United States Region III - Environmental
Environmental Protection Services Division (Phila., PA.) and
Agency Office of Policy Analysis (Wash., D.C.) July 1987
SEPA KANAWHA VALLEY
TOXICS SCREENING STUDY
FINAL REPORT
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION III
841 Chestnut Building
Philadelphia, Pennsylvania 19107
AUG 1 3 1987
TO ALL INTERESTED AGENCIES, ORGANIZATIONS AND INDIVIDUALS:
Enclosed is a copy of the Final Kanawha Valley Toxics Screening
Study Report which was prepared by the U.S. Environmental Protection
Agency in cooperation with West Virginia environmental agencies. The
Report presents findings from a unique screening study of several
environmental pollutants which are routinely emitted or present within
the Kanawha Valley under normal conditions. The purpose of the study
was to estimate the relative importance of these chemical pollutants
by assessing their potential for causing long-term human health risks.
The procedure used for estimating risks contains many uncertainties
and the numbers generated are by no means absolute. Nevertheless, the
results are useful for comparing pollutants and sources and are intended
to serve as a guide for EPA and West Virginia agencies to set priorities
for future assessments.
The Toxics Screening Study provided important preliminary information
for environmental agencies, industry, and the public. However , because
the scope was very limited and only a few of several hundred chemicals were
examined, EPA and other organizations will continue to study potential
health threats and reduce unnecessary risks in the Kanawha Valley.
Although not required to take action as a result of this Report,
several companies in the Valley which emit or discharge the pollutants
studied have recently taken commendable strides to reduce releases of the
higher risk chemicals. We hope these preventive actions will continue as
we learn more about environmental risks.
I thank all of those who participated or have shown interest in
this challenging exercise. If you should have any questions or comments
concerning this Report, please feel free to contact Mr. Greene Jones,
Director of Environmental Services Division, at the above address.
James M. Seif (y
Regional Administrator
Enclosure
-------
- FINAL -
KANAWHA VALLEY,
WEST VIRGINIA
TOXICS SCREENING STUDY REPORT
July, 1987
Prepared by:
U.S. ENVIRONMENTAL PROTECTION AGENCY
Environmental Services Division
Region III Office
Philadelphia, Pa.
and
Regulatory Integration Division
Office of Policy Analysis
Office of Policy, Planning, and Evaluation
Washington, D.C.
-------
ACKNOWLEDGEMENTS
This report on Kanawha Valley, West Virginia was jointly pre-
pared by the U.S. Environmental Protection Agency's Region III Office,
Environmental Services Division, in Philadelphia, and by EPA's Regula-
tory Integration Division, Office of Policy Analysis, in Washington, D.C.
Numerous individuals and organizations provided generous support
and valuable input to the Kanawha Valley Toxics Screening Study, which
is the major subject of this Report. The project could not have been
successfully initiated without the tireless efforts of Glenn Hanson of
Region Ill's Air Management Division and Alan Jones, formerly with the
Regulatory Integration Division. Both of these individuals exhibited
considerable skill and foresight in setting up the initial project
framework, and securing necessary State and EPA office support.
Key individuals who were responsible for study management and
content of the report include Greene A. Jones, Division Director, and
Thomas Slenkamp of Region Ill's Environmental Services Division; and
Samuel Napolitano and David Lee of the Regulatory Integration Division.
Credit should also be given Robert Currie, formerly with the Regulatory
Integration Division, as well as current Director Daniel Beardsley,
and Stanley L. Laskowski, Deputy Regional Administrator, who collec-
tively initiated study discussions within EPA and with the State of
West Virginia and outlined initial study direction.
Special thanks must go to the following Directors and staff mem-
bers of the West Virginia environmental agencies who served on the
Policy and Implementation Committees for the study, and who provided
technical and staff resources, and advice on study scope and direction:
Mr. Carl G. Beard, Director, Air Pollution Control Commission
Dr. David K. Heydinger, Director, Department of Health
Mr. Ronald Potesta, Director, Department of Natural Resources
Mr. Ronald A. Shipley, Department of Natural Resources
Mr. Robert P. Wheeler, Department of Health
Mr. Donald A. Kuntz, P.E., Department of Health
Mr. Lucas M. Neas, Department of Health
Several offices within EPA provided technical guidance and support
for the Toxics Screening Study and Report. Special recognition is due
to the Office of Air Quality Planning and Standards (OAQPS), particularly
Joseph Padgett for his major coordination and leadership role, as well
as Jack Farmer, Richard Rhodes, John O'Connor, Dr. Robert Schell and their
respective staffs. The work of Dr. Larry Zaragosa is especially appre-
ciated in the area of pollutant assessment.
-------
EPA's Office of Research and Development (ORD) conducted a signi-
ficant air monitoring program in the Kanawha Valley which was useful
in the study and is reflected in the Report. Gerald Akland, Thomas
Hartlage, Thomas Lumpkin, and Dr. William McClenney of the Environmen-
tal Monitoring Systems Laboratory (EMSL) were key individuals in this
effort. Also, Scott Baker and Charles Brunot of ORD in Washington,
D.C. provided valuable advice concerning scientific and technical
issues.
Within EPA Region III, the following individuals deserve recog-
nition: fot technical or programmatic support, and/or review of drafts
of the report - Denis Zielinski, Alan Cimorelli, Joseph Kunz, Daniel
Ryan and Israel Milner, Air Management Division; James Newsom, Dr. Roy
Smith, John Ruggero, Victor Guide, Charles Kanetsky and Elizabeth
Rhodes, Environmental Services Division; Jeffrey Burke, Yener Soylemez,
Bohdan Mykijewycz, and James Harper, Hazardous Waste Management Division;
and Jon Capacasa, Daniel Sweeney, Thomas Merski, and Robert Lange,
Water Management Division. Special credit also goes to Robert Runowski,
William Hagel, and William Hoffman for preparation of sections of
the report and review of numerous drafts. Special thanks is extended
to Virginia Watson, Cheryl Johnson, Gayl Solomon, and Donna Bostick
for exceptional clerical support.
Additionally, within EPA Headquarters, the efforts of the following
individuals are greatly appreciated: for technical support, Dr. Susan
Perlin, Rebecca Madison, Stephen Horn, Cathy Crane, Keith Hinman, and
Dr. Alan Wiedow, Office of Policy Analysis; Cliff Rothenstein of the
Waste Management Division; and for initial report review and editing
advice, Joan O'Callaghan, Office of Policy Analysis.
Members of the Science Advisory Board, Subcommittee on Integrated
Environmental Management, provided valuable direction concerning the
quality of science and technical approach, through their initial review
and comment on the design of the Kanawha Valley Toxics Screening Study.
Members of the Subcommittee are listed in Appendix C.
Finally, four private contractors provided significant technical
support for the Toxics Screening Study and deserve recognition:
Versar, Inc. - Michael Alford, David Sullivan, Dennis Hlinka,
Craig Koralek, Bill Breman, Joseph Martini, and Kevin Jameson
American Management Systems, Inc. - Jay Wind, Steve Hufford
ICF, Inc. - Baxter Jones, Timothy Kipp, Gary McKowan, Judith
Vreeland
Battelle, Inc. - Michael Holdrun
-------
KANAWHA VALLEY TOXICS SCREENING STUDY REPORT
TABLE OF CONTENTS
PAGE
EXECUTIVE SUMMARY -i-
CHAPTER ONE. INTRODUCTION 1-1
I. ENVIRONMENTAL ISSUES IN KANAWHA VALLEY 1-1
II. PROGRESS IN CONTROLLING POLLUTION 1-3
III. NEED TO EXAMINE TOXIC POLLUTANTS 1-7
IV. EPA's GOALS AND OBJECTIVES FOR KANAWHA VALLEY 1-13
V. REPORT ORGANIZATION 1-13
CHAPTER TWO. KANAWHA VALLEY TOXICS SCREENING STUDY - FRAMEWORK 2-1
I. BACKGROUND AND REASONS FOR STUDY 2-1
II. STUDY OBJECTIVES 2-2
III. ANALYTICAL SCOPE 2-3
IV. THE STUDY AREA 2-5
V. INSTITUTIONAL RELATIONSHIPS 2-7
CHAPTER THREE. GENERAL ANALYTICAL METHODOLOGY 3-1
I. LIMITATIONS AND UNCERTAINTIES 3-1
II. OVERVIEW OF METHODOLOGY 3-5
CHAPTER FOUR. AIR QUALITY ANALYSIS 4-1
CHAPTER FIVE. DRINKING WATER ANALYSIS 5-1
CHAPTER SIX. SURFACE WATER ANALYSIS 6-1
CHAPTER SEVEN. HAZARDOUS WASTE ANALYSIS 7-1
CHAPTER EIGHT. PUTTING RISK IN CONTEXT 8-1
REFERENCES R-l
APPENDICES
A. SUMMARY OF OTHER KANAWHA VALLEY ACTIVITIES A-l
B. KANAWHA VALLEY HISTORY, DEVELOPMENT AND EMPLOYMENT TRENDS B-l
C. SCIENCE ADVISORY BOARD REVIEW C-l
D. METHODOLOGY FOR EVALUATING HEALTH RISKS D-l
E. RISK ASSESSMENT FOR NON-CANCER EFFECTS E-l
-------
NOTICE
This Report has been independently reviewed in public session by
the Integrated Environmental Management Subcommittee of the Science
Advisory Board (SAB). The SAB is an independent public advisory group
providing scientific information and advice to the Administrator and
other officials of the Environmental Protection Agency. The Board
consists of nationally recognized experts representing a variety of
scientific disciplines, and is structured to provide a balanced expert
assessment of scientific matters related to issues facing the EPA.
The IEM Subcommittee's final comments on the Kanawha Valley Toxics
Screening Study, as well as a list of the Subcommittee members are
included in Appendix C.
-------
Progress and Compliance with Existing Environmental Laws
EPA and West Virginia agencies administer more than a dozen major
environmental laws collectively aimed at controlling environmental
pollution to acceptable levels. Considering the large number of
industrial facilities operating in Kanawha Valley today (close to 200
of all types); there is reasonably good compliance with these laws
(i.e., the large majority of facilities are operating in compliance
all or most of the time). This compliance success has resulted in
improved environmental conditions throughout the Valley:
o Air quality in the Valley consistently meets applicable
primary Ambient Air Quality Standards for the six criteria
air pollutants;
o The nineteen (19) major industrial facilities operate under
approved National Pollutant Discharge Elimination System
(NPDES) permits, and sixteen (16) of these are based on
Best Available Technology (BAT) limits;
° All public drinking water supply systems are consistently
in compliance with Maximum Contaminant Levels (MCLs) esta-
blished under the Safe Drinking Water Act;
o Eleven facilities are required to have permits under the
Resource Conservation and Recovery Act (RCRA) for handling,
treatment, storage, or disposal of hazardous waste. Only
one facility is currently not in compliance with RCRA regu-
lations .
While the trend is clearly toward generally improved environmental
conditions, there are occasional violations of established criteria
resulting in periodic stress to environmental quality. EPA and the
State of West Virginia will continue to vigorously enforce existing
environmental laws to minimize these violations.
THE KANAWHA VALLEY TOXICS SCREENING STUDY
Although there is generally good compliance with current laws,
many hundreds of chemicals routinely released within the Kanawha Valley
environment are not regulated by these laws, partly because information
is lacking concerning their toxicity, persistence, or fate (i.e., is
the population exposed to significant levels of these compounds?). In
order to help answer these questions, it was recognized that more
information was needed to begin to assess the possible risks posed by
some of these hundreds of chemicals.
-ii-
-------
EXECUTIVE SUMMARY
Introduction
This summary briefly explains the purpose, content and findings
of the environmental report on Kanawha Valley, West Virginia which
follows. The report was primarily prepared to describe a limited
study about toxic chemicals in the Valley undertaken by the U.S.
Environmental Protection Agency (EPA), in cooperation with three West
Virginia environmental agencies - the Air Pollution Control Commission,
Department of Health, and Department of Natural Resources. The report
also contains summary information concerning the area's environmental
compliance with current environmental laws, and includes, in Appendix
A, a brief description of some other ongoing activities concerning
Kanawha Valley environmental issues.
Why the Kanawha Valley Was Studied
The Kanawha Valley represents a unique geographical setting of
mountainous terrain with winding streams and valleys, and widely
variable meteorological conditions. It also includes a large concen-
tration of industrial facilities, with residential population often
located in close proximity to these facilities. These features in
combination make the protection of public health and the environment
a great challenge. During the past 25 years, this challenge has been
met by a cooperative effort of government and industry, and has resulted
in substantial improvement to the environmental quality of the Valley.
At the same time, in the Kanawha Valley, as in many other locations
nationwide, there remain complex environmental challenges which are
difficult to fully address, including the question of whether, and to
what degree, low levels of chemical pollutants present in the environ-
ment may cause unacceptable health and environmental risks.
Because the Kanawha Valley includes many chemical manufacturing
and handling facilities which routinely release small quantities of
potentially toxic chemicals to the outside environment, Valley residents
and officials have long been concerned with the possible health impacts
associated with long term exposure to these chemicals. Recent scien-
tific developments have for the first time enabled the Environmental
Protection Agency and other organizations to quantify and compare
potential health risks from certain chemical contaminants, using an
approach termed risk assessment. This capability was applied to the
Kanawha Valley to begin to assess some of the chemical exposures and
possible health risks which may occur within the Valley.
-------
Objectives of the Study
The Kanawha Valley Toxics Screening Study was designed to provide
an initial understanding of selected toxic pollutants which are present
in the ambient environment of the Kanawha Valley, to identify possible
sources of these pollutants, and to provide a sense of the nature and
relative priority of potential health risks associated with public
exposure to these chemicals.
Study Scope
Perhaps the most important point to understand about the study's
scope is that it was intended to provide only a limited understanding of
a few possible environmental risks within certain areas of the Kanawha
Valley. Our intent was not to comprehensively analyze all available
environmental data in order to determine whether or not there is a
significant cancer or other health risk from environmental pollution,
nor to identify all possible environmental problems which might be
present in the Valley. While we did analyze several issues in moderate
depth of detail, the study was not comprehensive in scope. It was
thus not a "screening" study in a broad or literal sense. That is, it
did not attempt to discover all possible toxic risks from all sources
within all potential exposure pathways.
What the study attempted to provide was an initial understanding of
potential risks from current ambient exposures for a few selected chemi-
cals, so that residents and officials could begin to understand the
nature of some risks present in the Valley and could set priorities
for further analysis of risks, and, where appropriate, control actions.
Omitted from this study were several types of environmental risks,
including those from cigarette smoking and diet, from indoor air exposures,
from occupational exposures, and from accidental releases of chemicals.
Although these are all potentially important exposure pathways, with
limited funding and the need to rely on existing data, we decided that
the screening study could best meet objectives by focusing on long
term health risks posed by average ambient exposure to selected chemicals.
The Toxics Screening Study examined potential community exposure to
several toxic pollutants through four exposure pathways by which contact
could conceivably occur: through (1) inhalation of ambient (outdoor)
air, (2) consumption of public drinking water, (3) consumption of fish
from the lower Kanawha River, and (4) consumption of groundwater which
could be contaminated by hazardous waste sites.
-iii-
-------
In the Air Analysis, twenty (20) known or suspected cancer-causing
(carcinogenic) chemicals (14 from point sources) were evaluated out
of more than 450 chemicals listed on an inventory of Kanawha Valley
air emissions prepared by the West Virginia Air Pollution Control
Commission. Potential community exposure to these 20 chemicals was
estimated through mathematical modeling within four distinct geographic
zones: Belle, Charleston/South Charleston, Institute, and Nitro.
For Drinking Water, nineteen (19) chemicals for which monitoring
is required were evaluated from the study area's fourteen public
water supply systems. Both potential cancer risks and non-cancer
effects were evaluated. Private wells and non-community supplies were
not included.
The Surface Water analysis focused on potential health effects,
both cancer and non-cancer, from the possible consumption of contami-
nated fish taken from the Kanawha River near Nitro. In addition,
concentrations of chemicals monitored within the River were compared
with water quality criteria established for the protection of human
health and aquatic life.
The Hazardous Waste analysis examined data from known hazardous
waste sites in the Kanawha Valley in order to devise hypothetical
hazardous waste settings (i.e., combinations of various facility types,
hydrogeology, and waste constituents) which could present a risk to
human health through public consumption of contaminated ground water.
It did not include any site-specific health risk estimates.
Chemicals were selected for study within each media analysis
primarily on the basis of the following criteria: (1) their known or
estimated quantities or concentrations within the Kanawha Valley; and
(2) sufficient information about their health effects to enable a
reasonable estimate to be made of potential risk. Pollutants studied
do not necessarily represent those posing the highest risks within the
Valley.
A large number of man-made chemicals are present or routinely
released to the environment of Kanawha Valley. However, because
information concerning the above criteria was often lacking or limited,
we were only able to roughly evaluate potential risks for a few of
those chemicals.
Because the study originated out of concern over air toxics, the
air portion of the study received by far the majority of attention
and allocation of available resources. We estimate that approximately
80% of study effort was devoted to the air analysis, 10% to the hazar-
dous waste analysis, and 5% each for the drinking water and surface
water analyses. This distribution of effort should not be construed
as correlating with the actual distribution of environmental problems
by media in the Valley.
-iv-
-------
Study Methodology and Limitations
The Kanawha Valley Toxics Screening Study used a risk assessment
screening methodology to evaluate and compare, in a limited fashion,
potential human health risks from current ambient exposure to a limited
set of pollutants. The purpose of this methodology was to permit a
comparison of one risk with another and to provide a general sense of
the risk a given chemical substance may present. This methodology
does not permit a definitive statement concerning the absolute risk
posed by a particular pollutant, source, or exposure pathway.
The two key elements in estimating risk are a chemical's
estimated potency, or toxicity, and human exposure to that chemical.
EPA estimates a chemical's toxicological potency on the basis of
available scientific evidence. Exposure to a chemical is estimated
by measuring or estimating the concentration at which a chemical is
present in the air or water, for example, and then making assumptions
about how much air a person breathes or water he or she drinks.
Potency and exposure estimates are combined to estimate individual
and population risks.
The Toxics Screening Study examined potential cancer risks, and,
in a more limited way, non-cancer risks where possible. For cancer
risks, available scientific data allowed for a quantitative estimate
of potential risks. For non-cancer effects, a qualitative analysis
was performed, simply indicating whether or not estimated exposure
appeared to be high enough to place an individual at increased risk of
an adverse health effect.
For this analysis, risk to an individual is defined as the in-
creased probability that an individual exposed to one or more chemicals
will experience a particular adverse health effect (e.g., will contract
cancer) during the course of his or her lifetime. The risk estimated
for a particular type of exposure is the incremental risk beyond that
which a person faces from exposure to other environmental or hereditary
causes of disease, sometimes referred to as the background rate of
disease. Calculated risks are considered plausible, upper-bound esti-
mates of potential risk rather than predictions of most likely risk.
For cancer effects, two types of individual risks were estimated:
(1) average individual risk, for a typically exposed individual, and
(2) for the air analysis only, risks to an individual who may be
particularly close to a source or who is likely to receive higher
than average exposure. Risk to the population as a whole is expressed
as the plausible, upper-bound increased incidence (number of cases),
above the background rate, within an exposed population.
-v-
-------
Despite the application of similar risk assessment techniques,
each media analysis differed in terms of geographic coverage, chemicals
studied, whether ambient concentrations were measured or estimated,
the potential health risks considered (cancer or non-cancer), and in
the means of expressing the risks studied (e.g., risk assuming uniform
ingestion of drinking water versus risk per pound of fish consumed).
Table 1 illustrates some of these differences.
TABLE 1
DIFFERENCES IN RISK ASSESSMENT APPROACHES1
BY MEDIA ANALYSIS
RISK ASSESSMENT APROACH
MEDIA STUDIED
PRIMARY ROUTE OF
EXPOSURE ANALYZED
Exposure Data
Health Effects Considered
Measured Modeled
Cancer Non-Cancer
Air Analysis
Inhalation of
X
X
Ambient Air
Drinking Water
Ingestion of Public
X
X X
Analysis
Drinking Water
Surface Water
Consumption of
X
X X
Analysis
Contaminated Fish
Hazardous Waste
Ingestion of
X
X X
Analysis
Contaminated Ground
Water
*In addition to these differences, the area of geographic coverage and types
of chemicals studied for each media analysis were different, as was the
means of expressing the risks studied - see text.
The application of risk assessment techniques in a complex geo-
graphic area such as the Kanawha Valley is difficult because there are
many scientific and technical limitations. For example, there are
uncertainties in estimating the amount of pollutants released from a
given source, in estimating the extent of exposure of the general
public to a particular pollutant(s), and in extrapolating health
effects evidence derived from animal studies to estimate corresponding
human responses. In general, the uncertainties introduced by the
exposure data are usually much smaller than those associated with the
dose response information used to estimate human health risks.
-------
Because of these uncertainties, EPA prefers to err on the side of
caution in making risk assessments. Therefore, following standard EPA
protocol, we use conservative estimates of risks that probably overes-
timate the actual risk from any one chemical. While the risks outlined
in the study should not be interpreted as a definitive prediction of
actual cancer risk to the general public in the Valley, these estimates
are useful in comparing the relative risks among various chemicals.
SUMMARY OF STUDY FINDINGS
Considering the approach and limitations noted above, following
is a summary of the study's major findings:
Air Analysis
A total of 20 chemical carcinogens were evaluated within the four
air zones (Belle, Charleston/South Charleston, Institute, and Nitro),
although only 14 of these are emitted from specific point sources
(e.g., industrial facilties). Not all chemicals are emitted in all
four zones. For the study area in general, point sources accounted
for the great majority of risks studied, while areawide, or non-point
sources (e.g., automobiles, dry cleaners, gasoline marketing) contri-
buted relatively little.
Within the Belle zone, the increased lifetime cancer risk for an
average individual from exposure to the chemicals studied was estimated
to be 2.5 x 10~4, or two and one half chances in ten thousand of con-
tracting cancer, aside from other causes. Within certain neighborhoods
nearby major industrial facilities, the increased individual risk could
be as high as 3 x 10~3, or three in one thousand. That is, at most, 3
out of every 1000 people living within these Belle neighborhoods could
contract cancer as a result of lifetime inhalation of the chemicals
studied. Table 2 portrays these risk results by zone.
Within the Charleston zone, the increased lifetime cancer risk to
an average individual from exposure to selected chemicals was estimated
to be 2.9 in 10,000. Within certain neighborhoods within the Charleston
zone, this risk could be as high as 5 in 1000. Within Institute, average
individual risk was estimated to be 1.1 in 1000, while in certain locations
individual risk could approach 1 in 100. For those chemicals studied
in the Nitro zone, estimated air risks were considerably less than in
the other three zones.
By assuming that all 20 of the air pollutants studied were evenly
dispersed throughout all four Valley zones, a box modeling approach
conservatively estimated that an additional 126 cancer cases could
result over a 70 year period of exposure.
-vii-
-------
**************************************
* Note: The numerical health risk estimates presented herein are based *
* on conservative assumptions which tend to produce an upperbound, or *
* maximum, value. Actual risks are not likely to be any higher; however,*
* they could be significantly lower, and may even approach zero. Because*
* of the many uncertainties and limitations in the methods used to *
* calculate risks, the risk estimates should not be considered defini- *
* tive predictions of actual health risk. The proper function of the *
* estimates is to help responsible officials select issues and set *
* priorities for future study. *
**************************************
TABLE 2
AIR ANALYSIS RESULTS BY ZONE
ZONE
Belle
Charleston/
South Chas.
Institute
Nitro
POPULATION AFFECTED POLLUTANTS
Total Nearby CONTRIBUTING
in Zone Facilities MOST TO RISK
ESTIMATED UPPER BOUND
INDIVIDUAL LIFETIME
CANCER RISK1
Nea rby
Average Facilities
15,530
51,750
22,390
9,990
99,660
600
2700
1300
1450
6050
Chloroform
<2.5xl0-4
Ethylene Oxide <2.9x10 4
Acrylonit rile
Ethylene Oxide <1.1x10"3
1,3-Butadiene
Trichloro-
ethylene
<2.6xl0"6
<3xl0-3
<5x10-3
<8xl0-3
<6xl0"6
1 Risks presented are a sum of the risks from individual chemicals studied
within a given zone
Source: Regulatory Integration Division,
Office of Policy Analysis, EPA, 1987
-viii-
-------
Drinking Water Analysis
Estimated individual cancer risks from certain chemicals studied
which are present in public drinking water supplies were approximately
1 in 10,000. For the total population served by these systems, this
risk would mean an additional 46 cancer cases over a 70 year period
of exposure. The primary chemical of concern is chloroform, a byproduct
formed during the disinfection process, and one of a class of compounds
called trihalomethanes (THMs). THMs are formed as the result of
chlorination of drinking water, which is used to prevent outbreaks
of waterborne bacterial diseases. All Kanawha Valley public water
supplies studied currently meet the EPA standard for THMs. If THMs do
not exceed the EPA standard, they are generally considered to be an
acceptable risk.
For non-cancer risks, public drinking water data available for
nineteen chemicals was compared with the corresponding threshold
levels (Reference Doses, or RfDs), below which adverse health effects
are unlikely to occur. None of the pollutants studied exceeded the
RfDs. In addition, an analysis of finished water for 129 priority
pollutants done for the West Virginia Water Supply Company in July,
1985 revealed no appreciable contamination. We did not have similar
data available for the smaller systems (<10,000 people served).
Among several limitations in this analysis was our inability, due
to a lack of data, to characterize risks posed by private wells, which
by one estimate number 22,000 in Kanawha County alone.
Surface Water Analysis
Fish tissue samples taken from the Kanawha River near Nitro in
September, 1985 showed elevated levels of several chemical compounds,
including PCBs, pesticides and dioxin. As a result, the State of West
Virginia issued a fishing advisory for a stretch of the Kanawha River
extending from the Coal River downstream to the Ohio River. Although
actual fish consumption rates are unknown, our study revealed that risks
could be significant if fish are regularly consumed from this area of
the Kanawha River. For example, consumption of from one tenth to one
pound per week of certain species over a lifetime could result in an
increased risk of contracting cancer of 1 in 1000.
-ix-
-------
Analysis of non-cancer risks was difficult, since very few pollutants
studied in the Surface Water Analysis have verified Reference Doses with
which to compare. However, based on limited data, we found that consump-
tion of between 3/10 and 2 pounds of fish per week, depending on the spe-
cies, could exceed the Reference Dose for chlordane, resulting in an in-
creased probability of chronic, non-cancer health risks. Also, consump-
tion of as little as 1 pound per week of certain species could result in
exceedance of the RfD for methyl mercury at the highest observed concen-
trations. For some species, these pollutants were only found at signifi-
cantly lower levels.
When we compared concentrations of metals measured in the Kanawha
River to water quality criteria for protection of aquatic life, we found
that lead and cadmium exceeded the criteria for 44 and 30 percent of the
observations, respectively. In general, however, the chemicals for which
monitoring data was available only infrequently exceeded any water quality
criteria.
Hazardous Waste Analysis
Although there are many potential hazardous waste sites in the
Kanawha Valley, preliminary risk assessments have been conducted at only
some of these sites. Active programs to clean up and manage hazardous
waste at specific sites are underway at EPA and the State.
For the Toxics Screening Study, representative hazardous waste site
conditions were formulated to gain a sense of some risks which could be
associated with different conditions. Based on limited data, the study
identified potential risks which could occur, under hypothetical condi-
tions, by consumption of ground water contaminated by hazardous waste.
Under the worst case analyzed, potential individual cancer risks could
approach one in a thousand. Risks were not estimated for any actual,
specific sites in the study area. The established EPA and State programs
for controlling hazardous waste are currently evaluating whether ground
water might be contaminated near any specific sites.
Overall Comparison of Risk Estimates
As indicated earlier, a precise estimate of risk from any one pollu-
tant is difficult to determine due to limitations concerning emission
levels, exposure, health effects, and other factors. However, the study
was valuable in quantifying potential cancer risks from selected pollutants
which allows for comparing these risks to get a sense of the relative
importance of the chemicals studied. Table 3 presents the chemicals
studied which resulted in the highest cancer risk estimates. Other
chemicals not studied and those associated with other than cancer effects
would not be shown here, but are nonetheless also of possible concern.
-x-
-------
Comparison of Risks with Other Areas of the Country
It is difficult to compare risks between different geographic areas.
Studies done for different regions with different objectives often have
exposure and risk measures that are not easily comparable. However,
EPA has estimated some ambient concentrations and risks associated
with certain chemicals in different areas of the country. Where those
chemicals and exposure assumptions were similar to those studied in
Kanawha Valley, we have attempted to present a comparison of estimated
levels and risks in Chapter Eight.
In general, the ambient concentrations and estimated risks from
many of the toxic chemicals which we evaluated in the Kanawha Valley
appear to be within the range we see in other areas where EPA has data
for similar individual pollutants. However, we see concentrations in
the Kanawha Valley at both the high and low end of this range.
The risk associated with drinking public water in the Kanawha
Valley appears to be about the same as in other cities in the United
States which also use surface water as a source. However, within the
air analysis, many risks estimated in Kanawha Valley appear to be
appreciably higher than comparable point source pollutant risks evalu-
ated by EPA in national air toxics studies.
Future Direction
The principal goal of this study was to assess some of the environ-
mental risks in the Kanawha Valley, in order to help set priorities for
future study and for managing those risks. A review of the analytical
findings and discussion within each media report suggests two major
priorities for future action: (1) a need to take steps to reduce the
significant risks identified, and (2) a need for continued analysis
of other chemicals, health effects and risks in the Valley.
Although there were many uncertainties and limitations in this
study, in certain instances the potential risks calculated are suffi-
ciently high to suggest that action to reduce the risks should be
initiated by EPA, the West Virginia regulatory agencies, and/or indus-
try. Where necessary, these risk management actions can be accomplished
through government regulation or through voluntarily negotiated reduc-
tions with the responsible industries.
Risk management is a process which involves residents, industry,
and government, and which includes many considerations beyond simply
the risk assessment results, such as available control methods, costs,
benefits, public opinion, and legal authority. All of these factors
must be evaluated before deciding what represents an acceptable level
of risk and what are the best means and tradeoffs for reducing an
existing risk.
-xii-
-------
TABLE 3
KANAWHA VALLEY CHEMICALS STUDIED
CONTRIBUTING MOST TO ESTIMATED CANCER RISKS
CHEMICAL
EXPOSURE
PATHWAY
GEOGRAPHIC
AREA
PROBABLE
SOURCE
Potential Risks*on the Order of 1 in 1000 (within specific locations):
1,3 - Butadiene
Ethylene Oxide
Chloroform
Ethylene Oxide
Air
Air
Air
Air
PCBs and Dioxin^ Surface Water
(Fish Tissue)
Institute
Charleston/
South Chas.
Belle
Institute
Nitro
Union Carbide/
Rhone-Poulenc
Union Carbide-South
Charleston
Occidental 2
Union Carbide/
Rhone-Poulenc
Unknown
Potential Risks*on the Order of 1 in 10,000 (within specific locations):
Acrylonitrile Air
Methylene Chloride Air
Chloroform (THMs) Drinking
Water
Acrylonitrile Air
Carbon Tetrachloride Air
Chloroform Air
Charleston/
South Chas.
Belle
Water Supply
Service Areas
Institute
Belle
Institute
Union Carbide-South
Charleston
Occidental^
Public Water
Suppliers
Union Carbide/
Rhone-Poulenc
Occidental 2
Rhone-Poulenc1
1 Formerly Union Carbide
2 Formerly Diamond Shamrock
3 Assumes consumption of 1/10 lb/week of contaminated fish over a lifetime;
Higher consumption rates would result in higher risks
*N0TE: BECAUSE OF SIGNIFICANT UNCERTAINTIES IN THE UNDERLYING DATA AND
ASSUMPTIONS, THESE ESTIMATES OF INDIVIDUAL RISK ARE ONLY ROUGH
APPROXIMATIONS OF ACTUAL RISK. THEY ARE BASED ON CONSERVATIVE
ESTIMATES OF EXPOSURE AND POTENCY, AND ARE MORE LIKELY TO OVER-
ESTIMATE RISKS THAN TO UNDERESTIMATE THEM. See Text.
Source: Regulatory Integration Division, EPA, 1987
-xi-
-------
EPA, in cooperation with West Virginia agencies, will continue to
reduce risks, where possible, by maintaining compliance with existing
environmental laws. Provisions for control of certain toxic chemicals
are available, for example, in the Clean Air Act for designated Hazar-
dous Air Pollutants (HAPs), through the Toxic Substances Control Act if
an "unreasonable" health risk is identified, by limiting toxic discharges
in point source surface water (NPDES) permits, and by enforcing the
Maximum Contaminant Levels (MCLs) set for certain drinking water pollu-
tants.
Where identified risks are not controllable through existing Federal
laws and programs, the State of West Virginia could independently pursue
a regulatory approach to control toxic pollutants in a variety of
ways (e.g., risk-based, control technology based, specified ambient
levels), or can continue to negotiate, with EPA assistance, with specific
industries on a voluntary basis to achieve further emission reductions.
In the past two years, major industries in the Valley have voluntarily
committed to reducing emissions of several toxic air pollutants, and
have recently moved to further decrease emissions based on results of
this Study.
The Kanawha Valley Toxics Screening Study can be viewed as an
initial step in a continuing process to assess and, where necessary,
reduce risks. Many chemicals, health effects, and types of risks not
considered in the Screening Study (e.g., acute health effects, acciden-
tal releases) may be individually or collectively important and may
merit further attention. The Study identified many scientific uncer-
tainties and data needs that may be appropriate research and management
priorities for followup by EPA and West Virginia agencies and other
groups. Future direction should include, in addition to reducing
risks from chemicals identified in the Study, further analysis of
these additional topics.
EPA is now working with West Virginia's environmental agencies to
develop specific plans for responding to the risks and needs identified
in this Study. Some of the possible activities include the following:
Air - in addition to the initial priority of reducing emissions for
those chemicals posing significant risks, the following activities
would contribute to a better understanding of air toxics risks in
the Kanawha Valley:
1. Improve the emissions inventory. Although the Kanawha Valley
inventory represents state-of-the-art estimation techniques,
some of the uncertainties can be further reduced.
(a) Ensure emission estimates reflect documented voluntary
reductions committed to by industry from 1984 production levels.
-xiii-
-------
(b) Apply recently developed and EPA approved plant-specific
methods for more accurately estimating contributions from
fugitive (dispersed) sources.
(c) Include estimates of short-term variable emissions to enable
future analysis of possible acute and sub-chronic health effects.
(d) Prioritize chemicals listed on the inventory into levels of
concern for future study, based on emission quantities, toxicity,
and potential for exposure.
2. Improve methods for measuring air toxics (especially for those
posing significant risks) and conduct further ambient monitoring
within selected areas to obtain more precise estimates of exposure
and health risks.
3. Develop an approach for estimating non-cancer health risks to
complement the quantitative analysis of cancer risks.
4. Perform additional risk assessments to include additional chemi-
cals and short-term effects, and using the most recent health
effects information.
5. Evaluate potential risks from accidental releases.
Drinking Water - the most pressing need appears to be for a survey of
the location and water quality of private drinking water wells in the
area, some of which could be impacted by hazardous waste sites. Also,
testing of smaller (<10,000 people served) public systems for trihalo-
methanes and additional toxic contaminants would be useful.
Surface Water - additional ambient monitoring of Kanawha River water
quality and further fish tissue analysis would be helpful to establish
whether the elevated concentrations of PCBs, dioxin, chlordane, methyl
mercury, and other chemicals found near Nitro are also present in other
River reaches. Knowledge of fish consumption rates and patterns would
allow for a better estimate of potential risks. West Virginia's Depart-
ment of Natural Resources' creel survey should provide this latter type
of information.
Hazardous Waste - the efforts to assess and cleanup specific sites
should continue under the established Federal and State hazardous
waste programs. From the Toxics Screening Study, sites in the shale/
sandstone hydrogeologic regime tend to present the greatest potential
health risks due to the increased probability of private well contami-
nation.
-xiv-
-------
In addition to the Toxics Screening Study, there are many other
programs and activities underway in the Valley and in the Nation col-
lectively aimed at better understanding and reducing risks to public
health and the environment. Some of these activities are described in
Appendix A.
Implementation of these efforts and the risk assessment analyses
presented in this report provide a resevoir of valuable information
for government, industry, and citizens to work toward continued environ-
mental improvement in the Kanawha Valley.
-xv-
-------
Chapter One
Introduction
-------
CHAPTER ONE
INTRODUCTION
I. BACKGROUND AND ENVIRONMENTAL ISSUES IN KANAWHA VALLEY
The Kanawha River Valley in central West Virginia, from Gauley
Bridge to Winfield Dam (see Figure 1-1), has often been an area of
environmental concern, principally due to its heavy concentration of
chemical and other industrial facilities situated in close proximity to
populated areas, and also because of unique topographical and meteoro-
logical conditions.
Beginning with the development of the chemical industry in the
Valley in approximately 1915, and continuing through several decades
of industrial expansion and accompanying population growth, the area
often experienced severe environmental stress.
Large volumes of municipal and industrial waste and other pollu-
tants were discharged to the Kanawha River resulting in gross pollu-
tion. At times, the River was unable to support even pollution tolerant
fish populations. Taste and odor problems were severe in public water
supplies obtained from the River. Beneficial uses of the River were
severely impaired.
Degradation of air quality was also a major problem. Emissions
of air pollutants from chemical plants, powerplants, and other manufac-
turing facilities, and from automobiles and localized sources in heavily
populated areas, were often trapped in the narrow Valley by adverse mete-
orological conditions. This often resulted in severe air pollution
episodes in the populated areas.
Large volumes of hazardous waste and other waste residuals were
disposed of in landfills, dumps, and surface impoundments that were not
properly designed, constructed, or maintained to adequately contain the
toxic substances present in the waste. As a result, toxic pollutants
were released in the air, to surface water, and to ground water.
Beginning in the 1960's, major new Federal, State, and local envi-
ronmental laws and programs were initiated, and municipalities and
industries invested major capital expenditures in plant improvements
and pollution controls. During the last 25 years, these combined
efforts have improved environmental quality in the Kanawha Valley and
elsewhere throughout the country. Many of the most obvious and pressing
problems were greatly reduced, or even eliminated, by removing large
amounts of conventional pollutants from the air and water. As a result,
the risks faced by the public and the environment from these highly
visible forms of pollution have been greatly reduced.
-------
StOflC
.¦J J
^ MintrsvHI*
g , »\HartfOfd Oam 31
Ch«ihjr. ^Jy2?!i*00d
m( Too«t»C Lm(MU -
- Grov*
Etizab«thVr-^o|^"
w aiy&£*
C Macf»lan ^m.thy^tl
Srofcard _
v" ' *" / TnurMay
MuAdly ^
i I
VI «--
.Point Pleasant
*(2>
(^SffWfton^
Ounenv1
Sai*Jyv.»t» Li»*booi'|~'|JJJ2*
,gt X
^prte^
•;$> S
,-'«ad#afl ©AgpjtGrowt
g£r - .
¦**>•• ~ v**-\
-yr \§2«y C'own Ciff\
'S1w»«i»y MJtoJJf G,-« Bottom
1 AthaJta (X* x
)u(jg
' RoM/taBuf f
\iH"~%->rvx/V""'
viei»»A r «—-»• jt
— . -=/ \ WlllAA J-K.
"OC-.
iCtua
f Liftwty^^! * y
STUDY AREA
ftrflf*..— u-
VIV 0^n«,V 4Corton
if*™"? £if*-/-pJLi.«> V r
.'' Vall#yforh"7
ffy vemon ayy i-VrtVu
^o;, Htilton/^'^H&UQSSIL* luitrtir^SW£?
~C&f5k yW^Birtours.il. \ Cullodw \ "'^eO^vSv^^Nnjtl
/ *Sbp«^
^ViXv""7_ c—<3r T°'n"°^^h*arleston'
v£ii I' ^VElkvitw ^
2gT
Ikvtew
Somi
,i:
: Oundon V*. Widfl ^ •
»r
"1 Creek.'"
G«Tft»s
Sand
——¦—"-3>^i3n*mjbC} ~~
is ^"<1^ ^ HamlinV
H V i' %f?] /Sweatiand^^ f Vawhay J >
1 Branchlandl~o p i -^*V
^ • fjVt S^SS-~-"T'M,dki?fte-'- ••'' V\ S™~""AW.vij X t^'^'MoraO (6
arleston0u"
^T^^^Crieo
iL^Cual ForifJ_Blount
-. ' L^*S^__BJekele|
h> - •-¦ ^.Pond
'^.Gap
u"nma'h
HI*-" Ioy_.
Londo!
Smil
\ \v Glan
\
, A Y Eno^i X
[H«rti4Ad 2/
V
Dad*
>n$UR1
^ Q^i s^Shltntr
Cowo Ga«^.
K.ah«viii«"^
% F'luton
%Ouahar
Qua, *v-' Eio.m / P#rrelfs^^<
/- BiQ Creait ^ j
Biii-"—" ««-
f
s«th©v?\ ; b IS; v 4^fy.ctof
.'*e»kdal« 3 Z •¦
^Madison "V®^^ ^l.**woodT PowaJlton / Backwith
v O O N E / 1(yow'^ r\u,iVa .. J Fayerttville :*
feufiMdj WLiiijmi f ;y_«i«h Ac»j,V, V. \ . § p'«'
nQu'fi'and
i#_Syi»#$tecrm -V KtncaidQ
I
_ Ojhio*
X
.6 Ooam
\\ '2 Ntv Cafawow^jo /
\ S^'. fer„, ^ ¦ <
v^WatobC^,} Wilsondaia • -*' ' f
( Oaco'la • Milbufn^T^—\
a3,on'- *?4R«'n«9*,-\ rff- lti.k 1"^'"^ — /, ®^^Mos$* ln9r*n,va,7^]UiW T it-'1 "Via u
7 T •-•¦< -• -. ¦ (,»N«%^oAy^c^
L N_ v 4, 1» : ^ Ottawa Q.ao6Wfbtt - rer-w'.-—sAmaaola ujn I . .^ ... ^r¥,*^« "¦'
sVC ¦' ^-T" ',,0— tfLifrv\p.ap'"h'
••• --w a A5 ' Ar &
—Vc-'V^" iv V ^— 3V_P..AheA-i N(!I?J '-UBafraM — —' ©sifclinaV A -,V WK*~ Thayat^t 5tt
yk\*.7:,r p,""y-j "''i''^- \ s,M,,r^io,1c,^ vrr
VM. [
V-, ^ ^
"H* ¦ /frV. * ~ LundaU
Chauncay S
G O', -< ,io°„ "¦^*rtn'n°'",in "i f
- - ^fT k'm f/-— - .
Figure 1. Location Map - Kanawha Valley Study Area
1-2
-------
Most of the pollution controls implemented to date have concen-
trated on removing or reducing releases of "conventional" pollutants
such as oxygen demanding substances in water and suspended particulates
in air. The Kanawha River now meets most applicable water quality
criteria most of the time. Sport fish are again present in the River.
Similar improvements in air quality have also been achieved, with
significant observed reductions of major ("criteria") air pollutants
(ie., particulate matter, carbon monoxide, sulfur dioxide, nitrogen
dioxide, lead, and ozone) within the Kanawha Valley Air Quality Control
Region.
Although much environmental progress has occurred and conditions
have generally improved, there remain several issues of concern to en-
vironmental agencies and the public which have not been fully addressed.
There remain today in the Valley nearly 200 industrial and manufactu-
ring facilities, about 20 of which are considered to be major sources
of toxic substances. Although most existing regulatory limits for
conventional pollutants are usually met, occasional violations do
occur, indicating the need for continued oversight and enforcement,
where warranted.
II. PROGRESS IN CONTROLLING POLLUTION
In July of 1986, EPA Region III prepared a draft report covering
existing environmental conditions and compliance with current environ-
mental laws within the Kanawha Valley, West Virginia. This compliance
status report was a follow-up to an August, 1984 report prepared by
EPA's National Enforcement Investigations Center (NEIC) entitled
"Overview of Environmental Pollution in the Kanawha Valley". Together,
these two reports evaluated the prevailing environmental conditions
in the Valley, identified pollutants of particular concern, and located
the major sources from which these pollutants were being emitted.
While these reports concluded that major improvements in environ-
mental quality were being made when "conventional" pollutants were
assessed, they also identified potentially significant concerns when
"non-conventional" pollutants (toxic substances and carcinogens) were
considered. It is largely as a result of these findings that EPA
decided to undertake the study of toxic pollution described later in
this Report. This Section will provide a synopsis of the findings
discussed in the July, 1986 compliance status report.
1-3
-------
Existing Environmental Programs and Compliance
As noted earlier, industrial development and population growth
in the Kanawha Valley sometimes resulted in environmental degradation.
Such degradation was common to large industrial centers throughout
the United States and was a major reason for the passage of new and
expanded environmental legislation in the 1960's and 1970's. As a
result of these control programs, major improvements in air and water
quality have been realized over the past 25 years.
For example, due to the development of National Ambient Air Quality
Standards (NAAQS), and State Implementation Plans, which outline
control strategies necessary to assure compliance with NAAQSs, present
air quality in the Valley is in compliance with all applicable primary
ambient air quality standards.
Similarly, as a result of permitting programs such as the
National Pollutant Discharge Elimination System (NPDES) under Section
402 of the Clean Water Act, water quality in the Kanawha River has
made a dramatic improvement from the gross pollution of the past. All
of the nineteen major facilities evaluated in EPA's 1986 compliance
report operate under approved NPDES permits, and sixteen are based on
Best Available Technology (BAT) limits.
Even though 100% compliance rates have not been achieved (ten
facilities were out of NPDES compliance at the time of the report),
water quality parameters necessary to support aquatic life have im-
proved. In addition, concentrations of arsenic, copper, cadmium,
cyanide, lead, silver, and zinc all declined significantly between
1970 and 1984.
However, certain water quality criteria are still exceeded (e.g.,
silver, arsenic, chloroform, carbon tetrachloride, dichloroethene, etc),
which indicates a continuing environmental problem. Also, even while
general conditions are improving, there are still major concerns.
One of the most significant recent findings has been high concen-
trations of dioxin, PCBs, and chlordane found in fish tissue near
Nitro, which prompted the State of West Virginia in March, 1986 to
issue a fish consumption advisory for a stretch of the lower Kanawha
River. These problems notwithstanding, when compared to the quality
of the River prior to the issuance of environmental regulations, there
has been significant improvement.
1-4
-------
Several environmental programs have been established which will
help to continue these improvements. The permit program established
under the Resource Conservation and Recovery Act (RCRA) helps to pro-
tect both groundwater and surface water from hazardous waste facilities
including landfills. Of the eleven facilities in the Valley required
to have a RCRA permit, only one facility is not in compliance with
RCRA interim status. The Comprehensive Environmental Response, Compe-
nsation, and Liability Act (CERCLA), amended in 1986, will play a
major part in enhancing environmental conditions in the Valley, where
100 potential hazardous dumpsites have been discovered.
Monitoring programs have also been established to assess compli-
ance with the various laws. For example, in addition to regular air
and water quality monitoring programs, five industries in the Valley
are also monitored for compliance with Toxic Substance Control Act
(TSCA) requirements concerning storage, use, or disposal of PCBs.
Likewise, in accordance with the Federal Insecticide, Fungicide, and
Rodenticide Act (FIFRA), eleven facilities in the Valley that produce
or hold pesticides for distribution are monitored for compliance with
EPA requirements. All facilities inspected at the time of the compli-
ance report were in compliance with TSCA and FIFRA regulations.
Table 1-1, taken from the 1986 compliance report, indicates the
compliance status of the major Kanawha Valley facilities with various
environmental programs (at the time of the report).
Current Issues
Recent environmental attention has focused on potentially toxic
chemical substances present in the environment in low concentrations,
for which few regulatory limits have been established. Various vola-
tile organic chemicals, some with known or suspected human health
effects at relatively low concentrations, are known to be emitted to
the atmosphere from area industries and have been detected in the
ambient air of the Valley. Collection of air monitoring data has
recently begun, but there are few Federal or State regulatory limits
for these chemicals from which a comparison can be made.
The two Kanawha Valley compliance reports noted above identified
fish tissue contamination by dioxin, PCBs, and chlordane. Up to 100
potential hazardous waste dumps were identified. Surface and ground-
water contamination by toxic substances was also suspected from the
data that had been reviewed, and releases of toxic airborne pollutants
were indicated. Although an estimated 10 percent of the population in
the Kanawha Valley depends on groundwater as a drinking water source,
the qualtiy of groundwater and possible threat of contamination is not
well known. All of these findings indicate potential health hazards
to residents in the Valley.
1-5
-------
TABLE 1-1
MULTI-MEDIA COMPLIANCE PROGRAM SUMMARY
AIR
WATER
RCRA
TSCA
FIFRA
FACILITY
(TSP & S02
Only)
4th Qtr
1986
IN
COMP.
OUT OF
COMP.
IN
COMP.
OUT OF
COMP.
IN
COMP.
OUT OF
COMP.
IN
COMP.
OUT OF
COMP.
IN
COMP.
OUT OF
COMP.
Allied Chemical,
Nitro
X
X
N/A
N/A
N/A
Appalachian Power,
St. Albans
X
X
N/A
N/A
N/A
Appalachian Power,
Glasgow
X
X
N/A
N/A
N/A
Coastal Tank Lines,
Nitro
N/A
X
N/A
N/A
N/A
Chemical Leaman,
Institute
N/A
X
N/A
N/A
N/A
CST/Fike Chemical (Artel)
Nitro (2 RCRA Permits)
N/A
X
X,X
N/A
X
Diamond Shamrock
(Occidental), Belle
N/A
X
N/A
N/A
N/A
DuPont,
Belle
X
X
X
N/A
X
Elkem Metals,
Alloy
X
X
N/A
N/A
N/A
FMC,
Nitro
X
X
X
N/A
N/A
FMC,
S. Charleston
X
X
X
N/A
X
Hatf ield-Henson,
Dunbar
N/A
X
N/A
N/A
N/A
Kinkaid,
Nitro
N/A
X
N/A
N/A
X
Mason & Dixon,
St. Albans
N/A
X
N/A
N/A
N/A
Monsanto,
Nitro
X
X
X
N/A
X
S. Charleston,
Sewage Treatment Plant
N/A
X
N/A
N/A
N/A
Union Carbide(Rhone-Polenc'
Institute (2 RCRA Permits)
X
X
X,X
X
X
Union Carbide, (2 RCRA
S. Charleston Permits)
X
X
X,X
X
X
Union Carbide, Tech. Ctr.
S. Charleston
X
N/A
X
N/A
N/A
1-6
-------
III. NEED TO EXAMINE TOXIC POLLUTANTS
Everyone, in all sections of the country, faces risks to their
health and welfare in everyday life. These risks come in many forms
and include risks from diet, smoking, driving, accidents, environmental
pollution, etc. When the Environmental Protection Agency was formed
in 1970, it initially set out to eliminate what Americans considered
to be the greatest environmental risks to their health and welfare.
These risks appeared in the form of "conventional" pollution such as
smog and sewage, which were choking the nation's air and waterways.
Traditionally, the largest and most visible national environmental
programs have addressed mainly risks from elevated pollutant levels in
the ambient environment^ (primarily outdoor air and surface waters).
Control of particulates and other relatively well understood
"criteria" air pollutants (sulfur dioxide, carbon monoxide, etc.) and
water parameters such as total suspended solids, biological oxygen
demand, and organic wastes, have enabled significant reductions of
risks posed by these gross and largely visible forms of pollution.
With the growth of environmental sciences and the refinement
of analytical techniques, however, environmental scientists and policy-
makers have discovered gaps in the programs that protect us against
environmental risks. Toxic and hazardous chemicals^ are present
thoughout the country, in our homes (paints, cleaners, etc.), schools
and buildings, outdoor environment, and work places. Unlike conven-
tional pollutants, which are more easily detectable and relatively
simple to clean up, these potentially toxic chemicals frequently pose
health threats in even small quantities. They also sometimes change
form or migrate, and can be very persistent in the environment.
1 The typical method of addressing these risks to the ambient
environment from specific point sources of pollution has been essen-
tially two fold: the first is to establish acceptable ambient
concentrations for particular pollutants and to calculate the pollu-
tion controls needed to attain emissions/discharge levels that
would meet the required ambient levels; the second method is to
prescribe specific control technology (many times the "best avail-
able technology") regardless of acceptable ambient concentrations.
Often the most stringent of the two approaches are used to control
air and water sources. Where risk from pollution is not caused by
a specific point source, other methods are used by EPA to control
these risks. Examples of this would be EPA requirements for
applicator training for use of pesticides, and requirements under
the Toxic Substances Control Act to test new chemicals before pro-
duction and distribution.
1-7
-------
Because of the pervasiveness and complexity of pollution from
toxic substances, and the limited resources of government and industry,
policymakers must try to determine which risks from which toxics may
be significant, and then reduce the worst and most controllable risks
first. However, the science of quantifying public risk from "environ-
mental" factors is very much in its infancy. Although the public and
scientists now widely accept a few risks as being well-defined (e.g.,
the risk of lung cancer from smoking), the impact of a large majority
of risks cannot be quantified with much certainty because of many
unknowns (e.g., the extent of an individual's actual exposure and
sensitivity to a given chemical).
Even defining which risks to include when considering "environ-
mental" risks in a specific geographic area raises many legitimate
questions. The long-term health effects of ambient air and water
pollutants have historically been logical candidates for inclusion in
risk studies. However, to be exhaustive in defining total environ-
mental risks, a study would have to address accidental releases of
chemicals from industrial plants, indoor air pollution, and non-cancer
health risks. And for a more comprehensive outlook on these risks,
the results of the study should be compared with the risks that workers
and residents in the study area are exposed to from smoking, poor
diet, hazardous occupations, individual accidents, and the like.
^ As used in this report, "toxic substances" are defined as
those chemical substances which pose a substantial potential or
actual hazard to human health or the environment when present even
in relatively small quantities or low concentrations. The term
"toxic substances" is often used indiscriminately to describe a
wide variety of chemical substances, many of which may not pose a
hazard but are perceived to pose a hazard. The toxicity of a given
chemical compound is dependent, in addition to its chemical proper-
ties, on the dosage received. For example, ordinary table salt can
become "toxic" in large enough dosages. Various references discussed
in this report use the term "toxic substances" in a broad context;
in such discussions, the term is used in the context explicity
expressed in the reference. Whenever possible, however, the report
uses specific teminology keyed to environmental laws to identify
the substances present in the environment. For example, when
discussing wastewater discharges or surface water or groundwater,
"priority pollutants" refers to the group of 129 organic and inorganic
chemical substances and heavy metals classified as toxic pollutants,
as defined by Section 307(a) of the Clean Water Act. "Hazardous
wastes", another term frequently misused, in this report refers to
specific types of industrial, municipal, and commercial solid or
liquid wastes that have been classified as hazardous wastes by EPA.
1-8
-------
The national environmental programs administered by the Environ-
mental Protection Agency in the past ten years have begun to also
address non-conventional pollutant risks, but only to a limited extent
(e.g., priority pollutants in drinking water, asbestos-in-schools,
radon). Some other risks that are faced by the public are regulated
by other government agencies such as the Occupational Safety and Health
Administration (workplace risks), Food and Drug Administration (food
intake risks), Department of Transportation (transportation related
risks) Consumer Product Safety Commission (product risks) and Federal
Emergency Management Agency (emergency actions).
Private industry also addresses these risks both because of the
need to comply with government regulations, and for other reasons
(e.g., protection of property from fire/explosions, potential liabi-
lity from injuries/illness caused, protection of their work force and
the public, and the need to be socially responsible "good neighbors").
Only in very recent years has EPA attempted to look beyond the
respective programs that Congress has mandated to consider more broadly
the wide variety of risks that we face. This so-called integrated
approach to managing environmental risks has taken a number of forms.
One approach has been to look at all major uses of a chemical (e.g., a
specific solvent) before regulating it, so that the major risks of the
chemical (and its potential substitutes) can be considered.
Another approach has been to identify environmental risks in a
particular geographic area so that the relative risks from various
sources can be compared and, if needed, control strategies for cost-
effective reductions adopted. These geographic studies to date have
largely focused on ambient air and water problems although other risks
(e.g., indoor air) are beginning to be included in some studies.
These studies typically have not included the consideration of risks
from accidental (sometimes called catastrophic) releases of toxic
pollutants nor have they included non-cancer health risks.
The Concepts of Risk Assessment and Risk Management
The concepts of risk assessment and risk management have been
endorsed by EPA during the past several years, and have played a large
role in developing the concept of integrated environmental management.
Since 1983, EPA has increasingly used risk assessment to identify and
measure toxic problems, and to compare them to one another to establish
priorities ("Risk Assessment and Management: Framework for Decision
Making," EPA, December, 1984).
1-9
-------
In essence, risk assessment allows one to evaluate the health
effects from exposures to toxic contaminants by estimating the risk
from that exposure. Quantitative estimates allow us to compare
risks from different chemicals, exposure pathways, and sources.
While risk assessment is rooted in scientific principles and
knowledge of toxicology, epidemiology, and environmental exposure,
it is inherently imprecise - necessary data are often limited, if
available at all. Consequently, the results of a risk assessment
depend heavily on assumptions about physical relationships and bio-
chemical transformations, among other factors.
Despite its uncertainties, risk assessment provides a useful
framework for comparing different environmental risks. As long as .
one under problems and establishing priorities for risk management.
In a 1983 speech to the National Academy of Sciences, former EPA
Administrator William Ruckelshaus distinguished between risk assess-
ment and risk management:
Scientists assess a risk to find out what the pro-
blems are. The process of deciding what to do about
the problems is risk management. The second procedure
involves a much broader array of disciplines, and is
aimed toward a decision about control. Risk management
assumes we have assessed the health risks of a suspect
chemical. We must then factor in its benefits, the
costs of the various methods available for its control,
and the statutory framework for decision.
Risk Assessment and Risk Management Help Set Priorities
There are thousands of chemicals present in commerce today, and
an unknown number of contaminants and unintended by-products. Scien-
tific evidence suggests that some of these chemicals can cause cancer
and other diseases. We are exposed to a complex mixture of these
chemicals through air, water, and food. These chemicals are proper
targets for Federal, state, or local regulation if they pose signifi-
cant risks to health or the environment.
Regulatory priorities are often set in response to public pres-
sures to address particular environmental issues of the day, rather
than on the basis of systematic analyses of risks, costs and benefits.
This would not be a problem if the issues of greatest public concern
also posed the greatest risk. At times, however, public perception
and actual risk diverge widely. We cannot know which problems are
the most serious in terms of public health without some kind of
rational, systematic evaluation of the issues. Risk assessments
help identify and set priorities among those pollutants posing signi-
ficant risks.
1-10
-------
Some priorities will always be set by Congress or state legisla-
tures, or through the press of emergencies. However, a consistent
analytic approach is helpful in setting an agenda for effective risk
management. Analyzing alternative strategies for reducing risk
provides a useful basis for such an approach. Time and resources are
best spent controlling those toxic chemicals to which people are most
exposed and for which practical controls are possible. It is less
productive to spend limited resources regulating chemicals, even highly
toxic ones, to which no one will ever be exposed, or for which there
is no capability to impose additional controls.
Given the level of public anxiety, extreme points of view often
emerge when regulatory actions may involve substantial health or
economic impacts in the face of scientific uncertainty. This situa-
tion may polarize debate and bring public policy to an impasse.
Consistent use of risk assessment and risk management can assist by
explaining the scientific basis for the risk estimates, including
the confidence we have in such numbers; by placing the risk reduction
expected from control measures in context with other risks and oppor-
tunities for their reduction; and by explaining the broader social
values which affect risk management.
Moreover, communication of policies based on well articulated
scientific principles is perhaps the most important element in crea-
ting a strong base of public understanding of environmental control
programs. Enhancing efficiency through risk management means exami-
ning the regulatory options and selecting those that reduce risk the
most for any given level of resources. The risk assessment analysis
presented in this report is intended to provide a better understanding
of some of the issues associated with toxic pollution in the Kanawha
Valley and give decision makers a sense for the relative priority
of these issues.
Risks in the Kanawha Valley
The types of potential environmental risks that exist in the
Kanawha Valley are not considered to be significantly different
from risks in other industrialized areas. They involve potential
air and water pollution problems, indoor air pollution, short-term
air or water releases, etc. Although the types of potential environ-
mental risks are not considered different from other industrialized
areas, the potential degree of risk has been questioned because of
the high concentration of major chemical facilities, the geography
and meteorology of the Valley, and the proximity of the residents.
1-11
-------
Though the total number of industrial facilities in the Valley
has declined somewhat in recent years, there remain nearly 200 of
these facilities today, many of which process, handle, or dispose of
a wide variety of chemical compounds. This geographic cluster of
chemical/ industrial complexes is thought to represent one of the
five most concentrated industrial areas in the United States. Given
that many of these sources are located in close proximity to residen-
tial areas, Valley residents and officials have long been concerned
with the possible health implications associated with long-term
exposure to these chemicals. These environmental consequences are
often more subtle in their effect and require more sophisticated
approaches to analysis and control than the pollution problems of
the past.
Previous environmental studies of the Kanawha Valley have pointed
to a need for further evaluation of potential health effects from
community exposures to these potentially toxic substances.
A limited 1981 pilot geographic study of Kanawha Valley performed
under contract by EPA suggested that several environmental toxicants
might be present at ambient levels which would warrant further investi-
gation. In 1982, the West Virginia Department of Health conducted a
brief comparison of cancer mortality rates and found that the incidence
of malignant neoplasms (tumors) in a North Charleston area exceeded
that which would be expected from national cancer mortality data.
This brief analysis did not attempt to examine possible causitive
factors, but did suggest the need for further study.
A 1984 report by EPA's National Enforcement Investigations Center
identified several possible sources of toxic contaminants within Kanawha
Valley, and concluded that while many sources, and sometimes specific
chemicals and quantities released, were known, information was lacking
on the relative threats these substances posed based on population expo-
sure and inherent toxicity. The report also pointed to difficulties in
assessing the impact of toxics in the absence of established standards
or guidelines for many of the substances.
Though numerous studies of the Valley have been completed
indicating the presence of a variety of chemical compounds, none has
attempted to link ambient concentrations with human exposure and
health effects.
EPA recognizes that the expense of quantifying all risks from
all pollutants in the Kanawha Valley would be prohibitive, and that
our ability to quantify these risks using even state-of-the-art
techniques is very limited. However, a screening of the relative
severity of some of these risks could be provided as an initial
step in characterizing some environmental risks in the Valley.
1-12
-------
IV. EPA's GOALS AND OBJECTIVES FOR KANAWHA VALLEY
The mission of the Environmental Protection Agency is to improve
and protect the condition of the environment - both to reduce human
health risks to acceptable levels and to protect and enhance the quality
of natural ecosystems. Consistent with that mission, and as a result
of the above Kanawha Valley background, EPA's Region III Office in
Philadelphia outlined three broad goals for its involvement in the
Kanawha Valley. These goals are:
1. To ensure compliance with existing Federal environmental laws,
and to enforce requirements of those laws as necessary;
2. To determine the extent and sources of toxic pollutants in the
Valley, and to estimate the risks to public health and the envi-
ronment from exposure to these pollutants.
3. To reduce to acceptable levels the risks to public health and
the environment from toxic pollutants, in the most effective
way possible.
Many public and private organizations, in addition to EPA, are ac-
tive in attempting to meet the above goals. And there are many activi-
ties underway and planned to assist in this endeavor. Many of these
efforts involve analytic approaches and techniques which are on the
forefront of environmental health protection.
The major purpose of this Report is to describe EPA's and West Vir-
ginia's initial efforts to date in meeting the above goals, primarily
Goal No.2, and to provide insights into activities of other organizations
who are also addressing in some fashion environmental health risks
in the Kanawha Valley.
V. REPORT ORGANIZATION
Most of this Report (Chapters Two through Seven) is devoted to a
detailed discussion of a limited screening study of selected toxic
compounds recently completed by EPA and West Virginia environmental
agencies. The Kanawha Valley Toxics Screening Study, though limited,
is a first step in attempting to roughly assess and compare potential
health risks from chemical exposures in the Kanawha Valley.
1-13
-------
Since this Screening Study represents only a very limited perspec-
tive on a few selected toxics issues, we have attempted to describe in
summary form some other environmental issues and activities of EPA and
other organizations in Appendix A of the Report. Other Report Appen-
dices provide information indirectly related or supporting the body of
the main Report.
Chapter Two, "Kanawha Valley Toxics Screening Study - Framework",
describes general information about the Study, including reasons for
the Study, analytical study scope, specific objectives, study area,
and management organization and communications.
Chapter Three, "General Methodology", describes the risk assessment
screening methodology used to analyze all issues studied on a common
basis. More importantly, it presents limitations and uncertainties in
scope, exposure data, and toxicological data which have a profound
bearing on the conclusions which can be drawn.
Chapters Four through Seven present detailed technical reports for
air, drinking water, surface water, and hazardous waste. Each of these
independent reports provides an analysis of several chemicals and possi-
ble human health risks associated with routine community exposures. A
list of technical support documents is provided at the end of each of
these Chapters. Copies of the support documents are available upon
request from EPA's Region III Office.
Chapter Eight, "Putting Risk in Context", presents a brief discus-
sion and some comparative tables which attempt to relate the Kanawha
Valley risk estimates with similar estimates which have been calculated
for other geographic areas or types of risks.
1-14
-------
Chapter Two
Kanawha Valley Toxics Screening Study — Framework
-------
CHAPTER TWO
KANAWHA VALLEY TOXICS SCREENING STUDY - FRAMEWORK
I. BACKGROUND
As noted in Chapter One, the Kanawha Valley has been the subject
of environmental concern because of a combination of industrial facili-
ties' location, nearby population, topography, and meteorologic conditions.
Several previous studies of the Valley have indicated a need to more
fully evaluate the health effects from chemicals released or present in
the Valley. A 1984 study by EPA's National Enforcement Investigations
Center stated:
Air and water quality defined by traditional pollutant para-
meters (such as dissolved oxygen in water and suspended parti-
culates in air) are generally better than applicable regulatory
limits. However, some regulatory limits are occasionally not
met indicating the need for further controls on traditional
pollutants. Of more concern, however, is the presence of va-
rious potentially toxic chemical substances in the environment
for which no regulatory limits have been established.
The Bhopal, India tradegy in December, 1984 further heightened
public, industry, and government concern, both nationally and in the
Kanawha Valley, over the issue of chemical plant safety and toxic air
emissions. Subsequent Congressional inquiries into Bhopal and the U.S.
chemical industry focused attention on Kanawha Valley due to its concen-
tration of chemical facilities, and the presence of a production unit
(MIC) similar to that connected with the Bhopal accident.
The increased public concern and Congressional attention over toxic
air emissions also accelerated EPA's development of a National Air Toxics
Strategy which was announced on June 4, 1985. The strategy included two
major components:
1. A plan to address sudden, accidental air releases of chemicals
which could cause immediate (acute), often life-threatening health
effects; and
2. A plan to address routine releases of chemicals, normally over
a prolonged period of time, where long-term health effects were of
primary concern.
The first component included a variety of recommended actions designed
to prevent, or respond, if necessary, to an accidental release of chemicals.
Some of these actions are described in Appendix A. The second component
included several recommended actions to help deal with the more subtle
health risks which can occur from continual low level chemical exposures.
The second component became the focus of the Kanawha Valley Toxics Screening
Study described herein.
2-1
-------
Concurrent with development of the second component of the Air
Toxics Strategy, EPA had been evaluating options for how to address
the above mentioned issues of long-standing concern in Kanawha Valley,
primarily the long term threat of public exposure to air toxics. In
addition, the West Virginia Air Pollution Control Commission had reques-
ted assistance from EPA in identifying and prioritizing volatile organic
chemicals emitted in the Valley that could pose a significant health
risk.
After weighing several options, the EPA decided that because of
uncertainty as to the potential magnitude of the perceived problems,
that, it would be prudent to first try to better define some air toxic
issues in the Valley by performing a brief, limited scoping, or "screen-
ing level" exercise, in order to gain a rough understanding of the
nature and potential seriousness of some toxics issues. This screening
level assessment would use available health effects data, exposure
models, and risk assessment techniques developed within EPA. It would
provide an initial appraisal of whether some problems may exist, and
what, if anything, should be done next. Explicit in this decision was
an understanding that we would not be able to do a comprehensive study,
or address, even on a screening level, all areas of environmental risk
in the Kanawha Valley.
It was agreed that this screening study should rely primarily on
existing data, and that the study would attempt to analyze multi-media
exposure pathways, recognizing that the air pathway was of greatest
interest, and should be the focus of study.
Based on these assumptions, State environmental officials were
briefed on the study concept and agreed to participate. A workplan
was developed, and the screening study began in the summer of 1985.
H. STUDY OBJECTIVES
Study objectives were shaped by previous Kanawha Valley studies'
recommendations, State advice, available analytical methods, budget
and resource constraints, and compatability with other ongoing
environmental activities in the Kanawha Valley. The screening study
is designed to begin to address one of EPA's three broad goals for
Kanawha Valley to identify the extent and causes of toxic pollution in
the Valley.
The overall objective for the screening study is to provide a
general sense of the nature and relative priority of some ambient
toxics pollution risks in the Valley, for the purpose of helping EPA
and other agencies to make decisions about the need for further study
and control of toxics in the Valley.
2-2
-------
Specific objectives of the Toxics Study included the following:
1. Using available data bases, identify many of the toxic
chemicals routinely released or present within
various human exposure pathways within the Kanawha
Valley.
2. Develop a sense of potential public health concerns
which these toxic pollutants may pose in various path-
ways, based on health effects and exposure information.
3. For a select number of pollutants, provide an initial,
conservative assessment of the potential cancer risks,
and other potential non-cancer health risks where possible,
from predicted or observed concentrations within exposure
pathways.
4. Identify data and information gaps, and outline needs
and options for future study direction to enable a more
detailed investigation of health issues where warranted.
III. ANALYTICAL STUDY SCOPE
Multi-media screening study should not be confused with "compre-
hensive" problem analysis - the Kanawha Valley Toxics Screening Study
is not a study of all potential toxic pollution in the Kanawha Valley.
Although the Kanawha Valley Study considers a variety of issues
and exposure pathways, it does not examine risks from all pollutants.
We decided to focus on "toxic" pollutants (i.e. those capable of
causing adverse health effects at relatively low level of exposure)
because such pollutants are relatively less studied and less controlled
then more 'conventional' pollutants, such as ozone precursors in air
and oxygen-depleting substances in water. Conventional pollutants may
also affect health, although generally not at the low levels at which
more toxic pollutants may be harmful.
We did not evaluate all toxic pollutants in the Kanawha Valley.
We decided which pollutants to study largely on the basis of their
known presence in the environment and available health effects data
(see the discussion on pollutant selection in Chapters Four through
Seven). The study did not attempt to identify the "most dangerous" or
"highest risk" chemicals in the Valley. We attempted to identify and
study those toxic pollutants which we believed may represent some
substantive health risks in the Kanawha Valley. Nevertheless, it is
likely that some of the pollutants we are not studying because of a
lack of data concerning their presence or toxicity may also represent
some potentially significant health risks.
2-3
-------
It was recognized from the outset that all types of environmental
risks could not be addressed. With limited funding and reliance on
existing data, it was decided that the screening study could best meet
objectives by focusing on potential long-term health risks from current
ambient exposures. This necessarily excluded other types of risk,
such as those from accidental releases of chemicals, from voluntary
intakes (such as cigarette smoking and diet), and from exposures to
indoor residential air, occupational chemicals, or toxic residues in
food.
A major environmental concern among Valley residents is the risk
posed by a sudden, accidental release of chemicals. There is extensive
work continuing in this area by organizations like the Kanawha Valley
Emergency Planning Council, the National Institute for Chemical Studies,
and the chemical industry itself. We did not feel that we could measu-
reably improve on these efforts or contribute to a better understanding
of the issue within our limited study focus. Nevertheless, accident
prevention and emergency response capabilities are recognized as very
important issues which demand continuing attention and vigilance.
The Toxics Screening Study examined potential public exposure to
several toxic pollutants through four routes by which contact could
conceivably occur: through ambient (outdoor) air, consumption of
public drinking water, fish consumption from the lower Kanawha River,
and consumption of groundwater which could be contaminated by hazardous
waste sites.
In the Air Analysis, twenty (20) known or suspected cancer-causing
(carcinogenic) chemicals (14 from point sources) were evaluated out
of more than 450 chemicals listed on an inventory of Kanawha Valley
air emissions prepared by the West Virginia Air Pollution Control
Commission. Potential community exposure to these 20 chemicals was
estimated through mathematical modeling within four distinct geographic
zones: Belle, Charleston/South Charleston, Institute, and Nitro.
For Drinking Water, nineteen (19) chemicals for which monitoring
is required were evaluated from the study area's fourteen public
water supply systems. Both potential cancer risks and non-cancer
effects were evaluated. Private wells and non-community supplies were
not included.
The Surface Water analysis focused on potential health effects,
both cancer and non-cancer, from the possible consumption of contami-
nated fish taken from the Kanawha River near Nitro. In addition,
concentrations of chemicals monitored within the River were compared
with water quality criteria established for the protection of human
health and aquatic life.
2-4
-------
The Hazardous Waste analysis examined data from known hazardous
waste sites in the Kanawha Valley in order to devise hypothetical
hazardous waste settings (i.e., combinations of various facility types,
hydrogeology, and waste constituents) which could present a risk to
human health through public consumption of contaminated ground water.
It did not include any site-specific health risk estimates.
IV* description OF THE STUDY AREA
The Kanawha Valley is the narrow, winding Valley of the Kanawha
River in west-central West Virginia surrounding the capitol city, .
Charleston. The Kanawha River traverses the western foothills of the
Appalachian Mountains for 97 miles from its origin at Gauley Bridge
at the confluence of the New and Gauley Rivers to its confluence with
the Ohio River at Point Pleasant, West Virginia, northwest of Charleston.
The general study area for the Toxics Screening Study encompassed a
roughly 60-mile section of the Valley between Alloy and Windfield Dam
(see Figure 2-1). The study area boundaries were selected to encompass
major possible sources of toxics pollution and the population most
likely impacted by this pollution.
Each media analysis - that is, those for air, drinking water,
surface water, and hazardous waste - included a somewhat different
geographic focus, although all were within the general Kanawha Valley
study area. For example, the air analysis focused on four major indus-
trial clusters at Belle, Charleston/South Charleston, Institute, and
Nitro. Figures 2-2 and 2-3 indicate the geographic coverage of the
first three media analyses. For the hazardous waste analysis, only
hypothetical scenarios were constructed, and so no particular geographic
area is shown.
The total population of the general study area is about 200,000
people. The four zones covered by the air analysis include approximately
100,000 people, while the drinking water anaylsis covered approximately
180,000 residents. Population densities are relatively low in the
upper and lower thirds of the Valley with most of the population concen-
trated in Charleston and adjacent communities in the central portion
of the Valley.
Because the Kanawha River traverses mountainous terrain, the
Valley is relatively narrow, reaching a maximum width of only about 1
mile in the study area. The area of the valley floor in the study
area is less than 50 sq. mi., much of which is developed for urban,
industrial, or residential uses. The elevations of flanking mountains
range from 300 to 1,300 feet above the valley floor. This particular
topography sometimes act to trap air pollutants from industrial and
municipal sources in the valley in close proximity to populated areas.
2-5
-------
FIGURE 2-1 LOCATION MAP - KANAWHA VALLEY STUDY AREA
K>
I
to* iufiifitm • ~
? Landing «.
£ifhte«
Confld«nc«o Parad
C^\Bancr
urfican*
fy //CM °9
' I
Swaziland
rjss^sr\i T~\cb
\ J) \_V^n° 7 \
Xdven/o Vj
oGaldtowb /\
'Walton
VCIcaron*
OS)
Sorfiance? _,,.ro
S ^y 73I
V\w , /J S . / I ^^—^fallint Rock \Quaan 7*.
^ o Rocky r /> jf-* ( <2 Ishoals o $£"«t
Fork I ri^vlawfrffl-tftlua Crefk '
; OcmmTjJ
Ot»«r°
o Kettle rA
Wallbaek
Ivydale
„ Cotton. 0
:^X)505(m)7
_ Maysel
Proclous o
itr„ °Cros%Lanes iEn, (/jpinch«\ ,„
oGuthrlyHills ^^„ *
- oco
O^Vtt* 4! Alum
%ICreek'
Rrlfftthavllljg/o Sod
Big Chimney
/is?
^MhAanc>
V^V^Tad'^\_-
o Rensfolj
iCoal Fork
2 Mala
STUDY AREA
o Lizemor
Buffalo,
c,ayV 015
Hartl
3 Bickmbre
«i*e
'f\ bca^.ffii v"fsn^~^7yh"2i
101 02v J ^ ^
.Julian
¦^Palarmo
Cimn-'
>Laat
rellsbura
.Ridgavic
,
Peyton
Aihto
Racine
/Rand \»'6<>>
A
k)3!
aBelle
Chelyan
Inay \}BlaK€le4 ^^"Greenda^e
/p.™» Bentre* *T
. .2-Gap . - -a ^aughmn
j Mimmor ^
N
p"
arori
Dawes3
fbedar 6rov
Pratt
Hamtey _
Montgome?
ro
fSmitheri
FerrisH
rjVanetta
Suley Budge
.Ansted
Foatar
5 Comfort (
Madison
005
N
) Or|a>«
lOhley
?Eskdal
V-eewood
apeeo»»Bter
Becty/itf
alton
Rurnwel
&Unaada o
Williams
Mountal"-^
? Acifie
< 1
r« "*H Moun,r)^ / Y vjag
Greenview^^Gord^ ^WN,.,^^^^ Klng^t
learfs Fo/k ^ Feyette
Page (F8y?ttBV,% ^
iKIncaidi, P" \& SV F"
;ot./\M^.„w\^^w^)n Q19
CunBrdo_)j
^pOak Hill
") MinH»n
d Poa C
c
«a^
ORar isey
:iirtyO
Hlco
Russ
Lookc
Scale 1:500,000 i—i I—I i—i Fr~F=r
10 0 10 20 30 40 Miles
1 inch equals approximately 8 miles 10 0 10 20 30 *0 50 Kilometers
ffTTHMMI I I— I I 5
-------
, frjac* CnH>n
//i
I A \ . . * Z-.ws«#.. J
KANAWHA VALLEY TOXICS SCREENING STUDY
Location of Study Areas
Or
NORTH
tCJMKI
¦?U*t
0«« Fj'Ml
»moy Cit* y
I Lanham
'* J9-'
Msekr Fwk
WCfOfT
«so»
S* Srt.«l ^
r«timt ftcc*.
««r ^
Ml MM MM - /
,/0*J
iQ^M" aV»<»
A%- JZL
"» ='•" /ur
.rt'rtIS foiaT?.11"" Mil>
Pmeh
/ CvUi
ittiRya ,-.T ' y ¦ "* *,<*¦
' 'Vl"»« ^ • *\
-»k? iSss-' S£
KSMtClllfl* p
Bit Chi*""
M>fli'ha£
amt Aloant
S
5»»* y«
Quiet" Cotlr
Uta>«n
Stndtr»o*
Mitfsra re
riy#mtU
¦/
fctrarrla. /o«a
SmTHiH
r;ro ft
rWiton „ ^ .
_ Kan**na C>u
SowBi r»f«
pring F«*b
J. SewiUr \
HlMUl \
\ i t-
\ -
?Otl AflAWI
kiar'«:i S«A?
AJumC
H.t»e
Font Ojd
¦ U«|0
5nlti*»«ilta
So'coi
MolUfturU
Owtsn «v €wMt>«*OnR)>
¦*V. ¦¦• •- "w.n »y«cC*r*'>
^ * V* '•** Sctoat *
¦ fli l«rf M|T«H
9mm
for* "
CnnUivtMHi ^ Srwsr
QnoD*
u*0 ' n) .
PUBLIC DRINKING WATER DISTRICTS
,«"V
jS
Brand
Lift G«vt*y ln«a«
nojrj
.
Ptyt®«
Glanhittt*
Martm
Fottar
Hownns
Nucuil
auula 1% qkrilt*
WfcitUfc'r
Lt*«ootf
Kollr G'*<"c»t
JOftMO*
u/n«rtli '
MacOi^n/
wr
OutK Mmmh
KMOJ
Q—rnt*
kcm* lUing
yt CoK>nvMl«*
/655 Wtlliamt
Mountain
North Carben \»Manan
'TTt".
Qumtand
Camao ml'
V*
S'i'
*(t . «£»
j «:'¦/ . ,-y>
•d®»" .• •» -
' •••' :
^CT Tartu Cim .•
FIGURE 2-2. STUDY AREA - DRINKING WATER ANALYSIS
2-7
-------
!s«-
¦vV
mt ,'A> \ \ •
' rl
*«/
> «.
.Charleston Air Analysis
I - *•' ji f\ ....
f« -.¦•¦ . Jzr t*. • r\2K ^ • >^X vT- »•—
r^zssvc-fSWjt c ¦ • t: •
, |M«>ai • - mt . t-,1 ViT^ t' iw^ J . / J. mm
Belle Air Analysis ;'
ul %
Mf
•¦• •" i
, !
J=?A . /T i
'4W!48 */
•r^V
• ' » '.«»
(/a
(-i
ux ?..
• *
-------
Streamflow in the Kanawha River can be highly variable ranging
from flooding conditions in late winter and spring to low flows in
late summer and fall. Extreme low flows range from about 600 cfs near
Belle to about 1,500 cfs at Windfield Dam. This contrasts with average
flows of about 11,000 cfs near Belle and 15,000 cfs at Windfield Dam.
These flows are influenced by releases from upstream storage resevoirs,
power generation, and the effects of the navigation dams in the study
area. The Kanawha River is navigable throughout the study area with
slack water provided by London, Marmet, and Windfield Dams. Major
tributaries that join the Kanawha River within the study area are the
Elk, Coal, and Pocatalico Rivers.
Public water supplies in the study area are obtained entirely
from surface water. A major portion of the study area (about 180,000
population) is in the Kanawha Valley District of the West Virginia
Water Company that obtains water from the Elk River. This system
serves most of the Valley between Belle and Nitro. There are eight
small water supply systems serving a population of about 20,000 in the
upper valley upstream of Belle. These systems obtain water from the
Kanawha River upstream of all major chemical manufacturing facilities.
St. Albans Water Department serves about 20,000 persons in St. Albans
with water obtained from the Coal River. St. Albans has an auxiliary
intake near the mouth of the Coal River in backwater from the Kanawha
River for use in times of extremely low streamflow in the Coal River.
There are no longer any public water supplies using Kanawha River
water downstream from Belle. The Nitro area was formerly served by a
West Virginia Water Company facility using Kanawha River water. Severe
taste and odor problems necessitated a high level of water treatment
including activated carbon columns. This plant has been closed and
service is now provided from Elk River sources.
The Kanawha River is used by the various large industrial facili-
ties for cooling water and some process water. Some industrial use of
ground water may also be occuring. Groundwater is also used for private
water supplies for scattered residences and small commercial facilities,
mostly in areas away from the main Valley floor.
V. INSTITUTIONAL RELATIONSHIPS
At the outset of the study, it was clear that numerous offices
within EPA and the State of West Virginia, as well as many other
organzations, would have an interest in the issues surrounding the
study. While a full public participation program was not established,
EPA's Region III Office in Philadelphia took responsibility for
communicating with a variety of groups about the nature and scope
of the project as it progressed. These groups included local mayors
2-9
-------
and officials, the National Institute for Chemical Studies, the Kanawha
Valley Emergency Planning Council, environmental and industrial repre-
sentatives. Each of these groups provided valuable perspectives,
advice, and feedback that greatly helped to improve the study.
To manage the study itself, two intergovernmental committees were
established. The first was a Policy Committee comprised of the EPA
Region III Administrator, West Virginia Governor's Office, and the
Directors of three West Virginia State Agencies: Air Pollution Control
Commission (APCC), Department of Natural Resources (DNR) and Department
of Health (DOH). The Policy Committee provided overall guidance for
study direction, approved topics for study, and reviewed and evaluated
major issues and findings of the study. They also retained ultimate
authority for all policy issues, and made all decisions regarding
allocation of their respective agencies' staff time and resources.
A second group, the Implementation Committee, was composed of
EPA's Region III Environmental Services Division Director and West
Virginia State Program representatives. The Implementation Committee
was primarily responsible for defining the technical scope of the
study, for guiding specific technical components, overseeing the
analyses performed, arranging adequate staff time and data necessary
as input to the analyses, assuring the quality of data used, and eva-
luating key issues and questions as they arose. The Implementation
Committee kept the Policy Committee informed of project status and all
areas of deliberation, making periodic recommendations to the Policy
Committee on courses of action for unresolved issues.
Major technical support and analyses for the study was provided
by EPA's Office of Policy Analysis (OPA) in Washington, D.C. In addi-
tion, specific workgroups for each media area of analysis (i.e., air,
water, and hazardous waste) were established composed of senior staff
from EPA and West Virginia programs. These groups monitored progress
in each media area and provided information concerning ongoing program
activities that could affect issues of the Toxics Study. Also, several
offices within EPA contributed heavily to study direction and content.
These included the Office of Air Quality Planning and Standards (OAQPS),
in North Carolina, and the Office of Research and Development (ORD).
2-10
-------
KANAWHA VALLEY INTEGRATED TOXICS STUDY
MANAGEMENT ORGANIZATION
Policy Committee
Regional Administrator,
Governor's Office,
State Agency Directors
HE
N>
I
H
H
Implementation Committee
EPA and State Program
Water
Sub-Group
Hazardous Waste
Sub-Group
^ Headquarters Technical and Methods Assistance ^
-------
Chapter Three
General Methodology
-------
CHAPTER THREE
GENERAL METHODOLOGY
The purpose of the Kanawha Valley Toxics Screening Study was to
compare potential health risks attributable to different sources (e.g.,
chemical plants, runoff, hazardous waste sites), pollutants (organics,
metals, etc.), and exposure pathways (air, drinking water, surface water,
and ground water) in order to set research and management priorities.
To make those comparisons, we needed to first estimate the risks quanti-
tatively, where possible. This section discusses our methods of estima-
ting and comparing risks. These methods are generally consistent with
the approach to quantitative risk assessment for toxic air and drinking
water contaminants that EPA has used in other recent studies.
I. LIMITATIONS AND UNCERTAINTIES
An understanding of the uncertainties and limitations that underlie
the Kanawha Valley Toxics Screening Study is critical to a proper inter-
pretation of its results. Limitations in the scope of what was studied,
as previously discussed, as well as scientific and technical uncertainties
in both the exposure and toxicological data (see below), caution against
using the results as absolute health risk estimates. Nevertheless, deci-
sionmakers must often act to protect against health threats from toxic
chemicals without always having absolute scientific certainty. The Kanawha
Valley analyses used the best information and techniques available today
to estimate potential health risks from toxics so that decisionmakers will
be as informed as much as possible.
Several general points concerning the limitations and uncertainties
in our risk assessment deserve consideration. First, the risk estimates
are based primarily on existing knowledge about pollutant releases, am-
bient conditions, and chemical toxicity; these data vary widely in quality
and are almost always incomplete. Second, the exposure estimates incor-
porate a series of simplifying assumptions; although these assumptions
are necessary, they remain open to question and may be controversial.
Third, the potency estimates are necessarily based on current knowledge
of the toxicological effects of various substances. Considerable contro-
versy exists about the degree of hazard posed by different pollutants,
and about whether some are hazardous at all. Finally, resource and time
constraints and the breadth of our focus prevent us from analyzing indivi-
dual issues in as much depth as might be possible. We have attempted to
strike a balance between the desire for exhaustive and definitive analysis
and the need for timely results at a reasonable cost.
3-1
-------
To assure that our data, assumptions and analytic methods are as
good as they can be, we have sought and will continue to solicit review
from a range of knowledgeable sources: EPA regional and headquarters
staff, Science Advisory Board reviews, state and local government offi-
cials, industry representatives, university faculty, and members of
public interest groups. Our risk, assessment procedures must be seen in
light of their basic objective: to allow us to compare one risk to ano-
ther, rather than to make definitive statements about the absolute risk
posed by a particular substance, pollution source, or exposure pathway.
Limitations in Scope
It is important to realize that our analysis does not directly exa-
mine disease incidence in the local population and attempt to link it
with environmental exposure. For many of the exposures and health effects
with which we are concerned, such epidemiologic study is difficult because
background incidence is high, exposures to particular individuals are
difficult to quantify, and a variety of factors—genetics, occupational
exposures, food, exposures from other places, and many others—might be
the cause of an observed effect. Also, many effects may take years to
manifest themselves, and would not be evident in the current population.
Because it is not an epidemiologic study, the risk assessment metho-
dology is not intended to and does not answer questions such as what
caused a statistically higher rate of cancer or birth defects in a
particular neighborhood. Instead, it attempts to evaluate what health
effects might result from current and future environmental exposures.
While different in approach and interpretation, risk assessments
and epidemiologic work are complementary. Risk assessment can help to
identify populations and geographic areas that appear to be at risk and
that therefore might be appropriate subjects for an in-depth epidemiolo-
gic study. Epidemiologic analysis increases scientific understanding
of the relationship between exposure and health effects, thereby streng-
thening the basis of risk assessment. Epidemiologic analysis may also,
in some cases, be useful for confirming specific hypotheses suggested by
risk assessment.
This analysis does not attempt to estimate the health risks from
all chemicals or types of risks that individuals may be exposed to in
their daily lives. We chose not to analyze exposure to and risks from
conventional pollutants in air and water (such as ozone and oxides of
nitrogen and sulfur in air, and oxygen-depleting substances and oil and
grease in water) because we believed we could make a more significant
contribution by concentrating on less well understood and less regulated
toxic chemicals (largely organic chemicals thought to be potentially
hazardous at low levels).
3-2
-------
Finally, the study did not estimate risks from possible infre-
quent, accidental releases of toxic chemicals, such as a major release
of a toxic gas. The probability and magnitude of such an accident is
very difficult to estimate, and the likely risk from such an event is
therefore difficult to quantify. (Appendix A does discuss some current
activities related to prevention and response to such accidents). The
omission of such events from this analysis does not imply that possible
accidents are not an important environmental and public health concern.
The uncertainty in the risk assessment assumptions is great enough
that results should be considered rough indicators of the probable
magnitude of effects, not as precise, site-specific predictions of
effects.
Limitations in Exposure Data
Beyond these intentional limitations in scope, the study's expo-
sure and toxicological estimates are uncertain in a number of potenti-
ally important ways. For exposure assessment, one limitation of the
analysis is that it did not exhaustively examine all sources and
pollutants. While the study has tried to identify and assess risks
from many significant sources and pollutants, it was unable to estimate
exposure to some chemicals because of a lack of data. However, we
recognize that chemicals not included in the study may also pose some
health risk.
Even where exposure data were available, those data varied signifi-
cantly in their quality. Thus, the resulting exposure estimates vary
in their reliability. Those based on extensive field monitoring, such
as for trihalomethanes and inorganic substances in drinking water,
are the best types of exposure estimates we have. Those based on less
extensive monitoring, such as for fish tissue analysis, are somewhat
more uncertain.
Exposure estimates derived from modeling also vary in their relia-
bility. Estimates of exposure to toxic organic chemicals in air, cal-
culated using a dispersion model, are primarily dependent on the quality
of the emissions estimates and other factors such as meteorological
data. In general, the uncertainties introduced by the exposure data
are usuailly much smaller than those associated with the dose-response
information used to estimate human health risks. The analysis of
future risks from groundwater contamination, which relies heavily on
engineering assumptions and modeling of future events, is more uncertain.
Where there are significant uncertainties, such as in the groundwater
exposure analysis, we have attempted to make assumptions that are
likely to err on the side of overestimating possible health impacts.
3-3
-------
Limitations in Toxicology Data
Estimates of the potential health effects of particular chemicals
are designed to be conservative (i.e., more likely to overestimate toxic
health risks than to underestimate them) in several ways. When evalu-
ating potential health hazards from a chemical, EPA scientists assume
that health effects observed in laboratory animals are a reasonable indi-
cator of potential effects in humans. In converting the animal data
to predicted human responses, and in extrapolating from high doses to
low doses, EPA uses models that yield a plausible upper-bound estimate
of potency rather than a "best guess" estimate.
Many substances of potential concern have never been evaluated
scientifically, or have not been evaluated in sufficient detail to
allow estimation of specific effects on humans. For example, the
element lead (present in air, water, and dust) is thought to pose a
health risk to children at low ambient levels; currently, however, EPA
has no accepted Agency method for quantifying individual risks or
numbers of possible cases from non-carcinogens, although much work on
this issue is in progress.
EPA is likely to be aware of the dangers from many of the most
potent chemicals, since the evidence for their toxicity will typically
be the most obvious; however, it is possible that some chemicals about
which we currently know little may someday be demonstrated to be toxic.
In addition, EPA currently has very little information on possible
synergistic or antagonistic effects among pollutants.
Because of the many uncertainties and potential omissions, it is
difficult to say whether the risk estimates presented here are over- or
underestimates of actual health risks from pollutants studied. For
those chemicals for which the study was able to make quantitative esti-
mates of exposures and risks, it is more likely that risks are over-
estimated than underestimated. To the extent that toxic chemicals about
which we currently know little have been left out, risks may be under-
estimated.
The value of the study methodology is that it allows an evaluation
and comparison of the potential health risk from chemicals and pollution
sources about which we have reasonably good information. Management
of these risks, based on the best current information, can proceed,
while research continues on the effects of chemicals about which little
is currently known.
3-4
-------
II. OVERVIEW OF METHODOLOGY
The Kanawha Valley study uses a risk assessment screening metho-
dology to evaluate and compare, in a very limited fashion, the poten-
tial health risks from exposure to a limited set of pollutants. The
purpose of this methodology is to permit a comparison of one risk with
another and to provide a general sense of the risk a substance may
present. This methodology does not make a definitive statement concer-
ning the absolute risk posed by a particular pollutant, source or
exposure pathway.
The following overview of the methodology for quantifying health
risks applies, in theory, to both cancer and non-cancer health effects.
However, the Kanawha Valley study contains quantitative risk estimates
for cancer effects only. A brief discussion of our method for cancer
risk assessment begins below. [See Appendix D for a more detailed
discussion].
The risk assessment methodology combines estimates of toxicolo-
gical potency, derived from laboratory and occupational studies, with
estimates or measurements of local contamination (or exposure) levels
to estimate risk to the local population. This approach involves a
relatively high degree of uncertainty, and, in most cases, cannot
easily be verified or disproven by observation. It is useful, how-
ever, because it is analytically straightforward and is the only method
available for estimating the future health effects of current and
future exposures.
The two key elements in estimating risk are a chemical's estimated
toxicological potency and an individual's exposure to that chemical.
This risk assessment methodology is illustrated in Figure 3-1.
The toxicological potency assessment involves (1) a qualitative
'hazard identification' to determine if evidence exists that a chemical
causes an adverse health effect; and if so, (2) a quantitative estimate
of the 'doseresponse relationship' or potency of a chemical, linking
specific quantities of a chemical to particular health effect levels.
Quantitative potency estimates typically relate given dose (equivalent
to exposure) levels to the percent disease incidence expected in a
population, or to a probability that an individual will be affected
during the course of his or her lifetime. (See Figure 3-2 for an
example of a dose-response curve.)
3-5
-------
FIGURE 3-1
RISK ASSESSMENT METHODOLOGY
EXPOSURE
ASSESSMENT
POTENCY
ASSESSMENT
CHEMICAL
Modeling of
Fate and
Transport
SOURCE
?LAB
EXPERIMENTS
Hazard
Identification
nnnnnnnn
IMITTTTTI
Monitoring
PATHWAY
TO EXPOSURE
::: EPIDEMIOLOGICAL
studies.
M
k
Ambient
Pollution
Concentrations
Quantitative
Potency
Estimates
m m
ESTIMATED HEALTH RISK
3-6
-------
Exposure is estimated by measuring or estimating the ambient
and making assumptions about the relationship between ambient condi-
tions and actual human exposures. These standard assumptions, called
exposure constants, account for the amount of air or water a typical
person takes in in a day, and the weight of an average person.!
We assume that an average person weighs 70 kilograms (or roughly
154 pounds), breathes 20 cubic meters of air each day (700 cubic feet),
and drinks 2 liters (2.1 quarts) of water each day.
Finally, exposure and potency estimates are combined to estimate
individual and population risks.
We define risk to an individual as the increased probability that
an individual exposed to one or more chemicals will experience a parti-
cular adverse health effect during the course of his or her lifetime.
(An average lifetime is assumed to be 70 years). Risk to the population
is the expected increased incidence (number of cases), above the back-
ground rate, of an adverse health effect in an exposed population.
For cancer risks, we present both types of quantitative risk estimates
in the Kanawha Valley study.
If we make the simplifying assumption that the dose-response curve
is linear, potency (risk per unit of exposure) is constant at any
level of exposure in which we are interested (unless the chemical has
a threshold, below which we believe there is no effect). Under these
assumptions, lifetime risk to the exposed individual is simply the
product of exposure and potency:
R = E x P
(individual risk) (exposure) (potency)
As discussed above, exposure is the product of ambient levels of the
pollutant in the medium of concern (air, drinking water, fish tissue)
and exposure constants (i.e., the standard assumptions about body
weight, intake and absorption):
E = Y x Z
(exposure) (ambient concentrations (exposure constants)
in medium of concern)
1 A further calculation of the actual dose likely to be absorbed into
the body involves an estimate of the body's propensity to absorb a
chemical to which it is exposed. This conversion of exposure to dose
is estimated for each chemical and exposure route, and is contained
within the potency estimate.
3-7
-------
figure 3& DOSE-RESPONSE CURVE
INCIDENCE:
% OF EXPOSED
POPULATION
U>
I
00
No Effect
Threshold
DOSAGE: MILLIGRAMS PER DAY
i
-------
Therefore, individual lifetime risk can be calculated by multiplying
the ambient concentration (Y) and exposure constants (Z) by potency (P):
R Y x Z x P
(individual (ambient concentration (exposure (potency)
risk) in medium of concern) constants)
For example, we might wish to calculate the lifetime risk of cancer
for an individual exposed to ten micrograms per cubic meter of benzene in
the air. Filling in equation (3) presented above, the ambient concentra-
tion (Y) is ten micrograms per cubic meter (which is the same as 0.01
milligrams per cubic meter or 0.01 mg/m^). Our exposure constants (Z),
as discussed above, assume a typical 70 kilogram person breathing 20
cubic meters of air each day (20m3/day).
Our potency value (P) is the cancer potency score for benzene inha-
lation developed by EPA's Carcinogen Assessment Group (CAG) of 0.029
(mg/kg/day)~l. This value indicates that an individual taking in one
milligram of benzene per kilogram of body weight per day (this would be
70 milligrams per day for a typical 70 kilogram person) for a lifetime
has an estimated chance of about three in a hundred of contracting cancer.
(If the units are confusing, think of the score as indicating cancer
risk for a given dose of benzene.)
Using this information to fill in the equation presented above, we
can estimate risk as follows [THIS IS ONLY AN EXAMPLE]:
(individual
risk)
0.01 m;
m
(ambient
concentration)
20 m3
day
1
70 kg
(exposure constants)
0.029
(mg/kg day)
(potency)
Working through the arithmetic, individual risk (R) in this case is 8 x 10~5,
or eight chances in 100,000 of contracting cancer over a lifetime.
To calculate the aggregate expected increased incidence of disease,
we multiply the average individual risk by the number of individuals
exposed:
I R x P
(aggregate expected (individual risk) (exposed population)
increased incidence
of disease)
3-9
-------
Following this example, suppose that we estimate that 100,000 people
are exposed to 10 micrograms per cubic meter (or 0.01 mg/m3) of benzene.
We would estimate incidence as follows [THIS IS ONLY AN EXAMPLE]:
I = 8 x 10~5 x 100,000 people
(aggregate incidence (individual risk) (exposed population)
of cancer)
The product of this equation is 8 cases of cancer over a 70 year
period (the assumed average lifetime in the individual risk calculations).
By convention, incidence (unlike individual risk) is reported in annual
cases, so we divide our product, 8 cases, by 70 to arrive at our estimate
of 0.1 cases per year, or about one case every ten years.
Despite the application of similar risk assessment techniques,
each media analysis differed in terms of geographic coverage, chemicals
studied, whether ambient concentrations were measured or estimated,
the potential health risks considered (cancer or non-cancer), and in
the means of expressing the risks studied (e.g., risk assuming uniform
ingestion of drinking water versus risk per pound of fish consumed).
Table 3-1 illustrates some of these differences.
TABLE 3-1
DIFFERENCES IN RISK ASSESSMENT APPROACHESl
BY MEDIA ANALYSIS
RISK ASSESSMENT APROACH
PRIMARY ROUTE OF
Exposure
Data
Health Effects Considered
MEDIA STUDIED
EXPOSURE ANALYZED
Measured
Modeled
Cancer Non-Cancer
Air Analysis
Inhalation of
Ambient Air
X
X
Drinking Water
Analysis
Ingestion of Public
Drinking Water
X
X X
Surface Water
Analysis
Consumption of
Contaminated Fish
X
X X
Hazardous Waste
Analysis
Ingestion of
Contaminated Ground
Water
X
X X
lln addition to these differences, the area of geographic coverage and types
of chemicals studied for each media analysis were different, as was the
means of expressing the risks studied - see text.
3-10
-------
The application of risk assessment techniques in a complex geo-
graphic area such as the Kanawha Valley is difficult because there are
many scientific and technical limitations. For example, there are
uncertainties in estimating the amount of pollutants released from a
given source, in estimating the extent of exposure of the general
public to a particular pollutant(s), and in extrapolating health effects
evidence derived from animal studies to estimate corresponding human
responses. In general, the uncertainties introduced by the exposure
data are usually much smaller than those associated with the dose
response information used to estimate human health risks.
Because of these uncertainties, EPA prefers to err on the side of
caution in making risk assessments. Therefore, following standard EPA
protocol, we use conservative estimates of risks that probably overes-
timate the actual risk from any one chemical. While the risks outlined
in the study should not be interpreted as a definitive prediction of
actual cancer risk to the general public in the Valley, these estimates
are useful in comparing the relative risks among various chemicals.
3-11
-------
Chapter Four
Air Quality Analysis
-------
KANAWHA VALLEY AIR QUALITY REPORT
I. INTRODUCTION AND PURPOSE
The Kanawha Valley Toxics Screening Study has four stated objectives
that have provided overall direction to this technical report. The four
objectives are:
(1) Using available data bases, identify many of the chemicals
routinely released or present within various exposure pathways;
(2) Develop a sense of the potential public health concerns these
toxic pollutants may pose in various pathways, based on health
effects and exposure information;
(3) For a select number of chemicals, provide an initial conservative
assessment of potential cancer risk and potential noncancer
health risk for predicted or observed concentrations within
exposure pathways; and
(4) Identify data and information gaps and outline needs and options
for future study directions to enable a more detailed
investigation of health issues where warranted.
These four objectives have guided the development of the methodology
for this air quality report. We reviewed available data bases of air
emissions within the study area to identify many of the chemicals
routinely emitted by point and county-wide area sources within the
Kanawha Valley. Although we cannot at this time assess the potential
health concerns all toxic pollutants within the inventories may pose in
the ambient air, we do provide a conservative assessment of potential
cancer risk from predicted exposures from a small set of pollutants.
Because of limitations in our exposure assessment and health effects
information, we do not assess potential noncancer health effects of these
pollutants.
The chapter is divided into seven sections. Section I is the
introduction. Section II explains our methodology for selecting a
limited set of pollutants emitted by facilities within a set of defined
air sheds. Section III describes the general approach to the analysis
including discussions of technical issues involved in determining
predicted ambient concentrations of pollutants within the Kanawha
Valley. Section IV presents the exposure assessment methodology and
exposure results. We examined two types of exposures: ambient exposures
of pollutants within pre-selected neighborhoods surrounding the major
facilities in this study area, and ambient exposures to the average
individual within the study area. Section V discusses the risk
assessment methodology and limitations, and Section VI presents the
results of this risk assessment methodology. The risk assessment results
4-1
-------
are based on modeled concentrations developed within the exposure
assessment section, standard EPA exposure assumptions, and carcinogenic
unit risk factors. The results, by pollutants and sources, are presented
in terms of lifetime incremental risk to the individual and incidence,
which is the predicted number of cancer cases per year during a lifetime
exposure from these pollutants. Finally, Section VII compares the
results across valley zones to identify the most important pollutants,
facilities, and release types, in terms of risk and incidence, within the
Kanawha Valley.
II. STUDY AREA DEFINITION AND POLLUTANT AND SOURCE SELECTION
Study Area Focus
The Kanawha River Valley is a 60-mile river stretch in a region of
rolling terrain with numerous ridges and valleys. The major industrial
areas that are the focus of this study are not distributed throughout
this region, but rather are concentrated along the valley floor of the
Kanawha River between Belle and Nitro, an area of roughly 45 km2. The
study area for the air analysis was therefore defined to include only
this key central area. Within this area we further divided the valley
into four valley zones: Belle, Charleston/South Charleston, Institute,
and Nitro. The selection of this study area was based on four
justifications:
(1)
(2)
(3)
(4)
The valley varies from a minimum of about 1 km in width at Belle to
as wide as 2 km in the South Charleston region; valley walls vary in
height from approximately 200 m to 300 m. Wind flow is complicated not
only by the changing orientation of the valley as the flow is channeled
The population of the Kanawha region is heavily clustered
within this 45 km2 area around the industrial facilities.
Based on the 1984 WVAPCC emission inventory, an inventory of
organic pollutants released into the ambient air by industry,
there are no major releases from Kanawha, Putnam, or Fayette
Counties of the potential pollutants of concern from industrial
facilities outside of the 45 km2 area between Belle and Nitro.
Current air dispersion models are unable to effectively assess
the entire 60-mile river stretch with its complex meteorology
and terrain.
Current air dispersion models can be applied within the valley,
provided the analysis is limited to evaluating concentrations
at receptors on the valley floor within limited distances of
the facilities (see Section III below).
4-2
-------
within it, but also by daily wind flow patterns. By dividing the valley
into four separate zones (see Figure 1) it was possible to estimate
ambient concentrations in neighborhoods close to the industrial
facilities contained in each zone and to determine exposure to the
average individual within each zone by applying existing air pollution
dispersion models without modification.
Pollutant Selection
The West Virginia Air Pollution Control Commission (WVAPCC) has
compiled an extensive emissions inventory for the Kanawha Valley.
According to that inventory, at least 570 pollutants are emitted by the
principal 17 point source facilities in the area. Since one objective of
this study is to estimate potential carcinogenic risk from a limited set
of pollutants and facilities, the modeling analysis focused primarily on
pollutants that have received a unit risk factor for carcinogenicity from
EPA's Carcinogen Assessment Group (CAG). In addition to these
pollutants, the study modeled ambient concentrations of two other
chemicals, formaldehyde and propylene oxide, which the West Virginia
Department of Health and EPA have identified as potentially carcinogenic
and which are emitted by major industrial facilities in the region.
Table 1 lists the 20 pollutants selected for evaluation in this
study. The list includes 12 CAG-scored organic compounds known to be
emitted by industrial sources in the valley, 11 CAG-scored compounds that
are commonly emitted on an area-wide basis from dispersed nonindustrial
sources (county-wide area source emissions), and the 2 compounds
identified by the West Virginia Department of Health.
Sources
Point Sources: The selected study area of the Kanawha Valley
includes distinct clusters of industrial sources. These clusters include
some of the largest chemical manufacturing facilities in the United
States. Table 2 lists seven industrial facilities that were part of the
1984 WVAPCC inventory and that emit one or more of the studied pollutants
addressed in this study; it also lists five area source categories.
These industrial sources emit 14 of the 20 study pollutants; area sources
emit the additional 6 pollutants. Table 3 presents the available
emissions data from each industrial point source and area source category
evaluated in this study.
The EPA report "Overview of Environmental Pollution in the Kanawha
Valley, West Virginia" (Vincent 1984) provides a summary of the
products produced by the major industrial facilities in the area.
Products produced in the Valley by these industries include
insecticides, pharmaceuticals, plastics, and fibers.
4-3
-------
hl*M, C
Grimms L«ndi
SM«n Churct
Claanor.
0»k Forest
Winticld
Ml SiiW CRurWt
FIGURE 1
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
THE LOCATION OF THE FOUR CHEMICAL FACILITIES IDENTIFIED
FOR THE KANAWHA VALLEY STUDY
itr.yy ':^jL
V -.
-;v
I) " < v^"T
CIRCLES REPRESENT LOCATIONS OF FACILrTY WITHIN VALLEY ZONES 's^r, V
9<7 »- • »•
iv S>la*fi
| < Stalls MIWs
\ / ¦ ' eini> ¦-'*
f RobertsbtfH , iiaiiiM cntpfl ' ..
y Jf stot*«fBiin«CnuFrt _ EmmV \
ju " si- cit.it> ... '•• i ^
/7ej-.,ic • ' Vwj- v_.. y*. "r
r / ¦ ¦
_ _ ' - v- - ^ ' ' Z«i»
Plum Orchartf ' .
• 1.; Bels'ows I ' /
l/raw
f" 0 . , /''Hiahyr '• • , " OsiJfltlvllt*
1 ' { j ' ''
* ' • t06*'J ¦ ¦•Hoaiwti '
>v "flteeher 1 - '*
5p*«4
f
.»»' „ J1 ''' '
io.h \ jgr-su.."
f~" N
• '¦ / T
\ . V' ,; ••=*. V ' i
• -J,.". .
'•¦' • '• \ . ^ .. y%l
¦. ,'j \ - .J*' •• -r W$»0*
) 1 ' • - t RarttaM / V ' '' -
... .. .J' '¦. ¦ •
-ufc™..vj;: •
/
¦i'lWuiif -"FteKr ,- ,. <• V ••'. r'."Z.ibli'.lS. I ri :"• j&r; k
.'• ;«»¦• v.-. f ./- \ ¦?&.>&¦ ¦... ¦rr'' ¦'¦ ^ -~4
.y : -0: ^• ..r1 "¦;•.^ •'/ v^--^^;!:-t-'
U"?-.-.;-: . t ¦- ',ciA:•;¦
.1 • . SA ^"1 j-
— «,,...- -
;B4=k 6me,„ : i. - <, ,..• JT,
l^T .=•
•*
c T«vtr .
v * • -3ir*}
\ ^SS^^"6 j 1
_^oj| rku>M> -
>¦ ' "
'trrji. I'V -"tUv^
zone 1
NITRO
• - • .-fcn^-" ' lW'T/C*
I^ignneh j CrK-eVt,
: ; I VALLEY
'. .nn MuwH OwrcW
o-MeUi«
CHARLESTON/SOUTH
CHARLESTON
'j^aw
ffdsnfipt "W® "" . V
:xc--- Wltn ;L " .
^G*2iK-> TH»J% r, ••- <
•' " % rt'ek«o»
.. C-;; -r^o \
i^^moffcj- ? -
-V
KANAWHA
VALLEY
•-SW!
AIRPORT
.-, Jpftng Hill
ZONE 2 !•'••-• y/ >-
VALLEY |si?
CjEofldTBf* ' Ci
r>Tt.3:x f::^X
-'-KMOI ." ' «f/ i. - ' ' ^ -
^ ... ,
VALLEY -&*\Y v '>¦ \
¦r V\ T^L«5P|ftffey V\ - St>«y)fc- —%
ZONE 4 I ' . . •
;<^rettjLSenr=L rtW(U? tZptL -^^rSi-
fiUTljca ~Xv
¦5»<« • ...
XWgrS i.---"''
jssl' y
Mon«rc+
."Suiny
CoilOu^
»C*nnSflo£;
^^fCWtoo3»te .
gcfwi-; -t?
•" _ irj. If •
too«nhBar?ugf>
^ ^ V J Sr^-~' N r> UZPi ~~ • - r* .' fr, .CW - , r^"1 •¦*>?,'/ '_>••- .'S''J%<- . •-i*t£ — _ . ^ .-'¦ V • • .U. .¦¦»•:•' • vwtt ¦-i •- - - •- —' • •// ~ 1 -- / ' / . Ks«w
•Ob6lTY
-' - -
1- -?If n —-^X-'-'"~ ¦ - l^WW
«tnc«un
r "«S|s "FiyentvilltV"'" -'
yy^W^SMl. ' ^ ~~ * • "*" '- . J-
/I •' '•- v" -V • -TBiWKfcia • '•
> " • rW,- \
- .^^***a«uPsf g«a^«.•
x" v-~-v" M, '4J *Sj r : *- ^pn - »uprp»ri»wy y
^w*va*^ .hpg-
-Cocpg*)^*.
*C«ffWtd
4-4
-------
TABLE 1
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
Pollutants Evaluated in This Study
Emitted by
Emitted by
Industrial
County-wide
Pollutant
Sources
Sources
Benzene
X
X
Perchloroethylene
X
Methylene chloride
X
X
Formaldehyde
X
X
Trichloroethylene
X
X
1,3-Butadiene
X
X
Ethylene chloride
X
X
Ethylene bromide
X
Arsenic
X
Benzo(a)pyrene
X
Cadmium
X
Beryllium
X
Carbon tetrachloride
X
Chloroform
X
Ethylene oxide
X
Acrylonitri le
X
Allyl chloride
X
Vinylidene chloride
X
Vinyl chloride
X
Propylene oxide
X
4-5
-------
TABLE 2
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
SOURCES OF THE 20 POLLUTANTS EVALUATED IN THIS STUDY
Industrial
Faci 1 ities
Locat ion
DuPont
Union Carbide
Union Carbide
Rhone Poulenc
Occidental
Monsanto
Artel
Bel le
S. Charleston/Charleston
Tech. Center
Institute
Bel le
Nitro
Nitro
County-wide Area
Source Categories
Solvent Use
Road Vehicles
Heating
Gasoline Marketing
Waste Oil Burning
4-6
-------
TABLE 3
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
SUMMARY OF EMISSIONS DATA FOR SELECTED POLLUTANTS
(METRIC TONS/YEAR)
ZONE: BELLE
Poi nt
Sources^
Area Sources
.3
Waste
Dupont-
Solvent
Road
Gas
Oil
Pol 1utant
Occidental
Belle
Uses
Vehi cles
Heati na
Marketi no
Burni na
Total s
Methylene chloride
83.0
2.9
7.0
0.0
0.0
0.0
0.0
92.9
Chioroform
70.8
1 .3
0.0
0.0
0.0
0.0
0.0
72.1
Carbon tetrachloride
20.1
2.8
0.0
0.0
0.0
0.0
0.0
22.9
Formaldehyde
0.0
8.1
0.0
4.0
2.5
0.0
0.0
14.6
Benzene
0.0
0.0
0.0
13.9
c*
X
o
1
l\J
0.4
4.0 x 10"5
14.4
Perchloroethylene
0.0
0.0
12.6
0.0
0.0
0.0
1
o
X
o
12.6
Trichloroethylene
0.0
0.0
1.7
0.0
0.0
0.0
1.0 x 10"
1.7
1,3-Butadiene
0.0
0.0
0.0
0.1
0.0
0.0
0.0
0.1
Ethylene chloride
0.0
0.0
0.0
0.0
0.0
3.7 x 10"2
0.0
3.7xl0~2
Ethylene bromide
0.0
0.0
0.0
X
o
1
ro
0.0
3.7 x 10"3
0.0
1 .8xl0~2
Arseni c
0.0
0.0
0.0
0.0
1.2 x 10~2
0.0
-4
8.0 x 10
1.3xl0~2
Benzo(a) pyrene
0.0
0.0
0.0
2.3 x 10~3
3.1 x 10~3
0.0
-5
2 x 10
5.4xl0~3
Cadmiurn
0.0
0.0
0.0
-4
8.0 x 10
3.4 x 10"3
0.0
1
O
X
CM
-3
4.4x10
Beryl 1 i urn
0.0
0.0
0.0
0.0
4.0 x 10~4
0.0
8.0 x 10 5
-4
4.8x10
^ Source: West Virginia Air Pollution Control Commission 1984 Emission Inventory, 1987.
^ Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
-------
TABLE 3 (CONTINUED)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
SUMMARY OF EMISSIONS DATA FOR SELECTED POLLUTANTS
(METRIC TONS/YEAR)
ZONE: CHARLESTON/SOUTH CHARLESTON
Point Sources^
Area Sources^
Waste
Carbi de-
Carbide-
Solvent
Road
Gas
Oil
Pol 1utant
S. Char.
Tech. Ctr.
Uses
Vehi cles
Heati na
Marketi na
Burni na
Total
Propylene oxide
44.6
^r
i
o
X
o
m
0.0
0.0
0.0
0.0
0.0
44.6
Benzene
0.0
0.0
0.0
40.3
0.2
1.3
-4
1.0 x 10
41 .8
Perchloroethylene
0.0
0.0
36.5
0.0
0.0
0.0
3.0 x 10"4
36.5
Ethylene oxide
35.7
0.4
0.0
0.0
0.0
0.0
0.0
36.1
Methylene chloride
5.0 x 10"4
4.2
20.4
0.0
0.0
0.0
0.0
24.6
Formaldehyde
0.0
0.0
0.0
11.7
7.3
0.0
0.0
19.0
Acrylontri le
17.3
0.2
0.0
0.0
0.0
0.0
0.0
17.5
Tri chloroethylene
0.0
0.0
4.9
0.0
0.0
0.0
4.0 x 10~4
4.9
Vinylidene Chloride
0.1
0.3
0.0
0.0
0.0
0.0
0.0
0.4
1,3-Butadiene
0.0
0.0
0.0
0.3
0.0
0.0
0.0
0.3
Ethylene chloride
Ethylene bromide
Arseni c
Benzo(a) Pyrene
Cadmi urn
Beryl 1i um
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-2
4.0 x 10
0.0
6.6 x 10
2.3 x 10"3
0.0
0.0
0.0
3.5 x 10~2
_3
9.0 x 10
-3
9.9 x 10
_3
1.0 x 10
0.1
1.1 x 10"2
0.0
0.0
0.0
0.0
0.0
0.0
2.2 x 10"3
4.0 x 10~5
-4
5.0 x 10
-4
2.0 x 10
0.1
5.1xl0~2
3.7xl0_2
-2
1.bxlO
-2
1.3x10
1.2xl0~3
^ Source: West Virginia Air Pollution Control Commission 1984 Emission Inventory, 1987.
^ Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
-------
TABLE 3 (CONTINUED)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
SUMMARY OF EMISSIONS DATA FOR SELECTED POLLUTANTS
(METRIC TONS/YEAR)
ZONE: INSTITUTE
Paoe 3 of 4
Point Sources^
Area Sources^
Waste
Solvent
Road
Gas
Oil
Pollutant
Rhone Ponlenc
Uses
Vphicles
Heatina
Marketi na
Burni na
Total
Ethylene oxide
97.5
0.0
0.0
0.0
0.0
0.0
97.5
Propylene oxide
71.2
0.0
0.0
0.0
0.0
0.0
71 .2
Chloroform
62.1
0.0
0.0
0.0
0.0
0.0
62.1
Acryl oni tri le
28.1
0.0
0.0
0.0
0.0
0.0
28.1
1,3-Butadi ene
25.2
0.0
0.1
0.0
0.0
0.0
25.3
Methylene chloride
15.3
9.1
0.0
0.0
0.0
0.0
24.4
Benzene
0.3
0.0
18.0
8.0 x 10"2
0.6
5.0 x 10~5
19.0
Perchloroethylene
0.0
16.3
0.0
0.0
0.0
2.0 x 10~4
16.3
Formaldehyde
0.1
0.0
5.2
3.3
0.0
0.0
8.6
Trichloroethylene
0.0
2.2
0.0
0.0
0.0
2.0 x 10"4
2.2
Ethylene chloride
0.0
0.0
0.0
0.0
4.8 x 10"2
0.0
4.8xl0~2
Ethylene bromide
0.0
0.0
CSJ
1
o
X
CO
0.0
-3
4.8 x 10
0.0
2.3x10~2
Arseni c
0.0
0.0
0.0
1.6 x 10
0.0
10.0 x 10~4
1.7xl0~2
Benzo(a) Pyrene
0.0
0.0
3.0 x 10"3
4.0 x 10~3
0.0
2.0 x 10-5
7.0xl0~3
Cadmi urn
0.0
0.0
1.0. x 10"3
4.4 x 10"3
0.0
-4
2.0 x 10
5.6xl0~3
Beryllium
0.0
0.0
0.0
-4
5.0 x 10
0.0
-5
10.0 x 10
6.0xl0~4
^ Source: West Virginia Air Pollution Control Commission 1984 Emission Inventory, 1987.
^ Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
-------
TABLE 3 (CONTINUED)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
SUMMARY OF EMISSIONS DATA FOR SELECTED POLLUTANTS
(METRIC TONS/YEAR)
ZONE: NITRO
Pol 1utant
Poi nt
Monsanto-
Polvmer
Sources^
Artel
Solvent
Uses
Area Sources
Road
Vehi cles
3
Heati na
Gas
Marketi no
Waste
Oil
Burni na
Total
T richloroethy1ene
22.5
0.0
0.8
0.0
0.0
0.0
7.0 x 10"5
23.3
Formaldehyde
4.7
0.6
0.0
2.0
1.3
0.0
0.0
8.6
-4
-2
-5
Benzene
0.0
O
X
o
0.0
7.1
3.1 x 10
0.2
2.0 x 10
7.3
Perchloroethylene
0.0
0.0
6.4
0.0
0.0
0.0
6.0 x 10"5
6.4
Methylene chloride
0.0
5.0 x 10"4
3.6
0.0
0.0
0.0
0.0
3.6
Allyl chloride
0.0
2.0
0.0
0.0
0.0
0.0
0.0
2.0
-2
-2
-2
Ethylene chloride
0.0
5.2 x 10
0.0
0.0
0.0
1.9 x 10
0.0
7.1x10
1,3-Butadiene
0.0
0.0
0.0
5.4 x 10"2
0.0
0.0
0.0
5.4xl0~2
-3
-3
o o „-3
Ethylene bromide
0.0
0.0
0.0
6.9 x 10
0.0
1 .9 x 10
0.0
8.8x10
-3
-3
Arsen i c
0.0
0.0
0.0
0.0
6.1 x 10
0.0
4.0 x 10
6.5x10
Acryl oni tri 1 e
0.0
5.9 x 10~3
0.0
0.0
0.0
0.0
0.0
5.9xlO~J
-3
-6
. „ nn-3
Benzo(a) Pyrene
0.0
0.0
0.0
ro
X
o
i
1.6 x 10
0.0
8.0 x 10
2.8x10
-4
-3
-5
o o ,n-3
Cadmi um
0.0
0.0
0.0
4.0 x 10
1.7 x 10
0.0
9.0 x 10
2.2x10
Carbon tetrachloride
0.0
-4
9.0 x 10
0.0
0.0
0.0
0.0
0.0
9.0xl0~4
Chl oro form
0.0
9.0 x 10"4
0.0
0.0
0.0
0.0
0.0
9.Ox10~4
Vinyl chloride
0.0
-4
5.0 x 10
0.0
0.0
0.0
0.0
0.0
5.0xl0~4
-4
-5
-4
Beryl 1i um
0.0
0.0
0.0
0.0
2.0 x 10
0.0
4.0 x 10
2.4x10
^ Source: West Virginia Air Pollution Control Commission 1984 Emission Inventory, 1987.
3 Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
-------
Although the study excluded no known sources of the 20 pollutants
addressed within the study area, it did not model any sources outside of
the study area, such as the Amos power plant in Poca and the Kanawha
River station power plant at Cedar Grove. Although these facilities are
a potential source of metals, they were excluded because emissions data
were unavailable and because the expected incremental impacts in the
residential areas closest to it within the study area were assumed to be
minimal.
Area Sources: Estimates were made of the emissions for pollutants
from five county-wide area source categories: (1) solvent use, (2) road
vehicles, (3) heating, (4) gasoline marketing, and (5) waste oil
burning. All available emissions data for pollutants addressed in this
study were compiled for these categories and then apportioned into a
modeling grid.
Emissions were first estimated at the county level, based on factors
such as vehicle miles traveled, population, and fuel consumption.
Emissions factors were used to estimate the mass of specific pollutants
emitted as a function of the above factors. The county-wide estimates
were then distributed among the 2.5 km grid squares that were within the
study area. Refer to Technical Appendix D for a detailed description of
county-wide area sources.
III. GENERAL METHODOLOGY
The two principal elements of this study's methodology are exposure
assessment and risk assessment. This section dicusses general study
design issues pertaining to these key elements. Details on
implementation of the general design for exposure assessment are
presented in Section IV, and risk assessment is discussed in Section V.
The general methodology is summarized as follows:
• The approach for exposure assessment is presented by
identifying five major complications specific to this study
area, and then indicating how these technical problems were
overcome.
• The risk assessment subsection then summarizes how the
exposure data were used to estimate risk on two scales:
neighborhood scale and average individual exposures for each
zone.
Exposure Assessment
The exposure assessment used in this study had two objectives:
4-11
-------
(1) To characterize concentrations of toxic air pollutants in
neighborhoods adjacent to the major industrial complexes, and
(2) To provide conservative assessments of average concentrations
of toxic pollutants in each of the four valley zones studied.
To achieve these objectives, the study relied principally on the use
of dispersion modeling analysis. A limited amount of monitored data was
gathered as part of the effort, but these data were used only as a
reference for evaluating the general magnitude of the results developed
through dispersion modeling. The estimated pollutant concentrations used
in the risk analysis were not modified in any way to reflect the results
of the monitoring program.
The study relied to the extent possible on standard methods and
procedures. A detailed modeling protocol for conducting Gaussian
diffusion modeling within the valley was written and submitted to EPA for
review and approval by various offices within EPA; it is reproduced in
Technical Appendix C.
The study's general exposure assessment methodology can best be
discussed in relation to five basic problems specific to evaluating air
quality within the industrialized sectors of the Kanawha River Valley:
(1) Estimating releases from the hundreds or thousands of
potential sources within each industrial complex;
(2) Accounting for the variability of emissions from the
facilities under study;
(3) Developing appropriate modeling approaches to deal with the
highly complex terrain of the valley;
(4) Obtaining acceptable meteorological data representative of
conditions in and around each major facility; and
(5) Correctly identifying population locations, especially for
those nearest the major facilities.
Each of the five points is discussed below.
Multiple Release Points: Many of the chemical complexes under study
release air pollutants from literally hundreds of separate stacks,
valves, pipes, flanges, vents, and other sources; release characteristics
from these different types of sources vary significantly. In addition,
4-12
-------
each of the industrial complexes covers many acres. Simplified modeling
techniques that characterize releases as originating at only one, or even
several, different release points would inevitably yield inaccurate and
misleading estimates of pollutant concentrations in the local area,
especially those concentrations for the residential neighborhoods closest
to the plants. It was therefore imperative to estimate release points in
as much detail as possible, both in terms of where pollutants are
released and in terms of the characteristics of each release point.
To respond to this problem, the study relied on the WVAPCC 1984 Toxic
Air Pollutant Inventory, which separately describes up to thousands of
individual release points within each chemical complex in the valley in
terms of source characteristics and substances released. Staff of the
WVAPCC aggregated these release points within each of the seven
facilities, typically identifying 25 to 50 separate source groups within
each facility to be used as input to the computer modeling effort. These
aggregated sources represented three basic types of releases: stacks,
building vents, and industrial area sources (such as tank farms). This
level of detail increases the resolution of estimated pollutant
concentrations in the neighborhoods adjacent to the chemical plants.
Source Variability: Another major difficulty in conducting air
exposure assessments in this area is accounting for the variability of
source releases over time. Chemical manufacturing processes often
operate in batches. Releases may therefore vary substantially from hour
to hour and from day to day. Since air flow in the valley also tends to
change significantly over each 24-hour period (with nighttime conditions
often being considerably more stable and less likely to vigorously
disperse air pollution than more turbulent daytime conditions),
characterizing source release variations is critical to understanding
exposures in neighborhoods surrounding the facilities and to the average
exposed individual.
To respond to this problem, the study relied on the data gathered by
the WVAPCC, which considers hours of operation of the various processes
in the plants. This permitted the development of considerably more
representative day/night air dispersion model treatments.
Topographic Effects: In their current state of development, air
dispersion models have only limited success in characterizing dispersion
over terrain defined as "complex," that is, where the release height of a
pollutant is at a lower elevation than a receptor point (such as a nearby
hill).
To accurately estimate concentrations of pollutants at all locations
through the valley, one approach would be to develop a site-specific air
dispersion model to predict concentrations under the unique conditions
4-13
-------
imposed by the local terrain. This would have been too expensive and
time consuming for use in this study, and the accuracy of modeling
transport in complex terrain on the order of 40 km to 50 km would still
have involved considerable uncertainty. Instead, the study relied on
available Gaussian dispersion models, but limited the area of concern to
the valley floor, where their predictions of dispersion over flat terrain
are likely to be reasonably reliable. This approach is especially
applicable to the estimation of concentration in neighborhoods in
proximity to industrial facilities, provided that available
meteorological data accurately represent each of the valley zones (see
below). The approach will satisfy the exposure assessment's first
objective, assessing concentrations in neighborhoods nearest the chemical
plants.
To meet the study's second objective over the full area of each zone,
ISCLT, a gaussian model, was used for industrial sources, and box model
estimates were used for county-wide area sources to determine exposures
to the average individual. This provided one estimate of exposures to
the average individual. Since this approach does not allow for interzone
transport of pollutants, another approach was developed to provide likely
conservative estimates of exposure and to allow interzone transport of
pollutants. For this second approach, we applied the box model to all
pollutant sources. Although the results of this box model analysis of
total emissions in the study area are not as detailed as those of
Gaussian dispersion models for each valley zone, the box model can
estimate conservative upper bounds of annual average concentrations to
which the population as a whole may be exposed. The box model is a
simple treatment, but true concentrations of pollutants in the area are
unlikely to be any higher than the concentrations it predicts when
conservatively applied. Section IV, "Sensitivity Analysis," presents
these conservative estimates of average concentrations to which the
population within the study area could be exposed.
Meteorological Conditions: Air flow within the Kanawha River Valley
is strongly channeled by the valley walls. The models used for
evaluating pollutant concentrations require estimates of wind direction,
wind speed, and atmospheric stability on an annual basis. The study
therefore established meteorological monitoring stations within each of
the four valley zones and operated the network continuously for a period
of six months in cooperation with the WVAPCC. In addition to recording
wind speed and direction, the system also continuously recorded turbulent
intensity at the Belle and Institute zones, thereby providing detailed
data to help document valley dispersion characteristics, including
dispersion under nighttime conditions, when stable air can lead to higher
concentrations of pollution within the areas downvalley of the chemical
facilities.
4-14
-------
Population Distribution: Since one objective of the study was to
characterize exposures in the neighborhoods adjacent to the industrial
facilities, it was necessary to obtain accurate data on the location of
residences near the chemical plants. Residences are usually located
through review of USGS and other maps of a local area; for this effort,
however, field studies were conducted to precisely identify the location
of residences in key areas so that modeling results could be as accurate
as possible. In many cases, the closest residences are within 200 m of
the property lines of the major facilities.
Figures 2 through 5 show the neighborhood-level grid used by the
modeling analysis to evaluate concentrations in critical subareas of the
four study zones. Each of these receptors is located in a residential
area. The modeling also produced estimates of general concentrations
within a larger receptor grid established for each valley zone; Figure 6
shows the 2.5 km grid and associated population data used to produce
estimates of population-weighted average concentrations.
Risk Assessments
In our exposure analysis, we modeled the selected pollutants and
sources within each zone for the two exposures, neighborhood exposures
and average individual exposures for the zone. For our carcinogenic
health effects analysis, we apply our risk assessment methodology to the
concentrations predicted from the exposure assessment.
For these two types of predicted concentrations we developed several
measures of risk. First, we estimated lifetime incremental individual
risk for these two exposures. To develop these risks, we assumed
standard EPA exposure assumptions and applied unit risk factors developed
by the Carcinogen Assessment Group (CAG) within EPA. From these
estimates of lifetime incremental risk, we further developed cancer
incidence estimates for the population. Incidence, for the purpose of
this report, is the estimated upper-bound number of cancer cases expected
over a 70-year exposure period. Incidence is calculated by multiplying
individual risk by the population number, and can be annualized by
dividing by 70, the assumed average life span, to obtain cancer cases per
year.
IV. EXPOSURE ASSESSMENTS
This section describes the exposure assessment methodology in
detail. Topics covered are the following:
•Simplifying assumptions;
•Model selection;
4-15
-------
u
FIGURE 2
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
APPROXIMATE LOCATIONS OF NEIGHBORHOOD RECEPTORS, MAJOR FACILITIES,
AND MONITORING SITES FOR THE BELLE COMPLEX
\ .y \ \
A "A \
Wen.'/
Simmn
* 4-2 lr5*^n,"v-
v \xV/x^>,x ^ / / Vy:;
: N \
/ V-V'' J ^
V\\^,. ( |\ \-« 7
-14 )\ \
-»v V \ V V \ v •» ^ ' ••'fjfr
—T «>Nw ^V
¦• "iS- - • Si'Sj; ^ \
¦ .v. V?! C.^Vl'N, \ \ vW'X^.V,
" /..¦ 'tOriv»-i^,v V v s.
. ... 'n air quality monitoring site
X (MARCH - APRIL 1986)
X«o»\ \ V' / '* >"
V X y/.J • ^ v. ,\ ¦
h \ ~ \f V «v "
J Btrj"
2z' „.v •
a.z
m \J"I- 4^
5S§*6-.(. Q<
, % \v V \
i / $ /\ V
1 KILOMCT£R
4-16
-------
FIGURE 3
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
APPROXIMATE LOCATIONS OF NEIGHBORHOOD RECEPTORS, MAJOR FACILITIES
AND MONITORING SITES FOR THE SOUTH CHARLESTON COMPLEX
• r
v
^9OC
V /' / /£>,•;
¦ L' f'j\ {
I Ch. / . ,tP- *\\7jj S \ K \ \ *"$- ^ ) 1
vvv (. ¦ ¦' - ^ p.V-C^
i"' 1 i**1 M (_¦¦ :
.¦? — - - a ; / -7%. -X .o METEOI
- /i ~.
.« ' METEOROLOGICAL
! /MONITORING SITE
oods >\
julim
Rkcrcat
CARBIDE
HOSfiitl
ITSefi
^4^
1 KILOMETER
4-17
-------
FIGURE 4
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
APPROXIMATE LOCATIONS OF NEIGHBORHOOD RECEPTORS, MAJOR FACILITIES,
AND MONITORING SITES FOR THE INSTITUTE COMPLEX
- N /
AIR QUALITY MONITORING SITE C.
'(MARCH-APRIL 1986)
1 km DOWN VALLEY
i ' ^
1 ' ¦
UNION CARBIDE 7""*
<• 2-15
/ ¦¦ ' -
It T air quality monitoring site
(MARCH - APRIL 1986)
'^T/2-7 " / / f .
^ ¦ *—• > * / / ±.
^ fuM*'**-**' *' /? \ flF? W-VIRGINIA .
>-X. . n .. dcuadii iTATtnu.fi"3 ctatc mi i CfiPX\ %
STATE COLLEGE/
BUILDING
' *i '»'
- -^Y
2-9 y. '/xV". -7
. - ' c.rkV-"^-N ^
ft.»Vs w?
^V.v7 S
\ s.
^ \"
. r.
,; I
"U NJ-ON. '.
r* A # -
°o ^ ^Hr ."V », / *•¦'
r \ ,
_ _ v \\ 2- 2 lx< ;• / ^
— - ¦*» —C ' - * V.„ -\ *\ v:. "• .
- , V ~S- . :-. ^ ,o>«b\ •"•¦V- V Vn^/
t~--' •¦**: I •" ~ - i ~- S * *TL. •. .^=^- -i x r x. * \ >% +,
f 0 ______ j "L£ \ ^
1000
1000 2000
3000 4000 5000 6000
7000 fitT
1 KILOMCTt*
4-18
-------
FIGURE 5
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
APPROXIMATE LOCATIONS OF NEIGHBORHOOD RECEPTORS MAJOR FAfll tttfc
AND MONITORING SITES FOR THE NITRO COMPLEX FACILiTIES'
..j
METEOROLOGICAL -
MONITORING SITE
>9
\TTt-
„ MONSANTOjr 1J' *
I ' *25 M
t •*
ALLIED v i-#^-
i-6 -JL
v^7»
COASTAL
4 FIKE >i J5
* 599
& \ at
* * /1 ; > " ' ..V
-ii j } \ .- - v
1000
2000 3000 4000 5000 6000 7000 ^Ef
1 KtLOMCTE*
4-19
-------
FIGURE 6
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
THE GENERAL RECEPTOR GRID AND CORRESPONDING 1980 POPULATION VALUES
TO BE USED IN THE KANAWHA VALLEY STUDY TO THE.AVERAGE INDIVIDUAL EXPOSURE
Li >¦' '¦ - V II
mssm
11
ornado
m,' ^
Ivydal*\ yjnjv
csvJ^
Ruth,
m
Pridftley
^ « ^Mon«rcti
»
Scale 1:250,000
UvTt*utfc'd
U«m
,VK .% V
20 Statute Miles
10
20
25
30 Kilometers
10
15 Nautical Miles
POPULATION FOR KftNAUHA VALLEY AREA
•
6RID NUMBER
A
B
C
D
TOTAL
1
1,360
2, BOO
2,790
2,290
9,240
2
-
-
-
-
7,B30
3
-
-
-
-
4,400
4
-
-
-
-
2,650
5
-
-
-
-
2,500
6
-
-
-
-
2,630
7
750
750
3,000
1,570
6,070
B
7,150
1,500
1,490
B70
11,090
9
2,160
7,720
2,070
6,170
IB, 120
10
2,450
5,640
5,960
9,640
23,690
11
7,570
6,510
710
2,500
17,290
12
-
-
-
-
2,120
13
-
-
-
-
3,280
14
-
-
-
-
3,730
15
-
-
-
-
6,500
16
-
-
-
-
7,920
17
8,890
1,660
3,780
1,760
16,090
IB
950
4,530
340
340
6,160
19
-
-
-
-
900
20
-
-
-
-
1,080
21
-
-
-
-
1,320
22
-
-
-
-
1,320
23
-
-
-
-
1,110
24
1,950
1,440
320
300
4,010
25
-
-
-
-
900
26
-
-
-
-
900
27
-
-
-
-
900
28
-
-
-
-
900
29
-
-
-
-
560
30
2,730
140
1,B70
3,010
7,750
TOTAL
172,960
Cross Hatch Indicates those areas modeled
•DATA TAKEN FROM U.S. DEPARTMENT OF COMMERCE BUREAU OF THE CENSUS. 1983.
-------
• Model inputs;
• Uncertainties and limitations;
• Results;
• Sensitivity analysis; and
• Ambient monitoring programs.
Simplifying Assumptions
Perhaps the most important simplifying assumption made in this study
is that estimates of pollutant concentrations in ambient outside air are
useful indicators of potential health risks. Such an assumption
implicitly assumes that people are continuously exposed to ambient
concentrations predicted for the neighborhood in which they reside.
Seven other major simplifying assumptions were used in the exposure
assessment:
Meteorological conditions within each of the study zones were assumed
to be homogeneous: While meteorological conditions actually vary within
each valley zone, conditions throughout each zone were considered to be
represented by data from the single station sited by this study within
each zone. (For a discussion of the design of the meteorological
monitoring system, see Technical Appendix B.)
Valley zones were assumed to have no influence on each other for the
characterization of concentrations in the neighborhoods in proximity to
the industrial facilities: In the Gaussian dispersion modeling,
pollution transport between zones was assumed to be zero. In the
sensitivity analysis, box model estimates, based on the total emissions
throughout the study area, were made to provide a conservative estimate
of annual average concentration to which the population is exposed,
considering interzone transport. If average concentrations were to be
addressed in more detail, however, extensive tracer studies would be
needed to document transport and dispersion among valley zones, as well
as model modifications to account for the restriction of horizontal plume
spread.
Countv-wide area source emissions were based on national estimates:
In addition to the pollutants emitted by the chemical facilities, the
Kanawha Valley, like any metropolitan area, is exposed to toxic air
pollutants associated with commercial and residential development. These
include emissions from automobiles and trucks, solvent use (such as in
dry cleaning), heating, gasoline marketing, and waste oil combustion.
Since it would have been impractical to compile a detailed inventory of
such sources, and since county-wide area source emissions are often quite
similar from one location to another, emissions were estimated at the
county level based on national averages. These estimates were then
apportioned within the study area for modeling purposes on the basis of
population density.
4-21
-------
Atmospheric decay of pollutants was, with one exception, assumed to
be zero: When released into the air, organic compounds can react with
both ambient air and surface substances, thereby decreasing the
compounds' concentration in the ambient air. Some pollutants are
affected more than others; Table 4 presents available half-life data for
the chemicals being modeled.
Half-life refers to the length of time (presented here in hours) that
it would take, on average, to reduce a given pollutant's mass within a
plume by 50 percent. As Table 4 indicates, the chemicals of interest to
this study have half-lives much longer than typical transport times for
the distances encountered in the valley, especially for the neighborhoods
adjacent to the plants. For example, assuming a typical wind speed of
2 m/sec, it would usually take only a few minutes for pollutants to reach
most nearby receptors and only about one to two hours to travel the 10 km
to 15 km length of the longest valley zones.
For the purposes of this study, each pollutant was therefore assumed
to have an infinite half-life. This term, therefore does not
significantly affect modeling results. Furthermore, this assumption
allows concentrations to be linearly scaled by emission rate, thereby
facilitating data processing. An exception was made, however, for allyl
chloride, which has a half-life of 4.8 hours (EPA, December 1985). The
effect of decay for even this pollutant is relatively minor.
Formaldehyde, on the other hand, which also has a relatively short half-
life, was modeled conservatively assuming zero decay based on the upper
end of the range shown in Table 4. (For a quantitative assessment of the
sensitivity of this term, see also "Sensitivity Analysis" below.)
The population within the study area is assumed to live on the valley
floor and not on the hillsides of the valley: A modification to the
modeling protocol would have been needed to estimate pollutant
concentrations at hillside receptors. Since it is likely that much of
the population that is exposed to relatively high concentrations of these
pollutants does live on the valley floor, no such modification was
attempted. Limited modeling analysis was performed, however, to
determine the magnitude of the concentration for hillside receptors.
This analysis, contained in the sensitivity analysis section, indicates
that predicted hillside concentrations are lower than those of the valley
floor.
When apportioning population to the modeling grid, this study assumed
that a census tract that overlapped into two or more grids could be
apportioned based on the percentage of areas of the census tract in each
modeling grid: Population enters the calculation of exposure in two
ways: (1) it was used to apportion county-wide area source emissions into
the modeling grids, and (2) the concentration of each grid is weighted by
4-22
-------
TABLE 4
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
Atmospheric Half-life Values (Hours)
for Pollutants Modeled
Half-Life (Hours!
Pollutants Low High
Acrylonitrile
93.6
-
Benzene
144.0
-
Carbon Tetrachloride
192720.0
-
Chloroform
1920.0
-
Ethylene oxide
96.0
-
Formaldehyde
2.4
19.2
Methylene chloride
1380.0
-
Vinyl chloride
28.8
-
Vinylidene chloride
48.0
-
Ethylene chloride
864.0
3048.0
Allyl chloride
4.8
-
Trichloroethylene
88.8
-
1,3-Butadiene
3.0
-
Propylene oxide
148.8
-
This value applies to atmospheric half-life in the presence of
full sunlight (Federal Register, October 10, 1985, 50 FR41466).
The use of this value directly would have acted to understate
risk. Zero atmospheric decay was therefore assumed for 1.3-
Butadiene in order to conservatively estimate concentration.
Source: Office of Emergency and Remedial Response, Office of Solid Waste
and Emergency Response, EPA, 1985.
4-23
-------
the population of that grid to estimate weighted average exposures.
Uniform population density was assumed within census tracts. For census
tracts that include portions of the valley wall, or beyond, this is not
likely to be a conservative assumption. Based on the apportionment
method, approximately 100,000 people are found to live on the valley
floor, between Belle and Nitro. It is not clear what percentage of the
residents actually live on the valley floor.
A number of source categories were excluded from the analysis because
thev were assumed to be relatively insignificant or we were unable to
Quantify releases from such sources.
These source categories include:
- Volitilization from Waste Water treatment plants.
- Volatilization from the Kanawha River.
- Emission of point source metals and other inorganic pollutants
not covered by the WVAPCC inventory.
- Emissions from abandoned hazardous waste disposal sites and, in
some cases, active waste management facilities.
- Emissions from small facilities that were not covered in the
WVAPCC inventory.
- Emissions associated with accidental releases.
Model Selection
Numerous EPA dispersion models could have been used for this study.
Based on a review of the features of the models included in the "User's
Network of Applied Modeling of Air Pollution" (UNAMAP) , ISCLT (Bowers
and Bjorklund 1979) was selected as the primary model to be used for this
study as the basis of risk assessment. ISCLT is EPA's recommended model
for complex industrial sources for flat terrain applications (EPA 1986b),
and since all receptors modeled in this application are on the valley
floor, complex terrain was not an issue in model selection. In addition,
however, LONGZ (Bjorklund et al. 1982) was used as a supplement to ISCLT
because it offers alternative treatments (discussed below) for modeling
area sources and handling turbulent intensity data. ISCLT's results are
reported in this volume; LONGZ's results can be found in Technical
Appendix A.
Both ISCLT and LONGZ fully meet the study's first objective of
characterizing concentrations in neighborhoods nearest the chemical
plants. They are less reliable in estimating annual average exposures
to the average individual due to possible transport and valley wall
UNAMAP V was the most current version available for use in this
study at the time the modeling was performed.
4-24
-------
reflection of pollutants throughout each valley zone the study's second
objective. For this reason, a generic "box model" approach was used in
addition to ISCLT and LONGZ to estimate upper bound average annual
concentrations throughout the valley to account for interzonal
transport.5
Area Sources: Since ISCLT uses a "virtual point" concept to
represent area sources, it treats emissions from a particular area source
as if they were all released at a location sufficiently upwind of the
source so that the horizontal spread of the plume would approximate the
dimensions of the area source at the location of the source. For this
reason, it cannot properly compute estimates of concentrations at
receptors located within an area source. Although it would be
technically possible to minimize this problem by repeatedly subdividing
area sources, it would be impractical to do this on as large a scale as
this study required. LONGZ, on the other hand, uses an alternative area
source procedure that offered the potential for better characterization
of area sources. For the risk assessment, we used box model
concentrations for exposure from county-wide area sources.
Turbulent Intensity Data: ISCLT is well suited for addressing,the
complex industrial sources on the valley floor, but it cannot evaluate
site-specific dispersion conditions. Although LONGZ uses similar
treatments for handling industrial sources, it also is able to accept
turbulent intensity data such as those collected by this study's
meteorological monitoring network. Site-specific turbulent intensity
data have the potential to improve model performance because, especially
during stable nighttime air flow, dispersion conditions within the valley
can vary markedly from flat terrain settings. ISCLT would not be able to
distinguish between these two conditions.'
Model Inputs
Principal inputs to air dispersion analysis are (1) meterological
data, (2) emissions data, (3) receptors, and (4) classification of land
use. This section discusses these inputs in turn.
Refer to Technical Appendix F for a discussion of the input data and
approach used for the box model analysis.
Ultimately, the box model's (Hanna 1982) characterization of area
sources were found to be the most conservative and were therefore
used in the risk assessment.
Unfortunately, LONGZ's empirical formulae for relating turbulent
intensity data to plume dimensions do not appear to be
representative of near-surface releases. This weakness results in
LONGZ's predicting concentrations lower than those predicted by
ISCLT by a factor of roughly 2 to 3 for Institute and Belle, where
turbulent intensity data were incorporated into the modeling
analysis. Refer to Technical Appendix A for further details on this
-------
Meteorological Data: Atmospheric stability data are used in air
dispersion modeling to estimate the width and vertical extent of a plume
as a function of downwind distance. They provide a measure of
atmospheric mixing, which acts to spread a plume along its horizontal and
vertical axes. Turbulent intensity data can be used in modeling to
provide a more direct measure of horizontal and vertical dispersion than
is possible through the use of default stability data.
To develop such data for the purposes of this study, wind speed, wind
direction, and the standard deviation of horizontal wind direction were
monitored in four valley zones at heights ranging from 10 m to 20 m.
Vertical wind speeds were measured in two valley zones, Belle and
Institute, to help characterize vertical turbulent intensity; these
measurements were used to estimate vertical dispersion coefficients for
the LONGZ model runs (see discussion above).
The primary recording system was a data logger located at the WVAPCC
headquarters. The data logger was connected to the remote sites by
telemetry, such that the data could be displayed on a real-time basis or
archived. At each remote station, data were also collected on strip
chart or paper tape as a backup.
As noted earlier, the study subdivided the study area into four zones
so that the valley orientation within each zone was relatively uniform.
A meteorological monitoring site was at a representative location within
each zone. A six-month data set was collected in all four zones from
December 6, 1985, to June 6, 1986. Technical Appendix B describes the
design and results of the meteorological monitoring program in detail.
A quality assurance plan that specified the tolerances and operating
procedures for the meteorological monitoring program was written and
approved (EPA 1986). Part of this plan specified that an independent
systems audit would be performed annually. The first audit was performed
in April 1986, with the system generally found to be operating within
tolerance.
Ambient temperature data input to the model was collected at Kanawha
Airport. This input was not generally a sensitive parameter because
plume rise is not a factor except for stack releases, which, in this
study, were found to be a relatively minor component. Although some
differences in ambient temperature would be expected between the airport
and the valley floor, even for stack emissions the effect on model output
of these differences should be small.
Emissions Data: All the point source emissions data used in this
study were developed by the WVAPCC in its 1984 survey of air emissions
from West Virginia manufacturers.
4-26
-------
The goal of this survey was explicity to obtain data to assess the
impact of toxic air pollutants on air quality. It included all sources
emitting more than 100 tons/year of criteria pollutants, with some
exclusions. Power plants were not surveyed because their organic toxic
pollutant releases are expected to be low. The most significant power
plant in the area is the Amos facility, located approximately 4 km
northwest of Nitro. Since its effective release heights are high and its
expected organic toxic emissions are relatively low, no effort was made
to estimate independent emissions data for this plant for the purposes of
this study. The WVAPCC inventory also excluded small processing plants
and blending facilities; they were omitted because an earlier WVAPCC
inventory found them to be insignificant in relation to the major
chemical facilities.
WVAPCC's questionnaire was sent to approximately 40 companies,
primarily chemical manufacturers, petroleum refineries, and miscellaneous
firms that potentially emit significant quantities of volatile organics.
Within the study area, 17 facilities were sent questionnaires, and all
responded. The seven facilities in the study area that emit pollutants
of concern for this study are shown in Table 2.
Three major categories of sources were inventoried by the WVAPCC
using a wide variety of methods:
Process emissions - Industries were allowed to calculate process
emissions by any technique that they felt would provide the best
estimate. A variety of approaches were used, including use of
ideal gas laws, computer model emissions estimates, and test
data. References were requested for any method used.
Combustion emissions - These emissions were based on the
"Compilation of Air Emission Factors" AP-42, (USEPA 1984),
supplemented with data to describe the specific volatile organic
emissions.
Fugitive emissions - Fugitive emissions are those that are not
released from specific vents or stacks, such as emissions from
leaking pipes and loading/unloading emissions. The WVAPCC
instructed industries to estimate emissions based on specific EPA
emission factors (USEPA 1986). Tank emissions were estimated by
standard formulas (USEPA 1984).
Of these, equipment leaks within fugitive emissions were the most
difficult to quantify in many cases. WVAPCC offered four different
options for approaching this problem of estimating equipment leaks. In
increasing levels of refinement, these were:
4-27
-------
Option 1 - The number of pumps are counted, and a standard ratio
of pumps to valves, flanges, and other components is assumed.
Loss of product is then estimated based on EPA emission factors
(EPA 1986), assuming a generic percentage of leaking components.
Option 2 - This option is the same as Option 1, except the
component counts include open ended lines, compressors, relief
valves, sampling connections, valves, and flanges in addition to
pumps.
Option 3 - Option 3 requires field measurements to determine the
percentage of sources within a facility that leak, i.e., leak/ no-
leak approach. Once the leak/no-leak percentage is determined for
each component, EPA emission factors (leaking and non-leaking) are
used to estimate emissions (EPA 1986).
Option 4 - Option 4 involves performing facility-specific testing
to determine the emissions estimates. This option would require
detailed pollutant-specific monitoring of numerous release points
and would be resource- and time-intensive.
Most facilities used Options 1 or 2 as the basis for the emissions
estimates. In many cases, Option 1 was used because accurate component
counts were not available. Only the Rhone Poulenc plant and Dupont used
the leak/no-leak approach (Option 3) and then only for a selected number
of units, totaling 14. None of the Option 3-generated data serve as the
basis for the estimate of emissions of the 14 point source pollutants
being evaluated in this study, i.e., all of the releases of the 14 point
source pollutants addressed in this study are based on emission estimates
using Option 1 or Option 2 data. Option 4 was not used for any emissions
within the study area.
The data generated by the WVAPCC survey are voluminous. For the 17
facilities in the study region, a total of 2,258 release points and 570
pollutants were included. Since it was not feasible to verify all of the
emissions data, the WVAPCC reviewed the data sets for reasonableness and
spot-checked computations for release points that emitted large
quantities of highly potent pollutants. For the Rhone Poulenc plant, the
WVAPCC also requested backup calculations for 15 to 20 release points.
An independent review was provided by the Office of Air Quality Planning
and Standards of EPA for this special case.
For the purposes of this study, WVAPCC's emissions data were divided
into daytime and nighttime releases if data were available for the hours
of operation. This is important for two reasons: (1) release rates
during the nighttime hours can be subject to poor dispersion conditions,
and (2) emissions from some key sources are greatly reduced or even
eliminated during the nighttime. This consideration is especially
4-28
-------
important for the Kanawha Valley because of the likelihood of significant
nocturnal drainage flows and stable conditions, which result in distinct
daytime and nighttime flow patterns. Separation of daytime and nighttime
emissions provides a more realistic treatment, especially from facilities
or processes not in operation during the nighttime hours.
There are two limitations to consider concerning the use of the
WVAPCC inventory for this study. First, the inventory includes recent
and future voluntary reductions proposed by industry that have not been
verified by WVAPCC or EPA (Farley 1986); current emissions of some
pollutants may therefore have been underestimated. Second, there is some
disagreement between industry and EPA concerning the accuracy of the
emissions factors used to estimate fugitive sources. Industry argues
that the emissions inventory overestimates some pollutant emissions by an
order of magnitude or more because it relies on inappropriate emissions
factors. To date, no supporting data have been received by the WVAPCC or
EPA to confirm this position (Beard 1986).
An important aspect of the modeling assessment was the grouping of
major facilities into logical modeling units. This permits air pollutant
concentrations to be directly attributed to various components within a
specific facility complex. For this study, every release point in the
1984 WVAPCC survey was assigned to a common process and modeling unit
based on data provided by the WVAPCC. At the process level, a manageable
number of release specifications were selected so that every release
point for a given process could be assigned to one set of release
specifications. For example, if a process had several stacks, a building
with numerous vents, and a product storage area, the modeling analysis
could be facilitated by modeling the stacks as separate sources, the
vents as one building source, and the storage area as one area source.
The computer software system 1 inked modeled data with the extensive
emissions inventory which has resolution to the release point level.
Figure 7 provides an example of how source groupings were made.
Since the WVAPCC inventory covers only industrial sources, this study
independently estimated county-wide area sources for the study area.
County-wide estimates were made for emissions of specific pollutants from
solvent use, road vehicles, gasoline marketing, heating, and waste oil
consumption. These emissions were then apportioned into a modeling grid
along the valley floor. (Refer to Technical Appendix D for details on
this methodology.)
Receptors: A limited number of receptors have been used in this
study. The maximum values shown for the receptors in this study will
probably be lower than those associated with a denser array of receptors,
because denser spacing of receptors would increase the probability of
locating a receptor near the offsite absolute maximum concentration.
4-29
-------
FIGURE 7
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
EXAMPLE OF SOURCE GROUPINGS INTO COMMON RELEASE SPECIFICATIONS
JP
O
(•S»'
I*]
©
jr 4? *
\ iT J ¦ K1
u n o n n
«r
n n n n n n
gboup SOURCE TYPE specific release pqiNTS
1 STACK A
2 STACK B
3 VOLUME SOURCE C.D.E.F.G.H
(BUILDING)
4 VOLUME SOURCE I.J.K.L.M.N
(BUILDING)
5 AREA SOURCE O.P.Q
-------
For each receptor within a valley zone, annual average concentrations
were estimated based on the impacts from each source that emits
pollutants addressed in this study. Then the impacts from each
industrial source within each valley zone were summed at each receptor to
estimate a total for annual average ambient concentrations. Two types of
receptors were modeled. First, a set of receptors was placed at pre-
selected locations in neighborhoods expected to be more affected by
releases. Second, a uniform grid was established in each zone as shown
in Figure 6. Four equally spaced receptors within a grid were used to
compute an average concentration for each grid cell (2.5 km by 2.5 km).
The exposure to the average individual is based on the concentrations
that were weighted by population in each grid. To determine the exposure
to the average exposed individual, these population-weighted
concentrations are added and the sum is divided by the total population
of the zone.
The approach used to link concentration fields with population is
consistent with the approaches used in the EPA Graphical Exposure
Modeling System (GEMS) and Human Exposure Model (HEM). One exception is
the treatment in HEM for receptors within 3.5 km of a source. Within
3.5 km, HEM assigns population to model receptors, a different approach
from the treatment used here or in GEMS. The effect of this difference
on population-weighted averages is likely to be small.
Classification of Land Use: For the purposes of dispersion modeling,
an area can be characterized as "rural" or "urban"; the distinction can
be made on the basis of either population density or land use. This
study used the preferred land-use method (Auer 1978) of characterizing
all areas within a 3 km radius of each industrial zone. The results are
as follows:
Valley Zone Percent Urban Percent Rural
Belle 6 94
Charleston 25 75
Institute 14 86
Nitro 14 86
Since all areas were considerably more than 50 percent rural, all areas
were classified as rural for modeling purposes.
Summary of Input Data: Technical Appendix E documents the specific
input files used for the ISCLT modeling analysis.
Uncertainties and Limitations
Modeled ambient concentrations provide the basis, with toxicological
information, for estimating potential life-time incremental risks to
4-31
-------
individuals and the population within this study. It is important to
understand the limitations of the modeled concentrations that provide the
basis for calculating these risk estimates.
Uncertainty in Emissions: The modeling results are critically
dependent on the emissions inventory compiled for 1984 by the State of
West Virginia. The inventory details the pollutants and the various
associated quantities for the numerous release points in the Kanawha
Valley with estimates of emissions based on generally accepted
engineering calculation procedures performed by the sources themselves.
Some comments have been made regarding the applicability of generic EPA
emissions factors and techniques to specific processes in the Kanawha
Valley. Fugitive emissions factors and techniques developed for the
chemical industry overall have been used for estimating emissions from
specific process units. In some instances, the resulting estimates for
fugitive emissions represent a large contribution to the modeled ambient
concentrations of the pollutants being studied. Industry has suggested
that, for some pollutants, emissions estimates for fugitive releases may
be an order of magnitude or more higher than actual emissions.
Fugitive emissions for these pollutants are, at times, the
predominant contributors to the predicted ambient concentrations. This
report attempts to separate out fugitive emissions for the major
pollutants within the study to understand the importance of such releases
for the risk assessment.
In addition, the emissions inventory is based on the calendar year
1984. Although the WVAPCC has updated the emissions for several
pollutants to reflect the more current operating status of the major
facilities, most emissions reflect operating status as of 1984.
Furthermore, voluntary reductions reflected in the inventory have not yet
been documented to EPA.
Uncertainty in Modeling: Gaussian modeling results are generally
considered to be accurate within a factor of two on a seasonal or annual
basis. For neighborhoods adjacent to the industrial complexes, a factor
of two would still appear to be reasonable. As downwind distance
increases, the Gaussian model treatment would act to systematically
underestimate concentration because of the lack of a term in the models
to represent valley wall reflection. Therefore, as distance increases,
it would be expected that the accuracy of Gaussian modeling would
decrease.
The Gaussian modeling analysis does not consider interzone transport
of pollutants. Although this does not significantly affect the analysis
of concentrations in neighborhoods surrounding the industrial facilities,
interzone transport could affect the estimate of the average individual
exposure throughout the study area. Valley wall reflection and flow
4-32
-------
between valley zones will act to provide higher concentrations than the
ISCLT and LONGZ modeling performed for each zone independently. By using
a box model approach, which assumes uniform concentration within the
volume of the full extent of the valley within the study area (i.e., 45
km long, by 1 km wide and 0.3 km high), it is possible to conservatively
estimate population-weighted concentrations. This study performed a
bounding analysis to determine estimates of upper-bound concentrations of
studied pollutants based on a box model approach, which accounts for such
transport.
Dispersion Modeling Results
This section presents the results of the annual average model
analysis. Tables 5 through 8 present the predicted annual average
concentrations from industrial facilities for each pollutant in
neighborhoods within each valley zone. Figures 2 through 5, previously
shown, identify the locations of each neighborhood receptor. Population-
weighted average concentrations for industrial sources are presented in
Tables 9 through 12 for each pollutant and valley zone. The minimum and
maximum concentrations based on this uniform grid system (four receptors
per 2.5 km grid zone) are also displayed.
Because ISCLT cannot adequately model county-wide area sources, which
in this case are 2500 m wide, the concentrations for county-wide area
source emissions are not given in Tables 5 to 8. Box model estimates
were used to compute conservative estimates of county-wide area sources.
Table 13 presents these results. Refer to Technical Appendix F for a
comparison of the conservatism in the box model compared to a Gaussian
treatment.
Because of the conservative assumptions used to specify the vertical
extent of the box, in this case 100 m, and the assumption of no loss of
mass across the valley walls, the box model results are likely to be
higher than actual impacts. LONGZ, however, can predict impacts for
county-wide area sources. LONGZ results can therefore be used to show
the relative contributions from industrial and county-wide area sources
based on Gaussian modeling. (Refer to Technical Appendix A.)
Sensitivity Analyses
Sensitivity analyses were conducted to explore the effects of model
inputs on predicted concentrations. The following were considered:
source release type (area source, building source, or point source),
atmospheric decay, and dispersion coefficients. Sensitivity analyses
were made only for the Belle, Charleston/South Charleston, and Institute
zones because concentrations in the Nitro zone are relatively low
compared to those in the other three. In addition, uncertainties with
4-33
-------
TABLE 5
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PREDICTED CONCENTRATIONS IN ug/m3
TOTAL CONCENTRATION AT NEIGHBORHOOD RECEPTOR SITE
BELLE
COMPOUND
ICA33CM TICH L0ROFOIFORMALDEIMETHYLENI
lETRACHLOlRM IHYDE IE CHLORII
4- 1
(ISCLT)
1 2.3C3S
17.16591
1.475676
19.24479
4- 2
(ISCLT)
12.59149
18.12297
1.529156
110.8371
4- 3
(ISCLT)
13.01067
19.68793
1.539595
1 13.094
4- 4
(ISCLT)
128.4684
199.8666
11.00269
1117.002
4- 5
(ISCLT)
13.70601
113.6867
1.405509
120.4641
4- 6
(ISCLT)
13.81577
113.3193
1.374677
117.4906
4- 7
(ISCLT)
12.84737
19.62734
I.301089
111.6548
4- 8
(ISCLT)
12.03969
16.84242
1.261475
18.21528
4- 9
(ISCLT)
1.956009
1 3.16S7
1.136497
13.94895
4-10
(ISCLT)
11.23694
14.14253
1.155636
15.00111
4-11
(ISCLT)
1 1.3549
14.61279
1.210001
16.10933
4-12
(ISCLT)
1.833441
12.69129
1.194972
13.53382
4-13
(I3CLT)
11.46117
14.97915
1.197985
16.31872
4-14
(ISCLT)
11.74765
1 5.6652
1.172773
17.02454
4-15
(ISCLT)
11.57092
|5.00544
10.28446
16.20416
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
-------
TABLE 5 (CONTINUED)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PREDICTED CONCENTRATIONS IN ug/m^
CONTRIBUTED BY FACILITY
Occldental
Dupont Belle
COMPOUND
CO
cn
1 CARBON TlCHLOROFOlMETHYLENl
1ETRACHLOlRM IE CHLORIt
k 1
ISCLT)
12.00777
17.02782
18.14668 1
<¦ 2
: isclt)
1 2.27151
! 7.9746
19.26035 1
'¦ - 3
(ISCLT)
12.71599
19.54677
111.1735 1
<•- 4
(ISCLT)
!28.0801
1 99.6S3
1116.706 I
4 5
1ISCLT)
13.51831
113.5332
120.1072 I
4- 6
(ISCLT)
13.64452
113.2317
117.2073 I
4 7
(ISCLT)
12.71233
19.56027
111.4977 1
4 8
(ISCLT)
11.92609
16.7.3704
18.08942 I
4- 9
(ISCLT )
1.6S9354
13.15667
13.36211 I
4- 10
(isclt)
1 1. 1658
14.10745
14. t38fl 17 1
4-.1
t isclt )
11.25706
14.56439
1 5.9!i41 |
4- .2
(ISCLT)
1.727527
12.63973
13.35816 1
4-13
(ISCLT)
I 1.391)2
14.94213
16.13353 1
4-14
(ISCLT)
11.594(15
15.59151
1 6.5572 I
4-15
(ISCLT)
11.40853
14.92877
15.72872 1
ICAPBOH TlCHLOROFOlFC""ALDE
1 ETRACHLO 1 RI1 I HYDE
METHYLEN
E CHLORI
4- 1
(ISCLT)
I .296034
10.13809
1 .475676
1.0981
4- 2
(ISCLT)
I.316305
1.148369
1.529156
1.57674
4- 3
(ISCLT)
10.29468
1.141163
1.589595
1.92049
4- 4
(ISCLT)
!.380304
1 .:?03617
11.00269
.295919
4- 5
(ISCLT)
1.187191
1.103467
1 .405809
.276911
4- 6
(ISCLT)
10.17125
1.CS7613
1 .374677
.203292
4- 7
(ISCLT)
1.135042
1 .067073
1.301089
. 157014
1
CD
(ISCLT)
1.113601
1 .055378
1.261475
.125859
4- ?
(ISCLT)
1 .066156
1 .032037
1.136497
0.08684
4-10
(ISCLT)
1 .071136
1 .035078
1.155636
.112937
4-11
(ISCLT)
1 .097043
1 .043399
1 .210001
.155726
4-12
(ISCLT)
1.105615
i .051557
1.194972
.175659
4-13
(ISCLT)
1.071074
1 .037021
1.197985
.185191
4-14
(ISCLT)
1.153607
1 .073685
1.272773
.467343
4-15
(ISCLT)
I.162336
1.076679
10.28446
.475445
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
-------
TABLE 6
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PREDICTED CONCENTRATIONS IN ug/m3
TOTAL CONCENTRATION AT NEIGHBORHOOD RECEPTOR SITE
CHARLESTON/SOUTH CHARLESTON
COMPOUND
IACRYLONII ETHYLENE IMETHYLENIPROPYLENIVINYLIDE
ITRILE I OXIDE |E CHLORIlE OXIDE |NE CL (1
3- 1
(ISCLT)
1.561643
12.82028
10.03708
15.24214
1.010551
3- 2
(ISCLT)
|0.343>7
11.70196
10.01786
13.44946
1.005825
3- 3
(ISCLT)
10.45022
11.40272
1 .017318
14.78489
1.005583
3- 4
(ISCLT)
I.891218
14.57533
1.019908
114.8434
1 .010041
3- 5
(ISCLT)
11.22263
13.66757
1 .025567
15.71892
1.015237
3- 6
(ISCLT)
1 0.9229
1 2.3589
1.025446
13.61081
1.010749
3- 7
(ISCLT)
1 .7218V+
11.70707
1.027619
12.26287
1.008678
3- a
(ISCLT)
I 10.12
13.14244
10.06748
13.31826
1.017876
3- 9
(ISCLT)
13.83297
14.87921
1.083375
14.06942
10.02458
3-10
(ISCLT)
14.39437
19.65421
1.053463
17.92983
1.038734
3-11
(ISCLT)
111.2708
127.2998
1.046699
115.0134
1.072409
3-12
(ISCLT)
113.42,38
146.0279
1 .035+04
140.7283
1 .286163
3-13
(ISCLT)
15.30795
116.1771
1.076571
110.4349
1 .062038
3-14
(ISCLT)
14.82406
127.0914
1 .048931
123.2062
1.089732
3-15
(ISCLT)
11.35267
11'+. 1306
1.028777
125.4862
1.025141
3-16
(ISCLT)
1 1.6677
111.2468
1.P39J84
114.7202
10.03229
3-17
(ISCLT)
11.82633
1 13.3352
1.067077
110.7268
1.034677
3-18
(ISCLT)
11.76376
11.48224
1.041647
11.79892
1.009362
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
-------
TABLE 6 (CONTINUED)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PREDICTED CONCENTRATIONS IN ug/m3
CONTRIBUTED BY FACILITY
Carbide South Charleston yech center
CO
-^1
CGMFO'JND
IACR Y LC1MI | ETHY LENEIMETHY LENI PROPY LENIVINY LID E
ITRILH I OXIDE IE CHf.ORIlE OXIDE |NE CL (1
3-
3-
3-
3-
3-
3-
3-
3-
3-
3-
3-
3-
3-
3
3
3
3
3
1 (ISCLT)
2 (ISCLT)
3 (ISCLT)
4 (ISCLT)
5 (ISCLT)
6 (ISCLT)
7 (ISCLT)
8 (ISCLT)
9 (ISCLT)
10 (ISCLT)
11 (ISCLT)
12 (ISCLT)
13 (ISCLT)
14 (ISCLT)
15 (ISCLT)
16 (ISCLT)
-17 (ISCLT)
-18 (ISCLT)
.560086 12.81666 I
.342473 11.70017 I
0.44942 11.40094 I
.890283 14.57326 I
I.22148 13.66516 I
.921744 12.35638 I
.720571 11.70417 I
110.1168 13.13463 I
3.82894 14.86983 I
14.39232 19.64551 I
111.2687 127.2955 |
113.4271 146.0241 I
15.30467 116.1697 |
14.82181 127.0562 I
II.35137 114.1275 I
11.66584 111.2424 I
11.82365 I 10.3286 I
11.76183 11.47758 I
7E-06 |E."^214 1.007705
6E-06 I 3.'(':: v46 i . 0!l44f.7
7E- 06 14.73483 I .004206
+ +
1E-C5 114.8434 10.00845
2E-05 15.71892 1.013261
2E-05 13.61081 I.008731
4E-05 12.26287 1.006419
OE-05 13.31825 1.011943
7E-05 14.06941 1.017295
2E-04 17.92982 1.034289
6E-05 115.0134 1-068757
3E--05 140.7283 |0.28327
3E-05 110.4349 1.056649
2E-05 123.2062 1.085817
1E--05 125.4862 1.022831
1E-05 114.7202 1.029014
1E-05 110.7267 1.029514
5E-05 11.79892 1.005782
COMPOUND
IACRYLONIIETHYLEK
I TRUE ( OXIDE
c I MHTHVLE.Nl PRCPYLENl VIHYLIDE I
IE CHLORIIE OXIDE |NE CL (II
3- i
3- 2 (1; I : ;
3- 3 (ISCLT)
3- 4 (ISCLT)
3- 5 (ISCLT)
3- 6 (ISCLT)
3- 7 (ISCLT)
3- 8 (ISCLT)
3- 9 (ISCLT)
3-10 (ISCLT)
3-11 (ISCLT)
3-12 (ISCLT)
3-13 (ISCLT)
3-14 (ISCLT)
3-15 (ISCLT)
3-16 (ISCLT)
3-17 (ISCLT)
3-18 (ISCLT)
, !K:1556 ! . "07
3E-04 I.
8E-04 I.
(itii >V2
001779
I.037073 I
1.017854 I
1.017311 |
4E-06
2E-06
2E-06
1.002846 I
1.001398 I
1.001377 I
9E-04 I.
001145 I.
.001156 I.
001273 I.
003169 I.
002076
00 2403
002521
002901
007814
I. 0.19097 I
I .025543 I
1.025422 |
I". 027583 I
1.067404 I
2E-06
3E-06
3E-06
3E-G6
8E-06
1.001591 I
1.001976 I
1.002018 I
1.002259 I
I .005933 I
004025 I.
002554 I.
002137 I.
001709 I.
003281 |.
009371
005702
004331
003798
007413
I.083504 |
I.053275 I
+ +-
I.046642 I
1.035379 I
1.076538 I
10E-06
6E-06
5E-06
4E-06
7E-06
1.007284 I
1.004445 I
1.003652 I
1 .002893 1
1.005439 I
002253 I
001296 I
001861 I
002883 I
001885 I
.005211
.003111
.004363
006619
0C4662
1.043912 |
I.028765 1
1.039073 I
+ + -
1.067066 I
1.041599 I
5E-06
3E-06
4E-06
7E-06
5E-06
1.003915 I
1.002311 I
+ 1
1.003276 1
+ +
I.005163 1
+ +
10.00358 I
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
-------
TABLE 7
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PREDICTED CONCENTRATIONS IN ug/m3
TOTAL CONCENTRATIONS
INSTITUTE
COMPOUND
lACRYLONII BENZENE
! TRILE 1
ICHLOROFOI ETHYLENE
IRM 1 OXIDE
FORMALDE1METHYLENIPROPYLEN
HYDE IE CHLORllE OXIDE
1,3-B'JTAl
OIENE 1
2- 1
(ISCLT)
16.56565
1 .031644
17.48184
14.4919
.012949
2.59974
110.5372
21.2178 I
2- 2
(ISCLT)
11.31098
10.00545
1 3. 034 08
6.48205
.007104
1.18378
15.11137
3.01951 1
2- 3
(ISCLT)
1.862691
1.004247
16.261C4
7.30372
.010967
3.00909
1 4.4901
1.4678 I
2- 4
tISCLT)
1 .721332
1.003876
16.38188
6.28709
.014694
2.42746
13.66392
1.C6844 I
2- 5
tISCLT)
10.62025
1.003911
14.66446
5.07301
.014976
1.4437
13.14118
.849235 1
2- 6
(ISCLT)
10.50318
1.003022
12.96-79
3.76416
.012407
1.28641
12.40557
.747657 1
2- 7
(ISCLT)
10.48574
1.003017
12.69C98
3.16921
.007407
.8793C8
12.09304
.679536 I
2- 8
(ISCLT)
1 .401021
1 .002667
I1.98C69
2.59141
.007036
.609127
11 .74668
.528887 I
2- 9
(ISCLT)
1.281138
1.002083
1 1.0 754
1.52315
.003418
0.31214
11.06631
.351693 I
2-10
(ISCLT)
1.250313
1.001876
1.897757
1.34982
0.00299
.275497
1.918914
.314766 1
2-11
(ISCLT)
1.199069
1.001539
1 .662=93
.978149
.001953
.200141
1 .632414
.248728 I
2-12
(ISCLT)
i0.3S929
1.002484
1 2.2374
2.44371
.006321
.712993
11.58022
.568193 1
2-13
(ISCLT)
!.843106
1.004155
13.04252
5.19491
.038778
.512306
14.42911
1.4104 1
2-14
(ISCLT)
11.02914
1.006388
12.58381
12.9568
.040425
.593078
1 3.8409
1.08941 1
2-15
(ISCLT)
12.48409
1 .024218
15.83403
18.977
.034426
1.35598
19.61468
1.12071 I
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
-------
TABLE 8
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
TOTAL CONCENTRATIONS
NITRO
COMPOUND
IACRYLONIIBHNZENE
ITRILE I
I CARBON TICHLOROFOlFORMALDE iMETHYLENlVINYL CH
IETRACHLOlRM I HYDE IE CHLORIILORIDE
1- 1
(ISCLT)
2E-04
3E-05
3E-05
3E-05
.668361
2E-05
2E-05
10E-04
.021722
1.06686 1
1- 2
(ISCLT)
IE-04
3E-05
3E-05
3E-05
0.3782
1E-05
1E-05
8E-04
.020097
0.78389 1
1- 3
(ISCLT)
3E-04
5E-05
5E-05
1 5E-05
.457446
3E-05
3E-05
.001785
.046332
1.63816 I
1- 4
(ISCLT)
7E-04
1E-04
1E-04
1E-04
.934112
6E-05
6E-05
.004265
.115492
4.87016 I
1- 5
IISCLT)
.001024
2E-04
2E-04
I 2E-04
.956076
9E-05
9E-05
.006249
.172869
4.71002 !
1- 6
(ISCLT)
1 .001607
3E-C4
3E-04
1 3E-04
0.71295
1E-04
1E-04
.010399
.301345
2.76505 I
1- 7
(ISCLT)
I.003S13
7E-04
7E-04
I 7E-04
.697067
3E-04
3E-04
.021522
.582253
1.36783 I
1- 8
(ISCLT)
.005915
.001067
.001067
1.001067
.859498
5E-04
5E-04
.033418
.893875
1.1818 I
1- 9
(ISCLT)
1.002759
5E-04
5E-04
5E-04
.449896
2E-04
2E-04
.020166
.689724
.877357 1
1-10
(ISCLT)
1 .001114
2E-04
1 2E-04
1 2E-04
.256381
9E-05
9E-05
.007882
.256527
.713345 1
1-11
(ISCLT)
! 6E-04
9E-05
1 9S-05
9E-05
.172398
5E-C5
5E-05
.003984
.129957
.553317 1
1-12
(ISCLT)
1 4E-04
6E-05
6E-05
1 6E-05
.136385
3E-C5
3E-05
.002699
.087008
.464994 1
1-13
(ISCLT)
4E-04
7E-05
7E-05
I 7E-05
.175402
3E-05
3E-05
.002772
.083565
.474901 1
1-14
(ISCLT)
! 5E-0'»
8E-05
! 8E-05
1 8E-05
.227896
4E-05
4E-05
.003549
.119677
1.543989 1
ETHYLENE IALLYL CHITRICHLOR
CHLORID|LORIDE IOETHYLEN
Source: Regulatory Integration Division, Office of Policy Analysi
lysis, EPA, 1987
-------
TABLE 8 (CONTINUED)
PREDICTED CONCENTRATIONS IN ug/m3
BY FACILITY
NITRO
Artel
Monsanto
o
COMFOUND
AC3YLQNI1 BENZENE
TRILE I
1 (ISCLT)
2 (ISCLT)
3 (ISCLT)
4 (ISCLT)
5 (ISCLT)
6 (ISCLT)
7 (ISCLT)
8 (ISCLT)
- 9 (ISCLT)
-10 (ISCLT)
-11 (ISCLT)
-12 (ISCLT)
-13 (ISCLT)
-14 (ISCLT)
ICAneON T
lETHACHLO
2E-P4 I
1E-04 j
3E-05
3E-05
I 3E-05
I 3E-05
3E-04 !
7E-04 I
5E-05
1E-04
i 5E-05
I 1E-04
001024 I
001607 I
2E-04
3E-04
I 2E-04
I 3E-04
.00381.3 I
+¦
.005915 I
7E-04
001067
I 7E-04
+
1.001067
002759 I
001114 I
5E-C4
2E-04
I 5E-04
I 2E-04
6E-04 I
4 E - 04 I
9E-05
6E-05
I 9E-05
I 6E-05
4E-04 I
n:-04 |
7E-05
8E-05
I 7E-05
I 8E-05
CHLOROFOlFORMALDEIMETHYLEN!VINYL CHI ETHYLENE 1ALLYL CH|
RM I HYDE IE CHLORIILORIDE I CHLORIDILORIDE I
3E-05
3E-05
5E-05
1E-04
2E-04
3E-04
7E-04
.001067
5E-04
2E-04
9E-05
6E-05
7E-05
8E-05
0.02133 I
.015996 I
2E-05 I
1E-05 I
2E-05
1E-05
032522 I
0.07637 I
3E-05 I
6E-05 1
3E-05
6E-05
.110699 I
.185525 I
9E-05 I
1E-04 I
9E-05
1E-04
398581 |
.614097 I
3E-04 I
5E-04 I
3E-04
5E-04
.204377 |
.1.10308 I
2E-04 I
9E-05 I
2E-04
9E-05
.056183 I
.038293 !
5E-05 |
3E-05 !
5E-05
3E-05
.040934 I 3E-05 I
I.055317 I
4E-05 I
3E-05
4E-05
10E-04
8E-04
001785
.004265
006249
.010399
.021522
033418
,020166
007832
003984
,002699
002772
.003549
.021722 I
.020097 I
.046332 I
.115492 I
.172869 I
.301345 I
.582253 I
.893875 I
.689724 I
.256527 |
.129957 I
.087008 I
.088565 I
.119677 I
1- 1 (ISCLT)
1- 2 (ISCLT)
1- 3 (ISCLT)
1- 4 (ISCLT)
1- 5 (ISCLT)
1- 6 (ISCLT)
1- 7 (ISCLT)
1- 8 (ISCLT)
1- 9 (ISCLT)
1-10 (ISCLT)
1-11 (ISCLT)
1-12 (ISCLT)
1-13 (ISCLT)
1-14 (ISCLT)
IFORMALDEITRICHLORI
IHYDE lOUTHYLENl
0.64703 11.06686 I
.362204 10.78389 I
.424924 11.61816 I
. 907742 14.87016 I
.£'+5377 14.71002 I
.527425 I 2.765C5 I
.29S437 11.36783 I
.245401 I 1.1818 I
.165519 1.077357 1
.146073 1.713345 1
.116215 I .*>53317 1
.098092 1.464994 I
.134468 1.474901 1
.172579 1.543989 1
-------
TABLE 9
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONCENTRATIONS TO THE AVERAGE EXPOSED INDIVIDUAL
(ug/m3)
BELLE
Population for Belle: 15,530
COMPOUND
MINIMUM
MAIN GRID
CONCENTRATION
AVERAGE EXPOSED
INDIVIDUAL
CONCENTRATION
MAXIMUM
MAIN GRID
(POPULATION-WEIGHTED) CONCENTRATION
Carbon tetrachloride 8.70E-02
Chloroform 2.78E-01
Formaldehyde 1.48E-02
Methylene chloride 3.16E-01
2.36E+00
7.86E+00
4.76E-01
9.19E+00
1.33E+01
4.50E+01
2.97E+00
5.22E+01
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
4-41
-------
TABLE 10
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONCENTRATIONS TO THE AVERAGE EXPOSED INDIVIDUAL
(ug/m3)
CHARLESTON/SOUTH CHARLESTON
Population for Charleston/South Charleston: 51,750
MINIMUM
MAIN GRID
COMPOUND CONCENTRATION
AVERAGE EXPOSED
INDIVIDUAL MAXIMUM
CONCENTRATION MAIN GRID
(POPULATION-WEIGHTED) CONCENTRATION
Acrylonitrile 8.40E-02
Ethylene oxide 1.50E-01
Methylene chloride 1.37E-02
Propylene oxide 1.80E-01
Vinylidene chloride
(1,1-Dichloroethene) 1.79E-03
9.28E-01
2.19E+00
5.01E-01
2.09E+00
7.53E+00
3.01E+01
1.09E+01
2.03E+01
2.76E-02
3.61E-01
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
4-42
-------
TABLE 11
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONCENTRATIONS TO THE AVERAGE EXPOSED INDIVIDUAL
(ug/m3)
INSTITUTE
Population for Institute: 22,390
MINIMUM
MAIN GRID
COMPOUND CONCENTRATION
AVERAGE EXPOSED
INDIVIDUAL MAXIMUM
CONCENTRATION MAIN GRID
(POPULATION-WEIGHTED) CONCENTRATION
Acrylonitrile 9.71E-02
Benzene 8.29E-04
Chloroform 2.55E-01
Ethylene oxide 4.13E-01
Formaldehyde 7.14E-04
Methylene chloride 6.94E-02
Propylene oxide 2.77E-01
1,3-Butadiene 1.03E-01
1.61E+00
1.74E-02
2.41E+00
5.81E+00
9.21E-03
6.67E-01
3.57E+00
1.28E+00
6.74E+00
8.90E-02
1.96E+01
3.06E+01
1.40E-01
2.94E+00
3.84E+01
6.53E+00
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
4-43
-------
TABLE 12
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONCENTRATIONS TO THE AVERAGE EXPOSED INDIVIDUAL
(ug/m3)
NITRO
Population for Nitro: 9,990
AVERAGE EXPOSED
MINIMUM INDIVIDUAL MAXIMUM
MAIN GRID CONCENTRATION MAIN GRID
COMPOUND CONCENTRATION (POPULATION-WEIGHTED) CONCENTRATION
Acrylonitrile
5.
02E-
¦05
8.40E-04
4,
.01E-03
Benzene
8.
52E-
¦06
1.47E-04
7,
.03E-04
Carbon tetrachloride
8.
52E-
-06
1.47E-04
7.
.03E-04
Chloroform
8.
52E-
•06
1.47E-04
7.
.03E-04
Formaldehyde
3.
95E-
-02
4.08E-01
1.
.59E+00
Methylene chloride
4.
26E-
¦06
7.36E-05
3,
.52E-04
Vinyl chloride
(chloroethene)
4.
26E-
-06
7.36E-05
3,
.52E-04
Ethylene chloride-EDC
3.
45E-
-04
5.23E-03
2,
.51E-02
Allyl chloride
9.
62E-
-03
1.51E-01
7,
. 39E-01
Trichloroethylene
1.
28E-
-01
1.77E+00
1.
.01E+01
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
4-44
-------
Table 13
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
COUNTY-WIDE AREA SOURCES BASED ON BOX MODEL ANALYSIS
Emissions (County-Wide Area Sources) Modeled Concentration
(metric tons) (ug/m^)
Box Model
Pol 1utant
Zone
Bel le
Charleston
Institute
Ni tro
Bel 1 e
Charl eston
Inst.
Ni tro
Benzene
14.4084
41.7199
18.7094
7.31173
3.0
4.8
2.4
0.9
Perchloroethylene
12.5941
36.4665
16.3535
6.39104
2.4
4.2
2.1
0.9
Methylene chloride
7.03098
20.3584
9.12978
3.56796
1.5
2.4
1.2
0.3
Formaldehyde
6.6573
19.0323
8.5351
3.33555
1.2
2.1
1.2
0.3
-1
Tri chloroethylene
1.67983
4.86398
2.18127
0.85245
0.3
0.6
2.9xl0_1
1.0x10"
1,3-Butadiene
0.10648
0.308316
0.138265
0.054035
2.lxl0_2
3.6x10 '
1.8xl0~2
6.6x10
-3
Ethylene chloride
0.037012
0.10717
0.048061
0.018782
7.2x10~3
-2
1.2x10
-3
6.3x10
2.3x10
-3
Ethylene bromide
0.017362
0.050272
0.022545
0.008811
3.3xl0~3
-3
5.7x10
3.0xl0~3
1 .lxio'
-3
Arseni c
0.012881
0.037296
0.016725
0.006536
2.6xl0~3
4.2xl0~3
2.2xl0~3
7.8x10
-4
Benzo(a)pyrene
0.005412
0.015671
0.007028
0.002746
-3
1 .1x10
1.7xl0~3
-4
9.0x10
3.0x10
-4
Cadmi urn
0.004375
0.012667
0.005681
0.00222
-4
8.7x10
-3
1 .4x10
-4
7.5x10
2.7x10
-4
Beryl 1i um
0.00040
0.001265
0.00060
0.0002
-5
7.8x10
-4
1 .4x10
-5
7.8x10
2.4x10
-5
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
-------
TABLE 13 (CONTINUED)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
INCREMENTAL CONCENTRATIONS FROM AREA SOURCE CATEGORIES
POI 1 IITANT
B*
ROAD VEHICLES
I* C*
N*
B
SOLVENT USAGES
I C
N
B
HEATING
I
C
N
Benzene
2.9
2.3
4.6
0.9
0.0
0.0
0.0
0.0
1.3xl0-2
1.OxlO-2
2.OxlO2
3.8xl03
Perchlorehylene
0.0
0.0
0.0
0.0
2.4
2.1
4.2
0.9
0.0
0.0
0.0
0.0
Methylene chloride
0.0
0.0
0.0
0.0
1.5
1.2
2.4
0.3
0.0
0.0
0.0
0.0
Formaldehyde
0.7
0.7
1.3
0.2
0.0
0.0
0.0
0.0
0.5
0.5
0.8
0.1
Tri chloroethylene
0.0
0.0
0.0
0.0
0.3
0.3
0.6
0.1
0.0
0.0
0.0
0.0
1,3-Butadiene
2.1xl0"2
1.8xl0-2
3.6xl0-2
6.6xl0~3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Ethylene chloride
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ethylene bromide
2.6xl0-3
2.4xl0-3
4.5xl0-3
8.6xl0-4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Arseni c
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.4x10-3
2.1x10-3
3.9xl0-3
7.3x10-4
Btnzo(a)pyrene
4.5x10-4
4.0xl0-4
7.6xl0~4
1.4xl0-4
0.0
0.0
0.0
0.0
6-OxlO-4
5.3x10-4
1 .OxlO-3
1.9x10-4
Cadmi um
1.6xl0-4
1.4xl0-4
2.5xl0-4
4.9xl0-5
0.0
0.0
0.0
0.0
6.8xl0-4
5.8x10-4
1 .lxlO-3
2.1x10-4
i Beryllium
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
7.8xl0~5
6.5xl0-5
1 . lxlO-4
2.4xl0-5
en
*B, I, C, N represent Belle, Institute, Charleston, and Nitro respectively.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
-------
TABLE 13 (CONTINUED)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
INCREMENTAL CONCENTRATIONS FROM AREA SOURCE CATEGORIES
GASOLINE MARKETING
WASTE OIL BURNING
POLLUTANT
B"
I
C
N
B
I
C
N
Benzene
9.2xl0~2
6.1x10
0.1
2.8xl0~2
8.3xl0~5
6.4x10
1,2xl0~5
2.5x10~6
Perchlorethylene
0.0
0.0
0.0
0.0
1.9xl0~5
2.6xl0~5
3.5xl0~5
-6
8.4x10
Methylene chloride
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Formaldehyde
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
T ri chloroethylene
0.0
0.0
0.0
0.0
1.8xl0~5
2.7xl0~5
4.9xl0~5
8.2xl0~6
1,3-Butadiene
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Ethylene chloride
7.2xl0~3
6.3xl0~3
1.2xl0~2
2.3xl0~3
0.0
0.0
0.0
0.0
Ethylene bromide
7.0xl0~4
-4
6.4x10
1.2xl0~3
-4
2.3x10
0.0
0.0
0.0
0.0
Arseni c
0.0
0.0
0.0
0.0
-4
1.6x10
-4
1.3x10
-4
2.5x10
4.8xl0~5
Benzo(a)pyrene
0.0
0.0
0.0
0.0
3.9xl0~6
2.7xl0~6
4.6x10~6
9.8xl0~7
Cadmi uin
0.0
0.0
0.0
0.0
4.0xl0~5
2.6xl0_5
5.5xl0~5
1 .lxl0~5
Beryl 1i um
0.0
0.0
0.0
0.0
1.6xl0~5
1.3x10-5
2.2xl0~5
-6
4.8x10
*B, I, C, N represent Belle, Institute, Charleston, and Nitro respectively.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987.
-------
respect to impacts on hillside receptors and the significance of
interzone transport were explored through additional quantitative
analysis using techniques described below.
For the purposes of the sensitivity analyses, a 100 m by 100 m
industrial area source was assumed to exist in the center of each zone.
A common set of upwind and downwind receptor locations was also specified.
The results of the sensitivity runs are summarized on Table 14. The
following can be inferred:
Release Types: For sensitivity evaluations of release types (area
source, building source, or point source) the following additional
release specifications were used.
Building source dimensions: 50 m x 50 m x 30 m.
Stack parameters: 5 m (height), 0.1 m (diameter), 298 K
(temperature), 0.1 m/sec (flow rate).
Within the first kilometer of plume travel, the study found that
treatment of source type can significantly influence model outputs. For
the examples shown here, modeling the test release as a building source
was found to produce nearby concentrations much lower than those derived
from the stack or area source treatments. This can be attributed to the
relatively high initial horizontal and vertical dispersion produced by a
building source. Based on these comparisons, modeling vent releases as
building sources rather than as stack sources can .reduce estimated
concentrations within 200 m of the source by as much as a factor of 3 or
4 for a large building. Past one kilometer, however, the importance of
the initial release conditions decreases.
Atmospheric Decay: Allyl choride, with a half-life of 4.8 hours, was
modeled with the fastest decay rate of any pollutant in this study. As
shown in Table 14, atmospheric decay for this chemical is, as expected,
relatively unimportant for the distances and decay rates associated with
pollutants under study here. The decay term produces little differences
in the results, especially for the neighborhoods within a kilometer of
the industrial clusters. A half-life of one hour or less would be needed
to make a substantial difference in the results for the scales of
interest in this study.
Dispersion Coefficients: Dispersion coefficients are used in
Gaussian modeling to estimate plume dimensions along the horizontal and
vertical axes as a function of downwind distance and atmospheric
stability. For annual averages, vertical dispersion coefficients can be
input, as they were here, on the basis of site-specific meteorological
4-48
-------
TABLE 14
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
SENSITIVITY RESULTS
USING ISLCT
CONCENTRATIONS IN UG/CU.
DOUNVALLEY DISTANCE FROM SORUCE !M)
UPVALLEY DISTANCE FROM SDURCE (M)
200
600
1000
2500
5000
200
600
1000
2500
5000
ZONE: BELLE
CLASSIFICATION
AREA
3.1
0.8
0.4
9.2E-2
3.3E-2
1.8
0.4
0.2
4.6E-2
1.6E-2
BUILDING
1.3
0.4
0.2
0.1
2.7E-2
0.8
0.2
0.1
3.6E-2
1.3E-2
POINT
4.1
0.9
0.4
0.1
3.4E-2
2.7
0.5
0.2
4.9E-2
1.6E-2
ATMOSPHERIC DECAY
(HRS)
00.5
2.6
0.6
0.2
3.2E-2
6. IE—3
1.5
0.3
0.1
1.9E-2
3.7E-3
05.0
3.1
0.3
0.3
8.1E-2
2.6E-2
1.8
0.4
0.2
4. IE—2
1.3E-2
25.0
3.1
0.8
0.4
3.9E-2
3. IE—£
1.3
0.4
0.2
4.5E-2
1.5E-2
ZONE: CHARLESTON
AREA
1.7
0.3
0.1
3.7E-2
1.3E-2
0.7
7.9E-2
3.0E-2
6.SE-3
2.2E-3
BUILDING
0.6
0.2
9.4E-2
2.8E-2
1.1E-2
0.3
4.9E-2
2.2E-2
5.7E-3
1.3E-3
POINT
2.0
0.4
0.2
3.3E-2
1.3E-2
0.5
8.5E-2
3.4E-2
7.2E-3
2.3E-3
ATMOSPHERIC DECAY
(HRS)
00.5
1.4
0.2
8.7E-2.
1.2E-2
1.3E-3
0.5
5.5E-2
1.8E-2
2.5E-3
4.3E-4
05.0
1.7
0.3
0.1
3.3E-2
1.0E-2
0.7
7.6E-2
2.8E-2
6.1E-3
1.3E-3
£5.0
1.7
0.3
0.1
3.6E-2
1.2E-2
0.7
7.8E-2
3.0E-2
6.7E-3
2. IE—3
ZONE: INSTITUTE
CLASSIFICATION
AREA
3.7
1.0
0.4
0.1
4. IE—2
0.7
0.1
4.8E-2
1.2E-2
4.0E-3
BUILDING
1.3
0.4
0.2
8.3E-2
3.3E-2
0.3
6.2E-2
3.2E-2
9. IE—3
3.3E-3
POINT
5.2
1.2
0.5
0.2
4.2E-2
0.7
0.1
5.3E-2
1.2E-2
4.1E-3
ATMOSPHERIC DECAY
iHRS)
00.5
3.0
0.6
0.2
3.0E-2
4.5E-3
0.6
8.8E-2
3.3E-2
5.4E-3
1. IE—3
05.0
3.6
0.3
0.4
0.1
3. IE—£
0.7
0.1
4.6E-2
1.0E-2
3.3E-3
£5.0
3.7
0.9
0.4
0.1
3.3E-2
0.7
0.1
4.8E-2
1. IE—£
3.8E-3
4-49
-------
data; horizontal dimensions of the plume are specified by assuming a
uniformly mixed plume within a 22.5 sector. In this study, vertical
dispersion coefficients were computed based strictly on stability data in
ISCLT, and turbulent intensity data in LONGZ, for valley zones with
available measured data. Technical Appendix A provides a description of
the sensitivity of model output to the method of treating vertical
dispersion.
Hillside Receptor Analysis: The impacts of plumes on the hillside
were not considered to be a major issue for the pollutants studied. To
evaluate this assumption, a limited analysis was performed in the Belle
zone.
This example was selected because (1) chloroform is emitted in
relatively high quantities in Belle, and (2) there are residential areas
up the hillside directly across from major releases of chloroform in
Belle. The LONGZ model was used. As shown in Figure 8, for these
emissions associated with area sources, the impacts along the hillside
are one to two orders of magnitude less than maximum concentrations along
the valley floor. This is attributed to two factors: (1) flow along the
valley floor is far more common than crossvalley flow, and
(2) crossvalley flow is generally associated with relatively strong winds
and moderate dispersion.
Interzone Transport: Valley wall reflection and flow between
valley zones will act to provide higher concentrations from the box model
than the ISCLT and LONGZ modeling performed for each zone independently.
By using a box model approach, which assumes uniform concentration within
the volume of the full extent of the valley (i.e., 45 km long by 1 km
wide by 0.3 km high), one can conservatively estimate average
concentrations. Table 15 shows the average concentrations based on a
simple box model approach for the full 45 km2 study area, in comparsion
to population-weighted average concentrations based on ISCLT. As shown,
the box model approach was consistently more conservative than ISCLT.
Refer to Technical Appendix F for a summary of the inputs to the box
model estimates and a comparison of the box model treatment with the
preferred approach for modeling county-wide area sources in flat terrain,
the CIimatological Dispersion Model (CDM). Based on Technical Appendix
F, it appears that the box model approach is more conservative than a
Gaussian model treatment, such as CDM.
Ambient Monitoring Programs
Three ambient air quality monitoring data sets are summarized in this
section: (1) preexisting data, (2) the preliminary field program, and
(3) the ORD field program.
4-50
-------
FIGURE 8
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
V \ \
HILLSIDE RECEPTOR CONCENTRATIONS FOR CHLOROFORM IN
THE BELLE ZONE (ug/nT)
l\ Low«f/7/ */^Z/$
i.'ABdle' , CO": ;
irm»
1.3
/ iKO^
/ -£•••
:\v* si1** • !**—
\ ^*i - Tr vt> V v/ 1
^ x« v»K^ vV x.v / .s» /¦-*•-
f^2.0
*34.7
1/
2.1 >-4-10
X/J9S
4-51
-------
TABLE 15
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA AIR QUALITY REPORT
BOX MODEL CONCENTRATIONS FOR FULL STUDY AREA
TOTAL EMISSIONS MODELED CONCENTRATION Ratio:
(ug/m3)
POPULATION-WEIGHTED Concentration (Box Model)
POLLUTANT kkg/yr BOX MODEL CONCENTRATIONS BASED ON ISCLT* Concentration
Benzene
82
3.6
-
-
Perchl oroethylene
71.8
3.2
-
-
Methylene chloride
145.7
6.4
-
-
Formaldehyde
50.9
2.2
-
-
T ri chloroethylene
32.1
1.4
-
-
1,3-Butadiene
25.8
1.1
0.3
3.7
Ethylene chloride
0.26
0.01
-
-
Ethylene bromide
0.099
0.004
- •
-
Arseni c
0.07
0.003
-
-
Benzo(a)pyrene
0.03
1.3xl0~3
-
-
Cadmi um
0.025
1.lxlO-3
-
-
Beryl 1i um
2.5xl0~3
1.lxlO-4
-
-
Carbon tetrachloride
22.9
1.0
0.4
2.5
Chl oroform
134.3
5.9
1.8
3.3
Ethylene oxide
133.6
5.9
2.4
2.5
Acryloni tri1e
45.7
2.0
0.8
2.5
All y 1 chloride
2.0
9.0xl0-2
1.5x10"2
6.0
Vinylidene chloride
.48
2.0xl0-2
1.4xl0~2
1 .4
Vinyl chloride
5.10-4
2.02xl0-5
7.4xl0~6
2.7
Propylene oxide
115.9
5.1
1.9
2.7
Average 2.8
Comparisons are valid only for pollutants emitted primarily by industrial sources because ISCLT was not used to
model county-wide area sources.
-------
Preexisting Data: Comprehensive measured data for noncriteria
pollutants were not available for the Kanawha Valley study area.
Previous studies (EPA 1977, Pellizzari 1976, Pellizzari 1978) have
collected limited data on selected volatile organics. Somewhat more
extensive data are available for selected metals (through SAROAD).
Technical Appendix G presents a summary of data collected during these
programs.
Preliminary Field Program: During the week of December 9, 1985,
EPA's Region III Environmental Services Division conducted an ambient air
quality screening program in the four identified Kanawha Valley zones of
Belle, Institute, Nitro, and Charleston/South Charleston. The screening
program was short, with a limited scope and modest objectives. Details
of the program are outlined in a summary report dated March 6, 1986
(Guide 1986) (refer to Technical Appendix H.l.)
This program was designed to provide real time data at selected sites
within each valley zone in order to rank the importance of various
pollutants and to help in selecting locations for a 45-day monitoring
program, which was conducted by EPA's Office of Research and Development
during March and April of 1986.
Samples were collected in 5-liter multilayer air bags and were
analyzed by a Photovac Model 10A10 portable gas chromatograph (GC).
Sites targeted for gas chromatograph/mass spectroscopy GC/MS confirmation
were also sampled with 5-liter multilayer bags, which were returned to
the EPA Central Regional Laboratory for analysis.
Unlike the later monitoring program conducted in March and April in
which four fixed sites were sampled, the December field program collected
samples at a large number of sites that were selected primarily to
evaluate the maximum expected concentrations for the meteorological
conditions present during each sampling period.
The samples that were analyzed by GC and GC/MS were not in general
agreement; also, no clear pattern emerged to distinguish upwind and
downwind concentrations. There is no obvious explanation for these
inconsistencies. Quality control records did not indicate any
significant problems in the program. Because of difficulty in
interpreting these results, they were used only in a qualitative manner
as input for the more detailed monitoring study that followed.
Parallel with EPA's work, the WVAPCC collected five samples during
the December field program, some of which were concurrent with the Region
III samples (Engle 1986). These samples were collected on Tenax and
analyzed by GC/MS after being concentrated on a NuTech thermal desorption
system. These samples showed much lower concentrations than the Region
4-53
-------
Ill data. The results are presented in Tables 16 through 20. These
results were from samples taken generally downwind of the Union Carbide
South Charleston plant and the Nitro industrial complex (refer to
Technical Appendix H.2 for further information).
QRD Field Program: During March and April 1986, EPA's Office of
Research and Development (ORD) conducted a more detailed monitoring study
at four selected sites -- two in Institute and two in Belle. The study
had five objectives:
(1) To qualitatively and quantitatively confirm the presence of
targeted compounds in the ambient air;
(2) To determine whether a more in-depth emissions verification of
facilities and processes is needed;
(3) To assist in model selection for potential future phases of the
study;
(4) To help determine whether a more detailed monitoring program is
needed at a later date; and
(5) To evaluate a newly designed 24-hour cannister sampling system.
Twenty-four hour integrated air quality samples were obtained for
each site in the network during this program every third day over the 45-
day monitoring period (USEPA 1986c). Upvalley and downvalley monitors
were placed approximately 1 km from the fencelines of the Belle and
Institute complexes. The locations of these monitoring sites are shown
in Figures 2 and 4. The following compounds were routinely measured and
detected:
Methylene Chloride
Ethylene Chloride
Benzene
Trichloroethylene
Perchloroethylene
o-Xylene
Vinylidine chloride
Chloroform
1,1,1-Trichloroethane
Carbon tetrachloride
Toluene
Chlorobenzene
Vinyl chloride
Ethylene bromide
Refer to Technical Appendix H.3 for detailed information concerning
this program.
The ORD monitoring program produced a data set that appears in
general to be reasonably consistent with available emissions data and
4-54
-------
TABLE 16
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
VOLATILE ORGANIC COMPOUNDS COLLECTED ON TENAX BY APCC
LOCATION: No. Charleston WIND SPEED: 6 mph
DATE: December 6, 1985
WIND DIRECTION: 300° WNW
TIME: 10:18 a.m. - 2:
:18 p.m.
FLOW RATE
: 200 cc/min
FLOW RATE:
100 cc/min
CHEMICAL COMPOUND
(Vol:
48 liters)
Vol:
24 liters
Nq*
Cone. uq/m3**
Ng*
Cone. uq/m3
2-Chlorobutane
Trace
<0.5
Trace
<0.5
Chloroform
ND
ND
Benzene
61
1.27
36
1.5
Carbon Tetrachloride
Trace
<0.5
Trace
<0.5
Toluene
73
1.52
50
2.08
Tetrachloroethene
ND
ND
Ethyl benzene
Trace
<0.5
23
0.96
Benzaldehyde
112
2.33
77
3.21
Mesi tylene
Trace
<0.5
Trace
<0.5
Dichlorobenzene
233
4.85
413
17.2
* Nanograms deteched after correction for blank.
** ug/nw = Ng/liter
NOTE: ND denotes Nondetectable
4-55
-------
TABLE 17
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
VOLATILE ORGANIC COMPOUNDS COLLECTED ON TENAX BY APCC:
LOCATION: Institute WIND SPEED: 5.5 mph
DATE: December 9, 1985 WIND DIRECTION: 250°-255° WSW
TIME: 3:00 p.m. - 4:00 p.m.
FLOW RATE: 400 cc/min FLOW RATE: 200 cc/min
CHEMICAL COMPOUND (Vol: 24 liters) Vol: 12 liters
Nq* Cone. uq/m3** Nq* Cone. uq/m3**
2-Chlorobutane
Trace
<0.5
ND
Chloroform
ND
ND
Benzene
45
1.87
19
1.58
Carbon tetrachloride
ND
ND
Toluene
161
6.70
43
3.58
Tetrachloroethene
ND
ND
Ethyl benzene
22
0.92
Trace
<0.5
Benzaldehyde
121
5.04
60
5.00
Mesitylene
Trace
0.50
Trace
<0.5
Dichlorobenzene
37
1.54
ND
* Nanograms deteched after correction for blank.
** ug/m^ = Ng/liter
NOTE: ND denotes Nondetectable
4-56
-------
TABLE 18
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
VOLATILE ORGANIC COMPOUNDS COLLECTED ON TENAX BY APCC
LOCATION: Belle WIND SPEED: 4.5 mph
DATE: December 10, 1985 WIND DIRECTION: 155°-165° WSW
TIME: 10:00 p.m. - 11:00 p.m.
FLOW RATE: 400 cc/min FLOW RATE: 200 cc/min
CHEMICAL COMPOUND (Vol: 24 liters) Vol: 12 liters
Nq* Cone. uq/m3** Nq* Cone. ug/m3**
2-Chlorobutane
13
0.54
32
2.66
Chloroform
36
0.50
Trace
<0.5
Benzene
71
2.96
33
2.75
Carbon tetrachloride
33
1.37
Trace
<0.5
Toluene
91
3.79
30
2.50
Tetrachloroethene
ND
ND
Ethyl benzene
ND
NDe
Benzaldehyde
61
2.54
62
5.16
Mesitylene
Trace
<0.5
Trace
0.54
Dichlorobenzene
ND
ND
* Nanograms deteched after correction for blank.
** ug/m^ = Ng/1iter
NOTE: ND denotes Nondetectable
4-57
-------
TABLE 19
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
VOLATILE ORGANIC COMPOUNDS COLLECTED ON TENAX BY APCC
\
LOCATION: Nitro+ WIND SPEED: 2-4.8 mph
DATE: December 11, 1985 WIND DIRECTION: 210°-220° WSW
TIME: 11:40 a.m. - 12:40 p.m.
FLOW RATE: 400 cc/min FLOW RATE: 200 cc/min
CHEMICAL COMPOUND (Vol: 24 liters) Vol: 12 liters
Nq* Cone. uq/m3** Nq* Cone. uq/m3**
2-Chlorobutane
303
12.6
98
8.16
Chloroform
206
8.58
113
9.42
Benzene
206
8.58
94
7.83
Carbon tetrachloride
Trace
<0.5
24
2.00
Toluene
1044
43.5
309
25.7
Tetrachloroethene
1111
46.3
613
51 .0
Ethyl benzene
422
17.6
75
6.25
Benzaldehyde
127
5.29
71
5.92
Mesi tylene
120
5.00
30
2.54
Di chlorobenzene
161
6.70
101
8.42
* Nanograms deteched after correction for blank.
** ug/m^ = Ng/1i ter
+ Gress Equip. Co., Inc., Hubb Industrial Park
NOTE: ND denotes Nondetectable
4-58
-------
TABLE 20
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
VOLATILE ORGANIC COMPOUNDS COLLECTED ON TENAX BY APCC
LOCATION: Nitro+ WIND SPEED: 2-4.8 mph
DATE: December 11, 1985 WIND DIRECTION: 210°-220° WSW
TIME: 11:40 a.m. - 12:40 p.m.
FLOW RATE: 400 cc/min FLOW RATE: 200 cc/min
CHEMICAL COMPOUND (Vol: 24 liters) Vol: 12 liters
Nq* Cone. uq/m3** Nq* Cone. uq/m3**
2-Chlorobutane
63
3.15
25
2.5
Chloroform
87
4.35
59
5.9
Benzene
57
2.85
46
4.6
Carbon tetrachloride
32
1.60
19
1.9
Toluene
87
4.35
41
4.1
Tetrachloroethene
ND
ND
Ethyl benzene
ND
ND
Benzaldehyde
28
1.40
45
4.5
Mesi tylene
Trace
<0.5
Trace
<0.!
Di chlorobenzene
ND
ND
* Nanograms deteched after correction for blank.
** ug/m"^ = Ng/liter
+ State Hygenic Lab, 167-llth Ave., S. Chas. WV
NOTE: ND denotes Nondetectable
4-59
-------
meteorological conditions for the four pollutants that were compared with
meteorological data for the sampling days. The measured data
generally show the highest concentrations during periods with drainage
flows, some days with persistent flow oriented along the valley also
showed relatively high concentrations. During days with persistent flow
that resulted in monitoring sites being upwind of the industrial
complexes, concentrations were generally low, as would be expected.
For the pollutants present in the highest concentrations, methylene
chloride, chloroform, toluene, carbon tetrachloride and "unknown
compound," Table 21 summarizes the measured data by site and provides an
interpretation of the meteorological data. The results shown in Table 21
indicate that the average concentrations at Site 4 were found to be
higher than those at Site 3. Site 3 is approximately 0.45 km downvalley
of the Belle industrial complex and should therefore be more affected by
stable drainage flows. Site 4 is located approximately 0.8 km upvalley
of the industrial complex. The higher concentrations found at Site 4 are
consistent with the available emissions data, which show emissions of
chloroform and methylene chloride to be towards the upvalley border of
the complex.
Conversely, the toluene and chloroform concentrations at Site 2,
which is located approximately 2.5 km downvalley of the edge of the
Institute complex, were generally higher than those at Site 1, which is
approximately 0.9 km upvalley. Table 22 presents a summary of the means
and maximum/minimum values across all sites.
In interpreting these data, it is important to consider the Way in
which the monitoring sites were selected for this 45-day program.
Although an important goal of this study is to characterize pollutant
concentrations in the neighborhoods closest to the plants, it was
undesirable to locate the monitoring sites directly in these
neighborhoods. Since the facility complexes are very large, a station
near a fenceline of a plant could easily fail to intercept plumes emitted
by sources located cross-wind of the monitor, especially since the
samples were collected during a relatively short period of time, i.e., 15
days over a 45-day period. By locating the stations an average of one
Drainage flow is a condition that can occur in the Kanawha Valley
during the night if the sky is clear and regional winds are light.
Under these conditions, the air in contact with the valley walls and
floor cools and sinks, flowing down the valley. Within this flow,
air near the ground surface is stable; pollutants emitted within it
are dispersed slowly.
Other pollutants are emitted in substantial quantities in the study
area, including 1,3-butadiene and acrylonitrile. These pollutants
were not included in the monitoring program, however.
4-60
-------
Table 21
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWNA VALLEY AIR QUALITY REPORT
DETAILED SUMMARY OF MEASURED DATA SET
(MARCH 4, 1986-APRIL 11, 1986)
DISPERSION CONDmONS
CONCENTRATIONS (PPB)
BEL
LE
INSTfTt
ITE
SAMPUNG
f3 METHYLEN
E CHLORIDE
14CHOROFORM
910 TOLUENE
"UNKNOWN COMPOUND" "
GENERAL OESCRPTION OF
0100-0500
1300-1700
01000500
1300-1700
DAY 1
| srrE
->¦ *
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
METEREOLOGICAL CONDmONS
O0 - Of
06 -
o0 - o+
0
1
©
0
3/4/86
0.50
0.24
0.S4
3.08
0.11
NF
0.06
0.55
0.72
0.86
0.95
1.10
9.30
REGIONAL FLOW GENERALLY
OUT Of 240° ¦ 270®
48.0, 8.9
56.0, 6.9
10.8, 4.8
33.4, 8.9
3/6/86
0.61
0.25
0.24
4.10
0.12
0.01
0.01
1.66
0.57
0.50
0.58
1.29
-
REGIONAL FLOW GENERALLY
250 270°
16.0, 3.6
37.6, 10.9
24.8, 5.5
11.9, 7.2
3/9/86
0.33
0.83
0.25
0.57
0.05
2.49
4.55
0.08
0.57
1.76
3.36
1.18
9.06
FLOW ACT OF 230 °- 260 °
DURING AFTERNOON,
DRAINAGE FLOW 2000-0900
8.9, 4.4
51.2, 7.1
14.5, 3.4
18.6, 7.0
3/12/86
0.88
1.78
25.04
0.40
0.60
3.61
4.89
0.06
1.77
5.45
4.01
0.97
6.90
13.0
REGIONAL FLOW GENERALLY
OUT OF 120°- 150®
61.9, 5.3
10.2, 3.6
18.2, 7.0
20.9, 5.5
3/16/86
0.72
0.29
0.60
2.B5
0.05
0.03
0.10
1.36
2.44
0.51
1.26
1.47
-
REGIONAL FLOW GENERALLY
340° - 030°
42.0, '
31.0, 4.5
19.9, 7.3
3/18/86
0.60
0.97
11.96
0.51
0.14
2.30
2.82
0.0S
2.71
5.11
2.03
0.64
5.90
9.76
4.90
REGIONAL FLOW GENERALLY
140®- 200° UNTL2200,THEN
240 330 °
9.8, 6.2
23.9, 8.9
10.3, 3.0
23.5, 4.0
3/21/86
0.59
0.27
0.60
1.25
0.46
0.10
0.18
1.34
1.97
0.80
1.34
2.03
3.80
4.70
FLOW OUT OF 270 0 - 330°
IN AFTERNOON WEAK DRAHAGE
FLOW 2100-0900, LOW WINDS
11.1, 3.7
39.6, 11.9
18.7, 4.1
40.4, 10.4
3/24/86
1.13
1.42
6.09
1.14
0.41
5.27
3.27
0.53
2.91
14.16
2.38
2.03
6.60
9.80
270°- 330° UNTIL2000,
DRAINAGE UNTL 1000
30.8, 6.1
37.6, 16.1
31.4, 5.6
29.3, 13.2
3/27/86
0.89
1.21
2.33
9.92
1.03
2.71
0.52
2.56
4.88
4.59
1.98
1.55
4.80
6.00
300». 3£0° UNTIL2100,
DRAINAGE UNTL 0000
26.8, 6.3
21.4, 8.9
34.5, 8.9
25.6, 8.9
3/30/86
1.72
1.46
8.74
3.31
2.89
2.14
3.74
0.96
7.38
7.98
4.70
3.26
11.1
11.4
GENERALLY 250 •- 270 0 UNTL
2000. DRAINAGE UNTIL 1000
7.1, 4.3
56.4, 15.5
6.5, 2.6
24.6, 11.3
4/2/86
1.02
0.60
4.82
3.52
0.B5
1.56
1.95
4.32
2.80
2.96
2.33
2.76
6.10
5.30
300°. 360° UNTL2400,
DRAINAGE UNTL 0000
54.0, 3.9
12.4, 9.5
49.0, 6.2
32.2, 13.6
4/4/86
1.43
2.60
4.61
3.21
0.76
5.19
1.77
1.29
3.32
9.48
2.45
2.50
-
270°- 300° UNTIL2000,
DRAINAGE UNTL0900
24.3, 2.3
41.3, 11.0
17.4, 4.8
21.5, 10.6
4/7/86
1.14
0.59
3.35
5.61
0.63
0.38
2.47
2.93
4.27
2.23
1.36
1.08
7.20
7.40
11.5
290°- 300° UNTL2000,
DRAINAGE UNTL 0600
12.4, 4.1
33.6, 12.4
47.7, 2.3
21.0, 9.9
4/9/86
1.29
0.20
0.24
2.95
0.98
0.01
0.02
0.75
3.96
0.39
0.38
0.54
-
-
-
_
REGIONAL FLOW 250 •- 300®
14.9, 6.7
23.5, 9.8
13.1, 8.5
27.0, 6.7
4/11/86
1.41
0.86
4.77
10.61
2.03
1.13
1.32
6.16
8.60
3.01
1.35
1.30
-
6.90
12.2
11.0
270°- 290° UNTH. DRANAGE
FLOW DEVELOPS AT 2200,
PERSISTENT UNTIL 0900
49.1, 14.4
36.5, 11.8
15.2, 9.3
15.3, 7.2
MEAN
0.95
0.90
5.99
3.54
0.74
1.80
1.84
1.64
3.25
3.99
2.03
1.63
4.7
6.3
7.4
6.9
* Data Unavailable
** Same retention time and peak shape as ethylene oxide,
however, a positive identification was not made.
Incomplete coverage is available for this compound.
PAGE 1 of 2
Source: Regulatory Integration Division, Office of
Policy Analysis, EPA, 1987
-------
Table 21 (Continued)
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
DISPERSION CONDITIOl
CONCtNIHAIlONS (PFti)
BELLE
INSTITUTE
SAMPLING
*8 CARBOtfElfTRACHLORIDE
GENERAL DESCRIPTION OF
0100 0500
1300-1700
0100 0500
1300-1700
| SITE 1
2
3
4
METEREOLOGICAL CONDITIONS
oB - o$
O0 -
0$
a8-
ae-
i
3/4/86
0.00
0.1 1
0.1 1
0.16
REGIONAL FLOW GENERALLY
OUT OF 240 o. 270 o
48.0, 8.0
56.0,
6.0
10.8,
4.8
33.4,
8.9
3/6/86
0.14
0.11
0.11
0.28
REGIONAL FLOW GENERALLY
250 «- 270°
16.0, 3.6
37.6,
10.9
24.8,
5.5
11.0,
7.2
3/0/86
1.00
0.13
0.88
0.13
FLOW ACT OF 230 0 • 260 6
DURING AFTERNOON.
DRAINAGE FLOW 2000*0900
8.0, 4.4
51.2,
7.1
14.5,
3.4
18.6,
7.0
3/12/86
0.15
0.17
1.70
0.00
REGIONAL FLOW GENERALLY
OUT OF 120 C. 1500
61.0, 5.3
10.2,
3.6
18.2,
7.0
20.0,
5.5
3/16/86
0.11
0.10
0.11
0.30
REGIONAL FLOW GENERALLY
340 c . 030 *
• , ¦
42.0,
•
31.0,
4.5
10.0,
7.3
3/18/86
0.13
0.15
1.05
0.11
REGIONAL FLOW GENERALLY
140 c . 200 0 UNTIL 2200, THEN
240 « - 330 0
0.8, 6.2
23.0,
8.0
10.3,
3.0
23.5,
4.0
3/21/86
0.12
0.13
0.11
0.22
FLOW OUT DF 270 0 - 330 c
IN AFTERNOON WEAK DRAINAGE
FLOW 2100-0000, LOW WINDS
11.1, 3.7
30.6,
11.0
18.7,
4.1
40.4,
10.4
3/24/86
0.16
0.21
0.45
0.17
270 e- 330c UNTIL 2000,
DRAINAGE UNTIL 1000
30.8, 6.1
37.6,
16.1
31.4,
5.6
20.3,
13.2
3/27/86
o!ia
0.17
0.26
0.53
300 c • 360° UNTIL 2100,
DRAINAGE UNTIL 0000
26.8, 6.3
21.4,
8.0
34.5,
8.0
25.6,
8.0
3/30/86
0.62
0.30
1.04
0.42
GENERALLY 250 e- 270 c UNTIL
2000, DRAINAGE UNTIL 1000
7.1, 4.3
56.4,
15.5
6.5,
2.6
24.6,
11.3
4/2/86
0.16
0.13
0.30
0.50
300 6 • 360 c UNTIL 2400,
DRAINAGE UNTIL 0900
54.0, 3.0
12.4,
0.5
40.0,
6.2
32.2,
13.6
4/4/86
0.18
0.20
0.43
0.23
270 e - 300C UNTIL 2000,
DRAINAGE UNTIL 0000
24.3, 2.3
41.3,
11.0
17.4,
4.8
21.5,
10.6
4/7/86
0.11
0.11
0.76
0.30
200 * • 300 e UNTIL 2000,
DRAINAGE UNTIL 0800
12.4, 4.1
33.6,
12.4
47.7,
2.3
21.0,
0.0
4/0/86
0.11
0.11
0.1 1
0.22
REGIONAL FLOW 250 0 • 300 0
14.0, 6.7
23.5,
0.8
13.1,
8.5
27.0,
6.7
4/1 1/86
0.15
0.1 3
0.54
1.44
270 C - 290C UNTIL DRAINAGE
FLOW DEVELOPS AT 2200,
PERSISTENT UNTIL 0900
40.1, 14.4
36.5,
11.8
15.2,
9.3
15.3,
7.2
MEAN
0.22
0.16
0.54
0.35
*Data Unavailable
**Same retention time and peak shape as ethylene oxide;
however, a positive identification was not made.
Coverage is incomplete for this compound.
Source: Regulatory Integration Office. Office of Polirv Anaivcic fpa iq«7
-------
TABLE 22
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
AVERAGE AND RANGE OF CONCENTRATIONS FOR TARGET
COMPOUNDS WITH ORD MONITORING PROGRAM
(March 4, 1986-April 11, 1986)
Average and range (PPB) of target compounds
mean minimum
maximum
dichloromethane 2.82 0.0 24.71
triehloromethane 1.58 0.0 6.11
1,2-dichloroethane 0.03 0.0 0.20
1,1,1-trichloroethane 0.34 0.2 0.72
benzene 0.89 0.4 1.64
tetrachlozomethane 0.30 0.1 1.70
trichloroethene 0.02 0.0 0.13
toluene 2.99 0.5 16.54
tetrachloroethene 0.07 0.0 0.28
chlorobenzene 0.04 0.0 0.32
o—xylene 0.34 0.0 0.84
Source: Report on the Air Monitoring in the Kanawha Valley, West Virginia,
Fitz-Simons, Lumpkin, McClenny.
4-63
-------
kilometer from the plants, the natural channeling of air flow in the
valley tended to mix the plumes and ensure a more reliable identification
of the full range of pollutants being emitted. Unfortunately, measured
data collected at these sites do not provide direct indications of
concentrations at the locations likely to experience maximum
concentrations. The one-kilometer distance was chosen as the best
practical tradeoff between (1) assessing plant-level emissions data via
comparisons of modeled and measured data, and (2) measuring
concentrations in neighborhoods most affected by these releases.
Quantitative Comparison of Measured and Modeled Data: One way of
comparing the general magnitude of measured and modeled concentrations is
to compare the limited measured data with six-month average modeled
values for the model receptors closest to the monitoring sites. Table 23
compares the six-month average modeled values with the limited measured
data. The closest receptors to the monitoring sites were averaged to
provide a point of comparison. Clearly, in order to definitively compare
the results, the specific sites and monitoring periods would need to be
modeled based on daily-specific emissions data. These rough comparisons
should not be interpreted as a model performance evaluation.
Modeled and measured values for chloroform, carbon tetrachloride, and
methylene chloride were found to be in reasonable agreement based on
these rough comparisons. These results are based on ISCLT. Only
industrial sources were modeled in these comparisons. If county-wide
concentrations and general background values were also included, the
modeled concentrations would likely increase by up to a few micrograms
per cubic meters for chloroform, methylene chloride, and carbon
tetrachloride.
The comparisons for the modeled ethylene oxide data and the unknown
measured compound (with the same retention time as ethylene oxide) showed
poor agreement. While the measured and modeled values for the "unknown
compound," which was hypothesized to be ethylene oxide, at Sites 1 and 2
were found to be in reasonably close agreement, still higher values were
found for Sites 3 and 4, where the inventory shows no emissions of this
compound. Based on this extremely limited and uncertain data set, it
appears that the unknown compound may not be ethylene oxide, or that
possibly there is an interfering compound in the Belle zone.
Alternatively, if the measured data reasonably represent the magnitude of
ethylene oxide concentrations, the emissions data in the Belle zone are
not including these releases.
The preceding comparisons provide a rough indication of how well the
modeling approach represents transport and dispersion within the study
area, and, for each pollutant, how reasonably the emissions data
represent actual concentrations via dispersion modeling. Dispersion
models and measured data are typically not used to assess emissions. For
4-64
-------
Table 23
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
COMPARISON OF MEASURED (3/86 - 4/86)
AND MODELED (12/85 - 6/86)
DATA BASED ON ORD MONITORING PROGRAM
Concentration^ (ug/m3)
Institute Belle
Site 1 Site 2 Site 3 Site 4
(Upvalley) (Downvalley) (Downvalley) (Upvalley)
Pollutant mod, mon. mod, mon. mod, mon. mod, mon.
Chloroform 1.9 3.6 4.5 8.7 8.1 8.9 11.5 8.0
Methylene Chloride 0.6 3.3 1.3 3.1 10.8 20.8 14.6 12.3
"Ethylene Oxide" 2.4 8.4 12.8 11.3 0 13.3 0 12.4
(unknown comp.)^
Carbon Tet. 0 1.4 0 1.0 2.6 3.4 3.3 2.2
Based on 1SCLT for the closest receptors: Site 1 (2-7, 2-8, 2-9);
Site 2 (8A.1 - 8A.4); Site 3 (4-2); and Site 4 (4-6, 4-7).
^ Ethylene oxide (unknown compound) averages are based on very limited
data: Site 1 (n=10); Site 2 (n=14); Site 3 (n=3); and Site 4 (n=4).
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987.
4-65
-------
criteria pollutants, it is often assumed that the emissions inventory is
accurate; comparisons of measured and modeled data are made to determine
the validity of the modeling analysis. For this study, where annual
average impacts at the residential areas near major facilities are a key
objective, model performance in terms of representing transport and
dispersion is expected to be within a factor of two on a seasonal or
annual basis. Substantially greater uncertainty exists in the emissions
data, especially for pollutants involving a large percentage of fugitive
releases. Evaluating emissions is therefore of prime importance for this
study.
The objectives of the ORD field program were generally achieved. The
following repeats the objectives and provides brief comments concerning
the achievement of each objective:
(1) To Qualitatively and quantitatively confirm the presence of
targeted compounds in the ambient air. In general, this
objective was met except for uncertainty regarding the
identification of ethylene oxide.
(2) To determine if a more in-depth emissions verification of
facilities and processes is needed. To date, only rough
comparisons have been made. Tentative results indicate, for
example, that modeled and measured results for chloroform,
carbon tetrachloride, and methylene chloride are reasonably
consistent for nearby receptors. However, measured data for an
unknown compound with same retention time as ethylene oxide did
not match the modeled values very well. If a more complete data
set were compiled, including process status data for sampling
days, a more quantitative comparison could be made.
(3) To assist in model selection for potential future phases of the
study. Quantitative analysis of measured data in comparison to
alternative model formulations has not yet been performed.
(4) To help determine if a more detailed monitoring program is
needed at a later date. The results suggest that monitoring
methods applicable for the ambient air are needed for ethylene
oxide, 1,3-butadiene, and propylene oxide. In addition, follow-
up long-term monitoring in the immediate vicinity of major
facilities would be helpful in more definitively estimating
ambient exposures.
(5) To evaluate a newly designed 24-hour cannister sampling system.
There appears to be a reasonable comparison between the results
of Tenax and the cannister sampling system.
4-66
-------
V. RISK ASSESSMENT
This section presents the results of the cancer risk assessment
screening methodology of the air quality analysis of the Kanawha Valley
Toxics Study. This methodology is based on the exposure modeling
analysis of a limited set of pollutants performed in support of this
study. This modeling analysis is documented in detail in the exposure
assessment (Section IV) of this report. For this risk assessment
analysis, we estimated upper-bound, lifetime, incremental carcinogenic
risks by combining estimates of individual and average population
exposures to a limited set of pollutants, based on our modeling analysis,
with estimates of the cancer potency of pollutants. We designed our
analysis to allow comparison of potential risks among this limited set of
pollutants and sources within this analysis. We can also compare these
potential risks from airborne pollutants to those risks posed by
pollutants within other media.
The estimates of upper-bound cancer risk are primarily based on three
factors: (1) modeled ambient air concentrations (annual averages);
(2) assumptions regarding the relation of actual exposure to the modeled
concentrations; and (3) cancer potency unit risk estimates of the
compounds studied. As noted in the exposure assessment, we estimated two
types of exposures for our risk assessment. We first evaluated the
exposure of selected neighborhoods immediately surrounding the major
facilities within the study area. We had pre-selected these
neighborhoods for analysis and believed them to be the most affected by
emissions from these facilities. Second, we performed dispersion
modeling to estimate the exposure to the average individual from selected
pollutants within valley zones.
Risk Assessment Methodology
This study uses a risk assessment screening methodology to evaluate
and compare, in a very limited fashion, the potential health risks from
exposure to a limited set of potentially toxic pollutants. The purpose
of this methodology is to permit a comparison of risks across pollutants,
sources, and pathways and to provide a general sense of the risk a
substance may present. This methodology does not allow one to make a
definitive statement concerning the absolute risk posed by a particular
pollutant, source, or exposure pathway.
This study is not designed to draw conclusions regarding the observed
cancer incidence in the community and environmental exposures. To draw
such conclusions, one would need to conduct an epidemological study and
control for factors such as smoking and socioeconomic class.
4-67
-------
Risk to an individual is defined as the increased probability that an
individual exposed to one or more pollutants will experience a particular
adverse health effect, in this case cancer, during his or her lifetime.
Several indices of carcinogenic risk are used in these analyses,
including increased estimated lifetime risk to an individual and
increased incidence of cancers in the exposed population, i.e., the upper-
bound projected increased cases across the entire population within the
study area.
This risk screening methodology involves both a qualitative and
quantitative assessment of the potential carcinogenicity of the selected
pollutants. As a screening study, this analysis employs both types of
assessments. EPA's Carcinogen Assessment Group (CAG) reviews the
evidence of carcinogenicity for selected pollutants and qualitatively
categorizes the chemical based on the weight of scientific evidence. It
also classifies pollutants as human carcinogens (Group A), probable human
carcinogens (Group B), possible human carcinogens (Group C), not
classifiable because of inadequate evidence (Group D), and not
carcinogenic to humans (Group E).
For those chemicals in groups A, B, and C, CAG provides quantitative,
conservative upper-bound estimates of carcinogenic unit risk factors. A
unit risk factor allows the calculation of the estimated increased
lifetime individual risk posed by exposure to a chemical, given standard
exposure assumptions. To calculate individual risk, a chemical's unit
risk factor is multiplied by the estimated concentration of the pollutant
to which an individual is exposed.
The concentration used in this calculation of risk is the predicted
annual average concentration of the chemical in the ambient air. For
this analysis, we used the predicted model concentration to calculate
individual risk. The unit risk factor that provides a measure of the
carcinogenic potency assumes that an average person weighs 70 kg and
breathes 20 m3 of air each day for 70 years.
4-68
-------
SAMPLE CALCULATION
POSSIBLE RISK OF CANCER ASSOCIATED WITH
INHALATION OF CHLOROFORM IN
AMBIENT AIR
Estimated Average
Individual Lifetime Risk
Cancer Unit Risk Factor
for the Inhalation of
Chloroform
Concentration
Estimated Average
Individual Lifetime Risk
Unit Risk Factor x Concentration
2.3 x 10"5 (ug/m3)"1
For this example assume 1 ug/m3
of chloroform in the ambient air
2.3 x 10"^ (ug/m3)"l x 1 (ug/m3)
2.3 x 10-5
This estimated average individual risk, based on an upper-bound unit
risk factor, indicates that the increased risk of cancer is not likely to
exceed a two in a one hundred thousand probability over a 70-year
lifetime of constant exposure to this concentration of chloroform. Given
the upper-bound nature of the unit risk factor, the actual risk could be
considerably lower. Table 24 presents the unit risk values for the
selected pollutants for the air analysis of the Kanawha Valley Toxics
Study.
Incidence, another index of risk, is the probable number of cancer
cases expected over a 70-year period when a population is exposed to the
average individual exposure concentration. Incidence is calculated by
multiplying risk to the average exposed individual by the number of
people exposed and can be annualized by dividing by 70 to obtain cases
per year.
Risk Assessment Limitations
There are several limitations to the risk assessment methodology. It
is important to emphasize these points to allow proper interpretation of
the estimate of risk presented in this report.
• The estimate of individual cancer risk and aggregate cancer
incidence from exposure to toxics should not be interpreted as
precise or absolute estimates of present or future health
effects. The simplifying assumptions and uncertainties in both
4-69
-------
TABLE 24
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
UPPER-BOUND CANCER UNIT RISK VALUES FOR
SELECTED TOXICS IN THE AMBIENT AIR
POLLUTANT
INHALATION
POTENCY VALUE
f ud/m3 r1
1,3-Butadiene 2.8 x 10
Carbon Tetrachloride 1.5 x 10
Chloroform 2.3 x 10
Ethylene oxide 1 x 10"
Perchlorethylene 4.8 x 10"
Acrylonitrile 6.8 x 10
Methylene chloride 4.1 x 10
Arsenic 4.3 x 10
Benzo(a)pyrene 3.3 x 10
Benzene 8 * 10"
Beryllium 2.4 x 10
Cadmium 1.8 x 10
Ethylene bromide
Ethylene chloride 2.6 x 10
Trichloroethylene 1.3 x 10
Allyl chloride 2.2 x 10
Vinylidene chloride 5.0 x 10
Vinyl chloride 2.6 x 10"
-4
-5
-5
-5
-6
-3
-3
-3
-3
2.17 X 10"4
-5
-6
-8
-5
GROUPING
BASED ON EPA CRITERIA
B 2
B 2
B 2
B 1
B 2
B 1
B 2
A
B 2
A
B 2
B 1
B 2
B 2
B 2
B 2
C
A
SOURCE: EPA Carcinogen Assessment Group, 1986. Potency values are
plausible upper-bound estimates of human health
effects.
4-70
-------
the toxicology and exposure components are simply too great to
justify a high level of confidence in the precision of the
results. The estimates of cancer risks are approximations of the
upper-bound potential risks to human health and are designed to
permit comparison across different sources, pollutants, and
exposure pathways.
• The cancer unit risk factors used in this study are
consistently conservative in the direction of overestimating
risk for the particular pollutants and exposure pathways we have
assessed.
• We may understate risks to the extent that we have not
considered all pollutants and sources and have not considered
workplace or consumer exposure. Synergistic or antagonistic
effects of pollutants are not considered, and there may be
carcinogens that have not received a potency value from CAG.
Also, the scope of the study does not include a consideration of
non-cancer health effects.
• The results from this risk assessment are based primarily on
existing knowledge. As new scientific data and techniques
become available, these results may change.
VI. RESULTS
In the following sections, we present the results of our risk
assessment. We first discuss the results of the individual risks at the
neighborhood sites identified within the vicinity of the facilities.
These results are based on modeled concentrations presented in Tables
5-8 and the unit risk factors shown in Table 24. For this analysis, we
identify those neighborhoods of relatively higher risks. For these
sites, we present the primary pollutants and facilities, in terms of risk
and percentage contribution to total risk, contributing to the estimated
risks within these neighborhoods. For the neighborhood with the highest
relative risk, we have determined the contribution of fugitive emissions
of the primary pollutant contributing to the total risk. Technical
Appendix J provides the risk and percentage contribution of fugitive
emissions for all the sites. The risk results presented for the
neighborhoods only reflect potential risk from point source facilities.
Risks from county-wide area sources are not included.
Second, we present the risk posed to the average exposed individual
from the selected point source pollutants. These results are based on
modeled concentrations presented in Tables 9 - 12. We then identify the
percentage contribution of pollutants and facilities to the risk to the
average exposed individual. We also calculate incidence by pollutant and
facility within the zone.
Finally, we present risks posed by county-wide area sources predicted
by the box model. These results are based on model concentrations
presented in Table 13.
4-71
-------
Belle Zone
We modeled air emissions from two facilities, Dupont and Occidental,
within the Belle zone.
These facilities emitted three of the studied pollutants: carbon
tetrachloride, methylene chloride and chloroform. Formaldehyde is also
emitted. The risk estimates for formaldehyde are presented in
Technical Appendix I. For these pollutants and these facilities, we
present the results for our risk screening analysis, first for the
exposures within selected neighborhoods and then for the exposure to the
average individual within the Belle zone.
Risks in Neighborhood Sites with Suspected Highest Exposures from
Point Sources: Table 25 presents the estimated upper-bound lifetime
incremental risk posed by each pollutant and the total estimated risk
from these pollutants at neighborhoods we selected as likely to have
relatively high exposures from point sources. The total risk is the sum
of the risk posed by each pollutant (we assume the cancer risks are
additive). Figure 9 maps the estimated risks for the neighborhoods
surrounding the facilities. For these sites, site 4-4 has the highest
estimated risk from these pollutants. The total estimated lifetime
incremental risk is 3 X 10"3 using upper-bound potency values. This is
a three in a thousand probability of contracting cancer during a 70-year
lifetime exposure from these concentrations of chemicals. There are
approximately 600 people living within this vicinity (1980 Census).
According to Table 25, chloroform contributes almost three-fourths of the
estimated risk. Carbon tetrachloride and methylene chloride contribute
about one-eighth and one-sixth of the total risk, respectively, at this
neighborhood.
The other surrounding neighborhood concentrations present upper-bound
lifetime risks between 9 x 10~5 to 5 x 10_<* individual risk from
these pollutants. This risk runs from a nine in a hundred thousand
probability to a six in a ten thousand probability of contracting cancer
during a 70-year lifetime exposure to these chemicals. At all sites,
chloroform is the most significant contributor to the total risk.
Table 26 presents the estimated lifetime individual risk by pollutant
contributed individually by the Dupont and Occidental facilities. The
total column for each facility is the sum of the risk posed by each
pollutant emitted by the facility. Occidental contributes the
overwhelming majority of the risk from these selected pollutants.
4-72
-------
TABLE 25
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME CANCER RISKS IN
NEIGHBORHOODS SURROUNDING FACILITIES FROM
POINT SOURCE POLLUTANTS
BELLE
COMPOUND
ICARBON T!CHLCSC.-OIMETHYLENI
1ETRACHLOlRM |E CHLORII
TOTAL
4- 1
(ISCLT)
1 3E-05 I 2E-04 1 4E-C5 |
2E-04
4- 2
fISCLT)
1 4E-05 I 2E-C4 1 4E-05 |
3E-04
4- 3
(ISCLT)
1 5E-05 I 2E-04 1 5E-05 I
3E-04
4- 4
(ISCLT)
1 4E-04 1.002297 1 5E-C4 I
003204
4- 5
(ISCLT)
I 6E-05 I 3E-04 1 SE-05 |
5E-04
4- 6
(ISCLT)
1 6E-05 I 3E-04 1 7E-05 I
4E-04
4- 7
(ISCLT)
1 4E-05 I 2E-04 1 SE-05 I
3E-04
4- 8
(ISCLT)
1 3E-05 I 2E-04 1 3E-05 I
2E-04
4- 9
(ISCLT)
1 1E-05 1 7E-05 1 2E-05 I
1E-04
4-10
(ISCLT)
1 2E-05 I I0E-05 1 2E-05 |
1E-04
4-11
(ISCLT)
1 2E-05 I 1E-04 1 3E-05 I
2E-04
4-12
(ISCLT)
1 1E-05 I 6E-05 1 1E-05 I
9E-05
4-13
(ISCLT)
1 2E-05 I 1E-04 1 3E-05 I
2E-04
4-14
(ISCLT)
! 3E-05 1 1E-04 1 3E-05 I
2E-04
4-15
(ISCLT)
1 2E-05 1 1E-04 1 3E-05 1
2E-04
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS HAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987 4.73
-------
FIGURE 9
Kanawha Valley Toxics Screening Study
Kanawha Valley Air Quality Report
Preliminary Risk Screening Results Intended for policy uevelopment
—-V V
Shaded areas represent neighborhoods where 7" \ • \ \
residents may be exposed to a limited set of pollutant
concentrations that could result in an upper-bound,
lifetime incremental cancer risk of 3 out of 1000
during a lifetime exposure(see text for details on
pollutants and exposure assumptions and limitations).
' -i
.iv- - • -
frit
Sitomo
\v
: I0
MONITORING SITE .
— / : > r e ; j_
_ CoM. ¦ .
- , •*. Lax* ~
\ X 4-2 v. N ' V'\
~ \ \ W4®. \
X V 4-3
\ v \*\ - r- w
*4^
QUALITY MONrTORING SITE
(MARCH - APRIL 1986)
I a-4
.a^Iev
¦f-
Chesapeaicff x
/ ^ / ' ! 4 • '7*
4-74
-------
TABLE 26
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME CANCER RISK CONTRIBUTED BY
FACILITIES IN NEIGHBORHOODS SURROUNDING FACILITIES
BELLE
DuPont
Occidental
COMPOUND
COMPOUND
ICAS3CN' TICHLCROFOIMETHYLEN1
1 ETRACHLOlRM |E CHI.ORl!
TOTAL
1CA33CN T1CMLC3CFO1METHYLEN1
iETRACHLOlRM |E CHLORII
TOTAL
4- 1
(ISCLT)
4E-CS 1
3S-06 1 5E-06 1
1E-05
4- 1
(ISCLT) I 3E-05 I 2E-04 I JE-05 I
2E-04
4- z
(ISCLT)
5E-06 I
3E-06 1 6E-06 1
1E-05
4- 2
(ISCLT) 1 3E-05 1 2E-04 ! 4E-05 1
3E-04
4- 3
(ISCLT)
4E-06 1
3E-06 1 8E-06 1
2E-05
4- 3
(ISCLT) I 4E-05 I 2E-04 | 5E-05 1
3E-04
4- 4
(ISCLT)
6E-06 I
5E-06 1 1E-06 1
1E-05
4- 4
(ISCLT) I 4E-04 1.002292 1 5E-04 I.
133192
4- 5
(ISCLT)
3E-06 1
2E-06 1 1E-06 1
6E-06
4- 5
(ISCLT) I 5E-05 I 3E-04 1 3E-05 I
4E-04
4- 6
(ISCLT)
3E-06 1
2E-06 I 8E-07 1
5E-06
4- 6
(ISCLT) I 5E-05 1 3E-04 I 7E-05 I
4E-04
4- 7
(ISCLT)
2E-06 1
2E-06 1 6E-07 1
4E-06
4- 7
(ISCLT) | 4E-05 I 2E-04 1 iiE-05 I
3E-04
4- a
(ISCLT)
2E-06 I
1E-06 1 5E-07 1
3E-06
4- a
(ISCLT) 1 3E-05 I 2E-04 | 3E-05 1
2E-04
4- 9
(ISCLT)
10E-CI7 I
7E-07 1 4E--07 1
2E-06
4- 9
(ISCLT) I 1E-05 1 7E-05 I 2E-05 I
1E-04
4-10
(ISCLT)
1E-06 I
8E-07 1 5E-07 1
2E-06
4-10
(ISCLT) I 2E-05 1 9E-05 1 2E--05 1
1E-04
4-11
(ISCLT)
1E-06 1
1E-06 1 6E-07 1
3E-06
4-11
(ISCLT) I 2E-05 I 1E-04 1 2E--05 I
1E-04
4-12
(ISCLT)
2E-06 1
1E-06 1 7E-07 1
3E-06
4-12
(ISCLT) I 1E-C5 1 6E-05 1 1E-05 1
9E-05
4-13
(ISCLT)
IE-OS 1
9E-07 1 UE-07 1
3E-06
4-13
(ISCLT) I 2E-!!5 1 1E-04 I 3E-05 1
2E-04
4-14
(ISCLT)
2E-06 1
2E-06 1 2E-06 1
6E-06
4-14
(ISCLT) 1 2E-05 1 1E-04 1 3E-05 1
2E-04
4-15
(ISCLT)
2H-06 1
2E-G6 I 2E-06 1
6E-06
4-15
(ISCLT) I 2E-05 I 1E-04 1 2E--05 I
2E-04
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-75
-------
We have isolated the risk posed by the fugitive emissions of
chloroform at site 4-4. Approximately five-sixths of the chloroform
emissions are fugitives at site 4-4. Technical Appendix J presents
fugitive contributions at the other neighborhood sites.
Risks to the Average Exposed Individual from Point Sources and in the
Zone: Table 27 presents the risks to the average exposed individual
within the Belle zone from these point source pollutants. According to
Table 27, the total lifetime individual risk is 2.5 X 10"4 from carbon
tetrachloride, chloroform, and methylene chloride. This is a two and one-
half out of ten thousand probability of contracting cancer during a 70-
year lifetime exposure to these concentrations of pollutants. Chloroform
contributes most of this risk. Table 28 cites risks (and percent of
total risk) by pollutant and facility. Chloroform contributes 71 percent
of the total risk. Carbon tetrachloride and methylene chloride jointly
contribute approximately 29 percent of the risk. According to Table 28,
Occidental contributes approximately 98 percent of the total risk from
these pollutants; Dupont contributes about two percent of the total risk.
Table 29 presents the contribution of fugitive and non-fugitive
sources to the total risk to the average exposed individual from point
sources. According to Table 29, fugitive sources contribute almost
85 percent of the risk posed by these pollutants to the average exposed
individual.
Tables 30 and 31 present annual incidence by point source pollutant
and by facility within the Belle zone. Belle has a population of 15,530
(Census 1980). We calculate a total annual incidence of .056 cases per
year within the Belle zone from point source pollutants. Chloroform is
the major pollutant contributing to this incidence; Occidental is the
major facility.
Risks to the Average Exposed Individual from Countv-Wide Area
Sources: Finally, in Table 32 we present lifetime, upper-bound risks to
the average exposed individual from area sources. We calculated these
risk estimates from the predicted concentrations from the box model. The
upper-bound risk to the average exposed individual is approximately 5.3 X
10"5, or five in a one hundred thousand probability of contracting
cancer during a 70-year lifetime exposure from these county-wide area
source pollutants. The incidence for county-wide area sources is .012
cases per year.
Charleston/South Charleston Zone
We modeled air emissions from two facilities, Union Carbide-South
Charleston and the Union Carbide Technical Center. These facilities
4-76
-------
TABLE 27
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
TOTAL UPPER-BOUND ESTIMATES OF LIFETIME CANCER
RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS POINT SOURCE POLLUTANTS
BELLE
UPPER-BOUND
POLLUTANT RISKS TO THE AVERAGE
(Weight of Evidence)^ EXPOSED INDIVIDUAL
Carbon Tetrachloride (B2) 3.5 x 10~®
Chloroform (B2) 1.8 x 10~4
Methylene Chloride (B2) 3.8 x 10"®
Total 2.5 x 10"4
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS .IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (B1 indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-77
-------
TABLE 28
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONTRIBUTION OF FACILITIES AND POINT SOURCE
POLLUTANTS TO TOTAL UPPER-BOUND ESTIMATES
OF RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
(BY RISK AND PERCENT OF TOTAL POINT SOURCE RISK)
BELLE
POLLUTANT
(Weight of Evidence)1
FACILITY
Dupont (%)
Occidental (%)
Total Risk (%)
Carbon Tetrachloride (B2)
Chloroform (B2)
Methylene Chloride (B2)
2.5x10® (1%)
1.9xl0"6 (1%)
1.0x10"® (0.4%)
3.3xl0"5 (12.9%)
1.8xl0"4 (70.4%)
3.7xl0"5 (14.4%)
3.6x10 (13.9%)
1.8xl0"4 (71.4%)
3.8xl0"5 (14.8%)
Total
5.4xl0"6 (2.4%)
2.5xl0"4 (97.7%)
2.5xl0"4 (100%)
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT
GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN
SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK
ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL CANCER
RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER: JN FACT. THEY COULD BE
ZERO. THE PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE
ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
1 Weight-of Evidence rating derived by CAG, based on EPA's classification system: A =
proven human carcinogen; B = probable human carcinogen (B1 indicates limited evidence from
human studies, B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not classifiable, and E =
no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
4-78
-------
TABLE 29
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONTRIBUTION OF FUGITIVE AND NON-FUGITIVE POINT
SOURCES TO TOTAL UPPER-BOUND ESTIMATES OF RISK
TO THE AVERAGE EXPOSED INDIVIDUAL BY RISK
AND PERCENT OF TOTAL POINT SOURCE RISK
BELLE
POLLUTANT FUGITIVE NON-FUGITIVE
(Weight of Evidence)1 SOURCES SOURCES
Carbon Tetrachloride (B2)
3.2xl0~5
(12.7%)
3.0xl0~®
(1.2%)
Chloroform (B2)
1.5xl0~4
(60.4%)
2. 7xl0"5
(10.8%)
Methylene Chloride (B2)
2.9xl0~5
(11.5%)
8.4x10®
(3.3)
Total
2.lxlO-4
(84.6%)
3.8xl0"5
(15.3%)
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS .IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
1 Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (B1 indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-79
-------
TABLE 30
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES
OF EXCESS ANNUAL CANCER INCIDENCE
ACROSS POINT SOURCE POLLUTANTS
BELLE
POLLUTANT UPPER-BOUND
(Weight of Evidenced ANNUAL CASES
Carbon Tetrachloride (B2) .008
Chloroform (B2) .040
Methylene Chloride (B2) .008
TOTAL .056
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS .IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (B1 indicates limited evidence from human studies,
82 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-80
-------
TABLE 31
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF EXCESS ANNUAL CANCER
INCIDENCE ACROSS FACILITIES
BELLE
UPPER-BOUND
FACILITY ANNUAL CASES
Dupont .001
Occidental .055
TOTAL .056
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-81
-------
TABLE 32
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME
CANCER RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS COUNTY-WIDE AREA SOURCES
BELLE
AREA SOURCE
UPPER BOUND
LIFE-TIME CANCER
RISK
Gas Marketing
Heating
Road Vehicles
Solvent Use
Waste Oil Burning
TOTAL
1.1 x 10
1.3 x 10
-6
-5
3.1 x 10
7.7 x 10
6.8 x 10
5.3 x 10
-6
-7
-5
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS. SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: JN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-82
-------
emitted four of the selected pollutants: acrylonitile, ethylene oxide,
methylene chloride and vinylidene chloride. Propylene oxide is also
emitted. Risk estimates for propylene oxide are presented in
Technical Appendix I. For these pollutants and facilities, we present
the results of our risk screening analysis, first for the exposure within
selected neighborhoods and then for the exposure to the average
individual within the Charleston/South Charleston zone.
Risks in Neighborhood Sites with Suspected Highest Exposures from
Point Sources: Table 33 presents the estimated upper-bound lifetime
incremental risk posed by each pollutant and the total estimated risk
from these point source pollutants at neighborhoods we selected as likely
to have relatively high exposure. The total risk is the sum of the risk
of each pollutant (we assume that cancer risks are additive). Figure 10
maps estimated risks for the neighborhoods surrounding the facilities.
According to Table 33, nine sites within South Charleston, south and
southwest of the Union Carbide-South Charleston facility, have predicted
ambient air concentrations of these select pollutants that should present
a total upper-bound risk estimate of 1 to 5 X 10~3 lifetime incremental
risk. This is a one out of a thousand probability to a five out of a
thousand probability of contracting cancer during a 70-year lifetime
exposure to these chemicals at these sites. These sites cluster in South
Charleston within a neighborhood that contains approximately 2,700 people
(1980 Census). The remaining neighborhood sites could be exposed to a
range of lifetime incremental risks of 2 X 10"^ to 8 X 10 .
For site 3-12 with the highest risk relative to other sites
(5.5 X 10~3), ethylene oxide contributes four-fifths of the total
risk. Acrylonitile contributes about one-sixth of the total upper-bound
incremental risk.
Table 34 presents the contribution of risk by the two facilities.
Within this neighborhood of site 3-12, Union Carbide-South Charleston
contributes nearly all of the risk.
For ethylene oxide and acrylonitile, we can estimate the contribution
of fugitive and non-fugitive sources to the risk posed by these chemicals
at site 3-12. Approximately one-fifth of the ethylene oxide
concentration is provided by fugitive sources. Fugitive emissions
contribute more than four-fifths of the risk posed by acrylonitile at
site 3-12. Technical Appendix J presents fugitive contributions to the
remaining sites.
4-83
-------
TABLE 33
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME CANCER RISKS IN
NEIGHBORHOODS SURROUNDING FACILITIES FROM
POINT SOURCE POLLUTANTS
CHARLESTON/SOUTH CHARLESTON |
COMPOUND
IACRYLONII ETHYLENE IMETHYLENIVINYLIDEI
ITRILE
1 OXIDE
IE CHLORIlNE CL (11
TOTAL
3- 1
(ISCLT)
1 4E-05 1 3E-04
1 2E-07 1 5E-07
3E-04
3- 2
(I3CLT)
I 2E-05 1 2E-04
1 7E-08 1 3E-07
2E-04
3- 3
(ISCLT)
1 3E-05
1 1E-04
1 7E-08 1 3E-07
2E-04
3- 4
(ISCLT)
1 6E-05
1 5E-04
1 8E-08 1 5E-07
5E-04
3- 5
(ISCLT)
! 8E-35
1 4E-04 I 1E-07 1 8E-07
5E-04
3- 6
(ISCLT)
1 6E-0S
1 2E-04 1 1E-07 I 5E-07
3E-04
3- 7
(ISCLT)
1 5E-05
1 2E-04
1 1E-07 1 4E-07
2E-04
3- 8
(ISCLT)
1 7E-04
1 3E-04
1 3E-07 I 9E-07
.001004
3- 9
(ISCLT)
1 3E-04
1 5E-04
1 3E-07 1 1E-06
8E-04
3-10
(ISCLT)
1 3E-04
1 10E-04
1 2E-07 I 2E-06
.001266
3-11
(ISCLT)
i 8E-04
10.00273
1 2S-C7 1 4E-06
0.0035
3-12
(ISCLT)
1 9E-04
1.004603
1 1E-07 1 1E-05
0.00553
3-13
(ISCLT)
1 4E-04
1.001618
1 3E-07 1 3E-06
.001982
3-14
(ISCLT)
1 3E-04
1.002709
1 2E-07 1 4E-06
.003042
3-15
(ISCLT)
1 9E-05
1.001413
1 1E-07 1 1E-06
.001506
3-16
(ISCLT)
1 1E-04
1.001125
I 2E-07 1 2E-06
0.00124
3-17
(ISCLT)
1 1E-04
1.001034
1 3E-07 1 2E-06
0.00116
3-18
(ISCLT)
1 1E-04
1 1E-04
1 2E-07 1 5E-07
3E-04
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-84
-------
FIGURE 10
Kanawha Valley Toxics Screening Study
Kanawha Valley Air Quality Report
Preliminary Risk Screening Results Intended for Policy Development
-sac"
"V
Shaded areas represent neighborhoods where ^
residents may be exposed to a limited set of pollutant j /*- - —^ ,—
concentrations that could result in an upper-bound, -y7""Sf
lifetime incremental cancer risk of 1 to 5 out of 1,000 / f ^-'P^-Tiannah:
during a lifetime exposure(see text for details on "r "C" \
pollutants and exposure assumptions and limitations). \\
•^.,r c 1A iv 4f
r.'
j . ¦ .'¦**•' i.. 7
- »...~¦• -
. - -*•
111 'i. s-»c • ^ —' W" _ , i; . ;| / > ; ••
• •: **J* "t/ y - •••".?
_ ' Ox ' ' 4 METEOROLOGICAL
C - „ if : /MONITORING SIT1
Woods >\
CARBIDE
3-16
lOSCiti
,VicfV „
aba
: ««l£
ooc «-cr
1 *!LONfCI
4-85
-------
TABLE 34
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME CANCER CONTRIBUTED BY
FACILITIES IN NEIGHBORHOODS SURROUNDING FACILITIES
CHARLESTON/SOUTH CHARLESTON
Union Carbide - South Charleston
Tech Center
3- 1 (ISCLT)
3- 2 (ISCLT)
3- 3 (ISCLT)
3- 4 (ISCLT)
3- 5 (ISCLT)
3- 6 (ISCLT)
3- 7 (ISCLT)
3- 8 (ISCLT)
3- 9 (ISCLT)
3-10 (ISCLT)
3-11 (ISCLT)
3-12 (ISCLT)
3-13 (ISCLT)
3-14 (ISCLT)
3-15 (ISCLT)
3-16 (ISCLT)
3-17 (ISCLT)
3-18 (ISCLT)
COMPOUND
ACRYLONII ETHYLENE IMETHYLEN!VINYLIDEI
TRILE I OXIDE IE CHLORIINE CL (11 TOTAL
4E-05 I 3E-04 I 3E-11 I 4E-07
2E-05 I 2E-04 I 3E-11 ! 2E-07
3E-05 I 1E-04 I 3E-11 I 2E-07
6E-05 I 5E-04 I 4E-11 I 4E-07
8E-05 I 4E-04 I 1E-10 I 7E-07
6E-05 I 2E-04 I 10E-11 I 4E-07
5E-05 I 2E-04 I 1E-10 I 3E-07
7E-04 I 3E-04 I 3E-10 I 6E-07
3E-04 I 5E-04 I :$E--10 I 9E-07
3E-04 ! 1CE-04 1 8E-10 I 2E-06
8E-04 10.00273 I 2E-10 I 3E-06
9E-04 I.0C4652 I 1E-10 I 1E-05
4Z-C4 1.051617 I 1E-10 I 3E-06
3E-04 1.002709 I 8E-11 I 4E-06
9E-05 I.001413 I 5E-11 i 1E-06
1E-04 1.001124 I 4E-11 i 1E-06
1E-04 1.001033 I 5E-11 I 1E-06
1E-04 I 1E-04 I 2E-10 I 3E-07
3E-04
2E-04
2E-04
5E-04
5E-04
3E-04
2E-04
.001002
7E-04
.001265
.003499
0.00553
.001981
.003041
.001506
.001239
.001158
3E-04
COMPOUND
IACRYLCNII ETHYLENE IMETHYLENIVINYLIDEI
ITRILE I OXIDE IE CHLORIINE CL (11
3- 1 (ISCLT) I 1E-07 I 4E-07 I 2E-07 I 1E-07 I
— • — + — — + — — — — — — + — — — — — — — — — — — — — +
3- 2 (ISCLT) I 5E-08 I 2E-07 I 7E-08 I 7E-08 I
3- 3 (ISCLT) I 5E-08 I 2E-07 I 7E-08 I 7E-08 I
3- 4 (ISCLT) I 6E-08 I 25-07 I 3E-08 I 8E-08 I
3- 5 (ISCLT) I 8E-08 I 2E-07 I 1E-07 I 10E-08 I
3- 6 (ISCLT) I 8E-08 I 3E-07 I IE-07 I 1E-07 I
3- 7 (ISCLT) I 9E-08 I 3E-07 I IE-07 j IE-07 I
+ + + + +
3- 8 (ISCLT) I 2E-07 I 8E-07 I 3E-07 I 3E-07 I
3- 9 (ISCLT) I 3E-07 I 9E-07 I 3E-07 I 4E-07 I
3-10 (ISCLT) I 2E-07 I 6E-07 I 2E--07 I 2E-07 I
3-11 (ISCLT) I IE-07 I 4E-07 I 2E-07
2E-07
3-12 (ISCLT) I IE-07 I 4E-07 I IE-07 I IE-07 I
3-13 (ISCLT) I 2E-07 I 7E-C7 I 3E-07 I 3E-07 I
+ - + '- + + +
3-14 (ISCLT) | 2E-07 I 5E-07 I 2E-07 i 2E-07 I
3-15 (ISCLT) I 9E-08 I 3E-07 I IE-07 I IE-07 I
3-16 (ISCLT) I IE-07 I 4E-07 I 2E-07 I 2E-07 I
3-17 (ISCLT) I 2E-07 I 7E-07 I 3E-07 I 3E-07 I
3-18 (ISCLT) I IE-07 I 5E-07 I 2E-07 I 2E-07 I
+ + + + +
TOTAL
8E-07
4E-07
-------
Risks to the Average Exposed Individual in the Zone from Point Source
Pollutants: We present risk to the average exposed individual for the
Charleston/South Charleston zone in Table 35. The total risk posed by
these point source pollutants, which assumes additivity of cancer risks,
is close to 3 X 10"^ upper-bound incremental risk during a 70-year
lifetime exposure from these pollutants.
Table 36 displays the contribution by pollutant and facility to the
risk to the average exposed individual. According to Table 36, Union
Carbide-South Charleston contributes nearly all the risk from these
pollutants.
For pollutants, ethylene oxide contributes three-fourths of the risk,
and acrylonitile contributes more than one-fifth of the risk.
Table 37 presents the contribution of fugitive and non-fugitive
sources to the total risk to the average exposed individual. According
to Table 37, fugitive sources contribute 36 percent of the total risk to
the average exposed individual.
Tables 38 and 39 present annual incidence by pollutant and facility
within the Charleston/South Charleston zone. Charleston and South
Charleston have a population of 51,750. We calculate a total incidence
of .211 cases per year within the Charleston/South Charleston Zone.
Ethylene oxide and acrylonitile from the Union Carbide-South Charleston
facility are the primary pollutants contributing to this incidence.
Risks to the Average Exposed Individual from Countv-Wide Area
Sources: Finally, in Table 40 we present lifetime upper-bound individual
risks from area sources. We calculated these risk estimates from the
predicted concentrations from the box model. Compared to the total risks
from the two point sources, the risks from area sources are lower than
the risks from point sources for the selected pollutants. The total risk
from these county-wide area sources is 8 X 10"5; the annual incidence
from county-wide area sources is .059.
Institute Zone
We modeled air emissions from one facility, Rhone Poulenc, within the
Institute zone. The facility emitted six of the selected pollutants:
acrylonitrile, benzene, chloroform, ethylene oxide, methylene chloride
and 1,3-butadiene. Formaldehyde and propylene oxide are also emitted.
These pollutants are discussed in Technical Appendix I. We present the
results of our preliminary risk screening analysis, first for the
exposure within selected neighborhoods and then for the exposure to the
average individual within the Institute Zone.
4-87
-------
TABLE 35
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
TOTAL UPPER-BOUND ESTIMATES OF LIFETIME CANCER
RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS POINT SOURCE POLLUTANTS
CHARLESTON/SOUTH CHARLESTON
UPPER-BOUND
POLLUTANT RISKS TO THE AVERAGE
(Weight of Evidence)1 EXPOSED INDIVIDUAL
Acrylonitrile (Bl) 6.3 x 10"^
Ethylene Oxide (Bl) 2.2 x 10"4
Methylene Chloride (B2) 2.0 x 10"®
Vinylidene Chloride (C) 1.4 x 10"®
Total 2.9 x 10"4
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (Bl indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-88
-------
TABLE 36
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONTRIBUTION OF FACILITIES AND POINT SOURCE
POLLUTANTS TO TOTAL UPPER-BOUND ESTIMATES
OF RISK TO THE AVERAGE EXPOSED INDIVIDUAL
BY RISK AND PERCENT OF TOTAL POINT SOURCE RISK
CHARLESTON/SOUTH CHARLESTON
POLLUTANT
(Weight of Evidence)
FACILITY
TOTAL RISK (%)
Union Carbide
Union Carbide
Tech-Center
Acrylonitrile (Bl)
Ethylene Oxide (Bl)
Methylene chloride (B2)
Vinylidene chloride (C)
Total
6.2xl0"5 (21.9%)
2.2xl0"4 (75.9%)
n-ll
9.7x10
(<- IX)
4.2x10'7 (.1%)
2.8xl0"4 (97.9%) 5.6x10"° (2%)
5.5xl0"7 (.2%) 6.3xl0~5 (22.1%)
2.1xl0"6 (.8%)
2.0xl0"6(.7%)
9.6x10"
1.3%)
2.2xl0"4 (76.7%)
(
-6
2.0x10® (0.7%)
1.4x10
;.4%)
2.9x10 (100%)
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED
AS AIDS JO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL CANCER RISKS JN
THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER: JN FACT, THEY
COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL
OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS
EXAMINED.
Weight-of Evidence rating derived by CAG, based on EPA's classification
system: A = proven human carcinogen; B = probable human carcinogen (Bl
indicates limited evidence from human studies, B2 indicates sufficient
evidence from animal studies but inadequate evidence from human studies);
C = possible human carcinogen; D = not classifiable, and E = no evidence
of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
4-89
-------
TABLE 37
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED
FOR POLICY DEVELOPMENT
CONTRIBUTION OF FUGITIVE AND NON-FUGITIVE POINT SOURCES
TO TOTAL UPPER-BOUND ESTIMATES OF RISK TO THE AVERAGE
EXPOSED INDIVIDUAL BY RISK AND PERCENT OF TOTAL POINT SOURCE RISK
CHARLESTON/SOUTH CHARLESTON
POLLUTANT
FUGITIVE
NON-FUGITIVE
(WEIGHT OF EVIDENCE)
SOURCES
SOURCES
Acrylonitrile (Bl)
3.6xl0"5
(12.6%)
2.7x10"® ( 9.5%)
Ethylene oxide (Bl)
6.8x10"®
(23.7%)
1.5xl0"4 (52.9%)
Methylene chloride (B2)
-
2.0xl0"6 ( 0.7%)
Vinylidene chloride (C)
9.9xl0~7
( 0.3%)
3.9xl0"7 ( 0.1%)
TOTAL
l.OxlO"4
(36.6%)
1.8xl0"4 (63.2%)
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Weight-of Evidence rating derived by CAG, based on EPA's classification
system: A = proven human carcinogen; B = probable human carcinogen (81
indicates limited evidence from human studies, B2 indicates sufficient evidence
from animal studies but inadequate evidence from human studies); C = possible
human carcinogen; D = not classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-90
-------
TABLE 38
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF
EXCESS ANNUAL CANCER INCIDENCE
ACROSS FACILITIES
CHARLESTON/SOUTH CHARLESTON
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES MERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS .IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES
IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES
FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
FACILITY
UPPER-BOUND
ANNUAL CASES
Union Carbide-South Charleston
Union Carbide-Technical Center
.207
.004
Total
.211
1987
4-91
-------
TABLE 39
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF
EXCESS ANNUAL CANCER INCIDENCE
ACROSS POLLUTANTS
CHARLESTON/SOUTH CHARLESTON
UPPER-BOUND
POLLUTANT (Weight of Evidence)* ANNUAL CASES
Acrylonitrile (Bl) .046
Ethylene Oxide (Bl) .163
Hethylene Chloride (B2) .001
Vinylidene Chloride (C) .001
Total .211
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (Bl indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-92
-------
TABLE 40
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME
CANCER RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS COUNTY-WIDE AREA SOURCES
CHARLESTON/SOUTH CHARLESTON
UPPER BOUND
LIFE-TIME CANCER
AREA SOURCE RISK
Gas Marketing 1.4x10"®
Solvent Use 1.3x10
Heating 1.3x10"^
Road Vehicles 5.1x10"^
Waste Oil Burning 1.2x10"®
Total 8.OxlO"5
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS. SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN, THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-93
-------
Risks in Neighborhood Sites with Suspected Highest Exposures from
Point Source Pollutants: Table 41 presents the estimated upper-bound
lifetime incremental risk posed by each pollutant and the total estimated
risk from these pollutants, which is the sum of the risks posed by each
pollutant. Figure 11 maps the estimated risks from point source
pollutants for the neighborhoods surrounding the facility.
According to Table 41, seven sites surrounding the Union Carbide
facility have predicted ambient air concentrations of these pollutants
that could pose an upper-bound risk estimate of 1 to 8 X 10~3 lifetime
incremental risk using upper-bound potency values. This is a one to
eight in a thousand probability of contracting cancer over a lifetime
exposure from these pollutants. For the most affected site, site 2-1,
there is an eight in a thousand chance of contracting cancer from a 70-
year exposure to these concentrations.
The total population within these most affected neighborhoods is
approximately 1,300 people (1980 Census). The modeled concentrations for
the remaining sites surrounding the facility pose a total risk on the
order of 10"' lifetime risks from these pollutants.
At site 2-1, the site most affected by these pollutants, 1,3-
butadiene contributes approximately three-fourths of the total risk at
this site. Ethylene oxide contributes close to one-fifth of the total
risk. Acrylonitrile, methylene chloride, benzene, and chloroform pose
the remaining risk.
Nearly all of the 1,3-butadiene risk is from fugitive emissions.
Risks to the Average Exposed Individual from Point Source
Pollutants: Table 42 presents the risks for the average exposed
individual within the Institute zone.
According to Table 42, the total lifetime individual incremental risk
is 1.1 X 10~3 from a 70-year exposure from the point source
pollutants. This is an upper-bound estimate of one in a thousand chance
of contracting cancer from these pollutants during a lifetime exposure to
these concentrations.
Ethylene oxide contributes over one-half of the total risk posed by
these pollutants. 1,3-Butadiene contributes close to one-third of the
total estimated risk. Acrylonitrile contributes approximately one-tenth
of the total risk and chloroform, methylene chloride, and benzene
contribute the remaining risk.
Table 43 presents the contribution of fugitive and non-fugitive
sources to the total risk to the average exposed individual. According
to Table 43, fugitive sources contribute 70 percent of the total risk to
the average exposed individual.
4-94
-------
TABLE 41
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME CANCER RISKS IN
NEIGHBORHOODS SURROUNDING FACILITIES FROM
POINT SOURCE POLLUTANTS
INSTITUTE
COMPOUND
IACRYLONII BENZENE ICH LOROFOI ETHYLENE IMETHYLEN11,3-BUTA|
ITRILE
1
IRM
1 OXIDE
IE CHLORII
DIENE
1 TOTAL
2- 1
(ISCLT)
1 4E-04
1
3E-07
1
2E-04
1.001449
1E-05 1
.005941
0.00802
2- 2
(ISCLT)
1 9E-05
1
4E-08
1
7E-05
1 6E-04
5E-06 i
8E-04
.001658
2- 3
(ISCLT)
1 6E-05
1
3E-03
1
1E-04
1 7E-04
1E-05 1
4E-04
.001356
2- 4
(ISCLT)
1 5E-05
1
3E-08
1
1E-04
1 6E-04
10E-06 I
3E-04
.001134
2- 5
(ISCLT)
1 4E-05
1
3E-08
1
1E-04
1 5E-04
6E-06 I
2E-04
9E-04
2- 6
(ISCLT)
1 3E-05
1
2E-08
1
7E-05
1 4E-04
5E-06 I
2E-04
7E-04
2- 7
(I3CLT)
1 3E-35
1
2E-08
1
6E-05
1 3E-04
4E-06 i
2E-04
6E-04
2- 8
(ISCLT)
! 3E-35
1
2E-08
1
5E-05
1 3E-04
2E-06 I
1E-C4
5E-04
2- 9
(ISCLT)
1 2E-05
1
2E-08
1
2E-05
1 2E-04
1E-06 1
10E-05
3E-04
2-10
(ISCLT)
I 2E-35
1
2E-08
1
2E-05
1 1E-04
1E-06 I
9E-05
3E-04
2-11
(ISCLT)
1 1E-05
1
1E-08
I
2E-05
1 10E-05
8E-07 1
7E-05
2E-04
2-12
(ISCLT)
1 3E-35
1
2E-08
1
5E-05
1 2E-04
3E-06 1
2E-04
5E-04
2-13
(ISCLT)
! 6E-05
1
3E-08
1
7E-05
1 5E-04
2E-06 I
4E-04
.001044
2-14
(ISCLT)
1 7E-C5
1
5E-08
1
6E-05
1.001296
2E-06 1
3E-04
.001733
2-15
(ISCLT)
1 2E-04
1
2E-07
1
1E-04
1.001898
6E-06 I
3E-04
0.00252
THE UNIT RISK FACTORS USED IN THIS »
ASSUMPTIONS THAT GENERALLY PRODUCE UPPE#-30UND ESflwrtS BECAUSE OF
LIMITATIONS IN DATA AND NETHOOS IN SEVERAL AREAS OF THE ^YSIS SUCH
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, 5155 E^IL—^
rai nil ATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACUJA^
iruFR. IN FACT THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
I^flMATErirroHELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-95
-------
FIGURE 11
Kanawha Valley Toxics Screening Study
Kanawha Valley Air Quality Report
Preliminary Risk Screening Results Intended for Policy Development
Shaded areas represent neighborhoods where ;
residents may be exposed to a limited set of pollutant -
concentrations that could result in an upper-bound, ~
lifetime incremental cancer risk of 1 to 8 out of 1,000
during a lifetime exposure(see text for details on
pollutants and exposure assumptions and limitations).
- v
-"C
AIR QUALITY MONfTORING SfTH ..
"\
(MARCH-APRIL1986}
1 km DOWW MLLEY
"... ^. «»-r
• v 4N.*
lri^Z
IT" AIR QUALITY MONITORING SITE
(MARCH-APRIL 1986)
mmmp.
UNION CARBIDE *
Cntwi
3IU «»
Mc&
•JrJl«s -
STATE COLLEGE
**» BUILDING P- \
' ^X-
2-9 A/VV -•
V'« N
REHABILITATION
BUILDING
•• . "A- r-"* ^
^KT* •
MONfTORING SITE
U NV«-0 >¦ -
¦ g„;
• •
-«eh ."*<
C /V'' V\ IJJNBaR
v~ - ¦^ ^
. •_ t.-\ \ v
%.X>' \ N
1000
.ax
200C
300C
7000 nv
i *'lo*ctc«
4-96
-------
TABLE 42
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
TOTAL UPPER-BOUND ESTIMATES OF LIFETIME CANCER
RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS POINT SOURCE POLLUTANTS
INSTITUTE
UPPER-BOUND
POLLUTANT RISKS TO THE AVERAGE
(Weight of Evidence)1 EXPOSED INDIVIDUAL
Acrylonitri le (Bl) 1.1 x 10"^
Benzene (A) 1.4 x 10"^
Chloroform (B2) 5.5 x 10"^
Ethylene Oxide (Bl) 5.8 x 10 ^
Methylene Chloride (B2) 2.7 x 10"®
1,3-Butadiene (B2) 3.6 x 10"^
Total 1.1 x 10~3
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS .IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
1 Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (Bl indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-97
-------
TABLE 43
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONTRIBUTION OF FUGITIVE AND NON-FUGITIVE POINT
SOURCES TO TOTAL UPPER-BOUND ESTIMATES OF
RISK TO THE AVERAGE EXPOSED INDIVIDUAL
BY RISK AND PERCENT OF TOTAL POINT SOURCE RISK
INSTITUTE
POLLUTANT
(Weiaht of Evidence)*
FUGITIVE
SOURCES (%)
NON-FUGITIVE
SOURCES (%)
Acrylonitrile (Bl)
2.9 x 10~5
(2.6%)
8.0 x 10"5
(7.3%)
Benzene (A)
-
1.4 x 10"7
(<-l%)
Chloroform (B2)
2.3 x 10~5
(2.1%)
3.2 x 10~5
(2.9%)
Ethylene oxide (Bl)
3.7 x 10~4
(33.3%)
2.1 x 10"4
(19.2%)
Methylene chloride (B2)
1.4 x 10"6
(.1%)
1.3 x 10"6
(.1%)
1,3-Butadiene (B2)
3.6 x 10"4
(32.1%)
2.5 x 10"6
(.2%)
Total
7.8 x 10"4
(70.2%)
3.3 x 10"4
(29.7%)
THE UNIT RISK FACTORS USEO IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS. TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (B1 indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-98
-------
Table 44 cites annual incidence by point source pollutant within the
Institute zone. Institute has a total population of 22,390 (1980 Census).
We calculated a total annual incidence of .354 cases per year, using upper-
bound potency values. Ethylene oxide is the most significant of the six
pollutants, contributing about one-half of the risk to the total population.
Risks to the Average Exposed Individual from County-Wide Area Sources:
Finally, we present life-time incremental individual risks from county-wide
area sources. We calculate these estimates from the predicted
concentrations for the box model. We present average individual risks from
county-wide area sources in Table 45. Compared to the total risks from
Union Carbide (1.1 X 10~3), area sources contribute an individual life-
time incremental risk of 5 X 10 , approximately two orders of magnitude
lower than the risk from the Rhone Poulenc facility. The incidence from
county-wide area sources is .015.
Nitro Zone
We modeled air emissions from two facilities, Artel and Monsanto,
within the Nitro zone. These facilities together emitted nine of the
selected pollutants: acrylonitrile, benzene, carbon tetrachloride,
chloroform, methylene chloride, vinyl chloride, ethylene chloride, allyl
chloride, and trichloroethylene. Formaldehyde is also emitted in Nitro.
Risk estimates for formaldehyde are presented in Technical Appendix I. For
these pollutants and facilities, we present the results of our risk
screening analysis, first for the exposures within selected neighborhoods
and then for the exposure to the average individual within the Nitro zone.
Risks in Neighborhood Sites with Suspected Highest Exposures From Point
Source Pollutants: Table 46 presents the estimated upper-bound incremental
risk posed by each point source pollutant. The total risk is the sum of
the risk of each pollutant (we assume that cancer risks are additive).
Figure 12 maps the estimated risks for the neighborhoods surrounding the
facilities.
According to Table 46 and Figure 12, ten sites have a total risk
ranging from 1 to 6 X 10"° individual lifetime incremental risk from
these pollutants. This is a one to six in a million probability of
contracting cancer during a 70-year exposure to these concentrations of
pollutants. Many of these neighborhoods are located on the east bank of
the Kanawha River between the two facilities and the valley walls.
Monsanto contributes most of the total risk at these sites, primarily from
the emission of trichoroethylene. Table 47 breaks out estimated risk
contributed by facilities.
Risks to the Average Exposed Individual from Point Source Pollutants:
We present the total risk to the average exposed individual within the
4-99
-------
TABLE 44
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF EXCESS
ANNUAL CANCER INCIDENCE
ACROSS POINT SOURCE POLLUTANTS
INSTITUTE
UPPER-BOUND
POLLUTANT (Weight of Evidence!1 ANNUAL CASES
Acrylonitrile (Bl) .035
Benzene (A) <.0001
Chloroform (B2) .018
Ethylene Oxide (Bl) .185
Methylene Chloride (B2) .001
1,3-Butadiene (B2) .115
Total .354
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
1 Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (Bl indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies): C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-100
-------
TABLE 45
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME
CANCER RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS AREA SOURCES
INSTITUTE
UPPER BOUND
LIFE-TIME CANCER
AREA SOURCE RISK
Gas Marketing 7.9x10"^
Solvent Use 6.3x10"®
Heating 1.3x10"®
Road Vehicles 2.6x10 ®
Waste Oil Burning 6.5x10"^
Total 4.7x10®
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES MERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES
IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES
FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-101
-------
TABLE 46
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME CANCER RISKS IN
NEIGHBORHOODS SURROUNDING FACILITIES FROM
POINT SOURCE POLLUTANTS
NITRO
COMPOUND
ACRYLONII BENZENE
TRILE I
I CARBON TICHLOROFOIMETHYLEN!VINYL CH
IETRACHLOlRM |E CHLORIILORIDE
!ETHYLENE IALLYL CHlTRICHLORl
I CHLORIDILORIDE lOETHYLENl
TOTAL
1E-06
IE-06
2E-06
6E-06
6E-06
4E-06
3E-06
3E-06
2E-06
1E-06
9E-07
7E-07
7E-07
8E-07
1- 1 t I5CLT )
1- 2 tI3CLT)
IE-OS I
10E-09 I
3E-10
2E-10
5E-10 I
4E-10 !
8E-10 I
6E-10 I
7E-11 I
5E-11 I
4E-11
3E-11
3E-08 I
2E-08 I
5E-10 I
4E-10 I
1E-06 I
+
1E-06 I
———+
2E-06 I
+
6E-06 1
—+
6E-06 I
4E-06 I
—+
2E-06 I
— -+
2E-06 !
-+
1E-06 I
— +
95-07 I
+
7E-07 1
—+
6E-07 I
—¦———¦+
6E-07 i
+
7E-07 I
—+
1- 3 (ISCLT)
1- 4 tISCLT)
2E-Q8 I
5E-08 I
4E-10
10E-10
8E-10 I
2E-09 I
1E-09 I
3E-09 I
1E-10 I
3E-10 I
7E-11
2E-10
5E-08 I
1E-07 I
1E-09 I
3E-09 I
1--5 (ISCLT)
1- 6 (ISCLT)
7E-08 I
lc-07 I
1E-09
2E-09
3E-09 I
4E-09 I
4E-09 I
6E-09 I
4E-10 I
6E-10 I
2E-10
4E-10
2E-07 I
3E-07 I
4E-09 I
7E-09 i
1- 7 (ISCLT)
1- 8 (ISCLT)
3E-07 I
4E-07 I
6E-09
9E-09
1E-08 I
2E-C8 I
2E-08 I
2E-08 I
1E-09 i
2E-09 |
9E-10
1E-09
6E-07 I
9E-07 I
1E-08 I
2E-08 I
1- 9 (ISCLT)
1-10 (ISCLT)
2E-07 I
8E-38 I
4E-09
1E-09
7E-09 1
3E-09 I
1E-08 I
4E-09 I
9E-10 I
4E-10 I
6E-10
2E-10
SE-07 I
2E-07 I
2E-08 I
6E-09 |
1-11 (ISCLT)
1-12 (ISCLT)
4E-03 I
3E-38 I
8E-10
5E-10
1E-09
10E-10
2E-09 I
1E-09 I
2E-10 I
1E-10 I
1E-10
8E-11
1E-07 I
7E-08 I
3E-09 I
2E-09 I
1-13 (ISCLT)
1-14 (ISCLT)
3E-08 I
3E-08 I
5E-10
6E-10
10E-10 I
1E-09 I
2E-09 I
2E-09 I
1E-10 I
2E-10 I
9E-11
1E-10
7E-08 I
9E-08 I
2E-09 I
3E-09 I
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-102
-------
Figure 12
Kanawha Valley Toxics Screening Study
Kanawha Valley Air Quality Report
Preliminary Risk Screening Results Intended for Policy Development
Shaded areas represent neighborhoods where
residents may be exposed to a limited set of pollutant
concentrations that could result in an upper-bound,
lifetime incremental cancer risk of 1 to 6 out of
1,000,000 during a lifetime exposure(see text for details / <
on pollutants and exposure assumptions and limitations). ' 0
METEOROLOGICAL
/MONITORING SITE
MONSANTO 'r y*
1-6 ' ,•/ T~\ L
i ¦ "*' — TT. * '&*»—
ALLIED \<
COASTAL
. » PIKE .i 5
i;;. m
/ 1% .*/*
- tBk- '/
1- 13 •
tRT.
*4ie
15,1-12- ^=-"7^"
:0OC rooo 30OC 'OOP »OC 6000
700C rtr
: hl0»cte«
4-103
-------
TABLE 47
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUNCI ESTIMATES OF LIFETIME CANCER RISKS CONTRIBUTED BY
FACILITIES IN NEIGHBORHOODS SURROUNDING FACILITIES
NITRO
(Page 1 of 2)
Artel
COMPOUND
ACRYLONI i EENZENE ! CARBON T l CHLORDFOtMETHYLENl VIWVL t5H!|'ETRYL£HE,U.iibY,L CH I
TRILE 1
lETi?ACKLOlSM
IE CHLORIILORICE
1 CHLORID1 LORIDE
1
TOTAL
1- 1
(ISCLT)
1E-08 I
3E-10
1 5E-10 I
8E-10
1 7E-11 I
4E-11
1 3E-08
5E-10
1
4E-08
1- 2
(ISCLT)
10E-09 I
2E-10
1 4E-10 1
6E-10
1 5E-11 1
3E-11
1 2E-08
4E-1D
1
3E-08
1- 3
(ISCLT)
2E-G8 I
4E-10
1 3E-10 I
1E-09
1 1E-10 I
7E-11
! 5E-08
1E-09
1
7E-08
1- 4
(ISCLT)
5E-08 1
10E-10
1 2E-09 I
3E-09
1 3E-10 I
2E-10
1 1E-07
3E-09
1
2E-07
1- 5
(ISCLT)
7E-0S 1
1E-09
1 3E-09 1
4E-09
1 4E-10 1
2E-10
1 2E-07
4E-09
1
2E-07
1- 6
(ISCLT)
1E-07 1
2E-09
1 4E-09 1
6E-09
1 6E-10 I
4S-10
1 3E-07
7E-09
1
4E-07
1- 7
(ISCLT)
3E-C7 1
6E-09
1 1E-08 1
2E-08
1 1E-09 1
9E-10
1 6E-07
1E-08
1
9E-07
1- 8
(ISCLT1
4E-C17 1
9E-09
1 2E-08 1
2E-08
1 2E-09 1
1E-09
1 9E-07
2E-08
1
1E-06
1- 9
(ISCLT)
2E-07 1
4E-09
1 7E-09 I
1E-08
1 9E-10 I
6E-10
1 5E-07
2E-08
1
7E-07
1-10
(ISCLT)
£E-0fi 1
1E-09
I 3F 09 1
4E-09
1 4E-10 I
2E-10
1 2F-07
6E-09
1
3E-07
1-11
(ISCLT)
4E-08 I
8E-10
1 1E-09 1
2E-0V
1 2E-10 i
1E-10
i 1E-07
3E-09
1
1E-07
1-12
(ISCLT)
3E-08 I
5E-10
] 10E-10 I
1E-09
1 1E-10 1
8E-11
1 7E-08
2E-09
1
1E-07
1-13
(ISCLT)
3E-08 I
55-10
1 10E-10 I
2E-09
1 1E-10 1
9E-11
1 7E-08
2E-09
1
1E-07
1-14
(ISCLT)
3E-08 I
6E-10
1 1E-09 I
2E-09
1 2E-10 I
1E-10
1 9E-08
3E-09
1
1E-07
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES HERE
CALCULATED AS AIDS TO P'LICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KAWWA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-104
-------
TABLE 47
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS INTENDED FOR POLICY
DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME CANCER RISKS CONTRIBUTED
FACILITIES IN NEIGHBORHOODS SURROUNDING FACILITIES
NITRO
(Page 2 of 2)
Monsanto
COMPOUND
ITRICHLORI
lOETHYLENl
TOTAL
1- 1
tISCLT)
1 1E-06 1
1E-06
1- 2
(ISCLT)
1 1E-06 1
1E-06
1- 3
tISCLT)
1 2E-R6 I
2E-06
1- 4
(ISCLT)
1 6E-06 1
6E-06
1- 5
tISCLT)
1 6E-06 1
6E-06
1- 6
(ISCLT)
1 4E-06 1
4E-06
1- 7
(ISCLT)
1 2E-06 1
2E-06
1- 8
(ISCLT)
I 2E-D6 1
23-06
1- 9
(ISCLT)
1 1E-06 I
1 c-06
1-10
(ISCLT)
1 9E-07 1
9E-07
1-11
(ISCLT)
1 7E-07 i
7E-07
1-12
(ISCLT)
1 6E-07 I
6E-07
1-13
(ISCLT)
1 6E-07 I
6E-07
1-14
(ISCLT)
1 7E-07 I
7E-07
—__
+ +
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PROOUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHOOS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis,
EPA, 1987
4-105
-------
Nitro zone in Table 48. The total risk presented by these pollutants is
2.5 x 10"°. This is more than a two in a million probability of
contracting cancer during a 70-year lifetime exposure from these
concentrations of point source pollutants. According to Table 49,
trichloroethylene from Monsanto contributes more than nine-tenths of this
total risk.
Tables 50 and 51 present the annual incidence by pollutant and facility
within the Nitro zone for point sources. Nitro has a population of 9,990.
We calculated a total annual incidence of .0003 cases per year from these
point source pollutants using upper-bound potency values.
Risks to the Average Exposed Individual from Countv-wide Area Sources:
Finally, we present lifetime, upper-bound incremental risks from county-
wide area sources in Table 52. We calculated these risk estimates from the
predicted concentrations from the box model. For area sources, we observe
a total risk of 1.4 X 10 , which is larger than the estimated risk from
the selected point source pollutants. Incidence from county-wide area
sources is approximately .002 cases per year.
VII. AIR ANALYSIS SUMMARY AND CONCLUSIONS
The Kanawha Valley Air Quality Report presents a conservative
assessment of the potential cancer health effects that a select set of
pollutants may pose to a limited population within the Kanawha Valley.
Based on an emission inventory for volatile organic compounds developed by
the West Virginia Air Pollution Control Commission and EPA generated
estimates of county-wide area source pollutants, our analysis develops
exposure concentrations for various populated areas within the study area.
These exposure estimates are the basis for our risk assessment analysis of
these chemicals. Using upper-bound cancer unit risk values with exposure
estimates, we developed upper-bound lifetime, incremental individual risk
estimates which assume a 70-year exposure to these pollutants.
We emphasize that our results are not predictors of absolute risk. We
have reviewed only a small set of pollutants of our risk assessments; a
complete review of potential health effects from all pollutants has not
been completed.
A summary of the air quality report and findings follows:
• The air analysis for the Kanawha Valley Toxics Screening study
was limited to four geographic areas or zones within the Kanawha
Valley: Belle, Charleston/South Charleston, Institute, and
Nitro. The valley terrain of the Kanawha Valley precluded
modeling exposures throughout the 60-mile river stretch. We
selected these four zones based on the geographic orientation of
the valley within
4-106
-------
TABLE 48
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
TOTAL UPPER-BOUND ESTIMATES OF LIFETIME CANCER
RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS POINT SOURCE POLLUTANTS
NITRO
UPPER-BOUND
POLLUTANT RISKS TO THE AVERAGE
(Weight of Evidence)* EXPOSED INDIVIDUAL
Acrylonitrile (Bl)
5.7
X
10-8
Benzene (A)
1.2
X
10~9
Carbon Tetrachloride (B2)
2.2
X
10"9
Chloroform (B2)
3.4
X
10"9
Methylene Chloride (B2)
3.0
X
10~10
Vinyl Chloride (A)
1.9
X
10"10
Ethylene Chloride (B2)
1.4
X
10~7
Allyl Chloride (B2)
3.3
X
10"9
Trichloroethylene (B2)
2.3
X
10~6
Total
2.5
X
10"6
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (B1 indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-107
-------
TABLE 49
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
CONTRIBUTION OF FACILITIES AND POINT SOURCE
POLLUTANTS TO TOTAL UPPER-BOUND ESTIMATES OF
RISK TO THE AVERAGE EXPOSED INDIVIDUAL BY
RISK AND PERCENT OF TOTAL POINT SOURCE RISK
NITRO
POLLUTANT
(Weight of Evidence)^ FACILITY
Fike
Acrylonitrile (Bl)
5
.7x10'
•8
(2.3%)
Benzene (A)
1.
.2x10"
-9
(<.1%)
Carbon Tetrachloride(B2)
2
.2x10"
¦9
(.IX)
Chloroform (B2)
3;
o
X
¦9
(.1%)
Methylene Chloride (B2)
3,
o
X
o
-10
(<.1%)
Vinyl Chloride (A)
1.
.9x10"
¦10
(< - IX)
Ethylene Chloride (B2)
1.
.4x10"
¦7
(5.4%)
Allyl Chloride (B2)
3
.3x10"
-9
(.1%)
Trichloroethylene (B2)
-
Total
2
.1x10'
¦7
(8.0%)
TOTAL RISK
Monsanto
5.7xl0"8 (2.3%)
1.2xl0~9 (<.1%)
2.2xl0~9 (.1%)
3.4x10"9 (.1%)
3.0xl0~10 (<.1%)
1.9xl0"10 (<.1%)
1.4xl0~7 (5.4%)
3.3xl0"9 (.1%)
2.3xl0"6 (91.9%) 2.3xl0~6 (91.9%)
2.3xl0"6 (91.9%) 2.5xl0~6 (100%)
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS
THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND
METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULATIONS AND
POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOPMENT.
NOT AS PREDICTIONS OF ACTUAL CANCER RISKS JN, THE KANAWHA VALLEY. ACTUAL RISKS MAY
BE SIGNIFICANTLY LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES
FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's classification
system: A = proven human carcinogen; B = probable human carcinogen (B1
indicates limited evidence from human studies, B2 indicates sufficient
evidence from animal studies but inadequate evidence from human studies); C =
possible human carcinogen; D = not classifiable, and E = no evidence of
carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
4-108
-------
TABLE 50
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF EXCESS
ANNUAL CANCER INCIDENCE ACROSS
POINT SOURCE POLLUTANTS
NITRO
UPPER-BOUND
POLLUTANT (Weight of Evidence)* ANNUAL CASES
Acrylonitrile (Bl) <.0001
Benzene (A) <.0001
Carbon Tetrachloride (B2) <.0001
Chloroform (B2) <.0001
Methylene Chloride (B2) <.0001
Vinyl Chloride (A) <.0001
Ethylene Chloride (B2) <.0001
Allyl Chloride (B2) <.0001
Trichloroethylene (B2) .0003
Total .0003
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (Bl indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-109
-------
TABLE 51
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF EXCESS
ANNUAL CANCER INCIDENCE ACROSS
FACILITIES
NITRO
UPPER-BOUND
Facility ANNUAL CASES
Artel
<.0001
Monsanto
.0003
Total
.0003
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT ANO EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-110
-------
TABLE 52
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF LIFETIME
CANCER RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
ACROSS COUNTY-WIDE AREA SOURCES
NITRO
UPPER BOUNO
LIFE-TIME CANCER
AREA SOURCE RISK
Gas Marketing 3.4x10"^
Solvent Use 1.8x10"®
Heating 4.2x10"®
Road Vehicles 9.8x10"®
Waste Oil Burning 2.4x10"^
Total 1.6xl0"5
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES MERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: JN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-111
-------
the zone, the major facilities located within each zone, and the
populations surrounding the facilities. The total population for
all four zones is approximately 99,700.
The list of pollutants considered for our analysis was provided
by the West Virginia Air Pollution Control Commission's 1984
emission inventory and by EPA estimates of county-wide area
sources. The 1984 emission inventory focused primarily on
volatile organic compounds. The inventory surveyed all
facilities within the four zones with the exception of small
processing operations and blending facilities. The county-wide
area source pollutants included both organic and inorganic
pollutants. Our final selection of sources included seven point
source facilities and five county-wide area source categories,
and was based on available cancer health effects information for
pollutants emitted by these sources.
For our cancer health assessment of a select set of pollutants,
we modeled pollutants that have a unit risk factor developed by
EPA's Carcinogen Assessment Group. Through this selection
process, 12 point source pollutants and an additional six county-
wide area source pollutants were modeled. In addition to these
18 pollutants, we modeled propylene oxide and formaldehyde, two
additional pollutants identified from the 1984 Emission Inventory
by the West Virginia Department of Public Health for exposure
analysis.
The pollutants we have modeled do not necessarily represent the
chemicals of most concern. The CAG has developed unit risk
values for only a limited set of pollutants. Consequently, we
have selected chemicals of known or suspected carcinogenicity for
which we have adequate information to develop risk estimates. We
do not know if we have modeled the pollutants posing the most
risks.
For those pollutants for which we did develop a preliminary
conservative assessment of cancer effects, we examined two types
of exposures: exposures in preselected neighborhoods surrounding
the facilities and exposures to the average individual with each
zone. The Model ISCLT provided predicted concentrations for
these two exposures from point sources for the risk assessment.
We also used LONGZ to incorporate turbulent intensity data
collected in two of the study zones. LONGZ results are presented
in Technical Appendix A. Results from our box-model analysis
provided concentration estimates for our county-wide area sources.
We were required to make several assumptions within our modeling
analysis to allow us to assess exposures.
4-112
-------
- We assumed that all residents in the study area live on the
valley floor
For the neighborhood exposures, all modeled receptors were on the
valley floor and, therefore, this assumption did not affect this
analysis. For exposures to the average individual, this
assumption is not always true, since people do reside on the
valley walls. Since the predominant releases are low level, we
do not expect people on the valley walls to be exposed to higher
concentrations than those on the valley floor.
We did perform preliminary hillside receptor modeling in the
Belle zone with the model LONGZ. We found that predicted
concentrations were lower on the hillside than on the floor of
the valley. Consequently, this assumption may lead to an over
prediction of risk to the average individual.
- We assumed that there is no valley wall reflection
For neighborhood exposures, it is unlikely that plume reflection
off the valley walls will significantly affect concentrations at
receptors surrounding the facilities. These neighborhoods are
more affected by plumes emitted directly from the facilities,
although plume reflection may affect neighborhoods further from
the facilities and closer to the valley walls. For exposures to
the average individual, exposure concentrations could be higher.
- For our Gaussian modeling, we assume that there is no interzone
transport of pollutants
• It is unlikely that interzone transport of pollutants
significantly affects concentrations within the neighborhoods
surrounding the facilities. Again, these neighborhoods are much
more likely affected by low level releases from the immediate
facilities than facilities outside of the zone. For exposures to
the average individual, interzonal transport may be an issue.
For this case, we applied a box-model approach that allowed all
pollutants to disperse evenly throughout the four contiguous
valley zones. This analysis provides an upper-bound
concentration value for these pollutants. We did not account for
pollutant concentrations from sources outside of the valley.
• For exposures within the preselected neighborhoods surrounding
the major facilities, we developed estimates of upper-bound,
lifetime incremental cancer risk using conservative unit risk
factors and assumptions of 70-year exposures from these modeled
concentrations. We present a summary of the highest ranges
predicted in neighborhoods in Table 53. According to our
analysis, approximately 4,600
4-113
-------
TABLE 53
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
COMPARISON ACROSS ZONES OF UPPER-BOUND
ESTIMATES OF LIFETIME, INCREMENTAL
CANCER RISK WITHIN NEIGHBORHOODS
SURROUNDING FACILITIES
ZONE
RANGE OF
HIGHEST LIFETIME
RISK WITHIN
NEIGHBORHOODS
ESTIMATED NUMBER
OF PEOPLE EXPOSED
TO HIGHEST
PREDICTED RISK
(1980 CENSUS)
PRIMARY
POLLUTANTS
CONTRIBUTING
TO RISK
(Weight-of-
Evidence)
Belle 3xl0"3
Charleston/South 1 to 5x10
Charleston
Institute 1 to BxlO
Nitro
1 to 6x10
-3
-3
6
600
2,700
1,300
1,453
Chloroform (B2)
Ethylene oxide (Bl)
Acrylonitrile (Bl)
Ethylene oxide (Bl)
1,3-butadiene (B2)
Trichloroethylene (B2)
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED
AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL CANCER RISKS JN
THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER: IN FACT. THEY
COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL
OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS
EXAMINED.
Weight-of Evidence rating derived by CAG, based on EPA's classification
system: A = proven human carcinogen; B = probable human carcinogen (Bl
indicates limited evidence from human studies, B2 indicates sufficient
evidence from animal studies but inadequate evidence from human studies);
C = possible human carcinogen; D = not classifiable, and E = no evidence
of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
4-114
-------
people may be exposed to an upper-bound, lifetime incremental
risk on the order of 10"^, or approximately one or more chances
in a thousand of contracting cancer from a lifetime exposure to
these pollutants.
• For exposures to the average individual, we developed estimates
of incidence or predicted cancer cases per year using
conservative unit risk factors and assumptions of 70-year
exposures from these modeled concentrations. Specific
conclusions for the average individual exposures follow.
- The total incidence from the pollutants across all zones is
approximately seven-tenths of a cancer case per year, or more
than one cancer case every two years within the study area if we
assume no interzone transfer of pollutants. (See Table 54).
Industrial point sources contribute almost 88 percent to the
total incidence. To account for interzone transport the box
model was used to develop conservative pollutant concentrations
across all valley zones. Table 15 presents these estimated
concentrations for the box model analysis. Table 55 presents
upper-bound incidence from predicted concentrations for all
studied pollutants, based on the box-model concentrations in
Table 15 and the upper-bound potency factors. The predicted
annual cancer incidence rate is 1.8 cases per year. Based on
both modeling analyses, the predicted annual cancer incidence
rate is between .7 cases (based on Gaussian model analysis) and
1.8 cases (based on box model analysis) per year.
- Table 56 presents incidence by point source pollutants across all
four zones. Ethylene oxide, emitted from Institute, provides
almost 56 percent of the total point source incidence rate in the
study area. 1,3-Butadiene contributes almost 20 percent,
acrylonitrile and chloroform each contribute approximately 10
percent. Carbon tetrachloride, methylene chloride, benzene, and
vinylidene chloride contribute the remaining 5 percent to the
study area incidence rate from point source pollutants.
- When we review contribution to incidence by specific facilities
in Table 57, we find that almost 50 percent of the total
incidence rate is contributed by Rhone Poulenc in the Institute
zone and almost 30 percent is contributed by Union Carbide in
South Charleston. County-wide area sources contributed more than
10 percent, while the remaining facilities, DuPont, Occidental,
Union Carbide Technical center, Artel, and Monsanto together
contributed more that 10 percent to the overall incidence rate.
4-115
-------
SUMMARY TABLE 54
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF EXCESS ANNUAL
CANCER INCIDENCE FOR POINT AND AREA SOURCES
(AND PERCENT CONTRIBUTION TO TOTAL
INCIDENCE)
BY
ZONE
ZONE
POINT SOURCES
AREA SOURCES
TOTAL
Belle
Charleston
Institute
Nitro
.056 (7.9%)
.211 (29.8%)
.354 (49.9%)
.0003 (<.1%)
.012 (1.7%)
.059 (8.3%)
.015 (2.1%)
.002 (0.3%)
.068 ( 9.6%)
.270 (38.1%)
.369 (52.0%)
.002 ( 0.3%)
Total
.621 (87.6%)
.088 (12.4%)
.709 (100%)
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES
IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES FOR
THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-116
-------
TABLE 55
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF ANNUAL CANCER INCIDENCE
CALCULATED FROM PREDICTED BOX MODEL CONCENTRATIONS
POLLUTANT INCIDENCE
(Weight of Evidence) (Box Mode 11
Benzene (A) 0.041
Perchloroethylene (B2) 0.002
Methylene chloride (B2) 0.037
Trichloroethylene (B2) 0.002
1,3-Butadiene (B2) 0.438
Ethylene chloride (B2) < 0.001
Ethylene bromide (B2) 0.001
Arsenic (A) 0.018
Benzo(a)pyrene (B2) 0.006
Cadmium (Bl) 0.003
Beryllium (B2) < 0.001
Carbon tetrachloride (B2) 0.021
chloroform (B2) 0.193
Ethylene oxide (Bl) 0.839
Acrylonitrile (Bl) 0.193
Allyl chloride (B2) < 0.001
Vinylidene chloride (C) 0.001
Vinyl chloride (A) < 0.001
Total 1.80
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS JO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Ueight-of Evidence rating derived by CAG, based on EPA's
classification system: A = proven human carcinogen; B = probable
human carcinogen (Bl indicates limited evidence from human studies,
B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not
classifiable, and E = no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-117
-------
SUMMARY TABLE 56
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF EXCESS ANNUAL CANCER
INCIDENCE FOR POINT SOURCE POLLUTANTS BY ZONE
PERCENT
POLLUTANT
(Weiqht of Evidence)*
BELLE CHARLESTON
INSTITUTE
NITRO
TOTAL
CONTRIBUTION
TO TOTAL
POINT SOURCE
INCIDENCE
Carbon Tetrachloride (B2)
.008
<.0001
.008
1.3%
Chloroform (B2)
.040
.018
<.0001
.058
9.3%
Methylene Chloride (B2)
.008 .001
.001
<.0001
.01
1.6%
Acrylonitrile (Bl)
.046
.035
<.0001
.081
13.0%
Ethylene Oxide (Bl)
.163
.185
.348
56.1%
Vinylidene Chloride (C)
.001
.001
.2%
Benzene (A)
<.0001
<.0001
<.0001
<.1%
1.3-Butadiene (B2)
.115
.115
18.5%
Vinyl Chloride (A)
<.0001
<.0001
<.1%
Ethylene Chloride (B)
<.0001
<.0001
<. 1%
Allyl Chloride (B2)
<.0001
<.0001
<.1%
Trichloroethylene (B2)
.0003
.0003
<.1%
Total Incidence by
Point Source
Pol lutants
.056 .211
.354
.0003
.6213
100%
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT
GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN
SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK
ESTIMATES MERE CALCULATED AS MDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL CANCER
RISKS Ul THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER: .IN FACT, THEY COULD
BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE
ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
* Weight-of Evidence rating derived by CAG, based on EPA's classification system: A =
proven human carcinogen; B = probable human carcinogen (B1 indicates limited evidence
from human studies, B2 indicates sufficient evidence from animal studies but inadequate
evidence from human studies); C = possible human carcinogen; D = not classifiable, and E
= no evidence of carcinogenicity.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA, 1987
4-118
-------
TABLE 57
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES OF EXCESS
CANCER INCIDENCES BY SOURCES
(AND CONTRIBUTED TO TOTAL RISK)
PERCENT
CONTRIBUTION
UPPER-BOUND TO TOTAL
SOURCES ZONES ANNUAL CASES INCIDENCE
Dupont
Bel le
.001
.1%
Occidental
Bel le
.055
7.8%
Union Carbide
South Charleston
.207
29.2%
Union Carbide-Tech Ctr
South Charleston
.004
.6%
Rhone Poulenc
Institute
.354
49.9%
Artel
Nitro
<.0001
< .1%
Monsanto
Nitro
.0003
< .1%
Total County-wide
Area Sources
.088
12.4%
Total Incidence from
Point and Area
Source Pollutants
.709
100%
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. 8ECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES MERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES
IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES
FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-119
-------
- Fugitive emissions are an important release type. Table 58
presents the percent contribution to incidence by fugitives and
more nonfugitives releases for the three most important zones.
In the Belle and Institute zone, fugitive emissions contribute
significantly to the level of risk posed by point source
pollutants emitted within the zones.
In this report we have presented a conservative assessment of the
potential cancer risks which a small, select set of pollutants may pose
within various populated areas of the Kanawha Valley. Again, we emphasize
that our risk estimates are to be used for policy development purposes to
set priorities in addressing the concerns and causes of the potential risks
posed by toxic pollutants. These estimates are not predictors of absolute
risk posed bv these pollutants within the Kanawha Valley. The assumptions
of the risk assessment do not allow one to make a definitive statement
concerning the absolute risk posed by a particular pollutant or source. In
this report we have investigated only 12 point source pollutants for their
potential carcinogenic risks. We have not addressed noncancer health
effects nor the effects that the remaining pollutants within the emissions
inventory may pose, if any, to the population.
4-120
-------
SUMMARY TABLE 58
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY AIR QUALITY REPORT
PRELIMINARY RISK SCREENING RESULTS
PERCENT CONTRIBUTION TO POINT SOURCE
INCIDENCE BY FUGITIVE AND
NON-FUGITIVE SOURCES WITHIN
BELLE, CHARLESTON, AND INSTITUTE
PERCENT
PERCENT CONTRIBUTION
UPPER-BOUND CONTRIBUTION TO INCIDENCE
ANNUAL CASES TO INCIDENCE WITHIN ZONE
FROM WITHIN ZONE BY NON-
ZONE POINT SOURCES BY FUGITIVES FUGITIVES
Belle .056 84.6% 15.3%
Charleston .211 36.6% 63.2%
Institute .354 70.2% 29.7%
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION. RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT. NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS JN THE. KANAWHA VALLEY. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER: IN FACT. THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET
PRIORITIES FOR THE TOPICS EXAMINED.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA,
1987
4-121
-------
REFERENCES
1. Auer, A. H. Jr. Correlation of Land Use and Cover With Meteorological
Anomalies. Journal of Applied Meteorology, Vol 17, pp. 636-643, May 1978.
2. Beard, C. Letter to Joe Padgett. Tables of Industrial Estimated Fugitive
Emissions. July 1986.
3. Bjorklund, J. R. and J. F. Bowers User's Instructions for SHORTZ and LONGZ
Computer Programs. Vol. I. Report Number EPA-903/9-82-004a. March 1982.
4. Bowers, J.F., J.R. Bjorklund, and C.S. Cheney. Industrial Source Complex
(ISC) Dispersion Model User's Guide. Vol. I. Report Number EPA-450/4-79-
030. December 1979.
5. Brodzinsky, R. and Singh, H.B. SRI International. Volatile Organic
Chemicals in the Atmosphere: An Assessment of Available Data. Prepared for
the U.S. Environmental Protection Agency under Contract 68-02-3452, April
1983.
6. Engle, R. Memorandum to Carl Beard. Subject: Analytical Results from
Tenax/GC Samples Collected December 6, 9-11, 1985. January 9, 1986.
7. Farley, D. Memorandum to Carl Beard. Subject: Chemical Emissions
Reduction at Kanawha Valley Chemical Manufacturing Facilities. March 1986
8. Fitz-Simons, T., Lumpkin T. and McClenny, W. Report on the Air Monitoring
in the Kanawha Valley. West Virginia. U.S. Environmental Protection
Agency, Environmental Monitoring Systems Laboratory, Research Triangle
Park, N.C. June 1986.
9. Hanna, S. R. Buggs, G. A. Hosker, R. P. Handbook on Atmospheric
Diffusion. Atmospheric Turbulence and Diffusion Laboratory. National
Oceanic and Atmospheric Administration. 1982.
10. Irwin, J. S. Estimating Plume Dispersion - A comparison of Several Sigma
Schemes. Journal of Climate and Applied Meteorology. Lancaster,
Pennsylvania 17604.
11. Irwin, J. S. Dispersion Estimate Suggestion No. 8 Environmental
Applications Branch. Environmental Protection Agency, Environmental
Sciences Research Laboratory.
12. Pellizzari, E.D. Identification and Estimation of N-Nitrosolimethvlamine
and other Pollutants in the Baltimore, Maryland and Kanawha Valley Areas.
Progress Report prepared by Research Triangle Institute, EPA Contract 68-
02-1228. January 1976.
4-123
-------
References (Continued)
13. Pellizzari, E.D. Analysis of Organic Air Pollutants in the Kanawha Valley.
West Virginia and the Shenandoah Valley, Virginia. Research Triangle
Institute, Resarch Triangle Park, N.C.
14. Robertson, J. Memorandum to Vic Guide. Subject: Kanawha River Air Sample
Results. January 10, 1986.
15. Sullivan, D. A. A Screening Methodology for Air Quality Analysis. Versar
Inc. Prepared for U.S. Environmental Protection Agency, Office of Policy
and Program Evaluation, Integrated Environmental Management Program,
Contract No. 68-01-7002. April 1985a.
16. Sullivan, D. A. Memorandum to A. Cimorelli. Subject: Exposure Height for
Belle and Institute Meteorological Systems. August 23, 1985b.
17. Sullivan, D. A. Versar, Inc. Evaluation of the Performance of the
Dispersion Model SHORTZ for Predicting Concentrations of Air Toxics in the
U.S. Environmental Protection Agency's Philadelphia Geographic Study.
Prepared for the U.S. Environmental Protection Agency, Integrated
Environmental Management Division, under Contract No. 68-02-3181.
January, 1985c.
18. U.S. Environmental Protection Agency. The Measurement of Carcinogenic
Vapors in Ambient Atmospheres. EPA-600/7-77-055.
19. U.S. Environmental Protection Agency. AP-42 Compilation of Air Pollutant
Emission Factors Volume I: Stationary Point and Area Sources. Fourth
Edition. 1984. U.S. Environmental Protection Agency. Office of Air
Quality and Planning and Standards. Research Triangle Park, NC.
20. U.S. Environmental Protection Agency. Draft Plan for Meteorological
Monitoring, Dispersion Modeling, and Data Processing for Phase I of the
Kanawha Valley Study, August 1985.
21. U.S. Environmental Protection Agency. Quality Assurance Plan for
Meteorological Monitoring Network in Kanawha Valley. Revision "0",
February 1986; Revision "1", April 1986a.
22. U.S. Environmental Protection Agency. Guideline on Air Quality Models
(Revised). Office of Air Quality Planning and Standards. Research Triangle
Park, N.C., July 1986b.
23. U.S. Environmental Protection Agency. Quality Assurance Plan for the
Meteorological Network in Kanawha Valley. West Virginia. 1986. U.S.
Environmental Protection Agency. Regulatory Integration Division.
Integration Environment Management Program.
4-124
-------
References (Continued)
24. U.S. Environmental Protection Agency. Draft Superfund Public Health
Evaluation Manual. U.S. Environmental Protection Agency. Emergency and
Remedial Response. Office of Solid Waste and Emergency Response.
25. Vincent, J. R. Overview of Environmental Pollution in the Kanawha Valley,
West Virginia. U.S. Environmental Protection Agency, National Enforcement
Investigations Center, Denver, Colorado. August 1984.
4-125
-------
KANAWHA VALLEY AIR QUALITY REPORT
Technical Appendices
(Copies available upon request)
Appendix Title
A Interpretation of LONGZ Model Output
A.l - LONGZ Model Results (all sources combined)
A.2 - LONGZ Model Results (Point and Area Source
breakdown)
A.3 - Population-Weighted Concentration by
Pollutant (Main Grid) for LONGZ
B Meteorological Monitoring Program
8.1 - Detailed Description of Meteorological
Monitoring Program
B.2 - Rationale for the Selection of 20 m as the
Preferred Monitoring Height for Wind Data
B.3 - Specifications for Meteorological
Monitoring Systems
B.3.1 - Specifications for Climatronics F460 and
R.M. Young Propeller Anemometer Wind
Systems
B.4 - Description of Data Recovery Rates and
Conformance to Data Quality Objective
B.5 - Monthly Wind Rose Data for all Four Zones
B.6 - Tables of Standard Deviation of Horizontal
and Vertical Wind Direction
C Kanawha Valley Meteorological Monitoring/
Dispersion Modeling Protocol
D County-wide Area Source Emission Estimates
E Input Data Files for ISCLT Dispersion
Modeling Analysis
F Comparison of Box Model Methodology with
Gaussian Modeling
G Summary of Archived Data (SAROAD)
4-127
-------
Results of Ambient Air Quality Monitoring
During the Toxics Screening Study
- Results of EPA Region III Monitoring
(December 9-12, 1985)
- Results of West Virginia Air Pollution
Control Commission Monitoring
(December 6-11, 1985)
- Results of EPA Office of Research and
Development Monitoring (March 4,
1986 through April 12, 1986)
Preliminary Risk Estimates for
Formaldehyde and Propylene Oxide
Fugitive Emissions Estimates for Primary
Pollutants at Neighborhood Sites
4-128
-------
Chapter Five
Drinking Water Analysis
-------
KANAWHA VALLEY DRINKING WATER REPORT
I. PURPOSE AND INTRODUCTION
The Kanawha Valley Toxics Screening Study has four stated
objectives that have provided overall direction for this technical
report. In this drinking water analysis, we attempted to
address these objectives. The four objectives are:
1. Using available data bases, identify many of the
chemicals routinely released or present within various
exposure pathways;
2. Develop a sense of the potential public health concerns
these toxic pollutants may pose in various exposure
pathways, based on health effects and exposure in-
formation;
3. For a select number of chemicals, provide an initial
conservative assessment of potential cancer risk and
potential noncancer health risk for predicted for observed
concentrations within exposure pathways; and
4. Identify data and information gaps and outline needs
and options for future study directions to enable a
more detailed investigation of health issues where
warranted.
To meet the first objective, we reviewed available monitoring
data from community water supply systems to determine pollutants
present within these systems. We relied solely on monitoring
information required by federal regulations, and only for public
water supply systems. Due to this limited set of pollutants which
are monitored by the State, we may have underestimated health
risks from drinking water since we did not account for all the
potential pollutants. For a small set of pollutants detected
within the water supply systems, trihalomethanes, we developed a
conservative assessment of potential cancer risk. For all monitored
pollutants for which we have health effects information, we assessed
the potential for noncancer health effects. Finally, we outlined
several datagaps which limit our ability to draw additional
conclusions of possible health effects posed by pollutants in the
drinking water systems within the Kanawha Valley.
5-1
-------
This report is divided into nine sections. The first two
sections outline the general methodology and the health effects
analyses used within the report. Section ill and IV provide
descriptions of the general study area and the monitoring
requirements for the drinking water systems. Section V outlines
our selection of pollutants for the exposure and health assessment.
In Section VI we develop exposure concentrations for the selected
pollutants. Section VII presents health effects information for
the pollutants studied. In Section VIII we present the results of
our risk calculations for a small set of pollutants within the
drinking water systems. Section IX presents our summary and
conclusions.
II. DESCRIPTION OF THE GENERAL METHODOLOGY
In this section we briefly outline the methods used to
estimate exposures to a limited set of toxic contaminants in
the public water supply systems and describe their potential
health effects from these exposures. We combine estimates of
human exposure levels to these chemicals with information on
contaminant toxicity to estimate potential excess lifetime
average individual risks and excess cancer incidence (i.e.,
risk to the population). We also estimate the number of people
exposed to the potential increased risk of noncancer health
effects.
We have reviewed data from public water supply systems
within the Kanawha Valley study area. These systems serve the
majority of the population within the study area. The State
does not collect water quality information for toxics for the
other two systems: non-community water supply systems and
private wells.
For the public water supply systems, we reviewed pollutants
regulated by the Safe Drinking Water Act. The Safe Drinking Water
Act requires public water supply systems to monitor regularly for
designated drinking water contaminants. These pollutants
cannot exceed Maximum Contaminant Levels (MCLs) determined by
EPA.
Although we only developed potential exposure concentrations
for MCL pollutants, we did review available priority pollutant
scans of the major public water system (the West Virginia water
5-2
-------
Company) which serves more than 73% of the public water supply
system population to determine the potential for other pollutant
exposures. These priority pollutant scans are in Technical
Appendix A. According to Technical Appendix A, the priority
polluant scans did not detect priority pollutants except for
several trihalomethanes during the monitoring.
From monitoring data, we analyzed possible exposure of the
population to disinfection products, inorganics, and pesticides in
the drinking water. These are the classes of MCL pollutants
monitored by the West Virginia Department of Health.
For disinfection products (trihalomethanes) formed from the
chlorination of drinking water, we developed average concentration
values based on monitoring data over a three-year period. For
these average concentrations, we assumed standard EPA exposure
assumptions (a 70-kg adult consumes two liters of water per day
over a 70-year period) to develop exposure concentrations. We
then multiplied upper-bound cancer unit risk values, developed by
EPA's Carcinogenic Assessment Group (CAG), by the exposure concen-
trations to obtain lifetime, incremental individual cancer risk
for trihalomethanes.
Due to the limited monitoring information for metals and
pesticides, we did not develop average system concentrations for
these pollutants, and did not assess potential lifetime excess
average individual cancer risk for inorganics nor pesticides.
Many observations for these pollutants were below the detection
1imit.
For noncancer health effects associated with exposure to
trihalomethanes, we compared our estimated average concentrations
to EPA developed Reference Dose (RfD) criteria to determine if
such criteria are exceeded. For metals and pesticides, we
compared the RfDs to the highest monitored value. Again, many
of these values are below the detection limit.
Health Assessments
Cancer Assessment: This study uses a risk assessment screening
methodology to evaluate and compare, in a very limited fashion,
the potential health risks from exposure to a limited set of
toxic pollutants. Risk to an individual ls defined as the increased
probability that an individual exposed to one or more chemicals
will experience a particular adverse health effect during his
or her lifetime. Several measures of carcinogenic risk are
used in these analyses, including estimated risk to an individual
and projected risk to the entire population (i.e., incidence).
This risk screening methodology involves both a gualitative
and guantitative assessment of the potential carcinogenicity
of the selected pollutants. As a screening study, this
5-3
-------
analysis employs both types of assessments. The Carcinogen
Assessment Group (CAG) reviews the evidence of carcinogenicity
for selected pollutants, and classifies pollutants as human
carcinogens (Group A), probable human carcinogens (Group B),
possible human carcinogens (Group C), not classified as
carcinogens due to inadequate evidence (Group D), and not
carcinogenic to humans (Group E). These classifications
accompany all quantitative risk estimates within this study.
For those chemicals in groups h, B, and C, CAG provides
quantitative, upper-bound estimates of carcinogenic unit risk
factors. A unit risk factor allows the calculation of the
estimated individual risk posed by exposure to a chemical given
certain exposure assumptions. To calculate individual risk,
a chemical's unit risk factor is multiplied by the estimated
concentration of the pollutant.
Individual = Unit Risk Factor (ug/l)-1 x Concentration (ug/1)
Lifetime Risk
The concentration is simply the concentration of the chemical
found in the drinking water. The unit risk factor assumes that
an average person weighs 70 kg and drinks 2 liters of water each
day (2 1/day).
5-4
-------
SAMPLE CALCULATION
POSSIBLE RISK OF CANCER ASSOCIATED WITH
INGESTION OF CHLOROFORM IN
DRINKING WATER
Estimated Average
Individual Lifetime Risk
Unit Risk Factor x Concentration
Cancer Unit Risk Factor
for the Ingestion of
Chloroform
2.3 x 10~6 (ug/1)"1
Concentration
For this example assume 1 ug/1
of chloroform in the drinking
water.
Esimated Average
Individual lifetime Risk
2.3 x 10*"6 (ug/1)"1 x 1 ug/1
2.3 x 10"6
This estimated average individual risk, based on upper-
bound unit risk factors, indicates that conservatively,
there is approximately a two in a million increased chance of
developing cancer over a 70-year lifetime, constant
exposure to this concentration of chloroform. The actual
risk could be considerably lower.
For the disinfection products such as trihalomethanes,
we present risks to the average individual based on average
concentrations monitored by the State. incidence, another
measure of risk, is the probable number of cancer cases
expected over a 70-year period when a population is exposed to
the average individual exposure concentration. incidence is
calculated by multiplying individual risk by the population
number and can be annualized by dividing by 70.
The overall uncertainty in the risk assessment assumptions
is great enough that results should be considered rough indicators
of the probable magnitude of the effects, not as precise, site-
specific prediction of effects. Furthermore, EPA policy dictates
that the most conservative approach be used in developing unit
risk factors. Consequently, EPA develops estimates that are
generally unlikely to underestimate the true potency of the chemical.
Indeed, in many cases the lower-bound estimate for a chemical
may be zero. It is important to remember the conservative nature
of our risk estimates and the limitations of these values
throughout the study.
-------
Noncancer Health Effects: No currently accepted techniques exist
for estimating the probability or incidence of noncancer effects.
Therefore, in evaluating the potential noncancer health risks
within this study, we rely on benchmark values or Reference Doses
(RfDs), expressed in units of mg/kg/day, which represent dose
levels below which adverse health effects are not likely to occur
in most people. Unlike cancer effects, noncancer health effects
are assumed to be threshold events. This threshold assumption
means that noncancer health effects are assumed more likely to
occur above these exposure concentrations. Exposures that are
less than the RfD are not likely associated with noncancer health
effects and therefore are less likely to be of regulatory concern.
In our analyses, we modify the RfDs to determine safe
ingestion exposure concentrations for a pollutant through ingestion
of drinking water. (In our analysis, we assumed the only exposure
route for the studied pollutant is drinking water ingestion. In
fact, all exposure routes must be considered when reviewing RfD
levels.) To determine safe exposure concentrations, the RfD is
converted to an exposure concentration in units of ug/liter.
This exposure concentration implies several assumptions. Again,
we assume that the average individual weighs 70 kg and drinks 2
liters of water a day, and the average individual is exposed to
these concentrations for 70 years. Once developed, these criteria
are then compared to ambient pollutant concentrations. When
ambient pollutant concentrations exceed the RfD-based criteria,
there is an increased probability of a noncancer effect occuring
within the exposed population. Currently we cannot quantify the
increased probability, but only note that there is an increased
probability when the criteria are exceeded.
Ill. DESCRIPTION OF STUDY AREA
The Kanawha valley study area contains three types of public
water supply systems: private wells, non-community water supply
systems, and community water supply systems. The community water
supply systems are the predominant water supply with 14 such systems
serving households within the study areas. TABLE 1 lists the
community water supply systems, the population served by each system,
and their water sources. Map 1 outlines the boundaries of these
systems. (Map boundaries are numbered in sequence according to
TABLE 1.)
5-6
-------
table 1
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
COMMUNITY WATER SYSTEM, POPULATIONS, SOURCES
WITHIN STUDY AREA
MAP
KEY
SYSTEM
POPULATION
SERVED
1. Kanawha Falls Public Service District 3,885
2. Amstrong Public Service District 2,125
3. Smithers Public Water Supply 2,500
4. West Virginia Water Coirpany (Montgomery District) 4,907
5. Pratt Crown Hill Public Water Supply 2,300
6. Cedar Grove Water Coirpany 3,600
7. West Virginia Water Coirpany (Kanawha Valley District) 190,000
8. Washington Public Service District 1,500
9. St. Albans Public Water Supply 16,200
10. Sissonville Public Water System 4,000
11. North Putnam Water Supply 2,600
12. Kanawha Orchard Public Service District 525
13. South Putnman Public Service District 4,900
14. Winfield Public Water Supply 1,500
WATER
SOURCE
Kanawha River
Armstrong Creek
Mine Impoundments/Wells
Kanawha River
Kanawha River
Kanawha River
Elk River
Coal River
Coal River
Pocatalico River
West Virginia Water Coirpany
(Kanawha Valley District)
South Putnam Public Service District
Surface Impoundments
Surface Impoundments
Total:
240,542
Source: Department of Health, West Virginia, 1987
-------
By comparison, the other two systems, private wells, and non-
community water suppy systems, serve a much smaller population.
While it is difficult to estimate the exact proportion of these
users to the number of persons on public water systems, estimates
from the Charleston Planning and Development Commission suggest
that there may be approximately 22,000 private wells in Kanawha
County (Personal Communication, McQueen, 1987). It would be difficult
to assess the quality of the water in these supplies since private
wells are not required to monitor for water quality in Kanawha
Valley and hence very little monitoring data exists.
Non-community water supply systems serve transient or
intermittent users and can include facilities such as campgrounds,
parks, schools, gas stations, and motels which have their own
water supply systems. Such systems monitor for coliform bacteria,
turbidity, and nitrate.
For this analysis we examined only the public water supply
systems. Again, we are limited to available monitoring data for
selected toxics within the public water supply systems. Although
the public water supply systems serve most of the population
within the study area, individual wells and non-community water
systems with wells may have toxic exposure concentrations through
contamination of the groundwater supply from hazardous waste
sites within the study area. Contamination of wells has been
detected within the valley; Superfund Preliminary Assessments
have detected well contamination from Heizer Creek Dump and
South Charleston Municipal Landfill (Draft Site Inspection
Report, NUS Corp., 1984). The lack of exposure information for
private wells serving households and non-community water systems
is a limitation to this drinking water analysis. Although the
number of residents obtaining drinking water from private wells
is relatively small, private wells and non-community water supply
systems could be a source of toxic exposure to individuals.
IV. REGULATORY OVERVIEW OF PUBLIC COMMUNITY WATER SYSTEMS
The National Interim primary Drinking water Regulations,
established under the Safe Drinking Water Act, set Maximum
Contaminant Levels (MCLs) and monitoring requirements for community
water supplies. These MCLs include inorganic chemicals, three
organic chemical groups (chlorinated hydrocarbon pesticides,
chlorophenoxy pesticides, and trihalomethanes) radionuclides,
microbes, and turbidity.
The State is responsible for monitoring MCL pollutants.
Inorganics are monitored once a year. Additional monitoring
occurs if contamination is believed to have occurred, pesticides
are also monitored once a year during the spring or fall season
when contamination, usually from run-off, may occur.
5-8
-------
rii' ji—*7 ,vnk^,vKiMZ?y',ic> vmry
9
w^mm
MAP 1
KANAWHA VALLEY DRINKING WATER REPORT
W;yj-i
PUBLIC WATER SUPPLY SYSTEMS
(upper Kanawha River)
»
f
len .
TWSj:. I *
332
rfhajnt
*$#&&
A
Jjj)
£
'fcintrk*
* f
n^ttflrt&ugh
£
-------
Ai/ t -¦" r ^ r—W.-^- V *1.1 -" ~'J z>f>-\ »* • imr»srT= xSSc^St ••> ®L ^ v-ilvyi'^
W> v.
'Irsr-A
r.-~j
^ 'M?-^rm§
^ 1 ' 'Xfctfrc
u »/ ci» ,-- 53t ^T/— — • ,."v *¦* ~
;»""•' l"^' Mm?7 C » , ^T fe! 7orfyS*+
"t T* -^. 'X' ".''V O ' .v-4?
iTyiw Sata^C* J^l3Sc«
tSne^
!9£g$
•''^fiW,* >_ j~
i Vvi 5-
s»' > v..-^,-^^ -a
,3it3s$M>K-
~ UncoIpP5:* 'Vi ,-3? ft ~- 'i'—
—Rfnjtfart
,2 S>,if . a.«»KM
yp?r~' .• - v^s ¦*-¦>*•» >\ v^ - •w--1l-»ji rus ri*'« + ^-*:*?*-+'.-*"»r..• cj s _/v
;yf3%&». «&2? -#w -p.hI"
P^fe^!S?X^Sl
^ * f;, ->-* j- 'X^^cL-— . \ - •>) c /*.
*****».>»•' .^yV' ^ ^ =»_ -Jr {(
Twz&-^:n ''
sssssa^
^ r . ir-i i -ij» \ i/^l 17ti fj ^1 r-. • .
*2^a^^rt^
^*1
"Vt^vST*
¦<®o^ ;
>_ ^r> 'k^v, ^4g>v,rr' J
-•" -• -?: i i ' "<"->' '
-^••^r v V '
MAP 1
KANAWHA VALLEY CHINKING WATER REPORT
PUBLIC WATER SUPPLY SYSTEMS
(lower Kanawha River)
*>?i£> v^"- ^
¦#¦-
' ^r rJ'^~
WM^&i
: ---V^U-/v_jv-A1^-,-iS1-
.W0^*§
+-y?-:-.P~z. r i-.«r
-------
Community water systems which serve a population of 10,000
or more individuals and which use chlorine to disinfect drinking
water must also monitor for trihalomethanes. Systems must monitor
at least four times a year. The MCL level pertains to the total
trihalomethane concentration, the sum of the concentrations of
chloroform, bromoform, bromodichloromethane, and chlorodibromo-
methane.
V. POLLUTANT SELECTION
We analyzed MCL pollutants for exposure concentrations
determined from monitoring information provided by the State.
These pollutants and their sources are reviewed below.
Disinfection Agents and By-products
Disinfection is a process in which water is treated with
chemical agents, most commonly chlorine, to kill microbial pathogens
(disease causing agents). Without disinfection, people could be
exposed to numerous infectious agents through their drinking
water and periodic outbreaks of diseases such as cholera or
typhoid could result. Disinfection techniques commonly maintain
a residual disinfecting agent (chlorine if the disinfection
process is chlorination) in the treated water, to continue
disinfecting water until it is used. Unfortunately, potentially
harmful by-products such as trihalomethanes result from the
reaction of chlorine with organic matter (such as plant and
animal matter) in the water.
We have analyzed the potential effects of chlorination.
Organic chemicals of concern created through this process include
the trihalomethanes:
° chloroform;
° bromoform;
° chlorodibromomethane;
° dichlorobromomethane.
Inorganics
A number of inorganic substances, primarily metals and
other minerals, are found in drinking water. Most of these
substances are probably from natural background sources (i.e.,
they are present in soils and are picked up by water traveling
through those soils). Some of these chemicals may be the result
of man-made contamination, but, for most substances, this is
difficult to determine and specific incidents that cause the
contamination cannot be identified.
5-9
-------
Metals and minerals for which we have reviewed available
monitoring data include:
0 arsenic;
0 barium;
0 cadmium;
0 chromium;
0 lead;
0 mercury;
0 selenium;
0 silver; and
0 zinc;
Pesticides
Pesticides could contaminate imported surface water primarily
through runoff from agricultural areas through which the water
travels; such contamination would not necessarily have to occur
in or near Kanawha Valley to affect local water. Pesticides
from local sources could cause ground-water contamination. The
State monitors for the following pesticides:
0 Endrin;
° Lindane;
0 Methoxychlor;
0 Toxaphene;
° 2,4-D;
0 Silvex (2,4,5-TP).
VI. EXPOSURE
Disinfection Agents and By-products
TABLE 2 presents average concentrations for monitored
trihalomethanes within three community water supply systems:
Saint Albans, West Virginia Water Company (Kanawha Valley
District), and the West Virginia Water Company (Montgomery
District) Federal regulations require these three systems to
monitor for trihalomethanes on a regular basis. Although the
West Virginia Water Company (Montgomery District) serves less
than 10,000, the system monitors for trihalomethanes as part
of the larger West Virginia Water Company. These monitoring
data are limited and subject to significant uncertainties
introduced by seasonal and sampling location. Seasonal variations
are due mainly to the variations in organic matter in the raw
water and the effects of temperature on the formation of trihalo-
methanes. As the distance from the chlorination point increases,
the reaction time and, therefore trihalomethane formation,
also increases giving rise to the sampling point variations.
EPA's Office of Water in their water quality programs does not
attempt to quantify this variation, but rather accounts for the
5-10
-------
TABLE 2
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
PRELIMINARY RISK SCREENINNNG RESULTS
ESTIMATES OF AVERAGE LEVELS OF INDIVIDUAL TRIHALOMETHANES
IN COMMUNITY WATER SUPPLY SYSTEMS BASED ON SIXTY OBSERVATIONS
SOURCE
Saint Albansl
West Virginia Water Company^
(Kanawha Valley District)
West Virginia Water Company
(Montgomery District)1
Small Systems^
POPULATION
UG/L
CHLOROFORM
DICHDOROBRO-
MOMETHANE
CHLORODIBRO-
MOMETHANE
BROMOFORM
THM
TOTAL
16,200
29
12
1
<1
42
192,600
72
4
1
<1
77
4,907
50
4
<1
<1
54
26,835
42-77
^Estimates for Saint Albans and West Virginia Water Conpany (Montgomery District) are based on data averaged
from 1983 to October 1, 1986.
^Estimates for the West Virginia Water Conpany are averaged with data from 1984-October 1, 1986. The West-
Virginia Water Conpany includes North Putnam Water Public Water Supply System.
^We applied a range of the lowest and highest observed THM concentration within the study area to the smaller
systems. The lowest observed average THM concentration is 42 ug/1 (Saint Albans), the highest concentration is
77 ug/1 (West Virginia Water Company, Kanawha Valley District).
Source: Department of Health, West Virginia, 1987
-------
the variation by averaging quarterly monitoring samples from four
different points within the delivery system.
Other small systems within the Kanawha Valley do not
monitor for trihalomethanes although such systems chlorinate
the supply systems (personal communication, Hodges, 1986). To esti-
mate trihalomethane totals for these systems, we assume a range of
trihalomethane concentrations for these systems. For these smaller
systems we used the lowest observed concentration of trihalomethanes,
which was 42 ug/1 (Saint Albans Water District) and the highest
observed concentration of trihalomethanes which was 77 ug/1 (West
Virginia Water Company, Kanawha Valley District). Based on these
observations, we assigned a range of 42 to 77 ug/1 for the smaller
systems.
It is important to note that these systems are beneath.the
MCL requirement for total trihalomethanes (100 ug/1), in some
cases by more than half.
Inorganics and Pesticides
TABLE 3 presents the highest observed concentrations for metals,
minerals, and pesticides regulated by the Primary Drinking Water
Regulations, inorganics and pesticide concentrations come from a
more limited data set than the trihalomethanes; Federal regulations
require only a yearly measurement for these pollutants.
To obtain a larger data set, we reviewed monitoring data from
1981 to 1985.
VII. TOXICOLOGY EVIDENCE
Cancer Effects
TABLE 4 and TABLE 5 summarize the available evidence on
the chronic toxicity of those substances for which we reviewed
monitored concentrations. TABLE 4 presents the estimates of
the carcinogenic potential and unit risk factors for disinfection
agents and by-products, and inorganics and pesticides. Evidence of
the carcinogenicity of trihalomethanes is strongest for chloroform.
EPA classifies it as a probable carcinogen (B2), and has calculated
a unit risk factor for it. Experiments indicate that chlorodibro-
momethane and dichlorobromomethane may also be carcinogens (NTP,
1984), and many toxicologists suspect that bromoform may also be
carcinogenic. Since EPA's Carcinogen Assessment Group (CAG) has
not developed a potency evaluation for these chemicals, we analyzed
risks from total trihalomethanes (THMs) as if all THMs are as
potent as chloroform.
5-12
-------
TABLE 3
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
WATER QUALITY DATA
Maximum
Contaminant
HIGHEST OBSERVED CONCENTRATION OF PESTICIDES IN TREATED WATER (1981-1935)'
(jig/1)
Level
Armstrong
Winfield
S. Putnam
Sissonville
Washington
Kanawha
Pratt
Cedar Grove
PSD
Falls
rsenic
50
<1
<1
<1
10
<1
5
10
10
arium
1000
9
10
100
90
7
<1
40 .
40
admium
10
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
hromium
50
<1
<1
<1
<1
<1
<1
<1
<1
ead
50
<1
<1
<1
<1
<1
<1
<1
<1
ercury
2
<.2
<.2
<.2
<.2
<.2
<.2
<.2
<.2
elenium
10
<2
<2
<2
<2
<2
6
<2
<2
ilver
50
<.2
<•2
<.2
<.2
<.2
<.2
<.2
<.2
inc
NA
NA
NA
NA
NA
<5
NA
NA
ndrin
0.2
<.01
<.01
<.01
<.01
<.01
<.01
<.01
<.01
indane
4
<.005
<.005
<.005
<.005
<.005
<.005
<.004
<.004
lethoxychlor
100
<.03
<.03
<.03
<.03
<.03
<.03
<.03
<•03
oxaphene
5
<.2
<.2
<.2
<.2
<.2
<.2
<.2
<.2
iilvex
10
<.04
<.04
<.04
<.04
<•04
<.04
<.04
<.04
!, 4 D
100
<.2
<.2
<•2
<.2
<.2
<2
<2
<2
'The "LESS THAN" Sign (<) indicates a concentration belov the detection limit.
Cedar Grove WV Water Co. WV/Montgomery St Albans Smither:
<1
<1
10
<1
<1
<1
40
<1
<0.1
<0. 1
0.2
<0.1
2
<1
<1
<1
<1
<1
1
<1
0.7
<.2
<.2
<.2
<2
<2
<2
5
<.2
<.2
<¦2
<.2
<5
10
<5
NA
<.01
<.01
<.01
NA
<.004
<.004
<.005
NA
<.03
<.03
<•03
NA
<.2
<.2
<.2
NA
<.04
<.04
<.04
NA
<2
<2
<2
NA
Source: Department of Public Health, West Virginia, 1987.
5-13
-------
TABLE 4
KANAWAH VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND CANCER UNIT RISK VALUES:
POLLUTANTS IN DRINKING WATER
POLLUTANT
Chloroform (tota1
trihalomethane
Arsenic
Lindane
Toxaphene
UNIT RISK FACTOR
INGESTION (uq/1)~1
2.3 x 10 -6
4.5 x 10"4
3.1 x 10"5
3.2 x 10-5
WEIGHT-OF-EVIDENCEJ
B2
A
B2-C
B2
1 Weight-of-Evidence rating derived by CAG, based on EPA's classification
system: A = proven human carcinogen; B = probable human carcinogen (Bl
indicates limited evidence from human studies, B2 indicates sufficient
evidence from anumal studies but inadequate evidence from human studies);
C = possible human carcinogen; D = not classifiable, and E = no evidence
of carcinogenicity.
Source: Carcinogen Assessment Group, 1986.
5-14
-------
Among the inorganics and pesticides, arsenic, lindane, and
toxaphene have been found by EPA's CAG to be potentially
carcinogenic. (EPA considers cadmium and hexavalent chromium
carcinogenic via inhalation of air, but not via ingestion from
drinking water). The evidence for arsenic's carcinogenicity by
ingestion is epidemiologic evidence correlating high drinking
water arsenic levels with a form of skin cancer which is rarely
fatal. However, the carcinogenicity of arsenic ingested through
drinking water is the subject of some dispute. Some scientists
believe that arsenic may not be harmful, and may even be beneficial
at low levels. EPA's Office of Drinking Water has concluded
that arsenic is not harmful at low levels, and proposed a
Maximum Contaminate Level (MCL) of 50 micrograms per liter.
Noncancer Effects
The estimated toxicity of water disinfection agents and by-
products for effects other than cancer is summarized in TABLE 5.
We present benchmark criteria based on Reference Doses, below
which exposure is not considered likely to develop adverse
noncancer health effects within the general population over a
lifetime.
Strong scientific evidence exists that airborne lead can
have adverse health effects, especially for children. Waterborne
lead has also been shown to have adverse health effects. The
health effects of lead are documented in Volume IV, 1986 "Air
Quality Criteria for Lead". Evaluating effects from chronic
lead exposure is a complex issue involving evaluation of the total
body burden from air, water, and dust exposure and is the subject
of current EPA research. No EPA reference dose has yet been estimated.
Because of the issue's complexity, the ongoing Federal initiatives,
and the lack of an estimated no-effect threshold, we have not
estimated health effects from lead exposure. All lead concentrations
are below the detection limit in our dataset.
Cadmium, mercury, silver, hexavalent chromium, and selenium
may pose risks of noncancer effects at low levels. Other sub-
stances, such as barium and zinc, are guite a bit less toxic.
Zinc, is considered beneficial in drinking water at some concentra-
tions but harmful at higher concentrations.
VIII. RISK ESTIMATES
Disinfection By-products
We have calculated individual lifetime risks, based on upper-
bound unit risk factors and aggregate incidence for trihalomethanes
exposure by assuming that all trihalomethanes have the same
carcinogenic potency as chloroform. Structural similarities and
recent studies (NTP, 1984) suggest that all four may eventually be
5-15
-------
TABLE 5
Pollutant
Chloroform
Bromoform
Barium
Chromium - III
Chromium - VI
Mercury (inorganic)
Methyl Mercury
Silver
KANAWHA VALLEY DRINKING WATER REPORT
NONCANCER HEALTH EFFECTS
POLLUTANTS IN DRINKING WATER
Health Effect
liver cysts
hepatic lesions
Bromodichloromethane hepatic lesions
Dibromochloromethane hepatic lesions
increased blood
pressure
none
observed
none
observed
renal and
kidney damage
CNS effects
argyria
Reference Dose'
(ug/D
350.0
210.0
17.5
21.0
1,750.0
35.000J0
175.0
70.0
1.0
105.0
Fluoride
2,4-D
dental fluorosis
liver effects
2,100.0
350.0
Study RfD
Based On
Heywood et al. (1979)
Chu et al. (1982)
Chu et al. (1982)
Chu et al. (1982)
Perry et al. (1983)
Ivankovic and
Preussmann
(1975)
MacKenzie et al.
(1958)
Fitzhugh et al.
(1950)
Clarkson et al.
(1973)
Gaul and Staud
(1935)
Blumberg and Carey
(1934)
East et al.
(1980)
Underwood(1977)
Hazelton Labs (1983)
'EPA/RfD indicates a benchmark value derived from no-effect threshold value.
Exposures greater than the RfD are assumed to be associated with an increase
in noncancer health effects. EPA reviews and verifies these thresholds internally
through the Reference Dose Workgroup.
5-16
-------
found carcinogenic. EPA has used this approach in the past in
developing drinking water regulations. TABLE 6 presents our
estimates of average individual risks from trihalomethanes, and
TABLE 7 presents our estimates of increased incidence for each
source of community water supplies.
We estimated that the upper-bound increased lifetime risk of
cancer for average exposed individual drinking water is approximate
one to two chances in ten thousand. This level is comparable to wh
we have seen in several other cities. For the entire study
area, we project an upper-bound increased incidence of more than
one-half cases per year for a 70-year exposure of this concen-
tration level of trihalomethanes.
In order to compare estimates of increased incidence for
individuals drinking public water supplies in the valley directly
to the increased incidence caused by the inhalation of selected
point source air emissions, we calculated the water related
incidence within the smaller study area of the air analysis by
multiplying the average individual risk involved with drinking
water from the West Virginia Supply Company (2xl0~4) by the
populations in each of the four separate air zones defined
in the "Kanawha Valley Air Quality Report" (Belle, Institute,
Charleston, and Nitro) and divided by 70. The total population of
these four zones is 99,660. We estimate that 93% of the population
in the four air zones is served by the West Virginia Water Company
( 100% in Belle and Charleston, 85% in Institute, and 60% in Nitro)
Nitro and Institute are also served by South Putnam PSD, Kanawha
Orchard PSD, and Saint Albans Public Water Supply. By these
assumptions the incidence related to trihalomethane concentrations
in these four zones is approximately .3 cases per year.
TABLE 8 compares the highest average levels of disinfection-
related contaminants found in drinking water with the estimated
reference dose criteria for chronic health effects other than
cancer. Unlike cancer, exposure to a contaminant at a dose below tl
estimated RfD are likely not associated with noncancer health
effects. All trihalomethane levels appear to be below estimated
thresholds for noncancer effects.
Measured trihalomethane exposures are comparable with those
in other areas where chlorinated surface water is supplied.
TABLE 9 presents information on chloroform levels in some other
cities.
Inorganics and Pesticides
CAG has developed potency values for arsenic, lindane, and
toxaphene. Arsenic is rated a human carcinogen (Group A),
while toxaphene is considered a probable carcinogen (Group B2).
Lindane is both B2 and C according to the CAG although the Office
of Water within EPA considers lindane in category C. All measured
concentrations for lindane and toxaphene are below the detection
limit and no risk calculations are provided for these pollutants.
5-17
-------
TABLE 6
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF
LIFETIME CANCER RISKS TO THE AVERAGE EXPOSED INDIVIDUAL
INTENDED FOR POLICY DEVELOPMENT
SYSTEM ESTIMATED RISK WEIGHT OF EVIDENCE
Saint Albans 1 x 10~4 B22
West Virginia Water Company 2 x 10-4 32
(Kanawha Valley District)
West Virginia Water Conpany 1 x 10~4 32
(Montgomery District)
Smaller Systems 1 x 10~4 - 2 x 10~4 b2
l"Ihe unit risk factors used in this analysis are based on conservative
assumptions that generally produce upper-bound estimates. Because of
limitations in data and methods in several areas of the analysis, such
as exposure calculations and pollutant selection, risk estimates were
calculated as aids to policy development, not as predictions of actual
cancer risks in the Kanawha Valley. Actual risks may be significantly
lcwer; in fact, they could be zero. Hie proper function of the estimates
is to help local officials select and evaluate issues and set priorities
for the topics examined.
2epa considered chloroform a Class B2, or probable human carcinogen.
EPA has not classified the other THMs for carcinogencity. This analysis
assumes similar unit risk factors for other THMs as chloroform.
Source: Regulatory Integration Division, 1987
EPA, Office of Policy Analysis
5-18
-------
TABLE 7
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES
OF ANNUAL EXCESS CANCER
INCIDENCE: TRIHALOMETHANES
IN DRINKING WATER
INTENDED FOR POLICY DEVELOPMENT
ESTIMATED ANNUAL EXCESS1
SYSTEM INCIDENCE (cases/year) WEIGHT OF EVIDENCE
Saint Albans
.02
B22
West Virginia Water Ccrtpany
.55
B2
West Virginia Water Company
.01
B2
(Montgomery District)
Smaller Systems3
.04 - .08
B2
TOTAL
.62 - .66
B2
Source: EPA, Office of Policy Analysis, Regulatory Integration Divison,
Noveirber 1986.
Ithe unit risk factors used in this analysis are based on conservative
assumptions that generally produce upper-bound estimates. Because of
limitations in data and methods in several areas of the analysis, such
as exposure calculations and pollutant selection, risk estimates were
calculated as aids to policy development, not as predictions of actual
cancer risks in the Kanawha Valley. Actual risks may be significantly
lower? in fact, they could be zero. The proper function of the esti-
mates is to help local officials select and evaluate issues and set
priorities for the topics examined.
2EPA considers chloroform a Class B2, or probable, human carcinogen.
EPA has not classified the other THM for carcinogenicity.
3nata on levels of THMs is not available. This analysis assumes that
the THM concentration for smaller systems could be from 42 to 77 ug/1.
5-19
-------
TABLE 8
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
PRELIMINARY RISK SCREENING RESULTS
COMPARISON OF TRIHALOMETHANE REFERENCE DOSE CRITERIA
TO HIGHEST ESTIMATED AVERAGE CONCENTRATIONS
TRIHALOMETHANE
HIGHEST CONCENTRATION (uq/1)
REFERENCE DOSE
CRITERIA (uq/1)
Chloroform
Bromoform
Dichlorobrornome thane
Chlorodibromomethane
72 (West Virginia Water Co.)
<1 (all monitored systems)
12 (Saint Albans)
1 (Saint Albans, West
Virginia Water Co.)
350
210
17.5
21
Source: Regulatory Integration Division, Office of Policy Analysis, 1987
5-20
-------
TABLE 9
KANAWHA VALLEY TOXICS SCREENING STUDY
KANAWHA VALLEY DRINKING WATER REPORT
AVERAGE CHLOROFORM LEVELS IN DRINKING WATER
FOR SELECTED CITIES
(population >500,000) 1
Average
Population
City
ug/i
1980
Boston, MA
3.7
562,994
Chicago, IL
14.5
3,005,072
Cleveland, OH
15.0
573,822
Columbus, OH
171.0
564,871
Dallas, TX
18.0
904,078
Detroit, MI
10.5
1,203,339
Houston, TX
123.0
1,595,138
Los Angeles, CA
32.0
2,966,850
Memphis, TN
2.5
646,335
Milwaukee, WI
8.9
636,212
New York, NY
22.0
7,071,639
Philadelphia, PA^
47.7
1,656,300
San Antonio, TX
0.1
785,880
San Diego, CA
43.5
875,538
San Francisco, CA
58.5
678,974
Washington, DC
47.0
638,333
1These cities were selected frcm a sample of 137 cities studied in the source
document.
^Reported chloroform concentrations are an average of the levels at Baxter,
Queen Lane, Belmont plants. Populations data as of July 1983. See U.S. EPA,
Final Draft Report of the Philadelphia Integrated Environmental Management
Project, June 1986.
Source: U.S. EPA, Office of Standards and Regulations, Chemical Control
Options Analysis for Chloroform, (Contract Nos. 68-02-3168,
68-01-6775, and 68-01-6715), December 1984.
5-21
-------
We did review the detection limit for these chemicals to determine
what upper-bound individual risk would exist at these levels. For
arsenic the detection limit within these systems is 1 ug/1. When
these detection limits are assumed exposure concentrations, this
limit will allow us to detect upper-bound individual lifetime
risks greater than 10~4. Concentrations of arsenic have been
detected higher than 1 ug/1. For toxaphene, the detection limit
is .2 ug/1 within these systems. At these concentrations, these
laboratory methods will detect risks greater than 10~6. The
detection limit for lindane is .005 ug/1. If we assume that this
is the exposure concentration, then the upper-bound risk would be
2 x 10"7.
The highest observed concentrations for selected metals and
pesticides in TABLE 5 are below their respective verified Reference
Dose for noncancer health effects.
IX. SUMMARY AND CONCLUSIONS
We examined data on monitored toxic pollutant concentra-
tions within the 14 public water supply systems regulated under
the Safe Drinking water Act within the Kanawha valley. The
pollutants reviewed were trihalomethanes, inorganics and
pesticides. We did not review potential exposure of toxics
within private wells and non-community water systems. The
State is not required to monitor regularly within these systems.
The larger systems, Saint Albans, and the West Virginia
Water Company, Kanawha valley and Montgomery Systems, serve a
population of approximately 211,000 people. For these systems,
we obtained monitoring data for trihalomethanes, and developed
an initial conservative assessment of potential cancer risks, as
well as an assessment of noncancer health effects for this pollu-
tant category. We extended this analysis to the remaining systems
assuming similar trihalomethane concentrations within these
systems. Limited monitoring data precluded developing a more
detailed health effects analysis of the other pollutants.
The specific findings and conclusions follow.
0 Fourteen public water supply systems serve a popula-
tion of approximately 240,000 people within the
Kanawha Valley. For most of these systems monitoring
data exists for 19 pollutants. Although we had a limited
set of priority pollutant scans for other pollutants,
we were confined in our exposure and health effects
analysis to these 19 pollutants (See TABLE 2 and
TABLE 3).
5-22
-------
° Our analysis of trihalomethane data for the three
major public water supply systems indicates that an
average individual could be exposed to an upper-bound,
incremental risk of 1 to 2 x 10~4 (a one to two out of
ten thousand) probability of contracting cancer over a
70-year exposure to these concentrations of trihalome-
thanes. (We assume that all trihalomethanes have the
same potency as chloroform, a B2 carcinogen.) We
assume that the remaining systems have the same levels
of trihalomethanes, and consequently risk, since many
of these systems draw water from surface water supplies
and chlorinate their drinking water also.
0 When we calculate incidence for the total population
served by the fourteen public water supply systems,
we estimate approximately six tenths of a cancer case
per year, assuming a lifetime exposure to these
concentrations of trihalomethanes. To facilitate
comparison of pollutant incidence within the analysis
of the Kanawha Valley Toxics Screening Study, we
calculated incidence posed by trihalomethanes in
drinking water within the air zones defined within
the air quality analysis: Belle, Charleston, Institute,
and Nitro. These systems are served primarily by the
West Virginia Water Company. The incidence from chloroform
in drinking water within these zones is approximately one
third of a case per year, (or one case every three years)
using upper-bound potency values, and assuming a 70-year
exposure from this concentration of trihalomethanes.
° We compared reference dose levels for trihalomethanes
to our observed average concentration level for
these pollutants. RfD levels were not exceeded for any
of the trihalomethanes. We also compared reference
dose levels for the highest observed concentration
for MCL inorganics and pesticides. For this limited
data base, we did not see any observed concentrations
exceeding references dose levels.
Datagaps
We obtained limited monitoring information for many of
the smaller systems and no information on private wells within
the study area. Our ability to determine potential exposure
for these drinking water systems is extremely limited.
In addition, our health effects information is limited. For
health assessments, our knowledge on health effects is continually
evolving. As we learn more on the potential health effects of
many of these pollutants, we should be able to determine, given
the availability of pollutant concentrations, which pollutant
concentrations may prove harmful.
5-23
-------
Additional monitoring would be useful in the following
areas:
Pollutant scans of potential contaminants for public
water systems in addition to those scans currently
conducted by the west Virginia Water Supply Systems,
may determine if the population is exposed to additional
pollutants not routinely monitored under the Safe
Drinking Water Act.
There are estimated 22,000 private well systems within
Kanawha County, one of three counties in the study areas.
Monitoring these supplies would determine if these people
are exposed to toxics entering groundwater systems.
5-24
-------
REFERENCES
Hodges, Jim. 1986. Personal Communication; Telephone conversation
concerning chlorination and source public water supply systems
in study area, November 1986.
McQueen, Charles. 1986. Personal Communication; Telephone conversa-
tion concerning number of private wells in Kanawha County,
January 1986.
National Toxicology Program. 1984. "NTP Technical Report on the
Toxicology and Carcinogenesis of Chlorodibromomethane," 1984.
NTP #83-065.
National Toxicology Program. 1984. "NTP Technical Report on the
Toxicology and Carcinogenesis of Bromodichloromethane," 1984.
NTP #86-321.
NUS Corporation. 1984. "Draft Site Inspection Report: South
Charleston Municipal Landfill," 1984.
NUS Corporation. 1984. "Draft Site Inspection Report: Heizer
Creek Dump," 1984.
US EPA. 1986. "Air Quality Criteria for Lead".(Vol. IV)
Prepared by ECAO-OED. U.S. - June 1986 EPA/600/83/028df.
5-25
-------
DRINKING WATER REPORT
Appendix A: Priority Pollutant Scan of West Virginia
Water Supply Company - July 1, 1985
(available upon request)
5-27
-------
Chapter Six
Surface Water Analysis
-------
KANAWHA VALLEY SURFACE WATER REPORT
I. INTRODUCTION AND PURPOSE
The Kanawha Valley Toxics Study has four stated objectives
which provided overall direction to these technical reports.
1. Using available data bases identify many of the chemicals
routinely released or present within various exposure
pathways.
2. Develop a sense of potential public health concerns these
toxic pollutants may pose in various pathways based on
health effects and exposure information.
3. For a select number of pollutants, provide an initial,
conservative assessment of potential cancer risk and other
potential non-cancer health risk from predicted or observed
concentrations within exposure pathways.
4. Identify data and information gaps and outline needs and
options for future study direction to enable a more
detailed investigation of health issues where warranted.
To attempt to meet these objectives we relied primarily on
available data. We reviewed information on pollutants detected in
industrial discharges, non-point source run-off, the water column
and fish tissue. To assess potential health effects of these
pollutants, we reviewed fish consumption as one exposure pathway
within the study area. Information on fish tissue concentrations
and water column concentrations provided limited data for this
preliminary assessment of potential health concerns. Where possible,
given limited tissue concentration data for select chemicals, we
estimated potential health effects if fish were consumed, although
actual ingestion rates are not known for the study area. Finally,
we outlined the many datagaps which prevent a full assessment of
health effects of toxic pollutants.
Within our analysis, we have focused on fish consumption to
assess the potential public health concerns toxics may pose.
Although we focused on this one exposure pathway we recognized
that other routes of exposure to toxic chemicals within the Kanawha
River may be important. We did not investigate waterfowl harvesting
and consumption, for instance, although this area represents an
important duck and goose hunting area for the State, swimming and
boating do occur, and therefore dermal contact is likely. Although
the State believes that water contact through recreational uses of
the river maybe a significant route of exposure to toxic chemicals
for a sizeable portion of the population, this issue is not considered
in this analysis. We could not estimate potential risk from these
activities due to the lack of information on exposure.
6-1
-------
Ingestion of water from the Kanawha River is another potential
route of exposure. Four of the 14 public water supply systems
identified in the drinking water analysis draw water from the
Kanawha River. The sources for these systems, however, are upriver
from the industrial facilities within the study area and are not
considered contaminated from toxic chemicals discharged by industry
into the Kanawha River (Personal Communication, Hodges, 1986.) The
river below Belle is not designated as a drinking water source.
Since this is an unlikely major route of exposure to toxics, we did
not consider the potential health effects from ingestion of drinking
water from the Kanawha River.
The consumption of contaminated fish tissue is also a route
of exposure for which exposure levels are very difficult to deter-
mine. Nevertheless, exposure levels are considered significant
because fishing has been observed along the Kanawha River within
the study area. We did not estimate exposure or consumption rates
within the study area in order to calculate risk. Instead, we
estimated consumption rates associated with various levels of
carcinogenic risk, and consumption rates which exceed Reference
Dose levels for noncancer health effects.
Although we believed that fish consumption may be the most
significant route of exposure of toxics from the Kanawha River,
consumption of fish from the lower Kanawha River has probably been
discouraged by a recent health advisory on fish consumption.
This health advisory applies to the Kanawha River below the Coal
River, downriver from Charleston, and extends to the Ohio River.
The State issued the advisory due to the detection of elevated
concentrations of dioxin within fish tissue. The advisory probably
has also discouraged fishing further upstream.
The West Virginia Department of Natural Resources conducted a
creel study, a study of the number and type of fish caught,
at Marmet and Winfield dams during 1986. The final report for this
project is expected to be completed in 1987 (Personal Communication,
Kain, 1986). While it is recognized that direct inferences can
only be made for the specific area studied, the results of this
data gathering effort will allow more accurate estimates of consumption
patterns on the Kanawha River as a whole.
II. DESCRIPTION OF THE GENERAL METHODOLOGY
We have divided this analysis into three categories of pol-
lutants: (a) Pesticides, Dioxin, and Polychlorinated Biphenyls
(PCBs); (b) Metals; and (c) Industrial Discharged Organics. These
categories are reviewed separately due to the different analyses
applied to the available data for each category of pollutants. In
many cases, we have extremely limited data from a few segments of
the Kanawha River. Consequently, our ability to draw conclusions
for the entire river within the study is limited.
6-2
-------
We did not review biological oxygen demand, total suspended
solids, oil and grease, coliform bacteria, pH, ammonia, phosphate,
and physical and biological factors such as temperature, turbidity
and bacterial concentration. While these conventional and non-
conventional pollutants have both potential health effects, through
bacterial contamination, and ecological impacts, we limited the
scope of the analysis to the effects of a limited set of toxics on
human health and to provide a general sense of the ecological
impacts of some toxic pollutants.
For each pollutant category, we attempted to identify the pollu-
tants which exist within the water column or fish tissue which may
pose potential health effects if consumed through fish consumption.
We then describe the potential sources, both point and non-point,
for these pollutants.
For the review of potential health effects we reviewed fish
tissue data if available, and estimated various levels of consumption
needed to exceed a designated cancer risk level or Reference Dose
(RfD) level for noncancer health effects. Where fish fillet data
were not available, we compared observed ambient water concentrations
to EPA derived water quality criteria for the protection of human
health. These values, based on pollutant specific bioconcentration
factors and a national average individual consumption rate, attempt
to provide water quality concentrations at which a pollutant may
bioaccumulate in fish to various tissue concentration levels, and
consequently to various risk levels. These water quality criteria
are set at several water concentrations which represent these
various levels of carcinogenic risk.
Finally, in order to determine if aquatic organisms are poten-
tially at risk from high concentrations of toxic chemicals in the
water column, we compared water concentrations with an additional
set of water quality criteria which establish ambient concentrations
above which aquatic organisms may be harmed. These are the water
quality criteria for the protection of aquatic life.
III. HUMAN AND AQUATIC HEALTH ASSESSMENTS
Cancer Assessment
This study employed a risk assessment screening methodology to
evaluate and compare, in a very limited fashion, the potential
health risks from exposures to a limited set of toxic pollutants.
Risk to an individual is defined as the increased probability that
an individual exposed to one or more chemicals will experience a
particular adverse health effect during his or her lifetime. For
the surface water report, we present our analyses in terms of the
estimated incremental lifetime risk to an individual given various
consumption rates and concentrations of toxic chemicals in water.
We did not calculate incidence, the projected risk to the entire
population, due to the lack of actual exposure information on
fish consumption.
The risk screening methodology involved both a qualitative
and quantitative assessment of the potential carcinogenicity of
the selected pollutants. As a screening study, this analysis
6-3
-------
employed both types of assessments. EPA's Carcinogen Assessment
Group (CAG) reviews the evidence of carcinogenicity for selected
pollutants, and classifies pollutants as human carcinogens (Group
A), probable human carcinogens (Group B), oossible human carcinogens
(Group C), not classified as carcinogens due to inadequate evidence
(Group D) and not carcinogenic to humans (Group E). These classifi-
cations accompany all quantitative risk estimates within this
study.
For those chemicals in Groups A, B, and C, CAG provides
quantitative, upper-bound estimates of carcinogenic unit risk
factors. The traditional approach to carcinogen risk assessment at
EPA has been to take the most conservative approach in developing
potency estimates. In doing so, EPA develops estimates that are
likely to overestimate the true potency of a chemical. We feel
that such an approach is appropriate, particularly for public health
protection and setting priorities, in order not to underestimate
potential human health impacts.
A unit risk factor allows the calculation of the estimated
individual risk posed by exposure to a chemical given certain
exposure assumptions. To calculate individual risk from the
consumption of fish tissue, a pollutant's unit risk factor is
multiplied by the fish tissue concentration of the pollutant and
the consumption rate of fish by an individual.
Individual = Unit Risk x Concentration x Consumption x Body Weight
Lifetime Factor Rate Assumption
Risk (mg/kg-day) (mg/kg) (kg/day) (kg)
Concentration refers to the concentration of the pollutant found
in the fish tissue. The average person is assumed to weigh 70 Kg.
Lacking consumption rates for fish by recreational fisherman
within the study area, we could not calculate the individual lifetime
risks for those categories of pollutants for which we have fish
tissue data. To provide some estimate of potential health effects
from fish consumption, we specified several levels of risk and
calculated a consumption rate associated with the specified risk
level.
-------
SAMPLE CALCULATION
CONSUMPTION RATES ASSOCIATED WITH SPECIFIED
LEVELS OF RISK FROM CHLORDANE BIOACCUMULATED
WITHIN FISH TISSUE
Average
Individual
Lifetime Risk
Unit X Concentration X Consumption X Average
Risk Rate Body Weight
Factor Assumption
* To solve for consumption rate we can rearrange the equation as follows:
Consumption =
Rate
Average Individual Lifetime Risk
Unit Risk Factor X Concentration X Body Weight Assumption
* Assume the following data exists for:
Individual Lifetime
Risk
For this example, set average individual risk
at 1 x 10-5
Unit Risk Factor for
Ingestion of Chlordane
= 1.3 (mg/kg-day)-l
Concentration
For this example, assume 0.003 mg/kg
of chlordane in fish tissue.
Average Body Weight
Assumption
= 70 kg
Consumption
Rate
1 x 10-5
1.29 (mg/kg-d)-l x 0.003 mg/kg x 1/70 kg
0.18 kg/day
We can convert this consumption rate easily to pounds per week by
multiplying this derived consumption rate by 2.2 lbs/kg and 7 days/week.
For this example, 0.18 kg/day equals 2.8 lbs/week.
6-5
-------
This calculated consumption rate, based on upper bound unit
risk factors, indicates that a consumption rate of 2.3 pounds per
week for seventy years potentially creates a one in 100,000
increased chance of developing cancer over a lifetime.
We can modify this equation to calculate the consumption rate
for a specified level of risk when the cancer health effects from
more than one chemical are considered. It should be noted here
that we are assuming that cancer risks are additive. Analyzing the
effects of a chemical in combination with other chemicals is an
extremely complex task, and such analysis has not been done for
most pollutants. In the absence of this information, the additivity
assumption seems most appropriate.
For two pollutants X, and Y, the equation becomes:
Consumption Rate ¦
I ' Individual Lifetime Risk
70 X / Unit Risk Factor x Concentrations ./Unit Risk Factor x Concentration \
Llfor Pollutant X of Pollutant X' ; for Pollutant Y of PollutantX!
For the industrial organic chemicals however, we did not have
fish tissue data to calculate consumption rates. For those organic
chemicals, we compared observed water concentrations to EPA derived
water quality criteria to protect human health from ingestion of
contaminated fish tissue. For suspected or proven carcinogens,
EPA provides a set of water concentrations associated with a range
of incremental cancer risk resulting from consumption of fish
tissue. it is assumed that toxics at various water concentrations
will bioconcentrate within fish tissue and can expose consuming
individuals to a health risk. For this report, we compared observed
water concentration to estimated water concentrations associated
with 1 x 10-7, 1 x 10-6 and 1 x 10-5 average individual
incremental cancer risks. These estimated water concentrations
are calculated based on lifetime ingestion of fish and assume that
a 70-Kilogram adult consumes 6.5 grams of fish daily for a seventy
year lifetime.
Noncancer Health Effects
No currently accepted techniques exist for estimating the
probability of incidence of noncancer effects. Therefore, in
evaluating the potential noncancer health risks within this study,
we relied on benchmark values or Reference Dose (RfDs), expressed
in units of mg/kg/day, which represent doses levels below which
adverse health effects are not likely to occur in most people.
Unlike cancer effects, noncancer health effects are assumed to be
threshold events. This threshold assumption means that noncancer
health effects are assumed more likely to occur above these exposure
concentrations. Exposures that are less than the RfD are not
likely associated with noncancer health effects and therefore
less likely to be of regulatory concern.
-------
To calculate an RfD, EPA scientists collect the available
animal and human data and note the various dose levels (in milli-
grams per day) at which different health effects are seen. The
scientists then compare the dose with effect information and identify
the NOEL (the No Observed Effect Level). The NOEL represents the
highest dose tested that did not produce observable results. The
scientists also try to define the LOEL (Lowest Observed Effect
Level), which is the lowest dose tested at which some type of effect
occurs. They also try to define the FEL (Frank Effect Level),
which is the dose that involves more serious health problems.
There will be many NOELs and LOELs for each chemical. These
levels depend on the doses selected and the health effects tested
for by the researchers. Research is very expensive and each
experiment cannot be exhaustive. Since EPA is dependent on available
research, there is some uncertainty in the definition of NOELs and
LOELs.
To calculate the RfD, EPA scientists select the most reliable
NOEL and divide it by safety factors. The selection of appropriate
safety factors is based on the nature of the study from which the
NOEL was derived. Most safety factors are multiples of ten, with
each one representing an extra degree of uncertainty to account for
extrapolation from animal data to the average human, from the
average human to the most sensitive subgroup, from subacute effects
to chronic effects, and the LOEL to the NOEL. In general, the
safety factor is larger than the expected differences in the exposure
levels producing these effects, and is therefore a conservative
estimate of the actual threshold.
For our analysis, we calculated the number of pounds per week
which, if consumed, would exceed the Reference Dose for the
pollutant. We only reviewed noncancer health effects of pollutants
for which EPA has established an RfD.
Water Quality Criteria Exceedances
We derive a sense of the potential for harmful chronic
ecological effects from pollutants by simply comparing ambient
water concentrations to EPA derived ambient water quality criteria
for the protection of aquatic life. These criteria are derived
from aquatic toxicology data for chronic effects on both plants and
animals. EPA derives, in most cases, the highest four-day average
concentration which, if exceeded, may result in the loss of sensitive
species or the imoairment of reproductive functions. In many
cases, these four day concentrations are made a function of such
water quality characteristics as pH or hardness if a correlation
between toxicity and the characteristic can be established.
For this analysis, we compare the percentage of observations that
exceed the water quality chronic criteria for the protection of
aquatic life. A high percentage of water quality criteria
exceedances over an extended period of time would indicate that
chronic toxicological impacts are likely and sensitive aquatic
species may be eliminated.
6-7
-------
IV. DESCRIPTION OF STUDY AREA
The Kanawha River begins at the confluence of the New and
Gauley Rivers and flows northwestward toward the Ohio River. The
Kanawha River within the study is bounded upstream by Alloy
and downstream by Winfield Dam. The distance between Alloy and
Winfield is approximately 60 miles.
The Kanawha River drains an area characterized by rugged hills
and mountains with deep valleys cut by rivers and streams. Three
major tributaries, the Coal, the Pocatalico and the Elk Rivers,
drain into the river within the study area. Three dams have
been built within the study area, and help maintain the Kanawha
River at a minimum width of 300 feet and a depth of nine feet.
Land use adjacent to the river is primarily urban and industrial
with other land uses including mining, recreation, transportation
and utility right-of-ways.
For this analysis, we identified 22 major industries which
discharge into the Kanawha River. The Kanawha Valley, similar to
other metropolitan areas, is likely to have non-point source run-off
characteritic of residential and commercial areas. The large
concentration of industrial facilities may also provide non-point
source loadings from run-off from industrial lots. Hazardous waste
sites and sediments may also contribute pollutant loadings deposited
from past industrial activity.
V. PESTICIDES, DIOXIN, AND POLYCHLORINATED 3IPHENYLS (PCBs)
Pesticides, dioxin, and PCBs are the first of three categories
of pollutants reviewed within the surface water analysis of the
Kanawha Valley Toxics Study. In this section we briefly outline
the methods used to estimate the potential health effects of
consumption of a limited set of pesticides, PCBs, and dioxin
bioconcentrated within edible fish tissue.
It is important to reemphasize that we do not currently know
recreational fish consumption patterns within the study area.
Indeed, consumption may be zero. The West Virginia Department
of Health has issued a health advisory for the lower Kanawha
River due to the recently detected elevated concentrations of
dioxin within fish in the Nitro area. Since exposure is not
knownL and may be zero, we have estimated the potential health
effects of several hypothetical consumption or exposure rates.
Despite the advisory, fish consumption from the Kanawha River
may be occurring.
To estimate potential exposure concentrations from contam-
inated fish tissue, we calculated potential fish tissue concentra-
tions for a small segment of the river. We reviewed the only
available data for fillet concentrations of a limited set of
pesticides, PCBs, and dioxin sampled in fish tissue in the Nitro area
6-8
-------
on the lower Kanawha River. We calculated mean concentrations
for selected pollutants from this limited dataset. Although these
observed concentrations may be adequate to determine fillet fish
tissue concentrations within the Nitro area, the limited sampling
territory precludes drawing inferences for the entire Kanawha
River within the study area. In fact, a 1986 survey by WV DNR
and Region III EPA suggests that dioxin levels at different sites
on the Kanawha River may vary with location (Personal Communicaton,
Smith 1986).
We reviewed toxicity data for our selected pollutants. We
calculated weekly consumption rates needed to exceed a defined level
of risk, in addition, we examined noncancer health effects by
calculating weekly consumption rates needed to exceed available
Reference Dose (RfD) levels.
Potential Pollutants
The Department of Natural Resources (DNR) has monitored for
pesticides and PCBs in whole fish tissue over the years. Recently,
DNR has monitored for dioxin in the Nitro area. Pesticide run-off
from agricultural and residential areas and accidental PCB spills
from industry over the years have led to probable contamination of
fish tissue. The source of the dioxin has not yet been determined.
Since these pollutants readily bioaccumulate within fish tissue,
West Virginia has sampled fish throughout the Kanawha River over
the past years to determine if elevated concentrations of these
pollutants occur.
The Department of Natural Resources has sampled for the
following pesticides and organics within whole fish tissue over the
years (DNR, 1986):
Aldrin
Dieldrin
Endr in
Hexachlorobenzene
Lindane
Methoxychlor
Toxaphene
Alpha-BHC
Beta-BHC
Delta-BHC
Arochlor 1254
Arochlor 1260
Trans-Chlordane
Cis-Chlordane
0,P'- DDD
P,P'- DDD
0,P'- DDE
P,P'- DDE
0,P*- DDT
P,P'- DDT
Trans - Nonachlor
Cis - Nonachlor
Endosulfan I
Endosulfan II
Heptachlor
Heptachlor Epoxide
Toxaphene is the only pollutant from this list which
has not been detected within whole fish tissue.
6-9
-------
Observed Fish Tissue Concentrations
In September 1985, the West Virginia DNR sampled fish fillet
tissue at Nitro, West Virginia. Nitro is the last of the industrial
areas on the Kanawha River within the study area. DNR sampled
for seven fish species including, but not limited to, those commonly
caught and harvested in the area. The fish were collected as a
follow-up to the National Dioxin Study and were intended to help
determine the degree of dioxin contamination in various species.
These species were Smallmouth Buffalo, Channel Catfish, Sauger,
Smallmouth Bass, Spotted Bass, Largemouth Bass, and Freshwater
Drum. It should be noted that we limited our analysis to fillet
rather than whole fish samples due to the difficulties and un-
certainties associated with estimating concentrations of pollutants
in fillets from whole fish samples.
Fish between 10 and 14 inches, which approximate the average
size caught by fishermen, were targeted for analyses. DNR analyzed
composite fillet samples containing between three and five fish
of an individual species for chlordane, PCB, dioxin, and dieldrin.
TABLE 1 presents a mean fillet concentration, calculated from
reported composite concentrations, for each pollutant by fish
species.
Mean concentrations of pollutants are the highest for the
Smallmouth Buffalo, Channel Catfish, Freshwater Drum, and Sauger
The three bass species have lower, often nondetected concentrations
for these pollutants. Differences in migratory habits, positions
within the food chain, and lipid concentrations which accumulate
these pollutants within fish tissue may account for these differences.
Potential Human Health Risks
Contamination by pesticides, PCBs, and dioxin of fish
tissue may present a health hazard if the bioaccumulated toxic
chemical is present in large enough concentrations and consumed by
individuals. Concentrations of pesticides, PCBs, and dioxin have
been detected in fish tissue of several species. Such concentrations
may present a health concern for the consuming public.
Due to the lack of exposure information for fish consumption,
we instead calculated weekly consumption rates which could expose a
consuming individual to a designated level of carcinogenic risk.
Also, we calculated a weekly consumption rate which will provide
toxic pollutant concentrations exceeding noncancer health effect
levels which are considered safe for pollutants with established
6-10
-------
TABLE 1
KANAWHA VALLEY SURFACE WATER REPORT
SEPTEMBER, 1985 OBSERVED PESTICIDE, DIOXIN, AND PCB CONCENTRATIONS1
IN FISH FILLET AT NITRO, WEST VIRGINIA
(mg/kg)
Fish Species
Number of
Composite .
Samples Chlordane PCBs^ Dieldrin Dioxin
Smallmouth Buffalo
0.199
1.127 0.015 0.000032 T
Channel Catfish
Sauger
Smallmouth Bass
Spotted Bass
Largemouth Bass
Freshwater Drum
2
2
1
1
3
2
0.104
0.655 0.005 0.000023
0.032 0.264
ND
ND
ND
ND
0.003 0.067
ND
ND
ND
ND
0.000006
0.000007
0.000013
0.000004
0.03
0.13
0.005 0.000008
Concentrations shown in this table represent the mean of composite
samples. Each composite contains between 3 and 6 fish specimens.
Specimens collected by DNR, West Virginia.
i
Chlordane concentrations are the sum of Cis-and Trans-chlordane and
Cis-nonachlor concentrations in fillets.
PCB concentration is the concentration of Arochlor 1260.
4r\i
Dioxin values were measured in picograms/gram and have been converted
to mg/kg for this table.
TThese mean values were calculated from 5 composite samples,
6-11
-------
Reference Doses. In both cases, we set a specified level of cancer
risk or a Reference Dose (RfD) level for noncancer health effects
and calculated the weekly consumption rate needed to exceed these
levels .
For carcinogenic risk, we calculated a series of consumption rates
which correspond to exposure levels associated with estimated
lifetime average individual risks of 1 X 10-^, 1 X 10~4, and 1 X 10~3.
These risk levels were chosen for policy development purposes only.
Chronic Toxicity Data: TABLE 2 presents the carcinogenic unit
risk factors and strength of evidence classifications for PCBs,
chlordane, dioxin, and dieldrin. The unit risk factors were
developed by EPA's Carcinogen Assessment Group, CAG, and represent
plausible upper-bound values.
For noncancer health effects, only chlordane has a Reference
Dose to date. The RfD is 0.00005 mg/kg-day. RfDs for PCBs, dieldrin
and dioxin have not yet been established by EPA's Reference Dose
Workgroup. RfDs are benchmark concentrations to determine the
potential for noncancer health effects. Exposures below
the RfD are not likely to be associated with noncancer health
effects.
Exposure Concentrations
For fish tissue exposure concentrations, we assumed
concentrations observed within the 1985 fish fillet data from the
Nitro area presented in TABLE 1 . As we have noted, these samples
are from a limited geographic area with in the study area and are
from a single point in time. They may not be representative of the
entire study reach or of contaminant levels (i.e., exposure levels)
over time. DNR has detected pesticides and PCBs pollutants in
whole fish tissue throughout the Kanawha River in the study area.
Consumption Rates
We calculated individual consumption rates for various fish
species which would exceed defined levels of lifetime incremental
risk. These consumption rates are based on the upper-bound potency
values in TABLE 2 and the assumed exposure concentrations from
TABLE 1. TABLE 3 presents the results of this analysis.
According to TABLE 3, consumption rates could range from one
tenth of a pound to one pound per week, depending on fish species,
for an individual to be exposed to an upper-bound lifetime risk of
1 X 10~3 (a one out of a thousand chance). Consumption rates are
lower for lower risk levels.
Dioxin and PCBs contribute more than 90 percent of the risk
for each species. For Smallmouth Bass and Spotted Bass, dioxin was
the only pollutant detected. At the observed concentration of
dioxin for these two species, an individual would need to consume
one half pound of Spotted Bass or one oound of Smallmouth Bass to
be exposed to an upper bound risk of one in a thousand.
When interpreting TABLE 3, the reader should keep in mind that
the traditional approach to carcinogen risk assessment at EPA has
been to take the most conservative approach in developing potency
estimates. In doing so, EPA develops estimates that are highly
6-12
-------
TABLE 2
KANAWHA VALLEY SURFACE WATER REPORT
PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND CANCER UNIT RISK VALUES:
POLLUTANTS IN FISH TISSUE
Unit Risk Factor for Inhalation Grouping Based
Pollutant (ma/kq-dav)"1 on EPA Criteria
PCB 4.34 B2
Chlordane 1.33 B2
Dieldrin 20.02 B2
Dioxin I.60E+05 B2
' Weight-of-Evidence rating derived byCAG, based on EPA's classification system:
A - proven human carcinogen; B - probable human carcinogen (B1 indicates
limited evidence from human studies, B2 indicates sufficient evidence from
animal studies but inadequate evidence from human studies); C = possible
human carcinogen; D ¦ not classifiable; and E « no evidence of carcinogenicity.
Source: Carcinogen Assessment Group, EPA, 1986.
6-13
-------
TABLE 3
KANAWHA VALLEY SURFACE WATER REPORT
PRELIMINARY RISK SCREENING RESULTS
WEEKLY RATES OF FISH CONSUMPTION
ASSOCIATED WITH VARIOUS RISK LEVELS
INTENDED FOR POLICY DEVELOPMENT
UPPER-BOUND ESTIMATES
CONSUMPTION RATES PRODUCING
% CONTRIBUTION OF
POLLUTANTS TO RISK
SPECIFIED RISK LEVEL 1
FISH SPECIES 2
10^#
(lbs/week)
10 #
-5
10 *
PCB*
CHLORDANE"
DIELDRIN°
DIOXIN
Small Mouth Buffalo
0.103
0.0103
0.00103 |
47%
2%
3%
48%
Channel Catfish
0.160
0.0160
0.00160 |
42%
2%
3%
53%
Sauger
0.496
0.0496
0.00496 |
53%
2%
0%
46%
Large Mouth Bass
1 .025
0.1025
0.00103 |
41%
1%
0%
57%
Freshwater Drum
0.509
0.0509
0.00509 j
27%
1%
9%
63%
Smallmouth Bass
0.960
0.0960
0.00960 |
0%
0%
0%
100%
Spotted Bass
0.520
0.0520
0.00520 |
0%
0%
0%
100%
THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT GENERALLY
PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS, CONSUMPTION RATES
ASSOCIATED WITH VARIOUS RISK LEVELS WERE CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT AS ACTUAL
CANCER-CAUSING CONSUMPTION RATES IN THE KANAWHA VALLEY. ACTUAL RISKS FROM THESE RATES
MAY BE SIGNIFICANTLY LOWER; IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES IS TO
HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
Fish samples obtained bytheWest Virginia Department of Natural Resources In September, 1985
at Nitro, West Virginia.
~Risk levels are arbitrarily chosen for demonstration purposes only.
'% Concentrations were developed from the concentration values In Table 1.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA.
-------
likely to overestimate the true potency of a chemical. We feel
that such an approach is appropriate, particularly for public health
protection and setting priorities, in order not to underestimate
potential human health impacts. We also calculated the weekly
consumption rate needed to exceed the chlordane RfD. Of the pollut-
ants studied in this section of the analysis, there is a Reference
Dose only for chlordane. According to TABLE 4, the weekly consumption
rates for Smallmouth Buffalo, Channel Catfish, Sauger, and Freshwater
Drum would range from one third of a pound to almost 2 pounds.
Since chlordane was detected in limited amounts, or not at all in
the bass species, calculated consumption rates are much higher
for these species (18 lbs/week for Large Mouth Bass).
It is stressed that these results only apply to a limited
segment of the river and are based on a one day sampling event.
We cannot say that these findings apply to the whole Kanawha River
within the study area.
VI. METALS AND INORGANICS
Metals and inorganics are the second of three categories of
pollutants reviewed within the surface water analysis of the
Kanawha Valley Toxics Study. In this section we briefly outline
the methods used to develop a sense of the potential health and
ecological effects posed by metals and inorganics.
Like the pesticides analysis, it is important to emphasize
that we do not currently know recreational fish consumption
patterns within the study area. Again, consumption may be zero.
Since exposure is not known, we have estimated the potential
health effects of several hypothetical consumption rates for cancer
and noncancer health effects.
We have calculated edible fish tissue mean concentrations for
a limited set of metals within fillet tissue at the Nitro area.
This is the same dataset we used for the pesticide, PCB and dioxin
analysis. Although these observed concentrations may be adequate
to determine fillet fish tissue concentrations within this area,
the limited sampling territory precludes drawing inferences for the
entire study area.
To determine potential health effects for these bioconcentrated
metals we reviewed health effects information for a small set of
inorganics. From these data we estimated weekly consumption rates
which may pose potential chronic health effects.
For potential harmful effects on aquatic organisms we simply
compared observed ambient water concentrations to EPA chronic
criteria values for the protection of aquatic life to determine if
such criteria were exceeded during the sampling period.
Pollutant Sources
Metals and inorganics are present in point source industrial
discharges, urban non-point source run-off, and possibly in
hazardous waste site run-off and contaminated groundwater discharge
to the surface water.
6-15
-------
TABLE 4
KANAWHA VALLEY SURFACE WATER REPORT
INTENDED FOR POLICY DEVELOPMENT
LEVELS OF WEEKLY FISH CONSUMPTION
THAT COULD EXCEED THE REFERENCE DOSE
FOR NONCANCER HEALTH EFFECTS FOR CHLORDANE*
Consumption (lbs/week)
0.3
0.6
2.0
1.8
Not Detected
Not Detected
18.0
TThe RfD for Chlordane is 0.00005 (mg/kg-day), based on Yonemura et al.
(1983) showing liver necrosis effects.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA.
Fish Species
Small Mouth Buffalo
Channel Catfish
Sauger
Freshwater Drum
Small Mouth Bass
Spotted Bass
Large Mouth Bass
-------
To determine industrial point source contribution to the
Kanawha River within the study area, we reviewed NPDES permit
applications, discharge monitoring reports, and priority pollutant
scans of the major industrial facilities for 1983, 1984, and 1985.
The list of metals detected in the discharge of the facilities
is presented in TABLE 5.
Non-point source run-off is an additional source of metals and
inorganics to the river. Unfortunately, we did not have site-
specific metal loadings from non-point source run-off. Consequently,
we developed non-point source estimates from the National Urban
Storm water Run-off Program (NURP, 1982). This program was conducted
from 1980 to 1982, with run-off samples collected from predominately
residential and commercial areas throughout the country. Although
these estimates are not site specific, and may be off by an order of
magnitude, they may be sufficient to determine the magnitude of
non-point source contributions for this screening exercise.
Technical Appendix A describes our technique used to estimate
non-point source loading.
Appendix A also outlines our technique for estimating point
source loadings. Our point source loadings are based on 1984
information. We collected discharge information for facilities
from priority pollutant scans, NPDES permits, and DMRs. For each
pollutant we totaled the loadings based on flow and concentration
information obtained from the database.
We were unable to quantify run-off from industrial lots nor
hazardous waste sites. Our analysis of generic scenerios of
hazardous waste site contamination of groundwater discharge (See
Kanawha Valley Toxic Screening Study "Hazardous Waste Site Analysis")
indicates that loadings from groundwater are probably minimal.
TABLE 6 presents loading estimates for selected metals from
non-point and point sources. For arsenic and nickel, there appear
to be equal loading rates to the river from both point and non-point
sources. Non-point source contributions for total chromium, copper,
lead, antimony, cyanide, and zinc appear to outweigh the contribution
from point sources. Cadmium is the only metal where point source
loadings are larger than a non-point source loading.
Ambient Water Concentrations
For water quality analysis, we divided the river into five
segments. These segments correspond to the Valley Zones we
established for the ambient air quality analysis and correspond to
industrial clusters in the Kanawha Valley. Segment 1 is the river
stretch from Alloy to Charleston (90-57 KRM). DNR maintains a river
monitoring station at Kanawha River Mile 74. We used data from
this station to develop the average water concentration for this
segment. Segment 2 is the Charleston area between River Mile 57
and 50. We obtained ambient water monitoring data from a sampling
station at Kanawha River Mile 52 to represent this segment. The
Institute area is Segment 3 stretching from Kanawha River Mile 50
6-17
-------
TABLE 5
KANAWHA VALLEY SURFACE WATER REPORT
f
MAJOR FACILITIES DISCHARGING METALS
MAP
ALUM-
ANTI-
ARSENIC
BARIUM
CAD-
CHROM-
COPPER
LEAD
MANG-
MER-
NICKEL:SELEN-:
2 IMC
KEY
FACILITY
INUM
MONY
MIUM
IUM
ANESE
CURY
: IUM :
i :
1
Allied Chemicals
v ; :
V
2
Appalachian Power
(Amos Plant)
V
V
V
V
V
V
V
-/
-/ ! V i
V
3
Appalachian Power
'
* J
(Kanawha River Plant)
V
V
¦/
V
V
V : :
V
4
Autex Fibers (plant closed)
\ J
5
Chemical LeamanTank Lines
V
J J
6
Costal Tank Lines
V
; ;
7
Occidental
¦/
V
¦/
V
V
-/ i :
V
8
Oupont
V
V
-/
V • :
V
9
Elkem Metals
V
V
¦/
V
V
¦/
V : :
V
10
FMC, Nitro
V
V
; ;
11
FMC, South Charleston
V
V
¦S
V
V
V
12
CST/Fike Chemicals (Artel)
V
V
¦/
V
V
-/
V
V
V
V : :
13
Kincaid Enterprises
: ;
14
Mason Dixon Tank Lines
• •
15
Monsanto
V
V
V
V
¦/
V
V
• •
16
S. Charleston Sewa9e
• •
Treatment Plant
¦/
i •
17
Rhone Poulenc
V
V
V
V
: : V
18
Union Carbide, Technical Center
V
V
; •
19
Union Carbide, S. Charleston
V
¦/
V
V
v ; v i
V
20
Hatfield-Henson
V
¦/
V
V : :
V
P.B. and S. Chemical Company
V
; ¦
Smith Fastner Co., Inc.
V
'¦
¦/
V
: ¦: v
'Chromium present In the waste stream Is mainly Chromlum-VI.
SOURCE: Discharge Monitoring Reports, Permit Applications and priority pollutant scans.
-------
|^-o»tra.
'• M,t!
ik'J
>%*£.
•*&&& t» v
• iSoif />T'^~^..
Ov
I
H-4
v©
r.tf/V"
£ ^5'M* ftSrif Al
rs>.w
¦u-
""a 1 ^-.,^^^t^#*'5'
¦«]/ \ •
y^jy
'. X_ J 4,. ,)* r>(20)ni^^
fctfreiW
- . &v-i _
i5i ; %'
•#fw^v -s^-'/ r
vM^M
Tft
Map 1
Industrial Facilities t
on the Kanawha River
0,-^jr
5?ii
!!*«<=
?v^
5">-6l*/
-------
TABLE 6
KANAWHA VALLEY SURFACE WATER REPORT
PRELIMINARY RISK SCREENING RESULTS
NON-POINT AND POINT SOURCE LOADINGS
FOR METALS AND INORGANICS
INTENDED FOR POLICY DEVELOPMENT
Estimated
Estimated
Non-Point Source
Point Source
Ratio of
Load'
Load ^
Non-Point
Metal
(Mrtrlc Tmt/Yr)
(Metric Tons/YrV
To Point
Antimony
6.3
t
0.4
21 : 1
Arsenic
1.0
1.4
1 : 1.4
Cadmium
0.4
33.0
1 :83
Chromium - Total
2.0
0.4
5: 1
Copper
11.7
1.7
7: 1
Cyanide
11.3
0.9
13: 1
Lead
42.4
1.7
25: 1
Mercury
0.6
<0.1
—
Nickel
3.7
3.5
1:1
Selenium
25.6
0.0
—
Silver
0.3
<0.1
—
Zinc
59.6
1.0
60: 1
'see Appendix 1 for derivation of estimates.
2Estimated point source loadings are estimated from NPDES permit applications,
Discharge Monitoring Reports, and priority pollutant scans. See Appendix 1
for estimation techniques.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA.
6 20
-------
to 45 with a monitoring station at KRM 46. Segment 4 corresponds
to the St. Albans to Nitro area, (Kanawha Mile 45 to 42). Segment
4 does not have a monitoring station for metals. Finally, Segment
5 corresponds to the Nitro and Winfield areas, KRM 42 to 32. We
used ambient water monitoring data from the monitoring station at
Kanawha River Mile 31 for this segment.
TABLE 7 presents mean concentration values for these segments.
We calculated these mean concentration values from monitored
concentrations from these designated monitoring stations within
each zone. All monitoring data for metals were obtained from STORET.
For many observations, the concentrations were below the detec-
tion limit. In cases where the observation was below the detection
limit, we averaged that value at the detection limit concentration.
Consequently, these mean values should be considered the maximum
possible average based on the available data.
Fish Tissue Concentrations
In September 1985, the Department of Natural Resources sampled
fish fillet tissue at Nitro, West Virginia. This is the same data-
base used for the pesticides, dioxin and PCBs analysis. These
observed concentrations are presented in TABLE 8 as the range
observed in composite samples.
Bioconcentrated metal concentrations appear relatively
comparable across these six species of fish with total chromium and
arsenic concentrations differing among these species by the largest
factors. Other mean metal concentrations differ by no more than a
factor of two across these species.
Although this dataset may be appropriate for estimating tissue
concentrations in the lower Kanawha River, the limited sampling
area precludes drawing inferences for the entire study area.
Potential Human Health Risks
For these metals and inorganics, inorganic arsenic is the
only metal which has received a unit risk factor for carcinogenicity.
Although inorganic arsenic is a proven carcinogen (weight-of -
evidence classification is A) , recent laboratoy analyses suggests
that arsenic may exist in organic form within fish tissue, a far-
less toxic form than inorganic arsenic. For this reason, we do not
estimate the likelihood of carcinogenic risk from arsenic in fish
tissue.
6-21
-------
TABLE 7
KANAWHA VALLEY SURFACE WATER REPORT
ESTIMATED AVERAGE AMBIENT WATER CONCENTRATIONS*FOR
METALS BY RIVER SEGMENTS FOR 1983 - 1985
INTENDED FOR POLICY DEVELOPMENT
CONCENTRATION (ng/l)BY RIVER SEGMENT (# OF OBSERVATIONS)
Metal 1 2 3 4 5 Range
Antimony
<28.0
(3)
<30.0
(4)
<40.0
(3)
NM
<22.5
(2)
<22.5-<40.0
Arsenic
3.0
(6)
1.3
(8)
1.6
(6)
NM
1.3
(4)
1.3-3.0
Cadmium
2.0
(21)
1.3
(8)
1.3
(6)
NM
1.0
(17)
1.0-2.0
Chromium - VI
NM
—
1.3
(6)
1.3
(6)
NM
<1.0
(14)
<1.0- 1.3
Chromium - Total
11.3
(3)
8.5
(4)
8
(3)
NM
10.0
(2)
8.0- 11.3
Copper
2.3
(6)
3.0
(8)
2.7
(7)
NM
3.7
(6)
2.3-3.8
Cyanide
1.0
(6)
3.0
(8)
2.0
(6)
NM
1.0
(6)
1.0-3.0
Lead
13.5
(21)
10.0
(8)
24.5*
(11)
NM
11.2
(17)
10.0-24.5
Mercury
0.1*
(5)
0.2*
(6)
0.1
(3)
NM
0.2
(2)
<0.1 -0.2
Nickel
10.0
(6)
<10.0
(4)
<10.0
(3)
NM
<10.0
(2)
<10.0
Selenium
1.0
(2)
1.1
(4)
<0.8
(3)
NM
1.0
(2)
<0.8- 1.1
Silver
<2.0
(3)
<2.0
(4)
<2.0
(3)
NM
<2.0
(2)
<2.0
Zinc
10.2
(6)
7.8
(4)
8.3
(3)
NM
10.2
(5)
7.8- 10.2
'Water concentrations are the average of monitored values at STORET
river stations within each river segment. All average concentrations
are averages of detected concentrations and the detection limit, if
the monitoring did not detect the inorganic. If all concentrations are
less than the detection limit, average concentration value appears
with a less than sign (<). Consequently, these values should be con-
sidered the maximum possible average based on the available data.
NM = no monitoring station within the river segment.
'Averages are based on 1980 - 1985 data.
6-22
-------
TABLE 8
KANAWHA SURFACE WATER REPORT
SEPTEMBER 1985 METAL CONCENTRATION RANGES
IN FISH FILLET COMPOSITES AT NITRO, WEST VIRGINIA
INTENOEO FOR POLICY DEVELOPMENT
(mg/kg)
FISH SPECIES* (NUMBER OF COMPOSITES IN RANGE)
LARGE SMALL SMALL FRESH
METALS
MOUTH
BASS(5)
MOUTH
BASS (1)
SAUCER (2)
CHANNEL
CATFISH (2)
MOUTH
BUFFALO (5)
WATER
DRUM (2)
Mercury (inorganic)
Tr i 0.02
Tr
Tr - 0.02
Tr - 0.02
Tr - 0.02
Tr - 0.02
Mercury (organic)
0.1 - 0.17
- 0.18
0.1*4 - 0.23
0.15-3.22
0.1-0.24
0.18-0.33
Cadmium
Tr - 0.05
Tr
Tr - 0.05
Tr - 0.05
Tr - 0.05
Tr - 0.05
Chromium
0-0.8
NO
Tr - 0.4
Tr - 0.4
Tr - 1.3
0.9 - 2.9
Copper
1.7-2.9
2.5
2.5 - 2.8
0
•
N
1
in
1.4 - 2.8
1.7-2.3
Lead
Tr - 0.7
0.5
0.5-0.7
Tr - 0.4
Tr - 0 .4
0.5
Silver
0.1
0.1
0.1
0.1
0.1
0.1
Zinc
8.5-12.8
7.7
8.4-8.9
4.8-8.7
8.3-8.9
5.1 -5.2
Areenlc
0.1 -0.14
(0.118)'
0.1
(0.1 )»
0.02 - 0.03
(0.025)*
0.01 - 0.02
(0.015)f
Tr - .02
(0.018)'
0.04 - 0.05
(0.045)'
Nickel
Tr- 1.1
Tr
Tr - 0.8
Tr - 0.5
Tr - 0.5
Tr -0.4
• Fleh sample* obtained by the West Virginia Oepartment of Natural Reeourcee In September, 1905 at Nltro, WV <
Tr ¦ 'Tract*. A 'tract* value Indicate* the presence of pollutant concentration belov the reliable detection 11m*
NO ¦ Not Oetected.
'Moan valuta war* calculated for Areenlc vhlch la considered a carclnogan.
Rlek eetlmatee for Areenlc In this analyele un thle mean.
6-23
-------
In TABLE 9 we present RfDs for possible noncancer health
effects for the selected pollutants. The RfDs are benchmark values
for the protection of human health from noncancer health effects.
Exposures below the RfD are not likely associated with noncancer
health risks, As the frequency of exposures exceeding the RfD
increases, as well as the magnitude of the exceedance, the probability
of adverse noncancer health effects increases.
TABLE 10 presents estimated weekly consumption rates for
mercury, silver,and chromium for the highest observed fish tissue
concentration from the Nitro Area dataset. To be conservative,
we have assumed that organic mercury is methyl mercury and total
chromium is hexavalent. It is likely that chromium does not exist
in its hexavalent form in fish tissue.
According to TABLE 10 an individual could consume one pound of
fish a week with the highest detected concentration of total mercury
and exceed the Reference Dose for mercury.
WATER QUALITY EXCEEDANCES
To determine if aquatic life could be affected by metals within
the water column, we determined the percentage of observations which
exceed ambient water quality criteria for the protection of aquatic
life.
Water quality criteria designate ambient water concentra-
tions which are not expected to produce harmful effects in
aquatic oganisms. Levels which exceed the criteria may produce
harmful effects. These criteria concentrations are often functions
of pH, hardness, and chronic toxicity of the pollutant to aquatic
organisms. This information is used to calculate the average
concentration that should not cause unacceptable levels of chronic
toxicity during a long term exposure.
We compared observed values for the years 1983-85 to the
chronic toxicity criteria for a pollutant. in many cases,
EPA has established 4-day average concentrations for metals; for
other metals, 24-hour average concentrations criteria have been
determined.
TABLE 11 presents the number of samples analyzed by metals
monitored within the Kanawha River, the chronic toxicity criteria
6-24
-------
TABLE 9
KANAWHA VALLEY SURFACE WATER REPORT
NONCANCER HEALTH EFFECTS
Pollutant
Chromium - III
Health Effect
none
observed
Reference Dose1
(mg/kg-day)
0.005
Study RfD
Based On
Ivankovic and
Preussmann
(1975)
Chromium - VI
none
observed
1.00
MacKenzie et al.
(1958)
Mercury (Inorganic)
renal and
kidney damage
0.002
Fitzhugh et al.
(1950)
Methyl Mercury
CNS effects
0.0003
Clarkson et al.
(1973)
Silver
argyna
0.003
Gaul and Staud
(1935)
Blumberg and Carey
(1934)
East et al.
(1980)
'EPA/RfD indicates a benchmark value derived from no-effect threshold value.
Exposures greater than the RfD are assumed to be associated with an increase
in noncancer health effects. EPA reviews and verifies these thresholds internally
through the Reference Dose Workgroup.
6-25
-------
TABLE 10
KANAWHA VALLEY SURFACE WATER REPORT
LEVELS OF WEEKLY FISH CONSUMPTION
THAT COULD EXCEED NONCANCER HEALTH EFFECTS
REFERENCE DOSES FOR METALS
INTENDED FOR POLICY DEVELOPMENT
Metal
Highest
Concentration
Species
lbs/week
to Exceed
RfD
Mercury( organic) 0.33
Fresh Water Drum
Chromium#
2.9
Fresh Water Drum
2
Silver
0.1
Fresh Water Drum*
32
~Concentration Is for Total Chromium but reference dose for Chromium-VI used to
calculate lbs/week.
TAU species showed the same concentration value.
Source: Regulatory Integration Division, Office of Policy Analysis, EPA.
6-26
-------
TABLE 11
KANAWHA VALLEY SURFACE WATER REPORT
PERCENTAGE OF SAMPLE OBSERVATIONS FROM 1983-1985
WHICH EXCEED CHRONIC WATER QUALITY CRITERIA
FOR THE PROTECTION OF AQUATIC LIFE
INTENDED FOR POLICY DEVELOPMENT
I OF ACTUAL 1
NUMBER OBSERVATIONS 4 DAY
OF EXCEEDING CHRONIC AVERAGES
INORGANIC OBSERVATIONS CRITERIA* (ug/1)
Antimony 12 0 1600
Arsenic 24 0 190 (2)
Cadmium* 52 44 0.66
Chromium (+VI) 26 0 II
Copper* 26 0 6.5
Cyanide 26 0 5.2
Lead* 46 30 1.3
Mercury II 3 33 0.012
T Concentrations used in this report are total concentrations 8nd in this table
are compared to dissolved concentrations used in chronic criteria documents.
* At 8 hardness of 50 mg/L of CaC03.
(1)4- day averages not to be exceeded more than once every 3 years.
(2) A water quality criterion exists only for Arsenic III. Observed
concentrations are for total Arsenic.
INORGANIC
Nickel
Selenium
Silver
Zinc
NUMBER
OF
OBSERVATIONS
12
11
12
18
X OF ACTUAL
OBSERVATIONS
EXCEEDING CHRONIC
CRITERIA'
0
0
0
0
24 HOUR
AVERAGES
(ug/D
56
35
0.12 (4)
47
(3) 24 hour not to be exceeded. These values are used here for
inorganics for which 4 day averages are not available.
(4)
Available data indicate that this is the concentration to be used
for chronic toxicity.
6-27 Source: Regulatory Integration Division, Office of Policy Analysis, EPA.
-------
adjusted for water hardness, and the percentage of those samples
analyzed which were actually observed to exceed the chronic
toxicity criteria. Detection limit values were not counted as
exceedances.
According to TABLE 11 three metals have exceeded their
criteria during several sampling events: cadmium, lead, and
mercury. More than 40 percent of the cadmium samples analyzed and
30 percent of the lead samples analyzed exceeded their respective
average chronic criteria. Only three samples were analyzed for
mercury; more data are needed to evaluate the potential ecological
impact of mercury.
This comparison is only a rough aoproximation of the potential
effect of toxic metal concentrations on agnatic life. We emphasize
that we do not know if a metal concentration exceeded a criterion
for the appropriate time period corresponding to the criterion.
Our observations were only grab samples taken from the Kanawha
River and were not designed to determine if such concentrations
persisted for a duration of time. Nonetheless, such a comparison
suggests that certain pollutants are possibly exceeding their
criteria and should be monitored for longer time oeriods.
Industry with the state is currently conducting a cooperative
monitoring project. This project has monitored for metals and
organics at six sampling locations from river mile 33 to river mile
91 within the study area. The purpose of this sampling is to
detect seasonal variability of water concentrations. Sampling has
occurred in May, August, November of 1986 and February 1987.
Preliminary results are available for the May and August sampling
period, (DNR, 1987). For metals we comoared our water concentra-
tions in Table 7 to the observed concentrations from the May and
August sampling of the cooperative monitoring project and found
the detected ranges comparable, except for cadmium and lead. The
cooperative monitoring project has not detected lead nor cadmium
the two metals we have identified in our analyses as chemicals
requiring further investigation, during the monitoring period.
The Environmental Research Laboratory conducted ambient toxicity
tests in the Charleston area of the Kanawha River in 1984. Test
organisms were exposed to ambient water from the Kanawha River and
were examined for young production and survival after a period of
exposure. Although several sample sites did indicate toxicity and
reduced young production, the overall results suggested very little
ambient toxicity (Personal Communication, Donald Mount, Senior
Research Scientist, to Bernard Goldstein, EPA, 1985), (Personal
Communication. Scott Heinritz, Aquatic Biologist, to Eli McCoy,
Division of Water Resources, DNR).
VII. ORGANICS
Industrial organic chemicals is the third pollutant category
reviewed within the surface water analysis of the Kanawha Valley
Toxics Study. In this section we briefly outline our methodology
6-28
-------
to estimate potential health and ecological effects of organic
chemicals in the water column.
The information for organic pollutants was even more limited than
the data for pesticides, dioxin, PCBg, and metals within the Kanawha
Valley study area. For example, between the years 1933 and 1985,
only two sampling stations, both within the St. Albans and Nitro
area, monitored for organic pollutants. No fish tissue data for
organic pollutants were available, and our ability to determine the
water quality for the Kanawha River within the study area was
extremely limited. We therefore focused our analysis
for organic pollutants on this limited segment of the Kanawha
River.
For these two stations, we reviewed monitored ambient water
concentrations for selected pollutants. These concent rat ions
were compared to EPA derived ambient water quality criteria for
the protection of human health at the 10~7, 10~6, and 10~5 risk
levels. In addition to human health criteria, we compared observed
organic chemical concentrations to organic chemical concentrations
known to cause chronic and acute effects in aquatic organisms.
Sources for Organic Pollutants
Although we have measured ambient water concent rat ions for
only a limited segment of the Kanawha River, we do know that
organic chemicals are discharged from the major industrial
facilities within the study area. We hav:^ reviewed discharge
monitoring reports, NPDES oermit applications and priority
pollutant reports for the major facilities discharging into the
Kanawha River for the years 1983-85. TABLE 12 lists organic
pollutants which have been detected In effluents of the major
facilities. Due to the limited data for organic effluent
discharges, we could not quantify organic loadings to the Kanawha
Ri ver.
Ambient Water Concentrations
The Ohio River Valley Water Sanitation Commission (ORSANCO)
maintains an organic pollutant monitoring station in the St. Albans
area, downriver from the major chemical facilities. Spill detection
is the major purpose of this monitoring station. TA3LE 13
presents the monitored chemicals, the concentration ranges of
those chemicals, the mean concentration value for the monitored
pollutants and the standard deviation. Nondetected concentrations
are averaged at the detection limit. The ORSANCO station data is
considered qualitative data, and may be in error by as much as
100%.
Tn addition to the ORSANCO river monitoring station, the
West Virginia Department of Natural Resources monitored organic
pollutants further down river at River Mile 38 for several years.
DNR discontinued monitoring in 1984 at this station. TA3LE 14
oresents the concentration ranges and the mean concentration for
6-29
-------
TABLE 12
KANAWHA VALLEY SURFACE WATER REPORT
MAJOR FACILITIES DISCHARGING ORGANICS TO THE KANAWHA RIVER
BENZO(K)- B ISC 2— CARBON
MAP FLUOR- CHLORO- BROMO- TETRA- CHLORO-CHLORO-CHLORO-
KEY FACILITY BENZENE ANTHENE ETHYL)ETHER FORM CHLORIDE FORM BENZENE ETHANE
7
Occidental
¦f
¦/
¦f
8
Dupont
¦S
11
FMC, South Charleston
¦/
¦S
¦/
12
CST/Fike Chemicals (Artel)
¦S
¦S
¦/
13
Kincald Enterprises
15
Monsanto
¦S
¦S
16
S. Charleston Sewage
Treatment Plant
¦S
¦S
¦/
¦/
17
Rhone Poulenc
¦S
19
Union Carbide, S. Charleston
SOURCE: Discharge Monitoring Reports, NPDES permit applications and priority pollutant scans (1983-85).
-------
TABLE 12 (continued)
KANAWHA VALLEY SURFACE WATER REPORT
MAJOR FACILITIES DISCHARGING ORGANICS TO THE KANAWHA RIVER
DI8R0M0- Dl- DICHLORO- 2,4-DI- DIETHYL Dl(ETHYL- ETHYLENE
MAP CHLORO- CHLORO- BROMO- CHLORO- PHTHA- HEXYL)- ETHYL DICHLOR-
KEY FACILITY METHANE BENZENE METHANE PHENOL LATE PHTHALATE BENZENE IDE
7
Occidental
8
Dupont
11
FMC, South Charleston
•/
•f
12
CST/Flke Chemicals (Artel)
•f
S
V
S
13
Klncald Enterprises
15
Monsanto
16
S, Charleston Sewage
Treatment Plant
¦f
s
17
Rhone Poulenc
19
Union Carbide, S. Charleston
SOURCE: Discharge Monitoring Reports, NPDES permit applications and priority pollutant scans (1983-85).
-------
TABLE 12 (continued)
KANAWHA VALLEY SURFACE WATER REPORT
MAJOR FACILITIES DISCHARGING ORGANICS TO THE KANAWHA RIVER
METHYL-
TETRA-
TRI-
1,1,1 TRI—
1,2,4 TR I—
VINYL-
MAP
ISOPHOR-
ENE CHLOR-
METHOXY-
NAPTH-
CHLORO-
CHLORO-
CHL0R0-
CHLORO-
VINYL
IDENE
KEY
FACILITY
ONE
IDE
CHLOR
ALENE
ETHYLENE
ETHENE
ETHANE
BENZENE TOLUENE CHLORIDE CHLORIDE
7
Occidental
¦/
8
Dupont
J
V
11
FMC, South Charleston
12
CST/F1ke Chemicals (Artel)
¦f
¦f \ S
V
13
Klncald Enterprises
~
15
Monsanto
V
i V
16
S. Charleston Sewage
Treatment Plant
V
V
~
17
Rhone Poulenc
~
J
19
Union Carbide, S. Charleston
V
SOURCE: Discharge Monitoring Reports, NPDES permit applications and priority pollutant scans (1983-85).
-------
TABLE 13
KANAWHA VALLEY SURFACE WATER REPORT
AMBIENT WATER ORGANIC CONCENTRATIONS AT KANAWHA RIVER MILE 42 (ORSANCO STATION)
1983 - 1985
OBSERVATIONS = 369
ORGANIC CONCENTRATION RANGE ESTIMATED MEAN STANDARD DEVIATION
(uq/1) (uq/1)
BROMOFORM
<.5 - 86
8.2
10.9
CARBON TETRACHLHORIDE
<.5 - 18
0.8
1.6
CHLOROFORM
<.5 - 230
3.1
2.7
1,1-DICHLOROETHYLENE
<.5 - 29
0.7
1.7
METHYLENE CHLORIDE
<.5 - 15
0.9
1.3
TETRACHLOROETHYLENE
<.5 - 747
3.6
38.6
TRICHLOROETHYLENE
<•5 - 30
1.6
9.3
1,1-DICHLOROETHANE
<.5 - 14
1.7
1.0
1,2-DICHLOROETHYLENE
<.5 - 7
0.8
0.8
1,2-DICHLOROETHANE
<.5 - 41
1.7
3.2
1,1,1-TRICHLOROETHANE
<.5 - 18
1.0
1.7
BRCMODICHLOROMETHANE
<.5 - 96
1.8
5.6
1,2-DICHLOROPROPANE
<.5 - 49
1.2
3.4
1,3-DICHLOROPROPENE
<.5 - 29
0.8
1.9
BENZENE
<.5 - 19
1.7
2.1
TOLUENE
<.5 - 137
6.5
16.9
CHLOROBENZENE
<•5 - 28
1.7
2.7
ETHYLBENZENE
<.5 - 39
1.5
3.0
1a less than sign (<) indicates that the concentration is less than the detection limit.
^The average is based on all sample concentrations between 1983 and 1985. Values below
the detection limit are averaged at the detection limit concentration.
SOURCE: Ohio River Organics Detetion System, ORSANCO, 1986
6-33
-------
TABLE 14
KANAWHA VALLEY SURFACE WATER REPORT
AMBIENT WATER ORGANIC CONCENTRATIONS AT KANAWHA RIVER MILE 38 (KRM38)
1983-1984
OBSERVATIONS =131
ORGANIC
CONCENTRATION RANGE
(uq/1)
ESTIMATED MEAN
(uq/1)2
STANDARD
DEVIATION
DICHLOROBROMOMETHANE
<.1 -
1.0
0.22
0.21
BROMOFORM
<.1 -
0.7
0.12
0.08
CHLORODIBROMOMETHANE
<•1 -
0.7
0.13
0.10
CHLOROFORM
<•1 -
9.5
1.32
1.78
CARBON TETRACHLORIDE
<•1 -
2.2
0.61
0.52
METHYLENE CHLORIDE
<.1 -
4.6
0.19
0.48
TETRACHLOROETHYLENE
<.1 -
6.5
0.29
0.63
TRICHOLROFLUOROMETHANE
<.1 -
5.8
0.17
0.52
1,1-DICHLOROETHANE
<.1 -
0.5
0.11
0.04
1,1-DICHLOROETHENE
<.1 -
4.8
0.36
0.57
1,1,1-TRICHLOROETHANE
<•1 -
6.2
0.35
0.75
1,2-DICHLOROETHANE
<.1 -
3.9
0.3
0.46
1,2-DICHLOROPROPANE
<.1 -
5.4
0.2
0.48
TRICHLOROETHYLENE
<.1 -
2.3
0.26
-
CHLOROBENZENE
< .1 -
3.9
1.4
0.42
1a less than sign (<) indicates that the concentration is less than the detection limit.
^The average is based on all sample concentrations between 1983 and 1984. Values
below the detection limit are averaged at the detection limit concentration.
SOURCE: Department of Natural Resources, West Virginia (Storet, 1986)
6-34
-------
the years 1983 and 1984. Again, nondetected concentrations are
averaged at the detection limit concentration. Therefore these
mean concentrations are likely to overstate the true mean value.
When the concentration ranges and mean concentrations from the
two monitoring stations are compared, we notice that concentrations
are higher at the ORSANCO monitoring station than the DNR station
further downstream. Volatilization of these organics from the
water column, diLution by tributaries as -organic chemicals travel
downstream, different sampling techniques, and time periods may
account for the concentration differences.
Human and Aquatic Health Effects
TABLE 15 presents water quality criteria for the protection
of human health from consumption of contaminated fish tissue, and
observed concentration values which have produced harmful effects
on aquatic organisms. The water quality criteria were developed
by the Office of Research and Development within EPA and published
by the Office of Water. These values, based on pollutant-specific
bioconcentration factors and a national average individual daily
consumption rate, attempt to provide water quality concentrations
at which a pollutant may bioaccumlate i.n fish to various concentra-
tions levels. These water quality criteria are water concentrations
representing various levels of 10-5, 10"®, and 10"7 risk. In TABLE 15
we present the water quality criteria which represent an estimated
conservative lifetime average risk of 10-® for fish consumption.
The values to protect aquatic life are not criteria. For many
organic chemicals there are not enough data to establish a water
quality criteria for protection of aquatic life. These values
represent the lowest observed concentrations to date that have
produced harmful effects to aquatic organisms. Only acute effects
have been studied for many organic pollutants.
We compared these criteria and observed effect concentrations
of TABLE 15 to the observed concentration ranges at the ORSANCO
station. TABLE 16 presents the percentage exceedances. We found
four pollutants for which the human health criteria set at the 10"^
risk level were exceeded; chloroform, tetrachloroethylene,
trichloroethylene, and 1,l-dichlorethylene. However, each of these
exceedances represents only one (or 0.3%) observation exceeding the
criteria. At the 10"® risk level, we find the criteria exceeded by
seven pollutants; the four mentioned above plus bromoform, carbon
tetrachloride, and bromodichloromethene. All of the pollutants
sampled exceeded the criteria set at the 10"^ risk level, some quite
consistently such as 1,1-dichlorethane (97.9%). Of the three
pollutants looked at for noncancer health effects, 1,1,1-Trichloroe-
thane (11.1%) and toluene (3.4%) showed exceedances of the threshold
levels. The observed toxic concentrations for aquatic orqanisms
were never exceeded.
6-35
-------
TABLE J 5
WATER QUALITY CRITERIA FOR THE
PROTECTION OF HUMAN HEALTH AND POTENTIAL TOXIC
WATER CONCENTRATIONS TO AQUATIC ORGANISMS
INTENDED FOR POLICY DEVELOPMENT •
Ambient Water Concentration
Associated With An
Lowest Observed Toxic
Water Concentrations
Individual Lifetime Risk of 10"
-6 weight of
Harmful To Aquatic
For An Assumed Level Of
Evidence
Organisms
Fish Consumption'
for
(ug/l)
Oraanic
(ua/1)
Carcinoaens
Acute
Chronic
Dichlorobromomethane
15.7
11,000
NA
Bromoform
15.7
11,000
NA
Chlorodibromomethane
15.7
11,000
NA
Chloroform
15.7
B2
28,900
1,240
Carbon Tetrachloride
6.94
B2
35,200
NA
Methylene Chloride
15.7
B2
11,000
NA
Tetrachloroethylene
8.85
B2
5,280
840
T richlorofluoromethane
15.7
11,000
NA
1,1 Dichloroethane
NC
NA
NA
1,1 Dichloroethylene
1.85
C
11,600
NA
1,2 Dichloroethane
243.0
B2
118,000
20,000
1,2 Dichloropropane
NC
C
23,000
5,700
Trichloroethylene
80.7
B2
45,000
21,900'
Benzene
40.0
A
5,300
NA
1,3 Dichloropropane
NC
23,000
5,700
1,2 Dichloroethene
NC
NA
NA
1,1,1 Trichloroethane
Toluene
Chlorobenzene
Ambient Water Concentration
Associated With
Noncancer Effects
For An Assumed Level Of
Fish Consumption
(tio/1)
1.03
424,000
488.0
18,000
17,500
250
NA
NA
NA
NC - No Criteria has been calculated for these chemicals.
NA - Not Available.
'The Office of Water has set these levels of ambient water concentrations to protect human health at the 10-6
level of risk from consumption of fish tissue, except where noted. Other concentration criteria are set at
higher and lower risk levels. To calculate a risk level one order of magnitude higher, increase the
concentration by one order of magnitude.
THE UNIT RISK FACTORS USED IN DETERMINING THESE CONCENTRATIONS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT
GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS, RISK ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL RISKS IN THE KANAWHA VALLEY.
ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER; IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATE
IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
'This concentration is a 27 day LC50 value.
p
'¦Chronic criteria has not been established; this concentration adversely affected the behavior of one species.
Source: Office of Wat«r. EPA.
6-36
-------
TABLE 16
KANAWHA VALLEY SURFACE WATER REPORT
THE PERCENT OF OBSERVATIONS
AT ORSANCO WHICH EXCEED WATER QUALITY CRITERIA
FOR THE PROTECTION OF HUMAN HEALTH AT 10-5, 10~6, 10~7
LEVEL OF RISK AND NONCANCER HEALTH EFFECTS
FROM CONSUMPTION OF FISH TISSUE
% OBSERVATIONS EXCEEDING HUMAN HEALTH CRITERIA SET AT:
RISK LEVELS
NONCANCER
HEALTH EFFECTS
THRESHOLD
ORGANIC
10"5
10~6
10"7
Bromoform
0%
19.0%
62.5%
Chloroform
0.3%
2.9%
35.9%
Tetrachloroethylene
0.3%
3.2%
29.0%
Carbon Tetrachloride
0%
1.3%
11.3%
Trichloroethylene
0.3%
1.6%
29.0%
Bromodichloromethane
0%
0.8%
22.2%
Methylene Chloride
0%
0%
12.9%
1,1-Dichloroethylene
0.3%
13.4%
97.9%
1,2-Dichloroethane
0%
0%
0.5%
Benzene
0%
0%
7.7%
1,1,1-Trichloroethane
11.1%
Toluene
3.4%
Chlorobenzene
0%
Source: Office of Policy Analysis, Regulatory Integration Division, EPA.
6-37
-------
Again, the reader Is reminded that conservative assumptions
were used in the determination of risk levels. Due to these assump-
tions and the limited nature of the ORSANCO station data, the
actual percentage of exceedances are likely to be lower than TABLR
16 indicates. Furthermore, these water concentration which
correspond to various risk levels must, in theory, be maintained
over a period of a lifetime to pose the associated cancer risk.
The percent exceedances do not indicate a level of risk, but only
serves as a way to compare pollutants to each other. The actual risks
posed by organics require much more detail. The final results of
the cooperative monitoring program should be reviewed, when available,
to determine the frequency of exceedances of these pollutants.
SUMMARY AND CONCLUSIONS
In the Kanawha Valley Surface Water Report we have attempted
to develop a sense of the potential public health concerns toxic
pollutants may pose within the Kanawha River. We emphasize that we
examined only one exposure pathway, fish tissue consumption, within
the study area. We also emphasize that our analysis is extremely
limited due to limited health effects and exposure information.
An assessment of potential cancer risk and other potential noncancer
health risk posed to a known population is not possible given our
inability to characterize actual exposure within the study area.
Furthermore, monitoring data are limited in time as well as area
with much of the information drawn from monitoring stations on the
lower Kanawha River in the study area. Our health effects analysis
of pesticides, dioxin, PCB's and metals is based on a one day
sampling of fish fillet in the lower Kanawha River while our analysis
of industrial discharged organics relies on observed water concentra-
tions from two stations below many of the industrial facilities on
the Kanawha River.
Despite these limitations, we believe our limited analysis may
provide a general sense of the potential health hazards and may
provide some direction and priorities in the collection of data and
information needed for assessment of the effects of toxics
within the Kanawha River.
We present the following findings from this analysis:
° DNR of West Virginia has detected both pesticides and metal
concentrations within fish tissue from samples taken
throughout the study area.
A review of current discharge permits indicate that manufac-
tured pesticides in the Kanawha are not discharged and,
therefore, pesticide concentrations within fish tissue are
likely a result of non-point source run-off.
Metal loadings occur from both point and non-point sources
within the study area. Based on our review of non-point
source information and discharge permits, our analysis
6-38
-------
indicates that non-point source run-off, is likely to
contribute larger loadings than point sources for most
metals. Cadmium is an exception. TABLE 5 presents 22
metals point source discharges.
Industrial discharged organics have been detected in the
effluents of 9 point source discharger in the study area
(See Table 12). In addition, organics have been detected
in the water column at two monitoring stations on the lower
Kanawha River in the study area. TABLE 13 and 14 present
organics detected.
0 Based on a one day sampling on the Nitro area of the Kanawha
River, we calculated a weekly consumption rate of .1 to 1
pound, depending on the fish caught (Smallmouth Buffalo,
Channel Catfish, Sauger, Largemouth Bass, Spotted Bass,
Smallmouth Bass, and Freshwater Drum) which would expose an
individual to an upper-bound incremental lifetime risk of
1 X 10~3 or one out of thousand chances of contracting cancer
over a lifetime exposure. Dioxin and PCBs are the pollutants
primarily responsible for the risks at these various hypo-
thetical consumption rates. We do not know if these concen-
tration levels occur throughout the river.
For organics, we compared water concentrations obtained from
the lower Kanawha River to various levels of water quality
criteria set to protect human health. These criteria assume
that the average individual consumed approximately one
tenth of a pound of fish per week, a rate calculated for
the analysis above. Two pollutants, bromoform and
1,1-dichloroethylene exceeded the criteria concentrations
to protect human health at the 1 X 10~® (one out of a
million) for more than 10 percent of the observations.
Only one observations for each of these pollutants exceeded
the concentration representing a 1 X 10~* (one out of one
hundred thousand) risk.
0 We calculated consumption rates required to exceed EPA
established reference doses for chronic non-cancer health
effect for chlordane, which was detected in a one day sampling
in the Nitro area. We found that consumption rates ranged from
.3 pound per week to eighteen pounds per week, for
certain species caught, which would exceed the RfD for
chlordane, a nonpoint pollutant. Chlordane was not detected
in Small Mouth Bass nor Spotted Bass.
For mercury, we assumed that the organic mercury was methyl
mercury. If an individual were to consume the highest
observed concentration for organic mercury, at a consumption
rate of one pound per week, the RFD would be exceeded.
6-39
-------
To develop a rough approximation of the potential effect of
toxic metal concentration on aquatic life we compared the
number of monitored ambient water concentrations of a limited
set of metals to their corresponding chronic water quality
criteria for the protection for aquatic organisms. During
the monitoring period, lead and cadmium concentrations
exceeded their criteria 44 and 30 percent of their observa-
tions respectively. Cadmium comes predominately from
industrial discharges while lead comes from non-point source
run-off.
We compared our limited set of observed organic water concent-
rations from the lower Kanawha River to the lowest observed
toxic water concentrations harmful to aquatic organisms, a
proxy for a water quality criteria to protect human health.
These levels were not exceeded by the detected concentrations.
DATA GAPS
We have investigated only one type of exposure, the potential
risk from the consumption of fish tissue. There are several
other routes of exposure which should be investigated. Dermal
contact, water ingestion, and consumption of water fowl are
other routes of exposure from pollutants within the Kanawha
River.
Our data on fish fillet concentrations is currently limited
to the Nitro Area within the study area. Since whole fish
data indicate potential bioaccumulation of many of these
pollutants, fiilet data should be obtained for the upper
reaches of the river. Our analysis suggests that fish
tissue contamination by PCBs, dioxin, chlordane, and mercury
may pose more of a concern than other pollutants for which
we have health effects data.
The consumption rate of fish tissue is not currently known.
The West Virginia DWR Creel Survey will provide consumption
information about fish caught from the Kanawha River.
Recently the State in cooperation with industry has
monitored for priority pollutants throughout the entire study
area. The database should provide much needed monitoring
data on organic and inorganic pollutants in the Kanawha
River.
Finally, health effects information of these pollutants
should be continually reviewed as new data and knowledge
are developed. Current Reference Dose values for noncaricer
health effects and cancer potency values may change, and
values for new pollutants may be developed. Due to the
evolving nature of the knowledge of health effects information,
exposure concentration should be continually reviewed with
new health effects information.
6-40
-------
REFERENCES
Department of Natural Resources, 1987. Preliminary data from the
Cooperative Monitoring Program. January, 1987.
Department of Natural Resources, 1986. Whole fish tissue
concentration data, 1980-1984. July, 1986.
EPA, 1982. "NURP Priority Pollutant Monitoring Program,
Volume I: Findings". September, 1982.
EPA, 1986. "Environmental Overview of the Kanawha valley". 1986.
Heinritz, Scott, EPA, 1985. Correspondence to Eli McCoy, WV DNR,
transmitting preliminary results of ambient toxicity tests. 1985.
Hodges, Jim, WV Department of Health, 1986. Personal Communication;
telephone conversation concerning drinking water sources
for public water supplies. November 1986.
Kain, Donald, WV DNR. Personal communication; Correspondence
concerning the DNR Creel Study. January, 1986.
Mount, Donald, EPA, 1985. Memo to Bernard Goldstein, Office of
Research and Development, EPA, concerning results of ambient
toxicity tests within the Kanawha Valley. March 1985.
Smith, Roy, Region III, EPA, 1987. Personal communication;
telephone conversation concerning fish tissue contaminated with
dioxin, February, 1987.
6-41
-------
Appendix A
Kanawha Point and Non-Po1nt Source Loadings
This document outlines the procedures followed 1n computing both the
annual non-point source loadings and point source loadings to the Kanawha
River. These results were presented 1n the Kanawha Valley Surface Water
Analysis Table 2 along with the ratio of non-point to point source loadings.
Non-Point Source Loadings
The non-point source loadings for each pollutant are estimated by
multiplying the volume of water that flows Into the river via runoff, by the
average concentration of each pollutant 1n stormwater runoff. These
calculations were based on data from both the 1981 Geographic Study of the
Kanawha Valley (Draft EPA 1981) and the National Urban Stormwater Runoff
Program (NURP) (EPA 1982). The NURP data Included the mean concentrations
(ug/1) of the most frequently detected priority pollutants 1n NURP urban
stormwater runoff. The average flow of runoff to each segment of the river
was taken from the 1981 Draft Geographic study.
Although our estimates for non-point source loadings are based on national
data, we believe these estimates provide a sense of the possible magnitude of
non-point source runoff. Furthermore, these are the best available date for
estimating loadings
The NURP Program was conducted from 1980 to 1982, with runoff samples
being collected from 51 predominantly residential and commercial areas.
Therefore, the results from the NURP effort cannot be attributed to runoff
from Industrial facilities. Attachment 1 presents the average concentration
of heavy metal contaminants from the NURP data set.
The major limitation of the NURP data 1s that urban areas, with different
land uses than 1n the Kanawha Study area, were sampled to generate the average
concentrations. These data probably do not adequately simulate the runoff
from the Industrial complexes that border the Kanawha River. In addition,
loadings of toxic organlcs such as pesticides and PCBs cannot be quantified
using NURP data. Also the large number of CERCLA hazardous waste sites 1n the
study area could contribute a large amount of these types of toxic organlcs to
the river by stormwater runoff. In addition, NURP data 1s only considered to
be accurate to within one order of magnitude.
6-43
-------
To develop run-off flows, we used the 1981 Draft Geographic study., In
this study, the river was divided Into 18 distinct reaches. The division of
the river Into these reaches reflected discontinuities due to dams,
tributaries, and the depth of the river. The direct runoff flow for these 18
reaches was determined as follows. The drainage area of the Kanawha, as well
as the total drainage area of tributaries, 1s known at various points along
the river. The drainage area 1s the watershed area that contributes to the
river. I.e. the area where the land slopes toward the Kanawha river and runoff
could reasonably be expected to flow Into the Kanawha river. The drainage
area of the uprlver station was subtracted from the downriver station, then
the tributary total drainage area was subtracted from that difference leaving
the drainage area for direct runoff. The average annual rainfall 1s known, as
1s the fraction of that rainfall that runs off (Instead of evaporating or
Infiltrating the ground). The volume of direct runoff 1s then calculated by
the following equation:
q = c1a (Merrltt 1968)
where 1 1s the rainfall, a 1s the drainage area and c 1s the fraction of rain
that runs off. This direct runoff was then divided among the 18 reaches
depending on the length of each reach. The runoff volumes per reach are
presented 1n attachment 2.
Since the eighteen reaches developed for the 1981 study do not correspond
to the five segments 1n the present analysis, we modified the runoff flow to
reflect the new river segment 1n the following manner. We assigned each of
the 1981 reaches to the corresponding river segment 1n the present study area
based on river miles covered:
Segment 1 - Reaches 1-5
Segment 2 - Reaches 6-9
Segment 3 - Reaches 10-11
Segment 4 - Reaches 12-13
Segment 5 - Reaches 14-16
Reaches 17 and 18 are outside the study area. We then divided each runoff
flow by Its corresponding reach leng-th to give a runoff per mile for each .
reach. These were then averaged over all the reaches 1n a segment to give
average runoff per mile for every segment. These numbers were then multiplied
by the segment length to give average runoff flow 1n m3/hr for each segment.
6-44
-------
To determine the loadings, the average runoff flow 1n each segment,
described above, was multiplied by the mean concentration of each pollutant
from the NURP stormwater runoff data (1n cases where a range of concentrations
was given, we used the midpoint of the range) along with a conversion factor
to give a loading 1n kkg/yr for each pollutant 1n each segment. The five
segment loadings for each pollutant were sunned to give the total non-point
source loading to the river of each pollutant.
Point Source Loadings
We developed average loadings for 1984. The only water data available
for the Kanawha study area comes from monthly discharge monitoring reports
compiled for the National Pollution Discharge Elimination System (NPDES). We
first determined, for each outfall which 1s monitored, the average flow for
each month. We then multiplied the reported concentrations 1n a month by the
average flow for that same month to get a loading. We then averaged all the
loadings together to get a yearly average. In cases where there are multiple
concentration readings for a pollutant 1n the same month, we kept these
distinct when computing the loadings. As a result, the yearly average may be
composed of more than 12 Individual readings; or conversely, 1n cases where
data for some months was missing, the averages may be composed of less than 12
Individual readings.
Attachment 3 lists the loadings by pollutant. Under each pollutant
heading 1t lists the facility outfall number and average yearly loading. We
then summed the Individual loadings for each pollutant to give the total point
source loading to the river.
References
1) U.S. EPA "An Integrated Geographic Study of Potential Toxic Substance
Control Strategies 1n the Kanawha River Valley, West Virginia" Interim
Draft Report, July 15, 1981.
2) U.S. EPA, Monitoring and Data Support Division, "NURP Priority Pollutant
Monitoring Program Volume I: Findings", September 17, 1982.
3) Merrltt, Frederick S. Standard Handbook for C1v1l Engineers. McGraw-Hill
Book Co., New York, 1968, P 14-7.
6-45
-------
Ati:achei»nt 1
Most Frequently Detected Priority
fei'iutants 1n MAP Urban Stor*»ater Runoff
Staples1
Pollutant Detection Frequency Nun Concentration*
W (m/D
Copper
M
18.1
Load
N
65.5
Zinc
K *
92.3
Arsenic
S8
1.60
Chroalue *"
57
3.13
Cadelta
55
0.57
Nickel t.'
48
5.80
•-Heiach lorocye lohesane
20
0.093
Selenlua ¦
19
(2-77)3
lerylllia
17
(1-49)3
Cyanides '
16
(2-33)3
Mercury
16
(0.5-1.2)3
Pentachlorophenol
IS
(1-115)3
Antlaony
14
(2.6-23)3
Hs( 2-ethylhe*yl)phthalate
13
(7-39)3
Silver
12
(0.2-0.8)3
12
(0.2-12)3
Phenanthrene
12
(0.3-10)3
P -Xesach lorocye lohexane
12
(0.52-0.1)3
(Lindane)
(0.8-2.3)3
Naphthalene
11
Pyrene
11
(0.3-10)3
Thai Hue
10
(1-14)3
Olcnloroaethane (aethylene
10 .
(5-14.5)3
chloride)
•
Fluorantftene
10
(0.3-12)3
Source: OON (1982).
' Pollutants detected 1n at least 1(B of all sanies analyzed (86).
* GeoaetHc Man concentration. Calculated only If pollutant Is
detected in at least 201 of the acceptable sanies. Non-detections
«ere included in calculation as one-tenth of the reported detection
Halt.
3 Mean concentration not calculated. Ranqe of detected values only
reported. For purposes of initial scan analysis the aid-point of the
range was used to estiaate annuai pollutant loans.
6-46
-------
Reach
(ml)
1) 73.7 - 69.4
2) 69.4 - 68.0
3) 68.0 - 67.7
4) 67.7-66.1
5) 66.1 - 57.8
6) 57.8 - 56.3
7) 56.3 - 54.3
8) 54.3 - 52.9
9) 52.9 - 49.6
10) 49.6 - 48.0
11) 48.0 - 45.5
12) 45.5 - 43.5
13) 43.5 - 41.4
14) 41.4 - 39.3
15) 39.3 - 32.0
16) 32.0 - 31.1
17) 31.1 - 18.0
18) 18.0 - 0.0
ATTACHMENT 2
Direct Runoff
(m3/hr)
6.66
X
103
2.17
X
103
4.64
X
102
2.48
X
103
1.28
X
104
2.32
X
103
3.09
X
103
1.49
X
io3
3.52
X
io3
1.71
X
io3
2.67
X
io3
2.13
X
io3
2.24
X
io3
2.24
X
io3
1.83
X
io3
2.25
X
io2
2.82
X
io4
3.88
X
io4
6-47
-------
Attachment 3
KANAUHA VOLLEY, W STUDY MCA
TOOL 191: MEW LOADINGS IN KGD BASED CN OMR DATA 0N.Y
YEAR«84 POLLUTANT-ANTIMNY (SB)
CODE FACILITY PIPE KGD
W0000078
CARBIDE-SUI DIVISION
24
•
W0000086
CARBIDE-INSTITUTE
01
0.10196
W0000086
CARBIDE-INSTITUTE
06
0.80186
W0000167
ELXE* ICTALS
08
0.00005
WV0001074
APPALACHIAN 1074-AN0S
04
0.01363
HW001074
APPALACHIAN 1074-AWS
11
0.06631
W0001074
APPALACHIAN 1074-AWS
27
0.00900
W0002399
DtPONT -BELLE
62
•
POLLUTANT 1.01289
YEAR»84 PQLLUTANT4WSENIC
CODE FACILITY PIPE KGD
UV0000086
CARBIDE-INSTITUTE
01
0.17112
W0000086
CARBIDE-INSTITUTE
06
0.26729
IM0000167
ELXEM ICTALS
06
0.00002
W0000400
FtC-NITRO
01
1.133B9
UV0000400
FtC-NITRO
OS
0.01730
W0000442
FC-S. CHARLESTON
29
0.96944
W0001066
APPALACHIAN 1066-KANAfcHA
01
0.02030
MW0001074
APPALACHIAN 1074-AN0S
01
1.36078
W0001074
APPALACHIAN 1074-AN0S
03
0.17209
W0001074
APPALACHIAN 1074-AM0S
04
0.01817
W0001074
APPALACHIAN 1074-AN0S
11
0.00303
W0001074
APPALACHIAN 1074-AN0S
27
0.01314
W0001631
FIKE OEMICAL-C.S.T.
01
0.02833
W0002381
HATFIELAUCNSON
01
0.14839
W0002399
DIPOKT -BELLI
062
•
POLLUTANT 19472*
YEAR'S* POLLUTANT"BARIUM
CODE FACILITY PIPE KGD
W0001651 FIKE OENICAL-C.S.T. 01 0.04994
6-48
-------
YEAR«84 POLLUTttfT*CAMIUH
CODE
FACILITY
PIPE
KGB
WI623382
SMITH FASTEfER
01
0.0000
W0000086
CARBIDE—IN5TITUTE
01
0.4758
W0000086
CARBIDE-INSTITUTE
03
1.6797
WOOCOOtt
CARBIDE-INSTITUTE
05
86.2047
W0000167
EU® METALS
06
0.0000
HV0001074
APPALACHIAN 1074-flMS
04
0.0227
W0001074
APPALACHIAN 1074-AWS
11
0.0076
UV0001074
APPALACHIAN 1074-AWS
27
0.0151
W0001651
FIKE DENICAL-C.S.T.
01
0.0029
POLLUTANT
90.4067
YEBR«64 POLLUTANT-CHROMl* TOTAL
CODE
FACILITY
PIPE
KGD
W0000078
CARBIDE-SUI
DIVISION
16
•
UV0000078
CARBIDE—SUI
DIVISION
17
•
W0000078
CARBIDE-SUI
DIVISION
18
•
W0000078
CARBIDE-SUI
DIVISION
19
•
UV0000078
CARBIDE-SUI
DIVISION
22
•
IM0000078
CARBIDE-SUI
DIVISION
23
•
WW0000078
CARBIDE-SUI
DIVISION
24
•
W0000078
CARBIDE-SUI
DIVISION
28
•
W000007B
CARBIDE-SUI
DIVISION
29
•
W0000078
CARBIDE-SUI
DIVISION
32
•
W0000078
CARBIDE-SUI
DIVISION
S3
•
W0000078
CARBIDE-SUI
DIVISION
74
•
W0000078
CARBIDE-SUI
DIVISION
80
•
W0000086
CARBIDE-INSTITUTE
02
•
W0000167
ELJ0I METALS
02
0.019491
W0000167
ELKEM METALS
03
0.000000
W0000167
ELKEK METALS
05
0.000000
W0000167
ELKEN PETALS
06
0.000000
W0001074
APPALACHIAN 1074-AN0S
04
0.136274
W0001074
APPALACHIAN 1074-AM0S
11
0.045425
W0001074
APPALACHIAN 1074-AN0S
27
0.090850
POLLUTANT
0.29204C
YEAR"84 POLLUTflMKHRWIl* KEXflVflLEKT
QBE FACILITY PIPE KGD
W0000400
nC-NITRO
01
0.000081
W0000400
FMC-NITRO
02
0.000000
W0000442
FHC-5. QWHESTON
22
0.014599
W0000442
FMC-S.CHARLESTON
23
0.009234
UV0000442
FMC-S.CHARLESTON
29
0.231558
wooooaea
MONSANTO
01
0.453600
W0001631
FIKE CHEMICAL-C.S.T.
01
0.003539
W0002372
OEM LEAMAN TAW LIfCS
01
0.045425
W0050130
COASTAL TAN( LINES
01
0.000325
POLLUTANT
0.758361
6-49
-------
YEAR-84 POLLUTANT-COPPER
CODE FACILITY
PIPE KB
MVI623382
SMITH FASTBER
01
0.00081
W0000078
CARBIDE—SUI DIVISION
16
•
W0000078
CARBIDE-SUI DIVISION
17
•
W0000078
CARBIDE—SUI DIVISION
18
•
W0000078
CARBIDE-SUI DIVISION
19
•
IW0000078
CARBIOE-SUI DIVISION
22
•
W0000078
CARBIOE-SUI DIVISION
23
•
W0000078
CARBIDE-5UI DIVISION
24
•
W0000078
CARBIDE-SUI DIVISION
28
•
W0000078
CARBIDE-SUI DIVISION
29
•
W0000078
CARBIDE-SUI DIVISION
32
•
W0000078
CARBIDE-SUI DIVISION
S3
•
W0000078
CARBIDE-SUI DIVISION
74
•
W000007B
CARBIDE-SUI DIVISION
80
•
W0001066
APPALAQ4IAN 1066-KANAUHA
01
0.04312
W0001074
APPALACHIAN 1074-AWS
01
2.54250
W0001074
APPALA04IAN 1074-AWS
03
0.60790
W0001074
WPALA04IAN 1074-AWS
04
0.13899
W0001074
APPALAOHAN 1074-AWS
11
0.03028
UV0001074
APPALAOUAN 1074-AWS
27
0.06057
UV00016S1
FIKE OOICAL-C.S.T.
01
0.01727
W0001716
PBtS OCMICAL
01
0.22908
W0002313
DIAfOC 94AM0CX OCNICALS CO.
03
0.055%
W0002381
HATFIELDtfCNSON
01
0.13370
W0002399
DUOTT -flELLE
017
•
W0002399
DIPONT -flELLE
17
0.00000
MV00Q2399
DUPWT -flELLE
62
0.29572
W0023116
S. CHARLESTON POTU
01
0.42813
P01UTMT
4.60603
YEAIW4 P01UTWT
-------
YEAR*84 POLLUTANT
-------
YEMM4 POLWTANT«SELENIUN
CODE FACILITY PIPE KB)
UV0000078 CARBIOE-SUI DIVISION 10 .
IW0001074 APPALACHIAN 1074-AWS 01 3.36099
UV0001074 APPALACHIAN 1074-AMS 04 0.01817
UV0001074 APPALACHIAN 1074-AWS 11 0.01514
UV0001074 APPALACHIAN 1074-AWS 27 0.00606
W0002399 OUPOKT -BELLE 62 .
POLLimwr 3.62036
YEAR-84 PQLLUTANT>SILVER
CODE FACILITY PIPE KB
W0000078 CMBIDE-SUI DIVISION 17 .
W0000078 CARBIDE-SUI DIVISION 18 .
UV0000078 CARBIDE-SUI DIVISION 24 .
IW0000078 CARBIDE-SUI DIVISION 28 .
UV000007B CARBIDE-SUI DIVISION 29 .
W0000086 CAUBIDE-II6TITUTE t)l 0.0142596
POLLUTANT 0.0142596
6-52
-------
YEMW4 POLLUTANWINC
CODE FACILITY PIPE NED
WI623382 SMITH FflSTDCR 01 0.036
W0000078 CARBIDE-SUI DIVISION 16
UV0000078 CARBIDE-SUI DIVISION 17
KV0000078 CARBIDE-SUI DIVISION IB
UV000007B CARBIDE-SUI DIVISION 19
W000007B CARBIDE-SUI DIVISION 22
W000007B CARBIDE-SUI DIVISION 23
W0000078 CARBIDE-SUI DIVISION 24
W0000078 CARBIDE-SUI DIVISION 28
W000007B CARBIDE-SUI DIVISION 29
W000007B CARBIDE-SUI DIVISION 32
W000007B CARBIDE-SUI DIVISION S3
W000007B CARBIDE-SUI DIVISION 74
W0000078 CARBIDE-SUI DIVISION 80
W0000086 CARBIDE-INSTITUTE 01 0.156
W0000086 CARBIDE-INSTITUTE 02 1.3S2
1*0000086 CARBIDE-INSTITUTE 03 0.764
UV0000167 ELKEM fCTflLS 06 0.000
W0001066 APPALACHIAN 1066-KANAUHA 01 0.276
W0001074 APPALACHIAN 1074-4N0S 01 1.289
UV0001074 APPALACHIAN 1074-AN0S 03 0.241
UV0001074 APPALACHIAN 1074-AN0S 04 0.731
W0001074 APPALACHIAN 1074-AN0S 11 0.174
W0001074 APPALACHIAN 1074-AN0S 27 0.167
W0002313 DIAMOND SHANROtt OGIICALS CO. 03 0.108
W0002381 HATFIELD1KNSGN 01 0.128
W0002399 DUPONT -BELLE 017
W0002399 DUPONT -BELLE 046
UV0002399 DUPONT -BELLE 62
W0002342 ALLIED OOICAL 03 0.341
POLLUTANT 3.764
6-53
-------
Chapter Seven
Hazardous Waste Analysis
-------
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
I. INTRODUCTION AND PURPOSE
The Kanawha Valley Toxics Screening Study has four stated objectives that
have provided overall direction to this technical report. In this hazardous
waste site analysis, we attempted to address these objectives to provide a
screening analysis for policy development purposes. The four objectives are:
• Using available data bases, identify many of the chemicals
routinely released or present within various exposure pathways;
• Develop a sense of the potential public health concerns these toxic
pollutants may pose in various pathways, based on health effects and
exposure information;
• For a select number of chemicals, provide an initial, conservative
assessment of potential cancer risk and potential noncancer health
risk for predicted or observed concentrations within exposure
pathways; and
• Identify data and information gaps and outline needs and options
for future study directions to enable a more detailed investigation
of health issues where warranted.
To meet the first objective, we have reviewed available information for
hazardous waste sites within the study area to determine the types of
chemicals present. To develop a sense of the potential public health concerns
these toxic pollutants may pose, we developed a general characterization of
the hazardous waste sites within the Kanawha Valley. By organizing sites
according to hydrogeology, waste, and facility classes, we developed general
categories of waste sites. From these general categories we modeled several
site scenarios and estimated potential ground-water contamination and
potential risk posed by such contamination. It is important to emphasize that
we modeled scenarios representative of sites within the study area, and not
specific hazardous waste sites. The limited information available does not
allow an accurate assessment of potential cancer risk from specific sites,
even for a select number of pollutants. Through our analysis we can only
develop a general sense of potential public health concerns that these toxic
chemicals may pose in the Kanawha Valley. Finally, to meet the fourth
objective we outlined information gaps where more data are needed.
The remainder of the report is divided into five sections. Section II
describes the basic site characterization and data collection, and Section III
describes the development of hydrogeology, waste, and facility categories. In
the fourth section, we discuss the quantitative modeling including:
development of representative generic scenarios, methods for exposure and risk
modeling, and the results of the analysis. Section V contains a discussion of
the limitations of the analysis and identifies areas in which further study is
needed; this section also contains a summary of uncertainty issues related to
ground-water modeling. The final section presents conclusions for
7-1
-------
this phase of the analysis. There are three technical appendices to this
report: Appendix A describes in detail our characterization of the
hydrogeology of the Kanawha Valley, Appendix B provides a detailed discussion
of uncertainty issues in the ground-water modeling, and Appendix C lists our
main sources of information.
Limited time and resources required this study to focus solely on
potential ground-water contamination. Although we do not review other
exposure pathways such as those resulting from chemical volatilization and
run-off from hazardous waste sites, these pathways should be considered in a
comprehensive assessment of potential effects from hazardous waste sites.
II. HAZARDOUS WASTE SITE CHARACTERIZATION
We characterized sites according to hydrogeologic setting, exposure
pathway, and predominant waste types, and grouped sites in three general
hydrogeologic categories, five facility types, and four waste categories.
While there are a large number of sites in the study area (numerous RCRA
generators, approximately 20 RCRA treatment, storage, and disposal (TSD)
facilities, and many potential CERCLA sites), we were able to obtain
information relevant to risk assessment for only 25 facilities. These
facilities include 8 RCRA TSD facilities, 7 RCRA generators, and 10 potential •
CERCLA sites. Our characterization of these facilities is summarized in Table
1. Figure 1 is a map of the study area showing the locations of the 25 sites.
As indicated in Table 1, each site was identified by name and location,
and an attempt was made to classify each site according to whether it was a
RCRA or CERCLA facility. While this was obvious in most cases, some sites are
active or inactive RCRA facilities that are under investigation as potential
CERCLA sites (e.g., Occidental, FMC Corp.). In one case an apparently active
RCRA facility, Artel, is also on the CERCLA National Priorities List (NPL).
In our characterization we assigned to the RCRA category facilities that had
applied for a RCRA permit or were known to generate RCRA wastes; sites used
for the dumping of wastes (e.g., municipal landfills) were classified as
potential CERCLA sites. Using this methodology, Artel is considered a RCRA
site in our characterization.
Based on its location, we assigned each site to one of three general
hydrogeologic settings that we developed for the region: (1) Sandstone/
Shale-Clymer/Dekalb/Gilpin soil setting, found in the southeast portion of the
study area; (2) Shale/Sandstone-Gilpin/Upshur/Vandalia Soil setting found in
the northwest portion of the study area; and (3) Alluvium-Kanawha Soil
setting, found along the Kanawha River and its major tributaries. These
hydrogeologic settings are described in greater detail in Appendix A,
"Hydrogeologic Characterization of the Kanawha Valley," and in Section III of
this report.
7-2
-------
TABLE 1
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZ ARDOUS WASTE SITES
RCRA AND POTENTIAL CERCLA SITES WITH SUFFICIENT INFORMATION FOR PRELIMINARY RISK ANALYSIS
SITE NAME
LOCATION RCRA/CERCLA
HYDROGEDLOGIC SETTING
POTENTIAL EXPOSURE PATHWAY
PREDOMINANT WASTES
CONSENTS
1. Heizer Creek Dump Nitro, WV
CERCLA
2. South Charleston South CERCLA
Municipal Landfill Charleston, WV
Shal e/Sand tone - Gilpin-
Upahur-Vandal ia (GUV)
Soils
Shale/Sandstone-GUV Soils
Consumption of contaminated
ground water-private wells
Consumption of contaminated
ground water-private wells
Phenol; N-nitrosodiphenylamine; Former disposal
cresol; tetrachlorobenzene; area for indus-
trichlorobenzene; methylene trial wastes;
ch lor ide ; 2,4 ,6-tr ich lorophenol; ground-water
cyanide; arsenic (possibly i users currently
dioxins) exposed
Dichlorobenzene; methylene
chloride; aldrin; PCBs; MEX;
phenol; cresol; toluene;
ch lordane
Possible expo-
sure at present
time to methylene
chlor ide ,
chlordane; for-
mer industrial
waste disposal
site
3. Smith Creek Dump South CERCLA
Charleston, WV
Shale/Sandstone.-GUV Soils
4. Elk em Hetals
5. Occidental
Alloy, WV
Belle, WV
RCRA Alluv ium-Kanawha Soils
generator
RCRA Alluvium-Kanawha Soils
generator/
CERCLA
Consumption of contaminated
ground water-private wells
Ground-water discharge to
Kanawha River-may affect
Boomer water supply
Ground-water discharge to
Kanawha River
Bis ( 2-chl oroe thoxy )rae thane;
naphthalene; PCBs; chromium;
lead; cyanide; TCE
Chromium; copper; lead; zinc
Former indus-
trial waste
disposal site
(near South
Charleston Muni-
cipal Landfil I)
Carbon tetrachloride; chloro- Former surface
form; methylene chloride; 2,4- impountfcnent;
and 2,6-dinitrotoluene; mercury; ground-water
chromium; hexachloroethane; pumping
hexachlorobenzene and treatment
in progress
6. E.I. DuPont
7 . Un ion Carbide
Belle, WV RCRA All uv ium-Kanawha Soils
South RCRA
Charles ton,WV
All uv ium-Kanawha Soils
Ground-water discharge to
Kanawha River
Ground-water discharge to
Kanawha River
Methylene dianil ine; nitroeo-
amines; methacrylate; alcohols;
thorium; arsenic; hexamine
Acrylonitrile; benzene; vinyl
chloride a/
Inactive disposal
ar eas
8. FWC Corp., Indus-
trial Chemicals
South RCRA/
Charleston, WV CERCLA
Alluv ium-Kanawha Soils
Ground-water discharge to
Kanawha River
Mercury; cadmium; lead; cyanide; Inactive land-
lindane; carbon tetrachloride; fills at facility
chloroform; carbon disulfide;
MEX; phenol; cresol;
trich I orobenzene
-------
TABLE 1 (Continued)
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
RCRA AND POTENTIAL CERCLA SITES WITH SUFFICIENT INFORMATION FOR PRELIMINARY RISK ANALYSIS
SITE NAME
LOCATION RCRA/CERCLA
HYDROGBOLOGIC SETTING
POTENTIAL EXPOSURE PATHWAY
PREDOMINANT WASTES
COMMENTS
9. Union Carbide, South RCRA
Technical Center Char les ton (WV
Sands tone/Shale-Clymer-
Dekalb-Gil pin (CDG)
Soils
10. Rhone-Poulenc Institute, WV RCRA Alluvium-Kanawha Soils
Consumption of contaminated
ground water-private wells
Grounds-water discharge to
Kanawha River
Highly variable as site is
a research and development
facil ity
An thra cen e; chl oro for m;
chlorobenzene; fluoranthene;
N-nitroeodiphenylamine; 1,1,
1-TCA; carbon tetrachloride;
toluene; ethylbenzene;
methylene chloride
Holz Pond
(surface im-
poundment )
located nearby
t
Includes Golf
Mt . Landfil 1;
Private Truck-
ing Operations
(PTO) located
nearby; PTO has
docutnen ted
ground-water con-
tamination
11. Chemical Leaman Institute, WV RCRA All uv ium-Kanawha Soils
Tank Lines generator
12. Big Scary
Creek Pond
Nitro, WV
CERCLA Alluviura-Kanawha Soils
13. FMC Corp., Nitro, WV RCRA
Specialty Chemical
All uv ium-Kanawha Soils
Ground-water discharge to
Kanawha River
Ground-water discharge to
Kanawha River
Ground-water discharge to
Kanawha River
Phenol; chromium
Zinc; unknown organics and
heavy metals
1,1,1-TCA; arsenic; phenol;
chromium
Tank cleaning
facility; no
on-site disposal
Viscose rayon
plant ;inactive
disposal sites
(unauth or ized )
at facility
2 inactive
landfil Is
on-si te
14. Monsanto Company Nitro, WV RCRA All uv ium-Kanawha Soils
Ground-water discharge to
Kanawha River
2,4,5-T; dioxin; acrolein;
TCE; PCE; cyanide; toluene;
xylene; chlorobenzene; heavy
metals
3 inactive
landfills
on-si te;
documented
ground-wa ter
con tamination
15. Artel
Nitro, WV RCRA Alluvium-Kanawha Soils
(permi t
denied) ;
CERCLA
(National
Priorities
List) ;
Ground-water discharge to
Kanawha River
Benzene; bis( 2-chloroe thyl) 4 "evaporation"
ether; bis (2-chloroisopropyl ) ponds; unlined
ether; chloroform; 1,2-dichlo- burial pits;
ethane; vinyl chloride; arsenic; highly conta-
e thylbenzene; PCE; carbon minated ground
tetrachloride; PA Ha; DDE; DDD; water
heptachlor; lindane; mercury;
isophorone; lead; cadmium;
cobalt; nickel
-------
TABLE I (Continued)
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
RCRA AND POTENTIAL CERCLA SITES WITH SUFFICIENT INFORMATION FOR PRELIMINARY RISK ANALYSIS
SITE NAME LOCATION RCRA/CERCLA HYDROGBOLOGIC SETTING POTENTIAL EXPOSURE PATWAY PREDOMINANT WASTES COCMENTS
16. Coastal Tank
Lines , Inc.
17. Chem Formula tors,
I nc.
Nitro, WV
Nitro, WV
RCRA All uv ium-Kanawha Soils
generator
CERCLA
Alluvium-Kanawha Soils
Ground-water discharge to
Kanawha River
Ground-water discharge to
Kanawha River
Phenol; chromium
Me thoxychl or; chloroneb;
phenol; maleic hydrazine;
anisole
Tank truck
cl ean ing
facility
Sm4H pesticide
manufacturer w/
on-site contami-
ation
18. Mason & Dixon
Tank Lines, Inc.
19. Nitro Sanitation
St. Albans, WV RCRA All uv ium-Kanawha Soils Ground-water discharge to
generator Kanawha River
Nitro, WV CERCLA AH uv ium-Kanawha Soils Ground-water discharge to
Kanawha River
Phenol; chromium
Arsenic ; phenol; heptone
Tank truck
cleaning facility
Mun i c i pa 1 1 and -
fill that
accepted
hazardous waste
20. Gary Harris Dump Lanham, WV
CERCLA
Shale/Sandstone-GUV Soils Consumption of contaminated
ground water-private wells
21. Western Kanawha
County Landfill
Crosslanes, WV CERCLA
Shale/Sandstone-GUV Soils Consumption of contaminated
ground water-private wells
1,1,1-TCA; benzene; ethylben-
zene; dinitrotoluene; carbon
tetrachloride; chloroform;
toluene; benzo(a )pyrene; PAHs
Heavy metals; phenol; acrolein;
toluene; MEK wastes; generally
not well characterized
Uncontrolled
disposal area
(several hundred
drums); high
potential for
exposure via
ground water
Very high expo-
sure potential;
Fike and other
compan ies
allegedly
disposed of
chemical waste
here
22. Poca Stripmine
Drum Site
Poca, WV CERCLA Shal e/Sands tone-GUV Soils Consumption of contaminated
ground water-private wells
1,1,1-TCA; cyanide; zinc;
chromium; nickel; toluene;
methylene chloride;
is opropanol
"Midnight
dumping"; emer-
gency removal of
most of the
wastes has been
carried out
23. Vimasco Corp.
Nitro, WV RCRA All uv ium-Kanawha Soils
generator
Ground-water discharge to
Kanawha River
Arsenic; cadmium; lead; mercury; Uhl ined burial
zinc; PCBs; toxaphene; xylenol; pit and three
phenol; MIBK; 1,1,1-TCA; carbon waste ponds
tetrachloride; ethylbenzene; on-site
benzene; toluene
-------
TABLE I (Continued)
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
RCRA AND POTENTIAL CERCLA SITES WITH SUFFICIENT INFORMATION FOR PRELIMINARY RISK ANALYSIS
SITE NAME
LDCATION RCRA/CERCLA HYDROGEOLOGIC SETTING POTENTIAL EXPOSURE PATHWAY
PREDOMINANT WASTES
COMMENTS
24. Monday(Mundy)
Hollow Landfill
Poca, WV CERCLA Shale/Sandetone-GUV Soils Consumption of contaminated Phenol; arsenic
ground water-private wells
84 ground-water
users identified
within 1/2 mile
of facility;
inun|icipal land-
fill that
accepted hazar-
dous waste
25. Markay Chemical
Co.
St.
Alb ana, WV RCRA
generator
Alluvium-Kanawha Soils
Ground-water discharge to
Coal River
Chiorobenzene; 1,2-dichloro-
propane; carbon tetrachloride;
TCE; chloroform; methylene
chlor ide
Facility operated
for 8 years w/o
RCRA notification
(illegally);
halogenated
solvents stored
in unlined
impoun dment
-J
1
^ a/ These chemicals are used or produced at the facility, and have been suggested as indicator chemicals for ground-water monitoring at the facility.
However, it is not known whether these chemicals have actually been disposed of on-site.
-------
f /r»uff»io i a \
/ i .... ft -N • . r / x. .. \
/ - ' ' . ' ,:7~ ¦
•- vS.; •vy-^\<.
Sgr* \ ".. *';s .-V V*'•• \;
FIGURE 1
MAP OF THE KANAWHA VALLEY STUDY AREA
X(>) *ic*otr
V- c.';V-
\\ • I IV «
V v ';>i> «l <*• ;;A> V O-
" NX- ~ V\' ?1 %-\ 'W. TV
N*" .'^Vy.Hofflfiijwi. . /)- ( '¦ —• / P s""'
k X \\. 'v •• Sl-qL- '! — "/~ ' - ' I I jr; Jj^vr--* f.
-. . * V^KPtrmo^~V • . # , © * I 1 -».%•> i- c .:v
« * " "«\Vi •.'n'A - * » * . •. IS* • v-"r . T
^5 *j j | ^ ^ P«*'»t'C3Vl .C.tf ^., J
\ i ' dfW v ' ' r v ^
f.°" l7 §&' *A V-' " : ~ ; ,0'/ ' V -
tew&JF /"%"• -ML,-.' LV- A
;'/•
t" W»lluiA
OilA/fjf
*NO» .
• • . . /< • v - ¦¦.. .. (.
.P.MM11. -o„..«,ii • »Que*n S*ioalf; ,
... \w»'r ; /V Y - R3U:V
,,0R wtHi '%/ ¦ » /^' ' .&o™«r » \
vr .# . 60.,~»h CKvid*
I
M.llcreili
^C>fXRlESTON :
* /
Sindeison'
o'o
RO»in*c 0»0
Ktndaha*
fo»>a«y "-* . .»
1 09/ " G*i''Cllt Bind
' C°
r <> j y j
So*^-- /»5 ^c
V
1" • Vu)k«l|lM OnhKl
1 M.|1 StkMl
• ' **X - /
,- .--"..''i- ^ —r~^l , .!?*' Tjh/ ¦ , ^vCinct) /
^ . ^-N South-* . ' & V-— ' -•• ¦/„..'¦
* , ' ....... Charl,:"°n
J«"'« .lh»» Souln f,,!. F«'k - °S,. .. . ! ' > V. • N 1
>P(y( A t v V '' • -\
v-- yt • ,tr7.- y~ . ;xt p*tPeY ^ \V 1
'. ,' ' . ,'^oiN, Biii..cr\
^ en»r *
Tr«»d . t '
/ -. . • -tlM. . ¦
is69 .,;
/
'M.O
cr"
- * >> i -: . 1 ' _w
,. • f • c AJwmvillt • - r "\ -3I
' i vT v.' . -¦¦. -r 1 ".•¦ I
i«o *v
tr«OI V/
Hitop '
G'iM>thi»ili«
*«utMr
/ w>
KNOl ;-3uT-f:0
f»•* , ^
. w And " n:; ~ .
..1- , M»m«»olh^/-V
* -- r" .' 1 /*"" .• »//
r-'"'. " W¦ '
ClMlKf
CfciTci > ~~s ^Holi/t>o»fct T^>koir-' •' ^->
- - fvL A-x, ,-oi - /
"•"<*^ > _i -^; c»«c..,.o.».<» .CVT|Cf.\(-. ,A» /" - : >8^ .
' • 1 1 • .'• ^' r>Ch*»4p««Qijffiond a
«.«..) / "{- - C"ck '"xE«,
.tllini: ; . V* it* ^ • MOItr*!
Sj»d Fork
ChvlcJi |«4 5thf5l ^Sro
Emmocsi*
LEGEND
m Hazardous Waste Site
X Surface Water Intake
, tit • C««»Cif«k / 1 •
V. V 7 ' f y rSckMI*. J
S C*.-.»0 ^ /V^Reck Creek " ^
lutj
- )
¦' (CR0'fcC'e^ - L.rk *
..iMotikios J V 1 —/ (^ovlei
:• Civet . ' ^^D*n < - r
Hu(ltnla«
H«ndl
O.V e,.nt^\_-v» ; A-<
lend. -> G""«h-'
cA
K\ D*«.«0| L»»in»Uo*.\
l . ^1-.' y tsi»(«)jfd
'.jrOo^dile
"»b«t* ( ^
Vjneitj
«e ' C«yle|r tridg
® ^ ;
«—*- * Gl.n F.ri.i
, V • /
F 0 »»e*»i 1 te^%
Ouih t'lid
^" V'^*ep W*1,
'• .*/ColuT>t>»3
• JOMHiOp
4, ' INOI
\Oukrirr
. \*\ »\ V ",l \ ^
- v\ 'ANorlh CiOort \*M^f *
¦b-k- XsX"*l
Robton
,.&Gf
( An»r
^ ,,}J'
B(<)A(tDO' ou(n
cofroN
Cotton Hill
Beck*
let Mjiwf
An»r«d /665 ,ae,
— ¦"T
^flf»'d» Fo
-------
In evaluating potential human exposures via ground water, we considered
not only the direct pathway of release, subsurface transport, and consumption
of contaminated ground water, but also identified the ground water to surface
water migration pathway that exists at many of the sites close to the Kanawha
River. —
Finally, we attempted to identify the predominant waste types and amounts
disposed of at each site. As is frequently the case, this waste
characterization proved to be extremely difficult for several reasons:
(1) many RCRA facilities perform only the minimum characterization required by
law for much of their waste (e.g., determinations of whether the waste is
ignitable, EP toxic, corrosive, or reactive); (2) many CERCLA sites have not
been extensively sampled; and (3) based on quality assurance (QA) reports,
many of the sample analyses from CERCLA sites were suspect. Thus, waste
information on the facilities was either missing or substandard in many cases.
III. SITE CLASSIFICATION
We classified the 25 hazardous waste study sites on the basis of
hydrogeological setting, facility type, and predominant waste types. These
three parameters are important in determining the likelihood of a facility
failure, the transport of the pollutant through the ground water, and the
potential health effects posed by the transported pollutant if consumed. The
classification of study sites according to these parameters provides a
qualitative assessment of the potential for ground-water contamination. These
general classifications are described below.
Hydrogeologic Setting
Our evaluation of the hydrogeology in the Kanawha Valley and the
development of hydrogeologic scenarios are described in detail in Appendix A.
Thus, the discussion here focuses mainly on the hydrogeologic settings as they
relate to the sites listed in Table 1.
Table 2 presents brief descriptions of the three general hydrogeologic
settings we developed for our analysis, and lists the hazardous waste study
sites that are found in each of the three settings. As illustrated, all of
the RCRA facilities except one and several of the most notable potential
CERCLA sites are located in the Kanawha Soil/Alluvium hydrogeologic setting.
There is probably no direct ground-water exposure pathway for chemicals
released from these sites because: (1) the ground water discharges directly to
surface water within a short distance, and (2) people in the Kanawha Valley
proper generally do not use ground water for potable water, but rather are
served by public water supply systems, which obtain their water from
tributaries to the Kanawha River. However, these facilities may be a source
of surface water contamination via the pathway of contaminated ground-water
discharge.
7-8
-------
TABLE 2
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
HYDROGEOLOGIC SETTING CATEGORIES
IN THE KANAWHA VALLEY
HYDROGEOLOGIC SETTING
DESCRIPTION a/
HAZARDOUS WASTE SITES IN SETTING
CIyme r-DekaIb-G iIp i n
Soil/Sandstone-Sha le
Bed rock
Well drained, acid soils.
Bedrock is predominantly sandstone
with interbedded shales
(Pottsville and Allegheny Groups).
Ground-water flow is unconfined b /
through fracture systems.
Characteristic of SE mountainous
portion of study area.
Union Carbide Technical Center
-j
I
2. GiIpin-Upshur-VandaIia
SoiI/Shale-Sandstone
Bedrock
Moderately to slowly permeable
silt loam and siIty clay loam
soils. Bedrock is predominantly
shale with interbedded sandstone
(Conemaugh, Monongehala, and
Groups). Unconfined flow
through fracture systems at 300-400
meters/yr. Characteristic of NW
mountainous portion of study area.
Heizer Creek Dump; Gary Harris
Dump; Western Kanawha County
Landfill; Poca Stripmine Dump
Site; Monday (Mundy) Hollow
Landfill; South Charleston
Municipal Landfill; Smith
Creek Dump
3.
Kanawha Soil/
AI Iuvi um
Moderately drained to well-
drained lime influenced soils.
Alluvium consists of sand, silt,
gravel, and clay; upper layer
of sandy clay with underlying
sand and gravel; many lenses and
small perched water tables. Total
thickness of alluvium is about
60 feet with water table at
about 15 feet below surface.
Ground-water flow is unconfined
through porous media.
EIkem Metals; Occidental;
E.I. DuPont; Union Carbide South
Charleston; FMC South Charleston;
Rhone-Poulenc; Chemical Leaman
Tank Lines; Big Scary Creek
Pond; FMC Nitro; Monsanto Company;
Artel; Coastal Tank Lines; Chem
Formulators; Mason and Dixon Tank
Lines; Nitro Sanitation; Vimasco;
Markay Chemical Co.
a/ For more details see Appendix A, "Hydrogeologic Characterization of the Kanawha Valley." Information was
gathered from USGS and state reports, RCRA Part B Applications, CERCLA studies, and academic studies (see
reference list given in Appendix C).
b/ An unconfined aquifer is one whose upper surface is a water table free to fluctuate under atmospheric
pressure.
-------
Facilities that are located in the mountainous areas outside the main
river valley are generally potential CERCLA sites. Of these, most are located
in the Gilpin-Upshur-Vandalia Soil/Shale-Sandstone hydrogeologic setting,
which covers the northwest portion of the study area. Only one study site
(Union Carbide~Technical Center) is located in the Clymer-Dekalb-Gilpin Soil/
Sandstone-Shale hydrogeologic setting, although two other sites (Diamond
Shamrock, E.I. Dupont) are located where the alluvium is not well developed
over the Sandstone-Shale hydrogeologic setting. People living outside the
main river valley could use ground water for their domestic supply; thus, a
direct exposure pathway may exist at some of the facilities in these areas.1-1
Facility Type
An obvious division can be made for the hazardous waste sites in the
Kanawha Valley by separating them into RCRA and potential CERCLA sites.
However, as noted earlier, some sites are active or inactive RCRA facilities
that are also under CERCLA investigation. Also, there is a significant amount
of diversity among specific sites in each of these two categories. Thus, we
believed that further facility classification was necessary.
We classified RCRA facilities into two major types: (1) permitted
treatment, storage, and disposal facilities, and (2) generators who do not
treat or dispose wastes and do not store wastes more than 90 days. For
potential CERCLA sites, we identified three major categories: municipal
landfills that accepted hazardous industrial wastes, inactive chemical plants,
and uncontrolled disposal areas2-1 (e.g., "midnight dumping" areas, open drum
dumps).
These five categories, the typical kinds of release sources for each
category, and the sites that are included in each category, are presented in
Table 3. It should be noted that these categories were developed to represent
the major types of hazardous waste sites in the Kanawha Valley for which there
was available information. The fact that one category contains fewer sites
does not necessarily indicate that it is less significant; rather, it reflects
our lack of information on the majority of hazardous waste sites in the
Kanawha Valley. For example, we have information indicating that the Dutch
Hollow Drum Site near Poca is an open dump containing 600 to 1000 drums, but
have no information that will allow us to characterize the wastes at this
1J In our analysis, we were able to identify ground-water users in proximity
to several hazardous waste sites in the shale/sandstone and sandstone/shale
hydrogeologic settings. Based on Superfund preliminary assessment reports,
there is evidence of private well contamination in two cases (Heizer Creek
Dump and South Charleston Municipal Landfill). However, an in-depth
evaluation of potentially exposed populations and current ground-water
contamination was beyond the scope of this analysis.
2J This category is by necessity somewhat of a catch-all group but mainly
addresses sites at which there appears to have been no effort whatsoever to
prevent release or migration of wastes.
7-10
-------
TABLE 3
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
GENERAL FACILITY TYPES FOR HAZARDOUS WASTE
SITES IN THE KANAWHA VALLEY
FACILITY TYPE
TYPES OF
RELEASE SOURCES
SITES
RCRA treatment, storage,
and disposal facilities'
RCRA generators
Potential CERCLA sites -¦
municipal landfills that
accepted hazardous
industrial wastes
Potential CERCLA sites -•
uncontrolled disposal
Potential CERCLA sites -¦
inactive chemical plants
Regulated land dispo-
sal units, such as
landfills and surface
impoundments (lined
or unlined); also, some
old landfills and dump-
ing areas (solid
waste management units,
or SWMUs)
Unregulated release
sources, such as old
landfills and dumping
areas
Municipal
landfills (unlined)
Uncontrolled drum
dumping areas
Old landfills and
dumping areas
Monsanto Company
E.I. Dupont
Union Carbide South Charleston
FMC South Charleston
Union Carbide Technical Center
Rhone-Poulenc
FMC Nitro
Artel
Markay Chemical
Elkem Metals
Occidental
Coastal Tank Lines
Chemical Leaman Tank Lines
Mason and Dixon Tank Lines
Vimasco Corp
Heizer Creek Dump
South Charleston Municipal
Landfill
Smith Creek Dump
Nitro Sanitation
Western Kanawha County Landfill
Monday (Mundy) Hollow Landfill
Poca Stripmine Drum Site
Gary Harris Dump
Chem Formulators
Big Scary Creek Pond
7-11
-------
site. As a result, this potentially important site was excluded from further
consideration in this study. Thus, given the available information, it would
be difficult to estimate the relative importance of each of the five
categories.
Waste Type
As expected, the hazardous waste sites we characterized reflect the
diversity of chemicals disposed in the Kanawha Valley. A myriad of organic
and inorganic chemicals are produced, used, and disposed of in the study area,
resulting in a wide variety of chemicals being identified at both RCRA and
potential CERCLA sites. A list of the chemicals identified at hazardous waste
sites in the study area is presented in Table 4.
After identifying the chemicals present, out next step was to group sites
into categories of similar waste types and to designate a set of model
constituents (i.e., representative chemical constituents to be modeled) for
each waste type. Given the screening nature of the analysis, we minimized the
number of waste groupings. Thus, the four groups presented in Table 5 are
probably the minimum number into which the 25 sites could be divided while
accounting for real differences in waste characteristics that have been
observed at the sites.
To group sites into general waste categories, two methods were used.
These were: (1) grouping by professional judgment, based on the presence of
one or two chemicals at high concentrations; and (2) grouping through the use
of a waste/site matrix, in which all of the 25 sample sites included in the
analysis were listed along one axis and all chemicals identified at these
sites were listed along the other axis. Sites with similar distributions of
chemicals were then placed in the same group.
Although the second method would appear to be more objective, both
classification approaches necessarily relied heavily on professional judgment,
due to the extremely diverse nature of the wastes found at the sites included
in the analysis. Thus, both methods used several predominant chemicals as a
basis for grouping sites into general categories, with one or more surrogates
selected to represent other chemicals found at sites in each group.
To identify model constituents for each group of Kanawha Valley sites, we
used the following criteria (roughly in descending order of importance):
• Number of sites in a group in which the chemical was
identified (i.e., frequency of occurrence);
• Concentration of the chemical at the sites at which it
was identified;
7-12
-------
TABLE 4
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
PREDOMINANT WASTE TYPES AND CHEMICALS
FOUND AT HAZARDOUS WASTE SITES
IN THE KANAWHA VALLEY
WASTE TYPE
CHEMICALS
Halogenated aliphatic solvents
Industrial intermediates, solvents
and products
Inorganics
Methylene chloride
1,1,1-Trichloroethane
Trichloroethylene
Carbon tetrachloride
Chloroform
Tetrachloroethylene
1,2-Dichloroethane
1,2-Dichloropropane
Phenol
Cresol
Benzene
Acrylonitrile
Toluene
Methyl ethyl ketone
Chlorobenzenes
Chlorophenols
Vinyl chloride
Methacrylate
Carbon disulfide
Acrolein
Ethylbenzene
Dinitrotoluene
Methylene dianiline
Arsenic
Mercury
Lead
Cadmium
Chromium
Zinc
Cyanide
Nickel
7-13
-------
TABLE 4 (continued)
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
PREDOMINANT WASTE TYPES AND CHEMICALS
FOUND AT HAZARDOUS WASTE SITES
IN THE KANAWHA VALLEY
WASTE TYPE
CHEMICALS
Chlorinated pesticides
Chlordane
Methoxychlor
DDE
DDD
Dieldrin
Aldrin
Heptachlor
2,4,5-T
Lindane
Highly persistent and toxic
by-products
PCBs
Dioxin
Hexachloroethane
Hexachlorobenzene
Industrial carcinogens
Bis(2-chloroethoxy)methane(BCEXM)
Bis(2-chloroethyl)ether(BCEE)
Bis(2-chloroisopropyl)ether(BCIE)
N-Nitrosodiphenylamine
Chlorinated azo compounds
Polynuclear aromatic compounds
Naphthalene
Benzo(a)pyrene
Fluoranthene
Chrysene
Acenaphthene
Carbamate pesticides
Aldicarb
Carbaryl
7-14
-------
TABLE 5
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
GENERAL WASTE CATEGORIES
FOR HAZARDOUS WASTE SITES
IN THE KANAWHA VALLEY
WASTE CATEGORY NUMBER MODEL CONSTITUENTS HAZARDOUS WASTE SITES
Carbon
tetrachloride
Mercury
2,4-Dinitrotoluene
Methylene chloride
Cyanide
Phenol
Chlordane
Occidental; Artel; Union Carbide
South Charleston; Rhone-Poulenc;
Union Carbide Technical Center;
Gary Harris Dump; Markay
Chemical Company; Vimasco
Corp; FMC-South Charleston
Chem Formulators; Heizer Creek
Dump; Smith Creek Dump; South
Charleston Municipal Landfill;
Poca Stripmine Drum Site;
Western Kanawha County Landfill;
Monsanto
Phenol
Arsenic
Chemical Leaman Tank Lines;
Mason and Dixon Lines; Coastal
Tank Lines; Nitro Sanitation;
FMC Nitro; Mundy Hollow
Landfill; E.I. DuPont
Chromium
Zinc
Elkem Metals; Big Scary Creek
Pond
7-15
-------
• Toxicity of the chemical, with potential carcinogens
weighted more heavily than noncarcinogens3 (carcinogens
were scored according to potency values developed by EPA's
Carcinogen Assessment Group); and
• Mobility and persistence in ground-water systems.
Waste categories and model constituents are described below. For identi-
fication of which sites fall into each general category, refer to Table 5.
Category 1: In this category, carbon tetrachloride, a highly toxic and
relatively mobile chlorinated solvent and chemical intermediate, was used as
the predominant chemical for grouping hazardous waste sites. Carbon
tetrachloride was present at all of the sites in this group at significant
concentrations; however, the other two model constituents in this scenario,
mercury and 2,4-dinitrotoluene, should be considered surrogates for a variety
of chemicals found at these sites. Mercury was selected based on its
appearance at several of the sites and its high toxicity, and was chosen as a
surrogate for other toxic inorganic contaminants found in lower concentrations
or less frequently at sites in this group (e.g., cadmium, chromium, cyanide).
2,4-Dinitrotoluene was identified at two of the sites in this group, and is a
potent carcinogen that is relatively mobile in ground water. Thus, 2,4-dini-
trotoluene was considered to be a surrogate for other relatively mobile
carcinogens, such as acrylonitrile, benzene, and N-nitrosodiphenylamine, that
were found at some sites in this group.
Category 2: Sites included in this category were extremely diverse, but
were grouped mainly on the basis of three classes of contaminants:
(1) chlorinated solvents and intermediates (e.g., methylene chloride);
(2) inorganics (e.g., cyanide); and (3) highly persistent chlorinated aromatic
compounds (e.g., chlordane). These classes of chemicals were identified at
essentially all of the sites in this group. The most common chlorinated
solvents and intermediates identified were 1,1,1-trichloroethane, methylene
chloride, and di-, tri-, and tetrachlorobenzenes. Inorganics included
cyanide, arsenic, chromium, lead, and nickel, while highly persistent
chlorinated organics included PCBs, dioxin, chlordane, heptachlor,
methoxychlor, lindane, and DDE. In addition to these compounds, phenols and
cresols were identified at many of the sites. Based on the criteria cited
earlier (i.e., frequency of occurrence, concentration, toxicity, mobility, and
persistence), we selected methylene chloride, cyanide, phenol, and chlordane
to represent the wastes found at sites in this group.
3 Potential carcinogens were weighted more heavily in the selection of model
constituents because: (1) in assessing hazardous waste mixtures that contain
potential carcinogens, cancer risks usually reach levels of concern before
noncancer risks do; and (2) the primary focus of the overall study, including
the air, surface water, and drinking water analyses, was on cancer risk.
7-16
-------
Category 3: Sites in this category were generally not well-
characterized, but the predominant chemicals identified were industrial
chemicals such as phenol and heavy metals such as arsenic, chromium, lead, and
zinc. Phenol and arsenic were selected to represent this group based on
frequency of_occurrence, concentration, toxicity, mobility, and persistence.
Category 4: Information on two sites indicated that the predominant
hazardous wastes generated at these facilities were heavy metal sludges. Zinc
and chromium were identified as the only hazardous constituents of these
sludges and therefore were selected to represent this waste category.
It is important to note that certain sites were grouped mainly on the
appearance of one or two chemicals (e.g., carbon tetrachloride was the
predominant chemical used to group sites into waste category 1), and in some
cases there are fairly wide variations in other chemicals identified among the
different sites in a particular group. This problem is inevitable in a broad
categorization of this type; an attempt was made to minimize the problem by
choosing other model constituents that, although they were not always present
at a large number of sites, were reasonably representative of the toxicity and
mobility of other compounds seen at the sites (i.e., surrogate compounds)..
For example, 2,4-dinitrotoluene was identified at only two of the nine sites
in category 1; however, it is a potent carcinogen that is relatively mobile in
subsurface environments, and thus may provide a reasonable conservative
representation for some other chemicals seen at other category 1 sites (e.g.,
acrylonitrile, benzene, vinyl chloride, and N-nitrosodiphenylamine) with
respect to mobility and toxicity. Chemicals selected as model constituents
for each waste category were chosen primarily based on the degree to which
they represented the wastes found at sites in the Kanawha Valley. However, in
selecting a surrogate compound to represent a group of chemicals, the most
toxic/mobile chemical was chosen.
Also, it should be noted that we did not select many chlorinated aromatic
compounds such as PCBs, dioxins, hexachlorobenzene, or dieldrin as model
constituents. These types of compounds, which tend to bioconcentrate in
aquatic life, have been detected in fish from the Kanawha River. However,
these compounds were not identified at high levels at most of the hazardous
waste sites for which we had information. Moreover, these compounds, due to
their hydrophobic nature, tend to sorb strongly to earth materials and move
very slowly through ground water.
IV. QUANTITATIVE MODELING
Two steps were necessary for the quantitative modeling: development of
model scenarios and the actual modeling of potential exposures and risks. We
discuss these two steps, along with the modeling results, in this section.
7-17
-------
Development of Model Scenarios
Although there are a very large number of hazardous waste sites in the
Kanawha Valley, sufficient data to perform even a preliminary exposure/risk
assessment are, available only for a few sites. Most sites have been studied
only at the screening level, and, at best, the information available on a
particular site consists of: (1) a site report based on a one-day field trip
reconnaissance; (2) analytical data on a few samples (usually ten or fewer)
collected from various media at the site (ground water, soil, drum, and
surface water samples); and (3) a crude hydrogeological characterization and
identification of potentially exposed populations for the site (available for
only a few sites). Thus, the information available is not adequate to support
a site-specific risk analysis, and an extensive data gathering effort would be
needed to perform an in-depth analysis of the problem. Therefore, the
development and analysis of model scenarios that are representative of
different groups of hazardous waste sites was considered the best available
method for gaining some sense of which types of site situations have higher
risks than others and also to begin to get some sense of the overall
environmental problem represented by these sites. Development of the model
scenarios is outlined below.
It is important to emphasize that this quantitative modeling effort is a
screening-level analysis that is based on limited data. Thus, the study has
significant limitations, and definite conclusions with respect to the problem
of ground-water contamination by hazardous waste sites cannot be drawn from •
the results of the analysis. The analysis is, however, considered useful for
comparing relative risks between model scenarios, and for providing a
quantitative overview of the problem for policy screening purposes.
Review of Preliminary Site Characterization and Classification: As
discussed in Section III, similarities exist among the different sites in the
Kanawha Valley that allow them to be grouped into categories based on source
type, waste characteristics, and hydrogeologic setting. A waste/hydrogeology
matrix combining the information in Tables 2 and 5 for the 25 hazardous waste
sites considered in this analysis is presented in Table 6, and a grouping of
sites according to facility type is presented in Table 3.
The grouping of hazardous waste sites into these categories is based on
similarities in source type, waste characteristics, and hydrogeologic
setting. For example, two municipal landfill sites that are located in the
shale/sandstone hydrogeologic setting and contain carbon tetrachloride wastes
would be grouped into "the same matrix category. However, it is important to
note that these groupings are based only on very general information. This is
particularly true for waste characteristics, where in most cases sites have
been grouped on the basis of information from only a few samples. More
detailed analyses may show that two sites that were grouped together based on
screening-level data are in fact quite dissimilar with respect to waste
characteristics (e.g., one site may contain carbon tetrachloride wastes, while
a second site that was thought to contain similar wastes may actually contain
7-18
-------
TABLE 6
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
HYDROGEOLOGY/WASTE MATRIX
FOR HAZARDOUS WASTE SITES IN
THE KANAWHA VALLEY
SITE CLASSIFICATION INTENDED FOR POLICY DEVELOPMENT PURPOSES
HYDROGEOLOGIC
SETTING
PREDOMINANT WASTE CATEOGRY a/
Sandstone/Shale
Union Carbide
Technica I
Center •
Shale/Sandstone Gary Harris
Dump
l
KO
Heizer Creek Dump; Mundy Hollow
Poca Stripmine Landfill
Drum Site; Western
Kanawha County
LandfiII; Smith
Creek Dump; South
Charleston Munici-
paI Landf i I I
AIluvium
Occidental; Artel;
Union Carbide
South Charleston;
Rhone-Poulenc;
Vimasco Corp.;
Markay Chemical Co.;
FMC South Charleston
Chem Formulators;
Monsanto Company
Chemical Leaman
Tank Lines; Coastal
Tank Lines; Mason
and Dixon Tank
Lines; Nitro Sani-
tation; FMC Nitro;
E. I. DuPont
EIkem MetaIs;
Big Scary Creek Pond
a/ Model constituents for the four waste categories are:
1 — carbon tetrachloride, 2,4-dinitrotoIuene, mercury
2 — methylene chloride, phenol, chlordane, cyanide
3 -- phenol, arsenic
4 -- chromium, zinc
-------
predominantly heavy metal sludges). Thus, the grouping of sites into
categories is based largely on screening-level information, and more detailed
investigations may reveal substantial differences among sites in the same
category. _
Despite these problems, the grouping of sites into source/waste/
hydrogeology categories was considered a useful method for evaluating the
potential effects of a large number of sites for which limited data are
available. The initial grouping served as the basis for development of model
scenarios used in the analysis.
Source/Waste/Hydrogeology Scenarios: Based on the characterization and
initial classification of hazardous waste sites outlined above, we developed a
group of seven model scenarios that are intended to be representative of
different groups of hazardous waste sites in the Kanawha Valley. These
scenarios were developed from a compilation of actual site data on source type
and size, waste constituents and concentrations, hydrogeologic parameters, and
potentially exposed populations. However, the following points should be
noted:
• Each scenario represents, in a broad sense, a type
of site that we believe exists in the Kanawha Valley,
but does not correspond directly to any specific site
or group of sites included in the study. Therefore,
risks estimated for a scenario are considered to be
representative of similar hazardous waste sites in the
Kanawha Valley, but should not be associated with any
specific individual site in the area.
• The development of all scenarios, while based on
available data from Kanawha Valley waste sites,
required the use of professional .judgment to
supplement this information. Thus, parameters that
may have a major effect on risk and environmental
impact (e.g., waste constituent concentrations) often
were estimated using a combination of available data
and professional judgment. The significant data
limitations introduce considerable uncertainty into
the results of the analysis.
In the following sections, we outline our methodology for developing scenarios
and for evaluating (semi-quantitatively) the effects of using estimated values
for key parameters in our analysis.
Estimation of Parameters for Model Scenarios: In order to use a
deterministic mathematical model to estimate risks and other environmental
impacts (e.g., the contribution of contaminated ground-water discharge to
surface water contamination in the Kanawha River), it was necessary to develop
scenarios in which individual source, waste, and hydrogeologic parameters were
explicitly identified. As noted above, data from actual waste sites were used
7-20
-------
in identifying these parameters; however, actual arithmetic or geometric mean
values taken from a group of sites for a particular parameter (e.g., waste
constituent concentrations) were not used. Rather, the range of values
observed for a given parameter at a group of sites was used to provide an
estimate of typical values. The final value for the parameter was based on
this observed range and professional judgment. The following example should
serve to illustrate this procedure:
For waste category 1, carbon tetrachloride has been identified
as a waste constituent. A range of concentrations has been
identified in leachate, soil, drums, and surface streams at
sites that serve as the basis for this scenario; however, carbon
tetrachloride has been identified only at five of the nine sites
in this category.'*-1 In addition, quality assurance/quality
control checks on analytical data show that some of the positive
values are questionable. Thus, rather than use a mathematical
average of incomplete data of varying quality from various
media, we estimated a value for carbon tetrachloride
concentration based on the reported range. Our estimated values
were close to the observed values in all cases. Site
concentration data on model constituents selected for the
Kanawha Valley Hazardous Waste Site Study are summarized in
Table 7.
This procedure was used not only for identification of waste constituent
concentrations, but for identification of other scenario parameters as well
(e.g., exposure distance, source size). Thus, parameter values for scenarios
do not represent any individual site, but are representative of what one would
expect to observe at sites in the Kanawha Valley (i.e., values are not taken
from national averages or from sites that are not within the study area).
Parameters for the seven Kanawha Valley source/waste/hydrogeologic setting
scenarios are summarized in Table 8. Waste characteristics and hydrogeologic
settings are presented in Tables 9 and 10, respectively.5-1 As indicated,
two sets of values were developed for each scenario, with scenarios designated
as "A" representing our best-estimate values for parameters and "B" scenarios
representing our conservative values for parameters. The development of
conservative and best-estimate scenarios is discussed in the following section.
(,J As noted in Section III, surrogate compounds were used in some cases to
represent groups of similar chemicals; thus, all sites within a particular
waste category do not necessarily contain all model constituents in that
category.
5J It should be noted that although our hydrogeologic characterization
indicated three different settings, we had information only on sites that were
located in the alluvium and shale/sandstone settings. Thus, we did not
develop model parameters for the sandstone/shale setting. Based on DRASTIC
scores (see Appendix A), it is likely that model parameters for this setting
would be very similar to those for the shale/sandstone setting.
7-21
-------
TABLE 7
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
SITE CONCENTRATION DATA FOR CHEMICALS
SELECTED AS MODEL CONSTITUENTS
WASTE
CATEGORY a/
MODEL CONSTITUENTS
FREQUENCY
OF DETECTION b/
SAMPLE MEDIUM
CONCENTRATION
RANGE
(mq/l or mq/kql c/
1
Carbon tetrachloride
5/9
Leachate
Soi 1
Drum
Stream
0.24
2.42
NQ d/
0.99
2,U-D i n i t roto1uene
2/9
Soi 1
Drum
22.50
9.00
Me rcury
6/9
Ground water
Soi 1
Drum
St ream
0.13
1.8-19.0
0.2-13.4
0.075
2
Methylene chloride
3/7
Ground water
Leacha te
Drum
4.57
12.0
10,000 - 50,000
Pheno1
2/7
Soi 1
NQ-22.6
Chlordane
1/7
Ground water
0.013
Cyanide
4/7
Soi 1
Drum
NQ-2,800
41 .8-100,000
3
Pheno1
6/7
NA e/
NQ
Arsen i c
5/7
NA
NQ
a/ See Section II for a description of the development of waste categories.
b/ Number of sites in waste category at which contaminant was detected divided by the total
number of sites in that waste category.
c/ For sample media in which constituent was detected only once, a single value is reported.
d/ NQ = not quantified. Constituent was detected at the site, but concentration was not reported.
e/ Information on sites in waste category 3 indicated that phenol and arsenic were suspected of
being present at nearly all the sites in the group; however, no quantitative sampling
information was available for any of these sites.
-------
TABLE 8
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
SUMMARY OF PARAMETERS FOR
SOURCE/WASTE/HYDROGEOLOGIC SETTING MODEL SCENARIOS
MODEL
SCENARIO
SOURCE TYPE
a/ OPERATING/ b/
PCC PERIOD
r/ WASTF d/
WASTE CATEGORY- AMOUNT-
GROUND-
WATER e/ SIZE
SETTING (meters)
EXPOSURE f/ PRIMARY
DISTANCE EXPOSURE PATHWAY
-J
I
M
U>
1A
IB
2A
2B
3A
3B
4A
4B
5A
5B
"Subtitle C"
Lined landfiI I
"Subtitle C"
Lined landfiI I
(conservative)
Unlined surface
i mpoundment
Unlined surface
impoundment
(conservative)
"Subtitle D"
Unlined landfiI I
"Subtitle D"
Unlined landfi11
(conservat ive)
"Subtitle D"
Unlined landfi11
"Subtitle D"
UnIi ned I andf iI I
(conservative)
"Subtitle D"
Unlined landfi
I
"Subtitle D"
Unlined landfiI I
(conservative)
20 yrs./30 yrs.
20 yrs./30 yrs.
20 yrs./.O yrs.
20 yrs./O yrs.
20 yrs./O yrs.
20 yrs./O yrs.
20 yrs./O yrs.
20 yrs./O yrs.
20 yrs./O yrs.
20 yrs./O yrs.
1--Ave rage
(so I id )
1--Conservat ive
(sol id)
1—Ave rage
(liquid)
1—Conservat ive
(Iiquid)
2—Ave rage
( leachate)
22,500
metric tons/
yr
50,625
metric tons/
yr
8000 cubic
meters
50,000 cubic
meters
N/A
2--Conservative N/A
( leachate)
3--Average N/A
( leachate)
3--Conservative N/A
(leachate)
1—Average N/A
( leachate)
1—Conservative N/A
(leachate)
300 x 300 200m
450 x 450 60m
40 x 40
50 x 50
50 x 50
200m
100 x 100 60m
200m
100 x 100 60m
200m
100 x 100 60m
150 x 150 400m
400 x 400 200m
Ground-water discharge
to surface water
t
Ground-water discharge
to surface water
Ground-water discharge
to surface water
Ground-water discharge
to surface water
Ground-water discharge
to surface water
Ground-water discharge
to surface water
Ground-water discharge
to surface water
Ground-water discharge
to surface water
Consumption of conta-
minated water
Consumption of conta-
minated water
-------
TABLE 8 (continued)
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
SUMMARY OF PARAMETERS FOR
SOURCE/WASTE/HYDROGEOLOGIC SETTING MODEL SCENARIOS
MODEL
SCENARIO
a/
SOURCE TYPE
OPERATING/ b/
PCC PERIOD
c/
WASTE CATEGORY
WASTE d/
AMOUNT
GROUND-
WATER e/
SETTING
SIZE
(meters I
EXPOSUREf/
DISTANCE
PRIMARY
EXPOSURE PATHWAY
6A
"Subtitle D"
Unlined landfiI I
20 yrs./O yrs.
2--Average
( leachate)
N/A
2
150 x 150
400m
Consumption of
minated wpter
conta-
6B
"Subtitle D"
Unlined landfi11
(conservative)
20 yrs./O yrs.
2—Conservat ive
(leachate)
N/A
2
400 x 400
200m
Consumption of
minated water
conta-
7A
"Subtitle D"
Unlined landfi11
20 yrs./O yrs.
3--Average
( leachate)
N/A
2
150 x 150
400m
Consumption of
minated water
conta-
7B
"Subtitle D"
Unlined landfi11
(conservative)
20 yrs./O yrs.
3—Conservative
( leachate)
N/A
2
400 x 400
200m
Consumption of
minated water
conta-
I
a/ "Subtitle C" landfills are considered to be in compliance with current RCRA liner requirements for hazardous waste landfills.
"Subtitle D" landfills are unlined.
b/ PCC = post-closure care.
c/ Solid = waste constituent concentration specified for solid material disposed at site;
liquid = waste constituent concentration specified for liquid material disposed in surface impoundments;
leachate = waste constituent concentration specified for infiltrating liquid (leachate) at base of landfill. See Table 9 for
characteristics of waste categories 1, 2, and 3.
d/ Specified only for model RCRA facilities ("Subtitle C" landfills and surface impoundments).
e/ Setting 3 = alluvium; Setting 2 = shale/sandstone. Insufficient information on the few sites located in Setting 1
(sandstone/shale) was available to develop a generic scenario for this group. See Table 10 for details about ground-water
settings.
f/ Distance from source of contamination to (1) Kanawha River (scenarios 1-4); or (2) drinking water well (scenarios 5-7).
-------
TABLE 9
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
WASTE CHARACTERISTICS FOR MODEL SCENARIOS
CONCENTRATION
CONCENTRATION
WASTE a/
MODEL
(BEST-ESTIMATE
(CONSERVATIVE
CATEGORY
CONSTITUENTS
SCENARIO (A)) (ppm) b/
SCENARIO (B))(ppm) b,
1 (solid)
Carbon tetrachloride
10
1000
2,4-Dinitrotoluene
1
100
Mercury
1
100
1 (liquid)
Carbon tetrachloride
300
3000
2,4-Dinitrotoluene
0.05
5
Mercury-
0.1
10
1 (leachate)
Carbon tetrachloride
0.25
25
2,4-Dinitrotoluene
0.01
1
Mercury
0.05
5
2 (leachate)
Methylene chloride
5
500
Pheno1
10
1000
Chlordane
0.01
1
Cyanide
2
200
3 (leachate)
Phenol
5
500
Arsenic
0.05
5
a/ See Section III for a description of the development of waste categories.
b/ For solid and liquid wastes, ppm refers to the constituent mass fraction
in the waste stream itself. For these wastes, the Liner Location Model
calculates the mass fraction in leachate as part of the modeling process.
For leachate wastes, ppm refers to the constituent mass fraction in
leachate at the bottom of the disposal unit.
7-25
-------
TABLE 10
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
GROUND-WATER CHARACTERISTICS
FOR HYDROGEOLOGIC SETTINGS a/
GROUND-WATER SETTING 2 - SHALE/SANDSTONE
Horizontal
Hydraulic Depth to Ground-Water
b/ c/ Fraction Organic b/ Conductivity c/ Ground c/ Velocity c/
Bulk Density
-------
Best-Estimate and Conservative Scenarios: As noted previously,
professional judgment was used in some cases to identify values for scenario
parameters. Some of these parameters have a major effect on the results of
the analysis. Therefore, it was considered desirable to have some estimate of
the degree of_uncertainty that was introduced to the analysis by estimating
these parameters based on professional judgment. However, formal sensitivity
analysis techniques (e.g., Monte Carlo simulation) were not feasible, given
the limited resources allocated for this portion of the overall study and the
major data gaps that exist for some parameters.6-1
Therefore, to gain some estimate of the degree to which our identification
of scenario parameters affected the outcome of the analysis, we developed
best-estimate and conservative scenarios. Best-estimate scenarios were
developed as outlined above, by using available data and professional judgment
to construct scenarios that we believe to be most representative of a
particular type of hazardous waste site found in the Kanawha Valley. In
conservative scenarios, we used the same data and professional judgment to
construct a conservative yet reasonable scenario for each type of hazardous
waste site we identified. For example, in Scenario 1A we estimated carbon
tetrachloride concentration in the waste stream to be 10 mg/kg, landfill size
to be 300 x 300 meters, and exposure distance to be 200 meters. In Scenario
IB, we increased carbon tetrachloride concentration to 1000 mg/kg, increased
landfill size to 450 x 450 meters, and decreased exposure distance to 60
meters. Thus, in Scenario IB there is a considerably larger volume of
material, a higher concentration of carbon tetrachloride, and a shorter
distance between source and exposure point, resulting overall in a situation
where one would expect higher risks than in Scenario 1A.
In using best-estimate and conservative scenarios, we have attempted to
develop what we believe are representative situations for different types of
hazardous waste sites in the Kanawha Valley. However, it is important to note
once again that these scenarios are based on limited information and are
intended to represent a type of sites rather than any individual site.
Although our conservative settings are generally representative of the
"highest risk" conditions for sites on which we had information, we do not
believe that the conservative scenarios necessarily represent actual
worst-case conditions that may be encountered at individual sites; therefore,
we would caution against considering risk estimates for conservative scenarios
to be upper-bound limits for hazardous waste sites in the Kanawha Valley.
Conditions at an individual site could be worse than conditions specified in
our conservative scenarios (e.g., much higher concentrations of chemicals,
shorter exposure distances), resulting in risks significantly higher than
those estimated in our analysis.
6J A sensitivity analysis using a technique such as Monte Carlo simulation
is useful only if valid probability distributions can be constructed for each
variable. The Kanawha Valley data were insufficient for this purpose in most
cases.
7-27
-------
In the next section, we present a description of the exposure/risk model
used in the analysis and discuss the migration and exposure pathways
considered in the study.
Exposure/Risk Modeling
In order to evaluate potential risks and environmental impacts for the
Kanawha Valley model scenarios, it was necessary to: (1) identify the most
important migration and exposure pathways for chemicals released from
hazardous waste sites in the Valley; and (2) choose a mathematical model to
quantitatively evaluate these pathways and estimate risks from chemical
exposures. In the following sections, we describe each of these steps.
Identification of Migration and Exposure Pathways: Risk assessment
scientists generally consider an exposure pathway to consist of four basic
elements: (1) a source and mechanism of chemical release to the environment;
(2) an environmental receiving and transport medium for the released chemical;
(3) a point of potential exposure by humans or other receptors; and (4) a
route of exposure to the contaminants. An exposure pathway is considered
"complete" if all of these elements are present. A migration pathway consists
only of a source and mechanism by which a chemical can be released and
transported (e.g., miscible transport in ground water) and does not describe
potential receptors or routes of exposure.
In this analysis, we have considered one exposure pathway and one
migration pathway. The exposure pathway consists of the following:
• Release of chemicals to ground water from hazardous
waste sites as a result of rainwater infiltration (or
liquid infiltration from surface impoundments), either
through failed liner systems or directly to soil in
the case of unlined facilities;
• Transport through the unsaturated zone to the water
table, followed by miscible transport in ground water
to nearby drinking water wells; and
• Human exposure resulting from consumption of
contaminated ground water from these wells.
Thus, this exposure pathway is considered "complete," and risks can be
quantitatively estimated based on predicted exposures.
There are a number of other potential exposure pathways for contaminants
escaping from hazardous waste sites in the Kanawha Valley, including:
(1) direct contact (ingestion or dermal contact) with hazardous waste or
contaminated soil; (2) inhalation of contaminated dust particles;
(3) inhalation of volatile chemicals; (4) consumption or direct contact with
surface water contaminated by run-off or ground-water discharge; and
(5) consumption of biota that have been contaminated with chemicals. However,
we have selected consumption of contaminated ground water as the exposure
pathway of most concern in this analysis for the following reasons:
7-28
-------
• A review of the data for sites in the Kanawha Valley
has revealed that there are people living close to
some sites who use ground water from private wells as
their source of drinking water.7-1 (This is therefore
- a complete pathway at some of the sites on which we
have information.);
• Based on our experience with Superfund and RCRA
sites, we have found that this pathway often results
in the highest potential human exposures; and
• The quantitative evaluation of other pathways
requires more site-specific information than is
available for this study.
In addition to the ground-water exposure pathway, we considered one
migration pathway in our analysis: miscible transport of contaminants in
ground water and discharge of this contaminated ground water to surface
water. This pathway was included because many of the chemical plants and RCRA
facilities considered in our analysis are very close to the Kanawha River or
its tributaries, and therefore it was considered important to evaluate the
potential contribution of contaminated ground-water discharge to the
contamination problem observed in the Kanawha River.
Use of the OSW Liner Location Model: To predict risks to ground-water
users and estimate mass loadings of contaminants from ground water to surface
water, we selected the Liner Location Risk and Cost Analysis Model (LLM)
developed by EPA's Office of Solid Waste (OSW). The LLM is an integrated
computer model that consists of release, transport, exposure, and risk
component submodels. It was designed to analyze generic scenarios
representative of conditions at land disposal facilities throughout the United
States. Thus, the model is appropriate for screening-level, generic analyses
such as the Kanawha Valley study, and it allows both the ground-water and
ground-water/surface water pathways to be quantitatively evaluated.
The LLM includes the following component submodels:
• A stochastic failure model that predicts time to
failure and leachate release volumes for various types
of lined and unlined landfills and surface
impoundments;
7J As stated previously, exposure through consumption of contaminated ground
water is more likely for sites located outside the Kanawha Valley proper.
People in the Kanawha Valley generally receive their drinking water from a
municipal supply, while people living in mountainous areas outside the valley
may obtain their water from private ground-water wells.
7-29
-------
• A mass balance model that predicts contaminant
concentrations in leachate released from a land
disposal unit;
• Unsaturated zone and saturated zone contaminant
transport models (modified versions of the
McWhorter-Nelson Wetting Front and Prickett-Lonnquist
Random Walk Models, respectively);
• An exposure model based on consumption of
contaminated ground water; and
• Carcinogenic and noncarcinogenic risk models.
The model also provides the user with the ability to calculate contaminated
ground-water discharge to surface water.
The LLM is described in detail in the Draft Liner Location Risk and Cost
Analysis Model Report (EPA, 1985). As with any hazardous waste risk model,
the LLM incorporates many assumptions. It is believed that the LLM generally
overestimates risk levels (i.e., is a conservative model) because many of its
assumptions, algorithms, and default values are conservative.
Results of the Screening-Level Analysis
This section presents our results for the Kanawha Valley model scenarios.
As noted, we used the LLM to estimate carcinogenic risks and exposures to
noncarcinogens resulting from consumption of contaminated ground water for
scenarios in which it was determined that a potentially exposed population
could exist (scenarios 5 through 7). We also estimated loadings and potential
effects on the Coal River for scenarios 5 through 7. For scenarios 1 through
4, we evaluated the potential contribution of contaminated ground-water
discharge to contamination in the Kanawha River. The results of these
analyses are discussed below.
Exposure/Risk Analysis for Ground-Water Users: For scenarios 5 through 7
we estimated carcinogenic risks and evaluated noncarcinogenic exposures for an
individual living relatively close to a hypothetical hazardous waste site
represented by each scenario. For our modeling effort, we used the waste
constituents and concentrations, ground-water settings, and exposure distances
outlined for each scenario in Table 8. LLM assumptions regarding contaminant
transport and fate were used, as were typical assumptions about human exposure
and chemical absorption. These assumptions are as follows:
• Contaminant transport in ground water is controlled
primarily by two factors: (1) the retardation factor
of the particular contaminant of interest; and (2) the
ground-water velocity of the aquifer in which
transport is occurring. Aquifer properties that
affect transport (e.g., velocity, fraction organic
7-30
-------
carbon content) are considered to remain constant over
the entire distance from source to exposure point, and
a specific retardation factor is used for each
_contaminant (based on organic carbon-water partition
coefficients for organics and distribution
coefficients for metals). The aquifer is considered
to be a homogeneous, isotropic porous medium.
Contaminants are assumed to degrade at a rate
specified by a single first-order constant that
accounts for all chemical and biological degradation
processes. Rate constants are identified from
literature estimates; chemicals for which degradation
data are unavailable are assumed to be infinitely
persistent.
• Humans are assumed to drink 2 liters of water per
day, and to have a body weight of 70 kilograms. One
hundred percent absorption is assumed for all
contaminants, and each individual is assumed to be
exposed for 70 years.
These assumptions are necessary in a screening-level, non-site-specific
analysis such the Kanawha Valley Toxics Screening Study, and they are
consistent with assumptions made in similar EPA policy analyses. However,
without detailed site-specific information the degree of uncertainty
introduced to the analysis by the use of these assumptions cannot be fully
evaluated.
The total modeling time used for the exposure/risk analysis was 100 years,
which is long enough to allow ground-water transport of moderately mobile
chemicals. Carcinogenic risks were estimated using the one-hit equation, EPA
Carcinogen Assessment Group (CAG) potency values for all contaminants, and a
70-year moving average lifetime dose.,J After a time profile of drinking
water concentration (in mg/1) was estimated, we followed three steps to
compute cancer risk: (1) convert each year's concentration to an equivalent
human dose (in mg/kg-day) by multiplying by 2 liters per day intake and
dividing by 70 kilograms average body weight; (2) average the annual doses
over each 70-year period being analyzed to obtain average lifetime doses; and
(3) calculate the risk using the one-hit equation (linear at low doses), the
average lifetime dose, and the appropraite CAG upper-bound potency value. To
evaluate noncarcinogenic risk, we compared yearly doses of noncarcinogenic
chemicals to risk reference doses (RfDs) for these chemicals. The toxicity
parameters used for specific chemicals included in the analysis are given in
Tables 11 and 12.
,J For a more detailed discussion of exposure and risk modeling used in the
analysis see the Draft Liner Location Risk and Cost Analysis Model Report
developed by EPA's Office of Solid Waste (EPA, 1985).
-------
TABLE 11
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
UPPER-BOUND UNIT CANCER RISK VALUES FOR
POLLUTANTS AT HAZARDOUS WASTE SITES
UNIT CANCER
RISK VALUE a/ WEIGHT-OF-EVIDENCE
CHEMICAL (mg/kg-davW RATING b/
Arsenic c/ 15 A
Carbon tetrachloride 0.13 B2
Chlordane 1.3 B2
Methylene chloride 0.0075 B2
2,4-Dinitrotoluene 0.31 B2
a/ Potency values are upper-bound estimates derived by EPA's Carcinogen
Assessment Group (CAG).
b/ Weight-of-Evidence rating derived by CAG, based on EPA's classification
system: A = proven human carcinogen; B = probable human carcinogen (B1
indicates limited evidence from human studies, B2 indicates sufficient
evidence from animal studies but inadequate evidence from human studies);
C = possible human carcinogen; D = not classifiable, and E = no evidence
of carcinogenicity.
c/ CAG has calculated a potency score for arsenic based on epidemiological
evidence of skin cancer. However, some scientists believe that arsenic
may not be carcinogenic and may be a necessary element at low levels.
7-32
-------
TABLE 12
KANAWHA VALLEY TOXICS SCREENING STUDY/
HAZARDOUS WASTE SITES
VERIFIED REFERENCE DOSES FOR NONCARCINOGENIC
EFFECTS OF CHEMICALS INCLUDED IN THE ANALYSIS
CHEMICAL
VERIFIED REFERENCE DOSE
(mg/kg-day) a/
Arsenic
Carbon tetrachloride
Chlordane
Chromium VI
Cyanide
Methylene chloride
2,4-Dinitrotoluene
Mercury
Phenol
1,1,1-Trichloroethane
Zinc
b/
0.0007
0.00005
0.005
0.02
0.06
b/
0.002
0.10
b/
b/
a/ All reference doses derived for oral exposures,
and verified by EPA's Reference Dose Workgroup in
1985 and 1986.
b/ Verified reference dose unavailable.
7-33
-------
The results of our analysis are summarized in Tables 13 and 14 and Figure
2. Table 13 presents maximum and average risks for an individual living
directly downgradient from a hypothetical waste site and using contaminated
ground water as his/her source of drinking water. Risk estimates for all
carcinogenic chemicals contributing to risk are presented for each scenario.
As illustrated, risks are significantly greater for Scenarios 5A and 5B than
for the other scenarios. This is a result of the wastes found in Scenario 5
(particularly 2,4-dinitrotoluene, a potent carcinogen);9-1 all risk-
influencing parameters other than waste type are constant across the three
scenarios.
Figure 2 is a bar graph comparing peak individual carcinogenic risks for
best-estimate and conservative scenarios. As indicated, risks vary from 2 to
4 orders of magnitude between best-estimate and conservative scenarios, with
the greatest difference between scenarios 7A and 7B, and the least difference
between scenarios 5A and 5B. These differences illustrate that the specified
waste constituent concentrations for a particular scenario have a major effect
on the outcome of the analysis. Thus, this comparison provides a qualitative
estimate of the degree of uncertainty that is involved in this screening-level
analysis. There are of course other elements of uncertainty.10-1
A summary of predicted exposures to noncarcinogenic chemicals for
scenarios 5 through 7 is presented in Table 14. RfDs are not exceeded for any
of the chemicals for which exposure is predicted to occur.
Analysis of Mass Loadings to Surface Water: For scenarios 1 through 4,' we
identified no potentially exposed populations for the ground-water exposure
pathway. The sites that were used to develop these scenarios are located
within the Kanawha River Valley proper, rather than in nearby mountainous
areas, and residents in the river valley generally obtain their drinking water
from municipal supplies. 11J However, many of these sites are located very
close to the Kanawha River, and it is likely that contaminated ground water
from these sites is discharging to the river. Also, the Kanawha Valley
surface water contamination study performed by EPA indicated high levels of
certain contaminants in fish tissues and sediments that could not be fully
explained from NPDES effluent monitoring reports. Thus, it was hoped that an
analysis of river contamination by ground-water discharge would provide some
insight as to the source of these chemicals.
9J Carbon tetrachloride, also found in scenario 5, is a fairly potent
carcinogen, but is assumed to degrade in ground water.
10J As noted previously, conservative scenario risk estimates should not be
regarded as upper-bound limits for risks from hazardous waste sites in the
Kanawha Valley. Risks at specific individual sites could conceivably be
greater than those estimated for a model scenario.
1lJ Municipal supply for the Charleston-Institute-Nitro area is obtained
mainly from the Elk and Coal Rivers.
7-34
-------
TABLE 13
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
PRELIMINARY RISK SCREENING: INTENDED FOR POLICY DEVELOPMENT PURPOSES
ESTIMATES OF LIFETIME INDIVIDUAL CANCER
RISK FOR HAZARDOUS WASTE SITE MODEL SCENARIOS a/
MODEL b/
SCENARIO
-4
I
00
Ui
5A
5B
6A
6B
7A
7B
POTENTIAL CARCINOGEN
(WE IGHT-OF-EVI PENCE 1
INITIAL PEAK
EXPOSURE CONCENTRATION
YEAR (mq/l ) c/
2,4-Dinitrotoluene (B2) 11
Carbon tetrachloride (B2) e/
2,4-Dinitrotoluene (B2) 11
Carbon tetrachloride (B2) e/
Methylene chloride (B2) 5
Chlordane (B2) e/
Methylene chloride (B2) 5
Chlordane (B2) e/
Arsenic (A) 9
-------
r i uunt c.
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
PRELIMINARY RISK SCREENING: INTENDED FOR POLICY DEVELOPMENT PURPOSES
COMPARISON OF PEAK INDIVIDUAL CANCER RISKS FOR MODEL SCENARIOS a/
1 . 0E-00
1.0E-02 -
8.0E—04
1.0E-04 ~
9.0E—07
u> > 1.0E-06
2.0E—07
¥
£ 1.0E-08
1.0E-10
z\
SCENARIO
1.OE-12
IZ71 BEST ESTIMATE
\\| CONSERVATIVE
a/ RISKS ARE LIFETIME INCREMENTAL RISKS ESTIMATED FOR INDIVIDUALS EXPOSED BY GROUND-WATER PATHWAY (I.E., CONSUMPTION OF
CONTAMINATED GROUND WATFR). THE CARCINOGENIC POTENCY FACTORS USED IN (HIS ANAIYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT
l.l Nl KAI l Y PRODUCE UPPER-BOUND ESIIMAIES. BECAUSE OF LIMIIAIIONS IN OA IA AND METHODS IN SEVERAL AREAS OF THE ANAIYSIS, SUCH AS
JXPOSliKI" CAl CUl AT IONS AND CONTAMINANT SELECTION, RISK E ST I MA IFS WLRF CAICUIAIFO AS AIDS TO POLICY DEVELOPMENT, NOT AS
I'ltlDICilONS 01 AC I UAL CANCER RISKS IN THE KANAWHA VALLEY. ACTUAl RISKS MAY Bl HIGHER OR LOWER; IN FACT, 1 HEY COULD BE ZERO.
1111 I'HOPfR UbL OF IHESE ESTIMATES IS TO HELP LOCAL OFFICIALS SELLCT ANO EVALUATE ISSUES AND SET PRIORITIES.
-------
TABLE 14
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
PRELIMINARY RISK SCREENING: INTENDED FOR POLICY DEVELOPMENT PURPOSES
SUMMARY OF NONCARCINOGEN EXPOSURES FOR
HAZARDOUS WASTE SITE MODEL SCENARIOS a/
DISTANCE FROM
MODEL SOURCE TO EXPSOURE
SCENARIO POINT (ml
5A
5B
6A
-J
I
u>
-J
6B
7A
7B
400
200
400
200
400
200
CONTAMINANT b/
RISK REFERENCE
DOSE (RfD)
f mq/kq-day)
2,4-Dinitrotoluene d/
2,4-Dinitrotoluene d/
Methylene chloride 6.0 x 10
Phenol 0.10
Cyanide 0.02
Phenol
Cyanide
PhenoI
Arsen ic
PhenoI
Arsen i c
-2
Methylene chloride 6 x 10
-2
0.10
0.02
0.10
c
0.10
d/
d/
PEAK
CONCENTRATION
(mq/l1
6 X 10
-4
-2
9 x 10
1 x 10
4 x 10
1 x 10
¦18
-4
9 x 10
-2
5 x 10
¦14
2 x 10
2 x 10
-4
-10
2 x 10
-2
2 x 10
-6
2 x 10
PFAk r /
EXPOSURE DOSE-
tmq/kq-day)
2 x 10
-5
3 x 10
-3
3 x 10
-8
1 x 10
-5
3 x 10
-20
3 x 10
-5
1 x 10
-3
6 x 10
-16
6 x 10
6 x 10
-12
6 x 10
-4
-8
6 x 10
EXCEEDS RfD?
(YES/NO)
'N/A
N/A
No
No
No
NO
NO
No
No
N/A
No
N/A
a/ BECAUSE OF LIMITATIONS IN DATA AND METHODS IN SEVERAL PARTS OF THE ANALYSIS, EXPOSURE ESTIMATES WERE
CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL EXPOSURES IN THE KANAWHA VALLEY.
b/ Only contaminants reaching the exposure point within the modeling period are listed.
c/ Calculated from peak predicted well concentrations of contaminants, assuming consumption of 2 liters of
water/day per person and an average body weight of 70 kg.
d/ Verified RfD not available for this contaminant.
SOURCE: Regulatory Integration Division, Office of Policy Analysis, U.S. EPA, 1987.
-------
To evaluate the importance of this contamination source to overall
contamination problems observed in the Kanawha River, we used the LLM to
estimate yearly mass loadings of contaminants from scenarios 1 through 4 to
the river. We then obtained average initial concentrations of these
contaminants J.n the river resulting from contaminated ground-water discharge
by dividing yearly mass loadings by an average annual Kanawha River flow of
10 6
1.5 x 10 m3 (based on an average hourly flow of 1.7 x 10 m3).
The results of this analysis are presented in Table 15. As illustrated,
only a few of the contaminants considered in the analysis were of sufficient
mobility and persistence to migrate to surface water during the 100-year time
period used for the study. Contaminants that tend to bioconcentrate in fish
tissues (e.g., chlordane) are generally very hydrophobic and thus are
relatively immobile in ground water because they sorb strongly to soil and
aquifer materials. Of the chemicals that did migrate to the river, mass
loadings were not sufficient to cause significant water quality problems in
the river (contaminant concentrations were below federal ambient water quality
criteria in all cases), and in most cases river concentrations of contaminants
resulting from ground-water discharge would be below detection limits.12-1
This is not surprising, given that most ground-water plumes are relatively
small and contain small quantities of contaminants when compared to flow
volumes in a large river such as the Kanawha. Also, it should be noted that
many of the RCRA land disposal units used to develop scenarios 1 and 2 have
ground-water monitoring systems that should allow contamination to be detected
before a plume could migrate to the Kanawha River. Detection would allow for
corrective action to be taken (e.g., ground-water pumping and treatment), thus
minimizing surface water contamination via this pathway.
We also estimated mass loadings to and initial concentrations in the Coal
River for scenarios 5 through 7. Table 16 presents these results. None of
the model results for these scenarios indicate a significant surface water
quality problem, even though the Coal River has approximately an order of
magnitude lower flow than the Kanawha River. In all cases concentrations were
below federal ambient water quality criteria, and in most cases the predicted
concentrations would be below analytical detection limits.
V. UNCERTAINTY ISSUES IN GROUND-WATER MODELING
In the hazardous waste part of the Kanawha Valley Toxics Screening Study,
we used limited data on site characteristics and contaminants at 25 hazardous
waste sites in the Kanawha Valley to develop a group of model scenarios that
we believe are representative of the different types of sites in the area.
Based on information from characterization surveys, sampling reports, and
hydrogeologic investigations of specific sites, we assigned model chemical and
hydrogeologic characteristics to each scenario and used the EPA Phase II Liner
Location Model (LLM) to evaluate: (1) exposure and risk for scenarios where
12J Our preliminary concentration estimates do not account for background
contamination in either the Kanawha or Coal River.
7-38
-------
TABLE 15
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
PRELIMINARY EXPOSURE SCREENING: INTENDED FOR POLICY DEVELOPMENT PURPOSES
LOADINGS TO THE KANAWHA RIVER FROM CONTAMINATED GROUND WATER a/
MODEL
SCENARIO
-¦J
I
GJ
CONTAMINANTS
1A 2,4-Dinitrotoluene
1B 2,4-Dinitrotoluene
2A 2,4-Dini tro toluene
2B 2,4-Dinitrotoluene
3A Methylene chloride
Ptieno I
3B Methylene chloride
PhenoI
4A PhenoI
4B PhenoI
AVERAGE MASS
LOADING (ko/yr)
PEAK MASS
LOADING (kq/vr1
DURATION OF MASS
LOADING OVER 100-
YEAR MODELING
PER I OP Ivrs.1
3.6 x 10
1.2
-4
1.5 x 10
+2
1.0 x 10
-3
6.2 x 10
-13
3.1 x 10
-14
2.1 x 10
-6
5.2 x 10
-9
1.6 x 10
-14
2.6 x 10
-9
1.4 x 10
4.7
-3
+2
2.8 x 10
1.7 x 10
-3
4.3 x 10
¦12
1.2 x 10
-13
3.4 x 10
-5
2.7 x 10
-8
-14
6.0 x 10
-8
1.3 x 10
22
39
70
87
68
69
79
40
69
40
RESULTANT INITIAL KANAWHA
RIVER CONTAMINANT
CONCENTRATION (mq/ll b/ c/
-11
I
2.4 x 10
1.0 x 10
-5
-11
6.7 x 10
8.0 x 10
-8
4.1 x 10
-20
-21
2.1 x 10
1.4 x 10
-13
3.5 x 10
-16
-21
1.1 x 10
¦16
1.7 x 10
a/ BECAUSE OF LIMITATIONS IN DATA AND METHODS IN SEVERAL PARTS OF THE ANALYSIS, SURFACE WATER LOADING ESTIMATES WERE CALCULATED
AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL EXPOSURES IN THE KANAWHA VALLEY.
6
b/ Based on average mass loading and average flow of 1.7 x 10 cubic meters/hour in the Kanawha River (equals approximately 1.5 x
10
10 cubic meters/year). These estimates assume dilution by uncontaminated water (background contamination in the Kanawha
River is not accounted for). The computation formula is average mass loading (kg/yr) times 1,000,000 mg/kg, divided by
{average flow (cubic meters/year) times 1000 liters/cubic meter],
-6
c/ None of the values exceed federal ambient water quality criteria either for aquatic life or human health (at 10 cancer risk
level). Assuming 2 liters per day drinking water exposure, the highest individual cancer risk would be approximately
-8 . "
9 x 10 (scenario 1B).
SOURCE: Regultory Integration Division, Office of Policy Analysis, U.S. EPA, 1987.
-------
TABLE 16
KANAWHA VALLEY TOXICS SCREENING STUDY/HAZARDOUS WASTE SITES
PRELIMINARY EXPOSURE SCREENING: INTENDED FOR POLICY DEVELOPMENT PURPOSES
LOADINGS TO THE COAL RIVER FROM CONTAMINATED GROUND WATER a/
MODEL
SCENARIO
CONTAMINANTS
AVERAGE MASS
PEAK MASS
DURATION OF MASS
LOADING OVER 100-
YEAR MODELING
RESULTANT INITIAL COAL
RIVER CONTAMINANT
5A
2,4-Din i trotoluene
-2
-2
*_=.•* " r v i J • « • i
I*r»i lun 111114/ ll u/ y/
-8
3.3
X
10
5.4
X
10
90
1.7
X
10
+1
+ 1
-6
5B
2,4-D i n i t roto Iuene
1.2
X
10
2.2
X
10
90
6.3
X
10 1
-5
-5
-11
6A
Methylene chloride
2.2
X
10
9.0
X
10
95
1.2
X
10
-3
-2
-9
Pheno1
9.0
X
10
3.6
X
10
68
4.7
X
10
-17
-17
-23
Cyan ide
2.2
x
10
9.0
X
10
97
1.2
X
10
t
-2
-1
-8
o
6B
Methylene chloride
2.3
X
10
2.2
X
10
95
1.2
X
10
+ 1
-6
Phenol
3.0
1.2
X
10
64
1.6
X
10
-12
-12
-19
Cyanide
1.2
X
10
4.8
X
10
97
6.3
X
10
-9
-8
-15
7A
Arsen ic
8.9
X
10
1.8
X
10
7
4.6
X
10
-3
-2
-9
Phenol
4.5
X
10
1.8
X
10
68
2.4
X
10
-4
-4
-11
7B
Arsenic
1.8
X
10
4.8
X
10
12
9.5
X
10
-7
Phenol
1.2
4.8
64
6. 3
X
10
a/ BECAUSE OF LIMITATIONS IN DATA AND METHODS IN SEVERAL PARTS OF THE ANALYSIS, SURFACE WATER LOADING ESTIMATES WERE CALCULATED
AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL EXPOSURES IN THE KANAWHA VALLEY.
5 9
b/ Based on average mass loading and average flow of 2.2 x 10 cubic meters/hour in the Coal River (equals approximately 1.9 x 10
cubic meters/year). These estimates assume dilution by uncontaminated water (background contamination in the Coal River is
not accounted for). The computation formula is average mass loading (kg/yr) times 1,000,000 mg/kg, divided by [average flow
(cubic meters/year) times 1000 liters/cubic meter].
-6
c/ None of the values exceed federal ambient water quality criteria either for aquatic life or human health (at 10 cancer risk
level). Assuming 2 liters per day drinking water exposure, the highest individual cancer risk would be approximately
-8
6 x 10 (scenario 5B).
SOURCE:
Regulatory Integration Division, Office of Policy Analysis, U.S. EPA, 1987.
-------
direct exposure to contaminated ground water was plausible; and (2) chemical
mass loadings to the Kanawha and Coal Rivers resulting from ground-water
discharge for site scenarios in the alluvial setting.
The quantitative analysis described in Section IV is limited by the
modeling assumptions used and the uncertainties inherent in using mathematical
models as predictive tools. In general, modeling assumptions tend to
oversimplify environmental fate and transport (which may result in
over-prediction or under-prediction of exposure point concentrations). Also,
risks are estimated based on a number of conservative assumptions, including
100 percent absorption of contaminants and the use of CAG upper-bound potency
values for carcinogens. Unfortunately, these multiple factors interact in a
complex manner, and although risk modeling assumptions and data limitations
introduce a significant amount of uncertainty to the analysis, it is
impossible to state categorically that the results of the risk analysis are
conservative (i.e., that risk estimates are higher than one would expect to
find at most sites in the Kanawha Valley).
Appendix B provides a detailed discussion of the uncertainty involved in
our quantitative ground-water modeling. Qualitative/semi-quantitative
evaluations of ground-water modeling results for the hazardous waste part of
the Kanawha Valley Toxics Screening Study indicate that the results have
substantial uncertainty associated with them. This uncertainty is partially,
attributable to assumptions and limitations inherent to the models used in the
analysis. However, the greatest contributor to uncertainty in the analysis is
probably the lack of comprehensive quantitative information on hazardous waste
sites in the Kanawha Valley. Because of this lack of key information
necessary for ground-water modeling, the scenarios we developed to represent
typical groups of hazardous waste sites were based only on limited data from a
relatively small number of sites. We are unable to quantitatively evaluate
the degree to which our model scenarios are representative of actual sites in
the Kanawha Valley.
From our evaluation of uncertainty issues related to ground-water modeling
in the Kanawha Valley Toxics Screening Study, which is included as Appendix B,
we have drawn the following conclusions:
• Uncertainty inherent to the LLM version of the
Random-Walk Model used in our ground-water analysis is
probably not a major contributor to overall
uncertainty in the study. For situations in which
model conditions and assumptions are met, the model is
not likely to be in error by more than an order of
magnitude.
• Application of the model to ground-water situations
not well matched to model conditions and assumptions
may contribute greatly to the overall uncertainty of
the ground-water results. Predicted concentrations
could be in error by several orders of magnitude for
situations where model scenarios do not accurately
reflect site conditions. Potential sources
7-41
-------
of error include incorrect matching of the model's
generic hydrogeologic settings with site hydrogeology
and mis identification of retardation factors and
degradation potential of chemical contaminants.
• Chemical release modeling algorithms used in the
analysis are simplistic representations of complex and
possibly transient phenomena. Even under the most
ideal circumstances, estimation of a source term for
hazardous waste sites is likely to be a substantial
contributor to uncertainty in ground-water modeling
and exposure assessment.
• Data on geohydrologic and chemical characteristics
of hazardous waste sites in the Kanawha Valley were
limited and of variable quality. As noted, this lack
of data makes it difficult to evaluate the
representativeness of our model scenarios, and is a
source of considerable uncertainty in the analysis.
The multiple factors in the model interact in a complex manner, and linear
propogation of errors cannot be assumed. For example, under-estimation of '
certain parameters and over-estimation of others may fortuitously produce
correct results. Thus, we are unable to estimate quantitatively the overalla
uncertainty associated with ground-water modeling results in this study. We
have described in Appendix B, however, the modeling areas that we believe are
responsible for most of the uncertainty.
VI. SUMMARY AND CONCLUSIONS
We structured this analysis to develop a sense of the potential public
health concerns related to ground-water contamination from hazardous waste
sites within the study area. A review of available information on sites,
collected by EPA Region III and the West Virginia Department of Natural
Resources, provided data on characteristics of and chemical contaminants
detected at 25 hazardous waste sites in the study area. From this
information, we developed a group of model scenarios that we believe are
representative of hazardous waste sites present in the Kanawha Valley, and
quantitatively evaluated exposure and risk for these scenarios. This modeling
exercise can only provide a general sense of the potential public health
concerns these sites may present, and accounts for only two potential exposure
pathways within the Kanawha Valley. We cannot provide a quantitative estimate
of potential cancer risk for specific sites from this analysis; however, we
believe our analysis of model scenarios may provide a general sense of
direction and priorities in the collection of data and information for such an
assessment.
The specific findings from our study are as follows:
7-42
-------
The data available for assessing the potential
health risks from hazardous waste sites in the Kanawha
Valley are limited. Our data base included only 25
RCRA and potential CERCLA sites, which are listed in
Table 1. By contrast, EPA Region III has identified
56 uncontrolled waste sites that require a Preliminary
Assessment/Site Inspection through the CERCLA program,
in addition to numerous RCRA facilities throughout the
study area.13J Our data base represents only a
subset of these CERCLA and RCRA sites for which
sufficient data to perform a characterization were
available.
Of the 25 sites for which data were available, 17
are located in the river valley alluvial setting and
seven are in the shale/sandstone bedrock setting in
the northwestern part of the study area. Only one
site is located in the sandstone/shale bedrock setting
in the southeastern part of the study area.
Most of the RCRA sites are located in the alluvial
setting, while many of the potential CERCLA sites are
in the non-alluvial settings outside the valley walls.
Site monitoring studies have detected many hazardous
chemicals present at the 25 sites reviewed in this
analysis, as presented in Table 4. These chemicals
represent a wide variety of hazardous substances and
include a number of potential carcinogens.
Residents living in the alluvium are served
primarily by public water supply systems. Some
residents living in the other two hydrogeologic
settings, shale/sandstone and sandstone/shale, receive
drinking water from private wells.
Because exposure to contaminated ground water is
possible in the non-alluvial regions of the study
area, we' developed hazardous waste site scenarios that
included drinking water exposure for the shale/
sandstone setting. For various chemical contaminants
typically found at hazardous waste sites in this
setting, we estimated a wide range of risks. Our
modeling analysis indicates that individual
13J The 56 sites either are on the ERRIS site inventory (a listing of all
known potential Superfund sites) or have been recommended for the inventory
(Source: NUS Corp., Kanawha Valley Study - Phase II Screen of Non-ERRIS
Listed Sites, EPA Contract No. 68-01-6699, December 10, 1986).
7-43
-------
-12 -3
incremental risks could range from 10 to 10
probability (a one in one trillion to a one in one
thousand chance) of contracting cancer during a
-lifetime as a result of consumption of contaminated
water. The exposure point concentration and chemical
type were the most significant variables affecting
risk levels. No noncancer effects due to chronic
exposure were predicted.
• Tributaries of the Kanawha River are the primary
sources of drinking water for the public water
systems. As these tributaries flow to the Kanawha
River, they could be affected by discharged ground
water contaminated by hazardous waste sites along
their banks. To investigate the potential effects of
hazardous wastes sites on tributaries, we modeled
ground-water discharge to the Coal River, which
supplies two public water systems. The highest
modeled concentration could result in an individual
-8
incremental risk of 6 x 10 probability (a six in one
hundred million chance) of contracting cancer over a
lifetime exposure. This level of risk is driven
primarily by chemical type and concentration, as well
as the dilution effect of the tributary. The Elk
River, another tributary that serves a larger public
water supply system, has a much larger flow than the
Coal. Consequently, we would expect risks to be lower
for the modeled scenarios discharging contaminated
ground water to the Elk River.
• The model scenario analysis suggests that mass
loadings of contaminants to the Kanawha River via
ground-water seepage from hazardous waste sites in the
alluvium are likely to be relatively small. In
general, loadings from site scenarios were less than
one kilogram per year of a modeled constituent
discharged to the river. One conservative scenario, a
RCRA-type landfill, yielded discharge of approximately
280 kilograms per year of 2,4-dinitrotoluene.
VII. STUDY LIMITATIONS, DATA GAPS, AND RECOMMENDATIONS
This Kanawha Valley hazardous waste site study is a screening-level
analysis based on representative model scenarios, and as such provides only
limited insight into the problem of hazardous waste contamination in the
Kanawha Valley. The study is constrained by a lack of complete data on most
sites, by the use of model scenarios to represent groups of what may in some
cases be dissimilar sites, and by uncertainties attributable to the
7-44
-------
methodologies used to predict exposure and risk. This section outlines
limitations of the study caused by information gaps and provides some guidance
for areas where further study is needed.
Specific limitations and data gaps in the study can be grouped into the
following general areas: (1) limitations in the analysis of exposure and risk
via the ground-water pathway; and (2) limitations resulting from focusing on
ground-water exposure pathways. These areas are described below.
Limitations and Data Gaps in the Ground-Water Analysis
Although many hazardous waste sites have been identified in the Kanawha
Valley, sufficient data for a detailed exposure/risk assessment are available
for very few sites. Information on local hydrogeology, types and amounts of
wastes present, and potentially exposed populations is lacking for many of the
sites used to develop the model scenarios used in the analysis; thus, the
quantitative results of the analysis must be viewed with caution. Specific
areas in which more data are needed to improve our knowledge of the potential
effects of hazardous waste on ground water include the following:
• Potentially exposed populations for the ground-water
consumption pathway have not been fully identified for
any of the hazardous waste sites in the Kanawha
Valley. In some cases, a single ground-water user or
a small group of users has been identified near the
site, but in other instances there is no population
information. It is very important that the number of
people potentially affected by these sites be
identified, so that the scope of the hazardous waste
problem can be better evaluated. A review of
potential ground-water users indicated that private
wells serve a population outside the valley walls,
away from the public water supply system. This study
suggests that a survey of private well users in the
Kanawha Valley would be a useful first step in any
further studies.
• The majority of potential hazardous waste sites in
the Kanawha Valley have been identified only by aerial
photography. These sites need to be verified as
hazardous waste sites or removed from EPA's list of
potential sites. The very large number of sites on
which there is essentially no information makes the
actual magnitude of the hazardous waste problem in the
Kanawha Valley difficult to determine.
• Although a number of parameters affect the risk
attributable to hazardous waste sites, none have a
greater effect than the types and concentrations of
contaminants present. Unfortunately, this is the area
7-45
-------
of our analysis in which the data are weakest. Future
studies at hazardous waste sites in the Kanawha Valley
should concentrate on gathering reliable chemical
_information, preferably from a number of media (e.g.,
surface soil, subsurface soil, ground water, surface
water, biota).
• Detailed hydrogeologic studies were available for
only two hazardous waste sites in the area (Artel in
the alluvium, and Smith Creek Dump in the shale/
sandstone). Both studies provide a great deal of
useful information on the hydrogeology of the area;
however, more studies are needed to determine whether
these sites are representative of general
hydrogeologic conditions in the Kanawha Valley.
• For the effects of ground water on rivers and
tributaries, results of the model analysis suggest
that sources of pollutants other than ground-water
discharge may be more significant and should be
investigated. Although we cannot draw firm
conclusions on the significance of the modeled
discharge rates relative to other sources, it is
likely that pollutant loadings from industrial
discharges, contaminated sediments, and non-point
source run-off, including run-off from hazardous waste
sites and industrial facilities, may be more important
as major loading sources to the Kanawha River.
Limitations and Data Gaps for Other Potential Exposure Pathways
Exposure via consumption of ground water contaminated by hazardous waste
site leachate was the primary exposure pathway considered in this screening-
level analysis (surface water contaminated by ground-water seepage was also
examined). Although we believe that this pathway is generally the most
important for many hazardous waste sites, it is possible that there are other
pathways resulting in significant exposure to contaminants released from the
Kanawha Valley sites. For example, chemicals may volatilize from surface
impoundments, or from contaminated soil or drums, and may be transported to a
human receptor by air.. Other potential pathways that were not considered in
the analysis include wind entrainment of contaminated dust particles,
migration of contaminants via surface run-off, and direct contact with
materials at the site.
Additional information on hazardous waste sites in the Kanawha Valley
would need to be gathered to allow a quantitative evaluation of potential
exposures and risks resulting from these other pathways. To evaluate air
pathways (volatilization and wind entrainment of dust), information on the
following parameters would be needed:
7-46
-------
Wind speed and direction;
• Site topography and vegetation:
• ""Particle size distribution and moisture content of
site soils;
• Climate (e.g., temperature, rainfall, stability
class);
• Identity and concentrations of chemicals; and
• Potentially exposed populations (number of people
living in proximity to the site).
Some of this information is readily available (e.g., climate information),
while obtaining other data (e.g., site topography and vegetation, potentially
exposed populations) would require site-specific investigations.
In evaluating the importance of surface run-off as a potential exposure
pathway, similar information would be necessary. Data on site topography and
vegetation, soil infiltration rate and run-off coefficient, run-off patterns,
and rainfall would need to be gathered, as would information on downstream
uses of surface water that could be contaminated by this run-off. Potentially
exposed populations would need to be identified, and the fate in surface water
of the various hazardous contaminants from the site would need to be
considered.
To evaluate exposure via direct contact, information on populations living
nearby would be needed, as would data on site security, physical barriers
surrounding the site, and the ages of people who may come in contact with the
site (children are much more likely than adults to be exposed at a significant
level via soil ingestion).
7-47
-------
REFERENCES
U.S., Department of Agriculture, Soil Conservation Service 1981. Soil Survey
of kanawha County, West Virginia
U.S. Department of Agriculture, Soil Conservation Service, 1975. Soil Survey
of Fayette and Releigh Counties, West Virginia
Putnam Co. Soil report - Not available
Wilmoth, B.M. 1966. Ground water in Mason and Putnam Counties, West.
Virginia. West Virginia Geological and Economic Survey, Bulletin 32.
Doll, W.L., B.M. Wilmoth Jr., and G.W. Whetstone, 1960. Water Resources of
Kanawha County, West Virginia. West Virginia Geological and Economic Survey,
Bulletin 20.
Doll, W.L., G. Meyer and R.J. Archer. 1963. Water Resources of West
Virginia. West Virginia Department of Natural Resources, Division of Water
Resources.
Johnson, P.W. and T.E. Williams 1969. Kanawha Basin Comprehsive Study.
Appendix C - Hydrology, Part Two - Ground Water. Kanawha River Basin
Coordinating Committee, prepared by U.S. Geological Survey Department of the
Interior.
West Virginia Geological and Economic Survey 1968. Geologic Map of West
Virginia. 1:250,000
U.S. Environmental Protection Agency (USEPA) 1985. DRASTIC: A standarized
system for evaluating ground water pollution potential using hydrogeologic
settings. Office "of Research and Development. Robert S. Kerr Environmental
Research Laboratory May 1985. EPA/600/2-85/018.
Heath, R.C. 1984. Ground-water regions of the United States. U.S. Geological
Survey Water - Supply Paper 2242.
G.M. Ferrell, Map. Natural Resources 1984. Groundwater Hydrology of the
Minor Tributary Basins of the Kanawha River, West Virginia.
7-48
-------
Appendix A
Appendix B
Appendix C
7-49
HAZARDOUS WASTE ANALYSIS
Technical Appendices
(available upon request)
Hydrogeologic Characterization of the
Kanawha Valley
Uncertainty Issues in Ground-Water
Modeling
Hazardous Waste Site Reference List and
Source Information
-------
Chapter Eight
Putting Risk In Context
-------
CHAPTER EIGHT
PUTTING RISKS IN CONTEXT
There are several ways to compare environmental risks, each provi-
ding a different perspective. One way is to compare risk estimates for
the Kanawha Valley to similar risk estimates in other geographic areas,
although one must exercise caution in doing so. Studies done for
different regions with different objectives often have exposure and
risk assumptions that are not easily comparable. However, EPA has esti-
mated some ambient Concentrations and risks associated with certain
chemicals in different areas of the country. Where those chemicals
and exposure assumptions were similar to those studied in Kanawha
Valley, we have attempted to present a comparison.
In general, the ambient concentrations and estimated risks from
many of the toxic chemicals which we evaluated in the Kanawha Valley
appear to be within the range we see in other areas where EPA has data
for similar individual pollutants. However, we see concentrations in
the Kanawha Valley at both the high and low end of this range.
The risks associated with the toxics studied in public drinking
water in the Kanawha Valley appear to be about the same as in other
urban areas in the United States which also use surface water as a
source. Trihalomethane concentrations measured in Kanawha Valley
public water supplies fall in the middle of our comparative range. A
comparison of ambient drinking water concentrations of some common
pollutants found in different urban areas is presented in Table 1.
For cancer risks from toxic air pollutants, we compared the upper-
bound risks estimated in the Kanawha Valley Toxics Study with risks
estimated in national studies (NESHAP analyses) conducted by EPA for
evaluating various chemicals as potential national Hazardous Air
Pollutants under Section 112 of the Clean Air Act. The NESHAP analyses
model all major industrial plants within the country emitting a speci-
fic pollutant, and estimate national cancer incidence and a range of
individual risks to populations surrounding these industrial plants.
For the Kanawha Valley Study, we estimated risks to individuals living
near industrial plants, and incidence, the increased number of cancer
cases attributed to a pollutant within the Kanawha Valley study area.
Table 2 presents a comparison of incidence rate estimates obtained
from NESHAP analyses to incidence estimates obtained from the Kanawha
Valley Study. Incidence rates are presented in terms of cases per
million per year. According to this comparison, seven pollutants
found in the Kanawha Valley are approximately equal to or exceed the
calculated incidence rate for their respective NESHAP analysis. Risks
estimated for three pollutants studied in Kanawha Valley - Ethylene
Oxide, 1,3-Butadiene, and Acrylonitrile - significantly exceed their
NESHAP analysis. Incidence rate estimates for five pollutants found
in the Kanawha Valley are less than their corresponding NESHAP analyses.
8-1
-------
Table 3 presents a comparison of increased risks to individuals
living within the vicinity of the plants. For the NESHAP analysis,
these are the considered most exposed individuals (MEIs), presented as
a range to indicate the MEI risks calculated for a set of processes
emitting a pollutant. For the Kanawha Valley Study, these estimates
are not necessarily for the most exposed individual, but for people
who likely received a higher ambient exposure than an average individual
within the Kanawha Valley. According to this comparison, several Kanawha
Valley pollutants are at the upper end (higher risk) of the range
estimated for their NESHAP analyses.
Another perspective on air risks is to consider what people may
actually be exposed to in a typical day, including residential, occu-
pational and outdoor air exposures. Since people generally spend a
majority of their time indoors, national estimates of the risks from
indoor air exposure have been compiled through several studies. These
studies have shown that indoor levels of many organic compounds are
frequently higher than levels found outdoors. However, these generalized
findings may not apply to several of the highest risk Kanawha Valley
air pollutants (of those studied), which originate primarily from
outdoor point sources.
A third perspective on risk is to compare the risks estimated from
environmental exposures to mortality and sickness (morbidity) attributable
to a variety of other causes. When considering this type of comparison,
keep in mind that mortality and morbidity statistics describe actual
occurrences of disease in the past. The risks calculated for the Kanawha
Valley Toxics Screening Study are plausible, upper-bound estimates and
may be overstated. For a limited number of pollutants, the Kanawha
Valley Study estimated 0.3 additional annual cancer cases from toxic
pollutants in drinking water, and 0.7 to 1.8 in air. This compares to
between 275 to 650 estimated cancer cases which occur annually in the
Valley from all causes.
Even when we have considered our risk estimates in the context of
other risks, and formed an opinion about the significance of those risks,
we are still faced with the question of whether or not the risks are
acceptable. Central to this question is how risk is perceived. For
instance, people may find it hard to relate to large, abstract estimates
such as a 1 in 10,000 chance of contracting cancer over a lifetime.
Table 4 presents some common risks we face. Risk may be influenced
by factors such as voluntariness, controllability, familiarity or by the
immediacy, magnitude, and reversability of the consequences of the risks
faced. For example, people may find some risks more acceptable if bene-
fits accrue directly to them for incurring this risk.
Having considered the severity and significance of the risks in
question still leaves the management decision of whether to act to control
the risk now, engage in further study, or accept the risks. The technical
feasibility of control, the availability of resources, and the economic
implications are some of the most important factors, particularly given
the conservative nature of the risk estimates. Public concern and percep-
tion will always weigh heavily in these risk management decisions.
8-2
-------
TABLE 1
DRINKING WATER
COMPARISON OF AMBIENT LEVELS OF SELECTED
TOXICS IN KANAWHA VALLEY PUBLIC WATER
SUPPLIES TO LEVELS IN OTHER AREAS
All Values in Micrograms/Liter (ug/1)
SUBSTANCE
SANTA CLARA
COUNTY 1
NATIONAL
STUDIES 2
KANAWHA
VALLEY 3
Trihalomethanes
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Zinc
4-80
0.3 - 15
<100 - <1000
0-10
1-8
1-40
0-6
1-10
1-20
1-90
NA
10 - 1000
0 - 172
NA
0 - 112
NA
0.1- 1.8
0-10
NA
NA
42-77
<1 -10
<1 -100
<.l-.2
<1-2
<1-1
<•2-.4
<2-6
<.2
< 5 -10
1 SOURCE: Santa Clara Valley Integrated Environmental Management Project;
Tables 4-11, Exposure to metals and minerals
2 SOURCE: National Organics Monitoring Survey, 1977
3 SOURCE: Kanawha Valley Tbxics Screening Study
4 NA : Not Available
8-3
-------
TABLE 2
COMPARISON OF ANNUAL CANCER INCIDENCE
FOR SELECTED AIR TOXICS IN
KANAWHA VALLEY WITH INCIDENCE RATES
FROM NATIONAL EMISSION STANDARDS
FOR HAZARDOUS AIR POLLUTANTS (NESHAP)
(Annual Incidence per Million Population)
NESHAP KANAWHA
POLLUTANT STUDY1 VALLEY^
Ethylene oxide
.23
3.48
Chloroform
.10
.58
1,3-Butadiene
.35
1.15
Acrylonitrile
.01
.81
Carbon tetrachloride
<.01
.08
Methylene chloride
.14
.1
Trichloroethy1ene
.02
<.001
Vinylidene chloride
.02
.01
Vinyl chloride
2.34
<.001
Benzene
>.2
<•001
Ethylene dichloride
.01
<.001
Allyl Chloride
Not Available
<.001
Source: Communication from Pollutant Assessment Branch, OAQPS.
Estimates developed frcm "Intent to List" Federal Register
Notices or equivalent information.
Source: Kanawha Valley Toxics Study
8-4
-------
TABLE 3
COMPARISON OF UPPER-BOUND
ESTIMATES OF LIFETIME CANCER RISKS
TO THE MOST EXPOSED INDIVIDUAL
POLLUTANT NESHAP Kanawha
Ranqe 1 Valley (Zone)2
Ethylene oxide
10-7-10"3
5
X
10"3
(S. Charleston)
Chloroform
10-6-10-2
3
X
10-3
(Belle)
1,3-Butadiene
lO-6-io-l
6
X
10-3
(Institute)
Methylene chloride
10-5-10-3
5
X
10~4
(Belle)
Carbon tetrachloride
10-6-10-3
4
X
10-4
(Belle)
Acrylonitrile
10-6-10-3
9
X
10-4
(S. Charleston)
Tr i chloroethylene
10-5-10-3
6
X
10-6
(Nitro)
Vinylidene chloride
10-5-10"4
1
X
10-5
(S. Charleston)
Vinyl chloride
10-8-10-3
2
X
10-3
(Nitro)
Benzene
10-8-10-3
3
X
10-7
(Institute)
Ethylene dichloride
10-7-10-2
9
X
10-7
(Nitro)
1 Source: Communication frcm Pollutant Assessment Branch, OAQPS. Estimates
developed from "Intent to List" Federal Register Notices or
equivalent information.
2 Source: Kanawha Valley Toxics Study
8-5
-------
TABLE 4
SELECTED INVOLUNTARY RISKS
Lightening Striking Source for Wilson Derivation
-10-6 risk of death in 2 years U.S. Census (1975)
10-5 risk of death in 20 years
Motor Vehicle Accidents
-10-6 risk of death in 1.5 days
-10-3 risk of death in 4 years
National Safety Council
(1950-78)
Sulfate Air Pollution
-10-3 risk of death in a lifetime
Derived from several
sources
Radon
-10-2 to 10-3 risk of lung cancer
in a lifetime in the overall
• population
Office of Policy Analysis
derived from EPA published
information
8-6
-------
GENERAL REPORT REFERENCES
Hinman K., Schwartz D., Soffer E. Santa Clara Valley Integrated
Environmental Management Project. U.S. Environmental Protection
Agency, Regulatory Integration Division. Revised Stage I Report.
May 30, 1986.
Johe S., Sole D. , Walters M. Charleston Report. Fisher-Harrison
Publications Inc., Greensboro, N.C., (No Date).
National Institute for Chemical Studies. Safeguards Report.
National Institute for Chemical Studies, Charleston, W.V.
November 1986.
U.S. Environmental Protection Agency. Environmental Overview Kanawha
Valley West Virginia (Draft). U.S. Environmental Protection Agency,
Region III Philadelphia, Pa. July 1986.
U.S. Environmental Protection Agency. Final Report on the Philadelphia
Integrated Environmental Management Project. U.S. Environmental
Protection Agency, Regulatory Integration Division. December 1986.
U.S. Environmental Protection Agency. Risk Assessment and Management:
Framework for Decision Making. December 1984. EPA 600/9-85-002.
Vincent J.R. Overview of Environmental Pollution in the Kanawha Valley.
U.S. Environmental Protection Agency, National Enforcement Investigations
Center, Denver, Co. A"gust 1984.
R-l
-------
Appendices
-------
APPENDIX A
SOME ENVIRONMENTAL ACTIVITIES NOT ADDRESSED IN THE
KANAWHA VALLEY TOXICS SCREENING STUDY REPORT
I. Accidental Release Prevention, Preparedness and Response
- Federal/State
° The Chemical Emergency Prepardeness Prograj r*AFQQl ^ as estab-
lished by EPA to provide guidance to local communities for
the preparation of contigency plans to handle accidental air
releases of toxic substances. The CEPP criteria document
includes the list of 402 acutely toxic chemicals, which
is used in planning operations.
° EPA is initiating The Accidental Release Information Program
(ARIP) designed to compile data on the nature and causes of
accidental releases and the measures taken by industry to
prevent them. The early stages of data gathering is being
executed by letters to selected facilities requesting inform-
ation on releases. However, the ultimate goal of the program
is to publish regulations requiring companies to report certain
data subsequent to a spill or release.
° The recently approved amendments (SARA) to the Superfund law
(CERCLA) includes a provision entitled Title III which
institutionalizes and expands CEPP by setting up reporting
requirements for facilities that handle hazardous materials.
The facilities will report to a Governer appointed State
Commission which oversees emergency prepardeness and prevention
activities. The State Commission will designate local emergency
planning districts and local emergency planning commities who
will collect information and data through the right-to-know
provisions of the statute.
° EPA has established an Emergency Systems Review Workgroup
to review, on a national level, emergency systems for
monitoring, detecting and preventing releases of toxic
substances. It is expected that a wide range of potential
accident prevention techniques will emerge from the workgroup
project.
A-l
-------
EPA Region III and The West Virginia Department of Natural
Resources are parts of a Regional Response Team (RRT),
established to respond to chemical emergencies that are a
imminent treat to the public health or the environment. The
RRT has limited contractor authority and can call on the
resources of various State and Federal agencies for the
purpose of mitigating a hazardous situation.
° The West Virginia State Legislature passed a Right-to-Know
Law which calls for certain reporting requirements for
facilities that handle hazardous materials. This information
is submitted to The State Department of Health in the form
of Material Safety Data Sheets.
° The prevention of transportation accidents is enhanced through
the hazardous substances handling and shipping requirements
administered by the U.S. Department of Transportation (DOT).
The DOT has set regulations on the classification of hazardous
substances as well as standards for packing, labeling, and
shipping of the different classes of materials.
° The National Institute for Chemical Studies (NICS) was formed
in early 1985 by West Virginia Governor Arch Moore to serve as
a "bridge" between the general public and the chemical industry,
and to assess the seriousness of known or perceived risks in the
Kanawha Valley to determine how effectively those risks are
being managed. NICS Safeguard Report, published and released
in November 1986, examines the onsite accident prevention
and risk management activities of Valley chemical companies.
- Local
° Kanawha County and the City of Charleston are cooperating
with the Kanawha Valley Emergency Planning Council (KVEPC)
in building a new, state-of-the-art, Metro Emergency Operations
Center, which will be the headquarters for all response
activities in the Kanawha Valley. The Center is scheduled
for opening in the Spring of 1987.
° The Kanawha County Office of Emergency Services has developed
one mile emergency planning zones (EPZs) around each major
chemical plant. These EPZs will become the focus of special
efforts in emergency response planning.
° Agreements have been reached with the Kanawha County Board of
Education and the Kanawha Regional Transit (KRT) to assure
their cooperation in the event of an emergency. Two way
radios have been placed on all KRT buses and 67 radios have
installed in Kanawha County schoolbuses.
A-2
-------
- Industry
0 The Kanawha Valley Emergency Planning Council (KVEPC) is an
area coalition of chemical companies whose purpose is to share
knowledge and resources for preventing and responding to
emergencies at member facilities. The KVEPC is also interested
in outside response in the Kanawha Valley and has produced the
"Path Forward" Report, which presents a point-by-point plan to
upgrade emergency response in the region.
0 In an attempt to prevent accidents before they occur, several
companies in the area are performing formal engineering risk
assessments at the design stage for each major high hazard
chemical process. In addition the American Institute of
Chemical Engineers Center for Chemical Process Safety (CCPS)
is planning to publish a manual of engineering guidelines for
detection, containment and isolation of accidental releases,
which could lead to some additional safety modifications and
retrofitting at Valley chemical plants.
0 Several accident preventative quality control programs are
being conducted by many companies in the Valley. The Process
Hazard Management Review Program at DuPont and Union Carbide
systematically identifies and evaluates the impacts of potential
process hazards. The Tentative Process Amendment Program at
Monsanto is used to plan and implement significant changes or
major modifications to exsisting chemical processes, and FMC
has commissioned a full scale corporated-led process hazards
review project at their facilities.
° Some chemical companies are effectively reducing the risk
of an accidental release occuring by reducing their inventories
of hazardous materials. Some plants utilize the "just-in-time"
method of scheduling raw material or product shipments in order
to minimize storage. Other chemical plants are specifically
designed to completely use up quantities of chemicals immediately
subsequent to receiving them.
° Since most accidents at chemical plants are caused by human
error, many facilities are upgrading their operator training
programs. Union Carbide has plans to implement an operator
training, testing, and certification program which utilizes
a simulator unit equivilant to those used by airline pilots.
Monsanto and DuPont are using on-the-job evaluation techniques
to rate operator performance. The results are communicated
to workgroups for corrections and follow-up.
A-3
-------
II. Indoor Air
o The Kanawha Valley Toxics Study did not include sampling of
indoor air. The EPA has studied indoor air in six U.S. cities
over the past few years and exposure data has been remarkably
consistent from city to city. Charleston was not involved
in these TEAM (Total Exposure Assessment Methodology) studies,
however, relative assumption indoor air quality may be made
due to the data consistency.
o The EPA TEAM Study is the largest of the personal exposure
studies performed to date. The study measured personal
exposure of about 600 residents of five geographical areas
between 1980-84. Concurrent outdoor air measurements were
made near 200 of the homes, allowing direct comparison of
indoor-outdoor levels for the TEAM target compounds, which
included many of the Kanawha target compounds. In essentially
all cases, the indoor levels of the target chemicals outweighed
the outdoor levels, often by factors of 2-5. These basic
findings have been repeated in a number of studies by different
investigators in different countries.
o The sources of the high indoor and personal exposures have
been identified for some carcinogenic target compounds (benzene,
chloroform, tetrachloroethylene) but not for others (trichloro-
ethylene, carbon tetrachloride). The major source of benzene
exposure is undoubtedly mainstream cigarette smoke, which
contains about 55 ug/cigarette. Other main sources of benzene
include environmental tobacco smoke and auto-related activities
such as extended commuting time and filling a gas tank at a
service station.
o The major source of airborne chloroform exposure appears to
be the use of hot water in the home. Recent unpublished work
by EPA/ORD indicates that washing clothes or dishes is corre-
lated with higher levels of chloroform in the home than
taking showers or tub baths. In addition, use of chlorine
bleaches may create and release considerable quantities of
chloroform. The major source of tetrachloroethylene is wearing
and storing dry-cleaned clothes. Levels in homes remain
elevated for at least a week following storage of freshly
dry-cleaned clothes in a closet.
o It would seem that, for the chemicals described above,
exposures from indoor sources and activities are normally
dominant over exposures from outdoor concentrations, 3ven
when one's residence is in the heart of a chemical manufacturing
district. With the exception of chloroform, the outdoor
levels observed in the Kanawha Valley during this study are
typical of those observed in many other areas of the country.
Therefore it may be concluded that the indoor concentrations
and resulting personal exposures of Kanawha Valley residents,
as in all other areas of the country so far studied, are
important in terms of long-term risks and should not be
taken lightly when considering total exposure risk reductions.
A-4
-------
Ill. Epidemiological Health Effects Study
o The Kanawha Valley Toxics Screening Study was not epidemio-
logical in nature; that is, it did not examine current rates
or patterns of disease incidence in the resident population
and attempt to link these health conditions with previous
levels of chemical exposure.
o The U.S. Congress through the work of Senator Robert Byrd
appropriated $250,000 to explore possible designs for an
epidemiological health effects study in the Kanawha Valley.
o The study design is being conducted by a consortium of the
National Institute for Chemical Studies (NICS), the West
Virginia Department of Health and the Harvard University
School of Public Health (HSPH) through a grant administered
by EPA.
o Phase I of this study was initiated in October, 1986 and will
be implemented for a period of approximately fourteen months
ending in December, 1987. This effort includes: identifying
chemicals of significant concern in the ambient atmosphere;
identifying potential health endpoints relating to chemical
exposure versus time; and defining exposure gradients in the
study population which would optimize the sensitivity and
validity of the study.
o Additional Phase I activities include: conducting ambient
air monitoring, and possibly indoor air and water quality
monitoring to better define an exposure gradient; organizing
two workshops convening experts in the fields of toxicology
and epidemiology to meet with local health officials to
review and interpret available health and exposure data;
and establishing a process to continually assess new health
and exposure information from a toxicological perspective.
IV. Occupational Risk
o The Occupational Safety and Health Administration (OSHA) has
the responsibility for assuring that workers in the Kanawha
Valley are provided with safe and healthful working conditions.
OSHA works toward this goal by conducting safety and health
inspections in the workplace.
o OSHA has established permissable exposure limits for approxi-
mately four hundred chemical substances. These permissable
exposure limits (PELs) are eight hour time weighted average
airbourne concentrations to which most workers may be exposed on
a daily basis without experiencing adverse health effects. This
simply means that a chemical may exceed a PEL for a short period
of time, however, the chemical may not exceed the PEL when taking
the average concentration over an eight hour period. For a
limited number of chemicals ceiling levels (levels which cannot
be exceeded for any period of time) have also been established.
A-5
-------
o In order to further address the unique potential hazards that
exist in the chemical industry, OSHA developed a Pilot Special
Emphasis Program for inspections of certain chemical establish-
ments. These inspections are designed to focus on process
safety, backup system, preventive maintenance programs and
emergency response procedures. OSHA developed and used a
High Hazard Chemical List to further concentrate on those
facilities that produce high risk chemicals. The Charleston
OSHA office was one of ten offices chosen to participate in
the pilot program which resulted in some notable inspections
conducted in chemical plants in the Kanawha Valley.
o Four seperate inspections were conducted at the Union Carbide
plant in Institute between December 1984 and April 1986. The
inspections covered such things as the methyl isocyanate
manufacturing and handling operations and five other process
units in the plant. Emergency response procedures were also
reveiwed by OSHA.
o Two inspections of the Union Carbide South Charleston Plant
were concluded in August 1985 and October 1986. The first
inspection was a result of a alled^ed release of mesityl
oxide and acetone in which OSHA recommended design improvements
that could be made to the distillation equipment. The second
inspection scrutinized the acrylonitrile processing unit in
which several respirator violations were cited, however,
there were no potential exposures noted from company sampling
data.
o The Artel Chemical Co. (formerly Fike Chemical) was inspected
between May - July 1986 because of an accidental release of
chlorine gas from the plant. Actual exposure levels could
not be documented and no OSHA sampling was conducted.
o The Occidental Chemical Corporation inspection ending in
December 1986 was initiated as a result of an accidental
release of chlorine, carbon tetrachloride, chloroform and
methylene chloride. Again actual exposure levels could not
be documented. Process safety deficiencies were documented
and cited during the inspection.
A-6
-------
APPENDIX B
HISTORY OF DEVELOPMENT AND EMPLOYMENT IN KANAWHA VALLEY1
The history of the Kanawha Valley is extremely interesting, involving
names such as Daniel Boone, George Washington, William McKinley, and
Rutherford B. Hayes. Events such as the French and Indian War, the
Revolutionary War, the Civil War, and the two World Wars also played
a major role in shaping the development of the Valley.
Although not established by an act of the Virginia General Assembly
until 1794, Charleston's origins date back to the pre-Revolutionary
War days. Originating as a military outpost in the Virginia system
of the frontier, Charleston and its surrounding communities developed
into a major industrial and commercial area strategically located near
the large urban areas of the East, South and Midwest.
Charleston's origins reach back to the closing days of the French and
Indian War. The first white man of record to view the site was Matthew
Arbuckle, who, in 1764, passed from the Greenbrier Levels (Lewisburg)
down the Kanawha Valley to a French trading post on the site of Point
Pleasant.
In April, 1788, a company of Virginia Rangers under Colonel Charles
Clendenin erected a stockade fort at the corner of the present Brooks
Street and Kanawha Boulevard, officially known as Fort Lee, which was
a part of the frontier defense until the close of the Indian wars in
1795. In 1794, forty acres of the Clendenin holdings were laid out
in lots, and a town was authorized by the Virginia Assembly to be
known as Charleston in honor of Charles Clendenin. On January 19,
1818, Charleston, by name, was officially established.
The early years of Charleston were much interwoven with the story of
the great salt industry of the Kanawha Valley. As early as 1796, salt
was manufactured at the mouth of Campbells Creek, almost directly
opposite the homeplace of Daniel Boone, who resided at the present
Kanawha City from 1790 until 1795. This industry rose to great pro-
portions. By 1860, salt was shipped to all parts of the U.S. from
some 49 great furnaces. In addition to the salt industry, natural
curiosities known as "burning springs" were found, from which exuded
natural gas. Two such areas, above the suburb of Maiden, were embraced
in a "military survey" owned jointly by General George Washington
and General Andrew Lewis.
During the Civil War, the town of some 1,500 population was much
divided, and many citizens served in the Federal Army and many in
the Confederate Army. It was an important military post and the
nominal head of navigation for riyer transportation of troops. The
Battle of Scary was fought a few miles below the city in 1861, and
B-l
-------
Camp Piatt was located a few miles above. Fort Scammon, located in
the city, was an important post, and the remains may still be seen.
Camp White, on the "South Side", was occupied by many notable men,
including William McKinley and Rutherford B. Hayes.
The Chesapeake & Ohio Railroad was completed through the city in the
spring of 1873. This was followed by construction of the Ohio Central
Railroad, between 1880 and 1883, which is now part of the Consolidated
Rail Corporation. The Coal & Coke Railroad was opened in 1906, from
Charleston to Elkins, using in part an earlier railroad constructed
in the Elk River Valley. This rail system is now part of the Baltimore
& Ohio System.
The industrial and commercial history of Charleston falls logically
into three phases since its founding by the location of Fort Lee in
1788. The first era, from the early time of the city to as late as
1859, centers around the great salt industry. The second era saw
the development of Charleston as a market and wholesale town to
serve the rapidly developing coal mining industry in southern West
Virginia. The third, or present era is marked by the development of
Charleston and environs as one of the great chemical centers of the
U.S., based upon the manufacturing use of the natural resources
which surround the city.
The present era is further divided into the gas and glass period,
the war period and the synthetic period. The gas and glass period
was a result of the extensive development of oil and gas fields
around Charleston. The war period saw the location and expansion of
a number of manufacturing plants which were designed to be converted
easily into war munitions plants. The synthetic period, or the
present period, is marked by the location of internationally known
chemical industries in the Valley. Like earlier industries, most of
these are based upon the salt brine and other natural resources
underlying the Valley.
The presence of manufacturing plants, the nearby location of natural
resources such as coal, oil and gas, and the provision of facilities
for transportation, led to the location of the U.S. Naval Ordnance
Plant at South Charleston and the U.S. Government's Nitrocellulose
Plant at Nitro, both in 1917. As part of the war production effort
in 1942-43, the nation's then largest installations for the production
of bulk synthetic rubber were built at Institute.
The old Naval Ordnance Plant complex in South Charleston has been
completely renovated and is now called the Charleston Ordnance Center,
providing space for over 30 small companies as well as the Charleston
Stamping Plant of Volkswagen of America, Inc. The nitrocellulose
plants have been cleared from their original sites at Nitro, but
FMC, Monsanto and other major companies have developed facilities in
the same area. The bulk synthetic rubber property at Institute
today belongs to the Union Carbide Corporation, the Kanawha Valley's
largest employer.
While major industrial activity and employment have declined in
recent years (see TABLE 1 on next page), the Valley remains one of
the largest industrial complexes in the United States.
B-2
-------
TABLE B-l
ANNUAL AVERAGE
EMPLOYMENT TRENDS 1972 - 1983
CHARLESTON SMSA - KANAWHA & PUTNAM COUNTIES
1972 1973 1974 1975 1976 1977 1978 1979* 1980* 1981** 1982*** 1983****
LABOR FORCE 102,800 105,900 107,500 111,400 113,800 117,500 118,800 122,600 127,300 121,100 119,100 118,000
EMPLOYMENT 97,600 100,300 102,800 105,100 106,200 112,800 114,300 117,900 119,700 112,900 110,300 106,300
UNEMPLOYMENT 5,200 4,600 4,800 6,400 5,300 4,700 4,500 4,700 7,500 8,100 8,900 11,700
% UNEMPLOYMENT 5.1% 4.3% 4.3% 5.7% 4.7% 4.0% 3.8% 3.9% 5.9% 6.7% 7.4% 9.9%
SELECTED
SECTORS
WHOLESALE &
RETAIL TRADE
MANUFACTURING
GOVERNMENT
SERVICES
20,900 21,900
17,500 17,800
15,700 16,100
14,200 14,700
22,300 22,650
18,400 17,600
17,200 17,500
15,700 15,800
23,300 25,300
19,000 19,700
18,600 18,700
17,000 18,500
25,500 27,300
19,600 19,600
18,800 21,600
18,500 20,100
26,700 26,200
18,100 16,700
23,000 20,600
20,400 20,700
25,700 25,600
16,100 14,300
20,400 20,500
21,300 21,200
SOURCE: West Virginia Department of Employment Security
Labor & Economic Research Section
~Revised to March, 1980 Benchmark
**Revised to March, 1981 Benchmark
***Revised to March, 1982 Benchmark
****Preliminary, based on March, 1982 Benchmark
-------
Somewhat surprisingly, a look at Charleston's employment structure
reveals that the most identifiable enterprises - manufacturing - actually
account for a relatively small portion of the region's total economic
activity. This holds for both the data given in Table 1 above, and for
1984 figures. In fact, out of all persons employed in the Charleston
SMSA on the average for 1984:
Wholesale and retail trade employed more persons than any other
sector of the local economy in 1984. Approximately 27,600
persons had jobs in retail establishments, warehouses, and
distribution centers. This is equal to 25.9% of all employed
persons in 1984.
The second largest sector was services, which employed 22,100
persons, or 20.7% of all those employed in the region in that
year. The major group within this sector was health services,
which employed 9,200 persons.
The next largest sector of employment was government, which em-
ployed 21,300 persons or 20% of the total employed. While much
of this was state government employment, since Charleston is
the capital city, it also included a substantial number of city,
county and federal employees.
The fourth largest economic sector was manufacturing, with 13,300
persons, or 12.5% of all those employed. Chemical manufacturing
dominated this sector with over 9,800 employees.^
In terms of average weekly wages, however, manufacturing has accounted
for the highest incomes. Table 2 breaks down the average weekly wages,
by industry or business, for Putnam and Kanawha Counties for 1983, 1984,
and 1985.3
TABLE B-2
Putnam County
Kanawha County
1983 1984 1985
1983
1984
1985
Mining
Construction
Manufacturing
Trade
Retail
Finance
Services
Government
N.A. 252.04 357.92
N.A. 374.02 362.48
493.14 539.69 568.81
N.A. 257.31 260.94
N.A. 213.26 211.25
N.A. 292.52 299.19
N.A. 267.42 310.96
N.A. 302.33 325.28
N.A. 588.27 600.23
416.68 421.90 428.23
N.A. 571.44 604.10
N.A. 266.04 267.63
N.A. 203.90 203.93
N.A. 352.92 374.44
N.A. 304.42 314.90
N.A. 323.22 346.60
B-4
-------
With respect to specific industries identified in the Kanawha Valley
study, available employment figures for industries in the two counties
are given in Table 3.^
TABLE B-3
Putnam County
Company Employees
Allied Chemical
FMC
Monsanto
71
160
548
Kanawha County
Company
Diamond Shamrock.
(Occindental)
Dupont
FMC
Union Carbide
(South Charleston)
Rhone-Polenc
(Institute)
Union Carbide
(Technical Center)
Employees
74
1061
806
1172
1408
3629
In terms of employment outlook, it has already been indicated that the
region has been declining in both labor force and employment since 1980,
as compared to increases in both categories from 1972 up to, and including,
1980. Compared to the 1984 data given above, however, it appears that the
decline in the trade, government, and service industries has been reversed.
By comparison, the manufacturing sector has continued to decline, based
on the available data. The trickle down effect of an improved national
economy may have a positive effect on the local manufacturing sector.
However, it might be presumptuous to expect marked improvement.^
B-5
-------
APPENDIX C
SCIENCE ADVISORY BOARD REVIEW
INTEGRATED ENVIRONMENTAL MANAGEMENT SUBCOMMITTEE
° Final Comments - May 27, 1987
° List of IEM Subcommittee Members
-------
" ^ \
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
%«,0^0
-------
2
The Kanawha Valley Toxics Screening Study identified several
potentially serious risks which suggest the need for further analysis
and possibly risk management actions. We are now working with the
appropriate West Virginia agencies to develop specific plans for
responding to the risks and needs identified in the Study. Industry
within the Valley has shown a willingness to take voluntary steps to
reduce emissions where they contribute to significant risks. Other
ongoing work will also contribute to a better understanding of environ-
mental risks and control options, including the health effects feasibility
study now underway through the National Institute for Chemical Studies.
Again, T appreciate the efforts, conclusions, and insights of the 1EM
Subcormittee concerning the Kanawha Valley Study. I also look forward to
receiving your report on the full Integrated Environmental Management
Program in the near future.
cc: A. ."laTnee Barnes
Terry Yosie
IFW Subcommittee Member?
Jame? Seif
John Campbel1
Lee M. Thomas
A
C-2
-------
APPENDIX C
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. D C. 20460
May 27, 1987
SAB-BC-87-031
Honorable Lee M. Thomas
_ , . . . . OFFICE OF
Administrator THt .DMin>st».
U. S. Environmental Protection Agency
401 M Street, S. W.
Washington, D. C. 20460
Dear Mr. Thanas:
The Science Advisory Board's Integrated Environmental Management
Subcommittee has completed its review of EPA's Draft Kanawha Valley Toxics
Screening Study and is pleased to transmit its final report to you. The
Subcommittee met in public session on March 16, 1987 in Philadelphia,
Pa., to review the study. During March 11-13, 1987, three representatives
of the Subcommittee visited the Kanawha Valley to become more familar
with its environmental problems.
The Subcommittee unanimously concludes that the Kanawha Valley study "
represents an important component of EPA's overall effort to develop
methodologies to define public health and environmental priorities.
Studies such as this provide valuable technical challenges and experiences
to EPA staff, particularly those working in regional offices. And,
finally, they provide a valuable means for developing closer working
relationships with state and local officials and the general public.
This letter is the Subcommittee's second ccnmunication to you. On
July 30, 1986 it expressed "many concerns about the ability of the current
study to satisfy a number of technical issues. A chief concern is the
incongruity between [the study's] .... objectives and the fact that the
study design itself is not an integrated multimedia effort, nor a response
to Bhopal."
Since the transmittal of that letter, EPA staff have modified the
study's objectives and technical design, and have conducted supplementary
analyses to support the revised objectives and design. In general, the
Subcommittee believes that the staff have made appropriate responses to its
major concerns. Hie study reaches a number of scientifically supportable
conclusions about health risks from cancer in the Kanawha Valley. The
study also points EPA and other interested parties in a direction for
conducting further analyses of problems related to accidental releases of
pollutants and acute health effects.
Specific issues addressed durirxg the Subcommittee's review include:
the study's objectives and scope; pollution sources; pollution transport
and fate by media; health effects; risk communication; and recommendations
for additional follow-up efforts. Attachment A presents additional, more-
detailed recommendations for modifying the current study and future
activities in the Kanawha Valley. Attachment B lists the Subcommittee
members.
C-3
-------
- 2 -
In general, the Subcommittee views the Draft Kanawha Valley Toxics
Screening Study as one step of a continuing process to assess risks. The
current study addresses chronic health exposures to carcinogens which
represent one of many public health concerns in the Valley. As a follow-up
to the current study, the Subcommittee recommends two additional steps
that include:
o Expanded monitoring of air toxics, and use of monitored values
to obtain more precise estimates of exposure and health risks.
o Greater focus on accidental releases and fugitive emissions as
areas of public health concern.
The Subcommittee appreciates the opportunity to conduct an independent
scientific review of these important public health issues in the Kanawha
Valley. We request that EPA formally respond to our scientific advice.
Sincerely,
Ronald Wyzga, Chairman *
Integrated Environmental
Management Subcommittee
Science Advisory Board
Norton Nelson, Chairman
Executive Committee
Science Advisory Board
C-if
-------
SAB-EC-87-031
MAJOR FINDINGS AND RECOMMENDATIONS
OF THE
INTEGRATED ENVIRONMENTAL MANGEMENT SUBCOMMITTEE
ON THE
U. S. ENVIRONMENTAL PROTECTION AGENCY'S DRAFT
KANAWHA VALLEY TOXICS SCREENING STUDY
INTEGRATED ENVIRONMENTAL MANAGEMENT SUBCOMMITTEE
SCIENCE ADVISORY BOARD
U. S. ENVIRONMENTAL PROTECTION AGENCY
May, 1987
-------
U. S. ENVIRONMENTAL PROTECTION AGENCY
NOTICE
This report has been written as a part of the activities of
the Science Advisory Board, a public advisory group providing
extramural scientific information and advice to the Administrator
and other officials of the Environmental Protection Agency. The
Board is structured to provide a balanced expert assessment of
scientific matters related to problems facing the Agency. This
report has not been reviewed for approval by the Agency, and
hence the contents of this report do not necessarily represent
the views and policies of the Environmental Protection Agency,
nor of other agencies in the Executive Branch of the Federal
government, nor does mention of trade names or canmercial products
constitute endorsement of recanmendation for use.
C-6
-------
Study Objectives and Scope
The objectives of the Kanawha Valley study are limited, but reasonably
well-defined. In most instances, the study seeks to derive an upper bound
for the health risks associated with airborne carcinogens for which
EPA's Cancer Assessment Group has derived potency estimates. Other
potential carcinogens are minimally considered, and the health risks of
non-carcinogens, including those risks associated with the accidential
release of chemicals such as occurred at Bhopal, are not considered.
Hence, the health assessment of airborne toxics is far fran complete, but
this is clearly articulated in the study report. Available resources did
not allow a more comprehensive assessment.
The study attempts "to develop a sense of potential public health
concerns" associated with carcinogens in drinking water, surface water
and hazardous wastes. The efforts are not multimedia efforts, but medium-
specific efforts based upon very limited data; thus, conclusions frcm
these efforts are subject to considerable uncertainty.
Sources
The air analyses depend very heavily upon an emissions inventory of"
some 450 substances developed by the West Virginia Air Pollution Control'
Ccnmission (APCC). The inventory is as extensive and comprehensive as
any other available information. Nevertheless, there exist seme
uncertainties in the inventory, particularly with respect to fugitive
emissions, which the study identifies as a major source of health risk in
seme Kanawha Valley communities. The possibility that the inventory is
incomplete is also suggested by the fact that ethylene oxide was not included
for either the Belle or Nitro communities despite some limited monitoring
evidence that it may be present. If a compound was not in the inventory
it was not included in subsequent EPA modeling. This discrepancy underlines
the need for including ethylene oxide in future monitoring programs.
The drinking water and surface water analyses depend upon monitored
levels of toxics in water supply systems and fish fillets, respectively.
Data are limited to a subset of all public water suppliers, with no private
well samples, and to a very small number of fish sampled fran only one
location for a very limited number of toxic substances. The hazardous
waste inventory is based upon a priority pollutant screening of inventories
for a subset of RCRA and potential CERCLA sites. No information was
available on the total quantity and overall composition of toxic wastes
that may be entering surface or ground water. For this reason alone, the
results of this part of the study are, at best, suggestive.
Transport and Exposure
The transport models used in the studies generally appear to be
congruent with the study objectives. The air transport modeling addresses
the concerns of the Subcommittee in its July 30, 1986 letter, although better
C-T
-------
- 2 -
documentation of this modeling is needed. There is a factor of two
uncertainty on point source air emissions and another uncertainty of a
factor of two in dispersion modeling. The current modeling efforts do
not address these potential uncertainties, although "worst case" scenarios
should recognize their existence. Drinking water exposure was estimated
by assuming that individuals consume two liters of the water delivered to
their neighborhoods. Similar assumptions are often made in risk assessments.
The surface water and hazardous waste studies are greatly hampered by a
lack of data, making large assumptions necessary to estimate exposure to
toxics.
Health Effects
The study evaluated 20 known or suspected cancer causing chemicals
from the West Virginia APCC inventory of more than 450 compounds. The
Subcommittee concludes that the current study provides useful information
on health effects from cancer and environmental loadings of these 20
compounds. After finalizing the current study, EPA should conduct additional
efforts that include:
o Using the APCC inventory and information on toxicity to evaluate^
the potential health effects of some of the remaining compounds.
Of the remaining 430 or so compounds, relatively few merit further
attention, but EPA and APCC should work together to identify
compounds that need additional evaluation. These should be
identified by defining the set of those compounds to which scxne
exposure may be likely at known toxic levels.
o Broadening the health endpoints of concern to include non-cancer
and acute effects. Concern about the potential effects from
acute releases is strong within the community; hence, some
priority should be given to addressing this issue. The
methodologies used to address these endpoints require further
development, particularly in estimating the effects of accidential
releases. Some fault-tree or alternative analysis should be
designed to address this possibility. Experts from other groups
within the EPA should be enlisted in this effort.
o Incorporating frequency plots of pollutant concentrations versus
time, in addition to stating average pollutant concentrations.
o Assessing the conversion of reference doses from the ingestion to
the inhalation pathway, where reference dose information for the
inhalation pathway is not available.
o Evaluating whether to develop or use biological markers for health
assessment.
o Comparing risks from high mass emissions of pollutants with low
toxicity, with low mass emissions of pollutants with high toxicity
as a means to identify priority risk management needs.
C-8
-------
- 3 -
o Exploring other potentially useful sources of data for compounds
of concern, including monographs prepared by the International
Agency for Research on Cancer, Health Effects Profiles developed
by the Office of Research and Development, Reportable Quantities
for hazardous compounds and gaining access to information through
the community right-to-know provision of Superfund.
Ccmunication of Risks
The Subcommittee encourages EPA to continue its*efforts of working
with officials and citizens of the Kanawha Valley to update them on the
sources and magnitude of risks they experience. In particular, EPA should
seek to further improve its presentation of technical information to
better enable lay persons to understand the results of technical analyses
and to ensure it is understood that the risk numbers reflect upper bound
estimates. Clarification of the latter issue is also needed in the
executive summary of the study.
It is important for citizens, scientists and public officials to
understand that the principal value of the Kanawha Valley study is as a
screening study of airborne carcinogens. As the study acknowledges,
a screening study should strive to ensure that all potential risks are
identified even at the expense of calling attention to risks that subsequent
analysis may not confirm, or will be less than indicated in the screening
study. Accordingly, assumptions in screening studies are conservative in
nature; assumptions should be avoided that might cause potential risks to
be ignored. Within the stated scope of the study, conservative assumptions
are made; for example, individuals are assumed to be exposed continuously
to ambient outdoor levels of industrially emitted toxics and upper bound
risk estimates are given. There are a few instances, however, where the
study did not rigorously pursue conservative assumptions. These include
potential uncertainties or omissions in the emissions inventory. The
study suggests that point estimates could be too small (or too large) by
a factor of two. For fugitive emissions it could be greater. It is
important that these uncertainties and their likely direction be clearly
articulated in the report along with a discussion about whether additional
scenarios are necessary to consider these uncertainties.
In addition, the air quality models are equally likely to under-and-
over predict ambient concentrations. The biases of the models are fairly
predictable. Exposures are likely underestimated at the peaks of ridges
where the river turns and when overlapping models were not used. On the
other hand, the use of the Box model probably overpredicts exposure in
seme neighborhoods on the Valley floor, which are not adjacent to emissions
sources. Although it is to the study's credit to have implemented two
different modeling approaches to estimate exposure, further discussion
in the report is merited on the potential model biases and on their
implications for the risk estimates.
p_n
-------
Attachment A
ADDITIONAL RECOMMENDATIONS TO THE AGENCY
A. Current Repofrt
1. The technical assumptions for the underlying transport models
should be documented and made accessible to readers of the
report.
2. Given the comprehensive nature of the airborne toxic risk
assessment in contrast to the rudimentary nature of the other
three studies, it may be desirable to more cle'arly separate the
air toxic studies from the others; moreover, the various studies
are undertaken for differently defined geographic areas.
3. The risk estimate bounds are probably more clearly defined than
in most similar documents; nevertheless, further clarification
may be necessary. Cases could be presented as
-------
The hazardous waste data considered are very limited. CERCLA
requirentents can perhaps provide sane useful information. Other
parts of EPA should be enlisted to improve the source inventory
for these data. Analysis of historical operations and land use
may also be useful to characterize the types of chemicals in
waste sites. The fundamental approach to consider risk from
hazardous waste should be replaced, however, by one that examines
specific waste sites.
Increased monitoring data can aid the analysis of drinking water,
surface water, and ground water. For chemicals of concern in the
Valley, such efforts should be instituted to help ensure that no
major problems are overlooked.
Health surveys and measurement of biological markers could provide
some validation of the estimated health profile of the Valley.
Such efforts will not, however, be useful when incremental risk
estimates are small.
C-U
-------
Attachment B
U. S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
' Integrated Environmental Management Subcommittee
Dr. Ronald Wyzga, Chairman
Electric Power Research Institute
3412 Hillview Avenue
Post Office Box 1041
Palo Alto, California 94303
Dr. Stephen L. Brown
Environ Corporation
1000 Potctnac Street, N. W.
Washington, D. C. 20007
Dr. Thomas Clarkson
University of Rochester Medical Center
Box RBB
575 Elmwood Avenue
Rochester, New York 14642
Terry F. Yosie, Director
Science Advisory Board
U.* S. Environmental Protection
Agency
401 M Street, S. W.
Washington, D. C. 20460
Dr. Thomas Burke
New Jersey Department of Health
Div. of Occupational & Environ.
Health CN 360
Trenton, New Jersey 08625
Dr. Yoram Cohen
Chemical Engineering Department
U. C. L. A. Room 553L
Boelter Hall
Los Angeles, California 90024
Dr. Herbert H. Cornish
830 West Clark Road .
Ypsilanti, Michigan 48198
Dr. Terry Davies
Wbrld Wildlife Fund
The Conservation Foundation
1255 23rd Street, N. W.
Washington, D. C. 20037
Dr. Robert Frank
Johns Hopkins University
Department of Environmental
Health Sciences - JHSHPH
615 North Wolfe Street
Baltimore, Maryland 21205
Dr. Rolf Hartung
School of Public Health
University of Michigan
Ann Arbor, Michigan 48109
Dr. James Gruhl
7610 N. Christie Drive
Tucson, Arizona 85718
Dr. Paul Lioy
University of Medicine and
Dentistry of New Jersey
675 Hoes Lane
Robert Wood Johnson
Medical School
Piscataway, New Jersey 08854
C-12
-------
- 2 -
Dr. Francis McMichael
The Blenko Professor of
Environmental Engineering
Department of Civil Engineering
Carnegie Mellon University
5000 Forbes Avenue
Porter Hall 123A
Pittsburgh, Pennsylvania 15213
Dr. Ellen Silbergeld
Chief Toxics Scientist
Environmental Defense Fund
1525 18th Street, N. W.
Washington, D. C. 20036
Dr. Warner North
Principal, Decision Focus, Inc.
Los Altos Office Center
Suite 200
4984 El Camino Real
Los Altos, California 94022
C-13
-------
APPENDIX D
METHODLOGY FOR EVALUATING HEALTH RISKS: A GENERAL OVERVIEW
EPA is charged with controlling the levels of exposure to thousands
of chemicals in the environment. Decisions about which chemicals to
control are based, in large part, upon the population risks associated
with exposure to those chemicals. This is because we want to control
first those problems that pose the greatest and most controllable
risks, and because the amount of resources we will apply to a problem
should bear some relationship to the health threat it presents.
Population risks in the simplest sense, are a function of two
basic factors: hazard and exposure. To cause a risk, a chemical has
to be both toxic (present an intrinsic hazard), and be present in the
human environment at some significant level (provide opportunity for
significant human exposure). Risk assessment interprets the evidence
on these two points, judging the probability of whether or not an
adverse effect may occur.
To judge whether a chemical may cause a particular health
effect, EPA must first consider the inherent toxicity of the chemical.
In practice, this has been done in two quite different ways, depending
on whether or not the chemical has been identified as a carcinogen,
a substance that may cause cancer in animals or people. Toxicological
research has led the Agency to assume that the exposure to any amount
of a carcinogen is associated with the possibility, however small,
that some person in the exposed population may develop cancer.
For many other health effects, in contrast, the Agency assumes
that there is some level of exposure that causes virtually no harm.
This level of exposure is called a "threshold." In almost all people,
exposures below the threshold should not cause adverse health effects,
while exposures above the threshold may lead to such effects in some
exposed ivdividuals.
For each of these groups of chemicals, the Agency has adopted
a standard process for estimating how potent a substance might be.
For non-carcinogens, the Agency uses the concept of the threshold to
determine Reference Doses, or RfDs (previously called ADIs).
D-l
-------
For carcinogens, the EPA Cancer Assessment Group (CAG) is
responsible for assessing dose-response relationships from
exposure to carcinogens. The problem is that most of the informa-
tion that we have about the effects of carcinogens relates to high
doses, in animal studies or human occupational exposures. Alsmost
all environmental exposures occur at much lower levels. For years,
the scientific community has struggled with the problem of extra-
polating effects at low doses from information about high doses.
Also, for many substances, the only information that we have is
from animal studies. Here a further uncertainty arises because
scientists are uncertain about how to apply quantitative information
from animal experiments to predict the effects of carcinogens on
people.
CAG must make assumptions based on the best available scienti-
fic evidence. It uses mathematical models that are based on scien-
tific understanding of biochemical mechanisms to "fill in" the
missing lowdose data. Intrinsically, the assumptions and applica-
tion of such models in developing estimates of low dose effects
results in some uncertainty. In general, when making a choice
among such assumptions, concern for the public health drives Agency
policy to select the one that generally turns out to be "conserva-
tive" or pessimistic, i.e., the one most protective of human health.
The Agency, therefore, has chosen to define the CAG.scores as a
"plaisible upper bound." That is, if the application of the so-called
"CAG numbers" predicts, say, three cancers from a certain exposure,
we state that it is not likely to be more than that, but may be
much less, or even zero.
CAG numbers undergo substantial review within the Agency.
In a similar fashion, the Agency has established a review process
for RFDs. CAG numbers are often reviewed further by EPA's Science
Advisory Board. In this process they are often modified as a
result of addiitional research findings. Over time, CAG's and
RFDs come to represent a consensus about what EPA thinks about the
tendency of various substances to produce adverse health effects.
As such they are used as the scientific underpinnings of regulatory
policy making, and although dose-response and risk assessment are
not perfect, they represent the best quantitative decision aid we
have for discriminating among risks.
-------
The Kanawha Valley study uses a risk assessment screening metho-
dology to evaluate and compare, in a very limited fashion, the poten-
tial health risks from exposure to a limited set of pollutants. The
purpose of this methodology is to permit a comparison of one risk
with another and to provide a general sense of the risk a substance
may present. This methodology does not make a definitive statement
concerning the absolute risk posed by a particular pollutant, source
or exposure pathway.
Risk to an individual is defined as the increased probability
that an individual exposed to one or more pollutants will experience
a particular adverse health effect during his or her lifetime because
of that exposure. Several measures of carcinogenic risk are used in
this analyses, including estimated risk to an individual, and incidence,
defined as the increased number of cases projected in a given population.
This risk screening methodology involves both a qualitative and
quantitative assessment of the potential carcinogenicity of the
selected pollutants. As a screening study, this analysis employs
both types of pollutants. The Carcinogen Assessment Group (CAG)
reviews the evidence of carcinogenicity for selected pollutants and
classifies pollutants as human carcinogens (Group A), probable human
carcinogens (Group B), possible human carcinogens (Group C), not classi-
fied as carcinogens due to inadequate evidence (Group D), and not
carcinogenic to humans (Group E).
For those chemicals in group A,B, and C, CAG provides quantitative,
conservatives upper-bound estimates of carcinogenic unit risk factors.
A unit risk factor allows the calculation of the estimate individual
risk posed by exposure by a chemical given certain exposure assumptions.
To calculate individual risks, a chemical's unit risk factor is
multiplied by the estimated concentration of the pollutant.
Individual Lifetime Risk = Unit risk factor (uq/m^)-! x Concentration (ug/m^)
The concentration is simply the predicted concentration of the
chemical in the ambient air. For this analysis, we use the long-term
average concentration predicted by the model. The unit risk factor
which provides a measure of the chemical potency for this analysis
assumes that an average person weighs 70 Kg and breathes 20 m^ of
air each day.
D-3
-------
SAMPLE CALCULATION
POSSIBLE RISK OF CANCER ASSOCIATED WITH
INHALATION OF CHLOROFORM IN
AMBIENT AIR
Estimated Average
Individual Lifetime Risk
Unit Risk Factor x
Average Concentration
Cancer Unit Risk Factor
for the Inhalation of
Chloroform
= 2.3 x 10-5 (ug/m^)-!
Concent ration
For this Example assume 1 ug/m^
of chloroform in the ambient air
Estimated Average
Individual Lifetime Risk
2.3 x 10"^ (ug/m^) x 1 (ug/m-*)
2.3 x 10"5
This estimated average individual risk, based on an upper bound
unit risk factor, indicates that, concervatively, there is approximately
a 2 in one hundred thousand increased chance of developing cancer
over a 70 year lifetime of constant exposure to this concentration of
chloroform. The actual risk could be considerably lower.
Incidence, another measure of risk, is the probable number of
cancer cases expected when a population is exposed. Incidence is
calculated by multiplying individual risk by the population number
ard can be annualized by dividing by 70 to obtain cases per year.
D-4
-------
APPENDIX F
RISK ASSESSMENT FOR NON-CANCER pFFECTS
REFERENCE DOSE (RfD):
DESCRIPTION AND USE IN HEALTH RISK ASSESSMENTS
PRINCIPAL AUTHOR:
Donald Barnes, Ph.D. (OPTS)
RfD WORK GROUP:
Donald Barnes. Ph.D. (OPTS)
Judith Bellin, Ph.D. (OSWER)
Christopher DeRosa, Ph.D. (ORD)
Michael Dourson, Ph.D. (ORD)*
Reto Engler, Ph.D. (OPTS)
Linda Erdreich, Ph.D (ORD)
Theodore Farber, Ph.D. (OPTS)
Penny Fenner-Crisp, Ph.D. (ODW)
Elaine Francis, Ph.D. (OPTS)
George Ghali, Ph.D. (OPTS)
Richard Hill. M.D., Ph.D. (OPTS)
* Co-Chair
Stephanie Irene, Ph.D (OPTS)
William Marcus, Ph.D. (OW)
David Patrick, P.E., B.S. (OAR)
Susan Perlin, Ph.D. (OPPE) .
Peter Preuss, Ph.D. (ORD)*
Aggie Revesz, B.S. (OPTS)
Reva Rubenstein, Ph.D. (OSWER)
Jerry Stara, D.V.M., Ph. D (ORD)
Jeanette Wiltse, Ph.D. (OPTS)
Larry Zaragosa, Ph.D. (OAR)
E-l
-------
CONTENTS
I. INTRODUCTION A-3
A. Background A-3
B. Overview A-3
II. TRADITIONAL APPROACH TO ASSESSING SYSTEMIC (NONCARQNOGENIC) TOXICITY A-4
A. The Traditional Aproach A-4
B. Some Difficulties in Utilizing the Traditional Approach A-5
1. Scientific Issues A-5
2. Management-related Issues . A-5
a. The use of the term "safety factor" A-5
b. The implication that any exposure in excess of the ADI is
"unacceptable" and that any exposure less than the ADI is
"acceptable" or "safe" A-5
c. Possible limitations imposed on risk management decisions A-5
d. Development of different ADIs by different programs A-6
III. EPA ASSE5SMENT OF RISKS ASSOCIATED WITH SYSTEMIC TOXICITY A-6
A. Hazard Identification A-6
1. Evidence A-6
a. Type of effect A-6
b. Principal studies A-7
c. Supporting studies A-7
d. Route of exposure A-8
e. Length of exposure A-8
f. Quality of the study A-8
2. Weight-of-Evidence Determination A-8
B. Dose-Response Assessment A-9
1. Concepts and Problems A-9
2. Selection of the Critical Data A-9
a. Critical study A-9
b. Critical data A-10
c. Critical end point A-10
3. Reference Dose (RfD) A-10
C. Exposure Assessment A-11
D. Risk Characterization A-12
IV. APPLICATION IN RISK MANAGEMENT A-12
V. OTHER DIRECTIONS A-13
VI. HYPOTHETICAL, SIMPLIFIED EXAMPLE OF DETERMINING AND USING RfD A-14
E-2
-------
I. INTRODUCTION
This concept paper describes the U.S. Environmental Protection Agency's principal approach to and
rationale for assessing risks for health effects other than cancer and gene mutations from chronic
chemical exposure. By outlining principles and concepts that guide EPA risk assessment for such
systemic* effects, the report complements the new risk assessment guidelines, which describe the
Agency's approach to risk assessment in other areas (carcinogenicity, mutagenicity, developmental
toxicity, exposure, and chemical mixtures.)
A. Background
Chemicals that give rise to toxic end points other than cancer and gene mutations are often referred
to as "systemic toxicants" because of their effects on the function of various organ systems. It should
be noted, however, that chemicals which cause cancer and gene mutations also commonly evoke
other toxic effects (systemic toxicity). Generally, based on our understanding of homeostatic and
adaptive mechanisms, systemic toxicity is treated as if there is an identifiable exposure threshold
(both for the individual and for the population) below which effects are not observable. This
characteristic distinguishes systemic end points from carcinogenic and mutagenic end points, which
are often treated as nonthreshold processes.
Systemic effects have traditionally been evaluated in terms of concepts such as "acceptable daily
intake" and "margin of safety." The scientific community has identified certain limits on some of
these approaches, and these limits have been borne out in EPA's experience. Nonetheless, EPA is
called upon to apply these concepts in making and explaining decisions about the significance for
human health of certain chemicals in the environment.
To meet these needs, the RfD Work Group has drawn on traditional concepts, as well as on
recommendations in the 1983 National Academy of Sciences (NAS) report on risk assessment, to
more fully articulate the use of noncancer, nonmutagenic experimental data in reaching decisions
on the significance of exposures to chemicals. In the process, the Agency has coined new terminology
to clarify and distinguish between aspects of risk assessment and risk management. EPA has tested
and implemented these innovations in developing consistent information for several recent
regulatory needs, for instance under RCRA.
B. Overview
This Appendix consists of four parts in addition to this introduction. In Section II, much of the
traditional information on assessing risks of systemic toxicity is presented, with the focus on the
concepts of "acceptable daily intake (ADI)" and "safety factor (SF)." Issues associated with these
approaches are identified and discussed.
In Section III, the Agency's approach to assessing the risks of systemic toxicity is presented in the
context of the NAS scheme of risk assessment and risk management in regulatory decision-making.
This approach includes recasting earlier ADI and SF concepts into the less value-laden terms
"reference dose (RfD)" and "uncertainty factor (UF)." A new term, "margin of exposure,"** as
utilized in the EPA regulatory context, is introduced to avoid some of the issues associated with the
traditional approach.
Section IV examines how these new concepts can be applied in reaching risk management decisions,
while Section V briefly mentions some of the additional approaches the Agency is using and
exploring to address this issue. Section VI provides a sample RfD calculation.
*ln this document the term "systemic" refers to an effect other than carcinogenicity or mutagenicity induced by a toxic
chemical.
"In this Appendix, the ratio of the NOAEL to the estimated exposure (often referred to as "margin of safety") is referred to
as the "margin of exposure (MOE)" in order to avoid confusion with the original use of the term "margin of safety" in
pharmacology (i.e., the ratio of the toxic dose to the theraputic dose) and to avoid the use of the value-laden term "safety."
E-3
-------
II. TRADITIONAL APPROACH TO ASSESSING SYSTEMIC (NONCARCINOGENIC) TOXICITY
The Agency's approach to assessing the risks associated with systemic toxicity is different from that
for the risks associated with carcinogenicity. This is because different mechanisms of action are
thought to be involved in the two cases. In the case of carcinogens, the Agency assumes that a small
number of molecular events can evoke changes in a single cell that can lead to uncontrolled cellular
proliferation. This mechanism for carcinogenesis is referred to as "nonthreshold," since there is
essentially no level of exposure for such a chemical that does not pose a small, but finite, probability
of generating a carcinogenic response. In the case of systemic toxicity, organic homeostatic,
compensating, and adaptive mechanisms exist that must be overcome before the toxic end point is
manifested. For example, there could be a large number of cells performing the same or similar
function whose population must be significantly depleted before the effect is seen.
The threshold concept is important in the regulatory context. The individual threshold hypothesis
holds that a range of exposures from zero to some finite value can be tolerated by the organism with
essentially no chance of expression of the toxic effect. Further, it is often prudent to focus on the
most sensitive members of the population; therefore, regulatory efforts are generally made to keep
exposures below the population threshold, which is defined as the lowest of the thresholds of the
individuals within a population.
A. The Traditional Approach
In many cases, risk decisions on systemic toxicity have been made by the Agency using the concept of
the "acceptable daily intake (ADI)." This quantity is derived by dividing the appropriate "no-
observed-adverse-effect level (NOAEL)" by a "safety factor (SF)" as follows:*
ADI (human dose) = NOAEL (experimental dose) / SF (1)
The ADI is often viewed as the amount of a chemical to which one can be exposed on a daily basis
over an extended period of time (usually a lifetime) without suffering a deleterious effect. Often,
the ADI has been used as a tool in reaching risk management decisions; e.g., establishing allowable
levels of contaminants in foodstuffs and water.
Once the critical study demonstrating the toxic effect of concern has been identified, the selection of
the NOAEL derives from an essentially objective, scientific examination of the data available on the
chemical in question.
Generally, the SF consists of multiples of 10, each factor representing a specific area of uncertainty
inherent in the available data. For example, an SF may be developed by taking into account the
expected differences in responsiveness between humans and animals in prolonged exposure studies;
i.e., a 10- fold factor. In addition, a second factor of 10 may be introduced to account for variability
among individuals within the human population. For many chemicals, the resultant SF of 100 has
been judged to be appropriate. For other chemicals, with a less complete data base (e.g., those for
which only the results of subchronic studies are available), an additional factor of 10 (leading to an
SF of 1,000) might be judged to be more appropriate. On the other hand, for some chemicals, based
on well-characterized responses in sensitive humans (e.g., effect of fluoride on human teeth), an SF
as small as 1 might be selected.
*A NOAEL is an experimentally determined dose at which there was no statistically or biologically significant indication of
the toxic effect of concern. In an experiment with several NOAELs, the regulatory focus is normally on the highest one,
leading to the common usage of the term NOAEL as the highest experimentally determined dose without statistical or
adverse biological effect. In some treatments, the NOAEL for the critical toxic effect is simply referred to as the NOEL. This
latter term, however, invites ambiguity in that there may be observable effects which are not of toxicologic significance;
i.e., they are not "adverse." In order to be explicit, this Appendix uses the term NOAEL and it refers to the highest NOAEL in
an experiment. Further, in cases in which a NOAEL has not been demonstrated experimentally, the formulation calls for use
of the "lowest-observed-adverse-effect level (LOAEL)." In order to focus on the major concepts, however, we will use
NOAEL as a general example.
E-4
-------
While the original selection of SFs appears to have been rather arbitrary (Lehman and Fitzhugh,
1954)*, subsequent analysis of data as reviewed by Dourson and Stara (1983) lends theoretical (and
in some instances experimental) support for their selection. Further, some scientists, but not all,
within the EPA interpret the absence of widespread effects in the exposed human populations as
evidence of the adequacy of the SFs traditionally employed.
B. Some Difficulties in Utilizing the Traditional Approach
1. Scientific Issues
While the traditional approach has performed well over the years and the Agency has sought to be
consistent in its application, observers have identified scientific shortcomings of the approach.
Examples include the following:
• By focusing on the NOAEL, information on the shape of the dose-response curve is ignored. Such
data could be important in estimating levels of concern for public safety.
• As scientific knowledge is increased and the correlation of precursor effects (e.g., enzyme
induction) with frank toxicity becomes known, questions about the selection of the appropriate
"adverse effect" arise.
• Guidelines have not been developed to take into account the fact that some studies have used
larger numbers of animals and, hence, are generally more reliable than other studies.
These and other "generic issues" are not susceptible to immediate resolution, because the data base
needed is not yet sufficiently developed or analyzed. Therefore, these issues are beyond the scope of
this Appendix. However, the Agency has established a work group to consider them.
2. Management-related Issues
a. The use of the term "safety factor"
The term "safety factor" suggests, perhaps inadvertently, the notion of absolute safety, i.e., absence
of risk. While there is a conceptual basis for believing in the existence of a threshold and "absolute
safety" associated with certain chemicals, in the majority of cases a firm experimental basis for this
notion does not exist.
b. The implication that any exposure in excess of the ADI is "unacceptable" and that any exposure
less than the ADI is "acceptable" or "safe"
In practice, the ADI is viewed by many as an "acceptable" level of exposure, and, by inference, any
exposure greater than the ADI is seen as "unacceptable." This strict demarcation between what is
"acceptable" and what is "unacceptable" is contrary to the views of most toxicologists, who typically
interpret the ADI as a relatively crude estimate of a level of chronic exposure not likely to result in
adverse effects to humans. The ADI is generally viewed as a "soft" estimate, whose bounds of
uncertainty can span an order of magnitude. That is, within reasonable limits, while exposures
somewhat higher than the ADI are associated with increased probability of adverse effects, that
probability is not a certainty. Similarly, while the ADI is seen as a level at which the probability of
adverse effects is low, the absence of risk to all people cannot be assured at this level.
c. Possible limitations imposed on risk management decisions
Awareness of the "softness" of the ADI estimate (see b. above) argues for careful case-by-case
consideration of the implications of the toxicological analysis as it applies to any particular situation.
To the degree that AD!s generated by the traditional approach are the determining factors in risk
'Lehman, A.J. and Fitzhugh, O.G. (1954). Association of Food Drug Officials. USQ Bulletin 18:33-35.
-------
management decisions, they can take on a significance beyond that intended by the toxicologist or
merited by the underlying scientific support.
Further, in administering risk/benefit or cost/benefit statutes, the risk manager is required to
consider factors other than risk (e.g., estimated exposures compared to the ADI) in reaching a
decision. The ADI is only one factor in a management decision and should not prevent the risk
manager from weighing the full range of factors.
d. Development of different ADIs by different programs
In addition to occasionally selecting different critical toxic effects, Agency scientists have reflected
their best scientific judgments in the final ADI by adopting factors different from the standard
factors listed in Table A-1. For example, if the toxic end point for a chemical in experimental animals
is the same as that which has been established for a related chemical in humans at similar doses, one
could argue for an SF of less than the traditional 100. On the other hand, if the total toxicologic data
base is incomplete, one could argue that an additional SF should be included, both as a matter of
prudent public policy and as an incentive to others to generate the appropriate data.
Such practices, as employed by a number of scientists in different programs, exercising their best
scientific judgment, have in many cases resulted in different ADIs for the same chemical. The fact
that different ADIs were generated (e.g., by adopting different SFs) can be a source of considerable
confusion when the ADIs are applied in risk management decisionmaking (see c. above). For
example, although they generally agree on the experimental data base for 2,3,7,8-TCDD, regulatory
agencies within the United States and around the world have generated different ADIs by selecting
different "safety factors"; specifically, 1000, 500, 250, and 100. These different ADIs have been used
to justify different regulatory decisions. The existence of different ADIs need not imply that any of
them is more "wrong"-or "right"-than the rest. It is more nearly a reflection of the honest
difference in scientific judgment.
These differences, which may reflect differences in the interpretation of the scientific data, can also
be characterized as differences in the management of the risk. As a result, scientists may be
inappropriately impugned,and/or perfectly justifiable risk management decisions may be tainted by
charges of "tampering with the science." This unfortunate state of affairs arises, at least in part,
from treating the ADI as an absolute measure of safety.
III. EPA ASSESSMENT OF RISKS ASSOCIATED WITH SYSTEMIC TOXICITY
In 1983, the National Academy of Sciences published a report which discusses the conceptual
framework within which regulatory decisions on toxic chemicals are made; see Figure A-1. The
determination of the presence of risk and its potential magnitude is made during the risk assessment
process, which consists of hazard identification, dose-response assessment, exposure assessment, and
risk characterization. Having been apprised by the risk assessor that a potential risk exists, the risk
manager answers the question: "What, if anything, are we going to do about it?"
A. Hazard Identification
1. Evidence
a. Type of effect
Exposure to a given chemical, depending on the dose employed, may result in a variety of toxic
effects. These may range from gross effects, such as death, to more subtle biochemical, physiologic,
or pathologic changes. The risk assessor considers each of the toxic end points from all studies
evaluated in assessing the risk posed by a chemical, although primary attention usually is given to
the effect exhibiting the lowest NOAEL, often referred to as the critical effect. For chemicals with a
limited data base, there may be a need for more toxicity testing.
E-6
-------
FIGURE A-1
Dose-response
_ Assessment (e.g. RfD)
^ \
Hazard
Idendification
Exposure
Assessment
X
Risk
Characterization
(e.g. criterion)
Regulatory
* Decision
(e.g. RgD, Standard)
//
Control /
Options /
Non-risk /
Analyses
b. Principal studies
Principal studies are those that contribute most significantly to the qualitative assessment of
whether or not a particular chemical is potentially a systemic toxicant in humans. In addition, they
may be used in the quantitative dose-response assessment phase of the risk assessment. These
studies are of two types:
(1) Human studies
Human data are often useful in qualitatively establishing the presence of an adverse effect in
exposed human populations. Further, when there is information on the exposure level associated
with an appropriate end point, epidemiologic studies can also provide the basis for a quantitative
dose-response assessment. Use of these latter data avoids the necessity of extrapolating from
animals to humans, and therefore, human studies, when available, are given first priority, with
cinimal toxicity studies serving to complement them.
In epidemiologic studies, confounding factors that are recognized can be controlled and measured,
within limits. Case reports and acute exposures resulting in severe effects provide support for the
choice of critical toxic effect, but they are often of limited utility in establishing a quantitative
relationship between environmental exposures and anticipated effects. Available human studies on
ingestion are usually of this nature. Cohort studies and clinical studies may contain exposure-
response information that can be used in estimating effect levels, but the method of establishing
exposure must be evaluated for validity and applicability.
(2) Animal studies
Usually, the data base on a given chemical lacks appropriate information on effects in humans. In
such cases, the principal studies are drawn from experiments conducted on non-human mammals,
most often the rat, mouse, rabbit, guinea pig, hamster, dog, or monkey.
c. Supporting studies
Supporting studies include information from a wide variety of sources. For example, metabolic and
other pharmacokinetic studies can provide insights into the mechanism of action of a particular
compound. By comparing the metabolism of the compound exhibiting the toxic effect in the animal
E-7
-------
with the metabolism found in humans, some light may be cast on the potential for the toxic
manifestation in humans or for estimating the equitoxic dose in humans.
Similarly, in vitro studies can provide insights into the compound's potential for biological activity,
although a definite connection to the human experience cannot be drawn. Under certain
circumstances, consideration of structure-activity relationships between the chemical under test and
the effects of structurally related agents can provide a clue to the biological activity of the former.
At the present time, these data are supportive, not definitive, in assessing risk. However, there is
focused activity aimed at developing more reliable in vitro tests to minimize the need for live-animal
testing. Similarly, there is increased emphasis on generating mechanism-of-action and
pharmacokinetic information as a means of increasing the fundamental understanding of toxic
processes in humans and nonhumans. It is expected that in the future these considerations will play a
larger role in our determination of toxicity of chemicals.
d. Route of exposure
The Agency often approaches the investigation of a chemical with a particular route of exposure in
mind; e.g., an oral exposure for a drinking water contaminant or a residue in food. Although the
route of exposure is oral in both cases, specific considerations may differ. For example, the
bioavailability of the chemical administered in food may differ from that when administered in
water or inhaled. Usually, the toxicologic data base on the compound does not include detailed
testing on all possible routes of administration.
In general, it is the Agency's view that the potential for toxicity manifested by one route of exposure
is relevant to any other route of exposure, unless convincing evidence exists to the contrary.
Consideration is always given to potential differences in absorption or metabolism resulting from
different routes of exposure, and whenever appropriate data (e.g., comparative metabolism studies)
are available, the quantitative impacts of these differences on the risk assessment are fully
delineated.
e. Length of exposure
The Agency is concerned about the potential toxic effects in humans associated with all possible
exposures to chemicals. The magnitude, frequency, and duration of exposure may vary considerably
in different situations. Animal studies are conducted using a variety of exposure durations (e.g.,
acute, subchronic, and chronic) and schedules (e.g., single, intermittent, or continuous dosing).
Information from all of these studies is useful in the hazard identification phase of risk assessment.
For example, overt neurological problems identified in high-dose acute studies tend to reinforce the
observation of subtle neurological changes seen in a low-dose chronic study. Special concern exists
for low-dose, chronic exposures, however, since such exposures can elicit effects absent in higher-
dose, shorter exposures, through mechanisms such as accumulation of toxicants in the organisms.
f. Quality of the study
Evaluation of individual studies in humans and animals requires the consideration of several factors
associated with a study's hypothesis, design, execution, and interpretation. An ideal study addresses
a clearly delineated hypothesis, follows a carefully prescribed protocol, and includes sufficient
subsequent analysis to support its conclusions convincingly.
In evaluating the results from such studies, consideration is given to many other factors, including
chemical characterization of the compound(s) under study, the type of test species, similarities and
differences between the test species and humans (e.g., chemical absorption and metabolism), the
number of individuals in the study groups, the number of study groups, the spacing and choice of
dose levels tested, the types of observations and methods of analysis, the nature of pathologic
changes, the alteration in metabolic responses, the sex and age of test animals, and the route and
duration of exposure.
2. Weight-of-Evidence Determination
As the culmination of the hazard identification step, a discussion of the weight-of-evidence
summarizes the highlights of the information gleaned from the entire range of principal and
E-8
-------
supporting studies. Emphasis in the analysis is given to examining the results from different studies
to determine the extent to which a consistent, plausible picture of toxicity emerges. For example, the
following factors add to the weight of the evidence that the chemical poses a hazard to humans:
similar results in replicated animal studies by different investigators; similar effects across sex, strain,
species, and route of exposure; clear evidence of a dose-response relationship; a plausible relation
between data on metabolism, postulated mechanism-of-action, and the effect of concern; similar
toxicity exhibited by structurally related compounds; and some link between the chemical and
evidence of the effect of concern in humans. The greater the weight-of-evidence, the greater one's
confidence in the conclusions drawn.
B. Dose-Response Assessment
1. Concepts and Problems
Empirical observation generally reveals that as the dosage of a toxicant is increased, the toxic
response (in terms of severity and/or incidence of effect) also increases. This dose-response
relationship is well-founded in the theory and practice of toxicology and pharmacology. Such
behavior is observed in the following instances: in quantal responses, in which the proportion of
responding individuals in a population increases with dose; in graded responses, in which the
severity of the toxic response within an individual increases with dose; and in continuous responses,
in which changes in a biological parameter (e.g., body or organ weight) vary with dose.
However, in evaluating a dose-response relationship, certain difficulties arise. For example, one must
decide on the critical end point to measure as the "response." One must also decide on the correct
measure of "dose." In addition to the interspecies extrapolation aspects of the question of the
appropriate units for dose, the more fundamental question of administered dose versus absorbed
dose versus target organ dose should be considered. These questions are the subject of much current
research.
2. Selection of the Critical Data
a. Critical study
Often animal data are selected as the governing information for quantitative risk assessments, since
available human data are generally insufficient for this purpose. These animal studies typically
reflect situations in which exposure to the toxicant has been carefully controlled and the problems of
heterogeneity of the exposed population and concurrent exposures to other toxicants have been
minimized. In evaluating animal data, a series of professional judgments are made that involve,
among others, consideration of the scientific quality of the studies. Presented with data from several
animal studies, the risk assessor first seeks to identify the animal model that is most relevant to
humans, based on the most defensible biological rationale, for instance using comparative
pharmacokinetic data. In the absence of a clearly most relevant species, however, the most sensitive
species (i.e., the species showing a toxic effect at the lowest admininistered dose) is adopted as a
matter of scientific policy at EPA, since no assurance exists that humans are not innately more
sensitive than any species tested. This selection process is made more difficult if animal tests have
been conducted using different routes of exposure, particularly if the routes are different from those
involved in the human situation under investigation.
In any event, the use of data from carefully controlled studies of genetically homogeneous animals
inescapably confronts the risk assessor with the problems of extrapolating between species and the
need to account for human heterogeneity and concurrent human exposures to other chemicals,
which may modify the human risk.
While there is usually a lack of well-controlled cohort studies that investigate non-cancer end points
and human exposure to chemicals of interest, in some cases human data may be selected as the
critical data (e.g., in cases of cholinesterase inhibition). Risk assessments based on human data have
the advantage of avoiding the problems inherent in interspecies extrapolation. In many instances,
use of such studies, as is the case with the animal investigations, involves extrapolation from
relatively high doses (such as those found in occupational settings) to the low doses found in the
environmental situations to which the general population is more likely to be exposed. In some
E-9
-------
cases, a well-designed and well-conducted epidemiologic study that shows no association between
known exposures and toxicity can be used to directly project an RfD (as has been done in the case of
fluoride).
b. Critical data
In the simplest terms, an experimental exposure level is selected from the critical study that
represents the highest level tested in which "no adverse effect" was demonstrated. This "no-
observed-adverse-effect level" (NOAEL) is the key datum gleaned from the study of the dose-
response relationship and, traditionally, is the primary basis for the scientific evaluation of the risk
posed to humans by systemic toxicants. This approach is based on the assumption that if the critical
toxic effect is prevented, then all toxic effects are prevented.
More formally, the NOAEL is defined in this discussion as the highest experimental dose of a chemical
at which there is no statistically or biologically significant increase in frequency or severity of an
adverse effect between individuals in an exposed group and those in its appropriate control. (See
also discussion in the footnote on page A-4). As noted above, there may be sound professional
differences of opinion in judging whether or not a particular response is adverse. In addition, the
NOAEL is a function of the size of the population under study. Studies with a small number of
subjects are less likely to detect low-dose effects than studies using larger numbers of subjects. Also,
if the interval between doses in an experiment is large, it is possible that the experimentally
determined NOAEL is lower than that which would be observed in a study using intervening doses.
c. Critical end point
A chemical may elicit more than one toxic effect (end point), even in one test animal, or in tests of
the same or different duration (acute, subchronic, and chronic exposure studies). In general, NOAELs
for these effects will differ. The critical end point used in the dose-response assessment is the one at
the lowest NOAEL.
3. Reference Dose (RfD)
In response to many of the problems associated with ADIs and SFs, which were outlined in Section II,
the concept of the "reference dose (RfD)" and "uncertainty factor (UF)" is recommended. The RfD is
a benchmark dose operationally derived from the NOAEL by consistent application of generally
order of magnitude uncertainty factors (UFs) that reflect various types of data used to estimate RfDs
(for example, a valid chronic human NOAEL normally is divided by an UF of 10) and an additional
modifying factor (MF), which is based on a professional judgment of the entire data base of the
chemical.* See Table A-1.
The RfD is determined by use of the following equation:
RfD = NOAEL/(UF x MF) (2)
which is the functional equivalent of Eq. (1). In general, the RfD is an estimate (with uncertainty
spanning perhaps an order of magnitude ) of a daily exposure to the human population (including
sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a
lifetime. The RfD is appropriately expressed in units of mg/kg-bw/day.
The RfD is useful as a reference point for gauging the potential effects of other doses. Usually, doses
that are less than the RfD are not likely to be associated with any health risks, and are therefore less
likely to be of regulatory concern. However, as the frequency of exposures exceeding the RfD
increases, and as the size of the excess increases, the probability increases that adverse effects may be
observed in a human population. Nonetheless, a clear conclusion cannot be categorically drawn that
all doses below the RfD are "acceptable" and that all doses in excess of the RfD are "unacceptable."
•"Uncertainty factor" is the new description applied to the term "safety factor" (see Page A-4). This new name is more
descriptive in that these factors represent scientific uncertainties, and avoids the risk management connotation of "safety."
The "modifying factor" can range from greater than zero to 10, and reflects qualitative professional judgements regarding
scientific uncertainties not covered under the standard UF, such as the completeness of the overall data base and the
number of animals in the study.
E-10
-------
TABLE A-1.
GUIDELINES FOR THE USE OF UNCERTAINTY FACTORS IN DERIVING REFERENCE DOSE (RfD)
Standard Uncertainty Factors (UFs)
Use a 10-fold factor when extrapolating from valid experimental results from studies using
prolonged exposure to average healthy humans. This factor is intended to account for the
variation in sensitivity among the members of the human population. [10H]
Use an additional 10-fold factor when extrapolating from valid results of long-term studies on
experimental animals when results of studies of human exposure are not available or are
inadequate. This factor is intended to account for the uncertainty in extrapolating animal data to
the case of humans. [1 OA]
Use an additional 10-fold factor when extrapolating from less than chronic results on experimental
animals when there are no useful long-term human data. This factor is intended to account for the
uncertainty in extrapolating from less than chronic NOAELsto chronic NOAELs. [10S]
Use an additional 10-fold factor when deriving a RfD from a LOAEL, instead of a NOAEL. This factor
is intended to account for the uncertainty in extrapolating from LOAELs to NOAELs. [10L]
Modifying Factor (MF)
Use professional judgment to determine another uncertainty factor (MF) which is greater than
zero and less than or equal to 10. The magnitude of the MF depends upon the professional
assessment of scientific uncertainties of the study and database not explicitly treated above; e.g.,
the completeness of the overall data base and the number of species tested. The default value for
the MFis 1.
SOURCE: Adapted from Dourson, M.L.; and Stara, J.F. (1983) Regulatory Toxicology and
Pharmacology 3:224-238.
(This is a consequence of the inability of either the traditional or the RfD approach to completely
address the question of dose-response extrapolation.)
The Agency is attempting to standardize its approach to determining RfDs. The RfD Work Group has
developed a systematic approach to summarizing its evaluations, conclusions, and reservations
regarding RfDs in a "cover sheet" of a few pages in length. The cover sheet includes a statement on
the confidence the evaluators have in the stability of the RfD: high, medium, or low. High
confidence indicates that the RfD is unlikely to change in the future because there is consistency
among the toxic responses observed in different sexes, species, study designs, or in dose-response
relationships, or the reasons for differences, if any, are well understood. Often, high confidence is
given to RfDs that are based on human data for the exposure route of concern, because in such cases
the problems of interspecies extrapolation are avoided. Low confidence indicates that the RfD may
be especially vulnerable to change if additional chronic toxicity data are published on the chemical,
because the data supporting the estimation of the RfD are of limited quality and/or quantity.
C Exposure Assessment
The third step in the risk assessment process focuses on exposure issues. For a full discussion of
exposure assessment, the reader is referred to EPA's recently published guidelines on the subject (51
Federal Register 34042-34054, Sept. 24, 1986). There is no substantive difference in the conceptual
approach to exposure assessment in the case of systemic toxicants and of carcinogens.
In brief, the exposure assessment includes consideration of the populations exposed and the
magnitude, frequency, duration and routes of exposure, as well as evaluation of the nature of the
exposed populations.
E-ll
-------
D. Risk Characterization
Risk characterization is the final step in the risk assessment process and the first step in the risk
management process. !ts purpose is to present to the risk manager a synopsis and synthesis of all the
data that contribute to a conclusion on the risk, including:
• The qualitative ("weight-of-evidence") conclusions about the likelihood that the chemical may
pose a hazard to human health.
• A discussion of dose-response and how this information, through the use of particular uncertainty
and modifying factors, was used to determine the RfD.
• Data such as the shapes and slopes of the dose-response curves for the various toxic end points,
toxicodynamics (absorption and metabolism), structure-activity correlations, and the nature and
severity of the observed effect. These data should be clearly discussed by the risk assessor, since
they may influence the final decision of the risk manager (see below).
• The estimates of exposure, the nature of the exposure, and the number and types of people
exposed, together with a discussion of the uncertainties involved.
• A discussion of the sources of uncertainty, major assumptions, areas of scientific judgment, and,
to the extent possible, estimates of the uncertainties embodied in the assessment.
In the risk characterization process, comparison is made between the RfD and the estimated
(calculated or measured) exposure dose (EED), which should consider exposure by all sources and
routes of exposure. The risk assessment should contain a discussion of the assumptions underlying
the estimation of the RfD (nature of the critical end point, nature of other toxic end points, degree
of confidence in the data base, etc.), and the degree of conservatism in its derivation. The
assumptions used to derive the EED should also be discussed. If the EED is less than the RfD, the need
for regulatory concern is likely to be small.
An alternative measure that may be useful to some risk managers is the "margin of exposure (MOE)"
(see footnote on p. A-3), which is the magnitude by which the NOAEL of the critical toxic effect
exceeds the estimated exposure dose (EED), where both are expressed in the same units:
MOE = NOAEL (experimental dose)/EED (human dose) (3)
In parallel to the statements above on EED and RfD, the risk assessment should contain a discussion
of the assumptions underlying the estimates of the RfD and the degree of possible conservatism of
the UF and MF. It can be noted that when the MOE is equal to or greater than UF x MF, the need for
regulatory concern is likely to be small.
Section VI contains an example of the use of the concepts of NOAEL, UF, MF, RfD, and MOE.
IV. APPLICATION IN RISK MANAGEMENT
Once the risk characterization is completed, the focus turns to risk management. In reaching
decisions, the risk manager must consider a number of risk factors, nonrisk factors, and regulatory
options that influence the final judgment. It is generally useful to the risk manager to have
information regarding the contribution to the RfD from various environmental media. Such
information can provide insights that are helpful in choosing among available control options.
However, in cases in which site-specific criteria are being considered, local exposures through various
media can often be determined more accurately than exposure estimates based upon generic
approaches. In such cases, the exposure assessor's role is particularly important. For instance, at a
given site, consumption of fish may clearly dominate the local exposure routes, while, on a national
basis, fish consumption may play a minor role compared to ingestion of treated crops.
RfDs should be apportioned by route of exposure. Where specific exposure analysis can be made,
such apportionment is readily performed. If exposure information is not available, assumptions must
be made concerning the relative contributions from different routes of exposure. At present,
different EPA offices use assumptions that differ to some degree. These assumptions are being
reviewed by an Agency risk assessment group.
E-12
-------
As illustrated in Figure A-1, the risk manager utilizes the results of risk characterization, other
technological factors, and nontechnical social and economic considerations in reaching a regulatory
decision. Some of these factors include efficiency, timeliness, equity, administrative simplicity,
consistency, public acceptability, technological feasibility, and legislative mandate.
Because of the way these risk management factors may impact different cases, consistent--but not
necessarily identical-risk management decisions must be made on a case-by-case basis. For example,
the Clean Water Act calls for decisions with "an ample margin of safety"; the Federal Insecticide,
Fungicide and Rodenticide Act (FIFRA) calls for "an ample margin of safety," taking benefits into
account; and the Safe Drinking Water Act (SDWA) calls for standards that protect the public "to the
extent feasible." Consequently, it is entirely possible and appropriate that a chemical with a specific
RfD may be regulated under different statutes and situations through the use of different
"regulatory doses (RgDs)".
Expressed in general terms, after carefully considering the various risk and nonrisk factors,
regulatory options, and statutory mandates in a given case (i), the risk manager decides upon the
appropriate statutory alternatives to arrive at an "ample" or "adequate" margin of exposure
[MOE(i)], thereby establishing the regulatory dose, RgD(i) (e.g., a tolerance under FIFRA or a
maximum contaminant level under SDWA), applicable to that case:
RgD(i) = NOAEL/MOE(i) (4)
Note that, for the same chemical (with a single RfD), the risk manager(s) can develop different
regulatory doses for different situations that may involve different exposures, available control
options, alternative chemicals, benefits, and statutory mandates. Also note that comparing the RfD
to a particular RgD(i) is equivalent to comparing the MOE(i) with the UF x MF:
RfD/RgD(i) = MOE(i)/UFxMF (5)
In assessing the significance of a case in which the RgD is greater (or less) than the RfD, the risk
manager should carefully consider the case-specific data laid out by the risk assessors, as discussed in
in Section III. D. 4. In some cases this may require additional explanation and insight from the risk
assessor. In any event, the risk manager has the responsibility to clearly articulate the reasoning
leading to the final RgD decision.
V. OTHER DIRECTIONS
While the Agency is in the process of systematizing the approach outlined in this Appendix, risk
assessment research for systemic toxicity is also being conducted along entirely separate lines. For
example, the Office of Air Quality Planning and Standards is using probabilistic risk assessment
procedures for criteria pollutants. This procedure characterizes the population at risk, and the
likelihood of various effects occurring, through the use of available scientific literature and
elicitation of expert judgment concerning dose-response relationships. The dose-response
information is combined with exposure analysis modeling to generate population risk estimates for
alternative standards. These procedures present the decisionmaker with ranges of risk estimates, and
explicitly consider the uncertainties associated with both the toxicity and exposure information. The
Office of Policy, Planning, and Evaluation is investigating similar procedures in order to balance
health risk and cost. In addition, scientists in the Office of Research and Development have initiated
a series of studies that should lead to future improvements in risk estimation. First, they are
investigating the use of extrapolation models as well as the statistical variability of the NOAEL and
underlying UFs as means of estimating RfDs. Second, they are exploring procedures for less-than-
lifetime health risk assessment. Finally, they are working on ranking the severity of toxic effects as a
way to further refine EPA's health risk assessments. While these procedures are promising, they
cannot be expected at this time to serve as a foundation of a generalized health risk assessment for
systemic toxicity in the Agency.
E-13
-------
VI. HYPOTHETICAL. SIMPLIFIED EXAMPLE OF DETERMINING AND USING RfD
Suppose the Agency had a sound 90-day subchronic gavage study in rats with the following data:
A. Experimental Results
Dose
(mg/kg-day)
0
1
25
Observation Effect Level
Control - no adverse effects observed -
No statistical or biological significant NOEL
differences between treated and control animals
2% decrease* in body weight gain (not NOAEL
considered to be of biological significance)
Increased ratio of liver weight to body weight
Histopathology indistinguishable from controls
Elevated liver enzyme levels
20% decrease* in body weight gain LOAEL
Increased* ratio of liver weight to body weight
Enlarged, fatty liver with vacuole formation
Increased* liver enzyme levels
* = Statistically significant compared to controls.
B. Analysis
1. Determination of the Reference Dose (RfD)
a. From the NOAEL
UF = 10H x 10A x 10S = 1000
MF = 0.8, a subjective adjustment based on the fact that the experiment involved an
astonishing 250 animals per dose group.
Therefore UF x MF = 800, so that
RfD = NOAEL/(UFx MF) = 5 mg/kg-day/ 800 = 0.006 mg/kg-day
b. From the LOAEL (i.e., if a NOAEL is not available)
if 25 mg/kg-day had been the lowest dose tested,
UF = 10Hx 10Ax 10Sx 10L = 10,000
MF = 0.8
Therefore UF x MF = 8,000, so that
RfD = LOAEL/(UF x MF) = 25 (mg/kg-day) / 8000 = .003 mg/kg-day)
2. Risk Characterization Considerations
Suppose the estimated exposure dose (EED) for humans exposed to the chemical under the
proposed use pattern were .01 mg/kg-day; i.e.,
EED > RfD
E-lU
-------
Viewed alternatively, the MOE is:
MOE = NOAEL/EED = 5 mg/kg-day/0.01 mg/kg-day = 500
Because the EED exceeds the RfD (and the MOE is less than the UF x MF), the risk manager will
need to look carefully at the data set, the assumptions for both the RfD and the exposure
estimates, and the comments of the risk assessors. In addition, the risk manager will need to weigh
the benefits associated with the case, and other nonrisk factors, in reaching a decision on the
regulatory dose (RgO).
E-15
-------
U.S. DEPARTMENT OF COMMERCE
Technology Administration
National Technical Information Service
Springfield, VA 22161 (703) 605-6000
------- |