&EPA
United States
Environmental Protection
Agency
Office of Health and Ecological
Effects
Washington DC 20460
EPA-600/5-78-012
June 1978
Research and Development
Systems for
Rapid Ranking
of Environmental
Pollutants
Selection of
Subjects for
Scientific and
Technical
Assessment Reports
-------
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the SOCIOECONOMIC ENVIRONMENTAL
STUDIES series. This series includes research on environmental management,
economic analysis, ecological impacts, comprehensive planning and fore-
casting, and analysis methodologies. Included are tools for determining varying
impacts of alternative policies; analyses of environmental planning techniques
at the regional, state, and local levels; and approaches to measuring environ-
mental quality perceptions, as well as analysis of ecological and economic im-
pacts of environmental protection measures. Such topics as urban form, industrial
mix, growth policies, control, and organizational structure are discussed in terms
of optimal environmental performance. These interdisciplinary studies and sys-
tems analyses are presented in forms varying from quantitative relational analyses
to management and policy-oriented reports.
I his document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
June, 1978
EPA-600/5-78-013
SYSTEMS FOR RAPID RANKING
OF ENVIRONMENTAL POLLUTANTS
EPA Contract No. 68-01-2940, Tasks 015, 023
Technical Monitor
Alan P. Carl in
Office of Health and Ecological Effects
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
DISCLAIMER
This report has been reviewed by the Office of Health and Ecological
Effects, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use. This report is
available for purchase from the National Technical Information Service,
P.O. Box 1553, Springfield, Virginia 22161. The order number is PB258168,
li
-------
PREFACE
This report was prepared by an interdisciplinary team under the
general guidance of Alan P. Carlin, the EPA technical monitor. The SRI
team consisted of Howard C. Bailey, David C. Bomberger, Stephen L. Brown
(project leader), Kristin M. Clark, Anthony V. Colucci (consultant),
Jerie L. Etherton, Peter C. Hall, Buford R. Holt, David H. Liu, William
R. Mabey, Kirtland E. McCaleb, David R. Myers, Thomas 0. Peyton, Dennis
E. Schendel, Lyle M. Schump, Robert V. Steele, Steven H. Traver, and
Rose M. Wright.
WARNING: THE DATA REPORTED IN THIS DOCUMENT WITH RESPECT TO VARIOUS
POTENTIALLY HAZARDOUS ENVIRONMENTAL AGENTS SHOULD BE USED ELSEWHERE ONLY
WITH THE UTMOST CAUTION. THESE DATA WERE GATHERED AS INPUT TO A PRIORITY-
SETTING PROCESS AND ARE THUS INCOMPLETE AND LARGELY UNVERIFIED.
111
-------
CONTENTS
PREFACE iii
LIST OF ILLUSTRATIONS „ ix
LIST OF TABLES ix
PART ONE--OVERVIEW 1
I INTRODUCTION 3
A. Background , 3
B. Objectives 6
C. Method of Approach 6
II SUMMARY 11
III CONCLUSIONS ........... 15
IV RECOMMENDATIONS 19
PART TOO--SYSTEMS FOR RANKING ENVIRONMENTAL POLLUTANTS. 21
V SELECTION OF SYSTEMS FOR DEVELOPMENT 23
A. Ranking Systems 23
B. Assumptions and Criteria 25
C. Systems Considered 27
D. Expert-Based System 27
E. Screening-Indexing System 30
F. Model-Based System 34
G. Selection Rationale 35
VI A PRIORITY RANKING SYSTEM BASED ON THE USE OF
EXPERT GROUPS 37
-------
A. Outline of the System 37
B. Relevant Information and Literature Sources. ... 38
C. Format for Presenting Data 38
D. Composition of the Expert Committee and
Criteria for Their Selection 39
E. Operating Procedures 39
F. Scientific Review Group Operations 40
G. Estimated Operating Cost's 41
VII DEVELOPMENT OF OBJECTIVE SUBSYSTEM 43
A. Selection of Parameters 43
B. Basic System Concepts 44
C. Development of Procedures 45
D. Operating Personnel and Costs 46
PART THREE--TEST OF OBJECTIVE SUBSYSTEM 51
VIII DESCRIPTION OF THE TEST 53
A. Definition of the Agents 53
B. Assignment of Tasks. 54
C. Information Sources 56
D. Summary of the Findings 57
E. Sensitivity Analyses 66
IX EVALUATION OF THE TEST 71
A. Criteria for Evaluation 71
B. Successes and Failures of the System 71
C. Suggested Improvements 75
APPENDICES
A DATA ELEMENTS FOR THE EXPERT SYSTEM 79
B RECOMMENDED FORMAT FOR PRESENTING DATA TO THE
EXPERT COMMITTEE. . 85
C RECOMMENDED CRITERIA FOR SELECTING MEMBERS OF
THE EXPERT COMMITTEE 101
vi
-------
D RECOMMENDED EVALUATION SHEET AND PROPOSED METHOD
OF USE BY MEMBERS OF THE EXPERT COMMITTEE 107
E PARAMETERS FOR USE IN THE OBJECTIVE SUBSYSTEM 113
F PROCEDURES FOR OBJECTIVE RANKING 135
G SAMPLE RANKING RESULTS FOR CARBON DISULFIDE AND
CYANIDES 229
H ABBREVIATIONS 279
I REFERENCES. 285
VII
-------
LIST OF ILLUSTRATIONS
1 Overall Structure of the Objective Subsystem . . 47
F-l Generalized System Flow Component 139
LIST OF TABLES
1-1 Published and Planned STAR Documents. 4
1-2 Typical STAR Outline „ 5
VI-1 Expert-Based System Costs ...... ... 42
VII-1 Personnel for Objective Subsystem 48
VII-2 Objective Subsystem Costs . 0 .............. 49
VIII-1 Nominated Agents for the Test .............. 53
VIII-2 Representative Compounds Used 55
VIII-3 General Information Compendia . ..... 58
VIII-4 Basic Documents ........ 59
VIII-5 Agent Ranks and Reasons .... 61
VIII-6 Valuation Range ...... . 67
VIII-7 Matrix of Value Systems ... . 68
VIII-8 Range of Rankings ............ ... 68
E-l Processes and Uses Leading to Release of Pollutants . . .
to the Environment. . 116
E-2 Data Sources and Units for Release Routes A-N ...... 117
E-3 Release Routes, Data Sources, and Units ......... 119
E-4 Time/Cost Estimates ................... 120
ix
-------
Part One
OVERVIEW
-------
I INTRODUCTION
A. Background
The Office of Research and Development (ORD) of the Environmental
Protection Agency (EPA) is in the early stages of producing a series of
Scientific and Technical Assessment Reports (STARs), each of which will
summarize the state of knowledge about an environmental pollutant. These
reports, which have extended the concepts embodied in earlier criteria
documents for air pollutants and in National Academy of Sciences mono-
graphs on specific pollutants, will be used as input to the regulatory
process.
The distinguishing characteristic of the STARs is that they assess
all information relevant to the EPA regulatory mission, about the be-
havior of the pollutant in all media and with respect to all targets.
A list of published and planned STARs is shown in Table 1-1. In examin-
ing this list, first it should be remembered that many of the early docu-
ments had a different format from that currently in force, and second
that the list is very dynamic; consequently, the picture presented in
Table 1-1 is a snapshot as of early 1976.
An outline for a typical STAR is shown in Table 1-2. As this out-
line indicates, a key feature of the STAR is to provide information that
will enable a decision maker to determine the benefits of a contemplated
EPA action with respect to the pollutant. This implies that if a STAR
is to make much impact the state of knowledge about the pollutant must
be relatively good and the selection of candidates for STARs must be
made with this fact in mind.
-------
Table 1-1
PUBLISHED AND PLANNED STAR DOCUMENTS'
Issued
Particulate polycyclic organic matter
Manganese
Cadmium
Vinyl chloride
Planned
t
Arsenic
Halomethanes
Nickel
Vanadium
As of early 1976.
Partial list.
EPA Report No.
600/6-75-001
600/6-75-002
600/6-75-003
600/6-75-004
Having limited resources, EPA can produce only a few STARs each year.
This constraint makes it exceedingly important to select as candidates
for STARs only those pollutants whose control by EPA would result in the
most significant benefits for the nation's environment.
If it is assumed that the selection is based solely on objective
(non-political) considerations, which may not always be the case, the
best order for STAR production would be to attack first the pollutant
with the highest potential for environmental harm avoidable through EPA
actions. Under this assumption, the ordering of STAR candidates would
correspond to a priority ranking of environmental pollutants needing EPA
regulatory attention.
-------
Table 1-2
TYPICAL STAR OUTLINE
1. Summary and Conclusions
2. Pollutant Characterization
2.1 Chemical and Physical Properties
2.2 Measurement Techniques
3. Environmental Occurrence and Transport
3.1 Concentrations
3.2 Transformation and Transport Mechanisms
4. Environmental Exposure and Undesirable Effects
4.1 Mechanisms of Exposure
4.2 Mechanisms of Response
4.3 Undesirable Effects
4.4 Environmental Exposure
5. Sources and Controllability
5.1 Sources
5.2 Control Technology and Controllability
5.3 Undesirable Intermedia Effects of Principal Control Measures
6. Overview, Benefits, and Institutional Problems of Control
6.1 Economic Benefits from Control
6.2 Societal/Institutional Constraints on Control
6.3 Overview
The process of ordering the candidates for STARs is far from simple
and requires considerable amounts of information about the candidates that
will later appear in the STARs themselves. This process consists, at a
miminum, of the following steps:
-------
• Nomination of candidate agents.
• Analysis of the importance of STARs for these agents.
• Selection of agents from the candidates.
At this point we have introduced the term agent as being somewhat
broader than pollutant. In the remainder of this report, we will use
agent to include chemical pollutants, thermal waste, radiation, noise,
and other entities affecting the environment.
The nomination step can be either passive, allowing any interested
party to nominate candidates, or active, soliciting nominations from ex-
perts and searching for agents of general concern.
The analysis should provide guidance to the decision makers in order-
ing the STARs. Accordingly, it is the systematic portion of the larger
process.
The selection step is ultimately the responsibility of the EPA deci-
sion makers who will use the STARs. They must use all the information
available to them, explicit and implicit, to decide upon the best order.
B. Obj ectives
The objectives of the study reported in this document were to examine
alternative systems for the ordering or priority ranking of agents for
STARs, to recommend a preferred system to develop this system for possible
implementation by EPA, to test a portion of the system developed on a
selected set of agents, and to correct system deficiencies identified in
the test. Thus, in effect, the overall objective was the development of
a system for rapidly ranking environmental pollutants.
C. Method of Study
In pursuit of these objectives, SRI assembled an interdisciplinary
team of scientists and analysts, including two chemical economists, a
6
-------
chemical engineer, two physical chemists, two environmental health sci-
entists, an ecologist, two toxicologists, a mathematician, two literature
specialists, and an environmental systems analyst.
The team met with the EPA technical monitor to confirm the direction
of the study, and then conducted a survey of various systems for ranking
hazards. Three study groups were formed to investigate three different
possible systems for the STAR ranking problem; each group included per-
sonnel familiar with releases of agents to the environment, fate of agents
in the environments, and effects on receptors.
The three systems, based respectively on the judgment of experts,
information screening and ordering, and mathematical models, were devel-
oped and presented to EPA/ORD in outline format highlighting their ad-
vantages and disadvantages. EPA and SRI then selected a hybrid system
for further development. This hybrid system is largely dependent upon
the systematized judgment of experts who are supported and balanced by
a more objective subsystem based on screening and modeling. This hybrid
system was then further developed by two study groups, one concentrating
on the expert system and the other on the objective subsystem.
After a careful review of the resulting recommended procedures
EPA/ORD determined that a test of the recommended system would both mea-
sure its utility and identify desirable modifications. At that time,
ORD was attempting to recommend candidates for STARs, and a rapid ranking
was seen as useful. Because establishment of the expert committees nec-
essary for the full development of the system would be time consuming,
it was decided to test only the objective subsystem on the following ten
agents:
-------
Antimony Cyanides
Beryllium Heat from manmade sources
Carbon disulfide Lithium
Carbonyl sulfide* Molybdenum
Cobalt Plutonium
For those agents that included more than one specific chemical com-
pound of concern, the test of the subsystem included at least two com-
pounds believed to be of the most concern.
The method of approach for the test was dictated largely by the
nature of the objective subsystem. That is, once the candidates were
nominated, the procedures outlined in the subsystem were followed as
faithfully as possible, from definition of the agent to ultimate compu-
tation of an environmental hazard index and agent ranking. During this
operation, records were kept of the steps of the procedure that were
undertaken, and notes were taken on difficulties encountered and solu-
tions achieved.
The principal activities of the test were as follows:
• Data Collection. Basic documents on each of the agents were
collected as available. These included abbreviated summaries
from data compilations (for example, The Toxic Substances
List), draft criteria documents (for example, WHO preliminary
review on molybdenum), EPA publications (for example, Plu-
tonium: Statement of the Problem), and other readily avail-
able documents. Only limited use was made of bibliographic
search techniques (for example, some TOXLINE searches were
made).
These two candidates were of special concern for current standard-setting
activities. A separate, more comprehensive, report was prepared on these
entitled, "Carbon Disulfide, Carbonyl Sulfide: Literature Review and
Environmental Assessment" (Stanford Research Institute, July 1975) (Draft)
8
-------
Subsystem Operation. The objective subsystem was followed
step-by-step as closely as possible. Where difficulties
arose, they were solved on an ad hoc basis.
Sensitivity Analysis. The principal organized sensitivity
analysis was of variations in the assignment of relative
values to the effects of the agents. Also, certain steps
were repeated by different operators to test the degree of
subjectivity inherent; in addition, the entire ranking was
reviewed during the report preparation phase, and numerous
minor and several major changes were made.
Appraisal of Test. The results of the test and the methods
of achieving them were reviewed by the project team to deter-
mine areas of difficulty, recommendations for improvement,
successes and failures, and so on.
-------
II SUMMARY
This report concerns the development and testing of a systematic
procedure (system) for priority ranking environmental agents as candidates
for Scientific and Technical Assessment Reports (STARs) and thus for pri-
ority ranking environmental pollutants needing EPA regulatory attention.
The development of this system is part of a larger process that will in-
clude the nomination of candidate agents, and the final selection by EPA
decision makers of agents for STARs.
In the first phase, a number of systematic procedures previously de-
veloped for related ranking purposes were surveyed. Selected elements
were incorporated into three different system outlines. A system based
on expert judgment was seen as being technically and economically feasible
as well as acceptable to EPA decision makers, if the expertise represented
were sufficiently high. An information screening system with ranking on a
combination of several indexes was seen as simple and economical. A sys-
tem based on a mathematical model was seen as being more objective, ex-
plicit, and reproducible. A combination of the expert system and a more
objective screening/model subsystem was selected as having the highest
potential for further development.
In the second phase, the expert system was developed to provide for
the selection of an Expert Committee by EPA with support from a contractor;
the compilation of data by the contractor for presentation to the experts;
the priority ranking of agents in four categories, by the experts; the
summarization of the results by the contractor for consideration by a
Scientific Review Group composed of independent, recognized members of
the scientific community; and final ranking by the Scientific Review
11
-------
Group. It was estimated that the system could be operated for about
$3,500 per agent or $120-130,000 annually for about 36 candidate agents.
The objective subsystem is designed to support and calibrate the
expert system, and consists of an explicit procedure that tests the
state of information about an agent and processes available information
on the basis of the outcome of the tests. The information base for the
objective subsystem is a subset of the information for the expert system,
with a few exceptions. The subsystem has several important subjective
elements, including the choice of processes to represent in the model,
the values to assign to the predicted effects of the agent, and the op-
tion to use ad hoc studies. It was estimated that the subsystem would
cost $50-60,000 per year to operate, in addition to the cost of the ex-
pert system.
In the third phase, the objective subsystem was tested on a sample
of 10 agents to determine the weaknesses of the system and to confirm or
deny the operating cost estimates. Specific agents representative of the
generalized agents on the list of candidates were identified. Antimony
was represented by antimony trioxide, beryllium by beryllium metal and
beryllium oxide, cobalt by cobaltous chloride and cobaltous naphthenate,
cyanides by hydrogen, sodium, and potassium cyanides, lithium by lithium
carbonate and lithium chloride, molybdenum by molybdenum oxide and molyb-
239
denum sulfide, and plutonium by plutonium (oxide). Carbon disulfide,
carbonyl sulfide, and heat were single agents.
Information sources for the priority ranking procedure included
both general information compendia covering such aspects as agent proper-
ties or toxicology, and basic documents on individual agents, such as
criteria documents or EPA reports.
The ranking was successfully completed, at a cost of about $1,500
per agent, with the following results:
12
-------
Rank Agent Principal Effect
1
2
3
4
5
6
7
Cyanides
Carbon disulfide
Beryllium
Lithium
Plutonium
Antimony
Heat
Accidental acute
Odor (aesthetic
toxicity in man
annoyance)
Lung cancer in man
Central nervous
Life- shortening
Heart disease in
Fish mortality
system disturbance
in man
man
8 Carbonyl sulfide Heart disease in man
9 Cobalt Toxicity in fish
10 Molybdenum Molybdenosis in cattle
An alternative ranking, that eliminated a critical assumption on the
distribution of higher-than-threshold doses, resulted in the following
ranked order: carbon disulfide, beryllium, plutonium and cyanides (tied),
and heat; all other agents tied with no effects.
A sensitivity analysis showed that the subsystem was not markedly
sensitive to assumptions about the relative values of various effects.
It was also fairly insensitive to other subjective inputs, such as the
choices of sources of release, with the exceptions of the assumption con-
cerning high dose distribution and the choice of what effects should be
considered. As an example of the system's sensitivity to effects, if
odor problems with carbon disulfide had not been considered, it would
have been ranked in seventh position.
The principal difficulties encountered in the test were in the col-
lection and use of information, and a few procedural difficulties. Data
were scanty on release factors to the environment, persistence and inter-
media transfer, transport and diffusion, and populations at risk. Inter-
pretation of toxicological information was also difficult. An initial
problem was overcoming operator unfamiliarity with the system. This prob-
lem extended into difficulties in dealing with necessary subjective judg-
ments not forced by the system. Potential improvements to the system were
13
-------
identified as a result of the test, some of which have already been in-
corporated in the system and are discussed in this report.
It was concluded, in general, that the subsystem is workable and
useful, and can be operated with modest resources. Its principal use is
in making assumptions and relationships explicit, identifying factors
limiting environmental hazards, and spotlighting areas of critical un-
certainty. However, the system would benefit from longer lead times and
more access to agent nominators. The reliance on subjective inputs is
greater than desirable, and the reproducibility and credibility of the
system are consequently degraded.
It is recommended that the subsystem be used only as an input to a
more comprehensive process such as the expert process described above.
If the system is used in this way, it should be improved modestly but
continuously. At least 3 months lead time is recommended, and operation
under the supervision of a competent and confident environmental gener-
alist is suggested.
14
-------
Ill CONCLUSIONS
In the first phase of this study we reached the conclusion that it
would be technically and economically feasible to operate a STAR ranking
system based on the use of an Expert Committee, and that the output of
such a system would be readily acceptable to decision makers if the de-
gree of expertise were sufficiently high. However, we also concluded
that the expert system would be much more effective if supported by a
more objective subsystem that processed some of the information for pre-
sentation to the experts, and further served as a calibration for the
judgments of the experts.
In the second phase, we concluded that the expert system should
consist of the following major elements:
• EPA, with contractor support, would select an Expert Com-
mittee on the basis of specified criteria.
• A contractor would make a systematic compilation of data
on about 10 candidates at a time, for submission to the
Expert Committee.
• The experts would assign priority ratings in three sub-
ject categories and a fourth overall category.
• A contractor would summarize results for submission to an
independent Scientific Review Group.
• The Scientific Review Group would decide on a final ranking.
The cost of ranking about 36 candidate agents per year was
estimated at $120-130,000.
The objective subsystem should be operated in parallel by the EPA
contractor and should use a subset of the data collected for the expert
system. (Selective augmentation of data in critical areas may be nec-
essary.) The subsystem contains explicit instructions for obtaining
and processing data, as well as decisions on the depth of the analysis
15
-------
needed. (The latter depends on the results of the state of information
tests.) However, the objective subsystem also contains important sub-
jective elements,, which include the processes chosen to be modeled, the
values assigned to various predicted effects of the agent, and the ne-
cessity for ad h-oc studies when critical information is not easily avail-
able. We estimate that the subsystem could be operated for about $1,600
per agent or $50-60,000 per year. The overall system could thus be oper-
ated for about $170-190,000 per year.
In the third (test) phase, we concluded that the subsystem is work-
able and useful, with important caveats. No unresolvable difficulty was
encountered in the system operation, and the desired ranking was accom-
plished. The chief use of the subsystem is in making explicit the assump-
tions and information about environmental hazard potential and thus
identifying the limiting factors and areas of uncertainty.
First among the caveats is the observation that the subsystem would
benefit from longer lead times, more direct contact with agent nominators,
and operator familiarity. The principal ranking effort was accomplished
in about 6 weeks, which prevented us from obtaining as many basic docu-
ments on the agents as we would have liked. Direct contact with the
nominators would have not only enhanced this information gathering pro-
cess, but would have directed us more accurately to the principal con-
cerns. A moderate amount of effort was expended in making the operators
familiar with the system.
Second, and in many ways more important, is the fact that many pieces
of information desired for system operation are unavailable, fragmentary,
or difficult to interpret. The major areas in which these deficiencies
were limiting were:
• Release factors to the environment
« Persistence and intermedia transfer
16
-------
• Transport and diffusion
• Populations at risk
• Toxicology interpretation.
This situation required more reliance on default values and ad hoc pro-
cedures than was anticipated.
Third, the reproducibility of the system, in terms of the necessity
for subjective inputs by the operator, is lower than had been hoped.
Consequently, the subsystem is probably of low credibility if examined
closely. However, if alternative methods of priority ranking were simi-
larly examined, the subsystem would compare favorably. Moreover, it was
not found particularly sensitive to variations in the relative valuation
of effects or other uncertainties in the inputs.
Finally, as suggested by the subjectivity observed, the success of
the subsystem undoubtedly depends markedly on the creativity and boldness
of the operators.
17
-------
IV RECOMMENDATIONS
On the basis of the development and testing of the rapid ranking
system for environmental pollutants, we recommend, with qualifications,
that EPA/ORD implement the expert system along with the objective sub-
system for ranking STAR candidates and for other priority-setting pur-
poses.
Foremost among the qualifications is that the complete expert system
has not been tested; therefore, any implementation should be accompanied
by an evaluation and an option for termination after a year's operation.
Provision for modifying the expert system procedures in response to the
evaluation should also be made.
The objective subsystem should be operated only with careful examina-
tion of the assumptions and procedures associated with each agent's rank-
ing. If the objective subsystem is used in conjunction with the expert
process as recommended, this examination should be automatic.
If the subsystem is to be used at all, it should undergo modest and
continuous improvements, spanning at least the first group of improvements
listed in Section IX C.
We further recommend that a candidate agent be introduced into the
system at least 3 months prior to a required decision on its priority
for a STAR, to allow collection of background documents and thoughtful
assessment of the data. The ranking should be conducted under the super-
vision of a confident and competent environmental generalist.
19
-------
Part Two
SYSTEMS FOR RANKING
ENVIRONMENTAL POLLUTANTS
21
-------
V SELECTION OF SYSTEMS FOR DEVELOPMENT
The first phase of research leading to this report consisted of an
examination of a variety of ranking systems that had purposes similar to
the priority ranking of agents for STARs; the development in outline
format of three representative systems, showing the advantages and dis-
advantages of each to be presented to EPA/ORD; and the selection of a
hybrid system for further development,
A. Ranking Systems
The purpose of any ranking system related to an action program is
to enable decision makers to do the most important things first. In the
case of the STARs, EPA desires to summarize first the scientific and
technical information on those environmental pollutants (agents) that
have the highest potential for harm, so that EPA's regulatory response
can achieve the greatest gains as early as possible.
The STAR priority ranking process is composed of three major steps:
• Nomination of candidate agents
• Systematic ranking of candidates
• Selection, using both the systematic ranking and factors
beyond its scope, of the agents for which STARs will be
prepared.
This report is concerned principally with the second of these steps.
However, the importance of the third step is emphasized by the following
caveats about systematic ranking procedures. First, systematic approaches
tend to be mistrusted because they sometimes give results that are not
intuitively evident to the decision maker, and because these results are
23
-------
sometimes wrong. Second, if completely accurate information were avail-
able on all agents to be ranked, the need for a systematic approach would
be minimal. Hence it is axiomatic that the need to rank implies consid-
erable uncertainty about the information used in the system. Third, this
uncertainty implies that the resulting ranking will be imperfect, with
some agents of little actual importance high on the list, and conversely,
with some important agents low on the list. Finally, every systematic
approach has some unavoidable subjective inputs, whether explicit or
implicit, and the system can be attacked on these subjective components.
Once a ranking system is accepted as a part of the overall process,
however, some fundamental ranking concepts become important. Basically,
the output of a ranking system is a list ranked according to some prior-
ity. To use such a list, however, one must make selections from it, for
example, one might take the first M item from a list of N. Obviously,
if one takes M=N (the whole list), it doesn't matter how the list is
ordered. In general, as the ratio of M to N decreases, it becomes more
and more important to the selection process to have the list ordered
correctly.
Ranking can be accomplished ordinally or by an index. In ordinal
ranking, paired comparisons are made: Is this item more or less impor-
tant than this other item? The resulting list is ordered correctly, but
no feeling for the relative spacing between adjoining items is generated.
This problem is solved by indexed ranking, in which each item (agent) is
assigned a quantitative index, and the ranking is achieved by sorting on
this index. With such a system, one can see whether item three is twice
as important as item one or only 107° more important. All of the systems
considered here are based on indexes, although the ways in which they
are derived differ considerably.
If the system is to achieve the purposes for which it was designed
the index must correspond well to the actual importance of the agents to
24
-------
the decision maker. That is, the index must represent a unifying value
system. For STARs, we attempted to devise a value system in which the
important variable was the degree to which EPA actions could improve
human health and welfare through control of agents in the environment.
Suggestions as to how this might be defined were taken from two National
•k f
Academy of Sciences (NAS) reports as well as from previous studies by
Battelle,* SURC,§ and SRI.**
Two very important components of a ranking system must also be rec-
ognized. First, the system must specify the methods and sources for
obtaining information, and second, it must define the methods for process-
ing and using this information. Both components must be present for suc-
cess; all too many systems have failed by concentrating on processing
methods and ignoring the specification of sources.
B. Assumptions and Criteria
In developing the outlines for the three candidate systems and in
choosing among them, we made several assumptions about STARs and the
criteria on which the choice should be made.
"Principles for Evaluating Chemicals in the Environment," National
Academy of Sciences, (1975). (NAS 1975b)
"Assessing Potential Ocean Pollutants," National Academy of Sciences,
(1975). (NAS 1975a)
t
"Identification Systems for Selecting Chemical Classes as Candidates
for Evaluation," EPA-560/1-74-001, Battelle Memorial Institute,
(November 1974). (BMI 1974)
§
"Establishing Environmental Priorities for Synthetic Organic Chemicals:
Focusing on the Next PCB's," Paper presented by P. H. Howard, Syracuse
University Research Corporation at Seminar on Early Warning Systems for
Toxic Substances, (February 1974). (Howard 1974)
^L.
"Research Program on Hazard Priority Ranking of Manufactured Chemicals,"
Stanford Research Institute (April 1975). (SRI 1975)
25
-------
Some of the following assumptions may seem trivial; however, they
are important to our design:
• STARs will be written and used for regulatory purposes.
The order in which STARs should be written should be a
direct reflection of the importance of the potential
regulatory actions that could be taken.
• Importance is defined in terms of beneficial effects on
human health and welfare and ecology as determined by
environmental quality.
• The feasibility of control will be addressed partly in
the final selection process and partly in the STAR
preparation process.
• Certain actions require a high state of knowledge about
the agent.
• These state-of-knowledge issues will be resolved outside
the systematic part of the priority ranking process.
• The universe of agents nominated for ranking will be small
but growing. We assume no more than 36 nominations per year.
• The annual rate of STAR production will be between 6 and 24.
In consultation with the EPA technical monitor, we agreed that the
following criteria were valuable in selecting among the proposed systems:
• Technical feasibility
• Economic feasibility
« Acceptability to decision makers
• Robustness with respect to uncertain information
• Simplicity and understandability.
The following criteria are somewhat less important than those above:
• Credibility to various interest groups
« Relative objectivity
• Relative explicitness
• Reproducibility and traceability.
26
-------
C. Systems Considered
Four conceptual types of systems were considered as possibilities
for the STAR ranking.
An expert-based system would use the knowledge of recognized experts
to choose among the nominated agents. Although the expert opinion might
be based on objective information, the processing of that information
would be largely subjective and implicit.
A screening-based system would consist of a series of questions to
be answered about the agent. Depending on the answers to the questions,
the candidate agents would be sorted into various groups, and the groups
ranked by subjective means. The sorting would be done on the basis of
objective data.
An index-based system would assign several indices to the agent,
each based on objective information about the agent. These individual
indices would be combined by subjective rules and weighting factors.
A model-based system would attempt to construct a mathematical model
of the processes that cause an agent to be hazardous to human health and
welfare and/or to ecosystems. The subjective elements of a model-based
system would include the processes to be emphasized and the values to be
placed on various predicted effects.
It was found that no pure system satisfied the selection criteria
very well. Several hybrids were examined, and the final three candidates
emphasized experts, screening-indexing, and models, respectively.
D. Expert-Based System
The expert-based priority ranking system was based on an Expert
Committee, an EPA contractor charged with providing input to the Expert
Committee and systematizing its output for presentation, and a Scientific
27
-------
Review Group, such as a suitable National Academy of Sciences committee,
which would do the actual priority ranking.
Under this system the EPA contractor would start by gathering avail-
able relevant information from a selected group of sources on the chem-
icals of interest and putting this information in standarized form for
consideration by a committee of experts selected on the basis of a pre-
established set of criteria. Each expert would be asked to give his
individual estimate of the severity of the potential environmental prob-
lems associated with each chemical. These estimates would be done on a
numerical scale (accompanied by explanatory supporting text) for certain
categories of information and on an overall basis for the chemical. They
would be combined by an EPA contractor into a composite estimate and ac-
companying text for submission to the Scientific Review Group which would
be asked to review the estimates and to make recommendations concerning
the priority of each chemical in the preparation of STAR documents.
The data to be gathered should include information such as that in-
cluded in the UN-sponsored International Register of Potentially Toxic
Chemicals (IRPTC), the European Economic Committees-sponsored Environ-
mental Chemical Data and Information Network (ECDIN), and the United
Kingdom Network of Data on Environmentally Significant Chemicals (DESCNET).
In view of the large number of potential sources of information,
it is imperative that the sources tapped be restricted to those most
likely to provide useful data without incurring a major expense for liter-
ature searching. The bulk of the needed information could probably be
obtained from the sources that were found most useful in SRI's recent
(1975) NSF project (see p. 13). In addition, useful data are available
in the NIOSH Toxic Substances List, the NLM Toxicology Data Bank (TOXLINE
•k
Other types of agents would be treated on an ad hoc basis.
28
-------
CANCERLINE, and so on) and the EPA Oil and Hazardous Material Technical
Assistance Data System (TADS).
The relevant data would be provided to the experts in a form similar
to that used in the NSF project and they would all be asked to review all
of the supplied data in the three areas of product release, environmental
transport, and toxic effects, and then provide their estimates of the
potential hazard represented by the chemical. This would be done by
selecting numbers from a scale such as the following:
Potential Hazard Scale
Very
None Little Moderate Major
0 123 4567 89 10
Each expert would be asked to supply an estimate for each of the
three categories (product release, environmental transport, and toxic
effects) and an overall estimate for the chemical. In addition, he would
be asked to provide a brief description of the major factors behind each
of his four numerical estimates. The contractor would consolidate the
separate estimates into a composite estimate for each chemical and pre-
pare descriptions of the major factors behind the composite estimates.
In the course of doing this, the contractor would go back to the individ-
ual experts to clear up any problems associated with their estimates. If
considered desirable, the contractor could point out additional informa-
tion to experts whose estimates represented extremes and permit them to
change their estimates. If funding and time constraints permitted, this
process might even be expanded into a formal Delphi technique.
The positive and negative features of the proposed expert-based
system are presented in the following table.
29
-------
EXPERT-BASED SYSTEM
Positive Features
Technically feasible
Relatively inexpensive
Relatively simple and
systematic with built-in
checks
Uses recognized experts (an
aid in gaining acceptability
to decision makers)
Negative Features
Highly dependent on capabili-
ties of the group of experts
Necessarily somewhat subjective
Credibility to various interest
groups will depend on experts
used
Consistency over time may be
difficult
E. Screening-Indexing System
The preliminary screening-indexing system entailed the computation
of indices for release rate, toxicity, and exposure, and the subsequent
aggregation of these into a single index for ranking. The perceived
advantages of the system were flexibility, simplicity, ease of execution,
and explicit statement of assumptions. The principal disadvantages were
the subjectivity involved in the selection of weights to be assigned to
the components of the toxicity and exposure indices and the equally sub-
jective weighting of interactions between the components of these two
indices,
The sections which follow briefly discuss the trade-offs perceived
in development of the three indices considered and summarize the recom-
mended strategy.
30
-------
1. Release Rate Index
Four classes of materials release were -recognized, the sum of
which equals the total annual release to the environment. These were:
* Emissions and wastes resulting from manufacturing
operations, including clean-up, disposal of off-
grade batches, and spills.
• Losses during transportation from producer to point
of use, including spills, evaporation, and clean-up
of shipping containers.
• Dispersive uses (uses in which the chemical or agent
is not changed).
• Unintentional production and subsequent loss by
combustion, use, or manufacture of other materials.
Quantification of these losses is fairly easy, but the question
of whether or not the raw release rates should be transformed by loga-
rithm into an index, to reduce the contribution made by release rates to
the ultimate ranking of candidates for STAR documents is not clear. This
question is closely related to subjective weightings of mortality and
various degrees of morbidity: It is not clear, for example, whether a sub-
stance which frequently kills, but is released in small quantities, should
have a higher ranking than a less deadly, but more abundant substance; nor
is it clear who should make such decisions.
2. Human Toxicity Index
Toxicity in its broadest sense entails a number of negative
impacts on individual organisms, including various sources of mortality
and forms of morbidity, such as teratogenicity, mutagenicity, allergenic-
ity, and carcinogenicity. Data are commonly available for lethal dosages
of various a, ;nts, but are slightly less available for indexes of carcino-
genicity, mutagenicity, and teratogenicity. Data concerning allergenicity
are considerably less common. Similar variations in the availability of
3-1
-------
data exists for modes of intake, "oral" data being more common than
"respiratory" data. The less commonly available data bias the ranking
toward the better known agents, but the bias could be appraised, if nec-
essary, by duplicate rankings—one with and one without data other than
lethality. Inclusion of all available data requires that the various
measures of toxicity be assigned index values to allow for aggregation
of data for respiratory and oral ingestion. The method of indexing is
unimportant as long as it is coordinated with the release rate index.
That is, the ranking of toxicity on a scale of one to ten would make the
contribution of toxicity to the ultimate ranking negligible if raw re-
lease rate data were used, because of the enormous range (>10 ) in the
release rates.
Weighting of biological species remained an unsolved problem;
the best solution seemed to be to weight nonhuman organisms equally, and
to assign man an exceptionally high weight.
3. Environmental Exposure and Damage
Estimates of the transport and accumulation of toxins within
the environment are subject to the greatest ambiguity. Rates of physical,
chemical, and biological degradation are rarely available and generally
are not expressed in forms which permit extrapolation to unstudied en-
vironments. Consequently, subjective judgement is both extensive and
unavoidable with respect to both the data manipulated and the methods of
manipulation, the latter involving questions of weighting of the media
through which man and other organisms are exposed, the severity of the
environmental damage, and the rapidity of repair.
4. Recommended Strategy
The screening-indexing system proposed in the initial explora-
tion of alternative ranking procedures incorporated three fundamental
32
-------
ass ump t i on s:
• Simplicity and ease of execution were of paramount
importance.
• Biases should be in the direction of overestimation
of hazard if unavoidable.
• In regulatory actions, injury to man generally carries
more weight than injury to other organisms and this
should be reflected in the screening methodology.
The proposed methodology entailed logarithmic weighting of re-
lease rates, the use of all available toxicological data, and differen-
tial weighting of media for localized or nonpersistent toxins. For
simplicity, it was recommended that index values be assigned to the most
hazardous modes of exposure for each toxicological response (such as
carcinogenicity) for use in the computation of an aggregate index of tox-
icity. Differential weighting of media (air, land, water) was recommended
on the grounds that the rapidity of immobilization or dilution varies
among media, and that the probability of biological contact consequently
varies. Nonhuman target organisms were assumed to have equal value and
were accordingly weighted inversely by their intrinsic rates of increase,
which are an approximation of the ability of these species populations to
recover from mass mortality. In routine screening, this methodology ig-
nored impacts on the structure and function of assemblages of organisms;
however, provision was made for consideration of these and other impacts
that are difficult to appraise, such as aesthetics and population at risk,
in the event of ties in the final ranking. It was recommended that the
components of each index be aggregated by summation, and that the indices
for release rate, human toxicity, and environmental exposure and damage
be summed to obtain a final ranking.
33
-------
F. Model-Based System
The model-based system also defines an environmental hazard index
for ranking agents. Ideally, this index is related to the totality of
adverse effects on man and his environment that are potentially control-
lable by EPA. However, to achieve this relationship, it is necessary to
subjectively weight various effects. For example, the model system tries
to predict the numbers of human cancers, incidences of aesthetic impacts,
and percentage of fish killed by an agent; the importance of these three
effects are combined by subjective value weights.
The index is computed from a model of the processes that relate the
use and occurrence of environmental agents to their end effects. The
critical issue in developing such a model is in selecting the important
processes and the manner of representing them. For example, a model that
ignored toxicity would be useless, but a model that added the half-life
to the release rate would be equally unacceptable because it does not
represent reality correctly. The selection of processes and representa-
tions is subjective, but is also based on the availability of information
to carry out the model computations. For example, synergism between two
agents, such as between tobacco smoke and asbestos, is clearly important
in some cases, but data is so rarely available that the possibility is
probably not worth inclusion.
The model includes five basic compartments—source, distribution,
fate, effects, and valuation and ranking — related to the ones used in the
expert and screening-indexing systems.
The source comparment compares human production (intentional and
unintentional) and natural production of the agent.
The distribution compartment examines the uses, unintentional re-
leases, and "ultimate" disposal of the agent, and predicts releases to
air, water, and land.
34
-------
The fate compartment traces the transport of the agent in the en-
vironment, accounts for transformations or losses of the agent to inac-
cessible reservoirs, predicts concentrations in media, and relates this
information to the exposure of humans, non-human organisms, and inanimate
objects.
The effects compartment develops dose-effects relationships, esti-
mates the dose distribution to populations at risk, and assesses the
frequency of various effects as a result.
The valuation and ranking compartment assigns value weights to each
effect, derives an aggregate environmental hazard index, and ranks the
agent with respect to other agents.
The model-based system is viewed as being outstandingly relevant to
the ranking objective; it is relatively objective, explicit, reproducible,
robust, and credible. However, it suffers from being less feasible,
technically and economically, less acceptable to decision makers, and
more complicated and difficult to understand than the other systems.
G. Selection Rationale
Based principally on feasibility, simplicity, and acceptability to
decision makers, the expert-based system was selected as the most attrac-
tive for further development.
However, it was also recognized that objective information gathering
was essential for the credible operation of the expert system. It was
seen as desirable that some of this information be presented to the ex-
perts in processed rather than (or in addition to) raw form. For example,
production, import, export, and intermediate usage information could be
combined into a prediction of dispersive use.
35
-------
This processing can be construed as a more objective portion of the
expert system. The model/screening/indexing systems can assist in the
determination of what processed information to present.
Because much of the necessary information would be gathered in any
case for the expert system, it is possible to operate a parallel "objec-
tive" subsystem for only small incremental costs. This subsystem could
be used to calibrate the expert results and to identify, for reappraisal,
unusual agents that might originally escape attention by the experts.
Consequently, the second phase of the study was directed toward
development of a hybrid system based on the expert evaluation of objec-
tive data inputs, and supported by an objective subsystem that combined
the screening and model systems.
*
The "objective" subsystem is only somewhat "more objective" than the
expert system, in that it makes its subjective inputs more explicit.
36
-------
VI A PRIORITY RANKING SYSTEM
BASED ON THE USE OF EXPERT GROUPS
This chapter describes the steps necessary to establish and imple-
ment a priority ranking system for evaluating chemicals or groups of
"&
chemicals in order to decide which should be the subject of EPA-ORD
STAR documents. The proposed system is based primarily on the use of
expert groups to review the available data and establish the priorities.
A. Outline of the System
Under the proposed system the EPA contractor would start by gather-
ing available relevant information on the chemical of interest from a
selected group of sources and putting it in standardized form for con-
sideration by a committee of nine experts who have been selected on the
basis of a preestablished set of criteria. Each expert will be asked to
give his individual estimate of the severity of the potential environ-
mental problems associated with each chemical. These estimates will be
done on a numerical scale (accompanied by explanatory supporting text)
for certain categories of information and on an overall basis for the
chemical. These estimates will be combined by an EPA contractor into a
composite estimate and accompanying text for submission to a Scientific
Review Group which will be asked to review the estimates, examine what
is known about feasibility of control, and make recommendations as to
the priority of each chemical in the preparation of STAR documents.
For agents other than chemicals, or for effects other than biological
ones, ad hoc procedures similar to those suggested for the objective
subsystem (Chapter VII) would be utilized.
37
-------
B. Relevant Information and Literature Sources
It is recommended that the information gathered on a particular
chemical include selected data elements which:
(1) Identify the pure chemical and the commercial chemical
satisfactorily.
(2) Describe the physical and chemical properties of the
chemical that are relevant to possible environmental
hazard.
(3) Indicate the possible extent of distribution of the
chemical to the environment.
(4) Describe the regulation provisions that presently
control the release of the chemical to the environment.
(5) Provide information on the major factors involved in
the transport and transformation of the chemical in
the environment.
(6) Indicate the toxic effects of the chemical on humans
and the environment.
Appendix A lists the data elements within these six categories that
are recommended for inclusion in the information gathering step. It also
indicates the primary sources (publications or organizations) that should
be checked, plus a few additional sources for some of the data elements.
Data elements other than those recommended in Appendix A may be of
special interest for a particular chemical. Where this is recognized,
information on such data elements should be gathered.
C. Format for Presenting Data
It is recommended that the data gathered be presented to the
expert committee in the following four categories.
• General data (identification, properties, and regulations),
® Data on distribution to the environment.
38
-------
• Data on transport and transformations.
• Data on toxic effects.
The recommended detailed format for presenting the information is
shown in Appendix B.
D. Composition of the Expert Committee and Criteria
for Their Selection
It is recommended that the Expert Committee should consist of nine
members. The nine members should be selected to provide three repre-
sentatives with expertise in each of the following areas of concern:
• The extent of distribution of the chemical to the
environment.
• The transport and transformations of the chemical in
the environment.
• The toxic effects of the chemical.
To ensure that the desired expertise is actually obtained, it is
recommended that the members of the Expert Committee be selected on the
basis of the criteria presented in Appendix C. (Although selection of
equal numbers of committee members from business, academia, and govern-
ment may be desirable, it is considered more important to achieve the
balance of disciplines outlined in Appendix C.)
E. Operating Procedures
The following procedure is recommended for getting the maximum bene-
fit from the use of the Expert Committee:
• EPA establishes a list of candidate chemicals (or groups
of chemicals based on a particular element) and publishes
an RFP to obtain a contractor to carry out the contractor
tasks described below.
-------
In cooperation with EPA, the contractor selects a small
group (5-10) of chemicals, preferably related in chemical
structure, in use pattern, or in toxicology, for consider-
ation by the Expert Committee.
The EPA contractor collects the relevant data on the
selected chemicals from the sources indicated in Appendix
A, and puts them into the appropriate format for presenta-
tion to the individual members of the Expert Committee
(Appendix B).
Concurrently, the contractor seeks out candidates for
the Expert Committee using the recommended criteria
(Appendix C) and submits a list to EPA.
EPA selects candidates from the list and invites them to
participate on the committee.
When the necessary nine members of the Expert Committee
have been obtained, the contractor mails the following to
the individual committee members:
- A brief description of the nature of the STAR
documents and the expected function of the Expert
Committee.
- The relevant data on the selected chemicals in the
prescribed format.
- An evaluation sheet on which the individual experts
can provide their estimates, on a numerical scale,
of the severity of the potential environmental prob-
lems associated with the chemical. (The recommended
form for this evaluation sheet and the details of
its use are given in Appendix D.)
When the composite evaluation sheets have been prepared
by the contractor (see Appendix D for details), these
are reviewed by EPA.
F. Scientific Review Group Operations
• Concurrently with the processing of the evaluation sheets
by the contractor, EPA-ORD establishes a Scientific Review
Group whose purpose is to rank in order of priority the
list of chemicals evaluated by the Expert Committee, on
the basis of the composite evaluation sheets prepared by
the contractor.
40
-------
• The contractor and the chairman of the Scientific Review
Group select a mutually satisfactory date for a meeting
and the contractor supplies a brief description of the
expected function of the group and copies of the composite
evaluation sheets to the members approximately 10 days
prior to the meeting.
• The chairman establishes the procedure by which the group
will rank the chemicals on the list. (It is expected
that this will be a system in which each member makes his
own ranked list or votes for each chemical separately
using some preselected numerical scale.)
• The contractor assists the Scientific Review Group by
answering any questions that arise during the meeting,
and, as needed, by supplying details of estimates and
data in the relevant information summaries submitted
to the Expert Committee.
• The chairman of the Review Group submits the priority-
ranked list of candidates for STAR documents to EPA-ORD.
G. Estimated Operating Costs
In terms of the number of data elements that could potentially be
submitted to the Expert Committee, the system is very ambitious. However,
it is recommended that data elements be entered as "NAVA" (not available)
whenever a reasonable effort at searching has produced no results. This
philosophy allows us to specify a maximum effort for data acquisition
which should not be exceeded except for especially significant candidate
chemicals.
The costs of operating the expert-based system beyond data acquisi-
tion fall into three categories:
• Contractor support activities
• Expert Committee activities
• Scientific Review Group activities.
Estimates of the costs of these activities, as well as those for data
acquisition, are shown in Table VI-1. These assume that the experts are
41
-------
able to make their ratings within a period of 20 hours for each group
of 10 chemicals, and that only five members of the Scientific Review
Group receive travel and consultant fees.
Table VI-1
EXPERT-BASED SYSTEM COSTS
(per chemical)
Activity
Chemical Identification
Physical -Chemical Properties
Release and Distribution
Regulations
Transport and Transformation
Toxic Effects
Contractor Support
Expert Committee
Scientific Review Group
Total
Cost Range
(dollars)
$ 50-$
50-
200-
100-
500-
500- 1
100-
500-
300-
$2,300-$4
150
150
600
300
900
,100
200
700
500
,600
Average Cost
(dollars)
$ 100
100
400
200
700
800
150
600
400
$3,450
Assuming that about 36 chemicals per year are nominated for priority
ranking, the system will require about $120,000-$130,000 to operate.
-------
VII DEVELOPMENT OF OBJECTIVE SUBSYSTEM
The objective subsystem is designed to supplement the expert system
discussed in the preceding chapter. Basically; it operates on a subset
of the information (parameter list) gathered for the experts. However,
some information (such as transport in the environmental media) that is
left implicit in the expert system must be made explicit in the objective
one.
There are two basic components of the objective subsystem: the
information gathering component and the information processing component.
Although these components generally proceed at the same time, they can be
discussed separately.
A. Selection of Parameters
All the parameters of the objective subsystem must be quantifiable
in some sense. Even when the basic information is presented as a binary
(yes-no) result or as explanatory, the substance must be transformed into
a number for use.
In selecting the parameters, we used two criteria. First, the para-
meter must clearly be of significance to and usable in the information-
processing framework of the system (described below). Second, data for
a reasonably high percentage of agents must be available in literature
sources or obtainable through reasonably simple computations or experi-
ments. Otherwise, the information will not contribute materially to the
quality of the ranking. The determination as to whether a given parameter
satisfied the first criterion was largely a subjective judgment on our
part; the determination as to whether it satisfied the second criterion
43
-------
was made principally on the basis of experience with the parameter in the
NSF study.
The parameters selected were classified into three areas—release
and distribution, transport and transformation, and effects--similar to
those in the expert system. Both of the selection criteria tended to
limit the number of parameters in comparison with the expert system.
However, more interpretation is needed in using the information associ-
ated with the parameters.
The principal parameters for the objective.subsystem are listed in
Appendix E, with accompanying explanatory texts.
B. Basic System Concepts
The objective subsystem is a hybrid of the screening-indexing and
model systems described in Chapter V. The underlying structure is a
model of the agent's behavior in the environment, with an environmental
hazard index as the output. However, numerous screening questions are
asked in the process of exercising the model, and the degree of the
model's complexity depends on the answers to them.
In this regard, we have attempted to steer a course between the most
common failures of other objective ranking systems. On the one hand, we
have attempted to avoid making the system overly complicated when there is
little or no data to support such complexities. On the other hand, we have
allowed more sophisticated arguments to come into play when information is
available to exercise them, thereby avoiding (to some extent) the tendency
to oversimplify. For example, intermedia transfers of agents are not
addressed by the basic model; however adjustments to model results can be
made when intermedia transfer information is known. Our judgments about
the relative importance of specific processes and about the complexity or
44
-------
simplicity of the modeling attempts are necessarily subjective. However,
we believe them to be reasonable.
The basic theme of the objective subsystem, like the expert system,
is the effects of chemicals on biological systems. This concentration is
justified by an examination of the agents that have been selected or sug-
gested for STARs; very few exert their harmful effects in any other ways.
However, the subsystem makes ample provision for other conditions. As
the prospect of such other conditions occurring becomes less likely, less
detail is supplied in the ranking procedure, and more reliance is made on
ad hoc procedures. We submit that we cannot build a system that will take
care of every eventuality; however, this procedure defines where ingenuity
must be exerted or outside help obtained.
C. Development of Procedures
In developing the procedure, we tried to be as explicit as possible
in defining the methods for gathering and using information. However, we
soon found that this ideal was difficult to achieve, and substantial judg-
ment had to be left to the operators of the system.
The procedure, as developed, consists of a series of steps arranged
in branches, only some of which are followed for any one agent. Each step
is either a state of information test or an instruction for information
processing. Each state of information step leads to one or two or more
possible outcome steps, depending on the result of the test. For example,
if an agent exerts its effects chemically, then one proceeds down the
chemical agent branch, but if not, then one branches off to the nonchemi-
cal agent branch. Each information processing step, on the other hand,
leads to only one further step, either a test or an instruction.
45
-------
The overall structure of the procedure is shown in Figure 1. The
procedure itself, with further instructions for use, is presented in
Appendix F.
D. Operating Personnel and Costs
It was our original hope to construct a procedure that, with the
exception of certain ad hoc studies, could be operated by junior-level
persons in a near-mechanical fashion. This hope faded, and we now recom-
mend a team with the characteristics shown in Table VII-1. The composition
of this team again implies that the "objective" subsystem requiring sub-
jective "expert" opinion, is a hybrid.
The estimated costs of operation of the objective subsystem are
shown in Table VII-2. This table assumes that the costs of information
gathering for the system parameters will be allocated totally to the
operation of the subsystem. In practice, an estimated three quarters of
the information gathering costs would be incurred in any case by the
operation of the expert system.
46
-------
A - MAIN FLOW
AGENT
IDENTIFICATION
/\
Y
/\
Y
CHEMICAL
CHARACTERIZATION
/\
Y
B CLARIFICATION
1
1 *
C NONCHEMICAL
1
' CA CE
, RADIOLOGICAL P^
1 1
J t
G, H, 1,
NONORGANIC
1 1
t t
1 CC
1YSICAL BIOLOGICAL
1 1
AA - AG
DEFAULT RULES
TRANSFORMATIONS
A
Y
*
J HALF-LIVES
1
*
K NONSTEADY
STATE
1
* 1
L INTERMEDIA
TRANSFERS
1
M SPECIAL
POPULATIONS
1
TRANSPORT
AND DILUTION
F DOSE-EFFECTS
O BIODEGRADATION
DOSE
DISTRIBUTION
S NON-
BIOLOGICAL
BIOLOGICAL EFFECTS
*
SA PHYSICAL
1
»
SB AESTHETIC
1
1
SC RESOURCES
1
V DEVELOP
VALUES
\
\
VALUATION
z
AD
HOC
STUDIES
RANKING
( \
FIGURE 1 OVERALL STRUCTURE OF THE OBJECTIVE PROCEDURE
47
-------
Table VII-1
PERSONNEL FOR OBJECTIVE SUBSYSTEM
Senior Level
General!st: Knowledge of mathematical modeling, environmental
science, chemical information, toxic and other
effects, and so on
Chemical Economist/ Knowledge of industrial practices and modes of
Engineer: release
Chemist/Kineticist: Knowledge of physical or physical organic chemis-
try in real environments
Toxicologist/Health Knowledge of toxic effects in man and nonhuman
Scientist: organisms
Junior Level
Analyst: General knowledge of literature sources and mathematical
techniques
Chemist: Knowledge of sources for and meaning of chemical param-
eters
Biologist: Knowledge of sources for and meaning of biological param-
eters
48
-------
Table VII-2
OBJECTIVE SUBSYSTEM COSTS'
(Per Chemical)
Cost Range Average Cost
Activity (dollars) (dollars)
Gathering Release Information $ 200-$400 $ 300
Gathering Fate Information 500- 700 600
Gathering Effects Information 450- 600 550
Processing Information and Ranking 400- 600 500
Ad hoc studies 0- 3,000 550
Totals $1,550-$5,300 2,500
Costs assume that a contractor would operate the system. If
operated by EPA in-house, the allocatable costs would be con-
siderably reduced. If the expert system is assumed to be
operating and obviating about three-quarters of the informa-
tion gathering costs, 36 chemicals could be ranked for a
marginal cost of about $50,000-$60,000.
49
-------
Part Three
TEST OF OBJECTIVE SUBSYSTEM
51
-------
VIII DESCRIPTION OF THE TEST
After an extensive review of the recommended expert-based system and
its objective subsystem, EPA-ORD decided to test the objective subsystem
with a sample of 10 environmental agents of current concern. The full
system was not tested because of the expected time required to recruit and
orient the two expert groups.
A. Definition of the Agents
A list of candidate agents was received from the technical monitor
as part of the Statement of Work. Table VIII-1 shows the list of agents,
as defined, and the stated EPA concern leading to their nomination.
Table VIII-1
NOMINATED AGENTS FOR THE TEST
Agent Name
Symbol
EPA Concern
Antimony
Beryllium
Carbon disulfide
Carbonyl sulfide
Cobalt
Cyanides
Heat from manmade
sources
Lithium
Molybdenum
Plutonium
Sb
Be
CS2
COS
Co
CN
Heat
Li
Mo
Pu
Ocean disposal
Ocean disposal, agricultural runoff
Air pollution from energy conversion
Air pollution from energy conversion
Drinking water contaminant
General water pollutant, industrial
Thermal pollution; system test*
Drinking water contaminant
Drinking water contaminant
Radiation; system test
Chosen to exercise unusual branches of the objective subsystem.
53
-------
The first step in the objective subsystem procedure, as directed by
Step Al (see Appendix F), was to ensure that the agents were well defined.
Only CS2 and COS passed this test unambiguously. The elements (Sb, Be,
Co, Li, and Mo) were well defined as such, but conceptually included a
wide variety of (mainly inorganic) compounds. To avoid wasting time on
unimportant compounds, one to three important representative compounds
were selected for each element. In most cases, these compounds were
selected to estimate the release and environmental fate parameters of the
model, whereas the toxicologically active principle was the element (or
its ion) itself. Similar arguments applied to the cyanides, since CN is
associated with other chemical species to form molecules. Heat from man-
made sources was defined to include waste heat discharged to the environ-
ment from the better known sources, such as power generation. Energy from
other human activities such as metabolism, end uses of energy, or forest
fires was excluded, as were secondary energy effects like changing the
albedo of an area of land or removing shade vegetation from the banks of
a stream. Only the radiation hazards of Pu were covered, using Pu as
an example. Past releases of Pu from weapons testing, and so on, were
excluded. No consideration of the chemical toxicity of Pu was made.
Table VIII-2 shows the representative specific agents used and the reasons
thereto.
B. Assignment of Tasks
As the interim results of this phase were desired within about 6
weeks of initiation, the subtasks were divided among a rather large group
of investigators, larger than would be necessary during normal operations
and larger than would be optimal for efficiency. The principal subtasks
and responsibilities were:
• Completion of agent identification and preliminary effects
checklists, and general literature support (literature
specialists)
54
-------
Table VIII-2
REPRESENTATIVE COMPOUNDS USED
Symbol Representative(s)
Sb Antimony trioxide
Symbol
Be
cs2
COS
Co
CN
Beryllium metal
Beryllium oxide
As is
As is
Cobaltous chloride
Cobaltous naphthenate
Hydrogen cyanide
Sodium cyanide
Potassium cyanide
Be
BeO
Reasons for Selecting Representative
Commercial significance as fire retar-
dant; high toxicity
Primary use as metal
Likely end product of many processes;
high toxicity
Heat Waste heat
Inks, and so on; soluble
CoNaph Paint drier; cobaltous ion is most
likely toxic agent
HCN Forms in acid/neutral solutions from
other CNs. More toxic than CN
NaCN Biggest contributor of CN" ion
KCN Second biggest contributor of CN° ion
Heat Convenient definition with respect to
reasonable controls
Li Lithium chloride
Lithium carbonate
LiCl Drug
Li.CO Large commercial use. Both yield Li
ion
Mo
Pu
Molybdenum trioxide MoO,
Molybdenum disulfide MoS
Plutonium-239
a-radiation
Pu
2
239
Important commercially; likely end prod-
uct
Lubricant
Long-lived product of nuclear power in-
dustry
A mixture of similar metallo-organic compounds; the symbol is a convenience
but is not generally recognized.
55
-------
• Completion of release worksheets for chemical agents (chemi'
cal economists)
• Completion of transport/transformation worksheet (chemists
and research analysts)
• Completion of toxicological worksheets (biologists and
toxicologists)
• Valuation sensitivity analysis (research analyst)
• Ad hoc study on heat (engineer and ecologist)
• Ad hoc study on Pu (biologist and systems analyst)
• Comprehensive study of CS COS (biologist)
• Integration (project leader).
In accomplishing their tasks, the investigators were asked to:
• Record the steps of the procedure undertaken during their
investigations
• Note the areas in which difficulties were encountered and
their methods of overcoming or bypassing them
• Comment on opportunities for improving the system to take
greater advantage of existing data
• Subjectively evaluate the usefulness and relevance of their
activities to the system's objectives
• Comment on the procedures used by other team members.
C. Information Sources
The general philosophy of the STAR ranking system provides that only
moderate resources can be devoted to the ranking procedure in comparison
with those devoted to the production of STARs. Accordingly, only the most
readily available, easily usable, and obvious information sources should
be utilized for the ranking, leaving the comprehensive literature search
and evaluation for the STAR itself.
There are two general categories of information sources that fit
these criteria. First, there are the general information compendia, which
56
-------
provide the same type of information about all, or at least several, of
the agents under consideration. These sources should be maintained by
the operating system for repeated use at each cycle of ranking. The other
category consists of basic documents relevant to a single agent. These
basic documents range from treatises on some special aspect, such as
chemistry or toxicology, to a complete monograph or criteria document,
such as those produced by the National Institute for Occupational Safety
and Health, the World Health Organization, or the International Agency
for Research on Cancer. This category also includes EPA position papers,
preliminary reports, and so on.
Table VIII-3 lists the compendia used in the subsystem test by cate-
gory of information, and Table VIII-4 lists the basic documents by agent.
Complete citations for these references are given in Appendix H.
D. Summary of the Findings
The test proved that it was possible to operate the objective sub-
system to the point of computing an environmental hazard index and ranking
the agent with respect to those already treated. Furthermore, this pro-
"jf
cedure consumed only a moderate number of personnel and economic resources,
and was accomplished in a relatively brief timespan. The evaluation of
the successes and difficulties of conducting the test is presented in the
next chapter.
The results of the test are summarized in Table VIII-5. The rank and
environmental hazard index for each agent are shown with a skeletal out-
line of the reasons for each ranking. The reasons for these rankings are
outlined somewhat further in the following summaries for each agent.
if
It is estimated that the ranking of 10 agents consumed about $15,000.
Symbols for representative compounds are shown in parentheses.
57
-------
Table VIII-3
GENERAL INFORMATION COMPENDIA
Agent Identification and Properties
CHEMLINE
Handbook of Chemistry and Physics (Weast, 1975)
Merck Index (Steiber, 1968)
Lange's Handbook of Chemistry (Dean, 1973)
Encyclopedia of Chemical Technology (Kirk-Othmer, 1963)
"Partition Coefficients and their Uses" (Leo,-1971)
Agent Release
Directory of Chemical Producers (SRI, 1974)
Chemical Economics Handbook (SRI, a)
Census of Manufacturers (Bureau of the Census, 1972)
Mineral Industry Surveys (Bureau of Mines, 1974)
Synthetic Organic Chemicals (International Trade Commission,
1973)
"U.S. Imports, General and Consumption" (Federal Trade Com-
mission, 1973)
"U.S. Foreign Trade, Exports, Commodity by Country" (Federal
Trade Commission, 1973a)
Toxicology
TOXLINE
Toxic Substances List (NIOSH, 1974)
Merck Index (Steiber, 1968)
Dangerous Properties of Industrial Materials (Sax, 1975)
Clinical Toxicology of Commercial Products (Gleason, 1969)
"Water Quality Criteria 1972" (EPA, 1972)
Documentation of the Threshold Limit Values (ACGIH, 1971)
Industrial Toxicology (Hamilton, 1974)
Survey of Compounds Which Have Been Tested for Carcinogenic
Activity (Shubik, 1940-1973)
Handbook of Poisoning (Dreisbach, 1966)
The Pharmacological Basis of Therapeutics (Goodman, 1970)
58
-------
Table VIII-4
BASIC DOCUMENTS
Antimony
Chemical Week. Vol. 113, No. 3, 18 July 1973 (Anonymous, 1973)
Behrens and Rosenblatt, Journal of Chemical Thermodynamics, Vol. 5,
No. 2, March 1973 (Behrens, 1973)
Beryllium
International Agency for Research on Cancer Monograph I (IARC, 1972)
Preliminary Air Pollution Survey of Beryllium and its Compounds
(Durocher, 1969)
The Analysts Journal, November 1958 (Boland, 1958)
Beryllium: Its Industrial Hygiene Aspects (Stokinger, 1966)
Reeves, et al. , Cancer Research, ^7_:439 (Reeves, 1967)
Carbon Bisulfide
"Carbon Bisulfide, Carbonyl Sulfide" (Peyton, 1976)
Characterization of Glaus Plant Emissions (Biers, 1973)
"Environmental Aspects of Fossil Fuel Conversion Processes: Liquefac-
tion" (Exxon, 1975)
Assessment of Catalysts for Control of No from Stationary Power
Plants (Kontsoukos, 1972)
Toxicology of Carbon Bisulfide (Teisinger, 1974)
Carbonyl Sulfide
"Carbon Bisulfide, Carbonyl Sulfide" (Peyton, 1976)
Characterization of Glaus Plant Emissions (Biers, 1973)
"Environmental Aspects of Fossil Fuel Conversion Processes: Liquefac-
tion" (Exxon, 1975)
Matheson Gas Products Information Sheet (Matheson, 1966)
The Chemistry of Carbonyl Sulfide, Chemical Review, ^57:621 (Firm, 1957)
Cobalt
Cobalt Monograph (Centre B'information du Cobalt, 1960)
59
-------
Table VIII-4 (Conclude^,
Cyanide
Handbook of Hazardous Wastes (Capener^ 1974)
Heat
Edinger, Duttweiler, and Geyer, Water Resources Research, 4_:1137,
October 1968 (Edinger, 1968)
Edinger, Water Resources Research, 6.: 1392, October 1970 (Edinger; 1970)
Novotny and Krenkel, Journal Water Pollution Control Federation, 45:
240, February 1973 (Novotny, 1973)
Biology and Water Pollution Control (Warren, 1971)
Comparative Animal Physiology (Prosser, 1973)
Biology Data Book (Altman, 1974)
Lithium
None
Molybdenum
"Environmental Health Aspects of Selenium, Tellurium, and Molybdenum:
A Preliminary Review, No. 3 Molybdenum" (Fishbein, 1974)
"National Emissions Inventory of Sources and Emissions of Molybdenum"
(EPA, 1973a)
Plutonium
"Plutonium: Statement of the Problem" (EPA, 1973b)
"An Assessment of Accident Risks in U.S. Commercial Nuclear Power
Plants" (USAEC, 1974b)
"Generic Environmental Statement—Mixed Oxide Fuel" (USAEC, 1974a)
Plutonium Information Meeting (Startton, 1974)
"Plutonium and the Other Transuranium Elements" (EPA, 1974)
60
-------
Table VIII-5
AGENT RANKS AND REASONS
Environmental
Rank Agent Hazard Index Dominant Source (s)
1 Cyanides
2 Carbon disulfide
3 Beryllium
4 Lithium*
5 Plutonium
6 Antimony
7 Heat
8 Carbonyl sulfide*
9 Molybdenum*
10 Cobalt*
6 x 106
4 x 10
5 x 105
2 x 105
4
1 x 10
3 x 103
2 x 103
4 x 102
3
2
Plating, other industrial
Solvents, fumigants
Industrial processes
Consumer products, e.g.,
drugs
Possibility of accidental re-
lease
Fire retardants, industrial
processes
Energy conversion
Sulfur reduction processes
Industrial processes
Consumer products, e.g.,
paint
Dominant Environmental
Process (es)
Dominant
Medium
Precipitation of insoluble Water
compounds
Oxidation Air
Oxidation; intermedia Air
transfers
Water solubility Water
Intermedia transfers,
radioactive decay
Intermedia transfers
Dissipation
Oxidation
Intermedia transfers
Intermedia transfers
Air
Water
Water
Air
Water
Water
Dominant Effect(s)
Accidental acute toxicity to man
and other organisms
Odor (aesthetic annoyance)
Cancer in man
Central nervous system disorders
in man
Life shortening in man
Heart disease
Fish kill and
Heart disease
Cattle disease
Fish toxicity
in man
reproduction loss
in man
(molybdenosis)
These agents all require extreme dose distribution assumptions to exceed threshold. Otherwise, would all rank tied for last with index 0.
t 4
An extreme assumption was also made for cyanides. Without it, CN would have an index of about 10 and be tied with plutonium. The result-
ing ranks would be CS , Be, CN + Pu, and Heat.
-------
The backup data and calculations are available for inspection at EPA/ORD
and at SRI; Appendix G presents examples of the calculations for cyanides
and carbon disulfide.
1. Antimony (Sb 0 )
Antimony finds its principal use in alloys with other metals.
Howevers its most important dispersive use is as antimony trioxide, a
flame retardant. This compound is assumed to be released to air, water,
and land, but to deposit out of the air and to precipitate out of water
into sediments and soils. No environmental chemistry is expected. Al-
though toxic effects in fish have been demonstrated, the principal concern
is with human inhalation which can lead to severe heart disease. However,
only if antimony is concentrated in hot spots will detectable environmental
effects be predicted by the subsystem.
2- Beryllium (Be, BeO)
Beryllium is found principally in alloys used by the nuclear
and aerospace industries. It is found in rocket exhausts. Beryllium
oxide is used in ceramics and glass, and as a catalyst. The oxide is the
most likely form of emission from other activities involving beryllium,
and is the usual end product of environmental chemistry. It is assumed
that the most important releases are to the air, but some releases to
other media probably occur. Beryllium also disappears from water and air
relatively fast. Berylliosis is a serious concern in occupational settings,
but the principal environmental concern is lung cancer. Although beryllium
in ambient air is under regulation, the subsystem predicts a considerable
hazard still exists.
62
-------
3. Carbon Bisulfide (CS )
Carbon disulfide is emitted during its manufacture and during
the intermediate production of rayon, cellophane, carbon tetrachloride,
and so on. It is dispersively released as a solvent, fumigant, and corro-
sion inhibitor. Carbon disulfide is produced also in sulfur-reducing
technologies such as Glaus plants and automotive catalytic converters.
It is expected in the effluent of stationary catalytic converters for NO
X
control. Carbon disulfide is very volatile and will enter the air, even
if discharged to water. In the air, it degrades fairly rapidly to car-
bonyl sulfide and other products by oxidation. At low concentrations,
the principal health effect is increased risk of heart disease. However,
the subsystem predicts that the aesthetic impacts from the odor of carbon
disulfide would be even more serious overall, putting it high on the list
of ten,
4. Carbonyl Sulfide (COS)
Carbonyl sulfide has few commercial uses but is produced in sul-
fur recovery operations. A gas, it enters the air and is oxidized at
moderate rates. Little is known about its toxicity. Even if it is
assumed that COS has half the effect of CS with respect to heart disease,
and that COS occurs in hot spots, the subsystem still predicts it to be
a low hazard chemical.
5. Cobalt (CoCl CoNaph)
Cobalt has a number of industrial and consumer usess and is
essential to life as a constituent of vitamin B . Cobaltous chloride is
found in dyes and inks, feed additives, and catalysts. Cobalt naphthenate
is a drier for paint. During these dispersive uses CoNaph is assumed to
enter air and water as well as to find land disposal. The cobaltous ion
63
-------
is believed to be the environmental form of cobalt normally encountered.
Cobalt is assumed to move at moderate rates from air and water to inac-
cessible reservoirs. Although pneumonitis and dermatitis in man, and
losses in domesticated animals and plants are possible, the most likely
environmental effect appears to be toxicity to fish, and this occurs only
if hot spots of cobalt occur. Thus, cobalt is rated low as a STAR candi-
date by the subsystem.
6. Cyanides (HCN, KCN, NaCN)
Hydrogen cyanide is used as a fumigant and, more importantly, as
an intermediate to organic chemicals such as dyes. The potassium and
sodium salts find application in electroplating and other metal treatments,
and again as intermediates. HCN enters air and the salts enter water
(according to subsystem assumptions), but HCN tends to transfer to water.
Cyanides are slowly removed from water by complexation. Their acute
toxicity to animal life is well known. It is reasonable to expect occa-
sional fish kills, but effects in man will occur only if there are hot
spots. Because of the latter circumstance, cyanides could be a prime STAR
candidate.
7. Heat
The forms of waste heat considered to be environmentally damag-
ing come from industry; electric power generation, transportation, and
household/commercial sources. Locally, waste heat can raise water tem-
perature a few degrees. Quasi-equilibrium conditions are reached through
transfer of heat to the atmosphere. Concentrations of higher-temperature
water are more likely in lakes than in rivers. The most likely effects
of waste heat are on fish mortality and reproduction. The subsystem pre-
dicts that fish mortality is of only moderate concern, and it does not
treat reproduction.
64
-------
8. Lithium (Li CO LiCl)
Lithium carbonate (glazes for ceramics, drugs) and lithium
chloride (antidepressant drug, heat exchange medium) are two compounds
that contribute lithium ion to the nation's waterways, where the ion is
believed to remain in solution for long periods of time. Although more
severe toxic effects are known, the most likely effects of lithium at
environmental levels are central nervous system disturbances. If lithium
is found in hot spots, these might place it as a pollutant of moderately
high concern.
9. Molybdenum (MoO , MoS )
The most prevalent molybdenum compounds of human origin are
molybdenum disulfide (which is also the ore) and molybdenum trioxide. The
sulfide is used as a lubricant and the oxide as a catalyst and in ceramics.
Molybdenum metal is used widely in alloys, but most releases would occur
in the form of the oxide. It is assumed that molybdenum enters all media,
and that it moves rapidly from air and water to inaccessible reservoirs.
Various obscure toxic effects are known, of which molybdenosis in ruminants
appears the least unlikely at environmental levels. Even if hot spots in
agricultural drinking water occur, the subsystem predicts low environmental
concern.
10. Plutonium (Pu-239)
EPA concerns about plutonium stem from low level leakage from
nuclear fuel reprocessing plants and from the possibility of accidental
release of far larger quantities from reactor melt-down. These are future
rather than present problems, and the probability of accident appears low.
If releases occur, the principal concern would probably be in air, although
settling and rainout of the particulate emissions is expected. Inhalation
65
-------
of plutonium can lead to lung and bone cancers and life shortening in man.
The subsystem predicts the greater overall harm to be the life shortening
and this places plutonium as a pollutant of moderately high concern.
E. Sensitivity Analyses
Two types of sensitivity analyses were conducted in conjunction with
the subsystem test. The first type was a formalized procedure to analyze
the subsystem's sensitivity to the purely subjective relative value scale
for effects. The second set of analyses included a wide range of informal
sensitivity tests to variations in assumptions and procedures.
1. Sensitivity to Valuation
In Appendix F it is recommended that the subsystem operator
(presumably EPA) develop a set of relative values to be placed on various
predicted effects. To provide a starting point, values are suggested for
11 types of effects, relative to an. arbitrary value of 1,000 for human
mortality, which is expressed in excess deaths per year. No distinction
is made for deaths occurring to different age groups or other population
categories. However, these values are extremely subjective, and their
ratios are probably not transitive. (That is, the ratio of values be-
tween deaths and minor illnesses times the ratio of values between minor
illnesses and aesthetic impacts is not necessarily equal to the ratio of
values between deaths and aesthetic impacts.)
The sensitivity to these value assumptions was tested by defin-
ing a range of values for each effect, attempting to reach extreme limits
within reason. One other "intermediate" value was suggested in addition
to the original ones. These values are listed in Table VIII-6 for the
most important effects of the 10 agents ranked.
66
-------
Table VIII-6
VALUATION RANGE
Effect Value Per Unit Effect
.*
No. Effect Units LO NOW ALT HI
0 Human mortality deaths/yr 1,000 1,000 1,000 1,000
1 Human morbidity
(serious disease) cases/yr 100 200 300 800
2 Human morbidity
(other disease) cases/yr 5 10 30 100
3 Human life
shortening yr/yr 15 50 50 300
4 Morbidity (domes-
tic animals) % of pop/yr 50 1,000 300 10,000
5 Mortality (other
animals) % of pop/yr 100 1,000 800 10,000
6 Aesthetic occurrences
annoyance per person/yr 0.01 1 1 50
*
LO = lowest relative value, NOW = present recommended value, ALT =
alternative recommended value, and HI = highest relative value.
Next, several value systems were defined, in which some of the
effects took on the high or low extremes. The matrix of these assumptions
is shown in Table VIII-7. Using these systems with the values from Table
VIII-6 and the predicted numbers of effects, a series of rankings was
developed under the various value systems. Both NOW and ALT values were
used to fill the blanks in Table VIII-7. The range of rankings achieved
for the 10 agents is shown in Table VIII-8. Also shown is the range of
rankings most frequently encountered.
67
-------
Table VIII-7
MATRIX OF VALUE SYSTEMS
Effect No.
Value System _l_ 2- _L A. JL JL
"High Death"
"Low Death"
"High Health"
LO LO LO LO LO LO
HI HI HI HI HI HI
HI HI HI
"Very High Health" HI HI HI LO LO LO
"High Agriculture" HI
"Low Agriculture" LO
"High Economics" HI LO LO
"Low Economics1'
"High Ecology" HI
"Low Ecology" LO
"High Aesthetics" LO HI HI
"Low Aesthetics" LO
Note: Blank elements can be filled either
with NOW or ALT.
Table VIII-8
RANGE OF RANKINGS
Range
Most Frequently
Rank of Ranks Encountered Ranks
Antimony
Beryllium
Carbon disulfide
Carbon sulfide
Cobalt
Cyanides
Heat
Lithium
Molybdenum
Plutonium
6
3
2
8
10
1
7
4
9
5
6-7
2-4
1-5
7-8
9-10
1-2
5-8
2-4
9-10
4-6
6
3-4
2
7-8
9-10
1
7
3
9-10
5
v
Incurred for at least 1/3 of the 25 rankings.
68
-------
From Table VIII-8 it is obvious that the sensitivity to valua-
tion was not particularly great. Half of the agents move through a range
of only two places. Three move through only three places, and only one
ranges through five. There is little problem in designating cyanides,
carbon disulfide, beryllium, and lithium as the top ranking agents.
2. Other Sensitivities
The remaining sensitivities were all tested informally and with-
out conscious effort to test particular assumptions. The only systematic
features of these tests were first that plutonium was deliberately ranked
by two operators independently and second that all agents were reexamined
during final report preparation. The following sensititivies were found:
• To changes in the assumption regarding hot spots. At
first, we disregarded any effects when the highest dose
calculation did not exceed the stated threshold. Later,
we decided to assume that there was some probability of
even higher doses occurring and this resulted in a "high
dose tail" to the dose distribution. This difference
can easily move an agent from the lower half to the upper
half of the list, for example, lithium.
• To the choice of effects to be considered. It is essen-
tial to identify the "most important" (in terms of largest
environmental hazard index) effect. For example, elimi-
nating from consideration the odor effect of carbon
disulfide would move it from the top of the list to the
bottom.
• To consideration of all sources of release. Sometimes
it is possible to miss an important release, for example,
from combustion. This can make a difference of several
orders of magnitude in ranking index (although usually
it does not). The resulting differences in rank may be
several places, but rarely involve a movement from bottom
to top or vice versa.
• To interpretation of toxicology. Toxicology is frequently
very qualitative, and it is rare for more than two points
on a dose-response curve to be known. The subsystem's
sensitivity to the assumption of a threshold or no
69
-------
threshold is similar to its sensitivity for hot spots.
On the other hand, sensitivity to changes in curve shape
not involving threshold shifts is not very great.
• To errors. When one is dealing with a complex series of
computations, it is relatively easy to make a mistake
that is not readily spotted by a checker. Obviously,
the significance of an error depends on its magnitude,
but errors of less than an order of magnitude tend to
be relatively unimportant to the rankings.
In summary, the objective subsystem is sensitive to changes in
assumptions, but much less sensitive than one might assume from the degree
of uncertainty expected.
70
-------
IX EVALUATION OF THE TEST
A. Criteria for Evaluation
The objective subsystem can be judged by the same criteria used to
select among alternative ranking systems in the original development
phase. These criteria (the first five of which are considered somewhat
more important than the last four) were:
Technical feasibility
Economic feasibility
Acceptability to decision makers
Robustness with respect to uncertain information
Simplicity and understandability
Credibility to various interest groups
(Relative) objectivity
(Relative) explicitness
Reproducibility and traceability
The degree to which the test addressed these criteria and the success of
the objective subsystem in satisfying them are discussed below; associated
issues are also raised and evaluated.
B. Successes and Failures of the System
1. Technical Feasibility
The system proved workable in an overall sense. No insurmount-
able difficulty was encountered that prevented the team from deriving an
environmental hazard index for the purpose of ranking. However, the infor-
mation desired by the system was often lacking or highly uncertain;
reliance on default values or assumptions occurred far too frequently for
comfort. Moreover, the procedures were ill-defined at points, and the
analyst had to provide his own interpretation.
71
-------
The technical difficulties of greatest concern seem to fall in
the following areas:
• Release factors—The fraction of processed materials
released by industry and the fraction of consumer
products reaching various media usually are not
known, and need to be estimated by default procedures.
The need for changes in the default procedures also
seems to be indicated.
• Persistence and intermedia transfer—Environmental
transformations of compounds are difficult to pre-
dict, especially for inorganics; intermedia transfers
seem to be more important than they were originally
thought to be, however, data are scanty and a more
robust procedure is needed.
• Transport and diffusion—Virtually no information on
transport and diffusion seems readily available, and
computations are suspect; however, verifiable predic-
tions were surprisingly accurate (well within the
correct order of magnitude).
• Populations at risk—This feature has not been ade-
quately worked out; geographical descriptions were
the only ones easily associated with the agents.
• Toxicology interpretation—Although the basic toxicology
was usually available, the interpretation of the infor-
mation into quantitative terms was fraught with uncer-
tainty. In addition, extreme assumptions about the
distribution of doses were necessary to predict effects
above threshold for half of the agents.
In general, the technical feasibility of the subsystem seems to
depend on the creativity and boldness of the operator in dealing with
deficiencies in information and with subtleties of procedure not treated
in detail by the system instructions. This observation implies that the
principal operator should be a good generalist on the environment, with
considerable ability in making educated guesses.
72
-------
2- Economic Feasibility
The subsystem proved perhaps even more economically feasible
than anticipated. Data gathering required an average of about 40 hours
of (mainly) junior professional effort per agent. Ranking required
approximately 8 hours of senior professional time per agent. Consequently,
the basic test required about $1,500 per agent. Of course, many of the
deficiencies in technical feasibility result from the decision not to
devote more financial resources, so these two criteria trade-off against
one another.
3. Acceptability to Decisionmakers
This criterion for the system has yet to be tested.
4. Robustness
The subsystem provides default values for many of its parameters,
and is robust in this sense. However, many inputs require subjective
evaluation of scanty or missing information. The treatment of these inputs
is sometimes critical to the overall ranking. However, uncertainties
rarely changed ranks by more than three places out of ten.
5. Simplicity and Understandability
The subsystem is quite complex from the point of view of even
a scientifically oriented layman. Many of the procedures, especially the
default steps, need to be taken on faith as they-are not explained in any
detail. However, the steps do make sense if closely examined.
6. Credibility
This criterion for the system has not been tested. The complex-
ity could deter credibility in skeptics, but enhance it for others.
73
-------
Although credibility to the operators of the system is not a necessary
criterion, most of the operators felt extremely uneasy in making the
assumptions, extrapolations, and interpretations of minimal information
that are necessary for the system operation. The system is probably less
credible to the operator than to outside evaluators.
7. Objectivity
The subsystem is relatively objective in comparison with most
other priority systems. However, it required more subjective inputs than
were expected before the test. Some of the subjective features are:
• Choice of the processes in the real world to model in
the system.
• Default values and procedures.
• Interpretation of conflicting, incomplete, or missing
information, especially in
- toxicology
- release factors
- intermedia transfers.
• Choice of effects to consider.
« Treatment of agents which showed doses below threshold
on the first computation of effects. In these cases
we made an assumption on the distribution of higher
doses so that the threshold was exceeded by at least
some doses.
8. Explicitness
Most of the reasoning in the subsystem is relatively explicit
and can be examined on the worksheets and accompanying notes. However,
time and space limitations prevented our making all assumptions explicit.
74
-------
9. Reproducibility and Traceabilitv
The system was designed to be traceable by worksheets and other
reports. Except for the occasional lack of explicitness in assumptions,
cited above, the system succeeds in this design. However, it is not as
reproducible as hoped because of the subjectivity mentioned earlier. Both
ranking by different persons and ranking at different times by the same
person gave different answers, sometimes by orders of magnitude.
10. Summary Statement
In general, however, the subsystem adequately meets the criteria
more often than not. Consequently, it should be considered a valuable
procedure in the systematic setting of priorities for STARs. It is par-
ticularly good at outlining a comprehensive picture of an agent's total
behavior in the environment; for example the subsystem identifies factors
that limit the agent's potential for environmental harm (e.g., short half-
life) and highlights areas of great uncertainty that are critical to
understanding that potential. Nevertheless, the subsystem's deficiencies
are sufficiently disturbing that the rankings it produces should be care-
fully examined before decisions to undertake STARs are made. In essence,
this conclusion is consistent with the previous recommendation to use the
objective subsystem only as an input to and monitor of an expert system
of STAR selection.
C. Suggested Improvements
In the course of operating the subsystem, many areas of needed improve-
ment were identified. These ranged from the correction of typographic
errors to the addition of major branches. Not all of these improvements
were clearly economically justified. We list below the improvements most
likely to be worth their cost, with the additional observation that every
75
-------
new agent put through the system will probably generate additional modi'
fications, and a continuous improvement should be expected with use of
the subsystem.
A group of simple improvements have already been incorporated into
the subsystem and are therefore not detailed here. Recommended further
improvements fall into three groups. The first group could be incorpo-
rated with relatively little additional effort:
• Development of an explicit' procedure for handling sub-
threshold effects predictions, or for assuming a high-dose
distribution (this might not be an acceptable improvement
to certain philosophical/scientific attitudes). A proposed
procedure is included in Appendix F.
• Revision of procedures for inorganic compounds to make
these procedures parallel to those for organic chemicals.
• Revision and expansion of the procedures for treating
nonsteady-state conditions (Branch K). Some revisions
have already been made.
The second group could be accomplished with modest effort:
• Provision of explicit instructions and worksheets for
describing populations at risk. (See Worksheet A37 in
Appendix F. This worksheet was not used in the test.)
• Provision of explicit instructions and worksheets for
describing geographical concentrations, of exposure. (See
Worksheet A22 in Appendix F. This worksheet was not used
in the test.)
The third group would require somewhat more effort:
• Addition of a set of procedures for dealing with ocean
and/or estuarine pollution such as river runoff, precipi-
tation, and dumping. The National Academy of Sciences
publication^ is recommended as a guide.
Assessing Potential Ocean Pollutants (NAS, 1975a)
76
-------
Addition of a methodology for land contamination and effects
(including direct exposures and transfers to ground and sur-
face water), for use with pesticides and similar materials.
Expansion of the procedure for dealing with transportation
releases and other spills, using Office of Hazardous
Materials techniques. Alternatively, drop transportation
branch as unworkable and under the responsibility of the
Department of Transportation.
77
-------
Appendix A
DATA ELEMENTS FOR THE EXPERT SYSTEM
79
-------
Appendix A*'1"
DATA ELEMENTS FOR THE EXPERT SYSTEM
Data Element
Primary Source
Other Sources
I Data Elements Needed to Identify the Pure Chemical and the Commercial Chemical
1. Chemical Abstracts Services Registry
Number
2. Molecular formula
3. Structural diagram
4. Synonyms
5. Trade names for commercial chemical
6. Composition of commercial chemical
CHEMLINE
TADS, TSL
CHEMLINE
Merck
CHEMLINE
CHEMLINE
TADS, TSL
SOCMA
TADS, TSL
TSL, NIOS1
FDA, EPA-OPP
Product Bulletins NIOSH, CEH
II Data Elements Needed to Describe Physical and Chemical Properties Relevant to
Possible Environmental Hazard
7. Melting point
8. Vapor pressure
9. Boiling point
10. Decomposition point
11. Combustion products
12. Flash point
13. Density
14. Flammability limits
15. Explosive limits
16. Solubility in water
17. Solubility in nonpolar solvents
TDB
TDB
TDB
HPC
TADS
TDB
TDB
TADS
TADS
TDB
TDB
TADS, HPC
HPC
TADS, FPG, HPD
TADS
TADS, FPG, HPC
TADS, HPC, FPG
FPG
TADS, HPC
TADS, HPC
Source abbreviations used in this section are explained in Appendix I.
See Appendix F for nonchemical agents and nonbiological effects.
t
81
-------
Data Element
Primary Source
Other Sources
III Data Elements Needed on Regulations that Control the Release of the Chemical
to the Environment
18. Environmental Regulations
a. Effluent limitations guidelines for
30 industries (water pollution) EPA-OWHM
b. Pollutant discharge permits
(water pollution) EPA-OEGC
c. Toxic pollutants standards
(water pollution) EPA-OWHM
d. Hazardous substances standards
(water pollution) EPA-OWHM
e. Drinking water standards EPA-OWHM
f. Ocean disposal regulations EPA-OWHM
g. Toxic substances regulations EPA-OTS
h. Pesticide registration regulations EPA-OPP
i. Pesticide residue tolerances EPA-OPP
j. Ambient air quality standards EPA-OAWM
k. Emission standards EPA-OAWM
1. Solid, waste regulations* EPA-OSWMP
19. Meat additive regulations USDA-MID
20. FDA Regulations
a. Food additive regulations FDA
b. Cosmetics regulations FDA
c. Drug regulations FDA
21. Transportation regulations DOT and USCG
22. Consumer product safety regulations CPSC
23. Occupational safety and health
regulations OSHA
IV Data Elements Needed to Indicate Possible Extent of Distribution of the
Chemical to the Environment
24. Annual U.S. production (P)
25. Percent losses during manufacture (F )
i Jj
Regulations are pending.
SOC (organics)
COM (inorganics)
Industry survey
26. Estimated annual losses during manufacture P(F )
PL
CEH
MY
82
-------
Data Element
Primary Source
Other Sources
27. Annual U.S. imports (I)
28. Annual U.S. exports (E)
29. Apparent U.S. consumption (C)
30. Percent of consumption to dispersive
uses (F )
31. Estimated annual dispersive uses
32. Release rate (R)
33. Pollution Control
a.
b.
Air pollution controls used at
producing plants
Water pollution controls used at
producing plants
34. Principal transportation methods
35. Estimated losses during transportation
36. Consumption pattern (amount to various
uses)
37. Disposal methods following major uses
FT-246
FT 410
C = P + I - E
CEH
IBCP
P(F
m
rLt
C(F
Industry survey
Industry survey
DOT
DOT
CEH
EPA (several
program offices)
EPA-OAWM
EPA-OWHM
TADS
TADS
Industry
associations
f Data Elements Needed on Major Factors Involved in the Transport and Transformation
of the Chemical in the Environment
38. Octanol-water partition coefficient
39. Biochemical oxygen demand
40. Chemical oxygen demand
41. Rate of oxidation in air and water
42. Hydrolysis rate (pH 7 at 20°-25°C)
43. Concentrations in environmental media
(air, water, land, sediments)
CR
JWPCF
JWPCF
Calculation by
expert from
literature data
CA
TDB
44. Concentrations in organisms (fish, mammals, TADS
birds, insects, micro-organisms, plants)
45. Uptake rates by environmental media TOXLINE
46. Uptake rates by organisms TOXLINE
47. Release rates for environmental media TOXLINE
48. Release rates for organisms TOXLINE
49. Residence times in environmental media TADS
CA
EPA (various
program offices)
EPA (various
program offices)
Calculation by
expert from
literature data
TADS, STORET,
SAROAD
BA
BA
BA
BA
STORET, SAROAD
83
-------
Data Element
Primary Source
Other Sources
50. Residence times in organisms TOXLINE
51. Ratio of concentrations in organisms to
concentrations in pertinent environmental
media TDB
52. Mode of entry to organisms (oral,
respiratory, dermal) BA
53. Site of storage in organisms (organs,
tissues, and fluids) XDB
54. Mode of release by organisms TDB
55. Biodegradation products (metabolites) of
the chemical in organisms XDB
BA
TOXLINE, BA
CA
BA, CA
BA, CA
BA, CA, MP
VI Data Elements Needed on Toxic Effects of the Chemical
56.
Animal Effects
a
LD50, LC50
Target organs in mortality
b. LDLo (acute, subchronic, and chronic)
Clinical observations of toxic ef-
fects (acute, subchronic, and
chronic)
c. Metabolic effects indicative of
disease
57. Human Effects
a. Excess mortality from acute or
episodic exposures
b. Excess mortality from chronic
exposures
c. Threshold limit values or no effects
thresholds for acute effects
d. Threshold limit values or no effects
thresholds for chronic effects
Clinical observations of toxic ef-
fects (acute, subchronic, and
chronic)
e. Metabolic effects indicative of
disease
58. Other Data (for inclusion if ascertained
in the course of searching for the above
information)
a. Significant interactions (for example,
antagonism, synergism) with other
agents in either animals or humans TOXLINE
b. Dose-response data TOXLINE
TDB
TOXLINE
TOXLINE
TOXLINE
MEDLINE
TOXLINE
TOXLINE
TSL
TSL
TOXLINE
MEDLINE
TOXLINE, TMIC
MEDLINE, TMIC
MEDLINE, TMIC
MEDLINE, TMIC
BA
MEDLINE, TMIC
MEDLINE, TMIC
TOXLINE
TOXLINE
MEDLINE, TMIC
MEDLINE, TMIC
MEDLINE, TMIC
84
-------
Appendix B
RECOMMENDED FORMAT FOR PRESENTING DATA TO
THE EXPERT COMMITTEE
85
-------
Appendix B
RECOMMENDED FORMAT FOR PRESENTING DATA TO
THE EXPERT COMMITTEE*
I General Identification Data
1. CAS No.: 2. Molecular formula;
3. Structural diagram:
4. Synonyms:
5. Trade names for commercial chemicals:
*
Listed in the same order as in Appendix A; the reasons for seeking these
data and possible sources are also shown in Appendix A.
87
-------
6. Composition of commercial chemical:
70 active ingredient:
70 and name of major impurities:
Content and names of minor impurities;
(%, ppm, ppb)
, (%, Ppm, ppb)
II Physical and Chemical Properties
7.
8.
9.
10.
11.
12.
14.
15.
16.
17.
Melting point: __
Vapor pressure:
Boiling point: _____
Decomposition point:
Combustion products:
Ma i or
°C
_mmHg at
°C at
mmHg
Minor
Flash point:
Flammability limits:
Explosive limits: _
Solubility in water:
13. Density: g/cc at
70 - % by volume in air
7 -
la
_7o by volume in air
parts in 100 parts at °C
Solubility in nonpolar solvents: parts in 100 parts
at °C
88
-------
Ill Regulations
18. EPA guidelines, standards, and regulations
a. Effluent limitations guidelines
i) for the industries producing most of the chemical:
ii) for the industries using most of the chemical:
b. Pollutant discharge permits issued for the chemical:
i) to producing companies
ii) to consuming companies
c. Toxic pollutants:
d. Hazardous substances:
e. Drinking water:
f. Ocean disposal:
g. Toxic substances:
h. Pesticide registration:
i. Pesticide residue tolerances:
j. Ambient air quality:
k. Emissions:
1. Solid wastes:
89
-------
19. USDA meat additive regulations
20. FDA regulations
a. Food additives:
b. Cosmetics:
c. Drugs:
21. Transportation regulations
a. DOT regulations:
b. USCG regulations:
22. Consumer product safety regulations:
23. Occupational safety and health regulations:
IV Data on Distribution to the Environment
24. Annual U.S. production (P): _ Kg
25. Percent losses during manufacture (F ):
PL
26. Estimated annual losses during manufacture (PxF ) _ Kg
PL
27. Annual U.S. imports (I): _ Kg
28. Annual U.S. exports (E): _ Kg
29. Apparent U.S. consumption (C=P+I-E): _ Kg
30. Percent of consumption to dispersive uses (F ): _ %
31. Estimated annual dispersive uses (CxF ): _ Kg
32. Release rate (R = PxF + CxF ): _ Kg
PL D
90
-------
33. Controls used at producing plants
a. for air pollution:
b. for water pollution:
34. Principal transportation methods:
35. Estimated losses during transportation: _Kg
36. Consumption pattern
% of Total
Use Consumption
37. Disposal methods following major uses
Disposal
Use Method
91
-------
V Data on Transport and Transformation
38. Octanol-water partition coefficient:
39. Biochemical oxygen demand (BOD):
40. Chemical oxygen demand (COD) :
41. Rate of oxidation in air and water:
42. Hydrolysis rate (pH 7):
43. Concentrations in environmental media
Medium Concentration Units
Air
Water •
Land
Sediments
44. Concentrations in organisms
Organisms Concentration Units
92
-------
45. Uptake rates by environmental media
Medium Uptake Rate Units
Air ________ _____
Water
Land
Sediments
46. Uptake rates by organisms
Organisms Uptake Rate Units
47. Release rates for environmental media
Medium Release Rate Units
Air ______
Water
Land ______________ ______
Sediments
48. Release rates for organisms
Organisms Release Rate Units
93
-------
49. Residence times in environmental media
Medium Residence Time Units
Air
Water
Land
Sediments
50. Residence times in organisms
Organism Residence Time
Units
51. Ratio of concentrations in organisms to those in media
Organism/Medium Ratio of concentration
52. Mode of entry to organisms
Organism
Mode of entry
(oral, respiratory, dermal)
94
-------
53. Site of storage in organisms
Site of storage
Organism (organs, tissues, and fluids)
54. Mode of release by organisms:
55. Biodegradation products (metabolites) in organisms
Organism Biodegradation Products
VI Data on Toxic Effects
56. Animal Effects
a. Mortality
LD50: mg/Kg
Animal species:
Route of administration:
Gross pathology of principal target organs;
95
-------
Microscopic pathology of principal target organ:
LC50: mg/m3
Animal species:
Calculated total dose (mg/m3 x time x intake rate)
Gross pathology of principal target organ:
Microscopic pathology of principal target organ:
b. Morbidity
Acute effects
LDLo: mg/Kg
Animal species:
Route of administration:
Clinical observations:
Subchronic effects
LDLo: rag/Kg
Animal species:
Route of administration:
Clinical observations:
Chronic effects
LDLo: ____mg/Kg
Animal species:
Route of administration:
Clinical observations:
96
-------
c. Metabolic effects indicative of disease
LDLo: mg / Kg
Animal species:
Route of administration:
Metabolic effects:
57. Human Effects
a. Excess mortality from acute exposures
occupational groups:
route of exposure:
excess mortality:
pathology of principal target organs:
general population:
route of exposure:
excess mortality:
pathology of principal target organs:
selectively vulnerable subgroups:
route of exposure:
excess mortality:
pathology of principal target organs:
b. Excess mortality from chronic exposures
occupational groups:
route of exposure:
excess mortality:
pathology of principal target organ:
97
-------
general population:
route of exposure:
excess mortality:
pathology of principal target organ:
selectively vulnerable subgroups:
route of exposure:
excess mortality:
pathology of principal target organs:
Threshold limit value (TLV) or no effects threshold (NET)
from acute exposures
occupational groups:
route of exposure:
TLV or NET:
clinical observations:
general population:
route of exposure:
TLV or NET:
clinical observations:
selectively vulnerable subgroups:
route of exposure:
TLV or NET:
clinical observations:
98
-------
d. Threshold limit value (TLV) or no effects threshold
(NET) from chronic exposures
occupational groups:
route of exposure:
TLV or NET:
clinical observations:
general population:
route of exposure:
TLV or NET:
clinical observations:
selectively vulnerable subgroups:
route of exposure:
TLV or NET:
clinical observations:
e. Metabolic effects indicative of disease
population exposed:
route of exposure:
TLV or NET:
metabolic effects:
58. Other Data
a. Significant interactions with other agents
(1) In animals
other agents:
animal species:
route of exposure:
clinical observations:
99
-------
(2) In humans
other agents:
population exposed:
route of exposure:
clinical observations:
b. Dose-response data
Incidence
or Response
(units)
Exposure,
(units)
100
-------
Appendix C
RECOMMENDED CRITERIA FOR SELECTING MEMBERS
OF THE EXPERT COMMITTEE
101
-------
Appendix C
RECOMMENDED CRITERIA FOR SELECTING MEMBERS
OF THE EXPERT COMMITTEE
Experts on the Extent of Distribution of the Chemical to the
Environment
Chemical Market Research
Academic Training
Relevant Experience
Peer Recognition
Master's degree in chemistry or chemical
engineering. Bachelor's degree in these
disciplines plus a Master's degree in
business administration. Experience
equivalent to these degrees would also be
satisfactory.
A minimum of 10 years experience, the last
5 of which have been in the area of chemi-
cal market research on the chemical (or
group of chemicals) of concern.
Membership in a principal scientific
society representing candidate's disci-
pline, preferably the Chemical Market Re-
search Association. Publication, during
the last 5 years, of at least one article
(in an industry-oriented periodical, en-
cyclopedia, or book) on the chemical (or
group of chemicals) of concern.
103
-------
II Experts on the Transport and Transformations of the Chemical
in the Environment
Environmental Engineering
Academic Training
Relevant Experience
Peer Recognition
Either a Ph.D., Sc.D., or equivalent
experience in an engineering discipline
related to pollution control.
A minimum of 10 years experience, the
last 5 of which have been in the area of
research, development, or teaching of
pollution control.
Membership in a principal scientific so-
ciety representing candidate's discipline.
Publication of at least 10 papers in ref-
ereed journals in the related field since
completion of academic training.
Ecology, Earth, and Life Sciences
Academic Training
Relevant Experience
Peer Recognition
Either a Ph.D., Sc.D. , or equivalent ex-
perience in the subdisciplines of aquatic
or terrestrial ecology, which include
fisheries, limnology, botany, geology,
meteorology, and oceanography.
A minimum of 10 years experience, the
last 5 of which have been in the area of
environmental pollutant transport, eco-
logical effects of pollutants, and/or
environmental impacts of pollutants.
Membership in a principal scientific so-
ciety representing candidate's discipline.
Publication of at least 10 papers in ref-
ereed journals in the related field since
completion of academic training.
104
-------
Environmental Health
Academic Training
Relevant Experience
Either a Ph.D., Sc.D., M.D., or equiva-
lent experience in environmental health,
public health, industrial hygiene, or
epidemiology.
A minimum of 10 years experience, the
last 5 of which have been in the areas of
pollutant transport, transformations, ex-
posure and/or health effects. An inter-
disciplinary background covering pollu-
tion control, ecology, and health effects
is desirable.
Peer Recognition
Membership in a principal scientific so-
ciety representing candidate's discipline.
Publication of at least 10 papers in ref-
ereed journals in the related field since
completion of academic training.
105
-------
Ill Experts on the Toxic Effects of the Chemical
Animal Toxicology
Academic Training
Relevant Experience
Peer Recognition
Either a Ph.D., Sc.D., D.V.M., M.D., or
equivalent experience in toxicology,
pharmacology, pathology, biochemistry, or
medicine.
A minimum of 10 years experience, the
last 5 of which have been in the area of
toxicology of environmental pollutants or
xenobiotics.
Membership in the principal scientific
society representing candidate's disci-
pline. Publication of at least 10 papers
in refereed journals in the related field
since completion of academic training.
Human Toxicology
Academic Training
Relevant Experience
Peer Recognition
Either a Ph.D., Sc.D., M.D., or equiva-
lent experience in epidemiology or occu-
pational medicine.
A minimum of 10 years experience, the
last 5 of which have been in the area of
epidemiology of environmental pollutants
or xenobiotics.
Membership in the principal scientific
society representing candidate's disci-
pline. Publication of at least 10 papers
in refereed journals in the related field
since completion of academic training.
106
-------
Appendix D
RECOMMENDED EVALUATION SHEET AND PROPOSED METHOD OF
USE BY MEMBERS OF THE EXPERT COMMITTEE
107
-------
Appendix D
RECOMMENDED EVALUATION SHEET AND PROPOSED METHOD OF
USE BY MEMBERS OF THE EXPERT COMMITTEE
The Evaluation Sheet submitted for each chemical to all nine members
of a particular Expert Committee would have the following format:
Evaluation Sheet for Estimating the Severity of the Potential
Environmental Problem Associated with the Chemical
Please circle the number which most closely represents your estimate of
the chemical's environmental significance in the following four catego-
ries and provide a short description of the major factors behind each of
your four estimates.
1. Estimate the significance of distribution of the chemical to the
environment.
POTENTIAL PROBLEM SCALE
None Very Little Moderate Major
0 123 4567 89 10
Major factors involved:
109
-------
2. Estimate the significance of transport and transformations of the
chemical in the environment.
POTENTIAL PROBLEM SCALE
None Very Little Moderate Major
0 123 4567 89 10
Major factors involved:
3. Estimate the significance of the chemical's toxic effects.
POTENTIAL PROBLEM SCALE
None Very Little Moderate Major
0 123 4567 89 10
Major factors involved:
4. Estimate the overall severity of the potential environmental prob-
lem associated with the chemical.
POTENTIAL PROBLEM SCALE
None Very Little Moderate Major
0 123 4567 89 10
Major factors involved:
It is recommended that the contractor process these evaluation
sheets in the following way upon receipt from the members of the Expert
Committee:
1. Review the estimates for instances in which one expert's
estimate and major factors differ markedly from the in-
formation supplied by the other experts. Telephone this
110
-------
expert to point out the difference and ask him to consider
changing his estimate or to provide more information if
he feels a change is not appropriate. In the latter case,
the information provided should be telephoned to the other
experts and they should be given an opportunity to change
their estimates.
2. Once the estimates have been finalized, compile a com-
posite evaluation sheet for each chemical by averaging
the estimates from the nine experts in each of the four
categories, and prepare a description of the major
factors involved using the descriptions provided by
the experts as guidance.
3. Select one of the chemicals from those considered by the
first Expert Committee for submission to subsequent Ex-
pert Committees along with the new chemicals submitted
to these committees. The estimates on the composite
evaluation sheet for this chemical can then be compared
with the estimates from the subsequent committees and
used to maintain consistency across the estimates sub-
mitted at different times. If the composite estimates
for the reference chemical from a particular committee
vary widely from the original estimates, then the com-
posite estimates for the new chemicals in the group
considered by the new committee may need to be adjusted
accordingly.
Ill
-------
Appendix E
PARAMETERS FOR USE IN THE OBJECTIVE SUBSYSTEM
113
-------
Appendix E
PARAMETERS FOR USE IN THE OBJECTIVE SUBSYSTEM
Three team members were asked to recommend parameters for inclusion
in the objective subsystem, using as criteria (1) that data for the
parameters should be available for a substantial fraction of the agents
and (2) that the parameter have reasonably high relevance to the predic-
tion of environmental effects from controllable sources of the agent.
This appendix presents the rationale behind the team members' selec-
tion of parameters in the categories of release, fate, and effects in
the environment. It also presents the rationale behind modifications
introduced to weave the parameters into a mathematical model.
I Release to the Environment
The processes and uses that can lead to the release of pollutants
to the environment were identified (Table E-l) as a first step in refin-
ing parameters, defining source data, and selecting standard units for
parameters for such releases. No attempt was made to develop equations
that would sum the listed routes of release.
Availability of data related to all Table E-l release routes A-N
would be the best possible case for summarizing quantitative information
for priority ranking decisions; Table E-2 describes parameters, symbols,
and information sources for these routes. The information source list-
ings are probably not all-inclusive and could be developed further.
Tables E-l and E-2 represent the level of detail that might be re-
quired for assessment of a group of "critical" chemicals. However, for
practical reasons it may not be reasonable to go to this level of detail.
115
-------
Table E-l
PROCESSES AND USES LEADING TO RELEASE OF
POLLUTANTS TO THE ENVIRONMENT
Activity
Manufacture
Transportation of Material
Direct Dispersive Use
(not formulated w/other materials)
Intermediate Use
T Interrelated
Release as By-Product or Impurity
(in manufacturing of other product)
Formulated Product Use
Nonintentional Production
Description of Release Routes to Environment
'"Normal" emissions or waste disposal during manufacturing
process, including clean-up
'Off-grade batch disposal
[Accidental plant release (spills, and so on)
^Release or waste from cleaning of bulk storage facilities
I Release during loading—spills, evaporation
Release during transport
Release during unloading—spills, evaporation
Release during clean-up of shipping container
(Direct release related to method of use
|"Sealed" dispersive use
(Loss during conversion to another material
((includes storage, transfer, processing at conversion site)
(Release during storage, transfer and formulation
(Release during use of the formulated product
(Nonintentional, nonmanufacturing release
(via chemical or physical processes
Symbol
B
C
D
E
F
G
H
I
J
!A-D
Apply with
quantities
at much
lower levels
L
M
-------
Table E-2
DATA SOURCES AND UNITS FOR RELEASE ROUTES A-N
Release
Route Symbol
B
C
D
E
F
"Published" Data Sources/Units
1) NSF (Organic Chemicals); (kg/yr)
2) EPA documents; (kg/yr)
3) NIOSH (kg/yr)
1) Department of Transportation (DOT)
1) U.S. Tariff Commission, f, I, E; (kg/yr)
2) Census of Manufactures; (kg/yr)
3) Minerals Yearbook; (kg/yr)
4) Chemical Economics Handbook; (kg/yr)
5) Chemical Origins and Markets; (kg/yr)
6) NLM/EEC Studies; (kg/yr)
7) NCI Data Bank; (kg/yr)
8) LRPS; (kg/yr)
9) CEH Clipping Files (i.e. trade lit.)
10) "Other" multiclient studies
Same as I
NSF
EPA (?)
EPA (?)
NIOSH (?)
Numerous sources which provide information on
use of chemicals, metals, etc. in products
(e.g. CEH, Minerals Yearbook, etc.)
1) EPA Monitoring Sources
2) Lit. search
"Other1' Data Sources/Units
1) Industrial survey; kg/yr
or °L of production (P)
2) Expert estimate; (% of
P in kg/yr)
Same as A
Same as A
Same as A
Same as A
1) and 2) same as A
3) Survey of shippers; °L of
quantity shipped in kg/yr)
1) , 2) and 3) same as F
4) Survey of users receiving
shipments; (7<, of quantity
received in kg/yr)
1) Survey of shippers; (% of
quantity shipped in kg/yr)
2) Expert estimate; (% of
quantity shipped in kg/yr)
Same as A
Same as A
Same as A
Same as A
117
-------
Therefore, Table E-3 reduces the usable release routes and parameters to
a more practical level. In effect, the routes of Table E-3 combine the
various release rputes of Table E-l, and reduce the level of detail and
precision to lessen time and cost requirements.
In general, release to the environment resulting from emissions and
waste during manufacture will vary widely between companies and even be-
tween plant locations within the same company. This was the case with
the organic chemicals surveyed for the NSF study. Therefore, on a
practical basis, release data for the various phases of manufacture can
best be expressed as some percentage of the total production of the
chemical or compound.
For the purposes of ranking, it should be possible to determine or
estimate the quantities of chemicals released by the routes listed in
Table E-3, thus, eliminating many of the gaps that would result from a
more complex scheme. Table E-4 gives time/cost estimates.
II Transport, Transformation, and Fate
a. Degradation
The degradation of a chemical in the environment is represented
quantitatively in terms of the disappearance of the chemical as a func-
tion of time:
d (chemical) .
~ = k (chemical) [x]
dt
The disappearance is expressed by two types of terms, the rate constant
(k) and the concentration(s) [x] of the reactant(s). The rate constant
is a measure of the energy of the reaction of the subject chemical with
some reacting/attacking species x. The latter may be a chemical oxidant,
hydrolytic reagent or microbial agent. The rate constant is independent
118
-------
Table E-3
RELEASE ROUTES, DATA SOURCES, AND UNITS
Release Route
Data Sources
Units
1) Emissions and waste (EWR) 1) Emission and waste expressed as a func- 1) kg/yr from (P) (x%)
resulting from mantlfac- M nn nf nrnHTirM nn fT>} Knu-r^on fnr /"PV whft-ra v is a-n ^fit—fm
resulting from manufac-
turing operations.
2) Transportation from pro-
ducer to point of use.
(TR)
tion of production (P). Sources for (P):
U.S. Tariff Commission
Census of Manufactures
Minerals Yearbook
CEH
Chemical Origins and Markets
NLM/EEC Studies
NCI
NSF
Percentage of emission and waste and
specific environment (a, w, s, o) to
to which released; will be available
from:
EPA data
NIOSH data
or by estimate.
2) Expressed as a function of noncaptive
consumption or production. DOT is
probably only source. Total release
may be negligible by this route.
where x is an estimate
of emission and waste,
if quantitative data
are not available
FU =
-------
Table E-4
TIME/COST ESTIMATES
(Per Chemical)
Time Cost
Release Route Information Need
1) Emission and waste 1-a) Emission quantities (kg/yr) 1-a)
from manufacturing
R 1-b) Waste quantities (kg/yr) 1-b)
1-c)
1-d)
1-c) Production (kg/yr) 1-f)
[Emission and waste could
be expressed as a function
of production (P)] 1-h)
1-i)
l-j)
1-k)
1-D
1-m)
1-n)
l-o)
1-p)
1-q)
Source
NSF Survey (contains estimates and
reported industry values)
EPA Datat
Estimate (as function of produc-
tion volume and method of manufac-
ture)
Survey of manufactures
U.S. Tariff Commission
Census of Manufactures
Minerals Yearbook
Chemical Economics Handbook
Chemical Origins and Markets
NLM/EEC Studies
NCI Mark II Data Base
IARC Monographs
NSF Study
Chem. Profiles (ref. CMR)
CEH clipping files
Expert estimate
(hrs) (dollars)
(RA)*
(RA)
(P)*
(P)
(RA)
(RA)
(RA)
(RA)
(RA)
(RA)
(RA)
(RA)
(RA)
(RA)
(RA)
(RA)
(P)
i 3
1 12
1 33
2 66
i
*
*
*
4
*
i
*
4
4
£
6
66
1
* 17
Research Analyst.
Would require contact with this agency to determine whether emission and waste data are available. Release to various environments
(A,W,S,0) should become evident upon review of data.
Professional.
-------
Table E-4 (Concluded)
Release Route
2) Transportation from
producer to point of
use (TR)
3) Dispersive use (DU)
Information Need
2-a) Reported losses during
transport (kg/yr)
OR
2-b) Estimated losses during
transport (kg/yr)
3-a) Situations where chemical
is released to the environ-
ment as a result of its use.
Can be expressed as:
DU = (P + I) - [E + (nonDU)]
P = U.S. production
I = Imports
E = Exports
NonDU = uses in which
chemical undergoes change.
Source
4) "Nonintentional"
production (release
resulting from chem-
ical or physical
processes)
4-a) Reported or estimated pro-
duction (kg/yr)
2-a) Department of Transportation An-
nual Report on Hazardous Materials
2-b) Estimate (by expert) expressed as
a function of production. See
If-lq for sources of production
data.
3-a) Published use information on a
chemical will generally give both
DU (and nonDU) data directly.
Such sources are:
Chemical Economics Handbook
NSF Study
NCI Data Bank
NLM/EEC Projects
Chem. Origins and Markets
IARC Monographs
CEH clipping files
Chem. Profiles (CMR)
3-b) If it is necessary to determine DU
from P, I and E data, the latter
are available from:
P--(see If-lq)
I--U.S. Imports (FT-246)
E--U.S. Exports (FT-410)
4-a) Monitor sources:
National Emissions Data System
(NEDS)
Storage and Retrieval of Aero-
metric Data (SAROAD)
OSHA (?)
EPA (?)
Time
(hrs)
(RA) *
(P) 2
Cost
(dollars)
66
(RA) 4
P-2
48
66
(RA) 6
72
(RA) 2
24
Not all sources would be required for a chemical. Actual cost per chemical would probably range $200-$400.
-------
of concentration but is temperature dependent. The above reaction is
referred to as a bimolecular reaction, and is first order in both the
chemical and in x. A bimolecular rate constant is given in liter mole"1
-I o 1 1
sec" or cm molecule" sec" (or other appropriate units). The concen-
tration terms (chemical and [x] are expressed in units appropriate to
I _0
the rate constant, for example, moles liter , molecules cm , respec-
tively.
In the environment, there are actually many species x. which
may react with a chemical, each having its characteristic rate constant
(energy relationship). If we assume only bimolecular processes operat-
ing (a simplification for discussion), the rate of disappearance of a
chemical is then
n
-d (chemical)
dt
= _> k. j~x. 1[chemical]
1
u
= [chemical] \ k fx "I
/__j iL iJ
i
It becomes clear that the major degradation pathways will be those for
which the product k.[x.] is largest.
An example of the complexity of such potential degradation
processes is oxidation in the atmosphere. Laboratory experiments have
identified and quantified many different oxygen-derived species which
may react with a pollutant. Hydroxyl radical, hydroperoxyl radical,
alkoxy radical, ozone, and singlet oxygen are all reactive agents with
rate constants ranging from 10 to less than 10 liter mole sec .
Since the rate of the oxidation [-d (chemical)/dt, as differentiated
from the rate constant] is dependent also on the concentrations of the
reactants, a complete, detailed kinetic expression for atmospheric oxi-
dation is an impossibility. If hydrolysis, microbial degradation, and
122
-------
other fates of the chemical in other environments, are included in the
analysis, the large number of unknown factors presents a similar situa-
tion.
For chemical degradation, the problem may be approached if
data is available on both the rate constant and concentration of the re-
acting species x for a given process. Calculation of the product k.(x.)
for a process may then provide some measure of chemical degradation. As
a result of the lack of complete data on all reactions (some of which are
not measured and some of which may still be unknown), such a number would
be a minimal degradation rate. However, the data which are used may
represent a dominant degradation process and therefore a more meaningful
number. For the consideration of environmental effects, a minimal rate
of degradation will at least provide a documentable basis for which to
work. Other competitive or more dominant processes would result in more
rapid degradation and therefore less of a problem than anticipated.
For the purposes of obtaining a half-life for a chemical in an
environmental medium, an assumption or assignment must be made as to a
constant concentration of reactant species x. Such an assumption is
reasonable if the environment is considered as an infinite sink of such
reactants at steady state concentrations (water and oxidants in aquatic
environments, ozone, hydroxyl radical, and other oxidants in air). The
product k.(x.) then represents a new constant value with the chemical
still of first order dependence, referred to as a "pseudo-first order
reaction." This rate constant is in terms of reciprocal time (min~ ,
sec"-'-, etc.). The half-life for a first order reaction is independent
of the concentration of a chemical and is given by
t1/2
123
-------
Here k' is the product k. (x. ) and the units of the half-life are defined
by the units of the k'. It can be seen that where data are available,
the half-life for a chemical undergoing a particular degradation may be
evaluated for that process, with the resulting value then representing
a maximum half-life in the environment under consideration. Where several
processes may be acting simultaneously, the process showing the smallest
half-life (greatest rate of reaction for the first order reaction kinetics)
is assumed to be dominant, and its half -life is taken as the upper bound
of the actual half -life of the subject chemical.
The parameters of degradation in air and water are described
more fully in Appendix F. Following are descriptions of other parameters
related to the transport and transformation of chemicals in the environ-
ment.
b. Vapor Pressure
The vapor pressure for a pure compound at 20°C would be ex-
pressed in units of mmHg. If the vapor pressure is not available,
knowledge of the boiling point, T (at 760 mmHg), allows approximation
of the heat of vaporization AH^ .
AH
P fo 21 cal/deg (Trouton's Rule)
bp
With the vapor pressure P-^ at another temperature T-, , the value at 20°C
may be calculated by
AH
(1/TL - 1/293)
If necessary, the vapor pressure may be approximated by reference to
another compound of similar molecular weight, structure, and boiling
point. Vapor pressure citations are given in:
124
-------
• CRC Handbook of Chemistry and Physics (HCP).
• Langes Handbook of Chemistry.
• J. Timmermans, Physico-Chemical Constants of Pure
Organic Compounds, Elsevier, N.Y. 1965.
• T. E. Jordan, Vapor Pressures of Organic Compounds,
Interscience, N.Y., 1954.
• Chemical Abstracts, under specific compound or vapor
pressure listings.
• Industrial data, manufacturers and suppliers.
c. Partition Coefficients
The partition coefficient is useful in considering the fate of
a chemical in water, and also for biological implications. The partition
coefficient is a unitless quantity and will be defined as the concentra-
tion ratio of the organic solvent to water phase. Partition coefficients
in various organic solvent systems are available with direct experimental
and some calculated values for octanol/water. The following also may be
consulted for semi-quantitative evaluations.
• A. Leo, C. Hansch and D. Elkins, "Partition Coefficients
and Their Uses," Chem. Reviews, 71 (6) 1971.
• Chemical Abstracts (under subject chemical and partition
coefficients).
d. Occurrence of Chemical in Environment
The occurrence of a chemical in the environment is probably
best obtained by EPA from its sources and programs (SAROAD, STORET). As
the literature is searched for information on the degradation of the
chemical, its occurrence may also be found under various subheadings
(for example, analysis of, in soil, water, air, aerosol, and so on). An
experienced chemist familiar with the literature would recognize useful
information.
125
-------
e. Biodegradation
Degradation by biological action is probably the dominant
process for most chemicals, and especially so in soil and water environ-
ments. The rate of degradation of a chemical will be a function of both
the specific organisms and the populations, and will vary with the
season, climate, and history of the environment (for example, mountain
stream, minor and major rivers systems exposed to limited industrial
waste disposal, or sewage plant waters). For a useful evaluation of
biodegradation of a chemical, we propose a three step effort.
1. A literature search by a person with appropriate
training, and an evaluation by a knowledgeable
expert. Either extensive studies or simple BOD
values available in the literature for some com-
pounds could be used.
2. A BOD experiment using a sewage sample if little
or no data is found in the literature. BOD tests
are to be determined at 4 inoculate concentrations
at times of zero, 15 minutes, and 5 days. Com-
parison of the differences between the BOD and a
sample containing no added chemical would then
provide a quantitative measure of biodegradation.
3. If little or no degradation is found in step two,
identical BOD measurements would be carried out on
an acclimated sewage sample. The need to resort
to this experiment would indicate a more recalci-
trant chemical.
Biodegradation information can be found in:
Chemical Abstracts, under specific chemical as well as
under heading of BOD, Biodegradation, Biological Treatment.
f. More Refined Data Treatments
As the subject chemical is being considered for a STAR, the
following paper may be of interest to provide further data:
126
-------
D. Mackay and A. W. Wolkoff, "Rate of Evaporation of
Low Solubility Contaminants from Water Bodies to
Atmosphere," Environ. Sci. Tech.. 7 (7) 611 (1973).
A half-life may be calculated for loss through evaporation,
with the use of available vapor pressures, molecular weights, and solu-
bility data. The calculation requires some reasonable assumptions, such
as the water depth and the rates of water evaporation. The calculation
may be easily carried out by persons familiar with kinetic calculations.
A second paper approaches the problem of estimating sorption
of the chemical onto soil particles.
S. M. Lambert, "Functional Relationship Between Sorption
in Soil and Chemical Structure," J. Agr. Food Chem., 15(4)
572.
This paper will be of use if some sorption data on the class
of compounds already exists. The parameter used is the parachor, which
can be calculated from surface tension, density, and molecular weight
data. Parachor can also be calculated from structural parameter consid-
erations. The latter is simpler and reliable, and is suggested for use
in this application. When the fate of chemicals absorbed on soil is
under consideration, other factors such as humidity, moisture content,
soil porosity, and temperature are also important. However, the parachor-
soil absorption information will allow at least a semi-quantitative as-
sessment to be made if other soil sorption data are available for com-
parison and calculation.
g. Time Required
With the use of the references cited, it would require approxi-
mately 1/2 day of effort at B.S. level in chemistry to acquire the basic
information (solubility- boiling point, vapor pressure,, Chemical Abstracts
listing, and so on), for each compound. Effort at the Ph.D.-expert level
127
-------
would require a maximum of 2 days per compound; this would include the
survey of Chemical Abstracts, literature location, evaluation, and cal-
culations. In many cases a chemical may be completed in less than a
day if appropriate information is readily available. It would expedite
the evaluations if as many compounds as possible were surveyed in a con-
certed effort, especially when the same reference sources were being used.
The degree of sophistication required to calculate or generate the desired
information should be left to the best judgment of the person evaluating
the literature. As the data to be obtained are biased toward establishing
a minimal degradation rate (maximum half-life), it is necessary that the
searcher also recognize any documentable arguments for more rapid degra-
dation that may be generally applicable (direct photolysis, singlet
oxygen reactions, catalytic effects in soil, and so on).
The literature search for information on biodegradation would
take several hours to locate and evaluate. Depending on the facilities
and experience of the contractor (or EPA if done in-house), the BOD
measurements would take about 1/2 day per compound. If an acclimated
sewage sample BOD run is required, another 1/2 day would be required.
Ill Toxicology
The toxicological data selected for use in the objective ranking
system for chemicals were chosen on the basis of estimated availability
and utility. Emphasis was placed on data resulting from controlled ex-
perimentation with laboratory animals rather than on data derived from
field observations—including epidemiological and human case history in-
vestigations—because data from controlled experimentation more closely
meet our selection criteria.
The types of data we have selected include threshold limit values for
workroom spaces, LD50 values, and minimum dosages demonstrated to produce
chronic toxicological effects--including carcinogenic, mutagenic, and
128
-------
teratogenic effects. Except for acute toxicity in aquatic organisms,
there is little information published on the toxicity of chemicals to
nonmammalian animals. In acute toxicity tests with aquatic organisms,
the LC50 values derived from survival data are often based on different
exposure times and thus are difficult to equate. For this reason, we
have not included nonmammalian toxicity data.
The most likely ways that man may be poisoned by an environmental
chemical are by ingestion, contact with the skin, or inhalation; conse-
quently, we recommend the use of data obtained in studies in which the
oral, dermal, or pulmonary route of exposure wa"s used. We also recommend
that for all in vivo studies the administered amounts be converted to
mg/Kg units. Exposure concentrations administered in inhalation studies
should be converted to weight units. Admittedly, this conversion is
subject to error. Average respiratory rates and tidal volumes required
for this conversion are presented for various animals in the Biology Data
Book (FASEB, 1964).
a. Threshold Limit Values
Threshold limit values (TLV) for workroom spaces have been
established by OSHA to minimize exposure hazard to workers. TLV's have
also been assembled for many chemicals by the ACGIH. The values are
based on animal and human toxicological data and are expressed in parts
o
per million (ppm) or mg/m . The list of values published by OSHA or
ACGIH represent 8-hour weighted averages or ceiling values that are well-
defined.
We recommend rating those chemicals for which TLV's are avail-
able in the following manner:
129
-------
Rating
1
2
3
4
5
6
TLV
(ppm)
> LOGO
500-1000
100- 500
50- 100
10- 50
< 1
b. Acute Toxicity
Acute toxicity tests are usually the first type of test per-
formed on a chemical of unknown toxicity. Such tests are usually in-
tended to provide survival data from which the dose or concentration that
will kill 50% of a test population may be estimated. This dose is called
LD50. A rating system that may be used for LD50 values follows:
Rating
1
2
3
4
5
LD50
(mg/Kg)
> 5000
500-5000
50-5000
1- 50
< 1
In that the toxicity of a compound often varies with the route
of administration, species of test animal, age, physiological state, and
other factors, we recommend that use be made of average LD50 values. If
no toxicity data are available for a chemical, we recommend, as a mini-
mum, that the LD50 (oral) be obtained experimentally using either the
rat or mouse. Such a test will cost less than $1000.
Sources of LD50 values, in the suggested search sequence, are
the Toxic Substance List, Handbook of Toxicology (Spector), TOXLINE/
MEDLINE/CBAC, the Merck Index, Farm.Chemicals Handbook (pesticides only),
Biological Abstracts, Index Medicus, and Chemical Abstracts.
130
-------
c. Repeated Dose Effects
Repeated dose toxicity studies are usually designed to deter-
mine the effects of long-term exposure to sublethal doses of a chemical.
The duration of exposure may be up to 90 days (subchronic toxicity test)
or may extend throughout the normal life span of the test animal (chronic
toxicity test). During the course of a repeated dose study, many differ-
ent observations may be made. These observations include body weight,
behavior, organ function, gross and histopathology, hematology, physiol-
ogy, biochemistry., metabolism, life span and others.
Some of the observed effects may have less significance in
terms of the well-being of the organism than others; hence, the operator
of the ranking system should attempt to categorize the reported effects
in order of increasing biological significance. An experienced toxicolo-
gist may be needed for this task. Categories that may be used are listed
below.
Rating Category
1 Chemical accumulates in tissues without apparent effect.
2 Chemical produces effects of uncertain biological significance.
3 Chemical produces effects related to disease.
4 Chemical produces a persistent disease state.
5 Chemical significantly reduces the life span of the organism.
Chemicals in each category should then be ranked according to
the minimum dose that produces an effect of significantly greater magni-
tude or frequency than observed in the controls. By studying the data
collected, the operator of the ranking system should be able to devise
a rating system similar to that suggested for ranking LD50 values.
Although there are numerous reports on subchronic and chronic
toxicity of chemical substances, there is no single comprehensive source.
The Toxic Substances List may be used as a guide in determining if a
chemical has been subjected to subchronic or chronic exposure evaluation;
131
-------
however, information provided in this document is limited. A search of
computerized data banks such as TOXLINE, MEDLINE, and CBAC should not be
omitted, as they provide ready access to the more recent literature. By
and large, the most useful sources are Biological Abstracts, Index Medicus,
and Chemical Abstracts. Searching these sources is usually very time-
consuming; however, the task can be greatly expedited by photocopying
pertinent sections of the cumulative indices, flagging the abstracts,
and, via computer, organizing the abstract numbers in ascending order
according to volume and journal for all chemicals of interest. Using the
resulting list, a relatively untrained person should be able to photocopy
the abstracts quickly and efficiently. The photocopied abstracts should
then be reviewed by the data extractor. Original articles should be re-
trieved only if, for a given study, information required by the objective
ranking system is not given in the abstract.
d. Carcinogenicity, Mutagenicity, and Teratogenicity
Chemicals for which carcinogenic, mutagenic, or teratogenic
action have been demonstrated should be placed at or near the top of a
list of chemicals ranked in terms of toxicity, regardless of other toxi-
cological information that may have been collected. If it becomes neces-
sary to rank those chemicals that produce any one of these effects in
terms of relative potency, the operator of the overall ranking system
may devise a rating system based on dose.
A major source of information on chemical carcinogens is PHS-
149, a listing of chemicals that have been tested. Information may also
be obtained by contacting the National Cancer Institute. A limited
amount of information can be found in the Toxic Substances List. The
Environmental Mutagen Information Center (EMIC) maintains a fairly up-to-
date list of citations pertaining to chemical mutagenesis and is prob-
ably the best single source of information on mutagens. The Center does
132
-------
not abstract the original article; hence, to obtain specific information,
the full-text document must be retrieved. Teratological information may
be found in the Catalog of Teratogenic Agents (Sheppard), and up-to-date
information may be obtained by consulting "Teratology Lookout," a monthly
newsletter of the Karolinska Institute.
e. Time Required
Personnel
Literature Specialist: Experienced in searching
computerized data bases, Chemical Abstracts, Bio-
logical Abstracts, and Index Medicus. Should be
knowledgeable on journal inventory of nearby
libraries.
Biologist, B.S. or M.A.: Physiology and/or
toxicology background.
Toxicologist, Ph.D.: Working experience in mam-
malian toxicology.
Duties
The major responsibilities of the literature
specialist are:
Conduct searches of computerized data bases.
Search indices of the various abstract journals
and organize abstract numbers.
Retrieve pertinent abstracts selected by the
biologist.
The responsibilities of the biologist are:
Extract, organize, and prepare data for use in
the ranking system. Assist literature specialist
in data search (i.e., supply supervision as
needed).
The toxicologist is responsible for the overall
supervision of the task as well as for devising
suitable rating systems for the various toxico-
logical parameters.
133
-------
Time Involvement (per chemical basis)
Literature specialist: 12-15 hours
Biologist: 16-20 hours.
Toxicologist: 3-5 hours.
IV Modifications for Mathematical Model
In the recommended lists of parameters, only a few changes were
made to fit the abilities of the objective procedure to model an agent's
behavior in the environment.
In the area of release, we eliminated data on other (0) disposal,
for instance to deep wells, as being too infrequently available for use
and too difficult to model.
In the area of fate, we could not find an easy way to express the
effect of the partition coefficient in determining the retention of
agents in biological systems. However, the information could be used in
ad hoc studies. In addition, we found it necessary to add essentially
non-parametric data on transport and dilution as a component of this
area.
In the area of toxicology- the model system requires a more defini-
tive relationship between probable exposures and likely incidences of
effects than are provided by the simple numerical ranking system suggested
The difficulties in translating biological data into predictive models are
well known. However, for the purposes of priority ranking, the uncertain-
ties are less important, because any information is more useful than no
information at all. Accordingly, the systematic procedure in Appendix F
recommends that the parameters suggested in the toxicology section be
translated into dose-response curves by as crude methods as are justified.
134
-------
Appendix F
PROCEDURES FOR OBJECTIVE RANKING
135
-------
Appendix F
PROCEDURES FOR OBJECTIVE RANKING
The objective subsystem defines procedures for acquiring information
about the agent under consideration, processing that information, and using
the results to rank the agent with respect to other agents already ranked.
We take the position that these procedures should be as explicit as possi-
ble for the most commonly expected types of agents. However, the system
cannot be designed to cope with all possible types of agents; in fact, it
would not be cost effective to provide complete procedures for agents that
will rarely be encountered as STAR candidates.
Examination of the lists of agents currently proposed for STARs and
of those for future consideration reveals that most exert their effects
through chemical action on biological systems. These effects include known
and suspected effects on human health and on nonhuman fauna and flora.
Some (for example, PCBs) are suspected of having more subtle ecosystem
effects, while others have additional effects on the nonliving environ-
ment (for example, sulfates, if sulfuric acid is included). Therefore the
objective subsystem is designed with a main thread that includes charac-
terization of the agent chemically; examination of its occurrence in the
environment from natural and manmade sources; prediction of its environ-
mental transport, transformation, and fate; estimation of its toxic
effects; and valuation of those effects. Less attention is given to the
effects of chemicals on the nonliving environment, the biological effects
of nonchemical agents, and other even less frequently encountered effects.
The system for objective ranking consists of a set of procedures
called branches that are labelled A through 1. Each procedure consists
137
-------
of a series of steps. Each step is either a state-of-information test
or an information processing step, as shown in Figure F-l, and is struc-
tured as follows:
Step Test or Processing Next
Number Instructions Step Number
(In the case of state-of-information tests, the next step depends on the
answer to the test.) To avoid repetitive questions some tests have mul-
tiple answers.
When an agent gives a positive response to the state-of-information
tests, it is processed through BRANCH A, the main branch of the system,
which gathers and processes information about the environmental release,
fate, and biological effects of chemicals. Negative responses take the
operator into other branches of the system, but when the proper informa-
tion has been gathered and processed, he returns to some point further
down the main branch.
Some steps of BRANCH A direct the operator to other branches, which
each start on a new page. The direction to the "next step" consists of
a branch index (A,B,. . . Z, AA, . . .) and a step number. Major branches
often have further minor branches, but the flow is always eventually
directed back to BRANCH A.
Two special symbols are used to indicate further instructions. The
* indicates that the system operators must exert more than minimal sub-
jective judgement in answering tests or processing information. The
frequency of *'s in the procedure demonstrates that the subsystem has
substantial subjective inputs and is thus only "more objective" than the
expert system, in that explicit decisions are made. The t indicates that
additional information for answering the test or processing the information
138
-------
STATE
OF
INFORMATION
TEST
INFORMATION
PROCESSING
FIGURE F-1 GENERALIZED SYSTEM FLOW COMPONENT
is given on additional pages labeled with the step number. For example,
work sheets to be completed are so indicated.
Economy must be kept clearly in mind in carrying out the procedure.
The first information found should be accepted as definitive, unless it
is clearly suspect. Complex branches should not be entered unless the
need is clear. The procedure is designed to get only a better ranking of
priorities, not a perfect one.
In the same spirit, an attempt should be made to go through the en-
tire procedure and identify all of the questions that are not readily
answered. Some of the answers that are available may suggest that some
of the questions need not be answered. For example, if half-lives in
139
-------
water are exceedingly short, it is unnecessary to look hard for aquatic
toxicological data. All the unanswered questions (the Zls) should be
collected before ad hoc studies are authorized. The combined cost of
all jad hoc studies should not exceed $3,000.
Note: There are no Branches N, 0, P, R, T, U, W, X, or Y at present.
140
-------
Ranking Procedure
BRANCH A—MAIN FLOW
Step . Next
Number (Stary Step Number
Al Start with a trace of the procedure worksheet.t
Is the agent well defined? Yes* A2
No Bl
A2 Prepare agent identification sheett and proceed to .... A3
A3 Is the potential for environmental harm reasonably
clear? Yes* A4
No B3
A4 Complete preliminary effects checklist,! acquire
basic documents* (criteria documents, legislative/
regulatory history, reviews, data sheets). Continue
to '. . . . A5
A5 Does the agent exert its environmental effects
through its chemical properties? Yes A6
No Cl
A6 Is it a compound or well defined mixture?
Yes A7
No >D1
A7 Eatrieve CAS number if not already known. Use
CHEMLINE, TSL, TO 1C, CAS, proceed to A8
A8 Is the chemical organic? A9
elemental? Gl
meta llo-organic? HI
other inorganic? II
A9 Look for production information in (1) SOC; (2) NSF;
(3) Census; (4) CEH; (5) NCI A9
A10 Production information found? Yes All
No AA1
141
-------
BRANCH A—Continued
Step Next
Number Step Number
All Look for import and export information in
(1) FT-246+FT-410; (2) NSF; (3) CEH A12
A12 Import and export information found? Yes A13
No AB1
A13 Look for intermediate and dispersive use information
in (1) NSF; (2) CEH; (3) NCI; (4) COM . . . A14
A14 Intermediate use information found? Yes A15
No AC1
A15 Express information in kg/yr. Compute dispersive
use, (DU) , by
DU=P+I-E-IU,
where P = production, I = imports, E = exports,
and IU = intermediate usage. Check with DU
information from step A13, and adjust as
appropriate.* A16
A16 Look for information on fraction of production to air
emissions, water effluent, and land disposal of solid
waste in (1) NSF; (2) EPA reports; (3) NIOSH reports;
(4) other basic documents .......... . A17
A17 Adequate information found? Yes A18
No ADI
A18 Look for information on fraction of dispersive use to
air, water, and land in (1) NCI; (2) NSF; (3) WPPMP;
(4) basic documents A19
A19 Adequate information found Yes. A20
No AE1
142
-------
BRANCH A--Continued
Step Next
Number Step Number
A20 Express information from steps A16-A19 either
directly as discharges to air, water, and land or in
terms of e^ (fractional emissions to air), ew
(fractional emissions to water), eL (fractional
emissions to land), f^ (fraction of dispersive use
to air) , f^y (fraction of dispersive use to water) ,
and fj (fraction of dispersive use to land). In the
latter case, compute releases to air, water, and
land by multiplying the e's by P and the f's by DU.
Complete release work sheet.t A21
A21 Is there occasion to believe transportation losses
might be significant? Yes API
No A22
A22 Examine the information on production location,
dispersive uses, transportation, and other releases
to the environment.t is there reason to believe that
the releases are concentrated in localized areas?
Yes. AG1
No A23
A23 Look for information on air oxidation rate
constants in (1) Wilson; (2) ACS; (3) Doyle;
(4) CA . . . A24
A24 Rate information satisfactory? Yes A25
No Jl
A25 Compute air half-life T byt
A
0.693
T =
10" k + 2 X 10 k
OH
where k and k.. are the rate constants, in
OH 03
(Yr)~1/Mole for hydroxyl and ozone oxidation ...... . . A26
143
-------
BRANCH A--Con.ti.nued
Step Next
Number Step Number
A26 Look for information on oxidation in aqueous systems
or hydrolysis in (1) Hendry; (2) CA A27
A27 Rate information satisfactory? Yes A28
No Jl
A28 Compute water half-life TW by t
0.693
T = —
W -10
10 k + k
RO h
2
where kRQ is the rate constant, in (Yr) /mole
fraction, for alkylperoxy radical oxidation and kh
is the pseudo first order rate constant for
hydrolysis at pH ~ 7 A29
A29 Look for parachor information in Lambert A30
A30 Rate information satisfactory?* Yes A31
No Jl
A31 Compute land half-life TL from rate information.!
Is biodegradation likely? Yes. Ql
No A32
A32 Compute steady-state inventories, SSI, of the
chemical in the media from equations of the form
SSI = R T 70.693,
A A A
where RA is (for example) the total release per
year to airt A33
144
-------
BRANCH A—Continued
Step Next
Number Step Number
A33 Examine the available information to determine*
whether significant departure from steady-state
conditions are likely. Look particularly at (1)
rapid growth or curtailment of releases' in past 10
years, (2) half-lives greater than 10 years, (3)
suggestion that disappearance is not clearly related
to amount present A34
A34 Adjustments indicated? Yes Kl
No A35
A35 Examine the available information to determine*
whether intermedia transfers are likely to be
significant. Look particularly for (1) volatility
(water + land -* air) or lack of it, (2) water
solubility (air + land -• water) or lack of it, (3)
affinity for sorption on particulate material (air +
water -• land) , or lack of it A36
A36 Adjustments indicated? Yes LI
No A37
A37 Examine dispersive uses, other production sites,
methods of disposal, physical/chemical parameters,
and so on, to determine* whether any populations
(not all human, but other animate and inanimate)
are likely to be uniquely exposed.t ... A38
A38 Special populations? Yes Ml
No A39
A39 Separate the steady-state inventories of the agent
in the various media by the dilution factors shown
on the transport work sheet (A32).t Calculate
corresponding concentrations .... A40
A40 Are there any biological effects? Yes A41
No SI
145
-------
BRANCH A—Concluded
Step Next
Number Step Number
A41 Prepare a biological effects checklist.! Select* no
more than five species at risk. For man, select* no
more than three dominant effects; for other species,
no more than one A 42
A42 Effects related to a "dose" to species.
Yes A43
No El
A43 Look for dose-response relationships in (1) basic
documents; (2) limited search of literature, e.g.,
TOXLINE A44
A44 Any data? Yes A45
No Fl
A45 Express* information in terms of a graph of
incidence of effect (probability/yr) versus dose
(kg/yr or other natural units). Integratet
incidence by distribution of dose over population
of targets (human, biological, and other). For
example, if the same dose applies to all members of
a population N, and the corresponding incidence is I,
then the expected number of cases is NI. Repeat for
each effect A46
A46 Effects valued in accompanying table?t
Yes A47
No VI
A47 Multiply cases by values to obtain ranking index for
effects. Add indexes for all effects to obtain the
environmental hazard index for the chemical.t A48
A48 Compare environmental hazard index to those of
agents already ranked. Insert agent in list at
appropriate rank order, and reorder ranks of
lower-ranking agents END
146
-------
BRANCH AA--NO PRODUCTION INFORMATION
Enter from A10, 12
Step Next
Number (start) Step Number
AA1 Is the chemical known to be produced as part of a
class whose production is known? Yes AA2
No AA3
AA2 Divide production of class by number of chemicals in
class. Adjust* upward or downward if chemical is
known to be a major or minor contributor, All
respectively or
13
AA3 Is the chemical known to be in commercial production?
Yes AA4
No AA5
AA4 Set production provisionally at 10,000 kg/yr.
Reject this figure if dispersive use information All
later overweighs it or
13
AA5 Are the significant sources of the chemical
anthropogenic, even though unintentional?
Yes AA6
No AA9
AA6 Estimate* release to environmental media from known
sources, from basic documents AA7
AA7 Any problems? Yes Zl
No AA8
AA8 Complete release worksheet A20 A22
AA9 Are there controllable human activities which
influence the movement and distribution of the
chemical in the environment? Yes AA10
No B4
147
-------
BRANCH AA--Concluded
Step Next
Number Step Number
A410 Estimate* contribution of human activities to excess
inputs to the environmental media from basic
documents AA11
AA11 Any problems? Yes Zl
No AA12
AA12 Complete release worksheet A20 A22
148
-------
BRANCH AB--NO IMPORT/EXPORT1 INFORMATION
Enter from A12, 13
Step
Number
Next
Step Number
AB1 Any suggestion* that imports/exports are significant?
Yes AB2
No AB5
AB2 Estimate* imports/exports
AB3 Any problems?
Yes,
No ,
AB4 Go to
AB5 Imports/exports = 0
AB3
Zl
AB4
A13
14
A13
14
•"•Complete for imports only, exports only, or both as appropriate.
149
-------
BRANCH AC—INTERMEDIATES
Enter from A14, 14
Step
Number
AC1
AC 2
AC3
AC 4
AC5
Next
Step Number
Is the intermediate use suspected* to be:
Major AC2
Minor ACS
Negligible AC4
Unknown ACS
IU = 0.9(P+I-E)
IU = 0.3(P+I-E)
IU = 0
IU = 0.5(P + I - E)
A15
15
A15
15
A15
15
A15
15
150
-------
BRANCH AD--PRODUCTION LOSSES
Enter
from
All
Step
Number
^^MM>^^^«^»
ADI
ADZ
AD3
AD4
AD 5
AD 6
AD7
ADS
AD9
AD10
AD11
(gtar^
Is the chemical a gas or relatively volatile
(vapor pressure > 80 mm Hg)? Yes. ......
No .......
Is there information to suggest the chemical is re-
leased to the air as a particulate? Yes. ......
No .......
e - 0
A
e = 0. 05 .......
A
Examine basic documents and adjust* e downward if
£\
justified. .......
Is the chemical produced in a process that has a
significant chance of water discharges?
Yes. ......
No ......
e = 0. 05 ......
W
Examine basic documents and adjust* e downward if
W
justified. ......
e = 0 ......
e = 0.05 ......
L
Examine basic documents and adjust* e downward if
Next
Step Number
. . .AD4
. . .AD2
. . .AD4
, . . .AD3
. . .AD6
, . . .AD5
, . . .AD6
„ . . .AD7
. . . .AD9
. . . .AD8
. . . .AD10
. . . .AD10
. . . .AD11
. . . .AD18
1Use the default values from this sheet only for those media with inade-
quate information.
151
-------
BRANCH AE--DISPERSIVE RELEASE1
Enter from A19
Step
Number
Next
Step Number
•
AE1 Estimate fraction fr that accumulates in a rela-
tively inaccessible reservoir (e.g., dyes in glass
products) AE2
AE2 Is the chemical a gas or quite volatile
(vapor pressure > 120 mm Hg)? Yes AE3
No AE4
AE3 f = 1 - f AE7
A r
AE4 Is it somewhat volatile
(vapor pressure > 40 mm Hg)? Yes. . AE5
No AE6
AE5 f = 0.3 (1 - f ) AE7
A r
AE6 f = 0 AE7
A
AE7 Is it known to be disposed of in water (e.g.,, an
ingredient of soaps and detergents)? Yes AE8
No AE9
AE8 f = 1 - f - f . . .AE12
W r A
AE9 Is it known not to be disposed of in water?
Yes AE10
No AE11
AE10 f = 0 AE12
W
AE11 f = 0.5 (1 - f - fA) ......... .AE12
W r A
AE12 f = i - f _ f f _ .A20
L r A W
Use the default values from this sheet only for those media with inade-
quate information.
152
-------
BRANCH AF--TRANSPORTATION LOSSES
Enter
from
A21
Step
Number
AF1
Next
Step Number
Obtain information from TADS in EPA and from the DOT
Office of Hazardous Materials, if available. Express
the losses in kg/yr, average, to air, water, and land.
Enter information on Worksheet A22. Add to releases
calculated in A20
AF2 Any problems?
Yes.
No .
.AF2
.Zl
.AF3
AF3
Return to
.A22
153
-------
BRANCH AG--LOCALIZED RELEASES
Enter
from
A22
Step Next
Number (staru Step Number
AGl Examine information on localization with particular
attention to release into closed systems (e.g., homes,
holding ponds, dumps), in short, any system that
would sharply limit the dispersal into the general
environment. Develop a description of these limited
environments that includes (1) media involved and
3 ?
quantities (m air, 1 water, m land); (2) popula-
tions of targets (human, other living and nonliving)
in the local environment; (3) routes of escape to the
general environment. Enter data on Worksheet A22.
Save information for Step A39 AG2
AG2 Any problems? Yes Zl
No AG3
AG3 Return to A23
154
-------
BRANCH B--EARLY CLARIFICATION
Enter
from Al
Step
Number
Bl Request further definition and identification from
nominator
Next
Step Number
B2
Was clarification satisfactory?
Yes.
No .
.82
,A2
,B3
B3
B4
Enter Also from A3
Request identification from nominator. . . B4
Is the nature of the agent and its potential hazard
now clear? Yes
No
,A4
.END
155
-------
BRANCH C—NONCHEMICAL AGENTS
Enter
from A5
Step Next
Number (Start) Step Number
Cl Determine whether the agent's action is
Radiological CA1
Physical CBl
Biological . .CC1
Other C2
C Reexamine agent definition and justification for
consideration C3
C3 Same conclusions? Yes C4
No Cl
C4 Check Worksheet A2 carefully. Prepare materials on
the ranking procedure, with several diverse examples.
Identify expert in area, inside or outside EPA. Re-
quest ad hoc ranking relative to several indicator
agents. Budget #3,000 or less. Return to A48
156
-------
BRANCH CA--RADIOLOGICAL AGENTS
Enter from Cl
Step Next
Number (Start) Step Number
CA1 Define the limits of the agent carefully. For example,
is the radiation associated with a particular radionu-
clide? Is it ultraviolet radiation, but only that as-
sociated with decreases in the ozone shield? If micro-
wave radiation, what are the frequency limits? CA2
CA2 Is the radiation ionizing? Yes CAS
No CAA1
CA3 Describe the radiations, the mode of their release,
and the targets in the environment .CA4
CA4 Can dose distributions, in terms of rads or rems to
organs or the whole body, be described easily from
basic document data Yes CAS
No CABl
CAS Determine the at-risk populations and doses (rad/yr)
they face. Assume linear dose-response relationships
with slopes as given in Table.* Compute incidences
of effects and integrate ....... ,CA6
CA6 Effects valued in Table A46? Yes CA7
No VI
CA7 Multiply cases by values to obtain ranking index for
effect. Repeat for other effects. Add indexes for
all effects to obtain the environmental hazard index
for the agent. Return at A48
157
-------
BRANCH CAA--NONIONIZING RADIATION
Enter from CA2
Step
Number
CAA1
Next
Step Number
Determine whether radiation is (1) ultraviolet;
(2) visible; (3) infrared; or (4) other. Obtain
relevant literature, which should be rather limited.
fj
Attempt to quantify field intensities (joules/m ),
durations (e.g. ^ hours per year)^ and exposed popu-
lation.
CAA2 Any problems?
Yes.
No .
.CAA3
CAA3 Determine effects due to intensity distributions as
estimated. For example, there is a 50% probability
of blindness associated with Z joules/m of ruby
laser light if delivered in under 1 second. Inte-
grate incidence by distribution of dose over popula-
tion of targets. Use basic documents and collected
literature .....................
CAA4 Any problems?
CAA5 Effects valued in Table A46?
Yes.
No „
Yes.
No .
.CAA4
.Zl
.CAA5
.CAA6
.VI
CAA6 Multiply cases by values to obtain ranking index for
effects. Repeat for other effects and add indexes
for all effects to obtain the environmental hazard
index for the agent. Return to. ...
,A48
158
-------
BRANCH CAB—RADIATION DOSE DISTRIBUTION
Enter
from
CA4
Step Next
Number (start) Step Number
CABl Determine whether the radiation is
Alpha CABA1
Beta. ........ .CABB1
X or gamma. ........ .CABC1
Other. ........ .CAB2
A combination. ........ .CAB4
CAB2 Define the release and distribution of the radiation
and/or its carrier as well as possible from basic
documents. .............. ° ••>•>«•«•> ° ° .CAB3
JL,
CAB3 Can dose distributions be estimated?
Yes. ........ .CA5
No ......... .21
GAB4 Repeat from Step CABl for each radiation
involved .......... ......•««•• ° ° « • »CA5
159
-------
BRANCH CABA--ALPHA RADIATION
Enter from CABl
Step
Number
CABA1
Next
Step Number
Determine release and distribution of carrier (for
example, radon) in environment as for chemicals. De-
termine radioactive half-life (TR, yr) and energy
(E, Mev) of decay. Set relative biological effec-
tiveness (RBE) to 10. Estimate annual atoms of
intake for exposed populations, using exposure
factors in Table A45.1. Determine biological half-
life, TB, if available from basic documents or
Miller. If not available assume it is 50 years.
Compute net half-life (TN) as
N
B
CABA2 Estimate body burden (B) of carrier as
T
B = R
N
0.693
and annual disintegrations (R'), as
R' - 0.693 -a-
R R
where R is the annual rate of intake. Estimate the
annual effective dose as
1.6 x 10
-8 rad
(Mev/g)
(RBE)
R'E
m
where m is the mass (g) of the organ of concentration.1"
Use the whole body if not known. Return to. ....
,CA5
160
-------
BRANCH CABB--BETA RADIATION
Enter from CABl
Step Next
Number (start^ Step Number
CABBl Determine whether radiation is associated with a
carrier (e.g., tritium) or with some other source of
electrons (e.g., accelerators). If the latter, de-
vise ad hoc procedure for estimating dose distribu-
tion from basic documents or go to Zl. Otherwise,
determine release and distribution of carrier in
environment as for chemicals. Determine half-life
(TR, yr) and average energy (E, Mev) of decay. Set
RBE to 1. Estimate annual atoms of intake for ex-
posed populations, using exposure factors in Table
A45.1. Determine biological half-life, TB, avail-
able from basic documents or Miller. If not avail-
able, assume it is 50 years. Compute net half-life
TN as
Continue as in « CABA2
161
-------
BRANCH CABC--X AND GAMMA RADIATION
Enter from CABl
Step Next
Number (starly Step Number
CABC1 Determine whether radiation is associated with
A carrier (e.g., Cesium 137) CABC2
Other source of photons CABC4
CABC2 Determine—on the basis of modes of release and ex-
posure, affinity for body organs, and so on—whether
the carrier is likely to be significant as a source
of radiation
Externally to the body CABC4
Internally CABC3
CABC3 Set RBE equal to the fraction of dose remaining in
the organ (from Miller) CABBl
CABC4 Devise ad hoc procedure for estimating dose distri-
bution from basic documents CABC5
CABC5 Successful? Yes CABC6
No Zl
CABC6 Return to ........... ... .CA5
162
-------
BRANCH CB--PHYSICAL AGENTS
Enter
from Cl
Step Next
Number (Start) Step Number
CBl Quantify agent's discharge to environmental media.
For example, if the agent is waste heat, quantify
the energy flow (cal/yr) into air and water. If it
is particulate matter or solid waste, quantify the
kg/yr to air, water, and land. Use units appropriate"
to effects of concern. Use basic documents where pos-
-v
sible. If results unsatisfactory,' do limited liter-
ature search, or as a last resort, contact knowledge-
able people in government and industry CB2
CB2 Estimate the environmental half-life (Tg, yr) in the
media of concern. Use any information in the basic
documents, computation, limited literature survey, or
(as a last resort), expert opinion CB3
CBS Estimate dilution or dissipation factor(s) appropriate
to various media and populations affected. For ex-
ample, particulates should be assumed diluted in air
over a volume bounded by the aggregate areas of metro-
politan regions affected and the height of the mixing
layer, say 100-200 meters. Use basic documents to
determine regions affected. Units as appropriate,
e.g. (liters)"^ for water. See also Tables in A39.
CB4
CB4 Any Problems? Yes Zl
No CBS
CB5 Estimate the steady state concentration, Cs, in the
media by Cs = RD TE/0.693, where R is the release rate
in kg/yr, D is the dilution factor, and TE is the half-
life. Adjust as appropriate* if steady-state unlikely
(R not constant, Tp very long, and so on) CB6
163
-------
BRANCH CB—Concluded
Step Next
Number Step Number
CB6 Are effects related to a "dose" to target?1
Yes CB7
No .El
CB7 Dose-effect relationship known? Yes CBS
No Fl
CBS Express information in terms of a graph of incidence
of effect (probability/yr) versus dose (kg/yr or other
natural units). Integrate (A45) incidence by distri-
bution of dose over populations of targets. For ex-
ample, if the same dose applies to all members of a
population N, and the corresponding incidence is I,
then the expected number of cases is NI. Use basic
documents, or other known source of such information
CB9
CB9 Are effects valued in Table A46? Yes CB10
No VI
CB10 Multiply cases by values to obtain ranking index for
effect. Repeat for other effects. Add indexes for
all effects to obtain the environmental hazard index
for the agent. Return to A48
Relate dose to concentration through exposure factors in various media
for targets. For example, human water intake 500 1/yr, and
dose = Cwater * 500. See exposure factors in Table A45.1, or estimate
from very limited literature search. Little information generally
available.
164
-------
BRANCH CC--BIOLOGICAL AGENTS
Enter
from Cl
Step Next
Number; fstaru Step Number
•& i
CC1 Develop an ad hoc procedure, within budget limit of
$3,000, that would incorporate some of the following
ideas:
Division by class: viral, bacterial,
rickettsial, protozoan, higher forms
Reference to monitored values of concen-
tration, e.g., from STORE! for coliform
in water
Consideration of vectors for transmission
Valuation on basis of disease potential
,A48
1Enough is known about biological agents that this branch could be devel-
oped to the detail of Branch CA for a few hundred dollars. However, with
the possible exception of viruses in drinking water, EPA has little juris-
diction over such problems and the likelihood of a STAR nomination is
low.
165
-------
BRANCH D--CHEMICAL DEFINITION
Enter from A6
Step ^—^ Next
Number v*tarV Step Number
Dl Determine whether the agent is
a mixture of isomers (e.g., cresols) or
of similar compounds (e.g., long chain
fatty acids) DAI
an element1 and its important compounds
(e.g., mercury), or all compounds with
a similar functional group (e.g., sul-
fates or cyanides), or a group of chemi-
cals with similar uses (e.g., oil dis-
persants) D2
a natural product . DBl
other D2
D2 Define chemicals included in group . D3
D3 Is the group sufficiently* homogeneous that one member
can be chosen as representative Yes D4
No D5
D4 Choose* representative A7
D5 Complete Worksheet 05.* For each important member of
the group, complete Worksheets A2 and A4. Then go suc-
cessively to A7-A47; add environmental hazard indices
for all chemicals to obtain index for groups. Return
at A48
1See also Branch G.
166
-------
BRANCH DA—HOMOGENEOUS MIXTURES
Enter
from Dl
Step
Number
Next
Step Number
DAI Are there any significant differences among com-
pounds in the mixture? Yes DA2
No D4
DA 2
Define several representatives D5
167
-------
BRANCH DB—NATURAL PRODUCTS
Enter from Dl
Step
Number
DB1
Search for information on composition of product in
Merck, CTCP, primary literature
DB2 Any problems?
Yes.
No .
Next
Step Number
.DB2
.DB3
,D5
DB3 Attempt to continue from A7 using agent as a defined
but not chemically characterized mixture A7
168
-------
BRANCH E--EFFECTS UNRELATED TO DOSE
Enter from A42, CB6
Step Next
Number (staru Step Number
El Reexamine definitions to see if a dose-effects rela-
tionship could be constructed E2
E2 Same conclusions? Yes E3
No E5
E3 Define effects through known concentrations in media
and descriptions of targets and responses in basic
documents E4
E4 Any problems? Yes Zl
No E5
E5 Return to next step A43
CB7
169
-------
BRANCH F--DOSE/EFFECTS
Enter from A44, CB7, SA3
Step Next
Number (Eitart) Step Number
Fl Dose-effect relationships are generally rather diffi-
cult and expensive to develop unless quite a bit of
information is available. Devote a very limited time
to searching the literature for such information.
Otherwise, assume a linear relationship with no
threshold unless empirical or a priori information
suggests otherwise. The only parameter needed then
is the slope. This can be determined by estimating
the dose at any specified incidence, e.g., LD50. The
estimate will be more conservative as the specified
incidence approaches 1.0. Use 1.0 incidence if the
dose exceeds that for 1.0 incidence F2
F2 Return to next step A45
CBS
SA4
170
-------
BRANCH G--ELEMENTS
Enter
from A8
Step Next
Number (startt Step Number
Gl This branch applies to elements like mercury or
selenium. In most cases, the description of agents
that go by the name of elements actually imply ele-
ments and their compounds (see Dl). When possible, it
is best to separate each compound out and run it through
the appropriate branch (e.g., branch H for methylmer-
cury, Branch I for mercuric or mercurous species).
There is then no need to discriminate between the ele-
mental form and inorganic compounds G2
G2 Separation made?
(metallo-organic) Yes Hi
(inorganic) Yes II
No G3
G3 If it is infeasible to separate individual compounds
for ranking, consider the element as an entity in all
of its forms. Clearly, the element is neither created
nor degraded, and the total amount is fixed (disregard-
ing radioactive transformation). What is important, is
only the redistribution by man. This can be estimated
from MY (for production, imports, and exports), CEH
(for dispersive uses), and the basic literature--
especial ly NAS monographs and NSF trace contaminants
studies. Census and NCI may also be useful in this
regard. Fractions to various media from extraction,
processing, chemical conversions, and use can sometimes
be estimated from emissions factors for trace sub-
stances and the above sources G4
G4 Any problems? Yes G5
No G7
171
-------
BRANCH G--Concluded
Step Next
Number Step Number
G5 Attempt to go through main organic chemical branch
Steps A10, A12, A13, A15, A17, A19, And A20, branch-
ing to the default values where necessary; then re-
turn to G6
G6 Any problems? Yes Zl
No G7
G7 Perform Steps A21 and A22 G8
G8 Estimate,* from basic documents and limited litera-
ture search, the half-lives in air, water, and land
for the element. The removal processes are transfers
to relatively inaccessible reservoirs like binding to
sediments or soils, or burial in impermeable land-
fills G9
G9 Perform Steps A32-A40 GlO
G10 Complete a biological effects checklist (A41). For
agents specified only to the elemental level, it will
be difficult to apply a dose-response relationship
because of the wide range of effective levels of the
various compounds included. Use general descriptive
outlines of the element's effects (CTCP, Merck, basic
documents) to generate* a dose-response curve for all
compounds. In some cases, thresholds may be set by
looking at levels necessary for life A45
172
-------
tfKANCH H--METALLO-ORGANICS
Enter
from A 8
Step Next
Number (start) Step Number
Hi This branch was defined only to bring out the dif-
ferences between metallo-organics and the more usual
organic compounds H2
H2 Are the sources of the compound principally from
human manufacturing and use of the compound per se?
Yes A9
No H3
H3 Some metallo-organics, e.g., methylmercury, are
formed in the environment following the release of
metals. From basic documents and limited literature
search, estimate the rate constants, Xmo, for pro-
duction of the metallo-organic from the metal in air,
water, and land. Apply these to the steady state in-
ventories of metals (SSI ) as estimated in Step A32,
to obtain the release rates for the metallo-organic:
R = X SSI H4
mo mo m
H4 Any problems? Yes Zl
No H5
H5 Similarly, estimate the media half-lives of the
metallo-organic with respect to both removal to rela-
tively inaccessible reservoirs and transformation
back into metallic or other forms A32
173
-------
BRANCH I—INORGANICS
Enter from A8
Step Next
Number (startt Step Number
II This branch was defined to bring out the differences
between inorganics and organics 12
12 Production information is usually found in MY, Census,
CEH. Otherwise go to AA1 and return to 13
13 Similar instructions apply to imports and exports
(see also ABl) .14
14 Similar instructions apply to intermediate and dis-
persive uses (see also AC1) .15
15 Complete Steps A15-A20, using above sources and
Anderson (1973) ... .16
16 Complete Steps A21 and A22 .17
17 Any problems in Steps 12-16? Yes Zl
No 18
18 Examine information on transformations of the com-
pound in the environment. For example, sulfur diox-
ide passes through several stages of transformation
to become relatively innocuous neutral sulfates
(e.g., ammonium sulfate). Estimate environmental
half-lives on the basis of these transformations and
other movements to relatively inaccessible reser-
voirs A32
174
-------
BRANCH J--HALF-LIVES
Enter from A24, A27, A30, L5
Step _ Next
Number (Start) Step Number
Jl For the medium in question, determine from the basic
documents and chemical reasoning whether
the chemical is unlikely to disappear
rapidly ...................... J2
the chemical is likely to disappear
rapidly ...................... J3
there is no basis for judgment ........... J4
J2 T = 10 yr.1 .......... J5
x
J3 T = 0. 1 yr. .......... J5
X
J4 T = 1 yr.
x
J5 Return at exit point + 2 .......... A26
A29
A32
L6
1x = A, W, or L; or AW, WL, and so on.
175
-------
BRANCH K--NONSTEADY-STATE
Enter
from A34
Step _ Next
Number (start) Step Number
Kl This branch makes imprecise but potentially signifi-
cant adjustments in the "steady-state" inventory esti-
mate for conditions that depart substantially from the
steady state. The estimates are based on approxima-
tions that should lead eventually to estimates of the
average concentrations over the next 5 years. The de-
parture can be determined on the basis of several con-
ditions. Is the determination based on
Growth of releases? .......... K2
Curtailment of releases? .......... K3
Long half-life? .......... K4
Other? .......... K5
K2 Estimate the doubling time, T.^, in years, and the time
that releases have been extant, T. Select correction
factor (F) from graph.' ................... K6
K3 Estimate the time for which releases have been cur-
tailed, T, and previous release rate, R . Compute
correction factor as:1
K4 Estimate the time for which releases have been rela-
tively constant, T. Compute correction factor:
K6
K5 Is the adjustment obvious? Yes K6
No Zl
K6 Multiply SSI by CF to obtain new SSI A35
= half-life,T T or T • X = 0.693/T .
c\ W J_, X
176
-------
BRANCH L—INTERMEDIA TRANSFERS
Enter
from A36
Step
Number
LI
Estimate, from basic documents a limited literature
search, the transfer rates between the various media.
" j.
and calculate better values for the SSIs.T
Next
Step Number
,L2
L2
Any success?
Yes.
No .
L3 Use the methods described in the instructions
(see discussion Ll) to estimate the Xs from basic
considerations, and perform the calculations
,L7
.L3
,L4
L4
Any success?
Yes.
No .
,L7
.L5
L5
L6
L7
Use the default values in Branch J then return
to L6
Make the calculations,
Return to
.L7
.A37
177
-------
BRANCH M--SPECIAL POPULATIONS
Enter
from
A38
Step Next
Number (StarO Step Number
Ml Here, look for populations at risk that obtain their
exposures through means other than general exposure to
air (e.g., breathing), water (e.g., drinking), and
land (e.g., plant uptake). Some obvious examples of
special populations for mercury are fish eaters, cos-
metic users, and painters. However, the first two are
controlled by other agencies (FDA) and possibly the
third also (OSHA). We, on the other hand, are trying
to identify special populations at risk that are ex-
posed through means controllable by EPA. For example,
the PCB problem was highlighted by the poisoning of
chickens eating food contaminated by a heat-exchanger
leak. Authority over such incidents is not clear, but
conceivably belongs to EPA M2
M2 Describe, from basic documents and literature survey,
any special populations by their size (Nsp) and the
distribution of exposures by route of exposure. For
example, 2070 of Ngp might have exposures averaging
X/2, 60% at X, and 20% at 2X M3
M3 Any problems? Yes Zl
No M4
M4 Return to A39
178
-------
BRANCH Q--BIODEGRADATION
Enter
from A31
Step
Number
Ql Determine, by literature search or simple experiment,
the BOD and COD for the chemical
Q2 Estimate the biochemical half-life in years by
Next
Step Number
Q3
where COD is expressed in mg/1 oxygen per mg/1 chemi-
cal concentration, and BOD^ is expressed in the same
units (in the limit as concentration goes to zero)
Recompute the water and soil half- lives by
T T T
W W B
T T T
L L B
,Q3
.A32
179
-------
BRANCH S—NONBIOLOGICAL EFFECTS
Enter from A40
Step Next
Number (Starn Step Number
SI Determine whether the effects are
physical/chemical corrosion, abrasion,
and so on, of man-made things SA1
aesthetic in nature, affecting such prop-
erties as visibility, color, odor, taste,
texture, scale, and so on SB1
impairment of resources (e.g., excess
salinity for industrial uses of water) SCI
other S2
S2 Are the description and quantification of the effects
Yes
No Zl
obvious?* Yes S3
S3 Prepare quantitative description of effects by number
and kind S4
S4 Effects valued in Table A46? Yes S5
No VI
S5 Multiply numbers of effects by values to obtain rank-
ing indexes. Add to obtain hazard ranking index.
Return at A48
180
-------
BRANCH SA--EFFECTS ON PROPERTY
Enter from Si
Step Next
Number (start) Step Number
SA1 Construct* a model of the distribution of the items
affected by their exposure to the agent. For example,
most statuary is in big cities, and most sulfuric acid
mist is associated with urban industries. Thus the
appropriate dilution factors are relatively small.
SA2
SA2 Any problems? Yes Zl
No SA3
SA3 Is there enough information in the basic documents to
suggest a dose-effect relationship?
Yes SA4
No Fl
SA4 Express information in terms of a graph of fractional
damage versus dose (e.g., kg/m ). Integrate damage
by distribution of dose from model SA5
SA5 Does the property in question have clear economic
value? Yes SA6
No SA7
SA6 Multiply the economic value of all property by the
integrated percent damage A46
SA7 Is the value of the damaged property essentially
noneconomic, e.g., art? Yes SA9
No Zl
SA8 Return to A46
SA9 Use the percent damage/yr
,A46
181
-------
BRANCH SB—AESTHETIC EFFECTS
Enter
from
SI
Step Next
Number (staru Step Number
SB1 From basic documents or limited literature search,
establish the threshold of concentration above which
the presence of the agent is significantly distasteful
(e.g., ppm of SCK for visibility, ppm of phenol for
taste, size factor for scale) SB2
SB2 Any problems? Yes Zl
No SB3
&
SB3 Construct a model of the number of people annoyed
X times per year, as a function of X. Integrate the
people over the frequency distribution to get the num-
ber of cases per year SB4
SB4 Any problems? Yes Zl
No A46
SB5 Return to A46
182
-------
BRANCH SC--RESOURCE EFFECTS
Enter from SI
Step Next
Number fstarn Step Number
SCI From basic documents or limited literature search,
establish the threshold of concentration in the re-
source that prohibits effective use of the resource
(e.g., ppm dissolved solids in water prevent its use
as wash water) . .SC2
SC2 Any problems? Yes. ........ .Zl
No ......... .SC3
SC3 Construct a model of the percent of the resource use
denied by the distribution of concentrations. Deter-
mine the total economic value of the resource use.,
e.g., from Manufacturing Value Added (MVA) estimates
of the Census of Manufactures. Determine economic
loss by multiplying these quantities ........... . SC4
SC4 Any problems? Yes. ........ .Zl
No ......... .A46
SC5 Return to ......... .A46
1For example, Water Quality Criteria.
183
-------
BRANCH V—VALUATION
Enter from A46, CA6, CAA5, CB9, S4
Step
Numbe r
VI
V2
V3
Valuation is an inescapably subjective component of
this system. EPA must set the value of any given
effect relative to others, or at least concur with
the operator's evaluation. A table of values (see
Table A46) has been developed for some of the com-
monly incurred pollutant effects. For the effect
under consideration, determine a value per case rela-
tive to those in the table. For example, the value
of a given decrement in visibility due to particulate
matter in the air would presumably be about equal to
that for the visibility effects of NC^. Or one might
compute the economic penalties of hair loss, say, by
looking at physician diagnostic costs and the price
of hairpieces, and then comparing them with the eco-
nomic penalties of crop losses from air pollution to
arrive at a value. It will be difficult to fix a
value within an order of magnitude, but it is impor-
tant to try
Any insurmountable problems?
Yes.
No .
Add value(s) to Table A46.
at next stop
Return to former branch
Next
Step Number
.V2
.Zl
.V3
.A47
CA7
CAA 6
CB10
S5
184
-------
BRANCH Z--AD HOC STUDIES
Enter from AA7, Mil, AB3, AF2, AG2, CAA2,
CAA4, CAB3, CABC5, CB4, E4, G6, H4, 17, K5,
M3, S2, SA2, SA7, SB2, SB4, SC2, SC4, V2
Step
Number
Next
Step Number
Zl Prepare brief document describing problem area. Pre-
pare materials on appropriate ' portions of ranking pro-
cedure, with examples. Identify expert in area, inside
or outside EPA. Request ad hoc provision of informa-
tion leading to next step. Budget $1,000 or less
,Z2
Z2
Return to next step
.AA8
AA12
AB4
AF3
AG3
CAA3
CAA5
CAB 4
CABC6
CB5
E5
G7
H5
18
K6
M4
S3
SA3
SA8
SB 3
SB 5
SC3
SC5
V3
185
-------
WORKSHEETS AND EXPLANATORY MATERIAL
FOR OBJECTIVE PROCEDURE
(Keyed to the Step in the Procedure
to Which They Apply)
187
-------
AGENT NAME
BRANCH
Enter step numbers encountered in ranking agent. If sequential, use nomen-
clature like A3-A7. Use each column for a different branch. It is often
useful to place Branch A near the center.
A1 TRACE OF PROCEDURE
189
-------
A2--AGENT IDENTIFICATION
Fill in applicable sections
Agent Name
Common Synonyms
/ / Chemical / / Physical / / Radiological
/ / Biological / / Other
For chemical agents:
/__/ Compound / / Element / / Mixture
CAS Registry No.
1
Molecular Formula
2
Structural Formula :
3 o
Melting Point C
3 o
Boiling Point C
3
Vapor Pressure @20°C mmHg
3 o
Density @ 20 C Kg/1
3 o
Water Solubility, 20 C Kg/1
4
Partition Coefficient
For physical agents:
/_/ thermal / / particulate / / solid waste /_/ Other
Description/definition: ^
Temperature C (for thermal agents)
Density Kg/1 )
/ (for particulates)
Particle Size microns )
190
-------
A2--Concluded
For radiological agents:
/ / alpha / / beta / / gamma or x-ray /_/ mixed ionizing
/ / ultraviolet / / visible / / infrared and microwave
For non-ionizing': Frequency range Hz
For ionizing: Mean quantum energy Mev
For biological agents:
/ / viral / / bacterial / / rickettsial
/ / protozoan / / higher forms
Species included
Typical Size microns
Notes:
1 CHEMLINE, TADS, TSL
2 Merck, SOCMA
3 TDB, HPC, Lange, If not found, estimate to one decimal
accuracy by comparison
4 CR
191
-------
A4--EFFECTS CHECKLIST
Agent Name
Check the suspected effects
Human: / / Mortality through
Plant:
/ / Serious disease or injury
/ / Other disease or injury
/ / Physiological effects
/ / Aesthetic impact through
/ / Economic impact through
Animal: / / Mortality through
/ / Reproduction impairment through
/_/ Yield reduction through
/ / Mortality through
/ / Reproduction impairment through
/ / Yield reduction through
Other: / / Ecosystem disturbances through
Describe the principal concern:
This worksheet and worksheet A2 should be reviewed with the nominator
before major ranking steps are undertaken.
192
-------
A13--INTERMEDIATE AND DISPERSIVE USES
When a chemical is used in reactions to form other chemicals, such
use is called "intermediate use." All other uses, whether industrial,
commercial, or in consumer products, are dispersive. The distinction
becomes tenuous when the chemical is reacted before, during, or shortly
after its dispersive use, for example when adhesives polymerize and set,
or when bleaches oxidize colorants. However, these latter uses are
usually considered to be dispersive.
By definition, dispersive uses (DU) and intermediate uses (IU) ac-
count entirely for the net disappearance of a chemical:
DU+IU=P+I-E
It is the dispersive uses that are of environmental concern, and they
can be estimated either directly or by determining P, I, E, and IU.
IU is estimated by examining all the processes that use the chemical
to make others, and by estimating the consumption in each. The most com-
prehensive source of this information, even though it covers only a few
hundred compounds, is the Chemical Economics Handbook (CEH).
However, DU can be estimated directly by examining all the products
and activities that the chemical is used for. This information is also
often found in the CEH. If the major uses all fall into the following
list, then the NCI data bank is also useful: food additives, drugs,
cosmetics, soaps and detergents, paints. Dispersive use estimates can
also be found for 80 compounds in NSF, and for others in COM and the
basic documents.
193
-------
A20--RELEASE WORKSHEET
Agent Name
/ / produced commercially / / other human sources
p
I
E
IU
DU
kg/yr
kg/yr
kg/yr
kg/yr
kg/yr
production
imports
exports
intermediate uses
dispersive uses
(P + I - E - IU l_l, or directly / /)
Releases during production Releases during use
(fractional) (fractional)
Water
Land
Release Rate Computation
Trans-
Production Use portation Other Total
kg/yr
kg/yr
kg/yr
Derivation of Estimates (Document with brief narrative)
Production Processes
Air P x e =
A
Water P x e =
Land P x e =
L
+ DU x f + +
A
+ DU x f + +
+ DU x f + +
L
Includes combustion products, release during other activities as waste,
and so on. May be available in basic documents or calculable from EF.
194
-------
A20--Concluded
Intermediate Uses
Dispersive Uses
Transportation Releases
Other Releases
195
-------
A22--LOCATION OF POTENTIAL RELEASES TO ENVIRONMENT
Production
Firm
Location
Quantity
Produced
per Year
% of Total
Consumption (Nondispersive Uses)
Firm/Industry
Location
Quantity Used
per Year
of Total
Produced
Consumption (Dispersive Uses)
Firm/Industry
Location
Quantity Used
per Year
7o of Total
Produced
196
-------
A22—Continued
Transportation
Quantity
Type Carried ?„ of Total
Firm/Industry of Carrier per Year Produced
Storage
Mean
Mean Quantity Residence
Firm/Facility Container Stored Time
Natural Sources
Quantity
Phenomenon/ Released
Ecosystem Location per Year
197
-------
A22--Concluded
Natural Sinks
Quantity
Phenomenon/ Absorbed
Ecosystem Location per Year
Complete outline maps if appropriate. See The National Atlas of the
United States (U.S.G.S.) for outline maps of administrative subdivisions
for individual federal agencies if necessary.
198
-------
A25--AIR OXIDATION
Oxidation and/or the photochemical degradation are the major sources
of degradation in the atmosphere. The dominant fate is oxidation, with
photochemistry usually resulting in oxidation products. Direct photolysis
is usually of minimal consideration, because most compounds absorb below
the 300 nm solar region cut-off. In the evaluation of oxidation under
environmental conditions for NSF-RANN, SRI considered ozone (Oo) and the
hydroxyl radical ( *OH) as the primary air oxidants; both have been demon-
strated to be important in air pollution modeling. Although concentrations
of these species are subject to environmental conditions (sunlight in-
tensity, other pollutants present); they can be estimated correctly
within an order of magnitude that is useful for kinetic predictions.
Ozone is formed by photochemical processes in nature and is also
derived from direct and indirect anthropogenic sources. The harmful
effects and reactivity of ozone are well demomstrated. Data on the rates
of reactions of ozone are reasonably available for calculations. We as-
_9
sign an ambient concentration of 0.05 ppm (about 2 x 10 )M).
The hydroxyl radical is a very reactive species and is of pivotal
importance in all air pollution modeling systems. Ambient concentra-
tions of 10° M are assigned for these calculations.
In air, the oxidation of a chemical may then be represented as
=
dt *OH °3
r l
L 3J
[chemical]
The values of k~TT and k (M sec" ) are based on available literature
UH O^
information, either for the compound directly, or by analogy to other
appropriate compounds for which data are available. A half- life (sec-
onds) in the atmosphere may then be approximated by:
199
-------
A25—Concluded
0.693
T = [OH] + k
OH 3
The half-lives should then be converted from seconds to years.
200
-------
A28--WATER DEGRADATION
Both oxidation and hydrolysis may degrade a chemical in the aquatic
environment. Oxidation in this phase was estimated for the NSF-RANN
study by assuming alkyl peroxyl radicals (RO *) as the active oxidant
species, because these radicals are readily regenerated in the presence
of oxygen. A concentration of 10"^ M was assigned, using a solar flux
-7 2
value of 2.2 x 10 einsteins/cm sec, with acetone as a representative
o
photosensitizer, and a product quantum yield of about 10. With this
concentration and the large amount of kinetic data on alkyl peroxyl
radical reactions, the oxidation in water can be estimated.
Hydrolytic degradation of a compound may be accelerated by both
acidic and basic conditions. For most compounds, a minimal rate of
hydrolysis will occur at pH ~ 1, and we suggest referencing all data to
this value. Extrapolation of data to this pH is reasonable, and some
data are available in the literature for this pH. However, no compre-
hensive review of literature data on hydrolysis is available at this
time. Although there is little information directly relevant to an en-
vironmental assessment, there is a large amount of data on hydrolysis
at higher temperatures and in various solvent systems. These data may
be extrapolated or rendered useful through use of various empirical and
theoretical calculation techniques currently accepted by physical organic
chemists.
As with the air oxidation degradation, the disappearance of a com-
pound in the aquatic environment may then be represented by
[chemical]
where k^Q is a biomolecular rate constant (M~ sec ) and k is the
pseudo-first order rate Constant (pH = 7) in sec"1.
201
d (chemical)
dt
kR02-
[*
cy] +
kh
-------
A28--Concluded
An aquatic half-life in seconds may then be calculated by
= 0.693
w k rR(vi+ ^
RVl 2J ^
The half-lives should then be converted from seconds to years.
202
-------
A31--COMPUTATION OF LAND HALF-LIVES
Simple and wholly objective predictors of residence times in soils
are either unavailable, or, as in the case of Lambert's use of parachor,
untested for many classes of agents. Hence, informed speculation remains
the only consistently applicable method for predicting residence times
in soils.
However, objective estimates can be obtained for selected groups of
chemicals. For example, sorption rates and parachor are strongly cor-
related for pesticides, and this correlation, if calibrated against
chemicals of known residence times, potentially could be used to obtain
an estimate of residence times.
Clearly, the more rapid the sorption, the longer the residence time
in unavailable forms is likely to be, and conversely the shorter the
residence time is in soluble forms.
2 3
With the surface tension, Y, in g/sec and the density P, in g/cm ,
the parachor, P, is given by
P = M Y1/4/p
where M is the molecular weight.
The equilibrium constant, K , between liquid and solid phase is
then estimated by
0.0125 P
K = 0.2e
e
The half-life for movement from the liquid (£,) to the solid (s)
phase is related to that for movement back into the liquid phase by
203
-------
A31--Concluded
T. = T ./K
1-* s s-» & e
Assuming that the concentration in the solid phase never reaches equili-
brium with that in the liquid phase, the net half-life in the liquid
phase is given by
T = -
L K " 1
e
We finally assume that T -*& is very Ipng, arbitrarily 50 years.
Therefore, T in years is given by
Li
50
T = K - 1
11 yy
Since only the liquid phase is environmentally available, T is the land
Li
half-life.
204
-------
A32--TRANSPORT/TRANSFORMATION WORKSHEET
Agent Name
1. Chemical Transformation (Steps A25-A31)
Air: k_ (Yr)" /mole kn (Yr)" /mole T Yr
OH "
Water: kRO
Soil: P
- j A
(Yr)" /mole k, (Yr)" /mole T,, Yr
h • " W ' '
K T Yr
2. Intermedia Transfer (Step A35)
Transfer rates, in (Yr) } from column heading to row heading
Air Water Land
Air
Water
Land
3. Steady State Inventory (Step A32)
Solve equations shown in discussion Ll
R.
A
R
kg/yr
kg/yr
kg/yr
SSI
SSI
W
SSI
kg
kg
kg
4. Non-steady State Correction (Step A33)
CFA
A
CFTT
W
CF
L
5. River/Lake Partition
Lakes f
Rivers f
SSI.
^
SSI
1 • W
SSI
L
(see discussion A39)
SSI
SSI
kg
kg
kg
kg
kg
205
-------
A32—Concluded
6. Concentrations (Step A39)
Dilution Factors, D Fraction of SSI, fQ Concentration1
t o o
Air 10 m kg/m
10 m kg/m
in10 3 i / 3
10 m kg/m
in12 3 i / 3
10 m kg/m
14 3 3
10 m kg/m
In16 3 3
10 m kg/m
Rivers 10 1 kg/1
1012 1 kg/1
1013 1 kg/1
1014 1 kg/1
1015 1 kg/1
Lakes 3 x 1Q6 1 kg/1
3 x 108 1 kg/1
3 x 1010 1 kg/1
3 x 1012 1 kg/1
3 x 1014 1 kg/1
3 x 1016 1 kg/1
f\ ? 9
Land 5 x 10 m kg/m
Q O O
5 x 10 m kg/m
10 2 2
5 x 10 m kg/m
5 x 10 m kg/m
If release rate and/or transformation rate data are inadequate, use
ambient concentrations observed, from SAROAD and STORET.
206
-------
A37--POPULATIONS AT RISK
Human
Estimate Size
Name of Group of Group
Geographic groups
(e.g., Northeastern U.S. central
city dwellers)
• Occupational groups1
(e.g., farmers)
• Avocational groups
(e.g., fishermen)
• Dietary groups
(e.g., Weight-Watchers)
Other
(e.g., socio-economic groups)
Prime responsibility generally is that of other agencies.
207
-------
AST—Concluded
Domestic or Captive Nonhuman
Name
Number at
Location Risk
• Livestock
(e.g., dairy cattle;
minks)
• Pets
(e.g., cats; gold-
fish)
» Captives
(e.g., lions, bears)
Wild Nonhuman
• Widespread rare or
endangered1 species
Geographically isolated
rare or endangered
species
•"•Endangered either by the agent of interest or by other means (e.g.,
changes in land use).
208
-------
A39--DILUTION FACTORS
Air
We assume that a typical diffusion and transport velocity is about
1 m/sec or 3 x 10 m/yr horizontally, and 10 m/yr vertically. We also
1 (\ "3
assume that the agent has about 10 m available for expansion (this
limits the expansion to about a 1 km layer over the United States). Thus
a puff of agent released would fill this volume within about 1/10 yr.
If the agent remained in the air a relatively long time (>1 yr), a
steady state would result in a relatively uniform distribution of concen-
trations, with only a slight peaking near points of release; this is be-
cause the total inventory would be large in comparison with the release
rate. For shorter half-lives, the gradient of concentrations away from
the release point would be larger, because the agent would not exist long
enough to diffuse to the limits.
A rough calculation results in Table A39.1, which shows the percent-
age of the steady-state inventory that is in various dilution volumes as
a function of half-life. Interpolation on log-log paper is permissible.
Table A39.1
PERCENTAGE OF SSI IN VOLUME INDICATED (AIR)
Half-Life (yr)
Dilution Fact
(m3)
10
10
10
10
10
10
6
8
10
12
14
16
-3
10
5 x 10"4
-2
4 x 10
2.1
37
60
1
1(
6 x
6 x
5 x
4 x
13
86
10- 7
-5
10
10- 3
-i
i fl-
ic
3 x
3 x
2 x
2 x
2
98
f1
10"
10
i fl-
ic
7
5
3
1
10°
1 x lO"8
-6
1 x 10
1 x 10*
-2
1 x 10
1
99
io1
1 x
1 X
1 X
1 X
I
99
10
10
10
10
-8
-6
-4
-2
209
-------
A39—Continued
Water
The same general principles hold for water as for air, but the likely
transport rates and dilution volumes are quite different. All agents re-
leased into water can eventually reach the ocean, but they are relatively
unlikely to move from one river system to another, for example.
The total flow of all rivers in the United States is about 2 x 10
1/yr; in a case of uniform steady state for long-lived agents, this would
be the dilution factor because all of the annual input would reach the
ocean. We assume that the velocity of turbulent diffusion is about 100
m/hr, or about 10 m/yr, and that river dimensions are typically 100 m
by 10 m. Therefore uniform mixing across the rivers occurs within hours,
and only extremely short-lived agents (T < 10" yr) would not be mixed.
On the other hand, flows are typically 5 km/hr or 5 x 10 m/yr, and river
lengths are typically 1,000 km (10 m). Consequently, a similar table
(A39.2) can be constructed for water dilution factors.
Table A39.2
PERCENTAGE OF SSI IN VOLUME INDICATED (RIVER WATERS)
Half-Life (yr)
uiiution race
(1)
lo11
12
10
13
10
14
10
15
10
-3
10
1 x 10"2
1
12
62
25
io-2
1 x IO"2
-2
9 x 10
1.3
16
82
10- L
1 x IO"2
-2
9 x 10
_ i
9 x 10
10
89
10°
1 x
9 x
9 x
9
90
io-2
-2
10
_1
10
210
-------
A39--Continued
Note that rivers of the size assumed here would empty 50 times per
13
year, for a total volume of 5 x 10 1. It would take 400 rivers of such
size to supply the total runoff indicated above. If it is suspected that
only N rivers are significantly contaminated, then the dilution factors
on Table A39.2 should each be multiplied by N/400 before further analyses
proceed.
The total volume of all lakes and ponds in the United States, includ-
ing all of the Great Lakes, is about 3 x 10 1. This represents an area
11 2
of about 4 x 10 m times an average depth of about 75 m. We estimate
that mixing occurs at a rate of about 10 m/hr horizontally and about 100
m/yr vertically. Since few lakes have dimensions greater than 100 km
wide or 100 m deep, complete mixing usually takes place within a year.
Transport in lakes is more like transport in air than it is like trans-
port in rivers, so we use a scaled version of Table A39.1 for lakes (see
Table A39.3).
Table A39.3
PERCENTAGE OF SSI IN VOLUME INDICATED (LAKE WATERS)
Half-Life (yr)
U1JLU1
3
3
3
3
3
3
:ion
(1)
x 10
x 10
x 10
x 10
x 10
x 10
Facto
6
8
10
12
14
16
-2
10
-4
5 x 10
4 x 10"2
2.1
37
60
1
6
6
5
4
13
86
-1
10
x 10"
x 10"
x 10"
x 10"
7
5
3
1
3
3
2
2
2
98
0
10
x 10
-7
x 10'5
x 10
x 10
-3
-1
1
1
1
1
1
99
1
10
-8
x 10
x 10'6
x 10~4
x 10~2
2
10"
1 x
1 x
1 x
1 x
1
99
-8
10
10- 6
io"4
10- 2
211
-------
A39--Concluded
If it is known that discharges occur only into lakes of total vol-
ume V, then the dilution factors should be scaled by V/3 x 10 and the
half-lives by (V/3 x 1Q16) .
Sometimes the relative discharges into lakes and rivers are known,
and the fraction, f, of the inventory in each can be assumed to be the
same. If this division is not known, assume it is 50-50.
212
-------
A41--BIOLOGICAL EFFECTS CHECKLIST
Agent Name
Species at risk: (select no more than 5)
rj Man
/ / Domestic animals
LJ
— Other animals
n
n
I I Crops
n
n
I~~J Other plants
AT
n
Effects to be considered: (select no more than 3 for men, 1 each for
other species)
Indicator
Species Effects Test3
Man
Selection should be based on suspected dominance of value once
ranking is complete.
O
For example, LD-Q (mouse), human epidemiology, or TLV.
213
-------
A45--DOSE-EFFECTS WORKSHEET
Agent Name
Effect
in species
Dose-response relationship:
D,
(units)
Justification:
Incidence, fractional, of effect at dose shown. Enter scale as
required (maximum 1.0). Excess over background is implied.
214
-------
A45--Concluded
N.
Number
at Risk
Concentration
/Units
Exposure2
Factor/Units
D=CxE
Dose
/Units
D
Converted
Dose/Units
Incidence
Cases
Ul
Total Cases
In some cases it may be simpler to express the exposure dose in-
formation as a distribution, dN/dD (number per unit dose) as a function
of D. The number of cases is then given by
NC =
Saturation
Threshold
(D) I (D) dD
1Absolute number or percent of total population (show with %). See also high dose distribution
discussion, A45.2,
2See Table 45,1^ BOB, Altman, Prosser, or Dill.
-------
A45--Concluded
Table A45.1
EXPOSURE FACTORS
l
Food Intake, Water Intake,3 Air Intake/
Species
Man
Monkey
Dog
Cat
Cattle
Horse
Sheep
Rabbit
Rat
Mouse
Chicken
Weight, kg
65
5
10
2
500
500
60
2
0,4
0,02
2
kg/yr
550
90
275
37
2/750
3,650
875
22
7
1
90
1/yr
450
__
350
17,000
8,000
1,000
125
12
2.5
_ .
m3/yr
7,000
850
2,000
800
60,000
35,000
8,000
450
280
35
600
"Use these values with caution, they are subject to many caveats.
Although they are adequate for the purposes of this ranking, they
should not be used for more demanding tasks, such as the
development: of the STARS,,
'Drinking, doesn't Include watei: in food or water of metabolism.
3Tldal Volume—not all is taken in by far (about 8% of the
oxygen in the tidal volume is).
216
-------
A45.2--HIGH DOSE DISTRIBUTION
Because the procedure in Step A39 computes single concentration
value for discrete populations, rather than a continuous distribution,
the highest concentration is often below the threshold for toxic effects.
Consequently, the hazard index is estimated as zero, and no discrimina-
tion is made between agents with concentrations near the threshold and
those with concentrations far below it. (When a zero threshold is as-
sumed, this problem does not arise.)
If the highest predicted concentration does, in fact, fall below
the threshold for a given effect, the possibility that smaller numbers
of the exposed population are receiving even higher doses should be
explored. This could occur, for example, during rare but significant
incidents of accidental exposure, such as chemical spills.
If such a possibility exists, the suggested procedure is to assume
that the number of people (or other organisms at risk) varies inversely
with doses above the highest dose predicted from Step A39. Let the
highest predicted dose by D , and the number exposed to that dose be N .
Then the number exposed to a higher dose D is estimated by
N = N D /D
ih h
If the threshold dose is D , choose two or three doses higher than
D at which to estimate N, Ordinarily, it is convenient to choose doses
spaced by a factor of 10, following the pattern of Step A39, then com-
bine with these dose/number combinations as in Step A45« In some cases
it may be desirable to extrapolate from some other high A39 dose D / and
compare the resulting N's for the same D's with those from the first
extrapolation.
217
-------
A46--VALUATION
Valuation is surely the most controversial part of an "objective" rank-
ing system. We recommend that the values suggested here be examined closely
by the operator and replaced where necessary. When in doubt, complete a sen-
sitivity analysis.
Effect
Examples
Units
Excess human mortality Terminal cancer, acute Deaths/yr
poisoning
Excess human morbidity Cancer, heart disease,
Serious disease chronic respiratory
disease, chronic kid-
ney or liver disease
Excess human morbidity Acute respiratory
Other disease
Life shortening
Physiological effects
of uncertain signifi-
cance
Aesthetic annoyance
Economic losses
disease, hay fever
dermatitis
Lower life expectancy
in irradiated popula-
tion
Proteinurea
Observation of solid
waste in water., odor
annoyance
Excess painting, loss
of industrial produc-
tion
Mortality in domestic1 LD
animals LC
for chickens,
Morbidity in domestic
animals
Mortality in other
animals
LCcQ for tuna
Molybdenosis
Bald eagle,
neomysis shrimp
Loss of yield in crops Corn, pines
Cases/yr
•Cases/yr
Yr/yr
Occurrences/yr
Occurrences per
person/yr
Dollars/yr
7> of population
per year
7o of population
per year
7, of population
per year
7> of yearly
harvest
Value/
Unit
1,000
200
10
50
.01
1, OOO2
1,000"
10, 000"
218
-------
A46--Concluded
Value/
Effect Examples Units Unit
o
Mortality in other Seaweed 70 of population 1,000
plants per year
Effects on sensitive Eggshell thinning Occurrences/yr 10
indicator species
1 Supplying major human food needs or significant useful work.
If over 50°/o, increase value by factor of 3.
If over 507°, increase value by factor of 10.
-------
A47--HAZARD RANKING WORKSHEET
(Page of )
Agent Name
Ranking Index
1. Effect 2. Cases/Units 3. Value (2. x 3.)
Page Total (Environmental Hazard Index)
Sum of pages (Total Environmental Hazard Index)
Date Rank2
Applicable only to last page of multi-page forms.
'Enter new rank each time it changes as a result of new entries to the
list.
220
-------
CA5--RADIATION DOSE-EFFECTS RELATIONSHIPS
The following relationships have been used for predicting long-
term human health effects of ionizing radiation.
Genetic Deaths
N , = 0.19 bl D/100
where N is the number of genetic deaths/yr
b, is the number of births/yr
and D is the annual absorbed dose in rad for the reproducing population.
Induced Neoplasms
* -fi D
N = N x 10 x ^- xP
d
where N is the number of neoplasms/yr
N is the number of spontaneous neoplasms per year per million
D is the dose in rads/yr
D, is the doubling dose in rads
and P is the population.
For thyroid neoplasms only,
, D. . D + D. "I
N = F N " -e + x + F N -^ ~ P x 10~6
A D D.
* e + i
F° N° Ddo
* Do + Di "
+ F N e 1
-J- £ j\]
y y D,
dy
where F and F are the fractions of the population over and
under 20 years of age, D and D. are the external and internal
doses, and the other parameters are as given below.
221
-------
CA5—Concluded
Neoplasm
Thyroid, o
y
Respiratory
system
Digestive system
Breast
Lymphatic/
Hematopoietic
Leukemia
All
*
N
40
4
294
482
143
83
72
1,500
Dd
100
10
175
230
100
70
50
175
Life Shortening
10 days per rad.
222
-------
CABA2--ORGAN MASS FOR HUMANS
(kg)
Organ Newborn 1 3 10 Adult
GI Tract
Bone
Liver
Kidneys
Thyroid
Total Body
.13
.20
.13
.02
.002
3.5
.31
.65
.32
.06
.002
11.0
.60
1.17
.48
.08
.004
15.0
.85
2.56
.83
.15
.009
32.0
1.64
7.00
1.60
.28
.020
65.0
223
-------
D5--MULTIPLE AGENTS
If agent is represented by several compounds, list them here.
Agent as initially designated or defined
Representatives Reason for Choice
224
-------
K2 CORRECTION FACTORS FOR NON-STEADY-STATE CONDITIONS
I I I I I I I I
225
-------
L1--INTERMEDIA TRANSFERS
Physical transfers between two media can markedly affect the steady-
state inventories and the concentrations to be expected in each. At first
order, we assume that transfer rates are linearly related to inventories
in the original medium through rate constants X , . For example. \
mm AW
would be the transfer rate constant from air to water, whereas X would
WA
be the constant for water to air (evaporation). (The transfer rate con-
stants are inversely related to the characteristic half- life for the pro-
cess: X = 0.693/T , and so on.)
The most common transfers are evaporation (water or land to air),
deposition (air to water or land), and sedimentation (water to sediments).
In this subsystem, both land and sediments are considered to be parts of
a single, relatively inaccessible, reservoir or sink for pollutants.
Evaporation can be estimated by the techniques of Mackay and Wolkoff
(1973) if very sweeping assumptions are made. The basic equation sug-
gested is
M P
is
where \ is in (yr) , M. is the molecular weight, C. is the solubility
in water (mg/£), and P. is the vapor pressure (mmHg at 20°C). This equa-
tion could also be used for land to air evaporation, but with even less
certainty. Usually land to air evaporation is ignored, but not always
with justification.
Sedimentation is not well understood, but probably varies linearly
with solubility, all other things being equal. The equation suggested is
226
-------
L1--Concluded
In general, deposition rates from air are also difficult to estimate.
Gases can be assumed not to deposit, although they can, in fact, adsorb
on soil or dissolve in water. Most solid particulates of size greater
than 10 microns can be assumed to deposit rather quickly (T or T of
AW AL
the order of a few hours to a few days). Aerosols and vapors would be
intermediate in rate of transfer.
For very low vapor pressure (less than 0.01 mmHg), assume
X = 2 (washout). For vapor pressures between 0.01 and 1, assume
AJ_i
X =0.7. For vapor pressures between 1 and 100, assume X = 0,07,
AJ.J Aij
For higher vapor pressures, X =0.
Aij
Because oceanic pollution is disregarded, the significant transfer
is assumed to be air to land, and X can usually be ignored.
Once the T's and the corresponding X's have been estimated, the sig-
nificant terms of the following three equations should be written down
and solved for the SSI's:
xA + xAT7 + XSSIA - X...SST. - XT ssi = R
A AW AL/ A WA W LA L £
- X SSI - X SSI + (X + X + X ISSI = R
AL A WL W \ L LA LW/ L L
Use these values instead of those calculated with the simpler equations
of Step A32.
227
-------
Appendix G
SAMPLE RANKING RESULTS
FOR CARBON DISULFIDE AND CYANIDES
229
-------
Carbon Bisulfide
231
-------
A2 - AGENT IDENTIFICATION
Fill in applicable sections
Agent Name
Common Synonyms C-ty&Bf^ &I
Chemical / / Physical / / Radiological
/ / Biological / / Other
For chemical agents:
/y/ Compound / / Element /_/ Mixture
CAS Registry No.
Molecular Formula _ (2, S _______
2
Structural Formula :
S *• c =
Melting Point - I J0*% °C
Boiling Point & bt, 3 c
3
Vapor Pressure @20°C ^v 3^^ mmHg
Density @ 20°C /»«^4 Kg/1
Water Solubility, 20°C^l^/e> Kg/1
4
Partition Coefficient jj0_0_\_l_
For physical agents:
/ / thermal / / particulate / / solid waste j^J other
Description/definition: _
Temperature C (for thermal agents)
Density
„ (for particulates)
Particle Size _ mic
233
-------
A4 - EFFECTS CHECKLIST
Agent Name
Check the suspected effects
Human:
Animal
Plant:
Mortality through
Serious disease or injury
/X/ Other disease or injury
/ / Physiological effects
Z>/5£45£r
/ / Aesthetic impact through
/ / Economic impact through
/X/ Mortality through ACUTE
/ / Reproduction impairment through
/_/ Yield reduction through
/ / Mortality through
/ / Reproduction impairment through
/ / Yield reduction through
Other: / / Ecosystem disturbances through
Describe the principal concern: fc&CEA/T
THAT '
/M-S
This worksheet and worksheet A2 should be reviewed with the
nominator before major ranking steps are undertaken.
234
-------
A20 - RELEASE WORKSHEET
Agent Name
/X/ produced commercially
other human sources
p
I
E
IU
DU
y to
_Kg/yr
_Kg/yr
_Kg/yr
_Kg/yr
KG/yr
production
imports
exports
intermediate uses
dispersive uses
(P + I - E - IU /^7, or directly
Releases during production
(fractional)
Air
Water
Land
ew
eL
Releases during use
(fractional)
production
Release Rate Computation
Trans-
Use portation Other
Air p x e
Water P x e
Land P x 6
/.f
+ DU x
+ DU x
+ DU x f
+
+
+
+
Derivation of Estimates (Document with brief narrative)
Production Processes
Total
Kg/yr
Kg/yr
Includes combustion products, release during other activities as
waste, etc. May be available in basic documents or calculable f •* 01 .•> KF,
235
-------
A20 (concluded)
Intermediate Uses
TET£ACJ4J.ORtT>&,
Dispersive Uses
WOOD
Transportation Releases
Other Releases
AMD
CATAL-jric, CWVt&TEm tPG&MPs fef~} AvD //(/ Ttffr
236
-------
A32 - TRANSPORT/TRANSFORMATION WORKSHEET
Agent Name &/}-/Q& S(^^F/J> E"
1. Chemical Transformation (Steps A25-A31)
V _ Wr/-*- V (Yr)-W.
3 —
Water: KRQ^ _ (Yr)~1/mole K^ _ (Yr)~1/mole TW _ Yr
Soil: P _ Ke ___ TL _ Yr
2. Intermedia Transfer (Step A35)
Transfer rates, in (Yr) , from column heading to row heading
Air Water Land
Air J.^ V,
Water
Land
3. Steady State Inventory (Step A32)
Solve [ft] •'• [i\] 4GDH Equations shown in discussion LI
Kg/yr
Rs -
4. Non-steady State Correction (Step A33)
CFA _ SSIA _ Kg
CFW __ _ SSIW _ Kg
SSI _ Kg
5. River/Lake Partition (See discussion A39)
Lakes f '^ SSI 9.>
-------
6. Concentrations (Step A39)
Dilution Factors, D Fraction of SSI, f_ Concentration
6 3 ~'f 3
io10
Rivers IO11 1
3 x IO12 1
3 x IO14 1
.16
Air 10 m fx/0 _ J.vlTX/0 Kg/m
io8 m3 >Tx
IO12 m3 3 y/£>° _ ,3./ V/
-------
A41 - BIOLOGICAL EFFECTS CHECKLIST
Agent Name
Species at risk: (select no more than 5)
Man
/ / Domestic animals
LJ
— Other animals
n
n
I I Crops
n
LJ
I~~T Other plants
rj
rr
Effects to be considered: (select no more than 3 for men, 1 each for
other species)
Indicator
Species Effects Test2
Man
Selection should be based on suspected dominance of value once
ranking is complete.
2
For example, LDcQ (mouse), human epidemiology, or TLV.
239
-------
A45 - DOSE-EFFECTS WORKSHEET
Agent Name £X4/£g
-------
N.
Number
at Risk2
^00
10
fo
/o
Concentration
/Units
Exposure
Factor/Units
A45 - (concluded)
D=CxE
Dose
/Units
D'
Converted
Dose/Units
-Off
-AS
m,
10
i.
Incidence
Total Cases
In some cases it may be simpler to express the exposure dose in-
formation as a distribution, dN/dD (number per unit dose) as a function
of D. The number of cases is then given by
Saturation
NC = / ||- (D) I (D) dD
Threshold
NC.
Cases
.24
Absolute number or percent of total population (show with %). See also high dose distribution discussion, A45-2.
3 See Table 45.1, BOB, Altman, Prosser, or Dill.
-------
Carbon Bisulfide--High Dose Distribution
Although the computed concentrations for CS~ are below the threshold
for cardiovascular effects even for the smallest populations, there could
be hot spots that might lead to increased risk. Let us assume that the
number of people exposed to higher concentrations varies inversely with
the concentration, starting with the 2 x 10 concentration of .007 ppm.
The results are shown in the lower part of Worksheet A45.
Aes the tics_
The odor threshold for CS2 appears to be around .025 ppm. This is
only slightly higher than peak concentrations predicted on a year-round
basis, so that odor episodes are likely. Let us assume that the smell
becomes noticeable 10 times per year near (< 2.5 km) Glaus Plants and
rayon, rubber, and chemicals manufacturing. There are approximately
250 such operations. Although some will be remote, most are probably
close to urbanized areas. Each might affect some 8 square miles at per-
haps 200 persons/square mile, or 8 x 200 x 250 x 10 = 4 x 10 instances
per year. There might be additional instances associated with large
stationary sources equipped with catalysts. It should be remembered
that these sources are also sources of H^S, although the latter disap-
pears more rapidly than CS~.
242
-------
A47 - HAZARD RANKING WORKSHEET (page / of _J_ )
Agent Name CsA&QCAJ
i Fffpr1- 2- Cases/ Ranking
•"•' trrecc Units 3. Value Index (2.x3.)
£4£EVl//4.$£^44/e /ff)A*J > L / O
Lf- v / 0* / J4 X /
Page Total (Environmental Hazard Index) */ V / V
Sum of / pages (Total Environmental Hazard Index) if. y / &
**
Date Rank
* Applicable only to last page of multi-page forms.
** Enter new rank each time it changes as a result of new entries
to the list.
243
-------
TRACE OF PROCEDURE
Agent Name
Branch
Enter step numbers encountered in ranking agent. If sequential, use nomen-
clature like A3-A7. Use each column for a different branch. It is often
useful to place Branch A near the center.
244
-------
Carbon Disulfide--Narrative
Carbon disulfide is produced in large quantities both for inter-
mediate uses (rayon and cellophane, carbon tetrachloride, other organic
chemicals) and dispersive ones (solvents, fumigants, corrosion inhibitors)
It is also produced in reduction of carbonaceous sulfur, as in Glaus
sulfur recovery plants, and possibly in automotive catalytic converters
and (in the future) stationary source converters. Most of the CS will
enter the air.
In the environment, CS~ degrades to carbonyl sulfide and other prod-
ucts through oxidation by free oxygen atoms in the air and (possibly)
photooxidation. This occurs within a few days. Even more rapid trans-
fers occur from water and land to air, and the air is the principal res-
ervoir .
A variety of toxic effects have been observed for CS^ at high con-
centrations (> 20 ppm), but at lower ones the most prominent health ef-
fect is increased risk of cardiovascular disease. The aesthetic impacts
of CS9 may be considerable, however, because it has an unpleasant odor at
very low concentrations.
If the value placed on aesthetic effects is comparable to that placed
on health and ecological impacts, carbon disulfide ranks near or at the
top of the candidate list. If aesthetic effects are ignored, it becomes,
at best, of average concern.
Carbon Disulfide--Procedural Difficulties
Carbon disulfide presented relatively little procedural difficulty
as our comprehensive study provided most of the necessary data. Diffi-
culties occurred in:
245
-------
• Not discovering the catalytic conversion source until it
was pointed out after the first draft of the CS?/COS paper.
• Needing to make a dose-distribution assumption to estimate
any health effects.
• Making subjective estimates of odor instances.
Carbon Disulfide--Recommendations
Additional sensitivif" studies on the value of aesthetic impacts
are suggested.
246
-------
Cyanides
247
-------
D5 - MULTIPLE AGENTS
If Agent is represented by several compounds, list them here.
Agent as initially designated or defined C yA*JI
Representatives Reason for Choice
TO
&IP-
t o/J
249
-------
A2 - AGENT IDENTIFICATION
Fill in applicable sections
Agent Name
Common Synonyms /-/ff A?gC.,4vC//C-
Chemical / / Physical /_/ Radiological
/ / Biological / / Other
For chemical agents:
/// Compound / / Element / / Mixture
CAS Registry No.
Molecular Formula
2
Structural Formula :
Melting Point — 73. «3^ _ °c
3 „_
Boiling Point ^4, ,
3
Vapor Pressure @20°C <-^
3 o
Density @ 20 C £> t 4
3 o
Water Solubility, 20 C
Partition Coefficient •••»
7
/ ff Q
-1
CO
* lt\
°c
mmHg
Kg/1
Kg/1
/
For physical agents:
/ / thermal / / particulate / / solid waste
Description/definition :
Temperature C (for thermal agents)
Density Kg/1 )
4 (for particulates)
Particle Size microns f
250
-------
A2 - AGENT IDENTIFICATION
Fill ±n applicable sections
Agent Name
Common Synonyms1
Chemical / / Physical / / Radiological
/__/ Biological / / Other
For chemical agents:
/V Compound /___/ Element / / Mixture
CAS Registry No. 6^$/73&8~£" °C
Boiling Point A/ft °c
3
Vapor Pressure @20°C /(/Vj mmHg
Density3 @ 20°C /> 4"^ Kg/1
Water Solubility, 20°C . j,^ Kg/1
Partition Coefficient /Vft
For physical agents:
/ / thermal /_/ particulate / / solid waste / / other
Description/definition:
Temperature C (for thermal agents)
Density
(for particulates)
Particle Size mi<
251
-------
A2 - AGENT IDENTIFICATION
Fill in applicable sections
Agent Name _ £g>2>/«/>?
Common Synonyms A^/P RbCL^AAJ/ C A(L)T> „
Chemical /_/ Physical / / Radiological
/_/ Biological /_/ Other
For chemical agents:
/"$ Compound / / Element / / Mixture
CAS Registry No.1 _ /4* 33
Molecular Formula
2
Structural Formula :
Melting Point3 v<>'"^3>7 °C
3
Boiling Point / t£ <
3
Vapor Pressure @20°C JL
Density @ 20°C /\/ '/
3 o
Water Solubility, 20 C
Partition Coefficient
?£
M
^
* ^f&
/i/A
°c
mmHg
Kg/1
' Kg/1
/^'c-
For physical agents:
/ / thermal / / particulate / / solid waste / / other
Description/definition:
Temperature °C (for thermal agents)
Density Kg/1 ")
t (for particulates)
Particle Size micronsJ
252
-------
A4 - EFFECTS CHECKLIST
Agent Name
Check the suspected effects
Human: /£/ Mortality through A~£,M,TEr
Animal
Plant;
Other:
/ / Serious disease or injury
/ / Other disease or injury
/ / Physiological effects
/ / Aesthetic impact through
/ / Economic impact through
Mortality through
/ / Reproduction impairment through
/_/ Yield reduction through
/ / Mortality through
/ / Reproduction impairment through
/ / Yield reduction through
/ / Ecosystem disturbances through
Describe the principal concern:
T£>
4-C-t^Tlsr
"FISH-
-
/?
A/
This worksheet and worksheet A2 should be reviewed with the nominator
before major ranking steps are undertaken.
253
-------
A20 - RELEASE WORKSHEET
Agent Name H ?&&><>&&*} &&&/ T> G
p
i
E
IU
DU
l\f/ produced commercially
/. JS~~y / 0 Kg/yr
Kg/yr
Kg/yr
/../ V / C? Kg/yr
P-.^V/P7 Kg/yr
/ / other human sources
production
imports
exports
intermediate uses
dispersive uses
(P + I - E - IU / /, or directly /__/)
Releases during production Releases during use
(fractional) (fractional)
\ 1-
Air e
Water e
W.
Land e,.
> 0 >
0
,0r"
Release Rate Computation
Trans-
Production Use portation Other Total
Air pxe = 7,>Ty//? + DU x f 9-, Ly/ p' + + =9*Ly/t
A ' — A^— — ~~ '-^
Water Pxe = /p+DUxf + + Q Kg/yr
w W~~~ "—
Land PxeL=?> fT^i/ ft** + DU x fT + + =7..ry/
-------
A20 (concluded)
Intermediate Uses
c>f=-
Z>V g-JS , & TC,
Dispersive Uses
Transportation Releases
Other Releases
255
-------
A20 - RELEASE WORKSHEET
Agent Name SffPfH/7]
P
I
E
IU
DU
/ y produced commercially
^•/ y/^7 Kg/yr
s-.i
c,<+
t't-
64
v /oL
«;*<-
tJO'
V /V7
Kg/yr
Kg/yr
Kg/yr
Kg/yr
/ / other human sources
production
imports
exports
intermediate uses
dispersive uses
(P + I - E - IU /T> or directly
Releases during production
(fractional)
Releases during use
(fractional)
Air e 0
•"•
Water ew , Q^~
Land e . a {.•""''
f.
A
f
f
Release Rate
Production
Air p x e A = 0
Water Pxe =/,/yja^
,
Land Pxe = /,/ y/^^
Use
+ DUx f.
+ DU x f ^
W
' + DU x f
b
W7
*~*
a
£>
f
/•>
Computation
Trans-
portation Other
+ +
+ +
+ +
Total
Kg/yr
Kg/yr
Derivation of Estimates (Document with brief narrative)
Production Processes
Includes combustion products, release during other activities as
waste, etc. May be available in basic documents or calculable from EF.
256
-------
A20 (concluded)
Intermediate Uses
Dispersive Uses
Transportation Releases
Other Releases
257
-------
A20 - RELEASE WORKSHEET
Agent Name TOTA"S*>1UM CX#-/f/ Zr~ ____
_/S(/_ produced commercially / / other human sources
4.
P _ / X / ^ _ Kg/yr production
I % / V /£> & _ Kg/yr imports
E _ Kg/yr exports
IU _ Kg/yr dispersive uses
(P + I - E - IU //~, or directly £7)
Releases during production Releases during use
(fractional) (fractional)
Air
f
A
Water e^ #_
Land e^ . n A
Release Rate Computation
Trans-
Production Use portation Other Total
Air P x e. = £> + DU x f ffi + + = & Kg/yr
A A ,, 77~
Water P x e = ^ + DU x fu
W w
Land P x eT. = ^y / Q + DU x \^
Derivation of Estimates (Document with brief narrative)
Production Processes
Includes combustion products, release during other activities as
waste, etc. May be available in basic documents or calculable from EF.
258
-------
A20 (concluded)
Intermediate Uses
Dispersive Uses
Transportation Releases
Other Releases
259
-------
A32 - TRANSPORT/TRANSFORMATION WORKSHEET
Agent Name /+Y D&O&GrtJ
1. Chemical Transformation (Steps A25 - A31)
k (Yr)-l/moie Tfl /. £> Yr
J A
Water: kj^ _ (Yr)~1/mole k^ _ (Yr)~1/mole T /. ^ Yr
Soil: P _ Ke _ TL /, £ Yr
To r^x/c
2. Intermedia Transfer (Step A35)
Transfer rates, in (Yr) , from column heading to row heading
Air Water Land
Air
Water J. 3 /.
Land
3. Steady State Inventory (Step A32)
Solve [ ft] i •• ' [ \ ] [ &&I ] equations shown in discussion LI
Kg/yr
Kg/yr SSIL J^V/g Kg
A. Non-steady State Correction (Step A33)
CFA _ SSIA _ K§
CFW _ SSIW _ Kg
CFL _ SSIL _ R8
5. River/Lake Partition (see discussion A39)
Lakes f -^ SSI /j^V/ f? Kg
Rivers f . A^ SSI rii~*/ O? Kg
260
-------
6. Concentrations (Step A39)
Dilution Factors, D Fraction of SSI, f_ Concentration
f o — / G> A
Air 10b mj / ~> g / g>? Kg/m3
/g "^ Sy/g^ Kg/m3
Rivers IO11 1 / 0 " ^"x /^"^ Kg/:L
io12 l 3 xVg "^ .Cx//r Kg/1
13 ~b r
— "ft^
6 1 ^ V/^» 7 f>S~+IO Kg/1
3 x io8 l 3 X / g?~ 7 S~.JTyc/(>' Kg/1
3 x io10 i .2 y / o "^ 3. ^y/g'V Kg/i
12
3 x IO
3 x io14 i A y/o'^ 3,L*/i>~ Kg/1
3 x IO16 1 . f g^ AiT*/^ Kg/1
o
Land 5 x IO6 m2 Kg/m^
5 x IO8 m2 Kg/1"2
5 x IO10 m2
o
5 x IO12 m2 Kg/m
If release rate and/or transformation rate data are inadequate, use
ambient concentrations observed, from SAROAD and STORE!.
261
-------
A32 - TRANSPORT/TRANSFORMATION WORKSHEET
Agent Name
1. Chemical Transformation (Steps A25 - A31)
Air: KOH (YrrVmole K (Yr)-l/oole T y Yr
J A f
Water: KRQ^ (Yr)"1/mole ^ (Yr)~1/mole TW / Yr
Soil: P Ke TL /_ Yr
2. Intermedia Transfer (Step A35)
Transfer rates, in (Yr)~ , from column heading to row heading
Air Water Land
Air
Water _ - /»
Land
3. Steady State Inventory (Step A32)
Solve H^-^^-{;V] [GGI-j- equations shown in discussion LI
RA 0 Kg/yr SSIA _^ _ Kg
\ 1,^*1 17 Kg/yr SSIW
R .v//Ke/yr SSI
4. Non-steady State Correction (Step A33)
CFA _ SSIA _ Kg
CFW _ SSIW _ K§
CFL _ SSI
5. River/Lake Partition (see discussion A39)
Lakes f <^> SSI A /' j./' 0 Kg
Rivers f « C^ SSI /. / V/ ft? Kg
262
-------
6. Concentrations (Step A39)
Air
Rivers
Lakes
Land
Dilution Factors, D Fraction of SSI, f
106m3
108m3
1010 m3
10
12
1014 m3
1016 m3
1011 1
1012 1
1013 1
1014 I
1015 1
3 x 10 1
3 x 10° 1
3 x 1010 1
3 x 1012 I
3 x 1014 1
3 x 1016 1
62*
5 x 10 in
5 x 108 m2
5 x 1010 ^
5 x 1012 m2
D
"f
T
y /
~
y /
v /
"--*
Concentration
Kg/m3
Kg/m3
/ y / 0
3-
3.
Kg/
3
Kg/
Kg/m
Kg/m
Kg/1
Kg/1
Kg/1
Kg/1
Kg/1
Kg/1
Kg/1
Kg/1
Kg/1
Kg/1
Kg/1
Kg/m
Kg/
Kg/m
.
Kg/m
If release rate and/or transformation rate data are inadequate, use
ambient concentrations observed, from SAROAD and STORET.
263
-------
A41 - BIOLOGICAL EFFECTS CHECKLIST
Agent Name C
Species at risk: (select no more than 5)
>W Man
/V/ Domestic animals SH~S
n
Other animals
AT
n
/ / Crops
rj
n
I~~T Other plants
rr
n
Effects to be considered: (select no more than 3 for men, 1 each for
other species)
Species Effects Test
Indicator
2
Man A&M.TE TPt/C.tT'i C4S£-
Selection should be based on suspected dominance of value once
ranking is complete.
2
For example, LD5Q (mouse), human epidemiology, or TLV.
264
-------
A45 - DOSE-EFFECTS WORKSHEET
Agent Name ///P/&P&&1/
Effect
in species
Dose-response relationship:
(see
note 1)
Justification:
(units)
Incidence, fractional, of effect at dose shown. Enter scale as
required (maximum 1.0). Excess over background is implied.
265
-------
A45 - (concluded)
N. C E D=CxE D' I NC.
Number Concentration Exposure Dose Converted
at Risk2 /Units Factor/Units /Units Dose/Units Incidence Cases
i + / Q f'S */o~ jup/t. /^£/PA^I ?yi
, 0-
ON
V
Total Cases
In some cases it may be simpler to express the exposure dose in-
formation as a distribution, dN/dD (number per unit dose) as a function
of D. The number of cases is then given by
Saturation
NC = / Sg. (D) I (D) dD
Threshold
&"" O
/(poo
2
Absolute number or percent of total population (show with %). See also high dose distribution discussion, A45.2.
3 See Table 45.1, BDB, Altman, Prosser, or Dill.
-------
A45 - DOSE-EFFECTS WORKSHEET
Agent Name ^ //^^/tPg /^A^
Effect /&2/ivr£r 7~&X/<2I T y in species
Dose-response relationship:
(see
note 1) l~
Justification:
S&VSIT/I/E- TO
el.
(units)
/S
Incidence, fractional, of effect at dose shown. Enter scale as
required (maximum 1.0). Excess over background is implied.
267
-------
ho
00
A45 - (concluded)
N. C E D=CxE D' I NC.
Number Concentration Exposure Dose Converted
at Risk /Units Factor/Units /Units Dose/Units Incidence Cases
II £/£. 0 GL
A
C •*.//) b /. D
Total Cases
In some cases it may be simpler to express the exposure dose in-
formation as a distribution, dN/dD (number per unit dose) as a function
of D. The number of cases is then given by
Saturation
NC = / || (D) I (D) dD
Threshold
2
Absolute number or percent of total population (show with %). See also high dose distribution discussion, A45.2.
3 See Table 45.1, BOB, Altman, Prosser, or Dill.
-------
A45 - DOSE-EFFECTS WORKSHEET
Agent Name
Effect
/£> g- /
77>y/^/7~y in species
Dose-response relationship:
(see
note 1)
D , ^^lk^
(units^
Justification:
Incidence, fractional, of effect at dose shown. Enter scale as
required (maximum 1.0). Excess over background is implied.
269
-------
^ A*.
-------
Cyanide ion--High Dose Distribution
The predicted Levels are dangerously close to lethal. Let us assume
that the population exposed to higher levels varies inversely with the
concentration squared. The results are shown in the lower part of
Worksheet A45.
Hydrogen Cyanide—High Dose Distribution
The average dose of cyanides is well below the threshold. However,
in the real world, concentrations could accidentally rise above this
level. Let us assume that the number of people exposed to larger con-
centrations varies inversely with dose squared. The results are shown in
the lower part of the A45 worksheet.
271
-------
AA7 - HAZARD RANKING WORKSHEET (page / of
Agent Name
, .,,,. . 2. Cases/ Ranking
1. fcttect Units 3. Value Index (2.x3.)
Tty/csry /MA*) £<
Page Total (Environmental Hazard Index)_
Sum of / pages (Total Environmental Hazard Index)
**
Date Rank
* Applicable only to last page of multi-page forms.
** Enter new rank each time it changes as a result of new entries
to the list.
272
-------
A47 - HAZARD RANKING WORKSHEET (page JL of
Agent Name (2 y&A//£> £~ / t
i pffert- 2' Cases/ Ranking
"recc Units 3. Value Index (2.x3.)
ACTUTE
Page Total (Environmental Hazard Index) /j 00
Sum of ^ pages (Total Environmental Hazard Index) £ X//?
Date Rank
* Applicable only to last page of multi-page forms.
** Enter new rank each time it changes as a result of new entries
to the list.
273
-------
TRACE OF PROCEDURE
Agent Name
Branch
Enter step numbers encountered in ranking agent. If sequential, use nomen-
clature like A3-A7. Use each column for a different branch. It is often
useful to place Branch A near the center.
274
-------
TRACE OF PROCEDURE
Agent Name C y'AAJ'/'£> Zr.5
Branch
Enter step numbers encountered in ranking agent. if sequential, use nomen
clature like A3-A7. Use each column for a different Branch. It is
often useful to place Branch A near the center.
275
-------
TRACE OF PROCEDURE
Agent Name
Branch
Enter step numbers encountered in ranking agent. If sequential, use nomen
clature like A3-A7. Use each column for a different branch. It is
often useful to place Branch A near the center.
276
-------
Cyanides--Narrative
Cyanides were represented by the precursor (and sometimes end prod-
uct) hydrogen cyanide and two salts, sodium and potassium. The acid is
used as a fumigant, but much more heavily as an intermediate to organic
chemicals, dyes, and so on. The salts are used extensively in electro-
plating and steel treatment, in addition to intermediate uses for organic
chemicals. The acid is assumed released to air and land, the salts to
water and land.
Once in the general environment, the cyanides tend to move to the
water medium (rivers and lakes assumed equal). In all media, they are
assumed to react slowly (1 yr) to less toxic forms (complexes, and so on).
The cyanides are acutely toxic if the concentration is sufficiently
high, but slightly lower chronic doses seem to lead only to fatigue and
weakness. Toxicities to man from drinking water, to trout from river
water, and to sheep from land-contaminated forage were examined. The
human risk is potentially of most concern, but average doses were below
threshold. Under certain conditions, cyanide deaths could be quite high.
The hazard to trout is much more likely than hazard for humans.
Cyanides seem to pose a moderate to very high hazard relative to the
other candidate agents.
Cyanides--Procedural Difficulties
Cyanides did not pose any unusual procedural difficulties except in
the attempt to treat accidentally high dosages in man. The environmental
hazard index for this effect could be as high as 10 , but probably is
much smaller (104 ?).
277
-------
Cyanides--Recommendations
Alternative distribution models for cyanides should be tested to
determine whether the high dose distribution is reasonable.
278
-------
Appendix H
ABBREVIATIONS
279
-------
Abbreviations
Reference. If Applicable
ACS American Chemical Society
BA Biological Abstracts
BOB Biology Data Book
BOD, Five Day Biochemical Oxygen Demand
CA Chemical Abstracts
CAS Chemical Abstracts Service
CBAC Chemical Biological Activity Catalog
CEH Chemical Economics Handbook
Census Census of Manufactures
CHEMLINE Chemline
CMR Chemical Marketing Reporter
COD Chemical Oxygen Demand
COM Chemical Origins and Markets
CPSC Consumer Product Safety Commission
CR Chemical Reviews
CTCP Clinical Toxicology of Commercial Products
DCP Directory of Chemical Producers
Dill Handbook of Physiology
DOT Department of Transportation
Doyle G. J. Doyle, e_t al
EB Environmental Biology
EEC European Economic Community
EF Emission Factors for Trace Substances
EMIC Environmental Mutagen Information Center
EPA-OAWM Environmental Protection Agency, Office of Air
and Waste Management
EPA-OECG EPA, Office of Enforcement and General Council
EPA-OHM EPA, Office of Hazardous Materials
EPA-OPP EPA, Office of Pesticide Programs
EPA-OSWMP EPA, Office of Solid Waste Management Programs
EPA-OTS EPA, Office of Toxic Substances
EPA-OWHM EPA, Office of Water and Hazardous Materials
FCH Farm Chemicals Handbook
FDA Food and Drug Administration
FPG Fire Protection Guide on Hazardous Materials
FT-246 Federal Trade Commission
American Chemical Society, 1968
*
Altman, 1974
Sawyer, 1971
SRI a
Bureau of the Census, 1972
**
***
Sawyer, 1971
SRIb
Leo, 1971
Gleason, 1969
SRI, 1974
Dill, 1964
Doyle, 1975
Altman, 1966
****
Anderson, 1973
•ft*
Berg, 1975
NFPA, 1975
FTC, 1973b
281
-------
Abbreviations
Reference. If Applicable
FT-410
HBT
Hendry
HCP
IBCP
IM
JWPCF
Lambert
Lange
LC50
LD50
LDLO
LRPS
MERCK
Miller
MP
MY
NA
NCI
NEDS
NIOSH
NLM
NSF
OSHA
PHS 149
Prosser
SAROAD
Shepard
SOC
SOCMA
STORE!
Sax
TADS
Federal Trade Commission
Handbook of Toxicology
Journal of Physical Chemistry Reference Data
Handbook of Chemistry and Physics
Imports of Benzenoid Chemicals and Products
Index Medicus
Journal of the Water Pollution Control
Federation
Sorption in Soil
Lange 's Handbook of Chemistry
Concentration Lethal to 5070 of Exposed
Population
Dose Lethal to 50% of Exposed Population
Lowest dose reported as lethal
Long Range Planning Service
MERCK INDEX
Models of Radionuclides
Metabolism of Pesticides
Minerals Yearbook
Not Applicable
National Cancer Institute File
National Emissions Data System
National Institute for Occupational Safety
and Health
National Library of Medicine
National Science Foundation Study
Occupational Safety and Health Administration
Survey of Compounds Which Have Been Tested for
Carcinogenic Activity
Comparative Animal Physiology
Storage and Retrieval of Aerometric Data
Catalog of Teratogenic Agents
Synthetic Organic Chemicals
SOCMA Handbook
Storage and Retrieval of Water Data
Dangerous Properties or Industrial Materials
Technical Assistance Data System
FTC, 1973a
Spector, 1956
Hendry, 1974
Weast, 1975
ITC, 1976
*
***
Lambert, 1967
Dean, 1973
NIOSH, 1974
NIOSH, 1974
NIOSH, 1974
SRIc
Steiber, 1968
Miller, 1963
Monzie, 1969
Bureau of Mines
National Cancer Institute
****
SRI, 1975
Shubik
Prosser, 1973
**•**
Shepard, 1973
ITC, 1973
Synthetic Organic Chemical Man-
ufacturer's Association, 1966
-A--.V-W-
Sax, 1975
****
282
-------
Abbreviations
TDB
TLV
TMIC
TOXLINE
TSL
USCG
USD A -MID
Toxicology Data Bank
Threshold Limit Value
Toxic Materials Information Center
TOXLINE
Toxic Substances List
U.S. Coast Guard
U.S. Department of Agriculture, Meat Ins
****
ACGIH,
**
**
NIOSH,
>pec-
1971
1974
Reference, If Applicable
tion Division
Wilson Journal of Chemical and Physical Reference
Data
WPPMP Water Pollution Potential of Manufactured
Products
WQC Water Quality Criteria
Wilson, 1972
Berkowitz, 1973
EPA, 1972
See Abstracting Services in References.
See Computer-Based Literature Search Services in References.
See Periodicals in References.
See Computer-Based Data Files in References.
283
-------
Appendix I
REFERENCES
285
-------
REFERENCES
Altaian, P. L. and D. S. Dittmer, ed., Environmental Biology, Fed. Amer.
Soc. Expt. Biol., Bethesda, Maryland. 694 p. (1966).
Altaian, P. L. and D. S. Dittmer, ed., Biology Data Book, 2nd ed. Fed.
Amer. Soc. Expt. Biol., Bethesda, Maryland. 2123 p. (1974).
American Chemical Society, Oxidation of Organic Compounds--III, Advances
in Chemistry Series, 77 (1968).
American Conference of Governmental Industrial Hygienists, 1971. Documen-
tation of the Threshold Limit Values for Substances in Workroom Air,
3rd Edition (1971).
Anderson, D., "Emission Factors for Trace Substances," EPA Report No. PB-
230-894 (1973).
Anonymous, "Hot Market for Antimony Oxide," Chem. Wk. . 113:21-22 (1973).
Battelle Memorial Institute, "Identification Systems for Selecting Chemi-
cal Classes as Candidates for Evaluation," EPA-5611-74001 (November
1974).
Behrens, R. G. and G. M. Rosenblatt, "Vapor Pressure and Thermodynamics of
Orthorhombic Antimony Trioxide (Valentinite)," J. Chem. Thermodynamics,
5:173-188 (1973).
Berg, G. L., ed., Farm Chemicals Handbook. Meister Publishing Co. (1975).
Berkowitz, J. B., G. R. Schimke, and J. R. Valeri, "Water Pollution Poten-
tial of Manufactured Products," 4 vols. Off. of Res. & Monitoring,
Environmental Protection Agency, EPA-R2-73-179A-D (1973).
Biers, W. D., "Characterization of Glaus Plant Emissions," EPA Report No.
EPA-72-73-188 (1973).
Boland, L. F., "Beryllium: Present and Potential Uses," The Analysts
Journal, p. 27-31 (1958).
287
-------
Brush Beryllium Company, "Literature on Industrial Hygiene Practices for
the Control of Air-Borne Beryllium," Mimeographed Notes (1956).
Bureau of the Census, Census of Manufactures (several series) (1972).
Bureau of Mines, "Mineral Industry Surveys: Beryllium in 1974" (December
1974).
Capener, E. R. M. Wright, and S. L. Brown, Handbook of Hazardous Wastes,
Federal Ministry of Interior (Germany) (1974).
Centre D/Information du Cobalt, Cobalt Monograph (Brussels) (1960).
Dean, J. A., ed., Lange's Handbook of Chemistry, llth ed. (McGraw-Hill
Book Company, New York, New York, 1973).
Dill, D. B., E. F. Adolph, and C. C. Wilber, ed., Handbook of Physiology:
Section 4, Adaptation to the Environment. Amer. Physiol. Soc. ,
Washington, D.C. 1056 p. (1964).
Doyle, G. J., A. C. Lloyd, K. R. Darnall, A. M. Winder, and J. N. Pitts,
Jr., "Gas Phase Kinetic Study of Relative Rates of Reaction of
Selected Aromatic Compounds with Hydroxyl Radicals in an Environmen-
tal Chamber," Environ. Sci. Tech.. £ (3) 237 (1975).
Dreisbach, R. H., Handbook of Poisoning: Diagnosis and Treatment, 5th
Edition, Lange Medical Publications, Los Altos, California (1966).
Durocher, N. L., "Preliminary Air Pollution Survey of Beryllium and Its
Compounds: A Literature Review," Litton Systems, Inc., Final Report,
HEW Contract No. Ph 22-68-25 (1969).
Edinger, J. E., "Vertical Temperature Structure and Water Surface Heat
Exchange," Water Resources Research, 6:1392-1395 (1970).
Edinger, J. E., D. W. Duttweiler, and J. C. Geyer, "The Response of Water
Temperatures to Meteorological Conditions," Water Resources Research,
4:1137-1143 (1968).
Environmental Protection Agency, "Water Quality Criteria 1972," EPA-R-73-
033 (March 1972).
Environmental Protection Agency. 1973a, "National Emissions Inventory of
Sources and Emissions of Molybdenum," EPA-4501-3-74-009 (May 1973).
Environmental Protection Agency, 1973b, "Plutonium: Statement of the
Problem," Office of Radiation Programs (mimeo), 73 p. (1973).
288
-------
Environmental Protection Agency, "Proceedings of Public Hearings: Pluto-
nium and the Other Transuranium Elements," Office of Radiation
Programs, ORP/CSD-75-1 (1974).
Exxon Research and Engineering Co., "Evaluation of Pollution Control in
Fossil Fuel Conversion Processes—Liquefaction: Section 2, SRC
process," EPA Report No. EPA-650/2-74-009-f (1975).
Federal Trade Commission, 1973a, "U.S. Foreign Trade, Exports, Commodity
by Country" (FT-410), Federal Trade Commission (1973).
Federal Trade Commission, 1973b, "U.S. Imports, General and Consumption,
Schedule A, Commodity and Country" (FT-246), Federal Trade Commission,
(1973).
Firm, R. J., "The Chemistry of Carbonyl Sulfide," Chem. Rev., 57:621-640
(1957).
Fishbein, L., "Environmental Health Aspects of Selenium, Tellurium and
Molybdenum: A Preliminary Review, #3-Molybdenum," World Health
Organization. Environmental Health Criteria Programme, Document
EHE/EHC/WP/74.3 (1974).
Gleason, M. N., R. E. Gosselin, H. C. Hodge, and R. P. Smith, "Clinical
Toxicology of Commercial Products," Williams & Wilkins (1969).
Goodman, L. S. and A. Gilman, The Pharmacological Basis of Therapeutics,
4th ed., The Macmillan Co., London (1970).
Hamilton, A. and H. L. Hardy, Industrial Toxicology, 3rd ed., Publishing
Sciences Group, Inc., Acton, Massachusetts (1974).
Hendry, D. G. , T. Mill, L. Piskiewicz, J. A. Howard, and H. K. Eigenmann,
"A Critical Review of H-Atom Transfer in the Liquid Phase: Chlorine
Atom, Alkyl, Trichloromethyl, Alkoxy and Alkylperoxy Radicals," J.
Phys. Chem., Ref. Data, 3, 937 (1974).
Howard, P. H., "Establishing Environmental Priorities for Synthetic Or-
ganic Chemicals: Focusing on the Next PCB' s from Seminar on Early
Warning Systems for Toxic Substances," (February 1974).
LARC, International Agency for Research on Cancer Monograph on the Evalu-
ation of Carcinogenic Risk of Chemicals to Man. Vol. 1:17-28 (1972).
289
-------
International Trade Commission, Synthetic Organic Chemicals (1973).
International Trade Commission, "Imports of Benzenoid Chemicals and
Products, 1974," Publication 762 (March 1976).
Jordan, T. E., Vapor Pressures of Organic Compounds, Interscience (New
York, 1954).
Kirk-Othmer, Encyclopedia of Chemical Technology, 2nd ed., Wiley (1963).
Kontsoukos, E. P., J. L. Blumenthal, M. Ghassemi, and G. Baverie, "Assess-
ment of Catalysts for Control of NO from Stationary Power Plants,"
EPA Report No. EPA-650/2-75-001a (1972).
Lambert, S. M., "Functional Relationship between Sorption in Soil and
Chemical Structure," J. Agr. Food Chem., L5 (4) 572 (1967).
Leo, A., C. Hansch and D. Elkins, "Partition Coefficients and Their Uses,"
Chem. Reviews, 7_1 (6) 525 (1971).
Mackay, D. and A. W. Wolkoff, "Rate of Evaporation of Low Solubility Con-
taminants from Water Bodies to Atmosphere," Environ. Sci. Tech. , 7_
(7) 611 (1973).
Matheson Company, Inc., "Carbonyl Sulfide," pp. 97-100, In Matheson Gas
Data Book, 4th ed., East Rutherford, N. J. (1966).
Miller, C. F. and S. L. Brown, "Models for Estimating the Absorbed Dose
from Assimilation of Radionuclides in Body Organs of Humans," SRI
Project IMU-4021 (May 1963).
Mines, Bureau of, Minerals Yearbook, latest edition.
Monzie, C. M., Metabolism of Pesticides, Wildlife No. 127, Bureau of
Sports Fisheries and Wildlife (1969).
Myers, D. R., "Airborne Thermal Imagery of Discharge Plume Dispersion--
Pilgrim Station," Research Report, Coastal Research Corporation
(1971).
National Academy of Sciences, "Water Quality Criteria, 1972," EPA Ecologi-
cal Research Series, EPA-R3-73-033, 594 p. (1972).
National Academy of Sciences, 1975a, Assessing Potential Ocean Pollutants
(1975).
290
-------
National Academy of Sciences, 1975b, Principles for Evaluating Chemicals
in the Environment (1975).
National Fire Protection Association, "Fire Protection Guide on Hazardous
Materials" (1975).
National Institute for Occupational Safety and Health, The Toxic Substances
List (June 1974).
Novotny, V. , P. A. Krenkel, "Simplified Mathematical Model of Temperature
Changes in Rivers," Journal of the Water Pollution Control Federation,
45:240-248 (1973).
Peyton, T. 0., R. V. Steele, and W. R. Mabey, "Carbon Disulfide, Carbonyl
Sulfide: Literature Review and Environmental Assessment," Stanford
Research Institute (January 1976).
Prosser, C. L., Comparative Animal Physiology, 3rd ed., W.B. Saunders Co.,
Philadelphia, Pa. (1966).
Reeves, A. L., D. Deitch, and A. J. Vorwald, "Beryllium Carcinogenesis.
I. Inhalation Exposure of Rats to Beryllium Sulfate Aerosol," Cancer
Research, 27:439.
Sawyer, C. N., and P- L. McCarty, Chemistry for Sanitary Engineers (1971).
Sax, N., Dangerous Properties of Industrial Materials. 4th ed., Van
Nostrand (1975).
Shepard, T. H., Catalog of Teratogenic Agents, Johns Hopkins (1973).
Shubik, P. and J. L. Hartwell, Survey of Compounds Which Have Been Tested
for Carcinogenic Activity, HEW/PHS, several volumes (1940-1973).
Spector, W. S., Handbook of Toxicology. Vol. 1, Saunders (1956).
Stanford Research Institute, a, "Chemical Economics Handbook Program," a
continuing marketing research service.
Stanford Research Institute, b, Chemical Origins and Markets, "Chemical
Information Services," Stanford Research Institute.
Stanford Research Institute, c, "Long Range Planning Service," a continuing
service to industry.
291
-------
Stanford Research Institute, Directory of Chemical Producers, an annual
publication (1974).
Stanford Research Institute, "Research Program on Hazard Priority Ranking
of Manufactured Chemicals," Report by Stanford Research Institute to
the National Science Foundation (April 1975).
Startton, W. R., ed., Plutonium Information Meeting for an Ad Hoc Subcom-
mittee of the Advisory Committee on Reactor Safeguards, U.S. Atomic
Energy Commission, Conf-740115 (1974).
Steiber, P. G., ed., The Merck Index, 8th ed., Merck and Co., Rahway, N.J.
(1968).
Stokinger, H. E. , ed., Beryllium: Its Industrial Hygiene Aspects, Academic
Press, N.Y. (1966).
Sundrom, T. R. and R. G. Rehn, "The Seasonal Thermal Structure of Deep
Temperate Lakes," Tellus. 25_:157-167 (1973).
Synthetic Organic Chemical Manufacturers' Association, SOCMA Handbook
(1966).
Teisinger, J., "1972 Yant Memorial Lecture—New Advances in Toxicology
of Carbon Bisulfide," American Industrial Hygiene Association Jour-
nal, 35:55-61 (1974).
Timmermans, J., Physico-Chemical Constants of Pure Organic Compounds,
Elsevier, N.Y. (1965).
U.S. Atomic Energy Commission, 1974a, "Generic Environmental Statement--
Mixed Oxide Fuel" (draft), Wash-1327, 4 vol. (1974).
U.S. Atomic Energy Commission, 1974b, "Reactor Safety Study: An Assess-
ment of Accident Risks in U.S. Commercial Nuclear Power Plants,"
Wash-1400-D (1974).
U.S. Geological Survey, 1970, The National Atlas: the United States of
America, Washington (1970).
Warren, C. E., Biology and Water Pollution Control. Saunders (1971).
Weast, R. C., ed., Handbook of Chemistry and Physics, 56th ed. , Chemical
Rubber Company; Cleveland, Ohio (1975).
292
-------
Wilson, W. E., Jr., "A Critical Review of Gas-Phase Reaction Kinetics of
Hydroxy (Radical)," J. Chem. Phys. Ref. Data, 1(2) 535 (1972).
293
-------
Abstracting Services
Biological Abstracts, Biosciences Information Services (BA)
Chemical Abstracts, American Chemical Society (CA)
Index Medicus, Government Printing Office (IM)
Computer-Based Literature Search Services
CHEMLINE, National Library of Medicine (CHEMLINE)
Chemical Biological Activities, Literature of Interactions of Chemical
Substances with Biological Systems, Chemical Abstracts System (CBAC)
Environmental Mutagen Information Center, Oak Ridge National
Laboratory (EMIC)
*
Toxic Materials Information Center, Oak Ridge National Laboratory (TMIC)
TOXLINE, National Library of Medicine (TOXLINE)
Periodicals
Chemical Marketing Reporter, Schnell Publishing Co., New York (CMR)
Journal of the Water Pollution Control Federation, Water Pollution
Control Federation, Washington (JWPCF)
Computer-Based Data Files
Commission of the European Communities, Economic Data on Carbamates,
Organo-Chlorine Chemicals, and Organo-Phosphorus Chemicals, SRI (EEC)
National Cancer Institute, Chemicals That Impact on Man and His
Environment (NCI)
National Emissions Data System, EPA-OAWM (NEDS)
Storage and Retrieval of Aerometric Data EPA-OAWM (SAROAD)
Storage and Retrieval of Water Data EPA-OWHM (STORET)
Technical Assistance Data System EPA-OHM (TADS)
Toxicology Data Bank NLM (TDB)
PF/D 294
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO
EPA-600/5-012
3 RECIPIENT'S ACCESSION NO.
PB 258168
4. TITLE AND SUBTITLE
SYSTEMS FOR RAPID RANKING OF ENVIRONMENTAL POLLUTANTS
Selection of Subjects for Scientific and Technical
Assessment Reports
5. REPORT DATE
June 1978 issuing date
6. PERFORMING ORGANIZATION CODE
'. AUTHOR(S)
Stephen L. Brown, Buford R. Holt, and
Kirtland E. McCaleb
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG'VNIZATION NAME AND ADDRESS
Stanford Research Institute
333 Ravenswood Avenue
Menlo Park, California 94025
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-01-2940
Tasks 015 and 023
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
401 M Street, SW
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This document reports the results of the development and testing of a system for
rapidly ranking environmental pollutants. One potential use for the system is in
choosing the most important candidates for Scientific and Technical Assessment
Reports (STAR). Of several possible approaches to ranking environmental agents, a
system depending on expert opinion but assisted by an objective subsystem was
selected for development. The system defines procedures for collecting, processing,
and evaluating data on production and use; environmental transport, transformation,
and rate; and human health and welfare and ecological effects. A test of the
objective subsystem confirmed the utility of the system. Of ten candidate agents,
the three highest ranked were cyanides, carbon disulfide, and beryllium.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Contaminants
Assessments
Criteria
Ranking
Systems Analysis
Ecology
b.IDENTIFIERS/OPEN ENDED TERMS
STAR Carbonyl Sulfide
Priorities Cyanides
Environmental Agents Heat
Hazards Plutonium
Antimony Beryllium
Molybdenum Lithium
Cobalt Carbon Disulfide
c. COSATI Field/Group
6F 6R 6T
7B 1C 7D
12B 18G
3. DISTRIBUTION STATEMENT
Document is available to the public
through the National Technical Information
Service.. Springfield. Virginia 22151
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
306
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
'i,U.S GOVtRNMENT PRINTING OFFICE 1978-260-880.73
295
------- |