that brand name; evaluation of a statistically
significant number of brand-name analyzers was not
possible;
(2) performance evaluation tests were designed and implemented
in a uniform manner whenever possible, so that no one
analyzer was discriminated against;
(3) test results presented are representative of the performance
of each analyzer during its best period of performance
(i.e., tests were sometimes interrupted due to failures,
modifications, etc., and required rerunning);
(4) test results were invalidated for any analyzer during
periods when the analyzer's performance was questionable;
(5) test results are dependent to a large extent on the
precision and accuracy of the calibration systems to
deliver or generate test atmospheres reliably over long
periods of time;
144
-------
cn
Z
O
U
w
PL,
O"
o
H
a
pi
H
§
CO
M
58
oo
0)
id
H
a
d« So
3 >* CO O
as < u M
O O Q 06
M " | 0 W
go ( g*
Pu PI W
; SB
§y
§So
"s? II
z- ~
M Q
J I W
cn Q
W
o
1 cn
I
! z Hix
M cn
H U
i u H cn
i
I
1 cn cn
S u
X U
u i cn
° i si
! 3*
1 -] H
1 3 W 0
u H cn
cn N
PC U
U
1 cn
1
-§§
M CO
t-4 1 J H
cn ^ i < en
M M 1 > U
CJ s-* W |H
* « CJ
2 M
S£ ggf
gfe £|~
Gz S
S S < M S
M H > H O
a a I w ** s
M U 1 CA CJ
ac8 i 2»
INSTRUMENT
r*-
§
o
CM
O
O
o
o
m
o
i
0
cn
o
**
cn
0
o
H
0
0
o
CM
8
0
8
0
in
§
o
cn
o
CM
a
a
-
§
O
CM
O
O
O
O
^
0
o
o
CM
O
o
0
o
0
o
*n
8
O
cn
o
m
8
M
m
M
§O
0
0 0
* CM
8 8
0 0
00 fN
0 0
1 1
rH <*
rH O
0 0
IN C-4
IA *O
0 rH
O 0
1 1
-» n to
o o
o o
0 0
REM 642 N02
THERMO ELECTRON 14 NO
o »
3 3
0 0
0 0
CN -»
8 8
O O
O CO
OO f-4
0 -*
1
m r,
CM 0
1-4
0 0
CM CM
vO CM
O 00
O iH
CM O
o m
0 0
o o
CM CM
O -*
o o
o o
o o
o o
0 0
o o
m m
8 8
o o
THERMO ELECTRON 14 N02
BENDIX 8300 TS
r~ <*%
o o
0 0
O 0
o o
O 0
CD
rr, sr
O CN
1 1
10 CM
(M a*
o o
CN CN
-» o
O 0
o o
i
CO rn
O 0
O 0
0
0 0
CN fN
§00
o
o
o o
§ 3
o o
m m
8 8
O 0
MELOY SA-185R TS
TRACOR GC 270-HA TS
en
8
O
1
0
en
CM
1
0
CO
o
CM
m
O
0
CM
O
O
o
o
CM
00
o
o
0
r-t
o
o
*n
8
O
CM
O
cn
1
O
CM
U
U
06
i
H
cn
8
O
1
0
*
o
_,
cn
o
CM
0
0
CM
O
0
CM
O
CM
cn
O
O
0
8
O
CM
g
o
cn
CM
o
CM
O
O
a:
H
145
-------
c
o
o
g
H
5!
o
M
PK
M
U
Ed
fr
cn
Q
O
H
to
H
to
H
w
oo
0)
iH
3
H
:RATIONAL
COD (DAYS)
o w
(X.
1?
s-
Id
§ ~
. E
g
s
CO ^-s
CJ '-'
?
W £
DETECTABL
RATION (PP
g g
M U
35 2
as o
i
i
i
i
!
t
!
1
1
1
I
I
I
1
i
j
|
i
I
!
i
i
i
i
j
i
j
i
i
i
i
j
fVALUATIOir
i
i
I
INSTRUMENT
-UATION
Id
c
%
*H
f?
P
Id
SB
M
>
CO
S
*
SUGGESTED
V)
RESULT
CO
Id
f-i
*4
Q
W
£3
g
H
3
CO
f?
H
CO
S
§
o;
V,
f
to
£i
8
r-
I
CO
<5
CO
3
PERIOD
o
[d
IX
CO
Q
d
IX
0
CO
i
0
W
CU
CO
IX
d
N
U
W
S5
CO
s
H
CO
H
0
Id
CO
g
1
U
M
ft
O
H
O
ft
O
o
o
o
o
rH
O
lO
O
1
^.
rH
0
fM
rH
O
1
^
0
o
2
o
r^
o
O
o
o
0
8
0
o
o
r-.
CO
i
«
o
CO
o
rH
O
0
o
o
o
rH
o
0
rH
O
CM
in
r-l
rH
o
CM
§
°
CO
o
0
CO
o
o
CM
o
o
0
o
o
o
0
o
g
v£»
CJ
o
g
«
o
CM
vD
0
o
o
o
oo
oo
o
^
-H
o
CM
tn
CO
o
1
r-
o
0
rH
rH
o
0
o
rH
o
o
o
0
-a-
g
tH
0
H
a
Q
146
-------
(6) resolution of data down to ppb levels is possible only by
digital data resolution and computer processing of data;
and
(7) test results depend to a large extent upon the design of
the experiment or test used to determine the instrument
performance specification; all data presented in Table 8.1
were obtained using the performance evaluation tests
described in Section 5.0.
Important observations that should be noted from Table 8.1
and discussed are as follows:
(1) A significant number of analyzers failed to meet the
performance criteria for minimum detectable concentration.
The minimum detectable concentration was determined using
a confidence level of 90%; however, if the confidence
level were reduced to approximately 60%, more than half of
the analyzers would meet the performance specification.
In addition to the above, an improvement in calibration
systems, especially the hydrocarbon and carbon monoxide
standards, would significantly improve the lower detection
limits.
(2) Most analyzers easily met the zero drift specification,
with only two analyzers demonstrating appreciable mono-
tonic drift. This can be observed by comparing the mean
value and the standard deviation. The larger the mean
value, the greater the departure from random drift.
(3) The number of analyzers meeting the span drift performance
standard were almost equally divided. Recent modifications
and incomplete test data on the Bendix sulfur analyzer
indicate a substantial improvement in its span drift
characteristics. The Tracor GC 270-HA analytical column
and detection systems are potentially capable of better
performance than demonstrated during this evaluation.
147
-------
With improved design and upgrading of the electronic
subsystems, span drift should be capable of being
reduced to acceptable levels.
(4) Over one-half of the analyzers exceeded the linearity
performance criteria. Several analyzers failed by
very small margins approaching the testing capability.
Retesting and/or very minor modifications would make most
analyzers meet linearity performance parameters.
(5) Signal output noise from the analyzers was not compared
since almost all exceeded the specifications for a
monitoring system.
(6) None of the instruments evaluated could meet the specifi-
cation of 14 to 30 days' unattended operation; however, a
7- to 15-day period of unattended operation may be
achieved.
Recommendations for selection of instrumentation for the RAPS Program
that can meet most of its requirements can better be accomplished by
verifying that instrumentation based on the specified measurement principle
exists, than by naming specific brands for purchase. Recommendation of
specific instruments would be hazardous, in that all available analyzers
for each category could not be included in the evaluation and improved and
updated versions of specific analyzers are occurring every day. For
instance, design changes suggested to several manufacturers months ago
are presently being incorporated into the design of the particular analyzer
and should contribute to improved performance for future models.
Close examination of the performance evaluation results indicates
none of the analyzers evaluated met all the required specifications, some
came close to meeting most of the specifications, and others failed to
meet many, if not all, of the specifications. In most cases where a
particular analyzer failed to meet the required specification, poor
performance could usually be attributed to minor deficiencies, such as
poor quality control, use of underdesigned components, cheap components
to save on initial costs, and poor design, rather than major flaws in the
148
-------
measurement principle. These shortcomings can be overcome when users of
equipment at the local/State/Federal level demand that manufacturers be
required to demonstrate that their equipment meets minimum performance
specifications as measured by a uniform, standardized testing procedure.
Until such a program is implemented, equipment manufacturers will continue
to market air quality instrumentation which is underdesigned for manu-
facturing cost-savings, has not been properly evaluated for field
monitoring prior to release, and is continually updated and/or
redesigned as users discover deficiencies through field use and/or
instrument evaluation programs.
149
-------
9,0
1. Ballard, L. F., J. B. Tommerdahl, C. E. Decker, T. M. Royal, and D. R.
Nifong. Field Evaluation of New Air Pollution Monitoring Systems:
The Los Angeles Study. Interim Report. Research Triangle Institute,
Contract No. CPA 70-101., National Air Pollution Control Administration,
1971.
2. Ballard, L. F., J. B. Tommerdahl, C. E. Decker, T. M. Royal, and L. K.
Matus. Field Evaluation of New Air Pollution Monitoring Systems:
St. Louis Study, Phase 1. Interim Report. Research Triangle Institute,
Contract No. CPA 70-101, Environmental Protection Agency, 1971.
3. Decker, C. E. , T. M. Royaj, J. B. Tommerdahl, and L. K. Matus. Field
Evaluation of New Air Pollution Monitoring Systems: St. Louis Study,
Phase II. Interim Report. Research Triangle Institute, Contract
CPA 70-101, Environmental Protection Agency, 1971.
4. Decker, C. E., T. M. Royaj, and J. B. Tommerdahl. Field Evaluation of
New Air Pollution Monitoring Systems. Final Report. Research Triangle
Institute, Contract CPA 70-101, Environmental Protection Agency, 1971.
5. Nederbragt, G. W., A. Van Der Horst, and J. Van Duijn. Nature. 206:
87-90, 1965.
6. Hodgeson, J. A., K. J. Krost, A. E. O'Keeffee, and R. K. Stevens.
Chemiluminescent Measurement of Atmospheric Ozone: Response Character-
istics and Operating Variables. Anal Chem. 42: 1795-1802, December
1970.
7. Fontijn, A., A. J. Sabadell, and R. J. Ronco. Homogeneous Chemilumi-
nescent Measurement :-.{ Xitii^ Oxiue with Ozont. Anal Chem. 42: 575-
579, May 1970.
8. Hodgeson, J. A., K. A. Rehiiu-, B. L. Martin, and R. K. Stevens.
Measurement for Atmospheric Oxides of Nitrogen and Ammonia by
Chemiluminescence. Presented at 65th Annual Meeting of Air Pollution
Control Association, .Hine 1972.
9. Brietenbach, L. P. arid M. Shelef. Ue\e Inherit ^f a Method for the
Analysis of NO,., and NH, by NO-Measuring Instruments. Technical Report
No. SR. 71-130, Scientific Research Stiff, Ford Motor Company. 1.971.
10. Brody, S. S. and J. F. Chaney. Flame Photometric Detector: Application
of a Specific Detector for Phosphorus and for Sulfur Compounds Sensitive
to Subnanogram Quantities. .1. Gas Chromatog. 4: 42-46, 1966.
150
-------
11. Stevens, R. K., J. B. Mulik, A. E. O'Keeffee, and K. J. Krost. Gas
Chromatography of Reactive Sulfur Gases in Air at the Parts-per-
Billion Level. Anal Chem. 4: 827-831, June 1971.
12. Hodgeson, J. A., R. K. Stevens, and B. E. Martin. A Stable Ozone
Source Applicable as a Secondary Standard for Calibration of
Atmospheric Monitors. IN: Air Quality Instrumentation, Scales, J.
(ed.). Instrument Society of America, 1972. p. 114-128.
13. Federal Resister. National Primary and Secondary Ambient Air Quality
Standards. Environmental Protection Agency. 36: 8186-8201, April 1971.
14. Federal Register. Ambient Air Quality Standards: Reference Method for
Determination of Nitrogen Dioxide. Environmental Protection Agency.
38: 17174-15183, June 1973.
15. O'Keeffee, A. E. and G. C. Ortman. Primary Standards for Trace Gas
Analysis. Anal Chem. 38: 760-763, May 1966.
151
-------
APPENDIX A
REOWE0ED CALIBRATION PROCEDURES
153
-------
APPENDIX A REOTENDED CALIBRATION PROCEDURES
Recommended calibration techniques and detailed calibration procedures
utilized for this evaluation program to provide standard calibration con-
centrations were presented in Section 6.0 of this report. The purpose of
this section is not simply to reiterate what was presented in Section 6.0,
but to expand on the subject area and discuss the philosophy and recommend
frequency of calibration, advantages, disadvantages, reliability, precision
and accuracy of current calibration systems, problems encountered in cali-
bration of particular instruments, reliability and accuracy of commercially
available calibration gases (i.e., cylinders), and quality control procedures
applicable to calibration and operation of continuous analyzers.
Dynamic calibration procedures are recommended for calibration of all
continuous air quality analyzers. Static methods were useful at one time
for performing detector system checks, but serve no useful purpose for
calibration of present day state-of-the-art instrumentation. This does not
mean, however, that the development of specific test point checks should
not be pursued for verification of subsystem component performance (i.e.,
test points to verify proper operation of electrometer amplifiers, power
supply systems, photomultiplier tubes, flow rates, etc.). Properly designed
test indicators could be invaluable aids for maintenance, operational, and
troubleshooting of ambient air analyzers.
Recommended calibration techniques, optimum frequency of calibration,
and precision and estimated accuracy of the calibration procedure are
presented for each set of analyzers (i.e., 0.,, N0x, SC^, HC, CO) in Table
A-l. The optimum frequency recommended for calibration of continuous air
quality analyzers consists of daily zero and span checks with multipoint
(zero plus four calibration points) calibration at biweekly intervals. This
frequency of calibration has been found to be sufficient to insure reli-
ability of data and characterize instrument performance. It is readily
apparent that the proper use of calibration procedures and optimum cali-
bration schedules are paramount ii issuring the collection of quality data.
154
-------
CO
2
p
Q
erf
PVl
§
H
^
PQ
H
J
O
g
i
CJ
W
3
*
rH
Jt
at
H
3
H
>>
U
cd
M
O TJ
££
«S
TJ 01
01 g
4J
§«W
0
H
4J
CO
H
<4H C
0 0
T4
a w
o
H M
« £
o
§
iH
4J TJ
cd o
rl J2
,0 4J
1*
o
4J
§
4J
rH
rH
O
(X,
X
m
+ 1
x
CM
+ 1
£
rH
.. ^
>> o»
rH 0)
H >
cd -H
TJ rQ
i3^
P.iH
m o
a.
« «rl
O 4J
rl rH
a) 3
N §
M
O
U H
cd M H .
i-i cd
a) o M
fl 4J
&£*,-<*
0) CJ v^ rl
§rj 01
0) TJ *J
H rl O CO
O 0) 43 iH
1 «*-! 4J CO
& « jj «J
t> Wi S w
01
g
N
O
X
CM
+ 1
X
CM
+ 1
>?
rH
.. ^
>,1B
rH 01
H >
T) rfl
S4J
C
P.iH
to o
*
iH
0 4J
rl rH
o> 3
N a
g
iH
4J
cd
01
(X<
0)
to ^
g 5
r5 H
0)
TJ
iH
X
O
rl
Q
3
IH
rH
3
co
K
tr>
+ 1
»-e
CM
+ 1
>,
rH
M rS^
>^ V
rH 0)
H >
cd -H
TJ ^a
Pi -u
cd C
O.iH
co o
a
> iH
O 4J
rl rH
0) 3
N S
0 0)
iH O
4-> C
Cd 0)
M t-i
4-1 CD
H M-l
H 01
tA
0)
CO rH
cd cd / N
JS rl TJ
PL, 0) O
TJ A
CO 0) 4J
>
rH rH
" ^4 " M
>, 0) >. O)
1-10) rH 0)
H S iH >
cd -3 cd IH
TJ J3 TJ J3
C 4J C 4J
cd a cd a
P..H a>iH
CO O 0) O
51 5*
-rl «H
O 4J O U
Jj rH M rH
01 3 J 01
Cd TJ
y B ti
0 OX
« «8
^ c3=g
TJ
0)
<4-t
H
rl
0)
0)
,0
3
3
O
«c
CO
g
H
4J
cd
t-l
>> w
U C
a v
0) U
g- §
ai a
i4
M-l TJ
g 2
3 2
P CO
cd
rl CD
.0
H rl
rH 0)
cd M
o 3
4J
i 3
i -2
- i
<§ 1
« *
*
155
-------
During the evaluation program, problems were encountered in the
calibration of certain analyzers utilizing the techniques described in
Section 6.0 and summarized in Table A-l. These problems along with
pertinent comments and observations are presented for each calibration
procedure (i.e., Q^> N0/N02, SO, HC, CO) in the following paragraphs.
No problems were encountered with calibration of ozone instrumentation
using the UV-ozone generator, which was referenced to the Federal Reference
Method (Neutral-Buffered KI Procedure). It is readily apparent, however,
to experienced personnel that the precision of generation of ozone concen-
tration by the UV-ozone generator, once calibrated, is far superior to the
precision and accuracy achievable with the manual Neutral-Buffered KI
Procedure for verification of the ozone concentrations. It has also been
demonstrated that once calibrated, the UV-ozone generator described in
Section 6.1 can be used reliably in the field for calibration purposes for
several months without further verification. Calibration concentrations
generated by this technique should be referenced to and verified by the
manual Neutral-Buffered KI Procedure at least once every two months over
the ozone measurement range normally encountered at the particular
monitoring site (i.e., full scale range setting of measuring instrument).
Although the precision of generation of ozone by this technique has been
shown to be better than + 2%, the accuracy of ozone measurement can be no
better than the accuracy of the reference method in the absence of a
primary standard for ozone. Therefore, the accuracy of the ozone
measurement should not be reported as being better than +5%. In the
absence of a primary ozone standard, absolute measurements of ozone
concentration in ambient air are not achievable at this time.
No problems were encountered in calibration of total sulfur analyzers
with calibration concentrations generated by S02 permeation tubes. Certified
SO^ permeation tubes can be obtained from the National Bureau of Standards
as a primary standard for S0_. The major problem is maintaining the
permeation tube at a constant temperature of + 0.1°C. Calibration apparatus
are available commercially that utilize constant temperature water or
thermoelectric-cooled permeation tube holders. The main precaution when
156
-------
using permeation apparatus that employs extremely low flow rates, Teflon
tube holders, and low-output permeation tubes is to always maintain a suffi-
cient flow rate over the permeation tube to prevent buildup of S0_ in the
permeation tube holder and entrapment of SO- in the flow-dilution system.
The precision and accuracy of this calibration procedure are quite adequate
to insure quality acquisition of S02 ambient air data.
Calibration of chemiluminescent NO-NO -N09 analyzers with the gas-
X £
phase titration of NO with ozone was recommended due to problems encountered
with field use of N02 permeation tubes. As yet, the National Bureau of
Standards has not been willing to certify NO- permeation tubes as a primary
standard for generating NO,, test atmospheres. The gas-phase titration
technique has been shown to be an acceptable, precise, and accurate way to
calibrate chemiluminescent NO-NO -NO- analyzers. The primary consideration
is, again, the accuracy of the reference method for the iodometric calibra-
tion of the UV-ozone generator incorporated into the calibration apparatus.
Accuracy better than + 5% is not achievable under these circumstances,
although the precision of generation of NO and N02 calibration concentration
is on the order of + 2%. This technique offers the advantage of being able
both to generate NO and NO- calibration and to allow the operator to
determine the conversion efficiency of the catalytic converter (reduces N09
to NO), whenever desired. The converter efficiency should be verified
during each calibration, both daily zero and span calibrations and multi-
point calibrations at two-week intervals. The accuracy and reliability of
N02 data obtained by the chemiluminescent technique are highly dependent on
maintaining a constant converter efficiency, preferably as close to 100% as
possible. Verification of the NO in N. cylinder concentration should be
performed at two-month intervals. Precautions should also be taken to
verify the calibration of the UV-ozone generator at two-month intervals.
One problem was encountered with the gas phase titration system and
involved using compressed, dry air filtered through charcoal for zeroing the
Bendix NO-NO -NO analyzer. An erroneous high zero was obtained, which
X £.
could be attributed to the absence of moisture in the zero air stream.
157
-------
The discrepancy between zero obtained from dry air versus air containing
some moisture as is encountered when using the Bendix zero air module or
when monitoring ambient air, was on the order of 2%. Negative numbers
often result when erroneous zeroes are obtained and ambient levels of
NO and NO., are near zero. Design changes in the flow system have been
suggested to Bendix to eliminate this problem. No other chemiluminescent
NO-NO -NO^ analyzers evaluated exhibited this effect.
Calibration of hydrocarbon and CO analyzers using cylinder of
compressed gas mixtures is the ideal method, in that the source of gas is
readily accessible, can be used by unskilled personnel, requires no
elaborate accessory equipment, produces the stated calibration concentration
on demand, and is inexpensive when compared to other calibration systems
(i.e., permeation tube calibration systems, gas phase titration, etc.).
Major disadvantages using cylinders in calibration of hydrocarbon or carbon
monoxide analyzers are: (1) size and weight of cylinders; (2) unavail-
ability of reference standards for comparison with manufacturer's
certification of component content; (3) stability factors for component,
especially at low concentrations (i.e., carbon monoxide in air); and
(4) lack of quality control among suppliers of blended calibration mixtures.
Several mixtures of CH and CO in hydrocarbon-free air and zero air
obtained from various manufacturers of speciality gases were utilized in
this program to determine performance specifications for analyzers having
the capability to measure non-methane hydrocarbons and carbon monoxide.
As a result of this effort, several observations became readily apparent;
(1) Zero air certified to contain less than 0.1 ppm
(65.4 yg/m ) hydrocarbons as CH= was not acceptable for
zeroing hydrocarbon analyzers trying to discriminate
non-methane hydrocarbons in the range 0-0.240 ppm
(0-157 ug/m3);
(2) The manufacturer's certification of analysis cannot be
accepted without further reference to some other standard
(i.e., several cylinders were received with the wrong
analysis);
158
-------
(3) Inaccuracies and unreliability of standard calibration
gases biased the evaluation results for each analyzer
unfavorably;
(4) Calibration of non-methane hydrocarbon analyzers with
CH, is not totally adequate, but unfortunately, no one
can at present specify the composition of non-methane
hydrocarbons in ambient air;
(5) Future procurement of CH, in hydrocarbon-free air should
specify an analysis for CH, and total hydrocarbons as
CH,; these two analyses are usually not the same and
cause extreme difficulties when calibrating instruments
having the capability to discriminate non-methane
hydrocarbons;
(6) Cylinder gas containing CH, in air is stable for at least
six months at concentrations greater than 1.0 ppm; cylinder
gas containing CO in air at low ppm levels is not as stable
and its concentration should be verified at monthly intervals;
(7) Blended air is not acceptable for producing CH, and CO
standard calibration concentrations, unless the oxygen
concentration can be maintained to that normally encountered
in ambient air (i.e., 21.0 + 0.5%). Oxygen enriched moisture
of CH, in air produce elevated response for total hydro-
carbon measurements as methane.
(8) Quality control procedures should be upgraded within the
speciality gas industry; and
(9) Primary reference standards for hydrocarbon and carbon
monoxide should be developed and marketed by the National
Bureau of Standards.
The accuracy, and hence the usefulness, of data obtained from air
monitoring instrumentation and surveillance programs are dependent to a great
extent upon the ability to calibrate the instrument under actual operating
conditions. This can only be accomplished by the acceptance and proper use
159
-------
of standardized calibration techniques and procedures and requires
implementation of quality assurance programs at state, regional, and the
federal levels to insure the orderly development of more precise and
accurate methods for calibration and also requires some mechanism to
insure the interrelatibility between data being generated by various
organizations. Otherwise, a large storehouse of information will become
unusable simply because no one can place any confidence in the numbers.
ALTERNATE PROCEDURE FOR GENERATION OF
CALIBRATION QUALITY ZERO AIR
Generation of zero air at monitoring sites is an attractive feature
that needs to be considered from the standpoint of cost and convenience.
As was previously stated, zero air can be commercially obtained in
pressurized cylinders and specified to contain less than 0.01 to 0.1 ppm
hydrocarbons as methane and 0.1 ppm carbon monoxide. Typical test results
conducted on zero grade air obtained from several manufacturers are as
follows:
Certified Cylinder Measured Value
less than 0.01 ppm THC 0.05 ppm
less than 0.1 ppm THC 0.150 ppm
less than 0.1 ppm THC 0.200 ppm
As an alternate means of generating zero air, one air purifier was
tested during this evaluation and yielded zero air of higher quality than
presently available in compressed gas cylinders. The quality of zero
air generated by this system was dependent upon flow rate and contamination
levels. To remove the contaminants from air, a catalytic oxidizer was used
to convert hydrocarbons and carbon monoxide to carbon dioxide and water.
Ozone is also converted to oxygen in the process. The oxidizer employs a
temperature-controlled, two-stage oxidation system which reduces hydro-
carbons to less than 0.1 ppm (measured as methane) in its primary catalyst
bed. A second catalyst bed reduces carbon monoxide to less than 0.1 ppm.
The oxidizer was tested using zero air from pressurized cylinders
certified to contain less than 0.1 total hydrocarbon as methane and
160
-------
0.1 ppm carbon monoxide. Using a flow rate of 200 cc/min for producing
hydrocarbon-free, zero air and 2.0 JZ./min for carbon monoxide, the total
hydrocarbon concentration was determined to be 0.02 ppm measured as
methane and the carbon monoxide concentration was 0.05 ppm. In all cases,
zero air of a higher quality was produced by using the catalytic oxidizer
system in conjunction with zero air purchased in cylinders.
Frequently, it is desirable to process ambient air to produce zero
air for use in calibration of ambient air analyzers. Ambient air usually
contains sufficient moisture to cause condensation or act as an interferent
in the calibration system used for generating test atmospheres. One method
of effectively removing moisture from air is through the use of a heatless
dryer. Such a dryer uses two packed columns; while one column is passing
dry air to be used, the other column is vented to the atmosphere to allow
backflushing to exhaust the unwanted water. Such a drying system is very
effective and can be used satisfactorily for supplying dry air; however,
the resulting air tends to be oxygen enriched.
Such oxygen enrichment may cause difficulty with ozone, gas-phase
titration of N02, and hydrocarbon calibrations should the air supply be
interchanged with a compressed zero air cylinder. The oxygen enriched
air will yield slightly higher ozone and nitrogen dioxide concen-
trations and give an elevated hydrocarbon response using flame ionization
detectors. Air purification systems are commercially available that
utilize the heatless dryer in conjunction with a catalytic converter to
produce both dry air for calibration purposes (i.e., zero and diluent air
for 0-,, SO-, and NO, N02 analyzers) and chromatographic-grade air for
combustion air for gas chromatographic-flame ionization detector analyses.
161
-------
APPENDIX B
RECOftENDED OPERATING PR3CEDURES
163
-------
APPENDIX B REOWOTED OPERATING PlflCEDURES
The purpose of this section is not simply to reiterate the instructions
supplied by the manufacturer for operating each analyzer, but to discuss
the philosophy and recommend, based on experience and experimental evidence,
optimum operating procedures for a total air monitoring system. Factors
which must be considered are specification of space requirements and
operating environment (temperature and humidity controls), power require-
ments, air sampling and inlet systems, and data acquisition systems. These
factors will be considered independent of recommended operating and cali-
bration requirements for each set of analyzers (i.e., 0_, NO , S09, EC,
j x z
CO). Recommendation of optimum operating procedures for each set of
analyzers can best be accomplished by (1) reiterating those procedures
specified by the manufacturer and (2) pointing out any additional procedures
that need to be incorporated into the operational procedure to improve per-
formance and prevent excessive downtime. For the sake of brevity, operating
procedures recommended by the manufacturer will not be presented; however,
the reader is advised to review in detail and follow the operational
procedures supplied in the instruction manual accompanying each analyzer.
Only those additional procedures that need to be incorporated into the
operating instructions will be presented in this section for each set of
analyzers.
Space and an optimum operat' ; environment are important considerations
for any monitoring system. Proper selection of shelter design and size is
required to provide the necessary housing for analyzers, data recording and
calibration systems, and personnel. Space requirements will vary with the
sophistication of the station in terms of monitoring requirements and data
acquisition capabilities. To adequately house instrumentation for the
measurement of 0 , NO-NO -NO , SO-, THC-CH -NMHC, and CO in ambient air,
J X £, £ *\
accessory equipment, data acquisition and calibration systems, and
associated test equipment, a shelter with a minimum of 250 sq. ft. is
recommended. Consideration should also be given to maintaining the
aesthetic quality of the surrounding environment or neighborhood and should
dictate the exterior design of the shelter.
164
-------
Control of temperature and humidity to close tolerances is required
for optimum operation of air quality analyzers, regardless of the manu-
facturer or model. Requirements will vary from analyzer to analyzer
depending on the principle of operation and other design factors.
Therefore, temperature and humidity specifications should be designed
around the analyzer or analyzers most affected by or subject to variation
in environmental extremes. The following temperature and humidity ranges
are recommended for optimum operation of air monitoring systems, air
quality analyzers, and automatic data acquisition systems with telemetry
and computer processing capabilities:
(1) ambient temperature - 72° + 3°F
(2) humidity 30 - 50% R.H.
Design of air conditioning/heating systems should anticipate heat load
requirements, provide good distribution of heat/cooling air flow to minimize
temperature extremes (i.e., hot/cold spots), and be capable of maintaining
the required temperature and humidity specifications within the stated
tolerances over an exterior ambient temperature range consistent with that
typically experienced in the area where the site is to be located.
Regulation of AC-line voltage is recommended for optimum operation of
all analyzers, regardless pf manufacturer or model, stripchart recorders,
and automatic data acquisition systems with computer processing capabilities.
Regulators are available which are capable of maintaining AC-line voltage
to 117 volts + 2 volts, while supplying 30 to 40 amps of power. This is
sufficient to satisfy the regulated power requirements for the entire air
monitoring system. Regulation of line voltage variation removes one more
variable that can affect proper operation of instrumentation under adverse
conditions of extremes in line voltage.
Air sampling, inlet manifold, inlet sampling lines, and all surfaces
with which the air sample comes into contact should be selected to maintain
the integrity of the sample and minimize losses due to wall absorption and
reaction with surfaces. To maintain the integrity of the air sample, inlet
manifold and air sampling systems constructed of Teflon and glass are
recommended throughout. Instrument inlet sampling lines should be
165
-------
constructed of Teflon and be as short as possible. As was stated in
Section 2.2, laminar flow is not required in the manifold to prevent losses
of pollutants due to wall adsorption or reaction. An air sampling system
consisting of 1- to 2-in (2.54-5.08 cm) O.D. glass tubing with bushings
and Viton 0-ring connectors and a flow rate of 50 to 100 liters per minute
through the manifold is sufficient for any air monitoring system. Larger
glass manifold tubing may be used but requires higher investment costs.
Location of the manifold inlet should be at least 1 meter above the roof
surface, and a flower pot or cane arrangement should be used to prevent
moisture and large particulates from settling into the manifold. It is
also advisable to include a particulate and condensation collection bottle
at the base of the inlet line. All glass inlet lines should be thoroughly
cleansed prior to use and cleaned periodically, as needed. For instance,
experience has shown that some conditioning of the surfaces of the manifold
inlet line is required with ambient air after cleansing to prevent initial
loss of ozone.
Additional procedures recommended for incorporation into the operating
instructions supplied by the manufacturers for each set of analyzers (i.e.,
0~, NO , S02> HC, CO) or procedures that need to be reiterated are
presented in outline form in the following paragraphs. These recommen-
dations cover many aspects that can be related to safety, to optimum
operation of the facility as well as the analyzer itself, and to general
rules of good practice for operation of an air monitoring system. Recom-
mended procedures applicable to the operation of instrumentation based on
different measurement principles are presented together, in cases where
analyzers with different measurement principles were evaluated in this
study. Thus, some recommendations may not apply to both measurement
techniques if two were evaluated. As stated before, the reader should
refer to and follow all instructions specified in the instruction manual
supplied with the respective analyzer.
166
-------
A. OZONE INSTRUMENTATION (Chemiluminescent Principle)
(1) Precautions should be taken to ensure a leakless system
for the ethylene supply. In addition, a catalytic
oxidation unit or some other device should be incorporated
to oxidize or destroy ethylene in the exhaust stream from
the analyzer. The exhaust line should be vented to the
outside exterior of the shelter. Destruction of the
ethylene is warranted from a safety standpoint and to prevent
contamination of the air sampling system with the ethylene,
a non-methane hydrocarbon. Precautions should be taken to
safely dispose of or destroy percentage quantities of ethylene.
(2) Never operate an ozone analyzer without a particulate filter
on the sample inlet line. Construction of the filter and
filter element should be of Teflon. Change the filter element
as required.
(3) Teflon or glass tubing are the only materials recommended for
use for the sample inlet line. Teflon tubing is preferred.
B. OXIDES OF NITROGEN (Chemiluminescent)
(1) Precaution should be taken to ensure a leakless system if
oxygen is used for the ozonator supply. Ozone must be
destroyed prior to venting to the atmosphere. Ensure that
the destructive agent (charcoal, thermal decomposition, etc.)
is replaced as required. Vent the exhaust line to the outside
exterior of the shelter.
(2) Include provisions for monitoring cell pressure in the
daily operational procedures. Replace vacuum pump if cell
pressure decreases by 2.54 cm of Hg or recalibrate analyzer,
whichever is more practical. Regardless, replace vacuum pump
when it can no longer maintain the required cell pressure
specified by the manufacturer.
167
-------
(3) Never operate NO analyzers without a particulate filter
X
on the sample inlet line. Construction of the filter holder
and filter element should be of Teflon. Change the filter
element as required.
(4) Teflon or glass tubing are the only materials recommended
for use as the sample inlet line. Teflon tubing is
preferred.
C . SULFUR DIOXIDE (Flame Photometry)
(1) Precautions should be taken to ensure a leakless system for
the hydrogen supply and plumbing system. Vent exhaust from
analyzer to the outside exterior of the shelter.
(2) Use of a particulate filter on the sample inlet line to
remove particulates is recommended, preferably one constructed
of Teflon with a Teflon filter element.
(3) Incorporate a heated silver gauze or wire filter into the
sample inlet system to remove hydrogen sulfide and mercaptans
if S02 data are desired. Otherwise, report the data as total
sulfur, measured as S0~.
(4) Teflon or glass tubing is the only material recommended for
use as the sample inlet line. Teflon tubing is preferred.
D. TOTAL SULFUR, SULFUR DIOXIDE, HYDROGEN SULFIDE
(GC-Flame Photometry)
(1) Precaution should be taken to maintain integrity of the
hydrogen supply and plumbing system. Vent the exhaust from
the analyzer to the outside exterior of the shelter.
(2) Use of a Teflon filter holder and filter element is recommended
on the sample inlet line to remove particulates.
(3) Teflon or glass tubing is the only material recommended for
use as the sample inlet. Teflon tubing is preferred.
E. TOTAL HYDROCARBONS, METHANE, CARBON MONOXIDE
(GC-Flame lonization)
(1) Precautions should be taken to maintain the integrity of
the hydrocarbon supply and the plumbing system. Vent the
168
-------
exhaust from the analyzer to the outside exterior of the
shelter. This is important, since sudden pressure changes
in the shelter interior caused by opening and closing the
door will affect the flame.
(2) Never operate a GC-analyzer on ambient air without a
particulate inlet filter. Glass fiber or membrane filter
elements are satisfactory, provided the pressure drop across
the filter element does not restrict air flow.
(3) Use dryers on all gas supplies (hydrogen, burner air, etc.)
as specified by the manufacturer.
(4) Use of two-stage, metal-diaphragm regulators is recommended
for all compressed gas cylinders (i.e., hydrogen, air,
calibration gases).
(5) Use of Teflon tubing for sample inlet lines is not recommended,
due to outgasing of the Teflon and permeation of hydrocarbons
through the Teflon walls. Use of chromatographic grade
stainless steel tubing is recommended.
F. CARBON MONOXIDE (GC-Flame lonization, NDIR, D.I. Fluorescence -
IR Detection)
(1) Precautions listed in Part E apply equally to measurement of
CO by GC-flame donization detection.
(2) Use of particulate filters on sample inlet lines is
recommended. Glass fiber or paper filters are satisfactory.
(3) Teflon, glass, stainless steel, copper, or polyethylene
tubing is satisfactory material for use as sample inlet
lines.
The above mentioned precautions and procedures are recommended for optimum
operation of air quality analyzers in conjunction with those specified by
the respective manufacturers but are by no means all inclusive.
Data acquisition capabilities and facilities will vary with the
sophistication of the air monitoring system and the monitoring requirements
for which the system was designed. This capability can range from a mere
169
-------
stripchart recorder for a single analyzer to an on-line computer with
telemetering capabilities for a full complement of analyzers. Best
practice dictates the use of stripchart recorders for all instrument
outputs, even though automatic data acquisition systems with computer
processing capabilities are available. The stripchart recording provides
a visual record of the instrument's performance over the past 24 hours,
serves as a backup for retrieval of data when failures occur in the
automatic data acquisition system or computer, and is an invaluable aid
during instrument calibration and checkout.
170
-------
APPENDIX C
REOWBOD [1AINTENANCE PRXEDUES
171
-------
APPENDIX C RECQTfENDED MAINTENANCE PROCEDURES
Recommendation of maintenance procedures for each analyzer can best be
accomplished by (1) reiterating those specified by the manufacturer and
(2) pointing out any additional procedures that need to be incorporated
into the maintenance schedule to improve performance and prevent excessive
downtime. In most cases the procedures recommended by the manufacturer will
suffice. For the sake of brevity, recommended maintenance procedures are
presented in outline form for each analyzer. It should be noted that these
procedures are recommended based on the particular instrument design as it
existed in early Spring 1973. Accordingly, the reader is advised to review
in detail and follow the maintenance procedures supplied in the instruction
manual accompanying each analyzer.
1. BENDIX ?;ODEL 8002 CHEMILUMINESCENT OZONE METER
(a) Replace Teflon element of the sample air filter as required
by particulate loading in the area of the monitoring site
(at least once/two weeks).
(b) Replace calibration sample filter chemicals (mole sieve,
silica gel, charcoal) as needed (at least once/three months).
(c) Replace stainless steel mesh element of ethylene filter as
needed.
(d) Replace ethylene cylinder when cylinder pressures decreases
to 100 psig.
(e) Examine reed valves of evacuation pump and perform maintenance
as specified in instrument manual if pump performance degrades
to the point where it will not maintain the proper sample plus
ethylene flow rate.
(f) Refer to "Troubleshooting Section" of the instruction manual
to isolate problems and determine the cause of failure.
172
-------
2. DASIBI MODEL 1003-AH UV-ABSORPTION OZONE ANALYZER
(a) Visually inspect interior and exterior instrument and remove
excessive dust accumulation.
(b) Clean the absorption chamber and the absorption chamber
windows and mirrors at least once/six months.
(c) Replace the ozone catalyst as required (at least once/six
months).
(d) Replace the particulate or aerosol filter as required (at
least once/month).
(e) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and determine the cause of
failures.
3. BENDIX MODEL 8101-B CHEMILUMINESCENT NO-NO -N09 ANALYZER
/\ £
(a) Replace Teflon element of the sample air filter as required
(at least once/month).
(b) Replace charcoal in the exhaust filter on a monthly basis.
(c) Replace oxygen cylinder when pressure decreases to 100 psig
and the stainless steel mesh element of the oxygen filter
as needed.
(d) Observe cell pressure on a daily basis by noting the vacuum
gauge reading. Replace pump if pump performance degrades to
a point that it will not maintain at least 23 inches of Hg
(58.42 cm) vacuum. Changes in cell pressure of 1 inch of Hg
(2.54 cm) require recalibration of the analyzer.
(e) Clean instrument interior of accumulated dust by using low
velocity air from a compressed cylinder.
(f) Clean the residue buildup off the ozonator electrode on a
monthly basis.
(g) Monitor the sample air inlet and vacuum pump outlet flow
rates on a weekly basis to insure proper gas flows.
(h) Refer to the "Troubleshooting Section" of the instrument
manual to isolate problems and determine causes of failures.
173
-------
4. MELOY MODEL 520 CHEMILUMINESCENT NO-NO -NO, ANALYZER
A £
(a) Replace particulate filters as required. Manufacturer
recommendes once/year; however, this may not be adequate
for all environments.
(b) Replace flow controller assembly as required. Manufacturer
recommends once/two years.
(c) Replace zero and span valves every 36 months and the sample
switching valve every 24 months or upon failure.
(d) Replace charcoal in de-ozonizer filter on a monthly basis.
(e) Replace vacuum regulator every 12 months and the vacuum pump
every 24 months or upon failure.
(f) Replace catalytic converter as required (at least once/year).
Converter is rated at 1000 ppm-hrs of NO .
(g) Replace or clean window/filter arrangement of detector as
required (at least once/year) and/or other components of
the detector/chopper unit as needed.
(h) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and determine the cause of failures.
5. THERMO ELECTRON MODEL 14 CHEMILUMINESCENT NO-NO-NO, ANALYZER
A C.
(a) Replace Balstron filter as required (at least once/year).
Replacement is signaled when white portion of the filter is
black.
(b) Replace or recharge desiccant in air dryer for ozone generator
as required.
(c) Recharge the activated charcoal de-ozonizer filter at
12-month intervals, or more frequently if needed.
(d) Check the converter efficiency at five points in the measure-
ment range on at least a monthly basis.
(e) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and determine the cause of failures.
174
-------
6. REM MODEL 642 CHEMILUMINESCENT NO-NOX-N02 ANALYZER
(a) Replace Teflon element of the sample air filter as required
by particulate loading in the area of the monitoring site
(at least once/month).
(b) Clean the instrument and flush with nitrogen at 30-day
intervals.
(c) Replace drierite dessicant as required.
, (d) Rejuvenate catalytic converter when efficiency of conversion
drops below 95%.
(e) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and to determine the cause of
failures.
7. MELOY MODEL SA-185R FLAME PHOTOMETRIC TOTAL SULFUR ANALYZER
(a) Replace Teflon element of the sample air filter as required
by the particulate loading in the area of the monitoring
site (at least once/month).
(b) Replace vacuum pump as required to maintain the required
flow rate through the gas flow system.
(c) Clean and maintain the air filter unit for the ventilating
fan and oil the circulating fan motor twice each year.
(d) Clean and maintain the burner chamber as prescribed in the
instruction manual after 6 months of continuous operation,
or sooner if required.
(e) Replace the dilution air orifice (#18 hypodermic needle
1.00 inch long) and the rubber septum as required.
(f) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and determine the cause of
failures.
175
-------
8. BENDIX MODEL 8300 FLAME PHOTOMETRIC TOTAL SULFUR ANALYZER
(a) Replace Teflon element of the sample air filter as required
(at least once/month).
(b) Replace hydrogen cylinder when cylinder pressure decreases to
100 psig, if used.
(c) Replace the hydrogen drier every second cylinder change and/or
sooner, if required.
(d) Check reed values if pump performance degrades to the point
that it will not maintain the proper sample flow rate.
(e) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and determine the cause of failures.
9. BECKMAN MODEL 6800 AIR QUALITY CHROMATOGRAPH (THC, CH4, CO) ANALYZER
(a) Replace the ambient air filter element at least once/month.
(b) Replace all supply air and hydrogen gas driers and filters at
six-month intervals, or sooner if required.
(c) Check tightness of all gas line connections on a monthly basis,
especially the hydrogen line.
(d) Check tightness of all terminal board connections at six-month
intervals.
(e) Clean interior of instrument of dust and accumulated dirt as
required.
(f) Verify proper gas flow rates and instrument performance by
checking flow rates and running manual chromatograms to verify
proper retention times on a weekly basis.
(g) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and to determine the cause of
failures.
176
-------
10. BENDIX MODEL 8201 AMBIENT HYDROCARBON ANALYZER
(a) Maintenance requirements are not specified in the preliminary
instrument manual provided with the analyzer. Good practice
dictates replacement of the filter element (preferably Teflon)
of the sample inlet filter as required (at least once/month).
(b) Replace the hydrogen cylinder, if used, when the cylinder
pressure decreases to 100 psig. Replace gas drier at six-
month intervals or sooner, if required.
(c) Clean instrument of dust and accumulated dirt, as required.
(d) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and to determine the cause of
failures.
11. BENDIX MODEL 8501-5FA CARBON MONOXIDE ANALYZER
(a) Inspect the input sample filter on a weekly basis and drain
any condensate. Replace the microfliter element at least
once/month.
(b) Inspect the gas filter monthly and replace at six-month
intervals.
(c) Perform maintenance on the sample pump as required. Foreign
particles under the reed valves can be the cause of degra-
dation of pump performance.
(d) Maintain interior of instrument as dust free as possible.
(e) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and determine the cause of
failures.
177
-------
12. ANDROS MODEL 7000 DUAL ISOTOPE FLUORESCENCE CARBON MONOXIDE ANALYZER
(a) Normal maintenance requires only replacement of the inlet
filter element and maintaining the fan filter on the rear
panel at periodic intervals (at least once/month).
(b) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and to determine the cause of
failures.
13. TRACOR MODEL 270-HA GAS CHROMATOGRAPH1C-FLAME PHOTOMETRIC ANALYZER
(a) Replace the ambient air filter element and inspect and clean
the dilution flow filter on a monthly basis.
(b) Replace all supply gas driers and filter at six-month
intervals, or sooner if required.
(c) Check tightness of all gas line connections on a monthly
basis, especially the hydrogen line.
(d) Check tightness of all terminal board connections at
six-month intervals.
(e) Clean interior of instrument of dust and accumulated dirt
as required.
(f) Verify proper gas flow rates and instrument performance by
checking flow rates and running manual chromatograms to
verify proper retention times on a weekly basis.
(g) Refer to the "Troubleshooting Section" of the instruction
manual to isolate problems and to determine the cause of
failures.
178
-------
APPENDIX D
GMPHICAL PRESENTATION OF ZERO AID SPAN DRIFT
179
-------
APPENDIX D GRAPHICAL PRESENTATION OF
ZERO AND SPAN DRIFT
The performance test for stability was conducted as outlined in
Section 5.2.1. The following data show zero and span drift in percent of
full scale from consecutive calibrations obtained on a daily basis excluding
weekends and holidays.
Each of the small circles on the graphs represents a complete multi-
point calibration and the drift occurring since the previous calibration.
A deletion of several days indicated by a broken time axis indicates the
absence of calibrations and an unadjusted operation status of the analyzer.
The inclusion of more than one graph of drift data for an instrument
indicates that either results of long-term drift tests or additional testing
after engineering modifications or repairs were performed on the analyzer.
Observation of the drift data shows that some analyzers experience about the
same percentage daily drift as weekly drift. Most analyzer drift observed
was generally random and not unidirectional.
180
-------
+8
+6
+4
+2
S +2
H
n
§ -2
w
-4
-6
-8
-10
+10
+8
+6
+4
8 +2
Q
§! -2
PM
C/3
-4
-6
-8
-10
1% = 0.005 ppm =9.8 yg/m3
-
-
-
-
-
-
ll II II
123456789
TIME (DAYS)
-
-
O
O
-
o 9
o I
°
A ^
6
-
-
i i i i i i i i i
123456789
TIME (DAYS)
BENDIX OZONE ANALYZER DRIFT DATA (RANGE
181
0.5 ppm)
-------
TJ.U
4-8
4-6
4-4
8+2
H
fe4
M 0
cd
Q
§ -2
w
N
-6
-8
-10
1% = 0.005 ppm = 9.8 yg/m3
-
-
_
r 00
-
-
-
1 1 1 1 1 1 1 1 1
123456789
TIME (DAYS)
4-10
4-8
4-6
-4
-6
-8
-10
I
1
I
l
l
I
I
123456789
TIME (DAYS)
BENDIX OZONE ANALYZER DRIFT DATA (RANGE =0.5 ppm)
182
-------
H
Pi
Q
§
W
+10
+8
+6
+4
+2
0
-2
-4
-6
-8
_i n
1% = 0.005 ppm =6.1 Mg/m3
o A n O
6^ 6
i i i i i i i i i
TIME (DAYS)
+10
+8
+6
+4
tS +2
H
P.,
3 o
Q
2 -2
en
-4
-6
-8
-10
1
1
1
1
1
1
1
1245 6789 10
TIME (DAYS)
BENDIX NO ANALYZER DRIFT DATA (RANGE =0.5 ppm)
183
-------
^*>
DRIFT
1
NJ
4-10
4-8
4-6
4-4
4-2
0
-2
-4
-6
-8
-10
- 1% - 0.005 ppm =9.4 ug/m3
_
9 ^ °
-
-
1 I 1 1 1 1 1 1 1
12456 789 10
12456 789 10
TIME (DAYS)
4-8
4-6
4-4
S 4-2
i °
% -)
PI
co
-4
-6
-8
10
-
-
-
0 / 0 9
. i' r
-
-
-
i i i i i i i i i
12456789 10
TIME (DAYS)
BENDIX NO ANALYZER DRIFT DATA (RANGE =0.5 ppm)
184
-------
TJ.U
+8
+6
+4
S +2
fn
% -2
w
-4
-6
-8
-10
+10
+8
+6
+4
S +2
Q
CO
-4
-6
-8
-10
1% = 0.005 ppm =6.1 yg/m3
-
- .1,1.
0 6 *
-
-
-
i i ii
1 2 3 47 8 9 10 11
TIME (DAYS)
y
0
0
O
-
6 ^
-
_
4
i i i i i ii
-L^ 3 4 7 89 10 11
TIME (DAYS)
MELOY NO ANALYZER DRIFT DATA (RANGE =0.5 ppm)
185
-------
> s
&«
<-s
DRIFT
§
w
N
^x
6^
^.-^
DRIFT
1
+10
+8
+6
+4
+2
0
-2
-4
-6
-8
-10
+10
+8
+6
+4
+2
0
-2
-4
-6
-8
-in
1% = 0.005 ppm =9.4 pg/m3
-
-
O
-
1 A 9 0
A *
O
-
1 1 1 1 1
1 2 3 47 8 9Q 10 11
TIME (DAYS)
-
_
O
-
- | 1
1 lit
*
-
-
o
-
Ill 1 1 1 1 i
4 7 89
TIME (DAYS)
10 11
MELOY N0? ANALYZER DRIFT DATA (RANGE =0.5 ppm)
186
-------
/x
DRIFT
§
w
-nu
+8
+6
+4
+2
0
-2
-4
-6
-8
-10
1% = 0.005 ppm = 6.1 yg/m3
-
-
-
-
n n 0 n> 099
0 ° ° aV 6
-
-
i i i i i i i i i
12 3 4 7 8 9 10 11 12
TIME (DAYS)
PC)
O
Pi
en
+10
+8
+6
+4
+2
0
-2
-4
-6
-8
-10
1
1
1
1
I
1
t 1
1
1
1
1 2 3 47 8 9 10 11 12
TIME (DAYS)
THERMO ELECTRON NO ANALYZER DRIFT DATA (RANGE =0.5 ppm)
187
-------
+10
+8
+6
^ +4
H +2
M
g
S °
w
N -2
-4
-6
-8
+10
+8
+6
^ +4
6^S
H +2
M
« 0
& _2
-4
-6
-8
-in
1% = 0.005 ppm =9.4 ug/m3
-
-
-
-
^ _
6 *\r ° °
iiii i i i i
1 2 56 7 P 9 10 11
TIME (DAYS)
-
-
-
]
- M i
6
-
i i i i i i i i i
1 2 5 6 7 8 9 10 11
TIME (DAYS)
THERMO ELECTRON N02 ANALYZER DRIFT DATA (RANGE =0.5 ppm)
188
-------
TJ.U
+8
+6
+4
S +2
H
5 o
Q
O
(X 9
w -2
N
-4
-6
-8
-10
+10
+8
+6
+4
^ +2
H
fn
M
Q 0
Q. __. O
wo L
-4
-6
-8
-10
1% = 0.010 ppm ' -
= 26.1 yg S02/m3
__
<
-
i
?
»
7
A
O
-
-
Ill II
]
<
-
-
Mi
L 2 3 4 5 6 7
TIME (DAYS)
i O 1
8 9
>
- o
B
( )
III 1 1 1 1
123456789
TIME (DAYS)
BENDIX TS ANALYZER DRIFT DATA (RANGE =1.0 ppm)
189
-------
+10
+8
+6
+4
P +2
*~s
H
H 0
o
§ -2
w
NI
-4
-6
-8
-10
1% = 0.010 ppm = 26.1 yg SO /ro3
I
I
I
I
I
I
12 13 14 28 29 30
TIME (DAYS)
37 38 39 57 58
-rj_u
+8
+6
+4
+2
0
-2
-4
-6
-8
-10
-
1
-
? A
-
-
1
1 2
<
<
V
1
1
1
>
A
1
A A
1
1 1 1 1 1
V
1
1 1 1
3 12 13 14 28 29 30 37 38 57 58
TIME (DAYS)
> BENDIX TS ANALYZER DRIFT DATA (RANGE =1.0 ppm) <
190
1
CO
-------
-nu
+8
+6
+4
^ +2
H
fa
M A
a! u
Q
O
ft5 9
W z
Nl
-6
-8
-10
1% =.0.10 ppm = 26.1 yg S02/m3
-
-
_
9 /> i
i * 4
"
-
-
1 1 1 1 1 1 1 1 1
1-23456789
TIME (DAYS)
+10
+8
+6
+4
P +2
H
fa n
M U
Q
1 "2
-4
-6
-8
10
-
O
-
1°
-
-
-
1 1 1 l 1 1 1 1 1
123456789
TIME (DAYS)
MELOY TS ANALYZER DRIFT DATA (RANGE =1.0 ppm)
191
-------
+10
+8
+6
+4
£ +2
*~s
H
M 0
PS
§ -2
w
-4
-6
-8
-10
1%
0.010 ppm = 26.1 ug S0?/m"
I
J_
I
_L
I
I
I
+10
+8
+6
+4
I +2
<
i 0
-2
-4
_2
-8
-10
5 14 15 28
TIME (DAYS)
29 36 27 50 51
57
1
I
1 23 4 5 6 14 15 28 29 36
TIME (DAYS)
MELOY TS ANALYZER DRIFT DATA (RANGE 1.0 ppm)
192
37 50 51 57
-------
-t-iu
+8
+6
+4
^ +2
H
M
g o
§
w _2
-4
-6
-8
-10
+10
+8
+6
1% = 0.010 ppm = 26.1 yg S02/m3
D 9 Ao
" o A! ~ w » w
1 1 1 | 1 III
1 2 3 4 15 16 17 18 19 2(
TIME (DAYS)
O
O
_
+4 -
B +2
g
! -2 h-
i
i
-4
-6
-8 -
-10
<
(
1
K
1 2
d
1
N
>
1
1
»
6
i
i i
3 4 15 16 17 18 19 2(
TIME (DAYS)
i
TRACOR GC TS ANALYZER DRIFT DATA (RANGE =1.0 ppm)
193
-------
s
H
M
Q
§
W
/->
H
+10
+8
+6
+4
+2
0
-2
-4
-6
-8
-10
+10
+8
+6
+4
+2
1% = 0.010 ppm = 26.1 Mg/m3
-
O w V w ^ °
-
-
-
1 1 1 1 1 1 1 1
1 23 4 15 16 17 18 19 20
TIME (DAYS)
_ O
0
-
_
-4
-6
-8
-10
-------
xs
DRIFT
§
w
TJ.U
+8
+6
+4
+2
0
-2
-6
-8
-10
1% = 0,010 ppra - 13.9 ug/m3
~
1 1 | 1 1 1 1 1 1
1 2 34 15 16 17 18 19 2
TIME (DAYS)
H
fc
M
04
Q
1
+1U
+8
-f6
+4
+2
0
-2
-4
-6
-8
-10
1
i
I
(
1
A P 9 P
V O
>
1 1 1 1 1 1
1 2 34 15 16 17 18 19 20
TIME (DAYS)
TRACOR GC H?S ANALYZER DRIFT DATA (RANGE =1.0 ppm)
195
-------
x-x
B^
^s
H
fe
M
Pi
«
s
PQ
W
N
+1U
4-8
+6
_1_ /
+4
+2
0
-2
4
-6
-8
-in
1% = 0.20 ppm « 228 ug/m3
-
-
0p ». / ft A
^ ^^^w w O
1 1 1 1 1 1 1 1 1
478
TIME (DAYS)
10 11
12
+10
+8
+6
S 0
Q
CO
-2
-4
-6
-8
-10
i
1 2 3 4 7 8 9 10 11 12
TIME (DAYS)
ANDROS CO ANALYZER DRIFT DATA (RANGE = 20.0 ppm)
196
-------
/x
H
M
OH
Q
8
W
N
t-iu
+8
+6
+4
+2
0
-2
-4
-6
-8
-10
1% = 0.050 ppm =57.0 yg/m3
-
-
-
9
** I A _
6
-
i r i i i i i i
12 3456789
TIME (DAYS)
OS
Q
O-i
CO
+10
+8
+6
+4
: +2
I
0
-2
-4
-6
-8
-10
i
i
i
8
1234567
TIME (DAYS)
BECKMAN GC CO ANALYZER DRIFT DATA (Range =5.0 ppm)
197
-------
+10
+8
+6
+4
S +2
H
w
-2
-4
-6
-8
-10
ii
1% = 0.20 ppm = 228.0 Mg/nr
Ml
?
1 T T
I I I I I I I
3 47 8 9 10 11 12
TIME (DAYS)
-1-10
+8
1
rt 9
1
-2-
-£
-10
i
4i-Hi
J i !>
1
1
10
11 12
3 4
TIME (DAYS)
BENDIX CO ANALYZER DRIFT DATA (RANGE =20.0 ppm)
198
-------
/-N
H
M
g
§
w
M
E
M
0
5
CO
+10
+8
+6
+4
+2
0
-2
-6
-8
1% = 0.050 ppm = 32.7 ug CH /m3
-
-
- f . T .
-I
-
-
1 1 1 1 1 1 1 1 1
123456789
TIME (DAYS)
1 1 0
T J.U
+8
+6
+4
+2
n
v/
-2
-4
-6
-8
-in
-
-
h 1
1w 50
-
1 1 1 1 1 1 1 1 1
123456789
TIME (DAYS)
BECKMAN GC THC ANALYZER DRIFT DATA (RANGE =5.0 ppm)
199
-------
H
fu
+10
+8
+6
+4
+2
0
§
w -2
-4
-6
-8
-10
rr
1% = 0.050 ppm = 32.7 yg CH
I
I
I
I
I
10 11
12 13 41
TIME (DAYS)
42 45 46
52
53 54
H
l-u
to
+8
+6
+4
+2
0
-2
6-
8-
I
I
I
I
J_
I
I
I
I
I
I
I
I
42
45 46 52 53 54
45 10 11 12 13 41
TIME (DAYS)
BENDIX GC THC ANALYZER DRIFT DATA (RANGE =5.0 ppm)
200
-------
M
DRIFT
§
w
N
TJ.U
+8
+6
-1-4
+2
0
-2
-4
-6
-8
-10
1Z = 0.050 ppm =32.7 ug/m3
-
-
-
w ^ 0 °
-
i i i I i I i I i
123456789
TIME (DAYS)
^
H
M
oi
Q
^
PH
t J.V/
+8
+6
+4
+2
0
-2
-4
-6
8
-in
-
o
?
d ° 6
-
-
-
i i i i i i i i i
456
TIME (DAYS)
8
BECKMAN GC CH ANALYZER DRIFT DATA (RANGE
201
5.0 ppm)
-------
H
fn
+10
+8
+6
+4
\
3
' +2
0
§ -2
w
NI
-4
-6
-10
1% = 0.050 ppm =32.7 pg/nf
I
I
I
I
10 11 12 13 41 42
TIME (DAYS)
45 46 52 53 54
+10
+8
+6
+4
s
' +2
_2
-4
-6
-8
-10
g
CM
CO
I
I
I
I
I
I
I
I
I
I
10 11
12 13 41 42
TIME (DAYS)
BENDIX GC CH, ANALYZER DRIFT DATA (RANGE 5.0 ppm)
4 202
45 46 52 53 54
-------
APPENDIX E
SUMMARY OF CATA FORWTS
(HP-COMPUTER OUTPUTS)
203
-------
FEBRUARY 20, 1973
1 -.105920
2 -.105950
3 -.000550
4 -.002770
5 .004550
6 .003040
7 .793650
8 .045270
9 .026620
10 -.744970
11 -.108650
12 -.548480
13 .002010
14 .002300
15 .638510
16 15.C26998
17 -15.077000
18 -.000450
19 -.002070
20 .000020
21 -.034740
22 .004030
23 -.028810
24 .030690
25 .011160
26 .036540
27 .010750
28 -.024930
29 .003230
30 .045160
31 .003590
32 .003220
33 .023660
15:20 HOURS
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
.000010
.000010
.093640
-.007300
.205060
4.669600
3.525200
-.001150
-.004450
.001250
-.000870
-.000680
-.105920
-.895610
-.896400
-.897790
-.106630
-.105930
-.105970
-.105950
-.610430
-.106040
-.105900
-.105970
-.105890
-.106130
-.512430
-.513720
-.513350
-.512640
-.513900
-.512690
Printout of Voltages Measured
by Data Acquisition System
(Long Format)
204
-------
MEL. NO
MEL.N02
MEL.NOX
BEN. NO
BEN.N02
BEN.NOX
BEN.03
BEN.S02
TRAC.TS
TRAC.H2S
TRAC.S02
MEL.S02
BEK.THC
BEK.CH4
BEK.CO
BEK.THC
BEK.CH4
BEK.CO
DIAL 1/3
DIAL 2/4
MEL. NO
MEL.N02
MEL.NOX
BEN. NO
BEN.N02
BEN.NOX
BEN.03
BEN.S02
TRAC.TS
TRAC.H2S
TRAC.S02
MEL.S02
BEK.THC
BEK.CH4
BEK.CO
BEK.THC
BEK.CH4
BEK.CO
DIAL 1/3
DIAL 2/4
$$$$$$$$$$
$$$$$$$$$$
$$$$$$$$$$
.0131
.0089
.0198
.0461
.0013
.0092
.0000
.0039
.0043
.0389
-.0020
.0830
4.6034
3.9192
-.0882
-.0033
-.0000
.50000
.50000
.50000
.50067
.47600
.49730
32915
12518
18479
16929
19732
98331
40000
40000
40000
03635
12646
01638
-19.56100
-4.77980
4.
2.
1.
1.
1.
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X +
* X 4-
* X +
* X +
* X +
* X +
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
ROTAT
ROTAT
.00000
.00000
.00000
-.00057
.00024
-.00013
.00104
-.00664
-.00050
-.00057
-.00058
-.01932
.00000
.00000
.00000
-.20802
-.07433
-.08711
-.01408
-.01329
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
ROTAT
ROTAT
Printout of Transfer Coefficients
When Applied to the Voltage
Output of the Instruments Yields
the Measured Gas in Engineering
Units (PPM). These Coefficients
are Updated Automatically During
Recalibration Procedures.
m ^^*~ b
y » tax + b *
where
y is the concentration in ppm
x is the voltage output of the
instrument
205
-------
FEBRUARY
-.106
-.123
-.026
.006
.009
-.106
-.514
MEL. NO
MEL.N02
MEL.NOX
BEN. NO
BEN.N02
BEN.NOX
BEN. 03
BEN.S02
TRAC.TS
TRAC.H2S
TRAC.S02
MEL.S02
BEK.THC
BEK.CH4
BEK.CO
BEK.THC
BEK.CH4
BEK.CO
DIAL 1/3
DIAL 2/4
20, 1973
-.106 -.001
-.782 .000
.004 -.029
.003 .022
-.003 .000
-.106 -.106
-.513 -.513
$$$$$$$$$$ L
$$$$$$$$$$ L
$$$$$$$$$$ L
.0127
.0082
.0187
.0463
-.0006
.0056
.0004
.0000
.0023
.0351 T
-.0041 T
.0798 T
1.9071 T
.0407 T
-.0783 T
-.0031
-.0000
/
/
Printout of the
Voltage Measured on the
65 Channels of the
Data Acquisition System
-.003 .001 .000
.802 .037 .027 -.750
.002 .639 15.026-15.076 -.000 -.003 .000
.026 .017 .038
.000 -.000 .088 -
-.001 -.001 -.106 -
-.610 -.106 -.106 -
.010 -.025 .003 .033
.010 .200 2.041 .102
.896 -.896 -.896 -.107
.106 -.106 -.106 -.512
-.514 -.513
PPM
PPM
PPM
PPM
PPM *
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
PPM
ROTAT
ROTAT
Printout of up to 20 Instruments.
Output in Engineering Units.
$$$ Indicate Instrument Output Is
Offline.
206
-------
Hourly Averages
Printed Once per Hour Automatically
R.T.I. INSTRUMENT
FEBRUARY
TIME
15 0
15 5
1510
1515
1520
1525
1530
1535
1540
1545
1550
1555
20, 1973
BEN.S02
PPM
.001
.001
.002
.001
.000
.001
.000
.001
.001
.001
.001
.001
EVALUATION
TRAC.TS
PPM
.007
.008
.009
.009
.008
.008
.009
.008
.008
.009
.008
.008
STUDY
TRAC.H2S
PPM
-.000
.000
.000
.000
.000
-.000
.000
.000
-.000
.000
.000
.000
Code for Status of Instrument
(Blank for Ambient Air Data Point)
TRAC.S02
PPM
.000
.002
.002
.004
.000
-.000
.000
.002
-.000
.002
.002
.002
MEL. SO 2
PPM
.004
.005
.005
.004
.004
.004
.004
.005
.005
.005
.004
.004
BEK.THC
PPM i
r
.037T
.038T
.038T
.039T
.037T
.037T
.038T
.038T
.038T
.037T
.037T
.037T
BEK.CH4
PPM >
-.0021
-.0011
-.0021
-.0021
-.0031
-.0021
-.0011
-.0031
-.0031
-.0041
-.0031
-.0021
AVERAGE
.001
.008
.000
.001
.004 $$$$$$$ $$$$$$$
$$$ - Indicates Less Than 9
Valid Data Points During Past
Hour, Hence No Average Is Computed
t
i
207
-------
Hourly Averages
Printed Once per Hour Automatically
R.
T.I. INSTRUMENT
FEBRUARY
TIME
15 0
15 5
1510
1515
1520
1525
1530
1535
1540
1545
1550
1555
20, 1973
MEL. NO
PPM i
EVALUATION STUDY
MEL.N02
' PPM ^
MEL.NOX
PPM ' '
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$L $$$$$$$L $$$$$$$L
$$$$$$$! $$$$$$$L $$$$$$$L
Code for
Status of
(Blank for Ambient
BEN. NO
PPM
.013
.014
.013
.013
.015
.013
.012
.014
.012
.013
.013
.013
BEN.N02
PPM
.005
.006
.007
.009
.006
.004
.005
.007
.006
.007
.008
.008
Instrument
Air Data Point)
BEN.NOX BEN.
03
PPM PPM
.017
.018
.019
.020
.018
.018
.017
.019
.018
.018
.019
.019
047
048
047
046
048
048
047
046
046
046
046
046
AVERAGE $$$$$$$ $$$$$$$ $$$$$$$
1
k t
i, t
k
$$$ - Indicates Less Than 9
Valid Data Points During Past
Hour, Hence No Average Is Computed
.013
.007
.018
.047
208
-------
Hourly Averages
Printer Once per Hour Automatically
R.T.I. INSTRUMENT EVALUATION STUDY
FEBRUARY 20, 1973
BEK.CO BEK.THC BEK
TIME PPM v PPM u P]
15 0 .080T 4.534T 3
15 5 .079T 4.560T 3
1510 .081T 4.599T 3
1515 .083T 4.603T 3
1520 .082T 4.631T 3
1525 .080T 4.660T 3
1530 .080T 4.667T 3
1535 .084T 4.693T 3
1540 .083T 4.720T 3
1545 .082T 4.712T 3
1550 .080T 4.742T 3
1555 .079T 4.740T 3
Code for Status of Instrument
(Blank for Ambient Air Data Point)
.CH4 BEK.CO DIAL 1/3 DIAL 2/4
'M ^ PPM ROTAT ROTAT
.942T -.088T -.003 -.000
.908T -.087T -.003 -.000
.919T -.089T -.003 -.000
.919T -.088T -.003 -.000
.897T -.088T -.003 -.000
.874T ~.088T -.003 -.000
885T -.088T -.003 -.000
885T -.088T -.003 -.000
.863T -.088T -.003 -.000
863T -.087T -.003 -.000
.908T -.088T -.003 -.000
863T -.088T -.003 -.000
AVERAGE $$$$$$$ $$$$$$$ $$$$$$$ $$$$$$$
-.003
-.000
t t t
$$$ - Indicates Less Than 9
Valid Data Points During Past
Hour, Hence No Average Is Computed
209
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO
EPA-650/2-74-019
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Development and Testing of an Air Monitoring System
5. REPORT DATE
December 1973
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
C. E. Decker, T. M. Royal and J. B. Tommerdahl
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG "\N1ZATION NAME AND ADDRESS
Research Triangle Institute
Research Triangle Park, N. C. 27709
10. PROGRAM ELEMENT NO.
1A1003
11. CONTRACT/GRANT NO.
68-02-1011
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
National Environmental Research Center
Research Triangle Park, N. C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Contract Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The purpose of this contract was to test and evaluate instrumentation based on
specifed measurement principles selected for the Regional Air Pollution Study (RAPS),
where an extensive network of air monitoring stations equipped with state-of-the-art
monitoring instrumentation and sophisticated data acquisition and computer pro-
cessing systems will be required. The primary objective of the program was to
equip an experimental trailer with selected instrumentation and to evaluate these
instruments at a non-urban site to: determine if the monitors selected for
the study can meet the required performance specifications or need to be modified:
determine the operating environment needed to obtain optimum performance from
these monitors; evaluate the latest calibration techniques and select calibration
procedures to provide the most reliable measurements; and recommend, based on the
results of the evaluation program, instrumentation for use in the RAPS program.
Instrumentation for the measurement of ozone, sulfur compounds (sulfur dioxide,
hydrogen sulfide, total sulfur), nitric oxide, nitrogen dioxide, hydrocarbons (total
hydrocarbon, methane, non-methane hydrocarbon) and carbon monoxide in ambient air
were included in the program. The evaluation of each instrument included in the
program was a systematic comparison of its ability to obtain reliable data with
emphasis placed on implementation, if possible, of design changes to instrumentation
that could not meet the operational requirements specified for this program.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
Instrument Evaluation
Performance Specification
Air Monitoring System
Data Acquisition System
Calibration
Maintenance
COSATI Field/Group
IS. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
214
22. PRICE
EPA Form 2220-1 (9-73)
210
-------
o
-o
n
2 ?
z «?
5 S
x o)
1"
m
f I
s s-i
lilt
-10 O 3
« (5 0) 5
(D c n
153
»is3
Z 2 -. m
413
pfsi
N SS
N g
CD
n
-<
ID
(Q -
------- |