&EPA
United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Reaearch Triangle Park NC 27711
EPA-600/4-79-022
March 1979
Reaearch and Development
Sources of
Emissions of
Polychlorinated
Biphenyls into the
Ambient
Atmosphere and
Indoor Air
-------
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-600/4-78-022
March 1979
SOURCES OF EMISSIONS OF POLYCHLORINATED BIPHENYLS
INTO THE AMBIENT ATMOSPHERE AND INDOOR AIR
by
Kathryn E. MacLeod
Analytical Chemistry Branch
Environmental Toxicology Division
Health Effects Research Laboratory
Research Triangle Park, N.C. 27711
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
HEALTH EFFECTS RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
DISCLAIMER
This report has been reviewed by 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.
-------
FOREWORD
The many benefits of our modern, developing, industrial society are
accompanied by certain hazards. Careful assessment of the relative risk of
existing and new man-made environmental hazards is necessary for the estab-
lishment of sound regulatory policy. These regulations serve to enhance the
quality of our environment in order to promote the public health and welfare
and the productive capacity of our Nation's population.
The Health Effects Research Laboratory, Research Triangle Park, conducts
a coordinated environmental health research program in toxicology, epidemio-
logy, and clinical studies using human volunteer subjects. These studies
address problems in air pollution, non-ionizing radiation, environmental
carcinogenesis and the toxicology of pesticides as well as other chemical
pollutants. The Laboratory participates in the development and revision
of air quality criteria documents on pollutants for which national ambient
air quality standards exist or are proposed, provides the data for registra-
tion of new pesticides or proposed suspension of those already in use,
conducts research on hazardous and toxic materials, and is primarily respon-
sible for providing the health basis for non-ionizing radiation standards.
Direct support to the regulatory function of the Agency is provided in the
form of expert testimony and preparation of affidavits as well as expert
advice to the Administrator to assure the adequacy of health care and
surveillance of persons having suffered imminent and substantial endanger-
ment of their health.
This report represents a research effort in the field of monitoring for
airborne chemical pollutants. The emphasis of the project was to determine
the sources of atmospheric emissions of polychlorinated biphenyls (PCB).
Such data is required for the continued development of criteria for ambient
air quality standards.
F. G. Hueter, Ph.D.
Director
Health Effects Research Laboratory (MD-51)
m
-------
ABSTRACT
Polychlorinated biphenyls (PCB) have been identified in air samples
from many parts of the world since the 1960s. This study was undertaken to
identify and compare different sources of PCB in indoor and outdoor air.
Almost all sampling was performed in central North Carolina. The suspected
sources that were tested were fluorescent light ballasts, landfills, electrical
substations, a transformer manufacturer, and the sites of illegal dumping.
Defective light ballasts emit large quantities of PCB and are an
important indoor source. Capacitors in small electrical equipment may also
be an important source. In general, indoor air levels of PCB were at least
one order of magnitude higher than outdoor levels. The data indicate that
the landfills and electrical substations tested are not major sources of
PCB. The transformer manufacturer had elevated levels of PCB in the immediate
area of the plant but did not contribute greatly to the levels found off
the property. The spill sites also had elevated levels of the contaminant
in their immediate area, but the levels 50-100 m away were normal for rural
areas.
IV
-------
CONTENTS
Page
FOREWORD 1ii
ABSTRACT iv
LIST OF FIGURES vi
LIST OF TABLES vii
SECTION I: CONCLUSIONS 1
SECTION II: INTRODUCTION 2
Survey Sites 5
SECTION III: MATERIALS AND METHODS 10
Air Sampling 10
Analytical Methods 12
Quality Control 15
SECTION IV: PCB IN INDOOR AIR 26
Laboratories and Offices 27
Houses 28
SECTION V: LANDFILLS 32
SECTION VI: ELECTRICAL SUBSTATIONS 39
SECTION VII: MANUFACTURING SITE 43
SECTION VIII: PCB SPILL 55
SECTION IX: REFERENCES 68
-------
LIST OF FIGURES
No. Page
1. Sampling Area Within North Carolina. 7
2. Chromatograms of Ambient Air Sample Extract and Aroclor
Standards 19
3. Chromatograms Showing Peaks Used for Quantifying PCB,
SE-30/OV-210 20
4. Chromatographs of Personal Monitoring Air Extract Collected at
Inez, N. C. SE-30/OV-210 21
5. Locations of Durham Landfills Sampled 35
6. Location of Raleigh City Landfill Sampled 36
7. Location of Old Demolition Site, Goldsboro, N. C 37
8. Locations of Electrical Substations Sampled 41
9. Sampling Sites in the Vicinity of a Transformer Manufacturer ... 47
10. Annual Wind Rose for Goldsboro, N. C 48
11. Soil Sampling Sites in the Vicinity of Transformer
Manufacturer 49
12. Chromatograms of Extracts of Samples Collected Near
Transformer Manufacturer . . . 50
13. Area of North Carolina Containing PCB Spills Along Highways. ... 58
14. Location of the Practice Dig Near Inez, N. C 59
15. Chromatograms of Personal Monitoring Sample Extracts and
Aroclor 1260 Standard, SE-30/OV-210 60
16. Chromatograms of Spill Sample Extract and Aroclor 1260
Standard, OV-101 61
17. Chromatograms of Aroclor 1260 Standard and Ambient Air Sample
Extract from Highway PCB Spill Site, SE-30/OV-210 62
VI
-------
LIST OF TABLES
No. Page
1. Physical and Chemical Properties of Aroclors 8
2. Some Production and Use Data for Aroclors 9
3. Recovery of PCS from Spiked Soil Samples 22
4. Efficiency of PCB Removal From Foam Plugs During Extraction ... 23
5. Efficiency of Extraction and Cleanup 24
6. Replicate Samples of Indoor PCB 25
7. Decline of Airborne PCB in a Room After Burnout of Light
Ballast 30
8. Airborne PCB Levels in Private Dwellings 31
9. PCB Concentrations in Air Over Landfills. 38
10. PCB Data from Electrical Substations 42
11. Ambient Concentrations of PCB in the Vicinity of a Transformer
Manufacturer 51
12. PCB Concentrations as a Function of Distance from Ground
Level 52
13. Concentrations of PCB in Soil in the Vicinity of a Transformer
Manufacturer 53
14. Meteorological Data During Sampling at a Transformer
Manufacturer 54
15. Aroclor 1260 Data from Roadside Spill Sites 63
16. Profiles of Ambient Aroclor 1260 Concentrations Over
Spills 64
17. Indoor Levels of Aroclor 1260 Near Highway Spill Sites 65
18. Levels of Aroclor 1260 in Breathing Zones of Workers During Test
PCB Spill Removal 66
19. Airborne PCB Concentrations Near Test Removal Site in Warren
County, North Carolina 67
vii
-------
SECTION I
CONCLUSIONS
Defective light ballasts manufactured prior to 1972 can release large
quantities of polychlorinated biphenyls (PCB) into the air and appear to be
an important source of indoor air pollution. Capacitors in small electrical
equipment may also be an important source. In general, indoor air levels
of PCB are at least one order of magnitude higher than outdoor levels. The
data indicate that the landfills and electrical substations tested are not
major sources of atmospheric emission of PCB. Manufacturing and spill
sites were found to have elevated levels of the contaminant in the air
immediately surrounding the areas, but the levels 500 m away were normal
for rural areas.
-------
SECTION II
INTRODUCTION
The polychlorinated biphenyls (PCB) are a family of chemicals that
have been known to chemists since the 19th century and have found many
industrial uses during this century. They were first commercially available
in this country in 1929 and were manufactured by the Monsanto Corporation
between 1957 and mid-1977 in vast quantities as mixtures under the trade
name Aroclor. (PCB imports only accounted for about 1% of the total used
2
in the United States up to 1972 but have been rising. ) Monsanto marketed
at least nine different Aroclors which were distinguished by the percent
chlorination and had physical properties varying from clear mobile oil to
yellow, sticky, viscous resin to white powder. These different mixtures
found industrial uses in such diverse products as textile dyes, printing
inks, fireproofing agents, pesticide extenders, hydraulic fluids, dielectrics
in capacitors and transformer fluids. Their desirable qualities include
chemical and thermal stability, non-flammability, electrical insulating
properties, adherence to smooth surfaces and solubility in most common
organic solvents. They are hydrophobic and resistant to acids, bases and
oxidation.
By 1966, Monsanto was producing almost 30 million kg of these mixtures
in a single year. This was the same year that the New Scientist contained
3
a note about the discovery of PCB in the Swedish environment. In October,
1968, rice-bran oil manufactured in Kitakyusha City, Japan, was accidentally
contaminated by large quantities of Kanechlor 400, a Japanese commercial
mixture of PCB containing 48% chlorine. At least 1000 people of all ages
-------
were poisoned by consuming this contaminated oil. Symptoms of the disease,
which has come to be called "Yusho" included chloracne, eye discharges,
hyperpigmentation of skin and nails, swelling of upper eyelids, and still-
4
birth. Over the next few years, several investigators reported finding
PCB in everything from Antarctic snow to mothers' milk to Icelandic
plants and animals. The very properties that made PCB desirable for
industrial uses assured that they would not break down quickly in the
environment. PCB were on the way to becoming as universal a pollution
problem as DDT.
After the Yusho incident and other PCB accidents in this country, the
general population became more concerned about the environment and more
aware of the adverse effects of pollution at this time. In 1971, with the
mounting evidence against PCB, Monsanto voluntarily cut back production of
the Aroclors, ceased manufacturing the most highly chlorinated mixtures and
restricted sales to uses in closed systems only. At this time they also
introduced a new PCB mixture, Aroclor 1016, which contained significantly
less of the more highly chlorinated isomers than the Aroclor 1242 which it
was designed to replace. In October, 1977, they ceased making all PCB.
The problem of universal contamination is still present, however, with
o
atmospheric PCB concentrations being reported in the ng/m range from such
8 9 10
diverse places as Rhode Island, Bermuda and Southern California.
The existing U. S. Occupational Safety and Health Agency standard for
3
a safe workplace environment is 1 mg/m time-weighted-average permissible
air concentration for an 8 to 10 hr. workday, 40 hr. work week. In light
of recent evidence suggesting that PCB may be potential carcinogens, the
National Institute for Occupational Safety and Health has recommended
-------
a new criterion of 1 ug/m (based on the supposed analytical detection
4
limit) for all carcinogens including PCB.
Since PCB were used so enthusiastically in the 1960s and early 1970s,
there is still a large reservoir of the material capable of entering the
atmosphere from different sources. Every step in the process from the
manufacturing of PCB themselves to the final disposal of the worn out
product is a potential source of loss of PCB into the atmosphere. In their
1972 paper, Nisbet and Sarofim11 estimated that between 1930 and 1970, 2.7
4 4
x 10 metric tons of PCB had been released into the air; 5.4 x 10 metric
tons into lakes, rivers and other bodies of water and that an additional
2.7 x 10 metric tons were in dumps and landfills. They based their cal-
culations on losses during the manufacture of PCB and PCB-containing products,
leaks from the products during use, lifetime of the products and method of
disposal of scrap, waste and used products.
Since PCB are no longer manufactured in the United States, several of
the sources suggested by early investigators such as vaporization of PCB
containing paints and plasticizers are less important. Today's most likely
sources are considered by many to be: incinerators, dumps and landfills,
leaking transformers and capacitors, manufacturing sites and deliberate
spills. While many people are aware of the presence of PCB in the air, up
to this time very little actual sampling has been undertaken to confirm the
sources. After cataloging the various suspected sources of PCB emission
into the atmosphere, this study was undertaken to survey a number of these
sources and to determine the aerial PCB concentrations in their vicinity.
-------
Survey Sites
Dumps and landfills have been estimated to contain 2.7 x 10 metric
tons of PCB and, therefore, are a very likely source of atmospheric con-
tamination. One landfill that is presently in operation and three that
were in use prior to 1971, but are now closed, were chosen as sites for
sampling in this study. The landfill presently in use is the Durham, North
Carolina, city landfill. The three closed landfills are the old Durham,
Raleigh and Goldsboro, North Carolina, municipal landfills. Of these four,
only the one at Goldsboro has had known dumping of PCB.
Transformers accounted for 19% of the domestic use for Aroclors between
1960 and 1971 and with capacitors, the only use since 1972. There are
literally thousands of electrical substations all across the United States
with large transformers containing an average of 689 liters (1131 kg) of
askarel, the PCB containing fluid. These transformers are near buildings
and transportation facilities and are sitting out exposed to the elements
12
during their 25-30 year lifetimes. These seemed to be excellent potential
sources of contamination, so five of these substations were surveyed during
this study.
Another potential source of PCB is the manufacturing sites across the
country where PCB are being used or have previously been used. Prior to
1971, little care was taken in the handling, storage, cleanup or disposal
of PCB at these sites. One of the thirteen transformer manufacturers in
the country has a plant in Goldsboro, North Carolina. The company allowed
the area around the plant to be tested for PCB emissions.
Perhaps the most dangerous source of emissions is the site of deliberate
illegal dumping of PCB or PCB-containing wastes. The highways of North
-------
Carolina have recently been subjected to just such an occurrence. Used
transformer oil containing large quantities of Aroclor 1260 was dumped
along approximately 500 km of highways in a dozen counties surrounding
Raleigh during July and August 1978. Advantage was taken of the opportunity
to take air samples along the spills and the surrounding areas.
i
One use of PCB has been in the ballast of fluorescent lights. Because
of their long life and economy, fluorescent lights are used quite often for
institution, factory and business lighting. For that reason, fluorescent
light ballasts and rooms containing flourescent lights were included in
this study.
For use as controls, two sites with no suspected PCB contamination
were also monitored. These were a small goat farm near Apex in Wake
County and land behind the U. S. Environmental Protection Agency
Environmental Research Center at Research Triangle Park in Durham County.
Several potential sources were not surveyed during this study for
various reasons. Incinerators were not sampled because there are no municipal
13
incinerators in use in this area. Recent data indicate that incinerators
can indeed release large quantities of PCB. Data were collected at sewage
sludge incinerators in Kansas and Missouri, an industrial incinerator in
Texas and municipal refuse incinerators in Florida. The investigators
quantified the PCB as decachlorobiphenyl in all cases. The refuse incin-
erators released 2 to 34 [ig/m into the air while the industrial incinerator
3 3
released from 0.7 ng/m to 0.3 ug/m although that incinerator was burning
fluids containing 10-17% PCB at the time of sampling. The Missouri sewage
-------
sludge containing a significant industrial component released 200-300 ug/m
PCB while the Kansas sludge which was mostly domestic waste released only 4
o
ug/m when incinerated.
Landfills near industrial centers would be more likely to contain
large quantities of PCB than the ones which form part of this study. In
14
fact, one study has shown that the vent gases from an above ground landfill
q
in a Chicago suburb does contain ug/m quantities of PCB. Considerations
such a$ time, money and location made it impossible to include these in
this study. Neither a capacitor manufacturer nor capacitors themselves
were used for this study, because the above considerations made it
inconvenient to do so. The investment casting industry still uses PCB in
their lost wax process. This process would appear to be very prone to
losses of PCB (mainly decachlorobiphenyl) to the environment. However,
there was not a manufacturing site in the geographical area under study.
Figure 1. Sampling area within North Carolina.
-------
Table 1
Physical and Chemical Properties of Aroclors'
00
Aroclor
1221
1232
1242
1248
1254
1260
1262
1268
1016
Appearance Percent Cl
clear, mobile oil
clear, mobile oil
clear, mobile oil
clear, mobile oil
light yellow viscous liquid
light yellow, soft sticky resin
light yellow, sticky viscous resin
white to off-white powder
clear, mobile oil
21
32
42
48
54
60
62
68
41
Average N°C1 /molecule
1.15
2.04
3.10
3.90
4.96
6.30
6.80
8.70
2.96
Vaporization rate
(g/cmVhr)
0.00174
0.000874
0.000338
0.000152
0.000053
0.000009
0.000013
--
--
Technical Bulletin 0/PL-306A, Monsanto, Organic Chemicals Division, St. Louis, Missouri, 1970.
-------
Table 2
Some Production and Use Data for Aroclors*
Aroclor
U. S. production, 10 kg
1969 1971 1974
Major uses
1221
1232
1242
1248
1254
1260
230 2,007 26 capacitors, plasticizer,
adhesives
124 78 0 hydraulic fluids, plasti-
cizer, adhesives
20,678 9,991 2,821 capacitors, transformers,
hydraulic fluid, plasti-
cizer, adhesives, wax
extenders, carbonless
copy paper, heat transfer
2,568 97 0 vacuum pumps, hydraulic
fluid, plasticizer,
adhesives
4,464 2,119 2,811 capacitors, transformers,
vacuum pumps, hydraulic
fluids, plasticizer,
adhesives, pesticide
extender, inks, lubri-
cants, cutting oil
2,018 784 0 transformers, hydraulic
fluids, plasticizer
1262
1268
1016
324
136
0
0.4
0
1,515
0
0
9,980
plasticizer
plasticizer,
capacitors
wax extender
See reference 12.
-------
SECTION III
MATERIALS AND METHODS
Air Sampling
Both low- and high-volume air samples were taken in this study.
Low-volume collections were carried out with Mine Safety Appliance (MSA)
Portable Pump Model S personal samplers. These small battery-operated
pumps were run at 2.5 liters per minute and were used to sample total air
volumes ranging from 100 to 3000 liters. Larger air volumes (3.4 to 200
m ) were sampled with Bendix Hurricane dual speed pumps at flow rates of
0.1 to 0.5 m /minute. The Bendix pumps were operated in the field by
gasoline powered generators.
The air flow through the MSA pumps was indicated by a rotameter graduated
in liters per minute. Two methods were used to calibrate the pumps to
within ±5%. A bubble flow meter of 1.0 liter volume was used to measure
the volume of air moved per unit time. Some of the pumps were also calibrated
by displacing water from an inverted 2 liter graduated cylinder during a
measured length of time. On the Bendix high volume pumps, the exhaust air
was forced through a restricting orifice (supplied with the pump) and the
resulting back pressure was measured by a gauge placed directly ahead of
the orifice. The gauge was calibrated by the manufacturer to read directly
in ft /min air flow.
Polyurethane foam has been shown to be an efficient collector of
heavily chlorinated hydrocarbons, particularly the higher molecular weight
Aroclors. It is also inexpensive, easy to handle, and can be effectively
10
-------
Q
cleaned to provide low background levels. The polyurethane foam (polyether
o
type, density 0.0225 g/cm ) was received in a large sheet and plugs of the
proper sizes were cut from it. Prior to their use in the field, these
plugs were Soxhlet extracted twice for periods of at least seven hours in
5% diethyl ether in hexane. The solvent from the second extraction was
concentrated, eluted through an alumina column and analyzed by gas liquid
chromatography (GLC) to determine the blank value. Each foam plug was
given an ID number which remained the same for the entire survey period.
The plugs were individually wrapped in hexane-rinsed aluminum foil for
storage between uses.
Glass fiber filters, MSA #CT-75428, were cut to size, wrapped loosely
in aluminum foil, heated to 315°C in a muffle furnace overnight to remove
any organic material, and placed in a desiccator until use.
The 10 cm filter holder that is standard equipment on the Bendix
Hurricane pump was modified by epoxying a cylindrical chamber, 25 cm long x
5 cm i.d., behind the filter holder. Two foam plugs, 5.5 cm diameter x 8
cm thick, were placed in the chamber and a 10 cm diameter glass fiber
filter was placed in front of them in the filter holder. The ends of the
holder were covered with hexane-rinsed aluminum foil for transport to the
sampling site. At the site, the sampler was connected to the Bendix pump
(R)
by a 7.6 m length of Flexaust CWC hose and located upwind of the pump and
generator exhaust.
A glass tube, 2 cm in diameter and 7.5 cm long (tapered the last 3
cm), was used for the sampling cartridge with the MSA pumps. A small foam
plug, 4 cm long x 2 cm dia. was placed in this tube and the entire tube was
11
-------
wrapped in hexane-rinsed foil for transport. For sampling, the tube was
®
connected to the MSA pump by a length of Tygon tubing.
After the sampling period was completed the glass fiber filter and the
foam plugs were individually wrapped in aluminum foil until analysis.
Plugs and filters carried to the field along with those used for sampling
served as controls.
Analytical Methods
Air Samples. Upon returning to the laboratory, the large foam plugs and
filters were placed individually into Soxhlet extractors with 300 ml of 5%
diethyl ether in hexane and extracted for approximately 16 hours (3 cycles/
hour). The small plugs were placed in smaller Soxhlets with approximately
200 ml of hexane and were extracted for at least seven hours (6 cycles/hour).
The extracts were concentrated to a volume of less than 1 ml using a Buchii
®
rotary evaporator and an Organomation N-evap and were cleaned up by alumina
column chromatography.
Basic alumina (60 mesh), obtained from Alfa Products, was weighed into
a flask, distilled water equal to 6% by weight was added and the flask was
stoppered and shaken well. The adsorbent was allowed to equilibrate for at
least 15 hours before use. Any of this mixture not used within two weeks
was discarded. For chromatography a wad of pre-extracted glass wool was
®
placed in the bottom of a Chromaflex column (7 mm i.d. x 22 cm long,
Kontes #42100) and washed with 10 ml hexane. The column was then packed to
a depth of 17 cm with alumina. The sample was added to the column, eluted
with 15 ml hexane and the eluate was concentrated to a suitable volume for
GLC analysis.
12
-------
When necessary, sample extracts were fractionated by silicic acid
column chromatography to separate PCB from chlorinated pesticides (espe-
cially from technical chlordane residues which were present in many samples).
Mallinckrodt AR silicic acid (100 mesh) was prepared for use by heating for
at least seven hours at 130°C and then cooling to room temperature in a
desiccator. After cooling the silicic acid was weighed into a bottle and
3% distilled water was added. The bottle was sealed tightly, shaken well
and placed back in the desiccator for at least 15 hours. Any of this
deactivated silicic acid not used within a week was discarded. Columns
®
were prepared by placing 3 g of silicic acid in a Chromatoflo column (9 mm
i.d. x 25 cm long, Pierce #29020) equipped with a Teflon mesh support
membrane (Pierce #29268), lower end plate, adapter and a 500 ml solvent
reservoir (Ace #5824-10). The column was washed with hexane, the sample
which had been concentrated to less than 1 ml was placed on the column and
130 ml of hexane was added to the reservoir. Nitrogen pressure was applied
to the column to increase the flow to approximately 3 ml per minute. The
eluate was collected in three fractions. The first (0-30 ml) contained all
of the Aroclor 1254 and most of the Aroclor 1242. The second fraction
(30-50 ml) contained the remainder of the Aroclor 1242 and the early eluting
peaks of technical chlordane. The final 80 ml fraction contained the rest
of the technical chlordane, including all of the cis- and trans-chlordane.
The volumes of these fractions were adjusted and the samples were analyzed
by GLC.
The hexane used throughout the study was pesticide quality obtained
from Burdick and Jackson. The diethyl ether was analytical reagent grade
(Mallinckrodt #0850, containing 2% ethanol).
13
-------
All samples were analyzed by electron capture GLC. The instrument was
a Tracer 222 with a linearized Ni detector. The columns used included:
3% OV-225 on Supelcoport® 80/100 mesh; 3% OV-1 on Chromosorb® W HP 80/100
mesh; and 4% SE-30/6% OV-210 on Gas-Chrom® Q 100/120 mesh. All columns
were 6 mm i.d. and 1.8 m long. Off-column injection was practiced at all
times. The detector temperature was 295°C, the oven temperature was 200°C
and the transfer lines and inlet were maintained at 220°C. Nitrogen was
used as the carrier gas at a flow rate of 85 ml/min for the SE-30/OV-210
column and 70 ml/min for the other two columns.
Aroclor standards were obtained from the Pesticide Repository, U. S.
Environmental Protection Agency, Health Effects Research Laboratory, Research
Triangle Park, North Carolina. The standards were made by dissolving the
Aroclor in benzene and making dilutions with hexane. Stock solutions were
stored in brown bottles at -10°C. Working standards were remade periodically
from these and were stored under refrigeration when not in use.
Soil Samples. Surface (top 7 to 10 cm) soil samples were collected in
the vicinity of the transformer manufacturer and placed in pre-cleaned
glass and metal containers. Upon returning to the laboratory, the soil was
dried in a vacuum oven overnight at 110°C, put through a 120 mesh sieve and
weighed into extraction thimbles. The samples were extracted in Soxhlet
extractors using either 300 ml of 1:1 hexane/acetone as recommended by the
U.S. EPA Manual of Analytical Methods16 or 300 ml of 5% diethyl ether in
hexane. Re-extraction of the samples and extraction of soil fortified with
PCB standards showed that the 5% ether extraction was just as effective at
removing PCB from the soil as the acetone/hexane but did not produce as
14
-------
many interfering peaks in the gas chromatograms. Duplicate analyses showed
that with the acetone/hexane extraction 47 ± 3% of the Aroclor 1242 and 93
± 6% of the Aroclor 1254 were recovered. Using the ether/hexane mixture,
recoveries were 65 ± 3% for Aroclor 1242 and 99 ± 12% for Aroclor 1254.
PCB Quantification. Since Aroclors are combinations of many isomers
and congeners, the GLC elution gives a multi-component chromatogram. This
makes quantification a much more complicated task than would be the case
for a single component pollutant. The method chosen for identifying and
quantifying the PCB sample extracts was based on the total peak height of
the largest peaks present in the gas chromatographic "fingerprints." The
number of peaks used varied from seven for Aroclor 1254 to thirteen for
Aroclor 1260 (Figure 3).
The elution pattern for a sample extract that was not taken near a
source of PCB emission resembled that of a combination of Aroclors. Since
different Aroclors contain many of the same isomers, it was necessary to
assign a cut-off for one Aroclor and a beginning for another. For example,
the last two major peaks of Aroclor 1242 have the same GC retention time as
the first two major peaks of Aroclor 1254 on the SE-30/ OV-210 column.
These last two peaks were generally quantified as part of the Aroclor 1254.
Quality Control
Foam plug cleanup. Before being used for high volume samples, each
individual large polyurethane foam plug was extracted and the extract was
eluted through an alumina column and analyzed by GLC to determine its blank
value. The alumina removed early GLC eluting components present in the
diethyl ether and components present in the foam itself which interfered
15
-------
with the analysis of the Aroclors. Figure 4 shows an air sample extract
before and after alumina cleanup. At various times during the course of
this survey the plugs were re-extracted after use to determine whether
there was PCB buildup in the foam. When sampling at the location of a PCB
spill, it was not unusual for the first plug in line to collect over 5000
ng of PCB and in one case over 2 mg was found. If even a small percentage
of this amount remained on the plug until the next extraction it could add
as much as an order of magnitude to what actually would be collected during
that sampling. Table 4 shows the amount of Aroclor 1242 and Aroclor 1254
present on several plugs before and after sampling. The average amount of
PCB was less after re-extraction but it is not a statistically significant.
It can be seen that there is negligible PCB carryover.
The small foam plugs used with the MSA samplers were also pre-extracted
with 5% diethyl ether in hexane. They were then re-extracted, this time
with hexane alone, and the extracts were concentrated and analyzed by GLC.
The smaller size of these plugs resulted in lower blank values than the
large plugs had, but this advantage was lost in sampling due to the much
smaller volume of air pulled through them during a sampling period. Besides
having lower values, these plugs also had much less variation in the back-
ground level (9 ± 1 ng Aroclor 1242, 12 ± 2 ng Aroclor 1254 and 12 ± 5 ng
Aroclor 1260). Therefore, only the average blank was subtracted. Plugs
re-extracted after use indicated total removal of the PCB during a single
extraction.
Recovery efficiency. The efficiency of the recovery process was
checked by spiking foam plugs with Aroclor standards and then carrying them
16
-------
through the sample extraction and cleanup procedure. A syringe or pipette
was used to fortify the small plugs with a known amount of Aroclor 1242 and
Aroclor 1254 in hexane. After allowing the hexane to evaporate, the plugs
were extracted in the usual manner and the extracts were concentrated, run
through alumina and analyzed. Four small fortified plugs gave 77% + 11$
recovery for the Aroclor 1242 and 100% ± 18% for Aroclor 1254 when spiked
with 100 ng of each of these standards.
The large foam plugs were fortified with a mixture of five different
PCB isomers that are components of commercial Aroclor mixtures. After the
mixture was syringed into the plugs, the samples were Soxhlet extracted,
cleaned up by alumina chromatography and analyzed by GLC. The recoveries
of the Soxhlet extractions, the alumina column procedure and a combination
of the extraction and concentration steps were also checked to determine
any losses during the separate steps of the procedure. The results of
these studies, shown in Table 5, indicate that the greatest loss (least
accuracy) was during the rotovap step.
Sampling precision. The accuracy and precision of the final values
generated depend on the accuracy with which the actual air flow rate is
known, the collection efficiency of the sampling medium and the analytical
accuracy in the laboratory. The flow rate can, and usually does, change
during the sampling period due to particulate build-up on the filter, power
fluctuations in the generator or pump, and battery drain in the case of the
battery operated MSA samplers. Through calibration, the air flow rate of
the pumps can be accurately determined to within 5% and readings can be
taken during operation, so the greatest error or deviation is in the analytical
17
-------
procedure and not the sampling step. Billings et ah found that the
relative standard deviation of the analysis of triplicate air samples was
about the same as the precision obtained in analyzing test mixtures of PCB
and pesticides. When four MSA samplers were run overnight side by side in
the same laboratory and the samples were carried through the analytical
procedure at the same time, the results showed that the overall method had
a relative standard deviation of 17% (Table 6).
18
-------
I-
X
u
LU
I
LU
O.
LU
LU
cc
10
15
TIME, minutes
20
25
30
A. AROCLOR 1242, 0.50 ng
B. AROCLOR 1254, 0.50 ng _
C. AMBIENT AIR SAMPLE EXTRACT, 5 p.\ INJECTION FROM 1 ml
Figure 2. Chromatograms of ambient air sample extract and aroclor standards, SE-30/OV-210.
10
-------
I-
I
C3
UJ
I
UJ
Q.
UJ
>
<
_l
LU
CC
I
10
15
TIME, min
20
25
30
A. AROCLOR 1242, 0.50 ng
B. AROCLOR 1254, 0.50 ng
C. AROCLOR 1260, 0.50 ng
Figure 3. Chromatograms showing peaks used in quantifying PCB, SE-30/OV-210.
20
-------
X
C3
UJ
z
UJ
Q.
HI
cc
B
0 5 10 15
TIME, minutes
A. BEFORE ALUMINA CLEANUP, 1.3 ml, 5 jul INJECTION
B. AFTER ALUMINA CLEANUP, 1.2 ml, 5 jul INJECTION
20
25
Figure 4. Chromatograms of personal monitoring air sample taken at Inez, N.C., SE-30/OV-210.
21
-------
Table 3
Recovery of PCB from Spiked Soil Samples
Sample
Acetone: hexane,3 unspiked
(1)
(2)
X
Acetone: hexane , spi ked
(1)
(2)
X
Ether: hexane, c unspiked
(1)
(2)
X
Ether: hexane, spiked
(1)
(2)
X
ng Aroclor 1242
1044
305
674
1564
1665
1614
729
541
635
1979
1874
1926
ng Aroclor 1254
1592
957
1274
3253
3029
3141
2250
1607
1928
4161
3655
3908
a50% acetone in hexane
bFortified with 2000 ng each Aroclor 1242 and Aroclor 1254
C5% diethyl ether in hexane
22
-------
Table 4
Efficiency of PCB Removal from Foam Plugs During Extraction
Plug I.D.
fP 22
fP 20
fP 14
fP 17
fP 4
fP 15
fP 1
X
s
ng Aroclor 1242
Before Run3 After
62
69
31
47
67
120
73
67
27
1600 43
1400 25
83 48
720 32
8100 37
900 50
56
42
11
ng Aroclor 1254
Before Run3 After
97
100
45
180
160
120
53
110
49
2700
1600
97
320
5400
1700
140,000
120
35
72
41
45
54
150
75
45
3PCB collected during sampling period.
23
-------
ro
Table 5
Efficiency of Extraction and Cleanup
PCB
4
2
2
2
2
,4'
,4'
,2'
,2'
,2'
DCB
,5
,5
,4
,4
KB
,5' TCB
,5,5' PCB
,4', 5,5' HCB
ng of
10,
1,
1,
1,
1,
spike
000
000
000
000
000
% recovery,
Total
procedure (8)
88
90
88
88
86
mean3
Extraction (4)
116
110
108
108
109
Extraction
and Rotovap (4)
83
85
85
87
94
Alumi
column
107
106
104
100
88
na
(4)
aNumber of replicate samples given in parentheses
-------
Table 6
Replicate Samples of PCB Collected Simultaneously
Sample
Total air volume
ng/m Aroclor 1242
ng/m Aroclor 1254
1
2
3
4
2.3 irT
2.4
2.4
2.4 nT
143
186
198
211
109
118
135
159
X
s
RSD
184
29.5
16%
130
22
17%
25
-------
SECTION IV
PCB IN INDOOR AIR
As was noted in the introduction, fluorescent lighting has been and
continues to be a very popular method of providing light in factories,
businesses and institutions. Fluorescent fixtures manufactured before 1972
often incorporated PCB-containing ballasts and because these ballasts have
a very long life (estimated at 12 years by the IES Lighting Handbook,
18
1966), many of these are still in use at the present time. However, very
little work has been done on measuring the levels of PCB in indoor air.
19
Staiff et aJL made some measurements in 1974, using two different
methods of trapping the PCB. One method was to treat nylon screens with
ethylene glycol and expose them for a given time period. PCB collected in
o
this manner was measured in ug/m of fabric area and only very rough
correlations with air concentration could be made. The second, more accurate,
method used was drawing air through ethylene glycol in midget impingers.
3
This method had a lower limit of detection of 0.5 ug/m for PCB. Neither
of these methods gave very satisfactory results, especially considering
that the proposed NIOSH criterion of 1.0 ug/m time-weighed average for
4 3
maximum exposure in workplace air. They measured approximately 100 ug/m
of Aroclor 1242 one meter below the ballast immediately after burnout,
3
approximately 40 ug/m three days later, and "several weeks later" the
concentrations were below the detection limit.
26
-------
During this study air samples were taken in laboratories with covered
and uncovered fluorescent fixtures, in offices with covered fixtures and in
several homes in rooms with and without fluorescent lighting. There was
also an opportunity to monitor a room containing a burned out PCB-containing
light ballast.
Laboratories and Offices
The two buildings sampled were the Monsanto Triangle Park Development
Center (site A) at Research Triangle Park, North Carolina, and the Physical
Sciences Center, University of South Carolina, Columbia, South Carolina
(site B). The samples were collected during 1977 and the first half of
1978.
The ambient level of PCB outside of site A (laboratories with uncovered
3
fixtures) averaged 18 ng/m calculated as Aroclor 1242 plus Aroclor 1254.
o
Inside two different laboratories the level averaged 217 ng/m and contained
approximately equal amounts of Aroclor 1242 (55%) and Aroclor 1254 (45%).
Levels in the laboratory with the covered fixtures (site B) averaged 208
ng/m while the ambient air at that location averaged approximately 4
3
ng/m . The two offices tested were also located at site A. PCB concen-
trations within the offices averaged 96 ng/m .
On March 31, 1977, a fluorescent light ballast burned out in an office
at site A. The ballast was replaced almost immediately but was left in the
room overnight while an air sample was collected. The ballast was removed
and disposed of the next morning. At regular intervals after this, air
samples were taken overnight in the same location in that room. The results
19
are shown in Table 7. While the ballast studied by Staiff et aJL contained
27
-------
only Aroclor 1242, the one in this study released a PCB mixture resembling
equal amounts of Aroclor 1242 and Aroclor 1254.
These measurements indicate that when fluorescent light ballasts burn
out, quantities of PCB are released which can elevate the levels in indoor
air for periods of several weeks. After three, or at most four, months the
levels receded to what must be considered background for indoor air. The
fluctuations in PCB concentration after that time are probably due to daily
variations or, more likely, to limits in sampling precision. For further
discussion of this point, see the section on quality control.
Houses
The houses used in this study located were in Raleigh, Durham, Cary,
and Lumberton, North Carolina. The samples were collected between April
and August, 1978. Five of these houses had fluorescent lights been installed
prior to 1971 in the kitchens while the other four contained only incandescent
lighting.
The data from the nine houses included in this study (Table 8) showed
some unexpected results. The PCB levels in the kitchens of these houses
ranged up to six times those found in offices. The kitchens also con-
sistently had higher concentrations than the other rooms of the same house.
3 3
The house with mean levels of 530 ng/m in the library and 620 ng/m in the
kitchen was sampled on two separate occasions and the results were consistent.
Even though 95% of the PCB present resembled Aroclor 1242 there did not
seem to be an obvious point source producing these levels. The average
level of PCB, again quantified as Aroclor 1242 plus Aroclor 1254, was
slightly higher for kitchens with fluorescent lighting, but this difference
was not statistically significant.
28
-------
The results of the air sampling show that indoor levels of PCB are
considerably higher than normal outdoor levels. They also show that the
levels in laboratories tend to be higher than those in offices. Since the
laboratory with covered fixtures had a level comparable to that in the
other laboratories and not to those in the offices, it appears that the
difference may not be related to the lights but instead to the equipment
present. The offices tested did not contain any electrical equipment which
might contain capacitors whereas all three of the laboratories did. It
might be noted that the ratio of Aroclor 1242 to Aroclor 1254 in the rooms
was the same as that for the burned out ballast. This may not be indicative
of the source, however, since Aroclor 1242 and Aroclor 1254, as well as
Aroclor 1221, have been used in the manufacture of capacitors. The similar
levels present in the houses with and without fluorescent lighting also
indicate that the other appliances and electrical equipment found in kitchens
rather than the lights may be contributing the PCB to the atmosphere.
The question of air circulation (type of air conditioning, if any)
might come up in trying to determine the difference in PCB levels. The
only house included in the study that did not have central air conditioning
was the one in Lumberton. The kitchen of this house contained PCB concen-
trations slightly below the median for the other kitchens monitored.
29
-------
Table 7
Decline of Airborne PCB in a Room After Burnout of Light Ballast
3 3
Day Aroclor 1242 ng/m Aroclor 1254 ng/m
0 = burnout 5740 5860
31 530 730
61 250 249
,95 116 111
122 64 64
283 119 93
334 114 98
377 73 88
30
-------
CO
Table- 8
Airborne PCB Levels in Private Dwellings
House
Gary
house 1
house 2
house 3
house 4
Lumberton
Durham
house 1
house 2
house 3
Raleigh
All quantities are calculated as Aroclor 1242 plus Aroclor 1254
3 333
Room ng/m Room ng/m Room ng/m Room ng/m
kitchen
kitchen
*kitchen
*kitchen
kitchen
kitchen
*kitchen
kitchen
*kitchen
476
184
250
212 living room 39
241 attic <50 bedroom 174 basement 121
146 garage 64
576
255
620 library 530
*Indicates rooms with fluorescent lighting.
-------
SECTION V
LANDFILLS
Since publication of Nisbet and Sarofim's 1972 paper, several investi-
gators have suggested that landfills and dumps are potential emission
10 12 20
sources for PCB. ' ' Soil, groundwater and leachate have been shown to
21
be contaminated with PCB in and around landfills, but it was not until
1977 that data was published confirming emissions into the air from landfills.
14
Murphy and Rzeszutko collected two samples on the same day from one
landfill and found that the mean concentration of PCB was 3240 ng/m .
These samples were taken at the vent pipes of a sealed above ground landfill
in the highly industrialized Chicago area. In November of the same year,
22
Hetling et aj. sampled the air above two landfills that were dump sites
for General Electric capacitor facilities near the Hudson River. The
3 3
ambient levels at these sites ranged from 930 ng/m to 5900 ng/m calculated
as Aroclor 1016. These results indicate that some landfills do indeed emit
very high concentrations of PCB. On the other hand, landfills in New
3
Bedford, Massachusetts which contain over 22 x 10 kg of PCB were emitting
3
only 21 ng/m in January 1978 while the ground was frozen and light snow
was falling. The previous summer (June 1977) 400 to 1300 ng/m of PCB had
been measured at the same location. As Nisbet and Sarofim had suggested,
the lighter Aroclors seemed to predominate at these sites.
32
-------
Landfills for predominantly residential areas that contain household
trash and debris instead of industrial wastes have not been monitored up to
this time. Many of the former uses of PCB (plasticizers, flame retardants,
adhesives, inks, etc.) would point to their disposal at least in limited
quantities in this type of landfill. This is also the most abundant type
of landfill found in the area under consideration.
The present Durham, North Carolina city landfill was opened in
December 1974 and was the only one in the study still in operation at the
time of sampling. The old city landfill, one of three closed ones studied,
was located on property adjacent to the present one. It was in operation
from 1960 until 1974 when the new one was opened. The Raleigh, North
Carolina landfill, which ceased operation in 1973, was next to the state
mental hospital and directly across the highway from the state penitentiary
in downtown Raleigh. The air samples at this site were taken at a spot
that was known to have been leaking methane. The only landfill surveyed
that had known PCB dumping was the was the Old Demolition Site in Goldsboro,
North Carolina. The PCB waste material (Aroclor 1254) from the transformer
manufacturer surveyed in this study was dumped here. The locations of
these landfills are shown in a series of maps (Figures 5-7).
The data collected at the landfills are shown in Table. 9. Also shown
are data taken at two sites in Durham and Wake Counties that were used as
controls. As can be seen, the PCB concentrations in the air above these
landfills are slightly higher than those in the control area, but are
orders of magnitude less than those in the Chicago and Hudson River studies.
In contrast to the measurements made by Hetling et aj., Aroclor 1254 was
almost as abundant as the more volatile Aroclors.
33
-------
The landfill in Goldsboro did not show higher concentration of PCB
than the other landfills, and actually had the lowest ambient concentration
of all four landfills studied. The actual area of PCB dumping in the
landfill could not be determined, so it is possible that the areas sampled
toward the front half of the 200 m long landfill were not where the dumping
occurred. (An automobile demolition operation prevented the sampling of
the back half of the area.) The low results were still unexpected, considering
the magnitude of PCB found by the other investigators.
The Raleigh landfill had at least one third more Aroclor 1242 than any
other location. There is not enough data available to determine whether
this is significant or just a matter of sampling variation. The Wake
County levels of Aroclor 1242 are higher than the other locations, but the
two control locations are actually closer together than either is to the
landfills.
34
-------
Figure 5. Locations of Durham Landfills sampled.
35
-------
\
Figure 6. Location of Raleigh City Landfill sampled.
36
-------
/ SEYMOUR JOHNSON
f AIR FORCE BASE
Figure 7. Location of Old Demolition Site, Goldsboro, N.C.
37
-------
Table 9
PCB Concentrations in Air Over Landfills
Landfill
Durham, Durham Co.
Old Durham, Durham Co.
Old Raleigh, Wake Co.
Old Demolition Site, Wayne Co.
Control
RTP, NC (Durham Co.)
Shadowfax Farm (Wake Co. )
3
ng Aroclor 1242/m
9
11
18
4
4
12
3
ng Aroclor 1254/m
5
10
9
8
9
2
All samples were collected in March 1977, except those from Wayne Co., which
were collected in October 1978. All samples were taken 50 cm above the
ground with the intake horizontal to the ground.
38
-------
SECTION VI
ELECTRICAL SUBSTATIONS
Apparently the only monitoring near an electric power substation up
23
until this time was done by Environmental Science and Engineering, Inc.
in northern Florida in 1975. Their sampler was approximately 100 m from
the substation but only 20 m from a transformer storage facility. No
mention was made in their report concerning orientation of sampler, sub-
station and storage facility or of wind direction so there was no way to
discriminate between emissions from the substation and the storage facility.
The results were reported by chemical species (mono-, di-, trichlorobiphenyl,
etc.) instead of as a particular Aroclor mixture but by adding the concen-
o
trations reported, the total appears to have approximated 52 ng/m of
o
Aroclor 1016. This is higher than the 3-36 ng/m (calculated as decachloro-
biphenyl) that the same investigators reported in urban Jacksonville,
Florida, air.
Five electrical substations located in Durham County, North Carolina,
were chosen to be part of this survey. Even though only about 5% of all
transformers contain PCB, those that do are located primarily around buildings
12
or transportation facilities. The substations studied were chosen because
of the proximity of buildings or heavily traveled highways. The five sites
are indicated on the map shown below.
39
-------
For this survey, one sampler was located upwind of the substation
being monitored and another sampler was located downwind. Both samplers
were placed outside the fence protecting the transformers and capacitors
and so were approximately 30-40 m from them. At several of the sites,
samples were lost on different occasions due to failure of the generators
or pumps, thus reducing the amount of useful data collected. The data that
was obtained are shown in Table 10.
The results indicate that there is no large scale leakage of PCB from
these particular transformers and capacitors. This does not preclude power
substations as emission sources, however, since the number sampled for this
survey represents an almost infinitesimal percentage of the total number.
40
-------
Figure 8. Locations of electrical substations sampled.
41
-------
Table 10
PCB Data from Electrical Substations
Upwind air Downwind air
Concentration, Concentration,
Site Date (ng/m3) (ng/m3)
A
B
C
D
E
5-5-77
5-12-77
6-10-77
5-31-78
6-1-78
6-9-78
7-12-78
6-10-78
7-13-78
8-24-78
7-31-78
8-29-78
8-1-78
9-26-78
b
16
10
b
c
11
c
1
c
42
10
8
c
19
47
8
17,18
25
17
8
14
32
41
21
20
10
18
6
aAll amounts are calculated as Aroclor 1242 + Aroclor 1254.
All samples were taken 50-200 cm above the fround with the intake
horizontal to the ground.
bSamples not taken.
cSamples lost due to equipment failure.
42
-------
SECTION VII
MANUFACTURING SITE
For some time it has been recognized that the sites of manufacture of
both PCB and PCB-containing products have the potential for introducing
large quantites of these chemicals into the environment. Since 1969,
14 24 25 26
investigators ' ' ' have been reporting measurements of PCB concen-
trations in the soil and water surrounding these plants and have confirmed
that they do release large quantities of PCB. Lake Hartwell in South
27 22 28
Carolina and the Hudson River in New York ' have both been grossly
contaminated by factories on their shores. In 1976 and 1977, measurements
were made of PCB levels in the air around the facilities on the Hudson
River. These facilities ceased using PCB in June 1977 and almost immediately
the PCB levels dropped significantly. While PCB was being used, the ambient
o
air levels ranged from 500 to 1700 ng/m . The average levels from July to
3 22
November 1977 were below 400 ng/m .
Hevi-Duty Electric, a subsidiary of Solar Basic Industries, located in
Goldsboro, North Carolina, is one of the thirteen manufacturers of trans-
formers in the United States today. They have been making and repairing
large transformers at their present site since 1968. By 1971, they had
begun to realize the threat of PCB to the environment and to the health of
their employees and to control much more carefully the storage, use and
43
-------
disposal of the large quantities of askarel used in the production of their
products. Prior to 1971, they had stored the askarel in drums in an open
area behind the plant. Empty drums were also kept in this area and spills
in the plant area were cleaned up with rags which were then thrown into
these drums. The waste material from the plant was disposed of in the Old
Demolition Site, a landfill run by the city. This landfill is now buried
and no PCB emissions from it were detected in this work.
Since 1971, the company has tried to reduce the contamination of PCB
to a minimum. The askarel is now kept in a closed system under a nitrogen
atmosphere. The soil in the old storage area which was found to contain
large amounts of PCB, was dug up to a depth of 0.6 m, encased in plastic
and buried in another spot on company property.
For this study, air samples were taken over a 16 month period in the
area of the old PCB storage site and at 10 other sites in the vicinity of
the plant. These sites are shown in Figure 9. The ideal situation would
have been to collect samples at every site at the same time; however,
equipment limitations prevented this. During sampling, measurements were
made of soil and air temperatures. A wind rose of the area, Figure 10, was
obtained from Seymour Johnson Air Force Base 6 km away, and 17 surface soil
samples were collected to determine dispersal of contamination from the
site and whether it could be correlated to wind direction. Tables 11
through 13 show the analytical results while Table 14 gives the meteoro-
logical data.
44
-------
The data indicate that there is still a large amount of PCB present
even after the major cleanup attempts at the site of the original contami-
nation. As far as 150 to 300 m to the north and east (Sites C and H, Table
11) the aerial concentrations of Aroclor 1254 was in the 100 ng/m range.
At the same time, two other sites (D and K) approximately 300 m distant
showed an order of magnitude less PCB, approaching background levels found
elsewhere. The only air samples taken further than 300 meters were on
private property about 500-600 meters to the east. The levels here all
also approached "normal" levels. A series of samples taken vertically over
Site A from 0.02 m to 1.8 m above ground showed no appreciable difference
in PCB levels with height. This is probably due to the widespread contami-
nation of the area, soil, air, and perhaps interior of the building nearby.
No correlation could be made of PCB concentration with air or soil
temperature using the amount of data generated during this study. The
levels varied greatly at the individual sites near the plant, but this may
have been due to the eddying effect of the wind close to the building or to
gusting on certain days.
The soil samples collected away from the immediate plant area contained
larger concentrations of Aroclor 1254 than had been expected, but this was
also true of the one control sample (#18). All of the values were at least
one order of magnitude lower than those reported for similar distances from
pc
Monsanto's Aroclor production plant at Sauget, Illinois.
Generally, the further from the plant site the lower the PCB level in
the soil. The major exception was #17, which was the Old Demolition Site
where PCB wastes from the plant had been dumped previously. Two other
45
-------
sites were also higher than the other samples taken at comparable distance:
the sewage pumping station, site #7, which also had higher aerial levels
than the other site the same distance from the plant, and site #3, which
was located at the corner of the driveway used to deliver chemicals to the
plant.
Overall, the results indicate little or no contamination of the air or
soil beyond the property boundaries that can be traced to the askarel used
by the company.
46
-------
NOTE: THE SITES ARE INDICATED
BY THE AVERAGE PCB
CONCENTRATIONS FOUND
(ng/m3). COMPARE TO
DATA IN TABLE 11.
Figure 9. Ambient sampling sites in the vicinity of a transformer manufacturer.
47
-------
CALM
19.5
PERCENT
7.5
7.2
SCALE, percent
Figure 10. Annual wind rose for Goldsboro, N.C. (Mean wind speed in knots
is given at barb ends, frequency of occurrence is given by barb length.
48
-------
NOTE: THE SITES ARE INDICATED BY
THEIR PCB CONCENTRATIONS
(ng/g). COMPARE TO TABLE 13.
Figure 11. Soil sampling sites in the vicinity of a transformer manufacturer.
49
-------
X
C3
UJ
X
UJ
a.
UJ
<
UJ
cc
10
TIME, minutes
15
20
A. AMBIENT AIR EXTRACT FROMSITE H, 5/d INJECTION FROM 100 ml.
B. SOIL EXTRACT FROM SITE 9, 5 Ml INJECTION FROM 180 ml.
Figure 12. Chromatograms of extracts of samples collected near a transformer
manufacturer, SE-30/OV-210.
56
-------
Table 11
Aerial Concentrations of PCB in the Vicinity of a Transformer Manufacturer
All samples were taken approximately 50 cm above the ground and horizontal to it
Date
4/14/77
4/15/77
5/6/77
6/15/77
6/16/77
6/17/77
8/2/77
8/5/77
3/13/78
3/15/78
3/17/78
X =
A B
2,700a
1,900 2,000
1,500
430
520
2,900
1,200
790 3,700
3,400
740
1,600 2,800
C
900
39
78
790
390
40
560
300; 170
360
D
67
12
72
17
18
37
E F G H I J
12
20
2,800
3,400
210 410 720 160
340 87 1,400 120 5,900
180 39 1,600
240 180 2,000 140 5,900 16
K
25
25
aAll values are ng/m Aroclor 1254
-------
Table 12
PCB Concentrations as a Function of Distance from Ground Level
Site A, August 8, 1978
m above ground pg/m Aroclor 1254
0.02 2.5
0.30 2.2
0.60 2.6
1.20 2.9
1.80 1.9
52
-------
Table 13
Concentrations of PCB in Soil
in the Vicinity of a Transformer Manufacturer
Site
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17b
18b
m from
source
0
150
150
200
300
300
300
, 300
500
500
500
500
600
800
800
800
4,000
80 km
direction
from source
__
E
SE
S
S
SW
N
NW
E
S
W
NE
NW
SE
W
N
NE
ng/g Aroclor 1254
17,800
96
610
60
19
50
839
26
23
27
17
136
33
101
48
42
173
74
a01d Demolition Site - landfill
Investigator's backyard, Raleigh, N. C.
53
-------
Table 14
Meteorological Data During Sampling
at a Transformer Manufacturer
Date
Cloud
Cover
Wind
Speed
km/hr
Wind
Direction
Soil
Temperature
°C
Ambient
Temperature
°C
4/14/77
4/15/77
5/6/77
6/15/77
6/16/77
6/17/77
8/2/77
8/5/77
3/13/78
3/15/78
3/17/77
CLR
CLR
CLR
SCT
HAZY
OVC
SCT
BKN
OVC
SCT
SCT
3-8
3-8
3-8
0-3
2-6
0-10
11-18
13-21
0-16
6-29
from ENE
20°-60°
60°
130°-150°
230°-290°
220°-260°
80°-110°
270°-320°
270°-320°
30-39
33-40
12-16
10-20
12-16
•x-32
27-35
21-27
17-28
22-29
23-30
28-33
13-17
11-20
4-10
54
-------
SECTION VIII
PCB SPILL
During the period June 29, 1978 to August 8, 1978, 340 km of highway
in 12 different counties of North Carolina had chemical wastes systematically
spilled along them. Analysis of soil, plant and water samples by both the
U. S. Environmental Protection Agency and the North Carolina Department of
Human Resources indicated that the substance was predominantly Aroclor
1260.
On August 11, 1978, air samples were taken up and downwind at the site
of a spill on State Road 1346 in Chatham County. The next day another
spill site, on NC 210 in Johnston County, was sampled. At this spill,
which had been covered with sand the previous week, the air on-site and in
a house 30 m from the spill site was monitored. On August 16, 1978, a
third site, on US 421 in Chatham County, was monitored. At this location,
samples were collected at varying levels directly above the spill from 2 cm
up to 1.8 m.
The North Carolina Department of Transportation placed a layer of
charcoal covered by road tar over the spills in an attempt to hold down
contamination of the surrounding areas until a permanent solution to the
problem could be found. The Johnston County spill site was covered with
this mixture on August 24 and was revisited for sampling on August 26 and
September 9.
After much discussion of the various alternatives for handling the
cleanup of the spill sites, it was decided to try digging up a 0.5 m wide
55
-------
strip 10 cm deep and removing the dirt for disposal. Before this was done
on a large scale, a 1.5 km stretch of road was to be tested to determine
the feasibility of this method of removal or any hazard to residents of the
area or to the personnel doing the removal work. The site chosen was NC 58
in the vicinity of Inez, Warren County.
The personnel involved in the removal wore disposable coveralls,
boots, disposable plastic gloves, goggles and two stage (dust and organic
vapor) respirators. Some of the workers also carried MSA personal monitors
with PVC filters and polyurethane foam to trap dust and PVC vapors from
their breathing zones during the dig.
In preparation for the practice dig, both indoor and ambient air at
the site was monitored on September 19 and October 4. On October 5, 1978,
the contaminated soil was removed and transported to a temporary holding
area where it was encased in four layers of heavy duty polyethylene plastic.
For the removal to proceed with as little dust as possible being generated,
a hydroseeder wet down the area just before the motorgrader removed the
®
dirt. The dirt was then collected by an Athey conveyor loader which
dumped it into trucks for transport to the holding area. The sweeper,
preceded by another hydroseeder, swept any remaining dirt on the road back
into the trench and the trench was refilled with uncontaminated soil from
another location.
During this operation, MSA samplers were monitoring three of the
buildings along the route of the digging operation. High volume air samples
were also taken in the fields 30 m on either side of the road and at the
holding area. One week after the dig, October 12, 1978, another set of
56
-------
ambient and indoor air samples was collected. All of the results are shown
Tables 16-18.
There is definitely an increase in aerial Aroclor 1260 concentration
in areas very close to the spills along the roads. The house in Johnston
County and at least one of the buildings in Warren County also had elevated
levels. There is no real indication that the charcoal/tar treatment reduced
the amount of Aroclor 1260 escaping into the air, through the amount seemed
to go down with time since the spill. Monitoring at the dig site showed no
increase of airborne PCB due to the digging. Indications are that there is
no danger from air pollution to the inhabitants of the area from a cleanup
operation of this type.
Personnel breathing zone measurements are given in Table 19. The
highest levels of PCB to which the workers were exposed were less than the
3
1.0 [jg/m PCB for an eight to ten hour day allowed under the proposed NIOSH
4
criteria for the workplace.
57
-------
INDICATES SAMPLING SITES
Figure 13. Area of North Carolina containing PCB spills along highways.
58
-------
NORTH
START
DIG
INEZ
WIND
DIRECTION
DURING
DIG
Figure 14. Location of the practice dig near Inez, N.C.
59
-------
O
UJ
I
1U
a.
ui
oc
I
I
10 15 20
TIME, min
25
30
A. EXTRACT FROM PCV FILTER WORN BY 2ND HYDROSEEDER DRIVER DURING PCB SPILL CLEANUP,
1.3 ml, 5 Ml INJECTION
B. EXTRACT FROM FOAM PLUGS WORN BY 2ND HYDROSEEDER DRIVER DURING PCB SPILL CLEANUP,
1.4 ml, 5 Ml INJECTION
C. AROCLOR 1260, 0.50 ng, 5jul INJECTION
Figure 15. Chromatograms of personal monitoring sample extracts and Aroclor 1260
standard, SE-30/OV-210.
60
-------
I-
x
UJ
Q.
LU
UJ
cc
10
15
20
25
30
35
TIME, min
A. SAMPLE EXTRACT OF ROADSIDE DIRT AND ROCKS FROM NC~87~iN CHATHAM COUNTY, lOg sample
in 1.08 x105 ml, 5 Ail INJECTION
B. AROCLOR 1260,0.50ng^5//I INJECTION .
Figure 16. Chromatograms of spill sample extract and Aroclor 1260 standard, OV-101.
61
-------
LU
I
LU
O.
UJ
cc
I
I
I
10 15 20
TIME, min
25
30
A. EXTRACT OF SAMPLE COLLECTED OVER SPILL ON NC 210 ON 8/11/78,12 ml. 5 /zl INJECTION
B. AROCLOR 1260,0.50 ng, 5 M> INJECTION
Figure 17. Chromatograms of Aroclor 1260 standard and ambient sample extract from
highway PCB spill site, SE-30/OV-210.
62
-------
Table 15
Aroclor 1260 Data from Roadside Spill Sites
Site Date Location (jg/m Aroclor 1260
SR 1346
NC 210
NC 210
NC 210
8/11/78 directly over spill
6.1 m upwind
12.2 m upwind
7.6 m downwind
15.2 m downwind
8/12/78 " directly over spill
30 m downwind
46 m downwind
8/26/78 directly over spill
7.6 m upwind
15.2 m upwind
7.6 m downwind
15.2 m downwind
9/9/78 directly over spill
7.6 m upwind
15.2 m upwind
7.2 m downwind
15.2 m downwind
10.8
0.22
0.01
0.54
0.24
4.8
0.04
0.01
2.1
0.28
0.46
0.05
0.03
0.39
0.06
0.17
0.07
0.05
aThese upwind sites were located next to a driveway. There could be a
secondary source of PCB emission further upwind causing these elevated
levels.
63
-------
Table 16
Profiles of Ambient Aroclor 1260 Concentration Over Spills
Date
Location
Distance above ground surface
0.02 m 0.3 m 0.6 m 1.2 m 1.8 m
Comments
8/16/78 Chatham Co. 57
(US 421)
10/4/78 Warren Co. 0.9
(NC 58)
10/12/78 Warren Co. 0.05
(NC 58)
11/7/78 Warren Co. 0.14
2.1 3.2 0.52 0.45 no charcoal
0.09 0.02 0.01 <0.01 charcoal,
before dig
0.14
<0.01 after dig
0.05 <0.01 0.05 0.03 after dig
All amounts are in ug/nT
64
-------
Table 17
Indoor Levels of Aroclor 1260 Near Highway Spill Sites
Date
8/12/78
8/26/78
9/19/78
10/4/78
10/5/78
10/12/78
Location*
Johnston Co. kitchen
bathroom
bedroom
Johnston Co. kitchen
Warren Co. Inez Country Store (3)
Thome house (4)
Thorne barn (5)
Thompson Grocery (1)
Farm building (7)
Thompson house (2)
Warren Co. Fleming house (6)
Warren Co. Inez Country Store (3)
Fleming house (6)
Farm building (7)
Warren Co. Inez Country Store (3)
Fleming house (6)
Farm building (7)
ug/m
0.19
0.04
0.08
0.07
0.01
<0.01
<0.01
0.10
0.01
0.02
0.10
0.01
<0.01
<0.01
<0.01
<0.01
<0.01
*The numbers refer to locations indicated on Figure 14.
65
-------
Table 18
Airborne PCB Concentrations Near Test Removal Site
in Warren County, North Carolina
Site
A
B
-
C
D
Dump site 1
Dumps ite 2
Date
10-4-78
10-4-78
10-12-78
10-4-78
10-5-78
10-12-78
10-4-78
10-5-78
10-12-78
10-4-78
10-5-78
10-12-78
10-4-78
10-5-78
10-12-78
10-4-78
10-5-78
10-12-78
ug/m
_ _
0.064
0.010
—
0.011
0.013
0.022
0.007
0.008
0.014
0.023
0.006
0.024
0.034
0.012
0.011
0.014
0.007
aData courtesy M. D. Jackson, Health Effects Research Laboratory, Research
Triangle Park, North Carolina.
66
-------
Table 19
Levels of Aroclor 1260 in Breathing Zones of Workers
During Test PCB Spill Removal
3
Position ug/m
1st hydroseeder operator <0.4
Motorgrader operator <0.1
Athey Loader operator <0.1
Dump truck drivers - 1 <0.4
- 2 <0.4
- 3 0.15
- 4 0.14
Dump truck coverer 0.10
2nd hydroseeder operator 0.57
Sweeper operator 0.82
Dumpsite worker 0.34
Supervisors - 1 <0.1
- 2 0.23
67
-------
SECTION IX
REFERENCES
1. 0. Hutzinger, S. Safe, and V. Zitko, The Chemistry of PCB's, CRC
. Press, Cleveland, Ohio, 1974.
2. R. L. Durfee, "Production and Use of PCB's in the United States."
Paper presented at the National Conference on Polychlorinated Biphenyls,
Chicago, Illinois, November 19-21, 1975.
3. Anon.. New Scientist. 32:612 (1966).
4. NIOSH, Recommended Criteria for Occupational Exposure to Polychlorinated
Biphenyl, DHEW (NIOSH) Publication No. 77-225, Washington, D. C.,
1977.
5. R. Risebrough, B. DeLappe, and Walker, in Marine Pollutant Transfer
(H. L. Windom and R. A. Duce, eds.), D. C. Heath and Company, Lexington,
Massachusetts, 1976.
6. E. P. Savage, J. D. Tessari, J. W. Malberg, H. W. Wheeler, and J. R.
Bagby, Pest. Monitoring J., 7:1 (1973).
7. S. A. Bengtson and A. Sodergren, Ambio., 3:84 (1974).
8. T. F. Bidleman, C. P. Rice, and C. E. Olney, in Marine Pollutant
Transfer (H. L. Windom and R. A. Duce, eds.), D. C. Heath and Company,
Lexington, Massachusetts, 1976.
68
-------
9. T. F. Bidleman and C. E. Olney, Bull. Environ. Contam. Tox., 11:442
(1974).
10. V. E. McClure, Environ. Sci. and Tech., 10:1223 (1976).
11. C. T. Nisbet and A. F. Sarofim, Environ. Health Perspectives, 1:21
(1972).
12. B. Fuller, J. Gordon, and M. Kornreich, Environmental Assessment of
PCBs HI the Atmosphere. Report No. EPA 450/3-77-045, U. S. Environ-
mental Protection Agency, Research Triangle Park, N. C., 1977.
13. C. L. Haile and E. Baladi, Methods for Determining the Total Poly-
chlorinated Biphenyl Emissions from Incinerators and Capacitor- and
Transformer-Filling Plants. Report No. EPA 600/4-77-048, U. S.
Environmental Protection Agency, Research Triangle Park, N. C., 1977.
14. T. J. Murphy and C. P. Rzeszutko, Polychlorinated Biphenyls HI
Precipitation i_n the Lake Michigan Basin, Report No. EPA 600/3-78-071,
U. S. Environmental Protection Agency, Duluth, Minnesota, 1978.
15. R. G. Lewis, A. R. Brown, and M. D. Jackson, Analytical Chemistry,
49:1668 (1977).
16. J. F. Thompson, ed., Analysis of Pesticide Residues HI Human and
Environmental Samples, Pesticides and Toxic Substances Effects Laboratory,
U. S. Environmental Protection Agency, Research Triangle Park, North
Carolina, 1974.
17. W. N. Billings, T. F. Bidleman, and C. G. Simon, Precision of Air
Sampling and Analysis for Chlorinated Hydrocarbons Using Solid Adsorbants,
presented at 29th Southeastern Regional ACS Meeting, Tampa, Florida,
1977.
69
-------
18. J. E. Kaufman, ed., Lighting Handbook, 4th Ed., Illuminating Engineers
Society, New York, 1966.
19. D. C. Staiff, G. E. Quinby, D. L. Spencer, and H. C. Starr, Jr., Bull.
Environ. Contam. Tox., 12:455 (1974).
20. T. F. Bidleman, personal communication.
21. United States Environmental Protection Agency, Analytical Quality
Control Laboratory, Cincinnati, Ohio, unpublished data.
22. L. Hetling, E. Horn, and J. Tofflemire, Summary of Hudson River PCB
Study Results, NYS Department of Environmental Conservation, Albany,
N. Y., 1978.
23. C. L. Stratton, S. A. Whitlock, and J. M. Allan, A Method for the
Sampling and Analysis of Polychlorinated Biphenyls (PCBs) i_n Ambient
Air, Report No. EPA-600/4-78-048, U. S. Environmental Protection
Agency, Research Triangle Park, N. C., 1978.
24. T. W. Duke, J. I. Lowe, and A. J. Wilson, Jr., Bull. Environ. Contam.
Tox.. 5:171 (1970).
25. G. D. Veith and G. F. Lee, Water Research, 5:1107 (1971).
26. C. L. Stratton and J. B. Sosebee, Jr., Environ. Sci. and Tech., 10:
1229 (1976).
27. W. N. Billings, T. F. Bidleman, and W. B. Vernberg, Bull. Environ.
Contam. Tox., 15: 215 (1978).
28. R. J. Nadeau and R. A. Davis, Bull. Environ. Contam. Tox., 16:436
(1976).
70
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/4-79-022
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
SOURCES OF EMISSIONS OF POLYCHLORINATED BIPHENYLS INTO
THE AMBIENT ATMOSPHERE AND INDOOR AIR.
5. REPORT DATE
March 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Kathryn E. MacLeod
1GANIZATION NAME AND ADDRESS
5. PERFORMING ORGANIZATION NAME M
Analytical Chemistry Branch
Environmental Toxicology Division
Health Effects Research Laboratory
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
1EA615
RTP, NC
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Health Effects Research Laboratory
Office of Research and Development
US Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-600/11
15. SUPPLEMENTARY NOTES
16. ABSTRACT
TRACT
Polychlorinated biphenyls (PCB) have been identified in air samples from many parts
of the world since the 1960s. This study was undertaken to identify and compare differ-
ent sources of PCB in indoor and outdoor air. All sampling was performed in central
North Carolina. The suspected sources that were tested were fluorescent light ballasts,
landfills, electrical substations, a transformer manufacturer, and the sites of illegal
dumpings.
Defective light ballasts emit large quantities of PCB and are an important indoor
source. Capacitors in small electrical equipment may also be an important source. In
general, indoor air levels of PCB were at least one order of magnitude higher than out-
door levels. The.data indicate that the landfills and electrical substations tested are
not major sources of PCB. The transformer manufacturer had elevated levels of PCB in
the immediate area of the plant but did not contribute greatly to the levels found off
the property. The spill sites also had elevated levels of the contaminant in their
immediate area, but the levels 50-100 m away were normal for rural areas.
KEY WORDS AND DOCUMENT ANALYSIS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI field/Group
DESCRIPTORS
jolychlorinated biphenyls
indoor air
air monitoring
07, b,d
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
21. NO. OF PAGES
78
20. SECURITY CLASS (Thispage)
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
71
------- |