vvEPA
   tfJPFp
   f. :  „* f »
   fr,' " ' *• , ';»*'$*
              United States
              Environmental Protection
              Agency
              Environmental Sciences Research
              Laboratory
              Research Triangle Park NC 27711
EPA-600/4-79-031
May 1979
              Research and Development
Regional  Air
Pollution  Study

Quality Assurance
Audits

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U S Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology  Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are.

      1.  Environmental  Health  Effects Research
      2  Environmental  Protection Technology
      3  Ecological Research
      4.  Environmental  Monitoring
      5.  Socioeconomic Environmental Studies
      6  Scientific and Technical Assessment Reports (STAR)
      7  Interagency Energy-Environment Research and Development
      8  "Special" Reports
      9.  Miscellaneous Reports

This  report has been assigned to the ENVIRONMENTAL MONITORING series.
This  series describes research conducted to develop new or improved methods
and  instrumentation for the identification and quantification of  environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                             EPA-600/4-79-031
                                             May 1979
         REGIONAL AIR POLLUTION STUDY

           Quality Assurance Audits
                      by
                  0. Klein
                 F. Littman
            Air Monitoring Center
           Rockwell International
         11640 Administration Drive
            Creve Coeur, MO 63141
           Contract No. 68-02-2093
                Task Order 106
               Project Officer

             Stanley Kopczynski
      Organic Pollutant Analysis Branch
 Environmental Sciences Research Laboratory
      Research Triangle Park, NC  27711
 ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
     OFFICE OF RESEARCH AND DEVELOPMENT
    U.S.  ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711

-------
                                 DISCLAIMER
     This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendations
for use.
                                      ii

-------
                                   ABSTRACT
     RAPS Quality Assurance audits were conducted under this Task Order
in continuation of the audit program previously conducted under Task
Order No. 58.  Quantitative field audits were conducted of the Regional
Air Monitoring System (RAMS) Air Monitoring Stations, Local Air Monitoring
Stations (State of Illinois, St. Louis City, and St. Louis County), RAPS
helicopters and various measurement systems employed in the RAPS intensive
studies.  Audit results are reported for systems measuring NO, NO , 03,
S02, total sulfur, total hydrocarbons, CH^ and CO.               x

     An investigation was conducted on the effect of Teflon particulate
filters on NO, NOp, Oo, and S02 concentrations in sampled air.  Measured
sample losses are reported for synthetic pollutant -aair mixtures sampled
through new and used filters under both dry and humid conditions.  The
investigation also revealed effects of humidity on the response of
analyzers to the various pollutants.

     The accuracy of S02 calibration mixtures prepared with the commercial
dynamic calibration system employed in the audits was investigated.  As
a result of this investigation and experence gained during the audits
the calibration system was modified to improve performance under field
conditions.  The air scrubber and dilution air flow measurements and control
were modified to eliminate the undesirable effects of uncontrolled ambient
temperature and humidity.  A new permeation tube holder was designed and
constructed to eliminate a recurrent air leakage problem.
                                      m

-------
                                   CONTENTS

Abstract                                                                 111
Figures                                                                   vi
Tables                                                                   vii
1.0  Introduction                                                          1
2.0  Summary                                                               2
3.0  Quality Assurance Audits                                              4
     3.1  Instrumentation                                                  4
     3.2  Van Preparation                                                  5
     3.3  Van Operation                                                    7
     3.4  Bendix Portable Calibrator                                       7
     3.5  Audit Standards                                                 24
     3.6  Results and Discussion                                          26
          3.6.1   Data Analysis Methods                                   26
          3.6.2   Audit Results                                           35
4.0  Filter Study                                                         38
     4.1  Experimental                                                    38
     4.2  Results and Discussion                                          40
5.0  Calibrator Operating Variables; an S02 Study                         57
     5.1  Equipment                                                       57
     5.2  Permeation Tube Variables                                       60
     5.3  Temperature Excursions                                          62
     5.4  Results and Discussion                                          62
Appendices
     A.   Audit Procedures for RAPS Instrument Systems                    71
     B.   Modified Bendix Model 8861 Portable Calibration System          99

-------
                                  FIGURES
Number                                                                  Page
  1          Winnebago Wiring                                              6
  2          Installation of Instruments in Winnebago                      8
  3          Flow Calibration Equipment                                   10
  4          Orifice No.  1 Calibration Curve                              15
  5          Orifices 1 and 2 Calibration Curve                           16
  6          Orifices 1 and 3 Calibration Curve                           17
  7          Orifices 1 and 4 Calibration Curve                           18
  8         Orifices 1 and 5 Calibration Curve                           19
  9          Orifices 1,  2 and 3 Calibration Curve                        20
 10         Orifices 1,  2, 3 and 4 Calibration Curve                     21
 11          Orifices 1,  2, 3, 4 and 5 Calibration Curve                  22
 12         NO Capillary Calibration Curve                               23
 13         Typical Audit Data Record                                    27
 14         Schematic Layout for Filter Study                            39
 15         NO Pollutant                                                 45
 16         N02 + 03 Pollutant                                           47
 17         Ozone Pollutant                                              49
 18         S02 Pollutant  (Tracer)                                       51
 19         S02 + 03 Pollutant, S02                                      53
 20         S02 + 03 Pollutant, 03                                       54
 21         S02 Pollutant  (Meloy)                                        56
 22         Initial  Equipment Configuration for S02 Study                53
 23         Final Equipment  Configuration for S02 Study                  59
 24         Meloy S02 Analyzer Calibration                               61
 25         Typical  Strip  Chart Trace,  S02 Study                         66

-------
                                   TABLES
Number                                                                  Page
  1          Bendix 8861  Orifice Calibration Data                         11
  2         Data Summary - Rams Stations                                 29
  3         RAPS Helicopters                                             31
  4         St.  Louis City and County Stations                           32
  5         Illinois Stations                                            33
  6         Miscellaneous Audits                                         34
  7         Summary of Results for Particulate Filter Study              41
  8         Data Summary, NO Pollutant Monitor Labs Instrument           44
  9         Data Summary NOp + Ozone Pollutant Monitor Labs
              Instrument                                                 46
 10         Data Summary, Ozone Pollutant Monitor Labs Instrument        48
 11          Data Summary, SO- Pollutant Tracer Instrument                50
 12         Data Summary, Sulfur Dioxide Plus Ozone Pollutant
              Tracor and Monitor Labs Instruments                        52
 13         Data Summary, S02 Pollutant Meloy Instrument                 55
 14         Effect of Permeation Chamber Flow Rate on Signal Response    63
 15         Effect of S02 Permeation Tube Length                         65
 16         Effect of Cooling on Modified Bendix 8861P Calibrator        67
 17         Effect of Heating on Modified Bendix 8861 Calibrator         68
                                     v.n

-------
                              1.0  INTRODUCTION

     The primary purpose of this task order was the support of the RAPS
quality assurance effort.  This task order was a logical  extension of Task
Order 58 under contract 68-02-1081.   Under Task Order 58, quality assurance
audits were conducted in support of the 1975 Summer Intensive on the RAPS
helicopters, the MRI and Battelle airplanes, RTI and EMI  vans, EPA aerosol
trailer, RAMS stations, portable 0, and CO monitors, and  the gas chromatog-
raphy laboratory.
     In the initial  scope of work of this task order, emphasis was placed
on the preparation and deployment of various resources to perform calibra-
tions and audits of air monitoring systems employed in the RAPS 1976 winter
and summer field exercises.  These resources consisted principally of a
Winnebago Mobile Laboratory Van equipped with a complete  complement of
ambient air analyzers and a Bendix 8800 series portable calibrator.  Sub<-
sequent modification to the task order included:  an extension to provide
audit coverage to the 1976 Fall Intensive, selected audits of local agency
monitoring stations, a detailed study of the effect of the particulate
filters on the RAMS analyzers, and a detailed study and modification of the
Bendix portable calibrator.

-------
                                2.0  SUMMARY

     The work performed under this task order was divided into four major
areas:  first, preparation and calibration of the instrumentation and support
eouipment; second, audit of RAPS related instrumentation; third, modifica-
tions to the Bendix Model 8800 series calibrator; and fourth, special studies
concerning the particulate filters on the RAMS analyzers intake lines.  Also
included as an appendix are the audit procedures developed for this task
order.
     Preparation of the Winnebago Mobile Van consisted primarily of modifi-
cations to the electrical system and improved apparatus for securing the
instrumentation.  Calibration curves were prepared for the Bendix 8861D
calibrator.  The system was first leak-checked and then calibrated using
rising soap-film bubble meters.  Data were converted to STP (0°C, 760 mm Hg,
dry).  Standards for analyzer calibrations were obtained either directly
from the National Bureau of Standards (NBS) or traceable to NBS.
     Field audits were performed in support of the RAPS 1976 Winter and
Summer Intensives and of selected RAMS stations.  Audit results are given in
the text for RAMS stations, RAPS helicopters, Illinois agency sites,
St. Louis City and County sites and other miscellaneous audits.  In general,
most instruments were good (errors within 15%); however, in selected cases
results were poor (errors > 20%).
     An extensive study was also conducted on the effects of the Teflon
particulate filters placed in the inlet of the RAMS instruments measuring
atmospheric pollutants.  Both new and aged filters were studied under con-
ditions of low and high relative humidity.  Results indicate that:  moist
air alters the instrument response for NO, 0.,, and S02; fresh filters tend
to attenuate the results, particularly for ozone; and humidity did not
affect instrument response to zero air.

-------
     A thorough investigation of the Bendix 8861 calibrator's performance
in the delivery of SC^ was conducted.  NBS permeation tubes were used in
studying flow rate over the tube, tube length, attitude of the permeation
tube chamber and ambient temperature.  A Meloy SA185 analyzer was used to
monitor the results.  As a result of this study, a new chamber was designed,
fabricated, and installed in the calibrator to minimize leakage and provide
adequate purge air temperature control.

-------
                        3.0  QUALITY ASSURANCE AUDITS

     The St. Louis Regional Air Pollution Study was conducted to develop,
evaluate and validate air-quality simulation models on both regional and
local scales covering urban and rural areas, and stationary and mobile pollu-
tion sources.  For this reason, a comprehensive, accurate and readily
retrievable data base of pollutants was developed for future model testing
and validation.  To guarantee the quality of the data, it was necessary to
check the performance of the instruments used to make these pollution
measurements.
     The quality assurance program provided an in-depth audit of the instru-
mentation by checking the instruments with gases of known composition.  Any
discrepancy between the known gas composition and the concentration indicated
by the instrument was logged and reported within twenty-four hours.
Instruments which were checked included those at selected RAMS stations and
those of various RAPS Special Experiments, such as the RAPS helicopters.

3.1  INSTRUMENTATION
     The following instruments were used in this study:
          1)  Bendix 8861P
          2)  Bendix 8861D
          3)  Monitor Labs Model 8440 Oxides of Nitrogen Analyzer
          4)  Tracor Model 270HA Atmospheric Sulfur Analyzer
          5)  Meloy Model SA 185-2 Total Sulfur Analyzer
          6)  Bendix Ozone Monitor
          7)  Monitor Labs Model 8410 Ozone Analyzer
          8)  MRI Integrating Nephelometer
          9)  Rikadenki five pen recorder
         10)  Beckman 6800 Gas Chromotograph

-------
     Out of the above list, the following instruments were mounted in the
Winnebago Van:
          1)  Monitor Labs Model 8440 Oxides of Nitrogen Analyzer
          2)  Tracer Model 270HA Atmospheric Sulfur Analyzer
          3)  Monitor Labs Model 8410 Ozone Analyzer
          4)  MRI Integrating Nephelometer
          5)  Rikadenki five pen recorder

3.2  VAN PREPARATION
     To prepare the Winnebago Laboratory Van for field use it was first nec-
essary to strip out all wiring and accessory internal batteries.   It had been
demonstrated during the previous intensive effort period that the use of
batteries and an inverter to generate 115V power was unsatisfactory due to
excessive current drain.  The exterior-mounted ONAN power generator had been
installed but power connections were incomplete.
     All wiring was traced out and labelled and a color coded circuit diagram
was prepared (Figure 1) showing all of the breaker panels, generators and
switches necessary to the operation of the electrical system of the van.
The system was made operational using either the self-contained 115 volt
ONAN generator or 115 volts from an external source.
     With this current configuration, a switch-over from generator to an ex-
ternal source (and vice versa) is awkward because of the power interruption.
The system could be improved by the incorporation of a shorting-type transfer
switch which would provide for an uninterrupted power transfer.   This would
eliminate the necessity of relighting burners and awaiting restabilization
of the instruments which invariably drift when power is interrupted.
     The Winnebago Laboratory Van was put in dependable running order.  Both
gas tanks were removed and cleaned, all fuel lines were blown clean and
dried.  The carburetor was rebuilt and readjusted.  The carpet and draperies
were cleaned and a vinyl runner was installed to decrease carpet  wear and
finally, the cabinets were covered with walnut grain contact paper to improve
the appearance of the interior.  The continuous gas analyzers were installed
on a four inch thick styrofoam mattress pad and secured in place  using a two

-------
UJ
0£.
CO
o
at.
           s:
    r—     D- O- C
    -t->            o  c
    3     a)  oi c_j  ro
    o     13 -a     c
 OJ     O) T- -i- O O
 i- i —  S_ (/J  V) Q
 n3 i —  rx3 -P -P — 3
 Q. 03  CL 3  3 O :*£
CO 3 CO O O 
 I   I   I   I   I   I   I
i — CM co «* L
 S-
 o c
+J «3
 H3 C
 i- O
 0)
 C 3
 O> ii
ta LO
     I
    00
                                      O)

-------
inch square unistrut.  A compression spring was incorporated in such a way
that the instruments were held in place by spring compression (Figure 2).
In addition to the continuous gas analyzers, the van was equipped with a wall
mounted integrating nephelometer on separate manifold.  A Bendix 8861 cali-
brator, a Meloy S0? analyzer, and an alternate ozone analyzer were deployed
at various times during the course of audit activities on an as needed basis
and configured accordingly.  The Bendix 8861 calibrator was completely
portable and was removed from the van and relocated on occasion as audit
requirements dictated.

3.3  VAN OPERATION
     The EPA Winnebago Mobile Laboratory Van was deployed to various field
sites in the St. Louis area to perform audits of air monitoring systems
employed in the RAPS 1976 Winter and Summer field expeditions and of selected
RAMS stations.  These audits were performed under the technical direction
and, in many cases, in the presence of the EPA Task Coordinator.  Scheduling
of the audits was provided by the EPA Task Coordinator.
     The audits were conducted using the Bendix 8861 portable calibration
systerr,.  All spare equipment such as tubing, instrument parts, fittings,
test instruments (DVM, portable recorders) were conveniently carried in the
racks below the instrument shelf.  The van was also used as a mobile labora-
tory when it became necessary to audit/calibrate instruments brought in for
a check.  When fully operational, audit/calibrations were possible for NO,
N02, NOX, SOp, total sulfur, ozone, CH4> THC, CO and light scatter.  Audits
of CH^, THC and CO analyzers were generally accomplished utilizing standard
gas cylinders and/or Teflon bags.  When access to an (audit) analyzer was
required, a Beckman 6800 located in the RAPS laboratory was deployed.  Instru-
ment output was observed using a Hewlett-Packard DVM and continuously moni-
tored by connecting the output of each instrument to the Rikadenki recorder.

3.4  BENDIX PORTABLE CALIBRATOR
     The equipment used during calibration/audit work consisted of a Bendix
model 8861 portable calibration system capable of providing precise concen-
trations of 03, SO^, NO and N02-  The 8861 system is composed of a model
8861P permeation tube assembly and a model 8861D dilution air/ozone generator
                                      7

-------
o
CO
CO



CO
                            to
                            O

                            O
                                ce.
                                o_

                                ">

-------
assembly.   The 8861  system is entirely self-contained within two aluminum
carrying cases.
     The 8861P permeation tube assembly consists of a permeation tube
chamber, a pressure regulator/capillary flow system, temperature control
circuitry, and a battery pack.  Chamber temperature was maintained at 30°C
+_ .01°C either by line power during calibration work or by battery pack
during transport.  Chamber purge flow was provided by compressed air during
calibration work or by a small aquarium pump during transport.
     The 8861D dilution air/ozone generator assembly consists of a pump,  an
air drier, an air scrubber, pressure regulators, a series of temperature
controlled flow orifices, mixing chambers, and a temperature controlled,
constant current, ozone generator.  Prior to use of the 8861D,  it was neces-
sary to develop calibration curves for the flow controlling orifices.
Duplicate calibrations were performed at all available instrument pressures
from 30-100% of gauge on each of the five orifices.  Several orifice combina-
tions were also calibrated yielding eight calibration curves in all.
     All orifice calibrations were performed using the calibration train  out-
lined in Figure 3.  Initial leak checking of the 8861D was performed by
sealing the outlet line of the 8861D and pressurizing.  Leaks were located
and eliminated with the aid of an in-line mass flow meter.  Dry compressed
air was then allowed to flow through the 8861D orifice under calibration  and
the exiting airflow rate was measured with a moving bubble flow meter.
Since the soap film would add an unknown amount of water vapor to an
otherwise dry air stream and introduce measurement error, the exiting air
stream was saturated with water vapor before measurement.  Following the
determination of observed flow using the moving bubble flow meter, flow
rates were corrected to standard conditions of 0°C and 760 mm Hg, dry
basis.  All calibration curves generated were maintained in the Bendix
8861D instruction manual.
     Orifice calibrations were performed repeatedly during the course of
the audit program to insure the integrity of the orifice calibration data
in current usage.  Typical calibration data is presented in Table 1.
Typical calibration curves are presented as Figures 4-12.
                                     9

-------
                                         25 PSIG
   HUMIDIFIER
      AP
      MANOMETER
SOAP FILM
BUBBLE METER
  FIGURE  3.  FLOW CALIBRATION  EQUIPMENT
                   10

-------
           TABLE 1.  BENDIX 8861 ORIFICE CALIBRATION DATA

% Of
Scale
30
40
50
60
70
80
90
100
*Correction
ORIFICE
Time For
1000 cc,
seconds
122.6
113.0
104.2
97.4
91.8
86.8
82.8
79.0
Factor = 0.88029
NUMBER 1
Observed
Flow
cc/minute
498.4
530.9
575.8
616.0
653.6
691.2
724.6
759.5


Flow at
Standard
Conditions
438.7
467.3
506.9
542.3
575.4
608.4
638.8
668.6

  CF =
  1°
   T
                   P-Pw
where
To
T
P
Po
Pw
'K)
                                                             (continued)
Standard Temperature (273.2
Ambient Temperature; °K
Ambient Pressure, mm Hg
Standard Pressure (760 mm Hg)
Partial pressure of water at ambient temperature T, mm Hg
       Flow at Standard Conditions = CF x Observed Flow
                                 11

-------
TABLE 1 (continued)
I Of
Scale
30
40
50
60
70
80
90
100
30
40
50
60
70
80
90
100
30
40
50
60
70
80
90
100
ORIFICE NUMBER 1
Time For
1000 cc,
seconds
61.2
56.0
52.0
48.4
45.6
43.4
41.0
39.4
ORIFICE NUMBER 1
40.8
37.4
34.8
32.6
30.8
29.2
27.8
26.6
ORIFICE NUMBER 1
122.4
111.4
102.0
95.8
90.0
85.8
80.4
76.2
AND 2
Observed
Flow
cc/minute
980.4
1071.4
1153.8
1239.7
1315.8
1382.5
1463.4
1522.8
AND 3
1470.6
1604.3
1724.1
1840.5
1948.1
2054.8
2158.3
2255.6
AND 4
2451.0
2693.1
2941.2
3131.5
3333.3
3496.5
3731.3
3937.0
Flow at
Standard
Conditions
863.0
943.1
1015.7
1091.3
1158.3
1217.0
1288.2
1340.5
1294.5
1412.2
1517.7
1620.2
1714.9
1808.8
1900.0
1985.6
2157.6
2370.7
2589.1
2756.6
2934.3
3077.9
3284.6
3465.7
                                      (continued)



        12

-------
TABLE 1  (continued)
% Of
Scale
30
40
50
60
70
80
90
100
30
40
50
60
70
80
90
100
30
40
50
60
70
80
90
100
ORIFICE NUMBER
Time For
1000 cc,
seconds
120.6
110.4
102.0
96.0
91.0
86.2
82.0
78.6
ORIFICE 1, 2
152.8
140.0
129.4
121.6
114.2
109.0
103.4
99.2
ORIFICE 1, 2,
77.0
70.6
65.0
61.2
58.0
55.2
52.6
50.0
1 AND 5
Observed
Flow
cc/minute
2487.6
2717.4
2941.2
3125.0
3296.7
3480.3
3658.5
3816.8
AND 3
1963.3
2142.8
2318.4
2467.1
2627.0
2752.3
2901.4
3024.2
3 AND 4
3896.1
4249.3
4615.4
4902.0
5172.4
5434.8
5703.4
6000.0
Flow at
Standard
Conditions
2189.9
2392.1
2589.1
2750.9
2902.0
3063.7
3220.5
3359.9
1725.1
1882.8
2037.1
2167.8
2308.3
2418.4
2549.4
2657.2
3423.4
3733.8
4055.4
4307.3
4544.9
4775.5
5011.5
5272.1
                                     (continued)
        13

-------
                        TABLE 1  (continued)
                     ORIFICE 1,  2,  3,  4 AND 5
                    Time For           Observed           Flow at
x Of                1000 cc,             Flow            Standard
Scale               seconds            cc/minute        Conditions
 30                  52.0               5769.2            5069.3
 40                  47.4               6329.1            5561.3
 50                  44.0               6818.2            5991.0
 60                  41.2               7281.6            6398.2
 70                  39.8               7537.7            6623.2
 80                  36.6               8196.7            7202.2
 90                  35.0               8571.4            7531.5
100                  33.4               8982.0            7892.3
                                 14

-------
   100





    90-





    80-





    70-
LU
_l
«=t


"  60-
U_
O



T  50-


•z.


§  40-
QL

UJ

§   30

=c
CD
    10-
      400    50    500   50    600   50    700
                     CC/MIN.
         FIGURE 4.   ORIFICE NO. 1 CALIBRATION CURVE
                            15

-------
    100 -




     90-




     80 -




     70 -




0    60 -
o
oo
CD
     50 -
a    40 -j
LU
CD
Z3

SJ    30-





     20-




     10-
        700   800    900   1000   1100   1200   1300   1400


                          CC/MIN.







        FIGURE  5.  ORIFICES  1 AND  2  CALIBRATION  CURVE
                              16

-------
UJ
CD
Z
V—4
Q
ct
UJ
o:

LU
CD
100



 90 -



 80 -



 70-



 60 -



 50 -



 40 -



 30 -



 20 -



 10-
       1200  1300  1400  1500  1600  1700  1800  1900
                           CC/MIN.
        FIGURE 6.   ORIFICES 1  AND 3 CALIBRATION CURVE
                              17

-------
   100 -

    90 -

    80 -

^   70 -
o
oo
fe   60 H

o   50 -
•z.
t—(
I   40 -
LU
CD
I   30  -

    20 -

    10 -
       2000  2200  2400  2600  2800   3000   3200   3400
                          CC/MIN.
        FIGURE 7.  ORIFICES  1 AND  4  CALIBRATION  CURVE
                               18

-------
   100 -


    90


    80
uj   70
«=c
oo
u_   60
    50-
§   40
o;
    30


    20


    10
         ^i'i'i•r^i'|•|
      2000  2200   2400  2600  2800  3000  3200  3400

                         CC/MIN.
       FIGURE 8.  ORIFICES 1 AND 5 CALIBRATION CURVE
                             19

-------
LU
O
oo
  100  -



   90  -



   80  -



   70  -
u.  60  -
o
 1  50  H
CD
LU
CD
CD
40  -



30  -



20  -



10  -
              T     I      I      I      |      1
      1700    1900   2100   2300  2500 2700  2900
                       CC/MIN.
       FIGURE  9.   ORIFICES 1,  2 AND 3 CALIBRATION CURVE
                             20

-------
UJ
o
oo
O

<
UJ
CD
TOO




 90




 80




 70




 60




 50




 40




 30




 20




 10
      3400  3600  3800  4000  4200  4400  4600  4800  5000   5200



                             CC/MIN.
      FIGURE 10.   ORIFICES 1,  2,  3 AND 4 CALIBRATION CURVE
                               21

-------
  LU
  oo
  LJ
  CD
100 -


 90 -


 80 -


 70 ~


 60 -


 50 -


 40 -


 30 -


 20


 10 -1
        5000  5500  6000  6500  7000  7500  8000
                         CC/MIN.
FIGURE 11.   ORIFICES 1, 2, 3, 4 AND 5 CALIBRATION CURVE
                           22

-------
  TOO  -





   90  -





   80  -





   70  -
o
oo

u_   60  ~
o
50  -




40  -




30  -




20  -




10  -
Q

<
a:


LU
CD
                    234



                        CC/MIN.
                                1      r
                                        T
                                        6
     FIGURE 12.   NO CAPILLARY CALIBRATION CURVE
                         23

-------
3.5  AUDIT STANDARDS
     The calibration standards employed during the course of the audit pro-
gram were obtained either from the National  Bureau of Standards (NBS)  direct-
ly or from vendors supplying certification of analysis.   In the latter case
the vender's analysis was certified by the Environmental  Protection Agency
before the standard was submitted for use.  The following specific gas
sources and/or techniques were used during the program to assure quality in
the audit standards.

Sulfur Dioxide (SQ?)
     NBS Standard Reference Materials (SRM)  1625, 1626,  and 1627, were used
exclusively throughout the project.  The choise of a 10 cm, 5 cm, or 2 cm
permeation tube was governed by the operating range and sample flow rate of
the analyzer audited.  The source gas permeation tube was maintained at con-
stant temperature (30.00°C +_ ,05°C) during use.  Audit concentrations  were
prepared by quantitative dilution of the source gas with ultrapure air.

Nitric Oxide (NO)
     NBS SRM 1684, 100 ppm NO in nitrogen, was typically employed.  This
cylinder was equipped with a low internal volume, stainless steel regulator
which was routinely evacuated and purged prior to withdrawal of cylinder con
tents for audit gas preparation.  The cylinder was used either directly by
quantitative dilution and subsequent introduction of the audit gas to the
analyzer under test or it was used to certify a working standard.  In the
latter case, the NBS cylinder was used to calibrate an oxides of nitrogen
analyzer and a vendor-supplied NO standard was then certified by analysis
under identical instrument operating conditions.  The working standard, with
NBS traceability, was then used during field audits.

Nitrogen Dioxide
     Nitrogen dioxide was generated by gas phase titration of nitric oxide.
A constant-current, constant temperature, ozone generator within the portable
calibrator served to provide the ozone for the titration.  The procedure
employed to accomplish the gas phase titration is described in Appendix A.
                                     24

-------
Ozone (03)
     Ozone was generated within the portable calibrator as described above.
The concentration produced was a function of lamp current and was routinely
certified by gas phase titration of nitric oxide.  The oxides of nitrogen
analyzer aboard the van served to monitor each titration performed for the
purpose of certifying the ozone concentration being delivered from the
portable calibrator during an audit.  Frequent pre-audit and post-audit
certifications were performed to insure the integrity of the results
obtained in the field.

Total Hydrocarbons (THC). Methane (CH^). and Carbon Monoxide (CO)
     THC, CH., and CO audit gas concentrations were derived from cylinders.
Generally, the cylinders were vendor supplied and EPA certified.  The
standards were low level and prepared in air so that the cylinder contents
could be introduced to an analyzer without prior dilution.  The cylinders
were either employed in the field with the contents introduced directly to
the analyzer being audited, or used to prepare an audit gas sample in a
Teflon bag.  In the latter case, the bag was taken to analyzer to be audited
and the contents introduced to the analyzer.
     NBS SRM 1681, 1000 ppm CO in nitrogen was used on occasion to certify
the contents of vendor supplied CO standards.  NBS SRM 1658, 1659, and 1660,
CH. standards in air, were not available during the course of the study.

Zero Air
     The source of zero air most commonly used during audits was vendor
supplied cylinder air certified by the vendor to meet the following specifi-
cations:
                           THC (as CH4)   < 10 ppb
                           CO             < 10 ppb
                           NOX            < 10 ppb
                           S02            <  1 ppb

     This zero air was also used routinely as diluent air during the prepara-
tion of audit gas concentrations where dilution of a source gas was required.
                                     25

-------
On occasion, when an alternate source of dilution air was available, the
alternate source was substituted as diluent after quality was verified by
direct comparison to the cylinder gas.   This practice was limited to audits
involving S02, NOX, and O.^.   Audits involving THC, CH.,  and CO were performed
using zero air obtained only from the zero air cylinders.

3.6  RESULTS AND DISCUSSION
     Various field sites in the St. Louis area were audited.  Audits of air
monitoring systems employed in the RAPS 1976 winter and summer field
expeditions and of selected RAMS stations were also performed.  RAMS stations
were audited in accordance with schedules provided by the EPA Task Coordi-
nator.  A written report of each audit was prepared on the day following the
audit so that investigators could be advised of any errors discovered in
their instrument systems.
     Audits were run by supplying a carefully calibrated source gas to the
instruments under test.  Audits for ozone, NO-NO,, and sulfur dioxide were
performed using zero and four non-zero concentrations.  Audits for CO, CH.
and total hydrocarbons were performed using gases of known concentration
supplied from a certified cylinder.
     The results were recorded in tabular format.  A typical data record is
shown in Figure 13.  For each instrument, the concentration of the calibra-
tion gas was recorded in the "Auditor Value" column.  The corresponding in-
strument reading (in volts) was recorded in the next column.  This voltage was
translated  into concentration using the applicable instrument transfer
equation, which is of the general form
                                ppm = MX + b
      In the case of RAMS stations, the coefficients M and b were obtained
from the daily auto-calibration data reported in  the station printouts.
     For county and city stations, as well as helicopter instruments, the
corresponding conversion factors were obtained from the  respective personnel.

3.6.1  Data Analysis Methods
     The instrument values were then compared to  the audit  values by calcu-

                                     26

-------
                                   STATION 111
            NO
                                         NOv
                                              6/25/76
                                                  03
. .350
.250
.150
j zero
3.2324
2.2680
1.4184
.0683
N0/N02 CONVERTER EFF
Instrument ppm = A+B
A = -0.005 B
.326
.226
.138
-.002
ICIEN'CY:

.350
.250
.150
zero
96%
3.3129
2.3364
1.4697
.0683
.334
.233
.144
-.011



(Auditor ppm)
0.951 A = -o.Oll 3 = 0.996
r2= 0.999 Sv v= 0.005
SO,
i Auditor
i Value
1 (PP'-O
j





Instrument
Reading
(volts)





Instr.
Reading
(ppm)





|
A = B
r2= S v=
CH4
Auaitor
Value
Instrument
Reading
Instr.
Reading
;
2
r = 0.999 Sv x= 0.006
TOTAL SULFUR
Auditor
Value
(ppm)
.700
.500
.300
.150
zero

Instrument
Reading
(volts)
3.3077
2.2680
1.4111
.6420
-.0048

Instr.
Reading
(pnm)
.614
.420
.261
.117
-.004

i = -0.009 B = 0.880
r2= 0.999 Sv. v= 0.009
TOTAL rfM'ROC -\R30NS
Auditor
Value
Instrument
Reading
Instr.
Reading


.142
zero


1.1938
.0708


.128 !
.000 '

j
A = .or," B = . 2?.c,
r2= .9998 Sy x= .002
CO
Auditor
Value
fppra)
40.60
20.15
3.43
5.29
zero

Instrument
Reading
(volts)
Instr.
Reading ;
(ppr )
3.8330 | 4C.9& ,
1.9213
1.6943
2.2705
.0463

20.50 :
3.54 ;
4.78 .
C. 018
1
A = -0.120 B =1.014
r2= 0.9997 Sy v= c.3195
XEPHEI/OMETER
Auditor
Value
Instrument
Reading
Instr. ;
Reading :
Auaitor
Value
(ppm)
2.20
zero




Instrument
Reading
(volts)
.9057
.0268




Instr.
Reading
(ppra)
1.89
-.062




Auditor
Value
(ppm)
2.20
zero



	
Instrument
Reading
(vclts)
.8276
.0170




Instr.
Reading
1.68
-.041




Auditor
Value
fnp-1
. .00-. 03
.00-. 03
1.00±. 03
4.00t.25


Instrument
Reading
fvolts)
.36
. C16
2.566
10.61


Instr. !
Reading :
;
*
i
i
t
j
j
i
t
A = -.062
r2=
              Sy.x=
 .887    A =  -.041     3      .782
	r2=	3>'.x=	
                                                         A =
B
Sv.x=
                     FIGURE 13.   TYPICAL  AUDIT DATA RECORD
                                         27

-------
lating the coefficient of a  linear  regression equation between the two sets
of values using a linear regression program of a Hewlett-Packard 65 cal-
culator.  The equations used were
                                 Zy.      /EXA
                             A.-,,!  .B(V-)
and

                                ^y.;  —J—L
                             B  =
                                         (Zx.)2
                                  Zxi
where
             A   =   intercept of  regression line
             B   =   slope of regression  line
             x.  =   audit gas concentration i
             y.  =   instrument reading  i
             n   =   number of observations
     As used here, the symbol Z means  summation of values following the
symbol for index values, i, from 1  to  n.
                                                             2
     In addition, the square of the correlation coefficient, r , was deter
mined, which measures the degree of fit  of the points to the least squares
                      2
straight line.  When r  =1, the correlation is said to be perfect, with
                                                 2
all points falling on the regression line.  When  r  = 0, the relationship
between the two variables, x and y, is completely random.
                                                   2
     The equation used by the program to determine r  is
\        Exi^i
LEVi  -  -F—
                    r2-
     The coefficient B,  the  slope of the regression curve, is an indication
of the accuracy of the  instrument under test (assuming that the auditor
values are unbiased).   A slope  of 1.000 indicates complete correspondence,  a
slope of 1.0500 or .9500 indicates a concentration or instrument dependent
error of +5% and -5% respectively in the instrument under test.

                                     28

-------
oo
co
CSJ
o
CO

m
O


Q£
UJ
UJ
3
CSJ
o
z

X
z



i


_l
_J
o
a.
CSJ
CO
CSJ
CO
<

M
o
u.
u_
UJ
CSJ
l_

CO
<

CSJ
i.
CO
^




csj in c co r-.
LO vo cn ^ co
cn co en en cn
cn en cn cn cn
O VD O O O ^
O CTI Is* O »— r
r-. o o csj o c
r-. csj o o CTI c
Csj CT* *4D VO ^O
r— ^- o r^. csj
o o o csj o
o o o o o
1 1 1
CSJ CSI CSJ
II II II 111 II II 1 II
lllllllllillllll 1
oooooooooooooooo
oooooooooooooooo
1

S8<*§o,c* o a>OTm^c.a>
cn cn cn cn cn CT^ cn cn cn cn cn cn o 0*1 cn
"
CT> O CT^ «~~ o cn o r~~ cn o co cn CTI o o '
OOOOOO OOOOOOOOO
III II II 1 1 1 1
cn CT* CT* CT> CTI cn cn cn en cn cn cn cn cn cn
LCI TO •-" ^ o o cn vo ^ i— LO co to o vo *
LOLOLOOLon LOCsiCTiocno«*csji — <
cnocnooo oocoocoocnoo
nLOr— csjr^^o cncsj<3-cooor-ocn
SOOOO-— OOOOOOOOO
OOOOO OOOOOOOOO
1 1 t 1 1 1 1 1 1 I 1 1
^"^CSJCSJCSJCSJCOr- CSJi— p- t— ,— -^C^^.^
^******* *J J =
^CSJOr-r-r-OOCSJ.— ^-^Ot— CSIOO 3
™
n oo
- CSJ
S^o
.—

5 *t
y» o





•^. in
sj CO
~> o



=a- csi
* n
3% IT)
T. O


ANDARD
DEVIATION
* ft

                                                                                 -a
                                                                                 
-------
        ror^-ococnco       o        com       en   «3-  r-*
        !^Q   CT>   CTi   CTi  CT>   CTi       O        CTt  i—•       O"i   CO  CO









    r-x  CTt   O   CD   i—  O   O       O"i   O   CTi  O  CO   CO   CT>  O







        §vOCMCOCMi—   I—       O1—   CTt  ^D  Ul   O   ^~  M
        O   -—   O   CM  .—   CM       OOOOOOOOJ

              III         I                            -—- I     1.—* I
                                                               cxi       CM



















    rOOOOOOO        OOOO^OCTtCO'X)

    O  VO  IO   lO   <—  CMLf>        i—  CMOCMF-x^-fOi—
    OCMOOOOO        OOOOOOOO

          III              I         I              II         I






    cr>  cr>                                                 cT*  CT>  cr>   o






    ff)  VO  O   O   O   O  O        O  f~~  CO   ^"  CO  O\  *f   i—
    •—  CMCMLDi—   CT>UD        COCTti—   i—  ^CMCOQ
    ijOoco^r^-LOuo        cMr-»LOor^-i—  cno









    OOOOOOO       OOOOO-—  OO

          II                   I                   II













    LT)  VO   CTN   LO            *O       fl       I«O  O   00       r"   LO







    '—  *^   S   ^            ^       Lf>       NlOCM       O^*


    8888            8       8       080       80

          III            ^_l       ^_^       ^^  ^_,  ^-^       ^^.  	

             CMCM            CM       CM        CMCMCM        CMCM



                                                                  .—  CM   CO
    ^OLncocno            cr>r^cocriLni—  i—   i—
    i—  CMCMCMCM<"Oi—   CMi—   i—  "—  '—   ^^   "^  ****.   "^




              *    *             *                                *         *
0-^-K-K****                           **         *

         Oi—   i—  i—   OCCMi—   i—  •—  Oi—   CMOO
•a
 OJ
 13
 c
 o
 u
CM
CO
OCA
                                                                                                           iO    CO

                                                                                                           1    «
           _

           o
NDARD
EV I AT
                                                            30

-------
oo
Q.
o
co

UJ
,_r
z
i


0




=>
R
_J
D



n
0


VERIER
g
CM


r





o
z





,_
H-
=

O-
K-
e»
z
CM

CO
«t

^

00
-a:


t-
03


*e
criai e
I llisl Isll
•— I CM ^- f*l r- r-
CM o«9"or^oicoficOLnPO^-aiLf)
r^, cOr— voaicTicococoaicrico^DCO
o
o ijonorocMO"— o> — aiinoo ^
i i i . ^
a
O ooooooooooooo
• . i i . i

SSsSsSslslSssi
_a J3^i-Q -Q^i^3J2^1ia^J3JD^
8^~OOOO^-OOOOOOOO
ooooooooooooo
II 11 1
OOO oooo"»r-^{sjr-*.io
'•'§g§!'|S|S;SSS§
III II III II 1 II 1 1

Iliillillllllil 1
ooooooooooooooo
1 1 1 1 II

Iliillillllllil

||l||||g|||||5| I
OOOOOOOOi — i — OOOOO
OOOOOOOOOOOOOOO
111 111 r i i
uj«3^«r - * ^ r^ OMr-r.co-
^•^^^-^-^-^.^-^'^.•^.-^Of-r- t— t—

Of
SlsssssKsgSsssss

|^ 000000000000 = 00 L




i a*
1 t~^
j ro
1 O



? s
i S!
3 O
1 £



' U3






ii





i -
K S


z
l-l
eC
S
o
o;
g
? ^
y g
£ (^
                                                                                    OJ  1/1  «—
                                                                                    I-  o  3 :
                                                                                    a. a,  
-------
                                                                 Ol£J«—


                                                                 CT>O^CT>
                                     oocrm
                                     «j  CSI   OC  «3
                                     Lnc^^-
                                     O  •—   M  »—
                                                                 cooocch
                                                                 OC^l/llOU")
                                                                 c   f—  i —   CTich^.
                                                                 O   •—  m   »—  O   CO
                            OOOCOO
                            cocrccojccn
                            «3LT)cnr^cri^-
                                                                 OC3O
                                                                          OOOO
                                                                          •—   o    r—  o
                 jl
                 k
to

o
to

I
o

o
 o

 to
 I-H

 o
 CO
 CO
                            oooo
                            «*fOO>CM
                            OCOO
                            CD  O   C  O
                            sags
                                                                 ooo
                                                                 LT>   (7i  »—
                                                                 o   o  o   o
                                                                 o   o  c   o
                                                                 u">   O  O   O
                                r-   co  ,=•  co
                                CTI   CTI  o-  cr
OOOO

 LO   r--  CTI   r--.

 CTI   i—  •—   «a-
                                     O  O  10

                                     CM  cr  <£>
                                                                 cxjr-r-
                                                                   -
                                                   IT)  lO
                                                   o  o
                                                   o  o
                                                                                   c   c
                                                                                    CO  •—
                                                                                    CVJ  CO
                                                                                    VO  CM
                                                                                    CO   f^>
                                          in  c
                                          cc  r-^

                                          c  o
                                                                  CD  O   O  O
                                                                  CD  r~-   r-.  o
                                                                  CM  LT1   i—  O
                                                                  CM  r-   O  C
                                                                  OCOO
                                     ^  CO  CM
                                                                  OCOO

                                                                  •3-   i£j  m   CM
                                      ooo

                                  ooooo
                                                        o
                                                        z    >-
                                                                   ooooo
                                                                                                                    ^  £   ~
                                                                                                                    o       >~
                                                                                                                    .    „   i
                                                                                                                    CO*
                                                                                                                             ••-   O
                                                                                                                             •4-J  •(—
                                                                                                                             O  -M

                                                                                                                             J-   3
                                                                                                                         O   «   T)

                                                                                                                     N
                                                                                                                             >   N
                                                                                                                    >,   1-   r-   >>


                                                                                                                    "(O   N   C   (O
                                                                                                                    C   X  «t  ^

                                                                                                                    "*  'ro    CM



                                                                                                                    O
                                                                                                                    N   OJ   C   C


                                                                                                                         O   W   U
                                                                                                                    X   N   "-  -r-
                                                                                                                    -r-  O   C   C

                                                                                                                    -a       -c  £
                                                                                                                    C  (->   U   U


                                                                                                                    CO  Z   h-  h-
                                                                    32

-------
oo
z:
o
I-H

 o co o~» ^"
\o o os r-*. r^.
EC- E E E
«d" r- ^ CXD CM tO V£J
O 00 O> CT> O^ O^ O
0% o m n PX.
O ^" Ch CT> ^~
CO VO OJ CM ^>
00 O O O
1 1
1 1
•O -D
CO O CM *d-
o o o •*
o cc en •-
O CTi CTi O
CM O
o o
o o
O 10
3 g
o o o «a-
co cyi i — in
i — «d" CO CM
CO CO 00 O
§o
CO
\O CO
*!• CO
CM CM



o
UJ O O h-
(— i— *— CO CO i— r-  -C
                                                                                  II  II  II

                                                                                 -o E a-
                                        33

-------
CO
t-H
o
co
o
UJ
O
co
<£>
CQ
                                     34

-------
3.6.2  AUDIT RESULTS
     Tables 2-6 give the results for the audits of RAMS station instruments,
RAPS helicopter instruments, St. Louis City and County stations, Illinois  EPA
stations, and miscellaneous audits.
     Audits of the RAMS station instruments indicate that
     1.   The Monitor Lab NO NOX instruments show an average slope of .9944,
          with a standard deviation  of only .0532 for NO.  For NO^ the slope
          is 1.0247, the SD is somewhat larger at .0815.  With one exception,
          NOp converter efficiency was found to be 97% or better.
     2.   The Monitor Labs ozone instruments show a somewhat lower slope
          with audit values .9287, with a SD of .0646.
     3.   The Tracer sulfur dioxide  instrument gave an average slope of
          .9815.  However, this number is somewhat misleading since individ-
          ual instruments were off by large amounts, as indicated by an SD
          of .1623.
     4.   The total sulfur measurments from the Meloy and Tracer instruments
          gave a rather low average  slope of .8706 with a SD of .1263.  The
          Tracer S02 analyses were generally in closer agreement with the
          audit test gas values.
     5.   The Beckman 6800 total hydrocarbon analyzers gave an average
          slope of .9205 with a SD of .0933.
     6.   The Beckman 6800 methane analyzers gave an average slope of
          .9251 with a SD of .0717.
     7.   The Beckman 6800 carbon monoxide analyzers gave an average slope
          of .9795 with a SD of .0905.

     Helicopter audits are summarized in Table 3.  The results are similar
to those obtained at the RAMS stations.
     1.   The Monitor Labs recorders gave an average slope of .9760 for NO,
          .9804 for NOX.  The standard deviations were .0817 and .0712,
          respectively.  These results compare well with the RAMS stations.
                                     35

-------
         Inspection  of the  data  indicates  a  negative calibration bias prior
         to 1  November,  which  changed  to a positive bias after that date.
     2.   The REM Scientific Company  ozone  analyzer showed an average slope
         of 1.0153,  with a  SD  of .1569.  The average slope  is somewhat
         better than that of the RAMS  sensors,  however  the  SD is greater.
     3.   The Meloy Total  Sulfur  recorders  gave  an average slope of  .9230
         with a SD of .2168.   This is  comparable to the performance of the
         RAMS Meloys.
     4.   The carbon  monoxide analyzers gave  an  average  slope of  .9255,
         with a SD of .0379.
     Helicopter audits were conducted  pre-flight and post-flight in an effort
to quantify the degree to which  instrument drift may have occurred as a
result of the rigors of flight.   The agreement  between  pre-flight and post-
flight audits was generally good with  the  exception of  the  ozone analyzers.
     St. Louis City  and County Stations use  a wide variety  of instruments,
as indicated on the  summary tables.  The data were, nevertheless, summarized
for each pollutant,  since any  future intercomparison of data will probably
not take instrument  types into account.
     A number of the City and  County sites employ relatively old analyzers
of the wet chemical  variety with such  long response times that multipoint
audits could not be  conducted  within a reasonable period of time.
     Data for NO indicate a large negative error  (average slope  of  0.4330),
with a large standard deviation (.3047).   Data  for NOp, taken only  at 2
stations, appear to  be unrelated to  actual concentrations.  Ozone data show
an average slope of  1.0867; however, the large  standard deviation of  .4099  is
an indication of the large  spread of actual  values.  Values for  total sulfur
indicate an average  slope of .9294,  but the  standard deviation  is again
large (.3586).
     The carbon monoxide analyzers gave an average  slope of 1.1288  with  a
rather large SD of  .1599.
     The total hydrocarbon analyzers gave  an average  slope  of  .7904 with
a SD of  .1993.
                                    36

-------
     Thus, the audit results indicate that data from these sources are
subject to large error and should be used with caution.
     Two Illinois State EPA stations were audited.   These stations have
Dasibi ozone analyzers, which gave a slope of .9902, with a standard devia-
tion of .0144.  The SO^ values obtained by Technicon IV  analyzers  yielded
a correlation of .9556 with a standard deviation of .0649.
     Miscellaneous audits included the audits of instruments used  on the
DaVinci balloon, and various aircraft and vans used by expeditionary groups.
The results, shown in Table 6, indicate slopes ranging from .460 to 1.2650.
     During the course of all audit work, a wide variety of analyzers,
support equipment, and monitoring configurations were encountered.  Addi-
tionally, the analyzers used for aerial studies were not always audited
in place in the aircraft.  In a number of instances, ambient temperatures
differed significantly from normal room temperature.  In all cases the
responsible investigator was consulted prior to the audit in regard to
instrument specifications relevant to the audit and to insure that the
introduction of audit gas would simulate the introduction of the monitored
pollutant during normal operation.  When the introduction of audit gas
through normal sampling lines was not possible, as  in the case where high
by-pass flows existed in sample lines, audit gases  were  introduced through
auxiliary ports specifically designed for calibration.
                                    37

-------
                            4.0  FILTER STUDY

     The primary purpose of this part of the task was to conduct a study of
the effect of particulate filters on measurements by RAMS analyzers.   The
pollutants of concern were sulfur dioxide (SCL),  nitric oxide (NO) and ozone
(03).

4.1  EXPERIMENTAL
     A study was made of the effects of new and used Teflon particulate
filters placed in the inlet of instruments measuring atmospheric pollutants.
The filters were 47 mm in diameter with an average pore size of 10 microns.
Both dry air and humidified air were used as diluent.  Gases studied  wereS02>
NO, Oo, N02 + Oo, SOp + DO-  NO was obtained by diluting the contents of an
NO in nitrogen compressed gas cylinder.  Ozone was obtained from a constant-
current ozone generator.  N0~ was obtained by the gas phase titration of NO
with ozone.  S02 was obtained from an NBS permeation tube.  Equipment was set
up in such a fashion as to allow rapid change from one filter to another and
from dry air to wet air (Figure 14).  The flow diagram shows airflow from the
cylinder to the Bendix calibrator, either directly or through a water bubbler.
Output from the calibrator entered a glass manifold for distribution.
Manifold air was sampled by an EG&G dew point hygrometer, the output of
which was monitored by a Leeds and Northrup recorder.  The dew point hygrom-
eter was checked by deriving relative humidity from outdoor dew point and
temperature measurements and comparison to relative humidity measurements
taken with a sling psychrometer.  Crosschecks were also made with relative
humidity measurements from the National Weather Service.   In no case was a
difference greater than 5% noted.
     The second line from the glass manifold led to a series of valves and
filters so that the gas stream could be shifted from one filter to another
(or  none) without  interruption or without opening the circuit.  The filters
were replaced during the changeover from wet to dry diluent air and again
                                    38

-------
                                                           LU  C3
                                                           >-  -z.
                                                                        o
                                                                          CM
                                                      o
                                                        C\J
                                                               CO
                                                               LU  CO  LU
                                                               —J   I   I—
                                                               h-  O  -Z.
                                                               O   O
                                                      O   LD  CO   Q.
                                                                             o:
                                                                             LU
                                                                             CO
                                                                             Q.
                                                                             CO
                                                                        CO  »-i
                                                                                      CO

                                                                                      a:
                                                           CO
                                                           LU
                                                       or

                                                       co
                                                                                      O

                                                                                      et


                                                                                      O
                                                                                      O
                                                                                      CO
                                                                                      «*


                                                                                      LU
                                                                                      o:
U-   Q.
     CVJ
      •
     o
                                     39

-------
during the changeover from pollutant to pollutant.   The "used" filters were
taken from station 120.  Filter 'age' varied from five to thirteen days.
All filters taken were obviously "used":  particulate matter was visible  on
all.
     Before the work was started a decision was made by the EPA Task Coordi-
nator to randomize the order of tests so as to remove any systematic bias as
the tests proceeded.  Furthermore, it was decided that two replicates were
required per test condition.  Proceeding in accordance with the above
directive the first randomization was made in the sequence of testing the
pollutants.  The order was as follows:
     1.   S02 (Tracor and Meloy)
     2.   N02 + 03
     3.   NO
     4.   03
     5.   S02 + 03
     The random sequence established for the individual filter tests is
obvious from the tables of results (Tables 8-13).  Nearly every test in-
volving a different pollutant and/or humidity condition was preceeded by
a  run using zero gas.  When the test was completed, the zero gas run was
repeated.  Concentrations of pollutants observed represent an average of
at  least 12 readings made 4 minutes apart after the system was judged to
have reached equilibrium.

4.2  RESULTS AND DISCUSSION
     Table 7 summarizes the results of  the particulate filter study.  In
comparing wet and dry  air results, it should be noted that no correction  for
the  volume added by water vapor has  been included.  This amounts to  1-1.5%
at the  relative humidities  encountered.  Reproducibility of duplicate runs
was  generally good, with standard deviations within a range of  .001  to .008
ppm.  Thus, changes of the  order  of  5%  are significant.
     The following  conclusions can be drawn  from the  results  for the pol-
lutants tested:
                                    40

-------
                                                    c\j       ,—
                                                  AA
        O       •—
          /\/\
        CM       ,—
             LP,      •—
                                                    C\J       ,—
                                                  /\/\
         i—      c\J
             ^      C-J
•— E    3
                  41

-------
Nitric Oxide (NO)
     An increase of relative humidity from 5 to 50% results in about 10% at-
tenuation of the response of the NO analyzer.   The presence of a particulate
filter appears to have no effect.

Nitrogen Dioxide NOp)
     Moist air appears to have no effect on the N02 reading obtained from
mixtures of N02 + 03<  A new particulate filter caused a 4% drop in the N02
reading for moist air, but had no effect on the N02 reading for dry air.

Ozone (0,)
     Moist air enhanced the response of the instrument to ozone by 10%.  A
fresh particulate filter resulted in a 23% decrease in ozone level.  A used
filter had a lesser effect (approximately 8%).  Analyzer response time was
significant.  As a result, some compromise in data quality occurred yielding
apparent inconsistencies such as two significantly different observations
taken on the same filter tested twice in succession.

Sulfur Dioxide  (SOp) - Tracer Analyzer
     Moist air  resulted in a significant signal attenuation (16%).  The in-
troduction of a  particulate filter did not affect the S02 level significantly.

Sulfur Dioxide  - Ozone Mixtures
     This mixture contained a much higher S02 concentration than the one
discussed above  (.252 ppm vs. .085 ppm).  However, the attenuation in  S0£
measurement due  to the presence  of moisture was about the same  (0.01 ppm).
The  particulate  filter produced  no significant change in the S02 measurement.
     No  change  in ozone level resulted from the use of moist air.  A new
particulate filter resulted in a 12% reduction of ozone level,  which dropped
to 5% when  a used filter was  introduced.

Sulfur  Dioxide  (SOp)  - Meloy  Analyzer
     This  instrument showed an enhanced  response to the introduction of moist
air  (11%);  as  before, the  introduction of a particulate filter  did not result

                                     42

-------
in significant changes.
     The results indicate:
     1.    Moist air alters  the instrument response to NO,  0,  and  S02-   This
          effect can be  alleviated by calibrating the instruments with  air
          of about the same relative humidity as that which will  be sampled,
     2.    Particulate filters, especially fresh ones, have a  pronounced
          attenuating effect on the ozone level.  This effect decreases as
          the filter becomes conditioned, but remains significant (5-10%).
     3.    None of the analyzers employed in this testing were sensitive to
          zero air humidity, with or without a filter.

     Data for individual tests are shown in Tables 8-13 and  Figures 15-21.
                                     43

-------
TABLE 8.  DATA SUMMARY, NO POLLUTANT
      MONITOR LABS INSTRUMENT
I. Wet
A.



B.
C.
D.
E.
F.
G.
H.



II. Dry

A.



B.
C.
D.
E,
F.
G.
H.



air, 51% R.H., R.T. 70°F, NO
Zero Gas
1. No filter
2. New filter
3. Used filter
Used filter, NO
New filter, NO
Used filter, NO
New filter, NO
No filter, NO
No filter, NO
Zero gas
1. No filter
2. New filter
3. Used filter
air, 5.5% R.H., R.T. 70°F, NO

Zero gas
1. No filter
2. New filter
3. Used filter
No filter, NO
New filter, NO
Used filter, NO
Used filter, NO
New filter, NO
No filter, NO
Zero gas
1. No filter
2. New filter
3. Used filter

Cone. PPM

.005
.005
.005
.135
.135
.135
.135
.138
.138

.003
.002
.002

Cone. PPM

.001
.001
.001
.152
.153
.155
.156
.156
.156

.002
.002
.003
                 44

-------
     ,16
                                               * = DRY GAS
                                               A = WET GAS
                                                          ..._•:.*.
    .15
Q.
Q-
    .13
                 NONE
NEW
USED
                        FIGURE 15.   NO POLLUTANT
                                   45

-------
     TABLE 9.  DATA SUMMARY, N0? + OZONE POLLUTANT
                   MONITOR LABS^INSTRUMENT
     Wet  air, 40.5%  R.H., R.T. 70°F, N02 +
    A.   Zero  Gas
                                         Cone. N00,  PPM
         1.   No  filter                          .005
         2.   New filter                         .004
         3.   Used filter                        .006
     B.   Used filter, N02  +  03                  .161
     C.   New Filter, N02 + 03                   .155
     D.   No  filter, N02 +  03                   .162
     E.   No  filter, N02 +  03                   .164
     F.   New filter, N02 + 03                   .156
     G.   Used filter, N02  +  03                  .165
     H.   Zero gas                          not  repeated

II.   Dry air, 5.5% R.H.,  R.T.  70°F,  N02 + 03
     A.   Zero Gas                         Conc' N02'  PPM
         1.   No  filter                          .002
         2.   New filter                         .002
         3.   Used filter                        .003
     B.   Used filter,  N02  +  03                 .161
     C.   No filter, N02 +  03                   .161
     D.   Used filter,  N02  +  03                 .162
     E.   No filter, N02 + 03                   .163
     F.   New filter,  N02  + 03                  .160
     G.   New filter,  N02  + 03                  .162
     H.   Zero gas                          not repeated
                            46

-------
       .17
                                                    <(> = DRY GAS
.16
                     *
                                             *
                                                               *
                                                               
OH
       ,14
                     NONE
                                      NEW
USED
                    FIGURE 16.  N02 + 03 POLLUTANT
                                  47

-------
                 TABLE 10.   DATA SUMMARY,  OZONE  POLLUTANT
                        MONITOR LABS INSTRUMENT*
I. Dry

A.



B.
C.
D.
E.
F.
G.
H.



II. Wet

A.



B.
C.
D.
E.
F.
G.
H.



air, 7.4% R.H., R.T. 70°F, 0,
o
Zero gas
1. No filter
2. New filter
3. Used filter
No filter, 03
New filter, 03
New filter, 03
Used filter, 03
Used filter, 03
No filter, 03
Zero gas
1. No filter
2. New filter
3. Used filter
air, 64% R.H., R.T. 70°F, 03

Zero gas
1. No filter
2. New filter
3. Used filter
New filter, 03
New filter, 03
No filter, 03
Used filter, 03
Used filter, 03
No filter, 03
Zero gas
1. No filter
2. New filter
3. Used filter

Cone. PPM

.005
.005
.005
.157
.120
.128
.144
.151
.163

.005
.006
.006

Cone. PPM

.005
.005
.005
.133
.137
.170
.157
.161
.180

.004
.005
.004
* Bendix ozone monitor in Winnebago exchanged for M.L.   Extreme instrument
  drift on Bendix.

                                    48

-------
   .17
    .16
    ,15
                                              4 = DRY  BAS
                                              A = WET  GAS
D-
Q_
 oo
O
    .13
    .12
                 NONE
NEW -
USED
                      FIGURE 17.   OZONE  POLLUTANT
                                   49

-------
TABLE 11.   DATA SUMMARY, SO- POLLUTANT
          TRACOR INSTRUMENT
I. Dry

A.



B.
C.
D.
E.
F.
G.
H.



II. Wet

A.



B.
C.
D.
E.
F.
G.
H.



air, 8.1% R.H., R.T 70°F, SO.
L.
Zero Gas
1. No filter
2. New filter
3. Used filter
Used filter, S02
No filter, S02
New filter, S02
No filter, S02
New filter, S02
Used filter, S02
Zero gas
1. No filter
2. New filter
3. Used filter
air, 71% R.H., R.T. 63°F, S09
C.
Zero Gas
1. No filter
2. New filter
3. Used filter
New filter, S02
New filter, S02
No filter, S02
No filter, S02
Used filter, S02
Used filter, S02
Zero gas
1. No filter
2. New filter
3. Used filter

Cone. PPM

.000
.000
.000
.087
.087
.081
.084
.081
.082

.000
.000
.000

Cone. PPM

.000
.000
.000
.070
.070
.072
.072
.071
.071

.000
.000
.000
                   50

-------
    .08
Q-
Q.
O
o
OL
    .0?
                                                      4> = DRY GAS

                                                      A = WET GAS
 
-------
TABLE 12.  DATA SUMMARY, SULFUR DIOXIDE PLUS OZONE POLLUTANT
             TRACOR AND MONITOR LABS INSTRUMENTS
I. Dry

A.



B.
C.
D.
E.
F.
G.
H.



II. Wet



A.



B.
C.
D.
E.
F.
G.
H.



air, 7.6% R.H., R.T. 70°F, S0« + 0
2 3
Zero gas
1. No filter
2. New filter
3. Used filter
Used filter
Used filter
New filter
New filter
No filter
No filter
Zero gas
1. No filter
2. New filter
3. Used filter
air, 64% R.H., R.T. 70°F > SO, + 0,
c. O


Zero gas
1. No filter
2. New filter
3. Used filter
No filter
New filter
No filter
New filter
Used filter
Used filter
Zero gas
1. No filter
2. New filter
3. Used filter

Cone.
sp_2
.002
.002
.002
.251
.251
.242
.242
.251
.251

.002
.002
.002

Cone.
SO
*J\J r\

.002
.002
.002
.241
.231
.241
.231
.241
.241

.002
.002
.002

PPM
P_3
.004
.004
.004
.168
.176
.159
.161
.180
.182

.006
.006
.006

PPM
n
-3

.005
.005
.005
.178
.156
.178
.157
.171
.172

.004
.003
.004
                             52

-------
       .26
  O
  w
S5 fe
M O
       .25
O K
       .23
                   NONE
                                                    * = DRY GAS

                                                    A a WET GAS
NEW
USED
                  FIGURE 19.   S02 +  03  POLLUTANT, S02
                                  53

-------
                                                        
-------
              TABLE  13.   DATA  SUMMARY, SCL POLLUTANT
                       MELOY* INSTRUMENT^
I. Dry

A.



B.
C.
D.
E.
F.
G.
H.



II. Wet

A.



B.
C.
D.
E.
F.
G.
H.



air, 7.3% R.H., R.T. 67°F, S00
C.
Zero Gas
1. No filter
2. New filter
3. Used filter
Used filter, S02
No filter, S02
New filter, S02
No filter, S02
New filter, S02
Used filter, S02
Zero gas
1. No filter
2. New filter
3. Used filter
air, 61% R.H., R.T. 67°F, SO,
L.
Zero Gas
1. No filter
2. New filter
3. Used filter
Used filter, S02
No filter, S02
New filter, S02
Ho filter, S02
New filter, S02
Used filter, S02
Zero gas
1. No filter
2. New filter
3. Used filter

Cone. PPM

-.008
-.008
-.008
.154
.151
.155
.160
.160
.163

-.009
-.018
-.016

Cone. PPM

-.004
-.007
-.006
.167
.169
.173
.173
.172
.173

-.008
-.018
-.018
*Meloy analyzer at RAMS station 102 was used to collect S09 data.
 This test was not run until all  of the other tests were completed.
                                55

-------
     .18
     .17
O-
D_
5-
o
      .16
                                          0 = DRT GAS

                                          A = WET GAS
 CM
O
oo
.15
                   NONE
                                     NEW
                     FIGURE 21.   S02  POLLUTANT  (MELOY)

-------
              5.0  CALIBRATOR OPERATING VARIABLES; AN S02 STUDY

     This work is an extension of Task Order 106 using the modified Bendix
8861 portable calibration system and National  Bureau of Standards S02 permea-
tion tubes in a study of the effect of several  variables on the accuracy of
prepared S02 calibration gas mixtures.  Variables investigated include flow
rate over the permeation tube, tube length, attitude of the permeation tube
chamber, and ambient temperature during the use of the 8861 calibration
system.

5.1  EQUIPMENT
     The following equipment was used during this phase of the Task Order:
     1.  A Bendix portable calibrator (described in Section 3.4)
     2.  A Meloy S02 analyzer
     3.  An environmental chamber
     The equipment employed during the first phase of testing was configured
as outlined in Figure 22.  Scrubbed ambient air was used during all tests.
Scrubbing consisted of passage through Drierite, activated charcoal, and
Ascarite.  Ascarite scrubbing of carbon dioxide proved necessary in order to
eliminate the variable suppression of S02 analyzer response due to changing
COp levels in the diluent air.  Tests were performed to verify that C02
removal was complete by comparing the span response of the Meloy S02
analyzer with bottled zero air as diluent to the span response with ascarite
scrubbed air as diluent.
     The equipment configuration during the second phase of testing is
displayed in Figure 23.  An environmental chamber was constructed using a
wooden framework and cardboard sheeting to enclose the Bendix portable
calibration system.  Components contained within the environmental chamber
are shown as being within a dotted line in Figure 23.
     The Meloy S02 analyzer was calibrated prior to testing following
                                      57

-------
       BENDIX 88fclP
                  SO2 TO DILUTHR
                  _V  '
               OUTPUT-^
    V
    I
            MELOY
INLET
 EXHAUST
   TO
ATMOSPHERE
                           BENDIX 8K,ID

                          -1	~
                                                         -AMBIENT AIR IN
                              nun
                               /
DRIERVTE
  ACTfVATED
  CHARCOAL.
     ASCARITE.
                                      J
                                   H-P  RECORDER
                                   O-5  V INPOT SIGNNL
               H-P DIGITAL VOLT
                     METELR
         FIGURE 22.  INITIAL EQUIPMENT CONFIGURATION FOR S02 STUDY
                                  58

-------
                                BENDIX8861D
                                 AIR SUPPLY
BENDIX
TEMPERATURE
SENSOR IN
BLOCK
                                                      THERMISTOR
                                                      QVEN TEMp
                                                      UVLII  i ti ir .
                                                      MONITOR
                                                      liumi IUt^
                              VARIABLE
                             .FLOW  RATE
                              0-100 cc/MIN
 c
                                             PERM.
y
  MOLECULAR
    SIEVE
   DRIERS
                  •Q TOGGLE
L_
                                                EXHAUST TO
                                                ATMOSPHERE
  Equipment inside dotted line enclosed in chamber
  for control  tests at three different temperatures
  *Provides air flow to purge permeation device
   during transit
                                                                    H.P.
                                                                 RECORDER
           FIGURE 23.  FINAL EQUIPMENT CONFIGURATION FOR S02 STUDY
                                      59

-------
standard procedures yielding the calibration data presented in  Figure 24.

5.2  PERMEATION TUBE VARIABLES
     An SO^ concentration of approximately 0.3 ppm was selected for testing.
At this concentration a 2 cm, a 5 cm, and a 10 cm S02 permeation tube could
be exchanged in turn within the permeation tube chamber without exceeding the
range of dilution airflow measurement within the Bendix 8861.   The appropri-
ate dilution airflow rates were selected for the particular tube in use so
that the S02 concentration prepared should remain constant.  Airflow over the
tube in use was varied using a mass flow controller with a range of 0-100
cc/minute.  The effect of permeation tube chamber position was  determined by
physically rotating the chamber through 90° between experiments.  Chamber
temperature was monitored during the course of the study in two ways.  The
Bendix platinum resistance thermometer embedded in the aluminum block sur-
rounding the Teflon permeation tube cylinder was monitored.  In addition, a
small glass thermistor was placed within the tube cylinder and  the actual
air temperature over the permeation tube was monitored by placing the therm-
istor in a voltage divider circuit and recording a voltage inversely pro-
portional to temperature.  Thus, a change in air temperature could be
measured directly and without the time lag associated with the  Bendix
temperature sensor.
     The Meloy analyzer output voltage was monitored using a Hewlett-
Packard DVM and a Hewlett-Packard strip chart recorder.  Once a stable trace
was obtained, readings were taken at 15 minute intervals and voltages were
converted to concentrations using the slope and intercept developed during
calibration of the Meloy analyzer.
     Throughout this phase of testing, the  sequence of operations was as
follows:
     1.   Permeation  tube chamber placed  in  a  horizontal position, purge
          flow  rate stabilized  (5 cc/min.,  10  cc/min., etc.), chamber
          temperature at  30.0°C.
     2.   Recorder  trace  stable,  analyzer  signal  level and  chamber temper-
          ature recorded.
                                      60

-------
     .3 -
     .2 -
 Q.
 O-
     .1 -
                    1           2
                        VOLTS
Cone.  SOpj, ppm
     0.00
     0.10
     0.20
     0.30
Meloy Output, Volts
       0.064
       0.664
       1.280
       1.905
     FIGURE 24.  MELOY S02 ANALYZER CALIBRATION
                         61

-------
     3.   Chamber placed in a vertical  position.   Step 2  repeated.
     4.   Chamber returned to a horizontal  position.
     The results obtained are presented in Tables 14 and 15.   A cypical  strip
chart record is included as Figure 25.

5.3  TEMPERATURE EXCURSIONS
     The second phase of testing involved the placement  of the Bendix 8861
portable calibrator in an environmental chamber.   Cooling of the system com-
ponents was achieved by exposing the chamber to a stream of outside air at
10°C.  Heating was achieved by use of a strip heater. A blower and a damper
were employed for regulation of chamber temperature.  This test was performed
with a 2 cm SO- permeation tube and at a purge flow rate over the tube of
5 cc/min.  The results obtained are presented in Tables  16 and 17.

5.4  RESULTS AND DISCUSSION
     A small effect was observed as the purge flow rate  was varied over the
2 cm and the 5 cm permeation tubes.  The data indicates  an increase in
analyzer response of 1-2% as the purge flow rate was increased from 5
cc/minute to 50 cc/minute.  Analyzer response remained constant at purge
flow rates above 50 cc/minute regardless of tube length.
     The effect of varying the permeation chamber position was to create an
abrupt transitory disturbance within the chamber with a resulting change in
S02 concentration observed at the analyzer.  The magnitude of the transitory
change in concentration ranged from 10-25% and appeared to depend on the
position of the tube within the chamber prior to the tilt.  The disturbance
in the SCL concentration  persisted for about 15-20 minutes with a 5 cc/
minute purge flow rate  and  for about 1 minute with  a 100 cc/minute purge
flow rate.  The equilibrium S0? concentration remained constant regardless
of the chamber  position as  the data in Table 14  indicates.  Thus, the
reservoiring of SO- within  the chamber would not appear to be a significant
source of error even  at a purge flow rate of 5 cc/minute.
     The data  obtained  in regard  to tube  length  are  anomalous.  Analyzer re-
sponse  increased  5%  after the exhange  of  a 5 cm  tube for a 2  cm tube.  A 4%

                                     62

-------
TABLE 14.  EFFECT OF PERMEATION CHAMBER FLOW RATE ON SIGNAL RESPONSE
A.
FLOW
CC/MIN
5.0
10.0
25.0
50.0
100.0
B.
5.0
10.0
25.0
50.0
100.0
recheck
5.0
2.0 cm permeation
OVEN
POSITION
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
5.0 cm permeation
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal

horizontal
tube
OBSERVED,
PPM
.306
.310
.312
.314
.314
.314
.317
.317
.317
.320
.320
.319
.320
.320
.320
tube
.330
.330
.328
.328
.330
.333
.339
.338
.336
.335
.335
.335
.327
.335
.335

.328

OVEN
THERMISTOR(V)
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549

.551
.551
.551
.550
.550
.550
.550
.550
.550
.550
.550
.550
.550
.550
.550



TEMPERATURE
(30.00°C)
(30.00°C
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)

(29.92°C)
(29.92°C)
(29.92°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)



BENDIX °C
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00

30.01
30.01
30.01
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00


                                                              (continued)
                                 63

-------
TABLE 14 (continued)
C. 10.0 cm permeation tube
FLOW
CC/MIN
5.0
10.0
25.0
50.0
100.0
OVEN
POSITION
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
horizontal
vertical
horizontal
OBSERVED,
PPM
.343
.345
.346
.345
.345
.343
.348
.346
.346
.346
.346
.344
.345
.346
.346
OVEN
THERMISTOR(V)
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
.549
TEMPERATURE
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
BENDIX °C
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
         64

-------
               TABLE 15.  EFFECT OF S0£ PERMEATION TUBE LENGTH
2.0 cm Tube
FLOW OVER
PERMEATION
TUBE
CC/MIN
5.0
10.0
25.0
50.0
100.0
(NBS 29-99)

S0?
REPORTED,
PPM
.309
.314
.317
.320
.320








PERMEATION OVEN TEMPERATURE
THERMISTOR, V
.549
.549
.549
.549 (
.549 (

30.00°C)
30.00°C)
30.00°C)
30.00°C)
30.00°C)
BENDIX
30.
30.
30.
30.
30.
°C
00
00
00
00
00
AVERAGE
.316
5.0 cm Tube (NBS 27-12)
     5.0
    10.0
    25.0
    50.0
   100.0

AVERAGE
.329
.330
.338
.335
.332

.333
.551
.550
.550
.550
.550
(29.92°C)
(29.96°C)
(29.96°C)
(29.96°C)
(29.96°C)
30.00
30.00
30.00
30.00
30.00
10.0 cm Tube (NBS 27-13)
     5.0
    10.0
    25.0
    50.0
   100.0

AVERAGE
.345
.344
.347
.345
.346

.345
.549
.549
.549
.549
,549
(30.00°C)
(30.00°C)
(30.00°C)
 30.00°C)
 30.00°C)
30.00
30.00
30.00
30.00
30.00
Repeat 2.0 cm Tube (NBS 35-16)

     5.0          .319
              ,551
              (29.92°C)
                   30.00
                                    65

-------
o


oo
LLJ
IE
O
•z.
I— I

CO
UJ
UJ
D-
oo
                                    .MOVE CHAMBER FROM
                                     VERTICAL TO HORIZONTAL
                                       MOVE  CHAMBER FROM
                                       HORIZONTAL TO VERTICAL
                    5.0 cm SO^ Permeation Tube

                        5 cc/minute purge flow rate


            FIGURE  25.   TYPICAL STRIP CHART TRACE, S02 STUDY

                                   66
                                                                  FULL SCALE

-------
TABLE 16.   EFFECT OF COOLING ON MODIFIED BENDIX 886IP  CALIBRATOR
TIME
1300
1315
1330
1345
1400
1415
1430
1445
1500
Mean
SD
1515
1530
1545
1600
1615
1630
1645
1700
Mean
SD
SO,
OBSERVED,
PPM
.317
.319
.318
.321
.318
.322
.317
.319
.320
.319
.002
.329
.335
.336
.334
.334
.331
.330
.330
.332
.003
5 CC/MIN. PURGE FLOW
ENVIRONMENTAL SULFUR OVEN
CHAMBER T°C THERMISTOR(V)
22.1
23.8
23.2
23.3
23.3
23.8
23.3
23.0
22.8


10.0
9.0
8.0
8.0
8.0
8.0
8.0
8.0


-
-
.551
.551
.551
.550
.550
.551


.552
.557
.561
.564
.566
.567
.568
.568


-
-
(29.92°C)
(29.92°C)
(29.92°C)
(29.96°C)
(29.96°C)
(29.92°C)


(29.87°C)
(29.63°C)
(29.45°C)
(29.28°C)
(29.18°C)
(29.14°C)
(29.08°C)
(29.08°C)


BENDIX °C
-
-
30.00
30.00
30.00
30.00
30.00
30.00


30.00
30.00
30.05
30.05
30.05
30.05
30.05
30.05


                              67

-------
TABLE 17.  EFFECT OF HEATING ON MODIFIED BENDIX 8861  CALIBRATOR
                      5 CC/MIN. PURGE FLOW
                            TRIAL 1

TIME
1000
1030
1300
1330
Mean
SD
1400
1415
1430
1445
1500
1515
1530
1545
1600
Mean
SD

0845
0900
0915
0930
0945
1000
1015
1030
Mean
SD
1045
1100
1115
1130
1145
1200
1215
1230
Mean
SD
S0?
OBSERVED,
PPM
.323
.321
.322
.321
.322
.001
.328
.325
.328
.327
.327
.318
.328
.327
.327
.326
.003

.340
.341
.341
.343
.342
.341
.337
.338
.340
.002
.336
.337
.337
.339
.338
.338
.337
.334
.337
.002
ENVIRONMENTAL
SULFUR OVEN
CHAMBER T°C THERMISTOR
25.0
24.1
24.6
24.4


38.9
36.2
41.8
39.0
37.8
37.8
37.7
38.2
38.1



23.7
23.9
24.1
24.1
23.9
24.1
24.0
23.9


32.0
33.0
33.4
33.4
33.6
32.5
32.3
32.1


.548
.548
.548
.548


.545
.541
.540
.538
.538
.538
.538
.538
.538


TRIAL 2
.550
.549
.549
.549
.549
.549
.549
.549


.548
.547
.545
.544
.543
.543
.543
.543


(30.05°C)
(30.05°C)
(30.05°C)
(30.05°C)


(30.20°C)
(30.40°C)
(30.45°C)
(30.55°C)
(30.55°C)
(30.55°C)
(30.55°C)
(30.55°C)
(30.55°C)



(29.96°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)
(30.00°C)


(30.05°C)
(30.10°C)
(30.21°C)
(30.25°C)
(30.30°C)
(30.30°C)
(30.30°C)
(30.30°C)



BENDIX °C
30.00
30.00
30.00
30.00


29.95
29.95
29.95
29.95
29.95
29.95
29.95
29.95
29.95



30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00


30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00


                               68

-------
increase in analyzer response was observed following the exchange of a 10 cm
tube for the 5 cm tube.   A recheck of the 2 cm tube agreed within 1%.   The
time allowed for equilibration of the permeation tubes after exchange ranged
from 16 hours to two days.
     Although the effect of permeation tube length on analyzer response
appears significant, conclusions do not seem warranted.   Of necessity, due to
equipment limitations, the study period associated with the exchange and
equilibration of permeation tubes was two weeks in duration.  The precision
of the measuring system over the two week period of time may be no better
than the magnitude of the apparent effect of permeation tube length on
analyzer response.
     The data in Table 16 indicate an adverse effect of cooling on the
Bendix calibrator.  The permeation tube chamber purge air temperature dropped
from the normal  control  temperature of 30.0°C to a low of 29.1°C as the
environmental chamber temperature decreased 15°C.  The SO,, concentration
observed at the analyzer, however, increased.  The reason for the increase
was not determined.  The dilution air flow controlling orifices in the cal-
ibrator, normally thermostated, were observed to be operating below the
normal control temperature during the cooling test as indicated by contin-
uous heater operation.  The dilution air mass flow meter, however, did not
indicate a change in dilution airflow.
     The only effect of heating on the Bendix calibrator that was observed
was a rise in the purge air temperature from the normal  control temperature
of 30.0°C to 30.6°C indicating an inefficiency in the thermoelectric cooling
assembly control.  No significant change in S02 concentration was observed
at the analyzer during the course of the test.  Trial 1  and trial 2 were
conducted under seemingly identical conditions several days apart.  The
only variable known to have changed was the air scrubber (Ascarite was
changed).  The difference in analyzer response at room temperature of 5%
between the two trials would appear to be an estimation of the precision of
the measuring system employed.
     During the course of this study, the most significant problem observed
in regard to the routine operation of the Bendix calibrator was the difficulty

                                     69

-------
in obtaining a complete seal  of the permeation tube chamber after the
replacement of a permeation tube.   Even a relatively minor leak was observed
to be of consequence with a purge  airflow rate of only 5 cc/minute.  The
problem was aggravated at high dilution airflow rates where significant
back pressure forced additional leakage.  A new permeation tube holder
designed to eliminate this problem and to provide adequate purge air temper-
ature control was designed, fabricated, and installed.  A description of the
holder is included in Appendix B.
     The study described in this section was performed during the final
phase of the Task Order.  The modified calibrator resulted from experience
gained during the course of the study.  It is apparent from the data
obtained that the Bendix calibration system, as originally supplied by the
vendor, may have had limitations in routine operation, particularly in
regard to audit work performed at low ambient temperatures.
                                     70

-------
                 APPENDIX A



AUDIT PROCEDURES FOR RAPS INSTRUMENT SYSTEMS
                     71

-------
                                 CONTENTS
Figures                                                                 73
Tables                                                                  73
A.I  Introduction                                                       74
     A.1.1  Preparation of Bendix Calibration/Audit Equipment           74
A.2  Calibration of Winnebago Instrumentation                           81
     A.2.1  Calibration of the Tracor Model 270HA Atmospheric
              Sulfur Analyzer                                           81
     A.2.2  Calibration of the Monitor Labs Model 8440 Oxides
              on Nitrogen Analyzer                                      84
            A.2.2.1  Ozone Generator                                    87
     A.2.3  Bendix Ozone Analyzer                                       89
     A.2.4  Calibration of MRI 1561 Nephelometer                        90
A.3  Audit Procedures                                                   92
     A.3.1  Sulfur Analyzer Audits                                      92
     A.3.2  Oxides of Nitrogen Analyzer Audit                           92
     A.3.3  Ozone Analyzer Audit                                        94
     A.3.4  MRI 1561 Nephelometer Audit                                 96
     A.3.5  Audit Procedure for Atmospheric Methane, Carbon
              Monoxide and Total Hydrocarbons                           96
A.4  Air Management                                                     98
                                     72

-------
                                 FIGURES
Number                                                                 Page
 A.l       Bendix 8861D Flow Diagram                                    75
 A.2       Bendix 8861P Flow Diagram                                    76
 A.3       Typical Calibration Curve for Bendix 8861D                   78
 A.4       Typical Interconnections for Introduction of Calibration     82
             Gas into Instrument to be Calibrated or Audited
 A.5       Conversion Efficiency Determination                          94

                                  TABLES
Number                                                                 Page
 A.I       Sulfur Dioxide Calibration:  Typical Bendix 8861
             Settings                                                   83
 A.2       Settings for NOX Analyzer                                    85
 A.3       Oxides of Nitrogen Calibration:  Typical Bendix
             8861D Settings                                             86
 A.4       Sample Tabulation for a NO - 0., Titration                    88
 A.5       Values for Ozone Calibration                                 89
 A.6       Sample Table to be Used in the NOX - NO Audit                93
 A.7       Settings to use on Bendix Calibrator to Produce
             Various Ozone Concentrations                               95
 A.8       Values for Ozone Audit at High Lamp Output                   95
                                    73

-------
                              A.I  INTRODUCTION

     The following procedures were established and tested during the initial
stages of work on Task Order 106.  The procedures are presented in logical
sequence and related to both routine audit procedures, as well as the routine
calibration work performed, utilizing the pollutant monitoring equipment
aboard the Winnebago van.

A.1.1  PREPARATION OF BENDIX CALIBRATION/AUDIT EQUIPMENT
     1.  Calibration curves used to read air and gas flows from the Bendix
instruments must be developed as follows (reference Figures A.I and A.2):
         a)  Break the connection at the brass "T" upstream of pressure regu-
             lator PR-2 on the 8861D and cap off the side of the "T" leading
             to the instrument filter.
         b)  Attach a zero air cylinder to the 8861D at the "clean air" port
             using a length of 1/4" tubing.  Set the cylinder regulator
             delivery pressure to 40 PSI.
         c)  Starting with pressure regulators PR-1 and PR-2 closed and
             orifice No. 2, 3, 4, 5 closed, connect the 8861D "out" port  to
             a 1000 cc bubble tube.
         d)  Open PR-2 to 10% of gauge as monitored by pressure gauge P2  and
             measure the air flow rate at the "out" port using the bubble
             tube.  Repeat this procedure at 20%, 30%, etc., to 100% of
             gauge.
         e)  Repeat d) above with orifice No. 1 and No. 2 open.
         f)  Repeat d) above with orifice No. 1 and No. 3 open.
         g)  Repeat for  the remaining orifices and all combinations of
             orifices, i.e., finish with 1+2+3+4+5 for maximum flow through

                                     74

-------
                   5
                   5
               o   5
                      2
                      O
                                               o
                                               00
                                               00
                                               X
                                               I— I
                                               Q
                                               UJ
                                               CO
                                               a:
                                               rs
                                               CD
75

-------
                                  CD

                                  i—i
                                  Q


                                  O
                                  VD
                                  CO
                                  CO
                                  
-------
             the 8861 D.
         h)  Repeat  the above procedure and calibrate  the  nitric  oxide  capil-
             lary using nitrogen entering  the  "NO  Inlet".   Flow from  the
             capillary to a  10 cc bubble tube  can  be taken at  the NO  "pull-
             to-test" valve.
         i)  Repeat  the above procedure and calibrate  the  capillary upstream
             of 8861P permeation oven.  Close  VI toggle  valve  (off position),
             introduce zero  air into  the "N,,"  port and with  V2 closed and  V3
             open, attach a  10 cc bubble tube  to the 8861 P outlet port  for
             the flow calibration.
         j)  After converting all of  the flows  to  cc/minute  for each  of the
             10% gauge increments, the flows must  be corrected to standard
             temperature of  0°C and standard pressure  of 760 mm Hg by multi-
             plying  each reading by a factor.   The correction  factor  is
             derived as follows:

             23.5°C  = temperature in  lab at time of calibration
             746.3 mm mercury, barometric  pressure in  lab  at time of  calibra-
             tion.

         Correction  _ _ 273 _   Bar. Press,  in  mm Hg
         Factor       273 +  Ambient Temperature            760

                     273           746.3
                 273 + 23.5        760

     Graphs are now prepared by plotting "Gauge Reading - % of scale"  as  the
ordinate and "cc/minute" along the abscissa.  This will facilitate  reading
any flow within the range of the capillary or orifice under study  (see
Figure A. 3).

     2.  With Bendix 8861 P set on "stand-by", the internal air pump  is opera-
tive and keeps a flow of air through the permeation oven.  If this  pump is
shut off, twelve to twenty-four hours are required to re-establish  equilib-
rium condition within the permeation oven.  Assuming an equilibrium  condition
with the permeation oven temperature at 30°C +_ 0.02°C (running for  twelve to
                                     77

-------
    90 -



    80 -



    70 -



    60 -
 '   50 -
CJ3
a:

UJ
C3
40 -



30 -



20 -



10 -
      4000  4500   5000  5500   6000  6500  7000  7500  8000

                              cc/tnin.


                   Bendix 8861D Orifice 2+3+4+5



      FIGURE A.3   TYPICAL CALIBRATION CURVE FOR BENDIX 8861D
                                78

-------
twenty-four hours on 115 VAC with power switch in ON position, VI toggle
valve in OFF position and PR-1 pressure regulator set to 100% of scale on
the 0-15 psig) prepare the 8861D for operation.
     3.  Turn the power switch on the Bendix 8861D to the ON position, turn
OFF V3, 4, 5, 6 (corresponding to orifices 2, 3, 4 and 5).  Connect a 1/4"
Teflon line between the 8861D "clean air" port and the N2 inlet on the 8861 P.
The 8861D is now providing zero air to the 8861P where it will flow through
the 5 cc/min capillary and purge the permeation oven.  At this point, the
permeation tube oven effluent will exit either through the "vent" or through
the "outlet" depending on the settings on V2 and V3 (V3 open, V2 closed,
gas goes to "outlet".  V3 closed, V2 open, gas goes through scrubber to
"vent").
     The Bendix 8861D can produce air of quality adequate for conducting all
of the standard audits except for the Beckman 6800.  Scrubbed air of good
quality is possible if scrubbers are installed as follows:  remove the "U"
tube connector between the air surge tank and the particulate filter.  At
the outlet of the surge tank install a short line leading to a "Drierite"
scrubber filled with indicating Drierite.  In series with the Driertie
scrubber install a second Drierite column charged with cocoanut shell
charcoal, 6-14 mesh.  The outlet from the charcoal scrubber is attached to
the Bendix particulate filter.
     4.  Before calibrations or audits actually begin it is necessary to
point out certain pitfalls which must be avoided.
         a)  Standards.  The NO gas cylinders used in this work must be
             certified as to exact composition, i.e., NBS or directly trace-
             able to an NBS cylinder.  For the CO, CH4, THC audits and
             calibrations, cylinders of known composition will be used and
             again, these must be certified as traceable to NBS or certified
             as to exact composition by RTP.  Sulfur dioxide permeation tubes
             purchased directly from NBS will be used in the Bendix 8861P
             and the permeation oven will be operated at 30°C +_ .02°C to
             insure an accurate output of S02.
                                     79

-------
    Since the FID response of most hydrocarbon analyzers is sensi-
    tive to oxygen concentration, it is necessary for zero air and
    matrix air (in the standard cylinders) to have the same oxygen
    content as ambient air.   Similarly, for TS analyses employing
    flame photometric detectors, the matrix air must have the same
    COp content as ambient air.
b)  Flow rates.  Every instrument audited has a specific sample flow
    rate and it is necessary to insure that the flow of calibration
    gas (from the Bendix 8861D) exceeds the analyzer demand.  Good
    practice dictates a procedure whereby a rotameter is used to
    measure the sample flow of any instrument when the demand is
    unknown.  The flow as measured by the rotameter must then be
    exceeded when preparing calibration gas to avoid taking in
    ambient air along with the calibration/audit gas entering the
    sample port of the instrument under test.
c)  Materials used.  Since SO^ and (L both are quite reactive it is
    mandatory to always convey these gases in glass and/or Teflon.
    Span gases containing nitric oxide, carbon monoxide, methane,
    or other hydrocarbons are less critical but nevertheless con-
    veyed in Teflon tubing.
d)  Distinction between "audit" and "calibration".  When the opera-
    tor is instructed to calibrate an instrument, known gas compo-
    sitions are introduced and the instrument is adjusted to respond
    in accordance with the known gas composition.  This is accom-
    plished in accordance with the instructions given in the manual
    for that instrument and involves changing instrument dial
    settings, potentiometer adjustments and any other adjustment
    necessary to cause the instrument to read the same value as
    the gas being  introduced.  An audit on the other hand involves
    absolutely no  instrument adjustment.  A known composition gas
    is introduced  and the instrument read-out is taken and recorded.
                             80

-------
                A.2  CALIBRATION OF WINNEBAGO INSTRUMENTATION

     The Winnebago van is equipped with a full complement of air quality
analyzers for use in preparation of the Bendix 8861 calibration system for
audit activities and for use in running "side-by-side" field comparisons
in proximity to existing air quality monitoring sites.  The following pro-
cedures relate to the calibration of this instrumentation for the above
mentioned purposes.

A.2.1  CALIBRATION OF THE TRACOR MODEL 270HA ATMOSPHERIC SULFUR ANALYZER
     Figure A.4 depicts the interconnection described in the following para-
graphs and should be referenced for clarity.
     1.  Connect a 1/4" Teflon line between 8861D "out" port and the normally
open port of the input solenoid valve of the Tracer.  Provide a "T" connec-
tion or a glass manifold in this line to vent excess gas delivered by the
8861D.
     2.  Provide zero air input to Tracer as follows:
         a)  A 1/4" Teflon line should be connected between the 8861D "out"
             and the Tracer input (through a "T" or manifold as previously
             described).  With the 8861D operating, adjust pressure gauge
             P-2 to 50% with valve (orifice) 5 in (M position.  Zero air at
             the rate of 3000 cc/min is now leaving the 8861D.
         b)  Set a DVM on 0-5 volt range.
         c)  Connect the DVM as follows:  one set of leads to "TS low" and
             "gnd" and another set of leads to "S02 low" and "gnd".  These
             connections are on the rear of the Tracer on the terminal
             strips marked "TB2".   The two sets of leads are alternately
             plugged into the DVM to read sulfur dioxide or total sulfur.
                                     81

-------
                                                   or
                                                   CQ
                                                   
                                                   00 CO
                                                   Z
                                                   o o
                                                   H-I H-

                                                   O (—
                                                    O ^>
                                                    O GC.
                                                    a: i—
                                                    LU oo
                                                    -•
                                                    HH
                                                    Du OO
                                                    >- •=£
                                                    H- CD
                                                    •sf
                                                      •

                                                    
-------
     3.  Allow  the analyzer  to  sample zero  air  for  fifteen minutes  and  then
take a voltage  reading.   Record sulfur  dioxide  and  total  sulfur  voltages  on
each channel.
     4.  Connect  a short  length of  1/8" OD  Teflon tubing  between the  8861P
"outlet" and  the  8861D HpS/SO,,  inlet.   Sulfur dioxide  gas is  now entering
the 8861D  so  that it may  be  diluted  to  the  desired  concentration.
     5.  Run  a  calibration recording voltages as explained above and  using
the following typical table  (Table  A.I) to  obtain the  desired dilutions.

    TABLE  A.I   SULFUR DIOXIDE CALIBRATION;   TYPICAL BENDIX 8861  SETTINGS
Dilution F * +
Concentration
Desired (PPB)**
0
50
60
70
90
100
150
190
300
500
700
900
-
7220 H
6015 H
5155 H
4008 -
3607 J
2402 H
1895 H
1197 H
715 H
509 H
395 H
FP

i- 7
H 7
i- 7
h 7
i- 7
h 7
H 7
h 7
H 7
H 7
i- 7
Orifice
Used

1,
1,
1,
1,
1,
1,
1,
1,
1
1
1

2, 3, 4, 5
2, 3, 4, 5
2, 3, 4, 5
4, 5
4, 5
4
3
2



Percent Station
of Voltage
Gauge S02 TS
-
69
43
26
25
16
30.5
72
14.5
91
35
12
        * Where        F  = cc/minute of S02 air delivered by 8861P
                            (through permeation tube oven at 30.00°C) and
                       Fd = cc/minute of air delivered by the 8861D diluter.
        ** Calculated using NBS permeation tube 21-71  with a permeation rate
           of .946 micrograms/minute at 30°C.
    The calibration is run by selecting four or more points from the table
and recording the voltages at each of the selected dilution settings.  A
graph is then plotted to determine response linearity.
                                     83

-------
     6.   In the event additional points are required, they are calculated
as follows:
                               r    = PR x K
                                ppm - Fp + Fd

         Where C    = desired concentration in parts per million
         PR         = permeation rate in ng/min (Ref. to oven temperature.)
         K (Ref. to 0°C and 760 mm Hg) - Mo1ecu1ar ^ perm. gas

         (22Al of an ideal gas = 1.00 mole at 0°C, 760 mm Hg)
         F          = cc/minute from the 8861P permeation device
         F,         = cc/minute from the 8861D dilution device

EXAMPLE:  The 2.0 cm S02 permeation tube is NBS certified to deliver 946
          ng/min at 30°C.  To calculate the dilution air flow rate corre-
          sponding to a desired gas concentration:

          Flow Rate           <946> < TT >
          (cc/min)   "  Desired Concentration
                                 ppm

     Desired flow rates are read off the calibration curves developed for
the orifices in these instruments.  These curves are kept with the service
manuals.

A.2.2  CALIBRATION OF THE MONITOR LABS MODEL 8440 OXIDES OF NITROGEN ANALYZER
     1.  Inspect rear panel of the analyzer unit and the sample conditioning
unit and check the following items:
         a)  Drier connected to brass air inlet
         b)  Teflon NO  line connected between modules
         c)  Teflon NOX line connected between modules
         d)  Teflon ozone line connected between modules
         e)  Teflon sample inlet  line connected to  glass sample manifold
              (see Figure A.4).
                                      84

-------
         f)  Glass sample manifold should be connected to the 8861D output
     2.  Connect the nitric oxide source gas (typically 100 ppm NO in
nitrogen) to 8861D NO inlet.
     3.  Turn on the main power and the ozone generator power on NO,,
analyzer.
     4.  Adjust vacuum gauge to 20 inches.  This results in rotometer flows
as follows:

             NOX - between 150-200 cc/min
             NO  - between 150-200 cc/min
             Ozone - between 100-150 cc/min

     5.  Connect a DVM to the DVM terminal blocks on the rear of the instru-
ment.  Use the NO terminals.
     6.  Connect a 0-1 volt recorder to the terminal posts on the front of
the instrument and adjust instrument settings as shown in Table A.2.
                    TABLE A.2  SETTINGS FOR NOX ANALYZER
Channel Sec.
NOY 20
A
NO 20

Range
2
2

Vac. Flow
20 175
225
Ozone = 100
     7.  Start the Bendix diluter and adjust for 3000 cc/min zero air.
Allow the analyzer to sample the zero air and then adjust both NO and NO,,
channels using the respective zero potentiometers to give zero volts on the
DVM (or a slight positive off-set if this is desirable).
     8.  Introduce each span gas concentration in turn selecting points and
conditions as shown in Table A.3.
                                     85

-------
TABLE A.3  OXIDES OF NITROGEN CALIBRATION:  TYPICAL BENDIX 8861D SETTINGS
Dilution
Concentration
Desired (ppm)
0.45
0.35
0.25
0.20
0.15
0.05
FP * Fd *
467
601
841
1052
1402
4208
Orifice
Used
1
1
2
2
3
3, 4, 5
Percent of
Gauge PI
38.5
38.5
38.5
38.5
38.5
38.5
FP
cc/min
2.0
2.0
2.0
2.0
2.0
2.0
?2 Gauge
Setting
34
72
24.5
50.5
37.5
22.5
 * Where F  = cc/min of NO from the cylinder
         F. = cc/min of air delivered through the 8861 D orifices
   Calculation of other points can be accomplished using the following
   equations:

   (NO cone, in cylinder, ppm) x F
   - = NO concentration, ppm
   For example, if a final concentration of .05 ppm NO,, is desired, and
   the concentration of NO in the cylinder is known to be 105.2 ppm, we
   can fix F  at 2 cc/min by setting P^ to 38.5% of gauge.  F^ can be
   determined from the above equation or:
   Since F  is set at 2 cc/min, Fd = 4208 cc/min.  This can be achieved
   by setting P« to 22.5% and opening orifice 3, 4 and 5.
                                   86

-------
     9.  Determination of efficiency of NCL to NO converter.  Proceed as
follows:
         a)  Starting with both channels properly zeroed, introduce diluted
             NO to the calibrated analyzer to read 1.5 - 2.0 ppm NO.
         b)  When equilibrium is achieved, observe that equal response
             obtained on the NO and NOX channels.
         c)  Reverse the channels to check for NO purity.
         d)  Turn on the 03 generator within the Bendix 8861D.  Wait 10
             minutes.
         e)  Adjust the Bendix 03 generator so NO reading is reduced to 50%
             of original reading.  Record stable NO and NOY values.  The drop
                                                          A
             in NO concentration is equal to the concentration of NOp formed
             in accordance with the reaction equation

                             NO + 03 = N02 + 02

         f)  Calculate efficiency
MX
                E =  1 -
                              (Orig.) - NOX (Final)
                          NO  (Orig.) - NO  (Final)

A.2.2.1  Ozone Generator
                            x 100
     Immediately following the calibration of the oxides of nitrogen analyzer
and prior to the calibration of the ozone analyzer, the ozone generator with-
in the Bendix 8861D should be calibrated.  To perform a gas phase titration
of nitric oxide within the Bendix 8861D, it is first necessary to establish
an equilibrium concentration of nitric oxide.  This may be done using the
techniques described in section A.2.2 in regard to the calibration of the
oxides of nitrogen analyzer.  The nitric oxide channel of the analyzer then
serves to monitor the results of the titration occurring within the Bendix
8861D.  All that is required to perform the titration is to turn on the
ozone generator in the calibrator.  Air flowing through orifice 1 will pass
over the ultraviolet lamp, undergo irradiation, and meet with nitric oxide
in a glass reaction chamber.  The reacted mixture leaves the reaction chamber

                                     87

-------
and is diluted in the mixing chamber.  The residual nitric oxide concentra-
tion after ozone titration is then observed using the nitric oxide channel
of the analyzer.  A step by step example is as follows:
     Proceed initially as explained in section A.2.2 to prepare the Bendix
8861D and a Monitor Labs 8440 oxides of nitrogen analyzer.  Establish zero
air flow either from the Bendix calibrator or from a cylinder to the sampling
manifold and subsequently to the sample inlet port of the 8440.  Zero the
analyzer using the zero potentiometer.  Establish an initial nitric oxide
concentration of 0.8 ppm.  A typical set of conditions using a cylinder
certified at 105.2 ppm NO is presented in Table A.4.

            TABLE A.4  SAMPLE TABULATION FOR A NO - 03 TITRATION
Desired p
Cone. , ppm 1
0.80 80%

Fp P2 Fd Orifice
4.0 49% 522 1
F + Fd = 526 cc/minute
Output,
NO
.800

Volts
NOV
A
.800

     Once a stable trace is established on the recorder, the ozone generator
is set to Range 500, Set Point 440 and the lamp is turned on.  After a short
time the NO channel will show a sharp decrease as ozone reacts with the NO.
After ozone addition the nitric oxide concentration  is measured again by
monitoring the NO channel of the analyzer.   In this  particular example, the
NO channel decreased to 0.400 volts.  The ozone output is now calculated
using the above information.

          0.800 (original voltage before 0^  addition)
          0.400 (final voltage after ozone lamp was  started)
          0.400 difference
     On  the analyzer used the slope was 1.00 ppm/volt
          0.400 x  1.00 = 0.400 ppm decrease  in NO concentration

      Since  NO reacts with 03  1:1  stoichiometrically, this means  0.400  ppm
ozone  reacted in  the diluted  gas  stream.   To convert this  figure to the
                                     88

-------
undiluted concentration delivered from the ozone lamp source we use
          microliters of ozone/min = ppm ozone x dilution air, liters/min
          or .400 x .526 = 0.2104
     The lamp is now calibrated at an air flow of 0.526 liters/minute and is
producing 0.2104 microliters of ozone per minute at standard temperature and
pressure.

A.2.3  BENDIX OZONE ANALYZER
     Once the ozone generator in the Bendix 8861D has been calibrated, an
ozone analyzer calibration may be performed.  In this case the pressure
setting on the capillary (No. 1) is not changed to insure a constant ozone
production as dilution air is varied.  Dilution air is added in this case
only by using additional capillaries.  To get the second point for the
calibration we now open capillary No. 2 and readjust P2 to 49%.  The table
below indicates the capillaries used and the resulting concentration.
After each additional  capillary is opened, P2 must be readjusted to 49% to
maintain a known standard air flow over the ozone lamp.  Typical flow
settings are shown in Table A.5.
                   TABLE A.5  VALUES FOR OZONE CALIBRATION
Ozone
Desired
Cone, ppm
0.40
0.20
0.10
0.083
0.040
Orifice
1
1 + 2
1 + 3
1 + 4
1 + 2 + 3 + 4
Dilution Air
LPM
0.526
1.055
2.104
2.550
5.320
P2
49%
49%
49%
49%
49%
sample calculation:
                                         lrr = 0<400 ppm ozone
                                     89

-------
     After the flow conditions have been established, perform the following
steps to complete the five point calibration of the ozone analyzer.
     1.  Attach the Teflon output line from the Bendix 8861D to the sample
input of the Bendix ozone monitor via a "T" or a glass manifold.
     2.  With power switch to "on" position (allow 120 minutes for warm-up),
adjust ethylene flow to 25 cc/minute.
     3.  Set mode selector to ambient, time constant to 10 seconds, and
range selector to 0.5 ppm.
     4.  The terminal block on the rear of the instrument may be used to
display a DVM response to the calibration gas.
     5.  Select each point in turn as illustrated in Table A.5 to complete a
five point calibration on the ozone instrument.

A.2.4  CALIBRATION OF MRI 1561 NEPHELOMETER
     All pots and test points referred to in this procedure are located on
the electronics board on the inside of the instrument front door.  In all
cases the voltages should be read with a digital volt meter connected to  the
red and black terminals located on the righthand side of the instrument.
The meter on the face of the instrument should be used only as a rough in-
dication of instrument behavior.  In RAMS stations the data acquisition
system digital display may be used with the analog signal appearing on
channel 37.
     The following steps should be followed in sequence.
     1.  Unhook the  inlet hose and plug the intake.
     2.  Move the "air - operate - Freon" switch to  the "air" position.
     3.  Plug the lower outlet and wait at least 15  minutes or until meter
reading and/or voltage output are steady.
     4.  Readout should be 0.00 VDC+  .03; if not, adjust R-42  (background)
for proper  reading.
     5.  Connect a jumper from TP-1  to TP-8 on the electronic board.  This
should cause the light to go out.  Readout should slowly return  to 0.00 +_

                                     90

-------
.03 VDC.  If not, adjust R-61 (zero) for correct reading.
     6.  Repeat steps 5 and 6 until interaction is out of circuit.
     7.  Place "air - operate - Freon" switch in the "Freon" positon.
Unhook the tubing coming from the top of the Y shaped tube fitting above  the
small clean air pump.  Insert the delivery tube from the Freon bottle  in  the
tubing and turn the valve on top of the bottle on.  A flow-limiting orifice
is in this valve so that proper flow occurs when the valve is fully open.
     8.  After readout becomes steady (15 minutes or more) the output  voltage
should be 1.00 +_ .03 VDC.  If not, adjust R-64 (Freon) for correct reading.
     9.  Connect jumper from TP-1 to TP-9.  Readout should slowly rise to
4.00 +_ .25 VDC.  If not, adjust R-57 (calibrate) for correct reading.
    10.  Remove jumper and observe that readout returns to 1.00 +_ .03  VDC.
If not, adjust R-64 (Freon) for correct reading.
    11.  Repeat steps 9 and 10 until there is no interaction.
    12.  Unit is calibrated.  Unplug air inlet and outlet, remove Freon con-
nection and reconnect tubing to the top of the Y shaped tube fitting in the
clean air system, reconnect inlet hose at top of instrument and move "air -
operate - Freon" switch to "operate" position.
                                     91

-------
                            A.3  AUDIT PROCEDURES

     The audit work performed during the course of this study was in most
part performed with the Bendix 8861 portable calibrator.  Additional audit
work was performed with standard gas cylinders introduced directly to the
analyzer audited.  All such procedures are outlined below.

A.3.1  SULFUR ANALYZER AUDITS
     Establish a flow of zero air from the manifold on the 8861D to the
sample air inlet of the sulfur instrument under audit.  The flow rate used
is determined by the instrument under test.  For example, an instrument such
as the Sign-X uses 2.5 LPM while a Tracor uses only 0.360 LPM.  Choose an
air flow from the calibrator in excess of the demand of the instrument under
audit.  Once a zero air flow is established and the instrument output
voltage is recorded, introduce the output from the 8861P into the dilution
air and establish the highest sulfur concentration to be used in the audit.
For example, when auditing a Tracor, set up to deliver 395 + 7 cc of zero
air.  (The 8861D is set at 12% of gauge (P2) and orifice 1 is open as per
Table A.I.)  Connect the 1/8" Teflon line (the permeation oven output) to
the input side of the dilution chamber on the 8861D.  Sulfur dioxide gas is
now being mixed in the dilution air in the proper proportions to give 0.900
ppm S0?.  After conditions have stabilized, read the voltage- corresponding to
0.900 ppm S02.  Next, using only orifice No. 1, set P2 to 35% of gauge.
The concentration of S02 is now at 0.700 ppm.  Continue to add more dilution
air as indicated in Table A.I until four or more points (plus zero air) have
been established.

A.3.2  OXIDES OF NITROGEN ANALYZER AUDIT
     Use a rotameter on the input of the instrument to check total sample
flow which should be about 500 cc/minute.
                                     92

-------
     Calculate the desired gas concentrations to be used to audit the
instrument using the following equation:

  (NO cone, in Tank, ppm) x  (cc/min thru V,)    ..n       .   ..
     (cc/min thru VJ) +  fcc/min thru Vg)  ]   = N0 concentration in ppm

     A table should be developed and used throughout this audit.  An example
of a calculation and an appropriate table is shown in Table A.6.

          TABLE A.6  SAMPLE TABLE TO BE USED IN THE NOX - NO AUDIT
Desired
Cone, ppm
0.40
0.30
0.20
0.10
0.05
zero
cPJ
4.0
4.0
4.0
4.0
4.0
0
PI
% Gauge
80.5
80.5
80.5
80.5
80.5
0
P2
cc
910
1213
1820
3640
7280
7290
P2
% Gauge
33.0
75.0
81.0
24.5
88.0
88.0
Orifice
No.
1+2
1+2
1+3
1+4+5
1+2+3+4+5
1+2+3+4+5
NOY
AVolts
.808
.604
.390
.194
.094
.004
NO
Volts
.808
.600
.390
.193
.092
.002
Sample Calculation:
                                         = 91° cc/min flow
     (91.0 ppm) = cone, of NO in NBS supply tank
          (4.0) = cc/minute flow thru capillary 1
         (0.40) = desired ppm of NO for first audit point

     The flow is established for the first point (0.40 ppm of NO) and after
the recorder indicates stable conditions the voltage is read on both channels.
Next, P2 is opened to 75% giving a flow of 1213 cc/minute.  The NO concentra-
tion now flowing to the manifold is 0.30 ppm.  Again, after conditions have
stabilized a voltage reading is taken after the last reading (0.05 ppm).  The
NO tank is turned off and a zero air reading is taken.
     The N0? converter in the oxides of nitrogen analyzer is challenged
during the course of the audit by the partial conversion of nitric oxide to
nitrogen dioxide prior to delivery to the analyzer audited.  The results
obtained by monitoring the nitric oxide (NO) and the oxides of nitrogen
                                     93

-------
(NOX) channels of the analyzer audited are used to determine the converter
efficiency.   As a typical example, the following audit data obtained at
RAMS station 114 is presented.
                    NO
Auditor
Value
(ppm)
.450
ozone
added
.350
.250
.150
zero
Instrument
Reading
(volts)
4.8413
1.5258
3.7792
2.6635
1.6479
.0341
Instr.
Reading
(ppm)
.473
.146
.368
.258
.158
-.001
Auditor
Value
(ppm)
.450
ozone
added
.350
.250
.150
zero
Instrument
Reading
(volts)
4.8217
4.7998
3.7792
2.6684
1.6074
.0512
Instr.
Reading
(m-1
.483
.480
.377
.254
.163
-.001
        N0/N02 CONVERTER EFFICIENCY:      99%
    Converter Efficiency =
(ppm NOY, 0, off) - (ppm NOY, 0, on)
_ _        .     . _
(ppm NO, 03 off)  -  (ppm NO, QS on)
inn
IUU
              FIGURE A.5  CONVERSION EFFICIENCY DETERMINATION

     The partial conversion of nitric oxide is achieved by energizing the
ozone generator within the Bendix auditor just after the establishment of
an equilibrium analyzer response to the first audit nitric oxide concentra-
tion delivered at .45 ppm (in this example).  The equilibrium response after
ozone addition is recorded for both the NO channel and the NOX channel.
Conversion efficiency is then calculated as indicated in Figure A05.

A.3.3  OZONE ANALYZER AUDIT
     Prior to the audit of every ozone analyzer,  it is advisable to check the
output of the ozone source in the Bendix 8861D by performing a N0-0~ titra-
tion.  A convenient time for this check will occur during the oxides of
nitrogen audit,  if the monitoring station being audited is equipped with an
oxides of nitrogen analyzer.  If it is not  so equipped, the ozone source
should be checked before and after the audit using the Winnebago instrumen-
tation and the  procedures described in section A.2.2.1.  After verification
of the ozone generation rate, develop the conditions for the audit such as
                                      94

-------
those indicated in Table A.7.
               TABLE A.7  SETTINGS TO USE ON BENDIX CALIBRATOR
                   TO PRODUCE VARIOUS OZONE CONCENTRATIONS
Fixed conditions:
!i !^
80% 4.0
Des i red
Ozone
Cone, (ppm)*
0.400
0.200
0.138
0.100
0.083
0.040

P2 Orifice
49% 1
cc Air/mi n
used to dilute
NO supply
526
1055
1530
2104
2550
5320


Orifice used
develop desi
air flow
1
1, 2
1, 3
1, 2, 3
1, 4
1, 3, 4, 5


to
red






                      *Lamp  at  440  set  point,  500  range

     The Bendix ozone lamp may be calibrated for a higher ozone output.  This
becomes necessary whenever high sample flow rate instruments are to be
audited.  For example, a Sign-X or a Dasibi uses a minimum of 2500 cc/minute.
This demand would only allow the use of the last two concentrations in the
above table.  An alternate scheme is given below for high ozone outputs.  In
this case, the lamp is at SET POINT = 500, RANGE = 1000.

            TABLE A.8  VALUES FOR OZONE AUDIT AT HIGH LAMP OUTPUT
Ozone
Desired
Cone, ppm
0.2550
0.2158
0.1420
0.110
Orifice
1 + 5
1 + 2 + 4
1+4+5
1+2 + 3 + 4+5
Dilution Air
LPM
3.364
3.975
6.044
7.730
                                      95

-------
     After establishing the appropriate Bendix 8861 settings for the analyzer
under audit, introduce each point in turn and record the analyzer response.

A.3.4  MRI 1561 NEPHELOMETER AUDIT
     A red and black binding post are located on the right hand side of this
instrument.  Connect a DVM to these terminals.  Do not use the meter on the
front cover to assess instrument behavior.  Next, proceed as follows:
     1.  Unhook the inlet hose and plug the intake.
     2.  Move the "air - operate - Freon" switch to the "air" position.
     3.  Unplug the lower outlet and wait at least fifteen minutes after the
machine is turned on or until meter reading and/or voltage output are steady.
     4.  Readout should be 0.00 VDC +_ .03.  Record this voltage.
     5.  Connect a jumper from TP-1 to TP-8 on the electronic board.  Record
the voltage after it reaches a steady value.  Remove jumper when step 5 is
complete.
     6.  Unhook the tubing coming from the top of the glass "T" above the
small clean air pump.  Insert the delivery tube from the Freon-12 bottle in
the tubing and turn the valve on the top of the bottle on.  A flow-limiting
orifice is in this valve, supplied by the vendor along with the bottle, so
that proper flow occurs when the valve is fully open.
     7.  After the readout becomes steady, record the voltage.
     8.  Connect a jumper from TP-1 to TP-9.  Readout should rise slowly.
When it is steady, record the voltage.
     9.  Remove the jumper and record the voltage after it again stabilizes.

A.3.5  AUDIT PROCEDURE FOR ATMOSPHERIC METHANE, CARBON MONOXIDE AND TOTAL
       HYDROCARBONS
     Establish a flow of zero air from a tank of known purity as follows:
connect a  1/4" Teflon line to the tank and terminate this line with a "T".
One side of the "T" goes to  the sample air inlet of the instrument under
test and the other side goes to a rotameter.  The outlet valve on the tanks
is regulated so as to keep excess air flowing through the rotameter.  If

                                     96

-------
the demand of the instrument under test exceeds the regulated output of the
tank the rotameter reading will fall to zero and the instrument under test
will take in ambient air.  After a voltage value is established for zero air,
replace the zero air tank with a tank containing a gas mixture of known
composition and obtain the analyzer response for the mixture.
     As an alternate procedure, the audit standards are prepared in Teflon
bags.   The bags are first purged and evacuated several times with zero air.
The bags are then filled with audit gas(es) and delivered for introduction
to the analyzer audited.  Sampling in this case is directly from the bag to
the analyzer through the analyzer sample line with no external venting.
                                     97

-------
                           A. 4  AIR MANAGEMENT

     Because of the fact that the Bendix calibrator does  not  produce  air  of
quality adequate for all audits, it is frequently necessary to  substitute
cylinder zero air.   Depending upon whether the audit is performed  at  a  RAMS
station using RAMS zero air, or at some other location, the choice of zero
air is somewhat determined by the pollutant under study.   A comparison  of
RAMS station zero air (No. 101 on 2/20/76) and Scott cylinder zero air
showed the following:
                                STATION AIR           CYLINDER  AIR
                            Analyzer Response,     Analyzer Response,
         ANALYZER               Volts/ppm              Volts/ppm
         03                     .0284/ .000            .0207/-.001
         NO                    -.0056/-.002            .0402/ .002
         NOX                    .0479/-.002            .0476/-.002
         TS Low                 .0634/-.002            .1171/-.001
         H2S Low                .0284/-.002            .1134/ .000
         S02 Low                .0284/-.002            .0976/ .000
         CO                     .1310/-.036            .1318/-.034
         CH4                    .1896/-.017            .0683/-.267
         THC                    .2368/-.198            .2575/-.149
      In general:
         1)  The Bendix auditor with charcoal scrubbers will  furnish air of
             sufficient quality to audit only the sulfur, NO, and ozone
             analyzers.
         2)  Ultra pure cylinder air should be used for CH.,  CO, and THC.
             The Bendix auditor is not capable of removing these gases  from
             ambient air.   RAMS station zero air is generally of high
             quality but  depends on the efficient operation of a catalytic
             oxidizer.
                                     98

-------
                      APPENDIX B




MODIFIED BENDIX MODEL 8861 PORTABLE CALIBRATION SYSTEM
                          99

-------
                                 CONTENTS

B.I   Introduction                                                     102
B.2  Modifications                                                    103
     B.2.1  Zero Air System                                           103
     B.2.2  Flow Measuring System                                     103
     B.2.3  Permeation Tube Chamber                                   115
     B.2.4  Monitoring Console                                        118
B.3  Instructions for Use of the Modified 8861                         123
                                    100

-------
                                 FIGURES

Number                                                                 Page
 B.I       Modified Bendix 8861P Flow Diagram                           104
 B.2       Modified Bendix 8861D Flow Diagram                           105
 B.3       MFM #1                                                       108
 B.4       MFM #2 (NO)                                                  110
 B.5       MFM #3 (Ozone)                                               112
 B.6       MFM #4 (Sulfur)                                              114
 B.7       Permeation Tube Holder                                       116
 B.8       Modified Bendix 8861P Permeation Tube Holder                 117
 B.9       Calibration of Fenwal GB41P2 Thermistor                      119
 B.10      Wiring Diagram                                               122

                                  TABLES
Number                                                                 Page
 B.I       Mass Flow Meter No. 1 (Main Dilution Air)                    107
 B.2       Mass Flow Meter No. 2 (NO)                                   109
 B.3       Mass Flow Meter No. 3 (Ozone)                                111
 B.4       Mass Flow Meter No. 4 (S02)                                  113
 B.5       Pin Assignments - Mass Flow Meters                           120
                                    101

-------
                              B.I  INTRODUCTION

     During the course of RAMS-RAPS audit work involving the routine use
of the Bench'x Model 8861 portable calibration system, a number of problem
areas were noted.  Experience gained during the use of the system enabled
the audit team to design, install, and test several modifications to the
calibrator to improve reliability in future use.  The modifications and the
instructions for use of the modified 8861 are reported herein.
                                     102

-------
                             B.2  MODIFICATIONS

     All modifications to the 8861 are referenced to Figure B.I, the Modified
Bendix 8861P Flow Diagram and Figure B.2, the Modified Bendix 8861D Flow
Diagram.  All modifications have been completed, tested, and documented.

B.2.1  ZERO AIR SYSTEM
     As delivered, the Bendix 8861D dilution air system is supplied with a
small and generally inadequate zero air scrubber system.  Only at relatively
low flow rates and low background pollutant levels was the supplied scrubber
found to be adequate.  The stock scrubber was removed and replaced with
Drierite columns as shown in Figure B.2.  Generally a single column con-
taining 6-16 mesh coconut shell activated charcoal was found to be adequate
for pollutant removal.  Two columns charged with Drierite were used for
water vapor removal.  Ascarite was employed only in cases where carbon
dioxide removal was required and this usage occurred only during the in-
vestigation of the effect of carbon dioxide on the response of flame
photometric sulfur dioxide analyzers.
     The series configuration of Drierite and charcoal was adequate for all
audit work conducted with the exception of audits involving carbon monoxide
and/or hydrocarbon analyzers.  Cylinder zero air was substituted in these
instances.

B.2.2  FLOW MEASURING SYSTEM
     Flow control within the Bendix 8861 as delivered is by pressure drop
across capillaries.  All five dilution air capillaries and the nitric oxide
capillary are housed in a thermostated chamber.   Purge flow through the
permeation tube chamber is controlled by capillary but this capillary is
not temperature controlled.
     During the course of work involving the 8861, it became obvious that
                                    103

-------
CD CX
                                                                                       o:
                                                                                       ts
                                                                                       o

                                                                                       o
                                                                                       00
                                                                                       00

                                                                                       X
                                                                                       I—I
                                                                                       o
                                                                                       •z.
                                                                                       LLJ
                                                                                       CO

                                                                                       O
                                                                                       O
                                                                                       O
                                                                                       02

                                                                                       UJ
                                            104

-------
105

-------
mass flow meters would be a desirable addition to the system.  The 8861 is
only conveniently flow calibrated in a laboratory under controlled conditions
of temperature and pressure.  Since the 8861 must be transported and used
under a variety of conditions involving significant departure from normal
temperature and pressure, the capability of measuring mass flow would repre-
sent an improvement.  Additionally, the mass flow meters are an invaluable
aid in leak checking the 8861 after field transport and prior to use.
     The first mass flow meter added to the system is shown in Figure B.2 as
MFM 1, a 0-5 liter per minute Tylan unit just upstream of pressure regulator
PR2.  Total dilution air flow is conveniently monitored at this point.  Addi-
tionally, the entire 8861D system is conveniently leak checked by sealing
the 8861 outlet port and monitoring the MFM 1 flow signal.  The calibration
data and calibration curve for MFM 1, Dilution Air, are presented as Table
B.I and Figure B.3 respectively.  The second mass flow meter is shown in
Figure B.2 as MFM 2, a 0-10 cc/minute Tylan unit just upstream of pressure
regulator PR1.  This flow meter monitors nitric oxide flow from an external
cylinder.  The calibration data and the calibration curve are presented as
Table B.2 and Figure B.4 respectively.  The third mass flow meter is shown
in Figure B.2 as MFM 3, a 0-500 cc/minute Tylan unit just upstream of the
ozone generator serving to monitor the air flow over the ultraviolet lamp
within the generator.  The calibration data and calibration curve are pre-
sented as Table B.3 and Figure B.5 respectively.  The fourth mass flow meter
is shown in Figure B.I as MFM 4, a 0-100 cc/minute Tylan mass flow controller
just upstream of the permeation tube chamber.  This mass flow controller
replaces the flow controlling capillary which, prior to removal, limited
flow through the permeation  tube chamber to 5 cc/minute.  The capillary
removed was not thermostated and variations in room temperature during
audits of sulfur analyzers caused  the purge flow to change and eventually
result in an analyzer trace  in oscillation  reflecting the cyclic flow
pattern through the permeation tube chamber.  The mass flow controller has
eliminated this problem and  enables an operator selected purge flow  rate up
to 100 cc/minute.  The calibration data and calibration curve for MFM  4  is
presented as Table B.4 and  Figure  B.6 respectively.
                                     106

-------
   TABLE B.I  MASS FLOW METER NO.  1  (Main Dilution Air)
MFM #1
VOLTS
0.500
1.000
1.500
2.000
2.500
3.000
3.500
4.000
4.500
5.000
5.500
6.000
P-2
% OF
GAUGE
47.
51.
18.
49.
88.
18.
32.
49.
68.
90.
41.
54.
5
0
5
0
0
0
5
0
5
0
0
5
TIME
FOR
1000CC
118.
58.
39.
30.
no.
100.
86.
75.
67.
60.
55.
50.
6 sec
0
2
0
2*
2*
6*
6*
4*
0*
0*
0*
OBSERVED
FLOW
CC/MINUTE
506
1034
1531
2000
2517
2994
3464
3968
4451
5000
5455
6000
FLOW
CC/MINUTE
AT ST'D
CONDITIONS
451
922
1364
1782
2243
2668
3087
3536
3966
4456
4861
5347
ORIFICE
1
1
1
1
1
1
1
1
1
1
1
1

+ 2
+ 2 +
+ 2 +
+ 2 +
+ 2 +
+ 2 +
+ 2 +
+ 2 +
+ 2 +
+ 2 +
+ 2 +


3
3
3
3
3
3
3
3
3
3





+ 4
+ 4
+ 4
+ 4
+ 4
+ 4 +
+ 4 +










5
5
*Time for 500 cc
All flows corrected to standard conditions  of 0°C  and  760  mm Hg.
Correction factor = Observed Flow x


                  =  0.891
P   273  1  [748 - 18.61
[273 + 2l|  I    760     I
                            107

-------
1000
2000     3000      4000
      FLOW, SCCM
5000
            FIGURE B.3  MFM #1
                    108

-------
TABLE B.2  MASS FLOW METER NO.  2 (NO)
P-2
MFM #2 % OF
VOLTS GAUGE
0.445
0.618
0.800
1.000
1.220
1.446
1.670
1.920
2.175
2.452
All flows
Correction
10
20
30
46
50
60
70
80
90
100
corrected to
TIME
FOR
4.0 CC
217.0
158.2
124.9
101.8
83.8
69.0
60.3
53.0
47.2
41.4
standard
factor = Observed Fit
OBSERVED
FLOW
CC/MINUTE
1.106
1.517
1.922
2.358
2.864
3.478
3.980
4.528
5.085
5.797
FLOW
CC/MINUTE
AT ST'D
CONDITIONS
0.99
1.36
1.72
2.11
2.56
3.11
3.56
4.05
3.55
5.19
conditions of 0°C and 760 mm Hg.
jw x f 273
M X [273 + 23.8
759.2 - 22.1
760
i. j
     = 0.892
                  109

-------
       567
     FLOW, SCCM
8
10    11
FIGURE  B.4  MFM #2 (NO)
        no

-------
           TABLE B.3  MASS FLOW METER NO.  3 (Ozone)
MFM #3
VOLTS
3.151
3.598
3.975
4.341
4.705
5.020
5.317
5.616
5.889
over
range
P-2
% OF
GAUGE
10
20
30
40
50
60
70
80
90
100
TIME
FOR
1000 CC
172.4
150.8
136.8
125.4
115.8
108.8
101.8
96.8
91.4
87.4
OBSERVED
FLOW
CC/MINUTE
348
398
438
478
518
551
589
620
656
686
FLOW
CC/MINUTE
AT ST'D
CONDITIONS
310
355
390
426
462
491
525
553
585
611
All flows corrected to standard conditions  of 0°C and 760 mm Hg.
Correction factor = Observed Flow x
                  = 0.891
  273
273 + 21
748 - 18.65
   760
                               in

-------
   6  -
   5 -
oo

5  4
   3  -
      300   350
 I
400
 »
450
 I
500
 I
550
600   650
                     FLOW cc/min.
           FIGURE B.5  MFM #3 (Ozone)
                       112

-------
                 TABLE B.4  MASS FLOW METER NO. 4 (S02)
This is a mass flow controller operating from a 10 turn command potenti-
meter.
POTENTI-
METER
READING
100
200
300
400
500
600
700
800
900
1000
MFM #4
VOLTS
0.486
0.975
1.467
1.956
2.445
2.938
3.426
3.918
4.410
4.905
TIME
FOR
10 CC
53.9
26.6
17.8
13.3
10.6
89.8*
78.2*
68.2*
60.7*
54.6*
OBSERVED
FLOW
CC/MINUTE
11.131
22.556
33.708
45.113
56.604
66.815
76.726
87.976
98.847
109.890
FLOW
CC/MINUTE
AT ST'D
CONDITIONS
9.939
20.141
30.099
40.283
50.543
59.661
68.511
78.556
88.263
98.124
OVEN
TEMPERATURE
(VOLTS)
0.548**
0.548**
0.548**
0.548**
0.548**
0.548**
0.548**
0.548**
0.548**
0.548**
*  Time for 100 cc

** 0.548 volts corresponds to 30.00°C


     All flows corrected to standard conditions of 0°C and 760 mm Hg.
   .  Correction factor = Observed Flow x
                       = 0.893
   273
273 + 22
753.1  - 19.8
    760
                                    113

-------
oo  3
h-
_i
o
        --".t:
            ---I
                        7t:
                              ;:.t;-
                         I- -:t-
                                     777t7
                                                    -t:;-;
                                                            1.17;
                                                                       :::-:^
        NIFM».OQS\/
         in::;:;
              10
20
 I

30
40     50     60


  FLOW,  SCCM
 i

70
80
 i

90
100
                             FIGURE  B.6   MFM #4  (Sulfur)
                                         114

-------
B.2.3  PERMEATION TUBE CHAMBER
     A platinum resistance thermometer which measures the temperature of the
permeation tube is part of the Bendix design.  It does not measure the air
temperature surrounding the permeation tube but rather the temperature of a
massive aluminum block surrounding the inner chamber.
     A redundant back-up system capable of measuring purge air temperature
seemed highly desirable and was installed.  A 3/4" diameter Teflon rod was
bored out, milled and threaded as shown in Figure B.7.  The interior was
enlarged to accommodate two 5.0 cm permeation tubes.  As a precaution
against future leaks, stainless steel bands were turned on the lathe and
press-fit over the ends of the Teflon permeation tube holder.  As Teflon
ages it has a tendency to spread, which could cause leaks at both ends of
the holder.  The steel bands will prevent this from occurring.
     A male run "T" was fitted to the rear of the holder and a previously
calibrated glass thermistor was sealed into the side of the "T".  The
thermistor was secured in place and the joint was made gas tight using
epoxy cement.  Finally, a five foot length of 1/8" copper tubing was
spirally wound around the Teflon permeation tube holder and the three
components were assembled as shown in Figure B.8.  The entire assembly is
contained within the permeation tube oven.
     The outlet of the permeation tube holder was plugged, 15 psi air was
introduced and the system was leak-tested.  The new design advantage be-
comes obvious at this point.  If the rear connections are leak-free it is
unlikely they will leak in the future because this connection is never
disturbed.  (In the old design, the permeation tube was inserted from the
rear.)  To reconnect the permeation tube holder to the purge air it is only
necessary to make a final connection between the 'air in1 and the outlet of
mass flow controller No. 4.  This connection is accessible for leak check-
ing.
     To change permeation tubes the front fitting (purge air exit) can be
removed and the permeation tube becomes accessible with forceps.  The
entire assembly is free to slide from the oven.  Tilting the assembly
discharges the permeation tube from the Teflon cavity.  After a new tube

                                     115

-------
                                     o:
                                     UJ
                                     o
                                     CO


                                     \-



                                     o
                                     KH


                                     fi
                                      CO

                                      UJ
                                      tx.

                                      3
116

-------
                                 UJ
                                 O
                                 _l
                                 O


                                 UJ
                                 CO
                                 UJ
                                 UJ
                                 Q-
                                 UD
                                 00
                                 00

                                 X
                                 I—4
                                 O
                                 z
                                 UJ
                                 CO

                                 Q
                                 O
                                 O
                                 CO

                                 GO
                                 UJ
                                 cc
117

-------
is inserted, all leak checks can be made external to the chamber.
     With the new design, the purge air temperature may be monitored directly
just prior to exposure to the permeation tube.  The leading copper tubing
insures that air temperature reaches the control temperature of 30.0°C re-
gardless of ambient temperature.  The data included in Table B.4 clearly
shows that this temperature is maintained throughout the 0-100 SCCM oper-
ating range of MFC 4.  The calibration curve for the Fenwal thermistor
installed in the tube holder is presented as Figure B.9.

B.2.4  MONITORING CONSOLE
     The wiring diagram for the monitoring console enabling an operator to
display any mass flow meter signal or thermistor signal is shown in Figure
B.10.  All pin assignments are given in Table B.5.  The console consists of
a 6000 count digital voltmeter mounted in a control panel.  The input to the
DVM is switched to display either the thermistor signal (position 1) or the
mass flowmeter signals (MFM 1-4, positions 2-5).  The command potentiometer
for MFM 4 is also mounted on the panel.  The pot is 10-turn and is linearly
proportional to a flow of 0-100 cc/minute through the permeation tube
chamber.
     All of the mass flow meters require a regulated +_  15  VDC supply for
operation.  Connections to Tylan power supply PS-14 are shown in Figure
B.10.  Detail for connections  (typical, 4 places) to MFM 4 is also shown in
Figure B.10.  The Fenwal thermistor mounted in  the permeation tube chamber
is  in a voltage divider across  the regulated +  15 VDC Tylan supply.
                                      118

-------
    31.0-
o   30.0-
    29.0-
LU
    28.0.
                                          30.O°C
               I     i      t      i      i     1      I      r
              .51   .52   .53    .54   .55   .56   .57    .58

                             INDICATED VOLTS
       FIGURE B.9  CALIBRATION OF FENWAL GB41P2 THERMISTOR
                               119

-------
TABLE B.5


Pin
Assignment
W
S
17
U
V


R
L
12
N
P


K
E
7
F
J
PIN ASSIGNMENTS - MASS
FLOW METER NUMBER 1
AIR
0-5LPM
Function
-15 V
Case Gnd.
Common
Output
+15V
FLOW METER NUMBER 2
NO
0-10 cc/min
-15V
Case Gnd.
Common
Output
+15V
FLOW METER NUMBER 3
OZONE
0-55 cc/min
-15V
Case Gnd.
Common
Output
+15V
FLOW METERS


Wire
Color Code
White
Braided shield
Black
Green
Red


White
Braided shield
Black
Green
Red


White
Braided shield
Black
Green
Red
120
                            (continued)

-------
                 TABLE B.5 (continued)
             FLOW METER/CONTROLLER NUMBER 4
                          so2
                      0-100 cc/min
   Pin
Assignment              Function
   25                   -15V
    Y                   Central
    X                   Case Gnd.
    a                   Common
    Z                   Output
    F                   +15V
   Wire
Color Code
White
Yellow
Braided Shield
Black
Green
Red
                           121

-------
                                            DETAIL-MFM. NO. 4
                                                   FC 26O
                                                   IOO ec/mio.
 Switch
Position
    1
    2
    3
    4
    5
                                                           CONTROL
                                                           veusw)
                                                           O-Sv/OC
 COMK.
(1BLACK)
 ounvr
                                                            COMM.
                                                           COMM.
AN
JER
PLY
14

1 	 v^
\O TUR
BOURN
POT.
Signal
Thermistoi
MFM 1 (Oil
MFM 2 (NO
MFM 3 (Oz<
MFM 4 (SO.
-D OUTPUT 1 VIOLET 	
- K otrrPUT 3 6REEN 	 ,
-A -t-Bv/OC PINK — |
-5 COP<^^^ON BiACK — •
-(t> COMMON SLOE — i
-7 -MSvOC ORANGE -i
Y ' V


N
IS
1 3;
5
«
ution)
( H5VAC
3ne)
,)
-i r-IBvDC
r«o>)
ZENER
TESTPT
4

6
D

F
VALVE
TE5TP

-ISv/D
(WHITE,
^
'• ^_^\_ . •
, Ti-


1 j
i
RED


3

^ *TJ :

• -t

re' THE
Tft2 i
A «

uj

i
DVM
TR?i '
4
                                                               FENVVA\_
                                                              THERMISTOR
                   FIGURE B.10  WIRING DIAGRAM
      (Interconnections  between Power Supply  - Mass Flow Meter
      and Control  Box mounted on the modified Bendix 8861P)
                               122

-------
               B.3  INSTRUCTIONS FOR USE OF THE MODIFIED 8861

     The following operating procedures pertain to the operation of the
modified Bendix 8861D and 8861P and involve the use of mass flow meters to
measure the flow of dilution air and/or calibration gas delivered by the
above.
     To set up for audit proceed as follows:
     1.  Open both cases, side by side and make the following intercon-
         nections;
         a)  Connect the 52" Teflon line between the "Clean Air" on the
             8861D and the input filter marked "air in" in the 8861 P.  This
             line supplies purge air to the S0£ permeation assembly.  Refer
             to Figure B.I, Modified Bendix 8861P Flow Diagram and Figure B.2,
             Modified Bendix 8861D Flow Diagram.
         b)  Turn the toggle switch VI, Figure B.I to "off" when the pump in
             the 8861D is started.
         c)  Make sure the air scrubbers are connected together as shown in
             Figure B.2 (i.e., drierite - drierite - carbon - ascarite).
             Connect the first drierite bulb to the air "out" and the outlet
             of the ascarite bulb to the "in" connections on the 8861D.
             These scrubbers will produce dry pollutant free air, free of
             carbon dioxide.  Under certain conditions it may be advisable
             to use charcoal alone depending on the analyzer being cali-
             brated or audited.
         d)  Connect the mass flow meter cables being careful to match the
             number on the plug with the number on the mass flow meter
             (i.e., No. 1, 5.0 LPM dilution air; No. 2, 10 cc NO; No. 3,
             500 cc ozone; No. 4 100 cc S02).  Use the Phillips head screws
             to secure the plug to the circuit board connection on the top
                                     123

-------
        of the mass flow meter.
    e)  Plug in the 8861D using the plug strip mounted in the 8861P
        and connect the plug strip to a 115V source.
    f)  Turn on the power switch on the 8861D.  This turns on the pump.
        The toggle valve mentioned in part b) above can now be turned
        to "off" and the "stand-by - off-on" switch on the 8861P can
        now be turned to "on".
The equipment is now in operation.
2.  Dilution flow calculations can be made as shown in the "Audit
    Procedures for RAPS Instrument Systems" and reviewed briefly below:
    For S02:
                 C = £R x £
        where:  C is concentration in PPM
                PR is permeation rate in ng/min.  (Ref. oven temperature)
                K is 22.4/M.W. of permeation gas  (M.W. S02 = 64)
                F  is permeation oven flow rate
                F. is dilution flow rate
                 d
    For NO:
                 r v  = r v
                 Lri   L2V2
        where:  C-, is concentration of NO source  gas
                V-, is the flow rate of NO source  gas
                C2 is final concentration
                V2 is total dilution flow rate.
3.  The quantity of dilution  flow can be read  from  the graphs  accom-
    panying this report.  The total dilution air  flow is  read  from mass
    flow meter No. 1, the NO  flow is read from mass flow  meter No.  2,
    the ozone is read from mass flow meter No. 3  and the  S02 is read
    from mass flow meter No.  4.  The volume of purge air  entering  the
    permeation oven is continuously variable between 2-100 cc/min  and
    care must be taken to insure that this volume is added to  F. when
                                                                d
                                124

-------
    the flow calculations are made.
4.  On shut-down, disassemble in the reverse order to set-up and observe
    the following precautions:
    a)  Turn the "stand-by, off, on" switch to stand-by.  The air pump
        is now running off the internal battery in the 8861 P.
    b)  The toggle valve VI is turned to "on".
    c)  The "air-in" filter in the 8861D is closed with a cap.  If  this
        is not done, air from the pump escapes through the filter and
        no purge air goes through the permeation oven.  As a final  check,
        turn the selector switch on the control panel to position 5 and
        note that a voltage is indicated on mass flow meter No. 4 show-
        ing that purge air is flowing through the permeation oven.
    d)  The Bendix 8861P should not be left on its stand-by battery for
        more than 24 hours.  High current drain will nearly totally
        discharge the battery in this time interval.
                                125

-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
1. REPORT NO.
   EPA-600/4-79-031
                                                          3. RECIPIENT'S ACCESSION"NO.
4. TITLE ANDSUBTITLE
  REGIONAL AIR POLLUTION  STUDY
  Quality Assurance Audits
            5. REPORT DATE
               May  1979
            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

  0. Klein and F. Littman
                                                          8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Rockwell International
  Air Monitoring Center
  11640 Administration Drive
  Creve Coeur, MO   63141
             10. PROGRAM ELEMENT NO.

               1AA603 AA-126 (FY-79)
             11. CONTRACT/GRANT NO.
               68-02-2093
               Task Order 106
12. SPONSORING AGENCY NAME AND ADDRESS
  Environmental Sciences  Research Laboratory - RTP,  NC
  Office of Research  and  Development
  U.S. Environmental  Protection Agency
  Research Triangle Park,  NC  27711
             13. TYPE OF REPORT AND PERIOD COVERED
               Final
             14. SPONSORING AGENCY CODE
               EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
       RAPS Quality Assurance  audits were conducted under  this  Task Order In contin-
  uation of the audit  program,  previously conducted under Task Order No. 58,  Quantita-
  tive field audits were  conducted of the Regional Air Monitoring System (RAMS) Air
  Monitoring Stations,  Local Air Monitoring Stations  (State  of Illinois, St, Louis City
  and St. Louis County),  RAPS  helicopters and various measurement systems employed in
  the RAPS intensive studies,   Audit results are reported  for systems measuring NO, NO
  03, S02, total sulfur,  total  hydrocarbons,
CH4  and  CO,
       An Investigation  was  conducted on the effect of  Teflon parttculate filters on
  NO, NOo, 0,, and  SQ2 concentrations In sampled air,   Measured sample losses are
  reported for synthetic pollutant T- air mixtures  sampled  through new and used filters
  under both dry and  humid conditions.  The investigation  also reyealed effects of
  humidity on the response of analyzers to the various  pollutants.

       The accuracy of S02 calibration mixtures prepared with the commercial dynamic
  calibration system  employed In the audits was investigated,  AS a result of this
  Investigation and experience gained during the audits, the calibration system was
  modified to Improve performance under field conditions,
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                             COS AT I Field/Group
  * Air Pollution
  * Quality Assurance
    Auditing
  St.  Louis, MO
  Regional  Air
        Pollution Study
13B
14D
05A
18. DISTRIBUTION STATEMENT
  RELEASE TO  PUBLIC
19. SECURITY CLASS (ThisReport)
  UNCLASSIFIED
                                                                         21. NO. OF PAGES
                                                                               134
20. SECURITY CLASS (Thispage)
  UNCLASSIFIED
                                                                        22. PRICE
EPA Form 2220-1 (9-73)
                                             126

-------