&EPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/600/R-94/177
September 1994
Chemical Hazard
Evaluation for
Management Strategies
A Method for Ranking and
Scoring Chemicals by
Potential Human Health and
Environmental Impacts
-------
-------
EPA/600/R-94/177
September 1994
CHEMICAL HAZARD EVALUATION FOR MANAGEMENT STRATEGIES:
A METHOD FOR RANKING AND SCORING CHEMICALS BY POTENTIAL
HUMAN HEALTH AND ENVIRONMENTAL IMPACTS
Gary A. Davis, Lori Kincaid, and Mary Swanson
Dr. Terry Schultz, Dr. John Bartmess, Barbara Griffith, and Sheila Jones
The University of Tennessee
Center for Clean Products and Clean Technologies
Knoxville, Tennessee 37996-0710
EPA Cooperative Agreement No. CR-816735-01-1
Project Officer:
Emma Lou George
Waste Minimization, Destruction and Disposal Research Division
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTlbN AGENCY
CINCINNATI, OHIO 45268
Printed on Recycled Paper
-------
DISCLAIMER
The information in this document has been funded wholly by the United States
Environmental Protection Agency under Cooperative Agreement CR #816735-01-0, to the
University of Tennessee's Center for Clean Products and Clean Technologies. It has been
subject to peer and administrative review, and has been approved for publication as an EPA
document. Mention of trade names or commercial products does not constitute endorsement
or recommendation for use. •• '••
11
-------
FOREWORD
Today's rapidly developing and changing technologies and industrial products and
practices frequently carry with them the increased generation of materials that, if improperly
dealt with, can threaten both public health and the environment. The U.S. Environmental
Protection Agency is charged by Congress with protecting the Nation's land, air and water
resources. Under a mandate of national environmental laws, the Agency strives to formulate
and implement actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. These laws direct EPA to perform
research to define our environmental problems, measure the impacts, and search for
solutions.
The Risk Reduction Engineering Laboratory is responsible for planning,
implementing, and managing research, development and demonstration programs to provide
an authoritative, defensible engineering basis in support of the policies, programs, and
regulations of the EPA with respect to drinking water, wastewater, pesticides, toxic
substances, solid and hazardous wastes, and Superfund-related activities. This publication is
one of the products of that research and provides a vital communication link between the
researcher and the user community.
This report, Chemical Hazard Evaluation for Management Strategies: A Method for
Ranking and Scoring Chemicals by Potential Human Health and Environmental Impacts,
funded through the Pollution Prevention Research Branch, is a major project in the area of
the Cleaner Products Program in researching methods to support the design and development
of products whose manufacture, use, recycle and disposal represent reduced impacts on the
environment.
This report presents a method for chemical ranking and scoring, designed for priority
setting, to identify specific chemicals as priorities for assessment of safer substitutes for
major uses. This methodology was developed for use in the project The Product Side of
Hazardous Waste Reduction: Evaluating the Potential for Safe Substitutes. A report on this
project is published under separate cover. Risk-based chemical ranking and scoring
combines an assessment of both the toxic effects of chemicals and the potential exposure to
those chemicals, to provide a relative evaluation of risk. Risk assessment is an integral part
of the environmental equation for successful protection and sustainability. The reader is
encouraged to contact the authors or project officer for more information concerning this
project and report.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
111
-------
ABSTRACT
Between 60,000 and 100,000 of the over than 8,000,000 chemicals listed by the
Chemical Abstracts Services Registry are commercially produced and are potential
environmental pollutants. Risk-based evaluation for these chemicals is often required to
evaluate the potential impacts of chemical releases, for priority setting for regulatory action,
for business decisions and to set priorities for pollution prevention. During the last decade,
there have been vast improvements in the methods used to assess chemical toxicity and
environmental fate and to interpret these data within a risk assessment framework. There is
still a need, however, for generally accepted and widely used tools for setting priorities and
providing consistency across environmental programs.
Risk ranking and scoring systems can be used to focus attention and resources on the
most significant hazards posed by industrial facilities, products or hazardous material sites.
Risk-based chemical ranking and scoring combines an assessment of both the toxic effects of
chemicals (human and/or environmental) and the potential exposure to those chemicals to
provide a relative evaluation of risk.
This method provides an approximate ranking of direct chemical hazards to human
health and the environment based on their relative toxicity and the potential for exposure.
The method does not include an evaluation of secondary global impacts such a ozone
depletion and global warming.
An algorithm has been developed to combine and weight evaluation criteria to. provide
a working tool that ranks chemicals according to their potential human health and ecotoxic
effects, and their potential environmental persistence and bioaccumulation. This report
presents methodology for doing ranking at a first, or screening-level, tier. . „ • '
This report was submitted in partial fulfillment of Cooperative Agreement
CR #816735-01-0 by the University of Tennessee's Center for Clean Products and Clean
Technologies, under the sponsorship of the U.S. Environmental Protection Agency. This
work covers a period from September 10, 1990 to September 9, 1994, and was completed as
of August, 1994.
IV
-------
TABLE OF CONTENTS
page
CHAPTER 1 INTRODUCTION , . 1
MAJOR RESEARCH TASKS 2
A PRIORI CONDITIONS 2
TIERED APPROACH 2
CHAPTER 2 SOURCES OF TOXICITY AND EXPOSURE DATA 5
THE USE OF STRUCTURE ANALYSIS 5
CHEMICALS SELECTED FOR EVALUATION . 6
INORGANIC CHEMICALS . 7
CHAPTER 3 DEVELOPMENT OF SCORING CRITERIA 9
HUMAN HEALTH EFFECTS 9
Acute Effects 9
Chronic Effects 11
ENVIRONMENTAL EFFECTS 12
Terrestrial Effects 12
Aquatic Effects 12
EXPOSURE PARAMETERS 13
Persistence 13
Bioaccumulation 14
Physicochemical Properties 15
CHAPTER 4 THE ALGORITHM 19
OVERVIEW 19
HAZARD VALUE 21
CORRELATION OF SCORING CRITERIA 22
WEIGHTED HAZARD VALUES . . . . 23
CHAPTER 5 RESULTS AND DISCUSSION . 25
DEMONSTRATION OF THE ALGORITHM 25
SENSITIVITY ANALYSIS 26
Effect of Missing Data 26
Excluding "Other Specific Effects" 28
Effect of Varying the Weighting of Endpoints 30
UNCERTAINTIES 30
SELECTION OF PRIORITY CHEMICALS ; 35
CONCLUSIONS AND RECOMMENDATIONS 36
REFERENCES
39
-------
TABLE OF CONTENTS, continued
APPENDIX A DATA SELECTION AND DETERMINATION OF HAZARD
VALUES
A.I HUMAN HEALTH EFFECTS , . ........ A-l
A. 1.1 Acute Effects . . .,..., . . .... . . A-l
A.1.2 Carcinogenicity A-5
A. 1.3 Other Specific Effects . A-10
A.2 ENVIRONMENTAL EFFECTS . ...... . . „ . ...... , A-12
A.2.1 Terrestrial Effects , . . ., . . . . .. A-12
A.2.2 Acute Aquatic Effects ... . . . : . . . ... . . . . . . A-12
A.2.3 Fish Chronic Toxicity ,. . ...... . . .,.,,,,. . . A-14
A.3 EXPOSURE PARAMETERS. . . ........ . . . . - . . . • A-19
A.3.1 Persistence , . .,>.„.,*..,...-..-. - - - - A-19.
A.3,2 Bioaccumulation. .,....,,,..,....--- . . . A-22
A.4 WEIGHTING BY RELEASES ,..,.,.. . . . . , - . . .. A-24
A.5 REFERENCES .....,,,,... . . . . .. . . > • • • - • • A-26
APPENDIX B TRI CHEMICALS AND HIGH-VOLUME PESTICIDES . . . , B-l
APPENDIX C RANKING RESULTS: HORIZONTAL TABLES . . , , . . . . C-l
APPENDIX D RANKING RESULTS: CHEMICAL SCORES „..,..... D-l
VI
-------
LIST OF TABLES
Table No.
page
1. TRI Inorganic Chemicals and Surrogate Compounds . . 8
2. Toxicological Endpoints 10
3. Exposure Parameters. . 10
4. Simple Correlation Coefficients (r) for Final Value of Algorithm
versus Parameter. . . 23
5. Top 30 Ranked Chemicals from Algorithm (default HV to zero for missing
data) 27
6. Number of Measured, Estimated and Missing Data Points . 28
7. Top 30 Ranked Chemicals From Algorithm, Sensitivity Analysis for Missing
Data (weighted by releases) 29
8. Top 30 Ranked Chemicals From Algorithm, Sensitivity Analysis for "Other
Specific Effects" (weighted by releases) 31
9. Top 30 Ranked Chemicals for Various Endpoint Weightings (not weighted by
releases) 32
10. Chemicals With Missing Data . . 34
A-l. IARC Carcinogen Classification System A-5
A-2. 1986 EPA Carcinogen Classification System A-7
A-3. Comparison of EPA and I ARC Rating of 31 Carcinogens A-9
A-4. Carcinogenicity Hazard Values . A-9
A-5. Data Sources for "Other Specific Effects" Cited in Roadmaps A-11
VII
-------
LIST OF FIGURES
page
i.
2.
A-l.
A-2.
A-3.
A-4.
A-5.
A-6.
A-7.
A-8.
A-9.
A-10.
A-ll.
A-12.
Predicted K™ vs Experimental Kow . , . . . . . , . . , , . - . .... . , .. 16
Conceptual Model of Chemical Hazard Ranking Method , ... ..',.. . , ,20
Decision Tree for Oral-LD.JO Data Selection V. . . . . . . . , . . . ..','. . A-2
Decision Tree for Inhalation LC50 Data Selection•, . . . , . . . :'•. . .'.-•, . A-3
Decision Tree for Oral LD50 Hazard Value .,,.....>. A-4
Decision Tree for Inhalation LC50 Hazard Value . , -,- • • - - • • A-6
Decision Tree for Carcinogenicity Hazard Value. . . . . . / . . . . . ; . . A-8
Decision Tree for Fish L'C50 Data Selection ....,.......,,.-. A-13
Decision Tree for Aquatic LC50 Hazard Value . . . . . . A-15
Decision Tree for Calculating Fish NOEL. .............. . . . A-16
Decision Tree for NOEL Hazard Value. . . . ... ...... . . . .'. . . . A-18
Decision Tree for BOD Half-life Hazard Value . . A-20
Decision Tree for Hydrolysis Half-life Hazard Value ............. A-21
Decision Tree for BCF Hazard Value. ................... A-23
Vlll
-------
LIST OF ACRONYMS
ATSDR: Agency for Toxic Substances and Disease Registry
BCF: bioconcentration factor
BOD: biological oxygen demand
CLSES: Center for Lake Superior Environmental Studies
CMR: Critical Materials Register
DWCD: Drinking Water Criteria Documents (EPA)
EC50: median effect concentration; the concentration at which 50 percent of the
test population exhibit a specified response during a specified time period
EPA: United States Environmental Protection Agency
GENETOX: Genetic Toxicity Chemical Information System (on-line database)
HAD: Health Assessment Document (EPA)
HEA: Health Effects Assessment (EPA)
HEED: Health and Environmental Effects Document (EPA)
KEEP: Health and Environmental Effects Profile (EPA)
HSDB: Hazardous Substance Data Bank, National Library of Medicine (on-line
database)
I ARC: International Agency for Research on Cancer
IPCS: International Programme on Chemical Safety
K^: octanol-water partitioning coefficient
LC50: median lethal concentration; the concentration at which 50 percent of the test
population die during a specified time period
LD50: median lethal dose; the dose at which 50 percent of the test population die
during a specified time period
NOEL: no observable effect level
POTW: publicly owned treatment works
QSAR: quantitative structure-activity relationship
RTECS: Registry of Toxic Effects of Chemical Substances (on-line database)
RWF: release weighting factor
SAR: structure-activity relationship
SARA: Superfund Amendments and Reauthorization Act
SMILES: Simplified Molecular Input Line Entry System
TRI: Toxic Release Inventory
wHV: release-weighted hazard value
WMS: Wet Milieugevaarlijke Stoffen
WOE: weight of evidence
IX
-------
-------
CHAPTER 1
INTRODUCTION
Between 60,000 and 100,000 of the over
8,000,000 chemicals listed by the Chemical
Abstracts Services Registry are commercially
produced and are potential environmental
pollutants. Some kind of risk-based evaluation
for these chemicals is often required to evaluate
the potential impacts of chemical releases, for
priority setting for regulatory action, for
business decisions and to set priorities for
pollution prevention. During the last decade
there have been vast improvements in the
methods used to assess chemical toxicity and
environmental fate and to interpret these data
within a risk assessment framework. There is
still a need, however, for generally accepted
and widely used tools for setting priorities and
providing consistency across environmental
programs. To date, we have relied upon a
multitude of approaches, some lacking any
scientific basis. Chemical have been selected
for some regulatory programs, for example,
with little systematic evaluation.
Risk ranking and scoring systems can be
used to focus attention and resources on the
most significant hazards posed by industrial
facilities, products or hazardous material sites.
Risk-based chemical ranking and scoring
combines an assessment of both the toxic
effects of chemicals (human and/or
environmental) and the potential exposure to
those chemicals to provide a relative evaluation
of risk. Along with toxicity and exposure,
ranking and scoring systems may include some
measure of economic impact and/or societal
value.
Risk-based chemical ranking and scoring
combines an assessment of both the toxic
effects of chemicals and the potential exposure
to those chemicals to provide a relative
evaluation of risk.
The University of Tennessee Center for
Clean Products and Clean Technologies
developed the chemical ranking and scoring
method in this report under Environmental
Protection Agency (EPA) Cooperative
Agreement CR 816735, The Product Side of
Hazardous Waste Reduction: Evaluating the
Potential for Safe Substitutes. The method was
designed for priority setting, to identify specific
chemicals as priorities for assessment of safer
substitutes for major uses. The method
-------
provides an approximate ranking of direct
chemical hazards to human health and the
environment based on their relative toxlcity arid
the potential for exposure. The method does
not include an evaluation of secondary global
impacts such as ozone depletion and global
wanning.
MAJOR RESEARCH TASKS
In the development of the chemical ranking
method, three major research tasks were
performed:
» compiling available experimental data and
selecting estimation methods for those
instances when experimental data were
absent;
• formulating scoring criteria, which,
individually or in combination, could be
used to estimate the toxic effects of
chemicals and the potential for exposure;
and
« developing an algorithm to combine and
weight evaluation criteria to provide a
working tool that ranks chemicals according
to their potential human health and ecotoxic
effects, and their potential environmental
persistence and bioaccumulation.
The method was demonstrated using the
chemicals for which toxic chemical release
reporting is made in the Toxic Release
Inventory (TRI) as required under Section 313
of Title III of the Superfund Amendments and
Reauthorization Act (SARA) of 1986. Selected
high-volume pesticides, as determined by
annual pesticide usage data, were also included.
Chapter 2 of this report presents an
overview of the sources of toxicity and
exposure data used to develop the chemical
ranking and scoring method. Chapter 3
discusses the types of scoring criteria selected
for the model, including human health effects
criteria, environmental effects criteria and
exposure parameters. Chapter 4 is a detailed
description of the algorithm. Chapter 5
presents the results of the algorithm when it is
demonstrated on the TRI chemicals and high-
volume pesticides. Several a priori condition^
that were incorporated into the design of the
model, and the tiered approach envisioned for
the model are discussed below.
A PRIORI CONDITIONS
Several a priori conditions were
incorporated into the development of this
scheme. First, it was determined that whatever
the framework of the chemical ranking method,
the final tool was to be sufficiently flexible so it
could be modified as experience was gained
and the validation process progressed. Second,
at no time was the process to become so
mechanical as to be isolated from expert
judgment. Finally, although it was not
necessary to separate human health and
environmental effects according to different
endpoints, this would make the algorithm more
transparent. Because it would be easier to
categorize these endpoints on the front end, the
aggregation of information would be done late
in the processing.
The screening tier was designed to avoid
false negatives, including rather than
eliminating chemicals of possible concern.
The confirmation tier should be designed to
avoid false positives and identify only
chemicals of concern.
TIERED APPROACH
The quantity and intricacy of the infoririation
required for a complete assessment of each
chemical, as well as the time and resources
needed to procure and process this information,
can be prohibitive. Thus, a tiered approach
-------
CHAPTER 1: INTRODUCTION
negatives); the confirmation tier should be
designed to identify only chemicals of concern
(avoid false positives). The information used at
the screening tier was designed to be a partial
assessment and sufficient to prioritize
pollutants. Using the results of the screening
tier, one should be able to identify priority
chemicals and move them to the yet-to-be-
designed confirmation tier, or identify
chemicals of lower concern and remove them
from further consideration. The information
used at the confirmation tier should be designed
to be sufficient to render a full assessment or
flag the chemical as having insufficient data for
proper analysis. In this way, information used
in the screening tier will be carried forward to
the confirmation tier.
was adopted with the method presented here
being the first, or screening-level, tier. The
advantage of the tiered approach is that it
reduces the number of chemicals being
evaluated as the depth, breadth and quality of
the required information increases. While any
number of tiers may be employed, a two-tiered
approach that consists of a screening tier and a
confirmation tier should be sufficient.
The screening tier was formulated to rely on
more readily available and/or easily estimated
information, while the confirmation tier should
be formulated to define the potential for
specific human health, environmental and/or
global effects. Also, the screening tier was
designed to include rather than eliminate
chemicals of possible concern (avoid false
-------
-------
CHAPTER 2
SOURCES OF TOXICITY AND
EXPOSURE DATA
Although some data are available in the open
literature for many of the chemicals listed in the
TRI, complete quantification of even a limited
number of toxicological endpoints is rare. One
of the main obstacles to the development and use
of any chemical ranking or scoring system is the
problem of what to do about this missing data.
The types of scoring criteria used and the design
of the algorithm for combining the criteria
depend in large part on whether experimental
data are available or can be estimated with an
acceptable degree of accuracy.
One of the main obstacles to the development
and use of any chemical ranking or scoring
system is the problem of missing data.
Some chemical ranking and scoring systems
incorporate expert judgment to fill data gaps,
either through analytical tools based on structure
analysis, or through ad hoc expert judgment.
Other systems avoid the use of non-experimental
data to fill data gaps by defaulting to endpoints
for which data exists or relying upon the most
sensitive endpoint asi the measure of chemical
risk. (See Davis, et al., (1994b) for a
comparison of data selection approaches for
approximately 50 ranking and scoring systems.)
The latter method presents obvious problems
when little, or no experimental data exist on a
chemical. It also does not necessarily encourage
further testing of compounds, since it is unlikely
that a score for a chemical can be lowered by
filling data gaps.
The chemical ranking method described hi
this report relies on peer-reviewed experimental
data from sources such as the Hazardous
Substances Data Bank (HSDB) whenever
possible. In the absence of experimental data,
structure analysis is used to estimate missing data
to ensure that highly toxic chemicals do not
receive a low ranking simply because they have
not been tested. (For further information
concerning the database, contact the University
of Tennessee Center for Clean Products and
Clean Technologies)
-------
THE USE OF STRUCTURE ANALYSIS
Structure analysis can be qualitative or
quantitative in nature. Qualitative structure
analysis involves evaluation of the molecular
structure of a chemical and the use of qualitative
correlations of particular molecular substructures
and/or functional groups with specific effects. In
this model, qualitative structure analyses are
classified as structure-activity relationships
(SARs). SARs are often used to make judgments
concerning the potential health effects of a
pollutant. Quantitative structure analysis
involves the development and computerization of
quantitative structure-property,
property-property, structure-activity and
property-activity relationships, hereafter referred
to collectively as quantitative structure-activity
relationships or QSARs. QSARs are widely
accepted methods to estimate missing data for
many endpoints, particularly physicochemical
properties and environmental effects.
Both QSARs and SARs are based on the
working hypothesis of "Guilt by Association,"
which says chemicals with similar molecular
structures have similar physicochemical
properties and, therefore, similar biological
activities. Despite potential short-comings,
structure-activity is presently the best means of
supplying missing data. This is certainly the case
if the processing of information takes on a less
quantitative nature where an order or half-order
of magnitude is sufficiently accurate.
Additionally, data for some toxicological
endpoints, such as developmental or reproductive
effects, are often scattered and fragmentary,
which can necessitate a more qualitative (i.e.,
yes or no) evaluation. Such evaluations lend
themselves nicely to the use of qualitative
structure analyses where the presence of a
particular substructure and/or function group is
correlated with a specific toxicological endpoint.
Although relationships between chemical
structure and biological activity were noted as
early as the 1860's, work by Hammett in the
1930's, Taft in the 1950's and Hansch in the
1960's, provided the cornerstones on which
QSAR methodologies in the United States have
been built for the past 20 years. A number of
QSARs are available to estimate a variety of
physicochemical endpoints. While they vary in
accuracy, especially when dealing with
structurally complex molecules, they are, by and
large, of good quality. Because of the general
lack of both human health and environmental
effects data, predictions based on QSARs have in
recent years played an ever more important role
in the acquisition of these data.
There are a variety of published methods for
both qualitative and quantitative estimation of
toxicological endpoints. The specific methods
used in this model are discussed in Chapter 3 and
Appendix A of this report. The quality and
amount of data on which these relationships are
based is typically less than for physicochemical
properties. Therefore, the accuracy of such
predicted values varies markedly. Good
predictions can be attained, however, for many
endpoints, especially effects in the aquatic
environment. Structure analysis techniques have
not yet been developed for all toxicological
endpoints. Moreover, some existing techniques
have yet to be validated and/or are limited to a
narrow range of chemicals. Whenever used, the
limitations of these procedures with respect to
accuracy should be kept in mind.
CHEMICALS SELECTED FOR
EVALUATION
When this model was developed, 158
chemicals were selected for evaluation, 140 from
the 1989 TRI and 21 high-volume pesticides (3
of the pesticides happened to be already included
in the chemicals selected from the TRI). The
method of choosing which of the more than 270
chemicals in the 1989 TRI to use for
development of the algorithm was based on the
quantities released. This selection of chemicals
based on release and transfer quantities could be
considered a preliminary screening tier. The 21
pesticides were selected from annual usage
-------
CHAPTER 2: SOURCES OF TOXICITY AND EXPOSURE DATA
estimates of the largest volume conventional
pesticides in the United States, prepared by EPA
(EPA, 1988b; Aspelin, etal., 1992). Appendix
B presents a list of all TRI chemicals and
indicates which were selected for this evaluation.
Appendix B also lists the high-volume pesticides
that were selected.
The 1989 TRI gives chemical release and
transfer quantities in the categories: 1) fugitive,
or non-point, air emissions, 2) stack, or point,
air emissions, 3) water discharges, 4) land
releases, 5) underground injection releases, 6)
transfers to publicly owned treatment works
(POTW), and 7) transfers to other off-site
locations. In each of these seven categories, in
descending order based upon the amount released
or transferred in each mode, the pounds released
or transferred were summed until enough
chemicals had been added up to give 99 percent
of the total releases or transfers reported. The
chemicals in this 99 percent group became part
of the study group; those that contributed to the
last one percent of releases were set aside for
study at another time.
This procedure was done for all seven
categories as well as for total releases and
transfers. Any chemical that was part of the 99
percent of releases or off-site transfers in any
category was selected. Some chemicals
obviously qualified in several categories, but
such multiple selection was not used to bias the
further evaluation of any compound.
INORGANIC CHEMICALS
It is intended that the algorithm be suitable for
use with a wide variety of chemicals, including
inorganic chemicals. Inorganic chemicals,
however, present unique problems, both from
the method used to report inorganics in the TRI
and from the limitations of methods available for
estimating toxicity or exposure values.
First, several categories of inorganic
chemicals are reported in the TRI as
"compounds" (i.e., antimony, arsenic, barium,
cadmium, chromium, cobalt, copper,
manganese, nickel, and lead compounds) for
which the specific chemicals released were not
reported. The ranking and scoring model
depends, however, on specific toxicological
information for specific chemical compounds.
Therefore, an attempt was made to choose
surrogate compounds that represent the most
widely used forms of the inorganic chemical
categories for evaluation. If the TRI indicated
that the majority of the site releases were from a
specific industry or application, surrogate
compounds that are the major production form of
the chemical used in that industry (e.g., arsenic
pentoxide for the wood preserving industry) were
selected. For cadmium, chromium, nickel and
lead, however, no single surrogate was obvious.
In these cases, expert judgment was used to
select the inorganic salts produced in the greatest
quantity. Table 1 lists the inorganic chemicals
and chemical surrogates included in the model.
Second, some of the toxicity and exposure
data are calculated using QSARs based on 1-
octanol-water partitioning coefficients (Kow).
The inorganics were considered poorly fat-
soluble, which resulted hi a lack of reliable
methods for calculating missing data. Because
many of the ions involved have specific toxic
properties, they had to be individually evaluated.
Thus, a more extensive literature review was
performed to find published experimental data
for the inorganics. If data were still unavailable,
the missing data were estimated using an SAR.
If no SAR was available, the missing datum was
flagged and no hazard value was assigned to the
missing endpoint.
-------
TABLE 1: TRI INORGANIC CHEMICALS AND SURROGATE COMPOUNDS
TRI Inorganic Compound
Antimony Compounds
Arsenic Compounds
Barium Compounds
Cadmium Compounds
Chromium Compounds
Cobalt Compounds
Copper Compounds
Lead Compounds t
Manganese Compounds
Nickel Compounds
Zinc Compounds
'Surrogate Compound i
Diantimony trioxide (SbjO3)
Arsenic pentoxide (As^O^)
Barium chloride (BaCl2) !
Cadmium chloride (CdCy
Chromium oxide (CrO3)
Cobalt chloride (CoCl2) ;
Copper sulfate (CuSO4)
JLgad chloride (PbCl2) :
Manganese oxide
-------
CHAPTERS
DEVELOPMENT OF SCORING
CRITERIA
Scoring criteria have been divided into three
categories: human health effects,
environmental effects, and exposure
parameters. Table 2 presents the toxicological
endpoints included in the model to represent
human health and environmental effects. Table
3 lists the environmental effects and exposure
parameters used in the model. Each of these
scoring criteria are discussed below. Appendix
A presents a more detailed description of each
of these criteria, including data sources and a
description of how the data were scored.
human health effects data included quantitative
assessment of acute oral and inhalation toxicity,
Scoring criteria are divided into three
categories: human health effects,
environmental effects, and exposure
parameters.
HUMAN HEALTH EFFECTS
Human health effects are related to the
various toxic responses in humans caused by
exposure to a chemical. The screening tier
Human health effects data include acute
oral and inhalation toxicity, carcinogenicity,
and "other specific effects."
semiquantitative assessment of carcinogenicity,
and qualitative assessment of "other specific
effects" (i.e., mutagenic effects, developmental
effects, reproductive effects, neurotoxic effects,
and other chronic effects). For acute toxicity,
rodents were used as surrogate models. In the
confirmation tier, evaluation of potency and
specific organ/organ system effects could be
quantified.
Acute Effects
Acute human health effects can be
manifested by a wide range of adverse effects
through numerous routes of exposure. Two
toxicological endpoints are included in the
model to estimate the acute human health
-------
TABLE 2: TOXICOLOGICAL ENDPOINTS
Type of Effect
Human Health Effects
Environmental Effects
Terrestrial Animals
Fish
Type of Toxicity
Acute
Acute
Chronic
Chronic
Acute
Aeute
Chronic
Toxicological Endpoint
Inhalation LC50a
Oral LDsob
Carcinogenicity0
Other Specific Effects'1
Oral LlV
f ft. e
1^-50
NOELf
(a) The concentration of a substance in air that will kill half of a group of rodents when inhaled continuously for a
specific period of time. Data from tests of eight hours or less were used; these data were scaled on a linear basis
to be equivalent to a four hour test by: LCSO @4hr = (LC50 @t hrs)x(r hrs)/4 hrs.
(b) The concentration of a substance that will kill half of a group of rodents within 14 days when administered
orally as a single dose. (Dose is expressed as mass of chemical per mass of animal body weight.)
(c) Based on the EPA or International Agency for Research on Cancer (IARC) weight-of-evidence classification.
(d) Includes positive evidence of mutagenicity, developmental effects, reproductive effects, other chronic effects
and neurotoxicity.
(e) The concentration of a chemical in water that causes death hi 50 percent of the fish tested hi a 96-houi test.
(f) No observable effect level (NOEL): The highest dosage administered that does not produce toxic effects,
estimated from LC50 data.
TABLE 3: EXPOSURE PARAMETERS
Persistence
Biological oxygen demand (BOD) half-life3
Hydrolysis half-life0
Bioaccumulation
Aquatic bioconcentration factor (BCF)b
(a) The number of days required to biodegrade a chemical such that its BOD hi water is reduced by half, as
predicted using QSAR.
(b) The ratio of the concentration Of a chemical hi an aquatic organism to its concentration in water.
(c) The number of days required for the amount of a chemical to be reduced by half through hydrolysis hi water,
at pH 7» as predicted using QSAR.
-------
CHAPTER 3: DEVELOPMEOT OF SCORING CRITERIA
effects of chemical exposure: oral LD50 and
inhalation LC50. These endpoints are based on
the concentration of a substance that causes
death to 50 percent of the exposed population
either given a single-dose or continuously
exposed to a fixed concentration of a chemical
for a short duration. Test protocols for the oral
LD50 and inhalation LC50 have been fairly well
standardized and data for these endpoints are
often available. The model uses data from
laboratory studies of rodents as a surrogate for
acute toxicity to humans.
Chronic Effects
Chronic health effects in humans include
cancer, mutagenic effects, developmental
effects, reproductive effects, neurotoxic effects,
and other target organ effects. The
carcinogenicity of a chemical based on its
weight-of-evidence (WOE) classification was
included in the model. Data are generally less
available on the other specific chronic effects,
but these endpoints were evaluated qualitatively
and included in the model.
Carcinogenicily. Carcinogenic effects are
observed as tumors (i.e., neoplasms) induced in
an organism by exposure to a chemical, via a
genotoxic or epigenic mechanism. Several
schemes have been developed to classify
chemical carcinogens based on the WOE of
carcinogenicity. WOE classifications refer
only to the amount and adequacy of the
available evidence and not to the potency of the
carcinogenic effect or the mechanisms
involved. Potency refers to the dose required
to elicit a toxic effect, in the case of
carcinogens, tumors.
The model ranks the carcinogenicity of a
chemical using the WOE classification assigned
by EPA or IARC. These WOE classifications
were available for 48 of the chemicals. For the
remaining chemicals, SARs were used to assign
carcinogenic effects scores. Potency
ofcarcinogenic effects would be included in the
confirmation tier.
Other Specific Effects. Other specific
human health effects included in the model are
mutagenic effects, developmental effects,
reproductive effecte, neurotoxicity, and other
chronic effects. Positive data for mutagenicity
is often suggestive of carcinogenicity potential
via a genotoxic mechanism. In this assessment
scheme, however, it is considered a separate
health effect. Mutagenic effects are observed
alterations in the genetic material of the germ
and/or somatic cells induced by chemical
exposure. These alterations may take the form
of a gene-loci or point mutation or a clastogenic
event (i.e., rearrangements, gains or losses of
parts of or whole chromosomes). Since there
are few known human mutagens, the
correlation between the effects of a chemical on
the various mutagenicity test systems and its
potential mutagenicity in humans is very
difficult to assess. It has been assumed,
however, that experimental data even from in
vitro test systems is an indicator of mutagenic
risk to humans.
Developmental effects (e.g., teratogenic and
other embryotoxic effects) are observed as
damage to the embryo or fetus induced by
chemical exposure. Embryotoxic effects
include malformation, death and growth
retardation. Whole-mammal data is limited but
there is evidence for correlation between effects
in humans and other mammals. It has been
assumed that animal teratogenicity studies are
good indicators of human developmental risk.
In vitro teratogenicity tests appear to be valid
for direct acting teratogens. Teratogens
requiring metabolic activation, however, are
often missed as false negatives.
There are adverse effects of chemical
exposure on other aspects of reproduction.
These include but are not limited to effects on
fertility, gestation and lactation. It has been
assumed that adverse reproductive effects in
animal studies are good indicators of human
risk for like effects. Due to the nature of the
endpoints for reproductive effects, in vitro tests
have not been considered.
11
-------
Neurotoxic effects are adverse effects on the
nervous system induced by exposure to a
chemical. These include effects that are
structural and/or functional in nature as well as
behavioral alterations and learning disabilities.
The other chronic effects are the adverse
structural, physiological or biochemical effects
on various non-reproductive organ systems. Of
particular interest are effects on the immune
system. Immunotoxic effects are adverse
effects on the immune system which include
allergic sensitization. These endpoints in
general are poorly defined and test systems,
both in vivo and in vitro, are not well
documented or well accepted.
Data on the other specific effect endpoints,
when available, are difficult to interpret.
Furthermore, structure analysis reported for
some of the specific effects is limited to certain
chemical classes. Despite the limitations of the
data, each of these endpoints (e.g.,
mutagenicity, developmental effects,
reproductive effects, neurotoxic effects, and
other chronic effects) was evaluated
qualitatively (i.e., assigned a "yes" for positive
test results or a "no" for negative test results or
a lack of data) and combined into one endpoint
in the model. Expansion and quantification of
this criteria should be performed in the
confirmation tier of the human health effects
assessment.
ENVIRONMENTAL EFFECTS
Environmental effects are related to the
response of populations of organisms
representing different trophic levels and
different environments exposed to a chemical.
As with human health effects, environmental
effects data fall into a number of areas.
Environmental effects included in the model
were quantitative assessment of mammal and
fish mortality (fauna representing terrestrial and
aquatic environments, respectively), and the
aquatic subchronic endpoint of no observable
effect in fish. These are effects for which the
most experimental data are available and SARs
have been the most widely demonstrated. In
the confirmation tier, this could be expanded to
include responses for flora and species
representing different trophic levels.
Environmental effects include acute
mammal and fish mortality and chronic
sublethal effects in fish.
Terrestrial Effects
Terrestrial effects include toxic effects to
various components of the terrestrial
environment, including effects on mammals,
birds and higher plants. Rodent acute oral
toxicity (LD50) serves in the model as a
surrogate for terrestrial effects. This endpoint
was also used as a surrogate for human acute
oral toxicity. At the confirmation tier, bird
acute toxicity and higher plant phytotoxicity
could be added.
Aquatic Effects
Aquatic effects include toxicity to aquatic
organisms exposed to chemicals. Acute fish
mortality data (i.e., LC50) is one of the most
readily available endpoints and one that can be
estimated well by QSARs. The universality of
this endpoint makes it important in the
screening phase of an evaluation. Fish LC50
was selected as a scoring criteriori for the
screening tier model. Several trophic levels
and levels of biological complexity, including
microorganisms, algae and invertebrates, could
be included in the confirmation tier.
Chronic effects on fish are sublethal effects,
typically of longer term exposure and typically
measured as the "no observable effect level"
(NOEL). The NOEL is defined as the highest
dosage administered that does not produce toxic
effects. The most sensitive effects are observed
in reproduction and growth. The NOEL for
many fish has not been measured, but can be
12
-------
CHAPTER 3: DEVELOPMEP>JT OF SCORING CRITERIA
estimated from the acute toxicity data (i.e., the
lethal dose to 50 percent of the population) and
the octanol-water partition coefficient of the
chemical.
EXPOSURE PARAMETERS
Travis and co-workers have defined
exposure as the concentration of a chemical in
space and time at the interface with a target
population (Travis et al., 1983). A chemical
once released into the environment is subjected
to a variety of physical, chemical and biological
processes. These processes are relevant to the
amount and distribution of the chemical in the
different compartments of the environment and,
therefore, affect potential levels and routes of
exposure. Processes such as bioaccumulation
and persistence, including abiotic (i.e.,
photolysis and hydrolysis) and biotic
degradation (i.e., microbial transformation),
are important components of exposure
assessment since each has the potential of
affecting exposure levels. The amount of
chemical released to the environment, the
environmental medium of release (e.g., air,
water) and local environmental conditions will
also obviously affect the potential levels and
routes of exposure.
The screening tier exposure assessment
includes persistence and bioaccumulation
along with annual TRI releases as an overall
measure of potential exposure.
The screening tier exposure assessment data
includes quantification of persistence (i.e.,
biotic and abiotic degradation) and
bioaccumulation (i.e., aquatic BCF). These
exposure criteria are used in the algorithm with
the quantity of releases reported in the TRI as
an overall measure of potential exposure.
Physicochemical properties, such as the 1-
octanol-water partition coefficient (K^,), are
used indirectly in the model to estimate
exposure.
It should be noted that the purpose of this
ranking scheme is to address environmental
releases and subsequent exposures, thus the
emphasis on persistence and bioaccumulation.
If the model were applied to the workplace,
these factors would be less important to the
potential for occupational exposure.
A pivotal aspect of exposure assessment is
the use of fate and l:ransport models to quantify
the concentration of a chemical as it moves
from a source, through the environment, to the
target population. Several multi-media fate
models have been developed to predict the
distribution of a chemical in the environment.
The majority of non-site-specific fate and
transport models have been based on the
concept of fugacity (Mackay, 1979). Fugacity
models work by converting chemical
concentrations in the major environmental
compartments (i.e., air, water, soil, etc.) to
fugacity, a thermodynamic equilibrium
criterion which has units of pressure. This
method of calculation can be extended to a
variety of environmental media and has the
advantage of being easy to compile and
manipulate. Fugacity models have been
developed to reflect several levels of
complexity. In the confirmation tier, a
prediction of the environmental distribution of
the pollutant based on fugacity could be
included as a further refinement of potential
exposure. Site-specific exposure assessments
with more tailored chemical fate and transport
modeling could also be performed for specific
areas of interest, as are done for site-specific
risk assessments.
Persistence
Abiotic and biotic transformations/
degradations of a chemical affect its persistence
and concentration in the environment.
Transformation results in a modification of the
parent chemical and the subsequent formation
of an analogous or homologous derivative.
Degradation is the breakdown of the chemical
to water, carbon dioxide, ammonia and other
micromolecules. Chemical processes influence
13
-------
the amount of the chemical present in the
environment by regulating abiotic
transformations and degradations. Abiotic
alteration is primarily a result of the action of
light (photolysis), or the reaction of the
chemical with water (hydrolysis). Biological
processes also influence the concentration and
distribution of the chemical in the environment.
Biotransformation/biodegradation is a result
primarily of microbial action. All of these
processes act to reduce the persistence of a
chemical in the environment.
Parameters used to measure persistence in
the model are BOD half-life and hydrolysis
half-life. In general, BOD is the amount of
oxygen required by bacteria to reduce the
organic matter in water from a waste, typically
measured in a 5-day test. In this study, BOD
half-life was used as a measure of the number
of days required to reduce the BOD from a
chemical in water by half due to biodegradation
of the chemical. Hydrolysis half-life is the time
required to reduce the amount of a chemical in
water by half through hydrolysis reaction.
Both BOD half-life and hydrolysis half-life
were estimated using QSARs due to the wide
variability in experimental data. Photolysis
half-life is another important measure of
persistence but was not included in the model
due to a lack of data and of a reliable QSAR to
estimate missing data.
The confirmation tier could include an
evaluation of other measures of persistence that
might be included in the model. Another issue
is whether the BOD half-life and hydrolysis
half-life criteria (assuming the medium is
water) are preemptive and should be combined
into one score. If either of these criteria are
very short, it is likely that the other criterion
could be neglected.
Bioaccumulation
The term bioaccumulation is used to
describe the phenomenon by which a chemical
is taken up by an organism to a concentration
greater than in the surrounding environment.
When a chemical accumulates in an organism
to a high steady-state level, bioconcentration
has occurred. This is a result of the uptake rate
constant being larger than the elimination rate
constant. In contrast, biomagnification results
when oral uptake of a chemical leads to an
increase hi its concentration from one link to
the next in a food chain. BCF is the ratio .of
the concentration of a chemical in an organism
to its concentration in the test medium or
environment, typically water, at steady-state
conditions. This factor is a measure of the
chemical's ability to bioaccumulate and is
typically reported in log units.
Bioaccumulation is a function of the
physicochemical properties of a chemical,
especially the chemical's lipid solubility. The
Kow (described below) is commonly used as an
estimate of fat solubility. K^, is, in turn, used
to estimate BCF. The estimation of the BCF
from its relationship with Kow appears to be
accurate but varies in formulation depending on
the test system (Veith et al., 1983; Geyer et al.,
1991). Such estimation may be considered
reliable unless metabolic processes are
significant.
While bioaccumulation may occur hi both
aquatic and terrestrial organisms most of the
data relates to the former. Bioaccumulation hi
terrestrial species does not correlate well with
bioconcentration in aquatic species because it is
not as dependent on chemical lipid solubility.
Rather it depends more on the rate of
metabolism and other excretion mechanisms.
In aquatic food chains, biomagnification is
not a significant aspect of bioaccumulation
unless the Kow is greater than 1,000,000 (log
Kow greater than 6). The direct
bioconcentration of a chemical is often lower
than predicted from the Kow and
bioconcentration tends to decrease with
increasing Kow beyond a log Kow of 6 due to
increasing molecular size (Bintein et al., 1993).
The use of Kow alone as an estimation of
14
-------
CHAPTER 3: DEVELOPMENT OF SCORING CRITERIA
bioconcentration is limited to un-ionized
organic chemicals. Chemicals which dissociate
50 percent or more bioconcentrate significantly
less than predicted by Kow-based estimation
methods. When evaluating an ionized
chemical, consideration should be given to the
dissociation constant. These limitations were
not considered too significant for purposes of
the screening tier, and bioaccumulation, as it
pertains to aquatic ecosystems using the BCF,
was incorporated in the model.
Physicochemical Properties
The partitioning of a chemical in and
between environmental compartments (i.e., air,
soil and water) is governed by physicochemical
properties such as water solubility,
water/organic matter partitioning, vapor
pressure, acid dissociation, and soil/sediment
adsorption. In addition, physicochemical
properties are often used as surrogates for
human health and environmental effects.
Several of these properties, described below,
are used in the model.
Molecular Weight. Molecular weights can
be calculated directly from the molecular
formula. The major value in having this datum
is that it is used for conversion between
mass-based (mg/kg or mg/I) and molar-based
(moles/kg or moles/1) properties.
1-OctanoI/Water Partition Coefficient.
The 1-octanol/water partition coefficient, or
Kow, is defined as the ratio of a chemical's
concentration in the octanol phase to its
concentration in the aqueous phase of a two-
phase 1-octanol/water system at equilibrium. It
is typically reported in log units and is a pivotal
parameter in the investigation of environmental
fate, representing the distribution tendency of
organic chemicals between organic and aqueous
phases. It is the most often used single
parameter in toxicity QSARs.
Kow is related to lipophilicity (fat solubility),
water solubility, soil/sediment adsorption and
aquatic BCF. A chemical with a low Kow value
is considered hydrophilic and tends to have a
low fat solubility, high water solubility, small
soil/sediment adsorption coefficient and a small
BCF. The converse is also true. The
universality of Kow as a descriptor stems from
the fact that in reality it is a multicomposite
parameter representing a mixture of a wide
variety of molecular interactions (Dearden,
1990).
Kow values can be determined experimentally
by several methods including the standard
shaker flask method and the more novel slow
stir method. In addition, several good
estimation methodologies exist. The fragment
constant method (Rekker, 1977; Hansch and
Leo, 1979) has been used to calculate Kow from
substitute constants. Constants have been
tabulated for approximately a hundred
molecular fragments. The MedChem CLOGP
software estimates Kow values from an
algorithm developed from fragment constants
and structural factors. Other Kow estimation
schemes include those used in the studies of
Ghose and Crippen (1986).
Kow is a key input variable to the QSARs
used in this model to predict aquatic acute and
chronic toxicity, BOD half-life, and BCF, when
experimental data are not available. Thus,
experimental values of Kow, rather than
predicted values, were preferred. If
experimental values were not available, Kow
was predicted using the estimation scheme of
Ghose and Crippen (1986). For the pollutants
evaluated in this exercise, experimental Kow
values parallel the predicted ones suitably (see
Figure 1).
15
-------
a
cu
3
u
in
m
a
a.
01
a
CO
r~
o
ID
•H
in
01
O
O
ID
o
ID
in
ai
CD
ID
-o
4-J
a
x
03
2:
O
en
o
_J
01
i
ID
01
I
es
O)
B-
w
o
JBSb ' (M0>j)
16
-------
CHAPTER 3: DEVELOPMENT OF SCORING CRITERIA
Water Reactivity. Some functional groups,
such as acid chlorides, isocyanates, and
epoxides react with water in less than one day.
If a QSAR or SAR were used to estimate the
aquatic environmental effects (i.e., fish LC5q
and fish NOEL) of such compounds, the effects
were assumed to be those of the hydrolysis
products which were substituted into the algorithm.
17
-------
-------
CHAPTER 4
THE ALGORITHM
This chapter presents a description of the
algorithm. An exact description of the data
sources and scoring of each criteria is included
in Appendix A.
The method evaluates the potential hazard
of TRI releases to humans, terrestrial animals
and fish. A chemical hazard value is
calculated based on the toxicity of the
chemical, its persistence, and its potential
bioaccumulation in the environment.
OVERVIEW
The screening tier chemical ranking method
is illustrated conceptually in Figure 2. The
method evaluates the potential hazard of TRI
releases to humans, terrestrial animals and fish.
In the model, a hazard value is calculated for a
chemical based on the toxicity of the chemical
together with its persistence and potential
bioaccumulation in the environment. A
weighted hazard value is then calculated that
combines the hazard value, based on toxicity,
persistence and bioaccumulation, with the
weight of nation-wide releases reported in the
TRI. The persistence, bioaccumulation and
release data are used in the model as a measure
of the potential for exposure. The basic
algorithm is shown in the box below.
There were several assumptions made in the
development of me algorithm. Considering the
accuracy of many of the studies of toxicity, it
has been assumed that these data are generally
only accurate to within an order of magnitude.
This allowed some leeway to use QSAR or
SAR derived data when experimental data were
not available. In this screening-level analysis,
the objective in terms of this level of
uncertainty was to avoid false negatives. In a
more detailed analysis (a later tier), a more
thorough search for data could be performed
for a smaller number of chemicals.
Basic Algorithm:
Total Hazard Value = (Human Health Effects + Environmental Effects) x Exposure Potential
19
-------
CD
•5
to
o
DJ
X
UJ
CO
'€
CD
€
Du
CD
O
CD
CO
'co
CD
Q_
H-
c
,o
4-J
CO
-3
E
13
o
o
CO
o
, bo
&JD
I
fi
>
'o
*x
o
r~
CO
c
CD
|
O
>
L:
LLJ
•o
C
(O
c
CO
3
X
o 4.1 S2
•p o »-j
l"« 1 1 r *•••
6m J
i+_ CD
CD CO
t; o
o x-
CC uj
ingestion
inhalation
w_
CD
C
CO
o
1
CD
4-J
0
_CD
'•^
Z)
E
l I
T5 .^
CD ]0
> x
§
o
O
'c
O
i
o
r~
Affected
Populatioi
Humans
ingestion
i_
CD
4->
CO
_co
CD
13
O
Terrestrial
CD
CO
_co
CD
4->
O
<
o*
'c
o
O
Animals
x:
V)
u_
N
93
U
«+H
O
s
PN
O
O
c
p
U
20
-------
CHAPTER 4: THE ALGORITHM
Secondly, considering the range of toxicity
and physical properties of the pollutants of
interest, constructing the algorithm to give an
approximate logarithmic response has been
deemed appropriate. This allowed pollutants of
widely differing properties to be on the same
scale. The algorithm, however, does not
always adhere to a strict mathematical form for
this. For carcinogenicity, for example, the
algorithm takes the various WOE classifications
assigned by IARC or EPA and specifies a
numeric effect rating for each. The net result
was an approximate logarithmic (i.e., order of
magnitude) scale.
All of the additive terms within this
algorithm (i.e., oral LD50, inhalation LC50,
carcinogenicity, "other specific effects", fish
LC50 and fish NOEL) were given equal
weighting by assigning a hazard value to each
that could range on a scale from zero to five.
Cutoff values were chosen for the terms so that
the hazard value for very high or very low
toxicities would not exceed five or be less than
zero, respectively. (See Appendix A for
further details of assigning hazard values.)
Although the terms were weighted equally,
there were two aspects that created an implicit
weighting of the endpoints. First, as previously
noted, there were several different kinds of data
used as input to the algorithm: quantitative
toxicity levels (i.e., LC50 and LD50); semi-
quantitative levels (i.e., carcinogenicity WOE),
and qualitative assignments (i.e., yes/no
information on other types of chronic effects).
Therefore, assigning hazard values to the group
of qualitative chronic effects (e.g.,
neurotoxicity, mutagenicity) on the same zero
to five scale as quantitative toxicity levels for a
specific endpoint (e.g., acute inhalation
toxicity) gives greater weighting to the one
quantitative endpoint than to the group of
qualitative effects. Second, the choice of cutoff
levels for assigning maximum or minimum
hazard values adds a weighted judgment to the
algorithm.
Furthermore, the algorithm was to some
extent molded by the choice of the chemicals
used in the evaluation. The TRI is by its nature
a list of toxic chemicals. Of the 112 organics
drawn from the TRI, 40 of them, or 35
percent, showed some indication of
carcinogenicity. This is higher than the general
run of chemicals because selection of chemicals
for the TRI, being based on regulatory lists,
was affected by carcinogenicity data as well as
release amounts.
Each of the toxicological endpoints are
treated as additive effects. Persistence and
bioaccumulation are considered pivotal to the
potential for exposure and are included as
multiplicative factors rather than additive
effects.
HAZARD VALUE
The hazard value for each chemical is
derived from data on the seven toxicological
endpoints and the exposure parameters
described in Chapter 3. Each of the
toxicological endpoints are treated as additive
effects. Each additive endpoint receives a
hazard" value between zero (relatively nontoxic)
and five (extremely toxic) based on the results
of laboratory tests, carcinogen classification
schemes, or QSARs. Although there is implicit
weighting involved in assigning hazard values,
as discussed above, no additional scalar
weighting is currently applied to any
toxicological endpoint. The algorithm,
however, could be easily modified to do so.
Persistence and bioaccumulation are
considered pivotal to the potential for exposure
and are included as multiplicative factors rather
than additive effects in the algorithm. Each
multiplicative criterion is assigned a hazard
value between 1 (not persistent or does not
21
-------
bioaccumulate) and 2.5 (highly persistent or
high tendency to bioaccumulate) based on the
experimental data or QSARs.
should dominate the results. To determine
whether particular terms dominated the total
hazard value for the chemicals scored, a linear
The algorithm is:
Total Hazard Value — (Human Health Effects + Environmental Effects) x Exposure Potential
where:
Human Health Effects = HVoralLDSO + HVinkaIatimLCSO + HVcarcin + HVother (max. = 20)
Environmental Effects = HVoralLDSO + HVfishLCSO +.HVJiskNOEL (max. = 15)
Exposure Factor = HVBOD + HVhydrolysis + HVBCF (max. = 7.5)
and:
HVX = Hazard Value for endpointx
Human health effects have the potential of
being rated from 0 to 20 (e.g., ten points for
acute effects and ten points for chronic effects).
Environmental effects have the potential of
being rated from 0 to 15. Exposure parameters
can be rated from 1 to 7.5 (e.g., up to 2.5 for
BOD half-life, hydrolysis half-life and log
BCF). Using this scheme, the theoretical
maximum total hazard value would be 262.5
(i.e., (20 + 15) * 7.5). The actual maximum
for a chemical among those evaluated was
187.5. For 90 percent of the chemicals the
total hazard values were below 107. In the
algorithm, the final hazard values were
normalized to a scale from 0 to 100.
The program has been designed so that the
user can change the weightings of each
endpoint and determine the effect such a
weighting has on the chemical ranking. In this
manner, the algorithm may be utilized to obtain
a chemical ranking for different purposes. This
is discussed further in Chapter 5, where the
sensitivity of the algorithm to changes in the
endpoint weightings are examined.
CORRELATION OF SCORING CRITERIA
One objective in developing the algorithm
was that no one term scored in the algorithm
regression analysis was performed for total
hazard value (i.e., the sum of the individual
hazard values) versus subtotal hazard value by
area (human health, environmental and
multiplicative), as well as individual terms and
log Kow. The correlation coefficient (r) values
from this analysis are reported in Table 4. An
r that is neither extremely high nor extremely
low, ideally in the range of 0.4 to 0.6
(regardless of sign), would indicate an
appropriate level of importance for each term.
The results of this regression analysis show
that, overall, the model is operating as it should
in that no one term dominates the results.
The terms that make up the algorithm should
also be independent of each other. Correlations
between the some of terms were therefore
examined. Because some correlation between
carcinogenicity and the "other specific effects"
was expected, a linear regression was
performed on the hazard values for each term.
These results show r = 0.419 for the hazard
values for carcinogenicity versus those for
"other specific effects", which does not indicate
a strong correlation between the two terms.
WEIGHTED HAZARD VALUES
It was planned to use the chemical release
and transfer data reported in the TRI together
22
-------
CHAPTER 4: THE ALGORITHM
TABLE 4: SIMPLE CORRELATION COEFFICIENTS (r) FOR FINAL VALUE OF
ALGORITHM VERSUS PARAMETER
Parameter
log (inhalation LC50)
log (oral LD50)
carcinogenicity
other specific effects
log fish LC50
log fish no effect
BOD V4 life
hydrolysis l/i life
log BCF
all human health terms
all environmental terms
multiplicative exposure factors
log Kow
r
-0.50
-0.48
0.43
0.37
-0.65
-0.66
0.32
-0;06
0.53
0.70
0.76
0.42
0.37
Number of Chemicals
121
158
158
158
154
154
155
149
152
158
158
158
158
with chemical production and usage data to
calculate weighted hazard values. TRI data are
readily available and, although limited to data
for certain manufacturing sectors, are the best
resource for assessing the overall
environmental releases of the chemicals listed.
Production and usage data should also provide
some measure of the potential releases of a
chemical to the environment, particularly for
chemicals that are intentionally released, like
pesticides and herbicides. Unfortunately,
accurate and reliable chemical production and
usage data are not available for many of the
chemicals listed in the TRI. Thus, surrogates
for the total environmental release of a
chemical were limited to TRI data, except for
pesticides. Since pesticides are designed to be
released to the environment, usage data was
added to any TRI release data for
manufacturing of pesticides to estimate total
environmental releases of these chemicals.
Environmental releases of a chemical are
obviously pivotal to the potential for exposure,
and are thus included as multiplicative factors
in the algorithm. Some method of scaling
hazard values and release was needed,
however, to ensure that neither dominates the
algorithm. For example, chemical hazard
values calculated by the algorithm can
theoretically range from 0 to 262.5 before they
are scaled to 100. TRI releases, on the other
hand, ranged up to 546 million pounds for
ammonium sulfate solution in 1989. Simply
multiplying the hazard value by the release
would result in a. weighted hazard value
reflective of the magnitude of the release and
not of the toxicity or persistence of the
chemical. An option explored hi these studies
was to weight the hazard value by release data
using various schemes.
23
-------
Weighting schemes examined included:
(1) Multiply the final hazard value by the total
releases in pounds (i.e., total releases and
transfers reported in the 1989 TRI plus annual
usage for pesticides). This approach skews the
results almost totally toward the mass of
releases.
(2) Multiply the final hazard value by the
logarithm of the total releases.
(3) Multiply specific hazard values by the
releases to air, to water, or to the sum of air
and water.
(4) Multiply specific hazard values by the
natural log of the releases to air, water, pr the
sum of air and water.
The fourth option was selected (see Section
A.4 for details). Taking the natural log of the
releases provides weighted hazard values that
are not dominated by the weight of releases,
and does not understate the importance of the
release amount. To determine the release
amount assigned to air and water categories,
the following scheme was applied to the release
data. It was assumed that:
« stack and fugitive releases went to air;
» land, injection, water and POTW release
went to water;
• annual pesticide usage amounts were
assigned to the water release category;
• off-site transfers to an incineration facility
were assumed to be destroyed and transfers
to a recycling facility were assumed reused
and therefore not released to the
environment; and
• all other off-site transfers (land, injection,
etc.) were assumed released to water.
Incineration and recycling amounts were
subtracted from total off-site transfers to
determine the remainder of off-site transfers
released to water.
To determine the weighted hazard values:
• rodent oral LD50, fish LC50 and fish NOEL
were multiplied by the water releases;
• rodent inhalation LC50 was multiplied by the
air releases; and
• carcinogenicity and "other specific effects"
values were multiplied by the sum of air and
water releases.
Applying releases to the type of
toxicological endpoint which correlates to the
route of exposure adds a slight degree of
sophistication to the model that would not be
found if all endpoints were simply multiplied
by the total release and transfers. The
assumptions used to apply releases (e.g., land,
injection, water and POTW releases to water,
etc.) are simplistic, but they are appropriate
considering the level of analysis in this
screening tier. This component of the model
could be improved in the confirmation tier by
incorporation of some type of fate and transport
model, such as a fugacity model.
24
-------
CHAPTER 5
RESULTS AND DISCUSSION
This chapter presents a summary of the
results of the algorithm, an analysis of model
sensitivity, a discussion of uncertainties and
recommendations for future work. The full
results are presented in Appendix C and D. For
further information concerning the data used for
each chemical, contact the University of
Tennessee, Center for Clean Products and Clean
Technologies.
Different variations of the algorithm were run
to examine the effects of release weighting,
missing data and the "other specific effects"
score on the chemical ranking results. These
variations include:
• using or not using chemical release amounts
to weight hazard scores;
» using or not using the "other specific effects"
score;
• assigning a default hazard value of either zero
or five to chemicals with missing data for
acute inhalation toxicity, and acute and
chronic fish toxicity endpoints; and
• varying the endpoint weighting factors.
It should be noted that although the model
provides a numerical ranking of chemicals, the
ranking results do not represent any quantitative
measure of hazard or risk. In fact, given the
uncertainty and variability inherent to the data
used to score and rani: chemicals, the most
appropriate interpretation of the results would be
to consider groups of chemicals, i.e., the top 30
chemicals, the top 20 percent, etc., rather than
for directly comparing results of one chemical to
another.
The model was demonstrated on the selected
group ofl58 TRI chemicals and high-volume
pesticides.
DEMONSTRATION OF THE ALGORITHM
The model was demonstrated on the selected
group of 158 TRI chemicals and high-volume
pesticides. This demonstration shows that the
relative hazards of a large group of chemicals
can be scored and ranked on the same scale for
the purpose of priority setting. This
demonstration also highlights the need for a
25
-------
confirmation tier as well as the need tor expert
judgement in performing chemical ranking and
scoring.
The top 30 ranked chemicals (approximately
the top 20 percent) from release-weighted and
unweighted results are presented in Table 5.
These results are from the algorithm using a zero
hazard value tor missing data and including the
"other specific effects" score. These results will
be considered the baseline for comparison
purposes in the sensitivity and uncertainty
discussions in the next chapters.
Four general groups of chemicals appear hi
the top 20 percent: metals, pesticides, mineral
acids and ammonia, and other organic
compounds. The metals receive high ranking
generally because they are persistent, a number
are carcinogens and some exhibit high toxicity to
fish (e.g., copper). Manganese ranks high
despite its relatively low toxicity due to its
persistence and high release amounts. The high-
ranking pesticides generally are toxic via
inhalation and are toxic to fish. 2,4-D also is
persistent in the environment.
Mineral acids and ammonia receive high
ranking due to both high release amounts and
general toxicity. The high ranking of these
compounds highlights a problem in the screening
tier: they are not expected to be toxic within the
pH range found in ambient waters, but the model
does not account for any buffering reactions
following release to the environment. In fact,
many of the acid releases are to deep-well
injection where they would be unlikely to
contaminate surface water or directly impact
aquatic organisms. The other organic
compounds (e.g., formaldehyde, styrene) receive
high rankings due to various combinations of
toxicity, persistence and release amounts.
Table 5 also shows the effect of weighting by
chemical releases. Chemicals that rank high
in the algorithm when not weighted by releases
do so because of toxicity, bioaccumulation
and/or persistence, which are chemical-specific
properties. Chemicals that are high-ranking
when weighted by releases but not otherwise
(e.g., ammonia, sulfuric acid) are relatively less
toxic, but rank high because of high release
amounts.
SENSITIVITY ANALYSIS
Effect of Missing Data
The algorithm was developed to use a
database with a complete set of data for each •
endpoint. For those chemicals missing
experimental data, quantitative or qualitative
structure-activity relationships (QSARs or SARs)
were used to derive an estimate. There were
some endpoints, however, where no reliable
QSAR or SAR exist to estimate missing data.
These were left as missing data. The number of
measured, estimated, and missing data points is
presented in Table 6. As can be seen, missing
data were most significant for the inhalation LC^
and "other specific effects" endpoints.
Acute inhalation toxicity was especially
problematic; very little data exist for chemicals
with low vapor pressures that may nonetheless be
acutely toxic as a fume or aerosol. Instead of
estimating highly uncertain values, with little
ability to relate toxicity to chemical structure, it
was decided to assess the sensitivity of the
algorithm to the value assigned to this endpoint.
The algorithm was run both with default
hazard value scores of zero and five (the
minimum and maximum possible values) for
each missing data point for the acute inhalation,
acute fish and chronic fish toxicological
endpoints. The top 30 ranked chemicals from
these two variations in the algorithm are
presented in Table 7.
26
-------
CHAPTER 5: RESULTS AND DISCUSSION
TABLE 5: TOP 30 RANKED CHEMICALS FROM ALGORITHM
(default HV to zero for missing data)
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Weighted by Releases
chromium compounds3 (100)b
arsenic compounds3 (99)
lead compounds2 (95)
copper compounds3 (87)
terbufos (85)
2,4-D (85)
nickel compounds3 (84)
formaldehyde (84)
1,3-dichloropropene (78)
trifluralin (76)
cadmium compounds2 (75)
/ " ^^"S^^x^-' "^L v " ^' ' f ' *" "'"^^ '• " * <•**-"
arnniojiia ;^c^ ^ -^ ~ -^i^t^P^;, • 'Se^v^)-"
8tat^a ^-;*£esA(5S) "«
JlcMOJine "^'-ffewi; TKs^Kieti^iji; ' - v?^^i§6fe
manganese cdmptiuffl&2,-tv"''^$>iii' /»- (54)?^
chlorothalonil (54)
di(2-ethylhexyl)phthalate (53)
hexachlorobenzene (50)
naphthalene '"^'' '"•>->, ^* (48) /
phosphoric acid^.' * - -^ ' (48)/*
cobalt compounds3 (48)
- phenol " - •" * . - ' ' 7Z"" (4W^
Not Weighted by Releases
cadmium compounds3 (100)
arsenic compounds3
terbufos
r5S^^-,-v'I
trifiuralin
hexachlorobenzene
(82)
(81)
(63)
(62)
chromium compounds2 (61)
^^SpSpanS'lK^lfiS HF
formaldehyde
cobalt compounds2
lead compounds3
nickel compounds3
(60)
(59)
(59)
(59)
^teaceil'SSMtfil S itfsf '
SamKrtoieMSES- "2:Slfe^
hydrogen fluoride
(55)
di(2-ethylhexyl)phthalate (55)
chlorothalonil
2,4-D
1 ,3-dichloropropene
(53)
(53)
(52)
»d1nitr|d^SlX ¥^%'^sSi
.''ep'icMorQhydrin'i' , /riS','"^ "A*' '•- •. (52)12
^'f&A
'tftSS&y?: $,"i£9:'!f ^(sil
copper compounds3
hydrogen cyanide
styrene
^diblylphrnaktef :
(51)
(51)
(50)
'".« '-J$ "^ "(50) ;
f'2A$foaaHb£ br' • - '^f-f4W
(a) Table 1 lists the surrogate compounds used for the metal compounds in this evaluation.
(b) Number hi parentheses is the total hazard value for that chemical, normalized to a 0 - 100 scale,
(c) Shading indicates a chemical that is unique to one column in the table.
27
-------
TABLE 6: NUMBER OF MEASURED, ESTIMATED AND MISSING DATA POINTS
Endpoint
oral LD50
inhalation LC50
carcinogenicity
other specific effects
fish LCjo
fish NOEL
BOD half-life
hydrolysis half-life
BCF
Number of Measured
Data Points;
(% of total)
142 (90)
83 (53)
48 (30)
115 (73)
104 (66)
0
0
0
8 (4)c
Number of Estimated
Data Points;
(% of total)
16 (10): SAR
38 (24): SAR
110 (70): SAR
0
45 (28): QSAR
154 (97): QSAR"
133 (84): QSAR
139 (88): QSAR
142 (90): QSAR
Number of Missing
Data Points;
(% of total)
0
37 (23)
0
43 (27) a
4 (3)
4 (3)
2 (1)
1 (0.6)
5 (3)
(a) Source of data for "other specitic ettects" only includes posnive resi resuiis.
to negative results or lack of experimental data.
(b) Quantitative values based on the QSARs or experimental data for fish LC50.
(c) Measured data points used for inorganic chemicals only.
From Table 7, it can be seen that six
chemicals within the top 30 differ from the
algorithm variation with a default hazard value
of zero for missing data to the variation with a
default hazard value of five. The top 11 ranked
chemicals are the same for both variations, with
only small differences in relative rank,
indicating that the missing inhalation LC50 data
for the top-ranked chemicals (chromium, lead,
arsenic, copper and nickel compounds, and
2,4-D) make essentially no difference in the
results for these chemicals.
The missing fish LC50 and fish NOEL data
do impact the results for zinc (fume or dust)
and friable asbestos when a maximum hazard
value is assumed. This is also the case for
missing acute inhalation data for zinc and
barium compounds, phosphorus and maneb.
For these chemicals, the sensitivity analysis
indicates that the missing data points could be
important to the overall ranking results and
more effort in locating or estimating data for
these endpoints may be warranted. In order to
avoid possible false negatives, these six
chemicals should be considered for any
confirmation tier analysis.
Excluding "Other Specific Effects"
Also, the algorithm was run excluding the
"other specific effects" score to determine the
effect of this endpoint on the results. This is
the only endpoint where an attempt was not
28
-------
CHAPTERS: RESULTS AND DISCUSSION
TABLE 7: TOP 30 RANKED CHEMICALS FROM ALGORITHM, SENSITIVITY ANALYSIS
FOR MISSING DATA (weighted by releases)
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Default HV to 0 for Missing Data
chromium compounds3
arsenic compounds3
lead compounds3
copper compounds3
terbufos
2,4-D
nickel compounds3
formaldehyde
1 ,3-dichloropropene
trifluralin
cadmium
ammonia
sulfuric acid
hydrogen fluoride
nitric acid
hydrochloric acid
styrene
chlorpyrifos
hydrogen cyanide
tetrachloroethylene
trichloroethylene
chlorine
manganese compounds3
^gro^ja/^Sf ^^M
di(2-ethylhexyl)phthalate
5 •he^Sotole^M-SC^'^P
I BrSaSr"#f;'f?^;%::
'-SSS^vM-^^'^"'^
'4jg5j' M' : -tSSy-' " 'jS-frss^g^gi j ijsS&^v.,- •'& »»„
t- r%e^f ^ISS" Z-^i-Sfi&ff :"?-$mi-
Default HV to 5 for Missing Data (endpoint)
chromium compounds3 (inhal. LC,n)b
lead compounds3 (mhal. LCsn)
arsenic compounds3 (inhal. LC,n)
copper compounds3 (mhal. LGm)
nickel compounds3 (inhal. LC™)
2,4-D (inhal. LC,n)
terbufos
formaldehyde
1 ,3-dichloropropene
trifluralin
cadmium compounds3
ammonia
sulfuric acid
manganese compounds3 (inhal. LCsn)
hydrogen fluoride
di(2-ethylhexyl)phthalate (inhal. LCSO)
nitric acid
hydrochloric acid
'^.!&^%^a^^^^^^;JM{fMi^&^&&^t>^^
chlorpyrifos (mhal. LC,n)
styrene
hydrogen cyanide
tetrachloroethylene
.-.-plbosphoiT&Xy^llb'W'of vtte) .'„ (iflhal. lAn)"
'*"mu& ' - *"" » '^f ", ^^(mha^&if
trichloroethylene
chlorine
(a) Table 1 lists the surrogate compounds used for the metal compounds in this evaluation.
(b) Endpoint included in sensitivity analysis.
(c) Shading indicates a chemical that is unique to one column in the table.
29
-------
made to obtain data for every chemical.
Because only positive results were reported in
the data base used for this endpoint (Roadmaps
- described in Section A. 1.3), it has the effect
of penalizing chemicals that have been tested.
The top 30 ranked chemicals from the
algorithm both including and excluding "other
specific effects" are presented in Table 8.
From the table, it can be seen that only three
chemicals (nitric acid, manganese and
hexachlorobenzene) are ranked in the top 20
percent for the algorithm with the "other
specific effects" endpoint included that are not
in the top 20 percent with the endpoint
excluded. Alachlor, zinc compounds and
atrazine are ranked in the top 20 percent with
the endpoint excluded and not with the endpoint
included. Twenty-seven out of the 30 top-
ranked chemicals were the same in both cases,
although the actual ranking numbers may have
changed slightly.
Effect of Varying the Weighting of
Endpoints
As mentioned in Section 4.2, the weight
assigned to the endpoints in this algorithm can
be varied to assign greater or lesser importance
to certain endpoints. For selecting chemicals
for safe substitutes analysis, equal weighting
was assigned to each endpoint. To examine the
sensitivity of the algorithm to changes in the
endpoint weighting, the following additional
model runs were performed:
" the human carcinogenicity endpoint weight
was doubled;
• the human acute oral LD50 and inhalation
LCso weights were cut in half; and
" the weight assigned to environmental effects
endpoints (acute oral LD50, acute fish LC50,
and fish NOEL) were cut in half.
These results are presented in Table 9 and
Appendix D. Table 9 shows the top 30 ranked
chemicals, hot weighted by release amounts,
for even weighting and for each variation of the
algorithm listed above. The biggest difference
resulted from doubling the carcinogen endpoint
weight; there are five different chemicals in the
top 30 as compared to the evenly weighted
endpoint results. The other variations have
only two or three different chemicals ranked in
the top 30. These results indicate that the
algorithm is not very sensitive to endpoint
weights when changed by a factor of two.
Greater changes to the endpoint weights may be
appropriate in some cases, depending on the
particular purpose for which the algorithm
might be used.
UNCERTAINTIES
Uncertainties in the algorithm primarily
result from uncertainties in the.data base.
Because one goal of the screening tier is to
avoid false negatives, it is recommended that,
in general, chemicals with missing data for any
endpoint be considered for the confirmation
tier. Some exceptions to this might include:
• if the chemical does not rank near the top of
the list even with a default hazard value of
five assigned to the missing data for that
chemical;
• if there are no reported air releases for
chemicals with missing data for acute
inhalation toxicity; and
• if the physical/chemical properties of the
chemical indicate that it would not pose a
hazard in the environment.
Table 10 summarizes the'chemicals with
missing data.
30
-------
CHAPTERS; RESULTS AND DISCUSSION
TABLE 8: TOP 30 RANKED CHEMICALS FROM A
FOR "OTHER SPECIFIC EFFECTS" (we
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
,22
23
24
25
26
27
28
29
30
Default HV to 0 for Missing Data,
"Other Specific Effects" Included
chromium compounds"
arsenic compounds3
lead compounds*
copper compounds3
terbufos -
2,4-D
nickel compounds3
formaldehyde
1 ,3-dichloropropene
trifluralin
cadmium compounds3
ammonia
sulfuric acid
hydrogen fluoride
il&l^aclfi^CSi^^'1^:fe]^'!^^^^/^
hydrochloric acid
.styrene
chlorpyrifbs
hydrogen cyanide
tetrachloroethylene
trichloroethylene
chlorine
^yfi^^^^Bfi^^*a:"'''/r^
chlorothalonil
di(2-ethylhexyl)phthalate
naphthalene
phosphoric acid
cobalt compounds3
phenol
I.GORITHM, SENSITIVITY ANALYSIS
ghted by releases)
Default HV to 0 for Missing Data,
"Other Specific Effects" Excluded
chromium compounds3
terbufos
copper compounds*
1 ,3-dichloropropene
lead compounds3
nickel compounds3
formaldehyde
2,4-D
sulfuric acid
cadmium compounds3
nitric acid
trifiuralin
chlorpyrifos
hydrochloric acid
hydrogen cyanide
chlorine
hydrogen fluoride
styrene
phosphoric acid
chlorothalonil
cobalt compounds*
naphthalene
phenol
di-2(etfaylhexyl)phthalate
tetrachloroethvlene -
(a) Table 1 lists the surrogate compounds used for the metal compounds in this evaluation.
(b) Shading indicates a chemical that is unique to one column hi the table.
31
-------
I
ja
0)
'nr
1
I
6JJ
t
^
*J
C3
0
cu
i
1
M
•a
i
U
•e
r^
c
c
ov
d
3
03
-42^
wfes '§1
t|i
1 1
I
^i
u 'I
f S,
«2 g
•11
Half weigl
buman be
t* *•*
c2 «9
60 §,
Double weighl
carcmogen ei
1
•n
§
'a
t<
cS
I
i,
Even wei
•8
1
£
•s
1 Cadmium compc
R
t=H
N^>
•a
i
Cadmium comp
8
Cadmium compounds
g
*c/>
1
1
Cadmium co
£
•a
1 Arsenic compou
en
oo
i
Arsenic compoi
So
****
•c
c
1
1
c
1
00
N~x
c/
1
8
e.
c
1
CO
V
tN
P
0>
_8
1
Hexachloro-1, 3-
£
,23
S
S
Polychlorinated
ti
CO
Polychlorhlated biphen
oc
Terbufos
en
?
Terbufos
g-
o
•3
1
1 Hexachloro-1, 3
t-
(U
Hexachloro-1 ,3-butadi<
oo
VO,
CO
f
.8*
Polychlorinated
VO
E3
j
x-N
VO
Chromium compounds
fo
Trifluralin
VO
$
Formaldehyde
ft'
1
I Hexachloroben.
/->
VO
CO
T:
g
c
Jo
enzene
Hexachlorob
^
en
Diammotoluene
en
VO
3
1
I
trT
VO
Hexachlorobenzene
58
S
o
K
t
0
1
C-
t— 4
OO
^
1 2-Nitropropane
en
•a
1
| Chromium com
en
VO
Formaldehyde
vn
E
Chromium c
a\
vri
•3
o
1
g
c^
c.
cd
73
1
I Di(2-ethylhexyl
r?
VO
2-Nitropropane
vS
t
c
c
2
es
o
T-H
1
Chromium coin]
VO
g
^>
i
cs
R
VO
Lead compounds
3
n
Diammotoluene
ft
O
C
S
c
c
*c
•§
u
tN
T— 1
VO
Trifluralin
vn
•8
1
u
1
p
•n.
u
cd
Di(2-ethylhexyl)phthal
o\
>n
o
1
C
C
u
•T:
c
11
en
00
CO
Lead compound
S
1 Trifluralin
(FT
•n
3
o\
m
0
C
E
£
x
*H
c
Tf
OO
n
u
t
<
c
c
1
e
<
•n
VI
f
S
oo
1 Anthracene
en
u
1,2,4-Trichlorobenzen
£;
i
CO
Diaminotolu
VO
VO
i
1
| 1,2,4-Trichloro
t—
0
o
1
! Decabromodip]
fn
(n.
Anthracene
in
>n
a
c
c
(
t
t-
s?
Q
4
w
x
1
en
>n
| 2,4-Dmitrotoluene
10
n
5
fj?
*ijj.
"1
".'^
1
^^
en
>n
"c
.£
C
t
PN
v?
0
1
il
t
|
•*~<
Q
in
CO
o
f?
C-
I Epichlorohydrin
V
Q
c
o
8
1 Anthracene
£
I Diamhiotoluen
R
iii-
1 ,3-Dichloropropene
>n
co
1
j
T-H
r-H
tN
32
-------
CHAPTER 5: RESULTS AND DISCUSSION
I
5
I
I
CO
oa
•I
CO
cs
iM
a
I
•a
o\
I
•8
o\
t
'a-
I
Q
S-
M
•§! a
I
o<
«
1
f
§•
See
33
-------
TABLE 10: CHEMICALS WITH MISSING DATA
Endpoint
Chemicals with Missing
Data
Comments
oral LD«,
No missing data
inhalation LC.
•so
Alachlor
Ammonium nitrate (sol'n)
Ammonium sulfate (sol'n)
Antimony compounds'1
Arsenic compounds6
Barium compounds1"
Butylate
Butyl benzyl phthalate
Catechol
Chlorpyrifos
Chromium compounds'1
Cobalt compounds1"
Copper compounds'1
2,4-D
2,4-dinitrophenol
2,4-dinitrotoluene
Decabromodiphenyloxide
Bis (2-ethylhexyl) adipate
Di-(2-ethylhexyl) phthalate
Glyphosate
Hydroquinone
Lead compoundsb
Maneb
Manganese compounds'1
Metolachlor
Metribuzhi
Molybdenum trioxide
Nickel compounds'"
Nitrobenzene.
N-nitrosodiphenylamine
Di-n-octylphthalate
Phosphorus
Picric acid
Polychlorinated biphenyls
Terephthalic acid
Thorium dioxide
Zinc compounds'1
no reported air releases
low volatility3
low volatility3
low volatility2
low volatility2
low volatility3
no reported air releases
no reported air releases
low volatility3
low volatility3
low volatility3
high rank for HV=5C
no reported air releases
low volatility3
low volatility3
low volatility3;
high rank for HV=5°
no reported air releases
no reported air releases
low volatility3
low volatility3
no reported air releases
no reported air releases
low volatility3
low volatility3
carcinogenicity
No missing data - all
chemicals for which
EPA/IARC classification not
available were examined by
SAR
34
-------
CHAPTER 5: RESULTS AND DISCUSSION
Table 10, continued
Endpoint
other specific effects
fish LC50
fish NOEL
BOD half-life
hydrolysis half-life
BCF
Chemicals with Missing
Data
Aluminum (fume or dust)
Asbestos (friable)
Thorium dioxide
Zinc (fume or dust)
Aluminum (fume or dust)
Asbestos (friable)
Thorium dioxide
Zinc (fume or dust)
Maneb
Phosphorus
Maneb
Maneb
Aluminum (fume or dust)
Titanium tetrachloride
Thorium dioxide
Molybdenum trioxide
Comments
There are 43 chemicals
which received a zero for
oilier specific effects. The
data base used, Roadmaps,
lists references for positive
results only. Some of these
chemicals may have been
tested, with negative
results.
hi gh rank when H V = 5°
high rank when HV=5C
high rank when HV=5C
high rank when HV=5C
no Kj1
noE^o
noK™"
(a) No experimental data expected due to low volatility of chemical.
(b) Table 1 lists the surrogate compounds used for the metal compounds in this evaluation.
(c) When given a hazard value of five for missing data, this chemical ranks in the top 20 percent.
(d) QSAR could not be run for this chemical due to lack of KOW data.
SELECTION OF PRIORITY CHEMICALS
The chemical ranking and scoring method
was developed as a priority-setting tool, to
select priority chemicals for evaluating the
potential for safe substitutes (Davis, et al.,
1994a). Of the top 30 ranked (release-
weighted) chemicals identified in Table 5,
seven had been selected previously as priority
chemicals because they were either included in
the EPA 33/50 Program (for voluntary
reductions in TRI releases) or manufactured
from a 33/50 chemical. These chemicals
include chromium compounds, lead
compounds, nickel compounds, cadmium
compounds, hydrogen cyanide,
tetrachloroethylene, tirichloroethylene and
styrene (manufactured from benzene).
Excluding these chemicals from the selection
process, the top eight priority chemicals are:
• arsenic compounds
• copper compounds
• terbufos
35
-------
» 2,4-D
» chlorine
» manganese compounds
• di(2-ethylhexyl)phthalate
Reasons for not selecting some of the other top-
ranked chemicals are discussed below.
Terbufos, an insecticide/nematicide, and
2,4-D, an herbicide, were selected to represent
the range of top-ranked pesticides. Although
some of the other pesticides on the list may be
used in different applications (e.g., different
crops or pests), many of the safe substitute
approaches identified for terbufos and 2,4-D
will likely be applicable.
Ammonia and the mineral acids were not
selected because the types of releases of these.
chemicals reported in the TRI probably do not
result in exposures to humans or fish to acutely
toxic concentrations. It is likely that these '
chemicals would have a lower ranking in a .
more sophisticated model that considers the fate
of chemicals in the environment. This is not to
downplay the possible significance of the
effects of large releases or accidental spills of
these chemicals. Their high ranking
underscores the fact that their toxicity and high
release amounts are areas of potential concern.
Copper compounds and manganese
compounds were selected for further
evaluation. These chemicals, however, point to
one of the limitations of the TRI in that it
groups metal compounds. No data is provided
on the speciation of compounds released to
different environmental media, nor on the
particular compounds released in the largest
amount. The toxicity of metal compounds and
their availability to different types of organisms
is, of course, highly dependent on the
speciation of the compounds. The copper and
manganese compound surrogates (copper
sulfate, manganese oxide) used in the algorithm
are highly to moderately toxic to fish, but other
copper or manganese compounds may have
lower overall toxicity and would thus rank
lower in the model.
The model uses a risk assessment approach
to place chemical release data into perspective
by combining release amounts with the
potential for environmental persistence and
bioaccumulation, and potential human health
and ecotoxic effects from these releases.
CONCLUSIONS AND
RECOMMENDATIONS
The UT chemical ranking and scoring model
was found to be a useful tool for screening
purposes, and for putting the TRI data in a
more useful framework than simply pounds of
releases. The model uses a risk assessment
approach to place chemical release data into
perspective by combining release amounts with
the potential for environmental persistence and
bioaccumulation, and potential human health
and ecotoxic effects from these releases. The
model was developed as a tool to select priority
chemicals for safer substitutes assessment, and
with the concurrent use of expert judgement, it
serves as an improvement over previous
methods of prioritization. It is also flexible
enough for other applications. The UT model
has been used to make a preliminary
assessment of the comparative potential hazards
posed by the reported releases of toxic
chemicals in Tennessee, Texas, Louisiana,
Indiana and Ohio in 1990, the five states with
the greatest releases in that year's TRI (Kincaid
and Bartmess, 1993). It is also being modified
for use by a major chemical company in
prioritizing reduction efforts for TRI releases.
Recommendations for future work include:
• addressing in greater depth the issues
resulting from missing data, or considering
alternate sources of data;
36
-------
CHAPTER 5: RESULTS AND DISCUSSION
further developing the chronic human health
effects scoring, for example, by using
cancer slope factors and chronic reference
doses or other measures of potency rather
than semi-qualitative WOE data or
qualitative "type of effects" data;
using the algorithm on a site- or facility-
specific basis;
incorporating chemical fate and transport
modeling into the algorithm, i.e.,
considering the short-term and long-term
distribution of chemicals in environmental
media, which might include photolysis or
other degradation reactions, acid/base
buffering reactions,, and metals complexation
in the environment; and
• developing the second, or confirmation, tier.
Overall, this screening tier should be
considered a first step. It should be
remembered that this model is a screening tool
and was not designed to be removed from
expert judgement. In some aspects the model
was found to be lacking in sensitivity in that it
does not adequately represent chemical
behavior in the environment, but it was found
to put environmental release data in a more
useful framework for priority setting.
37
-------
-------
REFERENCES
REFERENCES
Aspelin, A.L., A.H. Grube and R. Torla (1992) Pesticide Industry Sales andUsage: 1990 and 1991
Market Estimates. U.S. Environmental Protection Agency, Office of Peisticides and Toxic
Substances.
Bintein, S., J. Devillers and W. Karcher (1993) Non-linear dependence of fish bioconcentration on n-
octanol/water partition coefficient. SAR and QSAR in Environmental Research, 1: 29-39.
Davis, Gary, L. Kincaid, D. Menke, B. Griffith, S. Jones, and C. Brown (1994a) The Product Side of
Hazardous Waste Reduction: Evaluating the Potential for Safe Substitutes, Center for Clean
Products and Clean Technologies, University of Tennessee, Knoxville, Tennessee.
Davis, Gary, M. Swanson, and S. Jones (1994b) Comprehensive, Comparative Evaluation of Chemical
Ranking and Scoring Methodologies, Center for Clean Products and Clean Technologies,
University of Tennessee, Knoxville, Tennessee.
Dearden, J.C. (1990) Physico-Chemical Descriptors, In: W. Karcher and J. Devillers, eds., Practical
Applications of Quantitative Structured-Activity Relationships (QSARs) in Environmental Chemistry
and Toxicology. Kluwer Academic Publishers, Dorrecht, The Netherlands, pp 25-59'.
Geyer, H.J., I. Scheunert, R. Bruggemann, C. Steinberg, F. KortandA. Kettrup (1991) QSAR for
organic chemical bioconcentration wDaphnia, algae, and mussels. Sci. Total Environ. 109:
387-394.
Ghose, A.K., and G.M. Crippen (1987) Atomic Physicochemical Parameters for Three-Dimensional-
Structure-Directed Quantitative Structure-Activity Relationships; 2. Modeling Dispersive and
Hydrophobic Interactions. /. Chem. Inf. Comput.Sci.,27(l):21-35.
Hansch, C., and A.J. Leo (1979) Substituent constants for correlation analysis in chemistry and biology.
Wiley and Sons, New York.
Hazardous Substances Data Bank (HSDB) (1992) The National Library of Medicine's Toxicology Data
Network (TOXNET) System.
Kincaid, L.E., and J.E. Bartmess (1993) Evaluation of TRI Releases in Indiana, Louisiana, Ohio,
Tennessee and Texas, Center for Clean Products and Clean Technologies, University of
Tennessee, Knoxville, Tennessee.
Mackay, D. (1979) Finding fugacity feasible. Environ. Sci. Technol. 13: 1218-1223.
Rekker, R.F. (1977) The hydrophobicfragment constant. Elsevier, New York, NY.'
Travis, C.C., C.F. Baes III, L.W. Barnthouse, E.L. Etnier, G.A. Holton, B.D. Murphy, G.P.
Thompson, G.W. Suter II, and A.P. Watson (1983) Exposure assessment methodology and
reference environments for synfuel risk analysis. ORNL/TM-8672, Oak. Ridge National
Laboratory, Oak Ridge, TN.
39
-------
Veith, G.D., D.J. Call, and L.T. Brooke (1983) Structure-toxicity relationships for the fathead
mimo\v,'pimephalespromelas: narcotic industrial chemicals. Can. J. Fish. Aquat. Sci. 40: 743-748.
40
-------
APPENDIX A
DATA SELECTION AND DETERMINATION OF HAZARD VALUES
-------
-------
__ _ APPENDIX A
A.1 HUMAN HEALTH EFFECTS
A. 1.1 Acute Effects
Definitions/Test Methods
Oral LD50: The concentration of a substance, expressed in mass of the substance per mass of the
animal, that will kill half of a group of rodents within 14 days when administered orally as a single
dose.
Inhalation LCSO: The concentration of a substance in air (gas or dust) that will kill half of a group of
rodents when inhaled continuously for 8 hours or less, scaled to 4 hours.
Data Selection
Figure A-l shows the hierarchy for Oral LD50 data selection. Figure A-2 presents the
hierarchy for Inhalation LC50 data selection. Experimental data are preferred for both oral and
inhalation data. The hierarchy for experimental data sources was 1) Hazardous Substances Data Bank
(HSDB, 1993), and 2) Registry of Toxic Effects of Chemical Substances (RTECS, 1992, 1993), both
on-line data sources. Additional data sources such as ambient water quality criteria documents were
used for the inorganic chemicals (Davidson, et al., 1987; Hose, et al., 1989; IPCS, 1990, IPCS, 1991;
EPA, 1980a; EPA, 1980b, EPA, 1980c; EPA, 1980d; EPA, 1980e, EPA, 1984a). Additional data
sources were also used for pesticides (Kidd and James, 1991; EPA, 1984b).
If experimental oral LD50 data are available for more than one species of rodent, the most
sensitive test result is selected. If experimental inhalation data are available for more than one test
duration, the datum is selected from the test with duration closest to 4 hours but not exceeding 8 hours.
Since the test durations for the inhalation toxicity tests differ, a linear scaling function was
incorporated into the algorithm. The EPA requires a minimum test duration of 4 hours (40 C.F.R.
798. 1 150). Other test durations were scaled to the 4 hour test by the following equation:
concentration, * time, = concentration *
If experimental data are not available, but can be estimated by way of a structurally similar
compound in the same physical state, an SAR is used to estimate the oral or inhalation value. If an
SAR is not available, the missing data is flagged in the database and the hazard value for the missing
endpoint is set to zero,
Calculation of Hazard Values
Hazard value scores for the inhalation and oral acute toxicological endpoints are based on the
Iog10 of the LC50 and LD50. Figure A-3 is a decision tree used to calculate the Oral LD^ hazard value.
A-l
-------
Experimental
Data
Flag data missing
setHV = 0
Yes
Yes
select most
sensitive rodent
test results
estimate
LD50
Figure A-l. Decision Tree for Oral LD50 Data Selection
A-2
-------
APPENDIX A
Experimental
Data
No
Flag data missing,
set HV = 0
select test with
duration closest to
4 hrs, and not
exceeding 8 hrs.
estimate
LC
50
Figure A-2. Decision Tree for Inhalation LC™ Data Selection
A-3
-------
LogLD5o>3./
(5,000 mg/kg)
Log LD50^0.7
(5mg/kg)
Yes
HV =
Yes
HV = 5
HV = (6.2- 1.7* log LD50)
Figure A-3. Decision Tree for Oral LD50 Hazard Value
A-4
-------
APPENDIX A
The log,0 of the LD50 is taken and assigned a score between 0 and 5, using a continuous, linear
function. Numerical cutoff values are based on commonly accepted values (Michigan CMR, 1987).
Figure A-4 is a decision tree of the method used to calculate the inhalation LC5() hazard value.
The log,0 of the LC50 is taken and assigned a score between 0 and 5, using a continuous, linear
function. Numerical cutoff values are based on commonly accepted values (Konemann and Visser
1988; O'Bryan and Ross, 1988; Weiss, et al., 1988).
A. 1.2 Carcinogenicity
Definitions/Classification Methods
IARC Classification: The IARC publishes a series of monographs entitled "IARC Monographs on the
Evaluation of the Carcinogenic Risk of Chemicals to Humans," which provide WOE classifications of
chemical carcinogenicity. Chemicals are classified by a working group of experts in chemical
carcinogenesis and related fields based on published information available at the time the working
group was convened. Table A-l presents the IARC classification scheme.
TABLE A-l: IARC CARCINOGEN CLASSIFICATION SYSTEM
Group
Definition
The agent is probably not carcinogenic to humans. - This classification is used when there is
evidence suggesting lack of carcinogenicity in both humans and experimental animals. (In some
circumstances, agents for which there is inadequate evidence of or no data on carcinogenicity in
humans but evidence suggesting lack of carcinogenicity in experimental animals, consistently
and strongly supported by a broad range of other relevant data, may be classified in this group.)
The agent is not classifiable as to its carcinogenicity to humans. - This classification is used
when agents cannot be placed hi any other group.
2B
The agent is possibly carcinogenic to humans. - This classification is generally used when there
is limited evidence in humans in the absence of sufficient evidence in experimental animals. (It
may also be used when there is inadequate evidence of carcinogenicity in humans or when
human data are nonexistent but there is sufficient evidence of carcinogenicity hi experimental
animals. In some instances, an agent for which there is inadequate evidence or no data in
humans but limited evidence of carcinogenicity in experimental animals together with supporting
evidence from other relevant data may be placed in this group.)
2A
The agent is probably carcinogenic to humans. - This classification is used when there is limited
evidence of carcinogenicity in humans and sufficient evidence in animals. (Exceptionally, an
agent may be classified into this category solely on the basis of limited evidence of
carcinogenicity in humans or of sufficient evidence of carcinogenicity in experimental animals
strengthened by supporting evidence from other relevant data.)
The agent is carcinogenic to humans. - This classification is used only when there is sufficient
evidence of carcinogenicity in humans.
source: McGregor, 1992
A-5
-------
Yes
Yes
HV = (8.0-2.0* log LQ0 )
HV=0
HV = 5
Figure A-4. Decision Tree for Inhalation LC50 Hazard Value
A-6
-------
APPENDIX A
EPA Classification: The Carcinogen Assessment Group in EPA's Office of Health and Environmental
Assessment prepares WOE classifications for carcinogens. EPA's classification scheme is based
largely on an earlier version of the IARC classification scheme which did not include Group 4 and the
criteria for Group 2A and 2B. Table A-2 presents the 1986 EPA carcinogen classification scheme.
EPA in 1988 began an effort to revise its carcinogen classification scheme (EPA, 1988).
TABLE A-2: 1986 EPA CARCINOGEN CLASSIFICATION SYSTEM
Group
Definition
E
Evidence of Non-Carcinogenicity for Humans. This classification is used when agents
show no evidence of carcinogenicity in at least two adequate animal tests in different
species or in both adequate epidemiologic and animal studies.
D
Not Classifiable as to Human Carcinogenicity. This classification is generally used
when there is inadequate human and animal evidence of carcinogenicity or when no data
are available.
C
Possible Human Carcinogen. This classification is used when there is limited evidence
of carcinogenicity in animals in the absence of human data.
B
Probable Human Carcinogen. This group is divided into two subgroups, Bl and B2.
Subgroup Bl is usually used when there is limited WOE of human carcinogenicity based
on epidemiologic studies. Group B2 is used when there is sufficient WOE of
carcinogenicity based on animal studies, but inadequate evidence or no data from
epidemiologic studies.
Human Carcinogen. This classification is used only when there is sufficient evidence
from epidemiologic studies to support a causal association between exposure to the agent
and cancer.
Data Selection and Calculation of Hazard Value
Figure A-5 is a decision tree of the hierarchy of carcinogenicity data selection. If EPA and
IARC have both classified the carcinogenicity of a chemical, both classifications are used to calculate
the chemical hazard value. When both IARC and EPA classifications are available for a chemical, the
hazard value is assigned for each and the average is taken for the overall hazard value unless the IARC
classification is a 3. In this case, only the EPA classification will be used to determine the hazard
value. Otherwise, the hazard value is based on the classification available. If neither IARC nor EPA
have classified the chemical as a carcinogen, the carcinogenicity of the chemical is evaluated using an
SAR in MICROQSAR Version 2.0 which is based on the unpublished work of Arcos. The SAR
assigns a positive carcinogenicity rating to a chemical if it contains one or more molecular substructure
that has been related to carcinogenicity, such as a polyaromatic hydrocarbon.
A-7
-------
IARC Rating
other than
3
HV = EPA + IARC
Evaluate Using SAR
Figure A-5. Decision Tree for Carcinogenicity Hazard Value
A-8
-------
APPENDIX A
Before assigning hazard values to the IARC and EPA carcinogen groups, 31 chemicals for
which WOE carcinogenicity ratings had been assigned by both EPA or IARC were reviewed to
evaluate the correlation between EPA and IARC ratings. Table A-3 presents a comparison of the EPA
and IARC ratings for the 31 chemicals. Based on the definitions provided in the EPA and IARC
classification systems, it appears that IARC Group 2A corresponds to EPA Group B2 and IARC Group
2B corresponds to EPA Group Bl. Based on actual ratings of chemicals, however, the strongest
correlation of IARC Group 2B is seen with EPA Group B2.
TABLE A-3: COMPARISON OF EPA AND IARC RATING OF 31 CARCINOGENS
EPA Rating
A
Bl
B2
B2
B2
C
IARC Rating
1
2A
2A
2B
3
3
Number of 'Chemicals
6
4 ' .
2
12
2
5
Thus, EPA Group B2 and IARC Group 2B were assigned equivalent hazard values in the ranking
method.
Table A-4 presents the hazard values assigned to IARC and EPA carcinogenicity ratings.
Chemicals assumed to be carcinogens based on SARs were assigned a hazard value of 3.0 or 1.0, based
on the molecular substructures present.
TABLE A-4: CARCINOGENICITY HAZARD VALUES
IARC Classification
Group
4
3
N/A
2B
2A
1
Hazard Value
0
oa
N/A
3.5
4.0
5.0
EPA Classification
Group
E
D
C
B2
Bl
A
Hazard Value
0
0
1.5
3.5
4.0
5.0
(a) The EPA classification alone is used in this case.
N/A = not applicable
A-9
-------
/V.I.3 Other Specific Effects
Definitions/Test Methods (from Roadmaps):
Mtttagenicity: Chemicals are indicated as possible mutagens in humans if positive results inbioassays
are reported in the reference source (ICF, 1989).
Developmental Toxicity: Chemicals are indicated as exhibiting developmental toxicity if data in the
reference source support concern that the chemical may cause embryotoxicity, fetotoxicity or
teratogenicity in humans (ICF, 1989).
Reproductive Toxicity: Chemicals are indicated as exhibiting reproductive effects if data in the
reference source support concern that the chemical has adverse effects on male or female reproductive
performance (ICF, 1989).
Chronic Toxicity: Chemicals are indicated as exhibiting chronic toxicity if adverse effects other than
cancer occur at doses less than or equal to 1 g/kg/day following inhalation, oral or dermal exposure for
more than 90 days (ICF, 1989).
Neurotoxicity: Chemicals are indicated as neurotoxic if chronic (at least 90 days) ir^alatiqn, oral or
dermal exposure to doses less than or equal to Ig/kg/day results in neurotoxic effects (ICF, 1989).
Data Selection and Calculation of Hazard Value
Data for the "other specific effects" endpoints were obtained from Roadmaps, a database
developed by EPA of sources of information on the SARA 313 chemicals (ICF, 1989). Roadmaps
contains information for the SARA 313 chemicals on health and environmental effects, federal
regulations, state air and water regulations and monitoring data, and state contacts. It also summarizes
the publicly available toxicity information from a number of data bases. Roadmaps indicates if there
was sufficient evidence that exposure to a chemical substance resulted in the indicated health or
environmental effect. It indicates that data is available relative to the effect, but it does not report
severity or validity of concern, and it does not report numerical test results.
Data in Roadmaps on other specific effects are divided into five categories: 1) chronic toxicity,
i e health effects (non-cancer) from long-term exposure, 2) developmental toxicity in humans, 3)
heritable genetic and chromosomal mutation in humans, 4) neurotoxicity from chronic exposure and 5)
reproductive toxicity. If data are available which imply exposure to a chemical has one of these five
toxic effects, Roadmaps flags the endpoint and lists the source of the data. It should be noted if an
effect is not flagged for a chemical it does not necessarily mdicate negative test results; it could mean
either that the available data did not support a concern or that data was unavailable. Table A-5 lists the
data sources that Roadmaps references for each of the five endpoints.
A-10
-------
APPENDIX A
TABLE A-5: DATA SOURCES FOR "OTHER SPECIFIC EFFECTS" CITED IN ROADMAPS
Endpoint
Data Sources
Chronic toxicity
Developmental toxicity
Mutagenicity
Neurotoxicity
Reproductive toxicity
DWCD, HAD, HEA, HEED, KEEP, HSDB
ATSDR, DWCD, HAD, HEA, KEEP, HSDB
ATSDR, GENETOX
HAD, HEA, KEEP, HSDB
ATSDR, DWCD, HAD, HEA, KEEP, HSDB
KEY: ATSDR
DWCD
GENETOX
HAD
HEA
HEED
KEEP
HSDB
Agency for Toxic Substances and Disease Registry (ATSDR) Toxicological Profile
Drinking Water Criteria Documents, EPA
Genetic Toxicity Chemical Information System (on-line database)
Health Assessment Documents, EPA
Health Effects Assessment, EPA
Health and Environmental Effects Documents, EPA
Health and Environmental Effects Profiles, EPA
Hazardous Substance Data Bank, National Library of Medicine TOXNET (on-line
database)
Roadmaps data are incorporated into the chemical hazard ranking method data base. A value
of 1 was assigned for each flagged endpoint. If the endpoint was not flagged, a value of 0 was
assigned. The values for the five endpoints were summed in "other specific effects", for a maximum of
5 for each pollutant.
A-ll
-------
A.2 ENVIRONMENTAL EFFECTS
A.2.1 Terrestrial Effects
Definitions/Test Methods:
Oral LD& The concentration of a substance, expressed in mass of the substance per mass of the animal,
that will kill half of a group of rodents within 14 days when administered orally as a single dose.
Data Selection and Calculation of Hazard Value
The rodent oral LDX data used for the acute human health effect endpoint was also used as a surrogate
to represent terrestrial organisms. The hazard values are assigned like those for acute human health
effects.
A.2.2 Acute Acmatic Effects
Definitions/Test Methods;
LCjo: The concentration of a chemical, in water, that causes death in 50 percent of the fish tested. Acute
effects on fish are measured as mortality to Pimephales promelas (fathead minnow) in 1) a flow-through
96-hr test and evaluated as log LC^ (in mg/1). If data from this test are unavailable, then data is selected
from 2) a flow-through 96-hr LCjo data for another fresh water fish, or 3) a static 96-hour fathead minnow
test or 4) a static 96-hour test for another freshwater fish. In two cases, 48-hour test data were used when
96-hour data were unavailable.
Data Selection
Figure A-6 presents the hierarchy for aquatic LCy, data selection. Some functional groups, such
as acid chlorides, isocyanates, epoxides, etc., react with water in less than one day. Experimental
measurement of the aquatic LC^ will, of course, be based on the LC50 of the byproducts. When
experimental data are absent, the aquatic environmental effects (i.e., fish LC^ and fish NOEL) ot such
compounds calculated using a QSAR were, therefore, taken as those of the hydrolysis products rather than
the pollutant itself and the appropriate data substituted into the algorithm. Experimental data from 1) Acute
Toxicities of Organic Chemicals to Fathead Minnows, Volumes I through 5 (CLSES, 1984, 1985, 1986,
1988 1990) and 2) HSDB (1993) are preferred for the acute aquatic toxicity data. Additional data sources
were'used for the inorganic chemicals, because no valid means of estimating acute aquatic toxicity values
for inorganics has been identified (Banerjee and Paul, 1993; Davidson, et al., 1987; Ellgaard and Gilmore,
1984- EPS 1984- Hose, et al., 1989; IPCS, 1986; IPCS, 1990; IPCS, 1991; Smith, et al., 1985; Spehar
etal' 1980; EPA, 1980a; EPA, 1980b; EPA, 1980c; EPA, 1980d; EPA, 1980e; EPA, 1984a; EPA,
1984c). Additional data sources were also used for pesticides (Kidd and James, 1991; EPA, 1984b).
A-12
-------
U'PENDIX A
Water Reactive
Use Fish LC5O
of Byproduct
Flow-through
96-hr
Fathead
Minnow
Experimental
Data
Construct SMILES
(organics only)
Selept Toxicity
Type
Flow-through
• 96-hr
Freshwater
Fish"
Standard
Toxicity
Type
Static
Fathead
Minnow
Select Nonpolar
Narcosis Toxicity
Type
Calculate QSAR LCEO
Static
Freshwater
Fish"
Certain
Functional
Groupb
QSAR log LC50-1
I°9 !-(%(,= QSAR log LC50
a excluding trout
electrophiles, good nucleophiles, strong acids, chemicals with an aromaitic ring, and certain
Figure A-6. Decision Tree for Fish LC50 Data Selection
A-13
-------
If experimental data are not available, but can be estimated by way of a structurally similar
comnound in the same physical state, a QSAR is used to estimate the acute toxicity value
SROQSAR 2 0). First, the chemical SMILES (Simplified Molecular Input Line Entry System) was
constructed to determine the mechanism of toxic action. SMILES is a chemical nomenclature used to
desc be organic molecules for computer entry. Based on the toxicity type e.g., ™^\™™ls'
narcosis polar narcosis, aniline toxicity, ester narcosis, respiratory uncoupling, aldehyde toxicity
toxtity, and reactive toxicity) the LC50 is calculated as a function of the log Kow using QSARs.
There are certain cases where the molecular structure of a pollutant does not suggest a specific
toxic mechanism, so a default nonpolar narcosis was used For certain functional gr oupe, , it is known
that their reactivity increases the toxicity beyond what a QSAR based on K would predict. These
groups include good alkylating agents (electrophiles), good nucleophiles or bases s rong ; acids and
certain reactive groups. For such chemicals, the toxicity was increased over that predicted by default
nonpolar narcosis by -1 log unit.
Calculation of Hazard Values
Figure A-7 is a decision tree of the logic used to calculate the fish LC50 hazard value. Hazard
value scores for the acute aquatic toxicological endpoint were based on the logw of the LC50 values
wl fch were assigned a score between 0 and 5 using a continuous, linear function. Numerical cutoff
va ues were basfd on commonly accepted cutoffs (Behret, 1989; Foran and Glenn 1993; Konemann
ami Visser, 1988; Michigan CMR, 1987). Chemicals with a log Kow greater than 6 were assigned a
hazard value of 0.
A .2.3 Fish Chronic Toxicitv
Definitions/Test Methods:
No Observable Effect Level (NOEL): The highest dosage administered that does not produce toxic
effects (Casarett and Doull, 1986).
Data Selection
Fieure A-8 is a decision tree which shows the method for calculating fish NOEL data.
Experimental data are generally lacking for the fish NOEL endpoint and were not used In, .the serening
tier The literature data on fathead minnow acute toxicity provided, in addition to 96-hr LC50 data, 96
hr median effect concentration (EC50) data (CLSES, 1984-1990). The EC50 values were defined as the
con^en ration causing 50 percent of the fish to show an adverse effect. A comparison of he reported
SS^ted to the formulation of the general rules shown in Figure A-8 for estimating the NOEL
of organic chemicals from the fish LC50 or the log Kow.
A-14
-------
APPENDIX A
Experimental
LC5o data?
log LC50*3
{1000mg/n
log LQo < 0
(1 mg/l)
HV =
5
HV = -1.67*logLC50+ 5.0
Figure A-7. Decision Tree for Aquatic LC50 Hazard Value
A-15
-------
Inorganic \Yes
NOEL = 1/20 Fish
Compound
log Kow<2
NOEL = 1 /4 Fish LC50
NOEL = Fish LC5Q/(5.3*log KpW- 6.6)
Figure A-8. Decision Tree for Calculating Fish NOEL
A-16
-------
APPENDIX A
The NOEL of the inorganic chemicals is based entirely on the fish LC50 since inorganics are
poorly fat soluble and their fish toxicity does not correlate to log Kow. Both the fish LQ0 and log K
are used to estimate the NOEL of organics. Organic chemicals with a relatively high log Kow (e g ""
greater than or equal to 5) are generally more toxic to fish and assigned a lower NOEL compared to
organic chemicals with a relatively low log Kow (e.g., less than or equal to 2). The NOEL for the
remaining organic chemicals is calculated using a continuous, linear function.
Calculation of Hazard Values
Figure A-9 is a decision tree which shows the method used to calculate the fish NOEL hazard
values. Hazard value scores for the subacute aquatic toxicological endpoint were based on the log,0 of
the NOEL values, which were assigned a score between 0 and 5 using a continuous, linear function
Numerical cutoff values were set one order of magnitude lower than the cutoffs for the fish LC™ hazard
values. °
A-17
-------
log NOEL > 2
(100mg/l)
log NOEL * - 1
(0.1 mg/l)
HV = 0
HV = 5
HV = 3.33 - 1.67 log NOEL
Figure A-9. Decision Tree for NOEL Hazard Value
A-18
-------
APPENDIX A
A.3 EXPOSURE PARAMETERS
A.3.1 Persistence
Definitions/Test Methods:
BOD half-life: The BOD half-life is the time (in days) required for a chemical to biodegrade such that
its BOD in water is decreased to half of the original amount.
Hydrolysis half-life: The hydrolysis half-life is the time (in days) required for the amount of a
substance to decrease to one-half of the original amount through hydrolysis reaction in water at pH 7.
Data Selection
The BOD half-life of each organic chemical was determined with the computer assisted version
(MICROQSAR 2.0) of the structural feature approach developed by Neimi, et al. (1987). This was
based on selected literature for 287 chemicals.
Hydrolysis half-life data for the organics, ammonia, chlorine dioxide and hydrochloric acid
were determined with the computer assisted version (MICROQSAR version 2.0) of the Hammett and
Taft substituent constant methods described by Harris (1981).
Metal compounds and certain other inorganic chemicals in highly oxidized states (e.g.,
molybdenum trioxide, thorium dioxide, sulruric acid, nitric acid, and ammonium salts) were assumed
to have infinite BOD and hydrolysis half-lives. Zinc and aluminum dusts were assumed to have half-
lives of 500 days based on the judgment that they would degrade (oxidize) eventually, although slowly.
Calculation of Hazard Values
Figure A-10 is a decision tree which shows the method used to calculate the BOD half-life
hazard values. Figure A-ll is a decision tree which shows the method used to calculate the hydrolysis
half-life hazard values. It was decided to use the same scoring criteria for both BOD and hydrolysis
half-lives. The criteria were based on the distribution of the half-life data and on the range of values
assigned for environmental degradation in other chemical ranking systems in the literature. A
maximum hazard value of 2.5 is assigned to BOD or hydrolysis half-lives greater than 500 days and the
minimum hazard value of 1.0 is assigned for half-lives less than 4 days. Between 4 and 500 days, the
hazard value is calculated between 1 and 2,5 based on a linear scale.
A-19
-------
HV = 1
HV = 2.5
HV = 0.311 In BOD Half-life + 0.568
Figure A-10. Decision Tree for BOD Half-Life Hazard Value
A-20
-------
APPENDIX A
Hydrolysis Half-life
4 days
Hydrolysis Half-life
> 500 days
Yes
Yes
HV = 1
HV = 2.5
HV = 0.311 In Hyrolysis Half-life + 0.568
Figure A-ll. Decision Tree for Hydrolysis Half-Ltfe Hazard Value
A-21
-------
/V.3.2 Bioaccumulation
Definitions/Test Methods:
Aquatic Bioconcentration factor (BCF): The ratio of the concentration of a chemical in fish to its
concentration in water at steady-state conditions. This factor is a measure of the chemical s ability to
bioaccumulate and is typically reported in log units.
Data Selection
The experimental BCF for aquatic organisms can vary by several orders of magnitude
depending on specific test parameters and intra- and inter-species differences (size, age, etc.).
Therefor?, log BCF of each organic chemical was determined using the QSAR^equation developed by
Bintein et al. (1993) which considers these differences: log BCF = 0.910 log Kow - 1.975 log (6.8 10
K + 1) - 0 786 Experimental log BCF data tabulated by EPA was used when available for inorganic
chemicals (EPA 1979) Numerical values for barium and cobalt compounds were based on ranges of
BCF values from HSDB. Otherwise, the log BCF endpoint was flagged as missing data and no hazard
value was assigned.
Calculation of Hazard Values
Figure A-12 presents the method used to calculate the log BCF hazard values. The BCF value
increases with increasing K^ until the log Kow reaches a value of approximately 6. Beyond a log K of
6 the BCF drops off. Using this model, the maximum possible log BCF is approximately 4.5. Based
on this range of BCF values, a maximum hazard value of 2.5 is assigned for log BCF * 4 and the
minimum hazard value of 1.0 is assigned for log BCF , 1, Between log BCF of 1 and 4, the hazard
value is calculated between 1.0 and 2.5 based on a linear scale.
A-22
-------
APPENDIX A
HV = 0.5 log BCF + 0.5
HV = 1
HV = 2.5
Figure A-12. Decision Tree for BCF Hazard Value
A-23
-------
A.4 \VEIGHTINGBYRELEASES
The final hazard value considers both the intrinsic properties of each chemical and the
likelihood of exposure. The hazard values assigned to the toxicological endpoints are multiplied by a
release weighting factor (RWF) based on the appropriate type of TRI releases or transfers to air or
water.
Definitions
Release weighting factor (RWF): A multiplicative factor used to weight toxicity hazard values for each
chemical according to the amount of its annual releases or transfers to air and water, where
RWF = In releases - 10
Data Selection
Data for the releases and transfers for the industrial chemicals were obtained from the 1989
TRI. Releases for pesticides, which are intentionally released, were obtained from annual usage
information for 1987 (EPA, 1988b), 1990 and 1991 (Aspelin et al., 1992).
All releases were classified as either air or water releases. It is assumed that stack and fugitive
releases went to air, and that land, injection, water, and POTW releases went to water. Off-site
transfers to an incineration Or recycling facility were assumed destroyed or not released to the
environment. All other off-site transfers were assumed released to water.
A method was developed for scoring the releases on a smooth scale from 1 to 10 on a
logarithmic basis. Using the natural log (In) gives the data a normal distribution. The natural log,
rather than the base-10 log, was used to attain a range of 10 integers over the range of release amounts.
Total releases for the chemicals scored ranged from 860 Ibs. to 545,989,541 Ibs. The equation for
calculating the RWF results in the assignment of a multiplier of approximately 10 for the highest
release and a 1 for anything that is 59,874 Ibs or less. By subtracting 10 from the natural log of the
releases, a cutoff of 60,000 Ibs. is set, below which the weighting factor is always equal to one.
Calculation of Hazard Values
Although it is understood that releases to one medium can result in exposure by multiple routes,
for the purpose of simplicity in this screening tier, it was assumed that air releases would result in
inhalatory exposure and that water releases would result in oral exposure as well as exposure to aquatic
organisms. Fugacity modelling in the next tier is expected to provide more realistic assumptions. The
RWF was applied in the following manner:
The weighting factor for air releases (RWFair) was applied to the hazard value assigned
for the inhalation rodent LC50.
The weighting factor for water releases (RWFwalcr) was applied to the oral rodent LD50,
fish LC50, and fish NOEL.
A-24
-------
APPENDIX A
• The weighting factor for the total air and water releases (RWF1()lal) was applied to the
chronic toxicological endpoints for carcinogenicity and "other specific effects."
Therefore, the final weighted hazard values (wHV) for the human health effects and the environmental
effects were obtained as follows:
Weighted Human Health Effects = (HVorall.D50)(RWFw:iwr)+(HV „ ,.C50)(RWFair)
+(HVcarci,, + HVolhcr)(RWF,olal)
Weighted Environmental Effects = (HVorall.D50 + HVfishl.C50 + HVfisllNOi;iJ(RWFwa,r)
The Exposure Factor remains unchanged where:
Exposure Factor = HVBOD + HVhydrolysis + HV,
BCF
The final total wHV for each chemical is obtained as follows:
wHV = (weighted Human Health Effects + weighted Env. Effects) * Exposure Factor
These numbers provide the basis for the ranking of all scored chemicals.
A-25
-------
A.5 REFERENCES
in A L A H Gruhe and R. Torla (1992) Pesticide Industry Sales and Usage: 1990 and 1991
Market Estimates. U.S. Environmental Protection Agency, Office of Pesticides and Toxic
Substances.
Banerjee, T.K. and V.I. Paul (1993) Estimation of acute toxicity of ammonium sulphate to the freshwater
catfish Heteropneustes fossils. Biomed. Environ. Set. 6 (1): 45-58.
Bchret, H. (Ed.) (1989) Existing Chemicals of Environmental Relevance. GDCh-Advisory Committee on
' Existing Chemicals of Environmental Relevance. New York: VCH.
Bintein S J. Devillers and W. Karcher (1993) Non-linear dependence of fishbioconcentrationonn-
' octanol/water partition coefficient. SAR and QSARin Environmental Research, 1: 29-39.
Casarett, L.J. and J. Doull (1986) Toxicology - Casarett andDoull's Toxicology: The Basic Science of
'poisons, 3rd edition. MacmiUan, New York.
Center for Lake Superior Environmental Studies (CLSES), (1984) Acute Toxicities of Organic Chemicals
to Fathead Minnows (Pimephalespromelas) Vol.1. (L:T. Brooke, D J. Call, D.L. Geiger and
C.E. Northcott, Eds.) University of Wisconsin-Superior, Superior, Wisconsin.
Center for Lake Superior Environmental Studies (CLSES), (1985) Acute Toxicities of Organic Chemicals
to Fathead Minnow (Pimephales promelas) Vol.2. (L.T. Brooke, DJ. Call, D.L. Geiger and
C.E. Northcott, Eds.) University of Wisconsin-Superior, Superior, Wisconsin.
Center for Lake Superior Environmental Studies (CLSES), (1986) Acute Toxicities of Organic Chemicals
to Fathead Minnows (Pimephales promelas) Vol.3. (L.T. Brooke, DJ. Call, D.L. Geiger and
C.E. Northcott, Eds.) University of Wisconsin-Superior, Superior, Wisconsin.
Center for Lake Superior Environmental Studies (CLSES), (1988) Acute Toxicities of Organic Chemicals
to Fathead Minnom (Pimephalespromelas) Vol.4, (L.T. Brooke, DJ. Call, D.L. Geiger and
C.E. Northcott, Eds.) University of Wisconsin-Superior, Superior, Wisconsin.
Center for Lake Superior Environmental Studies (CLSES), (1990) Acute Toxicities of Organic Chemicals
to Fathead Minnows (Pimephales promelas) Vol.5. (L.T. Brooke, D J. Call, D.L. Geiger and
C.E. Northcott, Eds.) University of Wisconsin-Superior, Superior, Wisconsin.
Davidson K A., P.S. Hovatter, C.F. Sigmon (1987) Water Quality Criteria for White Phosphorus. Oak
Ridge National Laboratory, Oak Ridge, Tennessee, AD-ORNL-6336.
Elluaard, E.G., and J.Y. Gilmore ffl (1984) Effects of different acids on the bluegill-sunfish, Lepomis
macrochirus Rafmesque. /. FishBiol., 25:133-137.
Environmental Protection Service (EPS) (1984) Ammonium Nitrate-Technical Information for Problem
Spills. Technical Services Branch, Ottawa, Ontario.
A-26
-------
APPENDIX A
ForanJ.A., andB.S. Glenn (1993) Criteria to Identify Chemical Candidates for Sunsetting in the Great
Lakes Basin. The George Washington University, Environmental Health and Policy Program
Department of Health Sciences, Washington, D.C.
Harris, J. C. (1981) Rate of hydrolysis. In Research and Development of Methods for Estimating
Physicochemical Properties in Organic Compounds of Environmental Concern, Final Report
Phase II. Part I, Chapter 7. Arthur D. Little Inc. '
Hazardous Substances Data Bank (HSDB) (1993) The National Library of Medicine's Toxicology Data
Network (TOXNET) System. y
Hose, I.E., D. Di Fiore, H.S. Parker, and T. Sciarrotta (1989) Toxicity of chlorine dioxide to early life
stages of marine organisms. Bull ofEnv. Comtam. Toxicol. 42:315-319.
ICF, Inc. (1989). SARA Section 313 Roadmaps DataBase User's Manual- Version 2.10, U.S. Department
of Commerce National Technical Information Service PB( )-174855.
International Programme on Chemical Safety (IPCS) (1986) Environmental Health Criteria 54 - Ammonia
World Health Organization, Geneva. ,
International Programme on Chemical Safety (IPCS) (1990) Environmental Heiilth Criteria 107 - Barium.
World Health Organization, Geneva.
International Programme on Chemical Safety (IPCS) (1991) Environmental Heiilth Criteria 108 - Nickel
World Health Organization, Geneva.
Kidd, H. and D.R. James (1991) The Agrochemicals Handbook, Third Edition. Royal Society of
Chemistry, Cambridge, England.
Konemann, H. and R. Visser (1988) Selection of chemicals with high hazard potential: Part 1: WMS-
Scoring System. Chemosphere, 17:1905-1919.
McGregor, D.B. (1992) Chemicals Classified by IARC: Their Potency in Tests for Carcinogenicity in
Rodents and their Genotoxicity and Acute Toxicity. Mechanisms of Cardnogenesis in Risk
Identification, H. Vainio, P.N. Magee, D.B. McGregor & A.J. McMichael, Eds., Lyon,
International Agency for Research on Cancer, pp 323-352.
Michigan Critical Materials Register (CMR) (1987). Criteria and Support Documents, Michigan
Department of Natural Resources.
Niemi, G.J., G.D. Veith, R.R. Regal and D.D. Vaishnav (1987) Structural features associated with
degradable and persistent chemicals. Environ. Toxicol. Chem. 6: 515-527.
O' Bryan, T.R. and R.H. Ross (1988) Chemical scoring system for hazard and exposure identification /
Toxicol. Env. Health, 1: 119-34.
A-27
-------
Registry of Toxic Effects of Chemical Substances (RTECS) (1983-84; Supplement) National Institute of
Occupational Safety and Health. U.S. Department of Health and Human Services, Public Health
Service, Center for Disease Control. Cincinnati, Ohio.
Registry of Toxic Effects of Chemical Substances (RTECS) (1992, 1993) National Institute of
Occupational Safety and Health. The National Library of Medicine's Toxicology Data Network
(TOXNET) System.
Sax, N.I. (1989), Dangerous Properties of Industrial Materials, 7th ed., Van Nostrand Reinhold.
Smith, R.L., T.M Holsen, N.C. Ibay, R.M. Block and A.B. DeLeon (1985) Studies on the acute toxicity
of fluoride ion to stickleback, fathead minnow, and rainbow trout. Chemosphere, 14 (9): 1383-
1398.
Spehar, R.L., R.W. Carlson, A.E. Lemke, D.I. Mount, Q.H. Pickering and V.M. Snarski (1980) Effects
' of pollution on freshwater fish. Journal Water Pollution Control Federation, 52 (6): 1703-1767.
United States Environmental Protection Agency (EPA) (1979) Water-Related Fate of 129 Priority
Pollutants, Vol. I. Office of Water Planning and Standards, Washington, D.C. EPA -440/4-79-
029a.
United States Environmental Protection Agency (EPA) (1980a) Office of Water Regulations and
Standards. Ambient Water Quality Criteria for Cadmium. EPA 440/5-80-025.
United States Environmental Protection Agency (EPA) (1980b) Office of Water Regulations and
Standards. Ambient Water Quality Criteria for Copper. EPA 440/5-80-036,
United States Environmental Protection Agency (EPA) (1980c) Office of Water Regulations and
Standards. Ambient Water Quality Criteria for Chromium. EPA 440/5-84-029.
United States Environmental Protection Agency (EPA) (1980d) Office of Water Regulations and
Standards. Ambient Water Quality Criteria for Lead. EPA 440/5-80-057.
United States Environmental Protection Agency (EPA) (1980e) Office of Water Regulations and
Standards. Ambient Water Quality Criteria for Nickel. EPA 440/5-80-060.
United States Environmental Protection Agency (EPA) (19801) Office of Water Regulations and Standards.
Ambient Water Quality Criteria for Zinc. EPA 440/5^80-079.
United States Environmental Protection Agency (EPA) (1984a) Office of Water Regulations and
Standards. Ambient Water Quality Criteria for Arsenic. EPA 440/5-80-021.
United States Environmental Protection Agency (EPA) (1984b) Office of Toxic Substances. Pesticide Fact
Sheet - Chlorpyrifos. EPA/540/FS-87/037.
United States Environmental Protection Agency (EPA) (1984c) Office of Research and Development,
Cincinnati, Ohio. Health Assessment Document for Manganese. EPA/600/8-83-013F.
A-28
-------
United States Environmental Protection Agency (EPA) (I988a). Intent to review "Guidelines tor
Carcinogen Risk Assessment." Fed. Register 53, 32656.
United S^er^n^m^tal Protection Agency (EPA) (I988b), Pesticide Industry Sales and Usage, 1987
^existing
A-29
-------
-------
APPENDIX B
TRI CHEMICALS AND HIGH-VOLUME PESTICIDES
-------
-------
CHEMICALS SELECTED FROM 1989 TRI BASED ON 99% OF TOTAL RELEASES
Selected Chemicals Original list of Chemicals
[Selected Chemicals Original list of Chemicals
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1,1,1-trichloroethane
1.1,2,2-tetrachloroethane
1.1,2-trichloroethane
1.1-dimethyl hydrazine
1,2,4-trichlorobenzene
1,2,4-trimethylbenzene (pseudocumeme)
1,2-butylene oxide
1,2-dibromoethane
1,2-dichlorobenzene
1,2-dichloroethane
1,2-dichloroethylene
1,2-dichloropropane
1,3-butadiene
1,3-dichlorobenzene
1,3-dichloropropylene
1,4-dichlorobenzene
1,4-dioxane
2.4,5-trichlorophenol
2,4,6-trichlorophenol ;
2,4-D
2,4-diaminoanisole
2,4-diaminoahisple sulfate
2,4-diaminotoluene
2,4-dichlorophenol
2,4-dimethylphenol
2,4-dinitrophenol
2,4-dinitrotoluene
2.5-dichloro-3-aminobenzoic acid
2,6-dinitrotoluene
2,6-xylidirte
2-ethoxyethanol
2-methoxyethanol
2-nitrophenol
2-nitropropane
2-phenylphenol
3,3'-dichlorobenzidine
3,3'-dimethoxybenzidine
3,3'-dimethylbenzidine
4,4'-diaminodiphenyl ether
4,4'-isopropylidenediphenol
4,4'-methylenebis(2-CI-aniline)(mboca)
4,4'-methylenedianiline
4,6-dinitro-o-cresol
4-aminoazobenzene
4-aminobiphenyl
4-nitrophenol
5-nitro-p-anisidine
Acetaldehyde
Acetamide
Acetone
Acetonitrile
Acrolein
Acrylamide
Acrylic acid
Acrylonitrile
X
X
X
X
X
X
X,
I
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Allyl chloride
Alpha-naphthylamine
Aluminum (fume or dust)
Ammonia
Ammonium nitrate (solution)
Ammonium sulfate (solution)
Aniline
Anthracene
Antimony
Antimony compounds
Arsenic
Arsenic compounds
Asbestos (friable)
Barium
Barium compounds
Benzal chloride
Benzamide
Benzene
Benzoic trichloride
Benzoyl chloride
Benzoyl peroxide
Benzyl chloride
Beryllium
Beryllium compounds
Biphenyl
Bis(2-chloro-1 -methylethyl) ether
Bis(2-chloroethyl) ether
Bis(2-ethylhexyl) adipate
Bis(chloromethyl) ether
Bromoform
Bromomethane
Butyl acrylate
Butyl benzyl phthalate
Butyraldehyde
C.i. basic green 4 (Malachite Green Oxal
C.i. basic red 1 (Rhodamin 6G)
C.i. direct black 38
C.i. disperse yellow 3
C.i. solvent yellow 14 (Sudan I)
C.i.food red 15
Cadmium
Cadmium compounds
Calcium cyanamide
Captan
Carbaryl
Carbon disulfide
Carbon tetrachloride
Carbonyl sulflde
Catechol
Chlordane
Chlorine
Chlorine dioxide
Chloroacetic acid
Chlorobenzene
Chloroethane
B-l
-------
CHEMICALS SELECTED FROM 1989 TRI BASED ON 99% OF TOTAL RELEASES
Selected Chemicals
Original list of Chemicals
X Chloroform
X Chloromethane
Chloromethyl methyl ether
X Chlorophenols o
X Chloroprene
X Chlorothalonil
Chromium
X Chromium compounds
Cobalt
X Cobalt compounds
Copper
X Copper compounds
X Cresol (mixed isomers)
X Cumene
X Cumene hydropefoxide,
Cupferron
Cyanide compounds
X Cyclohexane
X Decabromodiphenyl oxide
X Di(2-ethylhexyl) phthalate
X Diaminotoluene (mixed isomers) 2,4
Dibenzofuran
X Dibutyl phthalate
X Dichlorobenzene (mixed isomers)
X Dichloromethane
Dichlorvos
Dicofol
X Dlethanolamine
X Diethyl phthalate
Diethyl sulfate
X Dimethyl phthalate
Dimethyl sulfate
X Epichlorohydrin
Ethoxylated C10-C16 Alcohols
Ethyl acrylate
Ethyl chloroformate
X Ethylbenzene
X Ethylene
X Ethylene glycol
X Ethylene oxide
Ethylene thiourea
Fluometuron
X Formaldehyde
X Freon 113
X Glycol ethers (use tri)
Heptachlor
X Hexachloro-1,3-butadiene
X Hexachlorobenzene
Hexachlorocyclopentadiene
X Hexachloroethane
Hydrazine
Hydrazine sulfate
X Hydrochloric acid
X Hydrogen cyanide
X Hydrogen fluoride
~1 [Selected Chemicals Original list of Chemicals
X Hydroquinone
X Isobutyraldehyde
X Isopropyl alcohol (manufacturing,
Lead
X Lead compounds
Lindane
M-cresol
X M-xylene
X Maleic anhydride
Maneb
Manganese
X Manganese compounds
Mercury
Mercury compounds
X Methanol
Methoxychlor
Methyl acrylate
X .Methyl ethyl ketone
Methyl hydrazine
Methyl iodide
X Methyl isobutyl ketone
Methyl isocyanate
X Methyl methacrylate
X Methyl tert-butyl ether
Methylene bromide
X Methylenebis(phenylisocyanate)
Michler's ketone
X Molybdenum trioxide
X N,N-dimethylaniline
X N-butyl alcohol
X N-dioctyl phthalate
N-nitrosodimethylamine
X N-nitrosodiphenylamine
X Naphthalene
Nickel
X Nickel compounds
X Nitric acid
Nitrilotriacetic acid
X Nitrobenzene
Nitroglycerin
O-anisidine
O-cresol
O-toluidine
X O-xylene
P-anisidine
P-cresidine
X P-cresol
P-nitrosodiphenylamine
P-phenylenediamine
X P-xylene
Parathion
Pentachlorophenol
Peracetic acid
X Phenol
Phosgene
B-2
-------
CHEMICALS SELECTED FROM 1989 TRI BASED ON 99% OF TOTAL RELEASES
jSatoctad Chemicals Original list of Chemicals I
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Phosphoric acid
Phosphorus (yellow or white)
Phthalic anhydride
Picric acid
Polychlorinated biphenyls
Propionaldehyde
Propoxur
Propylene
Propylene oxide
Propyleneimine
Pyridine
Quinoline
i
Quinone
Quintozene
Saccharin (manufacturing only)
Safrole
Sec-butyl alcohol
Selenium
Selenium compounds
Silver
Silver compounds
Styrene
Styrene oxide
Sulfuric acid
Terephthalic acid
Tert-butyl alcohol
Tetrachloroethylene
Tetrachlorvinphos
Thallium
Thallium compounds
Thiourea
Thorium dioxide
Titanium tetrachloride
Toluene
Toluene-2.4-diisocyanate
Toluene-2,6-diisocyanate
Trichlorfon
Trichloroethylene
Urethane
Vanadium (fume or dust)
Vinyl acetate
Vinyl bromide
Vinyl chloride
Vinylidene chloride
Xylene (mixed isomers)
Zinc (fume or dust)
Zinc compounds
Zineb
other mixtures or trade names
HIGH VOLUME PBSTICIDnS SELECTED
Alachlor
Atrazine
Bulylatc
Captan
Carbaryl
Chlorpyrifos
Cyanazinc
1,3-Dichloropropene
EPTC
Glyphosate
Malalhion
Maneb
Melam-sodium
Meihyl parathion
Melolachlor
Metribuzin
Terbufos
Trifluralih
B-3
-------
-------
APPENDIX C
RANKING RESULTS: HORIZONTAL TABLES
-------
-------
111
a
cc
m
|
o
o
•= E
II
- Q.
to (O CO CO
000000000000000000
0.000-0001 O'OO- 0-0 <0*-
CO ^T
o — — o'
o — ooovqoja— oiaioc9O— oianao
— ca — no o'oi oi dc5 — oic>
T-*o oin
o o T-" cJ ^
eg eo W -^ *:
U> O ^T O €0 W f CO (0 CT
cy cy d ci ci d
8 in o o ;
CO O ^
OJ CM Is- v-
ooooooooooooooooooo
oooooinooiooooowooooooooioowioooooo
n " . <0
d d d do d i-
COOQQCQOOOO OOO
. •» i'-. «m
otooo 9099
dddd'cyddoJd
•^dd o'd-r^ddddddd -»-^^
§'§§§" I ^|K§§|§II5§'8S
odd do
— do" d d —" — — o" d — oi o" oi — o oi — co o"
So
^2.
o
€
lag
>- § s
II?-
TS S
'I*
o E
5-1
D O
j? e
>- 0) £ - ~
(3 2 S < 5
II
• 3
O -J
£, ^
o fe
1 s * g i
H 1^1
5 TI Op " c
E ,
fig
i. o s
gl E_
R 2 A
9)
.2
0) -C
iS
i
ItIIII
Q- x a. z a °>
f I
if
l|I
ill
II
I *
•D
S
C-I
-------
000000-OOCMCMO--001-OOIMO--
< 3 32-535-5.--S533
o^d- ".' o
,„. 0-00-
o.ooooooo, oo. oo, ooo
oo, oooooo
ooooo ooo-ooooooo3,o3o
.33.3000.333. a
. .0,.
5 SS 8-5 88 I
•*" T- u/
§•§115
•i- _ X) £• Jf
c >- f ^
O I Q. <-
C-1
-------
oc
o
p
g
i
81
CM eq CM *-"
-53 3? 3-?
§! 33 3 =
S3 5 ^3 33 -3353333 55 bs-S 3
33§
---
tn tn in in in in in ---tncoininpi^-ini-
CM CM CM C\i CM CM CM CM ^ CM CM *° CM CM
ininininincotoininintn'*-
CMCMCMCMOi^CMCMCMtMOJ
•r-inint-cointnininintninioin
CMCM 'JcMcyoicMcycMcJcMw
8 8 S! "
2 33 33 3 -3 55 53 53 33 33 3.35 S3:3^33335B335 33 3333 53 33 33 333 3'
K m
g n a, «
tn u>
8^1
•*- CM •»-
-88
~
*~ S "*"
88
COPJ O)O OOOO)
cptocpoocomo)
CM ' •«— CU CO *
2 ' S
s ° s
s s
CM •*-
5 | 8 8 S
3 3 3 3 5 3 »" 3
§O O O
ji ^r -^
TO CM "~
2 3 3-3 3
15 O -'
-------
*- «J O O
•»
•1
5 «
S3
C-4
-------
cc
o
o
S o
>
X
>
I
II
>
I
81
>
X
c E.
>
i
ia
E
oj co co co 50 03 o> cfi !•* in
ooooooo'oio
OOIOOOOOOO
3 a
3£-c'*"?00?0':a!cl>0«'1-ooi«>«>o>cooi
« n « 01 o ^ 01 - d •» d oi -: d
^ -T CO *-
SCO ^J- o
cp in o
§ *• 2 S
Soi
-------
_ 2
i*:f«5
5 ™
I- O •» Q •» 01 io. 5 •« g « «. •V10. 5 0>- 5 ™ S S <8 2 5
•°. t s s TO a s * s 2 s 5 s> s e q ": <8.
b ss « s - K s 3 S s-« " S 2 s. s.
|S I « o iouiCTe>ioo-ffm
oi co co d -r-
d co -«r cy -v
88S2^8^S8
— i « in t- cy to ••-
y^ O Ol «^ O O
^ cy o o o cy
co cy «o ^r
^— O V x? *
S 888 8
,J d — — O O
SSR888128
cocao—
T- -- CM
-r^ ci co co •*- cy
c«i co cy co o
S
„ I
" -
- S-i^S
» ^ o
S E
1 .g *
g
a. Si co x o
|S.,IIs|
C
!iii-ll§i2Illtllll
in
•5. 2
** W ~^ 3 ^» -L- J^« .^ O **
tSSZOOOIZI-<
°S SS-^9 E2-^-5
S-^0^^^^^^^1*1*-.^^
Slo:£o--oc3Gjos5DO
-------
sis- i
335
J
S
j?
£
li
>
X
'5S--!»*--0'g*--'-a'
e« co to iv n ro
« T' d -^ ci
9 - Y d d
5 S 2
3 * " T S 3
3 -3 33 3 -3 3 3 -3
3 -33333 33 3 33 3-
*-
X
a
d U) U> »— u)' ^ CV
o °
1
O
I
5
ooo 033000, oo 0,30 ««o'.jo-o3o3oo3,jo ooo 3- 3' 30 0303
00330,0
I
a>-r-aj. {0
£ ^>
i c I
P. 3 P
Ki I
•2
I
t
H S. 4
*£.
I
d
I
a
5-
E
Tl
O
>, c
I'll
»
:s
X
E
' -^ rtT
•g I g
a. >,
-------
2 2
_ * S
*- co CM
— w-OOlCT^WOI
SU5 *• t/J *• trt U> U>
oi oi J CM ci oi
to in in in in u> u>
OJCidCICMOJCMOl
in in o in to in
8 g
O
-------
Q
UJ
Q
D
O
Ul
o.
CO
a:
UJ
I
§
I
oc
I
It
5 E
5522
El I
E
*:
.
-o-in rr^-^ v-«f co^f " 10 •» •» >r rr "'•«•«•«•»
§5!"2^0"0^^^0?""8-"«Sfei55g5SS
0<*~ ^~ *~ ^- o i- o^- cj -r- cu •»- TT c5 eo ^^oui
K CM *-
8" s: Si
TO co
2 2
u> ri
d ddddd d do ^ ^ -^ -^
oooooooooooooooooooo
£8588. ^S^SSS
OOOOOaiOO.i-000<-OOOOOI00000000ieilOOO
r>- ^r cy •»-;
9 T 9 T
in in in T-
in o
O O -^ t*.
S ° S 2 § S ° 2 ° 5
® o ^ o cs o ^_ ^:«'0's-^:'-'^-a>tf)Oio^-cyoor*o>u)ey^-oa)oiovco£oo
o -r- o -r- o o T- oo o" o* -r: o* W ,o* -r-" d d do d -^ 2 ^
§ I f I m f I § § § § I s 1 1 § S I I § § i § i I | § § f | | 8 8
12
CO
oooot^ddd
9°
-^
-------
_ 2 *
8 8* S 5 5 ? 5
^ ^. to 10 o to T (o
o, *i e o i * « fc S 8] S 5 S 5 S
««V«)into*c CD in in o
T-OCMOOO-^OOOOIOOOOOOI
"*""*" "" "" " *" _ _ ^mni cn»-mco^ «w«;r°3«. -«. ^"ow2ToS3«S22o9-o-'T':''io'0'0 ^ °°
o«-*- w o»- •»-,
S'5SS5"-ss?s'asssss2S~sB'"?s'"ss?s^s'"s""s's"^ss'"sss
illlil"
§88 S
8 8 S 8 S 8
ooogoocococoaioicor-moiojcocoggcagoimor-cooocMocog
S- S S n 3 3 " 2 S " S S 3 5 " S S S S 3 " 3 S 2 5 5 S S S 25
h-m
o^cypey^^pjcypj oioi
tojDOinoioQcyocOQJgg
irtcyepcocqiotom
cy ol co -^ oi w
a a> co o g g
o do
r^-tnoiiomco-^-tDweocncocvjinor^^win^rin
.: * _: -^ CJCJ^ T^CJCJ cyCMCTOI "T
I
000000000000000000000
000000000000000
cMcoocococMoJ---oiinoi^v-ncqcqocM^mcorro)^;iqcoco^:cq
«>'«10, N, neo. •". oc!2SdScMdrid^cM So drioicj^codd oi--- ^ooira^-co-oi -'-
IE
S
8-
I
„ 2
01 CI
80,"
w ,- X
f I?
S S s
s •
!;
it *5*
11 5 -S.1
i 6 h x i
x o
Ull
i
,9
•" 0) J_ «U
S f 8 S
o 8 |
Si I
I
^f ISIS
H*s**
±:g
S S
. - -
2Q-KlUJ § 5 ?
E S
e fe a s iJ Q.
5f l£S 8
-------
o
I
i
I
CO
oc
I
I
I
a:
Is
fc S K
v:
S
oi-r-mb- „:
110
OOOOO
co ID '
OOO
o>ioco
CO O *~ O)
O) O) (O
'OOOOOOO*-
ooooo
*-••-*- *- S SQ *- o>
CO-r- M1 OJ - ^
-f- ^ u> •»-
CJ OJ
o o o c5
8 8 8
°s
S
53?232?=323?353323^3-232335^33233333332333333333^
a,a,^o, .»
33 £ T3
f gl :
f?l •& :
E iiilf
S f S I 1 § 9
C-ll
-------
2 2 _ S
8 8 § S 5 S S 8.
o o «> in o»
in — "I:;
« CM T
in in m, in in « u> in
oooooooo
oomcoinmoi
o **
o»oooa>ou>CT
o
I
I
jr
8
m
SS'S
ss ss
o o q m o in o
C-12
-------
APPENDIX D
RANKING RESULTS: CHEMICAL SCORES
-------
-------
TOTAL HAZARD VALUES AND CHEMICAL RANKS, WEIGHTED BY RELEASES
chemical Hazard Value (norms
default=0 c
Chromium cmpds 100
Arsenic cmpds . 93.7
Lead cmpds 953
Copper cmpds 86.7
Terbufos (tBuSCH2SP(=S)(OEt)2 85.3
2.4-D 84-6
Nickel cmpds 84.4
Formaldehyde 83.7
1 ,3-dichlbropropene 77.5
Trifluralin 761
Cadmium cmpds 74.9
Ammonia 72.5
Sulfuric acid 72.1
Hydrogen fluoride 67.2
Nitric acid 64.5
Hydrochloric acid 63.9
Styrene 62.2
Chlorpyrifos 60.3
Hydrogen cyanide 53.4
Tetrachloroethylene ' 53.3
Trichlproethylene 55.1
Chlorine 55.6 -.
Manganese cmpds 54.1
Chforothalonil 53.9
Di(2-ethylhexyl) phthalate 52.7
Hexachlorobenzene 49.9
Naphthalene 43.4
3hosphoric acid ' '48.3
Dobalt cmpds 48.2
3henol 471
Barium cmpds 46.6
'olychlorinated biphenyls 46
Benzene 44 6
Daptan 44.3
Vcrylamide 43.9
Machlor 436
Chloroform 43.2
Kphenyl 43.2
kcrylonitrile 42.5
,2,4-Trichlorobenzene 42.3
,2-Dichloroethane 42.1
inc compounds 41.7
ylene (mixed isomers) 41.3
trazine 4-1.1
,3-butadiene 39.4
ecabromodiphenyl oxide 38.4
,4-Dinitrotoluene 38.2
,1,1-Trichloroethane 38.1
lethyl Parathion 36.9
letam Sodium (MeNHCS2Na) 36.9
hosphorus (yellow or white) 36.7
alathion 35.7
thylene oxide 34.9
Jlized) Chemical Rank
efault=5 default-0 default-5
100 1 1
91.9 2 3
95.7 3 2
90 4 4
72.9 5 7
75.5 6 6
83.8 7 5
71.6 8 8
66.3 9 9
65.1 10 10
64 11 n
62 12 13
61.6 13 14
57.5 14 16
55.1 15 18
•54.6 16 19
'53.2 17 23
54.3. 18 21
49.9 19 25
49.8 20 • 26
48 , 21 29
47.5 22 30
61.3 23 15
46.1 24 32
56.1 25 17
42.7 26 35
41 .4 27 36
41.3 28 37
46.3 29 31
40.3 30 38
50.4 31 24
42.9 32 34
38.1 33 40
37.9 34 41
37.5 35 42
39.7 36 39
37 37 43
36.9 38 44
36.3 39 47
36.1 40 48
36 41 49
53.3 42 22
35.3 43 50
35.2 44 51
33,7 45 52
36.3 46 46
36.5 47 45
32.6 48 53
31.6 . 49 54
31.5 50 55
49.2 51 27
30.5 52 57
29.8 53 58
D-l
-------
TOTAL HAZARD VALUES AND CHEMICAL RANKS. WEIGHTED BY RELEASES
nhamlnal Hazard Value (norrru
default=0 c
Cyanazine 34-6
Carbaryl 33-3
Dibutyl phthalate 33-1
2,4-Dinitrophenol 33
Carbon disulfide 32-7
2-nitropropane 32-6
EPTC (ethyldipropylthiooarbamate) 32.3
Cyolohexane 32-1
Cresol (mixed isomers) 31-8
Metolachlor 31-7
DIchloromethane 31.1
Ethylbenzene 3°-7
Epichlorohydrln 30-5
Toluene 30-4
Vinyl chloride 29-8
Acetaldetiyde 29-7
Acrylic acid 28-9
DIamlnotoluene (mixed isomers) [24] 28.6
p-Cresol 28-4
Aniline 27-9
Anthracene 27-3
Maneb 25-9
Carbon tetrachloride 26-7
Chloroprene 26-5
Butylate 25-4
Picric acid 25-2
Molybdenum trioxide 25
Chlorobenzene 24-8
Cumene 24-4
1,2.4-Trimethylbenzene 24.3
Asbestos (friable) 23-9
1 ,2-DIchlorobenzene 23-3
4-nitrophenol 23
Propylene oxide 22-3
Zinc (fume or dust) 22-3
Methyl ethyl ketone 22-1
Isopropyl alcohol 21 -8
Methytenebis(phenylisocyanate) 21 .2
4,4'-lsopropylidenediphenol 2O.S
Bis(2-ethylhexyl) adipate 19-8
1 ,4-Dichlorobenzene "1 9-2
o-Xylene "O-1
Catechol 19-1
4,4'-Methylenedianiline 18-6
Glycol ethers 17-7
Terephthallc acid 17-7
1.1.2-Trichloroathane 17.6
Cumene hydroperoxide 17
Butyl benzyl phthalate 1 6.7
Hexachloro-1 ,3-butadiene 1 6.1
Chlorine dioxide 16.1
Chloromethane 15-9
2-methoxyethanol 15-4
ilized) Chemical Rank
efault=5 default=0 default=5
29.6 54 59
28.5 55 61
28.3 56 62
31 57 56
28 58 63
27.9 59 64
27.6 60 65
27.4 61 66
27.2 62 67
29.5 63 60
26.6 64 68
26.3 65 69
26 66 70
26 67 71
25.5 68 73
25.4 69 74
24.7 70 75
24.4 71 77
24.3 72 79
23.8 73 81
23.4 74 83
48.8 75 :28
22:9 76 84
22.6 77 85
24.2 78 80
24.4 79 78
25.7 SO 72
21 .2 81 87
20.9 82 88
20.7 83 89
54.4 84 20
19.9 85 91
19.7 86 92
19 87 93
63 88 12
18.9 89 94
18.6 90 95
18.1 91 97
17.7 92 99
20.1 93 9C
16.4 94 10C
16.3 95 101
18.6 96 9£
15.9 97 10C
15.2 98 10*
24.7 99 76
15 100 10E
14.5 101 10£
21.3 102 8C
13.7 103 1K
13.7 104 11
13.6 105 11!
13.2 106 11'
D-2
-------
TOTAL HAZARD VALUES AND CHEMICAL RANKS, WEIGHTED BY RELEASES
chemical Hazard Value (norrm
default=0 c
Vinyl acetate 15.2
p-Xylene 15
m-Xylene 14.8
Hydroquinone 14.7
N,N-Dimethylaniline 14.6
Methyl isobutyl ketone 14
Chloroethane 13.g
n-butyl alcohol 13.5
Bromomethane 13.1
Metribuzin 12.g
Nitrobenzene 1'2.8
Phthalic anhydride 12.6
Pyridine 12.6
Vinylidene chloride 1 2.4
Maleic anhydride 12.3
2-ethoxyethanol 12.2
Propionaldehyde 12.1
Titanium tetrachloride 1 1 .9
1 ,4-Dioxane 1 1 .8
Diethanolamine 1 1 .7
Chlorophenols [o] 1 1 .7
Methyl methacrylate 11.1
Ethylene glycol 1 0.8
Ammonium nitrate (solution) 10.6
Toluene-2,4-diisocyanate 10.5
Di-n-octyl phthalate 10.2
Antimony cmpds 10.1
Butyraldehyde 9.7
Dichlorobenzene (mixed isomers) 9.7
N-nitrosodiphenylamine 9.1
Dimethyl phthalate 8.9
Hexachloroethane 8.9
Diethyl phthalate 8.7
Aluminum (fume or dust) 8.4
Acetonitrile 7.9
Freon113 7.6
Acetone 7.5
Senzoyl chloride 7.2
Allyl chloride 7.1
Ethylene 6.9
3utyl acrylate 5.6
Wethanol 5.1
^monium sulfate (solution) 5.1
sobutyraldehyde 4.9
3arbonyl sulfide 4.4
Dropylene 3.7
1 ,2-Dichloropropane 3.1
ert-butyl alcohol 2.2
Fhorium dioxide 2
Slyphosate 1 .4
rtethyl tert-butyl ether 0.6
.ec-butyl alcohol 0.4
ilized) Chemical Rank
efault=5 default=0 default-5
13 107 115
12.9 108 116
12.6 109 117
14.8 110 106
12.5 111 118
11.9 112 120
11.9 113 121
11.6 114 122
11.2 115 123
13.9 116 109
13.2 117 113
10.8 118 124
10.7 119 125
10.6 120 126
10.5 121 127
10.4 122 128
10.3 123 129
10.2 124 131
10.1 125 132
10 126 133
10 127 134
9.5 128 13'5
9.2 129 136
23.4 130 82
9 131 137
17.8 132 98
16 133 102
8.3 134 139
8.3 135 138
10.3 136 130
7.6 137 140
7.6 138 141
7.4 139 142
45.2 140 33
6.7 141 143
6.5 142 144
6.4 143 145
6.1 144 146
6.1 145 147
5.9 146 148
4.8 147 149
4.4 148 150
14.7 149 107
4.1 150 151
3.8 151 152
3.2 152 154
2.7 153 155
1.9 154 156
12.4 155 119
3.5 156 153
0.5 157 157
0.3 158 158
D-3
-------
TOTAL HAZARD VALUES AND CHEMICAL RANKS. WEIGHTED BY RELEASES
Hazard Value (normalized)
'other specific 1 'other specific
effects' used (effects' excl.
Chemical Rank
'other specific
effects' used
'other specific
effects' excl.
Chromium cmpds
Arsenic cmpds
Lead cmpds
Copper cmpds
Torbufos (tBuSCH2SP(=S)(OEt)2
2,4-D
Nickel cmpds
Formaldehyde
1,3-dIchloropropene
Trifluralln
Cadmium cmpds
Ammonia
Sulfuric acid
Hydrogen fluoride
Nitric acid
Hydrochloric acid
Styrene
Chlorpyrifos
Hydrogen cyanide
Tetrachloroethylene
Trichloroethylene
Chlorine
Manganese cmpds
Chlorothalonil
Di(2-athylhexyl) phthalate
Hexachlorobenzene
Naphthalene
Phosphoric acid
Cobalt cmpds
Phenol
Barium cmpds
Polychlorlnated biphenyls
Benzene
Captan
Acrylamide
Alachlor
Chloroform
Biphenyl
Aorylonltrile
1,2,4-Trichlorobenzene
1,2-DIchloroethane
Zinc compounds
Xyiene (mixed isomers)
Atrazine
1,3-butadiene
Decabromodiphenyl oxide
2.4-DInitrotoluene
1.1.1-Trichloroethane
Methyl Parathion
Metam Sodium (MeNHCS2Na)
Phosphorus (yellow or white)
Malathion
Ethylane oxide
100
98.7
95.3
86.7
85.3
84.6
84.4
83.7
77.5
76.1
74.9
72.5
72.1
67.2
64.5
63.9
62.2
60.3
58.4
58.3
56.1
55.6
54.1
53.9
52.7
49.9
48.4
48.3
48.2
47.1
46.6
46
44.6
44.3
43.9
43.6
43.2
43.2
42.5
42.3
42.1
41.7
41.3
41.1
39.4
38.4
38.2
38.1
36.9
36.9
36.7
35.7
34.9
100
91
77.9
80.5
90.6
71.4
74.3
73.8
78.2
66.5
69.5
71.6
69.8
53
68.5
62.2
52.5
64
57.6
40.8
37.5
54.4
38.9
48.3
42.3
38.2
44.5
51.3
48.2
42.3
39.7
40.6
35.2
34.2
34.9
46.3
33.1
39.7
38.4
37.9
30.9
44.3
27.4
43.7
26.2
34.4
32.6
23.5
39.2
39.1
35.4
37.9
23.7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
1
2
6
4
3
10
7
8
5
14
12
9
11
19
13
16
20
15
17
30
41
18
36
22
29
38
25
21
23
28
33
31
44
48
45
24
51
32
37
39
55
26
61
27
68
46
52
73
34
35
43
40
72
D-4
-------
TOTAL HAZARD VALUES AND CHEMICAL RANKS. WEIGHTED BY RELEASES
chemical
Hazard Value (normalized)
'other specific
effects' used
'other specific
effects' excl.
Chemical Rank
'other specific
effects' used
'other specific
effects' excl.
Cyanazine
Carbaryl
Dibutyl phthalate
2,4-Dinitrophenol
Carbon disulfide
2-nitropropane
EPTC(ethyldipropylthiocarbamate)
Cyclohexane
Cresol (mixed isomers)
Metolachlor
Dichloromethane
Ethylbenzene
Epichlorohydrin
Toluene
Vinyl chloride
Acetaldehyde
Acrylic acid
Diaminotoluene (mixed isomers){24]
p-Cresol
Aniline
Anthracene
Maneb
Carbon tetrachloride
Chloroprene
Butylate
Picric acid
Molybdenum trioxide
Chlorobenzene
Cumene
1,2,4-Trimethylbenzene
Asbestos (friable)
1,2-Dichlorobenzene
4-nitrophenol
Propylene oxide
Zinc (fume or dust)
Methyl ethyl ketone
Isopropyl alcohol
Methylenebis(phenylisocyanate)
4.4'-lsopropylidenediphenol
Bis(2-ethylhexyl) adipate
1,4-Dichlorobenzene
0-Xylene
Catechol
4,4'-Methylenedianiline
Glycol ethers
Terephthalic acid
1.1,2-Trichloroethane
Cumene hydroperoxide
Butyl benzyl phthalate
Hexachloro-1,3-butadiene
Chlorine dioxide
Chloromethane
2-methoxyethanol
34.6
33.3
33.1
33
32.7
32.6
32.3
32.1
31.8
31.7
31.1
30.7
30.5
30.4
29.8
29.7
28.9
28.6
28.4
27.9
27.3
26.9
26.7
26.5
25.4
25.2
25
24.8
24.4
24.3
23.9
23.3
23
22.3
22.3
22.1
21.8
21.2
20.8
19.8
19.2
19.1
19.1
18.6
17.7
17,7
17.6
17
16.7
16.1
16.1
15.9
15.4
36.7
25.3
27.5
28.5
11.1
28.2
34.3
34.1
31
33.6
27.9
21.1
25
21
22.8
,31.5
26.9
26.3
30.1
26.5
26.7
20.2
18.5 .
17.5
26.9
26.8
20.8
20
'17.4
25.8
21.1
22.6
19.7
14.7
7.1
4.4
16.8
22.5
22
19
17.7
12
20.3
18.3
18.8
15.3
16.4
18
15.1
15
12.9
9
7.9
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
42
70
60
57
113
58
47
49
54
50
59
79
71
•'so
74
53
63
67
56
66
65
83
88
92
62
64
81
84
93
69
78
75
85
100
133
142
94
76
77
86
91
109
82
89
87
96
95
90
97
98
104
125
131
D-5
-------
TOTAL HAZARD VALUES AND CHEMICAL RANKS. WEIGHTED BY RELEASES
i • ~ lilj/ju-IValuo (norrmli'frfl | Chemical Rank
chemical Hazard value (norm
I 'other specific
(effects used
. — I
m 9
Vinyl acetate 10^
p-Xylene ™
m-Xylene 14'H
Hydroquinone 14-
N.N-DImethylaniline 14-6
Methyl isobutyl ketone 1 4
Chloroethane 13-
HOC
n-butyl alcohol 13-b
Bromomethane 1
1 ? Q
Metribuzln
198
Nitrobenzene '*-°
Phthallc anhydride 12-6
1 0 R
Pyridlne 1'£lb
Vinylldene chloride 12-4
H O *3
Malelc anhydride ^^
2-ethoxyethanol 12'2
•other specific 'other specific 'other specific
0«^t<=' ovr.l effects' used effects' excl.
12.9 107
9.3 108
10.5 109
13.4 110
1 1 .5 111
6.8 112
14.7 113
10.1 114
9.7 115
13.7 116
9.9 117
8.5 118
11.2 119
8.2 120
11.5 121
5.3 122
ma 123
Proplonaldehyde ^2-^ ^~'~ 124
Titanium tetrachloride 11-9 12 5 125
1.4-Dloxane ^'® 12'g 126
Diethanolamlne __ i • 124 127
Chlorophenols [o] "•' 3'2 128
Methyl methacrylate 11-1 ^ 129
Ethylene glycol 1°-8 ' 130
Ammonium nitrate (solution) 1U-D ' 131
Toluene-2.4-d!isocyanate 10-° R'n. 132
Dl-n-octyl phthalate 10-"
Antimony cmpds 10>
1 • 3 133
-, ma 134
Butyraldehyde ^' '"g 135
Dichlorobenzene (mixed isomers) a-' 136
N-nitrosodiphenylamine 9-1 ' 137
Dimethyl phthalate 8-^ 74 138
Hexachloroethane • ' 139
Dlethyl phthalate BJ Qg 140
Aluminum (fume or dust) 8-4 '
"7 Q 1 i *t i
Acetc ..itrile '•* 33 142
Freon113 _ 2g 143
Acetone _ 58 144
Benzoyl chloride '-^ ^^ 145
Allyl chloride 7
Q 3.1 146
Ethylene '" 6 147
Butyl acrylate ^ Q 148
Methanol
K A 149
Ammonium sulfate (solution) 5-1 ^ 1SO
Isobutyraldehyde • 'Q 151
Carbonyl sulfide 4- 3g 152
Propylene • 3'3 153
1.2-Dichloropropane ^ o 3 154
tert-butyl alcohol '
Thorium dioxide
"2 2.1 155
14 1.5 156
Glyphosate '" Q7 157
Methyl tert-butyl ether »•» nA 158
103
122
115
102
111
135
99
118
120
101
119
128
112
130
110
140
114
105
106
107
108
147
134
138
117
129
149
116
126
124
121
132
123
127
154
14,6
150
137
141
148
136
158
139
144
157
143
145
151
152
153
155
158
.o^.K.Hyl alnohol iLS ~
D-6
-------
CO
LU
CO
UJ
CC
Q
/EIGHTE
:>
i— _,
O
z
co"
\f
^^
<
DC -
<
g
s
LU
X
O
Q
2
CO
LU
ID
$
Q
or
X
_l
<
1-
o
1-
c
(i
"5
c
'E
a
6
0)
-I
(0
3
B
a
X
« 1 '
x §
c
(0
X E n
fl>
S "1 "i
o « 'i
Q o 5
> '55
LU §
2 ^
<8 C '
I o) 3
c
(Q d
•II
X i: §
o>
1|I
O (5 «^.<
D-7
-------
T-eoeowp'j-ogifcfc.Jfrg
§£
j=l
«= > -
CBgP>
I ft
lit
I o> 5
«cOCM«
3 3 3
oo N o> o
-------
£_ »8S6R858CRR85SRS885R88R8K68R8668
•S c £'
111
88fcB8R?:KRRRRfcRR85SBS8Sfc3SS5SSSB
S
i*8S88i5sS3 22558^32253
X x: l&8'ae>l-!5*l8wag888"fc
t§ §"i
o
Acrylonitrile
m
p-Cresol
Carbon tetrachloridi
o
I
*s
4,4'-lsopropylidene<
Chlorophenols [o]
Chloroform
Benzene
a.
c S
i 1 -8 1 I « ||l
« o g | 7 § | £ g
5 o =§ § S s « .§• f -o i
ii€?2-«*5-i«E^S
III!
T- o-XOm<£oooza:5
^
Manganese compou
Bromomethane
Ethylbenzene
as
CM
0 „.
•£ §
Metam Sodium (MeN
1,2,4-Trimethylbsnze
c
.
1
A
c/)
'£1
a>
c
0)
">, fl>
£ .S
2 'c
5 <
F
0>
.«
"S
X
|§
O)
c
0)
£
Acetaldehyde
Alachlor
Catechol
Propionaldehyde
D-9
-------
g8S5g§SS£8SSo?
co en o> (O i-_ t~~, o> •* .
D-10
-------
CO
LU
3
a
tu
IT
m
Q
o
2
i
o
Q
2
CO
UJ
AR
i
c $
-------
-------
-------
m
-o
V«> "O
o o m c
I
-<.
O c
•* CO
"O5"
1!.
<
.
-
I MI f
C«" ^« «^i ^*
I?
If
33 >
CD (O
CO CD
CD 3
a> ,9
o
cT
so
o'
o.=s 5* e-3
l O *^ » S S^
Ij
"
3
gj W
So
s-a
S 5"
Sx
3-m
S
-o
3
s
Ol "0
------- |