EPA-450/3-77-045
November 1977
ENVIRONMENTAL
ASSESSMENT OF PCBs
IN THE ATMOSPHERE
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/3-77-045
ENVIRONMENTAL ASSESSMENT
OF PCBs IN THE ATMOSPHERE
by
H. Fuller, J. Gordon, and M. Kornreich
The Vlilre (Corporation
McLean, Virginia
Contract No. 68-02-1495
EPA Project Officer: Donald Loke>
Prepared for
I'.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
November 1977 '^
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35), Research Triangle Park, North Carolina
27711; or, for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161 .
This report was furnished to the Environmental Protection Agency by
The Mitre Corporation, McLean, Virginia, in fulfillment of Contract
No. 68-02-1495. The contents of this report are reproduced herein as
received from The Mitre Corporation. The opinions, findings, and
conclusions expressed are those of the author and not necessarily those
of the Environmental Protection Agency. Mention of company or product
names is not to be considered as an endorsement by the Environmental
Protection Agency.
Publication No. EPA-450/3-77-045
-------
v
ABSTRACT
Polychlorinated biphenyls (PCBs) are highly toxic compounds whose
large scale use over the years in numerous and diverse applications
has resulted in their ubiquitous distribution throughout the environ-
ment. Since 1970, domestic sales of PCBs have been restricted to closed
system electrical applications where no suitable replacement as an in-
sulating fluid has been found. Imported PCBs are, however, presently
used for dispersive applications, especially by the investment casting
industry.
Loss of PCBs to the environment may occur during their production,
transport, storage, incorporation into products, use and disposal. Due
to evaporation and codistillation, loss to the atmosphere is possible
whenever loss to the hydrosphere or lithosphere takes place. The high-
est levels of PCBs in both the atmosphere and hydrosphere occur close
to urban and/or industrial centers. PCBs have been discovered in areas
far removed from their points of origin and atmospheric transport repre-
sents the major pathway by which PCBs are disseminated throughout the
world.
Although human exposure to PCBs is mainly through the diet, PCBs
will be inhaled whenever they exist in the atmosphere. They are readily
absorbed by both routes and accummulate in body tissue. In mammals,
the major site of toxic action is the liver, although severe dermal
effects, sensory nerve involvement, teratogenicity and carcinogenicity
have also been reported.
iii
-------
Although the possibility of accidental release to the environment
will continue to pose a threat as long as PCBs remain in use, the
implementation of available control technology and proper disposal
methods can minimize such losses.
ACKNOWLEDGEMENTS
We gratefully acknowledge our colleagues L. Thomas, R. Ouellette
and J. Golden for their many helpful suggestions and criticisms. In
addition, we wish to thank M. Jones and D. Lokey of the U.S. Environ-
mental Protection Agency for their support and assistance.
iv
-------
TABLE OF CONTENTS
PAGE
CONCLUSIONS, RECOMMENDATIONS, AND SUMMARY 1-1
A. Conclusions and Rec onunend at ions 1-1
B. Summary 1-4
(ll) PHYSICAL AND CHEMICAL PROPERTIES 2-1
A. Summary
B. General Description and Physical Properties 2-2
C. Specific Chemical Properties 2-16
III. METHODS FOR SAMPLING AND ANALYSIS OF POLYCHLORINATED 3-1
BIPHENYLS IN THE ATMOSPHERE
A. Summary 3-1
B. Introduction 3-2
C. Sampling for PCBs in the Atmosphere 3-4
1. Liquid Absorption Methods 3-4
2. Liquid Phases on Solid Supports 3-9
3. Solid Adsorbents 3-16
D. Extraction and Cleanup Procedures 3-20
E. Qualitative and Quantitative Analysis 3-22
F. Confirmation 3-26
SOURCES 4-1
A. Summary 4-1
B. Production and Use 4-2
1. Production 4-2
2. Import and Export 4-9
3. Uses and Replaceability 4-10
C. Losses to the Environment 4-24
1. Losses During Production, Transport, and 4-24
Storage
2. Losses During Use 4-25
3. Disposal 4-34
Vy MEDIA DISTRIBUTION, TRANSFORMATION, MD TRANSPORT 5-1
A. Summary 5-1
B. Atmosphere 5-2
C. Hydrosphere and Lithosphere 5-14
v
-------
TABLE OF CONTENTS (CONTINUED)
PAGE
D. Transport at Interfaces 5-35
1. Air/Water Interface (including precipitation) 5-35
2. Air/Soil Interface 5-46
3. Soil/Water Interface (including sediment) 5-50
E. Transport Models 5-62
CONTROL OF PCBs IN THE ATMOSPHERE 6-1
A. Summary 6-1
B. Introduction 6-2
C. Control of PCB Losses at Production Facilities 6-3
and by Consumer Industries
1. Methods for Control of Worker Exposure 6-3
2. Methods for Control of External Emissions 6-6
D. Control of PCB Losses in Use 6-12
1. Losses to Atmosphere 6-12
2. Losses to Lithsophere 6-13
3. Losses to Hydrosphere 6-16
E. Control of PCB Losses in Disposal 6-17
1. Losses to Atmosphere 6-20
2. Losses to Lithosphere 6-22
3. Losses to Hydrosphere 6-23
F. Recommendations in Control Strategies 6-24
1. Losses to Atmosphere 6-25
2. Losses to Lithosphere 6-25
3. Losses to Hydrosphere 6-26
IVII. ENVIRONMENTAL EFFECTS OF PCBs 7-1
A. Summary 7-1
B. Effects on Humans and Laboratory Animals 7-1
1. Absorption and Accumulation 7-1
2. Metabolism and Elimination 7-5
3. Lethality 7-6
4. Toxic Effects on Liver 7-6
5. Effects on Epithelium 7-8
6. Hematological Effects 7-13
7. Nervous System Effects 7-14
8. Carcinogenesis 7-15
9. Teratogenesis and Other Reproductive Effects 7-16
C. Other Environmental Effects 7-18
1. Introduction 7-18
2. Microorganisms 7-18
3. Aquatic Organisms 7-19
vi
-------
TABLE OF CONTENTS (CONCLUDED)
PAGE
4. Birds 7-20
5. Mammalian Wildlife 7-21
6. Role of Trace Contaminants in Toxicity of 7-23
Commercial PCB Mixtures
VIII. REFERENCES 8-1
vii
------- OCR error (C:\Conversion\JobRoot\0000062L\tiff\2000MDPW.tif): Unspecified error
-------
LIST OF TABLES (CONTINUED)
TABLE NUMBER PAGE
4.5 PCB MANUFACTURE AND SALES BY CATEGORY 4-14
4.6 PCB RESIDUES (PPM) IN SILAGE 4-28
4.7 PCB CONCENTRATIONS IN INDUSTRIAL EFFLUENTS 4-39
4.8 CONCENTRATION OF PCBs IN MUNICIPAL SEWAGE 4-40
TREATMENT PLANT OUTFALLS
4.9 PCB CONCENTRATIONS IN WISCONSIN TRMTMENT 4-42
PLANT EFFLUENTS
4.10 PCB CONCENTRATIONS IN SLUDGE AND EFFLUENT 4-45
FROM AN EXPERIMENTAL BIOLOGICAL OXIDATION
SYSTEM
4.11 CONCENTRATIONS OF PCB (|J.g/l as Aroclor 1254) 4-48
IN SOME RIVER SYSTEMS
4.12 CONCENTRATION OF PCBs IN SEWAGE SLUDGES 4-50
5.1 AVERAGE MONTHLY LEVELS OF PCB IN AIRBORNE 5-8
FALLOUT COLLECTED IN SOUTH SWEDEN (ng nT2
month~l)
5.2 CONCENTRATIONS OF POLYCHLORINATED BIPHENYLS 5-11
IN MARINE AND CONTINENTAL AIR
5.3 PCB CONCENTRATIONS OVER THE WESTERN NORTH 5-13
ATLANTIC
5.4 ESTIMATED ANNUAL MASS EMISSION RATES OF 5-17
POLYCHLORINATED BIPHENYLS TO THE SOUTHERN
CALIFORNIA BIGHT (kg/yr)
5.5 SUMMARY OF PCB RESIDUE DATA FOR SURFACE AND 5-23
GROUND WATER, JANUARY 1971 - JUNE 1972
5.6 CONCENTRATION AND ENRICHMENT FACTORS OF 5-27
POLYCHLORINATED BIPHENYLS (AS AROCLOR 1254)
IN SURFACE MICROLAYER SAMPLES FROM NARRAGAN-
SETT BAY, RHODE ISLAND
ix
-------
LIST OF TABLES (CONCLUDED)
TABLE NUMBER PAGE
5.7 POLYCHLORINATtD BIPHENYLS (AS AROCLOR 1260) 5-29
IN SARGASSO SURFACE MICROLAYER (SI-I) AND
SUBSURFACE WATER (SS)
5.8 CONCENTRATIONS OF PCBs IN NORTH ATLANTIC OCEAN 5-31
WATER
5.9 THE FATE OF 0.1 ppm POLYCHLORINATED BLPHENYL 5-41
(AROCLOR 1260) IN 150 ml WATER SAMPLES FROM
THE FRASER RIVER, THE NICOMEKL RIVER, AND
GEORGIA STRAIT HELD IN THE LABORATORY FOR '
12 WEEKS AT THE GIVEN TEMPERATURE
5.10 EVAPORATION PARAMETERS AND RATES FOR VARIOUS 5-44
AROCLORS AT 25°C
5.11 FATE OF 0.1 ppm AROCLOR 1260 IN 150 ml WATER 5-54
SAMPLES FROM FRASER RIVER AND FROM GEORGIA
STRAIT IN THE PRESENCE OF BOTTOM SEDIMENTS
FROM THE SAME SOURCES, INCUBATED AT 13°C
5.12 POLYCHLORINATED BIPHENYL CONCENTRATIONS 5-56
(parts per billion (ppb) of the dry weight)
IN DATED SEDIMENTS
5.13 PCB RESIDUES FOUND IN SAN FRANCISCO BAY 5-60
AREA STREAMS
5.14 SUMMARY OF PCB RESIDUE DATA FOR BOTTOM 5-61
SEDIMENTS, JANUARY 1971-JUNE 1972
5.15 GROSS ESTIMATES OF RATES IN INPUT AND ACCUMULA- 5-65
TION OF PCBs IN NORTH AMERICA IN 1970
5.16 PCB ACCUMULATION BETWEEN 1930 AND 1970 5-71
6.1 FACILITIES AND SERVICES FOR PCB DISPOSAL 6-18
7.1 SYMPTOMS OF YUSHO PATIENTS (89 MALES, 100 7-10
FEMALES, AS OF OCTOBER 31, 1968)
x
-------
LIST OF FIGURES
FIGURE NUMBER PAGE
1.1 ACKNOWLEDGED ROUTES OF PCBs INTO THE 1-9
ENVIRONMENT PRIOR TO 1970
1.2 PRESENT ACKNOWLEDGED ROUTES OF PCBs 1-11
INTO THE ENVIRONMENT
1.3 ESTIMATED MOVEMENT OF PCBs THROUGH THE 1-16
ENVIRONMENT
2.1 THE MOST COMMON SUBSTITUTION PATTERNS 2-5
FOR THE CHLOROBIPHENYLS FOUND IN PCB
PREPARATIONS
2.2
3.1
3.2
3.3
3.4
3.5
3.6
4.1
4.2
4.3
VAPOR PRESSURE OF DIFFERENT AROCLOR
PREPARATIONS
FILTER, IMPINGER, AND ADSORPTION SAMPLING
TRAIN
A TYPICAL FRITTED-GLASS BUBBLER USED IN
SAMPLING FOR GASES
RECOVERY OF AROCLOR 1242 FROM SPIKED GLASS
PLATE
HIGH VOLUME SAMPLER MODIFIED FOR HOLDING
SOLID MEDIA
COMPARISON OF ELECTRON CAPTURE CHROMATOGRAMS
FOR AROCLOR 1221, 1242, 1248, 1254, and 1260
PARTIAL MASS SPECTRA (GC-MS) OF AROCLOR 1254
SHOWING THE NUMBER OP CHLORINE ATOMS ON EACH
ION
U.S. DOMESTIC SALES OF PCBs BY GRADE
U.S. DOMESTIC SALES OF PCBs BY APPLICATION
HOURLY CONCENTRATIONS OF POLYCHLORINATED
2-14
3-6
3-8
3-15
3-19
3-25
3-29
4-7
4-16
4-43
BIPHENYLS IN THE INFLUENT AND EFFLUENT FROM
THE CEDARBURG, WISCONSIN, TREATMENT PLANT,
APRIL 15, 1971.
xl
-------
LIST OF FIGURES (CONCLUDED)
FIGURE NUMBER PAGE
5.1 THE LOCATION OF SAMPLING STATIONS FOR THE 5-7
DETERMINATION OF PCB RESIDUES IN AIRBORNE
FALLOUT
5.2 SAMPLING SITES OVER THE WESTERN NORTH 5-12
ATLANTIC
5.3 RELATIONSHIP OF ATMOSPHERIC PCB AND DISTANCE 5-15
FROM INDUSTRIAL SOURCES
5.4 CONCENTRATION OF PCB IN SOIL WITH DISTANCE 5-36
(ppm)
5.5 LOSS OF AROCLOR 1254 FROM ITSELF AS A FUNCTION 5-48
OF TIME
5.6 LOSS OF AROCLOR 1254 FROM AN OTTAWA SAND 5-49
5.7 PERCENT DECREASE IN THE CONCENTRATION OF 5-52
AROCLOR 1254 BY THE ADDITION OF INCREASING
AMOUNTS OF ADSORBENT
5.8 DEPOSITION OF PCB IN DATED SEDIMENTS OF THE 5-57
SANTA BARBARA BASIN
5.9 ENVIRONMENTAL TRANSPORT MODEL 5-64
xii
-------
I. CONCLUSIONS, RECOMMENDATIONS, AND SUMMARY
A. CONCLUSIONS AND RECOMMENDATIONS
Polychlorinated biphenyls (PCBs) are released to the atmos-
phere from a variety of sources. These include incomplete incinera-
tion and burning of PCB-containing wastes in municipal and private
incinerators as well as in garbage dumps; vaporization from plas-
ticizers, paints, and other PCB-containing coatings; vaporization
from soil, sand, and other terrestrial reservoirs; evaporation and
codistillation from natural bodies of water; and evaporation from
their own surfaces when these have been exposed to the open air as in
the case of accidental spillage, leaks, or wear and weathering of PCB-
containing products. Loss to the atmosphere is, therefore, ultimately
related to loss to other environmental compartments,and control of
the former necessitates control of the latter.
PCBs, upon release to the atmosphere, are adsorbed to particulate
matter and undergo transportation to areas far removed from their point
of origin. They are eventually redeposited to the lithosphere and
hydrosphere. As a result of aerial transport, an ever increasing
accumulation of PCBs in the atmosphere will not take place. However,
high atmospheric levels of PCBs may be expected to occur in urban
industrialized areas, especially near incinerator stacks or garbage
dumps. Unfortunately, only a limited amount of monitoring data is
available for such areas. PCS levels over Providence, Rhode Island,
in 1973 were as high as 9.4 nanograms (ng) per cubic meter (m^), or
1-1
-------
approximately 1 part per trillion (ppt) . However, levels on suspended
participates in four U.S. cities between 1968 and 1970 ranged from
27 to 230 micrograms (^g) PCBs per gram (g) of particulate, with a
mean concentration of 50 (og/g of particulate. Assuming a particulate
3 3
loading of 60 fig/m , this average amounts to 3 ng/m PCBs. The
extent to which release of PCBs into the atmosphere poses a threat
to the general urban population remains uncertain due to the scarcity
of data relating to the toxic effects of low-level PCB inhalation in
mammals. In one inhalation study, 1.5 milligrams per cubic meter
3
(mg/m ; 0.11 parts per million [ppra]) of Aroclor 1254 was sufficient to
produce noticeable changes in rat liver following exposure for several
weeks. The maximum atmospheric concentration per 8-hour workday allowed
by the American Conference of Governmental Industrial Hygienists (1973)
3
for Aroclor 1254 is 0.5 mg/m (0.037 ppm) , one-third the level at which
an effect was experimentally observed in rats. The standard for
3
Aroclor 1242 is presently set at 1.0 mg/m (0.094 ppm).
In view of the above stated facts, the following recommendations
appear warranted:
1. Extensive monitoring data must be obtained for ambient PCB
concentrations over urban areas. Emphasis should be placed on deter-
mination of levels near such possible sources as municipal incinera-
tors, private incinerators operated by PCB-using industries, and
garbage dumps.
1-2
-------
2. Since most tcucicity data relate to Ingestion of PCBs, studies
must be initiated that will elucidate, in detaJ], all toxic effects
of low-level inhalation in mammals. Only by combining ambient levels
with toxicity data may we come to a definite, meaningful conclusion
concerning the health hazards of present atmospheric PCB levels.
3. From what little is known about toxic effects at low atmos-
pheric concentrations, workplace standards should be reevaluated and
acceptable concentration levels should perhaps be lowered.
4. Until definite toxicity data becomes available, the latest,
most efficient control technology must be implemented, if necessary,
through legal actions. High temperature incinerators (operated above
2000°F) will completely degrade PCBs to benign substances. Scrubbers
should be required on those incinerators that do not reach the necessary
temperatures. Scrubbers that remove particulates as well as gaseous
effluents are available and are necessary. Mist eliminators (such
as the Brink Mist Eliminator used by Monsanto) may be installed in
vapor lines to condense and collect vapors in the plant. Carbon
adsorption system (such as VentSorb from Calgon) may be hooked up
to the vent of a storage tank or other point source and may, in
addition, be adaptable to ventilation systems.
Numerous techniques for minimization of release to the hydro-
sphere are also available and are discussed in detail in the report
that follows.
1-3
-------
5. Available substitutes for PCBs, especially in electrical
applications, should be considered. Two companies, Dow Corning and
Dow Chemical, claim to have developed suitable substitutes for PCBs
in transformers and capacitors, respectively. Dow Coming's substi-
tute available for transformers is a polydimethyl siloxane liquid
known as "200 fluid". It is thermally stable, less explosive than
PCBs but is slightly more flammable. The substitute to be used on
capacitors is an alkylated chlorodiphenyl oxide that Dow Chemical
hopes to commercialize by early 1976. Monsanto and General Electric
(G.E.) are also working on substitute materials. The G.E. product,
known as Econol, is based on a phthalate ester and is presently used
in G.E. exports to Japan where PCBs are now banned.
B. SUMMARY
Polychlorinated biphenyls (PCBs) are substituted derivatives of
the compound biphenyl in which anywhere from one to ten of the hydrogen
atoms have been replaced by chlorine. PCBs are synthesized commercially
as mixtures of isomers, with the largest proportion of component com-
pounds corresponding in composition to the average chlorine content of
the mixture. While most individual chlorobiphenyls are solids at room
temperature, the majority of the commercial mixtures are mobile oils.
The physical characteristics which render PCBs useful are their
thermal stability, non-flammability and excellent dielectric properties.
They are also resistant to acids, bases, oxidizing agents and other
chemicals unless treated under vigorous reaction conditions. PCBs are
1-4
-------
soluble in most of the common organic solvents but are only slightly
soluble in water. Aqueous solubility of both individual isomers and
mixtures, in general, has been found to decrease with increasing
chlorine content. Of the individual isomers studied, the one exhibiting
the greatest solubility in water was 2-monochlorobiphenyl (5.9 ppm).
The solubility of a commercial mixture containing 42 percent chlorine
by weight was 200 parts per billion (ppb). Vapor pressures are also
low and tend to parallel aqueous solubility with the preparations
containing a higher percentage of less chlorinated isomers being
most volatile. The vapor pressures of commerical mixtures containing
_o
32 percent and 54 percent chlorine by weight are 5 x 10 millimeters
of mercury (mm Hg), and 6 x 10 millimeters of mercury respectively.
PCBs tend to adhere to smooth surfaces and are thus present in the en-
vironment adsorbed to sediment and transported via airborne particulates.
One pathway by which they may be removed from the environment is
through photochemical decomposition in the presence of sunlight. Re-
ductive dechlorination is the major photochemical reaction. However,
in hydroxylic solvents or aqueous suspension, replacement by alkoxy-
or hydroxy groups also occurs. The ortho position is the preferred site
of attach in both instances. Evidence suggests that highly toxic chloro-
dibenzofurans are formed from 2-chlorobiphenyls under photochemical
conditions leading to the formation of oxygenated products. Since
chlorodibenzofurans are themselves photolabile, their accumulation
in the atmosphere as a byproduct of photochemical decomposition of
PCBs is unlikely.
1-5
-------
Sampling methods for atmospheric PCBs fall into three categories:
liquid absorption systems, liquid phases on solid supports, and solid
adsorbents. Liquid absorption systems are most popular, but are
only suitable when high concentrations are expected. Liquid phases
on solid supports are used primarily for static sampling when relative
levels of contamination rather than ambient concentrations are sought.
The most promising sampling material is polyurethane foam. Polyure-
thane foam is compatable with high volume collection, does away with
the need for interfering collection liquids, and appears relatively
specific for PCBs.
The most sensitive analytical method for PCBs is electron capture
gas chromatography. Unequivocal qualitative confirmation may be ob-
tained by mass spectrometry.
PCBs are manufactured commercially in many of the world's indus-
trialized nations, although production has been banned in Japan since
1972. The sole producer of PCBs in the United States is the Monsanto
Company, which markets its mixtures under the trade name "Aroclor."
A voluntary restriction of sales to closed system electrical applica-
tions was initiated by Monsanto in 1970, resulting in a 50 percent
decrease in production and sales by that company through 1974.
Prior to 1970, PCBs found use in a variety of applications which
relied upon their unique chemical and thermal stability and electrical
insulating properties. They were found, for example, in heat trans-
fer fluids, hydraulic fluids, high vacuum oils, epoxy paints, printing
1-6
-------
inks, sealers in water-proofing compounds, and many other types of
products. Their major use, however, was and continues to be as an
insulating fluid in capacitors and transformers. Following the dis-
covery of a wide range of adverse effects associated with the release
of PCBs into the environment, an attempt was made to curb their use
for dispersive applications through the aforementioned restriction of
sales. However, the possibility that PCBs might be obtained for such
applications from other sources (for example imports) continues to be
of significant concern to environmentalists, especially since little
evidence for a decrease in environmental levels of PCBs following the
sales restrictions were found. Some environmentalists (such as William
G. Turney of Michigan) feel that the only certain means of getting
PCBs out of the environment may be to ban them completely, and several
companies in anticipation of such action are seeking alternatives to
PCBs for all types of applications, including electrical. For example,
Dow Corning Company and Dow Chemical Company claim to have developed
suitable substitutes for PCBs as insulating fluids in transformers
and capacitors, respectively. Dow Coming's substitute available for
transformers is a polydimethyl solixane liquid known as "200 fluid."
It is thermally stable, less explosive than PCBs and only slightly
more flammable. The substitute to be used in capacitors is an alky-
lated chlorodiphenyl oxide which Dow Chemical hopes to commercialize
by early 1976.
1-7
-------
Loss of PCBs to the environment may occur during their production,
transport, storage, incorporation into products, use, and disposal.
Due to evaporation and codistillation, loss to the atmosphere is
possible whenever loss to the hydrosphere or lithosphere takes place.
The pathways by which PCBs entered the environment prior to 1970, and
by which they are believed to enter the environment at present are
summarized in Figures 1.1 and 1.2, respectively. Losses during pro-
duction at the Monsanto facilities are presently estimated to be in-
significant, resulting, for example, in a parts-per-trillion
concentration in the effluent reaching the Mississippi River. Losses
during transport and storage are mainly due to leaks and spills and
are also thought to be small. Release to the environment as a
consequence of use occurred maximally prior to 1970 and was, for the
most part, unintentional. The major pathways by which PCBs were lost
to the atmosphere, hydrosphere, lithosphere, and food supply during
use included vaporization from PCB-containing paints, coatings and
plastics; migration and leaching from surface coatings and packaging
materials; leakage from faulty heat exchange 'systems and partially
sealed hydraulic systems and burnout of PCB-containing ballasts in
fluorescent light fixtures.
The primary source of PCBs in all compartments of the environment
is undoubtedly the disposal of waste PCBs by Consumer industries and
of PCB-containing products by municipal treatment plants. Disposal
practices which lead to environmental contamination include: ( 1)
1-8
-------
(SCRAP HEAT A f
EXCHANGER J I
HEAT
EXCHANGER
iMNTAMINATED \
'RODUCT J
^ 26.7 x Id6 Jb-;.
' 70
13.9 x 106 Ibs.
70
)3.98 x 106 Ibs.
\A,t/
^ vaP
<^J
7.4 x 106 Ibs x
70
19.5 x 106 Ibs s
70 ^
0,70 ,
<10 x 103 Ibs.
69
1.6 x 106 Ibs.
70
.urce: Nisbet, I. C. T. and A. F. Saroflm, "Rates and Routes of
Transport of PCBs in the Environment," Environmental Health
Perspectives Exp. 1. 21-38, 1972. \W,T,A
PCB-CONTAINING_-NCINERATION
KEY
(^ J Reservoir
o
Flow Path
Route Into the air (A),
water (W), or terrestrial
(T) environment
FIGURE 1.1
ACKNOWLEDGED ROUTES OF PCBs INTO THE
ENVIRONMENT PRIOR TO 1970
-------
MANUFACTURE
OF PCBs
MANUFACTURE OF
PCB-PLASTICIZED
RODUCTS
PCB-CONTAINING
RODUCTS
disposal
eachln
open-burning
Remaining stock of PCB
containing products
(quantity unknown)
34.4 K 10 Ibs.
(1974)
FORMULATION
OF DIELECTRIC
FLUIDS
MANUFACTURE OF
CAPACITORS
Source: Adapted from Nisbet, I. C. T. and A. F. Sarofim, "Rates and
Routes of Transport of PCBs in the Environment," Environmental
Health Perspectives Exp. 1, 21-38, 1972.
Route into the air (A) ,
water (W), or terrestrial
(T) environment;
(D) - destroyed.
FIGURE 1.2
PRESENTLY ACKNOWLEDGED ROUTES OF PCBs
INTO THE ENVIRONMENT
-------
open burning or incomplete incineration of municipal and industrial
solid wastes; (2) allowing PCB-containing fluids to flow into water-
ways with municipal and industrial water water effluents; (3) dumping
of sewage sludge, solid waste, and dredge spoil at sea, and; (4) dumping
of sewage sludge and solid waste into sanitary landfills and dumps.
Since PCBs vaporize unless incinerated at temperatures greater
than 2000°F, loss to the atmosphere following destruction of PCB-
containing refuse at municipal incinerators is very likely. Although
no actual data on gaseous PCB emissions from municipal incinerator
stacks are available, residues were detected in two samples of fly ash
from such incinerators. Further studies dealing with incinerator
emissions are needed.
Discharge of PCB-containing effluents has resulted in serious
contamination of several rivers and coastal waters. General Electric
has admitted to dumping as much as 30 pounds of PCBs(.13.62 kilograms,
or kg) per day into the Hudson River, and effluents sampled from several
industrial plants throughout the United States contained concentrations
ranging from 2.5 to 275 parts-per-billion. Discharge into rivers from
selected municipal sewage treatment outfalls ranged from 0.002 pounds
(0.91 grams) to 213 pounds (96.7 kg) per day; municipal release has
recently been found to be the primary source of PCBs in the marine
ecosystem off Southern California. About 6.5 metric tons were dis-
charged during 1973-1974.
1-13
-------
As might be expected, the highest levels of PCBs in both the
atmosphere and hydrosphere occur close to urban and/or industrial
centers. In the atmosphere, PCBs may be found in the vapor phase, in
the form of an aerosol, or adsorbed onto particulate matter,, Levels
of PCBs on suspended particulate in four U.S. cities between 1968 and
1970 ranged from 27 to 230 ug/g of particulate with a mean concentra-
tion of 50 ug/g of particulate. Assuming a particulate loading of
3 3
60 ug/m , this average amounts to 3 ng/m PCBs. PCBs in dry aerial
—8
fallout within the Los Angeles Basin ranged from 7.8 x 10 gram/
— f\ 0
square meter/day to 1.7 x 10 g/m /day. PCBs, mainly in the vapor
phase, were detected over the western North Atlantic. Concentrations
were highest at sampling stations closest to the Boston-Hartford-
New York-New Jersey industrial complex but decreased 100 fold at a
station located 2000 kilometers (km) farther out to sea.
PCBs are found in the hydrosphere mainly adsorbed to particulate
matter and sediment. Due to their lipophilic nature, high concentra-
tions also tend to accumulate in the surface microlayer of the.ocean.
Due to low aqueous solubility, only small quantities are found natu-
rally in solution. Actual concentrations observed in some fresh water
systems range from 0.5 ppt for unpolluted waters (such as that: in the
GBto and Nordre Rivers in Sweden) to 15,800 ppt in the highly polluted
Great Miami River in Ohio. Concentrations as high as 4.2 parts-per-
billion (ppb) have been observed in the surface microlayer of Narragan-
sett Bay, Rhode Island. Subsurface water from the same location con-
tained 0.15 ppb.
1-14
-------
The discovery of significant concentrations of PCBs in the atmo-
sphere and hydrosphere far removed from urban industrial areas suggests
that the atmosphere represents the major pathway by which PCBs
are transported throughout the world. PCBs have also been detected
in precipitation (such as antarctic snow) in support of this hypothesis.
Analysis of accumulated evidence reveals a model in which the atmo-
sphere plays a central role. PCBs released directly into the atmosphere
adsorb onto particulates and are transported with the prevailing winds.
Transport in the vapor phase may also occur. Particle fallout and
precipitation deposit the PCBs onto land or into bodies of water.
Most PCBs reaching the hydrosphere are removed by particle scavenging
followed by sedimentation; a small amount remains dissolved and is
subject to re-evaporation. This phenomenon is particularly noticeable
in marine waters located in areas of high evaporation and low rainfall.
But evaporation from shallow, fast-flowing streams fed with PCB-con-
taining effluents may be significant enough to result in falsely
diminished estimates of hydrospheric contamination. Successive cycles
of evaporation, adsorption and deposition eventually carry the PCBs,
initially released on land, to the coast and to their ultimate sink,
the sediment of the oceanic abyss. A scheme depicting the movement
of PCBs through the environment including quantitative estimates of
the amounts of material involved at each step is presented in Figure
J.3. Due to differentia] vapor pressures, the majority of atmospheric
samples analyzed contain a higher proportion of lower-chlorinated
1-15
-------
Reilclence unest im,,tedd
tlmp
PARTICLE SCAVENGING AND
SEDIMENTATION
j x 10 tons/year*
(7.5 x 102 Cons/yr)
DREDGING AND DUMPING
20 tons/year
RKDISSOUITION
2 x 10 form/year"
«. Niabet and Saroflm (1972)
b. Biklenan and Olney (1974a).
c. baled on a half-life of ^ yearn.
d. dapenda on depth of water, rate of flow,
organic content, etc.
FIGURE 1-3
ESTIMATED MOVEMENT OF PCBi THROUGH THE ENVIRONMENT
1-16
-------
PCB isomers, whereas hydrospheric samples appear to be enriched in
the higher-chlorinated compounds. The latter effect is magnified
by the preferential biodegradation by microorganisms of tetrachloro
and lower isomers. It is therefore difficult to ascertain which
Aroclor was originally responsible for any given instance of pollution.
Although the possibility of accidental relen.se to the environment
will continue to pose a threat as long as FCBs remain in use, the
implementation of available control technology can minimize loss during
production, transport, storage, incorporation into products and dis-
posal. The use of closed or sealed systems is suggested at production
and industrial facilities wherever possible and also during ship load-
ing, unloading, storage and any transfer operations. Appropriate
materials and containers for all procedures involving handling of PCBs
are described in publications by Monsanto, the Institute of Electrical
and Electronics Engineers, and the American National Standards Insti-
tute. Release to the atmosphere may be controlled by a variety of
techniques. Monsanto, for instance, employs Brink Mist Eliminators
in all vapor lines to condense and collect vapors in the plant. In
addition, all air from the ventilation system is passed through a
high temperature incinerator (2000°F) which completely destroys the
PCBs. Calgon has developed an activated carbon adsorption system
known as "VentSorb" which may be hooked up to the vent of a storage
tank or other point source. It is possible that this unit may be
adapted to a ventilation system and adsorb PCBs from the air stream.
1-17
-------
The recommended method for disposal of waste PCBs and PCB-con-
taining products is through high temperature incineration (>2000°F).
Many disposal services which properly incinerate all suitable PCB-
containing waste materials are available to user industries for a fee.
General Electric has recently installed a special incinerator of its
own exclusively for liquid wastes (askarel). This incinerator operates
under a different principal requiring lower temperatures (1200° to
1800°F) and 50 to 100 percent excess oxygen and results in virtually
zero emission of PCBs to the atmosphere. Since municipal incinerators
probably do not operate under appropriate conditions for destruction
of PCBs and since they handle the unlabelled PCB-containing refuse
dispersed in the environment, these incinerators should be fitted
with scrubbers which can remove PCB vapors and PCB-containing particu-
lates prior to venting. A variety of such scrubbers is available.
Loss to the atmosphere can be reduced to acceptable levels only
if loss to the hydrosphere and lithosphere are controlled as well.
The primary pathway for loss to the hydrosphere is contamination of
sewers and subsequently of wastewater effluent. Ideally, all sewer
systems that may contain PCBs should be isolated from a plant's
effluent stream. Contaminated waste water may then be passed through
such devices as a carbon adsorption or solvent extraction system.
Calgon is planning a service called the Calgon Adsorption Service
which will install an activated carbon unit, monitor effluents, pre-
pare a monthly report, and guarantee a discharge in the parts-per-
1-18
-------
trillion range. Unfortunately, no methodology is presently available
which can guarantee zero discharge to the hydrosphere from a point
source effluent. This is an area where more research is essential.
The major loss to the lithosphere results from leaks and spills.
Prompt clean up, using appropriate absorptive materials is recommended.
Suitable materials include the bags, blankets, and boxes marketed by
Gedcor and produced from Dow "imbiber beads." These soak up the askarel
and allow it to be transported safely for proper disposal. Leakage
from malfunctioning transformers must also be contained, followed by
proper clean-up procedures. Placement of imbiber blankets in the con-
tainment areas under transformers is, for example, standard procedure
at Dow Chemical Company.
Adequate methodology is available for those industries wishing to
control the release of PCBs to the environment. There is, however, a
need for a single comprehensive publication describing this methodology
along with safeguards for worker protection, handling and shipping
procedures, techniques for clean-up of spills and proper disposal
practices. The EPA Oil and Hazardous Materials Spills Research
Bttanch in Edison, New Jersey, is now in the process of preparing such
a manual. A legal basis for compliance with any recommendations may,
however, be necessary in order to ensure cooperation.
Controls and emission standards for PCBs are necessary because
PCBs are highly toxic to most life forms, including man. Although
human exposure to PCBs is mainly through the diet, PCBs will be inhaled
1-19
-------
whenever they exist in the atmosphere. The compounds are readily
absorbed by both routes and tend to accumulate in body tissues. PCBs
are stored preferentially in those tissues having a high lipid content
and it is estimated that 41 to 45 percent of the general population
have PCB levels of 1 ppm or more in their adipose tissue. Ease of
metabolism seems to decrease with increasing chlorine content. The
lower-chlorinated isomers undergo hydroxylation followed by conjuga-
tion with glucuronic acid and excretion in the urine. The more highly
chlorinated isomers remain in the various tissues of. the body.
The 'major site of toxic action for PCBs is the liver. Effects
include increase in liver weight, induction of microsomal enzymes
(which play a role in metabolizing foreign substances), and structural
changes such as increases in lipid droplets within the cytoplasm and
in lysozyme content. Concentrations of Aroclor 1254 as low as 0.15 ppm
in the air were sufficient to produce noticeable changes in rat liver.
Ambient levels of PCBs in urban areas such as Providence, Rhode Island,
.are approximately 1.6 x 10 times lower than this; however, the limits
set for industrial exposure by the American Conference of Governmental
Industrial Hygienists are 0.05 ppm, or only three times lower. It is
obvious that more studies must be initiated which will elucidate the
effects of low concentrations of PCBs on the human liver. It is
suggested that the present permissible exposure levels be re-evaluated.
Dermal effects were among the earliest recognized signs of PCB
toxicity in industrial workers handling PCBs or PCB-containing products.
1-20
-------
An outbreak of PCB poisoning in Japan also resulted in widespread
skin disease. Symptoms included acneform eruptions, increased pig-
mentation of the skin, lips, gums, nails, and mucous membrane of the
mouth, itching of the skin and stiffening of the soles of the feet and
palms of the hands. Concentrations as low as 300 ppm in the entire
diet were sufficient to produce these symptoms in sub-human primates
within three months. Concentrations ingested with milk fat (28 ppm)
and fish (35 ppm) are only 10 times less than this. (Note these values
exceed current Food and Drug Administration [FDA] temporary tolerances.)
Gastrointestinal symptoms, including nausea and vomiting as well as
certain cytological changes indicative of eventual neoplastic trans-
formation, were also observed.
Other effects observed during the poisoning episode in Japan
included changes in blood composition and bone marrow histology and
sensory nerve involvement symptomized by numbness, pain, dulled sensi-
tivity to touch, and absence of reflexes. Teratogenicity was also
observed and took the form of abnormally early appearance of teeth,
calcification of the skull, wide fontanelles and abnormal protrusion
of the eyeballs. No direct instance of carcinogenesis in humans
attributable to PCBs exists, although the aforementioned possibly
pre-cancerous stomach lesions are suggestive of a carcinogenic poten-
tial for these compounds. Hepatocellular carcinomas have been produced
in mice exposed to 300 ppm Aroclor 1254 in the diet for 11 months.
1-21
-------
PCBs have also been found to be toxic to fish and birds and are
most likely toxic to mammalian wildlife. Reduced egg production,
hatching and survival of offspring are among the major effects observed
in non-mammalian organisms.
1-22
-------
II. PHYSICAL AND CHEMICAL PROPERTIES
A. SUMMARY
Polychlorlnated biphenyls (PCBs) are substituted derivatives
of the compound blphenyl In which from one to ten of the hydrogen
atoms have been replaced by chlorine. Synthesized commercially as
mixtures of isomers, PCBs are produced in the United States only by
the Monsanto Company, which markets them under the trade name Aroclor.
While individual chlorobiphenyls are solids at room temperature, the
commercial mixtures are predominantly mobile oils. Their thermal
stability, non-flammability and excellent electrical insulating prop-
erties give a wide variety of applications. Since PCBs are extremely
hydrophobic, they are soluble in most of the common organic solvents
and are only very slightly soluble in water. Aqueous solubility of
both individual isomers and mixtures in general decreases with in-
creasing chlorine content. Vapor pressures are also low and parallel
solubility, with the preparations containing a higher percentage of
less-chlorinated isomers being most volatile.
PCBs tend to adhere to smooth surfaces. This property, along with
their lipophilicity, is responsible for their presence in the environ-
ment adsorbed to sediment and transported via airborne particulates.
Polychlorinated biphenyls are resistant to oxidizing agents,
acids, bases, and a variety of other chemicals. They are, however,
susceptible to photochemical degradation at wavelengths equivalent to
those of ultraviolet (U.V.) light (-300 nanometers, or nm). The major
2-1
-------
photochemical reaction is reductive dechlorination with ortho chlorines
preferentially replaced. In hydroxylic solvents or aqueous suspension,
replacement by alkoxy or hydroxy groups also occurs. Evidence suggests
that, under photochemical conditions leading to the formation of oxy-
genated products, toxic chlorodibenzofurans may be formed from 2-chloro
biphenyls.
B. GENERAL DESCRIPTION AND PHYSICAL PROPERTIES
Polychlorinated biphenyls are substituted derivatives of
the compound biphenyl (I) in which anywhere from one to ten of the
hydrogen atoms have been replaced by chlorine.
32 6' 5'
Although 209 chlorinated biphenyl isomers are theoretically possible,
from both a statistical and mechanistic standpoint it is unlikely
that all are formed in the technical chlorination process. Certain
isomers, such as 2, 3, 4, 5, 6-pentachlorobiphenyl for instance, where one
ring is fully chlorinated and the other not at all, would be highly
disfavored (Hutzinger tt_ al. , 1974) .
PCBs are prepared industrially by the chlorination of biphenyl
with anhydrous chlorine using iron filings or ferric chloride as
catalyst. The resulting product is a complicated mixture of isomers
2-2
-------
with the highest proportion of compounds corresponding in composition
to the average chlorine content. Domestic PCBs, produced by the
Monsanto Company, are marketed under the trade name Aroclor. Each
Aroclor mixture is characterized by a four digit number in which the
first two digits, 12, indicate chlorinated biphenyl and the last two
give the weight (w/w) percent of chlorine in the mixture. An exception
to this nomenclature is a recent product, Aroclor 1016, which contains
41 percent chlorine by weight but reduced quantities of the penta-,
hexa-, and heptachlorobiphenyl isomers. Typical percent compositions
of some Aroclor mixtures are presented in Table 2.1. The most common
substitution patterns for the chlorobiphenyls found in PCB preparations
are presented in Figure 2.1. Due to the ready availability of the
Aroclor mixtures and their wide use as such, most determinations of
physical properties are performed on Aroclor as opposed to individual
PCB isomers. Some important physical properties of the various
Aroclor mixtures are presented in Table 2.2. One major difference
between the commercial mixtures and the individual isomers is their
solidification points. Most individual chlorobiphenyls are solids at
room temperature, whereas commercial mixtures are mobile oils
(Aroclor 1221, 1232, 1242, and 1248), viscous liquids (Aroclor 1254),
or sticky resins (Aroclor 1260 and 1262) due to the mutual depression
of melting points by the component isomers (Hutzinger e£ al., 1974).
The physical characteristics which render PCB mixtures useful are
their thermal stability, non-flatnmability, and excellent dielectric
2-3
-------
TABLE 2.1 TYPICAL PERCENT COMPOSITION OF SOME POLYCHLORINATED BIPHENYL PRODUCTS*
NJ
I
-P-
HOMOLOG
# Cl/BIPHENYL
0
1
2
3
4
5
6
7
8
— — "i
AROCLOR
1221
11
51
32
4
2
<0.5
ND
ND
ND
AROCLOR
1016
<0.1
1
20
57
21
1
<0.1
ND
ND
AROCLOR
1242
<0.1
1
16
49
25
8
1
<0.1
ND
AROCLOR
1254
<0.1
<0.1
<0.5
1
21
48
23
6
ND
*Percent (w/w) by GC/M.S (Gas Chromatography/Mass Spectrometry).
ND = none detected, <0.01%
Source: Hutzinger, 0., S. Safe, and V. Zitko, "The Chemistry of PCBs," CRC Press,
Cleveland, Ohio, 1974.
-------
* Only one phenyl-ring is shown. The most abundant tetrachlorobiphenyls,
for example, are those from the dichlorophenyl-moleties shown. One di-
and one trichlorophenyl- would give most abundant penta-chlorobiphenyl
etc.
Source: Hutzinger, 0., S. Safe, and V. Zitko, "The Chemistry of PCBs,"
CRC Press, Cleveland, Ohio, 1974.
FIGURE 2.1
THE MOST COMMON SUBSTITUTION PATTERNS FOR THE
CHLOROBIPHENYLS FOUND IN PCB PREPARATIONS*
2-5
-------
Table 2.2 PHYSICAL PROPERTIES OF VARIOUS AROCLOR MIXTURES
Average Molecular
weight
% Cl
Average number of
Cl'iaoleclue
Appearance
Specific gravity
Density
(Ibs/gal, 25<=C)
Distillation range
°C, corrected
(ASTM D-20, mod)
Evaporation loss, 2
100°C, 6 hours
163°C, 5 hours
(ASTM D-6, mod)
Vaporization rate
(g/cm2/hr)* at
100°C
(surface area -
12.28 cm2)
1221
192
20.5-21.5
1.15
clear
mobile
oil
1.182-
1.192
(25°C/
15.5°C)
9. 55
275-320
1.0-1.5
— b
1.74xl(f3
1232
221
31.5-32.5
2.04
clear
mobile
oil
1.270-
1.280
(20°C/
15.5"C)
10.55
290-325
1.0-1.5
8.74xlO~4
1242
261
42
3.10
clear
mobile
oil
1.381-
1.392
(25°C/
15.5°C)
11.50
325-366
0-0.4
3.0-3.6
3.38xlO~4
1248
288
48
3.90
clear
mobile
oil
1.405-
1.405
(65°C/
15.5°C)
12.04
340-375
0-0.3
3.0-4.0
1.52xlO~'4
1254
327
54
4.96
light
yellow
viscous
fluid
1.495-
1.505
(65°C/
15.5°C)
12.82
365-390
0-0.2
1.1-1.3
5.3xlO~5
1260
372
60
6.30
light
yellow
soft
sticky
resin
1.555-
1.566
(90°C/
15.5°C)
13.50
385-420
•0-0.1
0.5-0.8
9xlO'6
1262
389
61.5-62.5
6.80
light
yellow
sticky
viscous
resin
1.572-
1.583
(90°C/
15.5°C)
13.72
390-425
0-0.1
0.5-0.6
1.3xlO~5
1268
453
68
8.70
white
to off
white
powder
1.804-
1.811
(25°C/
25°C)
15.09
435-450
0-0.06
0.1-0.2
NR
1016
NR
41
NR
clear
mobile
oil
1.362-
1.372
(25°C/
15.5°O
NR
323-356
NR
XR
NR
*g/cm /hr = gram square centimeter/hour
-------
TABLE 2.2 (Continued)
Flash point
(Cleveland open
cup) °C
Fire point
(Cleveland open
cup"* °C
Refractive index,
v°
20°C
Viscosity,
seconds
Saybolt universal
(ASTM D-83)
37.8°C
54.48C
98.9°C
1221
141-150
176
1.617-
1.618
38-41
35-37
30-31
1232 _,
152-154
238
1.620-
1.622
44-51
39-41
31-32
1242
176-180
none to
boiling
point
1.627-
1.629
82-92
49-56
34-35
1248
193-196
none to
boiling
point
1.630-
1.631
185-240
73-80
36-37
1254
none to
boiling
point
none to
boiling
point
1.639-
1.641
1800-2500
260-340
44-48
1260
none to
boiling
point
none to
boiling
point
1.647-
1.649
3200-4500
72-78
1262
none to
boiling
point
none to
boiling
point
1.650-
1.652
600-850
(at 71°C)
86-100
1268
none to
boiling
point
none to
boiling
point
1016
170
none to
boiling
point
1.622-
1.624
(|325°C)
71-81
NR
NR
= not applicable
NR = not reported
ASTM = American Society for Testing Materials.
Source: Modified from Hutzinger, 0., S. Safe, and V. Zitko, "The Chemistry of PCBs," CRC Press, Cleveland,
Ohio, 1974.
-------
(electrical insulating) properties. The flash and fire points are
included in Table 2.2; the electrical properties of some Aroclors are
presented in Table 2,3.
V The most important physical properties of PCBs from an environmental
point of view are solubility and vapor pressure. Like the chlorinated
hydrocarbon insecticides, PCBs are extremely hydrophobic and would
therefore be expected to occur in lipid phases such as the organic
films sometimes found on the surface of natural waters (Crosby and
Moilanen, 1973). The oils and resins are readily soluble in most of
the common organic solvents whereas the crystalline isomers and mixtures
such as Aroclor 1268 are less soluble (Hubbard, 1964). All are in-
soluble in glycerol and the glycols (Hubbard, 1964; Hutzinger et al.,
1974). The solubilities of 21 PCB isomers in water as determined by
WallnBfer et_ al. (1973) are presented in Table 2.4. More recent
determinations (Haque and Schmedding, 1975) on four of these isomers
and one additional isomer are also given. It can be seen from this
table that the water solubilities of all PCB isomers are low and, in
general, decrease with increasing chlorine content. The values reported
by Haque and Schmeddimg are somewhat lower than those of WallnBfer et
al. These authors attribute this discrepancy to failure on the part
of the earlier workers to allow sufficient time for equilibrium: to
take place (WallnBfer e_t^ _al_. allowed only 2 hours, whereas complete
equilibrium, according to Haque and Schmedding, requires at least one
month). Thus, aggregates of PCBs which would have made their way to
2-8
-------
TABLE 2.3 ELECTRICAL PROPERTIES OF SOME AROCLORS
AROCLOR
1232
1242
1248
1254
1260
1268
10168
DIELECTRIC CONSTANT
Q
AT 1000 CYCLES
25°C 100°C
5.7 4.6
5.8 4.9
5.6 4.6
5.0 4.3
4.3 3.7
2.5
NR 4.9d
VOLUME
RESISTIVITY,
8 cm AT 100° C
500 V, de
above 500x10^
above 500x10^
above 500x10*
above 500x10
oe
above 500x10
DIELECTRIC
STRENGTH ,CkV
>35
>35
>35
>35
>35
POWER FACTOR,3
100°C, 1000
CYCLES, %
<0.1
<0.1
<0.1
<0.1
NRf
NJ
I
ASTM D-150-47T
bASTM D-257-46
°ASTM D-149-44
60 cycles
Inferred from measurement at 500 VDC, 100°C, 0.1" gap
Same as for 1242 at 60 hz
^Values for Aroclor 1016 obtained from Monsanto specification sheet
Source: Hutzinger, 0., S. Safe, and V. Zitko,
"The Chemistry of PCBs." CRC Press, Cleveland, Ohio, 1974.
-------
TABLE 2.4 SOLUBILITY OF CHLOROBIPHENYLS IN WATER
COMPOUND
SOLUBILITY (ppm)
(Wollnb'Fer et al. , 1973) (Ilaque am! Scbmeo'ding, 1975)
Monochlorobiphenyls
2-
3-
4-
Dichlorobiphenyla
2,4-
2,2'-
2,4'-
4,4'-
Trichlorobiphenyls
2,4,4'-
2',3,4-
2,2',5-
Tetrachlorobiphenyls
2,2',5,5'-
2,2',3,3'-
2,2',3,5'-
2,2')4,4I-
2,3',4,4'-
2,3',4',5-
3,3',4,4'-
Pentachlorobiphenyls
2,2',3,4,5'-
2,2',4,5,5'-
Hexachlorobiphenyl
2,2',4,4',5,5'-
Octachloroblphenyl
99' TT' AA' S S '-
^ » ^ » J » J iH,H ,J)J
necachlorobiphenyl
5.9
3.5
1.19
1.40
1.50
1.88
0.08
0.085
0.078
0.046
0.034
0.170
0.068
0.058
0.041
0.175
0.022
0.031
0.0088
0.0070
0.015
0.637 + 0.004
0.248 + 0.004
0.0265 t 0.0008
0.0103 + 0.0002
0.000953 + 0.00001
Sources: Haque.R- and Schmedding, D., "A Method of Measuring the Water
Soluoility of Hydrophoblc Chemicals: Solubility of Five Poly-
chlorinated Biphenyls." Bulletin of Env. Contamin. and Tox.
_14_, 13-18 (1975).
WoLlnHfer, P.R., Koniger, M., and llutzinger, 0., "The Solu-
bilities of Twenty-One Chlorobiphenyla in Water", Analab Res.
Notes 13, 14-16 (1973).
2-10
-------
the surface, may have been sampled along with dissolved material. In
addition, it is suggested that the water in the earlier study may have
been contaminated with organic impurities, falsely increasing the
effective concentration of dissolved PCB. The solubilities of some
Aroclor mixtures are presented in Table 2.5.
Another determination of the solubility of Aroclor 1254 in water
( Haque e£ al. , 1974) revealed a value of 56 ppb at room temperature, in
good agreement with the previous study. Once again, those mixtures
richer in the less-chlorinated species exhibit a greater degree of solu-
bility. Table 2.6^summarizes the relative gas chromatography (GC) peak
heights of a saturated aqueous solution of Aroclor 1254 (Hutzinger et^
al.. 1974). It can be seen that the first few peaks, which, in general,
correspond to the lesser-chlorinated isomers, are significantly higher
than the equivalent peaks in the standard Aroclor 1254. Determinations
of the solubilities of PCBs in water are complicated by the fact that
these compounds tend to adsorb to smooth surfaces such as glass,
metal, varnished or lacquered surfaces, and plastic (Hubbard, 1964;
Hutzinger e± al., 1974). This tendency, along with their lipophilicity,
is responsible for the ultimate accumulation of PCBs in sediment and
their transport through the environment via airborne particulates.
The vapor pressure of several Aroclor preparations determined over
a wide range of temperatures is presented in Figure 2.2 and the actual
values at 37.8°C are listed in Table 2.7. Vapor pressures parallel
2-11
-------
TABLE 2.5 THE SOLUBILITIES OF AROCLOR MIXTURES, 20°C
AROCLOR
1242
1248
1254
1260
SOLUBILITY
200
100
40
25
(ppb)
Source: Panel on Hazardous Trace Substances, "Polychlorinated
Biphenyls — Environmental Impact." Environmental
Research 5: 249 (1972).
2-12
-------
TABLE 2.6 RELATIVE GAS CHROMATOGRAPH PEAK HEIGHTS (PEAK 5 - 100) IN
SATURATED AQUEOUS SOLUTIONS OF AROCLOR 1254
PEAK NO.
j
2
"3
4
5
6
7
8
9
10
11
12
13
SATURATED AQUEOUS
SOLUTION (26°C)
172
91
47
14
100
33
57
5
21
8
4
11
6
SATURATED AQUEOUS
SOLUTION (4°C)
144
72
41
9
100
28
59
5
24
13
4
24
10
AROCLOR 1254 STANDARD
35
16
30
1
100
23
55
10
25
31
6
50
11
Source: Hutzinger, 0., S. Safe, and V. Zitko, "The Chemistry of
PCBs." CRC Press, Cleveland, Ohio, 1964.
2-13
-------
EC
g
8!
Aroclor 1260
1254
1248
1242
Aroclor 1232
1242
1248
1254
10 -
10
-7
300
Source: Hutzinger, 0., S. Safe, and V. Zitko, "The Chemistry of PCBs,"
CRC Press, Cleveland, Ohio, 1974.
FIGURE 2.2
VAPOR PRESSURES OF DIFFERENT AROCLOR PREPARATIONS
2-14
-------
TABLE 2.7 APPROXIMATE VAPOR PRESSURE OF AROCLOR PREPARATIONS AT 37.8°C
Aroclor 1232
Aroclor 1242
Aroclor 1248
Aroclor 1254
0.005 mm. Hg
0.001 mm. Hg
0.00037 mm. Hg
0.00006 mm. Hg
Source: Monsanto, "Aroclor Plasticizers," Monsanto Technical
Bulletin 0/PL-306A, undated.
2-15
-------
solubility in that the preparations containing a higher percentage of
the less-chlorinated isomers are more volatile. When 1 mg of Aroclor
1254 is heated with 300 ml of water (conditions of codistillation),
the earlier GC peaks corresponding to the less-chlorinated isomers
are seen to decrease more rapidly than those peaks representing the
more highly chlorinated compounds (Table 2.8). This difference is,
however, much smaller when the same amount of Aroclor is heated in the
absence of water (Hutzinger _e_t al. , 1974) . Other data bearing on the
volatility of Aroclors, such as the percent loss through evaporation at
100°C ( 6 hours) and 163°C ( 5 hours) and the vaporization rate in grams/
2
square centimeter/hour (g/cm /hr) at 100°C for a surface area of 12.28
2
cm , are given in Table 2.2.
Studies relating the pertinent physical properties of solubility
and volatility to occurrence (either potential or actual) in the environ-
ment will be discussed in the appropriate subsection of Section IV.
C. SPECIFIC CHEMICAL PROPERTIES
Polychlorinated biphenyls are generally considered to be
inert compounds, which renders them extremely useful in a wide variety
of applications. Along with the previously mentioned properties of
thermal stability and non-flammability, PCBs also exhibit marked
resistance to oxidizing agents, acids, bases, and other chemicals.
Under vigorous conditions, however, certain reactions do take place.
Treatment of decachlorobiphenyl (II) with aqueous alkali in an
autoclave at high temperatures results in the formation of octachloro-
2-16
-------
TABLE 2.8 PERCENT LOSS IN AREA OF SEVEN CHROMATOGRAPH PEAKS OF AROCLOR
AROCLOR 1254 AFTER HEATING ON A STEAM BATH
AROCLOR 1254 PEAK
1
2
3
4
5
6
7
25 MIN
34
59
78
60
86
100
100
% PEAK REMAINING
WITH WATER
60 MIN
17
26
27
46
49
85
67
AFTER HEATING
WITHOUT WATER
10 MIN
13
15
20
20
27
28
16
Source: Hutzinger, 0., S. Safe, and V. Zitko, "The Chemistry of PCBs."
CRC Press, Cleveland, Ohio, 1974.
2-] 7
-------
4,4'-biphenylol (III) (Hutzlnger et .al., 1974). The 4,4'-methoxy
derivatives are also formed under similar conditions using sodium
methoxide or under milder conditions using sodium methoxide in pyridine.
These nucleophilic displacement reactions occur preferentially at the
para position, possibly due to the resonance stabilized intermediate
that results from para attack.
Cl Cl Cl Cl cl cl cl /C1
//VV // VL_ NaOH __//\\ /~~~\__ OH
r.rVi r.r Cl
Cl
Cl
l
autoclave
Although a mixture of chromic anhydride and acetic acid is able
to oxidize mono-, di-, and trichlorobiphenyls to the corresponding
chlorobenzoic acids, the oxidation of PCB mixtures with averages of 5
or 7 chlorine atoms per molecule requires boiling nitric acid. Potassium
permanganate, chromic acid, and nitric acid treatment under milder
conditions proved ineffective.
Reductive dechlorination of decachlorobiphenyl to yield nona- and
octa- chlorobiphenyls in which the 4 and/or 4' chlorine is replaced can
be effected by use of lithium aluminum hydride, butyl lithium and
water, or the Grlgnard reagent and water.
Nitration is smother possible reaction which occurs in both the
highly chlorinated and lesser chlorinated PCB'isomers. With the latter,
the extent of nitration may be controlled by varying the reaction
conditions.
2-18
-------
The last reaction to be mentioned is the cyclization of 2- and
2,2'-substituted biphenyls to yield highly toxic dibenzofuran deriva-
tives. The close proximity of the 2 and 2' substituents facilitates
cyclization under a variety of conditions. Some examples are shown
below.
Cl
"In view of the remarkable chemical stability of polychlorinated
biphenyls, environmental breakdown initiated by the photochemically
active part of the solar spectrum is of particular interest" (Hutzinger
et al., 1974).
Photochemical degradation is probably the sole pathway (except for
biodegradation) for breakdown of these compounds Ln the environment,
and several studies dealing with the irradiation of pure chlorobiphenyl
isomers at wavelengths above 290 nm (the lowest wavelength of the
2-19
-------
radiation generally received from the sun at the earth's surface) have
been reported. Irradiation of PCBs in the vapor phase or in the solid
state as a thin film adsorbed onto a glass surface has been attempted
but the results for the most part are inconclusive. Irradiation of
orthochlorobiphenyl at 300-310 nm under normal lower atmospheric
conditions resulted in the formation of quantitative amounts of HC1
along with unidentified aldehydes and phenolic compounds. One probable
product was orthohydroxy biphenyl (Arnts, 1975). The only truly
definitive experiments in which products have been identified are those
conducted in solution. Due to the low aqueous solubility of PCBs,
these experiments have been carried out in organic solvents. The
studies are, however, environmentally significant since organic hydrogen
donors are abundant in nature, existing in the form of oil films,
cuticular waxes, the surface microlayer of the ocean, and so forth.
It is in these "mixed" aqueous and organic compartments that PCBs are
most likely to reside. In general, reductive dechlorination is the main
initial photochemical reaction of PCBs in organic solvents (Hutzinger
_e_t _al. , 1974; Ruzo et^ jal., 3974). The ratr of the reaction is more ra-
pid in polar, hydroxylic solvents (alcohols) than in non-polar solvents
and, in the former case, substitution of chlorine by'an^Lkoxy group
is also observed. In a recent study, Ruzo et_ al. (1974) selected six
highly toxic tetrachlorobiphenyls and determined the products formed
following irradiation at 300 nm for 10 to 15 hours in both hexane and
methanol. The results are summarized in Table 2.9. Between 90 and 95
percent of the starting material reacted. Trichiorobiphenyls and di-
2-20
-------
TABLE 2.9 PHOTO PRODUCTS OF IRRADIATION OF SIX TETRACHLOROBIPHENYLS IN HEXANE AND METHANOL
POLYCHLORINATED BIPHENYL
DECHLORINATED PRODUCTS
METHOXYLATED PRODUCTS
NJ
NJ
2,2',5,5'-Tetrachlorobiphenyl
2,2',4,4'-Tetrachlorobiphenyl
2,2',3,3'-Tetrachlorobiphenyl
3,3',4,4'-Tetrachlorobiphenyl
3,3',5,5'-Tetrachlorobiphenyl
2,2' ,6, 6'-Tetrachlorobiphenyl
2,3',5-Trichlorobiphenyl
3,3*-Dichlorobiphenyl
3-Chlorobiphenyl
2,4,4'-Trlchlorobipheny1
4,4'-Dichlorobiphenyl
4-Chlorobiphenyl
2,3,3'-Trichlorobiphenyl
3,3'-Dichlorobiphenyl
3,4,4'-Trichlorobipheny1
4,4'-Dichlorobiphenyl
3,3',5-Trichlorobiphenyl
2,2',6-Trichlorobipheny1
2,2'-Dichlorobiphenyl
Trichloromethoxybiphenyl
Dichlorodimethoxybiphenyl
Trichloromethoxybiphenyl
Dichlorodimethoxybiphenyl
Trichloromethoxybiphenyl
Dichlorodimethoxybiphenyl
Trichloromethoxybiphenyl
Trichloromethoxybiphenyl
formed in both hexane and methanol
formed in methanol only
Source: Ruzo, L.O., M.J. Zabils, and R.D. Schuetz, "Photochemistry of
Photo products and Kinetics of Polychlorinated Biphenyls," J.
199-202 (1974).
Bioactive Compounds:
Agricul. & Food Chem. 22,
-------
chlorobiphenyls were the main reaction products in both solvents, with
monochlorobiphenyls constituting less than 1 percent of the total
product formation after 10 hours. In methanol, the amount of methoxy-
lated products (of which the primary components were trichloromethoxy
biphenyls) did not exceed 5 percent of total product formation.
Methanol substitution took place at the same site at which chlorine
was lost. Each tetrachlorobiphenyl that was irradiated containing
chlorines in the ortho position yielded products arising from the loss
of these, while those compounds containing only meta- and para-chlorines
lost the meta-chlorines preferentially. This pattern was also observed
in the formation of the secondary products, the dichloroblphenyls in
which ortho- and meta- but not para-chlorines were lost. The formation
of monochlorobiphenyls in yields of less than 1 percent or not at all
was most likely due to the failure of the dichlorobiphenyls to absorb
sufficiently at the wavelengtfts employed. It is suggested that the
^v
dechlorination reactions are free radical in nature whereas the methoxy-
lation reactions probably occur via nucleophilic attack by methanol
followed by elimination of HC1. Of particular environmental interest
is the fact that the reaction rates for dechlorination were found to
increase considerably when solutions were degassed prior to irradiation,
that is, when oxygen was removed from the system. The postulated
excited state for the reaction is known as a "triplet." Oxygen is
known to act as a triplet quencher by accepting excess electronic
energy from excited molecules before any chemical change occurs. Thus,
2-22
-------
greater photochemical decomposition may be expected to occur under
anaerobic conditions.
Because of the toxicity of chlorinated dibenzofurans, the possibil-
ity of their photochemical formation from PCBs must be considered. Kut-
zinger .et al.. (1973) have speculated that chlorodiberzofurans may be
formed from chlorobiphenyls under photochemical conditions leading to
the formation of oxygenated products—the formation of a chlorinated
2-biphenylol constituting a necessary intermediate. Although forma-
tion of polychlorinated 2-biphenylols has not been reported, Crosby
et_ al. (1973) found that polychlorinated phenols formed chlorinated
catechols and resorcinols following irradiation in aqueous media at
wavelengths equivalent to sunlight. Displacement of the chlorine
ortho to the oxygen predominated. Irradiation experiments with five
pure 2-chlorinated biphenyls (5 milligrams/liter [mg/1] aqueous sus-
pensions) revealed traces of 2-chlorodibenzofuran, although only the
2,5-dichloro- and 2,5,2',5'-tetrachlorobiphenyls provided identifiable
amounts (a roughly steady 0.2 percent yield during a 7-day irradiation).
Hutzinger et al. (1973) reported that preliminary studies suggest the
formation of chlorodibenzofurans from 2,4,6,2',4',6'-hexachlorobiphenyl
when irradiation takes place in methanol. Since chlorodibenzofurans
are themselves photolabile, it seems unlikely that accumulation of
these compounds formed from chlorobiphenyls by photochemical reaction
in the environment will occur. Nevertheless, due to their toxicity,
even their transient existence becomes important.
2-23
-------
-------
III. METHODS FOR SAMPLING AND ANALYSIS OF POLYCHLORINATED BIPHENYLS
IN THE ATMOSPHERE
A. SUMMARY
Methods of atmospheric sampling for PCBs generally fall into
one of three categories: liquid absorption systems, liquid phases on
solid supports, and solid adsorbents. Liquid absorption systems,
while popular, are suitable only when high concentrations of PCBs are
expected. Air flow rates are limited in these systems, and a sample
volume insufficient for detection of low-level PCB contamination is
obtained. Liquid phases on solid supports usually involve use of some
type of oil. Difficulties in removal of the oil from the sample, re-
sulting in subsequent interference with analysis, is one major drawback
to this type of system. The third type of methodology for atmospheric
sampling employs a solid adsorbent. One promising material is poly-
urethane foam. A high flow rate is achieved and no interfering oils
are necessary. Furthermore, the system is relatively selective for
PCBs.
Following extraction of the PCBs from the sampling medium, re-
moval of interfering compounds may be achieved by column chromatography.
Silicic acid columns separate PCBs from DDT and other pesticides.
Alumina removes additional electron capturing interferences. DDT may
also be removed by saponification followed by extraction with sulfuric
acid.
The most sensitive analytical procedure for PCBs is electron
capture gas chromatography. Qualitative verification may be achieved
3-1
-------
by use of perchlorination or thin layer techniques. Mass spectrometry,
when used in conjunction with gas chromatography, provides unequivocal
confirmation of the presence of PCS residues and is the best procedure
for characterizing the molecular composition of the observed residues.
B. INTRODUCTION
An assessment of the hazards presented by polychlorinated
biphenyls in the atmosphere can only be as accurate as the methods
employed for their sampling, detection, and quantItation. Although
sophisticated analytical techniques for PCBs are available, the diffi-
culties inherent in atmospheric sampling continue to create a signifi-
cant obstacle to the establishment of an intensive, wide-range
monitoring program. The low concentrations of PCBs expected in the
ambient atmosphere far removed from their points of origin require
that a large volume of air be sampled. On the other hand, practical
considerations require that sampling must be completed within a
reasonable length of time. Furthermore, vapor phase PCBs, .as well
as those adsorbed to particulates, must be trapped efficiently since
both forms may contribute substantially to the total residues present.
In general, sampling of particulate-bound PCBs involves collec-
tion via filtration. Loss of PCBs from the filtered particulates may,
however, ensue with volatilization to the air stream the most: likely
pathway. The rate of air flow through the sampling apparatus, the
temperature, and the adsorptivity of the particulate surface are
factors which determine the extent to which such loss occurs (Seiber
jet a!. , 1975). The volatility of the Individual components would
3-2
-------
also affect their retention so that the PCB mixture finally subjected
to analysis might be significantly different from that initially
present in the atmosphere. Loss of small particulates through the
filter itself poses another problem. Use of glass cloth or glass
fiber filters capable of efficiently trapping particles as small as
0.1 micron in diameter have, however, minimized these losses (Seiber,
et al., 1975).
Trapping of PCB vapors is usually performed by liquid extraction
or absorption from the sampled air stream. Inaccuracies due to non-
representative sampling, incomplete extraction, and revolatilization
from the trapping medium frequently result.
In an effort to counteract these difficulties, attempts at
determination of "sampling efficiency" have been undertaken. These
attempts usually involve spiking the air with known amounts of PCBs
as dusts and/or vapors and determining the overall recovery. However,
difficulties in simulating natural sampling conditions render the
usefulness of such results questionable (Seiber ^t aJ., 1975).
Additional problems associated with the sampling of PCBs in the
atmosphere include: chemical degradation of the sampling medium;
interference of the sampling medium (or its degradation products) in
subsequent work up or analytical steps; difficulties with the quanti-
tative removal of PCBs from the sampling medium; and simultaneous
collection of interfering compounds (Seiber et al., 1975).
3-3
-------
The methods for sampling and determination of PCBs in the atmos-
phere are generally similar to those used for the analysis of pesticide
residues (Sherma, 1975). The entire operation involves the following
overall steps:
a. Sampling of the atmosphere
b. Extraction of the PCBs from the sample
c. "Cleanup" of the extract to remove interfering substances
d. Qualitative and quantitative, analysis
e. Confirmation by additional analytical techniques
While suitable sampling procedures for PCBs are still in the
experimental stages, methods for extraction, cleanup, and analysis are
fairly standardized.
C. SAMPLING FOR PCBs IN THE ATMOSPHERE
Presently, available sampling procedures generally fall under
one of three headings: liquid absorption methods, liquid phases on
solid supports, and solid adsorbents.
1. Liquid Absorption Methods
Liquid absorption methods usually involve the use of
some type of impinger or fritted gas bubbler in conjunction with a
vacuum pump. One popular procedure utilizes the Greenburg-Smith
Impinger and consists of drawing air by vacuum pump through a
trapping medium of ethylene glycol at a rate of from 10.3 to 30 liters
per minute (U.S. Environmental Protection Agency, 1974; Kutz and Yang,
1973; Enos et al., 1972; Enos, 1976). Hexylene glycol (Stanley et
3-4
-------
al_, , 1971), hexane (Hochheiser, 1976), toluene (American National
Standards Institute, 1974), and secondary butyl alcohol (American
Industrial Hygiene Association, 1965) have also been used, although
the first of these alternatives resulted in significantly greater
interference in subsequent analytical steps (Seiber et_ al. , 1975).
Immersion of the ethylene glycol impinger in an ice bath (Hochheiser,
1976) presumably inhibits vaporization losses. An ethylene glycol
impinger system is presented in Figure 3.1. The alumina adsorption
"backup" also shown in this diagram, may sometimes be employed,
although use of a backup adsorbant often results in additional analyt-
ical interference (Seiber et_ aJL. , 1975).
The method has been found to work reasonably well with a variety
of pesticides and has proven efficient for both vapors and dusts.
In those cases where an alumina backup was employed, significant
amounts of trapped chemicals were detected in all three sampler
components (the impinger solution, the glass filter, and the adsorp-
3
tion backup). Levels as low as 0.1 ng/m were detectable for many
compounds (Seiber et_ _al. , 1975).
This method is currently employed for PCBs by the Ecological
Monitoring Branch of the U.S. Environmental Protection Agency (Kutz
and Yang, 1975) and is advocated by the American National Standards
Institute (ANSI) (1974); however, the efficiency for PCB collection
has never been fully evaluated (Enos, 1976).
3-5
-------
glass cloth
filter
TO FLOWMETER
AND
VACUUM PUMP
alumina
Source:
ethylene glycol
(100 ml)
Adapted from Seiber, J.N., J.E. Woodrow, T.M. Shafik and
H P Enos 1975. "Determination of Pesticides and Their
Transformation Products in Air," in Environmental Dynamics
oj; Pesticides: 17-43. Plenum Press, New York and London.
Figure 3.1. VILTF,R, TMPTNGER, AND ADSORPTION SAMPLING TRAIN
3-6
-------
A typical fritted glass bubbler is shown in Figure 3.2. Both
particulates and vapors are also trapped by these devices. Particu-
lates adsorb to the fritted glass and the PCBs are extracted from the
adsorbed particulates by the organic solvent in which the fritted
glass is immersed. It is advisable to surge the solvent back and
forth through the fritted glass several times before a sample is with-
drawn for analysis (Hendrickson, 1968). The solvent most frequently
used for PCB collection is toluene and the method is also recommended
by ANSI (ANSI, 1974; Hochheiser, 1976).
As in the case of the ethylene g]ycol impinger, neither the
capacity nor the efficiency of the fritted gas bubbler for removal
of airborne PCBs has been experimentally evaluated. Therefore, to
obtain measurable amounts of PCBs, it is best to minimize the sampling
flow rate and maximize the sampling time period. When high flow rates
must be employed, or when a larger air volume is required, it is recom-
mended that several gas bubblers be used in tandem (ANSI, 1974).
Several limitations are inherent in liquid absorption systems.
The most important of these is the difficulty of collecting a suffi-
cient sampling volume. Air flows are limited to about 30 liters per
minute, so that a 24-hour sampling period results in the collection
of PCBs from a total of only 43 cu m of air (Kutz and Yang, 1975).
While an increased sampling period is possible, losses from the impinger
reportedly increase upon prolonged sampling and are greater with more
volatile compounds (Seiber, 1975). Samples collected by EPA (U.S.
3-7
-------
Source: Adapted from Hendrickson, E.R. 1968. "Air Sampling and
Quantity Measurement," in Air Pollution II. Second Edition.
Academic Press, New York and London.
Figure 3.2 A TYPICAL FRITTED-GLASS BUBBLER USED IN SAMPLING FOR GASES
3-1
-------
Environmental Protection Agency, 1974) usually of 400 ml of ethylene
glycol, representing the contents of four impingers of 100 ml each,
two of which are operated simultaneously for 12 hours, and the other
two for an additional 12 hours. The total air volume sampled is about
80 cu m (U.S. Environmental Protection Agency, 1974). An additional
shortcoming of this method is that the equipment is very expensive and
quite fragile. Finally, the system also traps numerous interfering
pesticides, thus requiring additional cleanup steps which inevitable
result in further loss of PCBs.
Liquid absorption methods appear most suitable for sampling at
sites where relatively high concentrations are expected, such as near
investment casting foundries, capacitor manufacturing plants, and low
temperature incinerator stacks. Due to the absence of data relating
to the efficiency and capacity of these systems, they may only be
considered accurate for the measurement of relative PCS concentrations
sampled under identical conditions. The methods are not appropriate
for general ambient sampling (Enos, 1976).
2- Liquid Phases on Solid Supports
Liquid phases on solid supports fall into one of two
categories. The first category includes those samplers designed as
an alternative to liquid absorption systems. These samplers are
used in conjunction with a vacuum pump and, due to the porosity of
the solid support, allow a flow rate approximately 10 times that
achieved by impinger style collectors. The second category includes
3-9
-------
those samplers designed for static air sampling. These samplers are
suitable for comparison of contamination levels from one location to
another, but do not permit the calculation of concentrations (Seiber
J-Ji -Si.' » 1975). The variety of such collection systems used to date
for PCBs and similar compounds (the organochlorine insecticides, for
example) are summarized in Table 3.1.
Of the high volume samplers, the only types which have been
specifically applied to PCBs are the silicone oil coated ceramic
saddle system of Harvey and Steinhauer (1974) and the glycerine-
florisil system of Wakimoto et^ jj. (1974). In the former case, a
3.0 urn glass fiber filter was inserted in front of the coated saddles
to trap particulates. It was expected that PCBs in the vapor and
aerosol phase would easily pass through the filter and be trapped in
the silicone oil, while those PCBs adsorbed onto solid particles would
be retained by the filter. However, in maintaining their equilibrium
vapor pressure with the surrounding air at the high flow rates used,
the particulate-bound PCBs may have been entrained in the flow and swept
into the vapor trap by the filters. Thus, even though only one percent
of the PCB collected was retained by the filters no conclusion as to the
relative concentration of PCBs in the particulate and aerosol-vapor
phase is possible.
Cottonseed oil-coated glass beads and paraffin coated Chromosorb
A have both been tested for a variety of pesticides. While the former
system appeared promising, prolonged sampling of air resulted in
partial breakdown of the cottonseed oil with the formation of electron-
3-10
-------
Table 3.1 LIQUID PHASES ON SOLID SUPPORTS
A. HIGH VOLUME COLLECTORS
Solid Support
0.64 cm ceramic saddles
0.3 cm glass beads
Florisil
Chromosorb A
B. STATIC COLLECTORS
Solid Support
0.05 cm nylon nets
0.02 cm nylon net
nylon nets (no dimension
given)
glass plates
Liquid Phase
0.25 percent OV-17
silicone oil
cottonseed oil
5 percent glycerine
5 percent paraffin
oil
Reference
Harvey and Stein-
hauer, 1974
Compton and
Biorkland, 1972
Wakimoto et al.,
1974
Seiber j3t_ al. ,
1975
Liquid Phase
Reference
30 percent glycerine Risebrough jet al. ,
in water
1968
silicone oil (SE-30) SBdergren, 1972
10 percent ethylene
glycol in acetone
mineral oil
Tessari and
Spencer, 1971
Young et al.,
1975c
3-11
-------
capturing interferences (Seiber jit jil. , 1975). Other liquid phases
(such as ethylene glycol) might prove more suitable. The trapping
efficiency of paraffin oil on Chromasorb A was greater than 47 percent
for most of the 23 pesticides tested in the vapor phase. There were,
however, exceptions. Aldrin, for instance, was not trapped at all by
thia system. Therefore, one cannot be certain that this sampler would
be appropriate for PCBs unless tests are conducted on specific PCS
mixtures.
With the exception of the ethylene glycol coated nylon nets, all
of the static collectors have been tested for PCBs.
The major advantages of static collection are that the methods
are noise-free, require no electrical hookups, and apparently trap
about four times as much material as do impingers (Seiber et al. , 1975).
However, in many cases, vapors and small particulates are not collected
(Risebrough, 1968; Young, 1975), and the prolonged sampling periods
required for adequate collection result in significant volatilization
losses (Young j_t al_. , 1975c). As was mentioned previously, in the
static collection methods, the amount of air sampled remains unknown
so that concentrations of PCBs at a given location cannot be computed.
When these samplers are employed out of doors, variation in wind direc-
tion and speed also renders comparison among different locations or
of the same location at different times quite difficult.
A type of static sampler presently in use Ls the mineral oil
glass plate system of Young et^ ai_. (1975c). This system has
3-12
-------
undergone intensive study in order to estimate the collection efficiency
for various PCB mixtures. In one experiment, a series of mineral oil
coated plates were exposed for a period of eight days on a roof near
Santa Monica Bay, Calfornia. The plates were divided into four
groups. One group was sampled and repositioned daily; the second
group was sampled and repositioned every second day; the third group
was sampled and repositioned once, on the fourth day; the final group
was sampled on the eighth day only. The samples were analyzed for a
variety of chlorinated hydrocarbons and a PCB mixture, corresponding
to Aroclor 1254, was detected. The average total quantities of
Aroclor 125A accumulated over the eight-day period are presented in
Table 3.2. The data indicate that collection efficiency decreased
upon prolonged, uninterrupted plate exposure. The losses most likely
resulted from volatilization. The fact that Aroclor 1242 was not
detected by this method suggested that the collection efficiency was
not uniform for all PCB components. In a second experiment designed
to test this hypothesis, 0.1 mg of Aroclor 1242 was added to a solu-
tion of mineral oil in hexane and placed on fallout sampling plates.
The plates were exposed (on the roof of the laboratory) for four days.
The percent recoveries for the various Aroclor components, as a function
of gas chromatographic (GC) retention time, are presented in Figure
3.3. Those components with longer GC retention times (generally the
least volatile) exhibited the greatest percent recovery. The overall
recovery was less than 40 percent. Since the chromatograms showed no
3-13
-------
Table 3.2 CUMULATIVE AVERAGE QUANTITIES OF
AROCLOR 1254 COLLECTED OVER AN 8-DAY PERIOD
ON FOUR SETS OF COLLECTION PLATES SAMPLED AT DIFFERENT FREQUENCIES
Total Number of Collections Aroclor 1254 (ng)
8 401
4 352
2 '327
1 195
Source: Adapted from Young, D.R., D.J. McDermott, T.C. Heesen.
1975. "Polychlorinated Biphenyl Inputs to the Southern
California Bight," Background paper prepared for the
National Conference on Polychlorinated Biphenyls, 19-21
November, Chicago, Illinois.
3-14
-------
40
w
H
30 -
20
OS
w
o
w
10
1.0 2.0 3.0 4.0
RELATIVE RETENTION TIME
5.0
6.0
7.0
Source: Adapted from Young, D.R., D.J. McDermott, and T.C. Heesen.
1975. "Polychlorinated Biphenyl Inputs to the Southern
California Bight." Background paper prepared for the
National Conference on Polychlorinated Biphenyls, 19-21
November. Chicago, Illinois.
Figure 3.3 RECOVERY OF AROCLOR 1242 FROM SPIKED GLASS PLATE
5-15
-------
decomposition peak, loss through U.V. decomposition seemed unlikely.
Thus, the mineral oil glass plate samplers do not adequately capture
and retain the more volatile low molecular weight PCBs in the atmos-
phere. It is these very contaminants, however, which are most likely
present in the greatest quantities.
The drawback common to all liquid-coating methods lies in the
extraction of relatively large amounts of oil along with the desired
compounds. These oils are difficult to separate out and interfere
in subsequent analytical steps (Seiber et al_. , 1975).
Although a variety of liquid-coated solids are available and in
use both for high volume and static air sampling, no one system
efficiently and accurately traps and retains both vapor phase and
particulate bound PCBs over a wide range of molecular weights.
3. Solid Adsorbents
A number of solid adsorbents have been examined in the
hope that one or more of these might allow a greater rate of flow
than impinger systems and less resultant analytical interference
than coated solids. The solids tested for PCBs and pesticides are
listed in Table 3.3. Silica gel and Chromasorb 102 were tested using
the same protocol and air flow rate (-23.5 liters per minute) employed
for ethylene glycol impinger sampling (Seiber et_ a^. , 1975). Trapping
efficiencies towards a variety of pesticides were greater than 47
percent in all cases. A serious problem, however, was the large
quantity of electron capturing background trapped from the air by the
3-16
-------
Table 3.3 SOLID ADSORBENT SAMPLING MEDIA
Adsorbent Reference
Florisil Yule e^ aJL. , 1971
Alumina Stanley _et al., 1971
Chromosorb 102 Seiber _et al., 1975
Polyurethane Bidleman and Olney, 1974
Silica gel Seiber et al., 1975
Source: Adapted from Seiber, J.N., J.E. Woodrow, T.M. Shafik,
and H.F. Enos. 1975. "Determination of Pesticides
and Their Transformation Products in Air," in
Environmental Dynamics of Pesticides: 17-43, Plenum
Press, New York and London.
3-17
-------
silica gel. This background probably reflects the greater trapping
efficiency of the silica gel for air constituents not trapped well by
ethylene glycol.
Preliminary tests using a variety of organochlorine pesticides
have also been conducted with silica gel and Chromosorb 102 at much
higher air flow rates (Seiber e_t al_. , 1975). The device employed
was a commercial high-volume sampler (Figure 3.4). The results were
encouraging in that many pesticides (including the DDT compounds and
the cyclodienes) were trapped with efficiencies greater than 50 per-
cent at a flow rate as high as 1 cu m per minute. The trapping
efficiency of silica gel for chlorinated hydrocarbons of high volatil-
ity was somewhat greater than that of Chromosorb 102; however, the
air background trapped by silica gel was also greater. These solids
have not been tested for PCBs per se.
A highly promising variation of the above sampling systems is the
polyurethane foam system of Bidleman and Olney (1974a and 1974b),
which has been tested for PCBs. In this system, an 8 x 10 inch glass
fiber filter for particulate collection is placed on top of a container
holding the solid trapping medium. The medium consists of a porous
polyurethane foam plug, pre-cleaned by washing with water and extrac-
tion with acetone and petroleum ether. The entire apparatus is fitted
to a high volume sampler. Since the foam offers little resistance to
air passage, high airflow rates (0.4 to 0.8 cu m per minute) are
easily obtained and hundreds of cubic meters of air per day may be
3-18
-------
J" x 10"
FILTER
100 MESH
WIRE SCREEN
FAN &
MOTOR
I
COLLECTION
ADSORBANT
Source: Seiber, J.N., J.E. Woodrow, T.M. Shafik, and H.F. Enos. 1975.
"Determination of Pesticides and Their Transformation Products
in Air," in Environmental Dynamics of Pesticides;17-43,
Plenum Press, New York and London.
Figure 3.4 HIGH VOLUME SAMPLER MODIFIED FOR HOLDING SOLID MEDIA
3-19
-------
sampled. The efficiency of this collection system for tri-, tetra-,
and pentachlorinated PCB vapors was found to be greater than 90 per-
cent (Bidleman and Olney, 1974a and 1974b). In addition, this system
appears to be relatively selective in that many organochlorine insect-
icides are not trapped by the polyurethane plugs (Enos, 1976). Se-
lectivity provides an obvious advantage since fewer separation and
cleanup procedures are necessary.
High volume sampling using some type of solid adsorption medium
appears to be the best alternative for ambient monitoring of PCBs.
The polyurethane foam system of Bidleman and Olney is the best solid
sampling media available at this time. This system, which allows
3
detection of PCB concentrations as low as 1 ng/m in just four hours
of sampling time, is presently undergoing further evaluation and opti-
mization studies under the auspices of the Office of Toxic Substances,
U.S. Environmental Protection Agency (Hochheiser, 1976).
D. EXTRACTION AND CLEANUP PROCEDURES
Removal of PCBs from the sampling medium is usually carried
out by means of extraction with an organic solvent. The solvent used
most frequently is hexane (SBdergren, 1972; ANSI, 1974; Harvey and
Steinhauer, 1974; Kutz and Yang, 1975), although use of petroleum
ether, acetone-hexane mixtures, and petroleum ether-diethyl ether
mixtures have been reported (Bidleman and Olney, 1974a and 1974b;
Risebrough et_ al., 1968; Hochheiser, 1976). Since all sampling pro-
cedures trap some interfering compounds along with PCBs, some type of
3-20
-------
cleanup procedure is usually required prior to analysis. The stability
of PCBs to alkali permits the removal of DDT from the extracts by
saponification (U.S. EPA, 1974). Treatment with alkali converts DDT
to DDE, which may then be removed by extraction with concentrated
sulfuric acid. Other chlorinated hydrocarbon Interferences may also
be removed by this treatment (ANSI, 1974).
Column chromatography using silicic acid (silica gel), magnesia
silica gel( e.g., Florisil), alumina, or some combination of these in
sequence is another frequently employed cleanup procedure (ANSI, 1974).
The silica gel method was developed in 1970 by Armour and Burke
(1970) and is a standard procedure for separation of PCBs from a
variety of organochlorine pesticides. Careful preparation of the
silica gel is required prior to use. The silica gel should be heated
at 130°C for a minimum of seven hours ( preferably 24 hours); the
water content must then be brought to 3 percent in order to achieve
the maximum margin of separation between PCBs and pesticides. Five
grams of Celite should be added to every 20 grams of prepared silica
gel in order to achieve a faster elution rate from the adsorbent
column. PCBs are separated from DDT and its analogs by elution with
petroleum ether; the DDT compounds, along with lindane, heptachlor,
heptachlor epoxide, dieldrin, and endrin, may be recovered completely
by a second elution with acetonitrile: hexane: methylene chloride::
1:19:80 ( Armour and Burke, 1970).
3-21
-------
Since only a small margin of separation exists between PCBs and
p,p'-DDE, it is imperative that the mixture of silicic acid-celite
be of a composition and activity identical to that reported in the
original paper (Enos, 1976).
Elution of an activated Florisil (heated to 130°C) column with
hexane removes over 92 percent of the PCBs, PCTs, DDE, heptachlor, and
aldrin present in an applied extract. DDT, ODD, dieldrin, lindane,
heptachlor epoxide, and endrin may then be eluted with 20 percent
diethyl ether in hexane (Sherma, 1975). Some other procedure (such
as saponification or silicic acid-Celite chromatography) is still, how-
ever, required to separate the DDE from the PCBs.
Alumina has been found to be a more effective, more reproducible
column substrate in the separation of PCBs from electron-capturing
interferences than either silicic acid or Florisil (ANSI, 1974). The
activity of alumina varies considerably with age and lot; therefore,
a defined amount of water (5 to 6 percent) should be added to activated
alumina (heated for over four hours at 400'C) to ensure a reproducible
activity (ANSI, 1974; Bidleman and Olney, 1974b). Elution of the
alumina column with hexane or petroleum ether removes PCBs from inter-
fering impurities.
E. QUALITATIVE AND QUANTITATIVE ANALYSIS
Polychlorinated biphenyls are usually determined by gas
chromatography (GC) using microcoulometric, electrolytic-conductivity,
or electron-capture detectors (U.S. EPA, 1974; Sherma, 1975).
3-22
-------
Microcoulometric and electrolytic-conductivity detectors are halogen
specific so that less cleanup of extracts is required for accurate
identification (Sherma, 1975; U.S. EPA, 1974). Electron-capture
detectors are, however, the most sensitive (although least selective)
of the three for chlorine-containing compounds. Use of electron-
capture detectors is, therefore, preferred when low concentrations of
PCBs (such as are present in ambient air samples) are expected. The
key to successful GC analysis of PCBs is selection of a proper column.
The column consists of two parts: the solid support and the liquid
phase. Solid supports recommended for this work include Chromosorb
W, Gas Chrom P, Gas Chrom Q, and others (ANSI, 1974; Sherma, 1975).
A variety of polar and non-polar liquid phases have been investigated.
The following have been found to provide adequate separation for use
in PCB analysis by electron capture: DC-200, SF-96, OV-17, SE-30
QF-1, and mixtures of these (ANSI, 1974; Sherma, 1975). It is usually
advisable to use at least two different columns, one polar and one
non-polar, for unequivocal identification (Sherma, 1975).
Qualitative identification of residues may be made by comparing
retention times for all chromatographic peaks with those of standard
or commercial PCB formulations. Retention times must be relative to
some arbitrary standard such as dieldrin. The various Aroclors exhibit
different GC patterns useful in distinguishing between them. These
differences may include the number of major peaks and the peak-height
ratios of certain peaks. The electron-capture chromatograms for five
3-23
-------
Aroclor formulations under specified conditions are presented in
Figure 3.5. Due to such factors as selective metabolism, photochemical
breakdown, and non-representative sampling and recovery of all Aroclor
components, the GC patterns obtained from environmental samples are
seldom identical to those of the Aroclor standards. Sometimes, the
chromatogram clearly indicates the presence of dominant interferences
or of the components of more than one type of Aroclor. In the former
case, an attempt at chemical or chromatographic cleanup should be
ventured. In the latter case, the PCB residues should be examined
separately using the appropriate Aroclor reference for the respective
portion of the chromatogram (U.S. EPA, 1974; Sherma, 1975).
Once the elution pattern has been subjectively matched to a
standard PCB mixture, quantitation may be conducted by one of several
methods. The total area of all accepted matching peaks in the elu-
tion pattern may be compared with the total area of the same peaks in
a standard mixture of known concentration; the concentration of PCBs
in the original air sample is calculated by taking the sample volume
into consideration (U.S. EPA, 1975).
Other techniques for quantitation are based on the area of one
or more selected peaks; total peak height, average or individual
heights of selected peaks; average electron-capture detector response
to biphenyls containing one to seven chlorine atoms; and determination
of chlorine content of different PCB peaks with the microcoulometric
detector (Sherma, L975).
3-24
-------
AROCLOR 1221
AROCLOR 1242
0 2 4 6 8 I0 I? 14 16 18 20 22 24 26 28 30 32 34 36 38
AROCLOR 12A8
AROCLOR 125A
Source: American National Standards Institute. 1974. "Guidelines for
Handling and Disposal of Capacitor and Transformer Grade
Askarels Containing Polychlorinated Biphenyls," ANSI, C107.1,
New York.
Figure 3.5 COMPARISON OF ELECTRON CAPTURE CHROMATOGRAMS FOR AROCLOR
J221, 1242, 1248, 1254, AND 1260
3-25
-------
F. CONFIRMATION
When the quantity of PCBs in an air sample is limited, the
individual GC peaks may be too small for accurate qualitative and
quantitative analysis. In this case, quantitation may be effected
by perchlorination of all the PCB residues to decachlorobiphenyl.
Quantitative perchlorination may be achieved by incubation with SbCl,.
for six hours at 175°C (Armour, 1973). The perchlorinated PCB extract
may then be analyzed by electron-capture gas chromatography. Quanti-
tation is accomplished by comparison of peak area with that of a known
concentration of pure decachlorobiphenyl.
Qualitative confirmation of the existence of PCBs may be obtained
by thin layer chromatographic (TLC) analysis. The variety of TLC
systems used for PCBs are listed in Table 3.4 (Sherma, 1975). The
ratio of the distance traveled by the unknown to the distance of sol-
vent front is compared to those of known Aroclor mixtures. Quantita-
tion may be achieved by eluting the residues from the thin layer
substrate and subjecting them to GC analysis as described above.
Although infrared and nuclear magnetic resonance spectroscopy
have been used on occasion to aid in the identification of PCB
mixtures, both of these methods suffer from two disadvantages:
(1) lack of sensitivity and (2) requirement for a relatively pure
compound (Garrison ej: al. , 1972).
Mass spectrometry, although not in itself a very sensitive method,
when coupled with gas chromatography is undoubtedly the best procedure
3-26
-------
Table 3.4 SYSTEMS FOR THIN-LAYER CHROMATOGRAPHY OF PCBs
Layer
Solvent
Detection
Remarks
I
NJ
Aluminum oxide G
incorporating
AgNO.
MN-Silica gel
G-HR/AgNO.
Aluminum oxide G
Liquid paraffin
(8%) on kieselguhr
Benzene-hexane, 5:95
n-Heptane; heptane-
acetone, 98:2
Hexane-anhydrous
deithyl ether,
40:0.8, or pure
heptane
Acetonitrile-acetone-
methanol-water mixtures
Ultraviolet light after
spraying with
phenoxyethanol-H^O^
Ultraviolet light
AgNO ..^-phenoxyethanol-
H-O,., spray followed
by ultraviolet light
1.7 g AgNO in 96%
ethanol spray followed
by ultraviolet light
DDE converted to DCBP
by oxidative treatment;
R PCBs=0.91-0.94,
DCBP=0.30, other
chlorinated pesticides
= 0.48-0.88
Two-dimensional develop-
ment; PCBs separated
from DDT and analogs
Aldrin and DDE overlap
PCB spots when develop-
ment is with heptar.e
Reversed-phase TLC;
PCBs appear as a nuE-
ber of distinct spots
separated from DDE,
DDT, and some other
pesticides
Source: Modified from Sherma, J. 1975. "Gas Chromatographic Analysis of Polychlorinated Biphenyls
and Other Non-Pesticide Organic Pollutnats," Advances in Chromatography 12:141-176, Marcel
Dekker, Inc.
-------
for confirming the presence of PCB residues and characterizing the
molecular composition of PCB formulations (Sherma, 1975). Coupling of
the gas chromatograph and the mass spectrometer is normally carried
out in such a way that a chromatogram-type readout analogous to that
produced by conventional GC detectors is produced. A sample GC-MS
chromatogram (that of Aroclor 1254) is presented in Figure 3.6. The
fragment ion of greatest mass, in most cases, corresponds to the mole-
cular weight of the compound, having been created by the loss of one
electron. The masses of important fragment ions provide clues to the
structure of parts of the molecule. Apparently, however, the primary
ion spectra of different PCB isomers containing the same number of
chlorine atoms are virtually undistinguishable. Therefore, the use
of mass spectrometry for structural studies of PCB contaminants is
fairly limited (Sherma, 1975).
3-28
-------
OJ
I
M 10°
in
w
M
w
>
E-i
w 20
2C1 M
Cl+-2Cl)
-
3C1
(M+-C1)
1
1
4C1
1
200
240
280
320
n/
4(
-
- 2C1
-
3C1
,1
:i
M+,5C1
.
-
4C1
5C1
1
i 1 1 i IB i ifl i
6C1
220 260 300 340 240 280 320 360
m/e m/e
Source: Modified from Garrison, A.W., L.H. Keith, and A.L. Alford.
1972. "Confirmation of Pesticide Residues by Mass Spectrometry
and NMR Techniques," in Fate of Organic Pesticide in the
Aquatic Environment:26-54, Advances in Chemistry Series III,
American Chemical Society, Washington, B.C.
Figure 3.6 PARTIAL MASS SPECTRA (GC-MS) OF AROCLOR 1254 SHOWING THE
NUMBER OF CHLORINE ATOMS ON EACH ION
-------
IV. SOURCES
A. SUMMARY
Polychlorinated biphenyls are manufactured in many nations
of the industrialized world. The sole producer in the United States
is the Monsanto Company, which markets them under the trade name
Aroclor. The peak of production occurred in 1970, when 7.31 x 10
pounds (3.32 x 10 kg) were sold domestically. A voluntary limitation
of sales by Monsanto, restricting the use of PCBs to closed-system
applications, resulted in a reduction of sales in 1971 through 1974 to
less than half the peak value.
Prior to 1971, PCBs were used in a wide variety of applications.
Aside from their incorporation into closed system electrical devices
where they served as an insulating fluid, they could be found in heat
transfer fluids, hydraulic fluids, paints, protective coatings, and
many other types of products where their chemical and thermal sta-
bility and non-flammability were essential. At present, the PCBs
sold in the United States by Monsanto are used only for electrical
applications where alternate or substitute materials are not yet
available. However, PCBs may still be available for use in dispersive
applications through imports, reprocessed PCB-containing waste oils,
or resale by Monsanto's customers.
The greatest losses to the environment also occurred prior to
1971, when precautions against contamination had not yet been put into
4-1
-------
effect. Loss of PCBs to the environment as a consequence of use
includes accidental spills, leaks, and leaching, as well as vaporiza-
tion resulting from their use as pesticide extenders and in paints
and plastics into which they are incorporated. As a result, several
instances of gross contamination of human and animal food have been
reported. Disposal of waste PCBs during production or incorporation
into products was and continues to be the primary route of global
contamination. PCBs are released to the hydrosphere in both municipal
and industrial effluents and to the atmosphere as a result of
incomplete burning or incineration.
The discontinuation of the use of PCBs in applications where the
possiblity of dispersal exists, the stringent controls at production
sites that have been initiated, and the imposed limitations on effluent
concentrations should, if adhered to, result in a significant decrease
in the amounts of PCBs released to the biosphere In the future. But
evidence of negligence resulting in episodes of localized gross con-
tamination continues to be uncovered, suggesting that even more control
measures may be warranted.
B. PRODUCTION AND USE
1. Production
PCBs have been available commercially since 1929
(Hutzinger, ejt^ aj_, , 1974) and are presently manufactured in many of
the world's industrialized nations. Table 4.1 lists the major
4-2
-------
Table 4.1 THE WORLD'S MAJOR PRODUCERS OF PCBs
Company
Monsanto
Bayer
Prodelec
Kanega fuchi*
Mitsubishi-Monsanto*
Caffaro
Flix
Sovol
Chemko
Country
U.S.A. and Great Britain
Germany
France
Japan
Japan
Italy
Spain
U.S.S.R.
Czechoslavakia
Tradename
Aroclor
Clophen
Phenoclor
and
Pyralene
Kanechlor
Santotherm
Fenclor
CD
(1)
-------
producers of PCBs. The sole producer in the United States is the
Monsanto Company, which markets them under the trade name Aroclor.
Each Aroclor mixture is designated by a four-digit number in which
the first two digits, 12, specify chlorinated biphenyl and the last
two indicate the approximate weight percentage of chlorine in the
mixture. Table 4.2 summarizes the total domestic production and
sales figures for PCBs from 1957 to 1974 and gives a breakdovm of
sales by type of PCS mixture for the same period. Sales figures
are presented graphically in Figure 4.1.
Over 450,000 short tons (400,000 metric tons) have been pro-
duced domestically since 1957. Although similar data for worldwide
production are not available, it has been estimated that U.S.
production represents approximately half of the world production
(Interdepartmental Task Force on PCBs, 1972). In 1972, Japan was
believed to have produced about 26 million pounds of PCB (12,000
metric tons) (Nisbet and Sarofim, 1972). However, a ban on production
in that country was allegedly put into effect subsequent to that date
(Anonymous, 1975a). From 1960 to 1970, U.S. domestic production and
sales increased an average of 8.7 percent and 6.1 percent per year,
respectively, so that both production and sales more than doubled in
this decade.
By 1970, it became apparent, through monitoring efforts in
both this country and abroad, that PCBs were widely distributed
in the environment and might possibly pose a serious threat to a
variety of life forms including man. In late 1970, Monsanto, in
4-4
-------
Table 4.2 PCB MANUFACTURE AND SALES BY GRADE
MONSANTO INDUSTRIAL CHEMICALS COMPANY 1957 thru 1964
(Thousands of Pounds)
U.S. Production
Domestic Sales (Ibs)
DOMESTIC SALES BY PCB GRADE
Aroclor 1221
Aroclor 1232
Aroclor 1242
Aroclor 1248
Aroclor 1254
Axoclor 1260
Aroclor 1262
Aroclor 1268
Aroclor 1016
1957
(1)
32299
23
196
18222
1779
4461
7587
31
—
•—- -
1958
(1)
26061
16
113
10444
2559
6691
5982
184
72
—
1959
(1)
31310
254
240
13598
3384
6754
6619
359
102
—
1960
37919
35214
103
155
18196
2827
6088
7330
326
189
—
1961
36515
37538
94
241
19827
4023
6294
6540
361
158
—
1962
38353
38043
140
224
20654
3463
6325
6595
432
210
—
1963
44734
38132
361
13
18510
5013
5911
7626
414
284
—
1964
n^^^^^f—
50833
44869
596
13
23571
5238
6280
8535
446
190
—
I
Ln
(1) Production figures unavailable during year indicated.
Source: Adapted from Monsanto Industrial Chemicals Company, 1974. "PCB Manufacture and Sales
Monsanto Industrial Chemicals Company - 1957 thru 1964"(unpublished data).
-------
Table 4.2 (continued) PCB MANUFACTURE AND SALES BY GRADE
MONSANTO INDUSTRIAL CHEMICALS COMPANY 1965 thru 1974
(Thousands of Pounds)
U.S. Production
Domestic
D6MgSTie
Sales
SALES BY
PCB GRADE
Aroclor
Aroclor
Aroclor
Aroclor
Aroclor
Aroclor
Aroclor
Aroclor
Aroclor
1221
1232
1242
1248
1254
1260
1262
1268
1016
1965
60480
51796
369
7
31533
5565
7737
5831
558
196
0
1966
65849
59078
528
16
39557
5015
7035
5875
768
284
0
1967
75309
62466
442
25
43055
4704
6696
6417
840
287
0
1968
82854
65116
136
90
44853
4894
8891
5252
720
280
0
1969
76389
67194
507
273
45491
5650
9822
4439
712
300
0
1970
85054
73061
1476
260
48588
4073
12421
4890
1023
330
0
1971
34994
34301
2215
171
21981
213
4661
1725
1
0
3334
1972
38600
26408
171
0
728
807
3495
305
0
0
20902
1973
42178
37742
35
0
6200
0
7976
0
0
0
23531
1974
40466
34406
57
0
6207
0
6185
0
0
0
21955
Source:
Adapted from Monsanto Industrial Chemicals Company, 1974. "FUJ Manufacture and Sales -
Monsanto Industrial Chemicals Company - 1965 thru 1974". (Unpublished data.)
-------
20,000
/ | M)?4! 1973 197A
Figure 4.1 U.S. DOMESTIC SAM'S OF PCBs BY GRADE
4-7
-------
response to the mounting evidence and concern, voluntarily reduced its
sales of PCBs to use only in closed systems (transformer and capacitor
applications) where chances of further contamination would be
minimized (Monsanto, 1970). As a result, total production in 1971
dropped to 41 percent of the 1970 figure and remained below 50
percent of that figure in 1974. Monsanto subsequently developed a
new grade of product, Aroclor 1016, which has largely replaced its
former biggest seller, Aroclor 1242 (Monsanto, 1975a). Aroclor 1016
is 41 percent chlorine by weight (as compared to 42 percent chlorine
in Aroclor 1242) but contains significantly reduced amounts of the
>
penta-, hexa-, and hepta-chlorinated isomers, which are known to per-
sist in the environment (Monsanto, 1972). Aroclor 1036 now accounts
for 64 percent of total domestic sales. As of 1973, no Aroclor of a
higher grade than 1254 has been sold. Monsanto's sales restrictions
were not in response to any legal restrictions on the use of PCBs for
dispersive applications.
Glenn E. Schweitzer, EPA Office of Toxic Substances, has
suggested that PCBs are possibly being resold by reprocessors and
by some of Monsanto's customers for unauthorized uses. (Anonymous,
1975a; Anonymous, 1975b). Although such allegations are scoffed at
by representatives of Monsanto who claim that their customers "have
too much at stake to do this" (Anonymous, 1975c), Thomas Kopp, a
chemist at EPA's Toxic Substances Office, claims to have "received
several letters confirming Schweitzer's charge." (Anonymous, 1975a).
4-8
-------
2. Import and Export
Information as to the nature and fate of PCBs im-
ported into this country is limited and conflicting. According to
Nisbet and Sarofim (1972), imports of PCBs are primarily comprised
of transformer oils and capacitor fluids in electrical devices and
are thought to be small. However, according to a more recent
report (Anonymous, 1975b) Glenn Schweitzer, head of the EPA Office
of Toxic Substances, claims that most of the PCB imports (amounting
to about 188 tons in 1974) are not used in closed electrical systems
but end up in investment casting processes, heat exchange fluids and
hydraulic fluids, where increased danger of loss to the environment
exists. The Yates Manufacturing Company, the sole producer of the
PCB-filled pattern waxes used by the investment casting industry,
imports between 300,000 and 500,000 pounds of decachlorobiphenyl
per year from Caffaro in Italy (Versar, 1975a). Prior to 1972,
many imported resins and adhesives may also have contained PCBs as
plasticizers, but the subsequent restriction of their use for such
applications has probably minimized this source of entry (Organiza-
tion for Economic Cooperation and Development, 1973). Another
source of imported PCBs in the past was food packaging materials and,
consequently, their contents. A routine survey for pesticide res-
idues in imported food stuffs revealed 10 ppm of what appeared to
be Aroclor 1242 in ground cashew nuts (Bailey e_t_ j3uL. , 1970). The
lacquered cardboard drum in which these nuts were packed was found
4-9
-------
to contain 200 ppm of PCBs. It is believed that the European paper
industry is no longer recycling PCB-containing paper for food packag-
ing materials (organization for Economic Cooperation and Development,
1973).
Domestic exports for the period 1963-1974 are presented
in Table 4.3. To.tal exports averaged 15 percent of domestic sales
during this period. Actual exports increased steadily from
3,647,000 pounds (about 1600 metric tons) in 1963 to a high of
13,651,000 pounds (62,000 metric tons) in 1970. In 1974, however,
only 5,395,000 pounds (2400 metric tons) left the country.
3. Uses and Replaceability
PCBs were initially manufactured to satisfy the
electrical industry's need for a fire-proof, explosion-proof insulat-
ing fluid (askarel) for transformers and capacitors. The unique physi-
cal and chemical characteristics that make them ideally suited for
such use include thermal stability, non-flammability, superior
dielectric properties, and resistance to hydrolysis, oxidation, acids,
bases, and other chemical agents (Hutzinger j2_t a_l. , 1974). These
same properties eventually led to the utilization of PCBs in a wide
spectrum of applications. By 1970, aside from their incorporation
into closed-system electrical devices, PCBs could be found in heat-
transfer fluids, hydraulic fluids, machine-tool cutting oil, high-
vacuum oils, specialized lubricants and gasket sealers, epoxy paints,
printing inks, waxes, synthetic adhesives, textile dyes, and protective
4-10
-------
TABLE 4.3 U.S.
SALES OF PCBs BY MONSANTO 1963-1974
YEAR
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
THOUSANDS OF POUNDS
3,647
4,096
4,234
6,852
8,124
11,231
10,624
13,651
6,388
8,346
5,395
Source: Adapted from Monsanto Industrial Chemicals Company, 1974.
"PCB Manufacture and Sales - Monsanto Industrial Chemicals
Company - 1957 thru 1974". (Unpublished data.)
4-11
-------
coatings for wood, metal, and concrete, as sealers in water-proofing
compounds and putty, and as plasticizers in synthetic resins and
chlorinated rubber. PCBs have also been incorporated into pesticides
as extenders to suppress their vaporization and prolong their "kill"
time. One major use of PCBs prior to 1970 was as a plasticizer in
carbonless copying paper (Fishbein, 1974). The grades of Aroclor
employed for each of a selected number of applications are presented
in Table 4.4. Monsanto's sales figures for the period of 1957-1974, by
application, are presented in Table 4.5 and summarized graphically in
Figure 4.2.
Between 1957 and 1970, 64 percent of all sales were for use
in closed-system electrical applications, that is, transformers and
capacitors. The remaining 36 percent were used for purposes where
containment proved difficult and resultant environmental contamination
was probable. The restriction of non-electrical sales by Monsanto
in 1970 is reflected in their sales figures. Over 16,000 tons
(14,000 metric tons) was sold for such uses in 1970; the figure
dropped to 4,500 tons (4,100 metric tons) in 1971 and to zero in 1973
and thereafter. At present, Monsanto's sales are solely for elec-
trical applications. Approximately 33 x 10 pounds per year are sold
domestically for such applications. About 70 percent of the usage
is in capacitor production, and, of this total, r>5 percent is for
small capacitors and 45 percent is for large capacitors. The re-
maining 30 percent (12 x 106 pounds per year), is consumed by the
4-12
-------
TABLE 4.4 THE USES OF PCB PRIOR TO 1970
USE
GRADE(S) OF AROCLOR
Electrical capacitors
Electrical transformers
Vacuum pumps
Hydraulic fluids
Plasticizer in synthetic
resins
Plasticizer in rubbers
Adhesives
Wax extenders
Pesticide extenders
Inks
Lubricants
Cutting oils
Carbonless copying paper
Heat transfer systems
1221, 1242, 1254
1242, 1254, 1260
1248, 1254
1232, 1242, 1248, 1254, 1260
1248, 1254, 1260, 1262, 1268
1221, 1232, 1242, 1248, 1254
1268
1221, 1232, 1242, 1248, 1254
1242, 1254, 1268
1254
1242
Source: Adapted from Hutzinger, 0., S. Safe and V. Zitko, 1974. "The
Chemistry of PCBs", CRC Press, Cleveland, Ohio.
4-13
-------
Table 4.5 PCB MANUFACTURE AND SALES BY CATEGORY
MONSANTO INDUSTRIAL CHEMICALS COMPANY 1957 thru 19/4
(thousands of pounds)
I
H'
.0
U.S. Production
Domestic Sales
DOMESTIC SALES BY CATEGORY
Heat Transfer
Hydraulics /Lubricants
Misc. Industrial
Transformer
Capacitor
Plasticizer Applications
Petroleum Additives
(1) Production figures and
1957
(1)
32299
1612
704
12955
17208
(1)
1958
(1)
26061
1549
755
5719
14099
3939
Plasticizer
1959
(1)
31310
2685
1569
5984
16499
4573
1960
37919
35214
2523
1559
7921
16967
6244
1961
36515
37538
4110
2114
6281
15935
9098
1962
38353
38043
157
3915
1681
7984
15382
8924
Applications figures unavailable
Source: Adapted from Monsanto Industrial Chemicals
Monsanto Industrial Chemicals Company, 1957
Company, 1974.
thru 1964".
1963
44734
38132
582
3945
1528
7290
15606
9181
1964
50833
44869
929
4374
1692
7997
19540
10337
during year indicated.
"PCB Manufacture
(Unpublished data.
and Sales,
)
-------
Table 4.5 (continued) PCB MANUFACTURE AND SALES BY CATEGORY
MONSANTO INDUSTRIAL CHEMICALS COMPANY 1957 thru 1974
(thousands of pounds)
U.S. Production
Domestic Sales
DOMESTIC SALES
BY CATEGORY
Heat Transfer
Hydraulics/
Lubricants
Misc. Industrial
Transformer
Capacitor
Plasticizer
Applications
Petroleum
Additives
1965
60480
51796
1237
4616
1841
8657
23749
11696
—
1966
65849
59078
1766
4258
1779
8910
28884
13481
—
1967
75309
62466
2262
4643
1426
11071
29703
13361
—
1968
82854
65116
2529
5765
1283
11585
29550
14404
—
1969
76389
67194
3050
8039
1079
12105
25022
16460
1439
1970
85054
73061
3958
7403
1627
13828
26708
19537
—
1971
34994
34301
3060
1552
1155
11134"!
14141J
3259
—
1972
38600
26408
752
0
0
25656
0
0
1973 1974
42178 40466
37742 34406
37742 34406
-C-
Source: Adapted from Monsanto Industrial Chemicals Company, 1974. "PCB Manufacture and Sales -
Monsanto Industrial Chemicals Company, 1965 thru 1974". (Unpublished data.)
-------
80,000
FIGURE 4.2
U.S. DOMESTIC SALES OF PCBs BY APPLICATION
4-16
-------
transformer industry with 90 percent of this amount consumed in the
production of new transformers and 10 percent in the service of old
units (Durfee, 1975). In 1972, Bayer of Germany followed Monsanto's
lead by discontinuing the production and sales of Clophen brand of
PCBs for use in all open-systems applications (Anonymous, 1973).
In 1973, the Organization for Economic Cooperation and Development
(OECD) agreed to restrict the use of PCBs to applications where their
non-flammability was an essential factor (OECD, 1973). Their
agreement is less restrictive than Monsanto's program since they
allow the use of PCBs for non-food-related heat-transfer fluid
applications and for hydraulic fluid in mining equipment (Monsanto,
1974c). Although the use of those PCBs produced domestically appears
to be restricted to non-dispersive applications, no such regulation is
in effect for imported PCBs. One major application of a specific type
of imported PCB (decachlorobiphenyl) is as a filler in investment
casting waxes. About 400,000 pounds of decachlorobiphenyl are
imported annually from Caffaro in Italy by the Yates Manufacturing
Company in Chicago, Illinois, which produces approximately 1.5 million
pounds of PCB-containing wax per year. Decachlorobiphenyl constitutes
approximately 30 percent of the weight of the wax (Durfee, 1975;
Versar, 1975a). Yates is the sole producer of this type of invest-
ment casting wax although four other companies produce waxes contain-
ing polychJorinated terphenyls (PCTs) (Versar, 1975a). There are
currently 1.35 Investment casting foundries in the United States, 25
of which use PCB-containing wax.
4-17
-------
The investment casting process IB a lost wax casting process.
A pattern is molded by the injection of the molten casting wax into
a metallic die where the wax cools and solidifies to form the desired
shape. The wax pattern is then surrounded by a slurry containing a
refractory ceramic (known as the investment) to form the final mold.
After the mold dries to an appropriate strength, the wax pattern is
melted in an autoclave and the wax is recovered for possible future
use or disposal. Residues of wax remaining in the pores of the
ceramics mold are burned out in a furnace at 1900 to 2000° F. Molten
metal may then be poured into the cavity of the ceramic mold to form
a casting (Versar, 1975a). Addition of fillers such as PCBs or PCTs
to investment casting waxes is a development of the last decade. By
reducing the wax content through addition of low shrinkage fillers,
volumetric shrinkage of the ceramic mold may be controlled (Versar
1975a).
A source of considerable concern was discovered with the identifi-
cation of PCB residues in a variety of human and animal food and feed
(Interdepartmental Task Force on PCBs, 1972). The use of coatings
containing PCB as silo sealants and the recycling of PCB-containing
carbonless carbon paper for food packaging are examples of the means
by which such contamination came about. The major European producer
of carbonless carbon paper (Wiggins Teape) discontinued the use of
PCBs in its product as of June 1970 (Lister and Bennett, 1972) and
the American producer (National Cash Register) followed suit
4-18
-------
exactly one year later (Trout, 1972). Although the levels of PCBs in
paperboard have decreased since these measures were taken, 6 percent
of samples of paperboard used in Canadian food packaging were found to
contain in excess of 10 ppm PCS when analyzed in 1973. About 20 per-
cent of the samples contained between ] and 10 ppm (Villeneuve et al.,
]973).
The Food and Drug Administration, with the support of the
Department of Agriculture, has tightened controls over the use of
PCBs for any purpose that might result, directly or indirectly, in
the contamination of the food supply. Surveillance measures have
been put into effect to ensure that PCBs are not used on farms, in
food plants, or in food-packaging material (U.S. Department of Health,
Education and Welfare, 1972).
Alternate or substitute materials are available that can replace
PCBs in many of their former uses (Interdepartmental Task Force on
PCBs, 1972). Tricresyl phosphate and other phosphate esters are
frequently used as hydraulic-fluid additives and serve adequately as
vacuum-pump oils. Non-flammable fluids, which can be employed in heat-
transfer applications, include fluorocarbons, water, and molten salts
and metals. A wide variety of plasticizers are available depending
on the specific property required, and the phthalate esters have
replaced PCBs as general-purpose plasticizers. Tn 1972, Bayer, in
Germany, announced the development of two experimental products known
as Leromoll L2279 and L2280, which are mixtures of entirely synthetic
4-19
-------
aliphatic and aromatic hydrocarbons of medium molecular weight with
properties similar to Clophen. Their intended use Is in the manufacture
of chemical-resistant paints, sealing and joint-filling compounds,
adhesives, and printing inks (Anonymous, 1973). Alternatives to the
use of decachloro biphenyl as a filler in investment casting wax
include use of a substitute filler or of an unfilled wax. The
properties required for a filler are: a melting point exceeding
300°C, a high heat-transfer coefficient, a low thermal coefficient of
linear expansion, and "zero" ash (.Versar, 1975b). Isophthalic acid, one
possible candidate, was found to leave an ash residue upon testing.
A recently commercialized grade of isophthalic acid exhibited a much
lower ash content; its cost is comparable to that of decachloro
biphenyl (Versar, 1975b).
Industry claims that reverting to the use of unfilled waxes will
increase production costs by about 10 percent. However, new types of
unfilled waxes are presently available whose properties are supposedly
equivalent to those of the PCB-filled waxes and whose cost is somewhat
lower. The exact formulation of these waxes is not known (Versar, 1975b)
The only application of PCBs for which no suitable substitute is
presently commercially available is also their major application—as
electrical insulating and cooling fluids in transformers and capacitors.
The PCB-containing fluids (known as askarels) are used when the
electrical equipment is installed in or near buildings, as they are
virtually free from the danger of fire and explosion. While mineral
4-20
-------
oil, another frequently used insulating fluid, has a flammability
rating of 10-20 (based on a value of 0 for water and of 100 for
ether), Aroclors 1242 and 1016 have a rating of 2-3 (Interdepartmental
Task Force on PCBs, 1972). Thus, when power surges occur, the resultant
arcs will vaporize and ignite mineral oil, whereas PCBs will safely
withstand temperatures of up to 1600°F. (Anonymous, 1971). Mineral
oil is, therefore, dangerous to use where failure of equipment and
resulting fire might present a potential danger to life and property.
Equipment containing PCBs is also more reliable and longer
lasting. Small PCB-impregnated capacitors have a life expectancy of
10 to 15 years, while^jlarge capacitors last between 20 and 25 years.
PCBs in transformers need to be replaced only every 25 to 30 years.
j/Only 5 percent of all transformers contain PCBs; these are located
primarily in or near buildings or transportation facilities.
The amount of askarel contained in a transformer varies with
transformer size. The Interdepartmental Task Force on PCBs (1972)
reported that the quantity ranges from 40 to 500 gallons (150 to 1890
liters) which weighs 516 to 6450 pounds (235 to 2932 kilograms), the
average being 235 gallons (890 liters) weighing 3032 pounds (1378
kilograms). However, the largest askarel-containing transformer now
In use in Richmond by Virginia Electric Power Company holds about 600
gallons (2271 liters) (Lewis, 1976), and a transformer of the Chicago
Transit Authority contains 800 gallons (3028 liters) (Willmore, 1976).
Furthermore, Hall (1976) mentioned several transformers that hold in
4-21
-------
excess of 1000 gallons (3785 liters) and one with a content of 1500
gallons (5675 liters). Hall and Haigh (1974) reported that transformers
can contain thousands of gallons of askarel.
PCBs are used in over 90 percent of the large power-type and
smaller industrial-type capacitors presently manufactured. PCB-
impregnated capacitors are more reliable, longer-lived, and one-sixth
the size, one-fifth the weight, and one-fourth the cost of similar
oil-impregnated capacitors. Capacitors used for lighting and air-
conditioning purposes contain 0.0005 to 0.09 gallons (0.002 to
0.34 liters) of PCB each. The largest power capacitors contain about
6.7 gallons (25 liters) of askarel. (Interdepartmental Task Force
on PCBs, 1972) It is claimed that, if PCBs were no longer available
for this equipment, the time, money, technological advances, etc.,
required to replace them would result in a very serious disruption
of the nation's electrical system.
Dow Corning Company and Dow Chemical Company claim to haive
developed alternatives to PCBs for use in both transformer arid
capacitor applications respectively. The Dow Corning substitute to
be used in transformers is a polydimethyl siloxane liquid known as
"200 fluid" (Anonymous, 1975c; Mason, 1975). The product has been
on the market for about four months and is extremely thermally stable,
less explosive than PCBs, but is slightly more flammable. The
Japanese railroad system has made use of "200 fluid" for the past
four years, and field evaluations are currently in progress
4-22
-------
in Scandinavia, Britain, and France. The substitute to be used in
capacitors is an alkylated chlorodiphenyl oxide which Dow Chemical
hopes to commercialize by early 1976. The diphenyl oxide is only
suitable for power-factor-correction capacitors of the type used by
public utilities. These capacitors constitute an extimated 25 percent
of the total domestic market for PCBs (McStrock, 1976). Although a
300 percent increase in the price per unit weight of fluid is expected
to accompany the substitution of the Dow Chemical product for PCBs
(Lokey, 1975), the cost of the capacitor itself would only increase
from 10 to 20 percent (McStrack, 1976); however, Dow claims that due
to better performance, total operating costs would be reduced. One
capacitor manufacturer (McGraw-Edison) has been testing this substitute
insulator for about a year. General Electric is using a phthalate
e^r as the base for a non-PCB impregnant for capacitors (Richel,
1974). The G.E. product, known as Econol, is presently used in that
company's exports to Japan where PCBs are now banned.
Monsanto also claims to be working on alternatives to PCBs for
electrical insulation (Anonymous, 1975c) and Japan's Nippon Petro-
chemical Co. is marketing various types of naphthalenes as replace-
ments for PCBs in capacitors (Anonymous, 1975a). Contrary to the fears
of many electrical-industry spokesmen, no accidents directly related
to the ban on PCBs in Japan have been reported since the ban went
Into effect in 1972 (Anonymous, 1975a).
4-23
-------
C. LOSSES TO THE ENVIRONMENT
1. Losses During Production, Transport, and Storage
Spills and losses of PCBs during production and trans-
portation prior to 1970 are difficult to estimate, although Monsanto
claims they have been negligible (Monsanto, 1970). However, the high
level of PCBs in catfish found in waters near Anniston, Alabama, one
of the two locations at which PCBs have been produced in the United
States, constitutes evidence that effluents from the plant contained
high levels of PCBs (Nisbet and Sarofim, 1972) .
Since 1970, Monsanto has initiated a comprehensive program to
control losses during production and transport (Monsanto, 19/3a). A
drainage system has been installed in the production area so that any
material accidentally spilled will eventually be collected In one of
two 3,000-gallon settling basins. Discarded materials are either
stored or incinerated at 2000°F, a process which reduces them to C02,
H 0, and HC1 (the latter is removed prior to venting). All rupture
disc lines and atmospheric vents have been rerouted through catch
tanks or redirected to settling basins. Traps for leaks have been
installed at all sampling points. All air and gas that escape the
plant are passed through a Brink Mist Eliminator and vents are contin-
uously monitored. In addition, process temperatures were lowered to
minimize vaporization. Underground sewers were replaced with above-
ground sewers and effluents are monitored with equipment accurate in
the parts-per-billion range.
4-24
-------
Losses from the company's production unit into wastewater
effluents are presently estimated at less than one half pound per
day and that water is subsequently treated by a municipal plant. By
the time the water reaches the Mississippi River, the amount of PCBs
remaining is in the parts-per-trillion (ppt) range. Monsanto is pro-
viding clothing for its workers and is incinerating the fluid used to
clean it (Anonymous, 1971). Loading and shipping safeguards have also
been established with specific instructions written on every container
shipped. Inspections of customers' unloading facilities are period-
ically made to ensure that adequate precautions against spills are
being taken, j^ft is unlikely that such comprehensive precautions are
taken by the manufacturers of PCB-containing products. Sampling of
ventilation stacks at a General Electric Co. capacitor manufacturing
plant which had been using Aroclor 1016 for one year showed PCB levels
of 0.4 ppm (about 470 u.g/m ) with the range being 0.2 to 0.6 ppm, A
detailed description of the ventilation system was not available. Samples
of air taken 2 to 3 miles from the plant showed no detectable PCB
levels but the limit of sensitivity is not known (Levitz e_t al. , 1973).
2. Losses During Use
Except for their application as pesticide extenders, PCBs
were not intended to be discharged into the environment, ,/fheir
presence in the environment as a consequence of use was entirely
accidental. Due to their immediate and obviously disasterous effects,
the most notorious episodes were those which resulted in contamination
4-25
-------
of the food supply. However, of greatest significance from the point
of view of atmospheric contamination are those losses resulting in
either direct release to the air or release to land or water where
evaporation and codistillation may occur (see Sections V.D.I and
V.D.2).
Examples of pathways by which PCBs entered the environment
prior to 1970 are:
(a) Through vaporization from PCB-containing paints, coatings,
and plastics, etc., Nisbet and Sarofim (1972) have estimated that
10 to 20 percent of the PCBs sold domestically for plasticizer
3
applications during 1970, or 1 to 2 x 10 metric tons, were lost to
the environment through vaporization. Assuming that sales distribu-
tions were approximately the same since 1930 and that cumulative
sales in the U.S. over the period 1930 to 1970 were about 500,000
tons (454,000 metric tons), close to 30,000 metric tons of PCBs have
been released to the atmosphere by this pathway. Some specific PCB-
containing surfaces that are distributed throughout the United States
and that may be a continuous source of atmospheric contamination
through vaporization in sunlight include asphalt pavements (black top)
made with PCB-contuining waste oil, old roofing tile also made with
scrap oil, and highways whose lane dividers were painted with PCB-
containing paints (Papageorge, 1975).
(b) Improper use of discarded transformer fluid (Interdepart-
mental Task Force on PCBs, 1972). Discarded transformer fluid was
4-26
-------
used as a vehicle for an herbicide sprayed along a power-line right-
of-way near Martinsburg, West Virginia. The PCBs contaminated a dairy
cattle grazing area and were subsequently found in milk. It is
probable that release to the atmosphere also occurred during this
episode.
(c) Migration and leaching from surface coatings and packaging
materials.
1. PCBs were frequently incorporated into the formulation
of paints and sealants used on silo walls. The silage became contam-
inated as a result of dislodged wall chips or PCBs leaching into the
silage juices. In one study (Skrentny e± &\_. , 1971), gas chromato-
graphic analysis of silo wall chips showed a pattern identical to that
of Aroclor 1254. Table 4.6 summarizes the results of analyses of silage
from three farms as a function of distance from the contaminated wall.
The greatest amount of contamination was found in material collected
within the first six inches of the wall. Residues declined rapidly
as the distance from the wall increased, with none being detected
four feet from the wall. On Farm C, newly stored silage showed only
slight contamination two to four inches from the wall. After six months,
the contamination in this area increased almost 50-fold, indicating
that leaching into silage juices had occurred.
A later study (Savage et^ al. , 1973) on pit silos (as well as
upright silos) covered with plastic sheets containing PCBs revealed
contamination in two out of 31 silage samples. Values ranged from
0.04 to 0.08 ppm PCB.
4-27
-------
TABLE 4.6 PCB RESIDUES (PPM)* IN SILAGE
Farm
A
B
C
**
C
Silo
wall chips
10
10,000
2,000
2,000
0-2
0.00
24.31
31.43
15.04
2-4
0.00
4.02
0.07
3.40
Inches
6-12
0.00
0.27
0.10
0.13
from Wall
12-24^
0.00
0.15
0.10
0.00
36-48
0.00
0.13
0.00
0.00
48-72
0.00
0.00
0.00
0.00
Limit of sensitivity - 0.02 ppm
**
Six months after silage placed in silo.
Source: Skrentny, R.F., R.W. Hemkin, and H.W. Borough. 1971. "Silo
Sealants as a Source of Polychlorinated Biphenyls (PCB)
Contamination of Annual Feed." Bulletin of Environmental
Contamination and Toxicology 6(5): 409-416.
4-28
-------
These PCBs eventually ended up in the human food supply. In
1970, milk from dairy farms in Ohio, Georgia, and Florida was found
to be contaminated as a result of PCBs from silo sealants leaching
into the cattle feed.
Willett has pointed out that residues continue to appear in
silage many years after the coating is applied and long after all
visible traces of coating are gone (Willett, 1974a,b). PCB residues
were observed in silage from a si]o coated 16 years previously. Thus,
even though paints with PCBs are presently prohibited from use in
silos, the danger of contamination from old silos still exists. The
Food and Drug Administration recognized this problem and established
a temporary tolerance of 2.5 ppm on a fat basis for PCB residues in
milk (U.S. Department of Health, Education and Welfare, 1972).
Dairymen must presently face the choice of losing money from unsal-
able milk or replacing their silos. Willett tested a series of coating
materials to determine if they could provide a barrier to PCB
residues. Hydraulic cement with an acrylic bonder and water-carried
epoxy were extremely effective. Wall scrapings and silage prior to
recoating with the former material contained 20,155 ppm and 103 ppm
PCB, respectively. After recoating, silage contained 0.24 ppm and
0.10 ppm on two successive analyses. Concentrations in milk from
cows fed these silages were well below the 2.5 ppm temporary tolerance
established by the Food and Drug Administration. One year after
coating, the concentration in milk was 0.075 ppm.
4-29
-------
2. A study by the Food and Drug Administration (Interdepart-
mental Task Force on PCBs, 1972) revealed that 67 percent of food
packaging samples tested were positive for PCB residues, with the
highest value reporte^1 at 338 ppm. Nineteen percent of the foods
from these packages contained PCB residues at an average concentration
of 0.1 ppm and an upper limit of 5 ppm. Derived from carbonless copy-
ing paper recycled into food-packaging material, the PCBs probably
migrated to the food through physical contact and vaporization. PCB-
containing food wraps managed Lo contaminate human food indirectly
when bakery goods so packaged were ground up, packaging included, and
fed to poultry intended for human consumption. PCB residues in the
poultry fat were as high as 26.9 ppm. The eventual destruction of
both the original carbonless copying paper and that recycled into food
packaging was, and is, probably through burning or incineration, thus
releasing large quantities of PCBs directly into the atmosphere.
3. Jensen (1972) has estimated that 80 to 85 percent of all
small boats and 50 to 60 percent of the larger ships in Sweden have
been coated with paint, containing 3 to 5 percent PCB, on surfaces
below the water line. Plankton samples collected.in the wake of such
a boat were strongly contaminated with PCB from the bottom paint
(Jensen et_ al^. , 1972). In one sampling, there was a 14-fold
difference in concentration of PCB between plankton taken in the wake
(170 ppm) and plankton taken from the side of the boat (12 ppm).
It is suggested that extensive and continuous loss of PCB from boats
is contributing significantly to the pollution of Swedish water.
4-30
-------
The waterways of the United States may also have experienced
contamination with PCBs from painted boats in the past. However,
evidence suggests that use of PCB-containing antifouling paints is
declining. Because of the extensive use of recreational, commercial,
and naval vessels off Southern California, the Southern California
Coastal Water Project conducted a study of the PCB content of paints
used in boats docked in marinas and harbors along this coast (Young
_et al_. , 1974). Of 28 samples of commonly used paint, seven contained
PCBs corresponding to Aroclor 1242 and/or 1254 with median concentra-
tions of 0.3 and 0.7 mg/liter, respectively. From estimates of the
number of boats docked in the area, the percentage painted each year,
and the quantity of paint applied per boat, it was calculated that
for the entire fleet less that 300 grams of PCBs are applied to boat
bottoms each year. The fact that much larger quantities were used in
the past was deduced from examination of paint scrapings from boats in
Southern California drydocks. Concentrations of Aroclor 1242 ranging
from 20.1 to 3,000 mg/dry kg and of Aroclor 1254 ranging from 0.3 to
150,000 mg/dry kg were detected. These boats are still releasing
PCBs to the hydrosphere and contaminating the food chain. Aroclor 1254
levels in bay mussels from the interior regions of harbors where
bottom scraping and repainting facilities are situated were as high
as 0.9 mg/wet kg as compared to 0.05 mg/wet kg from outer control
regions. Since vaporization from paints and painted surfaces re-
presents a recognized source of PCBs in the atmosphere, direct
4-31
-------
contamination by this route may be occurring continuously. Monsanto
has not sold PCBs to paint manufacturers since 1970. It is not known
whether the low level PCB-containing paint now in use is left over
stock from 1970, is imported, or is produced from PCBs obtained from
sources other than Monsanto.
(d) Leaks from heat exchangers and partially sealed hydraulic
systems.
1. Leakage from a heat-exchange system used for pasteur-
ization of fish meal at East Coast Terminal in Wilmington, North
Carolina, resulted in extensive contamination of the fish meal with
Aroclor 1242. As a result, poor hatchability of chicken eggs was
noted at Holly Farms, Wilkesboro, North Carolina, and one poultry
producer had to destroy over 88,000 contaminated broiler chickens.
2. In 1968, leakage from a heat-exchange system contaminated
rice oil in Japan with Kanechlor 400, a PCB mixture containing 48 per-
cent chlorine (Edwards, 1971; Jensen, 1972). About 1,000 people were
eventually affected and many of them became seriously ill. The oil
itself was found to contain levels of 2000 to 3000 ppm; the average
amount of oil ingested resulted in human exposure to a total dose of
2 g (Kuratsune _et al. , 1972).
3. Residues of Aroclor 1254 were identified in sediment,
oysters, fish, blue crabs, and shrimp in Escambia Bay near Pensacola,
Florida (Duke et al., 1970). The source was the effluent from a local
4-32
-------
industrial plant which was accidently contaminated through leakage of
*
hydraulic fluid from an air compressor .
(e) Certain types of ballasts used in fluorescent-light fixtures
contain PCBs. When they burn out, leaking and smoking occurs. Al-
though newer ballasts have thermal protective cut-out switches which
prevent overheating and the consequent leaking and smoking, the older
types are still in service in many offices, laboratories, and indus-
trial plants as well as some homes. A study to determine the extent
of contamination of the air surrounding a newly burned-out ballast
with PCS was performed (Staiff ^t _al., 1974). Highest values were
found in air directly below the fixture, at the nose-level of workers
in this particular laboratory. All values, however, were well below
1 ppm. PCBs were still detectable at very low levels three days
after the burn-out. Analysis of the oily liquid which dropped from
the ballast revealed 6.2 x 10 ppm (62 percent) Aroclor 1242.
At present, domestic sales of PCBs are restricted to closed
system applications where chances of loss to the environment through
use are slim. Most of the sources of accidental loss cited above are
presumably no longer in existence or are in the process of being
phased out.
Losses in use continue to occur in the investment casting industry
(see Section IV.B.3). In the manufacture of the wax, the decachloro-
*The article states that the leak came'from a heat exchange system but
further evidence revealed that an air compressor was at fault (Nisbet
and Sarofim, 1972).
4-33
-------
-4
biphenyl is added in powdered form to the wax base. Losses of
PCB dust to the atmosphere are expected to take place both in the
process of powdering the PCBs and in the mixing itself (Versar, 1975a).
The nature of the casting operations also provide a great poten-
tial for environmental loss. During mold production, the wax is melted
and injected into metal dies. Significant quantities of wax dust and
fumes may escape during this process. The melting of the wax in a steam
autoclave also results in emission of PCBs to the atmosphere. Further-
more, if the wax is reclaimed, the water imparted to the wax during
the autoclave procedure is removed by evaporation at high temperature,
further contributing to contamination of the atmosphere.
Another operation leading to atmospheric contamination is the
removal of wax residues from the ceramic mold by firing in a furnace
at 1900 to 2000°F. Vapors from all sources leave the plant with the
stack gases or through the foundry air exhaust system and result in
contamination of the ambient air as well as the industrial environ-
ment (Versar, 1975a). Losses to the hydrosphere^ possibly through
contact with cooling water, may also occur. An effluent sample from
an investment casting plant in Michigan was found to contain 2.5 (j-8/1
of PCB (Hesse, 1975).
3. Disposal
Disposal of waste PCBs and products containing them
is undoubtedly the major source of environmental contamination.
Various methods of disposal, which may result in significant loss
to the environment, include:
4-34
-------
(a) Open burning or incomplete incineration of municipal and
industrial solid wastes.
(b) Allowing fluids containing PCBs to flow into waterways with
municipal and industrial waste effluents, treated as well as untreated.
(c) Dumping of sewage sludge, municipal and industrial solid
waste, and dredge spoil at sea.
\f (d) Dumping solid waste and sewage sludge on land in sanitary
landfills and dumps.
it
I^Zfue to vaporization and dodistillation, all of these disposal
methods are potential sources of PCBs in the atmosphere.
The unique chemical stability of PCBs prevents their being destroy-
ed by the usual waste incineration methods (Gustafson, 1970). Under
these conditions, PCBs do not burn, but vaporize. They are subsequently
carried into the atmosphere where they adsorb onto particulates and
return to the surface of the earth. It is believed that the PCBs
detected in the waxy layer on the needles of pine trees around
Stockholm are a result of vaporization due to incomplete incineration
or open burning of PCB-containing waste (Jensen, 1972). Certain PCB-
containing products are especially dangerous to burn because of atmos-
pheric contamination. For instance, PCBs are detectable 100 miles
downwind from a burning car. Incomplete incineration of recycled PCB-
containing paper and food packaging would be expected to release
large amounts of these chlorinated hydrocarbons (Young, 1975). Incin-
eration at 2000°+ F. for two seconds, however, will completely destroy
4-35
-------
PCBs. To meet the problem of disposal of scrap liquids, Monsanto has
set up a special incinerator and made incineration services available
to all of its customers (Monsanto, 1973; ANSI, 1974). The incinerator
can accomondate 10,000,000 pounds (4,500,000 kilograms) per year. One
year after the announcement that such a service was planned, 500,000
pounds (227,000 kilograms) of waste PCBs had accumulated at the disposal
site (Nisbet and Sarofim, 1972).
Municipal incinerators, which may unknowingly handle PCB-contain-
ing refuse, most likely do not reach the extreme temperatures necessary
to decompose these compounds and may be releasing them to 'the atmosphere
in large amounts. PCB residues were detected in two samples of fly ash
from particulate control devices of incinerators in Pennsylvania and
New Orleans indicating incomplete destruction (Games e_t al_. , 1973).
Thus, PCBs may also enter the atmosphere adsorbed to particulate
matter emanating from incinerators with inefficient or no particulate
controls.
Combustion of coal represents another possible route of PCBs into
the environment. PCBs were recently detected in the flue gases of a
representative coal-fired utility boiler (Cowherd et_ al_. , 1975). The
test facility was a 125-MW boiler at the Tennessee Valley Authority
Widows Creek Steam electric generating station. According to Ronald
Venezia of the EPA, there exists some question as to whether these PCBs
were formed in the combustion process, resulted from the dumping of
waste transformer oil on coal piles, or were generated as an artifact
4-36
-------
of the analytical system. Further analytical work is planned in other
power plants to confirm these preliminary results. In addition, the
Widows Creek samples will be subjected to additional analysis at
another laboratory to verify the original findings (Venezia, 1976).
Should subsequent findings indicate that PCBs are indeed formed in the
combustion process, coal-fired power plants would represent a signifi-
cant source of atmospheric PCB pollution.
The discharge of PCB-containing liquid wastes into inland and
coastal waters constitutes a serious source of contamination. Both
municipal sewage-treatment outfalls and industrial waste-water effluents
are responsible. An additional minor source is the dumping of sewage
sludge at sea (Interdepartmental Task Force on PCBs, 1972). Approxi-
mately 4,000,000 tons (3,630,000 metric tons) of sludge per year are
deposited into the Atlantic Ocean and the Gulf of Mexico, and estimates
of sewage sludge contamination ranging from 2.5 to 15.6 ppm have been
reportedC Interdepartmental Task Force on PCBs, 1972; Hesse, 1975).
Based on these estimates, it may be calculated that between 10 and
62.4 tonsC 9 to 57 metric tons) of PCBs reach the ocean annually by
this route. Nisbet and Sarofim (1972) estimated that between 1930 and
1970, 60,000 tons (54,000 metric tons) of PCBs were deposited into the
waterways of the United States from all sources. Because of their
very low aqueous solubility, PCBs discharged into a river, lake, or
coastal water will accumulate in the sediment in high concentration and
redissolve very slowly ( Flshbein, 1974). Thus, they present an
4-37
-------
immediate threat to life at the bottom of bodies of water and a chronic
threat to life in the water. The highest concentrations of PCB resi-
dues in freshwater fish occur in rivers that are associated with indus-
trialized areas. Concentrations ranging from 13.4 ppm to greater than
100 ppm have recently been found in fish in the Hudson River (Richardson,
1975). The source of this gross contamination was traced to the
General Electric Company (G.E.) which operates two plants on the river
about 60 miles north of Albany. G.E. has admitted to dumping as much
as 30 pounds per day into the Hudson and claims that this figure may
have been even higher in the past. In response to pressure from both
the New York State Environmental Conservation Department (which seeks
zero discharge of PCBs from G.E. into the Hudson by September 30, 1976)
and the United States Environmental Protection Agency (which stipulates
a limit of 99.84 g/day by 1977), G.E. has, for the present, managed to
reduce its discharge to approximately 2000 g/day. However, Ogden Reid,
New York State Environmental Commissioner, says that irreparable and
irreversible damage may already have been inflicted on the river
(Richardson, 1975). Tables 4.7 and 4.8 summarize data that has been
collected on actual PCB concentrations in industrial waste-water ef-
fluents and municipal sewage-treatment-plant outfalls, respectively.
Veith and Lee (1971) claimed that the occurrence of significant concen-
trations of PCBs in municipal waste constitutes evidence that PCBs were
widespread and possibly present in consumer products.
4-38
-------
Table 4.7 PCS CONCENTRATIONS IN INDUSTRIAL EFFLUENTS
Locatiba
Saukville, Wise.
Ohio - Great
Miami River
Florida-Escambia
River
Kind of
Industry
Chemical Plant
Paper Coating
Company
Paper Treatment
Appliance
it
Chemical Plant
Date
3/70
1/71
1/71
1/71
1/71
4/69 -
10/69
Aroclor
Compound
Detected
1242
1242 and
1248
1242
1254
1254
1254
Concentration
in Effluent
(ppb)
2.50
27
430, 470*
5
18
2.5-275
Source of Data
Vieth and Lee,
1971
EPA data -
Analytical
Quality Control
Lab
M
tl
II
Duke^al. , 1970
Samples from treatment lagoon.
Spurce: Interdepartmental Task Force on PCBs. 1972. "Polychlorinated Biphenyls and the Environment '
(Com-72-10419), Washington, D.C.
-------
TABLE 4.8 CONCENTRATION OF PCBS IN MUNICIPAL SEWACE TREATMENT
PLANT OUTFALLS
Arorlor j,s( . peg
Collection Compound ug/l Flow per, Dls. harged Source
Collection Site Date Detected (ppb) Day Mgd ] per day (Ibs) of Data
Ohio-Miami River
Dayton
Hamilton
Mlddleton
Wlsconsin-
Milwaukee River
Welt Bend
Fredonla
Saukvllle
Grafton
California
East Bay
(Sen
Francisco)
San Francisco
Terminal Island
Orange County
Hyperion
White Point
Richmond
San Diefco
Oxnard
1/19/71 1254 17 48 62 EPA
0800 Unpb.
1/20/71 Data
0800
1/19/71 1248 10 8 06
0000
to
1/19/71
2400
1/19/71 ND*
0800
to
1/20/71
0800
3/26/7(1 1254 0.2', 1.4 0.00? Veithi
Lee, 1971
" 1254 0.12 0.1 0.002 "
" 1260 0. D 0.1 0.002 "
1254 0.0'* 0.8 0.002 "
12/70 1254 3.1-3.8 155 4.2 Schmidt
et al. ,
1971
1254 3.8-5.8 31.5 1.2
1254 5.8-12.8 9.3 0.7 "
" 1254 0.21-0.64 130 0.4 "
1254 0.16-0.37 i40 0.7 "
" 1254 76 )50 213 "
ND*
A
ND
A
ND
1 Mgd - million
Nl) - not detected
oer day (8,000,000 Ibn/diw)
Source: Interdppartmenta I Task Force on I'CBs, 1972. "I'olychlorinated Illphenyls and the
Environment", (Com-72-10419) , Washington. D.C.
4-40
-------
A recent study designed to determine the major sources of PCBs
in the marine ecosystem off Southern California revealed some surprising
results (Young et al., 1975a, 1975b). The primary source appears to
be the submarine discharge of municipal waste water effluents. All
major municipal waste water systems release the PCBs in similar
quantities ( 1 to 3 metric tons per year) and the total loss for the
year 1973-1974 was 6.5 metric tons. Evidently, direct release from
industrial sources in this area is insignificant. Also, 80 percent
of the PCBs released via municipal effluents consisted of Aroclor 1242,
the remaining 20 percent consisting of Aroclor 1254 (Young and Heesen,
1975).
Another study of waste water effluents from eleven Wisconsin
cities is summarized in Table 4.9 (Dube et al., 1974). One of these
cities, Cedarburg, was chosen for a more extensive study of concentra-
tion as a function of time throughout a 24-hour period. These results
are presented in Figure 4.3. The concentration of PCBs in the raw
waste water (Influent) began to increase at 8:00 a.m., the beginning
of the working day; the concentration rose from 0.54 ppb to a maximum
of 3.1 ppb at 4:00 p.m. (1600 hours). The concentrations in the final
effluent began increasing from 0.33 ppb at midnight (2400 hours) to a
maximum of 0.77 ppb at 2:00 p.m. (1400 hours). Levels in both
influent and effluent increased and decreased simultaneously through-
out the day. Thus, in sampling to determine the extent to which
waste effluents contribute to environmental contamination, it is
extremely important to choose an appropriate time of day.
4-41
-------
TABLE 4.9
PCB CONCENTRATIONS IN WISCONSIN TREATMENT PLANT EFFLUENTS
CITY
8eaver Djim
Port Wddli I fiy i an
Ornfton
Cedarburg
Rac ine
Burlington
Lake (lencva
Waluorth
Bcloit
Fort Atkinson
Portage
SAMPLING
Half
(1171)
1/20
t'/)9
\IJ.l
10/6
\UO
1/19
1/22
10/6
)/9
2/19
1/22
10/6
1/20
2/19
3/22
10/6
10/6
10/6
1/20
2/19
t/22
10/6
1/2(1
2/19
3/22
10/6
1/2(1
2/19
1/22
10/6
1/19
2/20
3/22
10/6
1/19
2/20
3/22
10/6
1/19
2/20
3/22
10/6
5/20
9/20
10/6
Time
(>: 10
6: 10
6: 10
b: 10
H: 10
Hi 10
H: JO
8i30
9: 10
9: 10
9: )0
9: 10
10:30
10:30
10:30
10:30
10:30
]0:30
8:30
12:00
1 2 : DO
l?:10
9: 10
I 1 : 00
1 J:00
13:00
10: 10
14:00
14:00
14:00
11:00
14:30
14: JO
14: 10
12:30
16:00
16:00
16:00
2:00
18:00
18:00
18:00
10:00
10:00
10:00
PCB font.! li-
ft
trat liuid
(|'X/D
O.O'i
0.0')
• [) ()')
• l).(l')
0 1',
0.1.'
(l./'i
0. 1')
0. 12
0.1 1
0.2 1
0.07
0.48
0.28
0.97
0.91
1.1
1.0
0.72
0.60
0. 7>,
0.81
0 . 1 4
0.09
O.OH
0. 1^
/».'.
i.l
2.8
2.4
0.17
0.21
0.34
0.18
0.11
0.07
0.06
0.14
0. 15
0.07
0.10
0.08
42
18
32
Avinfie
1 low
(uiK'U
i . yj
4.11
1. '<(>
2. 1'.
o. in
\.ll
I./.6
1 . HO
0.84
1 .00
0.78
0.84
1.0r)
1.17
2.r>7
1.95
I.")1)
1.95
16.45
25.06
23.01
20. ',1
1. 74
/.hi
1 .Vt
1. 15
0.56
0.66
0.78
0.54
3.89
11.75
7.56
4.42
1.24
2.48
1.85
1.30
0.79
1.1)0
0.78
f Nt iDWtlid M0BB
TrHnnj>ort**
f lir'lL I'CB/day)
') 1
-
_ _
- -
/ . 1
---
0.8
1.1
1.5
4.2
2.7
15
18
16
142
_-_
1. /
—
18
—
—
3.8
5.2
1.5
—
The chroma tn^r HIM fr4; tlic- chrornntogram from the Portage mimplfiB mo»t
closily ra»«nibled th<> cltromatourfitn tjf Amdor 1^/tH.
Md»* I l'iliN|iil<">
tnuULplliid ny nv«tA|i> d«
bnm-d
obnfrvccl dminy sampling
Note: MK
-------
CONCENTRATION
(in (Jig/1)
3.5
3.0
2.5
2.0
1.5
1.0
0.5
•- INFLUENT
— EFFLUENT
A
,".
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
f
i
JL
8 12 16 20
TIME (in hours)
24
Source: Adopted from Dube, D. J., G. D. Veith and G. F. Lee, 1974.
"Polychlorinated Biphenyls in Treatment Plant Effluents,"
Journal of Water Pollution Control Federation 46; 966-972.
FIGURE 4.3
HOURLY CONCENTRATIONS OF POLYCHLORINATED BIPHENYLS IN
THE INFLUENT AND EFFLUENT FROM THE CEDARBURG, WISCONSIN, TREATMENT
PLANT, APRIL 15, 1971
4-43
-------
In the Cedarburg study, it was noted that the PCB concentration
in the effluent was about 30 percent that of the influent and that
the concentration in sludge was 1,000 times higher than that of the
effluent. This treatment plant, therefore, removes 70 percent of the
PCBs prior to discharge of the waste waturs.
A laboratory study waa initiated to ascertain the effectiveness
of a secondary aerated biological oxidation system in removing high
concentrations of PCBs from the effluent of primary treatment at the
Jackson Pike (municipal sewage treatment) Plant in Columbus, Ohio
(Choi et al., 1974). The laboratory system consisted of an aeration
tank and a clarifier. The system was operated for 5 to 11 days in
order to reach a steady state condition prior to introduction of PCB
for a period of two to three days. Four to eight days of final
operation followed. The results are presented in Table 4.10. PCB
concentrations in feed, effluent, and sludge were 0.009,<0.001, and
0.07 ppm, respectively, prior to introduction of experimental levels.
Concentration in sludge was 3600 times that in effluent on Day 3, the
day of highest total concentrations. The PCB is probably dissolved
in fats present in sludge, adsorbed at the surface of the suspended
material in sludge or ingested by the microbial cells in sludge. It
was concluded that a tertiary treatment system such as activated
carbon might be necessary to further reduce the concentration in
secondary system effluents.
4-44
-------
TABLE 4.10 PCB CONCENTRATIONS IN SLUDGE AND EFFLUENT FROM
AN EXPERIMENTAL BIOLOGICAL OXIDATION SYSTEM
Day after PCB added
1
2
3
5
6
10
Concentration in
feed (ppm)
16.88
16.88
16.88
0
0
0
Concentration in
sludge (ppm)
36.2
109.6
152.0
105.6
112.0
84.0
Concentration in
effluent (ppm)
0.006
0.015
0.042
0.032
0.022
0.015
Source: Adapted from Choi, P.S.K., H. Nack and J.E. Flinn, 1974, "Distribution
of Polychlorinated Biphenyls in an Aerated Biological Oxidation Waste-
water Treatment System". Bulletin of Environmental Contamination and
Toxicology. 11(1): 12-17. '—
-------
The United States Environmental Protection Agency has proposed
a final water quality criteria level for PCBs in fresh water of
1 ng/1 (1 ppt) (Anonymous, 1975e), and it is believed that effluent
standards will be revised shortly in accordance with this proposed
criteria (Kopp, 1975).
The 1 ppt criteria is half the ambient level considered desirable
by the EPA in 1972 (2 ng/1 or 2 ppt) (U.S. Environmental Protection
Agency, 1973a). Based on the 2 ppt level, effluent standards were
proposed in 1973 requiring that the daily average PCB concentration
in effluents must not exceed 280 jag/1 or a total of 0.0294 kg per day
(0.0648 pounds per day) into streams and 0.0245 kg/day (0.0540 pounds
per day) into lakes (U.S. EPA, 1973b). Monsanto (1974d) argued against
this standard as unrealistic from an environmental as well as a tech-
nological and economic point of view. PCBs, they claim, bioaccumulate,
dissipate, and biodegrade at different rates and must, therefore, not
be clumped together as of equal environment concern. With the ex-
ception of direct release near points of manufacture or use (which
can be eliminated), the PCBs found in the environment are the penta-
chloro and higher isomers, even though 65 percent of all PCBs manu-
factured over the years contain four or fewer chlorine atoms.
Laboratory studies demonstrated that PCB isomers undergo degra-
dation at rates dependent upon the number of chlorine atoms (Monsanto,
1972), and that lower-chlorinated isomers disappear through metabolism
and/or excretion from the tissues of birds and mammals. Therefore,
4-46
-------
Monsanto argues, the lower-chlorinated species should be excluded when
effluent standards are calculated. They also suggest that background
levels already present in waterways be subtracted before a final effluent
concentration is decided upon. The final concentration that Monsanto
feels is reasonable is 1 pound (454 grams) in effluent for every 100,COO
pounds (45,000 kilograms) produced (Monsanto, 1973a).
Industrial effluent standards (other than at the manufacturing
plants) may be possible to meet now that many industries are no longer
using PCBs in their products. Evidence that Monsanto's restriction
of sales in 1970 may have had an immediate effect on some river systems
was obtained by Veith (1972) . These observations are summarized in
Table 4.11. These data on rivers discharging into Lake Michigan show
that PCB concentrations in two major tributaries decreased dramatically
to below the detection limit seven, months subsequent to the sales re-
striction.
In addition, the American National Standards Institute, in their
1974 guidelines for handling and disposal of PCB-containing askarels
(American National Standards Institute, 1974), prohibits the disposal
of waste askarel from capacitor and transformer factories down effluent
drains and sewers and recommends incineration as the best method of
disposal.
Since the highest concentration of PCBs from municipal sewage-
treatment plants ends up in sludge, the manner in which sludge is dis-
posed of is particularly important. A summary of data collected on
4-47
-------
Table 4.11 CONCENTRATIONS OF PCB (|jg/l as Aroclor 1254)
IN SOME RIVER SYSTEMS
River
Peshtigo River
Oconto River
Pensawkee River
Big Suamico River
*
Fox River
(as Aroclor 1248)
12/29/70
0.31
0.45
<0.01
<0.01
0.18
5/21/71
0.38
0.16
<0.01
<0.01
0.26
7/20/71
<0.01
<0.01
<0.01
<0.01
0.16
8/6/71
<0.01
<0.01
<0.01
<0.01
0.15
The Fox River is near a highly industrialized area.
Source: Adapted from G.D. Veith, 1972. "Recent Fluctuations of
Chlorobiphenyls (PCBs) in the Green Bay, Wisconsin Region"
Environmental Health Perspectives, Exp. 1: 51*54.
4-48
-------
actual concentrations of PCBs in sewage sludges in the United States is
presented in Table 4.12. Sewage sludge is disposed of by dumping it
into the sea, depositing it in landfills, spreading it on land, or by
incineration. In Britain, the high PCB concentrations detected in water
from the Firth of Clyde was attributed to the routine dumping of crude
sewage sludge from treatment plants in Glasgow and adjacent areas (Hoi-
den, 1970). In Britain, effluents from industrial sources are combined
with those from domestic sources for purification. Many sludge samples
so combined gave gas-liquid chromatograms similar to those of Aroclor
1254 and 1260 and ranged in concentration from <0.1 to 14 ppm (1 to 185
ppm in terms of dry material). Plankton samples were taken at Garroch
Head (near the site of dumping) and at progressively further distances
from the source and their PCB concentrations were measured. A marked
gradation was observed from the Clyde (0.5 ppm) to the open waters of the
Atlantic 400 miles from Scotland (0.041 ppm) (Williams and Holden, 1973).
Dumping at sea is therefore not the method of choice for disposal
of sludge. Whether dumping on land or into sanitary landfills is any
better is questionable. Leaching is theoretically possible when dis-
posal sites are above permeable sand, gravel, or creviced bedrock forma-
tions. Toxic substances entering shale or clay, however, most likely
do not migrate (Walker, 1973).
In a laboratory study (Tucker _et al., 1975) , distilled water was
percolated through columns packed with several types of soil coated with
Aroclor 1016 and the effluent waste was .analyzed for PCBs. Breakthrough
of PCBs into the effluent water appeared related to the clay content
4-49
-------
Table 4.12 CONCENTRATION OF PCBs IN SEWAGE SLUDGES
i
w
o
Collection Site
California
Hyperion (Los Angeles)
Barstow
Ohio
Dayton (Miami River)
Little Miami
(Cincinnati)
Mill Creek
(Cincinnati)
Lebanon (Turtle Creek)
Shayler Run
Virginia
Lorton
Indiana
Indianapolis
Date
12/70
7/21/71
Aroclor
Detected
1254
1254
1254
1254
1254
1254
1254
1254
1254
Concentration
>g/g dry solids)
*
851
(78.5-92.1)
1,400
105,000
32,000
12,700
2,500
3,200
1,200
3,800
Sludge
per day
tons/day
1
20,000
1.4
47.9
20.2
88.3
1.0
126.1
Est. PCB
Content
Ibs/day
3.2
.004
10.1
1.3
2.2
.005
1.03
Source
Schmidt, et al. , (13)
EPA Unpublished data
EPA Unpublished data
11
it
u
1
II
It
This number is based on outfall discharge and represents a relatively dilute sludge. The estimated PCB content
in Ibs/day is the important figure here.
Assumptions: Each nillion gallons of sewage contains about 1 ton of sludge. The daily output of sludge, then
is 150,000,000 sewered population x 130 gal. sewage per day = 19,500,000,000 gallons per day and 19,500 tons of
sludge per day nationwide.
Assuming 10 ppm of PCBs in sludge, the daily output would be 19,500 tons x 2000 Ibs. = .39 million Ibs. x
10 ppm = 390 Ibs. per day, at 1 ppm, 39 Ibs/day. These would be respectively 70 and 7 tons/year.
Source: Interdepartmental Task Force on PCBs, 1972. "Polychlorinated Biphenyls and the Environment",
(Com-72-10419), Washington, D.C.
-------
of the soil. As expected, soils containing higher percentages of clay
retained the PCBs more effectively. But even in the worst case, less than
O.C5 percent of the total Aroclor present was leached from the soil after
a four-month exposure to a total of 100 liters of water. [The soil column
height was 9 inches (23 centimeters); the diameter was 3 inches (8 centi-
meters)]. In addition, only the less-chlorinated, more easily degradable
isomers were leached. From the results of this study, it would appear
that leaching from landfills is not a significant source of environmental
contamination and that landfills might be one possible method for sludge
disposal. However, samples of surface runoff water from 5 landfills in
Michigan showed concentrations of PCBs ranging from 0.04 to 0.30 micro-
grams/liter (ppb). Concentrations of PCBs in runoff water from four other
Michigan landfills were below the 0.03 microgram/liter limit of detection
(Hesse, 1975).
The method of choice for sludge disposal is incineration. PCBs in
sewage sludge can be completely destroyed by this process (U.S. Environ-
mental Protection Agency, 1972).
Nisbet and Sarofim (1972) have estimated that 300,000 tons (over
290,000 metric tons) of solid waste (including sludge) were deposited
in dumps and landfills between 1930 and 1970. Analysis of stagnant water
close to Swedish landfills revealed no detectable levels of PCBs where
the limits of detection were 4 ppb (Lidgett nnd Vodder, 1970). At present,
the approved method for disposal of scrap capacitors and the hardware
from scrapped transformers is to seal them, in containers and bury them
4-51
-------
in sanitary landfills (American National Standards Institute, 1974).
This same method is used for any scraps of solid waste which are saturated
with askarel. Monsanto believes that increased use of Aroclor 1016 for
capacitor and transformer applications will significantly reduce the .
amounts of PCB isomers with six or more chlorine atoms released into the
evnivronment from scrapped apparatus (Monsanto, 1972). Most of these
PCBs will end up in landfills and the rest will be incinerated.
The high cost in investment casting wax has stimulated many foundaries
to reclaim the wax after use. Even so, it is considered probable that
large quantities of PCB-containing wax are disposed of, primarily in
landfills. Wastes from spills and equipment clean-out are included
with the unreclaimed wax and excess reclaimed wax in solid waste disposal
(Versar, 1975a).
4-52
-------
V. MEDIA DISTRIBUTION, TRANSFORMATION, AND TRANSPORT
A. SUMMARY
The ubiquitous polychlorinated biphenyls have apparently
found their way into virtually all compartments of the environment.
They were presumably released in highest concentrations prior to 1971,
primarily to the air and water near urban, industrialized areas.
However, data gathered subsequent to that date show significant
levels of PCBs in air, fresh water, and marine samples from both
industrialized and remote areas, suggesting that contamination is
still occurring. Field studies have revealed PCB contamination in soil
and laboratory studies indicate that certain types of soil readily
adsorb PCBs and most likely contain them.
Atmospheric transport is the major pathway by which PCBs
are carried to locations far from their point of entry into the en-
vironment. They are transported either as a vapor or adsorbed onto
particulates and are deposited on land or into bodies of water by
particle fallout and precipitation. Most of the PCBs reaching the
hydrosphere are removed by particle scavenging, followed by sediment-
ation, and are only slowly released; small amounts remain dissolved
and are subject to re-evaporation. Successive cycles of evaporation,
adsorption, and deposition eventually carry the PCBs initially re-
leased on land to the coast and to their ultimate sink, the sediment
of the oceanic floor.
5-1
-------
B-. ATMOSPHERE
Atmospheric transport plays a, if not the, major role in the
worldwide dissemination of polychlorinated biphenyls (SHdergren, 1972;
Oloffs jet al. , 1972; Mackay and Wolkoff, 1973; Harvey and Steinhauer,
1974). The virtually universal distribution of PCBs in the global
environment, including the Arctic and other areas protected from
alternative routes of contamination, constitutes reasonable evidence
for this phenomenon (Risebrough and deLappe, 1972).
The various pathways by which PCBs have entered the atmos-
phere over the past 40 years include incomplete incineration and burning
of PCB-containing wastes in garbage dumps; vaporization from plasti-
cizers, paints, and other coatings; vaporization from soil, sand, and
other terrestiral reservoirs; evaporation and codistillation from
natural bodies of water; and evaporation from their own surfaces
when these have been exposed to the open air as in the case of acci-
dental spillage, leaks, or wear and weathering of PCB-containing
products.
Nisbet and Sarofim (1972) claim that the contribution to atmos-
pheric levels made by re-evaporation of the PCBs deposited into
bodies of water will be negligible because of their low concentra-
tions. But recent studies suggest that this may indeed be a signifi-
cant route of contamination (see Section V.D.I). In addition, the
failure to account for the bulk of the more than 2.8 billion kilograms
(kg) of DDT dispersed in the environment since the 1940s, in
5-2
-------
conjunction with the observation that DDT vapor can be converted to
PCBs by irradiation with ultraviolet light of the same wavelength
present in sunlight in the lower atmosphere (Maugh, 1973), provides
substance for speculation as to other possible sources of PCBs in
the air. Evidence supporting the hypothesis that PCBs in the atmos-
phere may in fact be formed from DDT was provided in a recent study by
Young and McDermott (1975) in which PCBs in dry aerial fallout were
determined at various locations within the Los Angeles Basin. The
highest levels (up to 1.7 x 10~ g/square meter/day) occurred in the
vicinity of the Montrose Chemical Company, which manufactures DDT.
The second highest levels (up to 1.6 x 10~6 g/m2/day) were found
near the Kazarian landfill, the present disposal site for Montrose
waste. By comparison, the Rolling Hills landfill, the disposal site
for Montrose waste up until about 1972, showed levels ranging from
-7 -7 2
2.4 x 10 to 6.0 x 10 g/m /day. PCB levels at other stations
throughout the Basin ranged from 7.8 x 10~8 to 9.3 x 10~7 g/m2/day.
The period of greatest loss to the environment, including the
atmosphere, presumably occurred during the decade 1960-1970 when
PCBs were widely used in a variety of non-containable applications.
Nisbet and Sarofim (1972), in their attempt to formulate a generalized
description of the transport of PCBs through the North American
environment, estimated that 1500 to 2500 tons a year of PCBs were
being released directly into the atmosphere through vaporization from
plasticizers (mainly Aroclor 1248-1260) and incomplete burning or
5-3
-------
incineration (Aroclor 1242). These figures were based on the 1970
sales by category of the Monsanto Company and reflected the authors'
assumptions as to the useful service life of various PCB containing
products and their logical modes of disposal, j^ontinuing with their
proposed model, it was estimated that the cumulative input of PCBs
into the atmosphere from 1930-1970 amounted to about 3 x 10 tons.
Due to chemical decomposition and metabolic transformation, a large
proportion of the less-chlorinated isomers in the various Aroclor
mixtures are expected to have been degraded. Consequently, about
2/3 (2 x 10 tons) of the amount originally lost to the air is
believed to still remain in the environment, deposited on land, in
the sea, or in fresh waters.
It can be expected that the highest concentrations of PCBs in
the atmosphere will be found over industrialized urban areas. The
PCBs may be found in the vapor phase, in the form of an aerosol, or
adsorbed onto particulate matter. By analogy to DDT, Nisbet and
Sarofim assumed that most airborne PCBs would be adsorbed onto par-
ticulates. Assuming the average rate of urban particulate production
to be 2.67 x 10 tons per year, a rough estimate of 50 to 80 (o.g PCB/
g of particulate in urban areas was calculated. An unpublished study
by the U.S. Environmental Protection Agency of PCB levels on suspended
particulates in four U.S. cities between 1968 and 1970 showed concen-
trations ranging from 27 to 230 |ag PCBs per g of particulates
(Interdepartmental Task Force on PCBs, 1972) with a mean concentration
5-4
-------
at 50 ug/g of particulates. Assuming a particulate loading of
o
60 ug/m , this average amounts to 3 ng/m3 PCBs. These values are
consistent with the proposed model. In a more recent study (Kutz and
Yang, 1975), samples of ambient air were collected from suburban lo-
cations in three U.S. cities (Miami, Florida; Jackson, Mississippi;
and Fort Collins, Colorado) and analyzed for polychlorlnated biphenyls
by electron capture gas chromatography. Preliminary results for
samples taken in April, May, and June of 1975 show that PCBs were
present at all locations at an average concentration of 100 ng/m3.
The PCBs detected most closely resembled Aroclor 1254. No mention
was made, however, as to what percentage of the observed PCBs was
adsorbed onto particulates. Most of the PCBs adsorbed onto large
particulates would be expected to be deposited on the land and nearby
coastal waters within three days and would not be available for long-
range transport.
The observation that PCBs must travel a long way by air in
order to reach certain remote areas in which they have been found leads
one to conclude that transport on small particulates or as a vapor
must also take place to a significant extent as well.
Studies confirming the basic assumptions of the Nisbet and
Sarofim calculations have been reported. The rate of input into the
North American atmosphere estimated at 1.5 to 2.5 x 103 tons per year
corresponds to a continent-wide mean rate of fallout of 60 to 100 jog/
2
m /year. In 1971, Persson reported a total fallout of 78 ug/m2 in
5-5
-------
southern Sweden for a five month period in 1970 (Panel on Hazardous
Trace Substances, 1972). This value is at least of a similar order
of magnitude to that predicted by Nisbet and Sarofirn. A second,
rather detailed study on the monthly rate of fallout of PCBs in eight
sites in Sweden during 1970-71 was performed by SBdergren (1972).
In this study, the amount of PCBs entering the areas investigated
via airborne participates and precipitation was determined. The
locations of the fallout stations and the relative frequency of wind
directions during several months of the year are shown in Figure 5.1.
The most frequent direction is west to southwest. This means that
the study area is subjected to airborne contaminants originating
from the industrialized regions along the Swedish west coast as well
as from those in Copenhagen and West Germany. Stations A, D, E, F,
and G are located in agricultural districts; Station B is located in
the city of Malmo close to a municipal refuse disposal plant; Sta-
tion C is located in the city of Lund; and Station H is located in a
forest district. The results of this study are summarized in Table
2
5.1. Values ranged from 550 to 10,510 ng/m / month or from 7 to
2
126 |J.g/m year, once again in the range predicted by Nisbet and
Sarofim. The highest levels of PCBs were detected at Stations A and
B, in the western part of the region. Despite the differences in
type of location, PCB concentratLons at these stations were very
similar. Low to moderate levels of PCBs were observed at the more
easterly stations C, E, F, G, and H. The unexpectedly high level at
5-6
-------
1:300,000
10 20 30km
Source: Adopted from Sodergren, A. 1972. Chlorinated Hydrocarbon Residues
in Airborne Fallout , Nature 236, 395-397
FIGURE r>.l
THE LOCATION OF SAMPLING STATIONS FOR THE DETERMINATION
OF PCB RESIDUES IN AIRBORNE FALLOUT IN SOUTHERN SWEDEN
5-7
-------
Table 5.1 AVERAGE MONTHLY LEVELS OF PCB IN AIRBORNE FALLOUT COLLECTED
IN SOUTH SWEDEN (ng m~2 month-1)
Collecting
Site*
A
B
C
D
E
F
G
H
Period of Collection
January-March 1971
January-March 1971
January-March 1971
January-March 1971
August-October 1969
February-April 1970
May-July 1970
August -October 1970
November-December 1970
January-March 1971
January-March 1971
January-March 1971
January-March 1971
PCB
10,510
9,860
1,740
5,980
910
4,460
1,590
550
1,800
850
1,900
2,190
620
* The Symbols for the Collection Sites refer to Figure 5-1.
Source: Adapted from Sddergren, A., 1972. Chlorinated Hydrocarbon
Residue in Airborne Fallout. Nature 236: 395-397.
5-8
-------
Station D was probably due to some unidentified local source. During
the period February-April 1970, Station E experienced a significant
increase in PCB fallout. This period was characterized by a high
mean rate of precipitation. The evidence presented in this study
provides some support for a model in which PCBs are carried through
the air from industrialized regions to other areas in the path of the
prevailing winds. This journey is occasionally shortened when the
atmosphere is cleansed of particulates during precipitation.
Evidence that PCBs may reside in the atmosphere mainly as
vapors was provided in a study by Bidleman and Olney (1974a). With
the development of a new method for collection of polychlorinated
biphenyl vapors (Bidleman and Olney, 1974b), it became possible to
sample hundreds of cubic meters of air per day. Air was pulled at
3
0.4-0.8 m /rain, through a glass-fibre filter and then through a
plug of polyurethane foam. The filter is capable of removing 98
percent of particles having radii greater than 0.015 urn while PCB
vapors are trapped on the polyurethane foam with an efficiency of
greater than 90 percent. During the period from February to June
1973, PCB concentrations in the marine atmosphere of the Bermuda-
Sargasso Sea area of the North Atlantic were measured. The samples
were collected from a tower 20 meters high located at High Point,
Bermuda. Marine air was also sampled from a tower 8 meters high on
the bow of the R.V. Trident during a cruise from Bermuda to Narragansett,
Rhode Island, in June 1973. Finally, local collections of PCBs were
5-9
-------
made on the University of Rhode Island campus and in downtown Providence.
The results are presented in Table 5.2. Only a small amount (2 per-
cent) of the PCBs detected were found on the glass-fibre filter,
suggesting that, contrary to the prediction of Nisbet and Sarofim,
adsorption onto particulafces is not the major mode of transport of
PCBs in the atmosphere. However, it is possible that PCBs volatilize
from trapped particles, pass through the filter, and are subsequently
collected by the polyurethane foam. The gas chromatograms of the PCBs
in the atmosphere matched those of Aroclor 1242 and 1248, approximately
90 percent of which are tetrachloro or lower isomers. No concen-
tration gradients were observed in going from the open sea towards
Rhode Island although the urban concentrations were significantly
higher (2.1-9.4 ng/m in Rhode Island as compared to 0.21-1.6 ng/m3
at sea). Since no wind direction, etc., was given, the explanation
for the absence of a gradient ±s not obvious.
The question of whether or nor atmospheric gradients of PCBs
exist from industrialized areas to the open sea was investigated by
Harvey and Steinhauer (1974). Four stations over the western North
Atlantic were sampled between January and June of 1973. The loca-
tions of sampling stations are shown in Figure 5.2 and the analytical
results are presented In Table 5.3. The PCB concentrations were
highest in Vineyard Sound located about 150-250 kilometers (km) from the
Boston-Hartford -New York-New Jersey industrial complex. The concen-
trations over the Grand Banks, located over 2000 km from the north-
5-10
-------
TABLE 5.2
CONCENTRATIONS OF POLYCHLORINATED BIPHENYLS IN MARINE AND CONTINENTAL AIR*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Collect ion
dates
(1973)
2/12-2/13
2/13-2/14
2/15-2/16
2/16-2/18
2/19-2/28
2/29-3/9
4/8-4/11
4/11-4/17
6/4-6/5
6/5-6/6
6/7-6/8
6/8
6/9
1/18-1/19
1/21-1/22
2/4-2/5
5/8
Location**
Bermuda
Bermuda
Bermuda
Bermuda
Bermuda
Bermuda
Bermuda
Bermuda
33°20'N, 65°14'W
34°39'N, 66015'W
37°39'N, 68°12'W
38°48'N, 69°14'W
40°32'N, 70°20'W
U.R.I.
U.R.I.
U.R.I.
Providence, R.I.
Volume
of air
(m3)
1070
1320
918
1950
1740
732
1300
860
300
267
402
222
196
392
1071
744
76
PCB
(ng/m )
0.59
0.30
0.65
0.62
0.55
0.52
0.61
0.21
1.6
0.79
0.72
0.83
4.0
2.1
5.8
9.4
•The total retained by the glass-fiber niter and trapped on the polyurethane foam
™
Filter-
**Samples 9 through 13 were collected from the R.V. Trident while the ship was en route.
location marks the midpoint of the collection track.
Calculated as Aroclor 1242 or Aroclor 1248.
The
Source: Adopted from Bidleman, T. F., and C. E. Olney. 1974. Chlorinated Hydrocarbons in the
Sargasso Sea, Atmosphere and Surface Water. Science 183: 516-518.
-------
ioo° 9
-------
TABLE 5.3 PCB CONCENTRATIONS OVER THE WESTERN NORTH ATLANTIC
Station
Bermuda
(32°20tN;64°40lW)
(32020'N;64040'W)
(32°20'N;64°40'W)
Georges Bank
(41°40'N;67°30'W)
(41°40'N;67030'W)
(41°40'N;67°30fW)
(41°40'N;67°30'W)
(41°40'N;67°30IW)
Vineyard Sound
(41020fN;70050'W)
Grand Banks
(45°16'N;52°08'W)
(45°16'N;52008?W)
(45°16'N;52008'W)
(45°16'N;52°08'W)
Date
(1973)
12 February
13 February
14 February
15 February
10 April
13 April
15 April
17 April
19 April
21 April
13 April
30 April
25 June
26 June
27 June
28 June
29 June
Sample
volume
(m3)
560
480
820
500
105
675
660
655
640
650
105
224
780
960
840
940
540
.Wind
direction
WNW
W
Variable
S
NW
NW
NE
NW
SW
sw
SW
sw
ssw
sw
wsw
wsw
W
PCB ng m
(calc. as
Aroclor 1254)
0.5
0.4
0.16
0.15
1 A
0.82
0.58
0.61
0.80
i fin
-j q
C 5
f\ : i^.
0.07
0.10
0.16
0.05
Source: Harvey G R and „ G. Steinhauer. 1974. Atmospheric Transport
of Polyehlorobaphenyls to the North Atlantic. Atmospheric
Environment 8: 772-782. ^ •
-------
eastern industrial areas, were 100 times less than those over Vineyard
Sound. The two middle stations showed intermediate levels of atmos-
pheric PCBs. The PCB level reported over Bermuda agreed well with
those observed by Bidleman and Olney in the previous study. When
the average concentrations at each station were plotted against the
average distance of the site from major industrial centers, the sea-
ward decrease was seen to be exponential (Figure 5.3).
The method used for sample collection was one in which vapor
phase PCBs and those adsorbed onto particulates could be differen-
tiated. Once again, particulates were trapped on fiberglass
filters, while PCB vapors were collected in silicone oil traps. Anal-
yses of the fiberglass filters revealed PCB concentrations ranging
from undetectable in the Grand Banks samples to 0.04 ng/m over Vine-
yard Sound. Thus, only 1 percent or less of the total PCBs in the
atmospheric samples were adsorbed onto particulates, providing further
evidence for vapor phase transport.
C. HYDROSPHERE AND LITHOSPHERE
A
Nisbet and Sarofim (1972) estimated that roughly 6 x 10 tons
of polychlorinated biphenyls were lost to the fresh and coastal waters
of North America between 1930 and 1970. Metabolic transformation
and chemical decomposition are expected to have degraded approxi-
4
mately 50 percent of this amount, leaving about 3 x 10 tons of PCB
contamination in the North American hydrosphere. Of this remaining
4
total, 2/3 or 2 x 10 tons were estimated to have been deposited into
5-14
-------
I
I—1
Ln
500
1000 1500
KM FROM SOURCE
2000
Source: Harvey, G. R., and W. G. Steinhauer. 1974.
Atmospheric Transport of Polychlorobiphenyls to The
North Atlantic. Atmospheric Environment 8: 772-782.
FIGURE 5.3
RELATIONSHIP OF ATMOSPHERIC PCB AND DISTANCE FROM INDUSTRIAL SOURCES
-------
fresh-water systems via industrial and municipal waste-water efflu-
ents. Losses to the sea from such sources as river runoff, sediment
transport, sewage-sludge and dredge-spoil dumping, accidental leaks
from coastal industries, losses from ships, and deliberate dumping of
Industrial solid wastes presumably amounted to about 10^ tons. An
additional source of coastal contamination is aerial fallout and
4
rainout. Of the 2 x 10 remaining tons of PCBs originally lost to the
o
atmosphere, about 1/4, or 5 x 10 tons, is believed to have been
deposited into the sea, mainly into the Atlantic Ocean. The total
quantity of PCBs lost to North American coastal waters between 1930
and 1970 was, therefore, approximately 1.5 x 10 tons.
A recent study conducted by Young _et al. (1975b) investigated
inputs of chlorinated hydrocarbons into the Southern California
Bight and established the importance of aerial fallout as a source of
PCBs in the coastal waters of the Pacific. Dry aerial fallout was
analyzed from 13 coastal and 5 island stations throughout the Bight
during two 13-week periods. Samples consisted of duplicate 1-week
collections. The fallout was collected on glass plates sprayed with
a hexane-mineral oil mixture, a procedure developed by Dr. Vance
McClure of The National Marine Fisheries Service, LaJolla, California.
In addition, four other routes of input were investigated, including
municipal waste-water effluents, industrial waste-water effluents,
surface runoff, and antifouling paint from boat bottoms. The conclu-
sions are presented in Table 5.4.
5-16
-------
Table b.4 ESTIMATED ANNUAL MASS EMISSION RATES OF
POLYCHLORINATED BIPHENYLS TO THE SOUTHERN CALIFORNIA BIGHT
(kg/yr)
Route
Municipal
wastewater
Direct
industrial
Antifouling
paints
Surface
runo f f
Aerial
fallout
Total (Land)
Year
1974
1973-74
1973
1973
1973-74
1242 PCB
4300
—
-------
It can be seen that the major source of PCBs is the marine
discharge of municipal waste-water effluents. Surface runoff is of
minor importance and direct industrial discharge is more or less
insignificant. The second major source of PCBs in the Southern
California Bight is aerial fallout. Analysis was conducted only
for Aroclor 1254 and it was estimated that about 1500 kg/year of this
PCB mixture enters the Bight by this route. However, since the concen-
tration of Aroclor 1242 in surface runoff is of the same order of mag-
nitude as that of Aroclor 1254, and since aerial fallout is a major
contributor to surface runoff, it is estimated that aerial fallout of
Aroclor 1242 is approximately equal to that of Aroclor 1254. In view
of the severe air pollution problems in southern California, it is not
surprising that aerial fallout has been found to play such a signifi-
cant role in coastal contamination.
Some idea of the maximum possible levels of dissolved PCBs in
the hydrosphere (excluding those adsorbed onto particulates and sedi-
ment, see Section V.D.3) can be obtained from solubility studies.
The solubilities of 21 PCB isomers in water have been tested
(WallnHffer £t al., 1973). With the exception of a few isomers, the
more highly chlorinated the PCB, the lower its solubility in water.
Values ranged from 5.8 ppm for 2-monochloro to 0.0072 ppm for 2,3,4,
5,2' ,3' ,4',5'-octachloro biphenyl. Haque et al. (1974) studied the
equilibrium between Aroclor 1254 and water and noted the final solu-
bility. Equilibration was apparently complete after two months as no
5-18
-------
further increase in dissolved PCB was observed at six months. The
final concentration was 56 ppb. From the relative intensitie8 of the
gas chromatogram peaks, it was concluded, in agreement with the study
cited above, that the water solubilities of PCB isomers in general
decrease with increasing chlorine content. It must be emphasized that,
upon sampling natural waters for PCBs, sediment and suspected particu-
lates will also be collected and total PCB concentrations can appear
much higher. The observation of PCB concentrations as high as five
times the solubility (275 ppb) in the Escambia River due to accidental
leakage of hydraulic fluid from an air compressor was reported by
Duke _et al. (1970) (see Section IV.C.2).
Since municipal and industrial waste-water effluents contribute
the major bulk of PCBs to fresh water systems (see Tables 4.7 and 4.8),
one expects to find the highest concentration in rivers, streams, and
lakes serving industrialized, urban areas. Nisbet and Sarofim adopted
0.5 parts per trillion (ppt) as a plausible background level for PCBs
in unpolluted fresh water. They based this figure on the observation
that water entering a Swedish waterworks contained 0.5 ppt. The rela-
tive lack of pollution in Scandinavian fresh waters was recently
verified by Lunde (1975). Values ranging from 0.3 to 0.6 ppt were
observed in the fresh water supply and tap water at Maridalsvennet,
Oslo, and in the GBta and Nordre Rivers in Gothenburg. From scattered
reports on PCB concentrations in North American fresh waters, Nisbet
and Sarofim further estimated 5 ppt to be the mean concentration in
5-19
-------
the Great Lakes, 50 ppt in slightly polluted rivers (and also polluted
bays), and 500 ppt in highly polluted industrialized rivers. The
value for the Great Lakes was estimated from calculations based on the
observed 2DDT/PCB ratios in fish along with the known concentrations
of DDT in the body of water considered. It was assumed that fish con-
centrate both of these chlorinated hydrocarbons in an identical manner.
A gradient of ZDDT/PCB in the direction of increasing PCB concentra-
tions was noted in fish from four Great Lakes as follows: Superior
(ratio - 2.4) < Huron (0.8) - Michigan (0.4-1) •- Ontario (0.08-0.2).
These increasing PCB levels from west to east are consistent with the
relative industrialization of the areas. The PCB concentrations cal-
culated for Lake Michigan ranged from 1 to 7.5 ppt; an actual value
obtained in the summer of 1970 for this lake was 13 ppt (Interdepart-
mental Task Force on PCBs, 1972). Despite predictions that PCB con-
centrations in the Great Lakes would decrease as a result of Monsanto's
restrictive sales policy, high levels in fish continue to be observed.
Concentrations in lake trout have apparently increased from 12.5 ppm
in 1972 to 19 ppm in 1973 to 23 ppm in 1974 (Anonymous, 1975d).
Whether this increase is due to uptake of materials from past spills,
negligent disposal practices C including low-temperature incineration
followed by aerial transport), or use in unauthorized applications is
uncertain. However, William Turney, Water Management Bureau Chief
of the Michigan Department of Natural Resources, claims that if the
U.S. Environmental Protection Agency does not take steps to halt the
5-20
-------
disposal of PCBs in the environment, he will recommend that PCBs be
banned completely in Michigan (Mason, 1975). Values for slightly
polluted rivers and polluted bays were based on the report of 40 to
70 ppt in Green Bay, Wisconsin, and 10 to 20 ppt In some rivers dis-
charging into Lake Michigan (see Table 4.11) (Velth and Lee, 1971;
Veith, 1972). The 500 ppt level for highly polluted industrialized
rivers was based on determination of concentrations in the Milwaukee
River, which ranged from 30 to 2,700 ppt (. Veith and Lee, 1971).
Another highly polluted river, the Great Miami River in Ohio, exhibited
concentrations of PCBs ranging from undetectable to 15,800 pptC the
mean of 16 samples was 5,700 ppt) (Interdepartmental Task Force on
PCBs, 1972), indicating that the Nisbet and Sarofim estimate may have
been somewhat conservative.
A study by Veith and Lee (1971) concerning the fate of PCBs
deposited in the Milwaukee River revealed that variations in the
composition of the PCB mixtures occur as they travel downstream from
their sources. A more rapid decrease in the concentrations of the
less chlorinated isomers with respect to the more chlorinated isomers
was observed. For example, the apparent Aroclor 1242 concentration
in the lower river decreased from 2.07 ug/1 at Estabrook Park to
approximately 0.3 yg/1 just below the North Avenue dam, a seven-fold
reduction. In the same stretch of the river, the apparent concentra-
tion of Aroclor 1260 decreased from 0.1 to 0.05 yg/1, a two-fold
reduction.
5-21
-------
These observations provide support for the hypothesis that the
less-chlorinated isomers are preferentially removed from the hydro-
sphere by vaporization, codistillation, and more rapid metabolic
degradation.
A recent compilation of data gathered from monitoring activities
of the Geologic Survey, U.S. Department of the Interior (Crump-Wiesner
.§£ al. , 1974), testifies to the ubiquitous distribution of PCBs in the
aquatic environment. The major drawback of this survey is that only
one liter of water was collected at each site, imposing a lower limit
of reproducible detection of 0.1 ug/liter. Thus, concentrations of
less than 0.1 ppb were reported as "not detectable." Since Niebet and
Sarofim considered levels as low as 0.005 ppb to represent significant
pollution, the present study must be considered incomplete. Neverthe-
less, the sites at which levels of 0.1 ppb or more were reported are
representative of most regions of the country and constitute evidence
for widespread contamination. The results are presented in Table 5.5.
PCBs were recently discovered in Lake Anne, a 10-year-old man-made lake
in Reston, Virginia (Martell et_ al_. , 1975). Concentrations ranged from
50-200 ppb, in the range approaching the Nisbet and Sarofim estimates
for highly polluted rivers. Since there are no industrial or municipal
discharges into Lake Anne and indeed little if any industry in Reston,
the PCBs were assumed to originate from "diffuse sources as-
sociated with urban development and living." Aerial transport must
also be considered as a possible route of input in this case,
5-22
-------
Table 5.5 SUMMARY OF PCB RESIDUE DATA FOR SURFACE AND GROUND
WATER, JANUARY 1971 - JUNE 1972
STATE
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Hawaii
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
New Jersey
New Mexico
New York
North Dakota
Oklahoma
Oregon
Pennsylvania
Puerto Rico
South Dakota
Texas
Virginia
Washington
West Virginia
Wisconsin
Wyoming
NO.
SAMPLES
3
8
32
161
32
13
5
24
10
7
9
2
6
5
2
3
8
21
47
44
11
36
325
40
19
13
2
7
18
660
4
25
4
3
18
OCCUR-
RENCES
0
0
0
2
1
6
0
0
0
0
0
0
1
1
0
2
0
0
0
0
3
0
52
0
0
0
1
1
0
12
1
0
0
0
0
CONCEN-
TRATION,
Ug/LITER
ND
ND
ND
0.1,0.1
0.3
0.1-0.2
ND
ND
ND
ND
ND
ND
0.1
0.2
ND
0.1,0.3
ND
ND
ND
ND
0.1
ND
0.1-4.0
ND
ND
ND
0.2
0.1
ND
0.1-3.0
0.1
ND
NO
ND
ND
MEDIAN
CONCENTRATION,
Ug/LITER
ND
ND
ND
ND
ND
0.1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.1
ND
0.3
ND
ND
ND
ND
ND
ND
0.4
ND
ND
ND
ND
ND
NOTE: ND - Not Detected
Source;: Crump-WloHner, II. J., h. K. I'Vli/., and M. I.. Yates. 1974,
J'esticJdt'.s in Water: A Study of the Distribution of
Polyclilorinatud Biphenyls Jn t lie Aquatic Environment.
Pesti£idf Monitoring Journa^ 8 (')) : 157-161.
5-23
-------
Just as when dealing with atmospheric contamination, one must
also be receptive to possible insidious sources of PCBs in the hydro-
sphere. Evidence was recently presented indicating the possible
inadvertent manufacture of PCBs at a waste treatment plant. In
November 1972, the failure of a municipal sewage treatment plant was
reported. Reduction of biochemical oxygen demand was decreased by
up to 30 percent of the reduction previously achieved. Half of the
flow to the plant originated from a textile mill using at least 2 tons
per week of biphenyl as a dye carrier for synthetic fibre. At the
same time, the plant was subjected to significant concentrations of
heavy metals from another industry. Between 150-190 kgs/day of
chlorine gas was added to the system for influent odor control and
effluent "disinfection." A PCS content of 18 ppm was subsequently
noted on the filters of this plant (Gaffney, 1974),
It is generally accepted that the oceanic abyss represents an
ultimate sink for PCB residues (Nisbet and Sarofim, 1972} with local-
ized discharge and aerial fallout being the primary routes of con-
tamination. The observation that approximately 1 ton per year of
PCBs was being deposited into the Firth of Clyde in crude sewage
sludge prompted an investigation of PCB levels in the surface and
sub-surface waters of the Clyde Estuary (Holden, 1970). No PCB
residues were detected in this study. However, subsequent analysis
of plankton samples taken at a number of stations from the Firth of
Clyde out to the International Ocean Weather Station India, 400
5-24
-------
miles west of Scotland, revealed a gradient of PCB concentrations
terminating at a level 12 times lower than that found in the polluted
estuary (Williams and Holden, 1973).
Another study in which relative levels of PCB contamination in
the ocean were inferred by comparison with the levels in plankton
was reported by Harvey .et al. (1974). In this study, PCB analyses
were made on 53 plankton samples from both the North and South Atlantic,
In both areas, the mean concentration was 200 ug/kg of wet weight and
no discernable horizontal concentration gradients from coastal waters
to the open sea were noted. This study is in conflict with that of
Williams and Holden (1973) reported above. However, due to differences
in collection procedure and the type of plankton, the two studies are
not truly comparable.
The detection of significant quantities of PCBs in plankton is
not surprising when it is remembered that plankton receive a large
part of their nourishment from the surface layers of the sea. The
surface microlayer contains a variety of surface-active substances
such as fatty acids and fatty alcohols. These materials often form
a visible film or slick; but the absence of such a film does not
discount the likelihood of finding high concentrations of surface-
active materials in the surface layers. The source of these compounds
is the abundance of natural marine organic matter just below the
surface. Surfactants are probably concentrated by convection currents,
rising bubbles, and diffusion, with the more surface-active compounds
5-25
-------
displacing the less active ones. Many pollutants, especially the
lipophilic variety such as chlorinated hydrocarbons, may be concen-
trated in this layer. Once concentrated, they are readily accessible
to microorganisms, phytoplankton, and zooplankton residing at: the
surface. Thus, pollutants enter the food chain and are conccmtrated
by the higher organisms in the hydrosphere.
The enrichment of the surface microlayer of Narraganaetf. Bay,
Rhode Island ( an area basically free of industrial and municipal
effluents and major ship traffic), with polychlorinated biphenyls was
studied (Duce e£ al.. , 1972). Several surface microlayer samples,
along with subsurface samples (20 centimeters [cm] below the surface)
from each site, were collected and analyzed for polychlorinated
biphenyls. The results of two samplings are presented in Table 5.6.
Sample 1 was collected from a heavy slick and showed an enrich-
ment factor in the surface microlayer relative to the subsurface
waters of -28. Sample 2, from a less visible slick, showed an enrich-
ment factor of ~9. Of cource, collection of the surface microlayer
per se is impossible and the enrichment factors reported represented
those of the first 100 to 150 micrometers (p.m) of surface water. If
the surface layer is assumed to be monomolecular, it should have a
_3
thickness of about 2 x 10 |Jim; if the film thickness is assumed
_2
(conservatively) to be five molecular layers of 1 x 10 um and all
of the chemical enrichment is concentrated in this film, the true
4
concentration would be 1.5 x 10 times the concentration in the first
5-26
-------
Table 5.6 CONCENTRATION AND ENRICHMENT FACTORS OF POLYCHLORINATED
BIPHENYLS (AS AROCLOR 1254) IN SURFACE MICROLAYER SAMPLES FROM
NARRAGANSETT BAY, RHODE ISLAND
Sample 1
Concentration
(jag/liter)
Enrichment
factor
Surface Subsurface
4.2±1.0 0.15±0.04 28*10
Sample 2
Concentration Enrichment
(fig/liter) factor
Surface Subsurface
0.45±0.11 SO.05 >9
Source:
Adopted from Duce, R. A., J. G. Quinn, C. E. Olney, S. R.
Piotrowicz, B. J. Roy, and T. L. Wade. 1972. Enrichment
of Heavy Metals and Organic Compounds in the Surface Micro-
layer of Narragansett Bay, Rhode Island. Science 176 (4031)
161-163. '
5-27
-------
150 (jjn actually sampled. In Sample 1, the true concentration would,
therefore, be ~60 ppm or an enrichment factor of 4 x 105 for the
film layer.
Other more recent studies have confirmed the importance of the
surface microlayer in concentrating polychlorinated biphenyls.
Bidleman and Olney (1974a) sampled both the surface (first 150 u.m)
and subsurface (30 cm below the surface) waters of the Sargasso Sea
in a range of 80 to 320 kilometers south of Bermuda. Their results are
presented in Table 5.7. While the predominant gas chromatographic
patterns of atmospheric samples taken near Bermuda resembled Aroclor
1242 or 1248 (see Section V.B), the most abundant mixtures in the
water were Aroclor 1254 and 1260. This finding is in agreement with
results of fresh water studies and suggests that vaporization, co-
distillation, and metabolic degradation of the lower isomers result
in the concentration of higher isomers in the hydrosphere. Once
again, an enrichment of the surface microlayer was noted. Except
for one sample, that taken on 11 April 1973 when excessive windborne
dust was carried to sea, the enrichment factors (average « 5.2) were
lower than those noted previously by Duce et al. (1972) in Rhode
Island waters. This might be expected since the Sargasso Sea area
is one of high evaporation and low rainfall and, once again, stresses
the importance of atmoBpheric transport of PCBs to the sea.
In another study conducted from June through October of 1972
(Harvey et al., 1973), PCB concentrations in the open ocean water of
5-28
-------
Table 5.7 POLYCHLORINATED BIPHENYLS (AS AROCLOR 1260) IN SARGASSO
SURFACE MICROLAYER (SM) AND SUBSURFACE WATER (SS)
^"•"•" "•— — •— MM
Collection
date
(1973)
4/9
A/10
4/10
4/11 .
4/12
4/13
4/16
4/17
-•— ^— — •— — i
Location
29°56'N,64°40'W
30°45'N,66°50'W
30°34'N,66059'W
28°53'N,65°07IW
29°56'N,63°OOfW
30°00'N,64030'W
31°34'N,63°49'W
31038fN,63057'W
*Blank value was 0.9 and was
results.
**Sample taken
in a Sargassum
Sam-
ple
SM
SS
SM
SS
SM
SS
SM
SS
SM
SS
SM
SS
SM
SS
SM
SS
subtracted
windrow.
PCB*
(ng/liter)
11.2
1 f>
•J » \J
4.9
<0.9
8.3
1.0
42.6** 19.3
<0.9** <0.9
3.8
<0.9
5.6
] 6
-I. » \J
5.0
1 8
J~ • \J
8.4
<0.9
in calculating
Enrichment
factor
3.1
5.4
8.3
47.3
21.4
4.2
3.5
2.8
9.3
the above
Source: Adopted from Bidleman, T. F., and C. E. Olney. 1974.
Chlorinated Hydrocarbons in the Sargasso Sea, Atmosphere and
Surface Water. Science 183: 516-518.
5-29
-------
both the eastern and western North Atlantic were determined. The
results are presented in Table 5.8.
The widespread distribution of PCBs (mainly as Aroclor 1254)
once again is suggestive of atmospheric transport. Radically different
concentrations at points fairly close together were observed and were
attributed to seaslicks, localized discharge from ships, and localized
rainfall. Just as was the case with plankton (Harvey .et al. , 1974a),
no concentration gradients were noted from coastal waters to the open
sea. PCB concentrations decrease with depth, but they are still de-
tectable at levels greater than 1 ppt, 3000 meters below the surface.
Although a wide range of concentrations exists, the average concen-
trations in the northern North Atlantic surface waters and subsurface
waters (200 m below the surface) are 35 ppt and 10 ppt, respectively.
The surface waters of the Sargasso Sea had slightly lower surface
concentrations (average » 27 ppt) than those in other parts of the
North Atlantic but were 10 to 20 times higher than those reported by
Bidleman and Olney (1974) in April of 1973.
Continued monitoring of North Atlantic surface waters by Harvey
et al. (1974b) revealed that PCB levels had declined approximately
40-fold since the 1972 survey (Harvey ^it al. , 1973). Levels of
0.8 ng/1 were reported in samples from the Sargasso Sea to the New
York Bight and the New England Continental Shelf during February of
1974. These values wei a in good agreement with those of Bidleman and
Olney (average - 1 ng/3).
5-30
-------
Table 5.8 CONCENTRATIONS OF PCBs IN NORTH ATLANTIC OCEAN WATER
he limit of detection was 1 x 10~9 g/liter for a 19-liter sample)
— " " •— ~ •*- -ifc j-w ^ / J- .*. v, ^ J. A-'-'i. 0. O.-/ — J-O.UCA. D d J
POSITION
STATION
NORTH WEST
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
52°
44°
40°
36°
34°
35°
34°
38°
38°
38°
41°
43°
43°
43"
46°
52"
52°
55°
57°
60°
60°
60°
63°
41°
39°
37°
34°
35°
35°
35°
36°
36°
35"
34°
33°
28°
26"
25°
26°
28°
31°
55'
00'
33'
11'
02'
00'
47°
20'
23'
19'
09'
57'
20'
16'
31'
31'
35'
41'
22'
04'
09'
29'
03'
32'
40'
12'
32'
22'
17'
56'
05'
24'
37'
26'
41'
42'
M)'
56'
58'
26'
13'
35°
30°
29°
25°
22°
18°
14°
11°
11°
19°
20°
22°
21°
21°
21°
19°
19°
15°
12°
06°
05°
04°
02°
70°
70°
68°
67°
67°
68°
66°
67°
68°
f.7°
66°
65°
'.8"
'>!>"
'>4"
53°
53°
53°
08'
36'
16'
33'
50'
59'
57'
23'
11'
28'
46'
13'
57'
34'
43'
52'
53'
02'
01'
02'
36'
43'
22'
40'
03'
54'
01'
36'
28'
34'
27'
24'
49'
22'
44'
39'
38'
15'
57'
55'
49'
DATE
(1972)
6/30
7/3
7/5
7/7
7/9
7/11
7/13
7/15
7/22
7/23
7/24
7/25
7/27
7/29
7/30
8/1
8/2
8/4
8/5
8/6
8/7
8/7
8/8
10/2
9/21
9/22
9/23
9/24
9/24
9/25
9/26
9/27
9/27
9/28
9/28
10/3
10/5
10/6
10/7
10/8
10/9
DEPTH
(m)
0
0;
0;
0;
0;
0;
0;
0;
0;
0
0
0
0
100;
0
0;
1500
0
200
200
0
0
0;
0
0
0
0
1000
0
0
0
0
0
0
0
0
0
0
0
0
0
200
200
200
200
200
200
200
200
3000
100
200
PCB
CONCENTRATION*
(X 10-9 g/LITER)
150
35;
4-
^ t
11;
30:
-j\j ^
14:
j--r j
67:
u * t
19;
41-
"-•- >
77
< i
52
47
45:
T _/ |
10
45;
> it
3
97
42
82
23
21:
•*• >
30
29
36
12
1
2
5
j
11
22
9
12
1 "5
-J- J
26
36
27
42
88
68
10
7
3
6
5
39
2
13
> It
8
7
*The closest matching commercial mixture In all cases was that containlne
j>4 percent chlorine. Thus, the commercially available Aroclor 1254
(Monsanto) was used as the standard. Procedural blanks ranged from 0 to
3 ng/llter and were subtracted from the tabulated concentrations- DDT
and its metabolites, it present, were present at concentrations less
than 1 ng/Hter.
Source:
Harvey, G. R. , W. G. Steinhauer, and J. M. Teal.
^ N°rth Atlantlc Ocean
5-31
1973
Science
-------
The volume of the upper 200 m of the North Atlantic is 1Q18 liters.
If the average PCB concentration in that volume during 1972 was
— Q
2 x 10 g/liter (20 ppt), then a total of approximately 2 x 104 tons
of PCBs were present in the water at that time. This value is in
excellent agreement with the value of 1.5 x 104 tons predicted by
Nisbet and Sarofim (1972). However, to account for the 40-fold de-
crease in concentration observed between 1972 and 1973, the data re-
4
quire that almost 2 x 10 tons of PCBs were lost from the upper 200 m
in less than one year (Harvey _et _al. , 1974b). Longhurst and Radford
(1975) derived a mean residence time for PCB in seawater in the North
Atlantic based on a standing stock of 2 x 10 tons and the 40-fold
decline which reportedly took place in the year 1972 to 1973.
The model used for these calculations simulated the loss process
on a daily basis by reducing the standing stock (S) by a constant
(S/r). The value of the mean residence time, r, was calculated so as to
4
reduce S (-2 x 10 tons) by 97.5 percent (40-fold) in 365 days; r was
calculated to be just under 100 days. In order for this system to
sustain the reported standing stock of 2 x 10 tons, the annual input
to the waters must have been of the order of 7.3 x 10 tons. This
figure seems improbable since it necessitates that the entire world
production of PCBs was continuously dumped into the North Atlantic
(annual world production figures between 1967 and 1970 averaged
4
7.27 x 10 tons). Longhurst and Radford suggested that either the
analytical results are not correct, with the 1971 to 1972 data C21 to
5-32
-------
41 ug/1) being too high and the 1973 to 1974 data (0.8 to 2.0 ug/1)
being too low or the extrapolation from individually correct data to
a standing stock for the entire North Atlantic being unfounded. Harvey
and Steinhauer (1975) replied to this argument by reaffirming the
validity of their analytical results but admitting that their extrap-
olation to the entire North Atlantic was not justified. The large
decrease in concentration was observed in the North American basin
between 1972 and 1974 and most likely reflected changes in the North
American PCB input. Since continuous monitoring data for the same
period is not available for other areas, it is not possible to draw
any conclusions regarding trends in concentration.
According to Nisbet and Sarofim, approximately 1.5 x 104 tons of
PCBs were distributed over terrestrial North America between 1930 and
1970 (Nisbet and Sarofim, 1972). However, relatively little data on
concentrations of PCB residues in soil are available. In one study
(Carey _et _al. , 1973), soil and crops from 400 sites in 12 states
comprising the corn belt were analyzed for levels of organochlorine
pesticide residues (including PCBs) during 1970. PCB residues were
apparently not detectable in the soil itself as no data was given. How-
ever, Aroclor 1232 and 1242 were Identified in the crops. The PCBs were
found in cornstalk and hay samples from six of the 12 states studied.
Concentrations in 10 cornstalk samples (6.9 percent of those tested)
ranged from 0.53 to 6.25 ppm (average, 2.8); concentrations in three
samples of mixed hay (27.3 percent of those tested) ranged from 0.8 to
5-33
-------
2.94 ppm (average, 0.57). It was concluded that the PCBs entered the
crops through translocatlon from the soil and that the soil could,
therefore, be assumed to be contaminated.
A comprehensive program for monitoring levels of pesticide
residues in soils of the United States (The National Soils Monitoring
Program) was initiated in 1969 as part of the National Pesticide
Monitoring Program. One small segment of this program monitors five
urban areas each year and includes analyses for PCBs (Ruopp arid DeCarlo,
1975). PCBs were detected in three of the five cities sampled in each
year for which data are available (1971 to 1973). Of the 22 positive
samples, 17 were below 1 ppm. Pittsfield, Massachusetts, the home of
a large transformer plant which uses PCBs, was sampled during the 1973
survey. No PCB residues were detected in any of six sites located
within a one-mile radius of the plant. In another study, conducted
for the Office of Toxic Substances, soils were sampled near a plant in
Illinois using PCBs in the manufacture of investment casting wax
(Ruopp and DeCarlo, 1975). Samples taken within a quarter-mile radius
around the plant ranged in value from 0.77 to 5.2 ppm total PCB.
The PCBs were identified as Aroclor 1260 and decachlorobiphenyl.
Levels up to 1.8 ppm Aroclor 1260 were detected up to one and one-
quarter miles from the plant.
In March of 1975, sites were sampled in the vicinity of an invest-
ment casting company in Michigan and a PCB manufacturer in Illinois
(Ruopp and DeCarlo, 1975). Surface soil samples were collected at
5-34
-------
one-quarter mile intervals up to one mile from each plant in all
directions. PCBs were detected at both locations. Aroclors 1242 and
1260 as well as decachlorobiphenyls were detected near the PCB manufac-
turers but only Aroclor 1260 was detected near the investment casting
plant. Total concentrations ranged from 0.001 ppm to over 20 ppm.
Values were highest near the plant sites and, in general, decreased
with increasing distance from the site. The actual concentrations at
every sampling site for each of the three plants where soil contamina-
tion was observed are presented in Figure 5.4.
D. TRANSPORT AT INTERFACES
-1- Air/Water Interface (including precipitation)
"From a pollution standpoint, the air-sea interface is
perhaps one of the most important regions of the marine environment"
(Duce et al., 1972). In the case of PCBs, ample evidence exists to
support this statement. \^he observation of significant concentrations
of PCBs in the marine atmosphere, far removed from urban industrial-
ized areas, suggests that atmospheric transport represents the major
route of PCBs to the sea (Harvey and Steinhauer, 1974; Bidleman and
Olney, 1974a; see Section V.B). It follows logically that the sur-
face microlayer would be the first compartment of the hydrosphere to
receive the majority of PCBs deposited into the ocean. Surface-
active agents, including a high concentration of lipids, which com-
prise the surface slick are receptive to lipophllic compounds such as
the chlorinated hydrocarbons and thus tend to concentrate and
5-35
-------
<0.01
0.02.
Aroclcir 1260
Investment CaHtlriR Cumpnny
1 .
Aroclor ] 260
Investment Casting W/ix Manufacturer
<0.001
,• ' 0 .001
Ar
-------
stabilize them. The highest concentrations of dissolved PCBs would,
therefore, be expected to reside in the surface microlayer and indeed
have been found to do so (Bidleman and Olney, 1974a; Duce et al., 1972;
Harvey jet al., 1973; see Section V.C).
The microorganisms and plankton that live in the first 100
meters below the surface subsequently ingest and/or absorb the PCBs
dissolved in this surface microlayer and, by this route, the PCBs
enter the marine food chain. The high concentrations of PCBs in
plankton from waters that themselves contained undetectable levels
of PCBs (Williams and Holden, 1973) testify to this occurrence.
Bidleman and Olney (1974a) estimated the residence time of PCBs
in the atmosphere over the Sargasso Sea from the mean concentrations
in the air (5xlO~10 g/m3; Table 5.2 samples 1-8) and subsurface
waters (l.lxio"6 g/m3*) using the following assumptions:
(i) except for the surface microlayer, PCBs are uniformly
distributed in the top 100 m and have a resident time
there of 4 years, (ii) the atmospheric concentration is
constant up to the tropopause and zero above, (iii) the
major source of PCB input is the atmosphere.
Using the mean air and water concentrations given above, the
quantity of PCBs In a column of air 1 m2 extending up to the tropo-
pause, a distance equivalent to 6.3 kilometers after correcting to
standard conditions (0°C, 760 mm pressure), is (6.3xl03m3) (5xlO~10
*Thls value is obtained by averaging the values on Table 5.7 by two
different methods.
a) assume <0.9 - 0.9; average - 3.4 ng/1 - l.4xlO~6 g/m3
b) assume <0.9 - 0.0; average = 0.89 ng/1 - 0.89x10-6 g/m3
The average of these two results is l.lxlO"6 g/m*.
5-37
-------
g/m ) or 3.15x10 grams; that in a column of water 1 m extending
down to 100 m is (IxlO2 m3) (l.lxlO*6 g/m3) or 1.IxlO"4 g.
Let the residence time in air, Ta, equal N/(dN/dt) where N is
the quantity of PCB in the air (3.15xlO~6 g) and dN/dt is the rate
of loss from air into water. If the residence time in water, Tco, is
4 years
then d(PCB)/dt - 1.1x10" g/4 years or 2.8xlO~5 g/year
Ta then equals 3.15xlO~ g/2.8xlO"5 g/year or 0.11 years or
40 days.
Harvey et al. (1973) reported PCB concentrations in Sargasso
Sea surface waters 10 to 20 times higher. If these concentrations are
correct, the residence time would be an order of magnitude shorter.
'Some studies have been reported in which an attempt was made
to detect the presence of PCBs in precipitation. Bevenue ££ al.
(1972) analyzed rainwater in Hawaii for the presence of organochlorine
pesticide residue. While some pesticide residues were present, under
the conditions of this study, no PCBs were detected. The pesticides
observed, however, could be easily detected at the 1 ppt level; PCBs
on the other hand would have had to be present in concentrations of
20-40 ppt for minimum detection. One therefore cannot conclude that
PCBs are absent from Hawaiian rainwater. Tarrant and Tatton (1968)
examined rainwater collected continuously at seven widely distributed
sites in the British Isles chosen so as to represent a variety of
conditions and locations. Small amounts of PCBs were detected in all
the samples but were not quantified.
5-38
-------
Samples of snow were collected at seven locations in Wisconsin,
allowed to melt in the laboratory, and analyzed for PCBs. Three
samples (Green Bay, Grafton, and Kewaskum) were below the 0.01 ppb
level of detection. However, samples from Racine, Kenosha, Madison,
and Milwaukee displayed 0.17, 0.22, 0.24, and 0.20 ppb PCB, respective-
ly (Kleinert, 1976).
Peel (1975) took advantage of the unique characteristics of
Antarctica to study the role played by precipitation in the transport
of PCBs. The absolute lack of any nearby source of emission along
with the lessened probabilities of interference by other, more domi-
nant organochlorine compounds make Antarctic snow an ideal form of
precipitation to study. ln addition, the sub-zero temperatures and
negligible biological activity in the snow minimize the possibilities
of chemical alteration or diffusion once precipitation has been deposited.
Ten snow samples laid down between 1965 and 1969, 360 and 400 km from
the coast, were collected and analyzed. Low levels <5xl of
PCBs were found in the snow samples. These results agree with resi-
dent animal data from the area which show I DDT/PCB ratios much
larger than those found north of the Antarctic convergence. However,
since Antarctica is about as far as one can get from urban industrial-
ized areas, the fact that PCBs can be detected at all in snow samples
constitutes readable evidence for airborn, precipitation mediated
1'Cfl transport.
5-39
-------
Although Nisbet and Sarofim (1972) put little weight on transport
of PCBs from the surface of bodies of water back to the atmosphere,
evidence exists that such reentry does in fact occur. A laboratory
study was reported (Oloffs, et_ al_. , 1972> in which the fate of PCBs
(Aroclor 1260) in three natural water samples was demonstrated. The
sources were (a) the Fraser River at New Westminster in British
Columbia, 10 miles east of Vancouver, (b) two miles above the mouth
of the Nicomekl River at a point 20 miles southeast of Vancouver and
(c) the shoreline of the Georgia Strait, 15 miles north of Vancouver.
To 150 milliliters (ml) of each water sample in a 500-mJUErlenmeyer
flask, 0.1 ppm of Aroclor 1260 was added. The flasks were plugged with
glass wool, swirled, and allowed to stand for 12 weeks with swirling
once per day. The results are presented in Table 5.9.
PCBs concentrations were seen to decrease in the samples of
natural water even at temperatures as low as 7°C. However, the peak
heights in the gas chromatograms did not change relative to one
another and the overall gas chromatographic pattern remained identical
to that of Aroclor 1260. No new peaks were observed nor were any
metabolites detected. Thus, whatever the fate of the Aroclor 1260
mixture was, all of its isomers behaved in a similar manner. Signif-
icant amounts, up to 6.2 percent, of the added PCB were detected
in the glass wool indicating that evaporation had taken place. Since
the initial concentration of Aroclor 1260 was greater than its solu-
bility, it was postulated that the excess PCBs accumulated at the
5-40
-------
Table 5.9 THE FATK OF 0.1 ppm POLYCIILOKINATKU BII'HKNYL (AROCLOR 1260)
IN 150 ml WATER SAMPLES FROM THE FKASER JUVEK, THE NICOMEKL RIVER,
AND GEORGIA STRAIT HELD IN THE LABORATORY FOR 12 WEEKS AT THE GIVEN
TEMPERATURE
Recovery, %
Fraser River (7°C) Nicomekl River (16°C) Georgia Strait
Glass
Time (weeks) Water
0
6
6
12
12
95.
60.
60.
64.
74.
6
1
9
1
7
Wool
0
1.8
4.5
1.2
5.6
Lost
4.4
38.1
34.6
34.7
19.7
Water
96
73
74
33
56
.2
.8
.5
.0
.7
Glass
Wool
0
1.4
1.6
1.1
1.1
Lost
3.8
24.8
23.9
65.9
42.2
Water
93.5
70.7
40.7
37.7
38.7
Glass
Wool
0
1.3
6.2
2.4
0.8
Lost
6.5
28.0
53.1
59.9
60.5
Source: Adapted from OLoffs, P. C. , L. .). Albright, and S. Y. Szeto.
1972. Fate and Behavior of Five Chlorinated Hydrocarbons
in Three Natural Waters. Canadian Journal of Microbiology
18: 1393-1398.
5-41
-------
water/air interface and 8ubsequently evaporated. The PCBs unaccounted
for (see "% lost" column, Table 0.9) were assumed to have been lost
from the glass wool into the atmosphere. The authors neglected to
take into consideration the fact that PCBs tend to adsorb tenaciously
to glass surfaces (see Section V.D.3) and were probably lost, to some
extent, by this route. Nevertheless, their presence on the glass
wool is ample evidence for the plausibility of re-evaporation from
natural waters contributing to atmospheric contamination under certain
conditions. A similar study using the chlorinated hydrocarbon pesti-
cide chlordane revealed that the presence of materials that promote
the even distribution of these residues in natural waters (e.g.,
detergents) will minimize this accumulation at the surface and, in
turn, their evaporation. Because of similarities in properties,
such a phenomenon is applicable to PCBs as well and they therefore
should be expected to behave differently in sewage outfalls or in
bodies of water serving industrialized areas than in other natural
waters.
When considering the evaporation of PCBs from water, one tends
to assume that due to their high molecular weight and low vapor
pressure, evaporation should be slow. A factor that is overlooked,
however, is their high activity coefficients in water, which, when
multiplied by the molecular concentration, yield the active mass or
actual number of gram molecular weights per liter in solution or in
gaseous form. This results in unexpectedly high equilibrium vapor
5-42
-------
partial pressure and consequently high rates of evaporation. An
attempt was made to quantify the rates of evaporation of low-solubility
contaminants such as PCBs from water bodies to the atmosphere (Macday
and Walkoff, 1973). The results are given in Table 5.10.
The assumptions made in calculating these data are as follows:
(a) The contaminant concentration used is that of dissolved
PCBs and excludes those in suspended, colloidal, complexed
or adsorbed form. These forms may be included if they can
be converted to the dissolved form as evaporation takes
place.
(b) The vapor formed is in equilibrium with the liquid at the
interface.
(c) The water diffusion is fast enough to ensure that the con-
centration at the interface is representative of that in
the bulk of the water.
(d) The water evaporation rate is not affected by the presence
of the PCB.
From these assumptions one can see that the calculations deal with
situations in which perfect mixing and equilibration of the liquid
phase occur so that as the compound evaporates from the upper layers,
no concentration gradients appear. This situation may be approached
in reality. However, in many cases, the rate of evaporation will be
reduced due to the delay in transference of materials from the mass
of water to the depleted interface. Thus, while the calculations
5-43
-------
Table 5.10 EVAPORATION PARAMETERS AND RATES FOR VARIOUS AROCLORS AT 25°C
•C-
Aroclor
1242
1248
1254
1260
Solubility
Mg/liters
0.24
5.4 x 10~2
1.2 x 10~2
2.7 x 10~3
Vapor Pressure
mm Hg
4.06 x 10~4
4.94 x 10~4
7.71 x 10~5
4.05 x 10~5
Fff(Z)
9.7
46
39
72
T*
(for 1 meter depth)
5,96 hrs.
58.3 min.
1.2 min.
28.8 min.
^F = the fraction of the mass of PCB present which is evaporated when 0.01% of the water
is evaporated.
*T = the time required for the concentration of PCB to drop to one-half its initial value.
Source: Adapted from Mackay, D. , and A. W. Walkoff. 1973. Rate of Evaporation of Low
Solubility Contaminants from Water Bodies to Atmosphere. Environmental
Science and Technology 7: 611-614.
-------
may be suitable for a fast-flowing, shallow river, they fall short of
being representative for a deep, quiescent lake and, in fact, in all
cases represent the maximum rates of evaporation achievable. The
last column of Table 5.10 presents the calculated half lives (T) for
each respective Aroclor in a column of water 1 m2 by 1 m deep. The
first and second columns tabulate the solubilities and vapor pressures
of each mixture that was used to calculate the T value. ^ addition,
the fraction of PCB present which is lost to the atmosphere when 0.01
percent of the water evaporates is given in the column labeled F.
Thus, in a fast-flowing, shallow river 1 m deep where good mix-
ing occurs, values of T are quite low, e.g., 5.9 hours for Aroclor
1242 and one-half hour for Aroclor 1260. The authors noted an interest-
ing consequence of this high rate of evaporation, if a monitoring
station for municipal or industrial effluents were positioned one
hour downstream from a source of Aroclor 1242, by the time the water
reached the station, 10 percent of the Aroclor would already have been
lost to the atmosphere.
The situation in the Sargasso Sea reflects the importance of
evaporation in determining the concentration of PCBs in surface water
and atmosphere. The Sargasso Sea is an area of high evaporation and
low rainfall and has the lowest surface-water concentrations of PCBs
in the North Atlantic: (Harvey _et al. , 1973).^ should also be
mentioned that the surface layer of the ocean Is a major source of
atmospheric particulars. Participates are ejected, both from a central
5-45
-------
Jet and from the'ruptured bubble film, when surface bubbles break.
Evidence suggests that these particulates are more representative, in
composition, of the surface layer than of the bulk of water beneath.
Since PCBs concentrate in the surface layer and adsorb to particulates
(see Section V.D.3), they are also released to the atmosphere by this
mechanism.
2. Air/Soil Interface
The primary source of soil contamination appears to be
. \/it
fallout from the atmosphere (Nisbet and Sarofim, 1972). \/Lt has been
estimated that 1 to 2 x 10 tons per year of the PCBs are released
into the atmosphere, adsorb to particulates, and are redeposited on
land. By analogy to DDE (a metabolite of DDT), the half life of PCBs
in the soil is about five years. »/!! 2 x 10 tons of PCB originally
deposited into the atmosphere and remaining undegraded were thought
to have subsequently been deposited on land. However, through repeated
vaporization, readsorption onto particulates, and redeposition, approxi-
mately 1/4 of this amount (5 x 10 tons) eventually ended up in the
ocean. The pattern of deposition of the PCBs will depend to a large
extent on the size distribution of the particulates to which they
adsorb. The major fraction will end up in coastal waters, but a sig-
nificant amount will be transported to the open sea.
Some estimate of the rates at which PCBs can be expected to
evaporate from various surfaces was provided by Haque et al. (1974) .
Two types of investigations were carried out. In the first
5-46
-------
experiment, vapor loss was determined from Aroclor 1254 itself; in
the second experiment, the vapor loss was determined following adsorp-
tion onto Ottawa and Woodburn soil, two materials exhibiting markedly
different adsorbing capacities. The loss of Aroclor 1254 from itself
is shown in Figure 5.5. Losses at 26°C are very small, but losses at
60°C are substantial. The loss of PCBs from a sand surface is shown
in Figure 5.6. As expected, the percent loss through vaporization
decreases as the number of chlorine atoms in the isomers increase.
Vapor losses under wet and dry conditions were similar and comparable
in magnitude. An identical experiment performed with Woodburn soil
revealed that vapor loss from this soil was negligible (no quantitative
data presented). The high percentage of organic matter (3.1 percent)
in this soil contributes to its adsorbing capacity and is probably
responsible for the observed result.
From these experiments, some conclusions may be drawn about
transport at the soil/air interface. Adsorption to a surface will
be influenced by such factors as surface area, organic content of
the material, and nature of the material (e.g., soil, sand, etc.).
The transport of PCBs in the atmosphere will, in turn, be influenced
by such factors as temperature and vapor pressure. Tsignificant
amounts of PCB could be lost to the air at higher temperatures if
PCBs are present in the lithosphere free, loosely bound, or adsorbed
to a surface! |A PCB mixture containing a large percent of the lower-
chlorinated isomers will experience greater loss into the atmosphere
5-47
-------
*~
OD
26°C
-O~ OOOOO---OOO-O O
.002L.
Gms
.006
.010
—6 7
Loss =2.0 x 10 gms/day/on
PCB 1254 Vapor Loss
cxn 60°C
^s Lose = 8.6 x 10 gas/day /cm
Nx D
X
x
X
N
X
X
X
x.
_L
10
30
20
TIME (days)
Source: Haque, R., D. W. Schmedding, and V. H. Freed. 1974.
Aqueous Solubility, Adsorption and Vapor Behavior of
Polychlorinated Biphenyl Aroclor 1254. Environmental
Science and Technology 8: 139-142.
FIGURE 5.5
LOSS OF AROCLOR 1254 FROM ITSELF AS A FUNCTION OF TIME
-------
PCB 1254 Vapor Loss @ 26°C from Dry Sand
7C1
LOST
% LOST
. PCB 1254 Vapor Loss @ 26°C from Wetted Sand
Source:
2 3
TIME (weeks)
Haque, R. , D. W. Schmedding, and V. H. Freed 1974
Aqueous Solubility, Adsorption and Vapor Behavior of
Polychlorinated Blphenyl Aroclor 1254. Environmental
Science and__Technology 8: 139-142. ~
FIGURE '>.(,
LOSS OF AROCLOR 1254 FROM AN OTTAWA SAND
5-49
-------
than one containing the more highly chlorinated isomers.7 Strongly
bound PCBs, on the other hand, will be lost only to a slight extent.
3- Soil/Water Interface (including sediment)
Because of their low solubility, high specific gravity,
and tendency to adsorb to particulates, the majority of PCBs released
into the hydrosphere are expected to eventually end up adsorbed onto
sediment or resting as sludges at the bottom of rivers and lakes and in
the oceanic abyss (Nisbet and Sarofim, 1972). Nisbet and Sarofim
estimated that all of the 2 x 10 tons of PCBs discharged into fresh
water between 1930 and 1970 could be found in the sediment. Trans-
port in rivers is thought to take place by solution and readsorption
onto sediment and by the transport of sediment itself. Approximately
200 tons/year of PCBs are transported to the ocean by this pathway
and the sediment of the deep oceans represents the ultimate sink for
PCB residues.
Several laboratory studies have been reported that demonstrate
the tenacity of PCBs for a variety of surfaces and particulates.
Eichelberger (1971) added 10 (ig of each Aroclor ( from 1232 to 1268)
individually to 1 liter of river water in glass containers, sealed
each container with a teflon-coated cap, and sampled the water for
PCB concentration periodically for 16 weeks. Approximately 40 percent
of the PCBs remained in solution at this time. Evaporation was
eliminated as the pathway of loss because the bottles were tightly
sealed. Also, since the Aroclors were recovered intact, biodegradation
5-50
-------
could also be eliminated. It was concluded that irreversible
adsorption onto the walls of the sample container and/or the silt
present in the river water was responsible for the disappearance.
An interesting study performed by Haque et al. (1974) investi-
gated the behavior of a saturated aqueous solution of Aroclor 1254
(56 ppb) in the presence of increasing amounts of adsorbed materials.
The results are presented in Figure 5.7. Iliite clay and Woodburn
soil showed the greatest adsorbance; montmorillonite and kaolinite
clays were intermediate; [and sand and silica gel adsorbed very little
of the dissolved PCBs, even at high concentration^ The high adsorb-
ing capacity of the Woodburn soil was attributed to the presence of
organic matter. Thus, the concentration of PCBs in water will be
reduced whenever the water comes in contact with particulate matter.
. The extent of reduction will depend on the adsorption characteristics
of the particulate matter involved. A sand surface with few adsorbing
sites adsorbs relatively littleJJwhereas a soil surface with high
clay and organic content will adsorb significant amounts.
[These experiments are relevant not only to the partitioning of
PCBs between water and sediment but also to that between precipitation,
irrigation waters, or leachate and soils. If PCBs should come into
contact with the terrestrial environment by any of these routes, the
nature of the surface will determine whether the PCB will be adsorbed
and, therefore, contained or will be allowed to progress to other
areas. The study by Tucker at al. (1975a) , in which the extent to
5-51
-------
% ADS
Aroclor 1254 Adsorption
Del Monte Sand
Silica Gel
Kaolinite
Montmorillonite Clay
Clay
M
o
Woodburn Soil
80 -
50 75
GRAMS ADSORBENT
100
Source: Haque, R., D. W. Schmedding, and V. H. Freed. 1974.
Aqueous Solubility, Adsorption and Vapor Behavior of
Polychlorinated BiphenyJ Aroclor 1254. Environmental
Science and TechnoloRy 8: 139-142.
FIGURE 5.7
PERCENT DECREASE IN THE CONCENTRATION OF AROCLOR 1254
BY THE ADDITION OF INCREASING AMOUNTS OF ADSORBENT
TO A SATURATED (56 pob) ' SOLUTION
5-52
-------
which PCBs were removed from various types of soil by percolating
water (see Section IV.C) was investigated, in general confirmed the
above conclusions. Soils with a high clay content retained the PCBs
most effectively. Contamination of ground water supplies are, there-
fore, most likely when the PCBs travel through permeable sand and
creviced bedrock formations where they may progress at rates of
several thousand feet per year (Walker, 1973KJ,
A study of the behavior of PCBs (as Aroclor 1260) in natural
waters from two sources when their respective sediments were also
present was reported by Oloffs jet al. (1973). The sources of the
water and sediments were the Fraser River and the Georgia Strait,
British Columbia. Bottom sediment (100 g) from each source was
placed in a 500 milliliter (ml) Erlenmeyer flask. One hundred and
fifty ml of water from the same source containing 0.1 ppm PCB (Aroclor
1260) was carefully layered over the sediment. The flasks were plugged
with glass wool and allowed to stand for 12 weeks. Water and sediment
were analyzed for PCB at 0, 6, and 12 weeks. The results are presented
in Table 5.11. All detectable residues had moved out of the water by
six weeks and an average of 73.8 percent and 73.2 percent could be
detected in the sediment from the Fraser River and Georgia Strait,
respectively. No residues could be extracted from the glass wool plug,
suggesting that evaporation was negligible. The 25 percent of the
PCBs that was unaccounted for most likely was adsorbed to the glass
walls of the Erlenmeyer flask. This study shouJd be compared to that
5-53
-------
Table 5.11 FATE OF 0.1 ppm AROCLOR 1260 IN 150 ml WATER SAMPLES FROM
FRASER RIVER AND FROM GEORGIA STRAIT IN THE PRESENCE OF BOTTOM SEDI-
MENTS FROM THE SAME SOURCES, INCUBATED AT 13°C
Recovery in %
Incubation
time (weeks)
0
6
6
12
12
Fraser River
Water Sediment
100
0
0
0
0
0
70.4
67.1
72.8
74.7
Georgia Strait
Water Sediment
100 0
0 70.6
0 71.1
0 80.7
0 65.6
(No residues could be detected on the silanized glass wool plugs
used to stopper flasks during incubation. Controls (0-12 weeks)
were free from detectable residues.)
Source: Adapted fromOloffs, P. C., L. J. Albright, S. Y. Szeto, and
J. Law. 1973. Factors Affecting the Behavior of Five
Chlorinated Hydrocarbons in Two Natural Waters and Their
Sediments. Journal of Fisheries Research Board of Canada 30;
1619-1623. ———.
5-54
-------
of Oloffs £L al. (1972) in which it was shown that Aroclor 1260
escaped from the same natural waters into the atmosphere when incubated
under identical conditions in the absence of bottom sediment (see
Section V.D.I). Apparently, when PCBs are present in natural waters
not in contact with their bottom sediments, loss into the atmsophere
at the water-air interface is significant. However, if the water is
in contact with bottom sediment, adsorption onto the sediment occurs
preferentially. The depth of water probably influences the net out-
come of these two opposing processes.
The implications of these laboratory studies are borne out in the
environment. Horn et al. (1974) analyzed dated oceanic sediment from the
Santa Barbara Basin of the Southern California Bight for PCB residues.
The layered anaerobic sediments of this area provide a situation in
which the rate of deposition of PCBs throughout the years can be
determined. The results are presented in Table 5.12 and shown
graphically in Figure 5.8. The gas chromatographic profiles of all
samples resembled Aroclor 1254 which, consequently, was used as a
reference standard. Analytical background levels were approximately
30 ppb (limit of detection). It can be seen that deposition of PCBs
began about 1945 and increased consistently through 1967 with no indi-
cation of a leveling off taking place. Since about 0.39 cm of sediment
are laid down per year with a mass of 0.3 g per cubic cm (determined by
Horn et al., 1974), It can be calculated that the 1967 PCB concentration
of 103 ppb is i-quivnient to a deposition rate of:
5-55
-------
Table 5.12 POLYCHLORINATED BIPHENYL CONCENTRATIONS
(parts per billion Ippb] of the dry weight) IN DATED SEDIMENTS
Group of Years
1925,1927,1930
1932,1935,1937
1940,1942,1945
1947,1950,1952
1955,1957,1960
1962,1965,1967
Dry Weight
analyzed (g)
7.52
7.86
7.85
7.75
7.90
7.05
PCB
(ppb)
< 29
<21
<31
49
66
103
Source: Adapted from Horn, W., R. W. Risebrough, A. Soutor, and
D. R. Young. 1974. Deposition of DDE and Polybhlorinated
Biphenyls in Dated Sediments of The Santa Barbara Basin.
Science 184: 1197-1199.
5-56
-------
_
Ul
Analytical Background 30
1960
Source: Horn, W. , R. w. Risebrough) A. Soutor;
FIGURE 5.8
DEPOSITION OF PCB IN DATED SEDIMENTS OF THE SANTA BARBARA
BASIN
-------
-9 3 422
103 x 10 (ppb) x 0.3 g/cm of deposit x 0.39 cm/year x 10 cm /m =
-42 2
1.2 x 10 g/m /year = 120 ug/m /yr.
This figure is assumed to represent a minimum estimate of the rate
of loss of these compounds into the overlying waters. Deposition
into sediment is thus a major pathway for the removal of PCBs from
marine waters.
Duke et al. (1970) analyzed the water and sediment of Escambia
Bay, Florida, during and after leakage of Aroclor 1254 from an
air compressor which contaminated the water. The highest concentra-
tions of PCBs in the water occurred at the outfall in the Escambia
River in August 1969 (275 ppb) and decreased dramatically to 3 ppb
in September when the leakage was corrected. Slightly less than
3 ppb could still be found in October 1969. The continued presence
of PCBs in the water months after the leakage was stopped was attrib-
uted to leaching from the sediment. PCS residues reached 486 ppm in
August 1969 (1767 times that in water) in sediment samples taken
close to the outfall but have decreased since then (no value given).
Samples of water and bottom sediment were collected from six
sites on Lake Poinsett and Dry Lake, South Dakota, and analyzed for
PCBs (Greichus et al., 1973). The average value in the water from
six samples was <0.5 ppb, whereas that of the sediment was 6.4 ppb,
representing concentration in sediment by a factor of about 13.
Law and Goerlitz (1974) reported the results of a study initiated
by the U.S. Department ot the Interior in which contamination by
5-58
-------
chlorinated hydrocarbons (including PCBs) was investigated in the
numerous streams that discharge into the San Francisco Bay. Bottom
sediment from 26 streams (39 sites, total) was collected and analyzed
for PCBs. The results are presented in Table 5.13. PCB residues
were found in all of the streams tested and are, therefore, widely
dispersed throughout the San Francisco Bay area. Values ranged from
less than 1 ug/kg to 1400 ug/kg. Stevens Creek (180 ug/kg) and
Alamitos Creek (610 ug/kg) had values much greater than expected
since neither stream is located near an industrial or commercial
development.
Data gathered by the Geological Survey, U.S. Department of the
Interior, in the course of pesticide residue monitoring is seen to
reflect the widespread occurrence of PCB residues in bottom sediments
throughout the country (Crump-Wieaner et al., 1974). Samples were
taken from lakes and streams generally located away from industrial
centers. The results are presented in Table 5.14. Of samples
collected at random from 16 states, 13 states had positive samples
containing PCB residues at levels ranging from 5 to 2400 ug/kg. One
out of every five bottom sediments across the nation contained PCBs.
Data from Florida (not shown on the table) revealed that 12 out of 231
water samples (5 percent) contained PCB residues at levels of from
0.1 to 2.1 ug/kg (median concentration, 0.2 ug/k«). Of 118 samples of
bottom sediment analyzed, 50 (42 percent) contained PCBs at levels
of from 5 to 3200 u^/kg (median concentration, 30 ug/kg). Once again,
5-59
-------
Table 5.13 PCB RESIDUES FOUND IN SAN FRANCISCO BAY AREA STREAMS
SITE
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
31
36
37
'38
39
STREAM
Colmn Cretk
Colma Creek
Bflmont Greek
CordUlkircm Creek
CordtlleroH Creek
Redwood Creek
San Franc 1 Htjuito
Creek
San Krancisqulto
Creek
Loq francos Creek
Stevens Creek
Stevens Creek
Los Gatos Creek
Los Gatos CrecSk
Guadalupe River
GuadaJupe River
Alamltos Creek
Coyote Creek
Coyote Creek
Alameda Creek
Alameda Creek
Arroyo de la Laguna
Arroyo de la Laguna
San Lorenzo Creek
Wildcat Creek
Wildcat Creek
San Pablo Creek
Union Creek
Green Valley Creek
Napa River
Napa River
Napa River
Sanoma Creek
Petaluma River
Navato Creek
Miller Creek
San Rafael Creek
Corte Madera Creek
Cortr Madera Creek
Arroyo Corte Madera
del Persidto
PCB RESIDUE
CONCENTRATION
IN pG/KG1"2
3.9
12
52
14
6.0
25
1.2
430
"2l"
180
30
0.0
170
O)
2.7
610
14
12
11
30
160
33
25
21
43
27
140
5.3
8.8
1400
7.6
5.0
27
10
35
350
81
if
24
Based on oven-dry weight of stream bed material uncorrccted for
-percent recovery.
Underlined values Indicate mass spectrometric confirmation of residue
-identity.
The presence of 5!) iig/kg of polych lorinated naphthalenes (PCN)
obscured any PCBf, present. PCN w;u; identified hy gas chromato-
graph/mass Hpectromeler.
Source; Adapted From Law, 1..M. and Coerlltz, D.I'., 19M
Selected Ch lor inat eel Hydrocarbons In Hot lorn Material from
SlrcamH Tributary to San Kr.mt-l.scci Bay. Pesticide Monitoring
Journal 8:33-16.
5-60
-------
Table 5.14
SUMMARY OF PCB RESIDUE DATA FOR BOTTOM SEDIMENTS
JANUARY 1971-JUNE 1972 '
-'
State
1 •
Alaska
Arkansas
California
Connecticut
Hawaii
Georgia
Maryland
Mississippi
New Jersey
Oregon
Pennsylvania
South Carolina
Texas
Virginia
Washington
West Virginia
•
-
No.
Samples
—
~
3
13
1
12
11
8
12
4
16
11
293
10
10
2
MIIIIB*HM«'MMMM
Occur-
rences
"
4
3
i
X
10
5
2
10
2
11
8
23
8
0
1
— — — — — — —
Concen-
tration,
Mg/kg
•
ND
20-2,400
20-190
40
ND
10-1,300
10-1,200
50;170
8-250
15;140
10-50
30-200
7.9-290
5-80
10
NOTE: ND - not detected. ' "
Median
Concentration,
Vg/kg
•
ND
60
85
ND
ND
300
30
ND
20
ND
20
50
80
40
ND
ND
•
SOUrCe: SSeTiV' J" "• "• FeUZ' a"d "• L' Y«-' »".
Pesticides in Water: A Study of the DistrihuMnn nf r» (
chlorinated Biphenyls in the Aauatir rjf*ributlon of Poly-
Monitoring Journal 8= 1S7_^ ' ^ Envlronraent- Pesticide
5-61
-------
it can be seen that PCBe accumulate in sediment to anywhere from 2.4
to 32,000 times their concentration in water.
The affinity of PCBs for sediment has been invoked to explain
the absence of an in-shore to open-ocean concentration gradient in
the surface waters of the North Atlantic (Harvey et^ a_l., 1973; Harvey
et al., 1974) when such a gradient exists in the atmosphere (Harvey
and Steinhauer, 1974). Since more PCBs are being delivered to the
coastal regions than to the open ocean, via the atmosphere, higher
concentration would be expected in coastal waters. However, the
higher particulate concentration in the coastal waters (100 times
greater than that in the open ocean) along with a more rapid sedi-
mentation rate enables a larger proportion of those PCBs delivered
to coastal areas to be adsorbed and sedimented out of the water.
Therefore, while most of the PCBs delivered to the open ocean remain
in solution, a good percentage of those delivered to the coastal
regions are deposited in the sediment. The net result is the absence
of any discernable concentration gradient.
E. TRANSPORT MODELS
The function of a transport model for a chemical pollutant is
to provide us with information as to the fate of that pollutant in
the environment. The model must trace the substance in question
from its entry into each susceptible environmental compartment,
through its transfer (if any) between compartments, to its decomposi-
tion (if any) within compartments, and finally to its ultimate sink(s).
5-62
-------
An accurate model will successfully account for present levels of the
pollutant in each compartment and will correctly predict the alter-
ations in environmental concentrations resulting from a change in
input.
Assuming that PCBs are similar in behavior to DDT, Nisbet and
Sarofira (1972) proposed a transport model which has held up reasonably
well in the face of evidence, both experimental and analytical. All
of the evidence and most of the results have been presented in the
previous sections of this chapter and will only be briefly reviewed
at this time. The reader should refer to the appropriate section
for details as to quantities of material released and transported
by each mechanism and the relative importance of each pathway as a
source of contamination. The model proposed by Nisbet and Sarofim
is presented in Figure 5.9. This model was based on pathways into
the environment prior to 1971, some of which are presumably no longer
a source of concern. However, the significant concentrations of PCBs
found in the atmosphere and hydrosphere (see Sections V.B. and '
V.C) years after the supposed discontinuation of their use for dis-
persive applications suggest that these pathways may still be operative.
Gross estimates of the rates of input via each pathway based on the
Interpretation of 1970 sales figures and assumptions as to the useful
life span of PCB-containing products are summarized in Table 5.15.
PCBs are released directly to the atmosphere via vaporization
from products, evaporation from themselves, or incomplete
5-63
-------
leaks >
dumping
waporat
icinera
fal
/
__/
g-*
Ji
v__r
,y
1-S
— /
*-'
ion and-^
tion
Ldut
AIR
TERRESTRIAL
SYSTEMS
leaching T '
(irrigation
^dredging
Mredging
FRESH WATER
river
T ,
biotj
flux
' |
ESTUARIES
t
mixing
' I '
— photolysis——^,
^_....__
evapor
co-dis
/
b
>iota
:lux
y
^>
i
, /
t
biota
flux
' I
OCEAN
,y
^
at ion and
tillation
V
11 r
% ^
Vl1 »
V *.
decomposition
Source: Nisbet, C. T., and A. F. Sarofim. 1972.
Rates and Routes of Transport of PCBs in The Environment.
Environmental Health Perspectives Exp. 1: 21-38.
FIGURE 5.9
ENVIRONMENTAL TRANSPORT MODEL
5-64
-------
Ln
I
TABLE 5.15 GROSS ESTIMATES OF RATES IN INPUT AND ACCUMULATION OF PCBS
IN NORTH AMERICA IN L(&)
Category of Input
.
Vaporization of plasticizers
/Vaporization during
/ open-burning
Leaks and disposal of
industrial fluids
Disposal in dumps and
landfills
Accumulation in services
Rates
(tons/year)
1-2 x 103
4 x 102
4-5 x 103
1.8 x 104
7 x 103
PCB Grade
Mainly 1248 to 1260
Mainly 1242
1242 to 1260
1242-1260
1242-1254
••
Compartment
Affected
atmosphere
Ltmo sphere
hydrosphere
lithosphere
Source: Adapted from Nisbet, C. T. and A. F. Sarofim. 1972
and Routes bf Transport of PCBs in the Environment '
Environmental Health Pers^-i^o Exp. ±. 21_3g "^
Rates
-------
incineration. Nisbet and Sarofim estimated that 1500 to 2500 tons per
year are released into the enviornment by this pathway. The PCBs are
carried through the air predominantly in the vapor phase (Bidleman
and Olney, 1974a) or adsorbed onto particulates and may travel great
distances in this form. It. is thought that the atmosphere provides the
major route of transport for PCBs throughout the environment
(SBdergren, 1972; Oloff et_ _al, , 1972; Mackay and Wolkoff, 1973;
Harvey and Steinhauer, 1974). PCBs are then deposited on land or
into fresh and marine waters by particle sedimentation or rainout
(estimated total fallout, 1000 to 2000 tons/year). The greatest
loss to the atmosphere occurs in industrialized or urban areas, and
decreasing concentration gradients in air are observed from the coast
out to the open ocean (Harvey and Steinhauer, 1974). Fallout onto
land and water is also greatest in these areas, close to the sources
of emission. R^-evaporation from the lithosphere and hydrosphere
is a further source of atmospheric contamination. The extent of
vaporization from terrestrial surfaces will depend upon many factors,
including the nature of the surface. Losses will be greatest from
surfaces such as sand and small from soils with a high organic matter
content (Haque _et_ _a_l. , 1974). Evaporation and codistillation from
the surface layer of bodies of water where PCBs accumulate due to the
high lipid content of the surface microlayer may be significant.
Shallow, fast-flowing rivers and streams where depleted surface layers
are constantly renouriahed will experience the greatest loss of PCBs
by this route (Mackay and Walkoff, 1973).
5-66
-------
Dumping of wastes in the liquid form via industrial and municipal
waste water effluents (4 to 5 x 103 tons/year) or in the solid form
as sewage sludge, solid waste, and dredge spoil (maximum 400 tons/year)
are further sources of contamination to both lithosphere and hydro-
sphere. Accidental spills and leaks should also be included in this
category. Chough leaching from sanitary landfills and garbage
dumps (which accumulate 1.8 x 104 tons of PCBs annually) is theoreti-
cally possible, evidence suggests that losses by this route are, in
fact, negligible. Reaching from certain terrestrial surfaces such as
sand and creviced bedrock into groundwater supplies is, however, a
more likely source of contamination as is leaching from paints and
surface coatings in contact with moist food products, e.g., silos
painted with coatings containing PCB. The reverse type of contamina-
tion, that is, contamination of terrestrial compartments via irriga-
tion, is a distinct possibility, especially if the water is taken
from a highly industrialized area. Contamination by this pathway was
estimated to amount to about 15 tons per year.
The fate of PCBs delivered to both fresh and marine waters is
adsorption onto particulate matter and, ultimately, deposition into
bottom sediment. Transport of PCBs in rivers to estuaries and the
open sea takes place by solution and readsorption onto particulates
and by the transport of sediment itself and was estimated to amount
to about 200 tons per year. PCBs dissolved in relatively shallow
water, in which there is close contact with bottom sediment, will adsorb
5-67
-------
onto the sediment rather than move to the surface layer and evaporate
into the atmosphere. Only in deep waters does evaporation compete
successfully with particulate scavenging and sedimentation (Oloffs
jet ad. , 1973).
PCBs may also travel through the environment via biota flux
and food chains. PCBs entering food chains may eventually end up
in migratory species which may transport them to relatively uncon-
taminated areas.
In any transport model, pathways of decomposition and other
avenues of disappearance from the biosphere must be considered. Since,
in general, the water solubilities and vapor pressures of the
various PCB isomers decrease with increasing chlorine content, the
lower isomers will more readily be dissolved, evaporated, and co-
distilled. The net result is a fractionation of PCB isomers with
the higher isomers exhibiting less mobility than the lower ones.
This fact may contribute to the observation that gas chromatographic
profiles of PCBs from contaminated enviornmental samples are frequently
rich in the higher isomers and usually resemble Aroclor 1254 or
1260 (Veith and Lee, 1971). The lower isomers also undergo metabolic
degradation by microorganisms, probably the major cause of their re-
moval from the blo&phere. Photolytic decomposition of some PCB
isomers in natural sunlight has been reported (Risebrough et_ al. ,
1968; Safe and Hutzinger, 1971; Hutzinger and Sage, 1972b; see Section
II.C). Apparently some higher isomers are broken down more easily
5-68
-------
than the lower isomers , perhaps contributing to the observed relative
abundance of lower isomers in atmospheric samples ( Bidleman and Olney,
1974a). In addition, photolytic dechlorination is also expected to
give rise to lower isomers (Hutzinger and Safe, 1972). Due to the
high stability of PCBs, other non-biological forms of decomposition
are expected to be extremely slow or negligible.
Nisbet and Sarofim concluded that most PCB isomers with four or
fewer chlorine atoms released into the environment have been degraded,
primarily by microbial decomposition; consequently, about 75 percent
of the Aroclor 1242 released, 60 percent of Aroclor 1248, 20 percent
of Aroclor 1254, and 5 percent of Aroclor 1260 have disappeared.
Assuming that the proportion of PCB sales for different applications
remained constant between 1930 and 197C-and that the percentage lost
during this period was identical to that reflected in Table 5.15 for
the year 1970, Nisbet and Sarofim estimated that cumulative losses to
the environment were 3 x 104 tons to the atmosphere, 6 x 104 tons into
fresh and coastal waters, and 3 x 105 tons into dumps and landfills.
From the grade of PCB released into each compartment (Table 5.15)
and the estimated percentage decomposition of each grade, it can be
concluded that roughly one-third of the PCBs released into the atmo-
sphere and one-half of those released into the hydrosphere during the
period in question have now been degraded. Little can be said about
those PCBs disposed of in landfills since they w,-re often in sealed
containers and, therefore, not available for bacterial degradation.
5-69
-------
Thus, 2 x 10 tons of the PCBs released into the atmosphere and
4
3 x 10 tonw of the PCBs releaaed Into the hydrosphere between 1930
and 1970 remain in the environment. Of the remaining PCBs originally
o
released into the atmosphere, about one-quarter (5 x 10" tons) are
expected to have been transferred to the sea; the rest (1.5 x 10
tons) were deposited in the lithosphere. Of the remaining PCBs origi-
nally released into the hydrosphere, 2 x 10 tons are expected to have
accumulated in the sediment of rivers and lakes (a small amount,
10 tons, is in solution in fresh waters) and 10 tons to have
reached the sea. These results are summarized in Table 5.16.
It is believed that, due to precautions presumably being taken
by manufacturers and consumers and to the voluntary restriction of
sales for dispersive applications, the major input of PCBs into the
environment has already occurred. However, present atmospheric and
hydrospheric samples suggest that a significant amount is still
entering the biosphere. PCBs are widespread and chemically stable.
Thus, even if a total cessation of both their manufacture and use were
possible, it would be many years before the environment would rid
Itself of these compounds.
5-70
-------
Table 5.16 PCB ACCUMULATION BETWEEN 1930 AND 1970
Reservoir
PCB Accumulation (tons)
Soil (excluding dumps)
Oceans Cadjacent to North
America)
Fresh water (dissolved or
in suspension)
Fresh water sediment
Dumps and landfills
1.5 x 10
1.5 x !(/
102
2 x 104
3 x 105
4
Source:
Adapted from Nisbet, C. T. and A. F.
Sarofim. 1972. Rates and Routes of
Transport of PCBs in the Environment.
Environmental Health PerapprM^g £xp
1: 21-38.~~
5-71
-------
'•f- '
VI. CONTROL OF PCBs IN THE ATMOSPHERE
A. SUMMARY
Proper safeguards will minimize the loss of PCBs to the en-
vironment. All PCBs should be handled, shipped, and used in closed
systems. However, some escape of PCBs from such systems is inevitable
as the result of spills, leaks, accidents, and equipment malfunction.
Workers are protected by good housekeeping procedures in the
plant; these will minimize PCB loss to the working environment and
ensure containment and prompt cleanup of spills. Protective clothing
and adequate ventilation systems are specific controls that safe-
guard the worker.
Loss of PCBs to the environment can occur during manufacture, in
consumer industries during use, and during disposal. PCBs may enter
directly through the atmosphere, lithosphere, or hydrosphere but,
because of media distribution and transport patterns, these realms
are interconnected and PCBs are present in all three.
PCB losses to the atmosphere can be controlled rather effectively
if the PCBs are handled, used, and disposed of properly. PCB losses
to the lithosphere can theoretically be controlled, although accidents
will always be a prime source; a major problem is the installation of
controls for transformers and capacitors that are already installed.
Control of PCB losses to the hydrosphere is currently the least ef-
fective. Although there are some promising developments, apparently
there is now no technology that can ensure zero PCB discharge from
6-]
-------
a point-source effluent. Control is, therefore, best effected by
preventing contamination of sewage.
The controls that are currently available should be evaluated for
effectiveness, cost benefit, etc. New controls and technology should
be developed and evaluated and then made available as, aoon as possible.
There is a need for a manual that will detail the hazards of PCBs and
the recommended procedures for the handling, use, and disposal of PCBs
and PCB-contaminated materials. This manual should list controls,
methods, devices, materials, and technologies that are available for
controlling the loss of PCBs to the environment. This manual should
be distributed widely and should be readily available.
B. INTRODUCTION
Workers may be exposed to PCBs at production facilities, in
consumer industries, and wherever PCB-containing products are used.
If the American National Standards Institute (ANSI) guidelines (1974)
are followed, worker exposure should be nil inasmuch as the recommended
procedures entail use of closed systems for the manufacture, transport,
and handling of PCB-containing askarels (non-flammable dielectric in-
sulating fluids). Nevertheless, even when these guidelines are followed,
workers may still be exposed to PCBs as a consequence of equipment
malfunction or breakdown, leaks, spills, and various accidents.
The general populace is exposed to PCBs when PCBs enter the ambient
environment from the working environment or from use of products
containing PCBs. This can occur when the recommended procedures are
6-2
-------
not followed, when leaks and spills are not contained and cleaned up,
and when PCBs and PCB-contaminated items are not disposed of properly.
Any of these events introduce PCBs into the atmosphere and/or litho-
sphere and/or hydrosphere.
C. CONTROL OF PCB LOSSES AT PRODUCTION FACILITIES AND BY CONSUMER
INDUSTRIES
lt Methods for Control of Worker Exposure
The American Conference of Governmental Industrial
Hygienists (1964, 1973) has recommended that the maximum atmospheric
concentration per eight-hour workday be 1.0 mg chlorodiphenyl (42 wt %
CD per cubic meter of air and 0.5 mg chlorodiphenyl (54 wt % Cl) per
cubic meter of air. Air monitoring is probably not necessary inasmuch
as the odor of PCBs is strong, very unpleasant, and irritating at
concentrations that are much lower than the threshold value, and
consequently the odor threshold would provide ample warning of the
presence of PCB vapor (Papageorge, 1975). Nevertheless, Monsanto does
-niter for PCBs (levels are non-detectable in the working environment),
and it is also developing a portable, battery-operated detection
unit to be carried by each individual worker (Papageorge, 1975).
The American Industrial Hygiene Association (AIHA) (1965)
recommends the use of chemical cartridge respirators or gas masks
approved by the U.S. Bureau of Mines should worker exposure to higher
concentrations of PCBs be unavoidable (e.g., ln the evenfc of leaks
or spills of hot fluids). The ANSI guidelines (1974) cite.the
need for maintenance of gas masks, respirators, and replacement parts
6-3
-------
on a regular basis. Monsanto (1975b) specifies that when hot askarel
must be handled in a closed or confined area, the area should be
provided with mechanical exhaust ventilation or the workers should
wear an organic cartridge respirator approved by the U.S. Bureau of
Mines. The instructions of the U.S. Army Environmental Hygiene Agency
(1975) specify adequate ventilation of 150 cubic feet per minute per
square foot of area when hot askarels (>55°C) are handled.
Prolonged or repeated skin exposure to PCBs should be avoided
(AIHA, 1965). The Institute of Electrical and Electronics Engineers
(IEEE, 1974) recommends that no type of glove be used. On the other
hand, ANSI (1974) specifies that porous gloves be avoided and that
barrier creams or resistant gloves be used. Similarly, the use of
chemical gloves is recommended in the General Motors guidelines (1974)
and by the U.S. Army Environmental Hygiene Agency (1975). Exposed
skin surfaces should be washed thoroughly and immediately with soap
and water (AIHA, 1965),-and then cold cream (ANSI, 1974; IEEE, 1974)
or silicone-bearing hand lotion (IEEE, 1974) should be applied.
When the splashing of askarels is possible, workers should use
eye protection such as glasses or shields (IEEE, 1974) or safety
glasses with side shields or face shields (ANSI, 1974; General Motors,
1974). If askarel does contact the eye, then the recommended first
aid is flushing the eye with large quantities of water (Monsanf.o, 1975b)
for 15 minutes (AIHA, 1965; ANSI, 1974) followed by application of
castor oil (ANSI, 1974; IEEE, 1974; Monsanto, 1975b). The worker
6-4
-------
should see a physician who may prescribe an ophthalmic anesthetic
solution to relieve Irritation and inflammation (AIHA, 1965; Monsanto,
1975b).
In General Motors plants, workers use chemical aprons to protect
their clothing from askarel (General Motors, 1974). Monsanto provides
its workers with work clothing (Papageorge, 1975). Contaminated
clothing should be removed promptly (AIHA, 1965), and subsequently it
should be dry-cleaned (Monsanto, 1975b) or otherwise properly laundered
or disposed of (ANSI, 1974). The U.S. Army Environmental Hygiene Agency
(1975) instructions specify that lightly contaminated clothing can be
laundered in warm soapy water with the wastewater "discharged to the
sanitary sewer at a rate consistent with the 1972 Environmental Pro-
tection Agency proposals to keep PCB levels in rivers and lakes below
0.01 ppb," whereas saturated clothing should be stored in steel drums
prior to disposal by incineration.
Ingestlon of askarels Is not a problem in industry (ANSI, 1974).
Good personal hygiene habits will greatly reduce this possibility.
The ANSI guidelines (1974) state specifically that workers should wash
their hands with soap and warm water before eating, drinking, smoking,
and using toilet facilities.
The AIHA (1965) recommends that persons who are regularly or
repeatedly exposed to PCBs should be examined periodically for evidence
of skin irritation and/or liver damage. ANSI (1974) was less stringent;
Its guideline specifies that a person who develops a skin Irritation or
6-5
-------
respiratory tract irritation while working with askarel should be
placed under the supervision of a physician.
It should be noted that this section is limited specifically to
those aspects of control of worker exposure to PCBs that pertain to the
worker's person, clothing, and immediate working environment as well
as personal health and hygiene. Obviously, the conditions in the plant
and the general working environment (which are discussed below in
Sections VI.C.2, VI.D.I, and VI.D.2) have a direct bearing on the worker.
2- Methods for Control of External Emissions
The recommendations of both ANSI (1974) and Monsanto
(1975b) emphasize that proper handling of askarels containing PCBs
necessitates the use of closed and sealed systems that will prevent the
entry of PCBs into the environment. ANSI (1974) and Monsanto (1975b)
as well as the IEEE (1974) have published specifications for piping,
hoses, connections, gaskets, valves, packing, cements, storage and
shipping drums, and tank cars that are used in the handling/transport
of PCBs.
In addition, certain basic precautions should be taken (ANSI, 1974;
Monsanto, 1975b) which involve general good housekeeping procedures
based on the necessity for keeping all PCBs as well as all PCB-contami-
nated items separate and apart from non-contaminated materials. This
entails separation, containment, and proper disposal of all PCB-con-
taralnated material.
6-6
-------
Even if all these guidelines are followed closely, control method-
ology must be available because leaks, spills, and other accidents do
occur. Prompt and proper cleanup procedures will effectively keep
PCBs from the environment.
Although the primary subject of this study is PCBs in the
atmosphere, the transport and transformation of PCBs in other media
oblige us to consider as well control methodology for the lithosphere
and for the hydrosphere.
a. Losses to Atmosphere. Monsanto (1973b) has endeavored
to eliminate the emissions of PCBs to the atmosphere from the manufactur-
ing area. Rupture disc lines and atmospheric vents were rerouted
through catch tanks or to settling basins. Mist eliminators (Brink
Mist Eliminator, Enviro-Chem Systems Company) were placed in all vapor
lines; these serve to condense and collect the PCB vapors (Papageorge,
1975). Use of the Brink Mist Eliminator in the ventilating system
prevents escape of PCBs into the air from the manufacturing area to other
parts of the plant and to the atmosphere. Storage tanks are blanketed
in nitrogen to eliminate "breathing" (Monsanto, 1973b). In addition,
all air from the ventilating system is passed through the incinerator
(Papageorge, 1975).
General Electric has eliminated pouring and the use of open con-
tainers of PCBs in the process of filling transformers. Thus, all
PCBs are now handled in a closed system, and the askarel is drawn by
vacuum into the previously evacuated transformer (Bachli, 1975).
6-7
-------
Monsanto (1975b) specifies that all tank cars must be unloaded
through the top (dome fitting) by pumping or with controlled pressure
using dry air or dry nitrogen. This technique prevents loss of PCBs to
the atmosphere during the loading and unloading of tank cars. Loading
and shipping safeguards are included in specific written instructions
on every container (Monsanto, 1973b).
An activated carbon unit (e.g., VentSorb by Calgon Corporation)
hooked up to the vent of a storage tank would effectively prevent the
release of PCBs to the atmosphere from such a point source (Langston,
1975). Since the adaorptive properties of activated carbon depend on
its characteristics (pore number and size, surface area, etc.) as
well as on temperature, pressure, and rate of flow, it is possible
that an activated carbon unit attached to the ventilating system of a
plant could adsorb PCBs from the air stream; however, we could not
ascertain that such a system is currently in use.
Another device that might remove PCB vapor is a charged droplet
scrubber (CDS) such as that produced by TRW. Its effectiveness in
controlling emissions of fine particulates has been demonstrated
(see Section VI.E.I). In theory, it will also capture PCB vapors and
aerosols and, therefore, could be used effectively in ventilating sys-
tems, but field testing is needed to confirm its utility (Whitson, 1976).
Control of PCB losses to the atmosphere that result from vapori-
zation of spills and from codistillation is discussed in the sections
on the lithosphere (VI.C.2.b.) and the hydrosphere (VI.C.2.C.),
respectively.
6-8
-------
b' l£8sesjtoJLltho8£here. The ANSI guidelines (1974) mention
control measures such as making provision for containment around all
askarel processing areas and the prompt cleanup of spills by absorp-
tive materials or by trapping and pumping.
The steps taken by Monsanto (1973a) exemplify specific control
techniques. In the manufacturing area, which may be described as a
"concrete bathtub," all drainage is directed to trenches and piping
and then to a settling basin; thus all spills are contained and removed.
On every shift, all pump8 are checked for leakage, drip pans and trays
are emptied into scrap PCB drums, and all leaks are reported and docu-
mented so that corrections are made and the settling basins are observed.
Liquid spills and leaks are promptly cleaned up by absorption.
Recommended adsorption materials include rags, sawdust (ANSI, 1974;
Monsanto, 1975b) and adsorbent clay (Monsanto, 1975b). Imbiber polymer
beads (Dow Chemical Company) for the cleanup of PCB spills are being
marketed by Gedcor Environmental Protection Corporation in the form
of beads, bags, and blankets; they can imbibe (adsorb) 27 times their
volume and the resultant solid is handled more easily and can be
handled and transported without danger of spills and leaks (Taylor,
1975).
Adsorptive materials that are effective in cleaning up PCB spills
include activated carbon and urethane foam.
C> Losses to Hydrosphere. ANSI (1974) recommendations
specify the establishment of containment provisions to ensure against
6-9
-------
inadvertent loss of PCBs to sewer systems in the event of spillage,
leakage, or other uncontrolled conditions or events. Ideally, all
sewer systems and effluent streams that could be contaminated with
PCBs should be isolated. This has been achieved at the Monsanto
manufacturing facility where all sewers are now above ground and com-
bined (Monsanto, 1973b).
PCB-contaminated water need not be discharged from the plant.
General Electric reportedly recycles it or drums it for suitable dis-
posal (Anonymous, 1975b). Carbon adsorption, limestone beds, and
solvent extraction are techniques that can be used prior to recycling
(ANSI, 1974).
The problem of PCBs in effluent discharge has been the subject
of much attention (e.g., EPA, 1973; Courchaine, 1975; Anonymous, 1975a)
Various systems have been devised that do reduce significantly the
PCB concentration in effluents.
Once such system is the Calgon Adsorption Service (CAS) of the
Calgon Corporation (Louros, 1975). The Calgon adsorption unit uses
granular activated carbon, which has a very high capacity for adsorb-
ing PCBs, i.e., its adsorption efficiency is more than 99 percent for
dissolved PCBs. CAS will install, operate, and maintain the unit,
guarantee the effectiveness, monitor the effluent, make a monthly
report, and exchange the exhausted carbon as needed. Validation
studies are now in progress. On the basis of current data, Calgon
will guarantee that effluent PCB concentrations will not exceed 500
6-10
-------
ppt provided that the equipment is operated properly, the design and
operating limitations are not exceeded, and the built-in safety factor
is not exceeded. Two such units should be on-line for clients by
mid-1976. The exhausted carbon can be incinerated, which destroys
PCBs. Alternatively, it can be regenerated in a furnace; this process
volatilizes the PCBs which are then led to an afterburner where they
are destroyed.
A California-based company, Envirogenics, has developed a system
for the iron-catalyzed reduction of organochlorine compounds present
in the hydrosphere (Sweeney, 1975). Their system has been success-
fully applied to a variety of organochlorine pesticides and Velsicol
is planning to use the method to control the levels of endrin in the
production plant effluents. A pilot program to assess the feasibility
of catalytic reduction of PCBs was established in December 1975. The
method dechlorinates the PCBs but the exact nature of the products and
the extent to which dechlorination occurs is not yet known.
One unnamed PCB polluter has expressed interest in this form of
effluent control since it appears as if the cost will be lower than
for carbon adsorption type control devices. The reductive dechlorina-
tion system should be available commerically by the end of 1976.
The Gedcor Environmental Protection Corporation is experimenting
with the Dow imbiber beads (Taylor, 1975) in an attempt to perfect
a system that can be used for effluent filtration; at present, their
device functions as a safety valve that shuts off flow in the event
of a spill.
6-11
-------
Although as noted above, there are some promising developments
in controlling PCS release to the hydrosphere, apparently no method
has yet been devised that can guarantee zero concentration of PCBs in
effluent discharge.
D. CONTROL OF PCS LOSSES IN USE
Use of PCBs is currently restricted to the askarels that
are used in electrical equipment, primarily in transformers and capac-
itors. These are closed systems; consequently, the loss of PCBs
would occur primarily through leakage, equipment malfunction or fail-
ure, and accident.
1. Losses to Atmosphere
The principal source of PCB loss directly to the atmo-
sphere is the leakage from askarel-filled capacitors and transformers
which are located atop power poles (Klapp, 1975). The power-factor-
correction capacitors of the Forces Command of the U.S. Army that
are situated on top of 40-foot power poles are usually dry; i.e., air
serves as the insulator. Those which do contain dielectric fluids are
old units that hold about one pint (0.47 liter) of askarel each
(Klapp, 1975). The pole-mounted power-factor-correction capacitors
used by Virginia Electric and Power Company contain about one to two
gallons (3.8 to 7.6 liters) of dielectric fluid, which is usually
askarel. They are hung in racks, usually in banks of four or five
(Lewis, 1976). Pole-mounted distribution transformers contain as much
as 50 gallons (189 liters) of askarel, with the amount depending on
the size of the transformer (Lewis, 1976).
6-12
-------
If the leak is a slow one, the leaking askarel drips or runs
down to the ground and the leak might continue for some time before it
is noticed. If the unit bursts, the leak is rapid and the askarel
would be sprayed out and dispersed over a wide area. Either way,
vaporization is more likely from pole-mounted capacitors and trans-
formers because of their location and the droplet form of leakage.
Many capacitors are now being installed in banks of about a
dozen within a drip pan (Papageorge, 1975). Although this would
prevent direct contamination of the lithosphere, the askarel accumulat-
ing in the drip pan would be subject to vaporization, especially since
the location atop a power pole would preclude observation and frequent
cleanup of any leakage that might occur.
2' Losses to Lithsophere
The control of PCB loss from capacitors atop power poles
can be effected by the use of drip pans in which the capacitors are
set (Papageorge, 1975) (see Section VI.D.l.). Drip pans could also
be placed below pole-mounted transformers to catch and contain any
leakage.
On the other hand, pad-mounted surface-level transformers are
usually much larger units, and the largest contain more than one
thousand gallons (3,785 liters) of askarel (Hall and Haigh, 1974;
Hall, 1976). Therefore, in order to prevent PCB contamination of the
lithosphere in the event of spills, leaks, or equipment malfunction,
it is essential that control measures be applied wherever transformers
6-13
-------
are used. These controls are of three types: spill prevention, back-
up measures for containment, and cleanup procedures (Obold, 1975).
Safety valves on transformers shut off the flow of askarel
when the equipment malfunctions. One such device that uses imbiber
beads (Dow Chemical Company) is distributed by the Gedcor Environmental
Protection Corporation. It has been installed on transformers in Dow
Chemical Company plants (Taylor, 1975).
To ensure containment within a given area, all floor drains must
be sealed, the floor must be impervious, and diking must surround
the containment area (Courchaine, 1975; Obold, 1975; Taylor, 1975).
An alternative is to place the transformer within a vault.
Should a spill or leakage occur within a contained area, the
askarel can be pumped out (Monsanto, 1973b) or cleaned up. Thus, in
the Dow Chemical Company plants, imbiber beads or blankets are placed
in the containment areas under the transformers. The Salt River Project
in Phoenix, Arizona, uses imbiber-bead blankets to clean up spills in
contained areas (Taylor, 1975). The various other materials that are
used to clean up spills in manufacturing and production facilities
(see Section VI.C.2.b) are equally suitable for leaks and spills in
other contained areas; these include rags, sawdust, activated carbon,
urethane foam, and adsorbent clay.
Another source of lithosphere contamination is the runoff from
rainwater that washes outdoor transformers (Papageorge, 1975). A
Gedcor ecology protection system (Gedcor Environmental Protection
6-14
-------
Corporation) is now being installed at the General Motors Cadillac
plant in Detroit, Michigan, to solve this problem. There will be a
cement base under the transformer and a 12-inch high dike surrounding
the area. A 4-inch pipe coupling will pass through the dike, and to
it on the outside will be attached an imbiber bead valve (Taylor,
1975). This system will permit the drainage of rainwater and melted
snow from the transformer containment area, but PCS contaminant will
be stopped by the imbiber bead valve.
There have been at least two accidents with large askarel-con-
taining transformers that resulted in significant contamination of the
lithosphere. One occurred when an 800-gallon askarel-containing trans-
former was dropped while it was being installed in Chicago, Illinois,
(Willmore, 1976). The other accident occurred near Kingston, Tennessee,
during shipment of a 1400-gallon askarel-containing transformer (Hall
and Haigh, 1974; Hall, 1976). Such accidents cannot be completely
eliminated. The best method for controlling this type of PCB con-
tamination is rapid, effective, and thorough cleanup of the spill.
The EPA Oil and Hazardous Materials Spills Research Branch in
Edison, New Jersey, is now working on a pilot project to determine the
feasibility of a shroud for use in cleaning up land spills. It would
be rapidly deployable and would be suitable for use with tank cars,
transformers, etc. The shroud would have a wet scrubber on one end
and be capable of pulling 40,000 cfm. The unit would function as a
mixing chamber into which could be added water, activated carbon, ion
6-15
-------
drums (Goff, 1975). It is estimated that about 50 more gallons of
askarel are still recoverable (Blaszevich, 1975), and that dredging
will yield 50,000 cubic yards (38,320 cubic meters) of PCB-contaminated
sediment (Goff, 1975). The rest of the spilled askarel is probably
lost to the environment; some penetrated the soil, some migrated up-
it
stream with the salt wedge , some was dispersed throughout the harbor
(Blaszevich, 1975).
E. CONTROL OF PCB LOSSES IN DISPOSAL
The recommended method of disposal of PCBs and PCB-contami-
nated liquids and software (e.g., clothing, rags, sawdust) is by high-
temperature incineration, whereas contaminated hardware (e.g., trans-
formers) should be solvent-rinsed and buried in a sanitary landfill
(ANSI, 1974; U.S. Army Environmental Hygiene Agency, 1975). The
disposal services offered by various companies are listed in Table
6.1.
PCBs and PCB-contaminated materials are shipped to disposal
facilities. The ANSI (1974) and IEEE (1974) specifications for drum
material, construction, and fittings also apply to drums used for
waste materials. Gedcor Environmental Protection Corporation is
marketing vapor-barrier bags and transport boxes for the handling and
shipment of waste askarels; their Burn Box (Hedwin Cubitainer loaded
with an appropriate quantity of Dow imbiber beads) seems to be quite
suitable for the disposal of liquid askarels. The Imbiber beads
*A wedge-shaped mass of salt seawater which intrudes into the fresh
river water.
6-17
-------
TABLE 6.1
FACILITIES AND SERVICES FOR PCB DISPOSAL
Company
Chem-Irol Pollution
Services, Inc.
Gedcor Environmental
Protection Corp.
General Electric Co.
Monsanto Industrial
Chemicals Co.
Nuclear Engineering Co,
Rollins Environmental
Services, Inc.
U.S. Armv Forces
'.***
Command
Location
Model City, N.Y.
Westland, Mich.
Pittsfield, Mass.
St. Louis, Mo.
Sheffield, 111.
Bridgeport, N.J.
Baton Rouge, La.
Houston, Tex.
Army posts
Material
Handled
Liquids &
software
Hardware
Liquids
Liquids
Liquids
All liquids
& solids
Liquids S
software
Hardware
Liquid &
software
Method
Incineration
Burial
Incineration
Incineration
in liquid-
injection
incinerator
Incineration
Burial in
Class I
landfill
Incineration
Incineration
i in pathologi-
cal incinera-
tors
Reference
Taylor, 1975
Carnes, 1975
Papageorge,
1975
Koneval, 1975;
Sernyak, 1975
Klapp, 1975
*Gedcor Environmental Protection Corporation provides shipping containers (Burn Boxes) and
disposal services. Actual disposal is handled under contract with a commercial disposal
facility (or facilities).
**General Electric Company put this liquid-injection incinerator into operation in 1973; they
handle PCBs from other companies, too, but this service is not advertised.
***The pathological incinerators are for use only by the U.S. Army; these are not commercial facilities.
-------
solidify the askarel and consequently there is no danger of spills or
leaks during shipment to a disposal facility where the entire unit is
incinerated.
Under proper conditions of high temperature incineration, PCBs
are broken down. The: Important conditions are temperature, residence
DPR >2000°F
PCBS ~2-3% exce^-Qj-> C°2 + H2° + HC1
1.5-2.0 seconds
time, and the concentration of excess oxygen in the stack gas. The
specific conditions recommended by ANSI (1974) are two seconds at
2000°F with 3 percent excess oxygen or 1.5 seconds at 2700°F with
2 percent excess oxygen. (Scrubbers are needed to remove the HC1.)
It should be noted that the liquid-injection incinerator
developed by General Electric and James Zink operates somewhat differ-
ently. Waste liquid is injected into the horizontal incinerator
cylinder under considerable pressure and it therefore atomizes almost
instantly upon injection. Operating conditions are: residence time -
2 to 12 seconds; temperature - 1200 to 1800°F; and excess air - 50 to
100 percent (Bachli. 1975). Because of these modifications, emissions
are almost zero (Carnes, 1975). Five test runs were designed to
demonstrate the efficiency of the incinerator in destroying DDT. The
waste industrial oil burned to provide additional heat was contami-
nated with approximately 1.7 percent PCB. Therefore, these tests
also evaluated the effect of the incineration on PCBs. The destruc-
tion efficiency for PCBs ranged from 99.9921 percent to 99.9995
6-19
-------
percent, with an average of 99.9956 percent (Leighton and Feldman,
1975). Carnes of the EPA Solid and Hazardous Waste Research Labora-
tory in Cincinnati, Ohio, commented that this incinerator is a "good
and well-operated system, the most efficient incinerator" he has
seen (1975).
Lockheed Corporation in Palo Alto, California, is currently en-
gaged in a study on low-temperature plasma destruction for hazardous waste
disposal, and PCBs are included in this study. (This project is con-
tracted by USEPA, Solid and Hazardous Waste Research Division, Munici-
pal Environmental Research Laboratory, Cincinnati, Ohio, under EPA
contract 68-03-2190, "New Methods for Efficient Detoxification/Destruc-
tion of Hazardous Wastes," with Don Oberacher as the EPA Project
Officer.) The preliminary findings from this study are promising;
the economics are unknown (Carnes, 1975).
1. Losses to Atmosphere
If one assumes that the commercial disposal facilities
that handle PCBs do operate properly — e.g., the Rollins Environmental
Services facility in Bridgeport, New Jersey, is certified by the State
of New Jersey for waste industrial incineration with no restrictions
(Koneval, 1975) — then these incinerators would not be a source of
PCB contamination of the atmosphere. v/However, most municipal incinera-
tors probably do not operate under the designated conditions that are
necessary for PCB destruction. Consequently, they constitute sources
of PCBs, especially since they would be handling non-labelled PCBs
and PCB-contaminated wastes as well as other PCB-containing products.
6-20
-------
In their study of this problem, Carnes et al. (1973) detected
PCBs In the residue fines and also in the fly ash from particulate
control devices of several municipal incinerators. One fly ash sample
C from Media, Pennsylvania) was taken from the bottoms of a clarifier
of a wet-wall, baffle-type scrubber; a second sample (from New Orleans,
Louisiana) was from the hopper of a multjple cyclone-type collector.
Carnes concludes that this finding indicates that: (1) PCBs are heat-
resistant; (2) PCBs are probably associated with (have an affinity for)
particulates and fly ash; (3) PCBs enter the air in fly ash; (4) an
undetermined fraction of PCBs emitted by incinerators is removed by
y
particulate control devices; and (SfPCBs are probably also vaporized
and enter the atmosphere in the gas emissions (Carnes, 1975).
Various scrubbers are effective in removing complete and incom-
plete combustion products from emissions prior to venting from the
stacks. These include cyclones, venturi scrubbers, high-energy
scrubbers, electrostatic precipitators, and fiber filter bags. Charged
droplet scrubber systems (e.g., that manufactured by TRW) effect re-
moval of fine particulates in the micron and submicron size range
(TRW, 1975; Lear et al., 1975). Inmost incinerators, however, the
only scrubbing that occurs is when exit gas containing fly ash passes
into a baffle system that has water running down the wall (Carnes,
1975).
PCB vapors can probably be removed from the emissions by alkaline
scrubbing at pH 11 to PH 12 ( Carnes, 1975). Packed-bed scrubbers
6-21
-------
are also effective in eliminating vapors; the limiting factor is
economics.
The absorbent liquid from scrubbers would contain PCBs. This
includes the PCBs associated with the particulates that dissolve in
the supernatant liquid while the particles are settling in the clarif-ier
as well as the PCS vapor that is scrubbed. Discharge of PCB-containing
liquid from scrubbers into ambient waters or even into sewage systems
will contaminate the hydrosphere (see Section VI.E.3).
2. Losses to Lithsophere^
PCBs can contaminate the lithosphere when they are disposed
of improperly. Waste askarel is probably disposed of correctly be-
cause of the publication of guidelines for its handlingCe.g., ANSI,
1974; U.S. Army Environmental Hygiene Agency, 1975). However, PCB-
containing materials are often discarded directly in landfills. In
addition, residues and fly ash from incinerators and sludge from waste
water treatment plants, which do contain PCBs C Games et al. , 1973;
Hesse, 1975), are usually deposited in landfills.
Proper containers will prevent leaks and spills and also the
leaching of PCBs. ANSI (1974) recommends the use of sealed vapor-
barrier containers; the instructions of the U.S. Army Environmental
Hygiene Agency (1975) specify steel drums and also the inclusion of
absorbent material (e.g., vermiculite, sawdust) in the drum when any
liquid askarel is present. Whenever possible, all scrap hardware
6-22
-------
should be drained of askarel and then washed with solvents to re-
move all traces of PCBs prior to disposal by burial or recycling of
the metal (ANSI, 1974).
When burial is the method of disposal, a Class I landfill site
should be used. Such a site may accept any type of waste including
toxic and hazardous materials because the fill will not enter the
groundwater. Class I landfill sites are located in special geologic
areas (Baird, 1975) or in special structures such as deep wells (Mon-
santo, 1975b) or empty missile silos (Goff, 1975). Empty missile silos
in southern Idaho were used for disposal of the PCB-contaminated
sludge from the Seattle harbor (see Section VI.D.3) (Goff, 1975).
3. Losses to Hydrosphere
PCBs can enter the hydrosphere by runoff of rainwater
from improperly disposed-of PCBs and PCB-contaminated materials.
Leaching can transfer PCBs from lithosphere to hydrosphere, and PCBs
were detected in leachate from landfills (Hesse, 1975). See Section
VI.E.2 for a discussion of land disposal methods and landfill sites.
The absorbent liquid from scrubbers that are used to control PCS
emissions from incinerators (see Section VI.E.I) is a potential source
of PCS contamination of the hydrosphere. Even discharge into sewage
systems will not prevent this contamination since municipal waste
water treatment plants remove only about 70 percent of influent PCBs
with the sludge (Section < V.C.3 ). Carbon adsorption units now under-
going validation tests by Calgon Corporation can reduce PCS concentra-
tion in aqueous effluent to 500 ppt (see SectionVLC.Z.c ). There are
6-23
-------
calls about PCBs to an attorney (Rabb, 1975). This manual would
probably be useful to the. consumer industries, and it would certainly
be very important to architect and engineering firms for their design
work.
1. Lossesto Atmosphere
Existing technology is probably sufficient to control
PCB losses directly to the atmosphere. The proposed manual should
list the devices that effect control for specific situations — e.g.,
mist eliminators for ventilating systems, scrubbers for incinerators.
It would be difficult to secure compliance with recommendations if
there is no legal basis for enforcement.
New technology may eliminate some current sources of PCB contami-
nation. For example, can light-weight economical vaults be devised
for enclosing capacitors and/or transformers? Is low-temperature
plasma destruction of PCBs feasible, economically as well as techni-
cally, and is this method emission-free?
2. Losses to Lithosphere
The problem of leak/spill containment is rather simple.
However, clean-up procedures are haphazard. Materials should be
studied for suitability, effectiveness, efficiency, availability, and
cost-effectiveness. Data from this study should be included in the
manual. Alternative clean-up materials and procedures should also
be evaluated. The shroud (Section VI.I).2) should be fully evaluated.
6-25
-------
3. Losses to Hydrosphere
Although there are some promising developments, apparently
there is no technology currently available that can effect arid guarantee
zero discharge of PCBs from a point-source effluent. Although it is
important that PCBs be prevented from entering the industrial sewage,
some contamination can and does occur. It is concluded that proper
technology be developed to ensure zero PCS concentration in effluent
discharge into waterways.
6-26
-------
VII. ENVIRONMENTAL EFFECTS OF PCBs
A. SUMMARY
PCBs accumulate in body tissues since they are readily absorbed
by inhalation, ingestion, and dermal contact but are not easily
metabolized or excreted.
The chronic toxicity of PCB appears to be of greater significance
than acute toxicity. The major target organ of PCBs is the liver.
Repeated exposures of laboratory animals to low levels of PCBs results
in hypertrophy of the liver accompanied by induction of liver microsomal
enzyme activity and structural changes in liver cells. Dermatological,
gastro-intestinal, hematological, and neural symptoms have been observed
in humans and animals exposed to PCBs. PCBs have been implicated as
human teratogens. Carcinogenic and embryotoxic effects of PCBs have
been observed in laboratory mammals.
B. EFFECTS ON HUMANS .AND LABORATORY ANIMALS
!• Absorption and Accumulation
PCBs tend to accumulate in body tissues since they are
readily absorbed through ingestion, inhalation, or dermal contact, but
they are not easily metabolized or excreted. Although human exposure
is mainly through the diet, PCBs will be inhaled wherever they exist in
the atmosphere.
Absorption through inhalation has been measured in laboratory
mammals. Tomborgs (1972) reports a German investigation which found
very high absorption in rats after a single inhalation of a mixture
7-1
-------
of low chlorinated biphenyls. The concentration in body tissues depended
on the time since the end of aerosol exposure. Fifteen minutes after
exposure, the concentration in the liver reached 50 percent of the
maximum concentration attained after two hours (70 Kg/g tissue). After
30 minutes, the concentration in fat was 14 ^g/g tissue (27 percent of
liver concentration at that time) and the concentration in brain tissue
was only 9 ug/g tissue (17 percent of liver concentration). Treon et
al. (1956) exposed rats, mice, rabbits, guinea pigs, and a cat to PCB
vapors five days a week for several weeks. Aroclor 1242 at concentra-
tions of 1.9 and 8.6 pg/1 (mg/m3; the higher number is approximately
1 ppm) had no ill effects. Aroclor L254 at concentrations of 1.5 and
5.4 mg/m3 (0.11 and 0.41 ppm) caused enlarged liver in the rats.
Concentrations of PCBs tend to parallel lipid content so that PCB
storage in body tissues occurs in the following order (Curley et al., 1971)
fat > liver > feces > kidney > brain > plasma ~> urine
The concentration in adipose tissue is 10 to 100 times the concentration
found in other tissues both early after single doses (Grant et al., 1971)
and after prolonged intake (Curley et al., 1971). PCBs are widely
distributed in human adipose tissue. From human adipose tissue studied
in their laboratory, Price and Welch (1972) concluded that 41 to 45
percent of the general population haw PCB levels of 1.0 ppm or more in
their adipose tissue. Yobs (1972) reported results of analyses for
PCBs in adipose tissue of the general population of the United States.
Of 637 samples analyzed, half were found to contain PCBs. Of these,
198 (31 percent) contained more than 1 ppm but only 33 (5.2 percent)
7-2
-------
contained more than 2 PPm. Samples were collected from 38 cities in
18 states. Positive samples came from every city sampled. It, there-
fore, appears that PCBs are widely distributed in human adipose tissue
in the United States. Biros et al. (1970) in a qualitative analysis
by combined gas chromatography and mass spectrometry (G.C.-M.S.) of
two human adipose tissue samples found substantial quantities of PCBs
ranging from pentachlorobiphenyls to decachlorobiphenyls. The PCB
residues included at least 14 isomers and homologs. Price and Welch
(1972) found 10 PCBs in human adipose tissue ranging from hexachloro-
biphenyls to nonachlorobiphenyls .
Continuous low level exposure can lead to bioaccumulation of
PCBs in fatty tissues of mammals. Even after relatively short term
low-level exposures PCBs accumulated in adipose tissue can be retained
for a prolonged period. Allen et al. (1974a) fed six adult female
rhesus monkeys 25 ppm of Aroclor 1248 for two months. PCB concentra-
tions in samples of adipose tissue averaged 127 ^g/g fat for all animals,
At that time, the experimental diet was discontinued. Eight months
later the PCB content was 34 (j.g/g fat. Curley et_ al. (1971) fed
Sherman weanling rats a diet containing 100 ppm Aroclor 1254 for 58
days and for 240 days. Animals were sacrificed at various intervals.
A steady buildup of PCB in all tissues was observed over the 58 day
period. The rats stored more PCB in their tissues after 240 days than
at the end of 58 days. In experiments at two dose levels (100 ppm
and 500 ppm) for 240 days, Curley et al. (1971) observed that PCB
7-3
-------
storage was related to the daily dosage. Concentrations of PCB in fat
after 240 days at dietary levels of 100 and 500 ppm were 1,101 and
10,021 ppm, respectively. They did not reach a point of equilibrium
storage.
When small amounts of PCBs are ingested over an extended period,
tissue concentrations may eventually reach toxic proportions. Allen
et al. (1974a) observed increasing tissue levels for periods in excess
of one year when rats were fed a diet containing 100 ppm Aroclor 1248.
Storage of PCBs in body tissues will persist after exposure is
terminated. Three years after accidental PCB poisoning, large amounts
of PCBs have been found in human adipose tissue (Kuratsune et_ jd. , 1971).
Savage jet al.. (1973a; 1973b) found that 20 percent of human milk
samples from Colorado mothers contained PCBs with levels ranging from
40 to 100 ppb. Musial ejt al. (1974) found PCBs in all of 15 milk
samples from Canadian mothers. Mean PCB concentrations were 20 ppb
for whole milk and 1.7 ppm for milk fat. Berglund (1972) reported a
mean concentration of 16 ppb in milk from Swedish mothers. These
concentrations may be compared with the levels (154-397 ng/g of whole
milk) Allen (1976) found in the milk of rhesus monkeys on PCB-contami-
nated diets. The nursing monkeys suffered a 50 percent mortality rate.
A nursing infant consumes about 150 g of milk per day per kg of
body weight during the second and third months of life (Berglund, 1972).
For a PCB level of 16 ppb, this would correspond to 2.4 ug/kg body
weight per day. For the range found in 20 percent of milk samples from
Colorado mothers, nursing infants would receive 6 to 15 ug/kg/day.
7-4
-------
These levels are much greater than the 1 pg per kg per day the U.S.
Food and Drug Administration reported as the mean adult.human intake
from the total diet (U.S. Department of Health, Education and Wel-
fare, 1972b).
2. Metabolism and Elimination
Metabolism of certain components of commercial PCB mix-
tures is indicated by the frequent observation in gas chromatographic
analyses that PCB samples extracted from animals are relatively deficient
in early peaks when compared to commercial PCB mixtures. The early
peaks are assumed to represent isomers with fewer chlorine atoms. This
depletion of lower retention time peaks from tissues of animals fed
standard PCB mixtures has been observed in monkeys (Allen et al., 1974a),
rats (Curley e£ al., 1971; Kiriyaoa et al., 1974; Grant et al., 197la),
and humans (Zitko and Choi, 1971).
Ease of metabolism seems to decrease with increasing chlorine
content. Certain structural features such as positional isomerisms
can also influence metabolic behavior. Peaks with retention times
corresponding to di-, tri-, and tetrachlorobiphenyls are metabolized
to the greatest extent. Mammals seem to be able to also metabolize
or excrete penta- and hexachloro isomers to some extent (Panel on
Hazardous Trace Substances, 1972) .
Hydroxylation appears to be the main metabolic route for chloro-
biphenyls. These hydroxylated metabolites may then be conjugated with
glucuronic acid and excreted into the urine.
7-5
-------
Rate of elimination of PCBs from body tissues seems to decrease
with an increase in the degree of chlorlnation. Highly chlorinated
biphenyls appear to be hardly eliminated at all since they are not
metabolized to excretable forms. The greatest percentage of PCB elimi-
nation from body tissues in humans probably occurs through biliary
excretion into the gastrointestinal tract (Allen et^ a^., 1974). Very
little PCB derived material is found in the urine (Burse et_ al_., 1974).
3. Lethality
Chronic exposure to PCBs poses a greater risk to the
*
general population than acute exposure. Acute LD in laboratory
mammals for a single dose of PCBs can vary from about 2 to 11 g per kg
body weight (Interdepartmental Task Force on PCBs, 1972; Grant and
Phillips, 1974). Even though toxicity of PCBs is greater whan exposures
are repeated or continuous, human exposures are unlikely to ever reach
lethal levels.
4. Toxic Effects on Liver
The major target organ of PCBs is the liver. Repeated
exposures of laboratory animals to low levels of PCBs results in enlarge-
ment of the liver accompanied by induction of liver microsomal enzyme
activity and structured changes in liver cells.
Among the most readily observed effects is an increase in weight
of the liver. Litterst et al. (1972) observed increased liver-to-body
*LD_n - that dose of a compound which will produce death in 50% of the
animals.
7-6
-------
weight ratios for rats fed diets containing 50 ppm PC^s for four weeks.
When rats are maintained on PCB-contaminated diets, liver enlargement
is progressive. Allen and Abramson (1973) fed rats diets containing
1000 ppm for six weeks. Increase in liver size became obvious in one
day and the liver became larger as the experiment progressed.
The increase in liver weight is due mainly to proliferation of
smooth surfaced membranes of the endoplasmic reticulum, a network of
interconnected channels present in the cytoplasm of most animal cells.
The endoplasmic reticulum contains complexes of enzymes which play a
role in metabolizing foreign substances such as drugs or environmental
pollutants. When liver samples are homogenized, the endoplasmic
reticulum is broken up into particles referred to as microsomes and
the enzymes are often referred to as microsomal enzymes. The liver
microsomal enzymes responsible for metabolizing substances not normally
expected to be present in the human body are sometimes referred to as
drug-metabolizing enzymes. Norback and Allen (1972) observed a pfo-
liferation of smooth endoplasmic reticulum (SER) in rats fed PCBs for
one to five weeks. Similar proliferation of SER has been observed in
mouse and monkey livers (Nishizumi, 1970).
The proliferation of SER is accompanied by increased activity of
the raicrosomal drug-metabolizing enzymes. Disruptions in normal enzyme
activity have been observed even at dose levels which had no effect on
liver weights or on liver-body weights ratios (Litterst et al., 1972).
Chronic exposure to low dose levels of PCB might exact an effect on enzyme
7-7
-------
activity in human liver (Allen and Abrahamson, 1973; Litterst et al. ,
1972; Chen jj: al., 1973). As long as exposure to PCBs continues, in-
duced levels of microsomal enzyme activity persist (Norback and Allen,
1972). After termination of exposure, the enzyme inducing effect of
PCBs appears to be reversible. Litterst and Van Loon (1974) maintained
rats on a diet of 50 ppm PCB for seven days. Discontinuation of PCB
treatment resulted in a slow decay of the induced enzyme activity to
approximately control levels after ten days.
Structural changes in liver cells, in addition to proliferated
smooth endoplasmic reticulum, include: increase in lipid droplets
within the cytoplasm, the formation of numerous multi-layered concen-
tric membrane arrays, increase in numbers of microbodies and lysosomes
within cells, and reduction of rough endoplasmic reticulum.
Numerous liver lesions can be observed after exposure to PCBs.
Dietary exposure (500 ppm) of Sherman rats to Aroclor 1254 for six to
eight months produced extensive areas of adenofibrosis in the liver.
Fatty metamorphosis, accumulation of brown pigment in hepatic macro-
phages,and Kupffer cells are also seen (Kimbrough et^ aJ^., 1972a;
Kimbrough e£ al., 1973).
5. Effects on Epithelium
Dermal effects were among the earliest recognized signs
of PCB toxicity. During the 1930s and 1940s, large outbreaks of
chloracne occurred In a number of factories manufacturing PCBs or PCB
products. Jones and Alden (1936) reported one of the first outbreaks
7-8
-------
of chloracne occurred in a company that was engaged in manufacture of
PCBs. Twenty-three workers had an acne-form eruption on the face and
body.
Dermal effects in humans were observed during an outbreak of PCB
poisoning in Japan during October 1968. This epidemic was caused by
accidentally contaminated rice-bran oil produced by a company in
Kitakyusha City and known as K rice oil. Analysis of some of the
recovered rice oil showed that it contained from 2000 to 3000 ppm of
Kanechlor 400 (Shea, 1973). Kanechlor 400 is a commercial mixture of
PCBs containing 48 percent chlorine. The disease became known as
"Yusho" or oil disease.
The subjective symptoms of Yusho patients are listed in Table 7.1.
Most of these symptoms relate to the skin and eyes (Kuratsune e^ al.,
1972). Dermatological symptoms included acneform eruptions with
marked enlargement and elevation of the follicular opening. Increased
pigmentation of the skin, lips, gums, nails, and mucous membrane of the
oral cavity were noted. Itching of the skin and "stiffening" of the
soles of the feet and the palms of the hands were also observed. Cysts
formed on the sebaceous glands in the genital region. Other symptoms
included dry skin, excessive perspiration, abnormally excessive
growth of hair, and swelling of Montgomery's gland in the breast.
Many patients were still suffering from the disease after three years
of treatment (Kuratsune et al., 1971 and 1972).
7-9
-------
Table 7.1 SYMPTOMS OF YUSHO PATIENTS (89 MALES, 100 FEMALES,
AS OF OCTOBER 31, 1968)
Symptoms
Blackening of nails
Distinctive hair follicles
Excessive sweating in palms
Acne like skin eruptions
Red spots on limbs
Itching
Change in skin color
Swelling of limbs
Stiffened soles of feet and palms of hands
Pigmentation of mucous membranes
Increased eye discharge
Hyperemia of mucous membranes in eyes
Temporary failing of eyesight
Jaundice
Swelling of upper eyelids
Sense of weakness
Numbness of Climbs
Fever
Hearing difficulty
Spasms of limbs
Headaches
Vomiting
Diarrhea
Males
%
83.1
64.0
50.6
87.6
20.2
42.7
75.3
20.2
24.7
56.2
88.8
70.8
56.2
11.2
71.9
58.4
32.6
16.9
18.0
7.9
30.3
23.6
19.1
Females
%
75.0
56.0
55.0
82.0
16.0
52.0
72.0
41.0
29.0
47.0
83.0
71.0
55.0
11.0
74.0
52.0
39.0
19.0
19.0
8.0
39.0
28.0
17.0
Source: Kuratsune, Yoshimura, Matsuzaka,and Yameguchi, 1971. Yusho,
a Poisoning Caused by Rice Oil Contaminated with Poly
Chlorinated BiphenyIs. HSMHA Health Reports 86(12); 1083-1091.
7-10
-------
The- epithelial UHSUOH of the r,;ir,t rolni est (rial tract, can be affected
by PCBs. Allen and Norback (1973) observed hyperplasia and dysplasia*
of the gastric mucosa in subhuman primates fed PCBs for three months.
JUJU
Edematous thickening of the stomach wall and marked hypertrophy of
the pyloric and fundic gastric mucosa occurred. The increased cellu-
larity of the mucous glands with invasion of the muscularis mucosae
and accompanying inflammation, observed by Allen and Norback (1973),
is histologically described as a hypertrophic gastritis. Ulceration
in the gastric mucosa of some animals developed after erosion of the
mucosal epithelium or rupture of large' muclnous cysts (Allen e_t al. ,
I973a). More recently, Allen (1975) observed dermatological symptoms
in adult female rhesus monkeys which had been fed diets containing
2.5 and 5.0 ppm of Aroclor 1248 for one year. These monkeys developed
swelling surrounding the socket of the eye, loss of hair, areas of
diffused redness over the skin, and acneform lesions. These symptoms
developed in one to two months at dosage levels equal to or half the
FDA temporary tolerance of 5.0 ppm (or fish or poultry fat (U.S.
Department of Health Education and WeJfare 1972a). The same
pattern of dermatologicai symptoms in rhesus monkeys fed low dosage
levels of PCBs were also observed by .IcNulty (I'm;, who added 3 ppm
of ArocLor 1242 to regular monkey chow. Studies by Ik'11 (1976) of
the effects of PCBs on gastroinLesLln.il epithelial cells of rhesus
monkeys were based on both light and electron i,,i c rosropic observations.
* Dysplasia - abnormal development of t issues.
**!lypertrophy - increase in si/e ui ,m organ or strurLure usually
not resulting from increase in number of cells.
7-J
-------
After ingestion of 100 ppm PCBs for two months, the parietal cells,
which secrete hydrochloric acid, and the zymogenic cells, which secrete
enzymes, had totally disappeared and had been replaced by mucous-
secreting cells. These effects also took place at dose levels of 3,
10, or 30 ppm, but more time elapsed prior to onset of symptoms with
the lower dose levels.
The experimental concentration of PCBs (3 ppm) within the entire
diet, which was sufficient to produce alteration of the gastrointestinal
epithelium in subhuman primates, is far below the levels that have
occurred in samples of milk fat (28 ppm) and of fish (35 ppm) (Kolbye,
1973). This concentration is much less than levels that occurred in
Japanese rice oil during the Yusho incident (2000 to 3000 ppm)
(Kuratsune £t al. , 1972).
The magnitude of the chronic effects which could be produced in
humans by PCB-induced hypertrophic gastritis has not been determined.
During the 1968 Japanese rice oil poisoning incident, some PCS poison-
ed victims developed visceral symptoms without the characteristic
dermatological symptoms. Symptoms shown by these patients included
nausea, vomiting, colic-like attacks of abdominal pain, and diarrhea
(Umeda, 1972). Gastrointestinal symptoms were also seen in more
typical "Yusho" cases. Kuratsune j^t al. (1971) reported vomiting in
26 percent and diarrhea in 18 percent of Yusho victims. These symptoms
in humans as well as weight loss in humans and animals exposed to PCBs
may be related to compromised gastric function. Allen and Norback
7-12
-------
(1973) suggest that the increased cellularity, abnormal dysplastic
growth pattern,and invasion of adjacent tissue regions might indicate
eventual neoplastic transformation.
6. Hematological Effects
Changes in blood composition and bone marrow histology
have been observed in humans and animals after exposure to PCBs.
Decreases in hemoglobin and in red blood cell count and increases
in leukocytes were observed in the most severe Yusho cases. Total
blood serum lipids increased, reflecting mainly an increase in trl-
clycerides. Total serum protein levels and levels of most electro-
lytes were normal (Kuratsune, 1972). Hirayama et al. (1974) found
that serum triglyceride concentration correlated with serum levels
of PCBs. In follow-up studies four years after the Yusho incident,
over half the Yusho patients still showed hypertriglyceridemia.
Allen et al. (1973b) found that hematological changes developed
gradually in rhesus monkeys fed diets containing 300 ppm Aroclor 1248,
Over three months, there was a decrease in hemoglobin of approximately
2 g/100 ml and a decrease in hematocrit from 40 to 33 percent. Allen
et al. (1974b) speculate that PCBs may have an inhibiting effect on
the red-blood-cell-forming tissue of bone marrow.
Hypobilirubinemia, a lowered concentration of serum bilirubin,
has been observed in PCS poisoning patients. The serum bilirubin
*Hematocrit - the volume of red blood cells packed by centrifuga-
tion in a given volume of blood; expressed as percentage of total
blood volume which consists of red blood cells.
7-13
-------
concentration in Yusho patients correlated inversely with blood levels
of polychlorinated biphenyls (Hirayama _e£ al. , 1974). It is suggested
that the hypobilirubinemia in PCB poisoning is caused through either
inhibition of heme catabolism or augmentation of bilirubin elimination.
It seems more likely, however, that the decrease in concentration of
serum bilirubin, a product of heme catabolism, is due to reduced
hemoglobin production in PCB-poisoned victims.
7. Nervous System Effects
Effects of PCBs on sensory nerves were observed during
the 1968 Yusho epidemic in Japan. No symptom was found suggesting in-
volvment of cerebrum, spinal cord,or cranial nerves. Spinal nerve
symptoms, however, were present (Murai and Kuroiwa, 1971).
Symptoms suggesting sensory nerve involvement were observed in
10 out of 21 (48 percent) of the Yusho patients studied by Murai and
Kuroiwa (1971). These symptoms included numbness, pain, hypoesthesia,
**
and areflexia . Slowing of peripheral sensory conduction velocity
was also observed in 48 percent of the cases, but motor nerve conduc-
tion velocity was not affected except in one case. No patient revealed
muscular atrophy or weakness. These data suggest the predominant
involvement of the sensory nerves in cases of PCB poisoning in humans
(Murai and Kuroiwa, 1971).
*Hypoesthesia - dulled sensitivity to touch.
**Areflexia - absence of reflexes.
7-14
-------
8. Carcinogenesis
No epidemiological studies indicating a relationship
between PCBs and cancer in humans are available. A limited number of
bioassays of PCBs have been performed and additional tests for carcin-
ogenicity are being conducted.
Kanechlor 500 and Aroclor 1254 have been shown to be carcinogenic
in mice when administered orally. Nagasaki et al. (1972) fed male dd*
mice diets containing Kanechlor 300, 400, or 500 for 32 weeks. Mice
fed the diet containing 500 ppm Kanechlor 500 developed hepatocellular
carcinomas. All types of Kanechlor produce benign liver tumors in
mice (Ho et al., 1974). Kimbrough and Linder (1974) fed mice diets
containing 300 ppm Aroclor 1254. After 11 months on the diet, 9 out
of 22 mice developed hepatomas. All 22 mice developed adenofibrosis.
PCBs appear to have some type of co-carcinogenic activity. Ito
et al. (1973) found that PCBs have a promoting effect on malignant
lesions induced in mouse livers by isomers of benzene hexachloride.
Among groups of mice fed benzene hexachloride alone, only the group
receiving 250 ppm of the a-isomer developed hepatocellular carcinoma.
When the diet was supplemented with PCBs, hepatocellular carcinoma
were induced in mice receiving only 50 or 100 Ppm ^-benzene hexachloride.
Diets containing p-benzene hexachloride caused hepatocellular carcino-
mas only when the diet was supplemented with PCBs.
*dd - a strain of mice.
7-15
-------
Many potential carcinogens require biotransformation by microsomal
enzyme ayatema , of the type induced by PCBs, before they become active
carcinogens. On the other hand, metabolism of primary or ultimate
carcinogens may reduce their carcinogenic properties (Popper et al.,
1973; Vaino, 1974).
9. Teratogenesis and Other Reproductive Effects
PCBs have been implicated as human teratogens. Terato-
genesis is the induction of structural or functional development
abnormalities by exogenous factors during gestation. Even a substance
that is regarded as safe for adults can be harmful to the fetus. The
fetus is characterized by an active proliferation of cells. A alight
influence on these proliferating cells can produce an abnormal develop-
ment of the fetal body.
During the Yusho incident in Japan during 1968, damage to new-
born infants was observed. Because of dark brown staining of skin
and mucous membranes, the infants were referred to as "cola-colored
babies." More than 10 cases were reported in the northern part of
Kyushu. Many other fetal symptoms were observed, such as stillbirth,
substandard birth weight, abnormal pigmentation, desquamation of skin
and mucous membranes, abnormally early appearance of teeth, calcifi-
cation of the skull, wide fontanelles, exophthalmos, and gingival
hypertrophy. Intrauterine growth suppression was confirmed in six
cases (Funatsu et al., 1972).
7-16
-------
Transfer of PCBs from mother to fetus through the placenta was
confirmed by demonstrating the presence of PCB in adipose tissue and
skin of the stillborn infant (Funatsu et al., 1972). The mothers
were sometimes free of clinical disease (Fraumeni, 1974).
In addition to their teratogenic effects, PCBs can cause abortions,
stillbirths, and reduced conception rates in laboratory animals. Doses
of 12.5, 25.0, and 50 mg/kg/day induced abortions and stillbirths in
rabbits when PCBs were administered orally to rabbits during the first
28 days of gestation (Villeneuve .et al., 1971). Kato _et al. (1972)
administered doses of 25, 75, 225, and 675 mg/kg/day to rats daily from
the first day of pregnancy. Incidence of stillbirth or abortion in-
creased as dose increased. Allen (1976) observed a decrease in live
births from adult female monkeys fed low levels of PCBs for six months
prior to mating. The rate of conception was only marginally reduced
by PCB exposure. In monkeys receiving 5 ppm of PCBs, six conceptions
occurred in eight monkeys. All eight monkeys receiving 2.5 ppm of PCBs
and all 12 control monkeys conceived. The rates of abortions and
resorptions were, however, greatly increased in monkeys receiving PCBs.
Four of the six conceptions in the mothers receiving 5 ppm and three of
the eight conceptions in mothers receiving 2.5 ppm resulted in abortions
or resorptions. These abortions tended to occur early in pregnancy
(during the first 45 to 60 days). No abortions or resorptions occurred
in control monkeys. All 12 control monkeys had live births. Five of
the eight monkeys receiving 2.5 ppm had live births. One live birth
7-17
-------
and one stillbirth occurred in monkeys receiving 5 ppm PCBs in the
diet. Allen (1976) observed a 50 percent mortality rate among the six
infants nursing from mothers who continued receiving PCBs in their
diet for three to four months after delivery. The infant deaths in-
cluded the one baby born to a mother receiving 5 ppm and two out of the
five babies born to mothers receiving 2.5 ppm. The mothers' milk con-
tained an average of 16.5 p.g per gram of milk fat (16.5 ppm). The FDA
temporary tolerance for PCBs in milk fat is 2.5 ppm (U.S. Department of
Health, Education and Welfare, 1972a).
C. OTHER ENVIRONMENTAL EFFECTS
!• Introduction
Since PCBs are persistent and ubiquitous in the environ-
ment, exposure of living things to PCBs is inevitable. Estimation
of the effects of PCBs on plants and animals is, however, difficult.
The types of adverse effects likely to result from PCB contamination
of the environment Include histological, biochemical, and reproductive
effects. These are all much harder to detect than gross pathological
lesions or mortality. Also, rarely is an organism in its natural en-
vironment exposed to a single pollutant. When toxic effects are noted,
it is therefore difficult to attribute them solely to PCBs.
2. Microorganisms
Some microbial degradation of PCBs may occur in the
environment. The ability of bacteria to degrade PCBs depends on the
degree of chlorinatlon and structure of the isomer. Mono-, di-, and
trichlorinated blphenyls and PCBs with one unsubstituted ring are
most readily degraded (Baxter e_t _al. , 1975; Ahmed and Focht, 1973).
7-18
-------
The effect of PCBs on microbial growth can vary. Differing sensi-
tivities of bacteria to PCBs may cause changes in bacterial community
populations. PCBs are inhibitory to phytoplankton* which have an
important role in the ecosystem as primary producers, nutrient cyclers,
and oxygen suppliers. Protozoa and plankton also play a role in accumu-
lation and translocation of PCBs to higher trophic levels.
3. Aquatic^ Organisms
Aquatic organisms accumulate PCBs from water, dietary
sources, and sediment. Bioconcentration tends to increase with higher
PCB levels in the water, with duration of exposure, with lipid con-
tent of organism, and with trophic level. Fish can accumulate 200,000
times more PCBs in their flesh than are found in the surrounding
water (Nebeker, 1975). The indirect toxicity of PCBs to predators
through accumulation of PCBs in tissues of food organisms may cause
deaths from water concentrations that do not cause direct lethality
(Nebeker, 1975).
PCBs are toxic to most aquatic organisms. Among aquatic invert-
ebrates, PCBs are toxic to shrimp, crab, snail, and aquatic insect
larvae. Molluscs can concentrate and transport PCBs within the marine
environment.
At high concentrations, PCBs can be lethal to fish. Juvenile
fish are more sensitive than adults (Stalling and Mayer, 1972). Less
chlorinated PCBs are more toxic to fish than more highly chlorinated
*Phytoplankton - photosynthetic organisma carried passively by water
currents. J J
7-19
-------
PCBs (Stalling and Mayer, 1972). At sublethal concentrations, PCBs
can cause anemia and metabolic alterations such as hyperglycemia and
increased thyroid activity. PCBs decrease activity of ATPases,
enzymes associated with osmoregulation (Yap £t al., 1971; Cutkomp £t
_al., 1972). Exposure of fish to PCBs can result in decreased egg
production, hatching, and survival of fry. Prolonged exposures to
very low levels may be more harmful to fish than acute exposures to
higher levels (Nebeker, 1975).
PCBs are toxic to freshwater organisms at concentrations below
5 ug/1 (ppb) (Nebeker, 1975). Newly hatched fish and small insects
and crustaceans with short life cycles are most sensitive. Aroclor
1254 at 0.45 ug/1 (ppb) produced a 50 percent decrease in midge re-
production; 1.3 ug/1 (ppb) caused a 50 percent reduction in Daphnia
reproduction; and 1.8 ug/1 (ppb) produced 50 percent reduction in
fathead minnow reproduction.
Low levels of Aroclor 1254 affect reproduction of sheepshead
minnows, an estuarine fish (Hansen, 1975). When eggs from adult fish
exposed to 0.1 ug/1 (ppb) for four weeks were fertilized and placed
in PCB-free water, survival of fry was diminished. Mortality of fry
increased as PCB content of the eggs increased (Hansen, 1975).
4. Birds
PCBs accumulate in body tissues of birds exposed to PCBs
in the environment. PCBs can be lethal to birds depending on level
*0smoregulation - the regulation of salt content of body fluids.
7-20
-------
of exposure, degree of chlorination, and sensitivity of the bird.
Sensitivity of the bird will depend on its age, sex, and species. The
major effects of chronic low-level exposure of birds to PCBs in the
environment would probably be reproductive effects. PCBs can exert
adverse effects on egg production (Cecil et al., 1973; Dahlgren et_ a^. ,
1972), embryonic development (Cecil et al., 1974), embryo viability
(Chang and Stokstad, 1975), and hatchability (Platanow and Reinhart,
1973; Llllle et al., 1974b) .
In birds, as in other vertebrates, the liver is the prime target
organ of PCBs. Increases in liver size (Lillie _et ad., 1974b; Hurst
et_ ajL. , 1974) are often accompanied by increased lipid'content of liver,
a decrease in storage of liver vitamin A (Cecil et_ al_. , 1973), and
increased hepatic microsomal enzyme activity (Bitman et al., 1972).
Porphyria (Strik , 1973a; Vos et al., 1971; Sinclair and Cranict,
1974) and chick edema disease (Rehfeld e± al., 1971; Vos, 1972) have
been attributed to PCBs.
5. Mammalian Wildlife
Where mammalian wildlife is exposed to sufficiently high
concentrations of PCBs, they would probably be affected adversely.
The significance of PCBs to wild animals depends primarily upon sub-
lethal physiological effects. Since PCBs would be expected to affect
growth, behavior, and reproduction, the likely consequences would be
Blow changes in reproduction rates, population sizes, and relative
abundance of species. The changes are subtle and difficult to detect.
7-21
-------
Fish-eating mammals, such as wetlands mammals, would be especially
likely to ingest large amounts of PCBs. Reproductive failure and
mortality have been reported in mink fed PCB-contaminaied fish from
the Great Laken. Kit mortality, reaching 80 percent, appeared to
depend on the percentage of coho salmon in the minks' diet as well as
the duration of feeding the salmon-containing diet. Mink rations
that contained other species of Great Lakes' fish caused similar re-
productive complications, but to a lesser degree (Aulerich ^Q of mink for deaths
within 48 hours after a single intraperitoneal injection with Aroclors
1221, 1242, and 1254 are 250 to 500 ppm, 500 to 1000 ppm, and 1000
ppm, respectively. The PCB toxicity to mink varied inversely with
chlorine content (Aulerich jet al_. , 1973). Feeding mink on diets that
contained 10 ppm PCB depressed growth significantly. AIL mink fed
PCB-fiupplemented diets failed to produce offspring. Mink fed diets
supplemented with 10 or 30 ppm failed to produce offspring (Ringer
_et ail. , 1972; Aulerich et_ aJL , 1973).
7-22
-------
Platonow and Karstad (1973) investigated the effects of chronic
feeding of low levels of PCBs in mink. No live kits were produced
and all adult mink died during a 105-day period of feeding a ration
containing 3.6 ppm. At a level of 0.6 PPm, one of 12 mink produced
three kits, all of which died during the first day after birth.
6- _Role of Trace Contaminants In Toxlcity of Commercial PCS
Mixtures
The likehood of PCB contamination by toxic impurities has
been proposed by many Investigators (Voss jet al., 1970; Kimbrough,
1972; McNulty, 1975). In 1970, Vos et a]_. reported that Phenoclor,
manufactured by Prodelec in France, and Clophen, manufactured by Bayer
in Germany, produced a much higher incidence of edema and liver
necrosis in cockerels than did Aroclor from Monsanto. Analysis re-
vealed the presence of polar inpurities (hexa- and heptachloronaphtha-
lenes and tetra- and pentachlorodibenzofurans) in the two European
PCB mixtures (Vos et al., 1970). Since chlorinated naphthalenes were
shown to be far less toxic than chlorinated dibenzofurans, the incre-
ment in toxicity of European PCB mixtures over Aroclors was attributed
to chlorinated dibenzofuran contamination (Vos et_ al. , 1970). Until
recently, no contaminants had been reported in American commercial
PCB mixtures (Aroclors). However, Bowes _et al. (1975) found tetra-,
penta-, and hexachlorinated dibenzofurans in Clophen A60, Phenoclor
DP6, and Aroclors 1248, 1254, and 1260. No contaminants were
found in Aroclor 1016. The conclusions of Vos _et _al. (1970) were
7-23
-------
supported by the finding that the Clophen and the Phenoclor, respec-
tively, contained 11 and 17 times the dibenzofuran concentration found
in Aroclor 1260.
Because the dibenzofurans are structurally as well as toxicologi-
cally similar to the dibenzodioxins, the effects of the compounds have
often been compared. Both classes of chlorinated compounds display
high levels of toxicity when three to six chlorine atoms are present,
causing chloracne, liver damage, and chick edema disease (Kimbrough,
1972). These symptoms are all also characteristic of PCB toxicity.
McNulty (1976) reported that 2,3,7,8-tetrachlorodibenzo-p-dioxin
caused exactly the same constellation of pathological changes in
rhesus monkeys as did PCBs. Since contaminants are present in PCB mix-
tures at very low levels, if at all, they are not likely to account
for the observed toxicity of PCBs. Aroclor 1248 contained 2 ppm
dibenzofuran and Aroclor 1260 contained 0.8 ppm (Bowes e± al_. , 1975).
Chlorinated dibenzodioxins have never been reported present in
commercial PCB mixtures. Experimental data of McNulty (1975) have
indicated that 2,3,7,8-tetrachlorodibenzo-p-dioxin is about 10,000
times as toxic as Aroclor 1242 on a per gram basis. Tetrachlorodi-
benzofurans are about an order of magnitude less toxic than tetrachloro-
dibenzodioxins (Kimbrough, 1972). If Aroclors are contaminated at
levels from 0.8 to 2.0 ppm, there could not be enough dibenzofuran
present to account for the effects of the Aroclor.
7-24
-------
VIII. REFERENCES
Ahmed, M.t and D.D. Focht. 1973. "Degradation of Polychlorinated
Biphenyls by Two Species of Achromobacter." Canad. J. Micro 19: 47-52.
Allen, J.R., and L.J. Abrahamson. 1973. "Morphological and Biochemical
Changes in the Livers of Rats Fed Polychlorinated Biphenyls." Arch.
Environ. Cont. and Tox. 1: 265-280.
Allen, .J.R., and D.H. Norback. 1973. "Polyrhlorinated Biphenyl- and
Trlphei.yJ-lnduced Gastric Mucosa] Hyporplasia in Primates." Science
179 (4072): 498-499.
Allen, J. R., L.J. Abrahamson, and D.H. Norback. J973a. "Biological
Changes in Primates Produced by a Polychlorinated Biphenyl and Triphenyl "
Federal Proceedings 32(3 Part 1) : 235.
Allen, J.R., L.J. Abrahamson, and D.H. Norback. 1973b. "Biological
Effects of Polychlorinated Biphenyls and Triphenyls on the Subhuman
Primate." Environmental Research 6(3): 344-354.
Allen, J.R. , L.A. Carstens, and D.A. Barsotti. 1974a. "Residual
Effects of Short-term, Low-level Exposure of Nohnhuman Primates to
Polychlorinated Biphenyls." Toxicology and Applied Pharmacology
30(3): 440-451. 1UL
Allen, J.R., D.I!. Norback, and I.C. Hsu. 1974b. "Tissue Modifications
in Monkeys as Related to Absorption, Distribution, and Excretion of
Polychlorinated Biphenyls." Archives of Environmental Contamination
and Toxicology 2(1)- 86-95. " —————
Allen, J.R. 1975. Response of the Nonhuman Primate to Polychlorinated
Biphenyl Exposure. Federation Proceedings 34(3): 1675-1679.
Allen, J.R. 1976. Personal Communication.
American Conference of Governmental Industrial Hygienists. 1964.
"Threshold Limit Values for 1964." AMAArch. Environ. Health 9:' 545.
American Conference of Governmental IndustiraL Hygienists. 1973.
"Threshold Limit Values for Chemical Substances and Physical Agents
in the Workroom Environment." ACG111, Cincinnati, Ohio.
American Industrial Hygl i>m> Assodnt Ion. 1965. "Gh 1 orodiphenyls. "
Hygiene Guide Series, Jan.-Feb.
8-1
-------
American National Standards Insitute. 1974. "Guidelines for
Handling and Disposal of Capacitor-and Transformer-Grade Askarels
Containing Polychlorinated Biphenyls." ANSI, C107.1-1974. New York.
Anonymous. 1971. "Pollution Cop's Lot Not 'Appy." Chemical Week,
Apr. 21: 19-20.
Anonymous. 1973. "OECD Recommends Restricted Use of Polychlorinated
Biphenyls." European Chemical News - March 2: 21.
Anonymous. 1975a. "A fresh fight over PCB pollution." Business Week,
pp. 30-31, October 6, 1975.
Anonymous, 1975b. "EPA's Schweitzer Claims PCB Use May Be Broader Than
Previously Thought," Toxic Materials News 2: 137.
Anonymous. 1975c. "Chementator..." Chemical Engineering, July.
Anonymous. 1975d. "Chemicals Found in Lake Fish: State PCB Ban Urged."
Michigan Out—of-Doors, July.
Anonymous. 1975e. "PCB Water Maximum of I PPT is 'An Appropriate Goal',
EPA Says." Pesticide Chemical News. September 3- pages 12-13.
Armour, J.A. 1973. "Quantitative Perchlorination of Polychlorinated
Biphenyls as a Method for Confirmatory Residue Measurement and Identifi-
cation." Journal of the Associationof Official Analytical Chemists
56(4): 987-993.
Armour, J.A., and J.E. Burke. 1970. "Method for Separating Polychlo-
rinated Biphenyls from DDT and Its Analogs" . Journal of the Association
of Official Analytical Chemists 53(4): 761-768.
Arnts, R.R. 1975. U.S. EPA. Personal communication.
Aulerich, R.J., R.K. Ringer, and S. Iwamoto. 1973. "Reproductive
Failure and Mortality in Mink Fed on Great Lakes Fish." Journal of
Reproduction and Fertility 19 (Suppl.): 365-376.
Bachli, W. 1975. General Electric Co., Pittsfield, Massachusetts.
Personal communication.
Bailey, S., P.J. Bunyan, and F.B. Fishwick. 1970. "Polychlorinated
BiphenyL Residues." Chemistry and Industry 22: 705.
Baird, R. 1975. Los Angeles County Sanitation District, Los Angeles,
California. Personal, communication.
8-2
-------
v L ' ;«;' D'5- Gilbert' R-A- Lidgett, J.H. Malnprize, and H.A.
Vodden. 1975. "The Degradation of Polychlorinated Biphenyls by
Micro-Organisms. " The Science of the Total Environment 4m; 53-61.
r ' "Ultrastructural Features of Gastric Mucosa and Sebaceous
Glands After Ingestion of Aroclor 1242 by Rhesus Monkeys." Presented at
November ll-lT^* °* P°lychl°rinated BiP^nyls, Chicago, Illinois,
Berglund, F. 1972. "Levels of Polychlorinated Biphenyls in Foods in
eden- nvironnienalHe-- Fvp ]: 67-71.
Hevenue, A., J.N. Ogata, and J.W. Hylin. J972. "Organochlorlne
^estlcides In Rainwater, Oahu, Hawaii, J97J-J97?." Bulletin of
Environmental Contamination o_f_ToxlrpJnj£v 8(4): 238-241. '
Bidleman, T.F., and C.E. Olney. J974a. "Chlorinated Hydrocarbons in
the Sargasso Sea Atmosphere and Surface Water." Science 183(4124):
516-518. 2nd Copy. ""
Bidleman, T.F., and C.E. Olney. 1974b. "High-Volume Collection of
Atmospheric Polychlorinated Biphenyls." Bulletin of Environmental
Contamination and Toxicology 11(5): 442-"45oT ~~~ '
Biros, F.J., A.C. Walker, and A. Medbery. 1970. "Polychlorinated
Biphenyls in Hunan Adipose Tissue." .Bulletin of Environmental
Contamination and Toxicology 5(4): 317^323^"
Bitman, J., H.C. Cecil, and S.J. Harris. 1972. "Biological Effects
of Polychlorinated Biphenyls in Kals and Quail." Environmental
Health Perspectives Expi 1: 145-149. ~"
Blaszevich, J. 1975. EPA, Seattle, Washington. Personal communica-
tion.
Bowes, G.W., M J Mulvihill, B.R.T. Simoneit, A.L. Burlingame, and R.W.
Risebrough. 1975 "Identification of Chlorinated Dibenzofurans in American
Polychlorinated Biphenlys." Nature 256 (7/24): 305-307.
Burse, V.W., R.D. Kimborgouh, E.G. ViJlanueva, R.W. Jennings R E
Linder and G.W. Sovocool. 1974. "Polychlorinated Biphenyls. 'storage,
Distribution, Excretion, and Recover: Liver Morphology After Prolonged
Dietary Ingestion." Archives__of_ Knvironmenta 1 Health 29(6)' 301-3^7
Cnrc-y, A.E., G.B. Wiersma, H.Tal, ,uid W.G. Mitt-hell. 1973. "Orsano-
ch,,rJne Pesticide Residues in Soil and Crops of the Corn Belt Region,
United States-1970." J]^8jU1cJ:des_MonltorJ.n£ Journal 6(4): 369-376
2nd Copy. " ~"
3-'J
-------
Carnes, R.A. 1975. EPA, Cincinnati, Ohio. Personal communication.
Games, R.A. , J.U. Doerger, and H.L. Sparks. 1973. "Polychlorinated
Biphenyls in Solid Waste and Solid-Waste-Related Materials." Archives
of Environmental Contamination and Toxicology 1(1): 27-35.
Cecil, H.C., S.J. Harris, J. Hitman, and G.F. Fries. 1973. "Poly-
chlorinated Biphenyl-Induced Decrease in Liver Vitamin A in Japanese
Quail and Rats." Bulletin of Environmental Contamination and Toxicology
9(3): 179-185.
Cecil, H.C., J. Hitman, R.J. Lillie, G.F. Fries, and J. Verrett. 1974.
"Embryotoxic and Teratogenic Effects in Unhatched Fertile Eg;gs from
Hens Fed Polychlorinated Biphenyls (PCBs)." Bulletin of Environmental
Contamination and Toxicology 11(6): 489-A95.
Chang, E.S., and E.L.R. Stokstad. 1975. "Effect of Chlorinated Hydro-
carbons on Shell Gland Carbonic Anhydrase and Egg Shell Thickness in
Japanese Quail." Poultry Science 54: 3-10.
Chen, P.R., H.M. Mehendale, and L. Fishbein. 1973. "Effect: of Two
Isomeric Tetrachlorobiphenyls on Rats and Their Hepatic Enzymes."
Archives of Environmental Contamination and Toxicology 1(1): 36-47.
Choi, P.S.K., H. Nack, and J.E. Flinn. 1974. "Distribution of Poly-
chlorinated Biphenyls in an Aerated Biological Oxidation Wastewater
Treatment System." Bulletin of Environmental Contamination and
Toxicology 11(1): 12-17.
Compton, B., and J. BJorkland. 1972. "Desien of a High Volume Samples
for Airborne Pesticide Collection." (021-Pesticide Chem.) 163rd Meeting,
Division of Pesticide Chemistry, ACS. Boston, Massachusetts.
Courchaine, R.J. 1975. Dear Sir letter to Michigan industries.
Cowherd, C.,Jr., M. Marcus, C.M. Guenther, and J.L. Spigarelli, 1975.
"Hazardous Emission Characterization of Utility Boilers." EPA-650/2-75-
066.
Crosby, D.G., and K.W. Moilanen. 1973. "Photodecomposition of Chlorin-
ated Biphenyls and Dibenzofurans." Bulletin of Environmental Contamin-
ation and Toxicology 10: 372-377.
Crosby, D.G., K.W. MoiJanen, and A.S. Wong. 1973. "Environmental
Generation and Degradation of Dibenzodioxins and Dibenzofurans."
Environmental Health Perspectives Exp. 5: 259-266.
8-4
-------
Crump-Wiesner, H.J., H.R. Feltz, and M.L. Yates. 1974. "Pesticides
in Water: A Study of the Distribution of Polychlorinated Biphenyls in
the Aquatic Environment." Pesticides Monitoring Journal 8(3): 157-161
Curley, A., V.W. Burse, and M.E. Grim. 1973. "Polychlorinated Bi-
phenyls: Evidence of Transplacental Passage in the Sherman Rat "
Food and Cosmetics Toxicology 11: 471-476.
Cutkomp, L.K., H.H. Yap, D. Desaiah, and R.B. Koch. 1972 "The Sen-
sitivity of Fish ATPases to Polychlorinated Biphenyls." Environmental
Health Perspectives Exp . 1: 165-168. - ~~
Dahlgren, R.B., R.J. Bury, R.L. Linder, and R.F. Reidinger, Jr. 1972
Residue Levels and HistopathoJogy in Pheasants Given Polychlorinated'
Biphenyls. Journal of Wi 1 dl i f o Maiiagen^nt- 36(2): 524-533.
Dube, D.J., G.D. Velth, and G.F. Lee. 1974. "Polychlorinated Biphenyls
" Control Federa-
Duce R A., J G Quinn, C.E. Olney, S.R. Piotrowicz, B.J. Ray, and
T.L. Wade. 1972. Enrichment of Heavy Metals and Organic Compounds
176(403l" °layer ^ Narragansett Bay' Rhode Island." Science
J'' Land A'J- W "A
r «/ ' < ' nae
(Aroclor 1254) in the Water, Sediment, and Biota of Escambia
l*y> F1°^da; Bulletin of Envirqnmental_Contamination and Toxicolog
j{£) . 1/1— 180. ~~ • — ~~ — *"*
Durfee, R.L. 1975. "Production and Usage of PCBs in the United States."
laper Presented at the National Conference on Polychlorinated Biphenyls
Nov. 19-21, Chicago.
Edwards, R. J971. "Polychlorinated Biphenyls, Their Occurrence and
^?U o /o?°e: A Revlew'" Chejmlstri^njlJndustrx (No Volume) Issue 47:
1J40-8 (21 November 71)
Eichelberger, J. 1971. "Persistence of Aroclors in Surface Waters."
PCB Newsletter 3: 8-9. U.S. Environmental Protection Agency; National
Water Quality Laboratory, Duluth, Minnesota.
Enos H.F., ]976. Personal communication. U.S. Environmental Pro-
tection Agency.
Enos, H.F., J.F. Thompson, J,B. Mann, and R.F. Mosemon. April 3972
Determination of Pesticide Residues in Air" (022-Pesticide Chem )
Division of Pesticide Chemistry, 163rd Meeting, ACS, Boston, Massachu-
setts.
8-5
-------
Fvana, T. 1975. U.S. Army, Washington, D.C. Personal communication.
Fishbein, L. 1974. Chromatography of Environmental Hazards. Volume
2-Metals, Gaseous and Industrial Pollutants, Chapter 24-"Polychlorina-
ted Biphenyls." 529-577. North-Holland, American Elsevier.
Fraumeni, J.R. 1974. "Chemicals in Human Teratogenesis and Trans-
placental Carcinogenesis." Pediatrics 53(6): 807-812.
Funatsu, H., F. Hamashita, Y. Ito, S. Tsugawa, T. Funatsu, T. Yoshikane,
and M. Hayashi. 1972. "Polychlorobiphenyl (PCB) Induced Fetopathy.
I. Clinical Observation." The Kurume Medical Journal 19(1): 43-51.
Gaffney, P.E. 1974. "PCBs: Another Source?" Science 183 (4123):
367-368.
Garrison, A.W., L.H. Keith and A.L. Alford 1972. "Confirmation of
Pesticide Residues by Mass Spectrometry and NMR Techniques." in Fate
of Organic Pesticides in the Aquatic Environment: 26-54 Advances in
Chemistry Series til. American Chemical Society, Washington D.C.
General Motors Corporation. 1974. Energy Management Bulletin EM 74-6.
November 27.
Goff, G. 1975. U.S. Army, Alexandria, Virginia. Personal communication.
Grant, D.L., W.E.J. Phillips, and D.C. Villeneuve. 1971, "Metabolism
of a Polychlorinated Biphenyl (Aroclor 1254) Mixture in the Rat."
Bulletin of Environmental Contamination and Toxicology 6(2): 102-112.
Grant, D.L., and W.E.J. Phillips. 1974. "The Effect of Age and Sex
on the Toxicity of Aroclor 1254, a Polychlorinated Biphenyl,in the
Rat." Bulletin of Environmental Contamination and Toxicology 12(2):
145-152.
Hreichus, Y.A., A. Greichus, and R.J. Emerick. 1973. "Insecticides,
Polychlorinated Biphenyls and Mercury in Wild Cormorants, Pelicans,
Their Eggs, Food and Environment." Bulletin of Environmental Contami-
nation of Toxicology 9(6): 321-328. 2nd Copy. Original inhouse.
Gustafson, C.G. 1970. "PCBs - Prevalent and Persistent." Environ-
mental Science and Technology 4(10): 814-819.
Hall, R.H. 1976. Dow Chemical Company, Midland, Michigan.
Personal communication.
Hall, R.H.,-and D.H. Haigh. 1974. "Imbiber Beads that Soak Up Oil Spills."
The Mangineer 3(2): 5-8.
8-6
-------
Or» ?' »'p CB6; EffeCtS °n and Accumulation by Estuarine
Organisms. Presented at National Conference on Polychlorinated
BiphenyJs, Chicago, Illinois. November 19-21. ^rinated
Haque R and D.W. Schmedding. 1975. "A Method of Measuring the
Water Solubility of Hydrophobia Chemicals: Solubility of Five Poly-
"
^
^^ment^L^^ 139-142?
<-
organisms.
]__of_MfLrlne_ Research 32 (2) : 103-118
Stelnhauer Reply.
p^ «eaaureBent." in
Edition, Academic Press: New York and London.
et and Sources of
on Polychlorinated Bipheny]" PCBI n th F / National Conference
Illinois, November 19-2?? Environment - 1975. Chicago,
hpo " '
Biphenyls Poisoning."
^
'-aboratory. R.s ^ 17 '^ ' ^^^^ and Support
1220-1221.
-------
Holsinger, J. 1975. U.S. Army, Fort Lee, Virginia. Personal communication.
Horn, W., Riebrough, A. Soutar, and D.R. Young. 1974. "Deposi-
tion of DDE and Polychlorinated Biphenyls in Dated Sediments of the
Santa Barbara Basin." Science 184(4142): 1197-1199.
Hubbard, H.L. 1964. "Chlorinated Biphenyl and Related Compounds."
Encylcopedia of Chemical Technology, 2nd Edition 5: 289-297.
Hurst, J.G., W.S. Newcomer, and J.A. Morrison. 1974. "Some Effects
of DDT, Toxaphene and Polychlorinated Biphenyl on Thyroid Function in
Bobwhite Quail." Poultry Science 53(1): 125-133.
Hutzinger, 0., and S. Safe. 1972. "A General Method for the Prepara-
tion of Tritiated Polychlorobiphenyls of High Specific Activity: 2,2' ,
5,5'-Tetrachlorobiphenyl-[3R]D and 2,2', 4,4', 5,5'-Hexachlorobiphenyl-
[3H]1." Bulletin of Environmental Contamination and Toxicology 7(6);
374-375.
Hutzinger, O.S. Safe, B.R. Wetzell, and V. Zitko. 1973. "Photo-
chemical Degradation of Di- and Octachlorodibenzofuran." Environmental
Health Perspectives Exp. 5: 267-270.
Hutzinger, 0., W.D. Jamieson, S. Safe, L. Paulman, and R. Ammon. 1974a.
"Identification of Metabolic Dechlorination of Highly Chlorinated
Biphenyl in Rabbit." Nature 252(12/20-27): 698-699.
Hutzinger, 0., S. Safe, and V. Zikto. 1974. "The Chemistry of PCBs."
CRC Press, Cleveland, Ohio.
Institute of Electrical and Electronics Engineers. 1974. "IEEE Guide
for Acceptance and Maintenance of Transformer Askarel in Equipment."
IEEE Standard 76-1974. IEEE, New York.
Interdepartmental Task Force on PCBs. 1972. "Polychlorinated Bi-
phenyls and the Environment." COM-72-10419. 1-192. National Technical
Information Service, Springfield, Virginia (U.S. Dept. of Agriculture,
Commerce, Health-Education and Welfare, EPA, and other agencies.
Ito, N., H. Nagasaki, M. Arai, S. Makura, S. Sugihara, and K. Kirao.
1973. "Histopathologic Studies on Liver TumorIgenesis Induced in
Mice by Technical Polychlorinated Biphenyls and Its Promoting Effect
on Liver Tumors Induced by B«nzene Hexachloride." Journal of the
National Cancer Institute 51(5): 1637-1646.
Ito, N. , 11. Nagasaki, J. Makuira, and M. Arai. 1974. "Histopathological
Studies on Liver Tumorigenesis in Rats Treated with Polychlorinated
Biphenyls." Gann 65: 545-549.
8-8
-------
Jensen, S. 1972. "The PCB Story." AM BIO 1 (4) : ]23-131.
Jensen, S., L. Renherg, and M. Ol.s.son. 1972. "PCB Contamination from
Boat Bottom Paint and LcveJ s of PCB in Plankton Outside a Polluted
Area. Na t ur e 240: 358-360.
Jones, J.W., and H.S. Alden. 1936. "An Acnef orm Dermategosis. "
Archives of Dermatology and SypjiiTogy 33: 1022-1034.
M ' A' Arima' K" Takahashi, M. Shimomura,
. Mxyahara, M. Adachi, Y. Tashiro, and M. Matsumoto. 1972 "Poly-
chlorobiphenyls (PCB) Induced Fetopathy. II. Eoerimental Studies:
Possible Placental Transfer of Polychlorobiphenyls in Rats." The
Kurume Medical Journal 19(1): 53-59. -
Kimbrough, R.D. 1972. "Toxicity of Chlorinated Hydrocarbons and Related
Compounds. Archives of Environmental jjgajj-h ?Wa)- 125-131.
Kimbrough, R.D., R.E. Linder, and T.B. Gaines. 1972. "Morphological
Changes in Livers of Rats Fed Polychlorinated Biphenyls." Archives of
Environmental Health 25(5): 354-364. -
Kimbrough, R.D., R.E. Linder, V.W. Burse, and R.W. Jennings. 1973
*" ^ ^ Llver'" Archives of Environmental Health
Kimbrough, R.D., and R.E. Linder. 1974. "The Induction of Adenofi-
brosis and Hepatomas of the Liver in Mice of the BALC/cj Strain by
Polychlorinated Biphenyls (Aroclor 1254)." Journal of the National
C anc er Ins t i t ute . " ' ~ -
Kiriyama, S., M. Banjo, and H. Matsushima. 1974. "A Failure to "
Prevent Absorption of Polychlorinated Biphenyls by Concurrent Feeding
aC Mann " International 10(3):
Klapp, F.J. 1975. U.S. Army, Fort McPherson, Georgia. Personal
communication .
KleJiu-rt, S. 1975 Wisconsin Dept. of National Resources. Personal
communication,
Kolbyc, A.C. 1972. "Food Exposures to 1'olych] orinated Biphenyls "
.knyironmental Health Ptfrsj)ec_tj.ves Exp. 1: 85-88.
Koneva], L. 1975. Rollins Environmental Services, Bridgeport New
Jersey. Personal communication.
8-9
-------
Kopp, T. 1975. U.S. Environmental Protection Agency. Personal communi-
cation.
Kuratsune, M., T. Yoshimura, J. Matsuzaka, and A. Yamaguchi. 1971,
"Yusho, a Poisoning Caused by Rice Oil Contaminated with Polychlori-
nated Biphenyls." HSMHA Health Reports 86(12): 1083-1091.
Kurantsune, M.T. Yoshimura, J. Matsuzaka, and A. Yamaguchi. 1972.
"Epidemiologic Study on Yusho, a Poisoning Caused by Ingestion of
Rice Oil Contaminated with a Commercial Brand of Polychlorinated
Biphenyls." Environmental Health Perspectives Exp 1: 119-128.
Kutz, ^.W.^and H.S.C. Yang. 1975. "A Note on Polychlorinated Biphenyls
in Air." Presented at The National Conference of Polychlorinated
Biphenyls, Nov. 19-21, 1975, Chicago, Illinois.
Lafornara, J. 1975. EPA, Edison, New Jersey. Personal communication.
Langston, P.D. 1975. Calgon Corporation, Pittsburgh, Pennsylvania.
Personal communication.
Law, L.M., and D.F. Goerlitz. 1974. "Selected Chlorinated Hydrocarbons
in Bottom Material from Streams Tributary to San Francisco Bay."
Pesticides Monitoring Journal 8(1): 33-36.
Lear, C.W., W.F. Krieve, and E. Cohen. 1975. "Charged Droplet Scrub-
bing for Fine Particle Control." Journal of the Air Pollution Control
Association 25: 184-189.
Leighton, I.W., and J.B. Feldman. 1975. "Demonstration Test Burn of
DDT in General Electric1s Liquid Injection Incineration." USEPA,
Region I, Boston, Massachusetts.
Levitz, N, W. Podolsk!* and E. Carls. Undated. "Control Technology for
Polychlorinatef) Biphenyls (PCBs)." U.S. Environmental Protection Agency,
unpublished contract report.
Lewis, R.V., 1976. Virginia Electric and Power Company, Richmond,
Virginia. Personal communication.
Llllie.l.J., S.J. Harris, H.C. Cecil, and J. Bitman. 1974. "Normal
Reproductive Performance of Mature Cockerels Fed Aroclor 1248."
Poultry Science 53(4): 1604-1607.
Lister, R.E., and N.J.M. Bennett. 1972. "PCBs in Copying Paper."
Nature 237(6/16): 414.
8-10
-------
Litterst, C.L, T.M. Farber, A.M. Haker, and E.J. van Loon. 1972.
'Effect of Polychlorinated Biphcnyls on Hepatic Microsomal Enzymes in
tho Kat-" Toxicology and A^pjJjL^JMinjrmacoj^ogv 23(1): 112-122.
Litterst, C.L., and E.J. van Loon. 1974. "Time-Course of Induction
of Microsomal Enzymes Following Treatment with Polychlorinated Biphenyl . "
Bulletin of Environmental Contamination and Toxicology 11(3); 206-212.
Lokey, D. 1975. U.S. Environmental Protection Agency. Personal communi-
cation.
Longhurst, A.R.,and P.J. Radford. 1975. "PCB Concentration in North
Atlantic Surface Water." Na.ture 256: 239. **
Louros, B. 1975. Calgon Corporation, Pittsburgh, Pennsylvania.
Personal communication.
Lunde, G., J. Gether, and B. Josef sson. 1975. "The Sum of Chlorinated
and of Brominated Non-Polar Hydrocarbons in Water." Bulletin of
Environmental Contamination and Toxi_coT_o_gv 13(6): 656-661.
Mackay, D. , and A.W. Wolkoff. J973. "Rate of Evaporation of Low-
Solubility Contaminants from Water Bodies to Atmosphere." Environ-
mental Science and Technology 7(7): 611-614.
Kartell, J.M. , D.A. Rickert, and F.R. Siegel . 1975. "PCBs in Suburban
Watershed, Reston, Virginia." Environmental Science and Technoloey
9: 872-875. ~~ ~~ - *"•
Mason, T. 1975. "Local Chemists Find Safeguards, Replacements For
Contaminant." Midland Daily News, June.
Maugh, T.H..H. 1975. "DDT: An Unrecognized Source of Polychlorinated
Biphenyls." Science 180 (4086): 578-579.
McNulty, W.P. 1975. Paper presented at National Conference on
Polychlorinated Biphenyls, Chicago, Illinois, November 19-21.
McStrack, D. 1975. Mcf-raw- Edison. Personal communication.
Monsanto Industrial Chemical Company. Undated. "Aroclor Plasticizers "
Monsanto Technica L H'lllgtjjl 0/PL-306A.
Monsanto Industrial Chemicals Company. 1970. "Monsanto Cites Actions
Taken on Environmental Issue." (News release - E.V. John, Public
Rel at ions Dept) .
8-1
-------
Monsanto Industrial Chemicals Company. 1972. "Presentation to the
Interdepartmental Task Force on PCBs." Washington, B.C.
Monsanto Industrial Chemical Company. 1973a. "Polychlorinated Bi-
phenyl (PCB)." A Presentation to the Effluent Standards and Water
Quality Information Advisory Committee, November 29, 1973.
Monsanto Industrial Chemical Company. 1973b. Letter to Dr. Martha
Sager, July 13.
Monsanto Industrial Chemicals Company. 1974a. "PCB Manufacture and
Salea-Monsanto Industrial Chemicals Company - 1957 Thru 1964." (un-
published data).
Monsanto Industrial Chemicals Company. 1974b. "PCB Manufacture and
Sales-Monsanto Industrial Chemicals Company - 1965 Thru 1974." (un-
published data).
Monsanto Industrial Chemicals Company. 1974c. "Proposed Toxic Pol-
lutant Effluent Standards for Aldrin-Dieldrin, Et Al-FWPCA (307) -
Docket No. 1." Sworn statement by W.B. Papageorge. St. Louis,
Missouri, 14 March 1974.
Monsanto Industrial Chemical Company. 1975a. "A Comparison of
Proposed Specifications for Aroclor 1016 with Current Specifications
for Aroclor 1242." (unpublished data).
Monsanto Industrial Chemical Company. 1975b. "Transformer Askarel
Inspection and Maintenance Guide." Bulletin IC/FF-38R.
Murai, Y., and Y. Kuroiwa. 1971. "Peripheral Neuropathy in Chloro-
biphenyl Poisoning." Neurology 21(11): 1173-1176.
Musial, C.J., 0. Hutzinger, V. , ZItko, and J. Crocker. 1974. "Presence
of PCB, DDE, and DDT in Human Milk in the Provinces of New Brunswick and
Nova Scotia, Canada." Bulletin of Environmental Contamination and
Toxicology 12(3): 258-267.
Negasaki, H., S. Tomii, T. Mega, M. Marugami, and I. Nobuyuki. 1972.
"Hepatocarcinogenicity of Polychlorinated Biphenyls in Mice." GAtTO
63(6): 805.
Nebekev, A.V. L975."Summary of Recent Information Regarding Effects
of PCBs on Freshwater Organisms." Presented at National Conference
on Polychlorinated BIphenyla, Chicago, Illinois, November 19-21.
Nisbet, C.T., and A.F. Sarofim. 1972. "Rates and Routes of Transport
of PCBs in the Environment." Environmental Health Perspectives Exp 1:
21-38.
8-12
-------
Nishizumi, M. 1970. "Light and Electron Microscope Study of Chloro-
biphenyl Poisoning— in Mouse and Monkey Liver." Archives of Environ-
mental Health 21(111 : 620-632. " -
Norback D.H., and J.R. Allen. 1972. "Chlorinated Aromatic Hydro-
carbon Induced Modifications of the Hepatic Endoplasmic Retidulum:
Concentric Membrane Arrays." Environmental Health Perspectives Exp 1:
137-143. " ~ ~~ - • — £ - - —
comm75' Gllbert Associates' Beading, Pennsylvania. Personal
communication
OCED. 1973. "Polychlorinated Biphenyls: Their Use and Control "
Environment Directorate from the Organization for Economic Cooperation
and Development. 02046. Paris.
,-J' Albrl«ht' and S'Y- Szeto. 1972. "Fate and Behavior
of Five Chlorinated Hydrocarbons in Three Natural Waters." Canadian
Journal of Microbiology 18(9): 1 393-1 398. ~ -
Oloffs P.C L.J. Albright, and S.Y. Szeto, and J. Lau. 1973. "Factors
Affecting the Behavior of Five Chlorinated Hydrocarbons in Two Natural
Waters and Their Sediments." Journal of __Kisheries_Je search Board of
Canada 30(11): 1619-1623. ---- ~~~ - -
Panel on Hazardous Trace Substances. 1972. "Polychlorinated Biphenyls-
Lnvironmental Impact." Environmental Research 5(3): 249-362.
Papageorge, W.P. 1975. Monsanto, St. Louis, Missouri. Personal
communication.
-325.
"Organochlorine Residues in Antarctic Snow." Nature
- -
Persson, B 1971. "ChJorinated Hydrocarbons and Reproduction of a
South Swedish Population of Whitethroat Sylvia Communls." Qlkos 22(2):
/-ty\.j~~ £-ji) m - -~ — -
I'latonow, N.S., and L.H. Karstad. 1973. "Dietary Effects of Poly-
" Medicine
Platnnow N.S., and B.S. Reinhart. 1973. "The Effects of Polychlori-
nntod Blphenyls (Aroclor 1254) on Chicken Egg Production, Fertility
and Hntchability.1 C^^d^r^ourna^^^m^arat ±v e Medicine 37(4)-
8-13
-------
Popper, H., P. Czygan, H. Creim, F. Schaffner, and A.J. Garro. 1973.
"Mutagenicity of Primary and Secondary Carcinogens Altered by Normal
and Induced Hepatic Microsomes." Society for Experimental Biology
and Medicine Proceedings 142(2): 727-729.
Price, H.A., and R.L. Welch. 1972. "Occurrence of Polychlorinated
Biphenyls in Humans." Environmental Health Perspectives Exp, 1:73-78.
Rabb, E.L. 1975. General Electric Company, Pittsfield, Massachusetts.
Personal communication.
Rehfeld, B.M., R.J. Bradley, Jr., and M.L. Sunde. 1971. "Toxicity
Studies on Polychlorinated Biphenyls in the Chick. (1) Toxicity and
Symptoms." Poultry Science 50(4): 1090-1096.
Richardson, A.F. 1975. "PCBs are Found in More Waterways; New Curbs
Planned." Chemical Marketing Reporter 208(12): 5-20.
Richel, S.M. 1974. "40 CFR 129 (Subport I)." Letters of January 15
to Hearing Clerk, U.S. Enivronmental Protection Agency, from General
Electric Company, Schnectody, New York.
Ringer, R.K., R.J. Aulerich,and M.Zabik. 1972. "Effect of Dietary
Polychlorinated Biphenyls on Growth and Reproduction of Mink,." Pre-
print of paper presented before American Chemical Society, New York,
New York, August 27-September 1, 1972. Vol. 12(1): 149-154.,
Risebrough, R.W., P. Rieche, D.B. Peakall, S.G. Herman, and M.N. Kirven.
1968. "Polychlorinated Biphenyls in the Global Ecosystem," Nature 220
(12/14): 1098-1102.
Risebrough, R.W., and B. dcLappe. 1972. "Accumulation of Polychlori-
nated Biphenyls in Ecosystems." Environmental Health Perspectives
Exp, 1: 39-45.
Ruopp, D.J. and U.J. DeCarlo. 1975. "Environmental Levels of PCBs."
Personal communication from D. Lokey, U.S. EPA.
Ruzo, L.O., M.C. Zabik, and R.P. Schuertz. 1974. "Photochemistry of
Bioactive Compounds: Photoproducts and Kinetics of Polychlorinated
Biphenyls." Journal of Agriculture and Food Chemistry 22: 199-202.
Safe, S-, and 0. Hutzinger. 1971. "Polychlorinated Biphenyls: Photo-
lysis of 2,4,6,2',4',6'-Hexachlorobiphenyl." Nature 232 (5313): 50-52.
Savage, E.P., J.l). Tesrsari, and J.W. Malberg. 1973. "The Occurrence
of Polychlorinated Biphenyls (PCBs) in Silage Stored in Pit and Upright
Silos." Bulletin of Environmental Contamination and Toxicology 10(2):
97-100.
8-14
-------
Savage, E.P., j.D. Tessarl, J.W. Malberg, H.W. Wheeler, and J.R. Bagby.
1973a. A Search for Polychlorinated Biphenyls in Human Milk in Rural
Colorado. Bulletin of Environmental Contamination, and Toxicology 9(4):
222-226. ' ' ~~~ " • — — — ^^
' !rP" J'D- Tessari' J-w- Malberg, H.W. Wheeler, and J.R. Bagby.
1973b. Organochlorine Pesticide Residues and Polychlorinated Biphenyls
inhuman Jlilk, Colorado-1971-1972." Pesticides Monitoring Journal
Schmidt, T.T., R.W. Riscbrough, and F. Cress. 1971. "input of Poly-
fl-\]r\V'tr\'lt,irll}1-~:l.,, _,--!__ T. f"l -1 , l~ , •*
forma Coastal Waters from Urban Sev
HfinAAl _Co"^jqj-J'MJ-.0" gnd Toxicology
tj j - - -'--^^.j. i y/ i_ . xiipuu ux x^uxy—
chlorinated Biphenyls into California Coastal Waters from Urban Sewage
Outfalls." Hu lie fin of I'ni'
SeLbc-r, I.N., J.K. Woodrow, T.M. Sl.of If', ,-ind U.K. Knos. 1975 "Deter-
mination ol IVstlcldes and their Transformation Products in Air " in
Environmental Dynamics of_P^t^i^^^ 17-43. Plenum Press, New York
a-nd London .
Sernyak, R. 1975. Rollins Environmental Services, Bridgeport, New
Jersey. Personal communication.
Shea, K.P. 1973. "PCB." Environment 1.5(9): 25-38.
Sherma, J. 1975. "Gas-Chromatography Analysis of Polychlorinated
Biphenyls and Other Non Pesticide Organic Pollutants." Advances in
Chromatography 12: 141-176. Marcel Dekker Inc. - "
Sinclair, P.R., and S. Granick. J974. "Uroporphyrin Formation
-Induced by ChJorinated Hydrocarbons (Lindane, Polychlorinated Biphenyls
retrachlorodibenzo-p-DIoxin). Requirements for Endogenous Iron
Protein Synthesis and Drug-Metabolizing Activity." Biochemical and
Biophysical Research Communicatjjrvn^ M m • 124-133.
Skrentny, R.F., R.W. Hemken, and H.W. Dorough. 1971. "Silo Sealants
as a Source of Polychlorobiophenyl (PCB) Contamination of Animal Feed "
.Bulletin of Environmental Contamination_andjroxjLcologv fi(S); 490-416.'
SBdergren, A. 1972. "Chlorinated Hydrocarbon Residues in Airborne
bailout." Nature 236: 395-397.
SL.-iiff, D.C., G.K. Q.miby, D.L. Spencer, and H.C. Starr, Jr 1974
Polychlorinated Blphcnyl Emission from I'luoresrent Lamp Ballasts "
ril'H!!!^ Toxh-oJ_ojg2. 12(4): 455-463.
Mxp
ling, D.I,., .,„,! F.,,. Mciyt.r> jr. ,972_ "Toxh-ities of PCBs to
.iml l',nvln,nm<-iit.il Residues." _Kiwl_ro,imont,T 1 Health Perspectives
'- " ~ — — c -
-------
Stanley, C.W., J.E. Barney, M.R. Helton, and A.R. Yobs. 1971 "Measure-
ment of Atmospheric Levels of Pesticides." Environmental Sci. and Technol.
5: 430.
Strik, J.J.T.W.A. 1973. "Chemical Porphyria in Japanese Quail
(Conturnixc japonica) Enzyme." Journal of Enzyme Physiology and Path-
ology 16: 211-223.
Sweeny, K.H. 1975. Envirogenics, El Monte California. Personal
communication.
Tarrant, K.R., and J.O.G. Tatton. 1968. "Organochlorine Pesticides
in Rainwater in the British Isles." Nature 219: 725-727.
Taylor, J. 1975. Gedcor Environmental Protection Corporation, West-
land, Michigan. Personal communication.
Tessari, J.D., and D.L. Spencer 1971. "Air Sampling for Pesticides in
the Human Environment." Journal of the AOAC 54(6) 1376-1382.
Tombergs, H.P. 1972. "The PCS Situation in Germany." Environmental
Health Perspectives Exp.1; 179-180.
Treon, J.P., P.P. Cleveland, J.W. Gappel, and R.W. Atchley. 1956.
"The Toxicity of the Vapors of Aroclor 1242 and Aroclor 1254.
American Industrial Hygiene Association Quarterly 17 (3): 204-213.
Trout, P.E. 1972. "PCB and the Paper Industry-A Progress Report."
Environmental Health Perspectives Exp.l: 63-65.
TRW. 1975. "Summary of CDS Field Testing. Micron and Submicron Process
Applications." Redondo Beach, California.
Tucker, E.S., W.J. Litschgi, and W.M. Mees. 1975. "Migration of
Polychlorinated Biphenyls in Soil Induced by Percolating Water.
Bulletin of Environmental Contamination and Toxicology 13UJt ab-yj.
Umeda. G. 1972. "PCB Poisoning in Japan." AMBIO 1: 132-134.
U.S. Army Environmental Hygiene Agency. 1975. "Disposal Instructions
for Askarels." HSE-ES/WP, 23 October, 1975. Aberdeen Proving Ground,
Maryland.
U S Department of Health, Education and Welfare. 1972*. "Final
Environmental Impact Statement-Rule Making on Polychlorinated
Biphenyls." Food and Drug Administration. Sam D. Fine, Associate
Commissioner for Compliance.
8-16
-------
U.S. Department of Health, Education, and Welfare 1972b "News
?7le?o1o" U'S> F°°d and DrUg Admi*istration, Rockville, Maryland. March
l/, 1972. re: .proposed regulations .PCBs .limit exposure from foods.
U.S. Environmental Protection Agency. 1972C. "Sewage Sludge Incinera-
of " °f SeWage Sludge Incineration, Washington, D.C.
. 1973a' "Water Quality Criteria
U.S. Environmental Protection Agency. 1973b. "40CFR Part 129 Water
Program, Proposed Toxic PolJutant Effluent Standards." Federal
Register 38(247): 35388-35395.
U.S. Environmental Protection Agency. 1974 revised. "Analysis of
Pesticide Residues in Human and Environmental Samples." Pesticides
and Toxic Substances Effects Laboratory. Research Triangle Park, N.C.
Vainio, H. 1974. "Enhancement of Microsomal Drug Oxidation and
Glucuronidation in Rat Liver by an Environmental Chemical, Polychlori-
nated Biphenyl.' Chemical and Biological Interactions 9(5): 379-387.
Veith G.D and G.F. Lee. 1971. "Chlorobiphenyls (PCBs) in the
Milwaukee River." Water Research 5(11): 1107-1115.
Veith, G.D. 1972. "Recent Fluctuations of Chlorobiophenyls (PCBs) in
the Green Bay, Wisconsin, region. " Environmental Health Perspectives
Exp 1: 51-54. ~~ — — — - c -
Venezio, R.A. 1976. U.S. EPA. Personal comnuni cation.
Versar, Inc. 1975a. Task III. Draft Report Contract 68-01-3259
Development of a Study Plan for Definition of PCB Usage, Wastes
and Potential Substitution In the Investment Coating Industry.
Versar Inc. 1975b. "Substitutes in PCBs Status and Prospects." Personal
communication, U.S. EPA.
Villeneuve, D C., D L. Grant, K. Khera, D.J. Clegg, H. Baer, and W.E.J.
?A iP lor??1- The Fetotoxiclty of a Polychlorinated Biphenyl Mixture
H2) 67-71 the RabMt ^ ^ ^ Rat-"- Environmental Physiology
Villeneuve, D.C. L.M. Reynolds, G.H. Thomas, and W.E.J. Phillips
1973. "Polychlorinated Biphenyls and Polychlorinated Terphenyls in
Canadian Food Packaging Materials." Journal of the Association of
otticial Analytical Chemists 56(4): 999-1001. ~ - '
8-17
-------
Vos, J.G. , J.H. Koeman, K.L. van der Maas, M.C. ten Noever de Brauw, and
R.H. deVos. 1970. "Identification and Toxicological Evaluation of
Chlorinated Dibenzofuran and Chlorinated Naphthalene in Two Commercial
Polychlorinated Biphenyls." Food Cosmetics and Toxicology 8(6): 625-633.
Vos, J.G., J.J.T.W.A. Strik, C.W.M. van Holsteijn, and J.H. Pennings.
1971. "Polychlorinated Biphenyls as Inducers of Hepatic Porphyria in
Japanese Quail, with Special Reference to Aminolevolinic Acid
Synthesis Activity, Flourescence, and Residues in the Liver." Toxi-
cology and Applied Pharmacology 20(2): 232-240.
Vos, J.C. ]972. "Toxicology of PCBs for Mamma1s and for Birds."
Environmental Health Perspectives, Kxp.1 : 105-117.
Waklmoto, T., R. Tachikowa, T. Ogawo, and L. Watouahe. 1974. "Method
for the Quantitation of Organic Chlorine Compounds In Air by the Dry
Sampling Method." Japan Analyst 23(7): 790-793.
WallnBfer, P.R., M. Koniger, and 0. Hutzingcr. 1973. "The Solubilities
of Twenty-One Chlorobiphenyls in Water." Analabs, Inc. Research Notes
13(3): 14-16.
Walker, W.H. 1973. "Where Have All the Toxic Chemicals Gone?" Ground
Water 11(2): 11-20.
Whitson, F. 1976. TRW, Redondo Beach, California. Personal
communication.
Willett, L.B. 1974a. "Coatings as Barriers to Prevent Polychlorinated
Biphenyl Contamination of Silage." Jpurnal__of Dairy Science 57(7):
816-825.
Willett, L.B. 1974b. "Coatings for Silos Continue to be Effective
Barriers to PCBs." Ohio Agricultural Research and Development
Center Research Summary 76: 32-34.
Williams, R., and A.V. Holden. 1973. "Organochiorine Residues from
Plankton." Marine Pollution Bulletin 4(7): 109-111.
Wlllmore, C.R. 1976. National Electrical Manufacturers Association
Washington, B.C. Personal communication.
Yap, H.H., D. Desalah, and L.K. Cutkomp. 1971. "Sensitivity of
Fish ATPases to Polychlorinated Biphenyls." Nature 233(5314): 61-62.
Yobs, A.R. 1972. "Levels of Polychlorinated Biphenyls in Adipose
Tissue of the General Population of the Nation." Environmental Health
Perspectives F.xp 1 : 79-81 .
8-18
-------
Young, D.R., T.C. Heesen, D.J. McDermott, and P.E. Smoklev. 1974.
"Marine Inputs of Po]ychlorinated Biphenyls and Copper From Vessel
Antifouling Paints." Southern California Coastal Water Research
Proj ect TM 212, May.
Young, D.R. 1975. Personal communication.
Young, D.R., and T.C. Heesen. 1975. "Inputs of DDT and PCB."
Southern California Coastal Water Research Projjgjc_t. Annual Report.
Young, D.R., D.J. McDermott, T.C. Heesen, and J. Tsu-Kai. 1975a.
"Pollutant Inputs and Distribution Off Southern California." Preprint
of a paper presented at the 169th National Meeting of the American
Chemical Society, Philadelphia, Pennsylvania, April 8-10.
Young, D.R., D.J. McDermott, and T.C. Heesen. 1975h. "Polychlorinated
Biphenyls Off Southern California." Paper presented at the Inter-
national Conference on Environmental Sensing and Assessment, September
14-19, Las Vegas, Nevada.
Young, D.R., D.J. McDermott, and T.C. Heesen. 1975c. "Polychlorinated
Biphenyl Inputs to the Southern Calfornia Bight." Background paper pre-
pared in the National Conference on Polychlorinated Biphenyls, 19-21 Nov
Chicago, Illinois.
Yule, W.M., A.F.W. Cole, and I. Hoffman. 1971. "A Survey for Atmospheric
Contamination Following Forest Spraying with Lenithrothion." Bull.
Environ. Contam. Toxicology 6: 289.
Zitko, V., and P.M.K. Choi. 1971. "PCB and Other Industrial Halo-
genated Hydrocarbons in the Environment. "Fisheries Research Board of
Canada (Technical Report No. 272). St. Andrews, N.B.
8-19
-------
TECHNICAL REPORT DATA
/rau' rcixl InilriK linn.', on ///<• reverse he jure i
\ m r\ MII No
ia'A-4rjO/l-77-04rj
1 I I I 1 I AND :>UI! MILL
Environmental Assessment of PCBs in' the Atmosphere
7 AUTHOR(S)
B. Fuller, J. Gordon, and M. Kornreich
8 PERFORMING ORGANIZATION REPORT NO
9 PL RKORMING ORGANIZATION NAME AND ADDRESS
MITRE Corporation
1820 Dolly Madison Blvd.
McLean, VA 22101
'.I'ON'.OHItJG AM NCY NAMl ANUADUMtSS
U.S. Environmental Protection Agency
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
3 HFCIPIENT'S ACCESSIOt*NO.
5 REPORT DATE
April 1976
6. PERFORMING ORGANIZATION CODE
MTR-7210 Rev. 1
10 PROGRAM ELEMENT NO.
11 CONTRACT/GRANT NO
68-02-1495
13. TYPE OF REPORT AND PERIOD COVERED
F i n_a]_ ___ _______
14 s~~
coo
200/04
15 SUPPLEMENTARY NOTES
16. ABSTRACT
This report examines atmospheric aspects of environmental problems associated
with polychlorinated biphenyls (PCB). Subjects covered include: (1) physical
and chemical properties of PCB, (2) monitoring methods for PCB in air, (3)
sources of PCB emissions, (4) environmental distribution, transport, and
transformation of PCB, (5) methods of control of PCB emissions, and (6) biological
effects of PCB.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Chlorohydrocarbons Carcinogens
Air Pollution Capacitors
Physiological Effects Transformers
Industrial Plants Ecology
Air Pollution Control
h.lDFNT IFIE RS/OPFN INDEDTERMS
Polychlorinated bi-
phenyls
Askarels
Stationary Sources
Monitoring
c. COSATI Field/Group
21. NO. OF PAGES
Unlimited
286
20 SECURITY CLASS (Thu page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
------- |