EPA
United States
Environmental Protection
Agency
Office of Water
Washington D. C. 20460
Office of Research and
Development
Washington D. C. 20460
Technology Transfer
EPA/625/4-89/024
Seminar Publication
Risk Assessment,
Management and
Communication of
Drinking Water
Contamination
-------
-------
EPA/625/4-89/024
June 1990
Seminar Publication
Risk Assessment, Management and
Communication of Drinking Water
Contamination
Office of Drinking Water
Office of Water
U. S. Environmental Protection Agency
Washington, DC
Office of Technology Transfer and Regulatory Support
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
Notice
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
11
-------
Contents
Page
Acknowledgments v
1. Introduction 1
1.1 About This Publication 1
1.2 Purpose and Scope 1
1.3 Organization of This Publication 1
2. Development of Drinking Water Regulations and Health Advisories
by EPA's Office of Drinking Water 3
2.1 Background for Promulgation of Standards Under the Safe
Drinking Water Act and Amendments 3
2.2 Summary of the Regulations Specified Under the 1986 Amendments
to the Safe Drinking Water Act 7
2.3 Specific Phases of Regulatory Efforts by the Office of Drinking Water 9
2.4 Criteria Documents • • • 18
2.5 Health Advisory Program Administered by the Office of
Drinking Water 21
2.6 References 27
3. Principles of Risk Assessment 29
3.1 Introduction 29
3.2 Hazard Identification 32
3.3 Dose-Response Assessment 41
3.4 Human Exposure Assessment 45
3,5 Risk Characterization 45
3.6 References 46
4. Principles of Toxicology 47
4.1 Absorption, Distribution, Excretion, and Metabolism 47
4.2 Toxicology of Selected Substances 51
4.3 References 58
5. Risk Reduction 59
5.1 Overview of Risk Reduction and Control Strategies 59
5.2 Inorganics Treatment 63
5.3 Organics Treatment 71
5.4 References 82
111
-------
Contents (Continued)
Page
6. Risk Communication 85
6.1 What You Need to Know About the Media 85
6.2 Handling the Media 85
6.3 Conclusion 86
Appendices
A. Primary and Secondary Drinking Regulations A-l
B. Health Advisory Documents: Aldicarb, Atrazine, Triehloroethylene,
and Vinyl Chloride B-l
C. Sample Risk Assessment and Risk Management Exercises C-l
IV
-------
Acknowledgments
This document is a compilation of the material presented in a series of workshops on
"Assessment and Management of Drinking Water Contamination." These workshops
were sponsored by the United States Environmental Protection Agency's (EPA) Office
of Drinking Water (ODW), Office of Technology Transfer and Regulatory Support
(OTTRS) and Regional Offices. Many individuals contributed to the preparation and
review of this document. A partial listing appears below.
Development of Material
Chapter 2: Development of Drinking Regulations and Health Advisories by
EPA's Office of Drinking Water
Dr. Penny Fenner-Crisp, Director, Health Effects Division, Office of
Pesticide Programs, EPA, Washington, DC.
Dr. Edward V. Ohanian, Chief, Health Effects Branch, Criteria and
Standards Division, Office of Drinking Water, EPA, Washington, DC.
Chapter 3: Principles of Risk Assessment
Dr. C. Richard Cothern, Executive Secretary, Drinking Water
Committee, Science Advisory Board, EPA, Washington, DC.
Dr. Penny Fenner-Crisp, Director, Health Effects Division, Office of
Pesticide Programs, EPA, Washington, DC.
Chapter 4: Principles of Toxicology
Dr. Curtis Klaassen, Professor of Pharmacology and Toxicology and
Deputy Director, New Center for Environmental and Occupational
Health, University of Kansas Medical Center, Kansas City, Kansas.
Chapter 5: Risk Reduction
Stephen Clark, Chief of the Science and Technology Branch, Criteria
and Standards Division, Office of Drinking Water, EPA, Washington,
DC.
John E. Dyksen, P.E., Director of Water, Water Department, Village
of Ridge wood, New Jersey
-------
Acknowledgments (Continued)
Technical Reviewers
Dr. William R. Hartley Senior lexicologist, Health Effects Branch, Office of
Drinking Water, EPA, Washington, DC.
Ms. Ronnie Levin
Ms. Jennifer Orme
Mr. H. Douglas
Williams
Chief of the Water Team, Regulatory Support Staff, Office of
Technology Transfer and Regulatory Support, EPA,
Washington, DC.
Chief, Health Effects Assessment Section 2, Health Effects
Branch, Office of Drinking Water, EPA, Washington, D.C.
Physical Scientist, Center for Environmental Research
Information, EPA Cincinnati, Ohio.
Technical Direction/Coordination
Dr. Edward V. Ohanian Chief, Health Effects Branch, Criteria and Standards
Division, Office of Drinking Water, EPA, Washington, D.C.
Dr. James E. Smith, Jr. Senior Environmental Engineer, Center for Environmental
Research Information, EPA, Cincinnati, Ohio.
Writing, Editing, and General Preparation of Publication
Mr. Robert C. Goldberg Editor, JACA Corporation, Ft. Washington, PA.
Ms. Elizabeth Collins Senior Writer, Eastern Research Group, Arlington, MA.
Other Support
Ms. Marlene Regelski Project Manager, EGOS Management Criteria, Inc.,
Alexandria, Virginia.
VI
-------
Chapter 1
Introduction
1.1 About This Publication
Every week the news media bombard us with reports
of toxic wastes threatening our environment,
especially our drinking water supplies. In recent
years a whole new group of manmade drinking water
contaminants has emerged. Over 60,000 potentially
harmful chemicals are now being used by various
segments of U.S. industry and agriculture. These
substances range from industrial solvents and
pesticides to cleaning preparations and septic tank
degreasers. When used or discarded improperly,
these chemicals can pollute ground and surface
waters used as sources of drinking water. Subsurface
activities can also cause problems. Mining
operations, the injection of waste chemicals and
brines, and the storage of substances in underground
tanks have all been linked to the contamination of
ground and surface water. Not all problems of
drinking water quality originate with the surface or
ground-water supplies. Sometimes contamination
can occur during the treatment process itself. In other
cases it can occur while the water is in transit from
the treatment plant to your home.
This publication was largely developed from the
speakers' handouts for a series of EPA workshops
entitled "Assessment and Management of Drinking
Water Contamination." These workshops were
designed for those involved in the management of
drinking water contamination incidents — for
example, local and state officials, consultants, and
utility employees.
Each workshop attendee partici-pated in a hands-on
case-study exercise for specific chemicals of concern
(aldicarb, TCE, and vinyl chloride) for both a risk
assessment and risk manage-ment problem. Of the
approximately 125 people who attended each
workshop, 28 percent were from state agencies
(health and treatment technology), 28 percent were
from city and county agencies and utilities, 13
percent were from federal agencies, 18 percent were
consultants, and 13 percent were private or
industrial water supply personnel.
1.2 Purpose and Scope
This publication discusses how to identify, assess,
and manage the occurrence of potentially toxic
chemicals in drinking water. It presents a broad
range of information from the fields of toxicology,
chemistry, and engineering, and is intended to assist
the reader in assessing and managing drinking water
contamination problems in his/her region, state, or
locality.
Technical information is presented on U.S. EPA
programs, toxicology, chemistry, treatment
principles, and media coverage and risk
communication during an emergency. In addition,
exercises are included on particular risk assessment
and management problems that center around
specific EPA-ODW Health Advisory chemicals.
1.3 Organization of This Publication
This publication consists of six chapters and three
appendices. It progresses through the essential steps
in solving a drinking water contamination problem:
review of available standards and advisories,
assessment of risk, review of toxicology, review of
risk reduction options, and risk communication.
Chapter 2 discusses the development of standards
and health advisories, in addition to giving an update
on the regulatory process for the provision of safe
drinking water.
Chapter 3 goes through the various steps involved in
performing a risk assessment: hazard identification,
dose-response evaluation, human exposure
evaluation, and risk characterization. Chapter 4 then
describes the principles of toxicology, including
-------
absorption, distribution, excretion, and metabolism.
The toxicology of selected substances is also
described.
Chapter 5 reviews the various risk reduction
technologies available for removal of inorganic and
organic contaminants. Where possible, the processes
are described, equipment needed is presented, key
design factors are noted, typical performance data are
noted, and operation and maintenance data and costs
are given. Comparative technology tables appear in
this chapter.
Chapter 6 covers the management of media coverage
and communication with the public.
This publication contains three appendices. The first
provides the current (1989) drinking water
standards. The second contains examples of the latest
health advisories for aldicarb, atrazine,
trichloroethylene, and vinyl chloride. The third
appendix contains example exercises for risk
assessment and risk management/reduction.
-------
Chapter 2
Development of Drinking Water Regulations and
Health Advisories by EPA's Office of Drinking
Water
2.1 Background for Promulgation of
Standards Under the Safe Drinking
Water Act and Amendments
2.1.1 Safe Drinking Water Act
The Safe Drinking Water Act of 1974 (SDWA)
directed the U.S. Environmental Protection Agency
to identify and regulate substances in drinking water
that, in the judgment of the Administrator, may have
an adverse effect on public health (1). It included
interim regulations (National Interim Primary
Drinking Water Regulations - NIPDWRs) to be
established within 180 days of enactment of the
SDWA and National Primary Drinking Water
Regulations (NPDWRs) to be finalized over a period
of years, with the amendments promulgated in 1986.
The National Drinking Water Regulations were
subdivided into primary regulations, affecting public
health, and secondary regulations, affecting
aesthetic qualities relating to the public acceptance
of drinking water.
The most relevant criteria for this selection of
contaminants for regulation are the potential health
risks and the occurrence or potential occurrence in
the drinking water (2). For each of the substances or
contaminants that the EPA identifies, there are two
methods for developing regulatory measures. The
EPA must either establish a Maximum Contaminant
Level (MCL) or, if it is not economically or technical-
ly feasible to monitor the contaminant level in the
drinking water, specify a treatment technique to
remove the contaminant from the water supply or
reduce its concentration in the water supply.
The standards development process .involves an
intensive technological evaluation that includes
many factors: occurrence in the environment; human
exposure in specific and general populations; health
effects; analytical methods of detection; chemical
transformations of the contaminant in the drinking
water; and calculations of population risks of adverse
health effects, treatment technology, and costs.
The NPDWRs follow specific steps for promulgation
(see Figure 2-1). EPA first publishes a proposed
Drinking Water Priority List (DWPL) of contami-
nants for future regulation. Then the Agency accepts
public comment, publishes a final DWPL, publishes
proposed regulations and accepts public comment,
and firially publishes final regulations.
The SDWA Amendments mandate that EPA propose
the MCLs, which are enforceable standards, and
Maximum Contaminant Level Goals (MCLGs),
which are nonenforceable health goals,
simultaneously (3). For each contaminant, MCLG
development occurs first; and then the MCL is set as
close to the MCLG as is feasible, taking into
consideration analytical methods, treatment
technology, economic impact (costs), and regulatory
impact (see Figure 2-2).
2.1.2 Standards Development for Non-
carcinogens
For noncarcinogens, MCLGs are derived in the three-
step process described in Table 2-1. The first step is
calculating the Reference Dose (RfD; formerly called
Acceptable Daily Intake, or ADI) for each specific
contaminant.1 The RfD is an estimate of the amount of
a chemical ihat a person can be exposed to on a daily
basis that is not anticipated to cause adverse
systemic health effects over the person's lifetime..The
RfD is usually given in milligrams of chemical per
kilogram of body weight per day (mg/kg/day), has an
overall built-in uncertainty spanning perhaps an
-------
Figure 2-1. Regulatory development process
Federal Register
Notice:
Proposed
DWPL
»•
Public Comments
Public Meetings
Public Workshops
^
Federal Register
Notice:
Final DWPL
MCL/MCLG
Proposed
>
Public Comments
Public Meetings
Public Workshops
>
Federal Register
Notice:
MCL/MCLG
Promulgation
Figure 2-2. MCL/MCLG development
Risk Assessment Risk Assessment
Hazard Identification
Dose-Response
Assessment
Risk Assessment
Analytical Methods
The MCLG is determined by multiplying the DWEL
by the percentage of the total daily exposure contri-
buted by drinking water (relative source
contribution), as seen in Table 2-1. For non-
carcinogens, the MCLG will often equal the MCL.
(When determ-ining the MCL/MCLG, generally,
ODW rounds the final calculation to one significant
figure.).
Table 2-1. Determining the MCLG
Human Exposure
Assessment
Risk Characterization
MCLG
(Health-Based Goal)
Technology and Costs
Economic and Regu-
latory Impact
MCL
(The Legal Limit)
Determine RfD (Reference Dose) in mg/kg/day:
RfD = NOAEL or LOAEL in mg/kq/dav
Uncertainty Factor
Determine DWEL (Drinking Water Equivalent Level) in mg/L
assuming 100 percent drinking water relative source
contribution (RSC) and adult weighing 70 kg:
DWEL = (RfD) (70 kg )
(2 L/day)
Determine MCLG in mg/L:
MCLG = (DWEL) (Percent drinking water RSC)
Note: If actual exposure data are not available, ODW assumes an RSC of
20%. RSCs as high as 80% may be set for some contaminants when
exposure data show large contributions from drinking water.
order of magnitude, and takes into consideration
sensitive human subgroups.
The RfD is derived from a No- or Lowest-Observed-
Adverse-Effect Level (NOAEL or LOAEL), which is
calculated on the basis of data from a subchronic or
chronic scientific study of humans or animals. The
NOAEL or LOAEL is divided by an uncertainty,
factor to obtain the RfD. The uncertainty factor takes
into account intra- and interspecies diversity and
sensitivities, limited or incomplete data, significance
of the adverse effect, length of exposure, and pharma-
cokinetic factors. This uncertainty factor can range
from 1 to 10,000.
From the RfD, a Drinking Water Equivalent Level
(DWEL) is calculated by multiplying the RfD by an
assumed body weight of 70 kg for an adult and then
dividing by an average adult water consumption of 2
L per day. The DWEL assumes that 100 percent of
exposure to a substance will be from drinking water.
Preferred data for RfD and DWEL development (in
order of preference under each subheading) are
shown in Table 2-2.
Table 2-2. Preferred Data for RfD and DWEL Development
Duration of Exposure
- Chronic
- Subchronic
Route of Exposure
- Oral: drinking water,
gavage, diet
- Inhalation
- Subcutaneous or
intraperitoneal
Test Species
- Human
- Most sensitive species
- Animal model
Dose-Response Relationship
- NOAEL and LOAEL
- LOAEL or NOAEL
End-point of Toxicity
- Biochemical/pathophysio-
logical changes
- Body/organ weight changes
- Mortality
-------
2.1.3 Standards Development for Potential
Carcinogens
A separate health assessment system is used for
potentially "nonthreshold" no-effect-level chemicals
with carcinogenic potential. Carcinogenicity is gen-
erally assumed to be a nonthreshold phenomenon,
meaning that any exposure is assumed to represent
some finite level of risk in the absence of sufficient
negative information. Precedence was given to this
method in the House Report that accompanied the
SDWA of 1974, indicating that MCLGs for nonthres-
hold toxicants (i.e., carcinogens) should be set at zero
(4).
If toxicological evidence leads to the classification of
the contaminant as a human or probable human
carcinogen, the MCLG is set at zero. Mathematical
models are used to calculate drinking water
concentrations associated with estimated excess
cancer risk levels (e.g., 10-4, io-5,1Q-6). A lifetime
risk of 10-4, for example, indicates a possibility of one
additional case of cancer for every 10,000 people
exposed over a 70-year lifetime; a risk of 10-5 indi-
cates one additional cancer per 100,000 exposed
individuals.
The data used in these risk estimates usually come
from lifetime exposure studies in animals. To predict
the risk for humans, these animal doses must be con-
verted to equivalent human doses. This conversion
includes correction for noncontinuous animal expos-
ure, less than lifetime studies, and differences in
animal/human surface area and weight. The size
differential is assumed to be proportional to the
difference in body surface area, which is approxi-
mated by the cube root of the ratio of the animal and
human body weights. It is assumed that the average
adult human body weight is 70 kg.
For contaminants with carcinogenic potential,
drinking water concentrations are correlated with
the carcinogenic risk estimates by employing a
cancer potency (unit risk) value together with the
assumption for lifetime exposure via ingestion of 2 L
of water per day. The cancer unit risk is usually
derived from a linearized multistage model with a 95
percent upper confidence limit providing a low-dose
estimate. The true cancer risk to humans is not likely
to exceed this upper limit estimate and, in fact, may
be lower.
Excess cancer risk estimates may also be calculated
using other models such as the one-hit, Weibull,
logit, and probit models. (See Chapter 3, Section
3.3.2.2.) Given the current limited under-standing of
the biological mechanisms involved in
carcinogenesis, no one of these models can be said to
predict risk more accurately than another. Each
model is based on differing assumptions; thus, the
estimates derived for the various models can differ by
several orders of magnitude.
A number of uncertainties are associated with the
scientific database used to calculate arid support
cancer risk estimates. For example, cancer studies
are usually performed with experimental animals,
and extrapolating these data to humans is difficult
due to a lack of understanding of the biological
mechanisms involved. Insufficient knowledge
concerning the health effects of contaminants in
drinking water; the impact of the experimental
animal's age, sex and species; and the nature of the
target organ system(s) examined adds uncertainty to
the use of the database. Dose-response data are
gathered in animals at high levels of exposure rather
than at the lower levels typical of human exposure.
Finally, most exposures are to more than one
contaminant, and little is known regarding possible
synergistic or antagonistic effects between mixture
components. All of these uncertainties support use of
the generally more conservative linearized
multistage model for estimating cancer risk rates.
Using one model also fosters a consistency of
approach. However, some data suggest that other
models, such as the one-hit, may be more appropriate
than the linearized multistage model.
Several scientific groups have designed carcinogenic
chemical classification schemes on the basis of
weight of evidence for carcinogenicity, including the
EPA, National Academy of Sciences (NAS) Safe
Drinking Water Committee, and the International
Agency for Research in Cancer (IARC) (5). Table 2-3
describes EPA's classification groups and Table 2-4
shows EPA's three-category approach for setting
MCLGs (6).
Note that in Table 2-4 the MCLG for Category I
chemicals is set at zero as an aspirational goal. The
problem with setting the MCLG at zero is that a zero
level is unattainable and is below the analytical
detection level. MCLGs for Category II contaminants
can be calculated with the reference dose (RfD) ap-
proach and an additional safety/uncertainty factor of
1-10. If adequate systemic data are not available to
calculate an RfD, the MCLG may be represented by
10-5 to 10-6 excess cancer risk range. Category III
adheres to the RfD approach to accommodate for the
extrapolation of animal data to human risk, for the
existence of weak or insufficient data, and for
individual differences in human sensitivity to toxic
agents. The general guidelines used to calculate the
uncertainty factors are based on the NAS
recommendations, and are shown in Table 2-5 (7).
The guidelines help in determining risk, but
evaluating carcinogenic potential is controversial in
light of the divergent interpretations of the scientific
community.
-------
Table 2-3. EPA Carcinogenic Assessment Categories
A. Human carcinogen, based on sufficient evidence from epidemiological studies
B. Probable human carcinogen, based on at least limited evidence of carcinogenicity to humans (B1), or
usually a combination of sufficient evidence in animals and inadequate data in humans (B2)
C. Possible human carcinogen, based on limited or equivocal evidence of carcinogenicity in animals in the
absence of human data
D. Not classifiable, based on inadequate evidence of carcinogenicity from animal data
E. No evidence of carcinogenicity for humans (no evidence of carcinogenicity in at least two adequate
animal tests in different species or in both epidemiological and animal studies)
Table 2-4. EPA's Three-Category Approach
Category Evidence of Carcinogenicity
Class
MCLG Setting Approach
Sufficient evidence in humans or
animals
Limited or equivocal evidence in
animals
EPA Group A or B;
IARC 1, 2A, 2B
EPA Group C
Inadequate or negative evidence from EPA Group D or E;
animal data IARC 3, 4
1) RfD approach with
additional safety factor
2) 10-5 to 10'6 cancer risk
range
RfD approach
Table 2-5. Guidelines on the Use of Uncertainty Factors
Uncertainty Factor Guideline
1-10
100
1,000
1-10
When a NOAEL from a human study is used (to account for intraspecies diversity)
When a LOAEL from a human study is used, incorporating a factor of 10 to account for
lack of a NOAEL and a factor of 10 for intraspecies diversity; or, when a NOAEL from an
animal study is used, incorporating a factor of 10 to account for interspecies diversity and
a factor of 10 for intraspecies diversity
When a LOAEL from an animal study is used, incorporating factors of 10 each for lack of
NOAEL, interspecies diversity, and intraspecies diversity
Additional uncertainty factors, ranging from 1 to 10, may be incorporated on a case-by-
case basis to account for deficiencies in the database
2.1.4 Analytical Methods
In addition to criteria for health risk, EPA must
specify the analytical method best suited to detecting
the amount of a contaminant in drinking water. Set-
ting an MCL or MCLG for a chemical below the
smallest amount detectable is not feasible. Thus, for
each MCL or MCLG it derives, the Agency must
specify an analytical method to be used, such as
purge and trap gas, high performance liquid
chromatography, mass spectrometry, and
photoionization.
2.1.5 Feasibility and Best Available
Technology (BAT)
The SDWA Amendment directs EPA to set MCLs as
close to MCLGs as "feasible." "Feasible" means as
close as possible "with the use of best technology,
treatment techniques, or other means which the
Administrator finds are available (taking cost into
consideration) after examination for efficacy under
field conditions and not solely under laboratory
conditions" (8). Determining the feasibility of
controlling a contaminant requires an evaluation of
several factors:
-------
After each HA has been prepared and becomes
available for public use, it is entered into a computer-
based HA Registry, and executive summaries for
quick reference are entered into the Integrated Risk
Information System (IRIS), an electronic database for
information on risk assessment and risk
management throughout EPA. IRIS is available to
the public and EPA staff through the B.T. Tymnet
Electronic Mail Network and National Library of
Medicine's Toxnet.
Developing an HA is a step-by-step process
estimated to require approximately 18 months from
the time of identification of the chemical to issuance
of a final document. The development process
includes a minimum of four separate review steps by
individuals within and outside of EPA to ensure
quality and accuracy. After in-house review, a draft
HA is submitted to an external peer review panel of
recognized experts as well as to an Agency-wide
review before the final draft HA is prepared. This
final draft is then presented for public comment. A
final HA is not issued for public use until all phases
of the review process have been satisfied.
2.5.1 Legal Status of Health Advisories
Health Advisories are neither legally enforceable
standards nor are they issued as official regulations.
They may or may not lead to the issuance of MCLs.
The HAs do not condone the presence of
contaminants in drinking water; rather, they are
prepared to provide specific advice on the levels of
contaminants as they relate to possible health effects.
They describe, rather than prescribe, concentrations
of contaminants in drinking water at which adverse
noncarcinogenic effects are not anticipated to occur
following one-day, ten-day, longer-term, or lifetime
exposures. The HAs are subject to change and are
updated as new and better information becomes
available.
2.5.2 Content of Health Advisories
Health Advisories present, in a capsular or "bullet"
format, essential background information for
developing a concise, but complete, profile of a
chemical. The standard format for HAs is outlined in
Table 2-14.
Following a brief standard introduction explaining
the HA program, general information and properties
of the specific chemical are presented. To assist in
chemical identification, the physical and chemical
properties are given in table format along with the
various synonyms and uses. Occurrence and environ-
mental fate data are included as a means of
determining the extent of possible human exposure.
All available information on the pharmacokinetic
properties of thp chemical are included, with
particular emphasis on the chemical's absorptive
Table 2-14. ODW Health Advisory Content
I. General Introduction
II. General Information and Properties
--Synonyms -Uses -Properties
—Occurrence -Environmental Fate
III. Pharmacokinetics
-Absorption
-Metabolism
-Distribution
-Excretion
IV. Health Effects
Humans Animals
- Short-Term Exposure
- Dermal/Ocular Effects
- Long-Term Exposure
Developmental/Reproductive/Mutagenic/Carcinogenic
Effects
V. Quantification of Toxicological Effects
-One-Day Health Advisory —Ten-Day Health
Advisory
-Longer-Term Health Advisory -Lifetime Health Advisory
-Evaluation of Carcinogenic Potential
VI. Other Criteria, Guidance, and Standards
VII. Analytical Methods
VIII. Treatment Technologies
IX. References
properties and known metabolites. When data are
available on health effects in humans, all pertinent
details are reported, including dose and mode of
exposure and the effects resulting from acute and
chronic exposures.
Since the bulk of information on the toxic effects of
chemicals is usually derived from animal studies,
Section IV of an HA is often the most comprehensive.
If available, the animal data will include both short-
term and long-term exposure studies, reproductive
and developmental effects, and mutagenicity and
carcinogenicity data, as well as any available dermal
and ocular effects information. However, since an HA
is intended to be a brief guidance document, only
those studies deemed most pertinent to its presence
as a contaminant in drinking water are included.
Thus, the HA is not meant to review all available
data but rather only the best data available on the
specific chemical.
Section V, Quantification of Toxicological Effects,
presents the rationale used in selecting the studies
for development of the one-day, ten-day, longer-term,
and lifetime HA exposure values. These expo-sure
values are developed from data describing only
noncarcinogenic endpoints of toxicity. If a chemical is
a known or probable human carcinogen, the HA
document includes carcinogenic potency factors and
drinking water concentrations that are estimated to
represent excess lifetime cancer risks.
22
-------
consensus for risk assessment procedures for
estimating human risk levels via dermal exposure
(i.e., washing, bathing, etc.) from drinking water
contaminants. Dermal dose-response data for most
contaminants are lacking, especially for exposures
and toxic endpoints outside the occupational setting.
Whether effects from dermal exposures are
significant relative to other exposures is often
debatable, but the lack of data prevents resolution of
the issue for many contaminants.
ODW's extensive research needs transcend the needs
of other EPA offices (air, hazardous waste, toxic
substances, etc.), as well as of other regulatory and
health agencies at the federal, state, and local level.
While EPA has some limited research funding
capabilities, many of ODW's toxicity data needs must
be met by research supported by other private or
public sector sponsors. Where modest adjustments
can be made in toxicity study protocols to meet
multiple needs, innovative and cooperative funding
arrangements should be sought to support research.
Regardless of how individual research projects are
funded, ODW requests that consideration always be
given to no-cost or low-cost adjustments to testing
protocols that extend the utility of the results to
better meet ODW's needs.
Thus, the primary focus of future toxicological and
epidemiological research undertaken for EPA must
be to provide the essential quantitative dose-response
data needed to support policymaking and regulatory
decisions required under the SDWA Amendments.
Research efforts should be directed toward the
identification of the significant toxic endpoints and
the quantification of the dose-response relationships
needed to develop MCLGs.
2.5 Health Advisory Program
Administered by the Office of
Drinking Water
In addition to regulating drinking water contami-
nation, ODW provides nonregulatory guidance on
drinking water contamination through its Health
Advisory program. The Health Advisory (HA)
program provides documents on specific chemicals
being regulated or monitored for possible regulation.
The HA documents include information on health
effects, analytical methodologies, and treatment
technologies for assessing and managing contami-
nated drinking water. HAs specify nonregulatory
concentrations of contaminants in drinking water at
which adverse health effects are not anticipated to
occur over specific durations of exposure. These
numbers are derived from various human and animal
studies that are discussed in the HA documents. A
margin of safety is included in the HA values to
protect sensitive members of the observed population.
The HA values are not legally enforceable federal
standards but instead provide informal technical
guidance for federal, state, and local health officials.
The HA documents also discuss analytical methods
proven to be acceptable and treatment methods used
or under consideration.
The HA exposure values are developed from
appropriate short- or long-term data describing
noncarcinogenic endpoints of toxicity. For all
chemicals with sufficient data, one-day, ten-day, and
longer-term HAs are derived, with accompanying
explanations.
Health Advisories for lifetime exposures are not
calculated for human or probable human carcino-
gens. Rather, projected excess lifetime cancer risks
(unit risk) are provided to estimate the concen-
trations of the contaminant that may pose a
carcinogenic risk to humans. These hypothetical
estimates are usually presented as upper 95 percent
confidence limits derived from the EPA linearized
multistage model of risk extrapolation and are
considered unlikely to underestimate the actual
cancer risk. Excess cancer risk estimates may also be
calculated using the one-hit, Weibull, logit, and
probit models. Since these models are based on
different assumptions, the resulting risk estimates
may differ by several orders of magnitude. These
models are explained more fully in Chapter 3, Section
3.3.2.2.
When a draft Criteria Document (CD) (see Section
2.4) has been drafted by ODW, the HA is based on
this document. Individuals desiring further infor-
mation on the toxicological database or rationale for
risk characterization should consult the CD. Criteria
Documents and HAs are available for review at each
EPA regional office or Drinking Water counterpart
(e.g., Water Supply Branch or Drinking Water
Branch) or, for a fee, from the National Technical
Information Service, U.S. Department of Commerce,
5285 Port Royal Road, Springfield, VA 22161.
The HA program, as conducted by ODW, is an
ongoing, multifaceted program designed to provide
the most currently available information on potential
or known drinking water contaminants as soon as
such information is needed (i.e., in response to
contamination incidents). Earlier phases of the
program, which began in 1979, have resulted in the
preparation of HAs for approximately 50 drinking
water contaminants. Currently, HAs for approxi-
mately 60 pesticides have been finalized. HAs for 20
unregulated volatile organic chemicals (VOCs) and
30 inorganics, disinfectants, and disinfection by-
products are in the process of preparation. The
Department of the Army has also entered into a
cooperative program with the EPA to develop HAs for
various munitions chemicals that can contaminate
drinking water.
21
-------
6) Other Criteria, Guidance and Standards, 7)
Analytical Methods, 8) Treatment Technologies, and
9) References.
2.4.2 Quantification of lexicological Effects
The fifth chapter of the CD, covering quantification
of toxicological effects (QTE), integrates key health
effects information and provides the basis for the
proposed MCLG. Risk assessments described in this
section are designed to define the level at which no
known or anticipated adverse effects on human
health may occur while allowing for an adequate
margin of safety to protect more sensitive
individuals. The QTE for a chemical consists of
separate assessments of noncarcinogenic and
carcinogenic health effects. Chemicals that do not
produce carcino-genic effects are believed to have a
threshold dose below which no adverse
noncarcinogenic health effects occur. Carcinogens
are assumed to act without a threshold (i.e., there is
no exposure level that is assumed to be without some
level of health risk). For nonchemical drinking water
contaminants (e.g., microorganisms), the
organizational structure and risk assessment
procedures are modified, as required by the
contaminant's health effects properties. These case-
by-case exceptions to the QTE's structure, however,
are not addressed here.
2.4.3 Noncarcinogenic Effects
In the quantification of noncarcinogenic effects, a
Reference Dose (RfD) (formerly termed the
Acceptable Daily Intake - ADI), is calculated. The
RfD is an estimate (with an uncertainty spanning
perhaps an order of magnitude) of the daily exposure
to the human population (including sensitive
subgroups) that is likely to be without an appreciable
risk of deleterious health effects during a lifetime of
exposure. To calculate the RFD, see Section 2.1.
2.4.4 Carcinogenic Effects
Carcinogenic effects are quantified as described in
Section 2.1.3. Excess cancer risk estimates are
produced from lifetime animal exposure data,
epidemiological data, and mathematical models.
2.4.5 Development of MCLGs
Using the assembled information on noncarcinogenic
and carcinogenic effects, ODW employs a three-
category approach to develop the MCLG for each
contaminant. This approach is summarized in Table
2-4 and Section 2.1.3.
2.4.6 Research Needs For Developing
Regulation
ODW is conducting the most detailed and comp-
rehensive assessment of drinking water quality
specifications ever attempted. These efforts are
hampered by the lack of quantitative dose-response
information on noncarcinogenic and carcinogenic
effects for several of the contaminants to be
regulated. As a result, Drinking Water CDs are left
incomplete and attempts are thwarted to derive
meaningful drinking water standards. Of particular
concern are evalu-ations of the carcinogenic potential
of contaminants. The SDWA requires that a finding
of carcinogenicity (no safe threshold) for a particular
contaminant leads to a MCLG set at zero. Such
findings have widespread impact and must be
validated to withstand the scientific and legal
challenges encountered during the regulatory review
process.
ODW's first priority in research is dose-response
toxicity data from long-term, lifetime animal or
human studies. Even when such data are available
for a contaminant, however, duplicate data from the
same and other species are highly desirable.
Uncertainties associated with intra- and interspecies
diversity often force ODW to adopt conservative risk
assessment policies for deriving the MCLG values.
The more comprehensive the database for a chemical,
the greater the certainty in making risk
extrapolations to humans. While ODW strives to
estimate realistic, safe human exposure levels, the
office must retain the "margin of safety" principle
incorporated in the SDWA. This leads to criticisms
and controversy that could be reduced by more
comprehensive toxicity databases.
Major sources of uncertainty that are worthy of
emphasis here are concerns over the most sensitive
toxicity endpoint and most sensitive subpopula-
tion(s). A drinking water standard should provide
public health protection from all adverse effects. The
heavy reliance on toxicity data from the open
literature creates uncertainty as to whether the most
sensitive toxicity endpoints have been evaluated.
Toxicity research is usually focused within academic
disciplines (e.g., neurotoxicity, development effects,
renal effects), and does not evaluate other potentially
relevant toxicity endpoints. Filling these information
voids for drinking water contaminants, even with
negative data, would provide added confidence
during risk assessments. Similarly, comparative
toxicity or pharmacokinetic studies that identify
effects associated with different species, ages, gender,
etc., are valuable in selecting the most appropriate
human model and identifying the most sensitive
subpopulations that require protection.
ODW also strongly supports research on a variety of
scientific issues associated with performing risk
assessments, including: 1) improving the use of
pharmacokinetic information in evaluating dose-
response data; 2) improving procedures for
estimating human inhalation exposures due to
contaminants in drinking water (i.e., relevant to
showering, cooking, etc.); and 3) developing a
20
-------
Figure 2-3 Criteria Document development process.
Time (months)
Regulatory Process
CD Development Process
12
24
36
Public Comment Period
Public Meeting (s)
Public Workshop (s)
Public Comment Period
Public Meeting (s)
Technical Support
Document
Chemical Identification
Rough Draft CD
ODW Review
Rough External
Review Draft CD
ODW Review
Expert Review (as needed)
External Review
Draft CD
ODW Review
External Peer Review
Agency-wide Review
1
FR Notice
MCLG/MCL
Proposal
Technical Support
Document
1
Final Draft CD
Public Comments
ODW Review
Agency-wide Review
Final CD
2.4.1 Major Elements of a Criteria Document
By definition, the EPA is required to promulgate
NPDWRs for contaminants that are known or
anticipated to occur in drinking water and that may
cause an adverse human health effect. The primary
objectives of CDs are, therefore, to establish core
information based on health effects of chemicals in
drinking water and to compile and evaluate data for
providing the qualitative and quantitative health
effects basis for MCLGs. Each CD consists of nine
chapters : 1) Introduction, 2) General Information
and Properties, 3) Pharmacokinetics, 4) Health
Effects, 5) Quantification of Toxicological Effects,
19
-------
Ozone: Ozone is used extensively in water treatment
as a primary disinfectant. The use of this oxidant
will probably increase in the U.S. as the study of
chlorinated by-products continues. However, since
ozone does not leave a residual oxidant in the water
entering the distribution system, as chlorine does, its
use can pose a problem in maintaining water
quality. Regrowth of biological contaminants and
decreased effectiveness of disinfection may occur as
the water passes through the distribution system.
The mutagenic activity of ozone and its by-products
in water has been assessed. Ozone was not reported to
increase mutagenic activity in a number of bacterial
systems (29).
2.3.6 MCLs/MCLGs by January 1991: 25
Additional Chemicals
The sixth phase of regulation will begin the first
cycle of the Safe Drinking Water Act requirement to
regulate or reassess 25 additional substances every 3
years. The regulatory effort will also include
development of Drinking Water Priority Lists
(DWPLs) to be published every 3 years.
Substances for future regulatory consideration
include those chemicals listed pursuant to SDWA
Section 1428 (wellhead protection), other CERCLA
Section 101 substances, and substances not included
on the first Drinking Water Priority List because of
data limitations.
The first DWPL was proposed in July 1987 and
finalized January 22,1988. MCLs are required to be
set for at least 25 substances from this list within 36
months of finalization. This process is scheduled to be
repeated every three years. A number of organiz-
ations will be involved in identifying the best
candidate substances from the DWPL to meet this
schedule. These organizations include, within EPA
for example, the program offices for Superfund and
hazardous waste, ground water, water quality,
pesticides, and toxic substances. In addition, EPA
will consult with outside groups, such as the National
Toxicology Program.
Since the lists of additional chemicals will include
substances for which insufficient data are available,
EPA must consider how to fill the data gaps on
health effects, analytic methodology, occurrence, and
treatment technologies for those substances.
Researchers must look across the board at private
wells, additive substances, surface waters, waste
waters, environmental chemistry of substances, the
mobility of these substances in the environment, and
their mechanisms of entering drinking water (either
in the ground or on the surface). Also requiring study
are patterns of use of these compounds and their
production, properties of biodegradation and ,
absorption, and the amounts in which they are found.
These research needs could be used to set priorities
across EPA programs and throughout the Agency,
thus consolidating the decision-making processes of a
variety of programs.
2.4 Criteria Documents
Drinking Water Criteria Documents (CDs) are being
prepared for most contaminants to be regulated.
These provide the health effects basis for
establishing the MCLGs. Developing CDs requires
evaluation of pharmacokinetics, human exposure,
acute and chronic toxicity to animals and humans,
epidemiology, and mechanisms of toxicity. Emphasis
is placed on'data providing dose-response
information. Thus, while the literature search and
evaluation performed in support of each CD is
comprehensive, only the reports considered most
pertinent to the derivation of the MCLG are cited in
the CDs.
Figure 2-3 illustrates the process used to develop a
CD. During the first year of the process, the CD
rough draft and rough external review draft are
prepared. These drafts are presented to experts
within EPA for review and comment on the adequacy
of the database and risk assessments proposed as the
health basis for the MCLG. Before preparing the
final draft CD, EPA submits the external review
draft to a panel of recognized experts outside the
Agency for a thorough, critical review. Following this
revision, the final draft CD serves as the technical
support document covering health effects for the
proposed MCLG published in the Federal Register.
Inputs from the review and comment period on both
the final draft CD and the proposed MCLG are
considered in producing the final version of the CD.
The final CD serves as the technical support
document covering health'effects for the final rule
promulgating the MCLG.
During the regulatory process for a chemical, ODW
prepares other documents dealing with analytical
methods, treatment technologies, human exposure
potential and cost-benefit analyses for that
contaminant. These documents are used to derive the
proposed and final MCL, which is set as close to the
MCLG (i.e., health goal) as is feasible. Thus, by
providing the health basis for deriving the MCLG,
the CD serves as the keystone to the regulatory
process for drinking water contaminants. The MCL is
then derived from the MCLG by adjusting this health
goal level to a level that can feasibly be attained,
given the availability and cost of analytical methods
and treatment technologies. If at any time during the
CD and regulatory processes, new data show that
the proposed MCLG and/or MCL are inadequate to
protect human safety, these regulations may be
amended. Additionally, the relevance and adequacy
of the NPDWRs are to be reviewed at least every
three years and revised regulations promulgated
when appropriate.
18
-------
more expensive to use than chlorine, chlorine dioxide
is a good oxidizing agent and does not produce signifi-
cant amounts of chlorine by-products. It does,
however, produce milder oxidation products such as
aldehydes.
Chlorine dioxide has also proven to be an effective
disinfectant, with nearly 2.5 times the Oxidizing
power of chlorine. Chlorine dioxide degrades into
chlorite (6162') and, to a lesser extent, chlorate
(ClOa") during these processes.
The health effects of chlorine dioxide and its
conversion products are primarily hematological,
presumably due to its oxidizing nature.
The National Academy of Sciences (NAS) has calcu-
lated a Suggested-No-Adverse-Response Level
(SNARL) of 0.3 mg/L for chlorine dioxide and 0.02
mg/L for chlorite and chlorate, assuming a 20 percent
contribution from drinking water (22). However,
since these substances are almost uniquely found in
drinking water, the estimation of a 20 percent
relative source contribution is probably low.
Chloramine: Chloramine is an alternative to
chlorine for disinfecting drinking water. Chlora-
mines are less reactive than chlorine and, due to their
persistence, are best used as secondary residual
maintenance disinfectants rather than primary
pathogenic control agents. They do not treat
resistant organisms, such as viruses and Giardia, as
effectively as chlorine but are an inexpensive way of
quenching the formation of halomethanes and other
by-products. Chloramines also reduce unpleasant
tastes and odors connected with the formation of
chlorophenolic compounds. The primary toxic effect
of Chloramine in reported studies appears to be its
hematological effects. Persons on hemodialysis may
be at risk if chloramines are present in dialysate
water.
Further research is necessary to determine chronic
health effects. The NAS has estimated a SNARL of
0.5 mg/L for chloramines, assuming a 20 percent
relative drinking water source contribution (21).
2.3.5.3 Disinfection By-Products
Trihalomethanes (THMs): Trihalomethanes
regulated in drinking water include chloroform,
bromoform, bromodichloromethane, and dibromo-
chloromethane. These compounds are formed from
the reaction of chlorine with organic matter in the
water, such as humus, fulvic acids, and amides. Liver
and kidney effects due to THM exposure have been
observed in rats, mice, and dogs, as well as decreased
immune system functions in mice (23,24,25).
The most noted health effect reported to result from
exposure to THMs - and in particular chloroform - is
carcinogenicity. Chloroform has been found to be
carcinogenic in rats and mice. The National Cancer
Institute reported an increased incidence of kidney
tumors in male rats and liver tumors in male and
female mice when chloroform was administered by
gavage in a corn oil vehicle (26). Kidney tumors were
also reported in male rats exposed to chloroform in
drinking water (27) and male mice exposed to
chloroform in toothpaste (28). Liver tumors were not
reported to be significantly increased in the drinking
water or toothpaste studies. While chloroform has
been implicated in bladder, colon, and rectal cancers
in humans, the evidence is inconclusive.
The EPA currently has set an MCL of 0.10 mg/L for
total trihalomethanes. According to the NAS report,
this limit corresponds to an upper-bound increm-
ental lifetime cancer risk on the order of 1 in 100,000
(i.e., 10-5) (22). This MCL, based primarily on
treatment capabilities, was established as an interim
National Primary Drinking Water Regulation and is
under reevaluation.
Chlorinated Acids, Alcohols, Aldehydes, and
Ketones: The reaction of chlorine with organics in
water may yield various chlorinated acids, alcohols,
aldehydes, and ketones in addition to the THMs. EPA
is evaluating whether MCLGs should be developed
for:
Monochloroacetic acid
Dichloroacetic acid
Chloroacetaldehyde
Chloralhydrate
1,1-Dichloroacetone
1,3-Dichloroacetone
Dichloroacetaldehyde
Currently available toxicity information on the
health effects of these substances is limited.
Haloacetonitriles, Cbloropicrin, and Cyanogen
Chloride: Bromochloroacetonitrile (BCAN),
dibromoacetonitrile (DBAN), dichloroacetonitrile
(DC-AN), and trichloroacetonitrile (TCAN) are also
products of the reaction between chlorine and
organics in water.
Chlorophenols: Mono-, di-, and trichlorophenol
(CP, DCP, and TCP) are potential by-products of
chlorination formed when chlorine reacts with
phenolic materials. They pose common taste and odor
problems in addition to their possible toxic
properties.
2.3.5.4 Other Disinfectants
Other disinfectants or treatment practices have been
used in drinking water disinfection. These include
ozone, iodine, bromine, potassium permanganate,
silver, ferrate, high pH, ionizing radiation, and UV
light. The information available for these substances
and treatment practices is extremely limited.
17
-------
Table 2-12. Contaminants Scheduled for Regulation by 1990 Under the 1986 Amendments to the Safe Drinking Water Act
Methylene chloride Legionella
(Dichloromethane) '
Trichlorobenzene Standard plate count
Antimony
Beryllium
Cyanide
Nickel
Sulfate
Thallium
Adipates
Dalapon
Dinoseb
Diquat
Endothail
Endrin
Glyphosphate
Hexachlorocyclopentadiene
PAHs
Phthalates
Picloram
Simazine
2,3,7,8-TCDD (Dioxin)
1 ,1 ,2-Trichloroethane
Vydate (Oxamyl)
Radium-226
Radium-228
Radon
Uranium
Gross alpha particle radioactivity
Beta particle radioactivity
Photon radioactivity
compounds that occur in water. Each reacts individu-
ally and can exist in different forms depending on
dosages, pH, temperature, amount of organic
substances in the water, and oxidation reduction
processes that might have occurred.
More generally, disinfectants can be termed oxidants
because they oxidize the water and other substances
in it, e.g., nitrite. They also assist in floe formation
and removal of color from the water. The pH of the
water, which may be regulated to control corrosivity,
significantly affects the potency of some disinfect-
ants. All of these competing considerations are
involved in EPA's current analyses.
Proposed disinfection treatment requirements and
by-product regulations are scheduled for proposal in
1990 and promulgation in 1991. Much research
remains to be done before the database for these
comprehensive regulations is sufficient to formulate
MCLs and MCLGs. Table 2-13 shows the drinking
water disinfectants and disinfection by-products for
which EPA is considering the development of MCLGs
and MCLs (19).
Following is a brief summary of health effects and
issues of concern for each disinfectant and
disinfection by-product category being considered for
regulation.
2.3.5.2 Disinfectants
Chlorine: Chlorine has been the most widely used
disinfectant in the U.S. for more than 60 years (20).
Despite its long and widespread use, very little
information exists on the low-level health effects of
ingested chlorine; most laboratory studies have used
inhalation as the route of exposure to this chemical.
The acute toxicity of chlorine in amounts found in
drinking water appears to be relatively low.
Table 2-13. Disinfectants and Disinfection By-Products
Considered for Development of MCLGs and
MCLs
Disinfectants
Chlorine
Chlorine dioxide
Chloramine
Disinfection By-Products
Trihalomethanes:
Chloroform
Bromoform
Bromodichloromethane
Dibromochloromethane
Chlorinated acetic acids
Chlorinated alcohols
Chlorinated aldehydes
Chlorinated ketones
Chlorite and chlorate
Chlorophenols
Chloropicrin
Cyanogen chloride
Haloacetonitriles
Ozone by-products
n-Organochloramines
MX (3-chloro-4-dichloromethyl-5-hydroxy-2(5-H)-furanome)
Additional chronic data and resolution of the issues
concerning chlorine's carcinogenicity or cardio-
vascular toxicity are needed before an MCLG can be
determined.
Chlorine Dioxide, Chlorite, and Chlorate:
Chlorine dioxide (C1O2) has often been used in con-
junction with chlorine to control phenolic tastes and
odors (21). It was first used in the U.S. during World
War II when chlorine was in short supply. Although
16
-------
All surface water systems must filter unless they
meet source water quality criteria and site-specific
conditions.
Only qualified operators will be entitled to operate
the systems. All systems will need to achieve at
least 99.9 percent removal and/or inactivation of
Giardia cysts and 99.99 percent removal and/or
inactivation of enteric viruses.
Filtration is not required if a system meets:
• Source water quality criteria (coliform and
turbidity levels)
• The following site-specific conditions:
1. Achieves disinfection rate of 99.9 and 99.99
percent inactivation of Giardia and viruses
respectively
2. Maintains watershed control/satisfies on-site
inspection requirements
3. Has no history of waterborne disease
outbreaks that were not followed by treatment
corrections
4. Complies with the revised coliform MCL
(unless the state determines that the violation
was not caused by a treatment deficiency of the
source water)
5. Meets the total trihalomethanes (TTHM) MCL
(for systems over 10,000 people)
Finally, local water system operators must report to
their state governments monthly on their progress in
meeting federal rules and must report within 48
hours any waterborne disease outbreaks. Operators
of both filtered and unfiltered water systems must
meet federal requirements within 4 years of issuance
of the final rule.
2.3.3 MCLs/MCLGs by December 1988:
Radionuclides
The existing NIPDWRs include both natural and
manmade radionuclides. The standards for natural
radionuclides include a gross alpha particle standard
of 15 pCi/L and a combined radium-226 and radium-
228 standard of 5 pCi/L. Both radon and uranium
were excluded from the interim regulations because
of lack of data regarding their occurrence and
toxicity. The interim standard for manmade radio-
nuclides is a total dose equivalent to 4 millirems
(mrems) per year.
The MCLGs under development for radionuclides
apply to natural uranium, radon-222, gross alpha
particle activity (probably as a monitoring screen),
beta and photon emitters (manmade radionuclides),
and separate values for radium-226 and radium-228.
All these pollutants are estimated to pose carcino-
genic risks to humans. The House Report that accom-
panied the SDWA states that when there is no
threshold in the dose-response curve for a drinking
water contaminant (i.e., a carcinogen), the MCLGs
must be set at zero. This is because the MCLG must
be set at a level for which there are no known or
anticipated adverse health effects.
2.3.4 MCLs/MCLGs by June 1989: Other
Inorganic Chemicals, Synthetic Organic
Chemicals, and Pesticides
Contaminants slated to be regulated by June 1989
are shown in Table 2-12. Included are representatives
from all five categories of contaminants, including
the first NPDWRs for radionuclides. EPA may make
up to seven substitutions to this list if the Agency
determines that regulation of a different chemical is
likely to be more protective of public health.
2.3.5 MCLs/MCLGs by January 1991-.Disin-
fectants and Disinfection By-Products
The EPA is required to specify criteria for the disin-
fection of public water supplies. As EPA develops
regulations for disinfection and disinfection by-
products, it must consider the relationship between
the benefits of disinfectants and any adverse health
effects brought about by their use. More specifically,
since the SDWA requires that disinfection be speci-
fied as a treatment technique for all public water
systems, EPA must determine the conditions under
which disinfection must be used and the con-ditions
under which disinfectant residues do not adversely
affect public health.
2.3.5.1 Background
Public water systems use disinfection to control
pathogenic microorganisms and thus reduce the risk
of waterborne disease. The introduction of disinfect-
ants into the water supply, however, has resulted in
undesirable by-products with toxic properties that
have caused other health risks.
Trihalomethanes (THMs), one family of the
disinfection by-products, are currently regulated.
These compounds are formed in drinking water
during the reaction between chlorine, an effective
and widely used disinfectant, and organic matter
already in the water. In order to reduce formation of
THMs during water treatment, alternative disinfect-
ants are being used to replace free chlorine. These
alternatives, however, may produce other by-
products that can be toxic under some conditions.
Because disinfectants are chemically very reactive
substances, they react quickly with the many organic
15
-------
For inorganics, proposed BAT treatment techniques
may be found in Table 2-11.
Table 2-11. Proposed BAT for Inorganic Chemicals
Chemical BAT
Asbestos Coagulation/Filtration
Corrosion control
Barium Ion exchange
Lime softening
Reverse osmosis
Cadmium Ion exchange
Reverse osmosis
Coagulation/Filtration
Corrosion control
Lime softening
Chromium Coagulation/Filtration
Ion exchange
Lime softening (chromium III only)
Reverse osmosis
Mercury Granular activated carbon
Coagulation/Filtration and powdered
activated carbon*
Lime softening
Reverse osmosis
Nitrate/Nitrite Ion exchange
Reverse osmosis
Oxidation (Nitrite)
Selenium Activated alumina
Lime softening
Coagulation/Filtration (selenium IV only)
Reverse osmosis
*Mercury influent concentrations < 10 yJL.
The same BAT is specified for variances under
inorganics . Monitoring and reporting requirements
and compliance determination, analytical methods of
detection, lab certification criteria, monitoring for
unregulated contaminants, and regulatory impact
analysis are also stipulated.
2.3.2.2 Microbials and Surface Water
Treatment
Drinking water treatment in the U.S. is among the
best in the world. While treatment may be adequate
at the drinking water source, however, the condition
of the distribution system may permit regrowth of
microbial, bacterial, and viral contaminants. These
two treatment rules will standardize and upgrade
monitoring and treatment processes and disinfection
standards, thus eliminating thousands of cases of
waterborne disease. EPA's current standards, in
effect since 1977, protect for coliform bacteria and
turbidity.
In recent years, Legionnaire's disease and giardiasis
(also called backpacker's disease) have made the
public increasingly aware of waterborne disease
outbreaks. Local water suppliers throughout the U.S.
will now be directed by EPA to filter their water
and/or disinfect it under certain specified conditions
to protect against coliform bacteria, Giardia,
heterotrophic bacteria, Legionella, turbidity, and
viruses. These contaminants are described below.
Coliform bacteria come from human and animal
waste. While common in the environment and
generally not harmful themselves, bacteria indicate
that the water may be contaminated with disease-
causing organisms. Total coliform bacteria regu-
lations apply to all 200,000 public ground-water and
surface-water systems (both community and noncom-
munity supplies). The final rule of June 29,1989
(effective December 31,1990) bases compliance on
the presence or absence of total coliform in a sample
rather than on an estimate of coliform density, as per
the current regulations. For more information on
monitoring requirements for different systems, see
the published final rule in the Federal Register (17).
Giardia are protozoa that originate in human and
animal waste. Giardiasis, the disease they cause, has
flu-like symptoms that can be severe, causing diar-
rhea, nausea, and dehydration that can last for
months. Backpackers who drink from unfiltered,
nondisinfected mountain streams often contract
giardiasis. Because of their size, Giardia can be
filtered out of water or alternatively can be
inactivated by a rigorous disinfection process.
Heterotrophic bacteria are organisms that use only
organic materials as their food source. Turbidity is a
measure of the cloudiness of water, which is
indicative of excess organic material (including
animal or human waste). Therefore, testing for
heterotrophic bacteria and turbidity can point to the
presence of disease-causing microorganisms and can
provide information on the effectiveness of treatment
processes.
Legionella are bacteria that cause severe pneumonia-
like symptoms (i.e., Legionnaire's disease), especially
in a weaker population such as the elderly. Viruses
cause such diseases as hepatitis-A and gastro-
enteritis.
Many water supply systems do not filter or disinfect
their water. Of the 9,800 drinking water systems in
the U.S. using surface water, 3,000 systems serving
approximately 21 million people currently do not
filter. The final rule of June 29,1989 (effective
December 31,1990) sets criteria for the states to
determine which water systems will have to install
filtration or update existing filtration facilities
and/or disinfect (18):
All surface water systems will now have to
disinfect.
14
-------
Table 2-10. Proposed MCLs/MCLGs for Second Phase of Regulatory Efforts
Chemical Proposed MCL (mg/L) Proposed MCLG (mg/L)
1. Asbestos
2. Barium
3. Cadmium
4. Chromium
5. Mercury
6. Nitrate
7. Nitrite
8. Selenium
1. Acrylamide
2. Alachlor
3. Aldicarb
4. Aldtcarb sulfoxide
5. Aldicarb sulfone
6. Atrazine
7. Carbofuran
8. Chlordane
9. Dibromochloropropane
10. o-Oichlorobenzene
11. cis-1,2-Dichloroethylene
12. trans-1,2-Dichloroethylene
13. 1,2-Dichloropropane
14. 2,4-D
15. Epichlorohydrin
16. Ethyl benzene
17. Ethyiene dibromide
18. Heptachlor
19. Heptachlor epoxide
20. Lindane
21. Methoxychlor
22. Monochlorobenzene
23. PCBs
24. Pentachlorophenol
25. Styrene
26. Tetrachloroethylene
27. Toluene
28. Toxaphene
29. 2,4,5-TP
30. Xylenes (total)
Inorganic Chemicals
7 million fibers/L*
5
0.005
0.1
0.002
10(asNr
1 (as N)
0.05
Synthetic Organic Chemicals
Treatment technique
0.002
0.01
0.01
0.04
0.003
0.04
0.002
0.0002
0.6
0.07
0.1
0.005
0.07
Treatment technique
0.7
0.00005
0.0004
0.0002
0.0002
0.4
0.1
0.0005
0.2/0.00011
0.005/0.1"**
0.005
2.0
0.005
0.05
10.0
7 million fibers/L"
5
0.005
0.1
0.002
10(asN)**
1 (as N)
0.05
zero
zero
0.01
0.01
0.04
0.003
0.04
zero
zero
0.6
0.07
0.1
zero
0.07
zero
0.7
zero
zero
zero
0.0002
0.4
0.1
zero
0.2/0
zero/0.1*"
zero
2.0
zero
0.05
10.0
Fibers longer than 10 ym.
MCL and MCLG for total nitrate and nitrite = 10 mg/L (as N).
MCL of 0.1 mg/L and MCLG of 0.1 mg/L based on Group C classification and MCL of 0.005 mg/L and
MCLG of zero based on Group B2 classification.
Issue on cancer classification and quantitation.
13
-------
Table 2-9. Unregulated Contaminants Under SDWA Section 1445
List 1: Monitoring Required for All Systems
Bromobenzene
Bromodichloromethane
Bromoform
Bromomethane
Chlorobenzene
Chlorodibromomethane
Chloroethane
Chloroform
Chloromethane
o-Chlorotoluene
p-Chlorotoluene
Dibromomethane
m-Dichlorobenzene
o-Dichlorobenzene
trans-l ,2-Dichloroethylene
c/s-1,2-Dichloroethylene
Dichloromethane
1,1-Dichloroethane
1,1-Dichloropropene
1,2-Dichloropropane
1,3-Dichloropropane
1,3-Dichloropropene
2,2-Dichloropropane
Ethyl benzene
Styrene
1,1,2-Trichloroethane
1,1,1,2-Tetrachloroethane
1,1,2,2-Tetrachloroethane
Tetrachloroethylene
1,2,3-Trichloropropane
Toluene
p-Xylene
o-Xylene
m-Xylene
List 2: Monitoring Required for Vulnerable Systems
Ethylene dibromide (EDB) 1,2-Dibromo-3-chloropropane (DBCP)
List 3: Monitoring Required at the State's Discretion
Bromochloromethane
n-Butylbenzene
Dichlorodifluoromethane
Fluorotrichloromethane
Hexachlorobutadiene
Isopropylbenzene
p-lsopropyltoluene
/7-Propylbenzene
sec-Butylbenzene
terf-Butylbenzene
1,2,3-Trichlorobenzene
1,2,4-Trichlorobenzene
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene naphthalene
geological formations. Some are also consistently
found in drinking water supplies from manmade
sources; i.e., copper, lead, chromium, and asbestos
pipes and plumbing supplies. These metals either
leach into water sources naturally or as a result of
corrosion of the pipes and plumbing.
Lead, an inorganic metal of great concern when
found in drinking water, was originally handled
under the second phase of the regulations, but it has
now been addressed in a separate rule along with
copper. Lead contamination in water entering the
public distribution system is rare. Instead, lead
contamination is caused mostly by corrosion of lead
piping, solder, and flux in public water systems and
plumbing.
The proposed rule of August 18,1988 (for both lead
and copper) specified an MCL for lead of 0.005 mg/L
for water leaving the treatment plant (current NIP-
DWR = 0.050 mg/L) and an MCLG of zero. In
addition, another lead standard of 0.01 mg/L was
proposed for an average of a representative number of
samples from consumers' taps (16). In the proposed
rule, systems exceeding the at-the-tap limit will be
required to implement corrosion control and/or
corrosion inhibition. The proposed rule also includes
public notice requirements.
As of June 19,1986, the SDWA amendments
prohibited the use of lead piping, solder, and flux in
material in contact with potable water. The
amendments also required public water systems to
identify and provide notice to persons who may be
affected by lead contamination of their drinking
water.
Secondary MCLs (SMCLs) are aesthetic drinking
water standards based on color, odor, and taste.
SMCLs are being proposed under Phase II for
aluminum and silver.
The best available technology (BAT) for all synthetic
organic chemicals except acrylamide and
epichlorohydrin is granular activated carbon (GAC).
BAT for those two chemicals is polymer addition
practices (PAP). Packed tower aeration (PTA), in
addition to granular activated carbon, will be
specified for dibromochloropropane, 1,2-
dichloropropane, cis-l,2-dichloroethylene, trans-1,2-
dichloroethylene, o-dichloro-benzene, ethylene
dibromide, ethylbenzene, monochlorobenzene,
styrene, tetrachloroethylene, toluene, and xylene.
12
-------
persons (i.e., workplaces, offices, and schools) that
have their own water supplies and from which users
consume from one-third to one-half or more of their
normal daily water consumption.
The rule also specifies the monitoring of contami-
nants that are not regulated as NPDWRs, as required
by Section 1445 of SDWA. Each public water system
must monitor at least once every 5 years for
unregulated contaminants, unless EPA requires
more frequent testing. The monitoring data will
assist EPA in determining whether regulations for
these contaminants will be necessary, and if so, what
levels might be appropriate.
EPA has chosen 51 unregulated chemicals for moni-
toring (see Table 2-9) and separated them into three
groups (13):
• List 1: Chemicals for which monitoring is required
for all CWSs and NTNCWSs. Compounds can be
readily analyzed.
• List 2: Chemicals for which monitoring is required
only for systems vulnerable to contamination by
these compounds. Compounds have limited
localized occurrence potential and require some
specialized handling.
• List 3: Chemicals for which the state decides
whether a system must monitor. These are
compounds that do not elute within reasonable
retention time using packed column treatment
methods or are difficult to analyze because of high
volatility or instability. They are much less likely
to be present in drinking water.
The monitoring methods for the unregulated VOCs
are similar to those required for the regulated VOCs.
As a result, water systems will be encouraged to use
the same samples for all of the analyses, and to have
the analyses of the unregulated VOCs performed
with the analyses for the regulated VOCs, thereby
reducing costs for both sampling and analysis.
This list also outlines some of the disinfection by-
products that are scheduled to be promulgated as part
of Phase IV of the regulatory program. Other
disinfection by-products will be extracted from the
Drinking Water Priority List (DWPL).
Along with the VOC rule, two proposals were
announced: the list of changes on and off the original
list of 83 contaminants and a list of 25 additional
substances (14). The lists of both proposals were
added to the DWPL, the final version of which was
published January 22,1988.
2.3.2 MCLs/MCLGs by June 1988: Organics,
Inorganics, Microbials, and Filtration
The second phase of regulations is designed to
respond to the statutory requirements of the SDWA
and Amendments to set 40 MCLGs and MCLs (plus
the monitoring of 51 contaminants) by June 1988.
Also scheduled to be established by June 1988 were
microbial contaminants and filtration criteria, a
proposed rule for which was published November 3,
1987. The June 1988 statutory deadline has not been
met.
The list of chemicals proposed on November 13,1985
only included MCLGs. Since then the SDWA
Amendments have superseded this proposal,
stipulating that MCLs and MCLGs must be
promulgated simultaneously. Promulgation has been
slower for these chemicals because few data are
available on their occurrence in drinking water, or on
the treatment technologies required. However, there
is enough information, as specified by law, to
regulate these contaminants.
2.3.2.1 Organics and Inorganics
The second phase covers 30 synthetic organic
chemicals and 8 inorganic chemicals (see Table 2-10)
(15). These 38 chemicals represent a widely varied
group of contaminants, each causing a unique
problem. The synthetic organics may be found near
manufacturing; pesticides near agricultural
development; and the inorganics both in natural
geologic formations and in treatment and conveyance
mech-anisms for drinking water supplies and
sources.
The health effects produced by these chemicals are as
varied as their uses. Some are potent neurotoxins,
others are organ-specific toxicants, and some are
animal or human carcinogens. Thus, the approach to
setting MCLs and MCLGs for each chemical must be
very comprehensive.
Over half of the organics are pesticides, which have
been frequently detected in drinking water. Unlike
other synthetic organics used in manufacturing
products and as additives, pesticides are
manufactured to be toxic. They are applied directly to
the ground to kill pests or, in the case of herbicides
registered for aquatic applications, are applied
directly to water or migrate to drinking water sources
from runoff. Their widespread use and direct access
to water supplies make them of special concern for
drinking water contamination.
In general, inorganic chemicals are naturally
occurring contaminants prevalent in natural
11
-------
Table 2-8. VOCs: Final MCLGs and MCLs (in mg/L)
Contaminant
Vinyl chloride
Benzene
Trichloroethylene
Carbon tetrachloride
1 ,2-Dichloroethane
para-Dichlorobenzene
1 , 1 -Dichloroethylene
1,1,1 -Trichloroethane
Health Effect
Human carcinogen
Human carcinogen
Probable carcinogen
Probable carcinogen
Probable carcinogen
Possible carcinogen
Possible carcinogen
Liver, circulatory system,
and central nervous system
(CNS) damage
EPA
Class
A
A
B2
82
B2
C
C
D
Final
MCLG
zero
zero
zero
zero
zero
0.075
0.007
0.2
Final
MCL
0.002
0.005
0.005
0.005
0.005
0.075
0.007
0.2
(PTA) as best available technology (BAT) for
removing all VOCs , except vinyl chloride, for which
only PTA is designated BAT. These technologies
have 90-99 percent removal efficiency, are
commercially available, and have been used
successfully to remove VOCs in ground water from
both influents and ef-fluents in many locations
throughout the U.S.
Note that in Table 2-8, for all the chemicals with zero
MCLGs except vinyl chloride, the MCLs are set at
0.005 mg/L. This number represents the "feasible"
level taking cost into consideration. With an MCL of
0.005 mg/L, only 1,300 community water systems
(CWSs) need to install treatment capabilities to
satisfy the requirements, incurring a total capital
cost of $280 million.
The MCL for vinyl chloride - 0.002 mg/L - does
not result in any increased costs for public water
systems. Very few, if any, would have to install
treatment solely to control vinyl chloride. Systems
contaminated with any level of this chemical
virtually always contain one or more of the other
VOCs, since vinyl chloride is known to be a degra-
dation product of PCE or TCE.
PTA removes vinyl chloride to a 0.002 mg/L level.
Although this level may be harder to measure than
0.005 mg/L, EPA recognizes that vinyl chloride is a
human carcinogen of possibly higher potency than
the other VOCs listed on Table 2-8- Thus, the risk
posed by each unit of exposure could be higher than
the equivalent unit of any of the other VOCs with a
zero MCLG.
In addition to establishing MCLGs and MCLs for the
eight VOCs, this rule specifies the following condi-
tions for regulating those chemicals:
BAT for treatment for the purpose of variances to
be set when MCLs are set (as per SDWA Sections
1412 and 1415)
Monitoring requirements and compliance
determination ••..<•.
Public notification and reporting requirements
Laboratory certification criteria
Allowable point-of-entry (POE) and point-of-use
(POU) devices and bottled water uses to achieve
compliance
Variances and exemptions of control techniques
for VOCs
An additional definition was added for public water
systems for which directives of this rule apply. Public
water systems are divided into community and
noncommunity systems. A community water system
(CWS) is one that serves at least 15 connections used
by year-round residents or regularly serves at least
25 year-round residents. Noncommunity water
systems (NCWSs) are, by definition, all other water
systems and include transient systems (i.e.,
campgrounds, gas stations) and nontransient
systems (i.e., schools, workplaces, hospitals) that
have their own water supplies and serve the same
population over 6 months of a year.
EPA has promulgated a definition of a "nontransient
noncommunity water system" (NTNCWS) and
applied it to the NPDWRs for the eight VOCs in
addition to the already defined systems. A
noncommunity nontransient water system is "a
public water system that is not a regular community
water system and that supplies at least 25 of the
same people over 6 months per year" (12).
The purpose of the change was to protect
nonresidential populations of more than 25 people
who, because of regular long-term water usage, incur
risks of adverse health effects similar to those
incurred by residential populations. The change was
designed to include systems serving more than 25
10
-------
• The contaminant must have a documented or
suspected adverse human health effect.
• There must be sufficient information available on
the contaminant so that a regulation could be
developed within the statutory time frames.
Substances for which insufficient information for
regulation is available will be candidates for
subsequent priority lists.
Further information on the specific selection criteria
may be found in the Federal Register, 52 FR 25720
(10).
The seven contaminants substituted onto the original
list of 83 contaminants are aldicarb sulfoxide, aldi-
carb sulfone, heptachlor, heptachlor epoxide, styrene,
ethyl benzene, and nitrite.
The contaminants removed from the original list of
83 contaminants, as listed below, will now be placed
on the DWPL (see also Table 2-7):
• Zinc, sodium, vanadium, silver, molybdenum,
dibromomethane, aluminum
Monitoring requirements are to be set to ensure
compliance with the MCLs. In all but three cases,
states have the responsibility for enforcement of
MCLs. Public water systems must give public
notification of a violation of an MCL or monitoring
requirement.
Other priorities set by the SDWA, such as compli-
ance requirements, surface water treatment and
disinfection criteria, variances and exemptions, and
regulatory timetables and deadlines, are discussed
briefly in each of the regulatory phase sections below.
2.3 Specific Phases of Regulatory
Efforts by the Office of Drinking
Water
The Office of Drinking Water has outlined a six-
phase plan (see top of next column)
2.3.1 Volatile Organicc Chemicals -
Promulgated July 8, 1987
On July 8,1987, the final rule was published for
NPDWRs for eight volatile organic chemicals.
Monitoring for unregulated contaminants (11) was
also covered. The VOCs listed in this rule, plus
fluoride (promulgated April 6,1986), satisfied the
statutory deadlines of SDWA, which required the
establishment of the first 9 MCLs within 12 months.
MCLs and MCLGs for the eight VOCs are shown in
Table 2-8.
Phase
Substances
Expected
Promulgation
Date
III
IV
V
VI
Volatile organic chemicals
(VOCs)
Synthetic organic chemicals
(SOCs), Inorganic chemicals
(lOCs), Microbial and surface
water treatment (Filtration)
Lead/Copper (corrosion by-
products)
Radionuclides (proposal)
Disinfectants and disinfection
by-products
Other lOCs, SOCs, and
pesticides
Additional DWPL chemicals
JulyS, 1987
December 1990
December 1988
February 1991
January 1992
June 1991
January 1992
These six regulatory phases parallel the SDWA-
specified deadlines, listed below:
9 MCLGs and MCLs + monitoring
Public notice revisions
Filtration criteria
Monitoring for unregulated
contaminants
Final list of contaminants on DWPL
40 MCLGs and MCLs + monitoring
34 MCLGs and MCLs + monitoring
Disinfection treatment
25 MCLGs and MCLs + monitoring
June 19, 1987
September 19, 1987
December 19, 1987
December 19, 1987
January 1, 1988
June 19, 1988
June 19, 1989
June 19, 1989
January 1, 1991
The eight synthetic VOCs shown in Table 2-8 are
widely used in products such as unleaded gas
additives; household cleaning solutions; solvents for
removing grease from clothes, electronics, and
aircraft engines; air freshen-ers; and mothballs. They
are found frequently in drinking water from ground-
water sources. All have relatively low boiling points
and vaporize readily.
EPA proposed the MCLs for these chemicals based on
an evaluation of 1) the availability and performance
of treatment technologies for the VOCs; 2) the
availability, performance, and cost of analytical
methods; and 3) the costs of applying the various
technologies to reduce VOCs to various
concentrations.
In reviewing the different technologies available for
VOC removal, EPA considered the following criteria:
removal efficiency, degree of compatibility with other
treatment processes, service life, and ability to
achieve compliance.
Based on these criteria, EPA proposed granular
activated carbon (GAC) and packed tower aeration
-------
Table 2-6. Contaminants Required to Be Regulated Under
the Safe Drinking Water Act Amendments of
1986
Volatile Organic Chemicals
Trichloroethylene* Benzene*
Tetrachloroethylene* Monochlorobenzene
Carbon tetrachloride* Dichlorobenzene**
1,1,1 -Trichloroethane* Trichlorobenzene
Dichloroethane* 1,1 -Dichloroethylene*
Vinyl chloride* trans-1,2-Dichloroethylene
Methylene chloride cis-1,2-Dichloroethylene
Microbiology and Turbidity
Total conforms
Turbidity
Giardia lamblia
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Nitrate
Selenium
Silver
Fluoride***
Aluminum
Antimony
Endrin
Lindane
Methoxychlor
Toxaphene
2,4-D
2,4,5-TP
Aldicarb
Chlordane
Dalapon
Diquat
Endothall
Glyphosphate
Carbofuran
Alachlor
Epichlorohydrin
Toluene
Adipates
Viruses
Standard plate count
Legionella
Inorganics
Molybdenum
Asbestos
Sulfate
Copper
Vanadium
Sodium
Nickel
Zinc
Thallium
Beryllium
Cyanide
Organics
1,1,2-Trichloroethane
Vydate (Oxamyl)
Simazine
Polyaromatic hydrocarbons
Polychlorinated biphenyls
Atrazine
Phthalates
Acrylamide
Dibromochloropropane
1,2-Dichloropropane
Pentachlorophenol
Pichloram
Dinoseb
Ethylene dibromide
Dibromomethane
Xylene
Hexachlorocyclopentadiene
2,3,7,8-TCDD (Dioxin)
Radionuclides
Radium-226 and -228
Beta particle and photon
radioactivity
Gross alpha particle activity
Uranium
Radon
Promulgated July 8, 1987
MCL for p-dichlorobenzene has been published;ortho-
dichlorobenzene is on additional list for consideration.
Promulgated April 2, 1986
MCLs and MCLGs are to be proposed and pro-
mulgated simultaneously, thus shortening the
standards-setting procedure.
The EPA under the SDWA must maintain a
Drinking Water Priority List (DWPL) of
contaminants for future regulation (see Table 2-7).
The proposed list was published July 8,1987; the
final list was published January 22,1988 (9).
MCLs and MCLGs must be set for at least 25
contaminants on the DWPL by January 1,1991;
every three years thereafter, 25 more MCLs and
MCLGs must be set.
Table 2-7. Drinking Water Priority List (53 FR 1901,
Jan. 22, 1988)
1,1,1,2-Tetrachloroethane
1,1,2,2-Tetrachloroethane
1,1 -Dichloroethane
1,1-Dichloropropene
1,2,3-Trichloropropene
1,3-Dichloropropane
1,3-Dichloropropene
2,2-Dichloropropane
2,4,5-T
2,4-Dinitrotoluene
Aluminum
Ammonia
Boron
Bromobenzene
Bromochloroacetonitrile
Bromodichloromethane
Bromoform
Bromomethane
Chloramines
Chlorate
Chlorine
Chlorine dioxide
Chlorite
Chloroethane
Chloroform
Chloromethane
Chloropicrin
Cryptosporidium
Cyanazine
Cyanogen chloride
Dibromoacetonitrile
Dibromochloromethane
Dibromomethane
Dicamba
Dichloroacetonitrile
ETU
Hypochlorite ion
Isophorone
Methyl tert-butyl ether
Metolachlor
Metribuzin
Molybdenum
Ozone by-products
Silver
Sodium
Strontium
Trichloroacetonitrile
Trifluralin
Vanadium
Zinc
o-Chlorotoluene
p-Chlorotoluene
Halogenated acids
alcohols, aldehydes,
ketones, and other
nitriles
The chemicals listed in Table 2-6 include the seven
contaminants taken off the original list of 83;
disinfectants and disinfection by-products; the first
50 contaminants specified under Sect-ion 110 of the
Superfund Amendments and Re-authorization Act of
1986 (SARA); pesticides included as design-analytes
in the National Pesticide Survey (NPS, see Section
2.5.6.2); volatile organic chemicals (VOCs) reported
in Section 1445 of SDWA as unregulated contam-
inants to be monitored; and certain other substances
reported frequent-ly and/or occurring at high
concentrations in other recent surveys. Criteria for
place-ment on the Drinking Water Contaminant
Priority List are outlined below:
• The contaminant must occur in public water
systems, or its characteristics or use patterns must
be such that it has a strong potential to occur in
public water systems at levels of concern.
-------
• Technical and economic availability of analytical
methods that would be acceptable for accurate
determinations of compliance (i.e., practical
quantitation levels-see below), limits of analytical
detection, laboratory capabilities, and costs of
analytical techniques
• Concentrations attainable by application of best
technology generally available; levels of chemical
concentrations in drinking water supplies;
feasibility/reliability of removing contaminants to
specific concentrations
• Other factors relating to the "best" means of
treatment such as air pollution and waste disposal
from the treatment method itself, and possible
effects on other drinking water quality parameters
• Costs of treatment to achieve contaminant
removal (8).
One of the factors used in setting the laboratory
performance requirements for an MCL is the
minimum detection limit (MDL). The MDL is the
minimum concentration of a substance that can be
measured and reported with 99 percent confidence
that the true value is greater than zero. These MDLs
are measured by a few of the most experienced labs
under nonroutine and controlled conditions.
A second measurement used by EPA, the practical
quantitation level (PQL), is not lab- or time-specific
and can provide a uniform concentration
measurement for setting standards. The PQL is the
lowest measurement level that can be reliably
achieved within specified limits of precision and
accuracy during routine laboratory operating
conditions. PQLs are based on four factors: 1)
quantitation, 2) precision and accuracy, 3) expected
normal laboratory operations, and 4) the
fundamental need of the compliance and monitoring
program to have a sufficient number of labs available
to conduct analyses.
Evaluating treatment technologies is part of
regulating chemicals or groups of chemicals. Both
available treatment technologies and analytical
methods are key to analyzing the regulatory and
economic consequences considered for each
contaminant. How to monitor, measure, and treat for
a specific contaminant (or mixture of contaminants)
is published as an integral part of any standard that
is promulgated. Regulations for some chemicals must
specify the best available technology (BAT) for
treatment procedures.
2.2 Summary of the Regulations
Specified Under the 1986
Amendments to the Safe Drinking
Water Act
The Safe Drinking Water Act was amended in 1986
to require EPA to regulate 83 drinking water
contaminants by 1989. An overview of these
amendments and the Office of Drinking Water's
regulatory program follows.
Recommended Maximum Contaminant Level
(RMCL), now termed Maximum Contaminant
Level Goal (MCLG).
EPA is to set MCLGs, nonenforceable health goals,
and NPDWRs, which consist of MCLs and
treatment techniques, for 83 specific contaminants
and for any other contaminant in drink-ing water
that may have an adverse effect on human health
and that is known or anticipated to occur in public
water systems (see Table 2-6).
The Act requires EPA to regulate drinking water
contaminants according to the following schedule :
• 9 MCLs in 12 months: June 19,1987
• 40 MCLs in 24 months: June 19,1988
• 34 MCLs in 36 months: June 19,1989
EPA is allowed to substitute up to seven
contaminants for ones on the above list if they are
found to be more harmful to public health.
MCLs are to be set as close to MCLGs as is feasible.
The term "generally available" technology was
changed to "as is feasible." As discussed earlier,
feasible is defined as "with the use of the best
technology, treatment techni-ques, or other means
which the Administrator finds are available
(taking cost into consider-ation) after examination
for efficacy under field conditions and not solely
under laboratory conditions."
EPA is required to prepare a Report to Congress
comparing the benefits and risks of treatment
versus no treatment (final report submitted in
November 1988).
Granular activated carbon (GAG) is stated in the
SDWA as feasible for the control of synthetic
organic chemicals (SOCs), and any techno-logy or
other means found to be "best available" for
control of SOCs must be at least as effective in
controlling SOCs as GAC.
-------
Table 3-4. Summary of Toxicity Tests (4)
Acute (Oral
LD50)
• Gavage
* Mouso and
rat most
often used
(sometimes
also rabbit
and dog)
• Often starve
animals for
16 hours
before
exposure
• Usually
administer
constant
concentration
for various
doses rather
than
constant
volume
• Typical
observations:
- Observe
animals at
1,2, 4
hours and
daily for 14
days
- Record
body
weight at
14 days
.
- Minimal or
no histo-
pathology
or clinical
chemistry
(except in
the dog)
Acute
Acute Inhalation
Dermal (LC50)
• Albino • Similar to
rabbits acute oral
used LD50
• Area of • Typical 4-
application hour
is free of exposure
hair and
abraded
• If sub-
stance is
solid, is
moistened
with saline
• Kept in
contact
with skin
for 24
hours
• Observe
for 2 weeks
Primary
Skin
Irritation
• Rabbits
(Draize
test) used
•Hair
clipped
• 0.5 mL
liquid or
0.5 g
solid
• Covered
by gauze
and then
plastic
• t^Ar\+ iri
r\ept in
contact
with skin
for 4
hours
• Swelling
and
redness
scored at
24 and
72 hours
after
appli-
cation
Primary
Eye
Irritation
• Rabbits
used
• Place
liquid or
unmoisten
-ed solid
in one
eye (0.1
mL liquid
or 100
mg solid)
• Other eye
serves as
control
•Eye
irritation
graded
and
scored at
1, 2, 3, 4,
and 7
days and
every 3
days
thereafter
until
toxicity
subsides
Skin
Sensitization
(Allergies)
• Guinea pigs
used
• Tests used
include:
.
- Draize
- Buehler
occluded
patch
- Magnus-
son and
Kligman
maximi-
zation
Subacute
• To determine
dose levels
for
subchronic
study
• Typical
protocol:
- 14-day
duration
- In rodents,
4 doses; 10
animals per
sex per
dose; in
dogs, 3
doses, 3
dogs per
sex per
dose
- Observe
twice a day
- Perform
clinical
chemistry,
histopatho-
logy, etc.
Subchronic
• Typical
protocol:
- 90 days (13
weeks)
dura-tion
- At least 3
doses and
controls
- 2 species
(15 rats of
each sex
per dose
and 4 dogs
of each sex
per dose)
- Route of
intended use
or exposure
(usually diet)
• Typical
observations:
- Mortality
- Body weight
changes
- Urinalysis
- Hematology
- Clinical
chemistry
- Gross and
microscopic
examination
of several
parts of the
body
Chronic
• Similar to
subchronic
but longer
duration
• Duration
depends on
intended
period of
exposure in
man. Can be
as little as 6
months or as
long as
lifetime of
animal (i.e., 2
years in rats).
Start with 60
rats per sex
per dose to
ensure that
30 survive.
• For dogs,
often use 3
doses and 6
males and 6
females per
dose. Typical
duration is 1 2
months;
clinical
chemistry
performed
before
exposure and
at 1, 3, 6, 9,
and 12
months.
• Typical
observations:
QQQ
- oc@
subchronic
- In dogs,
often
perform
ophthalmic
examination
every 6
months
(continued)
MTD and one-half of the MTD are the usual doses
used in a NCI carcinogenicity bioassay.
The main reason cited for using the MTD as the high-
est dose in a bioassay is that experimental studies are
conducted on a small scale, making them statistically
insensitive, and that very high doses overcome this
problem. Due to cost considerations, experiments are
carried out with relatively small groups of animals.
Typically, 50 or 60 animals of each species and sex
will be used at each dose level, including the control
group. At the end of such an experiment, the exam-
ining pathologists tabulate the incidence of cancer as
a function of dose (including control animal incid-
ence). Statisticians then analyze the data to deter-
mine whether any observed differences in tumor
incidence (fraction of animals hav-ing a tumor of a
certain type) are due to random variations in tumor
incidence or to exposure to the substance.
36
-------
Selection of Animal Species: Rats and mice are the
most commonly used laboratory animals for toxicity
testing. They are inexpensive and can be handled
relatively easily, and, such factors as genetic back-
ground and disease susceptibility are well
established for these species. The full life spans of
these smaller rodents are complete in 2 to 3 years, so
the effects of lifetime exposure to a substance can be
measured relatively quickly.
Other rodents such as hamsters and guinea pigs are
also used, as well as rabbits, dogs, and primates such
as monkeys or baboons. Reproductive studies often
use primates because their reproductive systems are
similar to that of humans. Rabbits are often used for
testing dermal toxicity because their shaved skin is
more sensitive than that of other animals.
Dose and Duration: Determining the LDso is
frequently the first toxicity experiment performed.
After completing this effort, investigators study the
effects of lower doses administered over longer
periods to find the range of doses over which adverse
effects occur and to identify the NOAEL for these
effects (although the NOAEL is not always sought or
achieved). A toxicity experiment is of limited value
unless a dose of sufficient magnitude to cause some
type of adverse effect within the duration of the
experiment is achieved. If no effects are seen at any
dose administered, the toxic properties of the
substance cannot be characterized and the
experiment will usually be repeated at higher doses
or over a longer timespan.
Some substances with extremely low toxicity must be
administered at extremely high levels to produce
effects; in many cases, such high levels will cause
dietary maladjustments leading to an adverse
nutritional effect that confounds interpretation. The
highest level of a compound fed to an animal in
toxicity studies is 5 percent of the diet, even if no
toxic effect is seen at this level.
Studies are frequently characterized according to the
duration of exposure. Acute toxicity studies involve a
single dose, or exposures of very short duration (i.e., 8
hours of inhalation). Chronic studies involve
exposures for nearly the full lifetime of the
experimental animals. Experiments of varying
duration between these extremes are referred to as
subchronic studies.
Number of Dose Levels: Although many different
dose levels are needed to develop a well-characterized
dose-response relationship, practical considerations
usually limit the number to two or three, especially
in chronic studies. Experiments involving a single
dose are frequently reported, and these leave great
uncertainty about the full range of doses over which
effects are expected.
Controls: All toxicity experiments require control
animals that are not exposed to the substances in
question. Control animals must be of the same
species, strain, sex, age, and state of health as the
treated animals, and must be held under identical
conditions throughout the experiment. Indeed,
allocation of animals to control and treatment groups
should be performed on a completely random basis.
Other controls are historical; i.e., data on what has
happened in the past with that species and strain of
experimental animal.
Route of Exposure: Animals are usually exposed by
a route that is as close as possible to the route by
which humans will be exposed. In some cases,
however, the investigator may have to use other
routes or conditions of dosing to achieve the desired
experimental dose. For example, some substances are
administered by stomach tube (gavage) because they
are too volatile or unpalatable to be placed in the
animals' diets at the high levels needed for toxicity
studies.
Summary of Toxicity Studies: Table 3-4
summarizes the major types of toxicity tests
currently used. It lists key characteristics of acute
tests, chronic tests, and various reproductive system
tests. Table 3-5 shows typical costs of some of these
tests, which can be quite high. Also noteworthy is the
completeness of the various tests often performed on
laboratory animals after their exposure to a
chemical. These urinalysis, hematology, clinical
chemistry, and histopathological tests examine many
more parameters than even thorough human
autopsies (see Table 3-6).
3.2.7.4 Designing Tests for Carcinogenicity
One of the most complex and important of the special-
ized tests is the carcinogenesis bioassay. This type of
experiment is used to test the hypothesis of
carcinogenicity—that is, the capacity of a substance
to produce tumors.
In a National Cancer Institute (NCI) carcinogenicity
bioassay, the test substance is administered over
most of the adult life of the animal, and the animal is
observed for formation of tumors. The general
principles of test design previously discussed apply to
this testing, but one critical and controversial design
issue requires extensive discussion: how to use the
maximum tolerated dose (MTD). The MTD is the
maximum dose that an animal can tolerate for a
major portion of its lifetime without significant
impairment of growth or observable toxic effect other
than carcinogenicity.
Because cancer can take most of a lifetime to develop,
scientists widely agree that studies should be design-
ed so that the animals survive in relatively good
health for a normal lifetime. Whether the MTD, as
currently used, is the best way to achieve this
objective, however, is currently under debate. The
35
-------
increase of tumor incidence in treated versus control
animals; dose-related shortening of the time-to-
tumor occurrence or time-to-death with tumor; and a
dose-related increase in the proportion of tumors that
are malignant. The following sections describe
animal toxicity studies, including major areas of
importance in their design, conduct, and
interpretation. Particular consideration will be given
to the uncertainties associated with evaluating their
results.
3.2.7.2 Interpreting Manifestations of Toxicity
Toxic effects, regardless of the organ or system in
which they occur, can take various forms. First, the
severity of injury can increase as the dose increases,
as happens with some chemicals affecting the liver:
High doses kill liver cells, perhaps so many that the
liver is destroyed and some or all of the experimental
subjects die. As the dose is lowered, fewer cells are
killed, but they exhibit other forms of damage that
cause imperfections in their functioning. At lower
doses still, no cell deaths may occur and only slight
alterations in cell function or structure may be noted.
Finally, a dose may be achieved at which no effect is
observed, or at which there are only biochemical
alterations that have no known adverse effects on the
health of the animal. (Although some toxicologists
consider any such alteration, even if its long-term
consequences are unknown, to be "adverse," no clear
consensus has been reached on this issue.) One of the
goals of toxicity studies is to determine the No-
Observed-Adverse-Effect level (NOAEL), which is
the dose at which no adverse effect is seen; the role of
the NOAEL in risk assessment is discussed further in
subsequent sections.
Second, the incidence but not the severity of an effect
may increase with increasing dose. In such cases, as
the dose increases, the fraction of experimental
animals experiencing adverse effects (i.e., the
incidence of disease or injury) increases. At
sufficiently high doses, all experimental subjects will
experience the effect. Thus, increasing the dose
increases the probability (i.e., risk) that the
abnormality will develop in an exposed population.
Third, both the severity and the incidence of a toxic
effect may increase as the level of exposure increases.
The increase in severity is a result of increased
damage at higher doses, while the increase in
incidence is a result of differences in individual
sensitivity. In addition, the site at which a substance
acts (e.g., liver, kidney) may change as the dose
changes. Many toxic effects, including cancer, fall in
this category
Generally, as the duration of exposure increases, the
two critical doses (the NOAEL and the LOAEL)
decrease; in some cases, new effects not seen with
exposures of short duration appear after longer
exposures.
Toxic effects also vary in degree of reversibility. In
some cases, an adverse health effect will disappear
almost immediately following cessation of exposure.
At the other extreme, some exposures will result in a
permanent injury—for example, a severe birth defect
from fetal exposure to a substance that irreversibly
damaged the fetus at a critical moment of its
development. Further, some tissues such as the liver
can repair themselves relatively quickly, while
others such as nerves have no ability to repair
themselves. Most toxic responses fall somewhere
between these extremes. In many experiments,
however, the degree of reversibility is difficult to
ascertain.
The seriousness of a toxic effect must also be
considered. Certain types of toxic damage are clearly
adverse and are a definite threat to health. However,
other types of effects observed during toxicity studies
are not clearly of health significance. For example, at
a given dose a chemical may produce a slight increase
in red blood cell count. If no other effects are observed
at this dose, researchers cannot be sure that a true
adverse response has occurred. Determining whether
such slight changes are significant to health is one of
the critical issues in assessing safety.
There are several other important factors to consider
when examining toxic effects. A toxic effect can be
immediate, such as in poisoning, or delayed, as in
cancer. Indeed, cancer typically affects an individual
many years after continuous or intermittent
exposure to a carcinogen. An effect can be local (i.e.,
at the site of application) or systemic (i.e., carried by
the blood or lymph to different parts of the body). ,
Since the concentrations of substances found in
drinking water are usually too low to cause local
effects, systemic effects should be considered the key
concern in drinking water. Effects can also be
"idiosyncratic" - affecting people with a certain
genetic predisposition much more than others.
Finally, some substances - for example, the oil in
poison ivy - cause allergic or sensitization reactions
in which production of anti-bodies causes symptoms
such as inflammation.
3.2.1.3 Designing and Conducting Toxicity
Tests
Toxicity experiments vary widely in design and
protocols used. There are relatively well standard-
ized tests for various types of toxicity (i.e., National
Cancer Institute carcinogenicity bioassays) develop-
ed by regulatory and public agencies in connection
with the premarket testing requirements for certain
classes of chemicals. However, many other tests and
research-oriented investigations are conducted using
specialized study designs (i.e., carcinogenicity assays
in fish). This section describes a few of the critical
considerations associated with designing toxicity
experiments.
34
-------
benzene caused excess production of white blood cells
came from a series of case reports), seldom provide
the central body of information for risk assessment.
For this reason, and because they usually present no
unusual problems of interpretation, they are not
further reviewed here. Rather, our attention is
devoted to the two principal sources of toxicity data:
animal tests and epidemiological studies. These two
types of investigation can present interpretative
difficulties, some subtle, some highly controversial.
3.2.1 Animal Studies
Toxicity studies are conducted to identify the nature
of health damage produced by a substance 3 and the
range of doses over which damage is produced. The
usual starting point for such investigations is a study
of the acute (single-dose) toxicity of a chemical in
experimental animals. Acute toxicity studies are
used to calculate doses that will not be lethal to
animals used in toxicity studies of longer durations.
Moreover, such studies provide an estimate of the
compound's comparative toxicity and may indicate
the target organ system for chronic toxicity (e.g.,
kidney, lung, or heart). Toxicologists examine the
lethal properties of a substance and estimate its
(lethal dose for 50 percent of an exposed population).
A group of well-known substances and their LDgo
values are listed in Table 3-3.
LDso studies reveal one of the basic principles of
toxicology: Not all individuals exposed to the same
dose of a substance will respond in the same way.
Thus, at a dose of a substance that leads to the death
of some experimental animals, other animals will
get sick but recover, and still others will not appear
to be affected at all. Each of the many different types
of toxicological studies has a different purpose.
Animals may be exposed repeatedly or continuously
for several weeks or months in subchronic toxicity
studies, or for close to their full lifetimes in chronic
toxicity studies.
3.2.1.1 Using Animal Toxicity Data
Animal toxicity studies are based primarily on the
longstanding assumption that effects in humans can
be inferred from effects in animals. This principle of
extrapolating animal data to humans has been
widely accepted in the scientific and regulatory
communities. All of the chemicals that have been
demonstrated to be carcinogenic in humans (with the
possible exception of arsenic) are carcinogenic in
some, although not all, experimental animal species.
Table 3-3. Approximate Oral LOSO in a Species
of Rat for a Group of Weil-Known
Chemicals (3)
LD50
Chemical mg/kg (ppm)
Sucrose (table sugar)
Ethyl alcohol
Sodium chloride (common salt)
Vitamin A
Vanillin
Aspirin
Chloroform
Copper sulfate
Caffeine
Phenobarbital, sodium salt
DDT
Sodium nitrite
Nicotine
Aflatoxin B1
Sodium cyanide
Strychnine
29,700
14,000
3,000
2,000
1,580
1,000
800
300
192
162
113
85
53
7
6.4
2.5
The term substance refers to a pure chemical, to a chemical
containing impurities, or to a mixture of chemicals. It is clearly
important to know the identity and composition of a test substance
before drawing inferences about the toxicity of other samples of
the same substance that might have a somewhat different
composition.
In addition, the acutely toxic doses of many chemicals
are similar in humans and a variety of experimental
animals. The foundation of this inference of effects
between man and animals has been attributed to the
evolutionary relationships between animal species.
Thus, at least among mammals, the basic
anatomical, physiological, and biochemical
parameters are similar across species.
Although the general principle of making such
interspecies inferences is well founded, exceptions
have been noted. For example, guinea pigs are much
more sensitive to dioxin (2,3,7,8-TCDD) than other
laboratory animals. Many of these exceptions result
from differences in the ways various species handle
exposure to a chemical and to differences in
metabolism, distribution, and pharmacokinetics of
the chemical. Because of these potential differences,
it is essential to evaluate all interspecies differences
carefully when inferring human toxicity from animal
toxicologic studies.
In the particular case of long-term animal studies
conducted to assess the carcinogenic potential of a
compound, certain general observations increase the
overall strength of the evidence that the compound is
carcinogenic — for example, an increase in the
number of tissue sites affected by the agent or an
increase in the number of animal species, strains,
and sexes showing a carcinogenic response. Several
other factors affect the strength of the evidence,
including the occurrence of clear-cut dose-response
relationships in the data evaluated; the achievement
of a high level of statistical significance of the
33
-------
Table 3-2. Data and Assumptions Necessary to Estimate Human Dose of a Water Contaminant from Knowledge of Its
* Concentration (1)
Total Dose Is Equal to the Sum of Doses from Five Routes
Direct Ingestion Through Drinking:
-- Amount of water consumed each day (generally assumed to be 2 L for adults and 1 L for 10-kg child)
-- Fraction of contaminant absorbed through wall of gastrointestinal tract
-- Average human body weight
Inhalation of Contaminants:
-- Air concentrations resulting from showering, bathing, and other uses of water
-- Variation in air concentration over time
— Amount of contaminated air breathed during those activities that may lead to volatilization
-- Fraction of inhaled contaminant absorbed through lungs
-- Average human body weight
Dermal Exposure:
-- Period of time spent washing and bathing
-- Fraction of contaminant absorbed through the skin during washing and bathing
- Average human body weight
Ingestion of Contaminated Food:
-- Concentrations of contaminant in edible portions of various plants and animals exposed to contaminated ground water
-- Amount of contaminated food ingested each day
-- Fraction of contaminant absorbed through wall of gastrointestinal tract
-- Average human body weight
Skin Exposure for Contaminated Soil:
-- Concentrations of contaminant in soil exposed to contaminated ground water
— Amount of daily skin contact with soil
-- Amount of soil ingested per day (by children)
-- Absorption rates
-- Average human body weight
In general, toxicity studies in experimental animals
are of greatest value when experimental exposures
mimic the mode of human exposure. If both animals
and humans are exposed to a contaminant via
drinking water, it is generally assumed that the data
in animals can be applied directly to man. When
experimental routes differ from human routes (e.g.,
animal dose via injection; human exposure via
drinking water), a correction factor often must be
used to apply such data to human exposures.
3.2 Hazard Identification
In identifying hazards, two kinds of data are gather-
ed and evaluated: 1) data on the types of health injury
or disease that may be produced by a chemical, and 2)
data on the conditions of exposure under which injury
or disease is produced. The behavior of a chemical
within the body and the interactions it undergoes
with organs, cells, or even parts of cells may also be
characterized. Such data may be of value in
answering the ultimate question of whether the
form's of toxicity known to be produced by a substance
in one population group or in experimental settings
are also likely to be produced in humans. Hazard
identification is not risk assessment; this step simply
determines whether toxic effects observed in one
setting are likely to occur in other settings. In other
words: Are substances found to be carcinogenic or
teratogenic in experimental animals likely to have
the same result in humans?
Researchers obtain information on the toxic
properties of chemical substances through animal
studies, controlled epidemiological investigations of
exposed human populations, and clinical studies or
case reports of exposed humans. Other information
bearing on toxicity derives from experimental studies
in systems other than whole animals (i.e., in isolated
organs, cells, subcellular components) and from anal-
ysis of the molecular structures of the substances of
interest. These last two sources of information are
generally considered less certain indicators of toxic
potential, and accordingly they receive limited
treatment here.
Similarly, clinical studies or case reports, while
sometimes very important (the earliest signs that
32
-------
adults are assumed to consume 2 L of water each day
through all uses. Thus, if a substance is present at 10
mg/L (ppm) in water, the average daily individual
intake of the substance is:
10 mg/L x 2 L/day = 20 mg/day
Toxicity comparisons among different species must
take into account size differences, usually by dividing
daily intake by the weight of the individual. Thus, for
a man of average weight (usually assumed to be 70 kg
or 154 Ib), the daily dose of this substance is:
20 mg/day •*- 70 kg = 0.29 mg/kg/day
For a person of lower weight, such as a female or
child, the daily dose at the same intake rate would be
larger. For example, a 50-kg woman ingesting this
substance would receive a dose of:
20 mg/day •*- 50 kg = 0.40 mg/kg/day
Using the same equation, a child of 10 kg would
receive a dose of 2.0 mg/kg/day. However, children
drink less water each day than adults (say, 1 L), so a
child's dose would be:
10 mg/L x 1 L/day •*• 10 kg = 1.0 mg/kg/day
In general, the smaller the body size, the greater the
dose (in mg/kg/day) received from drinking water.
This is also true of experimental animals. Usually
rats or mice will receive a much higher dose of
drinking water contaminants than humans because
of their much smaller body size.
These sample calculations point out the difference in
measuring environmental concentrations and dose,
at least for drinking water. For air and other media,
the relationship between measured environmental
concentrations and dose is more complex. Table 3-2
lists the data necessary to obtain dose from data on
the concentration of a substance in water through
five routes or media.
Each medium of exposure must be treated separately
and some calculations are more complex than in the
above examples of dose per liter of water. A human
may be simultaneously exposed to the same
substance through several media (e.g., through
inhalation, ingestion, dermal contact). The "total
dose" received by an individual is the sum of doses
received through each individual route (see Table 3-
2).
In some cases, it may not be appropriate to add doses
in this fashion since the toxic effects of a substance
may depend on the route of exposure. For example,
inhaled chromium is carcinogenic to the lung, but it
appears that ingested chromium is not. In most cases,
however, as long as a substance acts at an internal
body site (i.e., acts systemically rather than only at
the point of initial contact), it is usually considered
appropriate to add doses received from several routes.
Many risk assessors use the terms exposure and dose
synonymously. In this chapter, however, dose means
the amount received by the subject and encompasses
several factors, including contact with a substance,
the size of the dose, the duration of exposure, and the
nature and size of the exposed population.
Two additional factors concerning dose and exposure
require special attention. The first is the concept of
absorption (or absorbed dose). The second is the
technique of extrapolation, or drawing inferences
from toxicities observed under one route of exposure
to predict the likelihood of toxicity under other
routes.
When a substance is ingested in the diet or in
drinking water, it enters the gastrointestinal tract.
When it is present in air (as a gas, aerosol, particle,
dust, fume, etc.), it enters the upper airways and
lungs. A substance may also come into contact with
the skin and other body surfaces as a liquid or solid.
Some substances may cause toxic injury at the point
of initial contact (skin, gastrointestinal tract, upper
airways, lungs, eyes). Indeed, at high concentrations,
most substances will cause at least irritation at these
points of contact. However, for many substances,
toxicity occurs after they pass through certain
barriers (i.e., the wall of the gastrointestinal tract or
the skin itself), enter blood or lymph, and gain access
to the various organs or systems of the body.
Chemicals may be distributed in the body in various
ways and then excreted. (However, some chemical
types - usually substances with high solubility in fat
such as DDT - can be stored for long periods of time,
usually in fat.)
Substances vary widely in extent of absorption. The
fraction of a dose that passes through the wall of the
gastrointestinal tract may be very small (1 to 10
percent for some metals) to substantial (close to 100
percent for certain types of organic molecules).
Absorption rates also depend on the medium in which
a chemical is present: a substance present in water
might be absorbed differently from the same
substance present in a fatty diet. These rates also
vary among animal species and among individuals
within a species.
Ideally, an estimation of a systemic dose should
consider absorption rates. Unfortunately, data on
absorption are limited for most substances, especially
in humans. As a result, absorption is not always
included in dose estimation (i.e., by default, it is
frequently considered to be complete). Sometimes
crude adjustments are made to dose estimations,
based on the molecular characteristics of a substance
and general principles for the estimation of
absorption rates.
31
-------
Table 3-1. Annual Risk of Death from Selected Common Human Activities (2)
Activity
Coal mining:
Accident
Black lung
Motor vehicle
Truck driving
Falls
Home accidents
Number of Deaths in
Representative Year
180
1,135
46,000
400
16,339
25,000
Individual Risk in
Representative Year
1.30 x 10-3 or 1/770
8 x 10-3 or 1/125
3.2 x 10-4 or 1/4,500
10"* or 1/10,000
7.7 x 10'5 or 1/13,000
1.2x 10-5 or 1/83,000
Lifetime
Risk
1/17
1/3
1/65
1/222
1/186
1/130
NOTE: Lifetime risk based on 70-year lifetime and 45-year work exposure.
3.1.1.2 Toxic Versus Nontoxic
The term "safe" commonly means "without risk."
Scientists, however, cannot ascertain conditions
under which a given chemical exposure is absolutely
without risk of any type. Zero risk is simply
immeasurable. On the other hand, they can describe
conditions under which risks are so low as to be
considered of no practical consequence to a specific
population. In technical terms, the safety of chemical
substances - whether in food, drinking water, air, or
the workplace - has typically been defined as a
condition of exposure under which there is a
"practical certainty" that no harm will result to
exposed individuals. (As described later, these
conditions usually incorporate large safety factors, so
that even more intense exposures than those defined
as safe may also carry extremely low risks.) Note that
most "safe" exposure levels established in this way
are probably risk-free, but science simply has no tools
to prove the existence of essentially a negative
condition.
Another fundamental concept is the classification of
chemical substances as either "safe" or "unsafe" (or
as "toxic" and "nontoxic"). This type of
classification, while common even among scientists,
can be highly problematic. All substances, even those
that we consume in high amounts every day, can be
made to produce a toxic response under some
conditions of exposure. In this sense, all substances
can be toxic. The important factor, then, is not simply
the degree of toxicity, but rather the degree of risk;
i.e., what is the probability that the toxic properties
of a chemical will be realized under actual or
anticipated conditions of human exposure? To answer
this question requires far more extensive data and
evaluation than the characterization of toxicity.1
3.1.1.3 Exposure and Dose
Humans can be exposed to substances in the
environment through air, water, or food. Other
circumstances may also provide the opportunity for
exposure, such as direct contact with a sample of the
substance or contact with contaminated soil.
Experiments for studying the toxicity of a substance
usually involve intentional administration to
subjects through the diet or inhaled air, or direct
application to skin. Experimental studies may
include other routes of administration: injection
under the skin (subcutaneous), into the blood
(usually intravenous), or into body cavities
(intraperitoneal).
The amount of a substance in the medium (air, diet,
etc.) in which it is present or administered is the
exposure concentration.The amount of the chemical
that is received by the target, or the dose, may be
different from the exposure amount.
The following example illustrates the difference
between these two measures. Suppose a substance is
present in drinking water. An individual's dose of
this substance depends on the amount present in a
given volume of water. For many environmental
substances, this amount ranges from less than a
microgram (ug) to greater than a milligram (mg).
The analyst will usually report the number of mg or
ug of the substance present in 1 L of water, i.e., mg/L
or ug/L. These two units are sometimes expressed as
parts per million (ppm) or parts per billion (ppb),
respectively.2
Given the concentration of a substance in water (say
in ppm) and the human consumption of water per
unit of time, it is possible to estimate the total
amount of the substance an individual will consume
through water. In most public health evaluations,
1 Some scientists will claim that carcinogens display their toxic
properties under all conditions of exposure, and that there is no
"safe" level of exposure to such agents. This problem receives
extensive treatment later in this chapter.
2 A liter of water weighs 1,000 g. One mg is thus one-millionth the
weight of a liter of water, and 1 ug is one-billionth the weight of a
liter of water.
30
-------
Chapter 3
Principles of Risk Assessment
3.1 Introduction
This chapter outlines the types of scientific data
needed and methods currently used to assess the
human health risks of environmental chemicals. It is
not intended as a complete discussion of the complex
topic of risk assessment. Instead, the chapter
provides a general overview of risk assessment as an
introduction to the process used to determine the
risks associated with chemical contamination (1).
Human health risk can be defined as the likelihood
(or probability) that a given chemical exposure or
series of exposures may damage the health of exposed
individuals. Risk assessment involves the analysis of
past chemical exposures, the adverse health effects of
which may or may not have already occurred. It also
involves prediction of the likely consequences of
future exposures.
Risk assessment is composed of four major
components: hazard identification, dose-response
assessment, human exposure assessment, and risk
characterization. A separate section is devoted to
each component for describing the methods and tests
used to gather data, the principles used for data
interpretation, and the uncertainties in both the data
and inferences drawn from them. Throughout these
discussions, key concepts (i.e., exposure, dose,
thresholds, and extrapolation) are defined and
extended descriptions provided.
Many of the principles discussed in this chapter are
widely accepted in the scientific community. Others,
such as thresholds for carcinogens and the utility of
negative epidemiological data, are controversial. In
such cases, various points of view are provided,
including the one most broadly adopted by public
health and regulatory officials.
Finally, the concepts and principles described here,
although broadly applicable, may not apply in
specific cases. In some instances, the data available
on a specific chemical may suggest that a general
principle (e.g., that data obtained in rodent studies
are generally applicable to humans) does not hold
true. In such instances, the usual approach is to
modify the risk assessment process to conform to the
scientific finding.
3.1.1 Concepts and Definitions
3.1.1.1 Risk
Risk is the probability of injury, disease, or death
under specific circumstances. It may be expressed in
quantitative terms, with values from zero
(expressing certainty that harm will not occur) to 1
(expressing certainty that it will). In many cases, risk
can only be described qualitatively, as "high," "low,"
or "trivial."
All human activities carry some degree of risk. Many
risks are known with a relatively high degree of
accuracy because enough data have been collected on
their occurrence. Table 3-1 lists the risks associated
with some common activities.
The risks associated with many other activities or
events, including exposure to chemical substances,
cannot be readily assessed and quantified. Although
considerable data have been gathered on the risks of
some types of chemical exposures (i.e., the annual
risk of death from intentional overdoses or accidental
exposures to drugs, pesticides, and industrial
chemicals), such data are generally restricted to
acute poisonings. In such situations, a single, very
high exposure results in an immediately observable
form of injury, thus leaving little doubt about
causation. Far more complex is risk assessment for
chemical exposure that does not cause immediately
observable forms of injury or disease (or only minor
forms, such as transient eye or skin irritation). These
types of exposure range from brief to extended and
continuous. This chapter focuses on risk assessment
for chronic, continuous exposure, although some
review of acute poisoning is included.
29
-------
Beagle Dogs. J. Env. Pathol. Toxicol. 2:835-851,
1981.
25. Munson, A., L. Sain, V. Sanders, B. Kauffman,
W. White, D. Page, D. Barnes, and J. Borzelleca.
Toxicology of Organic Drinking Water
Contaminants: Trichloroethane, Bromodichlor-
omethane, Dibromochloromethane, and Tri-
bromomethane. Env. Health Persp. 46:117-125,
1981.
26. Report on the Carcinogenesis Bioassay of
Chloroform. Washington, DC: National Cancer
Institute, 1976.
27. Jorgenson, R., E. Meierhenry, C. Rushbrook, R.
Bull, and M. Robinson. Carcinogenicity of
Chloroform in Drinking Water to Male
Osborne-Mendel Rats and Female B6C3F(1)
Mice. Fun.. Appl. Toxicol. 5:760-769,1985.
28. Roe, F., A. Palmer, A. Warden, and N. Van
Abbe. Safety Evaluation of Toothpaste
Containing Chloroform, I, Long-term Studies in
Mice. J. Env. Path. Toxicol. 2:799-819,1979.
29. Cotruvo, J., V. Simmon, and R. Spanggard.
Investigation of Mutagenic Effects of Products
of Ozonation Reactions in Water. Ann. New
York Acad. Sci. 289:124-140,1977.
28
-------
guidelines. Additionally, FSTRAC provides an
opportunity for states to discuss their individual
regulatory activities, methodology status, survey
progress, and research activities and priorities.
2.5.7.2 Workshops
ODW with assistance from OTTRS is conducting a
series of workshops in all EPA regions on assessing
and managing drinking water contamination. The
workshops are led by scientists and regulatory
officials directly involved in the implementation of
EPA's drinking water programs. The workshops,
conducted over a period of 2 to 3 days each, stress the
qualitative and quantitative risk assessment process.
Additionally, presentations on the principles of
pharmacokinetics, risk assessment, carcinogenicity,
and toxicology are provided for the various classes of
drinking water contaminants (i.e., inorganics,
synthetic organics, and pesticides). The workshops
focus primarily on the HA program, its development,
philosophy, and methodology. Analytical technology
and treatment techniques are discussed at length, as
well as the communication of potential or existing
health risks to the general public. Actual risk
management case-studies are presented to provide
hands-on experience to the attendees for specific
drinking water contaminants.
2.5.7.3 Emergency Response Network
The Emergency Response Network is a long-
established and important component of ODW's HA
program. It is designed to give state, local, and other
concerned parties rapid access to existing
information on drinking water contaminants. This
service is provided through a systematic access to
EPA experts, databases, HAs, and other criteria and
regulatory documents. Requests received by letter or
telephone from the concerned party (regional and
state EPA offices, state and local health departments,
local water treatment facilities, or other concerned
individuals or organizations) are logged in, classified,
and referred to a specific chemical manager within
the ODW Health Effects Branch. This staff member
has ready access to other staff scientists, HAs and
criteria documents, contractor support, and other
national experts to formulate a response to the
request. Depending on the nature of the request and
the degree of urgency, the response may be relayed to
the requesting party via letter, telephone, or
conference call.
2.6 References
1. The Safe Drinking Water Act, 42 U.S.C., 300f et
seq., 1974.
2. Federal Register 50 (219): 46941,1985.
3. The Safe Drinking Water Act Amendments of
1986. P.L. 99-339,1986.
4. House of Representatives Report No. 1185. 93rd
Congress, 2nd Session, 10,1974.
5. Federal Register 49 (227):46294,1984.
6. Ibid. 50 (219): 46947,1985.
7. Ibid., p. 46946.
8. Cotruvo, J.A., S. Goldhaber, and C. Vogt.
Development of Drinking Water Regulations for
Organic Contaminants in the U.S., Paper
presented at 2nd National Conference on
Drinking Water, Edmonton, Alberta, 1986.
9. Federal Register 53 (14):1901,1988.
10. Ibid., 52 (130): 25720,1987.
11. Ibid., p. 25731.
12. Ibid., p. 25695.
13. Ibid., p. 25710.
14. Ibid., p. 25720.
15. Federal Register 54 (97): 22062,1989.
16. Ibid., 52 (130): 31516,1987.
17. Ibid., 54 (124): 27544,1989.
18. Ibid., p. 27486.
19. Orme, J., C. Sonich-Mullin, and E. Ohanian.
Water Chlorination, Vol. 6, Chemistry,
Environmental Impact, and Health Effects.
Lewis Publishing, Chelsea, MI, 1989.
20. Drinking Water and Health. Vol. 2.
Washington, DC: National Academy of
Sciences, 1980.
21. Registry of Toxic Effects of Chemical
Substances. Washington, DC: National
Institute of Occupational Safety and Health,
1984.
22. Drinking Water and Health. Vol. 7.
Washington, DC: National Academy of
Sciences, 1987.
23. Jorgenson, T. and C. Rushbrook. Effects of
Chloroform in the Drinking Water of Rats and
Mice: Ninety-Day Subacute Toxicity Study.
EPA-600/1-80-030,1980.
24. Heywood, R., R. Sortwell, P. Noel, A. Street, D.
Prentice, F. Roe, P. Wadsworth, A. Warden, and
N. Van Abbe. Safety Evaluation of Toothpaste
Containing Chloroform, Long-term Study in
27
-------
Table 2-18. HAs for 50 Pesticides
Acifluorfen
Ametryn
Ammonium suifamate
Atrazine
Baygon
Bentazon
Bromacil
Butylate
Carbaryl
Carboxin
Chloramben
Chlorthalonil
Cyanazine
Dalapon
Dacthal
Diazinon
Dicamba
1,3-Dichloropropene
Oieldrin
Dimethrin
Dinoseb
Diphenamid
Disulfoton
Diuron
ETU
Endothall
Fenamiphos
Fluometuron
Fonofos
Glyphosphate
Hexazinone
Maleic hydrazide
MCPA
Methomyl
Methyl parathion
Metolachlor
Metribuzin
Paraquat
Picloram
Prometone
Pronamide
Propachlor
Propazine
Propham
Simazine
2,4,5-T
Tebuthiuron
Terbacil
Terbufos
Trifluralin
requirements have been established for a treatment
technique. Monitoring for these chemicals will help
EPA to determine whether VOCs should be
regulated. An additional factor that influences
potential regulation is the degree of toxicity of each
VOC. To define this degree of toxicity and to assist
those faced with immediate VOC drinking water
contamination problems, the ODW has prepared HAs
for most of the chemicals listed in Table 2-9. If the
toxicity data were adequate, these HAs were
finalized. HAs were not finalized for many of these
VOCs, since toxicity data were quite limited.
ODW is also preparing HAs for inorganics,
disinfectants, and disinfection by-products. These
documents will be available for review in FY 90.
2.5.6.4 Department of the Army Munition
EPA has entered into a Memorandum of Under-
standing with the Department of the Army to provide
support in the preparation of HAs on various
munitions chemicals having the potential to
contaminate drinking water during their production,
use, or disposal. Table 2-19 lists the munitions
chemicals currently identified for HA development.
The HAs for trinitroglycerol and nitrocellulose,TNT,
HMX, RDX, and DIMP have been completed and the
others are in various stages of preparation and
review.
Table 2-19. Army Munition Chemicals Scheduled for
Health Advisory Development
Trinitroglycerol (TNG)
Nitrocellulose (NC)
2,4,6-Trinitrotoluene (TNT)
Cyclotrimethylenetrinitramine
(1 -hexahydro-1,3,5-trinitro-1,3,5-triazine)(RDX)
Cyclotetramethylenetetranitramine
(octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazoline) (HMX)
Diisopropyl-methylphosphate (DIMP)
Zinc chloride
White phosphorus
Hexachloroethane
Nitroguanidihe
dimethylmethylophosphonate (DMMP)
1,3-Dinitrobenzene
2,4-and 2,6-Dinitrotoluene
In addition to these HAs, ODW has prepared toxicity
profiles for the additional munition chemicals listed
in Table 2-20. These chemicals are largely
contaminants in and/or by-products of munitions
manufacturing or waste disposal processes and may
or may not be considered for future HAs. The toxicity
profiles provide a brief survey of the properties of the
chemical and the status of the toxicity database as is
available from the published literature.
Table 2-20. Chemicals for Which Toxicity Profiles Have
Been Prepared for the Department of the Army
1-Nitronaphthalene
3,4-Dinitrotoluene
2,3-Dinitrotoluene
2,6-Dinitrotoluene
1 -Chloro-4-nitrobenzene
1 -Methyl-2-nitrobenzene
3,5-Dinitrotoluene
2,5-Dinitrotoluene
1 -Methyl-4-nitrobenzene
1,2-Dichloro-4-nitrobenzene
2.5.7 Other Facets of the Health Advisory
Program
2.5.7.1 Federal-State Toxicology and
Regulatory Alliance Committee
The Federal-State Toxicology and Regulatory
Alliance Committee (FSTRAC) is a working group
composed of EPA and state experts in the areas of
risk assessment and risk management for drinking
water contaminants. The goal of the committee,
which meets approximately twice yearly, is to allow
an exchange of information between federal and state
agencies and to foster cooperation and consistency in
the development of drinking water standards.
Activities of the FSTRAC meetings include
coordinating and updating the status of many EPA
programs, including ODW drinking water
regulations, HAs, NFS, and risk assessment
26
-------
below) is not calculated and the DWEL is provided
to give the risk manager a reference point for
evaluat-ing noncarcinogenic endpoints. This infers
that carcinogenicity should be considered the toxic
effect of greatest concern when lifetime exposure is
anticipated.
• Step 3. Lifetime HA: For noncarcinogenic
chemicals the Lifetime HA is determined in Step 3
by factoring in other sources of human exposure to
the chemical (e.g., air, food). Preferably, the
relative source contri-bution (RSC) from drinking
water is based on actual exposure data. If data are
not available, a value of 20 percent is assumed for
organic or inorganic chemi-cals.
These three steps can be summarized as follows:
1. Determine RfD inmg/kg/day:
Table 2-17. Completed Health Advisories
RfD =
NOAEL orLOAEL in mg/kg/day
Uncertainty Factor
2. Determine the DWEL in mg/L assuming 100
percent drinking water contribution:
DWEL =
(fi/D) (70 kgforanadult)
(2 LI day)
3. Determine Lifetime HA in mg/L:
Lifetime HA = DWEL X Percent drinking water
contribution
If the chemical is a known or probable human
carcinogen, Lifetime H As are not determined. (See
Section 2.1 for a general discussion of the ODW's
approach for carcinogenic effects.)
2.5.6 Health Advisory Development Status
2.5.6.1 Completed Health Advisories
Health Advisories for the chemicals listed in Table 2-
17 have been completed and are available for use by
any interested organization or individual.
2.5.6.2 National Pesticides Survey
The ODW has entered into a joint venture with EPA's
Office of Pesticide Programs (OPP) to monitor those
pesticides either known to have occurred in drinking
water or most likely to be found in ground water. This
joint venture is known as the National Pesticides
Survey (NPS). An important element of NPS is the
development of HAs for all pesticides anticipated to
be detected in water samples. These HAs will allow a
NPS manager to issue immediate health guidance
when any pesticides are discovered in drinking water
supplies. Thus, an early step in the NPS was to
Acrylamide
Alachlor
Aldicarb/sulfoxide/sulfone
Arsenic*
Atrazine
Barium
Benzene
Cadmium
Carbofuran
Carbon tetrachloride
Chlorobenzene
Chromium
Cyanide
2,4-D
DBCP
o,m,p-Dichlorobenzene
1 ,2-Dichloroethane
1 ,1 -Dichloroethylene
cis-1 ,2-Dichloroethylene
trans-1 ,2-Dichloroethylene
1 ,2-Dichloropropane
p-Dioxane
Dioxin
EDB
Legionella
Endrin
Epichlorohydrin
Ethyl benzene
Ethylene glycol
Heptachlor/Heptachlor epoxide
Hexachlorobenzene
n-Hexane
Lindane
Mercury
Methoxychlor
Methyl ethyl ketone
Nickel
Nitrate/Nitrite
Pentachlorophenol
Styrene
Tetrachloroethylene
Toluene
Toxaphene
1 ,1 ,1 -Trichloroethane
Trichloroethylene
Vinyl chloride
Xylenes
"Undergoing revision.
compile a list of chemicals to be evaluated during the
sampling and analysis effort and for which HAs were
needed. This list was compiled based on usage, water
'solubility, persistence in soil, and soil-water
adsorption partition coefficient information. HAs for
50 pesticides were prepared as a part of this effort
(Table 2-18).
Other aspects of the NPS monitoring program
already completed or nearing conclusion include
development of analytical methods, selection of a
hydrogeology scheme, finalization of sampling
techniques, and a pilot sampling survey. This survey
will ultimately involve approximately 1,500
groundwater wells, weighted toward areas of
probable occurrence as influenced by pesticide usage
and hydrogeology data.
2.5.6.3 Unregulated Volatile Organic
Chemicals
Section 1445 of the SDWA directs EPA to require
public drinking water systems to monitor for un-
regulated volatile organic chemicals (VOCs). These
are VOCs for which no primary drinking water
regulations specifying a Maximum Contami-nant
Level (MCL) have been developed and no
25
-------
Table 2-16. Standard Assumptions Used to Develop Health
Advisories
Protected Individual
One-Day HA: 10-kg child
Ten-Day HA: 10-kg child
Longer-Term: 10-kg child and 70-kg adult
Lifetime HA: 70-kg adult
Cancer risk estimates: 70-kg adult
Volume of Drinking Water Ingested/Day
10-kg child: 1 L
70-kg adult: 2 L
Relative Source Contribution
In absence of chemical-specific data: 20%
Uncertainty Factors*
10: NOAEL from human study
100: LOAEL from human study, NOAELfrom animal study
1,000:LOAEL from animal study, NOAELfrom animal study of
less than lifetime duration (when calculating Lifetime HA)
10,000: LOAEL from animal study of less than lifetime duration
(when calculating Lifetime HA)
"In some cases, an additional uncertainty factor of
1-10 may be used to account for scientific judgment.
Thus, derivations from these basic guidelines may be
required when the total database for a specific .
chemical is considered.
2.5.5 Calculation of Health Advisories
As previously stated, HAs are based on identification
of the adverse health effects associated with the most
sensitive and meaningful noncarcinogenic endpoint
of toxicity. The induction of this effect is related to
both a particular exposure level and a specific period
of exposure and is most often determined from the
results of experimental animal studies. The general
formula used to calculate HA values is as follows:
HA =
(NOAEL or LOAEL) X (bw)
(UF)X ( LI day)
= mgIL ( ug/L)
where
NOAEL or
LOAEL = No- or Lowest-Observed-Adverse-
Effect Level in mg/kg bw/day
bw = Assumed body weight of a child (10
kg) or of an adult (70 kg)
UF = Uncertainty factor (10,100, or 1,000)
L/day = Assumed daily water consumption of
a child (1 L/day) or of an adult (2
L/day)
If the available data are derived from inhalation
studies, the total exposed dose (TED) must first be
determined before calculating the HA. This is
accomplished by adjusting the exposure
concentration for the ventilation rate and body
weight of the exposed animal to achieve a dose of
mg/kg bw/day.
2.5.5.1 Calculation of One-Day and Ten-Day
Health Advisories
The preceding formula is used for the One-Day and
Ten-Day HAs by inserting the data for a 10-kg child
consuming 1 L of water per day, the appropriate UF,
and the NOAEL or LOAEL derived from a study of
appropriate duration.
2.5.5.2 Calculation of Longer-Term Health
Advisories
Two values are calculated for the Longer-Term HA,
using data for both the 10-kg child consuming 1 L
per day and the 70 kg adult consuming 2 L per day
along with the NOAEL and LOAEL from the study of
appropriate duration. In this case, a 90-day to 1-year
animal study representing approximately 10 percent
of an individual's lifetime and the appropriate UF for
the type of data available are employed.
2.5.5.3 Calculation of Lifetime Health
Advisories
The Lifetime HA represents that portion of an
individual's total lifetime exposure to the chemical
that is attributable only to drinking water. All other
HA values are calculated based on the assumption
that drinking water is the sole source of the contami-
nant. The Lifetime HA is derived in a three-step
process with the first two steps being mathematically
equivalent to the procedure used for all other HA
calculations. The third step in the calculation is
added to factor in the relative contribution from other
exposure sources of the chemical.
• Step 1. Reference Dose: Step 1 determines the
Refer-ence Dose (RfD), formerly called the
Acceptable Daily Intake (ADI) (see Section 2.1).
• Step 2. Drinking Water Equivalent Level: From
the RfD, a Drinking Water Equivalent Level
(DWEL) is calculated. A DWEL is defined as a
medium-specific exposure level (i.e., mg/L in
drinking water), as-suming 100 percent exposure
from that medium, which is considered to be
protective for noncarcino-genic health effects over
a lifetime of exposure. The DWEL is derived by
multiplying the RfD by the assumed body weight
of an adult (70 kg) and then dividing by the
assumed daily water consumption of an adult (2
L/day). For drinking water the DWEL is expressed
in mg/L or ug/L. If the contaminant is classified as
a Group A or B carcinogen, the calcula-tion is
halted at this point. The Lifetime HA (Step 3
24
-------
Other known criteria, guidance, or published
standards are also included in the HA document as
an additional means of evaluating the status of the
contamination. Finally, analytical methodology and
treatment technologies are included to assist the user
in making the appropriate public health
management decisions. Should the user require
additional information, a list of the cited references is
included. If the HA is based on an existing Criteria
Document, it is also referenced. Additionally, the
user may contact an EPA regional office as well as
The Office of Drinking Water, EPA Headquarters,
Washington, DC, for further assistance. Also, EPA
provides a toll-free Safe Drinking Water Hotline,
(800) 426-4791 or (for within area code 202) 382-
5533.
2.5.3 Preferred Data for Health Advisory
Development
In deriving the HA values, EPA defines specific types
of data as most pertinent to each phase of the process.
These data may be subdivided into three categories,
as indicated in Table 2-15. The following sections ex-
plain how these data are selected to derive each of the
HA values.
Tablo 2-15. Preferred Data for HA Development
Duration of Exposure
One-day HA: Up to 7 daily doses
Ton-day HA: Up to 30 daily doses '
Longer-term HA: Subchronic study
Lifetime HA: Chronic study
Subchronic study (with added uncertainty
factor)
Route of Administration
Oral: Drinking water, gavage, or diet
Inhalation
Subcutaneous or intraperitoneal
Test Species
Human
Appropriate animal model
Most sensitive species
2.5.3.1 Duration of Exposure
One-Day Health Advisory: The One-Day HA is
calculated for a 10-kg child and assumes a single
acute exposure to the chemical. It is generally
derived from a study of 7 days or less.
Ten-Day Health Advisory: The Ten-Day HA, also
calculated for a 10-kg child, assumes a limited
exposure period of 1 to 2 weeks. It is generally
derived from a study of up to 30 days duration.
Longer-Term Health Advisory: Longer-Term HAs,
which are derived for both a 10-kg child and a 70-kg
adult, assume a human exposure period of
approximately 7 years (or 10 percent of an
individual's lifetime). The longer-term HA is
generally derived from a study of Subchronic
duration.
Lifetime Health Advisory: The Lifetime HA is
derived for a 70-kg adult and assumes an exposure
period over a lifetime (approximately 70 years). The
Lifetime HA is generally derived from a study of
chronic duration (approximately 2 years in rodents
and other experimental animals), but Subchronic
studies may be used by adjusting the uncertainty
factor employed in the calculation.
2.5.3.2 Route of Administration
In all cases, the route of choice is oral exposure. The
preferred vehicle is drinking water, but
administration via gavage or the diet is acceptable.
Inhalation, subcutaneous, or intraperitoneal
administration data are used on a case-by-case basis
when no oral or other satisfactory data are available.
2.5.3.3 Test Species
The preferred species for assessing health effects is
humans. However, since data in humans do not
usually provide reliable dose-response information
(and since very few human exposure data exist),
selection of an appropriate animal model is usually
required. This selection is based on the model's
similarity to man in its pharmacokinetic handling of
the chemical under evaluation. When different
animal models vary considerably in their response to
a chemical, the most sensitive, relevant species is
selected. However, depending on the toxicity of the
chemical and the scope of the data available,
information from all sources may be used.
2.5.4 Assumptions Used in a Health Advisory
The HA values are presented under the "Quanti-
fication of Toxicological Effects" heading of the
document and are based on the assumptions listed in
Table 2-16.
For consistency in calculation, EPA considers the
protected individual to be either a 10-kg child — the
individual likely to be most adversely affected during
short-term exposure periods — or a 70-kg adult. The
Agency also assumes that the average drinking
water intakes for a child and an adult are 1 and 2 L
per day, respectively. Additionally, if actual exposure
data are not available, it is assumed that drinking
water accounts for 20 percent of a person's total
intake of organic or inorganic chemicals. ODW uses
this final assumption only when calculating the
Lifetime HA, for which exposures from other sources
(e.g., air or food) may be significant.
Standard uncertainty factors (UFs) are also assumed
during the HA calculation (see Table 2-5). Note that
the selection of UFs requires case-by-case judgments.
23
-------
dermatitis and keratosis of palms and soles (and
eventually skin cancer), enlargement of the liver,
kidney injury, central nervous system effects, and
aplastic anemia (inability to properly produce red
and white blood cells).
Arsine gas, which contains arsenic, has its own
distinct health effect. Exposure to this gas causes
hemolysis, a bursting of the red blood cells that leads
to anemia and other deleterious health effects.
4.2.1.7 Other Inorganic Ions
Fluoride in drinking water reduces dental caries
(cavities) at levels of 0.7-1.2 mg/L but can discolor the
teeth at levels above 2 mg/L, and can weaken bones
at extremely high doses.
As mentioned earlier in the discussion of absorption,
nitrates from agricultural run-off can be converted to
nitrite within the GI tract. In infants, acute doses of
nitrites can cause methemoglobinemia or "blue-
baby" syndrome, in which the blood cannot carry
oxygen to the body's tissues, resulting in possibly
fatal asphyxiation and a blue cast to the skin. In
adults, nitrites can combine with other compounds in
the GI tract to form carcinogenic nitrosamines.
The sodium ion occurs naturally in drinking water
and as a result of man's activities, such as salting icy
roads in the winter. Sodium in drinking water
contributes to the daily total intake of sodium, which
is already high in the U.S. from consumption of salty
food and is a focus of concern because of its known
association with hypertension. Some epidemiological
studies have shown an increased incidence of
hypertension in those consuming drinking water
high in sodium; other similar studies have failed to
show such a correlation. It is estimated that sodium
in drinking water contributes only about 10 percent
to the daily total intake. As a result, some scientists
suggest that sodium in drinking water is a problem
mainly for those who must eat a salt-restricted diet
(2).
4.2.1.8 Asbestos
Asbestos is the name of a family of fibrous minerals;
the most commonly found type is called chrysotile
asbestos. Most of the concern over asbestos centers on
its inhalation effects: asbestosis (impaired lung
function), lung cancer, and mesothelioma (cancer of
the membranes lining the chest and abdomen).
However, asbestos has also contaminated drinking
water supplies due to mining operations, geologic
erosion, degradation of asbestos cement pipes, and
atmospheric sources. Research has only partially
defined the health effects of ingested asbestos.
Although epidemiologic studies have shown an
increased incidence of GI tract cancer among those
consuming water contaminated with asbestos,
ingestion experiments with laboratory animals and
various forms of asbestos have not revealed a
reproducible, organ-specific carcinogenic effect (3).
4.2.1.9 Inorganic Radionuclides in Drinking
Water
Certain unstable elements spontaneously decay into
different atomic configurations, in the process
releasing radiation consisting of alpha particles, beta
particles, or gamma rays. These particles and rays •
can damage living tissue and/or cause cancer to
develop, with the degree of damage depending on the
type of radiation and means of exposure (i.e.,
inhalation, ingestion, or external radiation). As an
element undergoes radioactive decay, it progresses
through a series of atomic configurations, each one of
which is called an isotope. Isotopes are identified by
atomic weight, a number indicating the atom's
number of neutral and charged particles. Everyone is
exposed to some background radiation from both
cosmic rays and sources on earth, such as radioactive
soil and rock.
The three naturally occurring series of isotopes stem
from the decay of the isotopes uranium-238,
uranium-235, and thorium-232. With regard to
naturally occurring radionuclides in drinking water,
the uranium-238 series (for which decay isotopes
include uranium-234, radium-226, and radon-222)
and the thorium series (for which decay products
include radium-228) are of the greatest concern.
Manmade radioactive isotopes, such as strontium-90,
also pose health risks, but such isotopes generally
occur in lower concentrations in the environment
than the naturally occurring radionuclides. However,
site-specific contamination, such as leaking nuclear
waste disposal sites and nuclear power plant
accidents, can pose health risks due to the manmade
radionuclides.
The concern over radionuclides focuses on their
potential to cause cancer. In drinking water regu-
lations, the radioactivity of an isotope is expressed in
units called picocuries (pCi) that represent the
isotope's number of disintegrations per second; 1 pCi
is equal to 4.7 x 1010 disintegrations per second.
Radium-226 is perhaps the single most important
radioactive isotope found in drinking water. It is
deposited in bone and can cause bone cancer. EPA
compliance data indicate that about 500 U.S. public
ground-water supplies exceed the existing NIPDWR
limit for combined radium-226 and radium-228 of 5
pCi/L. However, many exceeded the limit only
slightly. Radium-228 is also a bone- seeking
carcinogen.
The most commonly found forms of natural uranium
are uranium-234 and uranium-238. Natural
uranium is also believed to cause bone cancer;
54
-------
where a concentration above 200 ug cadmium per
gram of tissue damages the organ.
The kidney injury is thought to be responsible for the
brittle bones and pain described above. Through
inhalation, cadmium can cause chronic obstructive
pulmonary disease and emphysema. Although
cadmium has been established as a probable human
carcinogen (Class Bl) through inhalation, ingested
cadmium has not been shown to be carcinogenic
(Class D for ingestion). Cadmium has also been
associated with high blood pressure in humans and
laboratory animals.
4.2.1.4 Iron
Iron in drinking water can result from naturally
occurring iron in the soil and from corrosive water
contacting iron water piping. Chronic iron toxicity is
not a significant problem with regard to drinking
water. However, iron-rich water can discolor clothes.
Acute iron toxicity, on the other hand, is a major
problem in that children often mistake iron
supplements for candy and accidentally consume
large amounts of iron. Such accidental ingestion can
cause GI tract problems, metabolic acidosis, and
cardiovascular collapse.
cavity cancer but is not considered a drinking water
problem because its occurrence in drinking water is
rare and its absorption through food and drinking
water intake is low. Other inorganics in this category
would include antimony, barium, beryllium,
phosphorus, and trivalent chromium. However, in
certain parts of the U.S., metals not occurring in high
levels in other areas have been a drinking water
problem. For example, in northern Illinois high
levels of barium in groundwater have required
treatment because of possible GI and cardiovascular
effects. In some areas, selenium contamination has
required treatment because of possible GI effects at
lower doses and liver damage at very high doses.
Other metals require further study. For example,
aluminum is generally considered to have a low order
of toxicity. As a result, aluminum is purposefully
used in water treatment, and many people consume
antacid medicines composed of aluminum hydroxide
without experiencing ill health effects. However,
aluminum may be associated with neuropsychiatric
disorders in patients receiving renal dialysis who
also receive aluminum as part of their therapy. Also,
laboratory animals exposed to aluminum have shown
growths called neurofibrils that have also been seen
in people suffering from Alzheimer's disease.
4.2.1.5 Other Metals
There are many other metals that can be found in
drinking water. Many of these metals (for example,
copper, zinc, chromium, cobalt, manganese, and the
nonmetal selenium) are considered essential min-
erals in the diet. Although potentially harmful to
health in very high doses (such as high occupational
doses or site-specific situations such as mine leach-
ate), these metals are not considered a major health
hazard in drinking water. Further, many foods
contain significant amounts of metals, far over-
shadowing the contribution from drinking water in
terms of the total dose. It is important to note,
however, that a certain subpopulation of people are
unusually sensitive to copper (i.e., Wilson's disease).
Some metals, such as zinc, copper, and manganese, in
the amounts usually found in drinking water
generally present only the same type of relatively
minor problem that iron does, i.e., staining and
objectionable taste. Silver produces "cosmetic" effects
like skin discoloration at the higher observed
drinking water levels. Others have (or may have)
chronic toxicities similar to the metals described in
the previous section, but occur very rarely in
drinking water and only exert a toxic effect at
extremely high levels. Such metals may also occur in
drinking water in a less harmful form than in
occupational or medical exposures.
For example, nickel in occupational exposures has
been shown to increase the risk of lung and nasal-
4.2.1.6 Arsenic
Arsenic occurs naturally in bedrock and soil and is a
waste product from the manufacture of products such
as pesticides and from smelting operations. Arsenic is
considered an essential dietary element, although in
very small amounts. There are four main types of
arsenic: organoarsenicals, pentavalent arsenic,
trivalent arsenic, and arsine gas.
Arsenic can be excreted relatively quickly; its half-
life is 2 days. Arsenic has a wide variety of chronic
toxic effects, but many of them stem from its ability
to increase the permeability of capillaries in various
locations in the body. This increased permeability
allows plasma to leak into the tissues, for example,
leading to severe diarrhea and kidney injury. (In the
past, women took arsenic to obtain a "milk and roses"
complexion; the arsenic broke capillaries in the ,
cheeks, creating a rosy complexion.) Arsenic
damages the central nervous system by inflaming
peripheral nerves and causing brain injuries, and the
liver by fatty infiltration and tissue necrosis. Also,
EPA has classified arsenic as a Type A human
carcinogen (human carcinogen based on
epidemiological studies), with skin and lung cancer
as the two principal types of cancer arising from
arsenic exposure.
Early signs of chronic arsenic poisoning include
diarrhea; skin pigmentation and texture changes;
edema of eyelids, face, and ankles; and a garlic odor of
the breath. As exposure continues, symptoms include
53
-------
as a B(2) substance, a probable human carcinogen
based on sufficient evidence from animal studies and
insufficient evidence from humans. Lead's
reproductive hazards have long been known since in
the past it was used to induce abortions.
The more subtle, long-term effects of lead are the
subject of much scientific interest. Some research has
suggested that subtle developmental problems in
children, such as learning disabilities, may be linked
to long-term, low-level exposure to lead both in
childhood and the prenatal stage.
In the diagnosis of lead poisoning, specific
concentrations of lead in the blood (ug lead/g blood)
correlate with specific symptoms. X-rays of long
bones also aid diagnosis by indicating the degree of
exposure. The presence of compounds called heme
precursors in the urine indicates that lead exposure
has interfered with the synthesis of the heme portion
of hemoglobin.
Finally, lead can also exist in organolead compounds.
Here, the symptoms are different than described
above. The central nervous system again is the main
target, but the symptoms are primarily
neuropsychiatric, including insomnia, nightmares,
irritability, and anxiety.
4.2.1.2 Mercury
In terms of toxicologic effects, mercury can be divided
into three types: elemental mercury (such as in
thermometers and other measuring devices);
inorganic mercury salts (used in products such as
skin cream, antiseptics, and diuretics); and
organomercurials (used in fungicides for grain).
Elemental mercury is not significantly absorbed
through the GI tract when it is consumed orally and
thus is, in this instance, essentially nontoxic.
However, elemental mercury as a vapor can be
readily absorbed through the lung, where it can enter
the bloodstream and subsequently the brain. Also,
elemental mercury released into the ocean is
converted into the more harmful methylmercury by
plankton, which are then consumed by fish.
The inorganic mercury salts, because they are
ionized and water soluble, do not readily pass
membranes in the body. Thus, these compounds do
not readily pass the blood/brain or placental barriers
and only about 10 percent of a given dose is absorbed
through the GI tract. The organomercurials, because
they are lipid soluble, readily pass membranes in the
body. Thus, these compounds readily pass through
the blood/brain and placental barriers and about 90
percent of a given dose is absorbed through the GI
tract. The half-life for both inorganic mercury salts
and organomercurials is about two months.
Acute mercury poisoning, common before World War
II, when many nonprescription medicines contained
mercury, is relatively rare today. Of more importance
for drinking water is chronic, long-term mercury
poisoning. Chronic exposure to vapor from elemental
mercury produces neuropsychiatric symptoms,
including depression, irritability, shyness, insomnia,
emotional instability, forgetfulness, confusion,
excessive perspiration, uncontrolled blushing
(erethism), and tremors. Less is known about the
effects of chronic exposure to inorganic mercury salts,
although the kidney has been identified as the target
organ of toxicity. Chronic exposure to methylmercury
(from grain and through the food chain) affects the
central nervous system but affects the senses rather
than emotions, as with elemental mercury vapors.
Effects include abnormal tingling sensations,
constriction of the visual field, hearing defects,
speech impairment, and ataxia (impaired muscle
coordination). Mercury is a teratogen and the fetus is
extremely susceptible to methylmercury. (A well- •
known case of mercury poisoning in Minimata Bay,
Japan resulted in birth defects and widespread
illness in 1968.) EPA has given mercury a Class D
carcinogenicity designation: i.e., inadequate evidence
of carcinogenicity from animal data.
In the diagnosis of mercury poisoning, the various
neuropsychiatric symptoms may be difficult to detect.
Often family members will note a pronounced change
in a relative's behavior. To reveal elevated blood
mercury levels, tests can be performed on blood and
urine, as well as hair, which also concentrates
mercury.
4.2.1.3 Cadmium
Cadmium has several sources or uses. It is a by-
product of lead and zinc mining; is used as a pigment
and in corrosion-resistant coatings and nickel-
cadmium batteries; and is released when fossil fuels
are burned. It can also enter drinking water when
corrosive water contacts certain types of water
piping. Cadmium's health effects were made public
by an incident in Japan in which rice paddies were
contaminated with cadmium-rich zinc mine
drainage. Rice grown on the paddies concentrated the
cadmium and those eating it suffered easily broken
bones and extreme joint pain; thus, acute cadmium
poisoning was referred to as Itai-Itai (ouch-ouch)
disease.
Since cadmium is water soluble, only 1-5 percent of a
given dose is absorbed in the GI tract (although 10-40
percent can be absorbed through the lung). Cadmium
distributes to the kidney and liver. Unlike the
previously discussed metals, cadmium has an
extremely long half-life of 10-30 years. Acute
cadmium poisoning causes GI disturbances. Chronic
cadmium poisoning affects the kidney the most,
52
-------
When considering metabolism, researchers also look
at species, strain, and gender differences. In
laboratory animals, these species, strain, and gender
differences can play an important role in the animal's
ability to metabolize a chemical; in humans, these
factors are less significant. However, age is an
important factor in both humans and laboratory
animals, with both the very young and very old more
susceptible to certain chemicals. Also, exposure to
certain chemicals can induce the production of the
enzyme P-450 and similar enzymes (for example,
phenobarbital, DDT, and dioxin for P-450), resulting
in a higher level of that enzyme in the body compared
to the general population.
4.2 Toxicology of Selected Substances
The following sections provide a general overview of
the toxicology of certain substances. This overview
focuses on health effects due to exposure to or con-
sumption of drinking water, as opposed to effects
from occupational exposure or air pollution. The
omission of a substance does not imply that exposure
to it in any setting is safe. Further, although some
substances have been identified by both chemical and
trade names, space will not permit listing all of the
trade names for commercially produced toxic
substances such as pesticides. Wherever possible,
toxic chemicals have been grouped according to
chemical structure.
For some of the following substances, an EPA
carcinogenicity class designation has been provided;
Table 4-1 summarizes EPA's carcinogenicity rating
system.
Tablo 4-1. EPA's Carcinogenicity Rating System
Class
Designation Definition
A Human carcinogen based on sufficient
cpidomiological evidence
B Probable human carcinogen based on at
least limited evidence of carcinogenicity to
humans (Bi), or usually a combination of
sufficient evidence in animals and inadequate
data in humans (B2)
C Possible human carcinogen based on limited
evidence of carcinogenicity in animals in the
absence of human data
D Not classified based on inadequate evidence
of carcinogenicity from animal data
E No evidence of carcinogenicity for humans
(no evidence of carcinogenicity in at least two
adequate animal tests in different species or
in both epidemiological and animal studies)
4.2.1 Inorganics
Described below are the health effects of lead,
mercury, cadmium, other metals, arsenic, inorganic
ions (fluoride, nitrate, and sodium), asbestos, and
inorganics such as radionuclides.
4.2.7.7 Lead
Lead in drinking water comes from several sources,
including lead materials in water piping; food; tetra-
ethyl lead in gasoline (which ends up in the air, soil,
and water); lead-based paint (pre-WW II paint);
improperly glazed earthenware; and occupational
sources such as smelters and lead-acid battery
manufacturing. Drinking water risk assessors must
consider the total dose of lead from all these sources,
not only from drinking water. For drinking water,
the most significant source of lead is lead solder and
piping in water distribution systems, particularly
when contacted by corrosive water. While the use of
lead solder and piping in repairs and construction of
water piping has been banned (SDWA Amendments
of 1986), much of the existing water distribution
system still contains lead materials.
Since lead is a water-soluble compound, it is
relatively poorly absorbed; in adults, only about 10
percent of the lead ingested through the GI tract is
absorbed into the bloodstream. Lead is initially
distributed to the kidney and liver and then
redistributed mostly to the bone (about 95 percent),
where it is visible in x-rays. Lead does not readily
enter the central nervous system in adults because
the blood/brain barrier can keep it out. However, in
children, the blood/brain barrier is not as developed;
such exposure in children can affect their mental
development.
Acute lead poisoning is rare. Instead, chronic lead
poisoning or "plumbism" is the effect of most interest
to risk assessors. An important chronic effect,
although not the most serious, is a GI tract condition
known as lead colic. This condition is more common
in adults and can be quite painful. Because of the
pain, lead colic often causes an exposed person to seek
medical help, thus discovering the lead exposure
before more serious problems can develop. Lead can
also affect the neuromuscular system by decreasing
muscle tone in the wrists and feet. This condition,
"lead palsy," was common among house painters
before World War II. The most serious effects of lead
are called lead encephalopathies. These effects are
more common in children and can be quite serious;
approximately 25 percent of children with lead
encephalopathies die and approximately 40 percent
of the survivors experience neurologic after-effects.
Exposure to lead also affects the body's blood-forming
system. Lead can interfere with the synthesis of heme
(part of the oxygen-carrying compound hemoglobin);
can cause anemia; and can damage red blood cells in
a condition known as basophilic stippling. Effects on
the kidney include injury and cancer in laboratory
animals. As a result, lead has been classified by EPA
51
-------
Figure 4.2. Several types of chemical reactions occurring in metabolism (1)
Examples of the general type of oxidation reactions catalyzed by the cytochrome P-450-containing mono-oxygenases:
• Aromatic hydroxylation R-
Aliphatic hydroxylation
RCH2CH2CH3 ——*• RCH2CHOHCH3
N, O, and S-Dealkylation
H
R-(N, O, S)-CH3
R(RNH2, OH, SH)
Epoxidation
O
i = CHR' > R-CH-CH-R1
Desulfuration
S
II
R,R2P-X
O
II
+ S
Sulfoxidation
O
RSR,
N-Hydroxylation
O
li
RIMH-C-CHg
O
II
R-NOHCCH3
Non P-450:
Epoxide hydrolase (closely associated with P-450)
Epoxide R -jf^i — OH
2 Hydrolase R~^*r — OH
Esterases and amidases
O
II
CH3 C-OC2H5
Ethyl Acetate
O
II
- CH3COH
Acetic Acid
Alcohol and aldehyde dehydrogenase
CH3CH2OH + NAD"
CH3CH2OH
Ethanol
O
II
-—»• CH3CH —
Acetaldehyde
O
II
CH3COH
Acetate
Several of the reactions do not involve P-450, but
rather rely on enzymes such as amine oxidase,
epoxide hydrolase, esterases, amidases, and alcohol
and aldehyde dehydrogenase (see Figure 4-2). The
alcohol dehydrogenase reaction serves well to explain
metabolism. If we could not metabolize alcohol and
relied on our kidneys and lungs to eliminate it
without metabolism, we would remain intoxicated for
months after drinking alcohol.
The Phase II reactions add various specialized
molecules to compounds to make them water soluble.
For instance, the body can add a glucuronic acid
molecule (similar to a sugar molecule, which has
several hydroxyl groups). Other Phase II reactions
add a tripeptide called glutathione (consisting of
glycine, cysteine, and glutamic acid) to create a
compound that is further metabolized and then
excreted. Still others add a sulfate to make a
compound more water soluble. There are also several
other types of Phase II reactions, but they play a less
important role in the metabolism of toxicants and
thus will not be described here.
50
-------
possible through the hair follicles, through the cells
of the sweat glands and sebaceous glands, and
through cuts or abrasions (which increase the rate
and degree of absorption). The sole means of
absorption through the skin appears to be passive
diffusion. The dermis is much more permeable than
the epidermis.
4.1.2 Distribution
Distribution of toxicants to various organs depends
on the ease with which it crosses cell membranes, its
affinity for various tissues, and the blood flow
through the organ. A toxicant's site of concentration
is not necessarily the target organ of toxicity. For
example, many lipid-soluble toxicants (such as the
chlorinated hydrocarbon insecticides) are stored in
fat, where they cause relatively little harm, and lead
can be harmlessly stored in bone. However, a contam-
inant stored in fat can be.released back into the
bloodstream during conditions such as starvation,
dieting, or illness when fat is consumed.
A number of anatomical barriers in the body are
thought to prevent or hinder the entrance of certain
toxicants into organs. The so called blood/brain
barrier does not prevent toxicants from entering the
central nervous system (CNS); rather, the
physiologic conditions at the blood/brain interface
make it more difficult for some toxicants to leave the
blood and enter the CNS. In general, lipid-soluble
toxicants can cross the blood/brain barrier but some
water-soluble toxicants cannot. Even less of a barrier
is the "placental barrier"; simply stated, any
chemical absorbed into the mother's bloodstream will
cross her placenta and enter the bloodstream of the
fetus to some degree.
4.1.3 Excretion
Chemicals can be excreted from the body in several
ways, but the two most important routes with regard
to drinking water are through the kidney and
through the biliary system (liver). The kidney
removes toxicants from the blood in the same way
that the endproducts of metabolism are eliminated
i.e., glomerular filtration, passive tubular diffusion,
and active secretion. Glomerular filtration simply
involves filtering of compounds below a certain
molecular weight (and thus size) through pores in a
part of the kidney referred to as the glomeruli. In this
way, the kidney acts as a filter for the blood, with the
heart providing the pumping power to move the blood
through the porous membrane. All toxicants with a
molecular weight less than 60,000 will filter through
the glomeruli unless they are bound to plasma
proteins (only a few do so). Most toxicants in drinking
water have a molecular weight of between 100 and
500 and thus easily pass through the glomeruli. The
toxicants then pass through collecting ducts and
tubules where, if they are lipid-soluble, they can
defeat the excretion process by moving (by passive
diffusion) through the tubule wall and back into the
bloodstream. In contrast, water-soluble compounds
continue on and are excreted through the urine. The
kidney also employs "carrier" processes to actively
transport some toxicants into the urine.
The liver eliminates toxicants from the body through
the bile, which passes into the intestine through the
gall bladder and bile duct, and finally out of the body
via the feces. As in the kidney, the transport
mechanisms used are passive diffusion and carrier-
mediated transport. Toxicants that have been
excreted into the intestine through the bile can be
reabsorbed (especially if they are lipid-soluble) into
the bloodstream while in the intestine. This
reabsorption process is called enterohepatic
circulation.
Toxicants are also excreted through several other
routes, including the lungs, GI tract, cerebrospinal
fluid, milk, sweat, and saliva. Of these, only milk is
significant for drinking water contaminants. Because
milk has a relatively high concentration of fat (3.5
percent), lipid-soluble compounds such as DDT and
PCBs can concentrate in it. -i>, because milk is
slightly acidic (with a pH of 6.5), basic compounds
may concentrate in it. In this way, toxicants may be
passed from mother to child or from cows to humans.
An important concept in excretion is a toxicant's half-
life, which is the time it takes for one-half of the
chemical to be eliminated from the body. Thus, if an
imaginary chemical A has a half-life of 1 day, 50
percent of it will remain within the body 1 day after
absorption. Two days after absorption, 25 percent will
remain; at 3 days, 12.5 percent will remain, and so
on.
4.1.4 Metabolism
Because lipid-soluble compounds are reabsorbed in
the kidney and intestine due to their ability to cross
cell membranes, the body metabolizes these toxicants
into water-soluble compounds, which can be excreted
easily. However, in some instances, metabolism of a
chemical creates a more toxic chemical or does not
change the chemical's toxicity.
Two types of reactions occur in metabolism:
relatively simple Phase I reactions (oxidation,
reduction, and hydrolysis) and more complex Phase II
reactions (conjugation and synthesis). All of these
reactions occur primarily in the liver. Oxidation is
the mechanism of metabolism for many compounds.
An important family of enzymes in oxidation is the P-
450 mono-oxygenases. Figure 4-2 illustrates several
types of oxidation reactions catalyzed by cytochrome
P-450 mono-oxygenases; such reactions create water
soluble compounds.
49
-------
Figure 4-1. Key routes of chemical absorption, distribution, and excretion.
Excretion
methemoglobinemia or "blue-,baby syndrome." Also,
nitrates in the GI tract of people of all ages may lead
to the formation of carcinogenic compounds called
nitrosamines.
Age is also an important factor affecting the
intestine's ability to act as a barrier to certain
toxicants. For example, leadis absorbed to a much
greater extent in newborns than in adults.
Even though a chemical has been absorbed through
the GI tract, it can still be excreted or metabolized by
the intestine or liver before it reaches the systemic
circulation. This first chance to eliminate the
chemical is known as the first-pass effect.
Drinking water pollutants can also enter the body
through the lung—for example, when volatile
organic compounds volatilize in a warm shower. For
some chemicals, such as the VOCs, absorption
through the lungs can be considerable.
The lungs are anatomically designed to absorb and
excrete chemicals, as is shown by the absorption of
oxygen and excretion of carbon dioxide. The alveoli
have a large surface area (50-100 m2), are supplied
with a high flow of blood, and the blood is very close
(10 um) to the air space within the alveoli. Toxicants
may have to pass through as few as two cells to travel
from the air into the bloodstream.
t
The skin is relatively impermeable to toxicants:
This barrier is over 100 cells thick. However, some
toxicants, such as carbon tetrachloride, can be
absorbed through the skin in sufficient quantities to
cause liver injury. Absorption through the skin is
48
-------
Chapter 4
Principles of Toxicology
In order to familiarize the drinking water risk
assessor with the basic physiology on which
toxicology is based, this chapter begins by describing
the absorption, distribution, excretion, and
metabolism of toxic substances. Then, the toxicology
of four broad categories of substances — inorganics,
pesticides, solvents and vapors, and other synthetic
compounds—is reviewed.
4.1 Absorption, Distribution, Excretion,
and Metabolism of Toxic
Substances
The body's response to a toxic chemical depends on
the dose administered. However, once a toxicant
enters the body, the interplay of four processes —
absorption, distribution, excretion, and
metabolism—determines the actual effect of a toxic
chemical on the "target organ," which is the organ
that can be damaged by that particular chemical. For
example, carbon tetrachloride affects the liver and
benzene affects the hematopoietic (blood-cell
forming) system. Figure 4-1 summarizes routes of
absorption, distribution, and excretion.
4.1.1 Absorption
Understanding absorption requires a review of the
two main mechanisms by which toxicants pass
through membranes within the body: passive
transport (simple diffusion and filtration) and active
transport (assisted chemical transport). Simple
diffusion—movement from an area of higher to lower
concentration—accounts for much of the transport of
chemicals within the body. Lipid-soluble compounds,
especially nonionized forms, readily diffuse through
the lipid part of cell membranes. Filtration can be
defined as the flow of a solute through pores in a
membrane. It comes into play in the kidney, where
these pores are relatively large, thus allowing
excretion of chemicals through the urinary tract.
Active transport involves certain carrier compounds
that move chemicals from areas of low concentration
to high concentration. (For more detail on active
transport, see Section 4.1.3.)
Absorption of toxicants across body membranes and
into the bloodstream can occur in the gastrointestinal
(GI) tract, lungs, and through the skin. For drinking
water, the GI tract is the key portal of entry. Most
chemicals, once they enter the GI tract, must be
absorbed to exert their toxic effect. Lipid-soluble,
nonionized compounds such as DDT and PCBs are
more readily absorbed by diffusion in the GI tract
than lipid-insoluble, ionized compounds such as lead
and cadmium. The GI tract also employs specialized
active transport systems for compounds such as
sugars, amino acids, pyrimidines, calcium, and
sodium; in general, these active transport systems do
not play a major role in absorption of toxicants. (Some
toxicants, however, can be absorbed in the GI tract
through active transport systems; for example, lead
can be absorbed through calcium's transport system
and thallium through iron's transport system.)
The effect of digestive fluids must be considered when
examining toxicants in the GI tract. For example, a
toxin like snake venom is nontoxic when
administered orally because it is a protein that
stomach enzymes break down into amino acids, much
in the same way that a hamburger is digested. In
newborns, the GI tract has a higher pH and a higher
number ofE. coli bacteria than in adults. These
conditions convert nitrate, a common drinking water
pollutant from agricultural run-off, into the more
toxic chemical, nitrite. The nitrite then interferes
with the blood's ability to carry oxygen, thus causing
47
-------
is not to specify an RfD, but to ascertain risk. There
are no means available to accomplish this for
noncarcinogens. The MOE is used as a surrogate for
risk; as the MOE becomes larger, the risk becomes
smaller. At some point, most scientists agree that the
MOE is so large that human health is almost
certainly not jeopardized. The magnitude of the MOE
needed to achieve this condition will vary among
different substances, but its selection would be based
on factors similar to those used to select safety factors
to establish RfDs.
The risk characterization process can result in very
different statements of risk. As shown in Table 3-9,
risk characterization for an imaginary Chemical A
produces three distinct statements. The first
statement indicates that 327 per 1 million exposed
people will die, using three significant digits to
estimate the risk outcome. The second statement
more cautiously gives only a range of people that will
die —100 to 1,000 people per 1 million people exposed.
Finally, the third statement can only suggest that an
assumption that the chemical in question is
carcinogenic to humans is prudent.
3.6 References
1. Principles of Risk Assessment: A Nontechnical
Review. Environ Corporation, Washington, DC,
1986. Report prepared for EPA.
2. Ibid., p. 2.
3. Registry of Toxic Effects of Chemical Substances.
Washington, DC: National Institute of
Occupational Safety and Health, 1979.
4. Klassen, C.D., M.O. Amdur, and J. Doull, eds.
Casarett and Doull's Toxicology, 3rd Edition.
Macmillan, New York, NY, 1986.
5. Workshops on Assessment and Management of
Drinking Water Contamination. EPA-600/M-
86/026, U.S. Environmental Protection Agency,
Center for Environmental Research Information,
Cincinnati, OH, 1987, p. 41.
6. Ibid., p. 38.
7. Application of Risk Assessment to Food-Safety
Decision Making. Regulatory Toxicology &
Pharmacology 3:275-307,1983.
8. Ibid.
Table 3-9. Three Different Statements Resulting from the
Same Risk Characterization Process
327 per 1,000,000 exposed people will die from lifetime
exposure to Chemical A.
Chemical A is carcinogenic in rats and mice. Application of
low-dose extrapolation models and human exposure estimates
suggests that the range of risks in humans is 100-1,000
deaths per 1,000,000 persons exposed.
Chemical A is carcinogenic in rats and mice and it is prudent
public health policy to assume it is also carcinogenic in
humans.
46
-------
After the known or expected human dose is estimat-
ed, carcinogenic risk can be characterized. Although
the models in use yield a wide range of dose-response
relationships for the same data, the projections of the
more protective models are not likely to underesti-
mate risk, at least to experimental animals. (They
may strongly overestimate it.) In a few cases, dose-
response data are available from human epidemio-
logical studies and may be used in lieu of animal data
for low-dose extrapolation.
Certain classes of carcinogens do not apparently
possess the capacity to damage DNA (i.e., they are
not genotoxic). Some scientists maintain that such
nongenotoxic carcinogens must operate under
threshold mechanisms. Many of the reasons for such
a hypothesis are sound, but no general consensus has
yet emerged on this matter. It is nevertheless
possible that some classes of carcinogens could be
treated in the same way as noncarcinogens for
purposes of establishing RFDs.
3.4 Human Exposure Assessment
Assessment of human exposure requires estimation
of the number of people exposed and the magnitude,
duration, and timing of their exposure. The
assessment could include past exposures, current
exposures, or exposures anticipated in the future. In
some cases, measuring human exposure directly,
either by measuring levels of the hazardous agents in
the ambient environment or by using personal
monitors, is fairly straightforward. In most cases,
however, detailed knowledge is required of the
factors that control human exposure, including those
factors that determine the behavior of the agent after
its release into the environment. The following types
of information are required for this type of exposure
assessment:
• The factors controlling the production of the
hazardous agent and its release into the
environment
• The quantities of the agent released, and the
location and timing of release
• The factors controlling the fate of the agent in the
environment after release, including its
movement, persistence, and degradation
(degradation products may be more or less toxic
than the original agent)
• Human contact with the agent, including the size
and distribution of vulnerable human populations,
and activities that facilitate or prevent contact
• Information on human intakes
The amount of information available varies greatly
from case to case. For some agents, fairly detailed
information is available on the sources of release into
the environment and on the factors controlling the
quantities released. However, for many agents little
information is available on the factors controlling
dispersion and fate after release. Measurements of
transport and degradation in the complex natural
environment are often difficult to conduct; thus, it is
more common to rely on mathematical models of the
key physical and chemical processes, supplemented
with experimental studies conducted under simpli-
fied conditions. Such models have been developed in
considerable detail for radioisotopes, but have not yet
been developed in comparable detail for other
physical and chemical agents.
In comparison with toxicology and epidemiology, the
science of exposure assessment is still at a very early
stage of development. Except in fortunate circum-
stances, in which the behavior of an agent in the
environment is unusually simple, uncertainties
arising in exposure assessments are often at least as
large as those arising in assessments of inherent
toxicity.
Once these various factors are known, human data
can be estimated, as described earlier. The dose, its
duration and timing, and the nature and size of the
population receiving it are the critical measures of
exposure for risk characterization.
3.5 Risk Characterization
The final step in risk assessment combines the
information gained and analysis performed during
the first three stages to determine the likelihood that
humans will experience any of the various forms of
toxicity associated with a substance. Risk is
generally characterized as follows:
1. For noncarcinogens and for the noncarcinogenic
effects of carcinogens, the margin-of-exposure
(MOE) is estimated by dividing the experi-mental
NOAEL by the estimated exposure dose.
2. For carcinogens, risk is estimated at the human
dose by multiplying the actual human dose by the
risk per unit of dose projected from the dose-
response modeling. A range of risks might be
produced, using different models and assumptions
about dose-response curves and the relative
susceptibilities of humans and animals.
Although risk characterization can be far more
complex than is indicated here (especially if problems
of timing and duration of exposure are introduced),
the MOE and the carcinogenic risk are the ultimate
measures of the likelihood of human injury or disease
from a given exposure or range of exposures. RfDs are
not measures of risk; they are derived by imposing a
specified safety factor (or, in the above language, a
specified MOE). The purpose of risk characterization
45
-------
Table 3-8; Lifetime Risks Derived from Different
Extrapolation Models (8)
Model Applied
Lifetime Risk (1.0 mg/kg/day)
One-hit
Multistage
Multihit
Weibull
Probit
6.0 x 1C-5 (1 in 17,000)
6.0 x 10-6 (1 in 167,000)
4.4 x 10-7 (1 in 2.3 million)
1.7x 10-8 (1 in 59 million)
1.9 x 10-1° (1 in 5.3 billion)
NOTE: All risks are for a full lifetime of daily exposure. The lifetime is used
as the unit of risk measurement because the experimental data
reflect the risk experienced by animals over their full lifetimes. The
values shown are upper confidence limits on risk.
be consistent with the data. In many cases, however,
such data are very limited, resulting in great
uncertainty in how to select a model for low-dose
extrapolation. At present, understanding of the
mechanism of carcinogenesis is still quite limited.
Biological evidence, however, does indicate a
linearity of tumor initiation, and consequently linear
models are frequently used by regulatory agencies.
The one-hit model always yields the highest estimate
of low-dose risk. This model is based on the biological
theory that a single "hit" of some minimum critical
amount of a carcinogen at a cellular target—namely,
DNA— can initiate an irreversible series of events
that eventually lead to a tumor.
EPA generally uses the linearized multistage model
for low-dose extrapolation because it usually yields
estimates of risk that are the most conservative,
representing a plausible upper limit for the risk. In
other words, the actual risk is unlikely to be higher
than the risk predicted under this model.
The probit model incorporates the assumption that
each individual in a population has a "tolerance" dose
and that these doses are distributed in the population
in a specified way. The other models (Weibull, multi-
hit, and logit) have more complex bases and are not
widely used. None of these models currently
incorporates a threshold dose for an exposed
population.
3.3.2.3 Slope of the Dose-Response
The toxicologist must also keep in mind the slope of
the dose-response. In Figure 3-2, dose-responses for
the imaginary chemicals A and B are shown. Note
that, although both chemicals have the same LDso,
the values for higher and lower doses differ greatly.
3.3.2.4 Interspecies Extrapolation
For the majority of agents, dose-response evaluation
primarily involves the analysis of tests that were
performed on laboratory animals. In extrapolating
Figure 3-2 Slope of the dose-response.
7.0
§ 6.0
3
-------
3.3.2.2 Potency and High-to-Low Dose
Extrapolation
Table 3-7 illustrates the need for high-to-low dose
extrapolation. Assume that a substance has been
tested in mice and rats of both sexes and has been
found to produce liver cancer in male rats. A typical
summary of the data from such an experiment might
be as follows:
Table 3-7. Incidence and Probability of Liver Cancer at Low
and High Doses (7)
Lifetime
Daily Dose
(mg/kg/day)
0
125
250
500
1,000
Lifetime
Incidence of
Liver Cancer
in Rats
0/50
0/50
10/50
25/50
40/50
Lifetime
Probability of
Liver Cancer
0.0
0.0
0.20
0.50
0.80
The incidence of liver cancer is expressed as a
fraction, and is the number of animals found to have
liver tumors divided by the total number of animals
at risk. The probability (P) of cancer is simply the
fraction expressed as a decimal (i.e., 25/50 = 0.50).
Although there is no effect at 125 mg/kg/day, the
response is nevertheless compatible with a risk of
about 0.05 (5 percent) because of the statistical
uncertainties associated with the small numbers of
animals used.
This experiment reveals that if humans and rats are
about equally susceptible to the agent, an exposure of
250 mg/kg/day in humans will increase their lifetime
risk by 20 percent; if 1,000 people were to be exposed
to this substance at this dose for a lifetime, then 200
of these people will be expected to develop cancer.
This is an extremely high risk and obviously one that
few people would sanction. However, it is near the
low end of the range of risks that can be detected in
animal experiments.
To continue with the illustration, assume that the
estimated daily dose of the chemical in the human
population is 1.0 mg/kg/day. It thus becomes of
interest to know the risk to male rats at 1.0
mg/kg/day.
However, a great difference lies between the doses
used experimentally (125 -1,000 mg/kg/day) and the
dose of interest (1.0 mg/kg/day). Figure 3-1
illustrates the difference between the dose-response
observed in experiments and the dose-response of
ultimate interest to toxicologists. The risks that
would exist at a dose of 1.0 mg/kg/day are quite small
and to determine whether they exist at all would
require enormous numbers of animals (perhaps
hundreds of thousands). In these circumstances,
scientists must rely on something other than
experimentation to estimate potential risk— i.e.,
mathematical models to estimate low-dose risks from
high-dose risks.
Figure 3-1. Dose-response observable in experiments and
range of inference for dose-response at low
doses.
Observable
Range
Range of
Inference
Dose
Such models describe the expected quantitative
relationship between risk (P) and dose (d), and are
used to estimate a value for P at the dose of interest
(in our example, the dose of 1.0 mg/kg/day). The
accuracy of the projected P at d is a function of how
accurately the mathematical model describes the
true, but practically immeasurable, relationship
between dose and risk at the low dose levels.
Various models may lead to very different
estimations of risk. None is chemical-specific; that is,
each is based on general theories of carcinogenesis
rather than on data for a specific chemical. None can
be proved or disproved by current scientific data,
although future results of research may increase our
understanding of carcinogenesis and help refine
these models. Regulatory agencies currently use one-
hit, multistage, and probit models, but regulatory^
decisions are usually based on results of the one-hit
or multistage models. They also use multihit,
Weibull, and logit models for risk assessment.
If several of these models are applied to the
hypothetical liver cancer data, several different
estimates of lifetime risk for male rats at the dose of
1.0 mg/kg/day can be derived (see Table 3-8). No
experimental basis is available for deciding which
estimate is closest to the truth. Nevertheless, it is
possible to show that the true risk is very unlikely to
be higher than the risk predicted by the various
models.
In cases in which relevant data exist on biological
mechanisms of action, the selection of a model should
43
-------
general population versus populations such as
workers expected to exhibit a narrower range of
susceptibilities). Safety factors of 10,100,1,000, and
10,000 have been used in various circumstances.
NOAELs are used to calculate the reference dose
(RfD, formerly called Acceptable Daily Intake, ADI)
for humans for chemical exposures. The RfD is
derived by dividing the experimental NOAEL, in
mg/kg/day for the toxic effect appearing at lowest
dose, by one of the safety factors listed above. The
RfD (or its equivalent) is thus expressed in
mg/kg/day. For example, a substance with a NOAEL
from a chronic toxicity study of 100 mg/kg/day may
be assigned an RfD of 1 mg/kg/day for chronic human
exposure.
This approach has been used for several decades by
EPA and other federal regulatory agencies such as
the Food and Drug Administration, as well as by such
international bodies as the World Health
Organization and by various committees of the
National Academy of Sciences.
Although some biological justification can be found
for using safety factors to protect the more sensitive
members of the human population, scientific support
for the specific safety factors used is limited.
However, evaluation of inter species and intraspecies
variability data indicates that the current approach
is protective.
There is no way to ensure that exposures at RfDs
estimated in this fashion are without risk. The RfD
represents an acceptable, low level of risk but not a
guarantee of safety. Conversely, there maybe a
range of exposures well above the RfD, perhaps
including the experimental NOAEL itself, that bears
no risk to humans. The "NOAEL-safety factor"
approach includes no attempt to ascertain how risk
changes below the range of experimentally observed
dose-response relations.
3.3.2 Effects That May Not Exhibit
Thresholds
At present, only agents displaying carcinogenic
properties are treated as if they do not display
thresholds (although a few scientists suggest that
some teratogens and mutagens may behave similar-
ly). In more technical terms, the dose-response curve
for carcinogens in the human population achieves
zero risk only at zero dose; as the dose increases above
zero, the risk immediately becomes finite and there-
after increases as a function of dose. Risk in this case
is the probability of producing cancer, and at very low
doses the risk can be extremely small (this will vary
according to the potency of the carcinogen).
3.3.2.1 The Carcinogenic Process
Cancer can be defined as an uncontrolled new growth
of cells, or "neoplasm," with a tendency to be
invasive and metastasize (or spread). In some cases,
neoplasms can also be benign, or slow to develop,
noninvasive, and local. The type of carcinogenesis
depends on the type of cell involved: Carcinomas are
malignant growths of epithelial cells; lymphomas
are usually malignant neoplasms in lymph tissue;
sarcomas are malignant neoplasms in bone, muscle,
or other connective tissue; and leukemias are
malignant growths of cells in blood-forming tissues.
By one theory of chemical carcinogenesis, the
condition proceeds in two stages: initiation
(irreversible cell damage) and promotion
(development of a neoplasm in tissue in which
initiation has already occurred) (1). Initiation can
occur and not immediate-ly proceed to cancer because
of the body's ability to repair or suppress the
carcinogenic process. Initiators are referred to as
genotoxic carcinogens because they bind to the
genetic DNA. Primary genotoxic carcinogens act
directly on the DNA, while secondary carcinogens
must be metabolized to another form to exert their
genotoxic effect.
Other cancer-causing agents - epigenetic carcinogens
- do not act directly on the DNA. These substances
can act by promoting a genotoxic effect or through
various other mechanisms. For example, inhaled
asbestos fibers cause cancer through a solid-state
epigenetic effect, in which the physical nature of the
asbestos fibers contacting lung and other tissues
causes cancer. Other epigenetic carcinogens include
hormones (which only cause cancer in high doses),
immunosuppressive agents, and cocarcinogens
(which increase the carcinogenicity of a genotoxic
agent when administered with it). Some carcinogens
appear capable only of initiating the process and thus
are termed "initiators." Others called promoters act
only at later stages, and some carcinogens may act at
several stages.
Some scientists postulate that a very small amount of
a carcinogen, even a single molecule, can affect the
transition of normal cells to cancerous cells at one or
more of the various stages, and that a greater amount
of the carcinogen merely increases the probability
that a given transition will occur. Under these
circumstances, an absolute threshold below which
there is no effect on the process (even though the
effect may be exceedingly small) is extremely
unlikely.
This theory of the carcinogenic process is still under
extensive scientific scrutiny and is by no means
established, though it has substantial support in the
scientific community. The "multistage" model, as the
theory is called, has influenced the development of
some of the models used for dose-response evaluation.
Before describing these models, the experimental
dose-response information obtained from bioassays
and the need for models of the dose-response
relationship must be discussed.
42
-------
When humans are exposed to two or more chemicals,
several results may occur. The chemicals may act
independently, that is, exposure to the additional
chemical(s) has no observable effect on the toxic
properties of the first chemical. Or, toxic effects of
chemicals may be additive; that is, if chemical A
produces 1 unit of disease and chemical B produces 2
units of disease, then exposure to chemicals A and B
produces 3 units of disease. Exposure to combinations
of chemicals may also produce a greater- than-
additive or synergistic effect; that is, exposure to
chemicals A and B produces more than 3 units of
disease. Chemicals can act as potentiators, in which
exposure to chemical A normally produces no disease
but greatly increases the effect of chemical B. Ethyl
alcohol (and other forms of alcohol) and carbon
tetrachloride are examples of such substances. (When
carbon tetrachloride was used widely as a stain
remover, those using it with hangovers - i. e., high
blood ethyl alcohol levels - sometimes suffered severe
liver damage.) Finally, chemicals may reduce the
degree of toxicity of each other (antagonism); that is,
exposure to chemicals A and B produces less than 3
units of disease. Hazard evaluation of such mixtures
of chemicals is complex and not standardized.
3.2.4 Hazard Identification: A Summary
For some substances, the available database includes
substantial information on effects in humans and
experimental animals, as well as information on the
biological mechanisms underlying the production of
one or more forms of toxicity. In other cases, the
database is highly limited and includes only a few
studies in experimental animals.
In some cases, all the available data may point in a
single direction, leaving little ambiguity about the
nature of the toxicity associated with a given
compound; in others, the data may include
apparently conflicting sets of experimental or
epidemiological findings. It is not unusual for toxicity
tests to show conflicting results on well-studied
compounds. If the tests were performed properly,
positive test results usually outweigh negative test
results. Confusion may be compounded by the
observation that the type, severity, or site of toxicity
may vary with the species of animal exposed.
Although results in animals are and have been useful
in predicting effects in humans, such notable
exceptions as the testing on thalidomide have
occurred. (Premarket testing of thalidomide on
animals did not reveal its teratogenic effects in
humans.) This complex issue, briefly mentioned here,
must be considered for each compound examined.
A proper hazard evaluation should include a critical
review of each pertinent data set and of the total
database bearing on toxicity. It should also include
an evaluation of the inferences about toxicity in
human populations who might be exposed. At this
stage of risk assessment, however, there is no
attempt to project human risk. To do so, at least two
additional sets of analyses must be conducted: the
dose-response assessment and the human exposure
assessment.
3.3 Dose-Response Assessment
The next step in risk assessment describes the
relationship between the amount of exposure to a
substance and the extent of toxic injury or disease.
Even where good epidemiological studies have been
conducted, reliable quantitative data on exposure in
humans are rarely available. Thus, in most cases,
dose-response relationships must be estimated from
studies in animals, which immediately raises three
serious problems: 1) animals are usually exposed at
high doses, and effects at low doses must be predicted
by using theories about the form of the dose-response
relationship; 2) animals and humans often differ in
susceptibility (if only because of differences in size
and metabolism); and 3) the human population is
heterogeneous, so some individuals are likely to be
more susceptible than average.
Toxicologists conventionally make two general
assumptions about the form of dose-response
relationships at low doses: for effects that involve
alteration of genetic material (including the
initiation of cancer) and for most other biological
effects. These assumptions are discussed in the
following subsections.
3.3.1 Threshold Effects
Commonly accepted theory suggests that most
biological effects of noncarcinogenic chemical
substances occur only after a threshold dose is
achieved. In the experimental systems described
here, the threshold dose is approximated by the
NOAEL.
Another widely accepted premise, at least in the
setting of public health standards, is that the human
population is likely to have much more variable
responses to toxic agents than the small groups of
well-controlled, genetically homogenous animals
ordinarily used in experiments. Moreover, the
NOAEL is itself subject to some uncertainty since, for
example, epidemiologists may not be sure that the
most serious effects of a substance have been
identified. For these reasons, standard-setting and
public health agencies divide experimental NOAELs
by large "safety factors" when examining substances
that display threshold effects. The magnitude of
safety factors varies according to the following: the
nature and quality of the data from which the
NOAEL is derived; the seriousness of the toxic
effects; the type of protection being sought (protection
against acute, subchronic, or chronic exposures); and
the nature of the population to be protected (i.e., the
41
-------
3.2.2 Human Studies
Information on adverse health effects in human
populations is obtained from four major sources: 1)
summaries of self-reported symptoms in exposed
persons; 2) case reports prepared by medical
personnel; 3) correlation studies (in which differences
in disease rates in human populations are associated
with differences in environmental conditions); and 4)
epidemiological studies. The first three types of
studies can be characterized as descriptive epidemiol-
ogy. Epidemiological studies compare the health
status of a group of persons who have been exposed to
a suspected agent with that of a comparable
nonexposed group. Although they cannot identify a
cause-and-effect relationship, they can draw
attention to previously unsuspected problems and
can generate hypotheses that can be further tested.
Most epidemiological studies are either case-control
studies or cohort studies. Case-control studies
identify a group of individuals with a specific disease
and attempt to ascertain commonalities in exposures
the group may have experienced in the past. The
carcinogenic properties of DES, a drug once used to
prevent miscarriages, were discovered through such
studies. Cohort studies examine the health status of
individuals known to have had a common exposure to
determine whether any specific condition or cause of
death is revealed to be excessive compared to an
appropriately matched control population. Benzene
leukemogenesis was established with studies of this
type. Generally, epidemiologists have turned to
occupational settings or to patients treated with
certain drugs to conduct their studies.
Convincing results from epidemiological investiga-
tions can be enormously beneficial because the data
provide information about humans under actual
conditions of exposure to a specific agent. Therefore,
results from well-designed, properly controlled
studies are usually given more weight than results
from animal studies. Although no study can provide
complete assurance that a chemical is harmless,
negative data from epidemiological studies of
sufficient size can assist in establishing the
maximum level of risk due to exposure to the agent.
Interpreting epidemiological results, however, can be
quite difficult. These points should be remembered:
• Appropriately matched control groups are difficult
to identify, because the factors that lead to the
exposure of the study group (e.g., occupa-tion or
residence) are often associated with other factors
that affect health status (e.g., lifestyle and
socioeconomic status).
• Controlling for related risk factors (i.e., cigarette
smoking) that have strong effects on health is
difficult.
• Few types of health effects other than death are
recorded systematically in human populations,
and even the information on cause of death is of
limited reliability. For example, infertility,
miscarriages, and mental illness are not as a rule
systematically recorded by public health agencies.
• Accurate data on the degree of exposure to
potentially hazardous substances are rarely
available, especially when exposures have taken
place in the past. Establishing dose- response
relationships is thus frequently impossible.
• When investigating diseases that take many years
to develop, such as cancer, epidemiologists must
wait many years to ascertain the absence of an
effect. Of course, exposure to suspect agents could
continue during these extended periods of time and
thereby further increase risk.
• The statistical detection power of epidemiological
studies depends on the use of very large
populations.
For these reasons, interpretations of epidemiological
studies are sometimes subject to extreme
uncertainties. Independent confirmatory evidence is
usually necessary, such as supporting results from a
second epidemiological study or supporting data from
experimental studies in animals.
Negative findings in epidemiological studies must
also be interpreted with caution. For example,
suppose a drinking water contaminant causes cancer
in one out of every 100 people exposed to 10 units.
The average time required for cancer to develop from
10 units of exposure is 30 years (not uncommon for a
carcinogen). After people have been exposed to the
drinking water contaminant for 15 years, an
epidemiologist decides to study its effects. He locates
the death certificates of 20 people exposed to the
contaminant, but finds little information on their
actual exposure. Some were exposed when the
contaminant first entered the water supply, others
several years later. The health records, which are
incomplete, reveal no excess cancer in the 20 people
when compared to an appropriate control group. Is it
then correct to conclude that the carcinogen is not
carcinogenic?
3.2.3 Chemical Interactions
The foregoing discussion of hazard evaluation was
predicated on exposure to a single toxic agent.
Humans are rarely exposed to only one substance,
however: commercial chemicals contain impurities;
chemicals are used in combinations; and lifestyle
choices (e.g., smoking, drinking) may increase
exposure to mixtures of chemicals.
40
-------
3.2.1.6 Categorizing Toxic Effects
Toxicity tests may reveal that a substance produces
either a wide or narrow variety of adverse effects on
different organs or systems of the body. Some effects
may occur only at the higher doses used, while only
the most sensitive indicators of a substance's toxicity
may occur at the lower doses.
The toxic characteristics of a substance are usually
categorized according to the organs or systems that
they affect (e.g., liver, kidney, nervous system) or the
diseases they cause (e.g., cancer, birth defects). (See
Chapter 4 for descriptions of toxic effects on various
organs and systems.)
Although uncertainties are associated with most
evaluations of animal toxicity data, special problems
arise with interpreting carcinogenicity data. These
problems are the source of much controversy, as
described in the rest of this section.
One area of uncertainty and controversy concerns
the occurrence of certain types of tumors in control
animals. In most animal experiments, control
animals also develop tumors, and interpreting the
results of such experiments requires comparing the
incidence of tumors in control animals with that
observed in treated animals. This comparison is not
always straightforward. For example, the lifetime
incidence of lung tumors in a certain strain of male
mice, untreated with any substance, may vary from a
low of about 2 percent to a high of about 40 percent;
the average rate is about 14 percent. Suppose that
these male mice treated with a substance exhibited a
35 percent incidence of lung tumors, and control
animals exhibited an incidence of 8 percent. Because
the initial analysis of these data showed that the
treated animals experienced a statistically
significant increase in tumor incidence, the
substance producing this effect was labeled a lung
carcinogen.
However, further analysis of the data took the
investigators to a different conclusion. The 35 percent
incidence observed in the exposed animals was
within the range of tumor incidence that is normally
experienced by male mice (i.e., from 2 to 40 percent),
and the particular group of male mice used as
controls in this experiment also exhibited an
incidence within the range, although at the low end.
Therefore, use of the simple statistical test of
significance was claimed to have erroneously led to
the labeling of the substance as a carcinogen.
Another major area of uncertainty lies in the
interpretation of experimental observations of benign
tumors. Some types of tumors are clearly malignant;
that is, they are groups of cells that grow in
uncontrolled ways, invade other tissues, and are
frequently fatal. No significant controversy
surrounds such tumors, and pathologists generally
agree that their presence is a clear sign that a
carcinogenic process has occurred. Other tumors are
benign at the time they are observed by pathologists,
and whether they should be considered indicators of
a carcinogenic process is not always clear. Some
tumors will remain benign for the lifetime of the
animal, but others will progress to malignancy.
Generally, when establishing the total tumor
incidence, scientists combine the number of animals
with benign tumors that are thought to be part of the
carcinogenic process with the number with
malignancies. Many pathologists disagree with this
approach. The issue has been especially
controversial in connection with tumors found in
rodent livers.
3.2.1.7 Using Short-Term Tests for
Carcinogens
The lifetime animal study is the primary method
used for detecting the carcinogenic properties of a
substance. In recent years, however, short-term
experimental techniques have become available.
Short-term tests for carcinogenicity measure effects
that empirically or theoretically appear to be
correlated with carcinogenic activity. These tests
include assays for gene mutations in bacteria, yeast,
fungi, insects, and mammalian cells; mammalian cell
transformation assays; assays for DNA damage and
repair; and in vitro assays (outside the animal) and in
vivo assays (within the animal) for chromosomal
mutations in animal cells. In addition to these rapid
tests, several tests of intermediate duration involving
whole animals have been used. These include the
induction of skin and lung tumors in female mice,
breast cancer in certain species of female rats, and
anatomical changes in the livers of rodents.
Other tests are used to determine whether a
substance will interact with the genetic apparatus of
the cell, as some well-known carcinogens apparently
do. However, not all substances that interact with
DNA have been found to be carcinogenic in animal
systems. Furthermore, not all animal carcinogens
interact directly with genetic material.
These short-term tests are playing an increasingly
important role in helping to identify suspected
carcinogens. They provide useful information in a
relatively short period, and may become critical
screening tools, particularly for selecting chemicals
for long- term animal tests. They may also assist in
understanding the biological processes that underlie
the production of tumors. However, they have not
been definitely correlated with results in animal
models. Regulatory agencies and other public health
institutions do not consider positive or negative
results in these test, as definitive indicators of
carcinogenicity or the lack thereof, but only as
ancillary evidence.
39
-------
Table 3-5. Typical Costs of Descriptive Toxicity Tests
(1988) (5)
Test Cost, $
Acute oral toxicity 2,000
Acute dermal toxicity 2,800
Acute inhalation toxicity 3,300
Acute dermal irritation 700
Acute eye irritation 450
Skin sensitization:
Draize test 6,700
FCAT (Freunds Complete Adjuvant Test) 3,900
Guinea pig maximization test 5,500
Split adjuvant test 3,200
Buehler test 3,500
Open epicutaneous test 3,200
Mauer optimization test 3,850
Repeated dose toxicity (oral gavage):
14-day exposure 10,200
28-day exposure 12,800
Genetic toxicity tests:
Reverse mutation assay (S. typhimurium) 1,000
Mammalian bone marrow cytogenics (in vivo) 13,000
Micronucleus test 2,000
Dominant lethal test 8,500
Host mediated assay 4,400
Drosophila 12,500
Subchronic mouse study (90 days) 65-75,000
Rat oncogenicity 1,000,000
Mouse oncogenicity 1,000,000
Reproduction 200,000
Teratology (2 species 45,000
Acute toxicity in fish (LC50) 1,250
Daphnia reproduction study 1,400
Algae growth inhibition 1,450
NOTE: The number of animals used for various types of tests
varies, as does the duration of the tests. See Table 3-4
for details.
Table 3-6. Typical Observations for a Subchronic or
Chronic Toxicity Test (6)
Mortality
Body weight changes
Diet consumption
Urinalysis: color, specific gravity, pH, albumin, sugar, leukocytes,
erythrocytes, epithelial cells, case, bacteria, crystals
Hematology: RBC, WBC, platelets, differential
Clinical chemistry: glucose, creatinine, BUN, uric acid, sodium,
potassium, CO2, chloride, calcium, phosphorus, cholesterol,
triglycerides, bilirubin, SCOT, SPGT, lactate dehydrogenase,
alkaline phosphate, iron, total protein, albumin, globulin.
Gross and microscopic examination: brain, heart, liver, kidney,
spleen, testes, thyroid, adrenal (and weigh the eight
aforementioned organs), aorta, bone, bone marrow smears, gall
bladder, esophagus, duodenum, jejunum, cecum, colon, lung,
lymph node, sciatic nerve, parathyroid, pituitary, salivary gland,
epididymis, prostate.
Program) use a definition of MTD that does not take
biological mechanisms into account.
3.2.1.5 Conducting and Interpreting Toxicity
Tests
To ensure the utility of results of toxicity tests, the
following questions must be asked (1):
1. Was the experimental design adequate to test
the hypothesis under examination?
2. Was the general conduct of the test in
compliance with standards of good laboratory
practice?
3. Was the dose of test compound correctly
determined by chemical analysis?
4. Was the test compound adequately
characterized with regard to the nature and
extent of impurities?
5. Did the animals actually receive the test
compound?
6. Were animals that died during the test
adequately examined?
7. How carefully were test animals observed
during the conduct of the test?
8. What tests were performed on the animals (i.e.,
blood tests, clinical chemistry tests) and were
they adequately performed?
9. If the animals were examined
histopathologically (i.e., detailed pathological
examination based on sections taken from
individual tissues), was the examination
performed by a qualified pathologist?
10. Was the extent of animal and animal tissue
examination adequate?
11. Were the various sets of clinical and pathology
data properly tabulated (i.e., tumors grouped in
accordance with NTP guidelines)?
12. Were the appropriate statistical tests used and
were they adequately performed?
13. Was the report of the test sufficiently detailed so
that these questions can be answered?
A proper evaluation would ensure that these types of
questions were examined and would include a list of
qualifications on test results in areas where answers
were missing or unsatisfactory.
38
-------
Table 3-4. (continued)
Fertility and Teratogenic
Reproduc ive (phase n)
(Phase 1) v
• Rats usually used
• Typical protocol:
- 2 or 3 doses
that produce no
maternal toxicity
- Male is given
the chemical 60-
80 days prior to
mating and
female at 14
days prior
- 25 rats per dose
• Typical
observations:
- Percent
pregnant
- Number of stilt-
born
- Weight, growth,
survival, and
general
condition of
offspring during
first 3 weeks of
life
• Rats (25 per
dose) and rabbits
(20 per dose)
used
• Typical protocol:
- Exposed during
organogenesis
(days 6-1 5 in
rats); equivalent
to human first
trimester
- Fetuses
removed by
cesarean
section 2 or 3
days before
normal delivery
• Typical
observations:
- Number of live,
dead, and re-
sorbed fetuses
- Fetuses
weighed,
measured and
examined
grossly
- Histological and
skeletal
examination
Perinatal and
Postnatal
(Phase III)
• Administer
chemical in rats
from day 15 of
gestation
throughout
delivery and
lactation
• Observe
birthweight,
survival, and
growth of offspring
during first 3
weeks of life
Multi-Generation
Reproductive
• Rats used (25
female and 25
male)
• Typical protocol:
- First generation
(FO) exposed
from 40 days of
age until
breeding at day
140; F1 thus
exposed in utero
and during
breeding and
development of
F2
- 3 dose levels
• Typical
observations:
- Gross necropsy
and
histopathology
- Number of
pregnancies,
stillborn,
livebirths, and
other
reproductive
indices
Mutagenic
• Study of ability to
change genetic
material in
nucleus of cell
• Tests used
include:
- Cytogenic
analysis of bone
marrow
- Dominant lethal
test in rodents;
exposed male
mated with
untreated female
- Salmonella
reverse mutation
(Ames) with
metabolic
activation
In an experiment of this size, assuming none of the
control animals develop tumors, the lowest incidence
of cancer that is detectable with statistical reliability
is in the range of 5 percent, or 3 out of 60 animals
developing tumors. If control animals develop tumors
(as they frequently do), the lowest range of cancer
incidence detectability is even higher. A cancer
incidence of 5 percent is very high, yet ordinary
experimental studies are not capable of detecting
lower rates and most are even less sensitive.
Advocates of using the MTD argue that inclusion of
high doses will compensate for the weak detection
power of these experiments. By using the MTD, the
toxicologist hopes to elicit any important toxic effects
of a substance and ensure that even weak carcino-
genic effects of the chemical will be detected. Critics
of the MTD do not reject the notion that animal
experiments may be statistically insensitive, but
rather are concerned about the biological
implications of such high doses. Their concerns can
be summarized as follows:
• The underlying biological mechanisms that lead to
the production of cancer may change as the dose of
the carcinogen changes.
• Current methods for estimating an MTD for use in
an experiment do not usually take such biological
mechanisms into account.
• The biological mechanisms at work under
conditions of actual human exposure may be quite
different from those at work at or near the MTD.
In general, observations at or near an MTD (as
determined by current methods) thus may not be
qualitatively relevant to conditions of actual human
exposure.
Many risk assessors agree that greater attention
should be .paid to developing data on the underlying
mechanisms of carcinogenicity and their relation to
dose. Also, a range of doses should be included in
carcinogenicity testing to assess whether
physiological mechanisms that would normally
detoxify the chemical are overwhelmed at an MTD.
These biological considerations have considerable
merit, but are frequently disregarded in designing
studies and interpreting data. Although some risk
assessors have attempted to develop a more
biologically relevant definition of MTD, most current
tests (those carried out by the National Toxicology
37
-------
Table 5-8. Carbon Usage Rates for Several Organics
Type
Usage Rate
Ib/MG
Volatile Organics
TCE
PCE
Vinyl chloride
Cis-1,2-dichloroethylene
Aldicarb
Chlordane
DBCP
Pesticides
200
70
NA
250
25
5
15
Chlorinated Aromatics
PCB
Dichlorobenzene
5
10
To obtain carbon usage rates, tests can be performed
in the laboratory or in the field. Laboratory tests
include isotherm and minicolumn tests. The isotherm
test employs the following equation, called the
Freundlich isotherm relationship:
x/m = k c1/n
(5-1)
where
x/m
= equilibrium capacity (mg contaminant/g
carbon)
= capacity at 1 mg/L contaminant
concentration
= contaminant effluent concentration
(mg/L
1/n = exponent
This equation can be plotted to obtain "isotherms"
that provide an indication of which compounds are
more adsorbable than others. Isotherms also provide
a rough estimate of the carbon usage rate. Figure 5-
13 provides isotherms for several compounds. The
compounds with a higher molecular weight, such as
pentachlorophenol, are more adsorbable and less
likely to break through; i.e., appear in the effluent-
treated water. The isotherm test also provides
information for gauging competitive effects of other
contaminants, comparative performance of different
carbons, and the effects of changing temperature and
pH. Isotherm tests can be performed relatively
quickly and cost approximately $1,000 - $3,000 per
test. Further, published isotherm data for many
contaminants are available (7).
Figure 5-13. Adsorption isotherms for several organic
compounds found in ground-water supplies.
100.0
s
o
10.0
|> 1.0
X
0.1
0.0001 0.001 0.01 0.1
Residual Concentration,
1.0
Numbers in parentheses indicate the molecular weight of the
compound.
In contrast, "dynamic column" field tests can cost
tens of thousands of dollars and can take 6-10 months
before breakthrough is achieved. These tests allow
many variables to be changed and provide accurate
information on carbon depth, loading rate, and usage
rate.
Between the isotherm test and dynamic column tests,
in terms of cost and complexity, is the "minicolumn"
laboratory test. This test uses a device that simulates
a GAC system on a small scale, incorporating raw
water storage; stainless steel and teflon tubing and
fittings to avoid contamination of the sample; and a
small tube of carbon through which a contaminated
sample is passed. This system is comparable to the
isotherm test in cost and duration (not including
equipment costs) but provides more information.
Further, the carbon usage rates obtained with the
minicolumn test are relatively accurate (more
accurate than isotherm tests, but less accurate than
field studies).
The contaminant concentrations and type of
contaminants both affect the carbon life. Figure 5-1.4
illustrates the relationship between carbon life and
influent concentration at three different desired
effluent concentrations for the contaminant
trichloroethylene. Note that, to bring an influent
concentration of 200 ug/L down to 1 ug/L, the carbon
life would be about 3 months; while at a higher
influent concentration, such as 500 ug/L, the carbon
life drops to 2 months.
Figure 5-15 illustrates the relationship between
influent concentration and carbon life at an effluent
concentration of 10 ug /L. Note that for the three
contaminants shown, the carbon life at an influent
concentration of 200 ug/L would be about 1 month for
74
-------
Figure 5-12. The granular activated carbon adsorption
process.
Water
Adsorbate
Attached
Adsorbate
Adsorbent
Table 5-7. Readily and Poorly Adsorbed Organics
Readily Adsorbed Organics
•Aromatic solvents (benzene, toluene, nitrobenzenes)
•Chlorinated aromatfcs (RGBs, chlorobenzenes, chloronapthalene)
•Phenol and chlorophenols
•Polynuclear aromatics (acenapthene, benzopyrenes
•Pesticides and herdtoides (DDT, aldrin, chlodane, heptachlor)
•Chlorinated nonaromatics (carbon tetrachloride, chloroalkyl ethers)
•High MW hydrocarbons (dyes, gasoline, amines, humics)
Poorly Adsorbed Organics
•Alcohols
•Low MW ketones, acids, and aldyhydes
•Sugars and starches
•Very High MW or colloidal organics
•Low MW aliphatics
increases adsorbability by increasing the overall
driving force for adsorption in that solution.
Fortunately, most of the common organic drinking
water contaminants are easily adsorbed by GAG (i.e.,
PCBs, benzene, toluene, carbon tetrachloride,
trichloroethylene) and those that are less easily
adsorbed (i.e., aldehydes and ketones) are less
commonly found in drinking water. When ozone
treatment is applied ahead of GAG, the ozone can
break down the more easily adsorbed compounds into
less easily adsorbed compounds.
The solution itself is also a factor to consider. The pH
of the solution can have an impact if organic acids
and bases are being removed; however, many of the
organic drinking water contaminants are neutral
compounds. Increased solution temperature can
reduce adsorption somewhat. Probably the biggest
impact of the solution is the presence of competing
contaminants. For example, naturally occurring
organics (represented by total organic carbon) can
compete for carbon sites with the contaminants. This
competition reduces the total capacity of carbon
available for the contaminant.
There are two forms of activated carbon: granular
and powdered activated carbon. The granular form is
used in a filtering mode in that water is passed
through a closed vessel in which contaminants are
adsorbed onto the carbon. Powdered activated carbon
(PAG) is simply pulverized granular activated carbon
that is added to water along with a coagulant. The
organic contaminants are adsorbed onto the PAC and
the carbon then settles out in a sedimentation basin
or clarifier.
5.3.1.2 Process Design Considerations
Important process design considerations include the
contaminants present, the EBCT, the carbon usage
rate, surface loading rate, and carbon depth.
The EBCT is the primary factor in determining the
size and capital cost of a system. The EBCT can range
from 5-30 minutes, although studies have shown that
times less than about 7.5 minutes usually do not
obtain good use of the carbon. Generally, EBCTs of
10-15 minutes are used for the organics typically
found in drinking water. In the case of removing
radon, however, an EBCT of 180-200 minutes must
be used to achieve removals of 95% or greater. While
such a long EBCT might be feasible for point-of-use
applications, it would drive up the cost of GAG
treatment for larger applications.
The carbon usage rate determines the operating cost.
Generally, 100-300 Ib of carbon is used per million
gallons of water treated for the types of organics
found in ground-water supplies. However, some of
the synthetic organic chemicals now under
consideration for EPA regulation have lower usage
rates of about 50 Ib per million gallons. Table 5-8
shows usage rates for three categories of
contaminants: VOCs, pesticides, and chlorinated
aromatics. These usage rates translate into
replacement intervals, at normal demand, for VOCs
of about 3 to 6 months, and for the other two
categories, 1 to 2 years.
73
-------
Figure 5-11. Activated alumina system: operating mode flow schematics.
Raw Water
Raw Water
Treatment and
Downflow Rinse
Waste
Backwash and
Upflow Rinse
Raw Water
Raw Water
Caustic
Waste
Upflow
Regeneration
Caustic
Waste
Downflow
Regeneration
Table 5-6. Removals Possible with Activated Alumina
> 90 Percent Removal > 70 Percent Removal
As(V)
F
Se (IV)
Se (VI)
area due to its high porosity. In fact, 1 gram of GAG
has about the surface area of a football field. Also
important is the reversibility of the adsorption force.
Since this force is mostly physical and partly
electrical in origin, it can be easily reversed (unlike
chemical reactions), thus allowing the contaminants
to be driven off the carbon to allow re-use. (It is
important to remember that regular charcoal is not
activated carbon and does not have activated carbon's
large internal surface area.)
Several factors affect the degree of success possible
with the adsorption process. First, the nature of the
contaminant itself is important. Compounds with a
branch-chained molecular structure are generally
more adsorbable than straight-chained compounds.
Compounds with higher molecular weights are more
adsorbable than those with low molecular weights.
Less polar compounds like PCE are generally better
adsorbed than more polar compounds like alcohols
and ketones. Table 5-7 lists both readily adsorbed and
poorly adsorbed organic compounds. A higher
concentration of the contaminant in the solution also
72
-------
presaturated with hydroxyl ions through a strong
sodium hydroxide solution. In the activated alumina
process, the anions of fluoride, arsenic, and selenium
are exchanged for hydroxyl ions on the surface of the
alumina. When the alumina is completely saturated
with contaminant ions, it must be regenerated with
the sodium hydroxide solution. Because the
regeneration process increases the pH of the water
within the unit, it must be neutralized with a 3-
percent sulfuric acid solution. Figure 5-11 illustrates
an activated alumina treatment system in four basic
operating modes.
Like the standard ion exchange technique, activated
alumina requires pretreatment to remove suspended
solids, although a high concentration of total
dissolved solids (and even some suspended solids) can
be tolerated. Competing ions are also an important
process design consideration.
This process is most effective for removal of arsenic,
fluoride, and selenium. It is not effective for removal
of barium, radium, or cadmium. Table 5-6
summarizes the removal percentages possible with
activated alumina.
Advantages of the activated alumina process include
relative insensitivity to flow and thus on-demand
operation; high tolerance for total dissolved solids;
and effective removal of arsenic, selenium, and
fluoride.
A significant disadvantage is the necessity of using
caustic acid and base solutions, which present a
hazard to operators. Also, sodium hydroxide must be
liquefied in heating units, even in warm climates.
The activated alumina process is slower than other
ion exchange processes; activated alumina has an
EBCT of about 5 minutes compared to about 2-3
minutes for other ion exchange processes. Further
adding to the cost is the fact that the sodium
hydroxide solution dissolves the activated alumina
medium; a typical unit might lose 20 percent of the
medium per year. Finally, the wastestream from the
regenerant will be high in contaminants and
aluminum, and thus must be disposed of properly.
5.2.4.7 Case Study: Activated Alumina
This case study describes the use of activated
alumina for fluoride removal from a ground-water
supply in Gila Bend, AZ (4,5,6). Raw water fluoride
levels ranged from 4 to 6 mg/L. The plant had a
capacity of 600 gpm (900 gpm max.) and included
activated alumina, caustic regeneration, acid
neutralization, and an evaporation pond for
regenerant waste treatment.
Fluoride levels achieved in the treated water
averaged approximately 1 mg/L.
5.2.5 Treatment Technologies for
Radionuclides
Treatment techniques usually used for inorganics,
and some techniques used for organics, can be used to
treat radionuclides (see Sections 5.2 or 5.3 for more
information).
Methods shown in the literature to be effective for
treatment of radium in drinking water are lime
softening, reverse osmosis, and cation exchange.
Methods shown to be effective for uranium removal
are lime softening at high pH, reverse osmosis, and
anion exchange. The two available methods for
removing radon from drinking water are granular
activated carbon and aeration. For manmade
radionuclides, ion exchange can be used.
EPA has developed preliminary cost estimates for
technologies that may feasibly remove radionuclides
from drinking water. The following estimates are in
1989 dollars. Depending on the amount of water
treated, estimated costs range from $0.30 to $0.80 per
1,000 gallons for cation ion exchange; $0.30 to $1.10
per 1,000 gallons for iron and manganese treatment;
and $1.60 to $3.20 per 1,000 gallons for reverse
osmosis. Preliminary cost estimates for aeration
techniques range from $0.10 to $0.75 per 1,000
gallons for systems serving about 100,000 people and
100-500 people, respectively.
Preliminary cost estimates for removing radon from
household drinking water systems with point-of-
entry treatment devices are a capital cost of $400 to
$800 for granular activated carbon and about $900
for aeration, with annual operating costs estimated
at $20 per year for activated carbon and $80 per year
for aeration.
5.3 Organics Treatment
The organics treatment techniques considered here
include granular activated carbon (GAG) and
aeration.
5.3.1 Granular Activated Carbon
5.3.1.1 The Treatment Process
GAG works on the principle of adsorption. Adsorption
is the transfer of a dissolved contaminant (adsorbate)
from a solvent (solution) to the surface of the
adsorbent (granular activated carbon). Thus, the
contaminant is not changed but instead is held on the
surface of the carbon.
The GAG itself is a very porous type of carbon with
the property of adsorbing dissolved contaminants
onto its surface (see Figure 5-12). Key to its success as
an adsorbent is its extremely large internal surface
71
-------
Figure 5-10. Ion exchange treatment system.
Salt Loading
Waste to
Disposal
Distribution
System
Well Supply
• Conductivity
Monitor
Table 5-5. Removals Possible with Ion Exchange
Cations >90 Percent Removal Anions >90 Percent Removal
Ba
Cd
Cr (III)
Ag
As(V)
Cr (VI)
N03
Se (IV)
Se (VI)
adsorbed ions like nitrate. Regeneration must be
performed frequently, sometimes as often as once per
day, creating a potential problem with disposal of the
waste regenerant. Ion exchange is not feasible at
high total dissolved solid concentrations (> 500-
1,000 mg/L) because the ion exchange medium will
foul, and the removal process is most effective at
lower ionic strengths.
5.2.3.1 Case Study: Ion Exchange
This case study describes the use of ion exchange to
remove nitrate from a ground-water supply in
McFarland, CA (3). Four wells were affected by
nitrate contamination from agricultural runoff (i.e.,
fertilizers, manure), with raw water levels of 6.8 to
22.1 mg/L as N. Ground water was the sole source of
drinking water available.
Ion exchange was chosen because it could be applied
at the wellhead with a minimum of operator atten-
tion. Plant capacity was 1 MGD. Treatment processes
included anion exchange, sodium chloride regener-
ation with slow rinse and resin declassification, and
aerated lagoons and spray irrigation for brine waste
treatment. The treated water was blended with raw
water at a ratio of 70 percent treated water to 30
percent raw water..An empty bed contact time
(EBCT) of 2.5 minutes was used.
The treated water had nitrate levels ranging from 2
to 5 mg/L and the finished (blended) water had
nitrate levels ranging from approximately 6 to 8
mg/L. About 2,000 Ib of salt were used per day during
continuous operation.
5.2.4 Activated Alumina
The activated alumina process is similar to the ion
exchange technique described above. Activated
alumina is a commercially available product that
acts as an ion exchange medium for selected
contaminants. This exchange medium is
70
-------
Figures 5-9. Reverse osmosis treatment system.
-i
H
'H
H
Product to
Degasifier and Reservoir
P0" x^~N
Feed / \ _,
o Wn
mrrt i, ^ -rftrr1
-
H
Permeators
$4 6
gh Service * Percolation
Pumps Pond
r-
h,
H
L^*-, 0 h
Control Panel Lc— >i
~"r) : and Pumps ~\)
1
L
1 <
i
cCil O Filter x-W !
Feed x_,x /^ \ J •
— ] /*S Mixing I /-
r-J VJ Tank v — /
1 Cleaning
QO Percent Removal >70 Percent Removal
As(V)
Ba
Cd
Cr (III)
Or (VI)
F
Pb
Se (IV)
So (VI)
Ag
As (III)
Hg
No3
Figure 5-10 illustrates an ion exchange treatment
system.
Because the ion exchange medium has small internal
pore spaces that can become clogged, suspended
solids must be removed through prefiltration.
Competing ions are also important; as in reverse
osmosis, competing ions such as in very hard water
can occupy sites on the exchange medium needed for
contaminant ions.
The resin exchange capacity is usually available as a
number in milligrams per unit volume of resin. This
resin capacity, along with the empty bed contact
time, can be used to calculate the resin "break-
through" times (i.e., the time when contaminants
begin to be detected in the treated water). Table 5-5
summarizes ion exchange's removal percentages for
selected anions and cations. The process is effective
for barium, radium, and other cations (for cationic
resins) and nitrate and selenium (for anionic resins).
While sodium chloride is an inexpensive and
convenient presaturant, its use adds sodium to the
treated water, possibly making the water unsuitable
for those on a salt-restricted diet. In response to this
problem, calciform resins (calcium chloride) are
currently being developed that do not add sodium to
the water.
The advantages of ion exchange are similar to those
of reverse osmosis, in that it is also relatively
insensitive to flow and thus can be operated on
demand. A large variety of resins are available for a
variety of applications. For example, resins with
higher affinities for specific contaminants (i.e.,
radium and nitrate) can be purchased. Ion exchange
can very effectively remove selected contaminants,
with near zero contaminant levels possible in the
effluent-treated water.
One disadvantage of ion exchange is a phenomenon
called effluent peaking: When competition happens
among ions, spikes in the contaminant levels occur as
the competing ions occupy sites on the exchange
medium. Peaking is especially a problem with poorly
69
-------
Figure 5-8. Two types of reverse osmosis membranes.
Feed
(Raw Water)
Filtrate
(Treated Water)
Courtesy of Millipore Corporation
Process Flow Through Spiral-Wound Reverse Osmosis Unit
Casing -^ ,_ Hollow Fiber
Treated
Water
1
i— MOIIOW t-ioer
., v.^?
»
Feed
yf—Z- (Raw Water)
Hollow-Fiber Reverse Osmosis Unit
5.2.3 Ion Exchange
Ion exchange refers to the process of substituting one
ion for another on the charged surface of a resin. An
exchange medium, often plastic resin, is charged and
presaturated most commonly with sodium chloride.
When water containing contaminant ions contacts
the exchange medium, the contaminant ions, which
are preferred by the medium, replace the sodium or
chloride, depending on whether cation or anion
exchange is desired. After the exchange medium is
completely saturated with contaminant ions, the
medium must be regenerated with an appropriate
brine solution to resaturate it with the desired ion.
The resulting waste or regenerant is heavily
contaminated and must be disposed of properly.
68
-------
become concentrated enough that the water left
behind can be considered a brine.
Figure 5-8 illustrates two types of reverse osmosis
units. In the spiral-wound membrane unit, sheets of
membrane material cover each side of a porous
water-conducting backing cloth. This arrangement
forms an envelope closed on all sides except for one
that communicates with a perforated center tube.
This membrane, along with a mesh spacer, is
wrapped into a tight spiral so that feed water can
contact the entire surface of the membrane. Water is
pumped into the mesh space outside the membranes
under pressure. Since only water can pass through
the membranes and collect in the center tube, the
water is demineralized and purified.
In the hollow-fiber unit, the semipermeable
membrane takes the form of many hair-like hollow
tubes. Contaminated water pumped under pressure
into the spaces between these tubes allows water to
pass into the inside of the tubes, leaving the
contaminant behind.
An advantage of the spiral-wound unit is that it is
less likely to clog when treating water high in
suspended solids. In fact, reverse osmosis usually
requires pretreatment (described more fully below).
The hollow fiber unit has the advantage of having a
much higher membrane area per unit of space
occupied by the device. The hollow fiber unit provides
about 1,000 ft2 of membrane per ft3 of membrane
module, while the spiral wound unit provides only
100 ft2 of membrane per ft3 of membrane module.
Important process design considerations include the
influent suspended solids concentration, the ionic
size of the contaminants present, and the membrane
type. Membranes are available in specific pore sizes
for specific contaminants, and with specific chemical
resistances such as to chlorine.
Because the membranes are expensive and subject to
fouling, various pretreatment processes are usually
necessary. Figure 5-9 shows a reverse osmosis
treatment system incorporating pretreatment.
Typical pretreatment processes include pH
adjustment to protect the membrane, prefiltration to
remove particulates capable of fouling the
membrane, and sequestration of hardness ions to
prevent membrane clogging. (Sequestering agents
keep high concentrations of minerals like calcium
and magnesium in solution so that they won't
precipitate out and clog the membrane)
Table 5-4 summarizes the removal percentages
possible for reverse osmosis treatment of inorganic.
In addition to inorganic, reverse osmosis also
removes trihalomethane precursors (humic and
fulvic acids), nitrates, pesticides, and microbiological
contaminants (viruses, bacteria, protozoa).
Reverse osmosis might be best suited to waters
containing high levels of inorganic chemicals,
organic chemicals, and total dissolved solids. Because
it is relatively insensitive to total dissolved solids, it
has been used successfully in the Gulf Coast area,
where total dissolved solids concentrations can be as
high as 2,600 mg/L (whereas the ion exchange
process, for example, cannot tolerate such concen-
trations). This process is also relatively insensitive to
flow; the device can be turned on and will begin
working almost immediately. Reverse osmosis very
effectively removes contaminants, with almost zero
levels possible in the effluent-treated water.
Especially effective removal is possible if a multi-
stage system is used in which the water is pumped
through additional reverse osmosis units arranged in
series.
A major disadvantage of reverse osmosis is its high
operating and capital cost; for a small plant (< 1
MGD), the operating cost can be a relatively high
$3-6 per 1,000 gallons of treated water. Pretreatment
(pH adjustment, prefiltration, sequestration of
hardness ions) also adds to the cost. In addition to the
fouling problems already discussed, membranes can
also foul due to biological growths within the unit
during periods of disuse, such as during the off-
season in a seasonal system like that used for a
trailer park. Another significant problem with a
reverse osmosis system is that the reject stream can
be a high percentage (20-90 percent) of the feed flow.
Finally, this reject water or brine, now high in
contaminants, must be disposed of properly. When
the brine contains high concentrations of
radionuclides such as radium-226, disposal can be a
problem.
5.2.2.1 Case Study: Reverse Osmosis
This case study describes the use of reverse osmosis to
remove radium-226 from several ground-water sup-
plies in Sarasota County, FL (2). The article describes
existing facilities rather than newly constructed pilot
plants. Raw water radium-226 levels ranged from 3.4
to 20.2 pCi/L. Plant capacities ranged from a low of
800 gpd to a high of 1 MGD. Both hollow-fiber and
spiral-wound units were used. Treatment processes
included pretreatment (cartridge filtration, pH
adjustment, ion sequestration), reverse osmosis, and
post-treatment (pH adjustment, degasification,
chlorination).
Of eight systems examined in detail, all achieved
reductions to levels below the regulatory limit of 5
pCi/L, with product water concentrations ranging
from 0.14 to 2.0 pCi/L. Reject water concentrations,
however, ranged from 7.8 to 37.8 pCi/L.
67
-------
Table 5-2.
Removal Possible with
Conventional Processes
Table 5-3. Removals Possible with Lime Softening
> 90 Percent Removal > 70 Percent Removal
Iron Coagulation
Alum Coagulation
90 Percent Removal
As (V) Cr (III)
Cd(pH = 8) Pb
Cr (III) (pH = 10.5)
Cr (VI) [using Fe (II) ]
70 Percent Removal
Ag(pH = 8.0) Hg
As (V) (pH = 7.5) Cd (pH = 8.5)
Figure 5-7 shows a treatment system incorporating
lime softening. Again, it is in the sedimentation
phase that removal of the inorganic contaminants
occurs. The recarbonation phase in the illustration
involves the addition of CO2 to the softened water to
lower the pH.
Figure 5-7. Lime softening treatment system.
I—Chemicals
Disinfectant
Rapid Mix Flocculation Recarbonation
Sedimentation Filtration
Sometimes, in order to remove both inorganic
contaminants (optimal pH of 10-10.5) and hardness
(optimal pH of 9-10), a dual-stage process must be
used; for example, chemical addition, sedimentation
to remove inorganic, recarbonation to lower the pH,
and then another sedimentation phase to remove
hardness.
Because lime softening to remove inorganic
contaminants requires greater addition of lime than
is required for softening alone, organic polymers may
be used. These polymers are added to increase the
rate of settling of the precipitate. Table 5-3 sum-
marizes the effectiveness of softening for the removal
of inorganic contaminants. In the table, note that for
three important inorganic contaminants (arsenic,
barium, chromium), the optimal pH range of > 10 is
above the optimal pH range for removal of hardness.
For high volumes of water (> 1-5 MGD), the cost of
conventional coagulation and lime softening is
relatively low. Also, if the water is very hard, lime
softening removal percentages will increase because
removal depends on enmeshment within the floe
particles formed in the softening process. Another
advantage is that water utilities have used both of
As(V) (pH> 10.8)
Ba(pH = 9.5-10.8)
Cd
Cr(lll) (pH> 10.5)
Pb
As(lll)(pH> 10.5)
As(V) (pH = 10.0-10.5)
Cr(lll) (pH> 10.5)
Ag
these techniques for many years and are quite
familiar with their use.
A disadvantage of both techniques is that chemical
costs increase when removal of inorganic contami-
nants is desired. For example, to remove'turbidity,
alum dosages of 20-40 mg/L are required; to remove
arsenic, upwards of 100 mg/L may be required. Thus,
chemical costs can be doubled when using these
techniques. Also, these techniques create
considerably more sludge than normal, which is more
difficult to dewater than normal. As mentioned
earlier, lime softening may require a two-stage
process to optimize removal at different pHs. Finally,
these techniques require that chemicals be fed on a
relatively continuous basis. Thus, they may be
inappropriate for small water supply systems in
which the flow of water is often small and
intermittent.
5.2.1.1 Case Study: Conventional Treatment
This case study describes the use of conventional
treatment for barium contamination in ground-water
supplies in northeastern Illinois (1). Small areas had
elevated barium concentrations, ranging from 0.4 to
8.5 mg/L. Treatability tests indicated that optimum
barium removal could be obtained with 75-175 mg of
gypsum (calcium sulfate) per liter of water at apH of
11.0. The pilot plant had a capacity of 1.5 MGD and
incorporated precipitation, direct filtration, and
polymer addition.
Barium reduction from 6 mg/L down to 0.5 mg/L was
achieved (91 percent removal). Chemical dosages of
100 mg/L gypsum and 0.25 mg/L polymer were used.
5.2.2 Reverse Osmosis
Osmosis is a phenomenon by which solutions of
different concentrations are separated by a semi-
permeable membrane; water moves from the less to
the more concentrated solution. In reverse osmosis,
pressure is applied to the more solution and the
direction of flow reverses, with the water moving
from the more concentrated solution to the less
concentrated solution. The semipermeable
membrane is permeable to water but not to dissolved
ions. Thus, the water passes through the membrane
and the contaminants are left behind and eventually
66
-------
Table 5-1. continued
Relative Treatment Cost*
$/l ,000 gal
Contaminant
Mercury
Inorganic
Organic
Nitrate
Treatment Method
Lime softening, above
pH 10.5
Granular activated carbon
Reverse osmosis
Coag^/filt. with powdered
activated carbon
Granular activated carbon
Ion exchange (anion resin)
Reverse osmosis
Removal
Percentage
<90
>95
<85
50-75
>95
>90
<90
0.3 MGD
—
1.52
3.18
2.19
1.52
1.11
3.18
1.0 MGD
0.59
0.59
2.01
0.94
0.59
0.75
2.01
50 MGD
0.41
0.21
1.21
0.37
0.21
0.48
1.21
Nitrite
Selenium
Breakpoint chlorination
Ion exchange {anion
resin)
Reverse osmosis
>90
>90
<90
0.05
1.17
3.18
0.03
0.85
2.01
0.01
0.59
1.21
SelV
(Tetravalent)
Se VI (Hexavalent)
S/7ver
Ferric sulfate coagulation/
filtration, pH 5.5-7
Activated alumina
Reverse osmosis
Lime softening
Activated alumina
Reverse osmosis
Ferric sulfate coagulation/
filtration, pH 7-9
Alum coagulation
filtration, pH 6-8
Lime or excess lime
softening
Reverse osmosis
Direct filtration
<80
>95
75-99
<50
>95
75-99
<80
<80
<85
>90
<60
2.45
0.78
3.68
—
3.91
3.68
1.40
1.40
—
3.18
0.90
0.78
0.43
2.28
0.64
3.05
2.28
0.54
0.54
0.59
2.01
0.33
0.22
0.23
1.34
0.41
2.32
1.34
0.18
0.18
0.41
1.21
0.11
Source:
Technologies and Costs for the Removal of Inorganic from Potable Water Supplies-Draft Reports. U. S.
Environmental Protection Agency, Office of Drinking Water, Washington, DC, 1983-1985.
•Based on constructing new facilities; cost may be lower if existing facilities may be upgraded or optimized.
Figure 5-6. Coagulation/filtration treatment system.
r-(
r
V
Ra
Ihomicals
^
^
^
pid Mix Rocculation Sed
kx
mental
Disin
ion F
fectant
\$is
i
iltration
floe particles formed. Iron salts provide a wider range
of pH effectiveness than aluminum salts (aluminum
sulfate or "alum"). Because coagulation for removal
of inorganic contaminants requires higher doses of .
coagulant than for removal of turbidity alone,
coagulant aids such as organic polymers may be used.
Table 5-2 summarizes the effectiveness of iron and
alum coagulation for removal of inorganic
contaminants. Note that coagulation is primarily
effective for metals. For nitrate, nitrite, radium,
barium, and sulfate, the process is virtually
ineffective.
Lime Softening: Water is softened to remove
"hardness" from ground and surface water, part-
icularly in the Midwest and Southeast. Hardness is
the sum of divalent cations in solution and is due to
calcium and, to a lesser extent, magnesium. Hard
water contains large amounts of dissolved minerals
that cause scaling and excessive soap consumption.
In lime softening, lime is added to the hard water,
causing the mineral ions and lime to precipitate out.
When excess lime softening is used (i.e., higher pH),
precipitation of inorganic contaminants can also
occur. (If the hardness is noncarbonate hardness,
soda ash is used instead of lime.)
65
-------
Table 5-1. Summary
Contaminant
Arsenic
As V (Arsenate)
As HI (Arsenite)
Asbestos
Barium
Cadmium
Chromium
Cr III (Trivalent)
Cr VI (Hexavalent)
Copper
Fluoride
Lead
of Treatment Technologies
Treatment Method
Alum coagulation/
filtration, pH 6-7
Iron coagulation/
filtration, pH 6-8
Excess lime softening
Activated alumina
pH5-6
Ion exchange
Reverse osmosis
Oxidation treatment of As
Ml to As V, then above
treatment methods
Conventional filtration
Direct filtration
Diatomaceous earth
filtration
Ion exchange
Lime softening, pH 1 1
Reverse osmosis
Ion exchange
Excess lime softening
Reverse osmosis
Iron coagulation
filtration, above pH 8
Iron coagulation/
filtration, pH 6-9
Alum coagulation/
filtration, pH 7-9
Excess lime softening
Ion exchange
Ion exchange
Reverse osmosis
Ferrous sulfate
coagulation/filtration,
pH 7-9.5
Ion exchange
Reverse osmosis
Ion exchange
Lime softening
Reverse osmosis
Alum coagulation/filtration
Activated alumina,
pH 5.5
Reverse osmosis
Lime softening
Iron coagulation/
filtration, pH 6-9
Alum coagulation/
filtration pH 6-9
Lime or excess lime
softening
Ion exchange
Reverse osmosis
Direct filtration
for Inorganic R
Removal
Percentage
>90
>90
>90
>95
<90 '
! <90
>95
>95
>95
>90
<95
>95
>90
>90
>90
<80
90-98
90-98
98
<90
>92
>90
<90
>90
<95
>90
>95
>50
>90
>90
<65
>95
>95 •
>97
<95
<95
<60
emoval
Relative Treatment Cost*
$/1 ,000 gal
0.3 MGD
1.75
1.75
3.05
1.22
0.83
3.32
1.41
1.13
1.43
0.80
-
3.18
0.80
-
3.18
1.42
1.46
1.46
_
0.51
3.18
1.46
0.80
3.18
0.80
-
3.18
1.40
0.47
2.06
-
1.75
1.75
2.98
0.92
3.32
1.34
1.0 MGD
0.44
0.44
0.63
0.62
0.51
1.64
0.54
0.40
0.74
0.44
0.63
2.01
0.44
0.59
2.01
0.54
0.55
0.55
0.59
0.29
2.01
0.55
0.52
2.01
0.44
0.59
2.01
0.54
0.27
1.21
0.59
0.44
0.44
0.60
0.36
1.64
0.33
50 MGD
0.19
0.19
0.40
0.51
0.42
1.29
0.19
0.13
0.35
0.22
0.41
1.21
0.22
0.41
1.21
0.18
0.19
0.19
0.41
0.15
1.21
0.19
0.32
1.21
0.22
0.41
1.21
0.18
0.14
0.67
0.4.1
0.19
0.19
0.40
0.23
1.29
0.13
64
-------
Figure 5-5. Interceptor well.
Ground-Water Divide
Possible Location of Disposal
Interceptor Well Area
Bedrock Floor'
month. However, POU treatment is not
recommended for microbiological contamination
because of the acute nature of the diseases caused by
such contamination. Also, as with bottled water,
POU treatment at a single drinking water tap does
not minimize inhalation hazards incurred during
washing and bathing. EPA does not consider systems
using POU to be in compliance with the VOC
regulations.
In conclusion, risk reduction requires an under-
standing of several factors: the overall hydrogeology
of the system, treatment level targets in terms of
feasibility and cost, and the various control strategies
available.
5.2 Inorganics Treatment
Inorganic treatment techniques considered below
include conventional treatment (iron and alum
coagulation), lime softening, reverse osmosis, ion
exchange, and activated alumina. Table 5-1
summarizes treatment technologies for the removal
of inorganic chemicals.
5.2.1 Conventional Treatment and Lime
Softening
Conventional Treatment: Conventional treatment
involves the addition of coagulants, most commonly
aluminum or iron salts, to remove color and turbidity
from surface waters. This coagulation process also
removes inorganic contaminants (if precipitated out
of solution) through adsorption and enmeshment in
clumps of coagulated sediment referred to as floe
particles.
Polyvalent cations and anions such as calcium,
magnesium, carbonate, and phosphate can be readily
removed by direct precipitation. Direct precipitation
is generally accomplished by increasing the pH, as in
lime softening, where the pH is increased using lime
and sometimes soda ash. Trace anions such as
selenium, arsenic, and fluoride are removed by
coprecipitation or sorption onto the surfaces of the
floe particles formed by iron and aluminum salts.
Figure 5-6 illustrates a typical conventional
treatment system that uses coagulation. Removal of
the inorganic contaminants occurs through settling
in the sedimentation basin after flocculation. A rule
of thumb is that coagulation of inorganic chemicals
works better with polyvalent cations and anions
(charge of ± 2 or greater) than with monovalent
cations and anions (charge of +1).
The pH of the water is an important process design
consideration because it affects the form of the target
species (i.e., whether the contaminant is soluble or
insoluble), the form of the coagulant, and the type of
63
-------
Figure 5-3. Drilling a new well.
Ground-Water Divide
Bedrock Floor'
Figure 5-4. Blending existing sources.
Well #1 Distribution System
Well #2
Well #3
interceptor well is then pumped to waste to remove
the contaminant from the aquifer before it can
contaminate the drinking water wells. This
technique is currently being used in several locations
in the U.S. In such locations, without interceptor
wells operating, VOC contamination levels were 50-
100 ug/L. With the interceptor wells operating, the
levels dropped to less than 50 ug/L. The major
disadvantage of this approach is that disposal of the
wastewater from the interceptor well poses problems.
Some recent EPA guidance indicates that the water
from the interceptor well may have to be treated to
drinking water standards prior to discharge, thus
eliminating any cost- savings gained by using this
technique.
5.1.3.2 Short-Term Strategies
Short-term strategies include the use of bottled water
and point-of-use devices. Two concerns with bottled
water are ensuring the quality of the bottled water
itself and distributing the water to the customer.
Some states certify the quality of bottled water; in
areas without state certification, a system for testing
and monitoring the bottled water quality may have to
be developed. The cost of bottled water can be
approximately $50 per household per month.
Finally, the use of bottled water does not mitigate
inhalation hazards such as from the volatilization of
VOCs or radon gas from contaminated water used for
washing and bathing.
Point-of-use (POU) devices treat water at a single tap
with various existing treatment technologies. Many
types are available, based on techniques such as
activated alumina, granular activated carbon
adsorption, reverse osmosis, and ion exchange. In
general, POU devices are suitable for short-term
emergencies during which their ability to quickly
provide uncontaminated water is an advantage. For
example, at a typical Superfund site, it may take six
months to install a centralized treatment system at
the water supply but only a few weeks to install
commercially available POU treatment devices at all
of the affected households. Costs for POU treatment
range from $20-$60 per household (one tap) per
62
-------
Figure 5-2. Example of contaminated ground-water supply.
Tetrachloroethylene
1,2-Dichloroethylene
Trichloroethylene
Chloroform
1,1,1 -Trichloroethane
245
10
5
2
2
Trichloroethylene
Tetrachloroethylene
1,2-Dichloroethylene
Chloroform
1,1,1-Trichloroethane
1,2-Dichloroethylene
Chloroform
Tetrachloroethylene
Trichloroethylene
1,1,1 -Trichloroethane
can be easily identified. The tank could be repaired or
removed and the contaminated well could be pumped
to waste until the contamination concentration
dropped to acceptable levels. The first disadvantage
of this approach is that identifying sources is often
difficult because solvents like trichloroethylene and
tetrachloroethylene were used carelessly for many
years and dumped in sites such as airports and
shopping centers. These chemicals can migrate long
distances from the original site to the drinking water
supply. Second, even if the source can be identified,
pumping to waste might have to continue for years if
the contamination is severe and has occurred over
several years. Thus, even after eliminating a source,
treatment may still be necessary.
Another option is locating a new source of supply—
for example, drilling a new well. If the current source
is a ground-water supply, it is unlikely that the new
source would be a surface water supply because such
supplies require state permits and impoundments,
and surface water generally requires more involved
treatment, such as coagulation or filtration. Figure
5-3 illustrates the potential for contamination of a
newly drilled well. If the new well can be drilled
beyond the divide shown in the illustration, it will
not draw tainted water from the contaminated zone.
This option depends on accurately determining the
location of the divide and obtaining land beyond it.
In communities served by groups of wells in
wellfields, contaminated water can be blended with
uncontaminated water to reduce the concentration of
the contaminant through dilution (see Figure 5-4).
The contaminated water can be treated to some
extent and then blended, or the contaminated wells
can be manifolded together and treated at a central
location. On the other hand, individual treatment
units like GAG units and aerators can be installed at
the wellhouse of each contaminated well.
Disadvantages of blending include:
• The ground-water system may not be flexible
enough to permit sufficient blending; i.e., the
system hydraulics may be such that different wells
serve different parts of the distribution system.
• The contaminant concentrations may be too high
to achieve an acceptable level via dilution.
• Some states prohibit the use of blending for certain
contaminants.
• Consumers may not accept this alternative
because it does not involve actual removal of the
compound from the water.
Another option is installing an interceptor well (see
Figure 5-5). An interceptor well can be drilled
upstream of contaminated drinking water wells; this
61
-------
Figure 5-1. Monitoring wells.
Recharge Area
Nested System
Background
Well
Continuous Slotted -
Water
Table
Discharge Area
Aquifer
Water
Table
Contaminant
Plume
Flow Lines
Other data that should be gathered are the type of
contaminant, contaminant levels, and the type of
water supply. Inorganic contaminants tend to be
naturally occurring and show up in ground water in
relatively constant levels. In contrast, many organic
contaminants enter drinking water supplies as a
result of man's activities and thus contaminant levels
can vary significantly over time. Other site-specific
data to consider include:
• Water quality criteria such as hardness and, for
some heavy metals, competing ions for the
treatment process.
• Information on contaminant levels, including
historical levels, the mix of contaminant, and the
relation of these levels to goals for design influent
and effluent levels.
• Characteristics of the water supply, including
whether it is surface or ground water, the number
and location of wells in a ground-water supply, and
the supply system configuration (i. e., reservoir,
booster pumps).
In fact, the most cost-effective treatment solutions
are usually site-specific. For example, existing basins
originally designed for storage might be modified to
provide aeration for VOC contamination.
5.1.3 Choosing Control Strategies
After initial data gathering, risk managers can
examine three types of control strategies : 1) source
control, 2) treatment combinations, and 3) short-term
strategies. These options could be used to handle a
contamination problem in a small supply, such as
that illustrated in Figure 5-2. Note the mixture of
contaminants; each well contains a number of VOCs
at different levels and mixtures with no one
contaminant predominating. Also, each well
discharges to a different part of the distribution
system (a fairly common practice). Thus, treatment
would require either manifolding all wells together
or installing costly treatment equipment at each
well. In the illustration, the only current treatment is
chlorination, which would not control the VOC
contamination.
5.1.3.1 Source Control
Source control involves the following actions:
eliminating a contaminant source, obtaining a new
source by drilling a new well, blending contaminated
water with uncontaminated water, and operating an
interceptor well.
Eliminating a contaminated source is possible if the
source, such as a leaking underground storage tank,
60
-------
Chapter 5
Risk Reduction
This chapter builds on the definition of drinking
water problems in previous chapters, by describing
the process of rectifying these problems (i.e., "risk
reduction"). Section 5.1 provides an overview of risk
reduction and control strategies. Section 5.2
highlights inorganic treatment and Section 5.3
outlines organics treatment. Note that risk reduction
does not involve further investigation of health risks,
but rather is based on the risks defined in EPA
Health Advisories and Criteria Documents.
5.1 Overview of Risk Reduction and
Control Strategies
5.1.1 General Considerations
Control strategies for drinking water problems must
take into account a variety of factors. Among the
most important of these is the cost feasibility of
various treatment options, especially for small water
supplies. Treatment strategies must also account for
the ultimate fate of the contaminant in the
environment. For instance, the degradation product
of some contaminants in the environment can be
more toxic than the original contaminant. The
chemicals dichloroethylene, trichloroethylene, and
tetrachloroethylene, for example, break down into
the gas vinyl chloride in ground water.
Another factor is scientists' ability to test the extent
of ground-water contamination, especially for
contaminants that exert their health effects at very
low levels. To determine compliance, EPA must be
reasonably confident that the reported value for a
ground-water sample is close to the true value. To
this end, EPA has established two measurement
parameters, the minimum detection limit (MDL) and
the practical quantification level (PQL). The MDL is
the minimum concentration of a substance that can
be measured and reported with 99 percent confidence
that the true value is greater than zero. These MDLs,
however, are measured by sophisticated labs under
nonroutine conditions; moreover, MDLs indicate the
presence of a chemical (i.e., detection) not
measurement. Therefore, the PQL is generally 5 to 10
times the MDL, so that a sufficient number of
laboratories can report results within a reasonable
range of the true value (generally + 40 percent at
low levels, ± 20 percent at higher levels). Any
treatment concentration goal must be determined in
terms of the ability to analytically determine
contaminant concentrations after treatment.
5,1.2 Taking the Initial Steps In Planning a
Risk Reduction Strategy
When a drinking water contamination problem is
identified, several important initial steps should be
taken. A reliable data base must be developed,
through routine monitoring of existing nearby wells
(private, industrial, USGS, and state geological
survey wells); drilling of additional monitoring wells;
and assessment of existing hydrogeological data.
Figure 5-1 illustrates a contaminated site with
various monitoring wells. Note that the background
well does not show the contamination and that the
nested well system reveals the extent of
contamination. The continuous slotted well is also
designed to determine the vertical extent of
contamination; but, because water enters along most
of its length, it detects the presence of contaminants
but not their depth.
In some cases, especially for ground water, these
data can be plugged into mathematical models to
determine the location of the source of contamination
and to project future conditions (e. g., the impact of
continued pumping or cessation of pumping).
59
-------
Table 4-2, Health Effects of Some Chlorinated Halogenated Hydrocarbons
CNS
Depression
Sensitization
of
Heart
Liver
Injury
Kidney
Injury
Cancer
Methanes
Carbon tetrachloride
Chloroform
Dichloromethane
(methylene chloride)
Ethanes
1,1-Dichloroethane
1,2-Dichloroethane
1,1,1 -Trichloroethane
1,1,2-Trichloroethane
1,1,2,2-Tetrachloroethane
Hexachloroethane
Ethylenes
Chloroethylene
(vinyl chloride)
1,1-Dichloroethylene
(vinylidine chloride)
1.2-Trans-dichloroethylene
Trichloroethylene
Tetrachloroethylene
(perchloroethylene)
4.3 References
1. Workshops on Assessment and Management of
Drinking Water Contamination. EPA-600/M-
86/026, U.S. Environmental Protection Agency,
Center for Environmental Research Information,
Cincinnati, OH, 1987, pp. 48-50.
3. Klaassen, C.D., M.O. Amdur, and J. Doull, eds.
Casarett and Doull's Toxicology, 3rd Edition.
Macmillan, New York, NY, 1986, pp. 846-848.
4. Drinking Water and Health. National Academy
of Sciences, Washington, DC, 1977, pp. 169.
2. Federal Register 50(219): 46980,1985.
5. Klaassen, C.D., M.O. Amdur, and J. Doull, eds.
pp. 840-841.
58
-------
4.2.2.4 Fungicides
Fungicides are used to control fungus diseases on
plants, seeds, and produce. The fungicides include
organomercurial compounds, dithiocarbamates,
hexachlorobenzene, and pentachlorophenol. These
chemicals are less of a chronic, long-term threat to
drinking water than those discussed previously, but
they can have toxic and even carcinogenic effects in
doses of sufficient duration and amount.
Pentachlorophenol is of concern because commercial
samples have been shown to be contaminated with
highly toxic dibenzodioxins and dibenzofurans.
4.2.3 Solvents and Vapors (Volatile Organic
Compounds)
This class of contaminants includes halogenated
hydrocarbons and aromatic hydrocarbon solvents.
Many of these chemicals, especially chlorinated
solvents like trichloroethylene (TCE) and
perchloroethylene (PCE), have contaminated
drinking water sources; for example, by leaching
from hazardous waste sites or as a result of TCE
being used in the past for cleaning septic tanks..
4.2.3.1 Halogenated Hydrocarbons
These chemicals are widely used because they are
effective yet relatively inflammable solvents, as
opposed to kerosene or gasoline. Also, halogenated
hydrocarbons are formed during the chlorination of
drinking water when chlorine combines with organic
material in the water.
The halogenated hydrocarbons tend to have similar
health effects; hence, carbon tetrachloride will be
used as an example. Carbon tetrachloride in high
doses causes CNS depression and, as a result, was
once used as an anesthetic. It can also sensitize the
heart muscle to catecholamines (hormones such as
epinephrine) and thus can cause heart attacks. It can
cause kidney injury, liver injury, and cancer in
laboratory animals (Class B2). As mentioned earlier,
high blood alcohol levels can act as a potentiator for
carbon tetrachloride's damaging effects on internal
organs.
Table 4-2 summarizes the health effects of other
halogenated hydrocarbons. Plus signs (+) indicate a
harmful effect, minus signs (-) indicate a lack of
effect, and both a plus and a minus indicates a less
significant effect. Among the methanes, chloroform is
a trihalomethane formed during drinking water
chlorination. Chloroethylene (vinyl chloride) receives
three plus signs under cancer because it has been
established as a human carcinogen, while the others
have been established as probable human
carcinogens based on results of animal studies. Also
noteworthy are trichloroethylene and
tetrachloroethylene (perchloroethylene). These
chemicals are very common contaminants of
drinking water; although they are listed as probable
carcinogens, they cause cancer in laboratory animals
only at very high doses. Also, researchers have
observed that when perchloroethylene and
tetrachloroethylene degrade naturally in ground
water, vinyl chloride has been formed as a
degradation product.
4.2.3.2 Aromatic Hydrocarbon Solvents
The aromatic hydrocarbon solvents include benzene
and toluene. Benzene enters drinking water supplies
as a component of gasoline, which most often
contaminates water supplies by leaking out of
corroded underground storage tanks used by gas
stations. Acute doses of benzene cause CNS
depression, but chronic doses are more important
with regard to drinking water. Chronic exposure to
benzene in drinking water can cause bone marrow
depression - an impairment of the bone marrow's
ability to produce blood cells. More importantly,
chronic benzene exposure can cause leukemia in
humans and has been classified as a human
carcinogen by EPA (Class A). In comparison to
benzene, toluene is a relatively safe solvent (Class D).
However, in acute doses it can also cause CNS
depression, liver dysfunction, and kidney
dysfunction, and has been associated with female
reproductive effects (4).
4.2.4 Other Important Synthetic Organic
Chemicals
Polychlorinated biphenyls (PCBs) are synthetic
compounds that were manufactured in great
quantities from 1929 until the late 1970s, largely for
use in electrical equipment as nonconductive heat
transfer fluids. PCBs are lipid soluble and thus
biomagnify in the food chain and are teratogenic and
carcinogenic in laboratory animals. PCBs first
gained notoriety after an incident in Yusho, Japan in
1968 in which about 1,300 people used rice oil for
cooking that had been heavily contaminated with
PCBs. These people suffered a variety of health
effects, including skin irritation, eye discharges, GI
tract disturbances, and reproductive and nervous
system disorders. Clearly, this incident was a highly
acute exposure to PCBs; the effects of long-term, low-
level exposure to PCBs, such as would be found in
drinking water, have not yet been fully determined
(5).
57
-------
Fortunately, both lab tests and antidotes are
available for acute poisoning with organophosphorus
pesticides. Although currently available
organophosphorus insecticides do not cause delayed
neurotoxicity, some of the previously available
pesticides did cause delayed neurological problems
such as weakness, lack of muscle coordination, and
sensory disturbances.
Carbamates: The carbamate insecticides, such as
carbaryl and aldicarb, have toxicities very similar to
the organophosphorus insecticides. Like the
organophosphorus pesticides, these widely used
chemicals also act by inhibiting cholinesterase. The
toxic effects of carbamates may be more easily
reversed than those of the organophosphorus
pesticides. Current evidence does not seem to suggest
carcinogenicity as a toxic effect of the carbamates.
Aldicarb used on potato crops in Long Island, NY, has
contaminated ground-water aquifers.
Botanical Insecticides: The botanical insecticides
are derived from various plants and include the
compounds nicotine, pyrethrum, and rotenone.
Insecticides used in the home usually contain
pyrethrum and generally have a low order of human
toxicity. Insecticides containing nicotine are
considered the most toxic of the botanicals.
4.2.2.2 Fumigants
Fumigants are pesticides in gaseous form that can be
used to treat difficult-to-reach areas such as insects
in grain stored in grain elevators and insects in the
soil.
Fumigants include cyanide compounds,
methylbromide, dibromochloropropane (DBCP), and
ethylene dibromide (EDB). Acute doses of
methylbromide cause CNS depression and
pulmonary edema; in California there have been
more deaths due to acute poisoning with
methylbromide than the highly toxic
organophosphorus pesticides. However, it is not as
much of a problem in drinking water as DBCP and
EDB.
DBCP has been used extensively in California for soil
fumigation and has contaminated several drinking
water sources. In Nebraska, it has migrated from
treated grain into drinking water supplies. EDB has
contaminated drinking water sources in Florida,
Georgia, Connecticut, Massachusetts, Hawaii, and
other states. Both EDB and DBCP depress the CNS
and cause pulmonary edema in acute doses, and have
been shown to be potent carcinogens in laboratory
animals. DBCP has also been associated with
testicular injury and sterility in workers using and
manufacturing it. As a result of their probable
human carcinogenicity and reproductive effects,
many uses of EDB and DBCP have been banned.
4.2.2.3 Herbicides
Herbicides, chemicals used to kill plants, are used in
greater quantities than insecticides in the U.S.
Herbicides are added directly to the soil and thus
readily enter the ground water.
Chlorophenoxy Compounds: The chlorophenoxy
compounds include 2,4-D, 2,4,5-T, and 2,4,5-TP
(silvex). These compounds can have toxic effects on
the liver, kidney, and CNS, but clinical reports of
poisoning are rare. The compounds are rapidly
excreted into the urine, with a half-life of 24 hours in
humans. The most well-known and controversial of
the chlorophenoxy compounds is 2,4,5-T, which was
combined with 2,4-D to create the defoliant Agent
Orange used in the Vietnam War. When 2,4,5-T and
2,4,5-TP are manufactured, a contaminant called
tetrachlorodioxin (TCDD, or dioxin for short) is
inadvertently produced. Dioxin can also be produced
during the combustion of certain substances. This
chemical is the most toxic manufactured chemical
known. In sufficient doses, dioxin is a potent
teratogen and carcinogen in laboratory animals
(Class B2), and causes liver injury and general tissue
wasting. At lower doses, it causes a form of acne
called chloracne, which concentrates between the
eyes and hairline.
Clinical reports of acute dioxin poisoning are rare,
and in humans chloracne seems to be the worst effect
seen so far. Dramatic interspecies differences exist
for the effects of dioxin; the LD50 for guinea pigs is
about 1/10,000 of the LDso for hamsters. Fortunately,
evidence collected so far indicates that man's reaction
to dioxin resembles that of the hamster more than
that of the guinea pig. However, dioxin
contamination must be considered a serious
environmental problem. In Times Beach, Missouri,
dioxin contamination forced the abandonment of 800
homes in 1984.
Other Herbicides: Other herbicides include
dipyridyl compounds, triazines, and amides. Para-
quat is an example of a dipyridyl compound that can
cause severe lung injury through inhalation or GI
tract absorption. The triazines include the compound
atrazine, which seems to have a low order of toxicity.
However, because this abundantly used herbicide has
been detected in drinking water with increasing
frequency, further study is required. Structurally
similar to the triazifies is the herbicide amitrole
(aminotriazole), which attacks the thyroid and causes
cancer in laboratory animals.
Among the amides is the chemical alachlor, which is
used extensively in the U.S., especially on corn.
Other examples of amides are propachlor and
propanil. Although the amides have a low acute
toxicity, they have caused severe skin irritation, and
have been established as carcinogenic to rats in
recent studies.
56
-------
although there is no epidemiological evidence to
support this conclusion, ingested uranium deposits in
the bone in the same manner as radium. Through
chemotoxic effects, uranium can damage the kidneys.
A study of uranium occurrence in the late 1970s
conducted by the U.S. Geological Survey provides the
basis for an estimate of the actual natural uranium
concentra-tions in public water systems. Analysis of
the data from this study indicates that approximately
20,000 public water supplies (both surface- and
ground-water) have elevated natural uranium
concentrations; the average was approximately 2
pCi/L. Research is ongoing to determine the excess
health risk due to this level of contamination.
The isotope radon-222 is the subject of great public
concern lately. Inhalation of the short-lived progeny
of radon-222 can cause lung cancer; less is known
about the risks of ingested radon. Although most of
the total amount of radon-222 that enters homes
comes through the soil, it can also enter homes by
degassing from a dissolved state in drinking and
washing water. This degassing occurs when water is
heated and/or aerated, such as during clothes and
dish washing and showering and bathing. If the
concentration of radon-222 in the water is high
enough, using it can bring indoor radon levels above
the EPA guideline of 4 pCi per liter of indoor air.
4.2.2 Pesticides
Pesticides encompass a wide variety of compounds
formulated specifically to destroy plant or animal life
including insecticides, rodenticides, fungicides,
herbicides, and fumigants. Pesticides are used widely
in the U.S. and often eventually contaminate
drinking water. EPA has banned many of the uses of
some of these pesticides.
4.2.2.1 Insecticides
Insecticides can be divided into organochlorine,
organophosphorus, carbamate, and botanical
insecticides. Within each group, the pesticides have
similar characteristics; risk assessors use such
categories to make sense of the bewildering array of
commercially available insecticides.
Organochlorine Insecticides: This category was
commonly used in agriculture in the past and
includes the chlorinated ethanes, chlorinated
cyclodienes, and other chlorinated compounds.
DDT is the most well-known of the chlorinated
insecticides and was used extensively from World
War II until 1972, when it was banned in the U.S. It
is a highly lipid-soluble compound and thus is stored
in fat. In fact, most Americans have DDT
concentrations in their fat of 5-7 ppm. DDT is
biomagnified in the food chain; i.e., smaller
organisms absorb the compound and then are eaten
by progressively larger organisms, until DDT attains
a relatively high concentration in organisms such as
fish, which are then eaten by animals and people.
DDT is not particularly toxic to humans and most
other higher animal life, except in extremely high
doses. As a result, it was applied in much greater
quantities than were necessary. Then effects started
to be noticed. For example, DDT caused certain bird
species (for example, the Peregrine falcon) to produce
overly fragile egg shells that broke before hatching.
The toxicology of DDT stems from its ability to
increase production of P-450 enzymes. These enzymes
(see Section 4.1.4) metabolize foreign compounds
within the body. Apparently, they can also
metabolize endogenous compounds like estrogens,
which can interfere with proper egg shell growth. In
addition, recent experiments with laboratory
animals exposed to DDT have shown an increased
incidence of liver tumors. However, whether this
observation can be extrapolated to humans remains
to be answered. Methoxychlor is a compound similar
to DDT, but it is not as persistent in the environment.
An important subgroup of organochlorines are the
chlorinated cyclodienes. These chemicals include
aldrin, dieldrin, endrin, heptachlor, and chlordane.
These pesticides are similar to DDT but are more
toxic and have caused many human fatalities. Like
DDT, they are lipid soluble and stored in fat, and
thus are biomagnified in the environment. They have
also caused cancer in laboratory animals. Thus, their
registration for use on agricultural crops was
suspended in the mid-1970s.
Also similar are other chlorinated hydrocarbons such
as lindane, toxaphene, mirex, and kepone. In general,
the organochlorine insecticides cause some CNS
stimulation, induce P-450 production, increase
cancer incidence in laboratory animals (lindane, less
so), and persist in the environment to some degree.
Organophosphorus Insecticides: These
insecticides have largely replaced the chlorinated
hydrocarbon insecticides because they do not persist
in the environment and have an extremely low
potential to produce cancer. However, they have a
much higher acute toxicity in humans than the
organochlorines. A typical example of an organ-
ophosphorus insecticide is parathion, which must be
metabolized to the compound paraxon to exert its
toxic effect. This toxic effect stems from the
compound's ability to inhibit the enzyme
cholinesterase, a crucial chemical for the regulation
of the nerve transmitter acetylcholine. Thus, acute
effects of poisoning with organophosphorus
insecticides include fibrillation of muscles, low heart
rate, paralysis of respiratory muscles, confusion,
convulsions, and eventually death. (Another
organophosphorus pesticide called malathion is less
toxic in acute doses than parathion.)
55
-------
Figure 5-14.
200
100
O
Effect of contaminant concentration on carbon
life.
Compound: Trichloroethylene
EBCT = 10 min
-—^r::zr-
0 100 200 300 400 500 600 700 800 900 1,000
Influent Concentration
1,1,1- trichloroethane, about 5 months for
trichloroethylene, and about 1 year for
tetrachloroethylene. Thus, when costing out a GAG
system, plant managers must consider the level and
type of contaminant.
Figure 5-15. Effect of compound on carbon life.
400 -
300
-o
.B
S 200
-------
Figure 5-16. GAG treatment system.
trrfM
Sand ," •.'
Wastewater Recycle
\
Filtered Water to \
Distribution System -^
Well
/Pump
Table 5-9. Regeneration Options
Option Carbon Exhaustion Rate, Ib/day
OK-siIo disposal
Olf-silo rogcnoration
On-sito rofjonoration
<500
500-2,000
> 2,000
5.3.1 A Operational Issues
An operational problem called desorption can occur
when contaminant influent levels drop; since the
GAG process is an equilibrium phenomenon, a drop
in the influent concentration can actually cause
sloughing off of organics from the carbon.
Another problem is replacement, which occurs when
one contaminant is more preferred by the GAG; the
more preferred contaminant can replace the less
preferred contaminant that has already adsorbed
onto the surface of the carbon.
Bacteria can grow on the carbon. The types of
bacteria, and the health effects of this growth within
the context of the entire treatment system, require
further study.
Another important operational issue is "mass
transfer," or how the zone of spent GAG moves within
the contactor vessel. If the zone proceeds with an even
wavefront (see Figure 5-17), the carbon is used
efficiently. If the wavefront is upset (i.e., mixing of
Figure 5-17. Wavefront within GAC contactor
Influent
^." Spent Carbon
Wavefront
GAC Contactor
virgin with spent carbon), premature breakthrough
can result. Changing influent concentrations,
changing organic compounds, and the manner of
back washing can all affect the movement of the spent
zone.
Finally, waste disposal is an important issue. In
addition to disposal of spent carbon (for those systems
not regenerating), the contractors must be back-
washed both initially to remove fines and, with
some ground waters, periodically to remove solids.
This backwash must be either treated or disposed of
properly.
76
-------
5.3.1.5 Cost
The capital costs of a GAG system always include
three major components: contractors, the GAG itself,
and piping. Many other site- specific costs can be
added to these major components, including:
• Raw water holding tank
• Restaged well pump due to excessive pressure drop
through contractors
• Contactor building (needed in cold climates to
prevent freezing)
• Chemical feeding equipment
• Clear well (storage tank) and pumps
• Backwash storage tank
Figure 5-18 shows the relationship between system
size and capital cost based on the results of a survey
of GAG system capital costs in the U.S. (4). In
general, costs rise linearly with system size; the
outliers represent systems that incurred the
additional costs described above. Thus, if the EBCT—
the main determinant of system size—can be
estimated, the expected total capital cost can also be
estimated. Estimation of operating and maintenance
(O&M) costs is not as easy. Figure 5-19 illustrates the
relationship between system size and O&M cost. Note
that no clear relationship emerges. Different areas
throughout the U.S. have different contaminants and
contaminant levels that produce different carbon
usage rates—the main determinant of O & M cost
5.3.1.6 Case Study: Granular Activated
Carbon
This case study describes the use of GAG to treat
VOC contamination of ground water in Washington,
NJ (10). The contaminated supply was a single well
with a capacity of 550 gpm (0.792 MGD). The well
was contaminated with PCE at 50-500 ug/L, TCE at
1-10 ug/L, 1,1,1- trichloroethane at 1-20 ug/L, and
carbon tetrachloride at 1-5 ug/L. The well was
operated about 9 hours each day; the contaminant
levels varied throughout the day..
A GAG system was installed that incorporated two
downflow, pressure contractors operating in parallel.
Due to high solids concentrations in the raw water,
backwashing was performed about once a month. The
washwater was sand-filtered and recycled. An EBCT
of 10.5 minutes was used. When the water was
treated to below the regulatory limit for PCE, the
carbon usage rate was about 100 Ib of GAG per
million gallons of water treated.
5.3.2 Aeration
5.3.2.1 The Process
The aeration or "air stripping" process mixes
contaminated water with air so that the
contaminants can volatilize and escape into the
atmosphere. (This air stream is sometimes also
Figure 5-18. Capital costs for GAC systems.
400
350
300
8
°- 250
ifi
w 200
150
100
50
0.5 1.0 1.5 2.0
System Size, mgd
2.5
3.0
Figure 5-19. Annual O&M costs for GAC systems.
o
o
o_
To
o
O
5
O
ra
15
C
c
"^
60
55
50
45
40
35
30
25
20
15
10
5
-
* *
_
-
-
-
-
•
—
-
~»
—
1 I I I I I
0.5 1.0 1.5 2.0
System Size, mgd
2.5
3.0
treated.) The more volatile a compound is, the easier
it is to remove it from water through aeration.
Treatment system designers use a constant called
Henry's Law to compare the volatility of compounds.
Table 5-10 gives Henry's Law constants for several
contaminants. The VOCs have relatively low
solubilities, high vapor pressures, and relatively high
Henry's Law constants. Note that vinyl chloride has a
very high Henry's Law constant. As a result, EPA
has designated the packed tower aeration (PTA)
process as best available technology for vinyl
chloride; whereas for other regulated VOCs, EPA has
designated both GAC and PTA as best available
technology. Also, vinyl chloride is very poorly
adsorbed onto GAC.
5.3.2.2 Equipment
The two main types of aeration equipment are
diffused air units and waterfall units. Diffused air
units inject air into a water basin (see Figure 5-20);
77
-------
Tablo 5-10. Monty's Law Constants for Organic Chemicals
Typo of Organic Chemical
Henry's Law Constant,
Dimensionless Units
VOCs
Vinyl chloride
TCE
PCE
Cis-1,2-dichloroethylene
Pesticides
Aldicarb
Chtordano
DBCP
Chlorinated aromatics
PCB
Dichtorcbonzcno
285
0.44
0.88
0.18
1X10'7
0.015
0.011
0.021
0.086
Figure 5-20. Diffused air basin.
Inlliusnt
D'Huser Grid
Effluent
in the design shown, air enters the basin at the
bottom and disperses into many smaller bubbles as it
passes through the diffuser grid and rises to the
surface.
Waterfall units allow water to fall through air. The
most appropriate waterfall type of unitifor the
removal of VOCs is the PTA system (see Figure 5-21).
In this system, contaminated water flows down
through a column containing specially shaped
packing material that is designed to disperse the
water into many small droplets. A blower forces air
from the bottom of the column toward openings at the
top. When this air contacts the dispersed water in the
packing, contaminants leave the water and enter the
moving air stream. This air stream can then be
treated to prevent atmospheric pollution.
Other less complicated waterfall types of units have
been used, including open cascade systems that
simply direct a stream of turbulent water over a
spillway, spray aeration units that direct water
-through many nozzles over an open basin, and
multiple or slat tray units that use slats instead of
packing material.
Two recent waterfall designs are the catenary grid
unit and the Higee system. In the catenary grid
design (see Figure 5-22), water falls through a series
of parabolic wire grids mounted within a cylindrical
column; the grids serve the same purpose as the
packing in PTA systems except that the grids
"fluidize" the falling water between them,
momentarily slowing the water's passage through
the column. The efficiency of a catenary grid system
is a function of the number of grids used. Although
the catenary grid design allows a lower column than
with standard packing materials, much higher air
flows, and thus energy costs, are required.
In the Higee system (see Figure 5-23), water is
pumped into the center of a spinning disc of packing
material. Centrifugal force moves the water outward
while air is pumped in a countercurrent flow toward
the center of the unit. Existing Higee units are best
suited to temporary short-term treatmentdess than 1
year) at flows less than about 100 gpm. However,
future designs may be able to handle higher flows at
permanent systems.
5.3.2.3 Process Design Considerations
The diffused air system can be improved by
increasing basin depth, producing smaller bubbles,
optimizing basin geometry, and increasing gas flow.
Process design considerations for PTA must take into
account the following considerations:
Type of compound
Concentrations (ug/L)
Type of packing material
Air/water ratio (ft3/ft3)
Liquid loading rate (gpm/ft2)
Packing height (ft)
Water temperature
Figure 5-24 illustrates the effect of the type of
compound on two important variables, packing depth
and air/water ratio. Note the marked differences in
the design variables for the three contaminants
shown. Figure 5-25 shows the relationship between
packing height, water temperature, and removal
efficiency. The higher the water temperature, the
lower the packing height required for high removal
efficiency. Heating the raw water is usually not cost-
effective for typical drinking water system flowrates.
However, in temporary treatment situations and low
flowrates, water has been heated to increase tower
efficiency.
78
-------
Figure 5-21. Packed tower aeration system.
To Atmosphere
t
Packed
Column
Spray
Header
Plastic
Media
High Service
Vertical
Turbine Pumps
Blower
Assembly
Finished Water
to System
Clearwell
Well
5.3.2.4 Facility Design Considerations for PTA
Several design considerations are site-specific . Site
constraints include zoning requirements, height
restrictions, and restrictions on the location of air
intake louvers (the louvers can present noise
problems). The addition of PTA will also require
system hydraulics changes such as restaged well
, pumps, matching of booster pumps to well pumps,
and repumping of the treated water to the
distribution system.
In general, the temperature of the water within the
column is close to the temperature of the influent
water, and therefore it is not necessary to house the
tower. In one installation, air and water
temperatures were collected during winter operation
to demonstrate the small drop in water temperature
even under air temperatures well below freezing (see
Figure 5-26). In cold climates, blowers and pumping
equipment may have to be housed to prevent
freezing. Housing this equipment also provides the
additional benefits of increased security, reduced
noise, and less frequent maintenance.
The column itself can be made of a variety of
materials, including fiberglass-reinforced plastic,
aluminum, stainless steel, and concrete. The
column's internal components include a mist
eliminator to prevent water from escaping through
the air exit ports, a liquid distributor to separate the
water flow into many small streams, a support grid
for the packing material, and the packing material
itself.
The air flow system must draw ambient air that is
free of contamination and must discharge air that
meets discharge regulations. Some states enforce
discharge rate regulations in Ib/hr or Ib/day. Others
use modeling to determine the aeration system's
contribution to ambient concentrations of
contaminants. If an aeration system violates ambient
discharge regulations, column modifications such as
increased height, air flowrate, and exit velocity can
79
-------
Figure 5-22. Catenary grid system.
Demister
Water Inlet
Treated Water
Sample Collector
Manometer for
Air Flowrate Measurement
Raw Water
Rotameter
Raw Water
Sample Tap
Figure 5-23. HIgee system.
Exhaust Air
Air In
f
Blower
Ground
Water
Filter
Product
Water
Pump
Figure 5-24. Effect of compound on packed column design.
100
80
I 60
Q
0)
1 40
20
1,2-Dichloroethane
Chloroform
95 Percent Removal
Temperature: 55° F
TCE
0 20:1 40:1 60:1 80:1 100:1 120:1
A:W Ratio
80
-------
Figure 5-25. Packing height vs. removal (TCE).
35
30
£ 25
.31
0
X 20
o>
1 15
" Liquid loading Rate = 30 gpm/ft2
_ A:W Ratio = 30:1
_L
JL
J
50 75 90 95 97.5 99
Removal Efficiency, %
Figure 5-26. Low temperature on aeration.
99.5 99.8
Water In
51°F
Air Out
1 ' ' — 1
J_! L
J_! 1
*,
1 ' .'
) '
i i i
, .' '
' i
J_, „
i
!
!
.. _ . _L,
. . . i .
. . - ' ' - -
i — i
H
i— H—
_ _ . ...'._ ' '
iT-
,l, ',
!_!_
,',,'
!_!_
.'.,.!.
H-
, ,, , '„ '
1 '
' '
ii 11
iii it
J k
.J.
K
i '
Ambient Air
18°F
Water Out
49°F
Air In
be used to propel the exit air higher into the
atmosphere, thus allowing greater dilution and
dispersion to reduce contaminant concentrations
near the ground. If this does not work, or if a system
violates discharge rate regulations, the exit air can
be treated with a vapor phase carbon system, as
shown in Figure 5-27. This system passes the exit air
through a carbon contactor to trap the organic
contaminants. Before entering the contactor, the air
passes through a heating element to reduce its
relative humidity because high humidity competes
for carbon sites with the organics.
To prevent clogging of the packing in the column,
pretreatment may have to be incorporated into the
system. Causes for clogging include high calcium,
iron, and solids concentrations as well as biological
growth.
In some cases, the treated water may be made more
corrosive by aeration because the aeration process
increases the water's dissolved oxygen (DO)
concentration. However, since aeration also reduces
the treated water's CO2 concentration, the increased
DO is often balanced out. If necessary, corrosivity can
be reduced with corrosion inhibitors.
5.3.2.5 Cost Of PTA
The basic capital cost components of a PTA system
include the column structure and internals, the
packing, blower(s), clearwell, booster pump(s), and
any associated piping. Site-specific costs can also be
added to these costs, including the following:
Raw water holding tank
Restaged well pump
Blower building
Chemical facility
Noise control installation
Air emissions control
Figure 5-28 shows the relationship between system
size (in terms of water flowrate) and capital costs for
PTA systems. No clear trend emerges because
column height, rather than water and air flowrate, is
the main determinant of PTA capital cost. Therefore,
column height, rather than system size in flowrate,
should be used to estimate PTA capital costs.
Figure 5-29 shows the relationship between system
size (in flowrate) and O & M costs. Here, a linear
relationship occurs because flow rate is the main
determinant of O & M cost. Thus, an estimate of
system size will give a relatively good estimate of
O&M cost.
Table 5-11 summarizes the relative costs for removal
of certain organic contaminants by PTA.
5.3.2.6 Case Study: Packed Tower Aeration
This case study describes the use of packed tower
aeration to treat VOC contamination of ground water
in Scottsdale, AZ (11). The water supply system
81
-------
Figure 5-27. Vapor-phase carbon system for treatment of aeration exhaust air.
Contaminated Air
Raw Water
Clean Air'
Blower
Treated Water
Treated Air
Blower
Figure 5-28. Capital costs for packed column systems.
Table 5-11. Relative Costs for Removal by Aeration
g
s>
H
o
3
o
1,000
900
800
700
600
500
400
300
200
100
^
•
-
^
• 0
• •• * •
y *
pititiiiiiii
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
System Size, mgd
Figure 2-29. Annual O & M costs for packed columns.
100
^ 80
1 60
40
20
1.0 2.0 3.0 4.0 5.0
System Size, mgd
Vinyl chloride
PCE
TCE
Carbon tetrachloride
1,2-Dichloroethane
DBCP
Least Costly to Remove
Most Costly to Remove
consisted of 24 wells with a combined capacity of 40
MGD. Two wells were contaminated with TCE, one
with levels of 18-200 ug/L and the other with levels of
5-43 iig/L.
A packed tower aeration system was constructed that
had the following design characteristics: 1,200 gpm
flow, 12-ft packing height, 50:1 A:W ratio, and 10 ft
column diameter. Effluent TCE levels of 0.5 to 1.2
ug/L were achieved, making the removal efficiency
above 99 percent.
5.5 References
1. Krause, T.L. and E.L..Stover. Evaluating Water
Treatment Techniques for Barium Removal.
JAWWA 74 (9) 478,1982.
2. Sorg, T.J., R.W. Forbes, and D.g. Chambers.
Removal of Radium- 226 from Sarasota County,
Fla. Drinking Water by Reverse Osmosis.
JAWWA 72(4) 230-237,1980.
82
-------
3. Lauch, R.P. and G.A. Cutter. Ion Exchange for 8.
the Removal of Nitrate from Well Water. J.
AWWA78(5):83,1986.
4. Rubel, F. and R.D Woolsey. The Removal of
Excess Fluoride from Drinking Water by
Activated Alumina. JAWWA 71(1):45,1979.
9.
5. Rubel, F. Design Manual: Removal of Excess
Fluoride from Drinking Water. EPA-600/2-84-
134. U.S. Environmental Protection Agency,
Cincinnati, OH, 1984.
6. Rubel, F. and R.D. Woolsey. Technical Report:
Removal of Excess Fluoride from Drinking 10.
Water. EPA-5 3 0/9-78-001. U.S. Environmental
Protection Agency, Washington, DC) 1978.
7. Dqbbs, R.A. and J.M. Cohen. Carbon
Adsorption Isotherms for Toxic Organics. EPA-
600/8-80-023, U.S. Environmental Protection
Agency, Municipal Environ-mental Research 11.
Laboratory, Cincinnati, OH, 1980.
Dyksen, J.E., K. Raman, R.F. Raczko, and R.M.
Clark. GAC Treatment Costs Minimizing
Them. In: J. AWWA Seminar Proceedings,
Granular Activated Carbon Installations:
Conception to Operation, American Water
Works Association, Denver, CO, 1987.
Westerhoff, G.P and R. Miller. Design of the
Nation's First Major GAC Facility for Drinking
Water Treatment at Cincinnati, OH, In:
Proceedings of JAWWA Annual Conference,
American Water Works Association, Denver,
CO, 1985.
Chrobak, R.S., D.L. Kelleher, and I.H. Suffet.
Full-Scale GAC Adsorption Per-formance
Compared to Pilot-Plant Pre-dictions. In:
Proceedings of JAWWA Annual Conference,
American Water Works Association, Denver,
CO, 1985.
Cline, G.C., T.J. Lane, and M. Saldamando.
Packed Column Aeration for Trichloro-ethylene
Removal at Scottsdale, Arizona. In: Proceedings
of ABA Annual Conference, American Water
Works Association, Denver, CO, 1985.
83
-------
-------
Chapter 6
Risk Communication
A safe drinking water supply is very important to the
public, and any drinking water emergency, perceived
or real, will focus a lot of attention on the people
managing drinking water supplies. Therefore, this
chapter describes risk communication with the
public, primarily through the media. General
background on the media is provided and then
specific steps to take when relating to the media are
outlined
6.1 What You Need to Know About the
Media
First, it important to realize that the media can
actually be your ally in a drinking water crisis. They
can quickly disseminate crucial information to the
public, such as where to obtain bottled water. If their
coverage of the crisis is accurate, they can also allay
unfounded fears and inspire confidence
On the other hand, there are several disadvantages of
the media's coverage of environmental risk. Perhaps
the most important is shallowness. In newspapers
and radio and television news, stories must be brief
and easily understood by a very broad spectrum of
people: this condensation and simplification of often
very complicated situations can misrepresent the
truth.
Further, most reporters are generalists (or "general
assignment" reporters). Such a person has very little
college-level science background and very little time
to complete a story. Media deadlines are extremely
tight, with one reporter juggling multiple stories in
one day. Thus, a typical general assignment reporter
does not have the inclination or time to dig up the
background information necessary to fully
understand a drinking water emergency. All of these
factors add up to a rushed, somewhat unprepared
reporter arriving at you door. You will serve as this
reporters sole source of key scientific background
information (more on this later).
It is also important to note that the relative degree of
environmental risk (in terms of the actual human
toxicity yis-a-vis other daily risks and exposure
routes) is usually not of great interest to the reporter.
Typically, after having established that the
contaminant in question is, for example, carcinogenic
to laboratory animals, a reporter will move on to
other questions such as who caused the
contamination, who will clean it up, and how much it
will cost.
Another major disadvantage of the media's coverage
of environmental risk is sensationalism. The public
craves bad news, not good news. News stories must be
produced every day and a reporter's natural tendency
is to amplify the seriousness of your drinking water
emergency, even if the actual problem may be
relatively minor.
Another important media practice is the
personalizing of stories. A contamination emergency
may elicit questions such as "Would you want you
family to drink this water?"
6.2 Handling the Media
You can use several specific strategies in relating to
the media. A first step is to select a primary and
alternate spokesperson from within your
organization, and have the telephone receptionist
direct all media inquiries to these two people. Since
reporters' questions can seem combative, a
spokesperson must be calm under stress and capable
of speaking well in public. This spokesperson must
have access to all of the pertinent in-house
information.
85
-------
6.2.1 Overcoming Shallowness
6.3 Conclusion
To overcome the media's potential shallowness, you
must educate the reporter. Use lots of facts and be
prepared to clearly explain such principles of risk
assessment as the use of animal data, different
exposure routes, and the assumptions that underlie
dose-response curves.
If necessary, use simple visual aids such as paper flip
charts. In such material, make sure that all units are
presented consistently. Don't express a single
contaminant in three different (yet mathematically
equivalent) ways: for example, 1 mg/L, 1 ppm, and
1,000 ug/L. Explain any acronyms used, such as MCL
(maximum contaminant level).
If you're lucky enough to be covered by a specialized
"beat" reporter, try to cultivate this person. Perhaps
you can meet with this person to outline the scientific
and regulatory background the public needs to fully
understand drinking water issues.
6.2.2 Overcoming Sensationalism
To overcome sensationalism, appeal to the reporter's
values. Most reporters (as well as water utility
employees) adhere to a strict code of professional
ethics. If you make it clear that you are presenting
the facts in as clear and straightforward a manner as
possible, the media will treat you fairly in return.
When answering questions, never lie and never
guess. If necessary, offer to get back to a reporter for
questions that cannot be answered immediately.
Don't withhold information and allow a story to leak
out gradually, thereby creating the impression that
hidden wrongdoing in inexorably coming to light in
the press. Rather, present as much information as
possible in the initial interviews.
6.2.3 Conducting Interviews
In interviews, remember that, for all practical
purposes, there is no such thing as an "off the record"
comment. Assume all microphones are turned on.
When preparing for this interview, decide in advance
what you want to communicate, and stress and
repeat it until you are satisfied that the reporters
have gotten your message. Be firm but not hostile.
Don't forget practical considerations such as setting
aside space and telephones for the reporters.
In summary, prepare for the interview, educate the
media when necessary, and be as forthcoming as
possible. Finally, after having accommodated the
media in a professional manner, remember to stop
speaking. Don't be lured into speaking about subjects
about which you have limited knowledge.
As a final aid to the risk communicator, Table 6-1
provides a crisis communication checklist and Table
6-2 summarizes the 10 most common mistakes of
crisis communication.
Table 6-1. Crisis Communication Checklist
—Be prepared. Review the facts
-Be honest. Tell the truth.
—Anticipate likely questions.
—Consider what the audience wants to know.
—Decide what you want to say.
-Consider if there are things you do not want to discuss.
-Compose concise, accurate answers.
—Avoid jargon
—Don't fly by the seat of your pants; you might crash.
-If you do not know the answer to a question, do not guess.
-Stay calm. Do not lose your cool.
—Speak up. Do not mumble.
—Be assertive, not arrogant.
—Do not argue with reporters, bystanders, activists. Do not
show fright
-Avoid flight
—Counter false assumptions in questions.
—When finished, remember to stop.
Source: Rowan and Blewitt Environmental Consultants,
Washington, DC.
86
-------
Table 6-2. Ten Common Mistakes in Crisis Communication
The first mistake is to underestimate the importance of the media at the onset of a crisis. The
media's dissemination of information is crucial. In most serious emergencies, the presence
of photographers and reporters is automatic. If early on the press feels like an unwelcomed
guest, it returns the cool reception by heating up the rhetoric.
The second mistake is to fail to understand the media's need for regular information
updates. In this day of mini-cams, failure to provide concise factual updates can result in
wild speculation.
The third mistake is to fail to establish a central place where information can be coordinated.
Without one, reporters may wander and talk with uninformed bystanders. Communications
must be coordinated to ensure accurate information.
The fourth mistake is to fail to take charge. The spokesperson must both answer questions
and disseminate information.
The fifth mistake is to fail to anticipate likely questions. The basic questions of journalism-
who, what, when, how-can be expected. Remember, people want to know, "Is it safe now?"
The sixth mistake is to be lured into hypothetical questions.Avoid "what-if" questions.When
asked to predict, stick to the facts and make projections, if any, based on what is known.
The seventh mistake is to accidentally use emotionally charged or sensational language in
response to questions. Don't contribute to hype.
The eight mistake is to assign blame for an accident. It's likely that litigation will last for years
anyway, so keep personal opinions in check.
The ninth mistake is to try to distort the facts.
The tenth mistake is to let questions get under your skin. Show by your demeanor and candor
that you will cooperate with courteous journalists. Keep cool.
Source: Rowan and Blewitt Environmental Consultants, Washington, DC.
87
-------
-------
Appendix A
Primary and Secondary Drinking Water
Regulations
Table A-1. Primary Drinking Water Regulations
Contaminant
Health Effects
MCL, mg/L
Sources
Microbiological
Total coliforms (coliform
bacteria, fecal coliform,
streptococcal, and other
bacteria)
Turbidity
Inorganic Chemicals
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Nitrate
Selenium
Silver
Fluoride
Organic Chemicals
Endrin
Lindane
Methoxychlor
Not necessarily disease pro- 1 per 100 mL"
ducing themselves but can be
indicators of organisms that
cause assorted gastroenteric
infections: dysentery,
hepatitis, typhoid fever,
chlolera, and others. Also
interfere with disinfection
process
Interferes with disinfection 1 -5 Tu
Dermal and nervous system 0.05
effects
Circulatory 1.0
Kidney effects 0.01
Liver/kidney effects 0.05
Central and peripheral 0.05"
nervous system damage,
kidney effects, highly toxic to
infants and pregnant women
Central nervous system 0.002
disorders, kidney effects
Methemoglobinemia ("blue- 10
baby syndrome")
Gastrointestinal effects 0.01
Skin discoloration (argyria) 0.005
Skeletal damage 4
Numerous system/kidney 0.0002
effects
Numerous system/liver effects 0.004
Numerous system/kidney 0.1
effects
Human and animal fecal matter
Erosion, runoff, discharge
Geological, pesticide residues,
industrial waste, and smelter
operation
Geological
Geological, mining, and smelting
Industrial
Leaches from lead pipe and lead
based solder pipe joints
Manufacturing, fungicides,
geological
Fertilizer, sewage, feedlots,
geological
Geological, mining
Geological, mining
Geological additive to drinking
water, toothpaste, and food
processed with fluorinated water
Insecticide used on cotton, small
grains, orchards (cancelled)
Insecticide used on seed and soil
treatments, foilage applications,
wood protection
Insecticide used on fruit trees
and vegetables
A-1
-------
Table A-1. (continued)
Contaminant
Health Effects
MCL, mg/L
Sources
Organic Chemicals (continued)
2,4-D
2,4.5-TP (Silvex)
Toxaphene
Benzene
Carbon tetrachloride
p-Dichtoropenzene
1,2-Dichtoroethane
1,1 -Dichloroethylene
1,1,1 -Trichlorothane
Trichloroethylene (TCE)
Vinyl chloride
Liver/kidney effects
Liver/kidney effects
Cancer risk
Cancer
Possible cancer
Possible cancer
Possible cancer
Liver/kidney effects
Nervous system
effects
Possible cancer
Cancer risk
Total trihalomethanes (TTHM) Cancer Risk
(chloroform, bromoform,
bromodichloromethane,
dibromochloromethane)
Radlonuclides
Gross alpha particle activity
Gross beta particle activity
Radium 226 & 228
Cancer risk
Cancer risk
Bone cancer risk
0.1 Herbicide used to control broad-
leaf weeds in agriculture, used
on forests, range, pastures, and
aquatic environments.
0.01 Herbicide, canceled in 1984
0.005 Insecticide used on cotton, corn,
grain
0.005 Fuel (leaking tanks), solvent
commonly used in manufacture
of industrial chemicals,
Pharmaceuticals, pesticides,
paints, plastics
0.005 Common in cleaning agents,
industrial wastes from
manufacture of coolants
0.075 Used in insecticides, moth balls,
air deodorizers
0.005 Used in manufacture of
insecticides.gasoline
0.007 Used in manufacture of food
wrappings, synthetic fibers
0.20 Used in manufacture of food
wrappings, synthetic fibers
0.005 Waste from disposal of dry
cleaning material and
manufacture of pesticides,
paints, waxes and varnishes,
paint stripper, metal degreaser
0.002 Polyvinyl pipes (PVC) and
solvents used to join them,
industrial waste from
manufacture of plastics and
synthetic rubber
0.10 Primarily formed when surface
water containing organic matter
is treated with chlorine
15 pCi/L Radioactive waste, uranium
deposits
4 mrem/year Radioactive waste, uranium
deposits
5 pCi/L Radioactive waste, geological
ThQ final mlo of June 29, 1989 (effective December 31, 1990) based compliance on the presence/absence of total coliform rather
lhan density per unit volume as per current regulation.
Tho proposed rule o( August 18, 1988 proposed an MCL for lead of 0.005 mg/L for water leaving the treatment plant. In addition,
another lead standard of 0.01 mg/L was proposed for an average of a representative number of samples from customers' taps.
A-2
-------
Table A-2. Secondary Drinking Water Regulations
Contaminant
Limit
Effect
pH
Chloride
Copper
Foaming agents
Sulfate
Total dissolved solids (TDS)
6.5-8.5
250 mrj/L
1 mg/L
0.5 mg/L
250 mg/L
500 mg/L
Water should not be too acidic or too basic
Taste and corrosion of pipes
Taste and staining of porcelain
Aesthetic
Taste and laxative effects
Taste and possible relation between
low
Zinc
Fluoride
Color
Corrosivity
Iron
Manganese
Odor
5 mg/L
2 mg/L (plus
notification)
15 color units
Noncorrosive
0.3 mg/L
0.05 mg/L
3 threshold
odor number
hardness and cardiovascular disease, also an
indicator of corrosivity (related to lead levels in
water), can damage plumbing and limit
effectiveness of soaps and detergents
Taste
Dental fluorosis (discoloration of teeth)
Aesthetic
Aesthetic and health related (in relation to
leaching of materials such as lead from pipes)
Taste
Taste
Aesthetic
A-3
-------
-------
Appendix B
Health Advisory Documents for Aldicarb,
Atrazine, Trichloroethylene, and Vinyl Chloride
I. Introduction
The Health Advisory (HA) program, sponsored by the
Office of Drinking Water (ODW), provides
information on the health effects, analytical
methodology and treatment technology that would be
useful in dealing with the contamination of drinking
water. Health Advisories describe nonregulatory
concentrations of drinking water contaminants at
which adverse health effects would not be anticipated
to occur over specific exposure durations. Health
Advisories contain a margin of safety to protect
sensitive members of the population.
Health Advisories serve as informal technical
guidance to assist federal, state and local officials
responsible for protecting public health when
emergency spills or contamination situations occur.
They are not to be construed as legally enforceable
federal standards. The HAs are subject to change as
new information becomes available.
Health Advisories are developed for One-Day, Ten-
Day, Longer-Term (approximately 7 years, or 10% of
an individual's lifetime) and Lifetime exposures
based on data describing noncarcinogenic end points
of toxicity. Health Advisories do not quantitatively
incorporate any potential carcinogenic risk from such
exposure. For those substances that are known or
probable human carcinogens, according to the
Agency classification scheme (Group A or B),
Lifetime HAs are not recommended. The chemical
concentration values for Group A or B carcinogens
are correlated with carcinogenic risk estimates by
employing a cancer potency (unit risk) value together
with assumptions for lifetime exposure and the
consumption of drinking water. The cancer unit risk
is usually derived from the linear multistage model
with 95% upper confidence limits. This provides a
low-dose estimate of cancer risk to humans that is
considered unlikely to pose a carcinogenic risk in
excess of the stated values. Excess cancer risk
estimates may also be calculated using the one-hit,
Weibull, logit or probit models. Current
understanding of the biological mechanisms involved
in cancer is insufficient to prove that any one of these
models is able to predict risk more accurately than
another. Because each model is based on differing
assumptions, the estimates that are derived can
differ by several orders of magnitude.
These HAs are based on information presented in
ODW's draft Health Effects Criteria Documents
(CDs). The HA and CD formats are similar for easy
reference. Individuals desiring further information
on the toxicological data base or rationale for risk
characterization should consult the CD. The CD is
available for review at each EPA regional Office of
Drinking Water counterpart (e.g., Water Supply
Branch or Drinking Water Branch), or for a fee from
the National Technical Information Service, U.S.
Department of Commerce, 5285 Port Royal Rd.,
Springfield, VA 22161. The toll-free number is (800)
336-4700; the regular number is (703) 487-4650.
B-1
-------
II. ALDICARB (Sulfoxide and Sulfone)
A. General Information and Properties
CAS Wo.: 116-06-3
Structural Formula:
CH3 0
I II
CH3-S-C-CH=N - OCNHCHs
I
CH3
2-methyl-2-(methylthio)propionaldehydeO-
methylcarbamoyl oxime
Synonyms: Temik®
Use: Pesticide (nematocide, acaracide)
Properties: (U.S. EPA, 1985)
C7Hi402N2S
190.3
White crystals
Decomposes above
100°C
100°C
0.05torrat20°C
1.195 at 25°C
6 g/L (room temp.)
Odorless to light sulfur
smell
Chemical formula
Molecular weight
Physical state (room temp.)
Boiling point
Melting point
Density
Vapor pressure
Specific gravity
Water solubility
Taste threshold (water)
Odor threshold (water)
Odor threshold (air)
Conversion factor
Occurrence
• EPA estimated that aldicarb production ranged
from 3.0 to 4.7 million Ib per year during 1979-
1981. Aldicarb is applied both to the soil and
directly to plants.
• Aldicarb is considered to be moderately persistent
as a pesticide. Aldicarb is metabolized rapidly by
plants after application to sulfoxide and sulfone.
Once in the soil, aldicarb is degraded by both
aerobic and anaerobic bacteria. Aldicarb has a soil
half- life of 2 to 6 weeks, with residual levels found
up to 6 to 12 months later. Aldicarb in pond water
was reported to degrade more rapidly, with a half -
life of 5 to 10 days. Aldicarb is expected to
hydrolyze slowly over months or years in most
ground and surface waters. Aldicarb and its
sulfoxide and sulfone degradation products do not
bind to soil or sediments and have been shown to
migrate extensively in soil. Aldicarb does not
bioaccumulate to any significant extent.
• Aldicarb has been reported to occur widely in
ground water at levels in the low ppb range. New
York, Florida, Wisconsin and Maine, among other
states, have restricted the use of aldicarb based on
its potential for ground-water contam-ination.
Aldicarb has not been measured in Agency
surveys of drinking water, and estimates of
national exposures are unavailable. Because of
aldicarb's relatively rapid degradation rate, it is
expected to occur more often in ground waters than
surface waters (U.S. EPA, 1983).
• Monitoring of aldicarb residues on foods have
found only occasional low levels of the pesticide
and its metabolites (U.S. FDA, 1984). The Agency
has set limits for residues that would result in an
adult receiving a daily dose of 100 ug/kg a day. For
drinking water exposures to exceed this dose,
concentrations that need to exceed 50 ug/L.
B. Pharmacokinetics
Absorption
• Aldicarb, as well as its sulfoxide and sulfone
metabolites, has been shown to be absorbed readily
and almost completely through the gut in a variety
of mammalian and nonmammalian species
(Knaak et al., 1966; Andrawes et al., 1967;
Dorough and Ivie, 1968; Dorough et al., 1970;
Hicks et al., 1972; Cambon et al., 1979).
• Dermal absorption of aldicarb has been
demonstrated in rabbits (Kuhr and Dorough, 1976;
Martin and Worthing, 1977) and rats (Gaines,
1969), and would be expected to occur in
unprotected humans in manufacturing and field
application settings.
Distribution
• Aldicarb is distributed widely in the tissues of
Holstein cows when administered in feed (Dorough
et al., 1970). Highest residues were found in the
liver. When aldicarb was administered at a lower
level, residues were detected only in the liver.
• In rats administered aldicarb orally, residues were
found in all 13 tissue types analyzed. Hepatic
residue levels were similar to those of many other
tissues (Andrawes et al., 1967).
• Aldicarb, in a 1:1 molar ratio of the parent
compound to sulfone, administered orally to laying
hens in a single dose or for 21 consecutive days,
resulted in similar patterns of distribution with
the liver and kidneys as the main target organs
(Hicks et al., 1972). Residues also were present in
both the yolks and whites of the eggs laid by these
hens.
B-2
-------
A7etabo//sm
• The metabolism of aldicarb involves both
hydrolysis of the carbamate ester and oxidation of
the sulfur to sulfoxide and sulfone derivatives,
which have been shown to be active cholinesterase
inhibitors (Andrawes et al., 1967; Bull et al.,
1967).
• Metabolic end products of aldicarb detected in both
the milk and urine of a cow included the sulfoxides
and sulfones of the parent compound, oxime and
nitrile, as well as a number of unknown
metabolites (Borough and Ivie, 1968).
Excretion
• Elimination of aldicarb and its metabolism
products occurs primarily via the urine, as
demonstrated in rats (Knaak et al., 1966), cows
(Borough and Ivie, 1968), and chickens (Hicks et
al., 1972).
• Excretion of aldicarb via the lungs as C02 has been
demonstrated as a minor route in rats (Knaak et
al., 1966) and in the milk of cows (Borough and
Ivie, 1968).
• Excretion of aldicarb is relatively rapid, with
reported 24-hour elimination values in rats and
cows of approximately 80% to 90% of the
administered dose (Knaak et al., 1966; Borough
and Ivie, 1968).
C. Health Effects
Humans
• In two related incidents in 1978 and 1979,
ingestion of cucumbers presumed to contain
aldicarb at about 7 to 11 ppm resulted in
complaints of diarrhea, abdominal pain, vomiting,
nausea, excessive perspiration, dyspnea, muscle
fasciculation, blurred vision, headaches,
convulsions and/or temporary loss of limb function
in a total of fourteen residents of a Nebraska town
(CBC, 1979; Goes et al., 1980). Onset of symptoms
occurred within 15 minutes to 2.25 hours, and they
continued for approximately 4 to 12 hours.
• Industrial exposure by a man bagging aldicarb for
one day resulted in nausea, dizziness, depression,
weakness, tightness of chest muscles, and
decreases in plasma and red blood cell
cholinesterase activity (Sexton, 1966). The
symptoms lasted more than 6 hours, but the
subject returned to work the following day without
symptoms.
• In a laboratory study, four adult males orally
administered aldicarb at 0.1 rag/kg experienced a
variety of cholinergic symptoms, including
malaise, weakness in their limbs, pupil
contraction and loss of photoreactivity, epigastric
cramps, sweating, salivation, nausea, vomiting,
and "air hunger" (Haines, 1971). These symptoms
did not occur at 0.025 or 0.05 mg/kg. Bepression of
cholinesterase activity occurred in a dose-
dependent manner, with values as low as 25% of
the control value measured in two subjects dosed
at 0.1 mg/kg.
• Fiore et al. (1986) studied the effect of chronic
exposure to aldicarb-contaminated ground-water
on the human immune function. The study was
performed on women between the ages of 18 to 70.
A group of twenty-three women were exposed to
low levels of aldiearb ( < 61 ppb) and another group
of 27 women were unexposed. The results of this
study suggest a potential association between
exposure to aldicarb and abnormalities in T-cells.
However, the statistical analysis of these data
indicates that additional studies are needed before
further conclusions can be made on the effect of
aldicarb on the immune function.
Animals
Short-Term Exposure
• NAS (1977) stated that the acute toxicity of
aldicarb is probably the greatest of any widely
used pesticide.
• Reported oral LBso values for aldicarb
administered to rats in corn or peanut oil range
from about 0.65 to 1 mg/kg (Weiden et al., 1965;
Gaines, 1969). Females appear to be more
sensitive than males. The oral LD$Q in mice is 0.3
to 0.5 mg/kg (Black et al., 1973).
• Oral LB$o values for aldicarb were higher when
using a vehicle other than corn or peanut oil. Weil
(1973) reported an oral LB5o of 7.07 mg/kg in rats
administered aldicarb as dry granules. Carpenter
and Smyth (1965) reported an LB5o of 6.2 mg/kg in
rats administered aldicarb in drinking water.
Bermal toxicity also is high with 24-hour
values of 2.5 and 3 mg/kg reported for female and
male rats, respectively (Gaines, 1969) and 5 mg/kg
in rabbits (Weiden et al., 1965).
The principal toxic effect of aldicarb and its
sulfoxide and sulfone metabolites in rats has been
shown to be cholinesterase inhibition (Weil and
Carpenter, 1963; Nycum, 1968; Weil, 1969).
Feeding studies of short duration (7 to 15 days)
have been conducted by various authors using
aldicarb and/or its sulfone and sulfoxide.
Statistically significant decreases in
B-3
-------
cholinesterase activity were observed in rats at
dosage levels of 1 mg/kg/day (the approximate
LDgo in rats) (Nycum and Carpenter, 1970) and at
2.5 mg/kg/day in chickens (Sehlinke, 1970). The
latter dosage also resulted in some lethality in test
animals.
• A NOABL has been determined for a mixture of
aldicarb oxidation products based on data reported
by Mirro et al. (1982), who administered aldicarb
sulfone and sulfoxide in a 1:1 ratio in the drinking
water of young rats for 8 to 29 days. Doses ranged
up to 1.67 mg/kg/day for males and 1.94 mg/kg/day
for females. Based on statistically significant
reductions in cholinesterase activity in brain,
plasma, and red blood cells (RBC) at higher dosage
levels, a NOAEL of 0.12 mg/kg/day was
determined.
Long-Term Exposure
• High dosages of aldicarb sulfoxide (0.25 to 1.0
mg/kg/day) or aldicarb sulfone (1.8 to 16.2
mg/kg/day) administered in the diets of rats for 3
or 6 months resulted in decreases in cholinester-
ase activity in plasma, RBCs, and brain (Weil and
Carpenter, 1968a,b). No increases in mortality or
gross or microscopic histopathology were noted in
any group, however. Data derived from the lower
dosage levels of this study have been used by the
World Health Organization Committee on
Pesticide Residues (PAO/WHO, 1980) to derive a
NOAEL of 0.125 mg/kg/day for aldicarb sulfoxide
in the rat. The NOAEL for aldicarb sulfone alone
was 0.6 mg/kg/day.
• Aldicarb administered for 2 years in the diets of
rats or dogs at levels up to 0.1 mg/kg/day resulted
in no significant increase in adverse effects based
on a variety of toxicologic end points (Weil and
Carpenter, 1965,1966a). In another 2-year study,
levels of up to 0.3 mg/kg/day resulted in no adverse
affects in rats (Weil, 1975).
• Feeding studies using aldicarb sulfoxide at 0.6
mg/kg/day for 2 years resulted in an increase in
the mortality rates of female rats (Weil, 1975).
Reproductive Effects
• No reproductive effects have been demonstrated to
result from the administration of aldicarb to rats
(Weil and Carpenter, 1964,1974).
Developmental Effects
• No teratogenic effects have been demonstrated
from the administration of aldicarb in rabbits
(IRDC, 1983) or chickens (Proctor et al., 1976).
• No adverse effects on milk production were
observed in studies of lactating cows or rats
(Dorough and Ivie, 1968; Dorough et al., 1970).
• Statistically significant inhibition of
acetylcholinesterase activity has been
demonstrated in the liver, brain, and blood of rat
fetuses when their mothers were administered
aldicarb by gastric intubation on day 18 of
gestation (Cambon et al., 1979). These changes
were seen at doses of 0.001 mg/kg and above and
were manifested within 5 minutes of the
administration of 0.1 mg/kg.
Mutagenicity
• Aldicarb has not been demonstrated to be
conclusively mutagenic in Ames bacterial assays
or in a dominant lethal mutagenicity test in rats
(Ercegovich and Rashed, 1973; Weil and
Carpenter, 1974; Godek et al., 1980).
Carcinogenicity
• Neither aldicarb nor its sulfoxide or sulfone
metabolite have been demonstrated to increase
significantly the incidence of tumors in mice or
rats in feeding studies (Weil and Carpenter, 1965;
NCI, 1979). Bioassays with aldicarb in which rats
and mice were fed either 2 or 6 ppm in the diet for
103 weeks revealed no tumors that could be
attributed solely to aldicarb administra-tion (NCI,
1979). It was concluded that, under the conditions
of the bioassay, technical grade (99 + %) aldicarb
was not carcinogenic to F344 rats or B6C3Fi mice
of either sex. A 2-year feeding study reported by
Weil and Carpenter (1965) also produced no
statistically significant increase in tumors over
controls when rats were administered aldicarb at
equivalent doses of 0.005, 0.025, 0.05, or 0.1 mg/kg
bw/day in the diet. Weil (1975) similarly reported
no adverse effects in Greenacres Laboratory
Controlled Flora rats fed aldicarb at 0.3 mg/kg
bw/day for 2 years.
• In the only skin-painting study available to date,
Weil and Carpenter (1966b) found aldicarb to be
noncarcinogenic to male C3H/H3J mice under the
conditions of the experiment.
• Intraperitoneally administered aldicarb did not
exhibit transforming or tumorigenic activity in a
host-mediated assay using pregnant hamsters and
nude (athymic) mice (Quarles et al., 1979).
B-4
-------
D. Quantification of Toxicological
Effects
The HAs for noncarcinogenic toxicants are derived
using the following formula:
HA =
(NOAEL or LOAEL) x (BW)
(UF)x(_ L/day)
.mg/L(
.ug/L)
where:
NOAEL or
LOAEL
BW
No- or Lowest-Observed-Adverse- ...
Effect Level in mg/kg bw/day.
assumed body weight of a child (10
kg) or an adult (70 kg).
UF = uncertainty factor (10,100, or 1,000),
in accordance with NAS/ODW
guidelines.
L/day = assumed daily water consumption of
a child (1 L/day) or an adult (2 L/day).
The available data suggest that the appearance of
cholinergic symptoms indicative of cholinesterase
enzyme inhibition is the most sensitive indicator the
effects of exposure to aldicarb. Adverse health effects
appear to be related primarily to the depression of
cholinesterase activity, as no other biochemical,
morphological, reproductive, mutagenic or
carcinogenic effects have been reported, even after
chronic dosing.
Given the nature of the primary toxicity (rapidly
reversible cholinesterase inhibition) of aldicarb and
its oxidative metabolites/degradation products, it is
apparent that the same NOAEL can be used,as the
basis for the derivation of acceptable levels over
virtually any duration of exposure. In addition, the
Health Advisories calculated in this document are
appropriate for use in circumstances in which
sulfoxide and/or sulfone may be the substance(s)
present in a drinking water sample. Depending upon
the analytical method applied, it may not be possible
to characterize specifically the residue(s) present. By
establishing Health Advisories based on data from
valid studies with the most potent of the three
substances, assurance is greater that the guidance is
protective to human health.
As described above, a NOAEL of 0.125 mg/kg bw/day
can be determined from the Weil and Carpenter
(1968b) and Mirro et al. (1982) studies. From this
NOAEL, all HA values can be determined for
aldicarb, aldicarb sulfoxide, or a mixture of the
sulfoxide and sulfone metabolite (however, if the only
contaminant is sulfone and a less conservative value
is thus appropriate, the NOAEL for the sulfone, 0.6
mg/kg/day, as determined in the Weil and Carpenter
(1986) study, can be used).
One-Day Health Advisory
For the 10-kg child:
(0.125 mg/kg/day) (10 kg)
One-Day HA -
(100) (1 L/day)
where:
0.125 mg
/ kg/day
10kg
100
1 L/day
= 0.012 mg/L (10 ug/L)
NOAEL, based on lack of significant
decreases in cholinesterase
activity in rats.
assumed body weight of a child.
uncertainty factor, chosen in
accordance with NAS/ODW
guidelines for use with a NOAEL
from an animal study.
assumed daily water consumption of
a child.
(NOTE: Using the NOAEL for the sulfone alone, the
HA value for this metabolite would be 0.06 mg/L (60
ug/L) if sulfone is the only contaminant.)
Ten-Day Health Advisory
Since aldicarb is metabolized and excreted rapidly
(>90% in urine alone in a 24-hour period following a
single dose), the One- and Ten-Day HA values would
not be expected to differ to any extent. Therefore, the
Ten-Day HA will be the same as the One-Day HA (10
ug/L).
Longer-Term Health Advisory
For the 10-kg child:
Longer— term HA =
(0.125 mg/kg/day) (10 kg)
(100) (1 L/day)
= 0.012 mg/L (10 ug/L)
where:
0.125mg
/kg/day
10kg
100
= NOAEL, based on lack of significant
decreases in cholinesterase activity
in rats.
= assumed body weight of a child.
= uncertainty factor chosen in
accordance with NAS/ODW
guidelines for use with a NOAEL
from an animal study.
1 L/day = assumed daily water consumption of
a child.
B-5
-------
(NOTE: Using the NOAEL for the sulfone alone, the
HA value for this metabolite would be 0.06 mg/L (60
ug/L) if the sulfone is only contamiant.)
For the 70-kg adult:
Longer—term HA =
(0.125 mg/kg/day) (70 kg)
(100) (2 Lt'day)
= 0.042 mg/L (40 pg/L)
where:
0.125 mg
/kg/day
70kg
100
= NOAEL, based on lack of significant
decreases in cholinesterase activity
in rats.
= assumed body weight of an adult.
= uncertainty factor chosen in
accordance with NAS/ODW
guidelines for use with a NOAEL
from an animal study.
2 L/day = assumed daily water consumption of
an adult.
(NOTE: Using the NOAEL for the sulfone alone, the
HA value for this metabolite would be 0.21 mg/L (210
pg/L) if sulfone is the only contaminant.)
Lifetime Health Advisory
The Lifetime HA represents that portion of an
individual's total exposure that is attributed to
drinking water and is considered protective of
noncarcinogenic adverse health effects over a
lifetime exposure. The Lifetime HA is derived in a
three step process. Step 1 determines the Reference
Dose (RfD), formerly called the Acceptable Daily
Intake (ADI). The RfD is an estimate of a daily
exposure to the human population that is likely to be
without appreciable risk of deleterious effects over a
lifetime, and is derived from the NOAEL (or
LOAEL), identified from a chronic (or subchronic)
study, divided by an uncertainty factor (s). From the
RfD, a Drinking Water Equivalent Level (DWEL)
can be determined (Step 2). A DWEL is a medium-
specific (i.e., drinking water) lifetime exposure level,
assuming 100% exposure from that medium, at
which adverse, noncarcinogenic health effects would
not be expected to occur. The DWEL is derived from
the multiplication of the RfD by the assumed body
weight of an adult and divided by the assumed daily
water consumption of an adult. The Lifetime HA is
determined in Step 3 by factoring in other sources of
exposure, the relative source contribution (RSC). The
RSC from drinking water is based on actual exposure
data or, if data are not available, a value of 20% is
assumed for synthetic organic chemicals and a value
of 10% for inorganic chemicals. If the contaminant is
classified as a Group A or B carcinogen, according to
the Agency's classification scheme of carcinogenic
potential (U.S. EPA, 1986), then caution should be
exercised in assessing the risks associated with
lifetime exposure to this chemical.
As discussed previously, the studies by Weil and
Carpenter (1968b) and Mirro et al. (1982) are used in
the following calculations. Both studies reflected a
NOAEL of 0.125 mg/kg/day.
Step 1: Determination of the RfD ,
RfD =
where:
(0.125 mg/kg/day)
000)
= 0.00125 mg/kg/day
0.125mg = NOAEL, based on lack of significant
/kg/day decreases in cholinesterase activity
in rats.
100 = uncertainty factor, chosen in
accordance with NAS/ODW
guidelines for use with a NOAEL
from an animal study.
(NOTE: With the NOAE.L of 0.6 mg/kg/day for the
sulfone alone, the RfD value for this metabolite
would be 0.006 mg/kg/day.)
Step 2: Determination of the DWEL
DWEL -
(0.00125 mg/kg/day) (70 kg)
(2 L/day) ~
where:
0.00125 mg
/kg/day
70kg
2 L/day
= RfD.
= assumed body weight of an adult.
= assumed daily water
consumption of an adult.
(NOTE: With the RfD for sulfone alone, the DWEL
for this metabolite would be 0.21 mg/L (210 pg/L)).
Step 3: Determination of the Lifetime Health
Advisory
Lifetime HA
where:
0.42 mg/L
20%
= (0.042 mg/L) (20%)
= 0.009 mg/L = 10 pg/L
= DWEL.
= assumed contribution of drinking
water to total exposure to
aldicarb.
B-6
-------
(NOTE: With the DWEL for sulfone alone, the
Lifetime HA value for this metabolite would be 0.042
mg/L(42ug/L).
In summary, the Lifetime HA values for aldicarb and
its metabolites are as follows:
aldicarb (parent compound)* : 10 ug/L
aldicarb sulfoxide* : 10 ug/L
aldicarb sulfone** : 10 to 42 ug/L
* The HA values for aldicarb and aldicarb sulfoxide
are the same because they have similar toxicity,
and the effects of the parent compound are likely
due to the sulfoxide (and, to a lesser extent, the
sulfone).
** The HA value for the sulfone ranges from 10 to 42
ug/L depending on the presence of other aldi-
carb/aldicarb sulfoxide residues; only if the
sulfone metabolite is present alone as a
contaminant can the HA value of 42 ug/L be used.
Evaluation of Carcinogenic Potential
• Since aldicarb has been found to be
noncarcinogenic under all conditions tested,
quantification of carcinogenic risk for lifetime
exposures through drinking water would be
inappropriate.
• The International Agency for Research on Cancer
(IARC) has not classified the carcinogenic
potential of aldicarb.
• Applying the criteria described in EPA's
guidelines for assessment of carcinogenic risk
(U.S. EPA, 1986), the Agency has classified
aldicarb in Group E: No evidence of
carcinogenicity in humans. This category is used
for agents that show no evidence of carcinogenicity
in at least two adequate animal tests in different
species or in both epidemiologic and animal
studies.
E. Other Criteria, Guidance, and
Standards
• The National Academy of Sciences proposed an
ADI of 0.001 mg/kg/day based on 2-year feeding
studies in rats and dogs (NAS, 1977). NAS
reaffirmed this ADI in 1983 (NAS, 1983).
• In addition, NAS also derived a chronic Suggested-
No-Adverse-Effect level (SNARL) of 7 ug/L, using
the studies mentioned above with an uncertainty
factor of 1,000 (1977). The SNARL is protective of
a 70-kg adult, consuming 2 L of water per day and
for whom drinking water is assumed to contribute
20 percent of the daily exposure to aldicarb
residues.
• EPA's Office of Pesticide Programs established an
ADI of 0.003 mg/kg/day based on the data from the
6-month rat feeding study with aldicarb sulfoxide
(U.S. EPA, 1981).
• The FAO/WHO proposed ADIs for aldicarb
residues of 0-0.001 mg/kg/day in 1979 and 0-
0.005 mg/kg/day in 1982 (FAO/WHO, 1979; 1982).
F. Analytical Methods
• Analysis of aldicarb is by a high performance
liquid chromatographic procedure used for the
determination of N-methyl carbamoyloximes and
N- methylcarbamates in drinking water (U.S.
EPA, 1984). In this method, the water sample is
filtered and a 400-uL aliquot is injected into a
reverse phase HPLC column. Separation of
compounds is achieved using gradient elution
chromatography. After elution from the HPLC
column, the compounds are hydrolyzed with
sodium hydroxide. The methylamine formed
during hydrolysis is reacted with o-pthalaladehyde
(OPA) to form a fluorescent derivative that is
detected using a fluorescence detector. The method
detection limit has been estimated to be
approximately 1.3 ug/L for aldicarb.
G. Treatment Technologies
• Techniques that have been used to remove
aldicarb from water are carbon adsorption and
filtration. Since aldicarb is converted into aldicarb
sulfoxide and sulfone, all three compounds must be
considered when evaluating the efficiency of any
decontamination technique.
• Granular activated carbon (GAC) has been used in
two studies of aldicarb removal from contam-
inated water (Union Carbide, 1979; ESE, 1984);
Both studies utilized home water treatment units
rather than large-scale water treatment systems.
Union Carbide tested the Hytest Model HF-1
water softener in which the ion exchange ion was
replaced with 38.5 Ib Filtrasorb® 400 (Calgon
GAC). The unit was operated at a flow rate of 3
gal/min. Water spiked with 200 ppb or 1000 ppb of
a mixture of aldicarb, aldicarb sulfoxide, and
aldicarb sulfone in a 10:45:45 ratio was treated.
Under these conditions, the total aldicarb residue
level was reduced by 99% to 1 ppb for the
treatment of 13,500 gallons of water with 200 ppb
of residues and 41,500 gallons with 1,000 ppb total
residues. No breakthrough of aldicarb occurred.
When the study was terminated, the carbon had
adsorbed 9 mg aldicarb residue per gram. This
value can be compared with an equilibrium
loading value of 21 mg per gram of carbon at 166
determined using 200 ppb aldicarb residues. In the
second study, ESE (1984) did a field study in
Suffolk County, NY. Nineteen units using type
B-7
-------
CW12 x 40 mesh carbon were tested. After 38
months of use, breakthrough of aldicarb occurred
to levels over 7 ug/L in eight units tested. The
range of usage values can be attributed to the fact
that the natural well samples contained a variety
of adsorbable substances in addition to aldicarb.
• Chlorination also appears to offer the potential for
aldicarb removal (Union Carbide, 1979). The
company reported that 1.0 ppm free chlorine
caused a shift in the ratio of aldicarb, its sulfoxide
and its sulfone so that all residues were converted
to the sulfoxide within 5 minutes of chlorine
exposure. Normal conversion of aldicarb to
aldicarb sulfone did not appear to be affected. On
stand-ing, the sulfoxide and sulfone decomposed.
The decomposition products were not identified.
However, should these be nontoxic, then
chlorination could be feasible as an aldicarb
removal tech-nique.
• Aeration or air stripping, which is commonly used
to remove synthetic organic chemicals, is not a
good technique for the removal of aldicarb (ESE,
1984). This is because aldicarb has a low Henry's
Law constant (2.32 x 10-4 atm).
H. References
Andrawes, N.R., H. W. Borough and D.A. Lindquist.
1967. Degradation and elimination of Temik in rats.
J. Econ. Entomol. 60(4) 979-987.
Black, A.L., Y.C. Chiu, M.A.H. Fahmy andT.R.
Fukuto. 1973. Selective toxicity of N-sulfenylated
derivatives of insecticidal methylcarbamate esters. J.
Agr. Food Chem. 21:747-751.
Bull, D.L., D.A. Lindquist and J.R. Coppedge. 1967.
Metabolism of 2-methyl- 2(methylthio) propionalde-
hyde 0-(methyl carbamoyl) oxime (Temik, UC-21149)
in insects. J. Agr. Food Chem. 15(4):610-616.
Cambpn, C., C. Declume and R. Derache. 1979. Effect
of the insecticidal carbamate derivatives (carbofuran,
primicarb, aldicarb) in the activity of
acetylcholinesterase in tissues from pregnant rats
and fetuses. Toxicol. Appl. Pharmacol. 49:203-208.
Carpenter, C.P. and H.F. Smyth. 1965.
Recapitulation of pharmacodynamic and acute
toxicity studies on Temik. Mellon Institute Report
No. 28-78. EPA Pesticide Petition No. 9F0798.
CDC (Centers for Disease Control). 1979.
Epidemiologic notes and reports: Suspected
carbamate intoxications — Nebraska. Morbid.
Mortal. Week. Rep. 28:133-134.
Dorough, H.W., R.B. Davis and G.W. Ivie. 1970. Fate
of Temik-carbon-14 in lactating cows during a 14-day
feeding period. J. Agr. Food Chem. 18(1):135-143.
Dorough, H .W. and G.W. Ivie. 1968. Temik-535
metabolism in a lactating cow. J. Agr. Food Chem.
16(3): 460-464.
Ercegovich, C.D. and K.A. Rashid. 1973.
Mutagenesis induced in mutant strains of
Salmonella typhimurium by pesticides. Abstracts of
Papers. Am. Chem. Soc. p. 43.
ESE. 1984. Environmental Science and Engineering.
Review of treatability data for removal of twenty-five
synthetic organic chemicals from drinking water.
Prepared for EPA's Office of Drinking Water.
FAO/WHO. 1979,1980 and 1982. References not
available.
Fiore, N.C., H.A. Anderson, R. Hong, R.
Golubjatnikov, J.E. Seiser, D. Nordstrom, L.
Hanrahanand D. Belluck. 1986. Chronic exposure to
aldicarb-contaminated groundwater and human
immune function. Env. Res. 4-1: 633-645.
Gaines, T.B. 1969. The acute toxicity of pesticides.
Toxicol. Appl. Pharmacol. 14:515-534.
Godek, E.S., M.C. Dolak, R.W. Naismith and R.J.
Matthews. 1980. Ames SaZmoneWa/Microsome Plate
Test. Unpublished report by Pharmakon
Laboratories. Submitted to Union Carbide June 20,
1980.
Goes, E.H., E.P. Savage, G. Gibbons, M. Aaronson,
S. A. Ford and H.W. Wheeler. 1980. Suspected food-
borne carbamate pesticide intoxications associated
with ingestion of hydroponic cucumbers. Am. J.
Epidemiol. 111:254-259.
Haines, R.G. 1971. Ingestion of aldicarb by human
volunteers: A controlled study of the effect of aldicarb
on man. Union Carbide Corp., Unpublished report
with addendum (A-D), Feb. 11,1971, 32 pages.
Hicks, B.W., H.W. Dorough and H.M. Mehendale.
1972. Metabolism of aldicarb pesticide in laying
hens. J. Agr. Food Chem. 20(1):151-156.
IRDC. 1983. International Research and
Development Corporation. 1983. Teratology study in
rabbits. Union Carbide Corporation.
Knaak, J.B., M.J. Tallant and L.J. Sullivan. 1966.
The metabolism of 2-methyl-2- (methylthio)
propionaldehyde 0-(methyl carbamoyl) oxime in the
rat. J. Agr. Food Chem. 14(6):573-578.
B-8
-------
Ruhr, R.J. and H.W. Borough. 1976. Carbamate
Insecticides: Chemistry, Biochemistry, and
Toxicology. CRC Press, Inc., Cleveland, OH. pp. 2-6.
103-112,187-190, 211-213, 219-220.
Martin, H. and C.R. Worthing, Eds. 1977. Pesticide
Manual. British Crop Protection Council,
Worcestershire, England, p. 6.
Mirro, E.J., L.R. DePass and F.R. Frank. 1982.
Aldicarb sulfone: aldicarb sulfoxide twenty-nine-day
water inclusion study in rats. Mellon Inst. Rep. No.
45-18.
NAS. 1977. National Academy of Sciences. Drinking
Water and Health Volume 1. National Academy
Press. Washington; DC. pp. 635-643.
NAS. 1983. National Academy of Sciences. Drinking
Water and Health Volume 5. National Academy
Press. Washington, DC. pp. 10-12.
NCI. 1979. National Cancer Institute. Bioassay of
aldicarb for possible carcinogenicity. National
Institutes of Health. U.S. Public Health Service.U.S.
Department of Health, Education and Welfare.
Washington, DC. NCI-CG-Ta-136.
Nycum, J.S. 1968. Toxicity studies onTemik and
related carbamates.Mellon Institute, unpublished
report 31-48, 5 pages.
Nycum, J.S., and C. Carpenter. 1970. Summary with
respect to Guideline PR70-15. Mellon Institute
Report No. 31-48. EPA Pesticide Petition No. 9F0798.
Proctor, N.H., A.D. Moscioni and J.E. Casida. 1976.
Chicken embryo NAD levels lowered by teratogenic
organophosphorus and methylcarbamate
insecticides. Biochem. Pharmacol. 25:757-762.
Quarles, J.M., M.W. Sega, C.K. Schenley and W.
Lijinsky. 1979. Transformation of hamster fetal cells
by nitrosated pesticides in a transplacental assay.
Cancer Res. 39:4525-4533.
Schlinke, J.C. 1970. Toxicologic effects of five soil
nematocides in chickens. J. Am. Vet. Med. Assoe.
31:119-121.
Sexton, W.F. 1966. Report on aldicarb. EPA Pesticide
Petition No. 9F0798, Section C.
Union Carbide. 1979. Union Carbide Agricultural
Products Company. Temik® aldicarb pesticide.
Removal of residues from water. Research and
Development Department.
U.S. EPA. 1981. U.S. Environmental Protection
Agency. 40 CFR180. Tolerances and exemptions
from tolerances for pesticide chemicals in or on
agricultural commodities: aldicarb. Federal Register
46 (224): 57047.
U.S. EPA. 1983. U.S. Environmental Protection
Agency. Occurrence of pesticides in drinking water,
food, and air. Office of Drinking Water.
U.S. EPA. 1984. U.S. Environmental Protection
Agency. Method 531. Measurement of N-methyl
carbamoyloximes and N-methylcarbamates in
drinking water by direct aqueous injection HPLC
with post column derivatization. Environmental
Monitoring and Support Laboratory, Cincinnati,
Ohio 45268.
U.S. EPA. 1985. U.S. Environmental Protection
Agency. Draft health effects criteria document for
aldicarb. Criteria and Standards Division. Office of
Drinking Water.
U.S. EPA. 1986. U.S. Environmental Protection
Agency Guidelines for carcinogen risk assessment.
Federal Register. 51(185)33992-34003. September
24.
U.S. FDA. 1984. U.S. Food and Drug Administration.
Surveillance Index for Pesticides. Bureau of Foods.
Weiden, M.H.J., H.H. Moorefield and L.K. Payne.
1965. o-(Methyl carbamoyl) oximes: A new class of
carbamate insecticides-acaracides. J. Econ. Entomol.
58:154-155.
Weil, C.S. 1969. Purified and technical Temik.
Results of feeding in the diets of rats for one week.
Mellon Institute, unpublished report 32-11, 6 pages.
Weil, C.S 1973. Aldicarb, Seven-day inclusion in diet
of dogs. Carnegie-Mellon Institute of Research,
unpublished report 33-36,4 pages.
Weil, C.S. 1975. Mellon Institute Report No. 35-72,
Section C. EPA Pesticide Petition No. 3F1414.
Weil, C.S. and C.P. Carpenter. 1963. Results of three
months of inclusion of Compound 21149 in the diet of
rats. Mellon Institute, unpublished report 26-47,13
pages.
Weil, C.S. and C.P. Carpenter. 1964. Results of a
three-generation reproduction study on rats fed
Compound 21149 in their diet. Mellon Institute
Report No. 27-158. EPA Pesticide Petition No.
9F0798.
Weil, C.S. and C.P. Carpenter. 1965. Two-Year
feeding of Compound 21149 in the diet of rats. Mellon
Institute, unpublished report 28-123,40 pages.
B-9
-------
Weil, C.S. and C.P. Carpenter. 1966a. Two- Year
feeding of Compound 21149 in the diet of dogs.
Mellon Institute, unpublished report 29-5, 22 pages.
Weil, C.S. and C.P. Carpenter. 1966b. Skin painting
in mice. No reference available.
Weil, C.S. and C.P. Carpenter. 1968a. Temik
sulfoxide. Results of feeding in the diet of rats for six
months and dogs for three months. Mellon Institute
Report No. 31-141. EPA Pesticide Petition No.
9F0798.
Weil, C.S. and C.P. Carpenter. 1968b. Temik sulfone.
Results of feeding in the diet of rats for six months
and dogs for three months. Mellon Institute Report
No. 31-142. EPA Pesticide Petition No. 9F0798.
Weil, C.S. and C.P. Carpenter. 1974. Aldicarb.
Inclusion in the diet of rats for three generations and
a dominant lethal mutagenesis test. Carnegie-Mellon
Institute of Research. Unpublished report 37-90, 46
pages.
III. ATRAZINE
A. General Information and Properties
CAS No.: 1912-24-9
Structural Formula:
ci
Properties: (Meister, 1987; Windholz, 1976)
-AA.t
H
I
H
2-Chloro-4-ethylamino-6-isopropylamino-l,3,5-
triazine
Synonyms:
• AAtrex; Atranex; Crisatrina; Crisazine; Farmco
Atrazine; Griffax; Shell Atrazine Herbicide; Vectal
SC; Gesaprim; Primatol (Meister, 1987).
Uses:
• Atrazine over the past 30 years has been the most
heavily used herbicide in the U.S. It is used for
nonselective weed control on industrial or
noncropped land and selective weed control in
corn, sorghum, sugar cane, pineapple and certain
other plants (Meister, 1987).
C8H14C1N5
215.72
White, odorless,
crystalline solid
175 to 177°C
1.187
3.0 x 10-7 mm Hg
70 mg/L
2.33 to 2.71
Chemical formula
Molecular weight
Physical state
Boiling point (25 mm Hg)
Melting point
Density (20 °C)
Vapor pressure (20°C)
Water solubility (22°C)
Log octanol/Water partition
coefficient
Taste threshold —
Odor threshold —
Conversion factor —
Occurrence
• In a monitoring study of Mississippi River water,
atrazine residues were found at a maximum level
of 17 ppb; residues were detected throughout the
year, with the highest concentrations found in
June or July (Newby and Tweedy, 1976).
• Atrazine has been found in 4,123 of 10,942 surface
water samples analyzed and in 343 of 3,208
ground-water samples (STORET, 1988). Samples
were collected at 1,659 surface water locations and
2,510 ground water locations. The 85th percentile
of all non-zero samples was 2.3 ug/L in surface
water and 1.9 ug/L in ground water sources. The
maximum concentration found in surface water
was 2,300 ug/L and in ground water, 700 ug/L.
Atrazine was found in surface water of 31 states
and in ground water in 13 states. This information
is provided to give a general impression of the
occurrence of this chemical in ground and surface
waters as reported in the STORET database. The
individual data points retrieved were used as they
came from STORET and have not been confirmed
as to their validity. STORET data are often not
valid when individual numbers are used out of the
context of the entire sampling regime, as they are
here. Therefore, this information can only be used
to form an impression of the intensity and location
of sampling for a particular chemical.
• Atrazine has been found also in ground water in
Pennsylvania, Iowa, Nebraska, Wisconsin and
Maryland; typical positives were 0.3 to 3 ppb
(Cohen etal., 1986).
Environmental Fate
• An aerobic soil metabolism study in Lakeland
sandy loam, Hagerstown silty clay loam, and
Wehadkee silt loam soils showed conversion of
atrazine to hydroxyatrazine, after 8 weeks, to be
38,40 and 47% of the amount applied, respectively
(Harris, 1967). Two additional degradates,
B-10
-------
deisopropylated atrazine and deethylated atrazine,
were identified in a sandy loam study (Beynon et
al., 1972).
• Hurle and Kibler (1976) studied the effect of
water-holding capacity on the rate of degradation
and found a half-life for atrazine of more than 125
days, 37 days, and 36 days in sandy soil held at
4%, 35%, and 70% water-holding capacity,
respectively.
• In Oakley sandy loam and Nicollet clay loam,
atrazine had a half-life of 101 and 167 days
(Warnock and Leary, 1978).
• Carbon dioxide production was generally slow in
several anaerobic soils: sandy loam, clay loam,
loamy sand and silt loam (Wolf and Martin, 1975;
Goswami and Green, 1971; Lavy et al., 1973).
• 14C-Atrazine was stable in aerobic ground water
samples incubated for 15 months at 10 or 25°C in
the dark (Weidner, 1974).
• Atrazine is moderately to highly mobile in soils
ranging in texture from clay to gravelly sand as
determined by soil thin layer chromatography
(TLC), column leaching, and adsorption/de-
sorption batch equilibrium studies. Atrazine on
soil TLC plates was intermediately mobile in loam,
sandy clay loam, clay loam, silt loam, silty clay
loam, and silty clay soils, and was mobile in sandy
loam soils. Hydroxyatrazine showed a low mobility
in sandy loam and silty clay loam soils (Helling,
1971).
• Soil adsorption coefficients for atrazine in a
variety of soils were: sandy loam (0.6), gravelly
sand (1.8), silty clay (5.6), clay loam (7.9), sandy
loam (8.7), silty clay loam (11.6), and peat (more
than 21) (Weidner, 1974; Lavy, 1974; Talbert and
Fletchall, 1965).
Soil column studies indicated atrazine was mobile
in sand, fine sandy loam, silt loam and loam;
intermediately mobile in sand, silty clay loam and
sandy loam, low to intermediately mobile in clay
loam (Weidner, 1974; Lavy, 1974; Ivey and
Andrews, 1964; Ivey and Andrews, 1965).
In a Mississippi field study, atrazine in silt loam
oil had a half-life of less than 30 days (Portnoy,
1978). In a loam to silt loam soil in Minnesota,
atrazine phytotoxic residues persisted for more
than 1 year and were detected in the maximum
depth samples (30 to 42 inches) (Darwent and
Behrens, 1968). In Nebraska, phytotoxic residues
persisted in silty clay loam and loam soils 16
months after application of atrazine; they were
found at depths of 12 to 24 inches. Atrazine
phytotoxic residues had a half-life of about 20 days
in Alabama fine sandy loam soil, although
leaching may partially account for this value
(Buchanan and Hiltbold, 1973).
Under aquatic field conditions, dissipation of
atrazine was due to leaching and to dilution by
irrigation water, with residues persisting for 3
years in soil on the sides and bottoms of irrigation
ditches, to the maximum depth sampled, 67.5 to 90
cm (Smith etal., 1975).
Ciba-Geigy (1988) recently submitted comments
on the atrazine Health Advisory. These comments
included a summary of the results of its studies on
the environmental fate of atrazine. This summary
indicated that laboratory degradation studies
showed that atrazine is relatively stable in the
aquatic medium under environmental pH
conditions and indicated that atrazine degraded in
soil by photolysis and microbial processes. The
products of degradation are dealkylated
metabolites, hydroxyatrazine and nonextractable
(bound) residues. Atrazine and the dealkylated
metabolites are relatively mobile, whereas
hydroxyatrazine is immobile.
Ciba-Geigy (1988) also indicated that field
dissipation studies conducted in California,
Minnesota and Tennessee show no leaching of
atrazine and metabolites below 6 to 12 inches of
soil. The half-lives of atrazine in soil ranged
between 20 to 101 days, except in Minnesota where
degradation was slow. A forestry degra-dation
study conducted in Oregon showed no adverse
effects on either terrestrial or aquatic
environments. Also, bioconcentration studies have
shown low potential for bioaccummulation, with a
range of 15 to 77X.
B. Pharmacokinetics
Absorption
Atrazine appears to be readily absorbed from the
gastrointestinal tract of animals. Bakke et al. (1972)
administered single 0.53-mg doses of l4C-ring
labeled atrazine to rats by gavage. Total fecal
excretion after 72 hours was 20.3% of the
administered dosed; the remainder was excreted in
urine (65.5%) or retained in tissues (15.8%). This
indicates that at least 80% of the dose was absorbed.
Distribution
• Bakke et al. (1972) administered single 0.53-mg
doses of 14C-ring-labeled atrazine to rats by
gavage. Liver, kidney and lung contained the
largest amounts of radioactivity, while fat and
muscle had lower residues than the other tissues
examined.
B-11
-------
• In a metabolism study by Ciba-Geigy (1983a), the
radioactivity of 14C-atrazine dermally applied to
Harlan Sprague-Dawley rats at 0.25 mg/kg was
distributed to a minor extent to body tissues. The
highest levels were measured in liver and muscle
at all time points examined; 2.1% of the applied
dose was in muscle and 0.5% in liver at 8 hours.
• Khan and Foster (1976) observed that in chickens
the hydroxy metabolites of atrazine accumulate in
the liver, kidney, heart, and lung. Residues of both
2- chloro and 2-hydroxy moieties were found in
chicken gizzard, intestine, leg muscle, breast
muscle and abdominal fat.
Metabolism
• The principal reactions involved in the metabo-
lism of atrazine are dealkylation at the C-4 and C-
6 positions of the molecule. There is also some
evidence of dechlorination at the C-2 position.
These data were reported by several researchers as
demonstrated below.
* Bakke et al. (1972) administered single 0.53-mg
doses of 14C-ring-labeled atrazine to rats by gav-
age. Less than 0.1% of the label appeared in carbon
dioxide in expired air. Most of the radio-activity
was recovered in the urine (65.5% in 72 hours),
including at least 19 radioactive corn-pounds. More
than 80% of the urinary radio-activity was
identified as 2-hydroxy-atrazine and its two mono-
N-dealkylated metabolites. None of the
metabolites identified contained the 2-chloro
moiety (which may have been removed via
hydrolysis during the isolation technique or by a
dechlorinating enzyme as suggested by the in vitro
studies of Foster et al. (1979), who found evidence
for a dechlorinase in chicken liver homogenates
incubated with atrazine).
* Bohme and Bar (1967) identified five urinary
metabolites of atrazine in rats: the two monode-
alkylated metabolites of atrazine, their carboxy
acid derivatives and the fully dealkylated deriva-
tive. All of these metabolites contained the 2-
chloro group. The in vitro studies of Deuterman
and Muecke (1974) also found no evidence for
dechlorination of atrazine in the presence of rat
liver homogenates.
• Similarly, Brad way and Moseman (1982) admin-
istered atrazine (50,5,0.5, or 0.005 mg/day) for 3
days to male Charles River rats and observed that
the fully dealkylated derivative (2-chloro-4,6-
diamino-s-triazine) was the major urinary
metabolite, with lesser amounts of the two mono-
N-dealkylated derivatives.
• Erickson et al. (1979) dosed Pittman-Moore
miniature pigs by gavage with 0.1 g of atrazine
(SOW). The major compounds identified in the
urine were the parent compound (atrazine) and
deethylated atrazine (which contained the 2-chloro
substituent).
• Hauswirth (1988) indicated that the rat
metabolism studies taken together are sufficient to
show that in the female rat dechlorination of the
triazine ring and N-dealkylation are the major
metabolic pathways. Oxidation of the alky}
substituents appears to be a minor and secondary
metabolic route. The total body half-life is
approximately 11/2 days. Atrazine and/or its
metabolites appear to bind to red blood cells. Other
tissue accumulation does not appear to occur.
Excretion
• Urine appears to be the principal route of atrazine
excretion in animals. Following the
administration of 0.5 mg doses of 14C-ring-labeled
atrazine by gavage to rats, Bakke et al. (1972)
reported that in 72 hours most of the radioactivity
(65.5%) was excreted in the urine, 20.3% was
excreted in the feces, and less than 0.1% appeared
as carbon dioxide in expired air. About 85 to 95% of
the urinary radioactivity appeared within the first
24 hours after dosing, indicating rapid clearance.
• Dauterman and Muecke (1974) have reported that
atrazine metabolites are conjugated with
glutathione to yield a mercapturic acid in the
urine. The studies of Foster et al. (1979) in chicken
liver homogenates also indicate that atrazine
metabolism involves glutathione.
• Ciba-Geigy (1983b) studied the excretion rate of
14C-atrazine from Harlan Sprague-Dawley rats
dermally dosed with atrazine dissolved in
tetrahydrofuran at levels of 0.025, 0.25,2.5 or 5
mg/kg. Urine and feces were collected from all
animals at 24-hour intervals for 144 hours. Results
indicated that atrazine was readily absorbed, and
within 48 hours most of the absorbed dose was
excreted, mainly in the urine and to a lesser extent
in the feces. Cumulative excretion in urine and
feces appeared to be directly proportional to the
administered dose, ranging from 52% at the lowest
dose to 80% at the highest dose.
C. Health Effects
Humans
Short-Term Exposure
• A case of severe contact dermatitis was reported by
Sehlicher and Beat (1972) in a 40-year-old farm
worker exposed to atrazine formulation. The
clinical signs were red, swollen and blistered
hands with hemorrhagic bullae between the
3-12
-------
fingers. Although it is noted that the exposure of
this patient may have been inclusive to exposure
to other chemicals in addition to atrazine, it is also
noted that atrazine is a skin irritant in animal
studies.
Long-Term Exposure
• Yoder et al. (1973) examined chromosomes in
lymphocyte cultures taken from agricultural
workers exposed to herbicides including atrazine.
There were more chromosomal aberrations in the
workers during mid-season exposure to herbicides
than during the off -season (no spraying). These
aberrations included a four- fold increase in
chromatid gaps and a 25-fold increase in
chromatid breaks. During the off-season, the mean
number of gaps and breaks was lower in this group
than in controls who were in occupations unlikely
to involve herbicide exposure. This observation led
the authors to speculate that there is enhanced
chromosomal repair during this period of time
resulting in compensatory protection. However,
these data may not be representative of the effect
of atrazine since the exposed workers were also
exposed to other herbicides.
Animals
Short -Term Exposure
• Acute oral LD50 values of 3,000 mg/kg in rats and
1,750 mg/kg in mice have been reported for
technical atrazine by Bashmurin (1974); the purity
of the test compound was not specified.
• Acute oral studies conducted by Ciba-Geigy (1988)
with atrazine (97% a.i.) reflected the following
LD50s: 1,869 mg/kg in rats and > 3,000 mg/kg in
mice.
• Molnar (1971) reported that when atrazine was
administered by gavage to rats at 3,000 mg/kg, 6%
of the rats died within 6 hours, and 25% of those
remaining died within 24 hours. The rats that died
during the first day exhibited pulmonary edema
with extensive hemorrhagic foci, cardiac dilation
and microscopic hemorrhages in the liver and
spleen. Rats that died during the second day had
hemorrhagic bronchopneumonia and dystrophic
changes of the renal tubular mucosa. Rats
sacrificed after 24 hours had cerebral edema and
histochemical alterations in the lungs, liver and
brain. It is noted that the dose used in this study
was almost 2 x the LD50 (Ciba-Geigy, 1988).
• Gaines and Linder (1986) determined the oral
LDso for adult male and female rats to be 737 and
672 mg/kg respectively and 2,310 mg/kg for pups.
It is, therefore, noted that young animals are more
sensitive to atrazine than adults. This study also
reflected that the dermal LDso f°r adult rats was
higher than 2,500 mg/kg.
• Palmer and Radeleff (1964) administered atrazine
as a fluid dilution or in gelatin capsules to Delaine
sheep and dairy cattle (one animal per dosage
group). Two doses of 250 mg/kg atrazine caused
death in both sheep and cattle. Sixteen doses of 100
mg/kg were lethal to the one sheep tested. At
necropsy, degeneration and discoloration of the
adrenal glands and congestion in lungs, liver and
kidneys were observed.
• Palmer and Radeleff (1969) orally administered
atrazine SOW (analysis of test material not
provided) by capsule or by drench to sheep at 5,10,
25,50,100,250, or 400 mg/kg/day and to cows at
10, 25, 50,100, or 250 mg/kg/day. The number of
animals in each dosage group was not stated, and
the use of controls was not indicated. Observed
effects included muscular spasms, stilted gait and
stance, and anorexia at all dose levels in sheep and
at 25 mg/kg in cattle. Necropsy revealed epicardial
petechiae (small hemorrhagic spots on the lining of
the heart) and congestion of the kidneys, liver and
lungs. Effects appeared to be dose related. A
Lowest-Observed-Adverse-Effect level (LOAEL) of
5 mg/kg/day in sheep and a No-Observed-Adverse-
Effect level (NOAEL) of 10 mg/kg/day in cows can
be identified from this study.
• Bashmurin (1974) reported that oral administra-
tion of 100 mg/kg of atrazine to cats had a hypo-
tensive effect, and that a similar dose in dogs was
antidiuretic and decreased serum cholinesterase
(ChE) activity. No other details of this study were
reported. Atrazine is not an organophosphate (OP);
therefore, its effect on ChE may not be similar to
the mechanism of ChE inhibition by OPs.
Dermal/Ocular Effects
• In a primary dermal irritation test in rats, atra-
zine at 2,800 mg/kg produced erythema but no
systemic effects (Gzheyotskiy et al., 1977).
• Ciba-Geigy (1988) indicated that its studies
reflected dermal sensitization in rats but not
irritation in rabbits' eyes.
Long-Term Exposure
• Hazelton Laboratories (1961) fed atrazine to male
and female rats for 2 years at dietary levels of 0,1,
10, or 100 ppm. Based on the dietary assumptions
of Lehman (1959), these levels correspond to doses
of approximately 0, 0.05, 0.50, or 5.0 mg/kg/day.
After 65 weeks, the 1.0-ppm dose was increased to
1,000 ppm (50 mg/kg/day) for the remainder of the
study. No treatment-related pathology was found
at 26 weeks, at 52 weeks, at 2 years, or in animals
B-13
-------
that died and were necropsied during the study.
Results of blood and urine analyses were
unremarkable. Atrazine had no effects on the
general appearance or behavior of the rats. A
transient roughness of the coat and piloerection
were observed in some animals after 20 weeks of
treatment at the 10- and 100-ppm levels but not at
52 weeks. Body weight gains, food consumption,
and survival were similar in all groups for 18
months, but from 18 to 24 months there was high
mortality due to infections (not attributed to
atrazine) in all groups, including controls, which
limits the usefulness of this study in determining a
NOAEL for the chronic toxicity of atrazine.
• In a 2-year study by Woodard Research
Corporation (1964), atrazine (SOW formulation)
was fed to male and female beagle dogs for 105
weeks at dietary levels of 0,15,150, or 1,500 ppm.
Based on the dietary assumptions of Lehman
(1959), these levels correspond to doses of 0,0.35,
3.5, or 35 mg/kg/day. Survival rates, body weight
gain, food intake, behavior, appearance,
hematologic findings, urinalyses, organ weights,
and histologic changes wore noted. The 15- ppm
dosage (0.35 mg/kg/day) produced no toxicity, but
the 150-ppm dosage (3.5 mg/kg/day) caused a
decrease in food intake as well as increased heart
and liver weight in females. In the group receiving
1,500 ppm (35 mg/kg/day) atrazine, there were
decreases in food intake and body weight gain, an
increase in adrenal weight, a decrease in
hematocrit and occasional tremors or stiffness in
the rear limbs. There were no differences among
the different groups in the histology of the organs
studied. Based on these results, a NOAEL of 0.35
mg/kg/day can be identified for atrazine.
• In a study by Ciba-Geigy (1987b) using technical
atrazine (97% ai.), 6-month-old beagle dogs were
assigned randomly to four dosage groups: 0,15,
150, and 1,000 ppm. These doses correspond to
actual average intake of 0,0.48,4.97 and 33.65
/33.8 (male/female) mg/kg/day. Six animals/sex/
group were assigned to the control and high dose
groups and four animals/sex/group were assigned
to the low- and mid-dose groups. One mid-dose
male, one high-dose male, and one high dose
female had to be sacrificed moribund during the
study period. Decreased body weight gains and
food consumption were noted at the high-dose
level. Statistically significant (p<0.05) reductions
in erythroid parameters (red cell count,
hemoglobin and hematocrit) in high-dose males
were noted throughout the study, as well as mild
increases in platelet counts in both sexes. Slight
decreases in total protein and albumin (p
< 0.05)were noted in high-dose males as well as de-
creased calcium and chloride in males and in-
creased sodium and glucose in females. Decreases
in absolute heart weight were noted in females and
increased relative liver weight in males of the
high-dose group. The mid-dose females reflected
an increase in the absolute heart weight and
heart/brain weight ratios. The most significant
effect of atrazine in this study was reflected in the
high-dose animals of both sexes as discrete
myocardial degeneration. Clinical signs associated
with cardiac pathology such as ascites, cachexia,
labored/shallow breathing and abnormal EKG
were observed in the group as early as 17 weeks
into the study. Gross pathology reflected severe
dilation of the right atrium and occasionally of the
left atrium. These findings were also noted
histopatho-logically as degenerative atrial
myocardium (atrophy and myolysis). In the mid-
dose group, two males and one female appeared to
be affected with the cardiac syndrome but to a
much lesser degree in the intensity of the noted
responses. Therefore, the LOAEL in this study is
4.97 mg/kg/day and the NOAEL is 0.48 mg/kg/day.
A 2-year chronic feeding/oncogenicity study (Ciba-
Geigy, 1986) was recently evaluated by the
Agency. In this study, technical atrazine (98.9%
a.i.) was fed to 37-to 38-day-old Sprague-Dawley
rats. The dosage levels used were 0,10,70, 500, or
1,000 ppm, equivalent to 0, 0.5, 3.5, 25, or 50
mg/kg/day (using Lehman's conversion factor,
1959). Twenty rats per sex per group were used to
measure blood parameters and clinical chemis-
tries and urinalysis. Fifty rats per sex per group
were maintained on the treated and control diets
for 24 months. An additional 10 rats per sex were
placed on control and high dose (1,000 ppm) diets
for a 12-month interim sacrifice and another 10
per sex (control and high dose, 1,000 ppm) for a
13-month sacrifice (the 1,000 ppm group was
placed on control diet for 1 month prior to sacri-
fice). The total number of animals/sex in the
control and HOT groups was 90 and 70 for the 10,
70, and 500 ppm groups. Histopathology was
performed on all animals. At the mid- and high-
dose, there was a decrease in mean body weights of
males and females. Survival was decreased in
high-dose females but increased in high-dose
males. There were decreases in organ-to-body
weight ratios in high-dose animals, which were
probably the result of body weight decreases.
Hyperplastic changes in high-dose males
(mammary gland, bladder and prostate) and
females (myeloid tissue of bone marrow and
transitional epithelium of the kidney) were of
questionable toxicologic importance. There was an
increase in retinal degeneration and in
centrolobular necrosis of the liver in high-dose
females and an in-crease in degeneration of the
rectus femoris muscle in high-dose males and
females when compared to controls. Based on
decreased body weight gain, the LOAEL for
nononcogenic activities in both sexes is 25
mg/kg/day and the NOAEL is 3.5 mg/kg/day.
B-14
-------
However, oncogenic activities were noted at 3.5
mg/kg/day (70 ppm) and above as reflected in the
increased incidence of mammary gland tumors in
females.
• A 91-week oral feeding/oncogenicity study in mice
by Ciba-Geigy (1987c) has been evaluated by the
Agency. In this study, atrazine (97% a, i.) was fed
to 5-week-old CD-I strain of mice, weighing
21.0/26.8 grams (female/male). The mice were
randomly assigned to five experimental groups of
approximately 60 animals/sex/ group. The dosage
tested were 0,10, 300,1,500, and 3,000 ppm; these
dosages correspond to actual mean daily intake of
1.4, 38.4,194.0, and 385.7 mg/kg/day for males,
and 1.6,47.9,246.9, and 482.7 mg/kg/day for
females. This study shows that there are dose-
related effects at 1,500 ppm or 3,000 ppm atrazine:
an increase in cardiac thrombi, a decrease in the
mean body weight gain at 12 and 91 weeks during
the study, and decreases in erythrocyte count,
hematocrit and hemoglobin concentration. Cardiac
thrombi contributed to the deaths of the group of
mice that did not survive to terminal sacrifice. The
LOAEL is set at 1,500-ppm based upon decreases
of 23.5% and 11.0% in mean body weight gain
found at 91 weeks in male and female mice,
respectively. Also, an increase in the incidence of
cardiac thrombi is found in female mice in the
1,500 ppm exposure group. None of the above
effects are found at 300 ppm; thus, the NOAEL is
set at 300 ppm (corresponding to 38.4 mg/kg/day in
males and 47.9 mg/kg/day for females).
Reproductive Effects
• A three-generation study on the effects of atrazine
on reproduction in rats was conducted by Woodard
Research Corporation (1966). Groups of 10 males
and 20 females received atrazine (SOW) at dietary
levels of 0, 50, or 100 ppm. Based on the dietary
assumption that 1 ppm in the diet of rats is
equivalent to 0.05 mg/kg/day (Lehman, 1959),
these levels correspond to doses of approximately
0,2.5, or 5 mg/kg/day. Two litters were produced
per generation, but parental animals were chosen
from the second litter after weaning for each
generation. Young rats were maintained on the
test diets for approximately 10 weeks in each
generation. The third-generation pups were
sacrificed after weaning. It is noted that the
parental animals of the first generation were fed
only half of the dietary atrazine levels for the first
3 weeks of exposure. There were no adverse effects
of atrazine on reproduction observed during the
course of the three-generation study. A NOAEL of
100 ppm (5 mg/kg/day) was identified from this
study. However, the usefulness of this study is
limited due to the alteration of the atrazine
content of the diet during important maturation
periods of the neonates.
• A recent two-generation study in rats by Ciba-
Geigy (1987a) was conducted using the 97% a.i.
technical atrazine. Young rats, 47 to 48 days old,
were maintained on the control and test diets for
10 weeks before mating. The concentrations used
were 0,10, 50, and 500 ppm (equivalent to 0, 0.5,
2.5, and 25 mg/kg/day using Lehman conversion
factor, 1959). Thirty animals/sex/group were used
in each generation; one litter was produced per
generation. The level tested had no effect on
mortality in either generation. Body weight and
body weight gains were significantly depressed
(p<0.05) at the highest dose; however, food
consumption was also decreased at this high-dose
level in parental males and females during the
premating period and for the first generation
females (Fi) on days 0 to 7 of gestation. Neither
histopathological nor other effects were noted
during gross necropsy in either parental
generation, with the exception of increased
relative weight of testes in both generations at
the high dose. In pups of both generation,
significant reduction (p < 0.05) in body weight was
noted; however, this effect was only dose-related in
the second generation (F2) at both the mid- and
high-dose levels on postnatal day 21. Therefore,
maternal toxicity NOAEL is 2.5 mg/kg/day; the
reproductive LOAEL is 2.5 mg/kg/day (reduced
pup weight in F2 generation on postnatal day 21),
and the NOAEL is 0.5 mg/kg/day.
Developmental Effects
• In the three-generation reproduction study in rats
conducted by Woodard Research Corporation
(1966) (described above), atrazine at dietary levels
of 50 or 100 ppm (2.5 or 5 mg/kg/day) resulted in no
observed histologic changes in the weanlings and
no effects on fetal resorption. No malformations
were observed, and weanling organ weights were
similar in controls and atrazine-treated animals.
Therefore, a NOAEL of 100 ppm (5 mg/kg/day) was
also identified for developmental effects in this
study. However, the usefulness of this study is
limited due to an alteration of the atrazine content
of the diet during important maturation periods of
the neonates.
• Atrazine was administered orally to pregnant rats
on gestation days 6 to 15 at 0,100, 500 or 1,000
mg/kg (Ciba-Geigy, 1971). The two higher doses
increased the number of embryonic and fetal
deaths, decreased the mean weights of the fetuses
and retarded the skeletal development. No
teratogenic effects were observed. The highest dose
(1,000 mg/kg) resulted in 23% maternal mortality
and various toxic symptoms. The 100 mg/kg dose
had no effect on either dams or embryos and is
therefore the maternal and fetotoxic NOAEL in
this study.
B-15
-------
• In a study by Ciba-Geigy (1984a), Charles River
rats received atrazine (97%) by gavage on
gestation days 6 to 15 at dose levels of 0,10,70, or
700 mg/kg/day. Excessive maternal mortality
(21/27) was noted at 700 mg/kg/day, but no
mortality was noted at the lower doses; also,
reduced weight gains and food consumption were
noted at both 70 and 700 mg/kg/day.
Developmental toxicity was also present at these
dose levels. Fetal weights were severely reduced at
700 mg/kg/day; delays in skeletal development
occurred at 70 mg/kg/ day, and a dose-related
runting was noted at 10 mg/kg/day and above. The
NOAELfor maternal toxicity appears to be 10
mg/kg/day; however, this is also the LOAEL for
developmental effects.
• New Zealand white rabbits received atrazine
(96%) by gavage on gestation days 7 through 19 at
dose levels of 0,1,5, or 75 mg/kg/day (Ciba-Geigy,
1984b). Maternal toxicity, evidenced by decreased
body weight gains and food consumption, was
present in the mid- and high-dose groups.
Developmental toxicity was demonstrated only at
75 mg/kg/day by an increased resorption rate,
reduced fetal weights, and delays in ossification.
No teratogenic effects were indicated. The NOAEL
appears to be 1 mg/kg/day.
• Peters and Cook (1973) fed atrazine to pregnant
rats (four/group) at levels of 0,50,100,200,300,
400,500, or 1,000 ppm in the diet throughout
gestation. Based on an assumed body weight of 300
g and a daily food consumption of 12 g (Arrington,
1972), these levels correspond to approximately 0,
2,4,8,12,16,20, or 40 mg/kg/day. The number of
pups per litter was similar in all groups, and there
were no differences in weanling weights. This
study identified a NOAEL of 40 mg/kg/day for
developmental effects. In another phase of this
study, the authors demonstrated that
subcutaneous (sc) injections of 50,100, or 200
mg/kg of atrazine on gestation days 3,6 and 9 had
no effect on the litter size, while doses of ^ 800
mg/kg were embryotoxic. Therefore, a NOAEL of
200 mg/kg by the sc route was identified for
embryotoxicity.
Mutagenicity
• Loprieno et al. (1980) reported that single doses of
atrazine (1,000 mg/kg or 2,000 mg/kg, route not
specified) produced bone marrow chromsomal
aberrations in the mouse. No other details of this
study were provided.
• Murnik and Nash (1977) reported that feeding
0.01% atrazine to male Drosophila melanogaster
larvae significantly increased the rate of both
dominant and sex-linked recessive lethal
mutations. They stated, however, that dominant
lethal induction and genetic damage may not be
directly related.
• Adler (1980) reviewed unpublished work on
atrazine mutagenicity carried out by the
Environmental Research Programme of the
Commission of the European Communities.
Mutagenic activity was not induced even when
mammalian liver enzymes (5-9) were used;
however, the use of plant microsomes produced
positive results. Also, in vivo studies in mice,
atrazine induced dominant lethal mutations and
increased the frequency of chromatid breaks in
bone marrow. Hence, the author suggested that
activation of atrazine in mammals occurs
independently of the liver, possibly in the acidic
part of the stomach.
• As described previously, Yoder et al. (1973)
studied chromosomal aberrations in the
lymphocyte cultures of farm workers exposed to
various pesticides including atrazine. During mid-
season a 4-fold increase in chromatid gaps and a
25-fold increase in chromatid breaks was observed.
During the off-season (no spraying), the number of
gaps and breaks was lower than in controls,
suggesting that chromosomal repair is enhanced
during the unexposed period.
• Recently, Spencer (1987) and Dearfield (1988)
evaluated several in vitro and in vivo mutagenicity
studies on atrazine that were recently submitted to
the U.S. EPA by Ciba-Geigy. They noted that most
of these studies were inadequate with the
exception of the following three tests: a Salmonella
assay; an E. coli reversion assay, and a host-
mediated assay. The first two assays were negative
for mutagenic effects; the results of the third assay
were equivocal.
• Ciba-Geigy (1988) indicated that Brusick (1987)
evaluated atrazine mutagenicity and that the
weight-of-evidence analysis he used placed the
chemical in a nonmutagenic status. The Agency
(Dearfield, 1988) evaluated Brusick's analysisand
determined that using the weight-of-evidence
approach is not appropriate at the present time.
The in vivo studies by Adler (1980) suggest a
positive response. These findings have not been
diminished by other atrazine studies. In addition,
Dearfield (1988) indicated that the scheme used by
Brusick in this analysis is flawed by the lack of
calibration of the chemical test scores to an
external standard and by the use of some studies
that are considered inadequate by design to
determine the mutagenic potential of atrazine.
Carcinogenicity
• Innes et al. (1969) investigated the tumorigenicity
of 120 test compounds, including atrazine, in mice.
B-16
-------
Two FI hybrid stocks (C57BL/6 x Anf) Fa and
(C57BL/6 x AKR) FI were used. A dose of 21.5
mg/kg/day was administered by gavage to mice of
both sexes from age 7 to 28 days. After weaning at
4 weeks, this dose level was maintained by feeding
82 ppm atrazine ad libitum in the diet for 18
months. The incidence of hepatomas, pulmonary
tumors, lymphomas and total tumors in atrazine-
treated mice was not significantly different from
that in the negative controls.
A 2-year feeding/oncogenicity study in rats by
Ciba-Geigy (1986) has been evaluated recently by
the Agency. Atrazine (98.9% a.i.) was fed to 37-to
38-day-old Sprague-Dawley rats. The dosage levels
used were 0,10,70, 500, or 1,000 ppm, equivalent
to 0, 0.5,3.5, 25, or 50 mg/kg/day (using Lehman's
conversion factor, 1959). The total number of
animals/sex in the control and HDT groups was 90;
and 70 animals/sex/group for the 10,70 and 500
ppm groups. Histopathology was performed on all
animals. In females, atrazine was associated with
a statistically significant increase in mammary
gland fibroadenomas at 1,000 ppm; in mammary
gland adenocarcinomas (including two
carcinosarcomas at the HDT) at 70, 500, and 1,000
ppm, and in total mammary gland tumor bearing
animals at 1,000 ppm. Each of these increases was
associated with a statistically significant dose-
related trend and was outside of the high end of the
historical control range. In addition, EPA (1986a)
indicated that there was evidence for decreased
latency for mammary gland adenocarcinomas at
the 12- month interim sacrifice that was already
submitted by Ciba-Geigy in 1985. This study was
also reported as positive in a briefing paper by
Ciba-Geigy (1987).
> A recent 91-week oral feeding/oncogenicity study
in mice by Ciba-Geigy (1987c) has been evaluated
by the Agency. In this study, atrazine (97% a. i.)
was fed to 5-week-old CD-I mice weighing
21.0/26.8 g (female/male). The mice were randomly
assigned to five experimental groups of
approximately 60 animals/sex/ group. The dosage
tested were 0,10, 300,1,500, and 3,000 ppm: these
dosages correspond to actual mean daily intake of
1.4, 38.4,194.0, and 385.7 mg/kg/day for males,
and 1.6,47.9,246.9, and 482.7 mg/kg/day for
females. The following kinds of neoplasms were
noted in this study: mammary adenocarcinomas,
adrenal adenomas, pulmonary adenomas, and
malignant lymphomas. However, no dose-related
or statistically significant increases were observed
in the incidences of these neoplasms. Therefore,
atrazine is not considered oncogenic in this strain
of mice.
D. Quantification of lexicological
Effects ;
The HAs for noncarcinogenic toxicants are derived
using the following formula:
HA =
(NOAEL or LOAEL) x (BW)
(UF)x(_L/day)
_mg/L ( ug/L)
where:
NOAEL or
LOAEL
BW
UF
L/day
No- or Lowest-Observed-
Adverse-Effect level in mg/kg
bw/day.
assumed body weight of a child (10
kg) or an adult (70 kg).
uncertainty factor (10,100 or 1,000),
in accordance with NAS/ODW
assumed daily water consumption of
a child (1 L/day) or an adult (2 L/day).
One-Day Health Advisory
No suitable information was found in the available
literature for the determination of the One-Day HA
value for atrazine. It is, therefore, recommended that
the Ten-Day HA value calculated below for a 10-kg
child of 0.1 mg/L (100 ug/L), be used at this time as a
conservative estimate of the One-Day HA value.
Ten-Day Health Advisory
Two teratology studies by Ciba-Geigy, one in the rat
(1984a) and one in the rabbit (1984b), were
considered for the calculation of the Ten-Day HA
value. The rat study reflected a NOAEL of 10
mg/kg/day for maternal toxicity, but this value was
also the LOAEL for developmental toxicity. The
rabbit study reflected a NOAEL of 5 mg/kg/day for
developmental toxicity and 1 mg/kg/day for maternal
toxicity. Thus, the rabbit appears to be a more
sensitive species than the rat for maternal toxicity.
Hence, the rabbit study with a NOAEL of 1
mg/kg/day is used in the calculations below.
The Ten-Day HA for a 10-kg child is calculated below
as follows:
(lmg/kg/d)x(Wkg)
(100)x(l L/day)
= 0.1 mgL(WQug/L)
where:
1 mg/kg/day = NOAEL, based on maternal
toxicity evidenced by decreased
body weight gain and food
consumption.
B-17
-------
10kg
100
IL/day
assumed body weight of a child.
uncertainty factor, chosen in
accordance with EPA or
ODW/NAS guidelines for use
with a NOAEL from an animal
study.
assumed daily consumption for a
child.
Longer-Term Health Advisory
No suitable information was found in the available
literature for the determination of the Longer-Term
HA value for atrazine. It is, therefore, recommended
that the adjusted DWEL for a 10-kg child of 0.05
mg/L (50 ug/L) and the DWEL for a 70-kg adult of 0.2
mg/L (200 ug/L) be used at this time as conservative
estimates of the Longer-Term HA value.
Lifetime Health Advisory
The Lifetime HA represents that portion of an
individual's total exposure that is attributed to
drinking water and is considered protective of
noncarcinogenic adverse health effects over a
lifetime exposure. The Lifetime HA is derived in a
three-step process. Step 1 determines the Reference
Dose (RfD), formerly called the Acceptable Daily
Intake (ADI). The RfD is an estimate of a daily
exposure to the human population that is likely to be
without appreciable risk of deleterious effects over a
lifetime, and is derived from the NOAEL (or
LOAEL), identified from a chronic (or subchronic)
study, divided by an uncertainty factor .(s). From the
RfD, a Drinking Water Equivalent Level (DWEL)
can be determined (Step 2). A DWEL is a medium-
specific (i.e., drinking water) lifetime exposure level,
assuming 100% exposure from that medium, at
which adverse, noncarcinogenic health effects would
not be expected to occur. The DWEL is derived from
the multiplication of the RfD by the assumed body
weight of an adult and divided by the assumed daily
water consumption of an adult. The Lifetime HA is
determined in Step 3 by factoring in other sources of
exposure, the relative source contribution (RSC). The
RSC from drinking water is based on actual exposure
data or, if data are not available, a value of 20% is
assumed. If the contaminant is classified as a Group
A or B carcinogen, according to the Agency's
classification scheme of carcinogenic potential (U.S.
EPA, 1986b), then caution should be exercised in
assessing the risks associated with lifetime exposure
to this chemical.
Three studies were considered for the development of
the Lifetime HA. A 2-year dog feeding study
(Woodard, 1964), a 1-year dog feeding study Ciba-
Geigy, 1987b) and a 2-year rat oral
feeding/oncogenicity study (Ciba-Geigy, 1986).
The first study in dogs (1964) reflected a NOAEL of
0.35 mg/kg/day and a LOAEL of 3.5 mg/kg/day that
was associated with increased heart and liver
weights in females. The new 1 year dog study (1988)
reflected a NOAEL of 0.48 mg/kg/day and a LOAEL
of 4.97 mg/kg/day based on mild cardiac pathology
intensified at the higher dose tested 33.65/33.8
(male/female) mg/kg/day. The 2 year rat study (Ciba-
Geigy, 1986) reflected a NOAEL at 3.5 mg/kg/day for
systemic effects other than oncogenicity; however,
this study indicated hat atrazine caused mammary
gland tumors at this dose level and above, no adverse
effects were observed at the lowest dose tested, 0.5
mg/kg/day.
The 1964 dog study was initially used for the
calculation of the RfD and the Lifetime HA.
However, this study was partially flawed by the lack
of information on the purity of the test material and
by the inadequate documentation of the
hematological data. Therefore, the recent 1-year dog
study (Ciba-Geigy, 1987b), using technical atrazine
(97% a. i.), is considered a more adequate study for
the calculation of the RfD and the Lifetime HA. The
NOAEL in this study, 0.48 mg/kg/day, is also
supported by the NOAEL of 0.5 mg/kg/day in the 2
generation reproduction study (Ciba-Geigy, 1987a)
and by the fact that no systemic effects or tumors
were noted at this dose level in the two-year chronic
feeding/oncogenicity study in rats (Ciba-Geigy,
1986). [Other studies: Woodard Research Corporation
(1966) and Hazelton Laboratories (1961) identified
long-term NOAEL values of 5 to 50 mg/kg/day and
were not considered to be as protective as the dog
studies for use in calculating the HA values for
atrazine.]
Step 1: Determination of the RfD
0.48 mg/kg/day
(100)
= 0.005 mg/hg/day
(rounded from 0.0048
mg/kg/day)
where:
0.48 mg
/kg/day
100
NOAEL, based on the
absence of cardiac pathology or any
other/adverse clinical, hematological,
biochemical and hlstopathological
effects in dogs.
uncertainty factor, chosen in
accordance with EPA or NAS/ODW
guidelines for use with a NOAEL
from an animal study.
B-18
-------
Step 2: Determination of the DWEL
0.0048 mglkglday (70-kg)
DWEL =
(2 LI day)
where:
0.0048 mg
/kg/day
70kg
2L/day
= 0.168 mg/L (200ug/L)
= RfD (before rounding off
to 0.005 mg/kg/day).
= assumed body weight of an adult.
= assumed daily water
consumption of an adult.
Step 3: Determination of the Lifetime Health
Advisory
Lifetime HA = (0.168 mg/L)(20%)/10
= 0.003 mg/L (3 ug/L)
where:
0.168 mg/L
20%
10
= DWEL (before rounding off to 0.2
mg/L)
= assumed relative source
contribution from water.
— additional uncertainty factor,
according to ODW policy, to
account for possible
carcinogenicity.
Evaluation of Carcinogenic Potential
• A study submitted by Ciba-Geigy Corporation
(1986) in support of the pesticide registration of
atrazine indicated that atrazine induced an
increased incidence of mammary tumors in female
Sprague-Dawley rats. These findings have been
further confirmed in a briefing by Ciba-Geigy
(1987) on this study.
• Atrazine was not oncogenic in mice (Ciba-Geigy,
1987c).
• Three closely related analogs —propazine,
terbutryn and simazine — are presently classified
as Group C oncogens based on an increased
incidence of tumors in the same target tissue
(mammary gland) and animal species (rat) as was
noted for atrazine
• The International Agency for Research on Cancer
has not evaluated the carcinogenic potential of
atrazine.
• Applying the criteria described in EPA's
guidelines for assessment of carcinogenic risk
(U.S. EPA, 1986b), atrazine may be classified in
Group C: possible human carcinogen. This
category is used for substances with limited
evidence of carcinogenicity in animals in the
absence of human data.
E. Other Criteria, Guidance, and
Standards
• Toxicity data on atrazine were reviewed by the
National Academy of Sciences (NAS, 1977), and
the study by Innes et al. (1969) was used to identify
a chronic NOAEL of 21.5 mg/kg/day. Although at
that time it was concluded that atrazine has low
chronic toxicity, an uncertainty factor of 1,000 was
employed in calculation of the ADI from that,
study, since only limited data were available. The
resulting value (0.021 mg/kg/day) corresponds to
an ADI of 0.73 mg/L in a 70-kg adult consuming 2
L of water per day.
• Tolerances for atrazine alone and the combined
residues of atrazine and its metabolites in or on
various raw agricultural commodities have been
established (U.S. EPA, 1986c). These tolerances
range from 0.02 ppm (negligible) in animal
products (meat and meat by-products) to 15 ppm in
various animal fodders.
F. Analytical Methods
• Analysis of atrazine is by a gas chromatographic
(GC) method, Method No. 507, applicable to the
determination of certain nitrogen-phosphorus
containing pesticides in water samples (U.S. EPA,
1988). In this method, approximately 1 L of sample
is extracted with methylene chloride. The extract
is concentrated and the compounds are separated
using capillary column CC. Measurement is made
using a nitrogen phosphorus detector. The method
has been validated in a single laboratory. The
estimated detection limit for the analytes in this
method, including atrazine, is 0.13 ug/L.
G. Treatment Technologies
• Treatment technologies that remove atrazine from
water include activated carbon adsorption, ion
exchange, reverse osmosis, ozone oxidation, and
ultraviolet irradiation. Conventional treatment
methods are ineffective for the removal of atrazine
from drinking water (ESE, 1984; Miltner and
Fronk, 1985a). Limited data suggest that aeration
would not be effective in atrazine removal (ESE,
1984; Miltner and Fronk, 1985a).
• Baker (1983) reported that a 16.5-incb C! AC filter
cap using F-300, which was placed on the rapid
sand filters at the Fremont, Ohio water treat-ment
plant, reduced atrazine levels by 30 to 64% in the
water from the Sandusky River. At Jefferson
B-19
-------
Parish, Louisiana, Lykins et al. (1984) reported
that an adsorber containing 30 inches of Westvaco
WV-G 12 x 40 GAG removed atrazine to levels
below detectable limits for over 190 days.
• At the Bowling Green, Ohio, water treatment
plant, PAC in combination with conventional
treatment achieved an average reduction of 41% of
the atrazine in the water from the Maumee River
(Baker, 1983). Miltner and Fronk (1985a) reported
that in jar tests using spiked Ohio River water
with the addition of 16.7 and 33.3 mg/L of PAC and
15-20 mg/L of alum, PAC removed 64 and 84%,
respectively, of the atrazine. Higher percent
removals reflected higher PAC dosages. Miltner
and Fronk (1985b) monitored atrazine levels at
water treatment plants, which utilized PAC, in
Bowling Green and Tiffin, Ohio. Applied at
dosages ranging from 3.6 to 33 mg/L, the PAC
achieved 31 to 91% removal of atrazine, with
higher percent removals again reflecting higher
PAC dosages.
• Harris and Warren (1964) reported that Amber-
lite IR-120 cation exchange resin removed atrazine
from aqueous solution to less than detectable
levels. Turner and Adams (1968) studied the effect
of varying pH on different cation and anion
exchange resins. At a pH of 7.2,45% removal of
atrazine was achieved with Dowex® 2 anion
exchange resin and with H2P04-as the
exchangeable ion species.
• Chian et al. (1975) reported that reverse osmosis,
utilizing cellulose acetate membrane and a cross-
linked polyethelenimine (NS-100) membrane,
successfully processed 40% of the test solution,
removing 84 and 98%, respectively, of the atrazine
in the solution.
• Miltner and Fronk (1985a) studied the oxidation of
atrazine with ozone in both spiked distilled and
ground water. Varying doses of ozone achieved a
70% removal of atrazine in distilled water and 49
to 76% removal of atrazine in ground water.
• Kahn et al. (1978) studied the effect of fulvic acid
upon the photo-chemical stability of atrazine to
ultraviolet irradiation. A 50% removal of atrazine
was achieved much faster at higher pH conditions
than at lower pH conditions. In the presence of
fulvic acids, the time needed for ultraviolet
irradiation to achieve 50% removal was almost
triple the time required to achieve similar
removals without the presence of fulvic acids.
Since fulvic acids will be present in surface waters,
ultraviolet irradiation may not be a cost-effective
treatment alternative. !
H. References
Adler, I.D. 1980. A review of the coordinated research
effort on the comparison of test systems for the
detection of mutagenic effects, sponsored by the
E.E.C. Mutat. Res. 74:77-93.
Arrington, L. R. 1972. The laboratory animals. In:
Introductory laboratory animal science. The
breeding, care and management of experimental
animals. Danville, IL: Interstate Printers and
Publishers, Inc., pp. 9-11.
Baker, D. 1983. Herbicide contamination in
municipal water supplies in northwestern Ohio.
Final Draft Report 1983. Prepared for Great Lakes
National Program Office, U.S. Environmental
Protection Agency. Tiffin, OH.
Bakke, J. E., J. D. Larson and C.E. Price. 1972.
Metabolism of atrazine and 2- hydroxyatrazine by the
rat. J. Agric. Food Chem. 20:602-607,
Bashmurin, A. F. 1974. Toxicity of atrazine for
animals. Sb. Rab. Leningrad Vet. Institute. 36:5-7.
(English abstract only)
Beynon, K. L, G. Stoydin and A.N. Wright. 1972. A
comparison of the breakdown of the triazine
herbicides cyanazine, atrazine and simazine in soils
and in maize. Pestic. Biochem. Physiol. 2:153-161,
Bohme, E., and F. Bar. 1967. Uber den Abbau von
Triazin-Herbiciden in tierischen Organismus. Food
Cosmet. Toxicol. 5:23-28. (English abstract only)
Bradway, D.E., and R.F. Moseman. 1982.
Determination of urinary residue levels of the n-
dealkyl metabolites of triazine herbicides. J. Agric.
Food Chem. 30:244-247.
Brusick, D.J. 1987. An assessment of the genetic
toxicity of atrazine: relevance to health and
environmental effects. A document prepared for
Ciba-Geigy Corporation (submitted to EPA as a part
of Ciba-Geigy comments on the HA). December 1988.
Buchanan, G.A., and A.E. Hiltbold. 1973.
Performance and persistence of atrazine. Weed Sci.
21:413-416.
Chian, E.S.K., W.N. Bruce and H.H.P. Fang. 1975.
Removal of pesticides by reverse osmosis.
Environmental Science and Technology. 9(l):52-59.
Ciba-Geigy. 1971. Rat reproduction study-test for
teratogenic or embryotoxic effects. 10/1971;
Teratology study of atrazine technical in Charles
River rats 9/1984, SCDFA, Sacramento.
B-20
-------
Ciba-Geigy. 1983a. Dermal absorption of 14C-
atrazine by rats. Ciba-Geigy Corporation, Greens-
boro, NC. Report No. ABR-83005, May, 1983.
Accession No. 255815.
Ciba-Geigy. 1983b. Excretion rate of 14C-atrazine
from dermally dosed rats. Ciba-Geigy Corporation,
Greensboro, NC. Report No. ABR-83081, October,
1983. Accession No. 255815.
Ciba-Geigy Ltd. 1984a. A teratology study of atrazine
technical in Charles River rats: Toxicology/pathology
report No. 60-84. MAID 00143008.
Ciba-Geigy Ltd. 1984b. Segment II. Teratology study
in rabbits: Toxicology/pathology report No. 68-84.
MRID 00143006.
Ciba-Geigy. 1985. Atrazine chronic feeding/oncogen-
icity study. One-year interim report. May 17,1985.
Ciba-Geigy. 1986. Twenty-four month combined
chronic oral toxicity and oncogenicity in rats
utilizing atrazine technical by American Biogenic
Corp. Study No. 410-1102. Accession Nos. 262714-
262727.
Ciba-Geigy. 1987. Briefing paper on atrazine.
December, 1986. Analysis of chronic rat feeding
study results. Ciba-Geigy Corp., Greensboro, NC.
Ciba-Geigy. 1987a. Two-generation rat reproduction.
Study No. 852063. MRID 404313-03.
Ciba-Geigy. 1987b. Atrazine technical — 52-week oral
feeding in dogs. Study No. 852008 and Pathology
Report No. 7048. MRID 40313-01.
Ciba-Geigy. 1987c. Atrazine technical—91-week oral
carcinogenicity study in mice. Study No. 842120.
MRID 404313-02.
Ciba-Geigy. 1988. Comments on the atrazine draft
health advisory. A letter from Thomas Parish to U.S.
EPA/ODW.
Cohen, S.Z., C. Eiden and M.N. Lorber. 1986.
Monitoring ground water for pesticides in the U.S.A.
In Evaluation of pesticides in ground water.
American Chemical Society Symposium Series. No.
315.
Cosmopolitan Laboratories. 1979. CBI, Document
No. 00541, EPA Accession No. 2-41725.
Darwent, A.L., and R. Behrens. 1968. Dissipation
and leaching of atrazine in a Minnesota soil after
repeated applications. In Proc. North Cent. Weed
Control Conf., December 3-5,1968, Indiana, pp. 66-
68.
Dauterman, W.C., and W. Muecke. 1974. In vitro
metabolism of atrazine by rat liver. Pestic. Biochem.
Physiol. 4:212-219
Dearfield, K. L. 1988. An assessment of the genetic
toxicity of atrazine; review of submitted studies and
document prepared by D. Brusick for Ciba-Geigy. A
memo (including an executive summary) from U.S.
EPA, Office of Pesticide Programs. April 26.
ESE. 1984. Environmental Science and Engineering.
Review of treatability data for removal of 25
synthetic organic chemicals from drinking water.
Environmental Protection Agency, Office of
Drinking Water, Washington, DC.
Erickson, N.D., C.W. Frank and D. P. Morgan. 1979.
Determination of s-triazine herbicide residues in
urine: Studies of excretion and metabolism in swine
as a model to human metabolism. J. Agric. Food
Chem. 27:743-745,
Foster, T.S., S.U. Khan and N.H. Akhtar. 1979.
Metabolism of atrazine by the soluble fraction
(105,000 g) from chicken liver homogenates.
J. Agric. Food Chem. 17:300-302.
Gaines T.B. and R.E. Linder. 1986. Acute toxicity of
pesticides in adult and weanling rats. Fundam. Appl.
Toxicol. 7:299-308.
Goswami, K.P., and R.E. Green. 1971. Microbial
degradation of the herbicide atrazine and its 2-
hydroxy analog.
Gzhegotskiy, M.I., L.V. Shklraruk and L.A. Dychok.
1977. Toxicological characteristics of the herbicide
zeazin. Vrach. Delo 5:133-136, In Pesticides Abstract
10:711-712,1977.
Harris, C.I., and G.F. Warren. 1964. Adsorption and
desorption of herbicides by soil. Weeds. 12:120-126.
Harris, C.I. 1967. Fate of 2-chloro-s-triazine
herbicides in soil. J. Agric. Food Chem. 15:157-162.
Hauswirth, J. W. 1988. Summary on some atrazine
toxicity studies submitted Ciba-Geigy (including
metabolism studies No. ABR-87116, 87048, 87087,
85104,87115 and AG-520). Memo from U.S. EPA,
Office of Pesticide Programs. May 3.
Hayes, W.J.,Jr. 1982. Pesticides Studied in Man.
Baltimore, MD: Williams and Wilkins.
Hazelton Laboratories. 1961. Two-year chronic
feeding study in rats. CBI, Document No. 000525,
MRID 0059211.
B-21
-------
Helling, C.S. 1971. Pesticide mobility in soils. II.
Applications of soil thin-layer chromatography. Proc.
Soil Sci. Soc. Am. 35:737-748.
Hurle, K., and E. Kibler. 1976. The effect of changing
moisture conditions on the degradation,of atrazine in
soil. Proceedings of the British Crop Protection
Conference-Weeds. 2:627-633,
Innes, J.R.N., B.M. Ulland, and M.G. Valerio. 1969.
Bioassay of pesticides and industrial chemicals for
tumorigenicity in mice: A preliminary note. J. A.
Cancer Inst. 42:1101-1114.
Ivey, N.J., and H. Andrews. 1964. Leaching of
simazine, atrazine, diruon, and DCPA in soil
columns. Unpublished study submitted by Ciba-
Geigy, Greensboro, N.C.
Ivey, N. J., and H. Andrews. 1965. Leaching of
simazine, atrazine, diruon, and DCPA in soil
columns. Unpublished study prepared by University
of Tennessee, submitted by American Carbonyl, Inc.,
Tenafly, NJ.
Khan, S.U., and T.S. Foster. 1976. Residues of
atrazine (2-chloro-4-ethyl-amino- 6- isopropylamino-
s-triazine) and its metabolites in chicken tissues. J.
Agric. Food Chem. 24:768-771.
Khan, S. U., and M. Schnitzer. 1978. UV Irradiation
of atrazine in aqueous fulvic acid solution.
Environmental Science and Health. B13.-299-310.
Lavy, T.L. 1974. Mobility and deactivation of
herbicides in soil-water systems: Project A-024-NEB.
Available from National Technical Information
Service, Springfield, VA; PB-238-632.
Lavy. T.L., F.W. Roeth and C.R. Fenster. 1973.
Degradation of 2,4-D and atrazine at three soil depths
in the field. J. Environ. Qual. 2:132-137.
Lehman, A. J. 1959. Appraisal of the safety of
chemicals in foods, drugs and cosmetics. Assoc. Food
and Drug U.S., Q.Bull.
Loprieno, N., R. Barale, L. Mariani, S. Presciuttini,
A.M. Rossi, I. Shrana, L. Zaccaro, A. Abbondandolo
and S. Bonatti. 1980. Results of mutagenicity tests on
the herbicide atrazine. Mutat. Res. 74:250. Abstract.
Lykins, Jr., B.W., E. E. Geldreich, J. Q. Adams, J.C.
Ireland and R.M. Clark. 1984. Granular activated
carbon for removing nontrihalomethane organics
from drinking water. U.S. Environmental Protection
Agency, Office of Research and Development,
Municipal Environmental Research Laboratory,
Cincinnati, OH.
Meister, R.G., ed. 1987. Farm chemicals handbook.
3rd ed. Willoughby, OH: Meister Publishing Co.
Miltner, R. J., and C.A. Fronk. 1985a. Treatment of
synthetic organic contaminants for Phase II
regulations. Progress report. U.S. Environmental
Protection Agency, Drinking Water Research
Division. July 1985.
Miltner, R.J., and C.A. Fronk. 1985b. Treatment of
synthetic organic contaminants for Phase II
regulations. Internal report. U.S. Environmental
Protection Agency, Drinking Water Research
Division. December 1985.
Molnar, V. 1971. Symptomatology and
pathomorphology of experimental poisoning with
atrazine. Rev. Med. 17:271-274. (English abstract
only).
Murnik, M.R., and C .L. Nash. 1977. Mutagenicity of
the triazine herbicides atrazine, cyanazine, and
simazine in Drosophila melanogastar. J. Toxicpl.
Environ. Health. 3:691-697.
NAS. 1977. National Academy of Sciences. Drinking
Water and Health, Washington, DC: National
Academy Press, pp. 533-539.
Newby, L., and B.G. Tweedy. 1976. Atrazine residues
in major rivers and tributaries. Unpublished study
submitted by Ciba-Geigy Corporation, Greensboro,
N.C.
Palmer, J.S., and R.D. Radeleff. 1964. The
toxicological effects of certain fungicides and
herbicides on sheep and cattle. Ann. N.Y. Acad. Sci.
111:729-736,
Palmer, J.S., and R.D. Radeleff. 1969. The toxicity of
some organic herbicides to cattle, sheep and chickens.
Production Research Report No. 106.
U.S.Department of Agriculture, Agricultural
Research Service: 1-26.
Peters, J.W., and R.M. Cook. 1973. Effects of atrazine,
on reproduction in rats. Bull. Environ. Contam.
Toxicol. 9:301-304.
Portnoy, C.E. 1978. Disappearance of bentazon and
atrazine in silt loam soil. Unpublished study
submitted by BASF Wyandotte Corporation,
Parsippany, NJ.
Schlicher, J. E., and V. B. Beat. 1972. Dermatitis
resulting from herbicide use — A case study. J. Iowa
Med. Soc. 62:419-420,
Smith, A. E., R. Grover, G. S. Emmond and H.C.
Korven. 1975. Persistence and movement of atrazine,
B-22
-------
bromacil, monuron, and simazine in intermittently-
tilled irrigation ditches. Can. J. Plant Sol. 55:809-
816.
Spencer, H. 1987. Review of several mutagenicity
studies on atrazone.
U.S. EPA, Office of Pesticide Programs' review of a
Ciba-Geigy data submission. Accession No. 284052.
MRID 402466-01.
STORET. 1988. STORET Water Quality File. Office
of Water. U.S. Environmental Protection Agency.
(Data file search conducted in March, 1988).
Talbert, R.E., and O.H. Fletchall. 1965. The
adsorption of some S-triazines in soils. Weeds. 13:46-
52.
Turner, M.A., and R.S. Adams, Jr. 1968. The
adsorption of atrazine and atratone by anion- and
cation-exchange resins. Soil Sol. Amer. Proc. 32:62-
63.
U.S. EPA. 1986a. U.S. Environmental Protection
Agency. Atrazine chronic feeding/oncogenicity study
preliminary incidence table of tumors regarding
possible section 6(a) (2) effect. Washington, DC: U.S.
EPA Office of Pesticide Programs.
U.S. EPA. 1986b. U.S. Environmental Protection
Agency. Guideline for carcinogen risk assessment.
Fed. Reg. 51(185):33992-34003. September 24.
U.S. EPA. 1986c. U.S. Environmental Protection
Agency. Code of Federal Regulations. Protection of
the environment. Tolerances and exemptions from
tolerances for pesticide chemicals in or on raw
agricultural commodities. 40 CFR 180.220. p. 216.
U.S. EPA. 1988. U.S. Environmental Protection
Agency. Method #507 - Determination of nitrogen
and phosphorus containing pesticides in ground
water by GC/NPD. April 14, draft.
IV. TRICHLOROETHYLENE
A General Information and Properties
CAS Wo.: 79-01-6
Structural Formula: C1-HC-C-C12
Trichloroethylene
Synonyms:
TCE, trichloroethene, acetylene trichloride, Tri,
Trilene
Uses:
Industrial solvent and degreaser for metal
components
Properties: (Torkelson and Rowe, 1981;
Windholtz, 1983)
Chemical formula
Molecular weight
Physical state
Boiling point
Vapor pressure
Density at 25°C
Water solubility
Odor threshold (water)
Odor threshold (air)
Organoleptic threshold (water)
Conversion factor
Occurrence
C2HC13
131.40
Colorless liquid
86.7°C
77 mm (25°C)
1.4g/mL
0.1 g/100 mL
(20°C)
0.5 mg/L
2.5-900 mg/m3
0.31 mg/L
(Amoore and
Hautala, 1983)
1 ppm = 5.46
mg/m3
Trichloroethylene (TCE) is a synthetic chemical
with no natural sources.
Production of TCE was 200 million Ib in 1982 (U.S.
ITC, 1983).
The major source of TCE released to the
environment is its use as a metal degreaser. Since
TCE is not consumed during this use, the majority
of all TCE production is released to the
environment.Most of the releases occur to the
atmosphere by evaporation. However, TCE that is
not lost to evaporation becomes heavily
contaminated with grease and oil and has been
disposed of by burial in landfills, dumping on the
ground, or into sewers. Because metal working
operations are performed nationwide, TCE
. releases occur in all industrialized areas. Releases
of TCE during production and other uses are
relatively minor.
Trichloroethylene released to the air is degraded
in a matter of a few days. Trichloroethylene
released to surface waters migrates to the
atmosphere in a few days or weeks, where it also
degrades. Photo-oxidation appears to be the
predominant fate of this compound (U.S. EPA,
1979). Trichloroethylene released to the land does
not degrade rapidly, migrates readily to ground
water and remains in ground water for months to
years. Under certain conditions, TCE in ground
water appears to degrade to dichloro-ethylene and
vinyl chloride. Trichloroethylene also may be
formed in ground water by the degra-dation of
tetrachloroethylene TCE, unlike other chlorinated
B-23
-------
compounds, does not bioaccumulate in individual
animals or food chains.
• Because of the large and dispersed releases, TCE
occurs widely in the environment. Trichloro-
ethylene is ubiquitous in the air, with levels in the
ppt to ppb range. Trichloroethylene is a common,
contaminant in ground and surface waters awith
higher levels found in ground water. Surveys of
drinking water supplies have found that 3% of all
public systems derived from well water contain
TCE at levels of 0.5 ug/L or higher. A small
number of systems (0.04%) have levels higher than
100 ug/L. Public systems derived from surface
water also have been found to contain TCE but at
lower levels. Trichloro-ethylene has been reported
to occur in some foods in the ppm range.
B. Pharmacokinetics
Absorption
• Data on absorption of ingested TCE are limited.
When a dose of 200 mg/kg of "C-TCE in corn oil
was administered to rats, 97% of the dose was
recovered during 72 hours after dosing (De Kant et
al., 1984).
Distribution
• Doses of 0,10,100, or 1,000 mg TCE/kg/day were
administered by gavage to rats 5 days/week for 6
weeks (Zenick et al., 1984). Marginal increases in
TCE tissue levels were detected in the 10
mg/kg/day and 100 mg/kg/day dose groups.
Compared to controls, a marked increase in TCE
levels in most tissues was observed in the highest
dose group. Trichloroethylene was distributed in
all tissues examined with the highest
concentrations in the fat, kidney, lung, adrenals,
vas deferens, epididymis, brain, and liver.
Metabolism
• Studies indicate that TCE is metabolized to
trichloroethylene oxide, trichloracetaldehyde,
trichloroacetic acid, monochloroacetic acid,
trichloroethanol, and trichloroethanol glucuronide
(EPA,1985a).
Excretion
• Trichloroethylene and its metabolites are excreted
in urine, by exhalation and, to a lesser degree, in
sweat, feces, and saliva (Soucek and Vlachova,
1959).
C. Health Effects
Humans
Short-term Exposure
• Oral exposure of humans to 15 to 25 mL (21 to 35
g) quantities of TCE resulted in vomiting and
abdominal bain, followed by transient
unconsciousness (Stephans, 1945).
Long-Term Exposure
• Studies of humans exposed occupationally have
shown an increase in serum transaminases, which
indicates damage to the liver parenchyma
(Lachnit, 1971). Quantitative exposure levels were
not available.
Animals
Short-Term Exposure
• The acute oral LD50 of TCE in rats is 4.92 g/kg
(NIOSH, 1980).
Long-Term Exposure
• Rats exposed to 300 mg/m3 (55 ppm) TCE 5
days/week for 14 weeks had elevated liver weights
(Kimmerle and Eben, 1973).
Reproductive Effects
• No data were available on the reproductive effects
ofTCE.
Developmental Effects
• No data were available on the developmental
effects of TCE.
Mutagenicity
• Trichloroethylene was mutagenic in Salmonella
typhimurium and in the E. coli K-12 strain,
utilizing liver microsomes for activation (Greim et
al., 1975,1977).
Carcinogenicity
• Technical TCE (containing epichlorohydrin and
other compounds) was found to induce a
hepatocellular carcinogenic response in B6C3Fi
mice (NCI, 1976). Under the conditions of this
experiment, a carcinogenic response was not
observed in Osborne-Mendel rats. The "time-
weighted" average doses were 549 and 1,097
mg/kg for both male and female rats. The time-
B-24
-------
weighted average daily doses were 1,169 and 2,339
mg/kg for male mice and 869 and 1,739 mg/kg for
female mice.
• Epichlorohydrin-free TCE was reported to be
carcinogenic in B6C3Fi mice when administered
in corn oil at 1,000 mg/kg/day, 5 days/wk, for 103 ,
weeks (NTP, 1982). It was not found to be
carcinogenic in female Fischer 344 rats when
administered in corn oil at 500 or 1,000 mg/kg/day,
5 days/wk, for 103 weeks. The experiment with
male rats was considered to be inadequate since
these rats received doses of TCE that exceeded the
maximum tolerated dose.
• TCE has been shown to be carcinogenic in mice
utilizing the inhalation as well as the oral route of
exposure. The National Cancer Institute (1976)
and the National Toxicology Program (1982) each
conducted an oral gavage study with TCE, one
contaminated with epichlorohydrin and the other
free of epichlorohydrin, respectively. In these
studies, as described above, B6C3Fi mice were
used, and the results were unequivocally positive,
showing liver neoplasms.
• In an inhalation study, Henschler et al. (1980)
reported dose-related malignant lymphomas in
female mice exposed to 100 or 500 ppm TCE vapor
6 hr/day, 5 days/wk, for 18 months (HANrNMRI
strain). However, the authors downplayed the
significance of this observation, indicating that
this strain of mice has a high incidence of
spontaneous lymphomas.
• Fukuda et al. (1983) found pulmonary
adenocarcinomas in female ICR mice on exposure
to TCE vapor.
• Henschler et al. (1984) tested Swiss (ICR/HA) mice
and .reported that when the animals were treated
by gavage with TCE in corn oil, no statistical
differences were observed in the incidence of
cancers. The results of this study can be questioned
because the dose schedule was often interrupted
even with half of the original dose. Therefore, it is
very difficult to assess the exposure. A slight
increase in tumors was found in all groups treated
with TCE but did not approach statistical
significance.
• The van Duuren study (1979) with skin
applications of TCE in ICR/HA mice does not
negate the positive findings with other strains of
mice and other routes of exposure.
D. Quantification of Toxicological
Effects
The HAs for noncarcinogenic toxicants are derived
using the following formula:
HA -
(NOAEL orLOAEL) x (BW)
(UF)x(_Llday)
_,mglL (_
where: > >
NOAEL or = No- or Lowest-Observed-Adverse-
LOAEL Effect Level in mg/kg bw/day.
B W = assumed body weight of a child (10
kg) or an adult (70 kg).
UF = uncertainty factor (10,100, or 1,000),
in accordance with NAS/ODW
guidelines.
L/day = assumed daily water consumption of
a child (1 L/day) or an adult (2 L/day).
One-Day and Ten-Day Health Advisory
Suitable data were not available to estimate One-Day
and Ten-Day Health Advisories.
Longer-Term Health Advisory
No suitable data are available from which to
calculate a Longer-Term Health Advisory.
Lifetime Health Advisory
The Lifetime HA represents that portion of an
individual's total exposure that is attributed to
drinking water and is considered protective of
noncarcinogenic adverse health effects over a
lifetime exposure. The Lifetime HA is derived in a
three-step process. Step 1 determines the Reference
Dose (RfD), formerly called the Acceptable Daily
Intake (ADI). The RfD is an estimate of a daily
exposure to the human population that is likely to be
without appreciable risk of deleterious effects over a
lifetime, and is derived from, the NOAEL (or
LOAEL), identified from a chronic (or subchronic)
study, divided by an uncertainty factor(s). From the
RfD, a Drinking Water Equivalent Level (DWEL)
can be determined (Step 2). A DWEL is a medium-
specific (i.e., drinking water) lifetime exposure level,
assuming 100% exposure from that medium, at
which adverse, noncarcinogenic health effects would
not be expected to occur. The DWEL is derived from
the multiplication of the RfD by the assumed body
weight of an adult and divided by the assumed daily
water consumption of an adult. The Lifetime HA is
determined in Step 3 by factoring in other sources of
exposure, the relative source contribution (RSC). The
B-25
-------
RSC from drinking water is based on actual exposure
data or, if data are not available, a value of 20% is
assumed for synthetic organic chemicals and a value
of 10% is assumed for inorganic chemicals. If the
cotaminant is classified as a Group A or B crcinogen,
according to the Agency's classification scheme of
carcinogenic potential (U.S. EPA, 1986), then caution
should be exercised in assessing the risks associated
with lifetime exposure to this chemical.
Trichloroethylene may be classified in Group B:
Probable Human Carcinogen, according to EPA's
weight-of-evidence scheme for the classification of
carcinogenic potential (U.S. EPA, 1986). Because of
this, caution must be exercised in making a decision
on how to deal with possible lifetime exposure to this
substance. The risk manager must balance this
assessment of carcinogenic potential against the
likelihood of occurrence of health effects related to
noncarcinogenic end points of toxicity. In order to
assist the risk manager in this process, drinking
water concentrations associated with estimated
excess lifetime cancer risks over the range of one in
ten thousand to one in a million for the 70 kg adult,
drinking 2 L of water per day, are provided in the
following section. In addition, in this section, a
DWEL is derived.
Neither the risk estimates nor the DWEL take RSC
into account. The risk manager should do this on a
case-by-case basis, considering the circumstances of
the specific contamination incident that has
occurred.
The study by Kimmerle and Eben (1973) is the most
appropriate from which to derive the DWEL. This
study evaluated the subacute exposure to
trichloroethylene via inhalation by adult rats for
some 14 weeks following exposure to 55 ppm (300
mg/m3), 5 days a week. Indices of toxicity include
hematological investigation, liver and renal function
tests, blood glucose, and organ/body weight ratios.
Liver weights were shown to be elevated while other
test values were not different from controls. The
elevated liver weights could be interpreted to be the
result of hydropic changes or fatty accumulation. The
No-Observed-Effect level was not identified since
only a single concentration was administered. From
these results, a LOAEL of 55 ppm (300 irig/mS) was
identified. With the LOAEL, the DWEL is derived as
follows:
^lep 1: Determination of the Total Absorbed Dose
(TAD)
TAD =
(300 mg/m3) (8 m3/day) (5/7) (0.3)
(70kg)
= 7.35 mg/kg/day
where:
300mg/m3 = LOAEL for liver effects in rats.
8 m3/day = volume of air inhaled during the
exposure period.
5/7
conversion factor for adjusting from 5
days/week exposure to a daily dose
ratio of the dose absorbed.
assumed weight of adult.
0.3
70kg
Step 2: Determination of the RfD
RFD =
7.35 mg/kg/day
(100) (10)
where:
7.35 mg/kg/day
1,000
= 0.00735 mg/kg/day
= TAD
= uncertainty factor, chosen in
accordance with NAS/ODW
guidelines for use with a
LOAEL from an animal
study.
Step 3: Determination of the DWEL
DWEL =
(0.00735 mg/kg/day) (70 kg)
where:
0.00735 mg
/kg/day
70kg
2L/day
2 Ll'day
= 0.26 mg/L (260 ug/L)
= RfD.
= assumed body weight of an
adult.
= assumed daily water
consumption of an adult.
The estimated excess cancer risk associated with
lifetime exposure to drinking water containing TCE
at 260 ug/L is approximately 1 x 10-4. This estimate
represents the upper 95% confidence limit from
extrapolations prepared by EPA's Carcinogen
Assessment Group using the linearized, multistage
model. The actual risk is unlikely to exceed this
value, but there is considerable uncertainty as to the
accuracy of risks calculated by this methodology.
Evaluation of Carcinogenic Potential
• IARC (1982) has classified TCE in Group 3.
• EPA has classified trichloroethylene in Group B2:
Probable Human Carcinogen. This classification
B-26
-------
for carcinogenicity was determined by a technical
panel of EPA's Risk Assessment Forum using the
EPA risk assessment guidelines for carcinogens
(U.S. EPA, 1986). This category is used for agents
for which there is "sufficient evidence" for human
carcinogenicity from animal" studies and for
which there is "inadequate evidence" or "no data"
from human studies.
• With the improved multistage linearized model,
estimates can be made that water with TCE
concentrations of 280 ug/L, 28 ug/L, or 2.8 ug/L
may increase the risk of one excess cancer per 104,
105, or 106 people exposed, respectively. These
estimates were calculated from the 1976 NCI
bioassay data, which utilized TCE contaminated
with epichlorohydrin. Since then, an NTP (1982)
bioassay utilizing epichloro-hydrin-free TCE has
become available; the data from this bioassay have
been reviewed and evaluated for carcinogenicity,
and epichlorohydrin-free TCE has been reported to
be carcinogenic in mice.
E. Other Criteria, Guidance, and
Standards
• ACGIH (1984) has recommended a threshold limit
value (TLV) of 50 ppm (-270 mg/m3) and a short-
term exposure limit (STEL) of 150 ppm (~805
mg/m3).
• The NAS (1980) recommended One- and Seven-
Day SNARLs of 105 and 15 mg/L, respectively.
• WHO (1981) recommended a drinking water
guidance level of 30 ug/L based on a carcinogenic
end point.
• The EPA (U.S. EPA, 1980) recommended a water
quality criterion of 6.77 mg/L for effects other than
cancer.
• The EPA (U.S. EPA, 1985d) has promulgated a
Recommended Maximum Contaminant Level
(RMCL) of zero based upon TCE's classification as
a known or probable human carcinogen and has
proposed a Maximum Contaminant Level (MCL) of
0.005 mg/L based on its RMCL and appropriate
feasibility studies.
F. Analytical Methods
• Analysis of TCE is by a purge-and-trap gas
chromatographic procedure used for the
determination of volatile organohalides in
drinking water (U.S. EPA, 1985b). This method
calls for the bubbling of an inert gas through the
sample and trapping TCE on an adsorbent
material. The adsorbent material is heated to
drive off the TCE onto a gas chromatographic
column. This method is applicable to the
measurement of TCE over a concentration range of
0.01 to 1500 ug/L. Confirmatory analysis for TCE
is by mass spectrometry (U.S. EPA, 1985c). The
detection limit for confirmation by mass
spectrometry is 0.2 ug/L.
G. Treatment Technologies
• Treatment technologies that will remove TCE
from water include granular activated carbon
(GAG) adsorption, aeration and boiling.
• Dobbs and Cohen (1980) developed adsorption
isotherms for several organic chemicals including
TCE. Fibrasorb® 300 carbon exhibited adsorptive
capacities of 7,1.6, and 0.4 mg TCE/gm carbon at
equilibrium concentrations of 100,10, and 1 mg/L,
respectively. USEPA-DWRD installed pilot-scale
adsorption columns at different sites in New
England and Pennsylvania. In New England,
contaminated well water with TCE concentrations
ranging from 0.4 to 177 mg/L was passed through
GAC columns until a break-through
concentration of 0.1 mg/L was achieved with
empty bed contact time (EBCT) of 18 and 9
minutes, respectively (Love and Eilers, 1982). In
Pennsylvania, TCE concentrations ranging from
20 to 130 mg/L were reduced to 4.5 mg/L by GAC
after 2 months of continuous operation (ESE,
1985).
• TCE is amenable to aeration on the basis of its
Henry's Law constant of 550 atm (Kavanaugh and
Trussell, 1980). In a full plant-scale (3.78 MGD)
redwood slat tray aeration column, a removal
efficiency of 50-60% was achieved from TCE initial
concentrations of 8.3-39.5 mg/L at an air-to-water
ratio of 30:1 (Hess et al., 1981). In another full
plant-scale (6.0 MGD) multiple tray aeration
column study, TCE removal of 52% was achieved
from 150 mg/L (Hess at al., 1981). A full plant-
scale packed tower aeration column removed 97-
99% of TCE from 1,500-2,000 mg/L contaminated
ground water at air-to-water ratio of 25:1 (ESE,
1985).
• Boiling also is effective in eliminating TCE from
water on a short-term, emergency basis. Studies
have shown 5 minutes of vigorous boiling will
remove 95% of TCE originally present (Love and
Eilers, 1982).
• Air stripping is an effective, simple, and relatively
inexpensive process for removing TCE and other
volatile organics from water. However, use of this
process then transfers the contam-inant directly to
the air stream. When consider-ing use of air
stripping as a treatment process, plant managers
must factor in the overall environmental
occurrence, fate, route of ex-posure, and various
other hazards associated with the chemical.
B-27
-------
H. References
ACGIH. 1984. American Conference of
Governmental industrial Hygienists. Documentation
of the threshold limit values. 4th ed. 1980-1984
Supplement, pp. 406-408.
Amoore, J.E., and E. Hautala. 1983. Odor as an aid to
chemical safety: Odor thresholds compared with
threshold limit values and volatilities for 214
industrial chemicals in air and water dilution. J.
Appl.Tox. 3:272-290,
DeKant, W., M. Metzler & D. Henschler. 1984. Novel
metabolites of trichloroethylene through
dcchlorination reactions in rats, mice and humans.
Biochem. Pharmacol. 33:2021-2027.
Dobbs, R.A., and J.M. Cohen. 1980. Carbon
adsorption isotherms for toxic organics. EPA 600/8-
80-023, Office of Research and Development, MERL,
Wastewater Treatment Division, Cincinnati, Ohio.
ESE. 1985. Environmental Science and Engineering.
Draft technologies and costs for the removal of
volatile organic chemicals from potable water
supplies. ESE No. 84-912-0300 prepared for U.S.
EPA, Science and Technology Branch, CSD, ODW,
Washington, D.C.
Pukuda, K., K. Takemoto, and H. Tsuruta. 1983.
Inhalation carcinogenicity of trichloroethylene in
mice and rats. Ind. Health, 212243-254.
Greim, H., D. Bimboes, G. Egert, W. Giggelmann and
M. Kramer. 1977. Mutagenicity and chromosomal
aberrations as an analytical tool for in vitro detection
of mammalian enzyme-mediated formation of
reactive metabolites. Arch. Toxicol. 39:159.
Greim, H., G, Bonse, Z. Radwan, D. Reichert and D.
Henschler. 1975. Mutagenicity in vitro and potential
carcinogenicity of chlorinated ethylenes as a function
of metabolic oxirane formation. Biochem. Pharmacol.
24:2013.
Henschler, D., W. Romen, H.M. Elsasser, D. Reichert,
E. Edcr and Z. Radwan. 1980. Carcinogenicity study
of trichloroethylene by long-term inhalation in the
animal species. Arch. Toxicol. 43:237-248.
Henschler, D., H. Elsasser, W. Romen, and E. Eder.
1984. Carcinogenicity study of trichloroethylene,
with and without epoxide stabilizers, in mice. J.
Cancer Res. Clin. Oncol. 104149-156.
Hess, A.P., J.E. Dyksen and G.C. Cline. 1981. Case
study involving removal of organic chemical
compounds from ground water. Presented at Annual
American Water Works Association Conference, St.
Louis, Missouri.
IARC. 1982. IARC monographs on the evaluation of
the carcinogenic risk of chemicals to humans,
Supplement 4,. Lyon, France.
Kavanaugh, M.C., and R.R. Trussell. 1980. Design of
aeration towers to strip volatile contaminants from
drinking water. JAWWA. December.
Kimmerle, G., and A. Eben. 1973. Metabolism,
excretion and toxicology of trichloroethylene after
inhalation. 1. Experimental exposure on rats. Arch.
Toxicol. 30:115.
Lachnit, V. 1971. Halogenated hydrocarbons and the
liver. Wien. Klin. Wochenschr. 83(41):734.
Love, O.T., Jr., and R.G. Eilers. 1982. Treatment of
drinking water containing trichloroethylene and
related industrial solvents. JAWWA. August.
NAS. 1980. National Academy of Sciences. Drinking
Water and Health. Volume 3. National Academy
Press. Washington, DC.
NCI. 1976. National Cancer Institute.
Carcinogenesis bioassay of trichloroethylene. U.S.
Department of Health, Education and Welfare,
Public Health Service, CAS No. 79-01-6, February.
NIOSH. 1980. Registry of Toxic Effects of Chemical
Substances. U.S. Department of Health and Human
Services. DHHS (NIOSH) 81-116.
NTP. 1982. National Toxicology Program.
Carcinogenesis bioassay for trichloroethylene. CAS
79-01-6. No. 82-1799. (Draft).
Parsons, F., P.R. Wood, and J. DeMarco. 1984.
Transformation of tetrachlore thane and
trichloroethene in microcosms and groundwater.
JAWWA, 26(2):56f.
Soucek, B., and D. Vlachova. 1959. Metabolites of
trichloroethylene excreted in the urine by man.
Pracoc.Lek. 11:457.
Stephens, C.A. 1945. Poisoning by accidental
drinking of trichloroethylene. Brit. Med. J. 2: 218.
Torkelson, T.R., and V.K. Rowe. 1981. Halogenated
aliphatic hydrocarbons.: In Industrial Hygiene and
Toxicology. 3rd ed, Vol. 2B. John Wiley and Sons,
New York. p. 3553.
U.S. EPA. 1979. U.S. Environmental Protection
Agency. Water-related environmental fate of 129
B-28
-------
priority pollutants, Office of Water Planning and
Standards, EPA-440/4-79-029.
U.S. EPA. 1980. U.S. Environmental Protection
Agency. Ambient water quality criteria document for
triehloroethylene. Office of Water Research and
Standards. Cincinnati, Ohio.
U.S. EPA. 1983. U.S. Environmental Protection
Agency. Triehloroethylene occurrence in drinking
water, food, and air. Office of Drinking Water.
U.S. EPA. 1985a. U.S. Environmental Protection
Agency. The drinking water criteria document on
triehloroethylene. Office of Drinking Water.
U.S. EPA. 1985b. Method 502.1. Volatile halogenated
organic compounds in water by purge and trap gas
ehromatography. Environmental Monitoring and
Support Laboratory, Cincinnati, Ohio 45268.
U.S. EPA. 1985c. Method 524.1. Volatile organic
compounds in water by purge and trap gas
chromatography/mass spectrometry. Environmental
Monitoring and Support Laboratory, Cincinnati,
Ohio 45268.
U.S. EPA. 1985d. U.S. Environmental Protection
Agency. National primary drinking water
regulations; volatile synthetic organic chemicals;
final rule and proposed rule. Federal Register
50(219):46880-46933.
U.S. EPA. 1986. U.S. Environmental Protection
Agency. Guidelines for carcinogenic risk assessment.
Federal Register 51(185): 33992-34003. September
24.
U.S. ITC. 1983. United States International Trade
Commission. Synthetic organic chemicals. United
States production, USITC Publication 1422.
Washington, DC. 20436.
van Duuren, B.L., B.M. Goldschmidt, G. Lowengart,
A.C. Smith, S. Melchionne, I. Seldman, and D. Roth.
1979. Carcinogenicity of halogenated olefinic and
aliphatic hydrocarbons in mice. J. Natl. Cancer Inst.
63: 1433-1439.
Vogel, T., and P. McCarty. 1985. Biotransformation
of tetrachloroethylene to triehloroethylene, dichloro-
ethylene, vinyl chloride, and carbon dioxide under
methanogenic conditions. Appl. Environ. Microbiol.
49(5).
WHO. 1981. World Health Organization. Guidelines
for drinking water quality. Vol. I. Recommendations.
Geneva, Switzerland, pp. 63, 66.
Windholz, M. 1983. The Merck Index. 10th edition.
Merck and Co., Inc. Rahway, NJ. p. 1378.
Zenick, H., K. Blackburn, E. Hope, N. Richards and
M.K. Smith. 1984. Effects of triehloroethylene
exposure on male reproductive function in rats.
Toxicology 31:237.
V. VINYL CHLORIDE
A. General Information and Properties
CAS Wo.: 75-01-4
Structural Formula:
H-C=C-CI
I I
H H
Synonyms:
• Monochloroethylene, chloroethene
Uses:
• Vinyl chloride and poly vinyl chloride (PVC) are
used as raw materials in the plastics, rubber,
paper, glass and automotive industries. In
addition, vinyl chloride and PVC are used in the
manufacture of electrical wire insulation and
cables, piping, industrial and household
equipment, medical supplies, food packaging
materials and building and construction products.
Vinyl chloride copolymers and PVC are
distributed and processed in a variety of forms,
including dry resins, plastisol (dispersions in
plasticizers), organosol (dispersions in plasticizers
plus volatile solvent), and latex (a colloidal
dispersion in water used to coat paper, fabric or
leather) (U.S. EPA, 1985a).
Properties: (U.S. EPA, 1985a):
Chemical formula
Molecular weight
Physical state
Boiling point
Melting point
Density
Vapor pressure
Specific gravity
Water solubility
Taste threshold (water)
Odor threshold (water)
Conversion factor (air)
H2C = CHC1
62.5
Gas
-13.3°C
2,530mmHgat20°C
0.91
1.1 g/L water at 28°C
3.4 mg/L*
1 ppm = 2.6 mg/m3
* Amoore and Hautala (1983)
B-29
-------
Occurrence
• Vinyl chloride is a synthetic chemical with no
natural sources.
• Since 1979, yearly production of vinyl chloride has
been approximately 7 billion Ibs (U.S. ITC, 1983).
Vinyl chloride is polymerized, and little is released
to the environment. Environmental releases will
be limited to the areas where vinyl chioride is
produced and used.
• Vinyl chloride released to the air is degraded in a
matter of a few hours (EPA, 1980a). Vinyl chloride
released to surface waters migrates to the
atmosphere in a few hours or days, where it
undergoes photochemical oxidation. Vinyl chloride
that is released to the ground does not adsorb onto
soil and migrates readily to ground water.
Evidence from laboratory studies suggests that
vinyl chloride in ground water may degrade to C02
and Cl- (Vogel and McCarty, 1985). Vinyl chloride
is expected to remain in ground water for months
to years. Vinyl chloride has been reported to be a
degradation product of trichloroethylene and
tetrachloroethylene in ground water (Parsons,
1984). Vinyl chloride does not bioaccumulate in
individual animals or food chains.
• Vinyl chloride does not occur widely in the
environment because of its rapid degradation and
limited release. Vinyl chloride is a relatively rare
contaminant in ground and surface waters with
higher levels found ground water. The Ground
Water Supply Survey of drinking water supplies
have found that less than 2% of all ground water
derived public water systems contain vinyl
chloride at levels of 1 ug/L or higher. Vinyl
chloride almost always co-occurs with
trichloroethylene. Public systems derived from
surface water also have been found to contain
vinyl chloride but at lower levels. No information
on the levels of vinyl chloride in food have been
identified. Based upon the limited uses of vinyl
chloride and its physical chemical properties, little
or no exposure is expected from food. Vinyl
chloride occurs in air in urban areas and near the
sites of its production and use. Atmospheric
concentrations are in the ppt range (U.S. EPA,
1979).
• The major source of exposure to vinyl chloride is
from contaminated water.
B. Pharmacokinetics
Absorption
• Vinyl chloride is absorbed rapidly in rats following
ingestion and inhalation (Withey, 1976; Duprat et
al., 1977).
• Using statistical modeling, Withey and Collins
(1976) concluded that, for rats, a total liquid intake
containing 20 ppm (wt/wt) vinyl chloride would be
equivalent to an inhalation exposure of about 2
ppm (vol/vol) for 24 hours.
Distribution
• Upon either inhalation or ingestion of 14C-vinyl
chloride in rats, the greatest amount of 14C
activity was found 72 hours after treatment in
liver followed by kidney, muscle, lung and fat
(Watanabe et al., 1976a,b). Another study of
inhalation exposure of rats to 14C-vinyl chloride
showed the highest 14C activity immediately after
treatment in liver and kidney, followed by spleen
and brain (Bolt et al., 1976).
Metabolism
• Bartsch and Montesano (1975) reported two
possible metabolic pathways for vinyl chloride, one
involving alcohol dehydrogenase, the other
involving mixed function oxidase. Hefner et al.
(1975) concluded that the dominant pathway at
lower exposure levels probably involves alcohol
dehydrogenase.
• Vinyl chloride metabolism is saturable (Hefner et
al., 1975; Watanabe et al., 1976a, Bolt et al., 1977).
• Chloroethylene oxide, presumably through mixed-
function oxidase, may be the main metabolite
capable of alkylating intracellular
macromolecules (Laib and Bolt, 1977).
Excretion
• Rats administered vinyl chloride by ingestion or
inhalation exhale greater amounts of
unmetabolized vinyl chloride as the dose is
increased (Watanabe et al., 1976a, b).
• Vinyl chloride metabolites are excreted mainly in
the urine. In rats, urinary metabolites include N-
acetyl-S-( 2-hydroxyethylcysteine) and
thlodiglycolic acid (Watanabe et al., 1976a).
B-30
-------
C. health Effects
Developmental Effects
Humans
• Cancer findings in humans are described under
Carcinogenicity.
• Mutagenic effects in humans are described under
Mutagenicity.
• Developmental studies in humans are described
under Developmental Effects.
• At high inhalation exposure levels, e.g., 40-900
pjim (104-2,344 mg/m3), workers have experienced
dizziness, headaches, euphoria and narcosis (U.S.
EPA, 1985a).
• Symptoms of chronic inhalation exposure of
workers to the vinyl chloride- polyvinyl chloride
industry include hepatotoxicity (Marstellar et al.
1975), acro-osteolysis (Lilis etal., 1975), central
nervous system disturbances, pulmonary
insufficiency, cardiovascular toxicity, and
gastrointestinal toxicity (Miller et al., 1975;
Selikoff and Hammond, 1975; Suciu et al, 1975).
Data on dose-responses in humans are scarce
because few measurements of ambient vinyl
chloride levels hi the workplace were made before
1975 (Mancuso, 1975).
Animals
Short-Term Exposure
• Inhalation exposure to high levels (ca. 100,000
ppm or 260,417 mg/m3) of vinyl chloride can
induce narcosis and death, and, to lower doses,
ataxia, narcosis, congestion and edema in lungs
and hyperemia in liver in several species (U.S.
EPA, 1985a).
Long-Term Exposure
• Administration of vinyl chloride monomer to rats
by gavage for 13 weeks resulted in hematologic,
biochemical and organ weight effects at doses
above 30 mg/kg (Feron et al., 1975).
• Inhalation exposure of rats, guinea pigs, rabbits
and dogs to 50 ppm (130 mg/m3) vinyl chloride, 7
hours/day, 130 exposures in 189 days, did not
induce toxicity as judged by appearance, mortality,
growth, hematology, liver weight, and pathology.
Rats exposed to 100 ppm (260 mg/m3) 2 hours/day
for 6 months, had increased liver weights
(Torkelson etal., 1961).
Reproductive Effects
• Potential effects on reproductive capacity have not
been studied.
• Infante et al. (1976a,b) reported an association
between human exposure to vinyl chloride and
birth defects and fetal loss, but this association
was contradicted by Edmonds et al. (1975) and
Hatch et al. (1981).
• Inhalation exposure of rats and rabbits to vinyl
chloride concentrations as high as 2,500 ppm
(6,500 mg/m3) on days 6 to 15 (rats) and 6 to 18
(rabbits) of gestation and mice to vinyl chloride
levels as high as 500 ppm (1,300 mg/m3) on days 6
to 15 of gestation did not induce teratogenie effects
but did increase skeletal variants in high-dose
mice (John et al 1977).
• A developmental effects study with vinyl chloride
in rats exposed by inhalation to 600 or 6,000 ppm
(2,160 or 21,160 mg/m3) 4 hours daily on gestation
days 9 through 21 was negative for teratogenicity
and inconclusive for fetotoxicity (Radike et al.,
1977).
Mutagenicity
• Chromosomal effects of vinyl chloride exposure in
workers is conflicting in that positive (Ducatmann
et al., 1975; Purchase et al., 1975) and negative
(Killian et al., 1975; Picciano et al., 1977) results
have been reported. Picciano et al. (1977) reported
exposures of 0.13 to 15.2 ppm (0.34 to 40 mg/m3,
time-weighted averages) for 1 to 332 months.
• Vinyl chloride is mutagenic, presumably through
active metabolites in various systems including
metabolically activated systems with S.
typhimurium (Bartsch et al., 1975); E. coli (Greim
et al., 1975); yeast (Loprieno et al., 1977); germ
cells ofDrosophila (Verburgt and Vogel, 1977);
and Chinese hamster V79 cells (Hubermann et al.,
1975).
• Dominant lethal studies with vinyl chloride in CD-
1 mice were negative (Anderson et al., 1976).
Carcinogenicity
• Increases in the occurrence of liver angiosarcomas
as well as in tumors of the brain, lung, and
hematopoietic and lymphopoietic tissues have
been associated with occupational exposure to the
vinyl chloride-poly vinyl chloride industry in
humans (IARC, 1979). The initial report of a link
between vinyl chloride exposure and cancer in
humans by Creech and Johnson (1974), as well as
subsequent reports by others, indicates the high
risk and specificity of association with liver
angiosarcoma, a very rare tumor in humans.
• Ingestion of vinyl chloride monomer in the diet by
rats at feeding levels as low as 1.7 and 5 mg/kg/
B-31
-------
day over their lifespan induced hepatocellular
carcinomas and liver angiosarcomas, respectively,
as well as other adverse hepatic effects (Feron et
al., 1981). Til et al. (1983) extended the Feron et al.
(1981) work to include lower doses and did not find
a significant (P < 0.05) increase in carcinogenic
effects at feeding levels as high as 0.13 mg/kg/day.
Administration of vinyl chloride monomer by
gastric intubation for at least 52 weeks resulted in
carcinogenic effects in liver and other, tissue sites
in rats (Feron et al., 1981; Maltoni et al., 1981).
• Chronic inhalation of vinyl chloride has induced
cancer in liver and other tissue sites in rats and
mice (Lee et al., 1977,1978; Maltoni et al., 1981).
D. Quantification of Toxicological
Effects
The HAs for noncarcinogenic toxicants are derived
using the following formula:
HA =
(NOAEL orLOAEL) x(BW)
(UF) x(_L/day)
mglL(_
where:
NOABL or = No- or Lowest-Observed-Adverse-
LOAEL Effect Level in mg/kg bw/day.
BW = assumed body weight of a child (10
kg) or an adult (70 kg).
UF = uncertainty factor (10,100, or 1,000),
in accordance with NAS/ODW
guidelines.
L/day = assumed daily water consumption of
a child (1 L/day) or an adult (2 L/day).
One-Day Health Advisory
There are insufficient data for calculation of a One-
Day Health Advisory. The Ten-Day HA of 2.6 mg/L is
proposed as a conservative estimate for a One-Day
HA.
Ten-Day Health Advisory
Inhalation data by Torkelson et al. (1961) were not
selected for the Ten-Day HA calculation because of
preference for studies with oral exposure. Feron et al.
(1975) reported a subchronic toxicity study in which
vinyl chloride monomer (VCM) dissolved in soybean
oil was administered by gavage to male and female
Wistar rats, initially weighing 44 g, at doses of 30,
100, or 300 mg/kg once daily, 6 days per week for 13
weeks. Several hematological, biochemical and organ
weight values were significantly (P<0.05 or less)
different in both mid- and high-dose animals
compared to controls. The NOAEL in this study was
identified as 30 mg/kg.
The Ten-Day HA, as well as the One-Day HA, for a
10-kg child is calculated as follows:
Ten Day- HA =
(30 mg/kg/day (6/7) (10 kg)
(100) (I LI day)
where:
30 mg/kg/day
6/7
10kg
100
1 L/day
= 2.6 mg/L (2,600 ug/L)
NOAEL based on absence of
biochemical and organ weight
effects in rats exposed orally to
vinyl chloride.
expansion of 6 days/week
treatment in the Feron et al.
(1975) study to 7 days/week to
represent daily exposure.
assumed body weight of a child.
uncertainty factor, chosen in
accordance with NAS/ODW
guidelines for use with a NOAEL
from an animal study.
assumed daily water
consumption of a child.
Longer-Term Health Advisory
The Longer-term HA can be calculated from the
lifetime feeding study in rats by Til et al. (1983). Til
et al. (1983) have extended the earlier work by Feron
et al. (1981) to include lower doses with basically the
same protocol used in the latter study. Carcinogenic
and noncarcinogenic effects were evident with a vinyl
chloride dietary level of 1.3 mg/kg/day. At dietary
levels of 0.014 and 0.13 mg/kg/day, increased
incidences of basophilic foci of cellular alteration in
the liver of female rats were evident. However,
basophilic foci by themselves are concluded not to
represent an adverse effect on the liver in the absence
of additional effects indicative of liver lesions such as
those found in the 1.3 mg/kg/day group; and a dose-
related increase in basophilic foci was not evident.
Therefore, the dose of 0.13 mg/kg/day is identified as
the NOAEL for noncarcinogenic effects for the
Longer-Term HA calculation.
Using the 0.13 mg/kg/day NOAEL from the Til et al.
(1983) study, the Longer-term HA for a 10-kg child is
calculated as follows:
Longer— Term HA =
(0.13 mg/kg/day) (10 kg)
(100) (I L/day)
= 0.013 mg/L (13 ug/L)
B-32
-------
where:
0.13 mg/kg/day = NOAEL based on absence of
adverse liver effects in rats.
10 kg = assumed body weight of a
child.
100 = uncertainty factor, chosen in
accordance with NAS/ODW
guidelines for use with a
NOAEL from an animal
study.
1 L/day — assumed daily water
consumption of a child.
The Longer-Term HA for a 70-kg adult is calculated
as follows:
Longer—Term HA —
(0.13 mg/kg/day) (70 kg)
(100) (2 L/day)
where:
0.13 mg/kg/day
70kg
100
2 L/day
= Q.Q46mg/L (46 ug/L)
NOAEL based on absence of
adverse liver effects in rats.
assumed body weight of an
adult.
uncertainty factor, chosen in
accordance with NAS/ODW
guidelines for use with a
NOAEL from an animal
study
assumed daily water
consumption of an adult.
Lifetime Health Advisory
The Lifetime HA represents that portion of an
individual's total exposure that is attributed to
drinking water and is considered protective of
noncarcinogenic adverse health effects over a
lifetime exposure. The Lifetime HA is derived in a
three step process. Step 1 determines the Reference
Dose (RfD), formerly called the Acceptable Daily
Intake (ADI). The RfD is an estimate of a daily
exposure to the human population that is likely to be
without appreciable risk of deleterious effects over a
lifetime, and is derived from the NOAEL (or
LOAEL), identified from a chronic (or subchronic)
study, divided by an uncertainty factor(s). From the
RfD, a Drinking Water Equivalent Level (DWEL)
can be determined (Step 2). A DWEL is a medium-
specific (i.e., drinking water) lifetime exposure level,
assuming 100% exposure from that medium, at
which adverse, noncarcinogenic health effects would
not be expected to occur. The DWEL is derived from
the multiplication of the RfD by the assumed body
weight of an adult and divided by the assumed daily
water consumption of an adult. The Lifetime HA is
determined in Step 3 by factoring in other sources of
exposure, the relative source contribution (RSC). The
RSC from drinking water is based on actual exposure
data or, if data are not available, a value of 20% is
assumed for synthetic organic chemicals and a value
of 10% is assumed for inorganic chemicals. If the
contaminant is classified as a Group A or B
carcinogen, according to the Agency's classification
scheme of carcinogenic potential (U.S. EPA, 1986),
then caution should be exercised in assessing the
risks associated with lifetime exposure to this
chemical.
Because vinyl chloride is classified as a human
carcinogen (IARC Group 1 and EPA Group A), a
Lifetime Health Advisory is not recommended.
Evaluation of Carcinogenic Potential
• Applying the criteria described in EPA's
guidelines for assessment of carcinogenic risk
(U.S. EPA, 1986), vinyl chloride may be classified
in Group A: Human carcinogen. This category is
for agents for which there is sufficient evidence to
support the causal association between exposure to
the agents and cancer.
« The IARC (1979) has concluded that there is
sufficient evidence to classify vinyl chloride as a
human carcinogen in its Category 1.
• EPA's Carcinogen Assessment Group (CAG)
recently has recalculated its excess carcinogenic
risk estimates resulting from lifetime exposure to
vinyl chloride through the drinking water (U.S.
EPA, 1985a). CAG based its preliminary revised
estimates on the Feron et al. (1981) study. The
total number of tumors, considering tumors of the
lung and liver, in rats exposed through the diet
was used to calculate the excess cancer risk. Using
the 95% upper limit [qi* = 2.3 (mg/kg/ day)-*] with
the linearized multistage model, they calculated
that consuming 2 L of water per day with vinyl
chloride concentration of 1.5 ug/L, 0.15 ug/L and
0.015 pg/L would increase the risk of one excess
cancer per 10,000 (10-4), 100,000 (10-5) Or
1,000,000 (10-6) people exposed, respectively, per
lifetime. The CAG is presently reassessing the
cancer risk estimate based on the Feron et al.
(1981) study by taking into account the more
recent data by Til et al. (1983) which, as described
previously, is an extension of the earlier Feron et
al. (1981) work to include lower doses.
• Maximum likelihood estimates as well as 95%
upper limits of cancer risks by the multistage
model are presented. Expressing risk as
cases/lifetime/person, examples would be 0.01.
mg/kg/day or 0.35 mg/L exposure associated with
B-33
-------
risks of 1.6 x 10-2 (MLE) and 1.9 x 10-2 (UL) and
0.001 mg/kg/day exposure associated with risks of
1.6 x 10-3 (MLE) and 1.9 x 10-3 (UL).
• Cancer risk estimates (95% upper limit) with other
models are presented for comparison with that
derived with the multistage. For example, one
excess cancer per 1,000,000 (10-6) is associated
with exposure to vinyl chloride in drinking water
at levels of 50 pg/L (probit), 0.5 g/L (logit), and 0.02
pg/L (Weibull). While recognized as statistically
alternative approaches, the range of risks
described by using any of these modeling
approaches has little biological significance unless
data can be used to support the selection of one
model over another. In the interest of consistency
of approach and in providing an upper bound on
the potential cancer risk, the EPA has
recommended use of the linearized multistage
approach.
E. Other Criteria, Guidance, and
Standards
• The National Academy of Sciences (NAS, 1977)
estimated a 10-6 risk (95% upper bound estimate)
from lifetime exposure to 1 ug vinyl chloride/L
drinking water with the multistage model and the
lifetime ingestion study in rats by Maltoni et al.
(1981).
• The final RMCL by the U.S. EPA Office of
Drinking Water is zero, the proposed MCL is 1
ug/L, and the practical quantitation level is 1 ug/L
(U.S. EPA, 1985b).
• Ambient water quality criteria (U.S. EPA, 1980b)
are 20,2 and 0.2 ug/L for risks of 10-5, 10-6, and
10-7, respectively, assuming consump-tion of 2 L of
water and 6.5 grams of contam-inated fish per day
by a 70 kg adult.
• A workplace standard of 1 ppm (time-weighted
average) was set by OSHA in 1974 based on the
demonstration of angiosarcoma of the liver in
vinyl chloride workers (Federal Register.
39:35890).
• The ACGIH (1982) has recommended a TLV of 5
ppm(10mg/m3).
F. Analytical Methods
• Analysis of vinyl chloride is by a purge and trap
gas chromatographic procedure used for the
determination of volatile organohalides in
drinking water (U.S. EPA, 1985c). This method
calls for the bubbling of an inert gas through a
sample of water and trapping the purged vinyl
chloride on an adsorbent material. The adsorbent
material is heated to drive off the vinyl chloride
onto a gas chromatographic column. This method
is applicable to the measurement of vinyl chloride
over a concentration range of 0.06 to 1,500 ug/L.
Confirmatory analysis for vinyl chloride is by mass
spectrometry (U.S. EPA, 1985d). The detection
limit for confirmation by mass spectrometry is 0.3
ug/L.
G. Treatment Technologies
• The value of the Henry's Law constant for vinyl
chloride (6.4 atm-m3/mole) suggests aeration as a
potential removal technique for vinyl chloride in
water (ESE,1984). Removals of up to 99.27% were
achieved at 90°C using a pilot packed tower
aerator. In similar studies, vinyl chloride removed
from ground water using a spray aeration system
with total VOC concentration was 100 to 200 ug/L
(ESE, 1984). Greater than 99.9% VOC removal
was obtained using a four-stage aeration system;
each stage employed 20 shower heads with a
pressure drop of approx-imately 10 pounds per
square inch. In-well aeration has also
demonstrated up to 97% removal of vinyl chloride
using an air-lift pump. However, practical
considerations are likely to limit the application of
this (Miltner, 1984).
• The concentration of vinyl chloride in southern
Florida ground water declined by 25% to 52%
following passage through lime softening basins
and filters (Wood and DeMarco, 1980). Since vinyl
chloride is a highly volatile compound, it is
probably volatilized during treatment (ESE, 1984).
• Adsorption techniques have been less successful
than aeration in removing vinyl chloride from
water. In a pilot study, water from a ground water
treatment plant was passed through a series of
four 30-inch granular activated carbon (Filtrasorb
400) columns (Wood and DeMarco, 1980; Symons,
1978); the empty bed contact time was
approximately six minutes per column. Influent
vinyl chloride concentrations ranged from below
detection to 19 ug/L; erratic removal was reported.
To maintain effluent concentrations below 0.5
ug/L, the estimated column capacity to
breakthrough was 810,1,250, 2,760 and 2,050 bed
volumes for empty bed contact times of 6,12,19
and 25 minutes, respectively. In addition, the
estimated service life of the activated carbon was
low. Similarly, poor removal of vinyl chloride was
achieved using an experimental synthetic resin,
Ambersorb XE-340, (Symons, 1978).
• Treatment technologies for the removal of vinyl
chloride from water have not been extensively
evaluated except on an experimental level.
Available information suggests aeration merits
further investigation. Selection of individual or
combinations of technologies to achieve vinyl
B-34
-------
chloride removal must be based on a case-by-case
technical evaluation, and an assessment of the
economics involved.
H. References
ACGIH. 1982. American Conference of
Governmental Industrial Hygienists. Threshold limit
values for chemical substances and physical agents
in the workroom environment. Cincinnati, OH.
Amoore, J.E., and E. Hautala. 1983. Odor as an aid to
chemical safety: Odor thresholds compared with
threshold limit values and volatilities for 214
industrial chemicals in air and water dilution. J.
Appl. Toxicol. 3:272-290.
Anderson, D., M.C.E. Hodge, and I.F. H. Purchase.
1976. Vinyl chloride: Dominant lethal studies in
male CD-I mice. Mutat. Res. 40:359-370,
Bartsch, H., C. Malaveille, and R. Montesano. 1975.
Human, rat and mouse liver- mediated mutagenicity
of vinyl chloride in S. typhimurium strains. Int. J.
Cancer. 15:429-437.
Bartsch, H. and R. Montesano. 1975. Mutagenic and
carcinogenic effects of vinyl chloride Mutat, Res.
32:93-114.
Bolt, H.M., H. Kappus, A. Buchter, and W. Bolt.
1976. Disposition of (1,2-14Q vinyl chloride in the
rat. Arch. Toxicol, 35:153-162.
Bolt, H.M., R.J. Laib, H. Kappus, and A. Buchter.
1977. Pharmaeokinetics of vinyl chloride in the rat.
Toxicol. 7:179-188.
Creech, J.L. and M.N. Johnson. 1974. Angiosarcoma
of the liver in the manufacture of poly vinyl chloride.
J. Occup. Med. 20:338-340.
Ducatman, A., K. Hirschhorn and I.J. Selikoff. 1975.
Vinyl chloride exposure and human chromosome
aberrations. Mutat. Res. 31:163-168.
Duprat, P., J.P. Fabry, D. Gradiski. and J.L.
Magadur. 1977. Metabolic approach to industrial
poisoning: blood kinetics and distribution of 14C-
vinyl chloride monomer (V.C.M.). Acta. Pharmacol.
Toxicol. Suppl. (Kbh) 41(1):142-143.
Edmonds, L.D., H. Falk, and J. E. Nissim. 1975.
Congenital malformations and vinyl chloride.
Lancet. 2:1098.
ESE. 1984. Environmental Science and Engineering,
Technologies and costs for the removal of volatile
organic chemicals from potable water supplies.
(Draft) ESE No. 84-912-0300. Prepared for U.S. EPA,
Science and Technology Branch, CSD, ODW,
Washington, DC.
Federal Register. 39:35890.
Feron, V.J., A.J. Speak, M.I. Williams, D. van
Battum, and A.F. de Groot. 1975. Observations on the
oral administration and toxicity of vinyl chloride in
rats. Fd. Cosmet, Toxicol. 13:633-638.
Feron, V.J., C.F.M. Hendrikson, A.J. Speek, H.P. Til,
and B.J. Spit. 1981. Lifespan oral toxicity study of
vinyl chloride in rats, Fd. Cosmet. Toxicol. 19: 317-
331.
Greim, H., G. Bonse, Z. Radwan, D. Reichert and D.
Henschler. 1975. Mutagenicity in vitro and potential
carcinogenicity of chlorinated ethylenes as a function
of metabolic oxirane formation. Biochem. Pharmacol.
24:2013-2017.
Hatch, M., J. Kline and Z. Stain. 1981. Power
considerations in studies of reproductive effects of
vinyl chloride and some structural analogs. Environ.
Health Perspec. 41:195-201.
Hefner, R.E., Jr., P.G. Watanabe, and P.J. Gehring.
1975. Preliminary studies on the fate of inhaled vinyl
chloride monomer in rats. Ann. NY. Acad. Sci.
246:135-148.
Huberman, E., H. Bartsch, and L. Sachs. 1975.
Mutation induction in Chinese hamster V79 cells by
two vinyl chloride metabolites, chloroethylene oxide
and 2-chloro- acetaldehyde. Int. J. Cancer. 16:639-
644.
IARC. 1979. International Agency for Research on
Cancer. IARC monographs on the evaluation of
carcinogenic risk of chemicals to man. Vol. 19. pp.
377-438. Lyon, France.
Infante, P. F., J.K. Wagoner, and R. J. Waxweiler.
1976a. Carcinogenic, mutagenic and teratogenic
risks associated with vinyl chloride. Mutat. Res.
41:131-142,
Infante, P. F., J. K. Wagoner, R. J. Waxweiler, A.J.
McMichael, and H. Falk.l976b. Genetic risks of vinyl
chloride. Lancet. 1:734-735.
John, J. A., F. A. Smith, B. K. J. Leong, and B.A.
Schwetz. 1977. The effects of maternally inhaled
vinyl chloride on embryonal and fetal development in
mice, rats and rabbits. Toxicol. Appl. Pharmacol. 39
497-513.
Killian, D.J., D.J. Picciano, and C. B. Jacobson.
1975. Industrial monitoring: A cytogenetic approach.
Ann. N.Y. Acad. Sci. 269:4-11.
B-35
-------
Laib, R.J. and H. M. Bolt. 1977. Alkylation of RNA
by vinyl chloride metabolites in vitro and in vivo:
Formation of l-N'-etheno-adenosine. Toxicology
8:185-195.
Lee, C. C., J.C. Bhandari, J.M. Winston, W.B. House,
R. L, Dixon, and J.S. Woods. 1977. Inhalation
toxicity of vinyl chloride and vinylidene chloride.
Environ. Health Perspect. 21:25-32.
Lee, C.C., J. C, Bhandari, J. M. Winston, W. B.
House, R.L. Dixon, and J.S. Woods. 1978.
Carcinogenicity of vinyl chloride and vinylidene
chloride. J. Toxicol. Environ. Health. 4:15-30,
Lilis, R., II. Anderson, W.J. Nicolson, S. Dauni, A.S.
Fischbein, and I.J. Selikoff. 1975. Prevalence of
disease among vinyl chloride and polyvinyl chloride
workers. Ann. N.Y. Acad. Sci. 246:22-41.
Loprieno, N., R. Barale, S. Baroncelli, H. Bartsch, G.
Bronzctti, A. Cammellini, C. Corsi, D. Frezza, R.
Nieri, C. Leporini, D. Rosellini, and A.M. Rossi. 1977.
Induction of gene mutations and gene conversions by
vinyl chloride metabolites in yeast. Cancer Res. 253-
257.
Maltoni, C., G. Lefemine, A. Ciliberti, G. Cotti, and
D. Carretti. 1981. Carcinogenicity bioassays of vinyl
chloride monomer: a model of risk assessment on an
experimental basis. Environ. Health Perspec. 41:3-
11.
Mancuso, T. F. 1975. Comments for opening of
discussion on "neoplastic effects". Ann. N.Y. Acad.
Sci. 246:251-254.
Marsteller, H. J., W.K, Lelbach, R. Muller and P.
Gedigk. 1975. Unusual splenomegalic liver disease
as evidence by peritoneoscopy and guided liver biopsy
among poly vinyl chloride production workers. Ann.
N.Y. Acad. Sci. 246:95-134.
Miller, A., A.S. Tiersten, M. Chuang, I. J. Selikoff,
and R. Warshaw. 1975. Changes in pulmonary
function in workers exposed to vinyl chloride and
polyvinyl chloride. Ann. N.Y. Acad. Sci. 246:42-52.
Miltner, R., 1984. Personal communication, U.S.
EPA Technical Support Division, ODW, Cincinnati,
OH. Cited in: Technologies and costs for the removal
of volatile organic chemicals from potable water
supplies by environmental science and engineering.
NAS. 1977. National Academy of Sciences. Drinking
Water and Health. Volume 1, National Academy
Press, Washington, DC. pp. 783-787.
Parsons, F., P.R. Wood, and J. DeMarco. 1984.
Transformation of tertrachloroethene and
trichloroethene in microcosms and groundwater,
JAWWA, Vol. 26 No. 2, pg 56f. ,
Picciano, D.J. R.E. Flake, P.C. Gay, and D.J. Killian.
1977. Vinyl chloride cytogenetics. J. Occup. Med.
19:527-530,
Purchase, I.F.H., C.R. Richardson, and D^ Anderson.
1975. Chromosomal and dominant lethal effects of
vinyl chloride. Lancet. 2(7931):410-411.
Radike et al. 1977. Transplacental effects of vinyl
chloride in rats. Annual Report, pp. 183-185.
USPHS-ES-00159. Center for Study of the Human
Environment, Dept. Environ. Health, University of
Cincinnati Medical Center.
Selikoff, I.J., and B.C. Hammond, eds. 1975. Toxicity
of vinyl chloridepolyvinyl chloride. Ann. N.Y. Acad.
Sci., Vol. 246.
Suciu, I., L. Prodan, E. Ilea, A. Paduracu, and L.
Pascu. 1975. Clinical manifestations in vinyl chloride
poisoning. Ann. N.Y. Acad. Sci. 246:53-69.
Symons, J.M. 1978. Interim treatment guide for
controlling organic contaminants in drinking water
using granular activated carbon. U.S. EPA Office of
Research and Development, MERL, DWRD, Cincin-
nati, OH. Cited in: U.S. EPA SNARL Document for
Vinyl Chloride (Draft) and in U.S. EPA May, 1983.
Treatment of volatile organic compounds in drinking
water, Report No. EPA-600/8-83-019, Office of
Research and Development, MERL, DWRD,
Cincinnati, OH.
Til, H. P., H. R. Immel, and V. J. Feron. 1983.
Lifespan oral Carcinogenicity study of vinyl chloride
in rats. Final report. Civo Institutes TNO. Report No.
V 83.285/291099.
Torkelson, R.R., F. Oyen, and V. K. Rowe. 1961. The
toxicity of vinyl chloride as determined by repeated
exposure of laboratory animals. Amer. Ind. Hyg.
Assoc. J. 22:354-361.
U.S, EPA. 1979, U.S, Environmental Protection
Agency. Water related environmental fate of 129
priority pollutants. Office of Water Planning and
Standards, EPA- 440/4-79-029.
U.S. EPA. 1980a. U.S. Environmental Protection
Agency. Vinyl chloride occurrence in drinking water,
food and air. Office of Drinking Water.
U.S, EPA. 1980b. U.S. Environmental Protection
Agency. Ambient water quality criteria for vinyl
chloride. Office of Water Regulations and Standards.
EPA 440/5-80-078.
B-36
-------
U.S. EPA. I985a. U.S. Environmental Protection
Agency. Final draft for the drinking water criteria
document on vinyl chloride (Office of Drinking
Water), TR- 540-162.
U.S. EPA. 1985b. U.S. Environmental Protection
Agency. National primary drinking water
regulations ; Volatile synthetic organic chemicals;
Final rule and proposed rule. Federal Register.
50(219) 46880-46933, November 13.
U.S. EPA. 1985c. U.S. Environmental Protection
Agency. Method 502.1, Volatile halogenated organic
compounds in water by purge and trap gas
chromatography. Environmental Monitoring and
Support Laboratory, Cincinnati, Ohio 45268. June
1985.
U.S. EPA. 1985d. U.S. Environmental Protection
Agency. Method 524.1. Volatile organic compounds
in water by gas chromatography/mass spectrometry.
Environmental Monitoring and Support Laboratory,
Cincinnati, Ohio 45268. June 1985.
U.S. EPA. 1986. U.S. Environmental Protection
Agency. Guidelines for carcinogenic risk assessment.
Federal Register 51(185):33993-34003, September 24.
U.S. ITC, 1983. U.S. International Trade
Commission, Synthetic organic chemicals: United
States production. 1982. USTIC Publication 1422.
Washington, D.C, 20436. 1983.
Verburgt, F.G. and E. Vogel. 1977. Vinyl chloride
mutagenesis in Drosophila melanogaster. Mutat. Res.
48:327-333.
Vogel, T. and P. McCarty. 1985. Biotransformation of
tetrachloroethylene to trichloroethylene, dichloro-
ethylene, vinyl chloride, and carbon dioxide under
methanogenic conditions. Appl. Environ. Microbiol.
49 (5).
Watanabe, P.G., G.R McGowan and P.J. Gehring.
1976a. Fate of (14C) vinyl chloride after single oral
administration in rats. Toxicol. Appl. Pharmacol.
36:339-352.
Watanabe, P.G., G. R. McGowan, E. O. Madrid and
P.J. Gehring. 1976b. Fate of (14Q vinyl chloride
following inhalation exposure in rats. Toxicol. Appl.
Pharmacol. 37:49-59.
Withey, J.R. 1976. Pharmacodynamics and uptake of
vinyl chloride monomer administered by various
routes to rats. J. Toxicol. Environ. Health. 1:381-394.
Withey, J.R., and B.T. Collins. 1976. A statistical
assessment of the quantitative uptake of vinyl
chloride monomer from aqueous solution. J. Toxicol.
Environ. Health. 2:311-321.
Wood, P.R. and J. DeMarco. 1980, Effectiveness of
various adsorbents in removing organic compounds
from water, 1: Removing purgeable halogenated
organics. In: Activated Carbon Adsorption of
Organics from the Aqueous Phase. Volume 2. Ann.
Arbor Science, pp. 85-114.
B-37
-------
-------
Appendix C
Sample Risk Assessment and
Risk Management Exercises
Introduction to Risk Assessment Case
Study on Vinyl Chloride Contamination of
Drinking Water
New information on the toxic properties of a widely
used chemical, vinyl chloride (VC), has just been
published in a major scientific journal. The uses of
VC place it under the jurisdiction of EPA, and a
senior agency policy maker must decide whether the
new information justifies regulatory action. As a first
step the policy maker must determine whether and to
what extent current uses of VC endanger the public
health. The senior policy maker thus assembled a
group of top Agency scientists from various
disciplines —epidemiology, toxicology, biochemistry,
pathology, statistics, chemistry — and posed the
following questions:
1. What types of health hazards might be associated
with VC, and how well are these known?
2. What is the magnitude of human exposure to VC,
and how is the exposure distributed in various
population groups?
3. What is the nature and magnitude of human risk
associated with the various sources of exposure?
The group of scientists collected data to conduct a risk
assessment. In particular, they developed informa-
tion to estimate the likelihood that VC will exhibit
one or more of its hazardous properties under actual
conditions of human exposure. At this stage the
senior policy maker is only concerned with under-
standing the risks of VC and the ways in which that
risk can be characterized. The senior policy maker is
not presently concerned with what has been referred
to as risk management, or the issue of how to regulate
VC if a risk has been identified. Hence, the senior
policy maker is not considering the commercial
importance of VC and the possible regulatory
consequences of reporting a significant health risk.
The senior policy maker believes strongly that it
would not be satisfactory to conclude that no risk
assessment could be performed, or that "more re-
search" had to be conducted before any conclusions
could be reached. Rather, the senior policy maker felt
it was essential that as definitive a statement as was
currently possible be made about the health risk of
VC, and that the uncertainties in the assessment be
identified. The senior policy maker knew it would
have to be decided how to handle the scientific
uncertainties in the risk management decision, but
for now the need was to understand and characterize
the current scientific knowledge of the risks of VC.
Your Role
For this exercise, you will play the role of the senior
policy maker. Your objective is to ensure that you
thoroughly understand the possible health risks
associated with various uses of VC and that you can
convey your understanding to other people. You are
not yet concerned with the ultimate regulatory
question of whether and to what extent these uses
should be controlled or eliminated; you are concerned
with the risk assessment, not risk management.
You will receive various sets of data and analysis
from the team of scientists you have assigned to the
problem. You will conduct an analysis of the
information and its implications for risk. You will
review and evaluate the contents of the document.
You will be asked to formulate some conclusions
based on the data and analysis.
Your review and evaluation will take place within a
small working group. After the various issues are
aired and discussed, the working group (which
collectively represents the senior policy maker)
C-1
-------
should reach a consensus on how best to characterize
the data and the risk. If a consensus cannot be
reached, the alternative views should be expressed.
The conclusions of each working group will then be
compared and contrasted.
Again, at this stage you are concerned only with risk
assessment, not with risk management.
Nature of the Data and Analysis to Be
Reviewed
The report contains a discussion of the nature and
uses of VC, and the known extent of human exposure
to it. The toxicologieal data on VC will be presented
in summary form. You will be asked to examine
several issues relating to the data and reach
conclusions regarding them. This section constitutes
the Hazard Evaluation.
The relationship between exposure to VC and the
risk of adverse health effects (Dose-Response
Evaluation) is the next subject. There may be several
scientifically plausible options for describing this
relationship in the region of human exposure, and
you will be asked to judge the relative merits of these
various options. That is, you will be asked to choose
among them, or formulate a better one.
The third section will contain a summary of data on
the Exposure of various population groups to VC.
Again, several issues arise concerning the
interpretation and use of this information, and it will
be necessary for the senior policy maker to formulate
appropriate conclusions.
In the final step (.Risk Characterization) you will be
asked to present your conclusions regarding the
human health risks posed by VC and the
uncertainties in your knowledge.
At each of the four major steps of this exercise, issues
and data will be presented, and alternative
conclusions will be listed. After discussion, you may
select the conclusion that seems most appropriate; if
none seems appropriate, you should offer your own.
Resource Material
Chapter 3 entitled Principles of Risk Assessment
provides background material needed to assist your
evaluation. You also were given some key principles
and additional background material in Chapters 2,4
and 6.
In addition, each of the following sections contains a
discussion of the key principles directly relevant to
the specific issues under consideration.
A. Background Information on Vinyl
Chloride
Uses of Vinyl Chloride
Raw Materials
• Manufacture of
- in plastics, rubber, paper,
glass and automotive
industries.
- electric wire, insulation
and cables, piping,
industrial and household
equipment, medical
supplies, food packaging
and building supplies.
Chemical & Physical Properties of Vinyl Chloride
• Structure
H-C = C-C1
I I
H H
• Physical State
• Stability
• Solubility in Water
Production in U.S.A.
Human Exposure
• General Population
- Gas
- Degrades rapidly in the
environment
- l.lg/Lat28°C
- 1983 - 7 billion pounds
Worker Populations -
Humans could be exposed
to vinyl chloride in
drinking water, food, and
air. Some people could be
exposed also through
occupational and consumer
usage.
Workers are exposed
during manufacture.
B. Hazard Evaluation
Some Principles for Hazard Evaluation
• The purpose of hazard evaluation is to identify the
types of adverse health effects that may be
associated with exposure to VC, and to
characterize the quality and strength of evidence
supporting this identification.
• The specific hazard of concern in this review is
cancer.
C-2
-------
• Epidemiological studies in exposed human
populations generally are considered to be the best
source of information for hazard identification.
Unfortunately, they are not available for most
substances. Moreover, establishing firm causal
links between exposure and human disease is very
difficult.
• Studies in experimental animals also provide
useful information for hazard identification. Such
studies can be controlled, and thus can more easily
establish causality. Results from such studies
suffer from the obvious limitation that
experimental animals are not the species of
ultimate interest!
• With one possible exception (arsenic), all known
human carcinogens also are carcinogenic in one or
more experimental animal species. Many animal
carcinogens have not been established as human
carcinogens, in most cases because of the lack of
adequate epidemiological data.
• There are biological reasons to believe that
responses in experimental animals could be
mimicked by responses in humans, a proposition
supported by considerable empirical data.
However, other data show that species differ in
response to the same agent.
• It is known that the specific site(s) of tumor
formation in humans may be different from that
observed in experimental animals.
• Data obtained by administering a substance by the
same route of exposure that is experienced by
humans are considered more predictive than data
obtained by a different route. But if tumors form at
internal body sites, the route of exposure may not
be important.
• In general, a varied response in experimental
animals — tumor formation in several species, both
sexes, at several different exposure levels with
increasing response at increasing exposure, and at
multiple body sites — provides more convincing
evidence of potential human carcinogenicity than
does a response that is limited to a single species or
sex, or to a single common site of tumor formation.
A number of studies have been conducted in rats,
mice and rabbits which show that vinyl chloride is
carcinogenic in these species. Statistically significant
increases in the numbers of tumors at a variety of
sites have been reported following both inhalation
and oral exposure.
During the risk assessment case study, we are
focusing upon the results of just one of those studies.
The reasons for this decision should become clear as
you become more acquainted with the data.
The Feron et al. Study
In 1981, an article by Feron et al. entitled "Lifetime
Oral Toxicity Study of Vinyl Chloride in Rats"
appeared in a respected scientific journal (Food and
Cosmetic Toxicology). This paper presented data on
the effects in rats of lifetime oral exposure to vinyl
chloride. The design of the experiments and the major
findings are presented in Tables C-l and C-2.
Remarks on the Feron et al. Study
1. As far as can be determined from the published
Feron et al. article, this study was carefully
conducted and there is no reason to doubt the
accuracy of the reported data.
2. VC increased the incidence of risk of tumors
(number of animals with tumors) in allgroups of
animals given VC in the diet, although the
increase in low-dose males was not statistically
significant.
3. Rats treated with VC by gavage developed the
same tumor types in liver as those treated with
VC in the diet, but lung angiosarcomas were not
apparent with gavage exposure.
4. Rats developed liver tumors following both
dietary and gavage exposures. Lung tumors were
produced only by dietary exposure.
5. Following dietary exposure to VC, females
showed more neoplastic nodules and
hepatocellular carcinomas, whereas males
showed more angiosarcomas in liver and lung.
6. Neoplastic nodules and hepatocellular
carcinomas were proportionally greater than
liver angiosarcomas with dietary exposure to VC,
whereas the opposite was evident with gavage
exposure to VC.
7. Liver and lung tumors observed in treated
animals are rarely formed in untreated (control)
rats of the strain (Wistar) used by Feron et al.
This is particularly important in the interpret-
ation of liver tumor data in rats treated with VC
by gavage as a treatment-related effect.
8. Neoplastic nodules are considered to be a
progression towards hepatocellular carcinomas
and are, therefore, included in the tumor
incidence table. These tumor types are of
different cellular origin, and are thus considered
distinct tumor types from liver angiosarcomas.
C-3
-------
Table C-1. Design of the Feron et al. Study
Species & Route Groups Receiving
of Exposure Vinyl Chloride # of Animals
Rat, dietary
Rat, gavage
Control
Low Dose
Mid Dose
High Dose
-
Male
60
60
60
60
60
Female
60
60
60
60
60
Ami. VC Reed.
Each Day1
0
1.7
5.0
14.1
300.0
Duration of Exp.
(weeks)
104
104
104
104
104
1 The units of "amount received "are milligrams of vinyl chloride (VC) per kilogram of the animal's
body weight.
Note: Gavage is the administration of a substance by means of a stomach tube.
Tablo C-2. Significant Findings from the Feron et al. Study
Following are the groups in which a statistically significant excess of tumors was found. Complete assessment of tumor
formation was made in each sex.
Tumor Incidence (number of
animals with tumors)
Study
Group
Rat,
dietary
Rat,
dietary
Sex Tumors Found3
Male : Liver
a- neoplastic nodule
b- hepatocellular
carcinoma
c-angiosarcoma
Lung
a- angiosarcoma
Female Liver
a- neoplastic nodule
b-hepatocellular carcinoma
c- angiosarcoma
Control
0
0
0
0
2
0
0
Low-
Dose
0
1
0
0
26<=
4
0
Mid-
Dose
7
2
6C
40
39o
IQc
2
High-
Dose
24<:
9<=
270
19°
440
29<=
9<=
Lung
a- angiosarcoma
Rat, Male Liver
gavage «• neoplastic nodule
b- hepatocellular carcinoma
°- angiosarcoma
Rat, Female Liver
gavage a- neoplastic nodule
b- hepatocellular carcinoma
°- angiocarcoma
Doseb
3
1
27
2
0
29
0 Tumors are described both in terms of target organ and tumor type within the target organ. There are three tumor types distinguished in liver.
b Thoro was no matched control group with the treated group given VC by gavage. Thus, statistical comparison could not be done.
c A statistically significant excess of tumors relative to untreated control animals. This means that the difference in tumor incidence between the
treated and control animals is not likely due to chance. Because the only difference between the control and treated animals was the
presence of VC, it is thus likely that the excess tumor incidence is due to this compound. Tumors were found at other sites in both control and
treated animals, but no others occurred in statistically significant excess.
C-4
-------
9. Identification of tumor types in each animal
individually was not given. Therefore, for the
purpose of quantitative risk assessment, animals
with hepatocellular carcinoma also are assumed
to have neoplastic nodules. Therefore, only
neoplastic nodules and angiosarcomas are added
together to derive total liver tumors.
Issues to Be Considered by the Senior Policy Maker
1. How do these data conform (or not conform) to the
principles laid out on pages C-2 and C-3
particularly the last one?
2. In view of these principles, is there any reason to
conclude that VC is not carcinogenic in rats of
both sexes (by dietary and gavage exposures)?
3. Is there any reason to believe that humans would
not be at risk of developing these various tumors,
assuming they were exposed to VC?
4. Is there any way to determine, from the data
given, whether responses in humans are likely to
be similar to those of rats? Males or females?
5. Should the liver tumors be considered relevant to
humans?
6. Should the data obtained by gavage treatment be
considered relevant to human exposure?
Conclusions Regarding VC Carcinogenicity
Which of the following conclusions best characterize
• the evidence you have seen?
1. VC is a human carcinogen.
2. VC is a probable human carcinogen.
3. .VC is a carcinogen at several sites in rats of both
sexes, by both dietary and oral routes of
administration. VC is thus a human carcinogen
and is expected to increase the incidence of lung
and liver tumors in the exposed human
population.
4. VC is a carcinogen at several sites in rats of both
sexes. VC is thus a probable human carcinogen,
although only humans exposed orally are likely
to be at risk. Data obtained when VC was
administered by stomach tube are not relevant to
any route of human exposure. Thus, exposure
through other routes has no identifiable risk for
humans.
5. Although VC is carcinogenic in rats, no data
suggest that it is carcinogenic in humans. The
animal data provide only weak evidence that VC
may be a human carcinogen.
6. Because of the extreme conditions u -)der which
tumors were produced in these animal
experiments, there is no reason to believe VC is a
possible human carcinogen.
7. Other (formulate your own conclusion).
C. Dose-Response Evaluation
The General Problem and Principles Guiding
Approaches to Its Solution
Because of the relative complexity of dose-response
evaluation, the following discussion is substituted for
a statement of key principles.
Recall that animal data showing that a chemical is
carcinogenic usually are obtained in the high ex-
posure region of the dose-response curve: Thus,
animal exposures were in the 1.7 to 300 mg/kg/day
ranges (Table C-l). Human exposure is in the range
of 0.03 to 2.0 ug/kg/day over a range of potential
drinking water concentration levels (Table C-5).
What can be said about risks in the range of human
exposure?
At least three general approaches to this problem
have been proposed by various experts.
Approach 1
Based on general theories of how carcinogens act to
produce cancer (largely derived from experimental
studies and epidemiological data), all finite exposure
levels will produce a finite risk. The magnitude of the
risk will decline as the magnitude of exposure
decline (this is even clear in the animal data).1
If the quantitative relationship between exposure
and risk were known for all exposures, risk to rodents
exposed at very low levels could be predicted from the
measured exposure-risk data. Risks to humans could
be predicted at these very low levels if the relation-
ship between rodent and human susceptibilities were
known. Although these relationships cannot be
known with accuracy, a plausible upper limit on
human risk can be predicted with sufficient accuracy
to be used as a guide to making risk decisions. Actual
human risk is not likely to exceed the upper limit
(although it may), and it may be less.
1 These two sentences are the proper formulation of the "no-threshold"
concept. The "no-threshold" concept does not mean that all finite
exposures will cause cancer; instead, it means that all finite exposures
will increase the probability that cancer will occur.
C-5
-------
Approach 2
The quantitative relationships between high
exposure and low exposure risks in rodents and
between rodent and human risk are not known with
sufficient reliability to be used in risk assessment.
Moreover, there is no reliable theory on which it can
be concluded with assurance that low-level human
exposure (i.e., exposure below the range producing
detectable risks) poses any risk at all. As with other
toxic effects, carcinogenicity will not be initiated
within an individual until a minimum threshold of
exposure is exceeded. In such circumstances, the only
reasonable course is to report the magnitude of the
margin of safety (MOS) by which humans are
protected. MOS is the maximum amount of exposure
producing no measurable tumorigenic response in
animals divided by the actual amount of human
exposure. MOS gives the risk manager adequate
information on which to decide whether exposures
must be reduced or eliminated to provide human
protection. A relatively large MOS is desirable
because it is likely that the threshold for the entire
human population is substantially lower than that
observed in small groups of experimental animals.
Approach 3
Although there is adequate theory and some evidence
to permit the conclusion that humans are at finite
risk at all finite exposure levels, there is insufficient
knowledge to allow prediction of the risk in quanti-
tative terms. The risk assessor should simply attempt
to describe risks qualitatively, perhaps coupling this
description with some information on the potency of
the compound and the magnitude of human exposure.
This type of presentation is adequate for the risk
manager, who should not be concerned with the
quantitative magnitude of risk in any case.
Each of these views, and perhaps others as well, has
some merit. It would seem that the first approach, if
correct, would provide the most useful approach for
decision making. Indeed, it is the approach now used
by EPA and other agencies as well. EPA and the
other agencies emphasize that the predicted
numerical risks are not known to be accurate, but,
because of the nature of the models used to predict
them, they are likely to be upper-bound estimates of
human risk. An upper-bound estimate is one that is
not likely to be lower than the true risk.
For this exercise we shall estimate low exposure risks
using the model currently used by EPA. A model is a
mathematical formula that describes the
relationships between various measures. Two models
are needed to predict low exposure risk:
• A high-to-low exposure extrapolation model is
needed to predict low exposure risks to rodents
from the measured high exposure-high risk data
(Table C-2). EPA currently uses a so-called
linearized multistage model for this purpose. This
model is based on general (not chemical-specific),
widely held theories of the biological processes
underlying carcinogenesis. Application of the
model to the rodent exposure risk data produces an
estimate of the lifetime risk for each unit of
exposure in the low exposure region. This is called
the unit cancer risk. The "linearized" model is used
to ensure that the unit cancer risk is an upper
bound on risk.
• An interspecies extrapolation model is used to
extrapolate from rodent unit risks to human unit
risks. There are empirical data and theory to
support EPA's current use of the assumption that
rodents and humans are at equal risk at the same
exposure measured in milligrams of carcinogen
per square meter of body surface area per day.
EPA's selection of these models is based on the
agency's view that they are the best supported for
purposes of deriving an upper Bound estimate of risk.
Alternative models are available for both these forms
of extrapolation and cannot be ruled out. In most
cases, but not always, use of plausible alternative
models will yield lower estimates of risk than those
predicted by the two described here. Differences can
sometimes be very large, but in most cases
differences are relatively small, especially when the
models are limited to those that are linear at low
exposure.
Further discussions of various models and their
plausibility can be found in the resource material.
Approach Taken for This Exercise
In this exercise we reveal the upper bound of unit
cancer risks predicted for VC using the models
currently preferred by EPA. The effect of using
alternative, plausible models is also described.
Estimates of Upper Bound, Lifetime Unit Cancer
Risks Using Current EPA Models
Application of the EPA models for high-to-low dose
and interspecies extrapolation to the measured
animal cancer data of Table C-2 yields the results
shown in Table C-3.
Estimates of Lifetime Unit Cancer Risks Using
Other Models
Application of other models for high-to-low dose
extrapolation usually yields unit risks equal to or
slightly lower than those in Table C-3, as long as the
other models incorporate the concept that risk
increases in direct proportion to exposure in the low
exposure region (linear models).
C-6
-------
Table C-3. Upper Bounds on Lifetime Unit
Cancer Risks Predicted from
Application of EPA's Preferred
Model to Tumor Data, Table C-2
Species,
Sex
Rat, male
Rat, male
Rat, male
Rat, female
Rat, female
Rat, female
Route of
Exposure
Diet
Diet
Diet
Diet
Diet
D'ft
Tumor Site
Lung
Liver
All tumors
Lung
Liver
All tumors
Unit
Cancer
Risk1
0.11
0.3
0.29
0.058
1.9
2.3
'Risk for an average daily lifetime exposure of 1 unit.
Units are same as those used earlier for describing
the animal exposure (Table C-1) and the human
exposure (Table C-5) (mg/kg bw/day). Risk is
obtained from unit risk by multiplying the latter by the
actual number of units of human exposure; the higher
the unit risk, the higher the risk.
Adoption of certain nonlinear models for high-to-low
dose extrapolation predicts risk about 1,000 to 10,000
times lower than those predicted by use of the EPA
model. The nonlinear models are not widely
recommended.
Dose-Response Evaluation Not Involving Formal
Extrapolation
For those who believe formal extrapolation beyond
the measurable dose-response data should not be
performed, it is important to identify the exposures at
which VC produces tumors and those at which no
tumor excess is found (the "No-Observed-Effect
Level" or NOEL). Table C-4 identifies NOELs from
data in Table C-2.
Issues to Be Considered by the Senior Policy Maker
I. Which of the three possible approaches should be
taken? Explicit estimate of risk? Quantitative
estimate of MOS? Qualitative descriptions only?
Should other approaches be considered?
2. If explicit estimates of unit risks are made,
should only EPA's currently preferred models be
used? Should the results of applying other models
also be displayed?
3. Which species/sex/tumor site data from Table C-3
should be used for unit risk assessment? All,
shown individually as in Table C-3? Only the
data set yielding the highest unit risk? A sum of
all? Other?
4. How should the uncertainties in use of models be
described?
5. Are the observed NOELs true "no-effect" levels?
Could they simply reflect the fact that in
experiments with relatively small numbers of
animals, the failure to observe a statistically
significant increase of tumors is an artifact of the
experimental design, and not a true absence of
biological effect? How should this uncertainty, if
it is real, be taken into account?
Alternative Conclusions Regarding Dose-
Response Evaluation
I. The unit cancer risks listed in Table C-3 are true
upper bound estimates. The true unit risk is not
likely to exceed those determined, may be lower,
and could be zero.
2. The same as the first conclusion, but add: The use
of alternative, plausible models yields unit risks
about 10 to 100 times lower than those from
Table C-3.
3.
4.
Unit risks should be reported for all plausible
models, and the full range of estimates should be
reported without bias.
There is no justification for calculating and
reporting unit risks. What is critical for
understanding the public health importance of
low level exposure to VC is the margin of safety
(MOS). Estimation of the is based on the NOELs
for its carcinogenic effects; these figures are
reported in Table C-4.
Table C-4. No-Observed-Effect Levels (NOELs)
for Chronic Exposure to VC
Study
Group
Rat, dietary
Rat, dietary
Rat, dietary
Rat, dietary
Rat, gavage
Rat, gavage
Sex
Male
Female
Male
Female
Male
Female
Tumor
Liver
Liver
Lung
Lung
Liver
Liver
NOEL'
1.7
None found
1.7
5.0
None found
None found
1 Units are expressed as mg/kg bw/day. "None found" means
that a measurable excess of tumors was not found at all levels
of exposure used in the experiment.
5. Neither unit cancer risks nor NOELs are reliable
indicators of human risk.and neither should be
considered for risk assessment. Dose-response
relations for the human population are not
known for VC; risk should be described in
qualitative terms only.
6. Other (formulate your own conclusion).
C-7
-------
D. Human Exposure Evaluation
Some Principles for Exposure Evaluation
• The purpose of the exposure evaluation is to
identify the magnitude of human exposure to VC,
the frequency and duration of that exposure, and
the routes by which humans are exposed. The
number of exposed people also must be identified,
along with other characteristics of the exposed
population (e.g., age and sex).
• Exposure may be based upon measurement of the
amount of VC in various media (air, water, food)
and knowledge of the amount of human intake of
these media per unit of time (usually per day)
under different conditions of activity.
• Some individuals may be exposed by contact with
several media. It is important to consider total
intake from all media in such situations.
• Because only a limited number of samples of
various media can be taken for measurement, the
representativeness of measured values of
environmental contaminants are always
uncertain. If a sampling is planned adequately, the
degree to which data for a given medium are
representa-tive of that medium usually can be
known.
• Sometimes air levels of pollutants can be
estimated by the use of mathematical models.
Although some of these models are known to be
predictive in many cases, they are not thought to
be so in other cases.
• Standard average values and ranges for human
intake of various media are available and
generally are used unless data for specific agents
indicate such values are inappropriate.
Available Information on Vinyl Chloride
The following information has been summarized
from the human exposure section of the Office of
Drinking Water Criteria Document on Vinyl
Chloride. Use this information in formulating your
risk assessment decision.
Humans may be exposed to vinyl chloride in drinking
water, air, and food. This analysis is confined to these
three media since they are considered to be general
sources common to all individuals. Some individuals
may be exposed to VC from sources other than those
cited here, notably in occupational settings and from
the use of consumer products containing vinyl
chloride.
Unfortunately, data and methods to estimate
exposure of identifiable population subgroups from
all sources simultaneously have not yet been
developed. To the extent possible, estimates are
provided of the number of individuals exposed to each
medium at various VC concentrations. The 70 kg
adult male is used for estimating intake.
Water
Cumulative estimates of the U.S. populations
exposed to various VC levels in drinking water from
public drinking water systems are presented in Table
C-5. Of the approximately 1.3 million people exposed
to levels ranging from 1 to 5 ug/L, 0.9 million (65%)
obtain water from surface supplies. All exposure to
VC in drinking water at levels above 5 ug/L is
expected to be from ground-wate/ sources.
No data were obtained on regional variations in the
concentration of VC in drinking water. The highest
concentrations are expected to be near sites of
polyvinyl chloride production.
Table C-5 also shows daily intake levels of VC in
drinking water estimated at various exposure levels.
The data in the table suggest that the majority of the
persons using public water supplies would be exposed
to intake levels below 0.028 ug/kg bw/day.
Table C-5. Estimated Drinking Water Intake of Vinyl
Chloride
, Persons using supplies
exposed at indicated levels
Exposure
level (ug/L)
a 1
> 5
> 10
> 50
> 70
Population
1,922,000
591,000
118,000
118,000
0
% of total
population
0,9%
0.3%
0.1%
0.1%
0
Intake
(ng/kg/day)
> 0.028
>0.14
>0.29
>1.4
>2.0
Assumptions: 70-kg adult male, 2 L of water per day
Diet
No data were obtained on levels of VC found in foods
in the United States. Therefore, no estimates of the
daily intake of VC from the U.S. diet could be made.
Air
Exposure to vinyl chloride in the atmosphere varies
from one location to another. The highest level of VC
reported in the atmosphere was 2,100 ug/m3. High
levels (> 15 ug/m3) have been detected in other
areas. Normal levels, however, are somewhat lower.
Brodzinsky and Singh (1982) calculated a median air
level of 0.0 ng/m3 (0.0 ug/m3) in each of three types of
areas: rural/remote, urban/suburban, and source-
dominated.
.
C-8
-------
The monitoring data are not sufficient to determine
regional variations in the exposure levels.
Table C-6 describes the daily respiratory intake of
VC from air as estimated using the assumptions
present-ed and the maximum and minimum ambient
levels reported above. Intake calculated using the
maxim-um VC level reported is 690 ug/kg/day; few, if
any, persons are believed to be exposed to that level.
Estimated daily intake under other circumstances is
estimated to be 0 ug/kg/day.
Table C-6. Estimated Respiratory Intake of Vinyl Chloride
Exposure (iig/m3) Intake (yg/kg/day)
Rural/remote (0.0)
Urban/suburban (0.0)
Source dominated (0.0)
Maximum (2100)
0
0
0
690
Assumptions: 70 kg adult male; 23 m3 of air inhaled/day (ICRP,
1975)
Issues to Be Considered by the Policy Maker
• Is there any reason to believe that animal data
obtained from continuous lifetime exposure should
not be used to characterize the risk to people
exposed intermittently?
Conclusions Regarding Human Exposure to Vinyl
Chloride
1. Although the estimates for air and water are
based upon different data and different
assumptions, these data are adequate for
assessing vinyl chloride risks. The risk manager
should be made aware of the uncertainties in
each of the data sets.
2. In addition to conclusion #1, it should be noted
that all the exposures should be added because
some people will be exposed to all sources of vinyl
chloride.
3. None of the exposure estimates is adequate for
use in risk assessment. The risk assess-ment
should describe exposure in qualitative terms
only. Such a qualitative description is
appropriate and adequate for characterizing risk,
which also can be done in qualitative terms only.
4. Other (formulate your own conclusion).
E. Risk Characterization
Purpose
In the last step of risk assessment, the information
collected and analyzed in the first three steps is
integrated to characterize the risks to humans. In
line with the alternative ap-proaches for describing
dose-response relations, at least three approaches can
be taken to this step.
1. Provide an explicit numerical estimate of risk for
each population group by multiplying the unit
risk times the number of units of exposure
experienced by each group:
(unit cancer risk) x (units of exposure) = risk
In this equation, risk is unitless - it is a
probability.
Equation:
Unit risk x Ingestion volume x Body weight x
Conversion of mg to ug x Unit(s) of exposure
2. Provide an estimate of the MOS for each group by
dividing the NOEL by the exposure experienced
by that group.
3. Describe risks qualitatively for each of the
population groups.
Risk characterization also might include some
combination of all three approaches, along with a
description of their relative merits.
It also is essential that the statistical and biological
uncertainties in estimating the extent of health
effects be described in this step.
Attached you will find a discussion of Unit Risk
Assessment for Vinyl Chloride. This document
describes the use of Feron et al. data for the
estimation of a unit risk for oral exposure to vinyl
chloride.
In Table C-7, the risks for each population group
using data from Table C-5 are reported. These risks
are based on the highest unit cancer risk described in
the attached discussion (ai* = 2.3 (mg/kg/dayH for
all tumors combined. If other unit risk figures from
Table C-3 had been used, somewhat lower risks
would result. And, if unit cancer risks derived from
other dose-response models had been used, the risks
shown may be 10 to 100 times lower. The risks in
Table C-7 are thought to be upper-bound lifetime
risks.
Issues to Be Considered by the Senior Policy
Maker
I. Are the results reported in Table C-7 an adequate
characterization of VC risks? What else should be
added?
2. Should risks derived from all the unit risks
reported in Table C-3, the attached discussion
and unit risks obtained using alternative models
also be reported?
C-9
-------
Tabto C-7. Risks In Each Population Group for Risk
Characterization
Source
Risk
Size of
Population
Group
Upper Bound on
Number of Cancer
' Cases over
Lifetime
Drinking Water alone
Oug/L, 0
220 million
Ipgrt-
5 ug.1-
lOpgfl.
50ug/u
70ug/L
7xlO-5
3x 10-"
7X10"<
3x10-3
5X10-3
1.9 million +
591,000
118,000
118.000
0
133
177
83
354
0
3. The risks and number of cases reported in Table
C-7 depend on the assumption that the number of
people exposed and their level of exposure will
remain constant over a lifetime. Is this a
plausible assumption? Can alternative
assumptions be used?
4. Is it important to distinguish routes of exposure?
Should unit risks obtained from the inhalation
data be used only for population groups exposed
by inhalation? Should gavage data be used at all?
5. Is it important to know whether a finite risk
exists at all exposure levels, or whether a
threshold exists?
6. Is it appropriate to estimate the number of cancer
cases expected by multiplying risk times
population size (last column of Table C-7)? What
is more important — risk to an individual, or risk
to a population?
7. What are the biological and statistical
uncertainties in estimating the number of
expected cancer cases? How should they be
estimated and described?
2. The risks shown in Table C-7, as well as those
obtained from use of all other plausible models
and all of the various tumor site data, should be
reported, and all estimates should be given equal
weight. Such a presentation affords the decision
maker a view of the uncertainty in the estimated
risks.
3. Upper-bound estimates of lifetime risks to
humans are those reported in Table C-7. Use of
all other animal data sets and alternative,
plausible risk models would result in prediction
of lower risks, perhaps up to 100 times lower.
These risks are conditional on the assumption the
VC is a probable human carcinogen, based solely
on observations of carcinogenicity in several
species of experimental animals. Uncertainties in
exposure and population estimates are those
described in the exposure assessment section.
4. VC is a probable human carcinogen, based on
observations of carcinogenicity in more than one
animal species. Exposures needed to produce
animal carcinogenicity are many thousands of
times higher than those to which humans are
exposed. The margins of safety by which humans
are protected are shown in Table C-7. Because a
NOEL has not been identified for all the various
carcinogenic endpoints, a greater than usual
MOS should be employed to protect human
beings.
5. VC is a probable human carcinogen, based on
observations of carcinogenicity in more than one
species of animals. Humans may be exposed
through air, water, and during employment. In
general, small numbers of people may be exposed
continuously to very low levels of VC, and a few
groups are exposed intermittently. The
individual risk in the general population is
probably low to moderate, but this translates to a
relatively large number of cancer cases because
of the large population size, etc.
6. Other? Some combination of the others?
Alternative Conclusions
1. Upper-bound risks to humans exposed to VC are
those reported in Table C-7. Although risks
obtained from the use of other models may be
lower, the risks could be as high as those reported
in Table C-7.
C-10
-------
Attached Discussion to Aid Risk
Assessors
Unit Risk Assessment for Vinyl Chloride
The data used to estimate a unit risk for oral
exposure to vinyl chloride are based on the Feron et
al. (1981) study. The statistically significant
increases reported for liver and lung tumors were
considered biologically significant. For the liver
tumors, neoplastic nodules were considered a
progression toward hepatocellular carcinomas, and
these are included in the analysis in Tables C-8 and
C-9. Extrapolations using the linearized multistage
model show values of q*i for the individual tumors
ranging from 8.8 x Id-2 to 1.3 x 10-1 for the males and
from 5.8 x 10-2 to 1.3 for the females. The value of q*i
based on males was 3.0 x 10-1 for liver tumors and 2.9
x 10-1 based on all tumors combined. For the females
the value of q*i based on liver tumors was 1.9 and for
all tumors combined was 2.3. All units of q*i are per
mg/kg/day.
Before proceeding with the unit risk estimates, we
will explain, the total tumor counts in Tables C-8 and
C-9. For the liver, all animals with hepatocellular
carcinomas were assumed to also have the neoplastic
nodules. Thus, only the neoplastic nodules and liver
angiosarcomas were added to derive the total liver
tumors. Otherwise, the totals would have exceeded
the number of animals examined. Also, in adding the
lung and liver tumors, the totals were not allowed to
exceed one less than the number examined.
The result of this latter restriction was to raise the
value of q*x slightly due to increased variance. In
fitting the response data in Tables C-8 and C-9 with
the human equivalent dosages, the human
equivalent dosages were derived by dividing the
corresponding animal dosages by (Wh/Wa)1/3. The
human weight (Wh) was assumed to be 70 kg; the
male rats were estimated to weigh 350 g and the
female rats were estimated to weigh 200 g (Figure C-
1). Thus, the corresponding human equivalent
dosages were 0, 0.29, 0.85, and 2.41 mg/kg/day based
on the male rats, and 0,0.24, 0.71, and 2 mg/kg/day
based on the female rats.
When the response and human equivalent dose data
were fit to the linearized multistage model, the 95%
upper limit on the largest linear term (Table C-9)
was:
q*l = 2.3 (mg/kg/day)-1
To derive an estimate of the 95% lower level of
concentration, d, corresponding to a 95% upper level
of risk, R, the following equation is used:
R = i-e -q'di
where d is the lower limit on dose in mg/kg/day. To
solve for d in ug/L, we use the transformation
1 mg/kg/day x (70 kg/2 L) x 1,000 yg/mg = 35,000 iig/L
If we set R = 10-5 then
d = (35,000/q*i) In (1-10-5) (Ug/L).
For the highest value of q*i = 2.3 (mg/kg/day)-1
(Table C-9), setting R - 10-5 yields a value of d =
0.15 ug/L. Setting R = 10-4 or 10-6 yields values of d
= 1.5 ug/L and d = 0.015 ug/L, respectively. For
comparison purposes only we compare the potency of
vinyl chloride by the diet versus the inhalation
routes. A previous memo we sent you estimated the
95% upper limit of potency for VCM as q*i = 1.7 x 10-2
(mg/kg/day)-1 based on an inhalation study showing
angio-sarcomas and other tumors in rats. That
potency estimate vas derived for water quality
criterion purposes. In that document an inhalation to
inges-tion by gavage relationship of 1 ppm inhaled
= 2.28 mg/kg/day ingested was derived for 200 g rats
based on VCM uptake study. Without that
adjustment for route differences, a direct
transformation based on a 70 kg human breathing 20
m3/day would have yielded a 1 ppm inhaled = 0.76
mg/kg/day relationship and a q*i = 5.2x10-2
mg/kg/day, still 44 times less than the estimate from
the diet study.
In summary, the VCM potency estimates are
reported in Table C-10.
Introduction to the Risk Management
Case Study
You are a group of experts called together by the
water supply manager of a small town to advise her
on a possible case of drinking water contamination.
You will be required to analyze the situation and
make a brief presentation of your findings at a public
meeting. Earlier you were presented with infor-
mation concerning the health risks associated with
exposure to the three compounds. You are aware
that, although the risk assessment is fairly complete,
there are a host of other factors that must be con-
sidered in implementing a permanent solution. These
factors will be a part of your risk management
problem. While risk assessment considers the nature
of the risk, risk management must consider taking
appropriate action to alleviate that risk.
Most of you probably are familiar with the work of
Dr. John Snow in London, 1854. Dr. Snow, through a
very thorough epidemiological study, proved that the
Broad Street pump was the source of an outbreak of
cholera. He did this by statistically correlating
incidence of disease with exposure to drinking water
at that well. This example was an early form of risk
assessment. Later, Snow removed the handle from
C-11
-------
Pablo C-8. Type and Incidence of Statistically Significant Treatment-Related Changes in the Liver and Lung of Male Wistar
Rats Exposed to VCM in the Diet. Values of q^ and Concentration from Multistage Extrapolation Model Included
Number of rats examined0
Liver
Neoplastic nodules
Hepatocellular carcinomas
Angiosarcomas
Total liver tumorsd
Lung
Angiosarcomas
Total animals with tumors6
95% Lower-Limit
Treatment Group Concentration Associated
(mg/kg/day) with Risk b
0 1.7 5.0 14.1 (mg/kg/day)-1 10'4 10'5 io-6
55 58 56 59
0 1 7 23 2-1 * TO'1 16.7 1.7 0.2
0 1 2 8 8.8 x 10-2 39.8 4.0 0.4
0 0 6 27 1.3 X10'1 27.0 2.7 0.3
0 2 13 50 3.0 X10-1 11.7 1.2 0.1
0 0 4 19 1.1x10-1 31.8 3.2 0.3
0 2 17 58 2.9x10-1 12.1 1.2 0.1
Tablo C-9.
» Human equivalent eft = q*, (a) (Wh/Wa>i in (mg/kg/day) -'.
b Concentration in pg/L = (-35,000/q*-, In (1 -R).
c Found dead or killed in extremis or terminally.
d Sum of neoplastic nodules and liver angiosarcomas..
• Total must be at least less than total examined.
Type and Incidence of Statistically Significant Treatment-Related Changes in the Liver and Lung of Female Wistar
Rats Exposed to VCM in the Diet. Values of q*1 and Concentration from Multistage Extrapolation Model Included
Treatment Group (mg/kg/day)
95% Lower-Limit
Concentration Associated
with Risk (iig/L)b
Number of rats examined0
i/ver
Neoplastic nodules
Hopalocellular carcinomas
Angiosarcomas
Total liver tumorsd
Lung
Angiosarcomas
Total animals with tumors6
0
57
2
' 0
0
2
0
2
1.7
58
26
4
0
26
0
26
5.0
59
39
19
2
41
1
42
14.1
57
44
29
9
53
5
56
qV .
(mg/kg/day>
1.3
5.0 x 10'1
8.8 x 10-2
1.9
5.8 x 10-2
2.3
10-4
2.7
70.0
39.8
1.8
60.3
1.5
10'5
0.3
0.7
'4.0
0.2
6.0
0.2
10-6
0.03
0.07
0.4
0.02
0.6
0.02
" Human equivalent q", = q , (a) WhWa)^ in (mg/kg/day) •'.
b Concentration in ug/L - (-35,000/q*, In(i-R).
c Found dead or killed in extremis or terminally.
d Sum of neoplastic nodules and liver angiosarcomas..
° Total must be at least less than total examined.
C-12
-------
Table C-10. VCM Potency Estimates
Route
Potency
q", (mg/kg/day) -1
95% Lower-Limit Concentration
Associated with Risk (ng/L)
10'4 10-s 10'6
Oral
Based on diet study
Based on inhalation study
Inhalation
Based on inhalation study
2.3 1.5 0.15 0.015
1.7x10-2 200 20.0 2.0
5.2 X 10-2 67.3 6.7
0.7
Figure C-1. Average Body Weights of the Extra Controls
Fed the 10%-PVC Diet ad Libitum (-) and of the
Rats Given 300 mg VCM/kg Body Weight in Oil
by Gavagea
Mean Body Weight, g
500
400 i
300
200 -
100
Males
Females
60 100 140
Duration of Experiment, wk
a The weight curves of the rats receiving 0, 1.7, 5.0, or 14.1 mg
VCM/kg body weight/day from the 10% PVC diets fed for 4 hours
each day all lie within the shaded area.
Adapted from Feron et al., 1981.
the pump and observed that, as the people drank
water from other sources, the incidence of cholera de-
clined. This later act was what we are calling risk
management. Dr. Snow took positive action to correct
the problem. Unfortunately, today's drinking water
contamination problems are not solved as readily.
Snow had a relatively simple problem to solve by
modern standards, but remember, he accomplished
this twenty years prior to the discovery of the germ
theory of disease by Koch and Pasteur. The public
health aspect of drinking water has come upon the
reverse of Snow's problem. He knew the risk of
drinking water from the Broad Street pump, but
could not identify the contaminant.
Today we can identify many more contaminants, but
are unable to determine the exact nature of the
potential adverse human health effects. Further,
quantifying those risks is itself a risky business.
Projection of human risk exposure from data on
animal carcinogens would appear to be straight -
forward. But, as you saw in the risk assessment
problem, even the "experts" cannot agree on validity
of extrapolation of animal data to human health
risks. Even the most experienced scientists cannot
predict the exact nature of the risk of exposure to
chemical contaminants.
In the problem described here, the risk assessment
would likely conclude that one contaminant is an
animal carcinogen, another, a human carcinogen,
and the third, a neurotoxin. Large uncertainties
surround the projection of human risks from animal
data. Six or more orders of magnitude (106 or one
million times) of uncertainty are associated with the
use of models extrapolating animal data to human
data. Everyone would feel more comfortable if there
were more certainty in the risk assessment, but there
is very seldom a straight answer to a chemical
contaminant safety issue. All of this uncertainty
becomes part of the evaluation and analysis
conducted in the process called risk management.
Your Role
You, as an expert consultant, must advise the town
manager and recommend an appropriate course of
action to protect the public health, both long and
short term. Specifically, you are concerned with
mitigating people's exposure to the toxic chemicals in
drinking water.
This case study focuses on your ability to use the
information presented in this course to solve a
drinking water contamination problem. The review
C-13
-------
and evaluation will take place with a group of 10 to
15 people. You will realize that there is no one right
or wrong answer and common sense should prevail.
The process by which you arrive at your conclusions
is very important. The group should attempt to come
to a consensus about what action can be taken. If you
cannot come to a consensus, present the alternative
views. The conclusions of each work group will be
compared and contrasted at a final plenary session.
Nature of the Material
You will focus on several types of information.
Results of the previously completed risk assessments
will be reviewed briefly. In addition, both qualitative
and quantitative information will be provided on
various courses of action. This information will
include political and social factors as veil as
treatment, economic and environmental data. You
must consider the interests of various economic and
public interest groups in your recommendation.
The case study package is divided into five sections.
and you have available the Health Advisory
documents for aldicarb, vinyl chloride and
trichloroethylene. The Health Advisory documents
contain occurrence, health effects, analytical
chemistry and treatment data on each chemical. Use
this information as appropriate in formulating your
response to the questions that appear in the latter
sections of the case study. The discussion of drinking
water regulations focuses on proposed rulemaking for
the volatile synthetic organic chemicals and some
pertinent legislative background. This information
should prove useful in organizing your thoughts, but
should not be viewed as providing the exact answer or
constraining your response. Remember, this is pro-
posed rulemaking and you are required to respond
immediately. The following three sections provide
site-specific information, questions to be answered,
and calculations to be performed. It might be helpful
if someone in each group could provide a calculator,
but this is not required. We also will provide a
facilitator for each group. He should not lecture, nor
should you look to him to provide answers.
The focus of this exercise is risk management and
risk communication. Try to use the conclusions from
your risk assessment of the relevant chemicals, as
well as the information provided here and in the
lectures.
A. Background Information on
Chemicals
The Health Advisories for aldicarb, trichloro-
ethylene, and vinyl chloride are located. Table C-ll
provides helpful information for working through
this problem. Additional information concerning the
chemicals will appear as appropriate throughout this
document and in some of the lecture outlines.
B. Drinking Water Regulations:
Statutory and Institutional Concerns
Introduction
In thinking about how to manage a drinking water
contamination incident, you should understand the
framework provided by the Safe Drinking Water Act
as amended through 1986. This Act provides a two
Step approach to setting drinking water standards.
The first step is to set a maximum contaminant level
goal (MCLG), formerly called the recommended
maximum contaminant level (RMCL). EPA must also
set the maximum contaminant level (MCL) as close to
the MCLG as is feasible. Simply put, MCLGs are
health-based goals and MCLs are technology-based
standards. Standards are enforceable and goals are
not.
MCLGs are nonenforceable health goals. MCLGs are
"set at the level at which no known or anticipated
adverse effects on the health of persons occur and
which allow an adequate margin of safety." The
House Report on the Safe Drinking Water Act
provides Congressional guidance on developing
RMCLs (MCLGs):
...the recommended maximum level must be set to
prevent the occurrence of any known or
anticipated adverse effect. It must include an
adequate margin of safety, unless there is no safe
threshold for a contaminant. In such a case, the
recommended maximum contaminant level should
be set at zero level.
The RMCLs (MCLGs) for a number of carcinogenic
volatile organic chemicals were proposed at zero
based on this language. Obviously, the MCL or
enforceable level cannot be zero since zero cannot be
measured. The MCL or enforceable level must be a
non-zero number.
The MCL must be set as close to the RMCL (MCLG)
as is feasible. "Feasible" means with the use of the
best technology, treatment techniques, and other
means available, taking cost into consideration. The
1986 Amendments include language indicating that
these technologies must be tested under field
conditions.'The Amendments also state that
technologies for the control of synthetic organic
chemicals (SOCs) must be at least as effective as
granular activated carbon.
The general approach used in setting MCLs for the
volatile organic chemicals (VOCs) or any other
contaminant is to determine feasibility. This requires
an evaluation of: (1) the availability and cost of
analytical methods, (2) the availability and
performance of treatment technologies, and (3) an
evaluation of the cost and feasibility of achieving
various levels. A brief nontechnical description of
each component of the regulatory analysis follows.
C-14
-------
Table C-11. Information from Health Advisories and Analyses of Water Wells
Chemical Aldicarb Trichlorethylene
Vinyl Chloride
Use/State
Health Effect
Class
Pesticide solid
Colinesterase
inhibitor/CNS
D
Solvent/degreaser
liquid
Animal carcinogen
B2
Manufacturing
additive/gas
Human carcinogen
A
Route of
Exposure/
Occurrence
Chemical
Concentration
Recommended
Standards
Detection Limits
Level Down to
Which to Treat
Short-term
Treatment
Longer-term
Treatment
Oral, dermal/ applied
to soil and crops,
found in ground water
30 ppb (ug/L)
1-Day HA = 10 ng/L
(for 10-kg child)
Lifetime HA with
rel. source cont.
of 20% = 10 ug/L
1.3 ng/L
Bottled water
point of use device for
truck in water
Granular activated carbon
(GAG) adsorption with sulfone
& sulfoxide by products taken
into consideration •
Ingestion, inhalation,
used widely to degrease
machinery, found in
ground water
60 ppb
2.8 U9/L concentration
would produce a cancer
risk of 1 x 10-6 or 1
in a million
0.2 ng/L
bottled water
carbon adsorption,
boil water with adequate
GAC/Aeration
Oral, inhalation.almost
always co-occurs with
TCE, found in ground
water
20 ppb
.015 ng/L concentration
would produce a cancer risk
of 1 in one million (1 x 10-6)
2.6 ng/L short term HA
0.3 pg/L
bottled water
ventilation
Aeration
Analytical Methods
The analytical method constraints include
considerations of precision and accuracy at low (ppb
or parts per billion) levels. The numbers produced by
the analyst must be within some reasonable
proximity of the true value (accuracy) and must be
reproducible (precision).
The analytical methods for the volatile organic
chemicals include gas chromatography (GC) with
either conventional detectors or a mass spectra-meter
(GC/MS). These analytical methods use the purge
and trap technique for extraction from the liquid
phase and concentration on a column containing a
sorbent. The higher-molecular-weight organic
chemicals (e.g., pesticides) generally require
extraction with a solvent (e.g., hexane or methylene
chloride). The sample or solvent extract is injected
into the entrance port of the GC column: Purging of
the volatile chemicals is accomplished using an inert
gas. The organic chemicals of interest are then sorbed
to the wall or special packing material within the
column. The compounds are desorbed from the
column by heating and backflushed into the head of
the GC column. This is followed by separation of
constituents in the GC column and measurement
with a specific detection system. Detection systems
include photoionization and electrolytic conductivity.
The detection system generates an electrical signal
which is amplified and transformed to a peak on a
stripchart recorder. The position and height of the
peak is then compared to internal standards for
identification and quantification.
Each step of this process is subject to some error.
These errors are expressed as precision and accuracy.
For the single lab this is sufficient. But, in developing
national standards, interlaboratory variability must
be considered. In general, EPA defines the method
detection limit (MDL) as the minimum concentration
of a substance that can be measured and reported
with 99 percent confidence that the true value is not
zero. This detection limit differs for different labs,
different instruments, different analysts, and is not
necessarily reproducible over time if all these factors
remain the same. Traditionally, quantification limits
are five to ten times the method detection limit. The
importance of this is that it is not possible to
determine compliance or noncompliance with an
C-15
-------
MCL unless there is reasonable assurance that the
reported value is close to the true value.
The remaining component of the use of analytical
measurements in solving drinking water
contamination problems is that of acceptable
laboratory performance. The criteria for EPA
certified labs for the types of gas chromatography
(GC) analyses under consideration in this problem
are ± 4O percent at concentrations under 10 ug/L
and ± 20 percent at concentrations above 100 ug/L.
Consider these limitations in determining what
levels will be acceptable in solving the case study
problem.
Treatment Technologies
Once the lowest level that can be quantified has been
determined, the next constraint for determination of
the MCL is the performance of the best available
technology (BAT). The obvious first step would be to
list all technologies that have ever been used to
remove a particular compound or class of
contaminants. For example, for the volatile organic
chemicals, there are data available on ozpnation,
ultraviolet irradiation, aeration and adsorption.
Conventional coagulation and softening treatment
provides little to no removal of these compounds.
However, there is limited evidence that pzonation
and ultraviolet irradiation can break down
chlorinated ethylenes and other organic molecules
with double bonds. The kinetics of oxidation of
organic contaminants is not understood well enough
to determine the cost of various levels of removal.
Packed tower aeration and, to a lesser extent,
granular activated carbon (GAG) adsorption, have
been shown to be highly effective (>99.9% removal)
for the removal of volatile organic chemicals. The
BAT determination for the volatile organic chemicals
is then based on these two processes.
Aeration Treatment
The performance potential of a properly designed
packed tower aeration system is quite good for VOC
removal. Both field and laboratory experiments and
theoretical calculations indicate that at the
concentrations generally found in drinking water (a
few hundred parts per billion or less), aeration can
produce treated water with sub-parts per billion
concentrations. Aeration processes provide a fixed
percent removal of contaminants. As a consequence
the concentration in the treated water can be affected
by fluctuations in the raw water concentration.
Volatile organic chemical contamination of ground
waters is generally due to poor waste disposal
practices and many times the exact source can never
be found. The hydrogeological factors affecting the
fate and transport of these chemicals are complex.
Modeling them is an inexact science. As a result,
historic information on changes in concentrations
should be considered in the design of an aeration
treatment system. Traditionally, a safety factor of
two times the raw water concentration has been used
in a conservative design. If these and other design
factors are properly considered, the treated water
should meet a concentration goal below the
analytical quantification levels.
Transfer of volatile organic chemicals from air to
water might be a concern depending on the proximity
to human habitation, treatment plant worker expo-
sure, local air quality, local meteorological
conditions, daily volume of water processed and the
concentration of the contaminant. EPA evaluated a
number of existing and planned packed tower instal-
lations using an air dispersion/human Exposure
model. The results of this evaluation indicated that
lifetime exposure to small amounts of carcinogenic
chemicals in air did not result in a significant in-
crease in individual risk of cancer (generally, less
than one in 106 or 107). These were the highest risks
and occurred for persons exposed to 70 years bf worst-
case air concentration conditions at less than 200
meters from the source. As the distance grows, the
population exposed increases, but the concentration
declines so rapidly that projected cancer risks become
very small. Using very conservative assumptions,
these kinds of analyses resulted in a projection of less
than one possible cancer incidence nation wide over
70 years. Since drinking water contaminated with
the carcinogenic chemicals of concern was the
projected cause of approximately 50 excess cases of
cancer, air emissions from aeration treatment ,
facilities are not a major national concern.
If necessary, control of volatile organic chemical
emissions from packed tower aeration installations is
feasible using air phase GAG adsorption. EPA
currently has full-scale field evaluations of this
technology under way. Preliminary evidence
indicates that installation of this equipment would
approximately double the cost of water treated by
packed tower aeration.
GAC Adsorption Treatment
GAG adsorption removal of most organic
contaminants from drinking water, especially ground
waters, is very good. There are a few exceptions,
including low-molecular-weight compounds such as
vinyl chloride. Experiments with this chemical have
shown removal from water to be erratic using GAC
adsorption columns.
The capacity of carbon for removing a contaminant
from water can be determined empirically. Gener-
ally, GAC adsorption removes the contaminant to
below its detection limit until the capacity of the
fixed bed adsorber is reached. The point at which the
contaminant is detected in the effluent water is
C-16
-------
termed breakthrough. After breakthrough the GAG
may remain in service for some time until the
treatment goal is reached. Carbon is replaced at
intervals of 3 to 6 months or longer in practice.
Background organics, sometimes measured as total
organic carbon or TOG, can increase the amount of
carbon required to treat a given volume of water.
This is especially a problem in surface waters. But,
since the volatile organic chemicals do not occur often
above one part per billion in surface waters, this may
not become a major issue. It also should be noted that
empirical determination of carbon usage rates at the
site takes into account the competitive effects of
background naturally occurring organics (i.e., TOG).
Once the treated water goal is reached by a GAG
treatment system, the carbon must be replaced or
reactivated. Small systems generally have a con-
tract with a supplier who delivers fresh carbon and
removes the spent carbon. The supplier may then
reactivate the carbon far use in waste water
treatment. Larger systems can reactivate the GAG
on-site using heat. Fluidized bed reactivation
furnaces are popular for this taste. This thermal
reactivation process can result in the discharge of
particulates and com-bustion products of both the
fuel and the adsorbed organics to air. Experiments at
Cincinnati, Ohio, revealed that toxic (carcinogenic)
dioxins were in the stack gases of the reactivation
furnace. After-burners typically installed with
reactivation furnaces remove the dioxins and other
air pollutants. These concerns are not likely to limit
the applicability of GAC adsorption as BAT for the
control of organic chemical contaminants in water.
Cost Considerations
The Safe Drinking Water Act requires EPA to take
cost into consideration in setting standards. The
maximum contaminant level will be set as close to
the goal (zero for carcinogens) as is feasible taking
cost into consideration. Tables C-12 and C-13 contain
cost estimates for 99% removal of nine volatile
organic chemicals using GAC and aeration. For
perspective, the average cost of treated drinking
water in the U.S. ranges from about 1 dollar to a
$1.50 per 1,000 gallons. Table C-14 shows the cost of
removing trichloro-ethylene to various
concentrations. Notice that the rate of increase of cost
does not change dramatically as the percent removal
increases, nor are the actual costs significantly
higher than that paid for treated water today. The
cost of removing volatile organic chemicals down to
the analytical quantification level is thererfore
seemingly reasonable.
At the national level, total national costs are an
obvious concern. Table C-16 presents a summary of
the national cost as a function of the selection of
maxi-mum contaminant level. These data show that,
as the level decreases, the total number of systems
required to treat increases and consequently the cost
increas-es. The total national cost was not the major
determ-inant in the selection of the maximum
contaminant level, but was considered in the overall
analysis.
Final Rule
The final rule promulgating maximum contaminant
levels for the nine volatile organic chemicals has not
been published. The EPA may change the numbers or
the methodology used in determining those numbers.
The solution to the risk management problem should
consider that regulations for trichloroethylene and
vinyl chloride are due out shortly and that a rule for
aldicarb and other pesticides is also forthcoming.
But, do not restrict your response to what EPA may
or may not do. In other words, you must take the
Health Advisory and risk assessment/management
problem data and develop your own solutions and
numerical goals.
C-17
-------
Tablo C-12. Cost for 99 percent removal (from
1983 dollars
Compound
Trichtoroethytene
Capital cost
Annual O & M cost
Total cost (0/1 00 gallons
Telrachloroethylene
Capital cost
Annual O & M cost
Total cost (0/1,000 gallons)
Carbon tetrachloride
Capital cost
Annual 0 & M cost
Total cost (0/1 ,000 gallons)
1,2-Dichloroethane
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
Vinyl chloride
Capital cost
Annual 0 & M cost
Total cost (0/1,000 gallons)
r, 1 -Dichloroethylene
Capital cost
Annual O & M cost
Total cost (0/1,000 gallons)
Benzene
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
p-Dichlorobenzene (1,000 ugli to 750 ugh)
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
1. 1,1-Trichloroethane (500 ug/1 to 200 ugh)
Capital cost
Annual O & M costs
Total cost (0/1,000 gallons)
500 pg/L to 5 ug/L) of the
100-500
(0.037 MOD)
69,000
1,400
79.0
67,000
1,200
75.0
66,000
1,200
75.0
84,000
2,400
101.0
60,000
900
66.0
64,000
1,000
71.0
74,000
1,700
86.0
51,000
700
56.0
52,000
700
57.0
nine VOCs using packed tower
Costs by System Size Category"
3,300-10,000
(0.95 MOD)
264,000
18,000
15.5
252,000
15,000
14.2
249,000
15,000
14.0
461,000
37,000
28.5
201,000
11,000
11.0
229,000
13,000
12.5
325,000
23,000
19.2
146,000
8,000
8.1
150,000
8,500
8.2
aeration in August
100,000-500,000
(36.8 MOD)
4,789,000
617,000
9.4
4,607,000
513,000
8.4
4,536,000
509,000
8.3
10,221,000
1,149,000
18.7
3,453,000
377,000
6.2
3,975,000
428,000
7.1
6,538,000
781,000
12.3
2,489,000
283,000
4.6
2,500,000
290,000
4.7
"Number of persons served and million gallons per day
C-18
-------
Table C-13. Cost for 99 percent removal (from 500
adsorption in August 1983 dollars
Compound
Trichloroethylene
Capital cost
Annual O & M cost
Total cost (0/1 00 gallons)
Tetrachloroethylene
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
Carbon tetrachloride
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
1 ,2-Dichloroethane
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
Vinyl chloride
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
7 , 1 -Dichloroethylene
Capital cost
Annual O & M cost
Total cost (0/1 ,000 gallons)
Benzene
Capital cost
Annual O & M cost
Total cost (0/1 000 gallons)
p-Dichlorobenzene (1000 ug/L to 750 ug/L)
Capital cost
Annual O & M cost
Total cost (0/1000 gallons) o
1,1,1-Trichloroethane (500 ug/L to 200 ug/L)
Capital cost
Annual O & M costs
Total cost (0/1 000 gallons)
yg/1 to 5 "g/1) of the
100-500
(0.037 MOD)
24,000
4,500
57.0
24,000
2,800
45.0
24,000
5,700
66.0
24,000
9,400 ,
93.0
NA
NA
NA
24,000
4,600
58.0
24,000
15,00
150
24,000
1,900
38.0
24,000
6,600
73.0
nine VOCs using granular activated
Costs by System Size Category*
3300-10,000
(0.95 MOD)
240,000
86,000
34.0
240,000
45,000
22.0
240,000
85,000
34.0
240,000
150,000
52.0
NA
NA
NA
240,000 .
90,000
35.0
236,000
258,000
83.3
.240,000
22,000
15.0
240,000
100,000
38.0
carbon
100,000-500,000
(36.8 MOD)
9,000,000
710,000
14.0
7,700,000
400,000
11.0
9,800,000
930,000
17.0
11,000,000
1 ,500,00
23.0
NA
NA
NA
9,100,000
740,000
15.0
17,200,000
2,800,000
37.6
5,100,000
230,000
6.9
10,000,000
1,100,000
18.0
"Number of persons served and million gallons per day
C-19
-------
Table C-14. Comparison of Various Levels of Removal of
Trichloroethylene (as percent versus total costs
(cents per thousand gallons)
Total Cost (cents per thousand
gallons)
% removed
50
90
99
Using packed
tower aeration
5.9
8.5
12.0
Using GAC
adsorption
18.5
22.7
25.3
Table C-15. Summary of Impacts of the Regulatory
Options for Controlling Volatile Organic
Chemicals (Federal Register, November 13,
1985, p.46927)
Regulatory Options
Number of Systems Impacted
Cost of Control
Tptal cost ($M)
Annual cost ($M)
Cost of Monitoring
Compliance ($M)
Unregulated ($ M) (1445)
Annual cost per Family ($)
Very small (25-500)
Small (501 -3300)
Medium (3301 -50k)
Large (>50k)
Annual Cancer Cases Avoided
1 ug/L
3,800
1,300
100
—
—
96
47
12
42
5pg/L
1,300
280
21
9
2
91
41
12
32
10 yg/L
800
150
11
—
—
90
40
11
31
Table C-16. Costs Impacts of MCLs at Various Levels
National Cost Annual Cost per Family per Size of System
MCL Opts.
M/L
1
5
10
Estimated #
Systems
Impacted
3.800
1,300
800
($ millions)
Total
capital
1,300
280
150
Annual
100
21
11
Very
small
96
91
90
(dollars per year)
Small
47
41
42
Medium
12
12
11
Large
8
3
1
C-20
-------
C. Background on the Contaminated
Water Supply System
Existing Water System
Population served: 30,000 people
Capacity: 5.1MGD
Average demand: 3.0 MGD
Maximum day
demand:
Source:
Storage:
Treatment:
Constructed:
4.2 MGD
•three wells approximately 500
feet deep
• capacity of each well is 1.8
MGD
•screened between 400 - 500 feet
with gravel pack
• 18" steel casing from 0 - 400 feet
•Portland cement grout from 0 -
200 feet
•all wells are pumped to a
common manifold that flows to
the water treatment plant
•soil profile: 0 -100 ft., sandy soil;
100 - 400 ft., sand clay mixture;
400 - 500 ft., wet sand and
gravel; 500 ft., bedrock
3.5 million gallons
Iron removal using chlorine
oxidation, alum coagulation,
sedimentation, and rapid
pressure sand filtration.
Disinfection (chlorine),
fluoridation and corrosion control
(lime and metallic phosphates)
are also practiced.
1957
Mechanical/Structural
Condition: Excellent
Indebtedness:
Rates:
Major Employers:
None
•$1.05 per thousand gallons —
commercial/industrial
•$0.85 per thousand gallons —
residential
•printing plant (50 people)
•potato farming (4000 acres)
•machinery manufacturing (20
people)
•shopping center (30 people)
•plastic bag manufacturer (10
people)
•soda bottler (50 people)
•US Air Force Base (10,000
including residents)
All of the above employers are on the town water
system (except the Air Force base) and are within
three miles of the water wells. The Air Force base has
its own drinking water treatment plant, which is
supplied by a surface water source.
Water Quality Results
The analyses on the following page were reported by
the State Health Department lab. Since then, repeat
samples have been analyzed and the results were not
found to be significantly different. The health officer
wants you to notify the public immediately, but will
not tell you what to say. He says that no one should
use the water because it contains carcinogens and
other toxic chemicals. This is not all that acceptable
to the town government, since they cannot provide an
alternate water supply in a short time frame.
D. Determining Human Exposure and
Risks
Exposure
In order for human health effects to occur as a result
of environmental contamination, the level of expo-
sure to the contaminant must be high enough to
reach the target organs in toxic concentrations. Some
systems have been designed to directly measure
human exposure to potentially harmful agents, but
they are not generally available for situations like
this. Exposure to possible toxins in drinking water
cannot be determined precisely in the general
population.
In the case at hand, we have three contaminants, two
of which are volatile synthetic organic chemicals
normally used in industry and one is an agricultural
pesticide. This opens up a number of possible means
and routes of exposure for various individuals. First,
a number of people might be exposed to trichloro-
ethylene in the work place, since it is frequently used
to degrease machinery parts. Agricultural workers
might be expose to aldicarb during application to the
fields. These are specialized subpopulations who
might be considered in determining the "safe" dose
for the general population. We might have to do some
research to find approximations for the exposures in
the work place.
• Should we consider occupational exposures in
determining a "safe" level in drinking water?
• Which people might be receiving occupational
exposure? (see major employers list). Why?
C-21
-------
Well #1
Well #2
Well #3
Parameter
Iron (mg/L)
pH
Alkalinity (mg/L)
Vinyl chloride (yg/L)
Trichloroethylene (ug/L)
Aldicarb (total) yg/L
Total Organic Carbon (mg/L)
raw
3.0
6.0
10
40
50
30
3.0
treat
0.05
7.8
110
20
60
30
1.0
raw
2.2
5.9
14
14
30
30
2.1
treat
0.05
7.8
110
20
60
30
1.0
raw
2.0
6.2
12
6
100
30
1.0
treat
0.05
7.8
110
20
60
30
1.0
Concentrating on exposure in the home, we have
three major routes of exposure: breathing, oral
consumption, and dermal exposure. We generally
assume that the average adult drinks 2 L per day and
breathes 20 cubic meters of air. Another standard
assumption for volatile contaminants is half of the •
exposure is due to volatilization.
• For which contaminants might sources of exposure
other than drinking water be a concern? Name the
sources. What are the routes?
• Would a 20% relative source contribution from
drinking water be a satisfactory assumption in
this case?
• Is there any way for the residents to mitigate some
of the exposure? Would boiling the water help?
How should the boiling be done?
• The town has a central sewer system with an
activated sludge treatment system. The activated
sludge process includes 4 to 5 hours of vigorous
aeration of the waste water. What is the ultimate
sink (air, water, or land) for each contaminant?
Risks
In the risk assessment case study and the risk
communication video tape you learned some basic
principles that now need to be applied to risk
management.
• In layman terms, describe the individual and
population risks incurred from various sources of
exposure. Describe the fate and transport of the
contaminants and the relationship of this to the
human risk of disease.
• How did you calculate individual and population
risks for this exercise?
• What are your target numbers for correction?
• How would you quantitate and articulate the
uncertainties surrounding your risk estimates?
V. Options Available for Reducing
Risk
Short-Term
• Point-of-use carbon treatment units @ $400 per
year per home.
• Bottled water delivered to the doorstep @ $600 per
home per year.
• Issue a boil water order @ $ 0 per year.
• Do nothing @ $ 0 per year.
Long-Term
• Regional water supply with the Air Force @
$500,000 per year (this water contains an annual
average concentration of 98 pg/L of total trihalo-
methanes).
• Drill new wells @ $200,000 per year (extensive
studies would be required to find an uncon-
taminated source).
• Install point-of-entry GAG adsorption treatment
units in each home @ $1,000,000 per year.
• Install central GAG treatment to meet the
following levels of trichloroethylene:
1.0 ug/L @ 19.5 cents per thousand gallons
5.0 ug/L @ 19.3 cents per thousand gallons
25.0 ug/L @ 19.0 cents per thousand gallons
C-22
-------
• Install central packed tower aeration treatment to
meet the following levels of trichloroethylene:
1.0 ug/L @ 5.0 cents per thousand gallons ,
5.0 ug/L @ 2.9 cents per thousand gallons
25.0 ug/L @ 3.7 cents per thousand gallons
• Install central packed tower aeration and GAG
adsorption to meet the following levels of
trichlorethylene and aldiearb:
1.0 ug/L @ 22.1 cents per thousand gallons
5.0 ug/L @ 20.0 cents per thousand gallons
10.0 ug/L @ 18.3 cents per thousand gallons
Questions
• Which short and long term option (one of each)
would you select? Why?
• What is the total annual cost of each selected
option?
• What are some possible secondary impacts of the
selected options?
*U.S. GOVERNMENT PRINTING OFFICE: 1 99 3 -750- 00960116
C-23
-------
-------
-------
(D
cn
0)
m
1
t\3
tn
^
c a. 3;
•D to ^
111
|-o|
3- o P
It I
st?
=1 tn o
|on
r* < o
Pleasei make all
detach or copy,
left-hand corner.
a
(D O
QI X
Tfl
a>
3-3
-* o
5?
I-
II
> me
"
3C/)
r-«
CD QJ
S «>
0) (/>
TJ
3
(D
O
= 3 DO
=: _ rn r-
nm^x
^\ If "I !
9 o>m
01 ;g
a
------- |