v>EPA
United States
Environmental Protection
Agency
Environmental Monitoring and Support EPA-600/4-78-025
Laboratory Way 1978
Research Triangle Park NC 27711
Research and Development
Monitoring
for Polychlorinated
Biphenyl Emissions
From an Electrolytic
Capacitor Disposal
Project
-------
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 PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
MONITORING FOR POLYCHLORINATED BIPHENYL EMISSIONS
FROM AN ELECTROLYTIC CAPACITOR DISPOSAL PROJECT
by
Charles E. Rodes
Environmental Monitoring Branch
Environmental Monitoring and Support Laboratory
and
Merrill D. Jackson and Robert G. Lewis
Environmental Toxicology Division
Health Effects Research Laboratory
ENVIRONMENTAL MONITORING AND SUPPORT 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 Monitoring and Support
Laboratory and the Health Effects Research Laboratory, U.S. Environmental
Protection Agency, and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for
use.
ii
-------
ABSTRACT
Three different air sampling methods were used simultaneously to monitor
for polychlorinated biphenyl (PCB) emissions arising from a pilot disposal
project involving electrolytic capacitors. Analytical results indicated that
the primary PCB material was Aroclor 1242, and that airborne concentrations
inside the building housing the grinders exceeded 5 mg/m3. The PCB air con-
centrations outside the building at a distance of 9 m were typically <1.0 pg/m3.
Measurements made with one low-volume and two high-volume air samplers are
compared.
iii
-------
ACKNOWLEDGMENTS
The authors wish to thank Paul J. Haskins, Willie T. McLeod, and Everett
Quesnell for their valuable technical assistance.
-------
SECTION 1
INTRODUCTION
In late March 1976, the U.S. Environmental Protection Agency (EPA)
undertook a monitoring project to determine the extent of air pollution asso-
ciated with a pilot-scale disposal of electrolytic capacitors in El Dorado,
Arkansas. The proposed full scale disposal operation was to have consisted
of pulverizing, then incinerating electrolytic capacitors. It was suspected
that, during the grinding operation, polychlorinated biphenyls (PCBs)
would be emitted, which could pose a potential health hazard. The capacitors
were known to contain an Aroclor dielectric liquid, although the exact mixture
was not identified.
At the inception of the study, several air sampling methodologies for
PCBs were known to be in the developmental stages. However, no method had
been evaluated sufficiently to suggest its exclusive selection for use. It
was decided, therefore, to employ three different methods simultaneously to
provide corroborating data. It was also felt that valuable comparisons could
be derived from this approach.
-------
SECTION 2
SAMPLING METHODOLOGY
The three methods selected were: (a) a modified high volume (Hi-Vol)
sampler under evaluation by the Office of Toxic Substances (OTS) of EPA (1);
(b) a modification of the Hi-Vol sampler designed for EPA by Syracuse Univer-
sity Research Corporation (SURC) to monitor airborne pesticides (2); and, (c)
the low-volume pesticide air sampler marketed by Micro-chemical Specialties
Company (MISCO), Berkeley, California. The latter device has been widely
used by EPA and others to sample ambient air for pesticides and PCBs (3).
METHOD I
The OTS sampling method for PCBs is based on polyurethane foam collection
but uses a standard Hi-Vol head with an extended throat (Figure 1). The
sampler is operated as a standard Hi-Vol with the addition of a plug of foam
90 mm (3.5 in.) in diameter and 76 mm (3 in.) in length which is placed into
the 15-cm (6 in.) throat extension. A standard 20.4- by 25.5-cm (8- x 10-in.)
glass fiber filter is used as a prefilter. The addition of the foam plug
reduces the normal flow rate to 0.70-0.85 m3/ndn (25-30 ft3/min). No mention
is made in the OTS procedure (1) concerning changing of the inlet opening to
provide dynamic similarity with the standard Hi-Vol.
METHOD II
The SURC sampler is a substantially modified Hi-Vol sampler utilizing a
special glass collection module attached to a standard Hi-Vol vacuum blower.
The glass module is assembled from commercially available process glass pipe
and is basically a 10- by 5-cm reducing section with stainless steel screens
-------
POLYURETHANE FOAM
PLUG LOCATION
Figure 1. Diagram of OTS sampler (Method I)
-------
to support a circular glass fiber filter at the intake and a vapor trap in
the smaller chamber. The SURC sampler was modified for this study and is
shown schematically in Figure 2. The intake air is first filtered through a
glass fiber filter 115 mm (4.5 in.) in diameter. The air is then filtered
through a 58 mm (2.25 in.) diameter by 76 mm (3 in.) thick plug of polyether-
type polyurethane foam (average density 0.02 g/cm3). The sampling rate is
0.20-0.22 m3/min (7-3 ft3 min). The modified SURC samplers used for this study
have an annulus inlet space (around the shelter roof) of 0.5 cm (0.2 in.) to
provide an inlet velocity of 0.6 m/s (2 ft/s) consistent with the Federal
Register (4) specifications for a standard Hi-Vol, except that the shelter
(General Metals GMW-65000) is specially designed for ease in sample changing.
The sampler has a Dixon pressure recorder to monitor flow performance con-
tinuously and a venturi meter to provide a more accurate flow measurement.
These samplers are not presently available from commercial sources. The cost
of materials for construction is approximately $600.
METHOD III
The general pesticide air sampler designed for the U.S. Public Health
Service (3) is available commercially as the MISCO Model 88 sampler. This
sampler is essentially a low flow rate (0.017-0.020 m /min) impinger system
with a prefilter. The flow diagram of this sampler is shown in Figure 3.
The impingers are standard 500-ml Greenburg-Smith impingers, and the pre-
filter is a 78 mm (3 1/16 in.) diameter glass fiber filter. The absorbing
medium is 100 ml of preextracted high-purity ethylene glycol. The sampler
is designed for gaseous or fine particulate sampling only, since no attempt is
made to size-fraction the aerosols at the inlet, or to sample isokinetically.
This inlet arrangement, however, provides a means for attaching a short line
for sampling air in a specific area. The MISCO sampler is capable of operating
four separate impingers in sequence by means of a built-in timer if desired.
Parallel operation of impingers is also possible but at lower flow rates.
-------
INLET
CALIBRATION
VENTURI
OUTLET
AIR
FLOW
GLASS FIBER
FILTER
1
MANOMETER
(4 in.-0-4 in. Hg)
SURC
SAMPLING
P HEAD
FOAM PLUG
MAGNEHELIC
GAGE, (0-100 in.)
TEST VALVE
(NORMALLY CLOSED)
NOTE: FOR NORMAL OPERATION, A 15 cm. i.d. BY 3 meter FLEXIBLE HOSE IS
ATTACHED TO THE BOTTOM OF THE HI VOL WITH A BAND CLAMP.
Figure 2. Schematic diagram of modified SURC sampler (Method II)
-------
AIR INLET
EXHAUST
PREFILTER
(FIBER GLASS)
FLOW METER
MOISTURE TRAP
(GLASS WOOL)
CONTROL
VALVE
RUBBER STOPPER SOLENOID
VALVE
IMPING ER
(ETHYLENEGLVCOL)
VACUUM PUMP
SEQUENCE CONTROLLER
NOTE: SAMPLER SS CAPABLE OF OPERATING FOUR BUBBLERS SEQUENTIALLY
Figure 3. Schematic diagram of MISCO sampler (Method III)
-------
Quality assurance measures employed during the sampling program were
concerned primarily with flow measurement since this is the critical parameter
in integrated sampler performance. All samplers were calibrated in the labo-
ratory and referenced to a calibrated Rootsmeter. After setup in the field,
audit devices (calibrated against the lab standard) were utilized to check
the sampler calibrations and for recalibration if necessary. A calibration
plot for the modified SURC sampler is shown in Figure 4. The SURC sampler
provided two independent measurements of flow rate, one of which was a flow
recorder to provide a permanent record. A programmable calculator/plotter
was used to provide an exponential regression fit with associated statistics
used to control the degree of scatter acceptable in a given calibration. After
the sampling was completed, all samplers were rechecked for possible changes
in flow calibration.
Potentially interfering background contaminants were removed from the
collection media by chemical extraction prior to deploying the samplers in
the field. The foam plugs were preextracted by successive 12-h Soxhlet ex-
traction with acetone and hexane (5). The ethylene glycol used was purified
by successive extractions with dichloromethane.
-------
CALIBRATION PLOT FOR MODIFIED SURC SAMPLER
10
30 40 50
MAGNEHELIC OR DICKSON READING
Figure 4. Calibration plot for modified SURC sampler.
-------
SECTION 3
FIELD STUDIES
The capacitor disposal operation was conducted by Pollution Controls,
Inc., El Dorado, Arkansas, under subcontract to TRW, Inc. TRW was under
contract to the EPA Solid Waste Program to gather data on a variety of solid
waste incineration methods. There were approximately 4500 kg (10,000 pounds)
of discarded capacitors to be pulverized and ultimately incinerated in this
test program. The pulverizing equipment consisted of two hammer-mills to
handle both large and small capacitors. The hammer-mills had collection boxes
into which the pulverized product was collected. After processing, special
plastic-lined shipping containers were provided for the storage and shipment
of the PCB-coated pieces.
The site location was an abandoned oil refinery in El Dorado. A detailed
map of the site area and location of the concrete pad on which the grinders
were located is shown in Figure 5. In order to study the prevailing winds,
background meteorological measurements (wind speed and direction) were made
from May 16 through May 18, 1976, to determine whether there was a prevailing
wind direction. The data for this period indicated a general wind direction
from the southeast. Because of the wind direction variability, however, it
was decided to orient the samplers according to the wind direction just prior
to grinding. During this background period, several sets of background samples
were collected using the three sampling methods.
After the samplers were set up in the arrangement shown in Figure 5, the
Arkansas Pollution Control Authority concluded that the grinders should be
enclosed. Accordingly, a 3.6- by 5.5-m (12- by 18-ft) temporary building was
constructed around the grinders. Enclosing the samplers resulted in two
-------
PCI OFFICES
•&
SITE
C
(NOT
USED)
| SITED ]
(NOT USED)
SMALL
SAMPLER GRINDER;
EXHAUSTS
TEMPORARY BUILDING
ON CONCRETE PAD
ill
SITES
METEROLOGICAL TOWER (8 ml
•3
SITEi2
• 1
PARKING
LOT
SAMPLER LEGEND:»1 OTS
•2 MISCO
•3 SURC
10
NORTH
I
I
SCALE, feet
20 30 40
I I . |_
50
5 10
SCALE, me ten
15
Figure 5. Sampling site in Eldorado, Arkansas.
-------
problems: (a) how to protect the grinder operators inside the building from
the expected higher local PCD emissions; and, (2) whether pseudoemission
measurements inside the building could be made with the air samplers. In order
to resolve the first problem, self-contained oxygen breathing canisters and
protective clothing were provided for the two operators. Two of each type of
sampler were placed inside the building. Special ducts were provided for the
modified SURC and OTS samplers to exhaust the samplers outside the building.
Because of the high total sampling rate of the six samplers (2.5 m /min), it
was estimated that the entire volume of air inside the building would be ex-
hausted every 30 min. In order to verify that a negative draft existed inside
the building, a smoke bomb was set off outside the building. Unfortunately,
an undesirable residue from the smoke test was deposited on the inside of
the samplers, which subsequently posed some problems in sample analysis.
The capacitors were ground in a 35-min test run, then for 5 h of full
scale operation. Because of the elevated temperature inside the building
during daylight hours, the grinding was carried out only at night. Since
the meteorological data (Table 1) during the majority of the grinding period
indicated calm wind speeds and absence of a predominant wind direction, only
background samplers at Sites A and B were operated.
TABLE 1. METEOROLOGICAL DATA DURING CAPACITOR GRINDING - 5/20/76
Time
Interval
0300-0400
0400-0500
0500-0600
0600-0700
0700-0800
0800-0900
Wind
Direction
-
-
-
-
50°
130°
Wind
Speed , mph
calm
calm
calm
calm
1.0
2.0
Relative
Humidity, %
100
100
100
100
100
83
11
-------
Visual inspection through an observation port revealed very dense clouds
of aerosol emitted from the grinders during operation. Some of this aerosol
apparently emitted in large diameter (>100 ym) droplets, since the fog dis-
sipated rapidly after the grinders were turned off. An odor characteristic
of Aroclors was very pronounced in the building after grinding; subsequent
analyses showed that all equipment inside the building, including the housings,
was coated with this material. A sample of the capacitor grindings was col-
lected for identification of the type of PCB mixture present.
12
-------
SECTION 4
SAMPLE ANALYSES
Upon return to the laboratory, the 14 OTS, SURC, and MISCO sample sets
(filter and vapor trap) were prepared for analysis. Due to the difficulty of
the analysis, only six sample sets were examined. The samples selected in-
cluded one from each of the three sampling methods located inside the building
adjacent to the grinder and simultaneously outside at Site A (see Figure 5).
The remaining samples were stored in a light-tight thermally insulated
storage container.
Analysis of the capacitor grindings collected on-site to verify the
type of PCB present established that the dielectric fluid used was Aroclor
1242, a PCB mixture containing 42% chlorine and consisting primarily of di-,
tri-, and tetrachlorobiphenyls (v.p. MD.001 mm Hg). Each sample consisted
of two glass fiber filters (one representing the first 35 min of the collec-
tion cycle and the other the remaining 300 min) and one polyurethane foam plug
or ethylene glycol sorbent. All samples showed gas chromatographic patterns
which closely or exactly matched that of Aroclor 1242.
Two extraction methods were required for recovery of PCBs. Foam plugs
and glass fiber filters were separately extracted for 24 h each with 700 ml
of 5% diethyl ether in hexane in Soxhlet extractors. The extract was reduced
to 1 to 2 ml in a Kuderna-Danish (KD) concentrator for cleanup on 10 cm x 7 mm
(i.d.) alumina column (activity grade 4, 6.5% water). The fraction eluted
from the column with 12 ml of hexane was then analyzed by electron gas chroma-
tography (EC-GC). The ethylene glycol was diluted to 450 ml with water (3.5:1)
and extracted three times in a separatory funnel with a total of 300 ml of
hexane. The hexane was concentrated to 10 ml in a KD concentrator and subjected
13
-------
to Florisil column chromatography according to the EPA Manual of Analytical
Methods (6). Only fraction I (which contained the PCBs) was collected, con-
centrated to 1 to 2 ml, and analyzed by EC-GC.
GC analysis was carried out on a 138 cm x 6 mm (i.d.) glass column packed
with 1.5% OV-17/1.95% OV-210 on 80-100 mesh Gas Chrom Q at 180°C. A 3HSc
electron capture detector was used for detection. Quantification was based
on comparison of the sum of the heights of the 15 major chromatographic peaks
in a standard of Aroclor 1242 with the corresponding peaks in the sample.
(All 15 peaks were present in every sample.) Since only Aroclor 1242 was
found in the capacitor fluid and all samples appeared to contain only this
mixture, confirmation by mass spectrometry or perchlorination was deemed
unnecessary.
Recoveries of Aroclor 1242 from polyurethane foam and ethylene glycol
were determined to be 96%, and 98%, respectively, by analysis of spiked media.
Therefore, it was not necessary to correct analytical data for losses in the
analytical scheme. Sampling efficiencies could not be determined because of
the multiplicity and complexity of the sampling procedures used. However,
the 24-h collection efficiency for Aroclor 1242 vapors by polyurethane foam
in the modified SURC sampler has been shown (5) to be 70-80% for effective
air concentrations of 18-98 ng/m sampled at 225 1/min.
The physical state of the airborne PCB mixture was apparently mostly
liquid aerosol inside the grinding shed, while both particulate and vapor
forms appeared to be present in the air outside the building. Sampling ef-
ficiencies, therefore, may have varied widely for each sampler and at the two
sites.
Blank values for filters, foam, and ethylene glycol carried to the sampling
site in sealed containers and returned for analysis showed only minimal con-
tamination by Aroclor 1242 (<0.3 to 8.8 yg per sample). These values cor-
responded to air concentrations of 0.007 to 1.1 yg/m3, depending on the sampler
used, and were not used for data correction.
14
-------
SECTION 5
RESULTS AND DISCUSSION
Sample identifications and analytical results are presented in Table 2.
Total air concentrations of Aroclor 1242 as determined by the three samplers
appear to be in reasonably good agreement except in the case of sample No. 6
(MISCO sampler, site A). Sample contamination is thought to be the cause of
the relatively high value of this sample. Lowest values were obtained with
the OTS sampler (Method I) at both sites and may reflect poorer sampling ef-
ficiency. Disagreement existed between Methods I and II as to the relative
amounts of material collected by the filters and foam plugs. Since the two
samplers are very similar in design and have similar air volume-to-filter
surface area ratios (2 1/min-cm2 for OTS, 2.8 1/min-cm2 for the modified
SURC), similar collection efficiencies would be expected. The relatively high
proportion of PCB material collected by the foam vapor trap of the modified
SURC sampler may be due in part to leakages observed around the prefilter.
Substantial vaporization from the prefilter to the foam trap would be expected
for both Method I and II samplers, however. No satisfactory explanation could
be made for the lack of PCB material in the Method I foam trap.
The levels of PCB measured inside the building were much higher than
expected, which demonstrated that uncontrolled grinding emissions may pose a
substantial "in-plant" health hazard. By comparison, the 8-h NIOSH occupa-
^
tional standard is set at 1 mg/m for Aroclor 1242 (7). Since the collection
capabilities of the sampling methods may have been exceeded at the mg/m level,
the measured concentrations inside the building may have been even higher than
indicated. The outside levels of <1 yg/m^ during grinding were probably much
higher than the true background that existed before grinding began, although
no pregrinding background samples were analyzed. Because of the large dif-
15
-------
ferences between inside and outside measurements during grinding, the magnitudes
of the three background samples become relatively unimportant. If resources
permit, further analyses of the stored samples may be attempted at a later
date to obtain information on the reproducibility of a given sampling method.
Cross-contamination and sample loss during storage could, however, complicate
interpretation of the low level results.
TABLE 2. ANALYSIS DATA SUMMARY - EL DORADO, ARKANSAS, PCB STUDY
Sample
No.
1
2
3
4
5
6
Site
E
E
E
A
A
A
Method
I
II
III
I
II
III
Prefilter
Collection*
(yg/m3)
1848
762*
28*
0.62
0.16
0.84
Foam/Glycolt
Collection
(yg/m3)
trace
5070
5210
0.03
0.69
7.52
Total
Collection
(yg/m3)
1848
5832
5238
0.65
0.85
8.36
*Concentrations expressed as yg/m3 Aroclor 1242.
tMethod I and II used polyurethane foam; Method III used ethylene glycol as
second collection stage.
^Indications of leakage by prefilter.
16
-------
REFERENCES
1. Environmental Sciences and Engineering Co. Methods for Multimedia Sampling
and Analysis of Polychlorinated Biphenyl in the Environment. Contract
68-01-2978, Report prepared for the Office of Toxic Substances, U.S.
Environmental Protection Agency, Washington, D.C., August 1975.
2. Bjorklund, J., B. Compton, and G. Zweig. Development of Methods for
Collection and Analysis of Airborne Pesticides. Contract CPT-70-15,
National Air Pollution Control Administration, Durham, N.C., 1970.
3. Stanley, C.W., J.E. Barney, M.R. Helton, II, and A.R. Yobs. Measurement
of Atmospheric Levels of Pesticides. Env. Sci. Technol., 5:430, 1971.
4. Federal Register. 36(84):8191-4, April 30, 1971.
5. Lewis, R.G., A.R. Brown, and M.D. Jackson. Sampling for Low Levels of
Airborne Pesticides, Polychlorinated Biphenyls and Polychlorinated
Naphthalenes. 173rd Nat. Meeting Am. Chem. Soc., New Orleans, Louisiana,
March 1977 (Anal. Chem., in press, 1977).
6. Thompson, J.F. Analysis of Pesticide Residues in Human and Environmental
Samples. U.S. Environmental Protection Agency, Research Triangle Park,
N.C., December 1974.
7. Christensen, H.E. and T.T. Luginbyhl, Eds. Registry of Toxic Effects
of Chemical Substances. National Institute of Occupational Safety and
Health, Rockville, Md., 1975.
17
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
MONITORING FOR POLYCHLORINATED BIPHENYL EMISSIONS FROM
AN ELECTROLYTIC CAPACITOR DISPOSAL PROJECT
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Rodes, C.E., M.D. Jackson, and R.G. Lewis
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Joint Effort EMSL/HERL
ORD
RTP, N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Field Project
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Three different air sampling methods were used simultaneously to
monitor for PCB emisssions arising from a pilot disposal project involving
electrolytic capacitors. Analytical results indicated that the primary
polychlorinated biphenyl material was Aroclor® 1242, and that airborne
concentrations inside the building housing the grinders exceeded 5mg/m3.
The PCB air concentrations outside the building at a distance of 9 m
o
were typically <1.0 yg/m . Measurements made with one low-volume and
two high-volume air samplers are compared.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
Ambient Monitoring
Source Monitoring
PCB
Capacitor Disposal
13 B
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report}
Unclassified
21. NO. Of PAGES
22
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
IB
------- |