&EPA
            United States
            Environmental Protection
            Agency
            Environmental Sciences Research
            Laboratory
            Research Triangle Park NC 27711
EPA-600 4-79-024
April 1979
            Research and Development
Regional Air
Pollution Study

Dichotomous Aerosol
Sampling System

-------
                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.  Socioecpnomic 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-600/4-79-024
                                             April  1979
        REGIONAL AIR POLLUTION STUDY

     Dichotomous Aerosol  Sampling System
                     by
                  E.  Nelson
           Rockwell  International
       Atomics International  Division
            Air Monitoring Center
         11640 Administration Drive
           Creve Coeur, MO  63141
           Contract No.  68-02-2093
               Task Order 102
               Project Officer

             Stanley Kopczynski
 Atmospheric Chemistry and Physics Division
 Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 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


     Ten of twenty-five stations making up the Regional  Air Monitoring System
(RAMS) in St. Louis were equipped with dichotomous samplers and high volume
filter samplers for aerosol  measurements.   The dichotomous samplers, built by
Lawrence Berkeley Laboratories (LBL), were designed for automatic operation
and were capable of collecting up to 36 samples in each of two size fractions
before filter stacks must be changed.  Most of the time, the samplers operated
to collect 12-hour samples (0000-1200, and 1200-2400),  except that at two
stations with high aerosol loadings, the sampling intervals were split into
two 6-hour intervals.  Sample filters were pre-weighed  by beta gauging before
exposure in St. Louis and subsequently returned to LBL  for determination of
aerosol mass in each size fraction,  as well  as determination by x-ray
flourescence of the concentrations of the following elements:  Al, Si, P, Cl,
K, Cr, Mn, Ga, Rb, Sr, Sn, Sb, Ba, Hg, S,  Ca, Ti, V, Fe, Ni, Co, Zn, As, Se,
Br, Cd, and Pb. Approximately 33,000 samples were collected between March 1975
and March 1977.  Analytic data are stored in the RAPS Data Bank at Research
Triangle Park, N.C.

     The operation of the samplers in the RAMS network  is described along with
problems encountered and procedures  used for preventive maintenance and
quality control.  Also described are two streaker samplers specially installed
for continuous aerosol measurements  and a silicon cell  pyranometer.
                                     iii

-------
                                  CONTENTS

Abstract                                                                 iii
Tables                                                                    vi
  1.0  Summary and Conclusions                                             1
  2.0  Scope of Work                                                       3
  3.0  Operations                                                          4
       3.1  Corrective maintenance                                         6
       3.2  Failure analysis                                              10
       3.3  Dichotomous sampling network master log                       11
  4.0  Recommendations                                                    14
  5.0  Streaker and Radiometer Operation                                  15
Appendices
  A.  Sampling Equipment                                                  17
  B.  Quality Assurance                                                   34

-------
                                   TABLES
Number                                                                Page
  1       Lawrence Berkeley Laboratories (LBL) Dichotomous
           Sampler Operation 1975-1977                                  7
  2      Sampler Related Failures (Total  Failures - 227)                8
  3      Miscellaneous Failures (Total Failures - 584)                 10
  4      Dichotomous Sampling Network Master Log                       12
                                       VI

-------
                        1.0  SUMMARY AND CONCLUSIONS

     Aerosols, due to the diversity of their sources, physical and chemical
properties, present a special challenge to air monitoring technology.  The
bimodal particle size distribution in the urban aerosol suggests the use of
a dichotomous sampling method with a cut point near two microns.  A sampler
of this nature was constructed by Lawrence Berkeley Laboratory, University of
California at Berkeley.
     In support of the aerosol modeling studies as part of the St. Louis
Regional Air Pollution Study (RAPS), a network of ten automatic dichotomous
samplers was set up and operated to collect particulate pollutant data.  To
facilitate analysis and data handling, the samples were collected on a filter
with computer readable digital labels.  Information on trace elements together
with concurrent chemical and meteorological data will serve to define the air
quality in the St. Louis Region and relate atmospheric composition to various
emission sources.  In addition, it will provide needed input for defining
aerosol growth, transport and decay equations.
     During the period covered by this task order the RAPS LBL Dichotomous
Aerosol Filter Sampling System was established.  Ten samplers were installed
and operated in the St.  Louis Regional Air Monitoring System (RAMS) network.
An elaborate system for filter validation and sampler monitoring was devel-
oped and proved to be invaluable and time saving (see Appendix B).  Thirty-
three thousand six hundred ninety-five filters were collected and there
were nine hundred thirty-four filter failures for an overall efficiency of
97.25%.  Failures caused by sampler malfunctions accounted for 0.67% of the
total failures with overload failures accounting for 0.37%.  Failures from
miscellaneous causes accounted for 1.73% (see Table 1).  It appears that the
virtual impactor, with its inherent advantages, has fulfilled the need for an
instrument to collect aerosols in two distinct size ranges.
     Upon termination of the LBL Dichotomous Aerosol Filter Sampler System,

                                      1

-------
all samplers were returned to the warehouse where eight of them were care-
fully inspected, cleaned and crated for shipment.  Samplers with serial
numbers 3, 6, 7, 8, 9, 10 and 12 were shipped back to Lawrence Berkeley
Laboratory.  Sampler number 5 was placed in station 105 to be shipped out
with the station and sampler number 2 was installed in station 111 for a
continuing sampling program.  The samplers with serial numbers 4 and 11 were
previously shipped back to Lawrence Berkeley Laboratory for modifications.
All of the maintenance log books remain with each sampler, wherever they may
go, as a historical record.
     The LBL Automatic Dichotomous Air Sampler (ADAS) is a very well engi-
neered piece of equipment.  The instrument is capable of monitoring and dis-
playing most of the conceivable failures encountered by this type of instru-
ment, e.g., a.c. power, insert, withdraw, no sample, flow control valve and
valve out of range.  The status monitor and display located on the flow con-
troller instrument panel in combination with the failure display provided a
valuable tool in analyzing instrument malfunction at a glance.  Proper inter-
pretation of the panel display provided for faster repair and reduced sampler
downtime.

-------
                             2.0  SCOPE OF WORK
     Aerosol samples were collected on filters in size ranges greater than
and less than two microns at ten RAMS stations.  Two separate filters sampled
simultaneously from two air streams, depositing particulates of 2 microns or
less on one filter and those larger than 2 microns with a cut point around
10 microns.  Sample mass was determined by beta gauge and the samples were
analyzed for trace elements (atomic number 13 and above) by X-ray fluores-
cence.  Temporal resolution was as fine as two hours.  The extent of this
task order was for the collection of filters, maintenance and operation of
the Automatic Dichotomous Air Samplers.  Mass measurements and analyses of
the samples were performed under a separate contract between the Environ-
mental Protection Agency and the Lawrence Berkeley Laboratory (LBL), Univer-
sity of California at Berkeley.  The data produced from this sampling network
were to become part of the aerometric data base of the Regional Air Pollution
Study.
     Government Furnished Equipment (GFE) were eleven LBL Automatic Dichot-
omous Air Samplers, filter trays and membrane filters to be periodically
supplied by Lawrence Berkeley Laboratory.  A comprehensive quality assurance
plan was submitted to the EPA Task Coordinator.  This report dealt with the
calibration, maintenance, operation of the sampling units; collection,
identification, and delivery of exposed samples; and sampling procedure
validation.  In addition, monthly progress reports were made which included
descriptions of quality assurance activities, status of filter samples,
operational performance of each sampler, major problems and planned correc-
tive action and plans for future efforts (see Appendix B).
     Task Order No. 102 is a continuation of effort previously performed
under Task Order No. 27, under Contract DU 68-02-1081.  Under Task Order
No. 27, the ten Automatic Dichotomous Air Samplers of the Regional Air
Monitoring System  (RAMS) were operated and maintained from 2 January 1975
to 29 August 1975.

-------
                               3.0  OPERATIONS

     In November 1974 eleven Lawrence Berkeley Laboratory Automatic Dichot-
omous Air Samplers were received at the Regional Air Pollution Study (RAPS)
facility.  The samplers were stored in the warehouse area and when uncrated,
each sampler was thoroughly inspected for possible damage during shipment,
then cleaned for a test run with sample filters.  Dr. B. W. Loo from Lawrence
Berkeley Laboratory arrived in January 1975 with calibration equipment and
tests revealed no variances in the calibration set-point adjustments.
     Ten of the samplers were scheduled for installation in the Regional Air
Monitoring System (RAMS) stations.  The eleventh sampler would be used as a
backup unit in case of an emergency.  Before the samplers were installed, two
modifications were made on each sampler:  (1) the virtual impactor was
extended four inches, and (2) the failure mode circuitry was modified.
Modifications were also made in the RAMS stations.  The RAMS modifications
consisted of the installation of an isokinetic intake manifold, installation
of a larger manifold blower and an electronic interface to the computer
control system.  The modifications required for the isokinetic manifold were
completed in January 1975 and the computer interfacing was completed in
February.  The RAMS stations chosen for sampler installation were 103, 105,
106, 108, 112, 115, 118, 120, 122 and 124.  A detailed description of the
theory of the sampler operation is given in Appendix A.
     In February 1975, initial operation of the network was devoted to
"debugging" the samplers and the computer interface, and to the development
of operational procedures.  The dichotomous sampler was engineered to be kept
under surveillance by the RAMS central computer facility.  However, because
of the large number of failure mode indicators provided by the sampler, all
output modes were tied together (logical or) to reduce the number of signal
lines required.  A failure in any mode would be flagged at the central
facility, notifying the computer operator, who in turn notified operating

-------
personnel for corrective action.
     Originally the normal sampling schedule for each of the ten samplers was
to be one 24 hour sample per day, beginning and ending at midnight, except
during periods of "intensive study".  During the "debugging" period it became
apparent very early that this sample frequency was too long, as many filter
samples failed from particulate overloading.  The samplers at stations 103,
105, 108 and 115 were changed to 12 hour sampling periods, and the samplers
at station 106, 112, 118, 120, 122 and 124 were left on the 24 hour sampling
period.  The "debugging" period used 700 filter samples.
     Formal operations began on 17 March 1975.   All samplers were loaded with
matched pairs of filter trays and were synchronized by the central facility
computer.  Station visitation occurred routinely every three days.  Weekend
maintenance occurred only during periods of intensive study.  During regular
sampling periods, when a malfunction was detected by the computer surveil-
lance system, maintenance personnel were immediately dispatched to get the
malfunctioning equipment back on line.  The frequency of sample filter
overload required another evaluation and the samplers at 103 and 105 had
their sample frequencies further reduced from 12 hours to 6 hours on 21 April
1975 and 14 May 1975 respectively.  Subsequently, reductions from 24 to 12
hour sampling frequency were made at station 120 on 18 May, 106 on 19 May,
112 on 22 May, 118 on 3 June, 122 on 24 June, and 124 on 25 June 1975.  The 6
hour sampling frequency at stations 103 and 105 had starting times at 0000,
0600, 1200 and 1800 CST and the 12 hour sampling stations had starting times
at 0000 and 1200 CST.
     Early in the program it was noted that the samplers were not remaining
synchronized and this was somewhat of a mystery because there were no appar-
ent reasons for loss of synchronism.  However,  the mystery began to unravel
when it was discovered that reprogramming the station computer caused the
sampler to abort and remain in standby with no failure flags back at central.
Activating the station CPU inhibit button would also cause a sampler to
abort.  Operating the station teletype would sometimes cause a sampler to
abort.  In an attempt to solve the problem of sampling abortions, when for
any reason the station computer had to be disturbed, the signal cable con-
necting the computer and the sampler was to be removed and replaced when the

-------
attending technician completed his task.  In many instances however, the
technician would forget to replace the plug, thereby causing the loss of
command and surveillance control at the central facility.  In order to
eliminate the problem completely the decision was made to disconnect the
command lines and manually synchronize the samplers whenever the need arose.
The surveillance lines would remain connected to the computer.  The problem
of unexplained abortions was solved when the command lines were disconnected.
Manually synchronizing each sampler required more effort but was well worth
the price because the string of uninterrupted samples was greatly increased.
     During the first week in July 1975 all  samplers were throughly inspected,
cleaned, lubricated and calibrated in preparation for the summer intensive
operation.  Samplers at stations 103, 105 and 112 were scheduled to operate
on a 2 hour sampling frequency.  Samplers at stations 106, 108, 115, 118,
120, 122 and 124 were operated on a 6 hour sampling frequency.  At the end of
the summer intensive all the samplers were returned to their original sam-
pling schedules except the samplers at stations 106 and 124 which were used
in support of a special study initiated by Dr. T. Dzubay, from EPA, Research
Triangle Park, North Carolina.  Dr. Dzubay's study ended the first week in
September and the samplers were returned to their regular sampling schedules.
     During the month of October 1975 all the carbon vanes were changed in
the vacuum pumps.  The sampler at station 108 was damaged by gunfire and
required extensive repair.  The electronic controller had to be returned to
Lawrence Berkeley Laboratory for repairs and the controller was taken from
the spare unit at the warehouse and installed in the sampler at 108.
     For the rest of the sampling program the samplers were held to their
regular sampling schedules except for the samplers at stations 108 and 115
during the month of July 1976 when both sampling frequencies were changed to
6 hours as a precaution against particulate overloading.  These samplers were
returned to their regular sampling frequencies at the end of August.  The
samplers at 106 and 124 were returned to Lawrence Berkeley Laboratory in
August 1976 for modifications and were never returned.

3.1  CORRECTIVE MAINTENANCE
     Regular maintenance is described in detail in the Quality Assurance
                                      6

-------
             TABLE  1.   LAURENCE  BERKELEY  LABORATORIES  (LBL) DICHOTOMOUS SAMPLER OPERATION  1975-1977

Total Samples Collected
Total Samples Failed
Total Uninterrupted
Synchronized Samples
Total Failures Due to
Sampler Malfunction
Total Failures Due to
Filter Overload
Total Miscellaneous
Failures
Percent Failures Due to
Sampler Malfunction
Percent Failures Due to
Filter Overload
Percent Failures Due to
Miscellaneous Causes
Percent Uninterrupted
Synchronized Samples
RAMS STATIONS
103
5874
116
5758
52
18
46
0.885
0.306
0.783
98.025
105
6023
126
5897
46
6
74
0.764
0.100
1.229
97.908
106*
2214
89
2125
14
12
63
0.632
0.542
2.845
95.980
108
2802
105
2697
18
20
67
0.642
0.714
2.391
96.252
112
3492
82
3410
6
16
60
0.172
0.458
1.718
97.652
115
3098
65
3033
30
6
29
0.968
0.194
0.936
97.902
118
2706
112
2594
25
22
65
0.924
0.813
2.402
95.861
120
2808
98
2710
7
13
78
0.249
0.463
2.778
96.510
122
2666
84
2582
16
9
59
0.600
0.337
2.213
96.849
124*
2012
57
1955
13
1
43
0.646
0.050
2.137
97.167
TOTALS
33,695
934
32,761
227
123
584




* The samplers at these stations were removed in August 1976  and returned to the Lawrence Berkeley Laboratories
  for electronic and mechanical  modifications.

-------
Program, Appendix B.  The following paragraphs describe additional measures
that were found to be very helpful in maintaining a smoother operation.  As
shown in Tables 1 and 2, the failures caused by sampler malfunctions were of
a low order of magnitude.  Experience has shown that samplers should be
cleaned immediately after any nearby major farming or earth moving activ-
ities.  During the 30 day cleaning schedule two preventative operations were
found very helpful in enhancing valid filter capture and these were:   (1)
check the adjustment of the micrometer clutch, and (2) check adjustment of
the Geneva wheel.

         TABLE 2.  SAMPLER RELATED FAILURES (TOTAL FAILURES - 227)

Flow Control Valve
Failed on Insert
Geneva Wheel Assembly
Improper Seal
Micrometer Valve Assembly
Null Switch
Vacuum Pump
Failed on Withdraw
Other Various Infrequent Causes
Unknown
NUMBER
45
23
68
8
24
14
10
6
9
20
BY PERCENT
19.82
10.13
29.95
3.52
10.57
6.17
4.41
2.64
3.96
8.81
     Before filter trays are placed  in  the  sampler the  trays  should be held
two or three inches above a tabletop and  turned  so that the open side of the
tray is turned towards the tabletop. ' The tray should  be agitated slightly so
as to cause any loosely held filters to drop  to  the tabletop.   The filter
retainer springs can then be adjusted to  hold the  filters properly.  The
filters should be tested by manually inserting and withdrawing them.
A filter retainer spring adjusted  too tightly can  cause the stackloader to
malfunction because the exposed  filters cannot be  fully tucked back in the
tray and will protrude, causing  the  stackloader  to jam as the mechanism
                                     8

-------
travels upward.  A filter retainer spring adjusted too loose will allow
filters to creep outward as other filters are inserted and withdrawn, and
also may cause jamming of the stackloader.  Filters should also be inspected
for cracks that may occur during shipment.  Filters are checked before
shipment but may be damaged in transit.  Such transit damaged filters will
often fail causing an interruption in the sampling schedule.  Another useful
check on new filters is one in which the backside of the filter is inspected
for extrusions probably left over when the plastic filter holder is molded.
A larger than normal extrusion can also cause sampler failure.  The sampler
checks each inserted filter before allowing the filter to actually start
sampling, and when a filter does not pass the filter seal test the sampler
automatically shuts down until corrective action is taken by operating
personnel.  An extension on the back of the filter holder may cause a
failure when the filter pusher inserts the filter fully forward and then
slips under the extrusion causing the rear end of the filter to be lifted,
thus destroying the filter seal.
     Samplers are normally given a visual inspection on each station visita-
tion for filter and time synchronization and any obvious malfunctions.  Other
checks found to be useful were:  (1) check the knurled knob holding the
pusher for looseness; (2) check the cooling fan for proper operation by
touching the panels in the area of the vacuum pump and feeling beneath the
sampler for exhaust air; (3) check stackloader for partially protruding
filters that may cause the stackloader to jam; (4) check intake sample air
manifold for fully mated sections; and (5) disconnect tubing that checks
pressure for constant sample air flow.  A momentary change in equalization
pressure should cause the instrument to send out a corrective signal to the
micrometer drive motor indicating proper null region adjustment.  When the
tubing is replaced the micrometer drive motor should return the micrometer to
its original setting.
     Maintenance logs were kept at each station where all corrective measures
were entered into the logbook thus giving a complete history of each sampler.
Calibration data entered in the logbooks were checked by noting the microm-
eter calibration set-point when a clean filter was inserted in the sampling
position.  The samplers were found to be extremely reliable in maintaining

-------
their set-points. Periodic checks with calibration equipment also proved the
stability of the sampler in maintaining calibration.

3.2  FAILURE ANALYSIS
     A list of the most prevalent failures other than sampler related failures
is given in Table 3.  The computer related failures were completely elimi-
nated once the command lines were disconnected.  Most of the a.c. power
failures were caused by adverse weather conditions interrupting the power
from the electric utilities except at station 112.  The a.c. power for
station 112 was supplied by the power station at Washington University.
Occasionally a sampler was found in a standby mode with no failure lights
indicated on the panel.  It is thought at this time that spurious signals
getting into the system caused the sampler to interpret them as withdraw
commands.

         TABLE 3.  MISCELLANEOUS FAILURES (TOTAL FAILURES - 584)
._.
A.C. Power Related Causes
Computer Related Causes
Filter Related Causes
Filter Tray Retainer Springs
Other Various Infrequent Causes
NUMBER
79
334
24
100
47
BY PERCENT
13.53
57.19
4.11
17.12
8.05
     Failures resulting from overloads occurred early in the program, mainly
from sampling frequencies that were too long.  Shorter sampling frequencies
solved most of the overload failures.  The occasional overload failures
occurring after the shorter sampling frequencies were mainly caused by
farmers harvesting grains or plowing nearby fields.  A very few overload
failures were actually caused by unusually heavy particulate loading re-
sulting from high pollution activity after the sampling frequencies were
shortened.
                                     10

-------
3.3  DICHOTOMOUS SAMPLING NETWORK MASTER LOG
     A Dichotomous Sampling Network Master Log was developed which gave, at a
glance, the past history of each sampler and the anticipated operation in the
future.  By knowing the future of each sampler, tray changes and optimum
maintenance routes could be planned in advance of each day's operation (see
Table 4).  The Dichotomous Sampling Network Master Log was maintained on a
daily basis and provides at a glance, the status of each sampler and expected
positions of each filter.  During each station visit, the Dichotomous Sampler
Check Sheet (see Appendix B) was filled out, and upon return to the main
office, the information was transferred to the Dichotomous Sampling Network
Master Log.  A completed log will show periods of uninterrupted data, inter-
rupted data and causes for failure, tray identification and period of expo-
sure time.  Table 4 represents a period of actual sampling history for Julian
day 161 through 167, and it shows that station 103 for this period had an
uninterrupted train of samples as represented by the continuous line through
the good section.  The number 21 on Julian day 161 shows that the station was
visited and number 21, the expected filter position, was in its correct
place.  There was a tray change on Julian day 163 before 1200 hours.  The
dichotomous sampler is constructed so that a tray may be changed without
interrupting the sampling filter.  The first filter position in an unexposed
tray is always left empty.  When the last filter in a sampled tray is reach-
ed the filter pusher may be removed, the old tray lifted out and the new tray
installed.  The last filter in the old tray can now be withdrawn into the
empty space in the new tray, thus the last filter in the old tray now becomes
the first filter in the new tray.  The tray change also identifies the tray
as 2060.  The numeral 2 identifies the sampler serial number and 60 the
number of sampled trays that this sampler has exposed since the network was
established.  Station 106 shows an uninterrupted train of samples with a
station visit on Julian day 161.  Station 108 shows a station visit on Julian
day 161, and the sampling tray had an interruption as shown by the broken
line on Julian day 163 with a filter overload on filter number 18 after 11:07
elapsed time.  The sampler was synchronized on Julian day 166 with filter
number 19.  Station 120 shows an interruption on Julian day 166 on filter
number 8 when there was an a.c. power failure at the station.  Filter number
                                     11

-------
                               TABLE  4.   DICHOTOMOUS SAMPLING  NETWORK MASTER LOG
RAMS
JULIAN
DAY
1976

161

162
163 20*°


164
165


166

167
103
O <£
0 0
O _j
Q CD «:
O UJ
0 O >
CD Z o

M

••
mm


••
•i


M

«

•2

l»
••
•A

m





-1

1


29








r

3061












105
o §
8 3
Q CD a:
O LU
0 0 >
CD z O

«•

••
«•


•
•




•

•>
•2
m*
i*
-

•



M



15


























106
o
a <:
8 3
Q CO OC
O LU
0 O =»
CD Z O



••
«•


«
^


^

^
10


•



V
•


»

m




























108
a
C^ ^f
0 0
O _J
O CO tt
BLU
0 5»
CO Z O

«•

M
•i

-\

mi
m+


—
••»


•I

•
•1

•

5


JCV
11:
E





9






18
,VOf
37
.S



son




112
Q
Q «£
O O
0 S «
O LU
O O >•
C3 Z O

••

•i



^H
^


•

M
2<


^
••


V
»


•^

•t










&

















115
0 CO a!
O LU
0 0 >
CO Z o



••
••


M
«


•

ml


»
!•


IW
m


m

m
-CV
10
LL










27
14
VOR
27
1

8022









118
O
Q 
O z: o

*•

••
Mi


«•
^


•»

^
•5
•


V
V
f

r
10
T



"'
,




1
a,c
id
FOW










K)27

ure


1 22
0 '§
8 3
Q CD (^
O LU
O O >•
CD z: . o

M

M
•


M
M


«

«•

•»

•i
•5
I»


»
•

J
•"i

m



29
6





6


4














1 24
0
Q 
CD 2E O

••

«•
•••


«MI
^


•i

•

•

w
7
^


mt
m


»

»







10




















ro

-------
9 failed on withdraw at midnight causing another interruption as indicated by
the broken line.  The filter numbers displayed represent actual station
visits and can be used as direct evidence to prove that the filters may be
verified as valid filters.  Two filter trays are sampled simultaneously in the
Automatic Dichotomous Air Sampler (ADAS).  One filter tray samples for the
small particles and the other tray samples for the large particles.  One
vertical line represents both trays because a failure in either sample stream
will cause the sampler to shutdown and signal a failure back at the central
computer facility.  The crosslines on the vertical line indicate sampled
filters and when the crosslines are counted for one Julian date the sampling
frequency of that sampler can be determined.
     All of the accumulated check sheets, filter tray record sheets and the
master log sheets will be turned over to EPA.  Copies of the sheets and
master log are on file at LBL for use in validating filters for analyses.
When a number of samplers are operated in a network, the Dichotomous Sampling
Network Master Log is an invaluable aid in validating filters and determining
sampler operation.
                                     13

-------
                            4.0  RECOMMENDATIONS

     The sampler was designed to either operate under control of a central
computer facility or from its own internal system.  When synchronization
problems arose, as experienced in the RAMS network, manual synchronization
became necessary requiring operating personnel to actually be present at the
sampling station.  When a number of these samplers are active in a network it
would be advantageous to have a countdown circuit to activate the sampler
automatically.  Overloads cause the sampler to shutdown and hold the elapsed
time on the clock until the sampler is again manually synchronized.  A
circuit to store the elapsed time and automatically restart the sampler on
the next scheduled filter would cause the sampler to maintain synchroniza-
tion.
                                     14

-------
                   5.0  STREAKER ANU RADIOMETER OPERATION

     The Jensen-Nelson streakers and the silicon cell  pyronometer originally
operated under Task Order No. 117.  Task Order No.  117 had a period of perfor-
mance commencing on 1  July 1976 and terminating on  30  November 1976.  Con-
tinued maintenance of the Jensen-Nelson streakers and  the silicon cell
pyronometer were incorporated in Change 3 to Task Order No.  102.   The follow-
ing paragraphs provide a summary of the effort expended toward streaker and
radiometer installation and maintenance.
     The two Jensen-Nelson streakers were installed, one each on  the towers
at the 10 meter level, at stations 120 and 124 in the  spring of 1976 by
representatives from Florida State University.  The streaker pumps were
Sierra constant flow pumps located on the roof of the  shelters.  Power and
vacuum lines were already in place, being left over from the 1975 summer
intensive study.  Filter frames were supplied by Mr. W. W. Berg,  Jr. of
Florida State University.  Exposed filters were shipped back to FSU for
analyses.
     One streaker motor was replaced at station 120 when it started showing
signs of intermittent operation.  The streaker samplers operated  almost
trouble free; however, the constant flow pumps exhibited early signs of
trouble.  The flow meter in the pump at station 120 had to be replaced.
Many of the electronic parts had to be replaced because of severe damage
caused by corrosion at station 120.  The pump failed at station 124 and
because of the inability to get replacement parts was  shipped back to FSU.
A heavy-duty Gast pump was used as a replacement and gave trouble free
operation throughout the rest of the sampling program.
     On 14 February 1977 RAMS station 124 was deactivated and the streaker
and all associated equipment was removed and transported to RAMS  station
125, where it was installed at the 10 meter level on the tower.  The
streaker was operable on 15 February 1977.

                                     15

-------
     The streaker filtering system ran through 8 April 1977 at which time
the system was shutdown and the equipment dismantled and packed for shipment.
After the equipment was packed it was turned over to the EPA Task Coordi-
nator.
     The silicon cell pyranometer was installed on the roof of RAMS station
124.  The sampling system consisted of the sensor and a strip chart recorder.
The sensor operated trouble free throughout the life of the sampling period
and required very little maintenance.  The maintenance required by the
sensor consisted of cleaning the diffuser and insuring that the strip above
the sensor was rotating.  The inking system on the strip chart recorder
gave considerable trouble.  The inking pen exhibited a tendency to easily
become clogged.  However, a new type of ink and chart paper apparently
solved the problem.  It was also noted that the sensitivity of the strip
chart recorder was probably too high, and the resulting jittery action of
the inking pen loosened fibers out of the chart paper and forced them up
into the tip of the inking pen.  The completed radiometer charts were
shipped to the Argonne National Laboratory.
     The pyranometer and strip chart recorder were removed from station 124
on 14 February 1977 and installed at station 125 on 15 February 1977, at
which date it became operable.  The system operated without incident until
8 April 1977 and on this date the system was shutdown, the equipment
removed and packed for shipment.  The pyranometer, strip chart recorder and
all unused supplies were turned over to the EPA Task Coordinator.
                                     16

-------
    APPENDIX A



SAMPLING EQUIPMENT
        17

-------
                                 CONTENTS

                                                                     Page
1.0   Sampling  Equipment                                               20
      1.1   Virtual  Impactor                                            20
      1.2   Flow Controller                                             22
      1.3   Filters                                                     22
      1.4   Sample Changer                                              22
      1.5   Flow Monitor                                               23
      1.6   Electronic  Controller                                       23
                                     18

-------
                                  FIGURES
Number                                                                 Page
  1       LBL Automatic Dichotomous Air Sampler with Dustcover           25
           in Place
  2      ADAS with Dustcover Removed                                    26
  3      Schematic of the Automatic Dichotomous Air Sampler             27
  4      Schematic of a Two-Stage Dichotomous Virtual  Impactor          28
  5      Details of the Virtual Impactor                                29
  6      Disassembled Components of the Virtual Impactor                30
  7      Flow Controller Instrument Panel                               31
  8      Slides and Slide Trays Used in ADAS                            32
  9      The Functional Time Sequence of a Sampling Period              33
                                     19

-------
                         1.0  SAMPLING EQUIPMENT (1)

     The LBL Automatic Dichotomous Air Sampler (ADAS) is designed to sample
ambient aerosols at a rate of 50 1/min and divide the incoming particles
into two aerodynamic size fractions; greater than 2.0 microns and less than
2.0 microns for unit density spheres.  The particles are deposited separately
on two 37 mm diameter cellulose membrane type filters for subsequent total
mass measurements via beta-gauge and elemental analysis via X-ray fluores-
cence analysis.  The effective area of the deposits oh the filters is
                  2
approximately 7 cm .  The inlet adapter is designed to sample isokinetically
from a vertically downward flow of 217 cm/sec.  Automatic control functions
in the sampler unit allow unattended operation for up to 36 sample exposures
for preset time intervals from 1 to 100 hours, Figures 1 and 2.

1.1  VIRTUAL IMPACTOR
     The size fractionation is accomplished by a two-stage virtual impactor
preceding the membrane filter collectors.  The schematic of the complete
virtual impactor design is shown in Figure 3.  The size separation is accom-
plished aerodynamically in a manner similar to that used in more conventional
impactors, with the exception that the streamlines of the flow are main-
tained in part without the use of a mechanical impaction surface.  Instead,
an open tube opposes the inlet orifice permitting the large particles to be
"impacted" into a low flow region within the tube from which they are
subsequently collected on the membrane filter.  The small particles follow

(1) Much of the information contained in this section was excerpted and
paraphrased from the operation and maintenance manual supplied with the
ADAS and publication LBL-3854, "Dichotomous Virtual Impactors for Large
Scale Monitoring of Airborne Particulate Matter", B. W. Loo, J. M. Jaklevic,
and F. S. Goulding, 1975.
                                    20

-------
the bulk of the flow around the tube and are collected by a separate
filter.  Since optimum operation requires that approximately 20 to 25% of
the small particles are drawn out of the flow with the larger particles,
the completed unit consists of two separation stages.  Thus, the idealized
output of the unit would be 100% of the large particles (greater than 2.0
microns) collected on filter A, together with 5% of the small particles,
and 95% of these small particles (less than 2.0 microns) collected on
filter B.  A cross section of the virtual impactor is shown in Figures 4,
and 5.  Air is drawn through the inlet jets (part 1) in parallel.  Their
protrusion into the first stage cavity is necessary to eliminate the
"backwall" losses on part 2 due to the spatial oscillation of streamlines
as found in the Environmental Research Corporation design, as well as in
some conventional impactors.  Part 3 forms the first stage cavity.  The
three small holes in this part are symmetrically located about the three
central axis but are offset 60° azimuthally with respect to the coarse
particle receiving tubes (parts 4) to minimize flow interference.  These
holes, in combination with the one in part 7, also govern the internal flow
distribution.  The Q-i/Qn for the first stage was adjusted to be 25% to
minimize wall losses in the cavity in Figure 4.  The tapered lips on the
tubes have no significant effect on the cut point although they do tend to
defocus the streamlines and reduce cavity losses.
     The three coarse particle jets are then converged by a 15° cone (part
6) onto the second stage of separation after passing through the drift tube
(part 5) Figure 5.  Parts 8 are three positioning rods which form an open
cavity for the second stage jets.  The ratio Q-j/Q0 (Figure 4) here is
chosen to be 20%  Thus, 2.5 1/min of air will pass through filter A carrying
all the coarse particles, along with 5% of the fine particles.  The fine
particle stream of the second stage will merge with that from the first
stage and be deposited on filter B.  Thus, 70% of the fine particles are
drawn from the stream in the first stage of separation and 25% in the second
stage.  The large particle filter will contain all of the large particles
along with 5% of the small particles.  In analyses, a correction for the 5%
contamination of filter A can be made based on the amount of the uncontam-
inated 95% of the fine particles on filter B.  The overall construction
                                     21

-------
utilizes all stainless steel .parts (excepting part 6) for mechanical
integrity and corrosion resistance, compression 0-ring seals and tie rods
(part 13) with thumb nuts (part 12) for easy disassembly.  Stragetic
corners are shaped to minimize losses.  Figure 6 shows the virtual impactor
disassembled.

1.2  FLOW CONTROLLER
     Flow regulation is essential  for precise measurement of the air volume
sampled and the maintenance of a fixed particle-size cut point.  This is
accomplished as shown in Figure 5 by sensing (through part 15) the pressure
differential P , Figure 3, between the inlet and the second stage of the
impactor with a diaphragm operated null switch (Dwyer Model 1640-5) which
in turn causes the opening in a motor driven valve to be increased or
decreased to maintain the preset null condition.  The valve is simply a 5.1
mm diameter orifice pierced by a traveling micrometer shaft with a 2°
taper.  A fixed orifice limits the flow through filter A to 2.5 1/min.  The
variable orifice and the null  switch thus form a feedback loop to compensate
any impedance change in filter B.   The carbon vane vacuum pump used (Gast
Model 0522-103-G18D) has adequate pumping power to overcome an increase of
                                                                   o
about 70% in impedance from a typical initial value of 26.2 torr-cm /1/min
(1.2 micron cellulose membrane filter manufactured by Nuclepore Corporation).
Figure 7 shows the flow controller instrument panel.

1.3  FILTERS
     The filters used are 37 mm discs of cellulose membrane filters supplied
and mounted in 5.1 cm x 5.1 cm plastic frames by the Nuclepore Corporation.
Up to 36 of these slides are carried in a linear array standard 35 mm slide
projector cartridges (Argus Camera).  Figure 8 shows such a pair of car-
tridges containing the digitally labeled filter holders.

1.4  SAMPLE CHANGER
     The function of the slide changer is to extract a matched pair of
filters from side by side slide trays corresponding to the A and B filter
stacks, Figure 3.  A horizontal shuttle manipulates the  slides into their
sampling positions where they are clamped in the output  tubes of the

                                     22

-------
virtual impactor.  Upon the completion of the sampling interval, they are
undamped and withdrawn back into the slide trays.  The over-travel of the
shuttle actuates a "Geneva wheel" which advances the stack by one vertical
increment to be ready for the next insertion.  A single motor drives the
shuttle which performs the function of transporting, clamping and unclamping
slides, together with advancing the trays with a single forward and return
stroke.

1.5  FLOW MONITOR
     Several types of out-of-range conditions in the flow circuit are
detected and indicated by the system.  Excessive travel of the micrometer
valve due to the presence of leaks or broken filters causes out-of-range
switches to be activated.  An auxiliary pressure sensing snap switch PI
(Figure 3) is used to detect improper clamping or a broken filter.  Since
the vacuum needed for the fixed limiting orifice results from the proper
flow condition in the fine particle stream, PI actually monitors the
conditions at filter B even before the micrometer valve reaches its limits.

1.6  ELECTRONIC CONTROLLER
     The selection of sampling intervals, execution of the sequential
steps, regulation of flow, detection of errors, monitoring and display of
the system status and communication to an optional remote computer are
performed via the control module shown in Figure 7.
     In order to maintain the synchronism of the samplers with the clock,
ten seconds are allowed for a sample insertion or withdrawal cycle, which
normally requires about seven seconds, to complete.  Figure 9 illustrates
the time sequence of a typical sampling period.  While the vacuum is turned
on continuously, actual sampling starts at the twenty second mark when the
solenoid valve is opened.  Another ten seconds are allowed for steady flow
conditions to be established before the flow controller is activated.
The right hand column of the figure indicates the sequence in which error
conditions are checked.  The maximum times allowed to complete a sample
transport and flow adjustment are ten seconds and twelve minutes respec-
tively.
                                     23

-------
     To ensure synchronization in the event of short a.c. power failures
(less than 10 min.) the elapsed time clock and logic control circuits are
automatically switched to a rechargeable battery.
                                     24

-------
                                                          '•
FIGURE 1.   LBL AUTOMATIC DICHOTOMOUS AIR SAMPLER WITH DUSTCOVER IN PLACE
                                  25

-------

FIGURE 2.   ADAS WITH DUSTCOVER REMOVED
                26

-------
        INLET


      w
          U
         A     ^
        ____  w

        FA  I  |
                <
F'el
                U
           FIXED
        yORIFICE
                   tu
                   U

                   <
                   >
                   Q
                   <

                   X
                   o
                   <

                   w
                                      <
                                      U
                                      HORIZONTAL SHUTTLE

                                SOLENOID
                                 VALVE
                                                   XBL7411-8540
FIGURE 3.  SCHEMATIC OF THE AUTOMATIC DICHOTOMOUS AIR SAMPLER
                          27

-------
                                  so l/m
               FILTER A
                                                       FILTER B
                                                       '(<2tim)
                  FLOW
                  LIMITING
                  ORIFICE
                                   TO PUMP
FIGURE  4.   SCHEMATIC  OF A TWO-STAGE  DICHOTOMOUS VIRTUAL IMPACTOR
                                   28

-------
     INTAKE
     (50 l/m)
FIGURE 5.  DETAILS  OF THE VIRTUAL  IMPACTOR
                   29

-------
                                               HREE HOLES
   VIRTUAL
   IMPACTOR
  •MOUSING
FIRST
STAGE
                                                          DRIFT!
                                                          TUBE
 BOTTOM SECTION
 OF IMPACTOR
 HOUSING
                                                         THREE
                                                   POSITIONING.
                                                          RODS
                                                     .^SECOND
                                                          TAGE
FIGURE 6.   DISASSEMBLED COMPONENTS OF THE VIRTUAL IMPACTOR
                          30

-------
                                                           21X1461 PI
         LBL  AUTOMATIC:SAMPLER CONTROL
     FAILURE






  0) AC POWER




  41 WITHDRAW




  • INSERT     ,?




  £ NO SAMPLE




     ROW CONTROL VALVE




     VALVE OUT OF RANGE
-
 I  STATUS   1





£  STANDBY.




0  INSERTING




    CLAMPED




£  VACUUM ON




^  FLOW CONTROL ON




  Q INCREASE FLOW




  Q DECREASE FLOW




    WITHDRAWING




    LAST SAMPLE
                                                "I    {..„,
                                            I 	 J    \
        I  ABORT


             .
                                            D  t: D
ELAPSED TIME
        SAMPLING

        HOURS

        PRESET
     POWER' '
        FIGURE 7.   FLOW CONTROLLER  INSTRUMENT PANEL
                              31

-------
FIGURE 8.   SLIDES AND SLIDE TRAYS USED IN ADAS
                     32

-------
 START
O
ui
UJ
O
UJ
(O
ex.
UI
     10
     20
     30
        WITHDRAW
         SAMPLES
           — UNCLAMP

           — CLAMP
           _ RAISE
             STACKS
            •STANDBY
          INSERT
         SAMPLES
           -UNCLAMP

           -CLAMP
— OPEN SOLENOID VALVE
          FLOW
         ADJUST
             ENABLE
             FLOW
             CONTROLLER
       I	
       I  FLOW    I
        __ADJUSJ._ J

       I
       I
END OF I CLOSE SOLENOID VALVE
PERIOD '	
                                -LAST SAMPLE

                                -WITHDRAWN
                                  K10 SEC.)
                         — INSERTED
                           «10 SEC.)
-PROPER VAC SEAL
  FLOW LIMITS
  EXCEEDED

 FLOW ADJUSTMENT
 «12  MIN. EACH)
 FIGURE 9.  THE  FUNCTIONAL TIME SEQUENCE OF A SAMPLING PERIOD
                          33

-------
    APPENDIX B



QUALITY ASSURANCE
          34

-------
                                  CONTENTS
                                                                       Page
1.0   Introduction                                                      37
2.0   On Site Quality Assurance                                         38
      2.1  Status Checks                                                38
      2.2  Preventative Maintenance                                     38
           2.2.1  Monthly                                               38
           2.2.2  Tri-Monthly                                           41
           2.2.3  Yearly                                                41
      2.3  Filter Validation                                            41
                                      35

-------
                                  FIGURES
Number                                                                 Page
  1       Filter Record 1                                                 39
  2      Check Sheet"!                                                  40
                                      36

-------
                            1.0  INTRODUCTION

     The quality assurance plan described the quality assurance activities
planned by AMC personnel  in support of the operation and maintenance of the
Lawrence Berkeley Laboratories (LBL) Dichotomous Sampling Units.   Under a
previous task order the AMC installed the samplers in the RAMS shelters and
operated them for approximately six months.  Under this task order the AMC
continued the operation and routine maintenance activity.
     Since LBL, unlike other vendors, assumed an active role in this work,
a brief description of the relative roles of the AMC and LBL is appropriate.
LBL supplied AMC with slide cartridges filled with beta gauged and serial-
ized filters.  The AMC processed these filters according to the procedures
described in this plan.  The slides and relevant data concerning their
exposure were then returned to LBL for analysis.  AMC personnel provided
preventative maintenance and normal corrective maintenance.  Catastrophic or
extraordinary failures were corrected by using parts from a spare sampling
unit and returning the malfunctioning parts to LBL.
                                     37

-------
                       2.0  ON SITE QUALITY ASSURANCE

     The following sections describe the procedures employed by AMC per-
sonnel in the treatment of filters and the operation and maintenance of
the sampling units to assure the quality of the dichotomous data.

2.1  STATUS CHECKS
     The RAMS data acquisition and control system was used for surveillance
of the sampling network to detect sampler malfunction.  A computer operator
prepared a check list three times daily indicating the operating condition
of each sampler.  This list indicated whether a sampler had failed or was
operating normally.  A master check sheet was kept at the main office indi-
cating the operating condition of each sampler, the number of the slide
tray and the estimated slide position, valid and invalid slides, and cause
of invalidation.  The information contained in the master check sheet was
obtained from check sheets filled out when each site was visited,  and a
visual inspection was made by the visiting personnel.  When these check
sheets were returned, they were checked and the data recorded on the master
check sheet.  The master check sheet, over a given period of time, shows the
operating history of all samplers in the sampling network, Filter Record 1
and Check Sheet 1 (Figure 1 and Figure 2).

2.2  PREVENTATIVE MAINTENANCE
     Normal maintenance of the system consists mainly of removing the
accumulation of dust from the virtual impactor.  A maintenance log was kept
at each sampler location and appropriate entries were made indicating
sampler malfunction and corrective measures.

2.2.1  Monthly
     At approximately 30 day intervals the virtual impactor was disassembled
and cleaned with a washing of ethyl alchol and de-ionized water and wiped
                                     38

-------
   FILTER RECORD
YEAR 75
                       SAMPLING  SITE  115
TRAY
Slide
Position
'












1
2
3
• 4
5
6
7
8
9
10
11
12
13
=, 14
n 15
5 16










1
17
18
19
20
21
. 22
23
24
25
26
27
INVALID 28
i







=
29
30
31
32
33
34
35
36
3
_i
4
>
A 07011
Serial
Number
05634
06452

























06478
J6479






06486

B 07011
Serial
Number
55634
56452

























b0478
56479






16486

Start
(x)
X
X


























A








Stop
(x)



























X







X

Time
(Hour)
1200
1800

























1041
1200








Period
(Hour)
06
06

























06
06








Date
(Julian Uay)
218
218
219



22C



221



222



223



224



275
225

-------
       CHECK  SHEET
0 = START
X = STOP
                                              =  CONTINUE
                                            *  =  FAILED
PAGE  29U
YEAR 7!>
J. Day 2681 Hrs. 0715
RUN STATUS
SAMPLING PERIOD
TRAY
POSITION
TRAY NUMBER
EXPECT.
OBSERV.

103
y
06
25

02029
105
y
06
17

03028
106
y
12
18

04014
108
X
12
1

05019
112
y
12
5

06020
115
y
12
9

07015
118
y
12
13

08011
120
V
12
34

09013
122
*
12
?

10012
124
y
12
2

11013
 NOTES: Status Check
 SITE  108-Instrument Off-Bullet Wound!
 SITE  122-Cause Unknown
J. Dav 268
Hrs. 1150
RUN STATUS
SAMPLING PERIOD
TRAY
POSITION
EXPECT.
OBSERV.
TRAY NUMBER
103





105





106





108





112



,

115





118





120





122
*
12
?
30
10012
124





 NOTES: Withdrawn, no failure indication
       probably station power failure at  00:23 hours Julian 268
J. Day268|Hrs. 1200
RUN STATUS

SAMPLING PERIOD
TRAY
POSITION
TRAY NUMBER
EXPECT.
OBSERV.

103





105





106





108





J12





115





118





120





122
Y/
12
1
1
10013
124





NOTES: Change tray in sync.
J. Day268
Hrs. 1200
RUN STATUS
SAMPLING PERIOD
TRAY
POSITION
EXPECT.
OBSERV.
TRAY NUMBER
103





105





106





108
V
12
2
2
05019
112





115





118





i?n





1??





124





NOTES:  Unit repaired.  Shafts on two inside rollers of the knife edge assembly  bent causing improper
       seal.

                                FIGURE 2.   CHECK SHEET 1
                                               40

-------
dry with Kimwipes.  The jets were cleaned with a washing of ethyl alcohol
and de-ionized water with Q-Tips.
     0-rings on the sliding seals on the inlet pipe were greased with Dow
Corning 33 silicone grease as well as were the two 0-rings which make the
top seal between the sample and the impactor.
     Knife edge clamps and screens were cleaned with alcohol and wiped dry
with Kimwipes.  Dry accumulations of dust were also removed from other parts
of the sampler changer with alcohol, water and Kimwipes.
     While the impactor was disassembled a visual inspection was made and
preventative maintenance was performed if necessary.

2.2.2  Tri-Monthly
     All cooling fan hub bearings that are capable of being lubricated were
lubricated with light oil as well as sleeve bearings on shuttle and stack
loader assemblies.  The limiting flow orifice was washed with alcohol and
water.
     The sampler was vacuumed and the slipping clutches adjusted as re-
quired.  The output filter on the vacuum pump was cleaned and replaced
when necessary.
     Flow calibrations were checked with the flow meters that had been cal-
ibrated at Lawrence Berkeley Labs, namely the total flow at the inlet and
the flow through the limiting orifice.  The vacuum pressure of the pump
was checked and the vacuum pump vanes replaced when pressure fell below
25" Hg.  The calibrations were extremely stable.

2.2.3  Yearly
     Flexible hose was replaced as well as all worn 0-rings.  The sensor
pressure was checked at 5 cm water column.

2.3  FILTER VALIDATION
     All filters were visually inspected to determine if they were in the
proper sequence and were not cracked or broken before being placed in the
sampler.  Located at each sampler site was a notebook for logging sampler

                                     41

-------
malfunctions and corrective action, as well  as a filter tray record sheet
indicating the sampling site number, tray number filter position, filter
start and stop time, sampling time, Julian day and a section for notes
detailing periods when a filter had failed to sample correctly.   When a
paired tray set had finished its sample run the filters were validated by
checking the log sheets filled out on each visit to the site, verifying the
Julian day start and ending dates and the master check sheet.  All valid
and invalid filters were indicated and when a suitable number of paired
trays were collected the filter tray record sheet, site visitation log sheet
and filters were shipped to LBL for mass determination and analysis, Filter
Record 1.  A copy of all log sheets and filter records were kept at the
Rockwell International Air Monitoring Center in St. Louis.  Dr.  B. W. Loo
was contacted by telephone for help in solving difficult problems.
                                     42

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
   EPA-600/4-79-024
                             2.
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE

 REGIONAL AIR POLLUTION STUDY
 Dichotomous Aerosol Sampling System
             5. REPORT DATE
                April  1979
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

 E.  Nelson
             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	
             11. CONTRACT/GRANT NO.

              68-02-2093
              Task Order  102
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
      Ten of twenty-five stations  making up the Regional Air Monitoring System (RAMS)
 in St.  Louis were equipped with dichotomous samplers and high  volume filter samplers
 for aerosol measurements.  The dichotomous samplers, built by  Lawrence Berkeley Labo-
 ratories (LBL), were designed for automatic operation and were capable of collecting
 up to 36 samples in each of  two size fractions before filter stacks  must be changed.
 Most of the time, the samplers operated to collect 12-hour samples  (0000-1200  and
 1200-2400), except that at two stations with high aerosol loadings,  the sampling
 intervals were split into two 6-hour intervals.  Sample filters were pre-weighed by
 beta gauging before exposure in St.  Louis and subsequently returned  to LBL for deter-
 mination of aerosol mass in  each  size fraction, as well as determination by x-ray
 flourescence of the concentrations of the following elements:   Al,  Si, P, Cl, K, Cr,
 Mn, Ga, Rb, Sr, Sn, Sb, Ba,  Hg, S, Ca, Ti, V, Fe, Ni, Co, Zn,  As,  Se, Br, Cd, and Pb.
 Approximately 33,000 samples were collected between March 1975 and  March 1977.
 Analytic data are stored in  the RAPS Data Bank at Research Triangle  Park, N.C.

      The operation of the samplers in the RAMS network is described  along with
 problems encountered and procedures used for preventive maintenance  and quality
 control.  Also described are two  streaker samplers specially installed for continuous
 aerosol measurements and a silicon cell pyranometer.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                           c. COSATI Field/Group
 *Air pollution
 *Aerosols
 *Particle size distribution
 *Sampling
 *Chemical analysis
                            13B
                            07D
                            14B
18. DISTRIBUTION STATEMENT
 RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
 UNCLASSIFIED
                           21. NO. OF PAGES
49
                                              20. SECURITY CLASS (Thispage)

                                               UNCLASSIFIED
                                                                         22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE
                                             43

-------