-------
                        Conceptual Approach for Contaminant Identification
       2.2.4   Questions for Stakeholders on the Identification of Pathogens

       The following are a series of question compiled for consideration by Stakeholders regarding
the identification of pathogens for the Contaminant Candidate List:

       1.      There are two main schools of thought for identifying and prioritizing pathogens, as
              follows.

              a)    EPA prepares a list of known and potential waterborne disease agents, and
                    has it peer reviewed by experts in microbiology and public health within
                    EPA, Centers for Disease Control (CDC), states, acadernia, etc. Selected
                    experts in microbiology and public health would meet to discuss each
                    pathogen. The group would determine which pathogens should appear on the
                    Contaminant Candidate List, and the relative priority of each.

              b)    Develop  a  conceptual risk-based  approach with weighted  criteria  for
                    identifying and prioritizing pathogens.

              Question: Is there  a model mat is both practical and beneficial for identifying and
              prioritizing pathogens? If so, what is its general structure?

              Question: If approach (a) is used,  what peer review, should be used on this list of
              identified organisms and their relative priority?

       2.      Question:  What role should states and the water industry have if approach (a) is
              used?

       3.      Question: To what  extent should the relative source; contribution  be used in
              identifying and prioritizing organisms (e.g., food vs. inhalation vs. drinking water)?
Working Draft                                22                               November 1996

-------
                        Conceptual Approach for Contaminant Identification
       2.3    Description of Products from Contaminant Identification Method

       The purpose of the CIM is to classify contaminants into several possible regulatory and non-
regulatory categories. These categories include the Contaminant Candidate List, health advisories,
other guidance, toxicology research, unregulated monitoring, or no action.

       A major factor in the determination of these options is the availability of data. As potential
contaminants are processed through the CIM, it will become apparent which contaminants have
adequate data to determine if they are a possible risk to public health through drinking water and
which are lacking data. Regulatory decisions, such as placement on the Contaminant Candidate List
or removal from  consideration,  will  only be made for contaminants with  adequate  data.
Contaminants for which adequate data are riot available to reasonably gauge risk will be considered
for toxicity research needs or unregulated monitoring. The types of products for the CIM are
described in detail below.

       2.3.1   Drinking Water Contaminant Candidate List

       Under the  1996 SDWA Amendments, the EPA is required to ". . . publish a list of
contaminants which, at the time of publication, are not subject to any proposed or promulgated
drinking water regulations, which are known or anticipated to occur in public water systems, and
which may require regulation ..."  EPA is calling this list the Drinking Water Contaminant
Candidate List. This list will be composed of the contaminants  that have adequate data to determine
that they pose a potential human health risk from exposure via drinking water.  One element in
determining risk will be data on the known or anticipated occurrence of contaminants in public water
systems.

       Future regulatory decisions will be made on the chemicals selected for the Contaminant
Candidate List. The First Contaminant Candidate List, described by two tiers (see Section 3), will
be published by February 1998. Subsequent Contaminant Candidate Lists must be published every
five years thereafter. The Contaminant Candidate List is the first step in the process for developing
future drinking water regulations; however, not all contaminants on the list will be regulated. The
list is a tool to prioritize the evaluation of contaminants for regulation under the drinking water
program. Once a contaminant is on the Contaminant Candidate List, a determination will be made
to regulate, or not to regulate but perhaps to develop a health advisory.

       The Contaminant Candidate List will be published in the Federal Register.  Prior to its final
publication, a proposed list likely will be published with opportunity for public comment. According
to the 1996 Amendments, the scientific community, including the Science Advisory Board (SAB)
must be consulted, and the National Occurrence Database  must be considered prior to final
publication of the Contaminant Candidate List. Also, according to the 1996 Amendments, the
decision to include a contaminant on the Contaminant Candidate List is not subject to judicial
review.

       2.3.2   Health Advisories
Working Draft                                23                               November 1996

-------
                        Conceptual Approach for Contaminant Identification
       A health advisory is an informal technical guidance for protecting public health when
contamination situations occur. These documents are not to be construed as legally enforceable
federal standards.  A health advisory describes non-regulatory concentrations of drinking water
contaminants at which adverse health effects  would not be anticipated to occur over specific
exposure durations, with a margin of safety. A drinking-water standard is a legally enforceable
requirement which includes a maximum contaminant level (MCL) for the protection of human health
which public water suppliers cannot exceed.

       Under the CIM, contaminants may be identified for which health  advisories  are most
appropriate.  These contaminants will be the priority contaminants under the Health Advisory
Program.  It may also be appropriate to develop health advisories  for other contaminants included
on the Contaminant Candidate List. Health advisories could provide Ihe guidance necessary for
contamination situations during the regulator development phase if the determination is made to
regulate.

       2.3.3  Other Guidance

       "Other guidance" consists of various innovative approaches that the EPA may be able to use
to control contamination before it occurs.  Other guidance refers to the application and use of other
program authorities outside of the drinking water program.  An example of other guidance would
be the use of the pesticides  program labeling restriction, or cancellation authority, to  reduce
contamination of drinking water sources for a particular pesticide. This category is one which EPA
has not used before, but would like to investigate and employ when and if feasible.  To produce such
a product from the CIM, OGWDW will have to work closely with other EPA program offices.

       2.3.4  No Action

       "No action" on the part of the Agency may occur in cases where a contaminant has adequate
data to determine risk, but no action is warranted due to lack of toxicity and/or occurrence. For
example, a contaminant does not pose an adverse health risk at concentrations found in drinking
water.

       2.3.5  Priorities for Toxicity/Health Effects Research

       The priorities for toxicity and health eJ'fects research will be identified from the contaminants
lacking toxicity information for risk assessment purposes. Research plans will be developed for
these contaminants to meet the needs identified through the CIM.  Once the research is completed
the contaminant would be returned to Stage I of the CIM for reconsideration during the next CIM
round.
       2.3.6  Priorities for Unregulated Contaminant Monitoring

       The priorities for monitoring will be identified from the contaminants lacking monitoring
data necessary  for risk assessment purposes.   Unregulated monitoring  requirements will be
developed for these contaminants to meet the occurrence data needs identified through the CIM. As
with the priorities for toxicity/health effects  research, once the  occurrence is gathered,  the
Working Draft                               24                               November 1996

-------
                         Conceptual Approach for Contaminant Identification
contaminant would be returned to the Stage I of the CIM for reconsideration during the next CIM
round.
Working Draft                                25                               November 1996

-------
                        Conceptual Approach for Contaminant Identification
3.     FIRST CONTAMINANT CANDIElATE LIST

       The 1996 SDWA Amendments specify more distinct, risk reduction criteria for regulating
chemicals in drinking water than previous SDWA Amendments, and allow EPA more realistic time
frames to accomplish the SDWA directives. Prior to the 1996 Amendments, EPA was required to
publish a list of contaminants which were "known or anticipated to occur in public water systems
and which may require regulation." This Drinking Water Priority List (DWPL) was to be updated
on a triennial basis. Additionally,  EPA was required to select 25 contaminants from the list for
regulation every three years. Disinfectant and disinfection byproducts (D/TDBP) and the unproposed
Phase VIb contaminants were derived from the 1991 DWPL. As a result of a program redirection,
EPA held stakeholder meetings  in 1995 to solicit opinions on many of the contaminants  being
considered for regulation. The results of these 1995 meetings will be used in the development of the
First Contaminant Candidate List.

       3.1    The New Contaminant Listing

       The 1996 Amendments charge EPA with developing, within 18 months of enactment, a
"listing of contaminants for consideration	which are known or anticipated to occur in public water
systems and which may require regulation." The drinking water Contaminant Candidate List is to
be updated every five  years. The 1996 Amendments require that EPA give priority to selecting
contaminants that present the greatest public health concern, including vulnerable populations such
as infants, the elderly, and those with serious illnesses.

       Other requirements for contaminant listing include:

       •      consideration of substances referred  to in  section 101(14) of the CERCLA and
             pesticides registered under the: FIFRA;
       •      consultation with the scientific community and the Science Advisory Board;
       •      notice and  opportunity for public comment;
       •      inclusion of sulfate on the list; and
       •      consideration of the contaminant occurrence data base, currently under development.

       3.2    Use of the First Contaminant Candidate List: Five Year Determination Cycle

       The Contaminant Candidate List will  be used for further screening for selection  of
contaminants for regulation, health advisories or further data collection or removal from the list. The
Act specifies that "for not fewer than five contaminants  included on the list," EPA will  make
"determinations of whether or not to regulate such contaminants" every five years.

       3.3    First Contaminant Candidate Process

       As shown in Exhibit 3-1, the process for developing the First Contaminant Candidate List
will include all the components identified in the law plus an assessment of the scientific basis for
including contaminants in such a list, and consultation with the scientific community. Stakeholder
meetings will be a significant means of obtaining public input to each step of the process.  The key
features being proposed for developing the 1998 list are:
Working Draft                                26                              November 1996

-------
                        Conceptual Approach for Contaminant Identification
       •      consideration of significant lists of contaminants in various media and any other data
              bases;
       •      identification of contaminants already found in ambient or drinking water for further
              consideration;
       •      threshold criteria applied to contaminants not yet found in water to screen out those
              that will not pose problem in drinking water;
       •      health screening for potential low level occurrence but high toxicity;
       •      assessment of the potential for occurrence of remaining contaminants in ambient or
              drinking water;
       •      contribution to listing 30 data-poor contaminants for unregulated monitoring;
       •      candidates for further health effects research;
       •      health effects screening with sensitive sub-populations in mind;
       •      consideration of CERCLA and F1FRA chemicals; and
       •      consideration of microbials.

       The proposed First Contaminant Candidate List process, as shown in  Exhibit 3-1, is
described in summary form in 7 steps:

       Step 1 - Assemble Comprehensive List of Potential Contaminants

       The identification of the initial contaminants to be considered for the First Contaminant List
will take advantage of the listing processes used for the 1991 Drinking Water Priority List, the
CERCLA list, the FIFRA pesticide registration list, other lists of contaminants potentially in drinking
water, data from the evolving Drinking Water Contaminant Occurrence Data Base, as well as
contaminants identified by the public and the scientific community.  This comprehensive list will
be evaluated for further consideration through the screen of Threshold Criteria, described below.
During the first step and throughout the process, the scientific community will be consulted through
formal meetings  with the Science Advisory Board and other scientific and industry groups.

       Step 2 - Select Known Water Contaminants for Further Consideration

       •      Which chemicals have been measured in drinking water (public water systems) at any
              level?
       •      Which contaminants have been measured at any level in ambient water?

       Step 3 - Apply Simple Threshold Criteria for Anticipated Occurrence

       Threshold criteria are designed to take a comprehensive list of potential candidate chemicals
or organisms and screen out those lacking the possibility of being contaminants in drinking water.
This step will enable EPA to sort out chemicals which have not been measured in water (perhaps due
to a lack of analytical method) to determine which  are likely never to be  drinking  water
contaminants.
Working Draft                                27                               November 1996

-------
  t
  g
III
 •o >
  &
1999 List of <30
Unregulated
Contaminants
to be Monitored
} U\ U
N
k.


.«
1
3
1
u
>.

•a
S £2
§ -8
11
2 U
;,< k
S S 3 S S V ^ u
- . . . . , .
e -S « «
"!3 cQ -^ 3
I € 1 1 3 II
" ^ 5 g >J ^ Q
> k
•B -g ~ « «
® 2 I « g
-1 O § S W
^ w u a
t ...
1*1
.1 -a J 2
t
VIHKXIH3
a^OHsaHH

t t tt
§1111
1 - « „
M-J '-3 3 «
.s t' < J j
•S .^ 3 < h
[ n i

€
1
cc
p

«



1

\
\
O
LI I J
t

i

a>





N
1
u/o 3
/



^





ii 8 «
S3 o3 3 _3
1 J o o
Q












-------
                         Conceptual Approach for Contaminant Identification
       •      Is the chemical soluble in water (parts per trillion?)
       •      If low solubility, then screen for high toxicity?
       •      Is the chemical stable in water? For example, is the half life more than 24 hours?
              Consider factors such as photolysis, hydrolysis, bacterial degradation, reactivity, and
              volatility.
       •      Is there a stable degradate?

       Step 4 - Apply Specific Criteria for Assessing the Likelihood of Occurrence in Drinking
Water
       •      Is the chemical discharged (e.g., TRI) annually to:
                     water i.e. >10001bs.?
                     land > i.e. 10,000 Ibs.?
                     air i.e. > 1,000,000 Ibs.?
       •      Is it a pesticide applied (produced) greater than, for example, 1,000,000 Ibs annually?
       •      Does it have a leaching potential (persistent and mobile) over a certain threshold?
       •      Is it used directly in the water?

       Step 5 - What are the health effects

       •      What are the toxic endpoints?
       •      Who are the affected populations?
       •      How good and complete is the data base?

       Step 6 - Intersect Occurrence with Health Effects to Find Those of Greatest Concern

       The end result of the intersection of known or anticipated  occurrence and health effects of
each chemical will be a list of candidates for regulatory consideration during the next three and a half
years. The list  will be prioritized in accordance with the SDWA to further reveal contaminants
which,  by virtue  of their exposure to the general public or subgroups at a level, frequency, and
duration of a certain magnitude, present the greatest public health concern.

       The intersection matrix will be used to separate out contaminants into distinct groups, shown
here as Tier I and II. Tier I characteristics include:

       •      Contaminants with known occurrence and with adequate health data sufficient to
              derive a valid MCL and to conclude that these contaminants are a public  health
              concern from drinking water exists.

       •      Contaminants with anticipated occurrence (meeting the above criteria) and with
              adequate health information to derive an MCL and to conclude that a public  health
              concern from drinking water exists.

       Tier n characteristics  include:
Working Draft                                29                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
      •      Contaminants with good health data, but inadequate occurrence data to characterize
             an exposure risk. If the contaminants have high toxicity then they will make the
             drinking water Contaminant Candidate List for unregulated monitoring.

      •      Contaminants with known occurrence without adequate health information are high
             priorities for health research.

      •      Contaminants  with  anticipated  occurrence  without adequate health information
             would be kept for further monitoring and toxicity research.
      •      Contaminants with low anticipated occurrence and high toxicity would be kept for
             unregulated monitoring.

      Step 7 - Listing in Federal Register

      The listing will occur in two stages:

      •      The preliminary list will be published in the Federal Register for public comment.

      •      After consideration of public comments on the preliminary list, the final 1998 list
             will be published in the Federal Register on or before February 6, 1998.
Working Draft                                30                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
4.     STAKEHOLDER INVOLVEMENT

       The EPA is committed to facilitate stakeholder involvement in regulatory activities.  In this
section, the EPA has prepared a series of issues and questions regarding the CIM for stakeholders
in defining their involvement and  has prepared  a  list  of potential  avenues for stakeholder
involvement.

       4.1    Issues and Questions for Stakeholder Involvement

       The EPA has prepared several questions that  they would to like to pose to stakeholders
regarding the CIM. Throughout the discussion of the CIM, EPA has presented issues from the
perspective of how stakeholders could be involved in the development of the CIM.  EPA has
assumed that  stakeholders would choose to be intimately involved  in developing  the CIM,
particularly in developing some of the decision criteria.  EPA is also looking to stakeholders to assist
the Agency in defining stakeholder roles. In this process, EPA has identified a number of questions
that would require feedback from stakeholders of  many different backgrounds.  EPA requests
assistance in areas where specific expertise is required.

       The topics below generally follow the four main stages of the CIM as depicted in Exhibit 2-1.
For each topic some background is provided and specific questions are stated.

•      Stage I - As stated earlier, the one pagers (see Appendix B) describe how EPA envisions
       each of the routes would be used to identify potential contaminants under Stage I, Initial
       Identification. Within the description of each  route of identification, the sources of input and
       the process used to collect and compile the information includes some of the proposed roles
       and responsibilities of the stakeholders.

       1.     From the perspective of the stakeholders, could the organization and alternatives of
              these descriptions be better presented?
       2.     Do  the  proposed approaches have  the potential to  collect the most  valuable
              stakeholders input?
       3.     Would stakeholders like to be involved  in creating and implementing the processes
              to make the suggested routes of identification a reality?  How do stakeholders
              envision their involvement?

•      Moving from Stage I to Stage II - Developing criteria to move contaminants from Stage
       I to Stage n (Section 2.1.2) is an area where stakeholder involvement is sought.  EPA is
       seeking feedback on whether limiting criteria should be employed or not.

       1.     If limiting criteria are employed, what are some of the specific criteria that should be
              considered, and what should be avoided?
       2.     Should separate criteria be applied  to each route of identification that leads to
              suggested contaminants,  or should  criteria  be  applied to  the  overall pool of
              contaminants once they are identified?
Working Draft                                31                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
      Stage II - Preliminary Screening

      1.     From the stakeholder experience what are the most recent, reliable models for
             calculating stability in water and fate using physical/chemical properties data that
             EPA could employ in Stage H, Preliminary Screening?
      2.     Are there data sources, databases, or data elements EPA may have missed that should
             be included in the Stage n preliminary risk assessment?
      3.     Do stakeholders have access to data not available in general technical literature that
             EPA could include in conducting the preliminary risk assessment?

      Do stakeholders have comments or suggestions relating to the development of contaminant
      profiles under Stage n and selecting contaminants for Stage m (Section 2.1.3)?

      Stage III - Ranking and Risk Assessment

      1.     Are there data sources, databases, or data elements we have missed that stakeholders
             feel should be included in our Stage m ranking analysis?
      2.     Do stakeholders have access to data not available in general lexicological literature
             that EPA could include in conducting the Stage HI ranking?
      3.     Under Stage ffl, what do stakeholders consider to be adequate data?
      4.     What data quality objectives should be defined?
      5.     What elements must be presented to judge data as adequate?
      6.     Do stakeholders want to be included on the team judging data adequacy?

      Microbial Contaminant Identification - Specific questions relating to the microbiological
      identification method as listed in Section 2.2.4 are repeated below:

      1.     There are two main schools of thought for identifying and prioritizing pathogens, as
             follows.
             a)     EPA prepares a list of known and potential waterborne disease agents, and
                    has it peer reviewed by experts in microbiology and public health within
                    EPA, Center for Dise.ase Control  (CDC), states, academia, etc.  Selected
                    experts in microbiology and public  health  would meet to discuss  each
                    pathogen. The group would determine which pathogens should appear on the
                    Contaminant Candidate List, and the relative priority of each.
             b)     Develop  a  conceptual risk-based approach with weighted criteria for
                    identifying and prioritizing pathogens.

             Is there a model that is both practical and beneficial for identifying and prioritizing
             pathogens? If so, what is its general structure?

             If approach (a) is used, what peer review, should be used on this list of identified
             organisms and their relative priority?

      2.     What role should states and the water industry have if approach (a) is used?
Working Draft                                32                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
       3.     To what extent should the relative source contribution be used in identifying and
             prioritizing organisms (e.g., food vs. inhalation vs. drinking water)?

•      General - What issues would stakeholders like to be involved with, and how do stakeholders
       want to proceed with future involvement?

       4.2    Methods for Stakeholder Involvement

       There are several mechanisms currently available for stakeholder involvement in the CIM.
Potential mechanisms include: stakeholder meetings, comment/response to Federal Register notices,
NDWAC, and FSTRAC. These mechanisms are described in Exhibit 4-1 including advantages and
disadvantages.
Working Draft                               33                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
                  Exhibit 4-1  Mechanisms for Stakeholder Involvement
Mechanism
Stakeholder
meetings



Publishing in
the Federal
Register




National
Drinking
Water
Advisory
Council
(NDWAC)



Federal-State
Toxicologica
1 and Risk
Analysis
Committee
(FSTRAC)


When,
How often?
As needed




As needed






Semi-annual,
or quarterly
meetings






Semi-annual
meetings






Operation
Provide information and
obtain feedback from a
diverse group> who have
interest and have the
opportunity to participate
Provide information and
obtain feedback from a
diverse group who have
interest and take the
opportunity to comment
on the information
provided
Agenda items; are
proposed in advance for
informational or
consultative briefings,
with the possibility of
projects being taken on ,
and feedback is obtained
from a water industry and
academic constituency
Agenda items; are
proposed in advance for
informational or
consultative briefings,
and feedback is obtained
mainly from a federal-
state regulatory
constituency
Advantages
Face-to-face
communication
, time to fully
discuss issues

Minimal cost,
no travel
required




Meeting
mechanisms
are already in
place





Meeting
mechanisms
are already in
place




Disadvantages
Travel costs, time
investment



Background material
may not be readily
accessible, large time
investment required,
over a defined period
some may not be
aware of the notice
Stakeholder
participation may be
limited to members
of the Council with
public comment




Stakeholder
participation limited
to members of the
Committee with
public comment



Working Draft
34
November 1996
-------
                        Conceptual Approach for Contaminant Identification
5.     OTHER RELATED ACTIVITIES IN THE DRINKING WATER PROGRAM

       There are several other related activities in the drinking water program that may have impacts
on the development of the CIM and First Contaminant Candidate List.  These activities as described
in the 1996 SDWA Amendments include: the national drinking water contaminant occurrence data
base; contaminant selection; and regulatory development.

       5.1    National Drinking Water Contaminant Occurrence Data Base

       The  1996 SDWA Amendments require EPA  to establish a national drinking water
contaminant occurrence data base.  This data base is to include data on regulated contaminants for
concentrations at or below their MCLs, data on unregulated contaminants for which public water
system (PWS) monitoring is required, and other reliable information from public and private sources.
The data base is to be operational by August 6, 1999,  and is  to have the capability to provide
information to the public in a readily accessible form. The law indicates that data from this data base
should be  used for analyzing which contaminants should be  placed on  the drinking water
Contaminant Candidate Lists to be considered for future  regulation.

       Currently, the Safe Drinking Water Information System (SDWIS) maintained by the EPA
has information on  compliance with MCLs for  most PWSs.  Concentrations of drinking water
contaminants in PWSs are routinely tested and these data are reported in local and/or State reporting
systems, but not to EPA. Additionally, testing results for 48 unregulated contaminants are required
to be reported to assist the EPA Administrator in determining whether they should be considered for
regulation.  The 1996 SDWA Amendments also require that EPA publish a list of not more than 30
unregulated contaminants every five years (beginning in 1999) to be monitored by PWSs, the results
of which are to be included in the national occurrence data base.

       The law requires that input be solicited from the SAB, states, and the public concerning the
data base structure, design, input parameters and requirements, and the use and interpretation of the
data. Additionally, officials of the National Academy of Sciences (NAS), states, and the public may
recommend contaminants to be included  in  the  data  base.  A schematic representing these
requirements and inputs is presented in Exhibit 5-1.

       5.1.1  Data  Base Development

       Because data base development is expensive, several options are being considered  to ensure
comprehensive consideration of contaminants potentially affecting PWS:

       •      Build on the existing SDWIS data base, since the reporting data will be for the same
             PWSs.
       •      Identify  reliable information from other national data bases  and establish an
             electronic connection with them; these data bases may include: STORET, National
             Water Information System (USGS), and Centers for Disease Control data bases.
       •      Identify  reliable information from  other data bases  and establish  an electronic
             connection.  These data bases may include: state and PWS data bases, university and
             research organizations data bases,  and interstate agency data bases.
Working Draft                                35                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
             Exhibit 5-1 Drinking Water Contaminant Occurance Data Base
    Scientific
   Community
    SAB States
    Public give
     input on
      design
     and use
     NAS States
       Public
    Recommend
   Contaminants
                            Regulated
                           Contaminant
                           Monitoring/
                         Reporting above,
                         at or below MCL
  Safe Drinking
     Water
   Information
     System
    (SDWIS)
by 1999 to include
 the Contaminant
   Occurrence
   Data Base
                                      089D-33
                  Link to
               Other Reliable
                Information
                   from
              Public & Private
                  Sources,
              STORET, USGS
                 PCS, CDC
                 States, etc.
                                                                     Contaminant
                                                                       Candidate
                                                                   List Development
                                                                    1998 and every
                                                                     5 years after
               Unregulated
               Contaminant
               Monitoring/
                Reporting
                  with
              representative
                sample of
            PWS< 10,000 pop.
Working Draft
             36
November 1996
-------
                        Conceptual Approach for Contaminant Identification
       To ensure the best scientific analyses of the data, data quality objectives will be established
to screen data for future use. Since unregulated contaminant data from only a representative sample
of medium and  small PWS that serve 10,000 or fewer people will be included in the data base,
factors that describe this "representative sample' will need to be carefully selected.  A flow chart of
activities for the occurrence data base development is discussed below.

       The December 1996 Stakeholders meeting serves as a starting point to obtain input from the
public and scientific community concerning the development and use of this data base and ways to
make the data readily accessible.

       5.1.2  Anticipated Activities

       Activities that need to be conducted to have an operational data base by mid-1999 may
include the following:

       •       Identify and obtain input and recommendations from key national, state, and local
              stakeholders concerned with the occurrence  data base.
       •       Determine which states currently have occurrence data in any format and which have
              them in an electronic format.
       •       Conduct a data needs analysis.
       •       Perform a feasibility analysis of enhancing SDWIS reporting capabilities.
       •       Hold a separate public meeting to obtain input on the design and use of the data base
              and which contaminants should be reported.
       •       Determine the feasibility of electronic exchange of data with other national, state,
              interstate and PWS data bases.
       •       Identify ways of making the data from the occurrence data base readily accessible to
              the public.

       5.2     Regulatory Development under the 1996 SDWA Amendments

       The 1996 SDWA Amendments repeals the 1986 requirement that EPA set standards for 25
contaminants every three years. The 1996 Amendments require EPA to determine whether or not
to issue regulations for at least five of the contaminants listed on the Contaminant Candidate Lists
every five years.  EPA is required to propose standards for each contaminant it chooses to regulate
within two years, and to issue final standards on the contaminants within 18 months  after the
proposal. The measure requires EPA to regulate only contaminants that exist or are likely to exist
in PWSs, and to utilize the "best available, peer-reviewed science" in proposing regulations.  The
1996 Amendments also require EPA to prioritize new regulation of contaminants that present the
greatest public health concern, including their effects on vulnerable populations such as infants,
children, pregnant women, the elderly, and those with serious illnesses.
Working Draft                               37                                November 1996
-------
                       Conceptual Approach for Contaminant Identification
6.    NEXT STEPS

      The next steps for EPA regarding the CIM and the First Contaminant Candidate List will be
based on the outcomes of the December 1996 Stakeholder meeting on the Conceptual Approach for
Contaminant Identification.  Activities by EPA will include reviewing the results of the stakeholder
meeting as they impact the development of the CIM and First Contaminant Candidate List.

      6.1    Follow-up Activities after the Stakeholder Meeting

      After the December 1996 Stakeholder meeting on the Conceptual Approach for Contaminant
Identification, EPA will generate a meeting proceedings document.  This proceeding document will
list the events of the meeting, summarize reviews and suggestions made by the stakeholders, and
provide a detailed discussion of follow-up activities for development of the Conceptual Approach
for Contaminant Identification.  EPA will also  identify stakeholder involvement in  the further
development of the CIM and First Contaminant Candidate List. It is anticipated that many of the
follow-up activities will be determined during the December Stakeholder meeting.

      6.2    Development and Implementation of the Contaminant Identification Method
             and First Contaminant Candidate List

      EPA will continue the development of the CIM and First Contaminant Candidate List after
the December  1996  Stakeholder  meeting on the  Conceptual  Approach  for Contaminant
Identification. EPA is required by the 1996 SDWA Amendments to produce the First Contaminant
Candidate List by February 1998 and to revise this list every five years thereafter. Due to these time
constraints, particularly for the First Contaminant Candidate List, EPA will proceed in developing
the CIM and the First Contaminant Candidate List in parallel. A schedule for these efforts will be
prepared which will include internal EPA deadlines, stakeholder involvement, and necessary reviews
(e.g., SAB).
Working Draft                               38                              November 1996
-------
                     Conceptual Approach for Contaminant Identification
                              Appendix A:

                           Section 102 of the
                       1996 SDWA Amendments
Working Draft
November 1996
-------
                         Conceptual Approach for Contaminant Identification
                              [This page intentionally left blank.]
Working Draft                                                                     November 1996
-------
                       Conceptual Approach for Contaminant Identification
                                    APPENDIX A
                      Section 102 of the 1996 SDWA Amendments

SEC. 102. GENERAL AUTHORITY.

(a) STANDARDS- Section 1412(b) (42 U.S.C. 300g-l(b)) is amended by striking Xb)(l)! and
all that follows through the end of paragraph (3) and inserting the following:
" (b)  Standards-
       " (1) Identification of contaminants for listing-
             v (A) GENERAL AUTHORITY- The Administrator shall, in accordance with the
procedures established by this subsection,  publish a maximum contaminant level  goal and
promulgate  a  national primary drinking water regulation for a contaminant (other than  a
contaminant  referred to in paragraph (2) for which a national primary drinking water regulation
has been promulgated as of the date of enactment of the Safe Drinking Water Act Amendments
of 1996) if the Administrator determines that-
                    " (i) the contaminant may have an adverse effect on the health of persons:
                    "(ii) the contaminant is known to occur or there is a substantial likelihood
that the contaminant will occur in public water systems with a frequency and at levels of public
health concern; and
                    Xiii) in  the sole judgment of the Administrator,  regulation  of  such
contaminant presents a meaningful opportunity for health risk reduction for persons served by
public water systems.
             v (B) REGULATION OF UNREGULATED CONTAMINANTS-
                    *(i) LISTING OF CONTAMINANTS FOR CONSIDERATION-
                           N (I) Not later than 18 months after the date of enactment of the Safe
Drinking Water Act Amendments of 1996 and every 5 years thereafter, the Administrator,  after
consultation with the scientific community, including the Science Advisory Board, after notice and
opportunity for public comment, and after considering the occurrence data base established under
section 1445(g),  shall publish a list of contaminants which, at the time of publication,  are not
subject to  any proposed or promulgated national primary drinking water regulation, which are
known or anticipated to occur in public water systems, and which may require regulation under
this title.
                           " (II) The unregulated contaminants considered under subclausc (I)
shall  include,  but  not  be limited  to,  substances  referred to in  section 101(14)  of the
Comprehensive Environmental  Response, Compensation, and Liability  Act  of 1980.  and
substances registered as pesticides under the Federal Insecticide, Fungicide, and Rodenticide Act.
                           "(Ill) The  Administrator's  decision whether or not  to  select an
unregulated contaminant for a list under this clause shall  not be subject to judicial review.
                    "(ii) DETERMINATION TO REGULATE-
                           " (I) Not later than 5 years after the date of enactment of the Safe
Drinking Water Act Amendments of 1996, and every 5 years thereafter, the Administrator shall,
after notice of the preliminary determination and opportunity for  public comment, for not fewer
than  5 contaminants  included on the list published under clause (i), make determinations of
whether or not to regulate such contaminants.
                           " (II) A determination to regulate a contaminant shall be based on
findings that the criteria of clauses (i), (ii), and (iii) of subparagraph (A) are satisfied.  Such
Working Draft                              A-l                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
findings shall be based on the best available public health information, including the occurrence
data base established under section 1445(g).
                           " (III) The Administrator may make a determination to regulate a
contaminant that does not appear on a list under clause (i) if the determination to regulate is made
pursuant to subclause (II).
                           " (IV) A determination under this clause not to regulate a contaminant
shall be considered final agency action and subject to judicial review.
                    v(iii) REVIEW- Each document setting  forth the determination for a
contaminant  under  clause (ii)  shall be available for  public comment  at such time as the
determination is published.
             " (C) PRIORITIES- In selecting unregulated contaminants for consideration under
subparagraph (B), the Administrator shall select contaminants that present the greatest public
health concern. The Administrator, in making such selection, shall take into consideration, among
other factors of public health  concern, the effect of such contaminants  upon subgroups that
comprise a meaningful portion of the general population (such as infants, children, pregnant
women, the elderly,  individuals with a histoiry of serious illness, or other subpopulations) that are
identifiable as being at greater risk of adverse health effects due to exposure to contaminants in
drinking water than  the general population.
             XD)  URGENT THREATS  TO  PUBLIC HEALTH- The Administrator  may
promulgate an interim  national primary drinking water regulation lor a contaminant without
making a determination for the contaminam: under paragraph (4)(C), or completing the analysis
under  paragraph  (3)(C), to address an urgent  threat  to public health as  determined by the
Administrator after  consultation with and written response to any comments provided by the
Secretary of Health and Human Services, acting through the director of the Centers for Disease
Control and Prevention or the director of the National Institutes  of Health. A determination for
any contaminant in accordance with paragraph (4)(C) subject to an interim regulation under this
subparagraph shall be issued, and a completed analysis meeting the requirements of paragraph
(3)(C) shall  be published, not later than 3 years after the date on  which the regulation is
promulgated and the regulation shall be repromulgated, or revised if appropriate, not later than
5 years after that date.
             v (E) REGULATION- For each contaminant that the Aidministrator determines to
regulate under subparagraph (B), the Administrator shall publish maximum contaminant level
goals and promulgate, by rule, national primary drinking water regulations under this subsection.
The Administrator shall propose the maximum contaminant level goal and national  primary
drinking water regulation for a contaminant not later than 24 months after the determination to
regulate under subparagraph (B), and may publish such proposed regulation concurrent with the
determination to regulate. The Administrator shall publish a maximum contaminant level goal and
promulgate a national primary drinking water regulation within 18 months after the proposal
thereof. The Administrator, by notice in the; Federal Register, may extend the deadline for such
promulgation for up to 9 months.
             %(F) HEALTH ADVISORIES AND OTHER ACTIONS- The Administrator may
publish health advisories  (which are not regulations)  or take other appropriate  actions for
contaminants not subject to any national primary drinking water regulation.
       % (2) SCHEDULES AND DEADLINES-
             " (A) IN GENERAL- hi the case of the contaminants listed in the Advance Notice
of Proposed Rulemaking published in volume 47,  Federal Register, page 9352, and in volume 48,
Working Draft                               A-2                              November 1996
-------
                       Conceptual Approach for Contaminant Identification
Federal Register, page 45502, the Administrator shall publish maximum contaminant level goals
and promulgate national primary drinking water regulations—
                    v (i) not later than 1 year after June 19,  1986, for not fewer than 9 of the
listed contaminants;
                    x (ii) not later than 2 years after June 19,  1986, for not fewer than 40 of the
listed contaminants; and
                    " (iii) not later than 3 years after June 19, 1986, for the remainder of the
listed contaminants.
              v (B) SUBSTITUTION OF CONTAMINANTS- If the Administrator identifies a
drinking water contaminant the regulation of which, in the judgment of the Administrator, is more
likely to be protective of public health (taking into account the schedule for regulation under
subparagraph (A)) than a contaminant referred to in subparagraph (A), the Administrator may
publish a maximum contaminant level goal and promulgate a national primary drinking water
regulation for the identified contaminant in lieu of regulating the contaminant referred to in
subparagraph (A). Substitutions may be  made for not more than 7  contaminants referred to in
subparagraph (A). Regulation of a contaminant identified under this subparagraph shall  be in
accordance with the schedule applicable to the contaminant for which the substitution is made.
              V(C)   DISINFECTANTS   AND   DISINFECTION   BYPRODUCTS-   The
Administrator  shall promulgate an Interim Enhanced Surface Water Treatment Rule, a  Final
Enhanced Surface Water Treatment  Rule, a Stage I Disinfectants and Disinfection Byproducts
Rule, and a Stage II  Disinfectants and Disinfection Byproducts Rule in accordance with  the
schedule published in volume 59, Federal Register, page 6361 (February 10, 1994), in table III. 13
of the proposed Information Collection Rule. If a delay occurs with respect to the promulgation
of any rule in the  schedule referred to in this subparagraph, all  subsequent rules shall be
completed as expeditiously as practicable but no later than a revised date that reflects the interval
or intervals for the rules in the  schedule.'.

(b) APPLICABILITY OF PRIOR REQUIREMENTS- The requirements of subparagraphs (C) and
(D) of section 1412(b)(3) of the  Safe Drinking Water Act as in effect before the date of enactment
of this Act,  and any obligation to promulgate regulations pursuant to such subparagraphs not
promulgated as of the date of enactment of this Act, are superseded by the amendments made by
subsection (a).

(c) CONFORMING AMENDMENTS-
       (1) Section 1415(d) (42 U.S.C.  300g-4(d)) is amended by striking * 1412(b)(3)' and
inserting  M412(b)'.
       (2) Section 1412(a)(3) (42 U.S.C. 300g-l(a)(3)) is amended by striking ^paragraph (1),
(2), or (3) of in each place it appears.
Working Draft                              A-3                             November 1996
-------
                         Conceptual Approach for Contaminant Identification
                              [This page intentionally left blank.]
Working Draft                                 A-4                                 November 1996
-------
                      Conceptual Approach for Contaminant Identification
                               Appendix B:

                      Descriptions of Stage I Initial
                           Identification Steps
Working Draft
November 1996
-------
                          Conceptual Approach for Contaminant Identification
                               [This page intentionally left blank.]
Working Draft                                                                       November 1996
-------
                       Conceptual Approach for Contaminant Identification
                              STAKEHOLDERS INPUT

DESCRIPTION:
       EPA has conducted and will continue to conduct public meetings with individual drinking
water stakeholders to solicit ideas, suggestions, and options for proceeding with specific activities
related to the drinking water program or to serve as the basis for strategic directions on program
activities. Drinking water stakeholders include individuals associated with states, water suppliers,
local governments, consumer groups, environmental organizations,  businesses and  industries,
academic institutions, the agricultural community, and  the public.  The intent of stakeholder
involvement is to provide EPA with a wide  array of viewpoints, ideas, and concerns held by
stakeholders that would be taken into consideration for activities under the drinking water program.
Specifically, for the Contaminant Identification Method, EPA is seeking input on the approach and
on potential candidates for the drinking water Contaminant Candidate List.

SOURCES OF INPUT:
       One of the questions posed to Drinking Water Stakeholders who attended the meeting on
Scientific Data Needs in March 1995 was, "What mechanisms would enable Stakeholders to submit
candidate contaminants for regulatory consideration?" The suggestions in response were  as follows:
       •      Formation of a Stakeholder advisory committee that would review and comment on
             EPA's preliminary candidate lists.
       •      Stakeholders, through the development of the candidate lists, could submit formal
             requests for candidate contaminants to be included in the Federal Register notice on
             the Contaminant Candidate Lists.
       •      A consistent series of Stakeholder meetings established with a variety  of groups,
             particularly with the public health/medical community.
       •      State drinking water administrators  should  be allowed  to  petition EPA for
             contaminants to be considered.
       In addition to stakeholder involvement, the 1996 SDWA Amendments require the Agency
to seek consultation with the scientific community, including the Scientific Advisory Board (SAB)
prior to issuing drinking water Contaminant Candidate List.

PROCESS:  Three possible mechanisms to elicit stakeholder involvement are:
1.     The Agency could hold at least two stakeholder meetings during the various stages of the
       Contaminant Candidate List process.  The first would be to invite comment and request
       additional data or recommendations.  The second would be part of the opportunity to
       comment on the Contaminant Candidate List and for the Agency to seek input on additional
       contaminants that stakeholders believe should be on the list based on available data.
2.     The Agency could provide a mechanism for stakeholders to submit written comments and
       suggestions regarding  contaminants for consideration.
3.     Periodically, likely  throughout the five years during the  Contaminant Candidate List
       development phase, the Agency could convene a  subcommittee to the National Drinking
       Water Advisory  Council (NDWAC)  that would include  various Stakeholders.  The
       Subcommittee would provide recommendations to the Agency on a set of contaminants to
       be considered  for the Contaminant Candidate List.  Formation  of  a subcommittee to
       NDWAC would eliminate any potential violations of Federal Advisory Committee Act
Working Draft                              B-l                             November 1996
-------
                        Conceptual Approach for Contaminant Identification
       (FACA), and would allow for the formation of a body of Stakeholders that could provide
       significant input.
Working Draft                                B-2                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
                      CONTAMINANTS OF PUBLIC CONCERN

DESCRIPTION:

       Emerging public concerns can help identify potential contaminants for the drinking water
Contaminant Candidate List. These concerns are especially important when consumers (or a PWS
acting on their behalf) identify specific contaminants in SDWA-regulated tap water that do not have
primary MCLs. By actively seeking input from concerned citizens, the Agency may be able to more
quickly identify emerging contaminants and take action before the problem becomes widespread and
more costly to control. Even in cases where the contaminant has not yet been found in tap water,
information that a potential source of contamination may pose a risk to a PWS source water may
provide important information for EPA's Source Water Protection programs.

SOURCES OF INPUT:

       There are several potential sources for identifying contaminants of public concern including:

       •      Routine search of published literature.
       •      Public concerns on specific contaminants mentioned on the Safe Drinking Water
             Hotline can be characterized and categorized.
       •      EPA Regional staff could be  periodically  contacted to see  what  emerging
             contaminants appear to be of most interest to the citizens in their region.
       •      The  OGWDW Web page could be updated to make it easier for Internet users to
             suggest potential candidates for the list.
       •      EPA staff could more routinely seek informal public input at the many conferences
             and public meetings they attend each year.

PROCESS:

       It is recommended that EPA form a Workgroup to identify contaminants of public concern.
This EPA Workgroup will determine the methods and procedures for sources of input. Once the
sources of input have been determined, the information generated by these sources will be reviewed
and compiled the EPA Workgroup.

       One example of data that may be required is presented by the data collection sheet on the next
page. This sheet  was developed to capture relevant information from calls into the Safe Drinking
Water Hotline. The information for this sheet would be captured by the EPA Safe Drinking Water
Hotline personnel. The data would be kept in a computer data base so that reports could be generated
periodically with  minimal effort. The Workgroup, however, needs to clear any method of collecting
data so that it is in line with OMB "survey" restrictions.
Working Draft                               B-3                             November 1996
-------
                      Conceptual Approach for Contaminant Identification
                      SAFE DRINKING WATER HOTLINE
                         DATA COLLECTION SHEET:

                  Data Requested from Hotline Calls Concerning
                       Contaminants without a Primary MCL
         1.  Date of call
         2.  Contaminant name
         3.  Has it been found in tap water?
                    If so, what concentrations (if known)?
                    Source of the contaminant (if known)?
         4.  If it has not been found in tap water, why is it a concern?
         5.  Is it likely to pose a threat to a drinking water source?
         6.   State/County/City/Zip-Code information (if the caller provides).
Working Draft
                                       B-4
November 1996
-------
                       Conceptual Approach for Contaminant Identification
              CHEMICALS REGULATED BY OTHER EPA PROGRAMS

DESCRIPTION:

       In developing the drinking water Contaminant Candidate List, EPA will review chemicals
currently regulated or being considered for regulation under other EPA and federal programs with
the potential for occurring in drinking water. EPA will access major internal EPA regulatory lists
and toxicity/hazard, production/exposure,  monitoring  and risk information  databases to help
determine other regulated contaminants.

SOURCES OF INPUT:

       To gather an initial list of potential contaminants, the Office of Pollution Prevention and
Toxic  Substances (OPPTS) Screening Information Systems/LAN (SIS/L) database can be
searched to identify what chemicals other EPA and federal programs are working on.  This database
can also be used to access toxicity/hazard, production/exposure, monitoring and risk information
relating to chemicals which are  likely to  occur in drinking water.  The  SIS/L  contains a cross
reference to major non-commercial databases that EPA and other federal agencies maintain for each
non-confidential chemical on the Toxic Substances Control Act (TSCA) Chemical Inventory.  For
many of the sources referenced, the system can pull data from  the system,  or at least pointer
information. A key database in SIS/L is the Register of Lists (ROL) which contains  33 major EPA
lists of regulated chemicals and other materials and the contact for each list. Some other examples
of lists and databases in SIS/L are:

•      CUS86 and  CUS90: Aggregate non-confidential production volume data  on the TSCA
       Inventory entered into the Chemical Update System for the 1986 and 1990 reporting period
       which updates the Inventory.
       TSCA Test Submissions (TSCATS) database references to unpublished 8(d) Health and
       Safety Studies and 8(e) Notices of Substantial Risks for chemicals with  new data that
       indicates risk.
•      NIOSH lists the results  of the National Institute  for  Occupational Safety and  Health
       Occupational Exposure Survey.
•      SIDS lists chemicals being  evaluated under the Screening Information Data Set program
       under the OECD.
•      302 lists chemicals under  the Community Emergency Preparedness Program from Section
       302 of Emergency Planning and Community Right-to Know Act (EPCRA), which are listed
       because of their known extreme acute toxicity.

PROCESS:

       The first step will be to check the ROL and request each list on which chemicals are likely
to occur in drinking water and to obtain data that supports this determination. The second step will
be a search of SIS/L for the chemicals which are identified for a cross reference list to other sources
of Agency and federal data. The chemical will then move to Stage n for a preliminary assessment
using the SIS/L cross reference list to dig deeper into the sources identified.
Working Draft                              B-5                              November 1996
-------
                       Conceptual Approach for Contaminant Identification
                  CONTAMINANTS WITH KNOWN OCCURRENCE

DESCRIPTION:

       "Contaminants with known occurrence" are contaminants which have been found in PWSs
and/or in source waters.  In many cases, all that is  known  about these contaminants  is their
occurrence in drinking water or source water and nothing is known about their toxicity or exposure.
       These contaminants may in fact pose substantial risk to human health due to the toxic effect
of the compound and/or the possible level of exposure to the chemical. Toxicity and exposure may
be due to several factors including the amount of a compound produced, amount of possible release
to source waters, the physical/chemical properties of a compound, and chemical fate. The SDWA
Amendments of 1996 requires that any compound that "occurs" during monitoring of unregulated
contaminants be reported to EPA.

SOURCES OF INPUT:

       Potential sources of occurrence data include:

       •      Search the Safe Drinking Water  Information System National Occurrence
             Database (SDWIS)  for all unregulated drinking water chemicals and regulated
             chemicals at or below the  MCL in  SDWIS. Search the Storage and  Retrieval
             System (STORET) to identify chemicals that appear in ambient source  water that
             could eventually become drinking water contaminants.
       •      Identify physical chemical properties - Search databases to obtain data on the
             physical chemical properties of the compounds to obtain water solubility data. This
             would be somewhat of a preliminary screen. Chemicals mat are highly insoluble may
             be of lesser concern that those that are more water soluble.
       •      Monitoring records of PWSs and state programs - A number of state drinking
             water programs and individual PWSs routinely monitor for other contaminants in
             addition to those with federal NPDWRs or unregulated  monitoring requirements.

PROCESS:

       Occurrence of chemicals that have been reported to the Agency can be tracked in SDWIS.
In addition, voluntary ambient monitoring results can  be tracked in STORET. The results of the
searches can be  cross referenced with the Contaminant Candidate List. Additional chemicals that
are  similar in structure and function (chemical analogs) can be identified by way of a chemical
substructure search.  Databases can be searched  to  obtain informalion on  physical chemical
properties and to obtain data on other possible existing EPA regulations. These data will help in
performing a preliminary screening  of the chemical(s) under Stage EL. If information is found to
support a decision that the contaminants will be of little risk, no further assessment will be made and
the compound will be considered a contaminant of no/low risk. Conversely, if there is no  regulatory
information found to support a conclusion of no/low risk,  other EPA, federal, and commercial
databases can be searched to obtain additional information to support the continued inclusion of the
chemicals for the Candidate List.  In addition, EPA Committees, workgroups, teams and  Agency
experts will be consulted to support decision-making.
Working Draft                              B-6                             November 1996
-------
                       Conceptual Approach for Contaminant Identification
                    CONTAMINANTS WITH KNOWN TOXICITY

DESCRIPTION:

       Periodically chemicals of known toxicity, but with no MCLs, have been identified by various
sources.  These sources include the professional judgement of EPA staff and concerns by other
countries about particular chemicals.  These chemicals of known toxicity may be a discrete chemical
compound, a mixture, a class of compounds, or a byproduct of another compound. When such
chemicals are made known to EPA, they can be incorporated into the chemical identification process
for the drinking water Contaminant Candidate list.

SOURCES OF INPUT:

       In is expected that there will be a variety of sources for contaminants of known toxicity.
Examples include: internal EPA professional staff, other EPA and federal programs (e.g., CERCLA
and FIFRA), various health institutions, analogs of toxic chemicals, and public concern.

PROCESS:

       Contaminants of known toxicity will be identified for their discrete CAS registry number.
Additionally, analogs, mixtures, and byproducts will be identified and checked for production and
use. The contaminants can be cross referenced with the Contaminant Candidate List, SDWIS and
STORET to ensure that none are currently being regulated, are on the unregulated monitoring list,
or are on the drinking water Candidate list.  The availability of possible ambient source water data
can also be sought.
       Databases can be searched to obtained information on physical chemical properties such as
solubility and to obtain data on other possible existing EPA regulations.  These data can help in
performing a preliminary screening of the chemical(s) under Stage n. If information is found to
support a decision that the contaminants will be of little risk, no further assessment will be made and
the compound will be considered a contaminant of no/low risk. Conversely, if there is no regulatory
information found to support a conclusion of no/low risk, other EPA, federal and commercial
databases will be searched to obtain additional information to support the continued inclusion of the
toxin(s) as contaminants for the  Contaminant Candidate List.  In addition, EPA Committees,
workgroups, teams and Agency experts will be consulted to support decision-making.
Working Draft                              B-7                             November 1996
-------
                       Conceptual Approach for Contaminant Identification
                         HIGH PRODUCTION CHEMICALS

DESCRIPTION:

      Annual production volumes of chemicals represents the amount available for commercial use
and is often used as an indicator of potential exposure.  Production volumes do not, however,
indicate the amount of the contaminant released into the environment nor are production volumes
a direct measure of exposure for chemicals.

SOURCES OF INPUT:

      Production data are available in several different sources which will require in-depth review
for comprehensiveness and completeness in the test phase of the Contaminant Identification Method.
These sources include:

      EPA/OPPTS United States High Volume Chemical List. This list is accessible through
      the EPA Office of Pesticides, Prevention and Toxic Substances (OPPTS) Screening
      Information System/LAN (SIS/L) System and lists every chemical with a CAS number
      from the OPPTS 1986 and 1990 Chemical Update System (CUS) which the EPA has
      determined is a "High Volume" chemical. Currently this means having an aggregate non-
      CBI (non-TSCA Confidential Business Information) production volume of greater than 1
      million pounds in either the 1986 or 1990 CUS.  This list, however, is considerably less than
      the complete list which includes CBI production volume data.  CBI data is unavailable. CUS
      contains chemical production information which is reported to EPA every four years to
      update the Toxic Substances Control Act (TSCA) Chemical Inventory.

PROCESS:

      These databases can be searched once during each cycle of the Contaminant Identification
Method process. In the case of establishing a different production volume than that found in the high
volume chemical list, OPPTS can conduct an aggregated search of CUS for selected contaminants
based on drinking water needs.
Working Draft                              B-8                             November 1996
-------
                        Conceptual Approach for Contaminant Identification
                            HIGH RELEASE CHEMICALS

DESCRIPTION:

       Release of chemicals to the environment usually is the result of that portion of chemicals
released during production, use, or disposal. Releases to water are discharges to different water
bodies as the result of waste disposal permits or production processes. Other discharges which may
impact the quality of ground and surface water are those that discharge to Underground Injection
Wells, landfills, and hazardous waste sites. The amount released tempered by the contaminants
environmental fate can serve as a rough indication of the potential of the contaminant to occur within
drinking water supplies.   Release data is often used with production volume and use data in
conducting risk assessments. All three (i.e., release data, production volume, and use data) offer an
additional screening level for further risk analysis.

SOURCES OF INPUT:

       Release data are available in several different sources which will require in-depth review for
comprehensiveness and completeness in the test phase of the Contaminant Identification Method.
These sources include:

•      Permit Compliance System (PCS). PCS is an information management system maintained
       by the Office of Wastewater Enforcement and Compliance (OWEC), to track the permit,
       compliance and enforcement status of facilities regulated by the NPDES program under the
       Clean Water Act.  PCS tracks information for over 65,000  permittees on wastewater
       treatment and pollutant measurements of industrial  and federal facilities discharges into
       navigable waters.  PCS includes data on the mass of pollutant discharged by  a facility if
       required under their permit and also data on the release of contaminants through wastewater
       discharges to surface water of the United States.

•      EPA  Toxic Chemical Release Inventory Database.  TRI is an EPA database which
       contains information on the annual estimated release of toxic chemicals to the environment
       as reported  by industry as mandated under Title HI of the Superfund Amendments and
       Reauthorization Act (SARA) of 1986.  The database currently contains 619 chemicals and
       includes amounts released to the air, land, water and transferred to waste sites.
PROCESS:

       These databases will be searched once during each cycle of the Contaminant Identification
Method process.
Working Draft                               B-9                              November 1996
-------
                        Conceptual Approach, for Contaminant Identification
                           DIRECT/INDIRECT ADDITIVES

DESCRIPTION:

       Additives are organic or inorganic chemicals which enter drinking water during the treatment
and distribution processes.  The term "additive" refers to the chemical product as  well  as to
impurities in the product.  Additives are divided into two categories: direct and indirect. Direct
additives are intentionally added to a water supply to improve its quality for public consumption.
Examples of direct  additives include alum, lime, coagulant aids, and activated carbon.  Indirect
additives are contaminants which can leach or dissolve into a water supply from materials in contact
with the water.  These materials include components of pipes, tanks and equipment coatings,
lubricant and sealants.

       In the past, the EPA assisted states and public water systems on the use of materials and
contact surfaces in drinking water through the issuance of advisory opinions on acceptability of many
additive products.  This federal advisory program for additives was terminated in October 1988, and
the activity was transferred to the private sector. The National Sanitation Foundation International
(NSF) Ann Arbor, Michigan and Underwriters  Lab,  in Northbrook,  Illinois,  offer  testing,
certification, and listing services for drinking water additives.  Discontinuation of the additives
program at  EPA,  however, did not relieve the Agency  of  its  statutory responsibilities.  If
contamination resulting from third party sanctioned products seems likely, EPA can address the issue
with appropriate drinking water regulations or other  actions authorized by the SDWA.

SOURCES OF INPUT:

       Specific sources in input on additives include:

       •      NSF, Underwriters Lab,  etc. - Contaminants could be  suggested by NSF,
              Underwriters Lab, and other organizations involved in testing and certification of
              additives based on testing results of the additives program.

       •      Literature  search - Searches could be conducted of the scientific literature to
              identify other drinking water additives of concern.

PROCESS:

       Periodically, during the Contaminant Candidate List development phase, the NSF and other
testing and certifying organizations would be consulted and a literature search would be conducted
to identify direct and indirect additives which may also be contaminants of concern.
Working Draft                              B-10                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
  PESTICIDES WITH HIGH LEACHING POTENTIAL/HIGH RUN-OFF POTENTIAL
DESCRIPTION:

       Leachability is usually employed to characterize pesticides. Leachability includes both
mobility and persistence factors. Leaching refers to the downward movement of pesticides into the
soil An EPA screening technique has been identified regarding leaching potential. This leaching
index was derived based on physical-chemical properties, including half-life in soils and partition
coefficients between soil organic carbon and water (K^.) using the Groundwater Ubiquity Score,
(GUS)4 as shown below:

                             GUS = log.oCt,/"1) x (4 - log^KJ)
       The leaching potential is described by the GUS using the following ranges:
Leaching Potential
Extra Small
Small
Medium
Large
GUS Range
<0
0-1.8
1.8-2.8
2.8-6
       Run-off potential is generally characterized by K^, alone, the 1^  values that would be
regarded as an indication of high run-off potential is yet to be determined.

SOURCES OF INPUT:

       The leaching and run-off potentials can be derived from readily available chemical  and
physical constants.  It is anticipated that the value of these constants for a given contaminant would
be made available from EPA's pesticide program or will be derived from available databases of
chemical and physical constants such as Merck Index.

PROCESS:

       FIFRA pesticides, without NPDWRs, identified as having high (large) leaching potentials,
based on GUS calculations or high run-off potentials using a yet to be determined scale could be
automatically included as contaminants to be considered.
        Gustafson, D I, Groundwater Ubiquity Score: A Simple Method for Assessing Pesticide Leachability,
Environmental Toxicology and Chemistry, Vol 8, pp 339-357, 1989.
Working Draft
B-ll
November 1996
-------
                       Conceptual Approach for Contaminant Identification
                             SURROGATE CHEMICALS

DESCRIPTION:

       Surrogate chemicals are chemicals that may have adverse effects on human health or the
environment based on  their similarity to other  chemicals with known  health effects and/or
environmental fate and transformation. A common measure of the similarity of chemicals is their
Structure Activity Relationship (SAR).

SOURCES OF INPUT:

       Sources for identifying chemicals based on their SAR include:

       •      Chemical structure/substracture/Chemical Abstracts Service (CAS) registry
             number search  - Search chemical structure databases! to identify substructures,
             analogous compounds and their corresponding CAS registry numbers. Conducting
             this search can help to identify compounds and chemical classes that may not have
             been previously monitored for, but due to chemical structure, may be functionally
             similar to chemicals that were identified in the monitoring process.

PROCESS:

       The process for identifying chemicals with similar structures would include: (1) identifying
a short list of contaminants that have known environmental effects; including chemicals already
regulated under the SDWA and chemicals on the Contaminant Candidate List; (2) identifying the
chemical surrogates and analogs using the  CAS registry number search; and (3) considering the
addition of these surrogate chemicals to the Contaminant Candidate List via EPA Workgroup
deliberations.
Working Draft                              B-12                             November 1996
-------
                       Conceptual Approach for Contaminant Identification
                             ENDOCRINE DISRUPTORS

DESCRIPTION:

       Concern that certain chemicals implicated as hormone disrupting may pose a substantial
threat to human health has now been widely recognized. These chemicals (often termed "endocrine
disruptors") can lead to immune, behavioral, and reproductive changes with risks for individuals,
populations, and wildlife species. While it was recognized as far back as 1962 that some synthetic
chemicals can cause reproductive problems in birds, it has only been since the early 1990s that the
scope of the potential hazard to humans has been understood.

SOURCES OF INPUT:

       EPA has funded a National Academy of Sciences  (NAS) study on "Hormone-related
Toxicants in the Environment." A multidisciplinary Committee is currently completing its report
which is expected in the summer of 1997. The NAS Committee members have been selected for
their expertise in human toxicology, wildlife toxicology, developmental biology, endocrinology,
pharmacology, physiology, biochemistry, environmental chemistry, chemical engineering, ecology,
biostatistics, and epidemiology.

PROCESS:

       The EPA team will use the findings of the 1997 NAS study along with other related Agency
reports and studies to identify contaminants that are not currently regulated and should be considered
for the Contaminant Candidate List. In the short-term, EPA can gather occurrence data for several
of the contaminants that are suspected of being endocrine disruptors.
Working Draft                              B-13                             November 1996
-------
                       Conceptual Approach for Contaminant Identification
                         CHEMICALS OF INTERNATIONAL
                        INTEREST AND GLOBAL CONCERN
DESCRIPTION:
       Chemicals of global concern include various contaminants have been released by human
activities which have contaminated surface water on a global basis at very low concentrations (e.g.,
tritium from nuclear weapons tests). As noted below, new contaminants continue to be recognized
as potentially posing global contamination of isurface water (and subsequent contamination of ground
water as well). To address these concerns, the EPA will continue to seek out global contaminants
that may potentially be significant for drinking water safety.
       Several chemicals have solicited international interest due to their toxic effects on the
environment.  In  1989 the Organization for Economic Cooperation  and Development (OECD)
spearheaded a systematic effort to investigate existing chemicals, those already in use (some for a
long time), whose safety had not been evaluated.  Most of the chemicals are traded throughout the
world, and many are produced by a small number of multi-national chemical companies. It was
decided to share the burden of compiling data, and testing chemicals to fill data gap among 16 of the
24 member countries. Chemicals that are produced in high volume globally are potentially those that
constitute those with the highest risk of exposure to  humans and the environment.  An initial
representative list  of  1,500 high production volume chemicals  was prepared and each year
approximately 50 are reviewed  and further testing conducted. The OECD countries developed the
basis battery of necessary data - the Screening Information Data Set (SIDS) - to provide baseline
information. The SIDS includes: chemical identity; physical-chemical data; sources and levels of
exposure; environmental fate and pathways; ecotoxicological data; and lexicological data. The result
is a SIDS Initial Assessment Report (SIAR) mat covers potential risks and identifies whether further
testing or government policies  to minimize risk are needed.

SOURCES OF INPUT:

       One source for chemicals of global interest includes the routine review of available published
literature:
       •     Published Literature - As  a matter of routine, EPA will continue to monitor the
             standard international scientific  publications  (e.g., Nature, New Scientist, and
             Science) for items that may be of particular interest.

       For chemicals of international interest, SIARs are collected and made available through two
sources which can both be searched for contaminants for the initial list:

             The International Register of Potentially Toxic Chemicals (IRTPC) Database.
              This is a United Nations Environment Programme (UNEP) database.
       •     The International Uniform Chemical Information Database. This is a European
             Union database.
PROCESS:
       EPA will continue to monitor published literature on an active basis for items that may be
of interest for chemicals of global interest. The comprehensive SIDS  cheitnicals and interim and final
reports will be reviewed during each cycle of the Contaminant Identification Method process.
Working Draft                              B-14                             November 1996
-------
                     Conceptual Approach for Contaminant Identification
                             Appendix C:

                       Toxicity Ranking System
                         for the Contaminant
                        Identification Method
Working Draft
November 1996
-------
                          Conceptual Approach for Contaminant Identification
                              [This page intentionally left blank..]
Working Draft                                                                       November 1996
-------
                        Conceptual Approach for Contaminant Identification
                                     APPENDIX C:

                      TOXICITY RANKING SYSTEM FOR THE
                    CONTAMINANT IDENTIFICATION METHOD:
       The Methodology Options document (USEPA, 1994) provided a historical perspective of
previous public and private efforts to develop chemical ranking schemes based on toxicity, and
highlighted some of the practical issues and limitations that impact the selection, modification or
development of such methodologies. A number of references considered particularly valuable for
their review of these considerations, and/or for their presentation of selected collections of ranking
schemes, were  noted (Davis et al., 1994; Environ Corporation, 1986; Foran and Glenn, 1993;
Hushon and Kornreich, 1984; Ross, 1994).

       One important attribute required of the selected methodology is that it be able to rank or
categorize chemicals that have very little available toxicity information. Under such circumstances,
toxicity rankings can identify chemicals of potential regulatory concern, although they can generally
be anticipated to provide an inadequate basis for direct regulatory action. Therefore, another useful
scheme attribute is the capacity to provide an indication of the completeness and/or the general
quality of the available toxicity data for each input chemical.  In conjunction with appropriate
exposure or production data, these output characteristics could assist  in prioritizing its research and
information-gathering needs.  The methodology derived in this report has incorporated both of these
key attributes.

       Another goal is that the ranking methodology be reasonably simple, understandable, and
readily adaptable in terms of accommodating desired adjustments in scoring endpoints, parameters
and criteria.  Furthermore, to expedite acceptance of the methodology by the scientific, regulatory
and regulated communities, the approach was to be largely derived from methodologies previously
accepted by one or more of those communities (i.e., from approaches published in the peer-reviewed
scientific literature, previously reviewed and/or utilized by EPA or other regulatory bodies, etc.).
This revised methodology substantially reflects these considerations, without at the same time
significantly compromising important performance objectives.

       Although a system that segregates chemicals into a small number of toxicity categories may
be adequate in many situations, such an approach comes at a cost of losing "prioritizing resolution."
This revised methodology, therefore, rank all input chemicals individually (additional criteria or
professional judgment can always group ranked chemicals into functional categories, while the later
ranking of categorized  chemicals  seems certain to be less readily accomplished). For similarly
practical reasons, the chosen scheme ranks all  input chemicals by toxicity in a single list, with
separate appended indications of data sufficiency and quality. The rejected alternative was to first
segregate input chemicals into data sufficiency/quality categories, and then to rank them individually
according to toxicity within such categories, thus generating multiple prioritized toxicity lists.

       As further described below, provisions for the use of SAR/QSAR techniques have been
incorporated into the methodology in order to permit minimizing the instances in which input
chemicals are unable to be ranked.  Similarly, while the revised scheme of systematizes and
Working Draft                                C-l                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
standardizes the ranking procedure through the use of specified endpoints, criteria, procedures and
guidance, it recognizes that professional judgment does or may play a significant role (e.g., in the
application of SAR/QSAR, the evaluation of severity-of-effect and data quality ranking, rejection
or qualification of potential input data, etc.).  When a factor, the methodology provides that the
exercise of professional judgment be evident and documented, so that if desired,  it can  be
independently corroborated.

       Although "potency" (P) is a key component of a chemical's toxicity ranking in this scheme,
for some toxicity characteristics it has been augmented or replaced by "weight-of-evidence" (WOE)
and/or "severity-of-effect" (SOE) standards.  As indicated in the Methodology Options document
(USEPA, 1994), these approaches are common to many of the considered options. Where the nature
of the toxicity characteristic considered or Ihe type of test  system utilized makes quantitation of
chemical potency difficult or of particularly questionable relevance to humans, the WOE approach
provides another valid means of establishing the level of  concern that should attach to a given
chemical. In effect, the greater the weight of evidence mat the toxic effect is real and is relevant to
or likely to occur in humans, the higher the chemical's WOE score amd the greater the level of
concern merited by the chemical.  The utility of employing severity of effect to rank the toxicity of
chemicals has been discussed at length by the Environ Corporation (1986).  This approach seeks to
remedy the  type of situation where two chemicals of equal potency receive equal toxicity rankings,
even though at the given dose one produces a mild effect (e.g., reversible narcosis), while the other
results in a very serious one (e.g., irreversible damage to the central nervous system). Despite the
complications of professional judgment thai:  are involved, a methodology capable of ranking the
latter chemical over the former is generally considered desirable.

       Weighting of toxicity characteristics/endpoints and/or routes of exposure is an  additional
scoring feature  utilized by some of the ranMng schemes reviewed in the Methodology Options
document (Life Systems, 1994). As detailed below, the revised methodology no longer incorporates
a blanket preference for oral and dermal data over all inhalation data (other routes are not explicitly
considered), nor does it  retain  the original degree of emphasis on  chronic/subchronic  (and
reproductive/developmental) effects versus acute effects. The later preference was conceptually and
practically interrelated with severity-of-effect  and level-of-exposure considerations. Incorporation
of these preferences had been predicated on what is anticipated to be the principal use of the
methodology - to assess human risk from exposure to low levels of chemicals in water, generally for
substantial periods of  time, with emphasis  on effects that are particularly adverse.  However,
principally  in recognition of the fact that level-of-exposure considerations are anticipated to  be
factored in  utilizing a separate "exposure" methodology during the overall prioritization process,
these biases in the originally proposed toxicity ranking methodology have, at least in a theoretical
sense, been substantially reduced or eliminated.  In other words, for the lowest dose category and/or
the most severe effect (or highest WOE category) for six of the eight toxicity characteristics, the
maximum obtainable score is now the same (150). Largely because of dose constraints, however,
it is anticipated that maximum  scores will be more difficult to  achieve  for  some toxicity
characteristics (e.g., those dealing  with acute toxicity)  than for others.  The rationale behind
assigning less weight to the Mammalian  Sldn/Eye Irritation or sensitization and the Reportable
Quantity (RQ) toxicity characteristics is disc ussed in the "Methodology Description" Section.
Working Draft                               C-2                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
       Because one of principal objectives is to utilize the ranking methodology as a preliminary,
prioritizing tool, an attempt was made to have it accommodate the types of information likely to be
available in various data bases and secondary literature sources. This should help to minimize the
needs for professional judgment and to categorize input chemicals as "unrankable."

The perfect ranking system, even for a defined purpose, probably does not exist. More than one
approach could be devised, each with distinct advantages and disadvantages, and each capable of
being a tool to provide useful results. The methodology described in this document has a solid
foundation in that  it is largely based upon schemes that have some degree of actual use and/or
acceptance in the scientific  community.  While other variations could certainly be proposed for
identifying the toxicity characteristics and their scoring schemes, it appears unlikely that they would
provide significant  advantages over the outlined approach. More meaningful choices probably lie
in how to manipulate the individual toxicity characteristic scores to arrive at a final ranking. For
example, a decision to force the entry of a score for each characteristic (using professional judgment
or default values when real or derived data are not available) would permit the ranking to be based
upon aggregate scores without sacrificing the attribute of characteristic weighting.

       However, this methodology appears sufficiently rational (given the inherent arbitrariness of
such ranking schemes) and reasonably straightforward. It has the flexibility to handle different types
of input data (real  and derived), but does not force unnecessary or unreasonable "guessing."  It
should also be reasonably adaptable, and provides not only a chemical ranking, but also indicators
of data sufficiency and quality. Finally, it has the arguable advantage of utilizing WOE and SOE
types of information to rationally replace or supplement potency data.

METHODOLOGY DESCRIPTION

       This  section provides the functional  description of the revised  ranking methodology,
including  specifics about operational procedures, toxicity characteristics, scoring criteria, the
handling of data gaps and uncertainties, and the scheme's several evaluational outputs.  As noted
previously, this methodology has been derived from a number of previously published schemes that
have already received peer and/or regulatory agency acceptance, and that have to varying degrees
undergone some measure of  validation or actual regulatory use. It is hoped that this approach will
economically maximize the benefit that can be gained from prior efforts  in the field of toxicity
ranking, as well as  facilitate  the proposed methodology's acceptance.

       In terms of the schemes presented in the Methodology Options paper (USEPA, 1994), this
methodology has been derived principally from portions of the Option Three schemes of Sampaolo
and Binetti (1989,  1986) and O'Bryan and Ross (1988). The former has  been  "included, as the
official method, in the Italian decree for the implementation of the European Community Directive
82/501" (Sampaolo and Binetti, 1989), while the latter is based on a method developed by the
Chemical Effects Information Task Group of the Oak Ridge National Laboratory in collaboration
with EPA's Office of Toxic Substances (OTS), and has been used to help implement the mandates
of the Toxic Substances Control Act (TSCA) (O'Bryan and Ross, 1988).

       Additionally, certain  scheme elements or concepts have been at least partly drawn from the
Option One  approach developed originally for the Ontario Ministry of the Environment (ARET,
Working Draft                               C-3                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
1993; CLC, 1992; OME, 1992, 1989, 1989), from the Option Four "Environmental Priority Setting"
(EPS) system of Germany (Klein et al., 1988; Weiss et al., 1988), from the severity-of-effects
document prepared for EPA's Environmental Criteria and Assessment Office (Environ Corporation,
1986), and to varying degrees from the four Option Two approaches: (1) the RQ scheme developed
for use with the Comprehensive Environmental Response, Compensation  and Liability  Act
(CERCLA) and the Clean Water Act (CWA) (DeRosa et al., 1984; USEPA, 1987, 1985), (2) the
Great Lakes Basin Sunsetting scheme (Foran and Glenn, 1993), (3) the Michigan Critical Materials
Register scheme (MDNR, 1987), and finally (4) EPA's Chemical Use Clusters  Scoring scheme
(USEPA, 1993).  As previously noted, this current version of the methodology also incorporates, a
number of comments and adjustments suggested by internal Agency reviewers (USEPA,  1996b), and
approved for incorporation by the EPA WAM (USEPA 1996a).

       General Procedures

       When undertaking an effort to rank an input list of chemicals according to their toxicity, one
of the first major decisions to be made concerns the input toxicity data.  Essentially, it must be
determined what level of effort and expense is to be expended searching for data, and  from which
sources.  This can, of course, vary according to the specific objectives of the ranking exercise in
question.  At the one extreme, a data search could consist of querying only one or  two on-line
information sources, while at the other, a fairly exhaustive literature search of printed and on-line
sources can be made.

       For a ranking exercise that is intended primarily as a preliminary, prioritizing tool, and where
the level of effort expended must be somewhat limited, information sources such as those presented
in the following non-inclusive  list are suggested:

     •   The Integrated Risk Information System (IRIS), on-line EPA database
     •   Health Effects Summary Tables (HEAST), USEPA printed database
     •   Registry of Toxic Effects of Chemical Substances (RTECS), on-line National Library of
         Medicine (NLM) database or the printed version (1985-86 and supplements) - compiled
         by the National Institute for Occupational Safety and Health (NIOSH)
     •   Hazardous Substance Data Bank (HSDB), on-line NLM/TOXNET database (scientifically
         reviewed and edited)
     •   Chemical Carcinogenesis Research Information System (CCRIS), on-line NLM/TOXNET
         database sponsored by the National Cancer Institute (NCI)
     •   EPA Genetic Activity Profile (GAP) and Toxic Activity Profile: (TAP) computer databases
         (critically peer-reviewed data and graphical profiles of genotoxicity and classical toxicity)
     •   Reportable Quantity (RQ) list (40 CFR 302.4, Table  302.4., supplemented by Federal
         Register  updates) of hazardous substances under CERCLA, CWA, the Resource
         Conservation and Recovery Act (R.CRA) or the  Clean Air Act (CAA)

If it is determined that insufficient information is derived from these sources, further searching of
additional standard references or on-line  databases such as those listed in Appendix  1 of the
Methodology Options document (USEPA, 1994) may be conducted. To help conserve resources,
search strategies could involve prioritized information sources within several search "elements", for
example:
Working Draft                               C-4                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      •  HSDB—>IRIS—>HEAST—>CCRIS—>RTECS (for carcinogenicity)
      •  HSDB—>EPA/GAP—>CCRIS/RTECS (for genotoxicity)
      •  HSDB—>IRIS—>HEAST—>EP A/TAP—>RTECS (for other toxicity)
      •  40 CFR 302.4, Table 302.4 (for RQs)

Searching could be halted at any point along these paths whenever the cumulated information was
deemed sufficient according to certain predetermined decision criteria.  Using the above "preliminary
prioritization"  example,  it  might  be  determined  that  a high  score  for  carcinogenicity,
reproductive/developmental toxicity or chronic toxicity based on data initially found in HSDB would
preclude looking for any further data beyond one more resource level (in this case, beyond IRIS and
EPA/GAP).

      Generally speaking, data from the most sensitive test system within a toxicity characteristic
will form the basis of that characteristic's score, unless alternate data is clearly more relevant to
human exposure. Where multiple data have been retrieved for a toxicity characteristic, professional
judgement may be used to eliminate certain entered studies from those that are available to serve as
the basis for the toxicity characteristic's score, or in some cases to enter a score  that combines
professional evaluation of the available data with other relevant professional expertise.  This could
be appropriate when a substantial portion of the data indicates a lower level of toxicity, and when
such data is equally relevant to human exposure as the most sensitive test system data (or is from
other apparently valid studies using the same system). Methodology input and output forms (see
Section 3.5 and Tables 8,9 10) will provide at least cursory documentation of most such professional
judgment that directly impacts upon toxicity characteristic scores, and a ranking exercise summary
report may always include any additional explanation that is considered  necessary.

      When  searching and data evaluation activities have been completed, with or without the
generation or acquisition of supplemental SAR/QSAR information, the selected data will be entered
onto appropriate data input forms (see Table 3-8).  From these forms, the toxicity  information is
subsequently  entered into a custom-developed database application that manipulates the input data
in order to generate both individual chemical toxicity scores and a listing of the input chemicals
ranked by toxicity score,  as well as  other associated output that is  further described below.  If
desired, it would be possible for the person evaluating the toxicity information to enter it directly into
the computer  database, thus eliminating the data form entry step.  This approach would simplify the
data entry procedure, but could also diminish the extent and/or accessibility of documentation
regarding data input selection and decision-making.

      Toxicity Characteristics (Endpoints)

      Although somewhat arbitrary, the  toxicity characteristics chosen  for evaluation in this
methodology reflect several considerations.  Because it is anticipated that many of the chemicals
selected for input into the model might have only sparse toxicity profiles available, it is desirable that
many different toxicity endpoints could serve as a basis for ranking. In terms of format, this could
be accomplished either by having very few but very broad characteristics (e.g., cancer plus genotoxic
endpoints, and  noncancer/nongenotoxic  effects), or by  having relatively many  but  narrow
characteristics (e.g., carcinogenicity, genotoxicity, immunotoxicity, neurotoxicity, teratogenicity,
hepatotoxicity, cardiovascular toxicity, etc.).  The selected approach lies in between these two
Working Draft                               C-5                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
extremes, with enough toxicity characteristics to facilitate at least some degree of effect segregation
and weighting. These toxicity characteristics and the current maximum toxicity scores that may be
attributed to them are as follows:

      •   Mammalian Carcinogenicity (whether "genetic" or "epigenetic"):  150 points
      •   Genotoxicity: 150 points
      •   Reproductive and Developmental Toxicity:  150 points
      •   Mammalian Chronic/Subchronic Toxicity (other than the above):  150 points
      •   Mammalian Acute Lethality: 150 points
      •   Mammalian Nonlethal Acute Toxicity: 150 points
      •   Mammalian Skin/Eye Irritation or Sensitization: 10 points
      •   RQ Category: 100 points

      Note that this list of scorable characteristics could readily be expanded, as has been done in
some systems, to include various ecotoxicity and/or chemical-physical parameters.  The emphasis
of this effort was to focus on those parameters more commonly considered as direct or indirect
indicators of human toxicity. This arrangement of toxicity characteristics recognizes endpoints that
are particularly important lexicologically and/or socially, that may be associated with data for
sparsely tested compounds or that provide readily available data that has already undergone an EPA
review and ranking process (i.e., RQ values).

      Scoring Criteria and Categories

      This section and its accompanying tables detail the scoring criteria and categories that are
currently suggested for use with each of the toxicity characteristics.  As, discussed further below, a
revised weighting scheme has resulted in two of these characteristics being assigned lower maximum
point values than the rest.  Virtually all the scoring ranges have been set higher than those typically
found in most other schemes. This was done in order to facilitate current and future weighting and
severity-of-effect  adjustments, while  also rmnimizing the occurrence of fractional scores. One
additional general point needs to be made in response to an Agency comment that "[t]here does not
appear to  be any provision  for  including; human epidemiological data  for cancer  or other
nonreproductive-developmental   endpoints"  (USEPA,  1996a).    While   the  use of human
epidemiological data  is explicitly expressed only  in the scoring schemes for genotoxicity and
reproductive-developmental toxicity (Tables 6 and 7), it must be emphasized that consideration of
such data is contemplated for all toxicity characteristics except that of Mammalian Acute Lethality:
LD5(/LC50 (human LD5(/LC50 is not typically available for obvious reasons). However, even in this
latter case, acute  human  lethality data is intended to be accommodated by the maximum SOE
category under the Mammalian Non-LD50/LC50 Acute Toxicity characteristic.  [Note that the latter
two  toxicity  characteristic's names have been slightly revised to better focus on the distinction
between LD50/LC50 and other types of acute mammalian lethality data.] With specific respect to
mammalian carcinogenicity, it is assumed  thcit available and reliable epidemiological data have been
incorporated into the WOE, Structure Activity Team Rank and Chemical Category Human Toxicity
Estimate categorization processes. Where the evaluator knows of such data that have  not been
considered, the methodology allows the exercise of professional judgment in  arriving at an
appropriate carcinogenicity score.
Working Draft                              C-6                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      Mammalian Acute Lethality: LD50/LC50

      As revised, the Mammalian Acute Lethality characteristic essentially has been adopted from
the RQ scheme (USEPA,  1985), with the six scoring categories for LD50s/LC50s being assigned
values ranging from 0 to a maximum of 150 points (Table 3-1).  This tripling of the "weight"
accorded to Mammalian Acute Lethality (the former maximum value was only 50 points) was made
to address concerns that the original scheme reduced the relative importance of acute toxicity on the
basis of what were, in effect, really exposure-based considerations. Nonetheless, the modest relative
scoring bias toward more potent chemicals, as well as the potency levels at which zero scores are
obtained, continue to reflect the operational assumption that chemicals less potent than these cut-off
levels will not be found in water at concentrations sufficient to merit concern.

      Mammalian Non-LD50/LC50 Acute Toxicity

      Following similar reasoning, the same six potency categories are utilized in the revised
evaluation of Mammalian Non-LD50/LC50  Acute Toxicity, except that scores range from 0 to a
maximum of only 15 (previously 10) (Table 3-2). To achieve a final score for this characteristic,
however, potency is multiplied by a severity-of-effect score that can range from 0 to a maximum of
10 (a doubling of the previous value of 5, done in order to bring it into accord with the adjustment
to Mammalian Acute Lethality described above). Therefore, a chemical's score for this characteristic
may be zero, or may range from 3 to a maximum of 150 (for life-threatening effects induced at very
low doses).  These  severity-of-effect categories are adapted from O'Bryan  and Ross (1988), are
difficult to define with any degree of precision and completeness, and will require the exercise of
some degree of professional judgment in their application to toxicity data from specific studies.
They are also presented in Table 3-2, along with some general guidance partially adapted from a
scheme derived from DeRosa et al. (1984). It should be noted that reproductive, developmental and
skin/eye irritation or sensitization effects are scored under their own toxicity characteristics, as
described below.

      Mammalian Skin/Eye Irritation or Sensitization

      The Mammalian Skin/Eye Irritation or Sensitization scoring scheme is summarized in Table
3-3. As indicated, this toxicity characteristic is evaluated primarily on a severity-of-effect basis
rather than potency.  In practice, the methodology also will be capable of segregating actual primary
dermal irritation index, Buhler test, ocular irritation and "percent responding" input data into these
scoring categories. Although it is likely that some chemicals with sparse toxicity profiles will have
this type of data available, the low scoring values (0 to  10 points) assigned to this characteristic
reflect the relatively minimal relevance these endpoints are  expected to have for real-world human
exposure to water contaminants.
Working Draft                              C-7                              November 1996
-------
                      Conceptual Approach for Contaminant Identification
    TABLE 1: MAMMALIAN ACUTE LETHALITY: LD5Q/LC50 CHARACTERISTIC
                 Category   Oral LD50,
                  Score       mg/kg
                  150

                  120

                   90

                   45

                   15

                    0
O.lto500
             Dermal LD50,    Inhalation
               mg/kg       LC50, ppm
             <0.04
0.4 to <4
10 to <100    4 to <40
100 to 500    40 to 200
>200
               <0.4
0.04 to <0.4     0.4 to <4
4 to <40

40 to <400

400 to 2,000

>2,000
Working Draft
             C-8
                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
    TABLE 2:  MAMMALIAN NON-LD50/LC50 ACUTE TOXICITY CHARACTERISTIC
                  Potency (P)        Oral         Dermal      Inhalation
                   Category     Dose, mg/kg  Dose, mg/kg   Dose, ppm
      SOE
 Category Score

       10

        8
Score
15
12
9
4.5
1.5
0

<0.1
O.lto500

<0.04
0.04 to <0.4
0.4 to <4
4 to <40
40 to 200
>200
-
<0.4
0.4 to <4
4 to <40
40 to <400
400 to 2,000
>2,000
Severitv-of-Effect (SOE) Description
       0
Life-threatening (death or pronounced life-shortening)

Severe (pronounced necrosis, atrophy, hypertrophy, metaplasia, etc.
with substantial organ dysfunction [e.g., significant increases in serum
or urine indicators of hepatotoxicity or nephrotoxicity]; neuropathy
with gross changes in or loss of behavioral or motor control or sensory
ability)

Moderate (necrosis, atrophy, hypertrophy or metaplasia with little or
no apparent decrement in organ function [e.g., minimal or no increase
in serum or urine indicators of hepatotoxicity or nephrotoxicity];
neuropathy with little or no measurable change in behavioral, sensor)'
or physiologic activity)

Intermediate mild-to-moderate  effects (hyperplasia, hypertrophy or
atrophy with/without organ weight changes; reversible cellular
changes: cloudy swelling, hydropic or fatty changes, etc.)

Mild (enzyme level, other biochemical, and/or quantitative or
qualitative subcellular organelle changes)

No observed effect
 Total Score =     P Score x SOE Score
Working Draft
                        C-9
November 1996
-------
                       Conceptual Approach for Contaminant Identification
      Mammalian Chronic/Subchronic Toxicity

      In order to evaluate Mammalian Chronic/Subchronic Toxicity effects that are not addressed
by the other, more specific toxicity characteristics, this proposed methodology has adopted the
approach of DeRosa et al. (1984), which also has been incorporated into the RQ Scheme (see Option
Two in USEPA, 1994).  The score utilized is referred to as the "composite score" (CS), which is
calculated as the product of two other independently determined factors.

      The first of these, called a "dose rating value" (RV(j), is essentially a potency factor that in turn
is derived according to a truncated logarithmic relationship (see Step 5, below) from the human-
equivalent "minimum effective  dose" (MED). This MED value, expressed in mg/day, represents the
lowest dose that the available  data suggest is capable of causing a particular adverse response in
humans, and may be approximated from coiresponding animal data according to the conversion
formula given below in Step 4.  The second factor utilized in calculating ihe CS is termed the "effect
rating value" (RVe), and it represents a numerical evaluation of the severity of the effect that is
associated with the MED.

      Originally, this ranking methodology directly adopted the point valuations of DeRosa et al.
(1984) for both the RVd and RVe factors, each having a range of values from 0 to 10, and thereby
permitting a CS range of 0 to 100 based  on both potency and severity of effect.  However, in
response to Agency comments (USEPA, 1996a, 1996b), the RVe scale has been recalibrated to a
maximum of 15 points.  This results in a higher maximum score obtainable  for this toxicity
characteristic (maximum CS =  150), one that better reflects its perceived relevance to the assumed
typical exposure of humans to water contaminants. A nine-step general description of the scoring
procedure follows,  with guidance for determining  RVe  values presented in Table 4.  The
methodology user will:

      1.    Identify chronic toxicity indicator levels (No Observed  Adverse Effect  Levels
           (NOAELs), Lowest Observed Adverse Effect Levels (LOAELs) or Frank Effect Levels
           (FELs)); note dose/concentration, exposure route and duration, and any toxic effects.

      2.    Convert all NOAELs/LOAELs/lFELs to mg/kg-day (from mg/day, ppm in water, diet
           or air, etc.) using USEPA methodologies (e.g., USEPA, 1980, 1988, 1989a, 1989b).

      3.    If based on subchronic exposure, divide values by 10.

      4.    Convert to MEDs (mg/day) based on a 70 kg human, using the following fourth-root-of-
           the-body-weight approximation for animal data, which is based upon the empirical
           observation that physiological processes among species tend to exhibit proportionality
           according to body weight raised to the 3/4 power (USEPA, 1992, 1996c)—this is a
           revision of the original 2/3 power relationship used by DeRosa et al. (1984):

                 (animal dose, mg/kg/day) x (animal wt., kg/70 kg)1/4 x (70 kg)
Working Draft                              C-10                             November 1996
-------
                        Conceptual Approach for Contaminant Identification
                   TABLES:  MAMMALIAN SKIN/EYE IRRITATION
                        OR SENSITIZATION CHARACTERISTIC
              Category
               Score

                10
             Effect Description
Corrosive to skin/eye; severe eye irritation, extreme or
severe skin sensitization

Severe skin irritation, moderate-substantial eye
irritation, moderate-strong skin sensitization

Mild-moderate skin irritation, mild eye irritation, mild
skin sensitization
                 0
Minimal to no observed effects
Working Draft
                 C-ll
November 1996
-------
                        Conceptual Approach for Contaminant Identification
               TABLE 4: RATING VALUES FOR TOXIC EFFECTS (RVe)5

         Rating  	Effect	

          0      No observed effect

          1      Enzyme induction or other biochemical change with no pathologic
                 changes and no change in organ weights

          3      Enzyme induction or other biochemical change and subcellular
                 proliferation or other changes in organelles but no other apparent
                 effects

          5      Hyperplasia, hypertrophy or atrophy, but no change in organ weights

          6      Hyperplasia, hypertrophy or atrophy with changes in organ weights

          7      Reversible cellular changes: cloudy swelling, hydropic change or
                 fatty changes

          9      Necrosis, or metaplasia with no apparent decrement in organ
                 function (e.g., no increases in serum or urine indicators of
                 hepatotoxicity or nephrotoxicity). Any neuropathy without apparent
                 behavioral, sensory or physiologic changes

         11      Necrosis, atrophy, hyperlrophy or metaplasia with a detectable
                 decrement of organ function (e.g., minimal-to-small increases in
                 serum or urine indicators of hepatotoxicity or nephrotoxicity). Any
                 neuropathy with a measurable change in behavioral, sensory or
                 physiologic activity

         12      Necrosis, atrophy, hypertrophy or metaplasia with definitive organ
                 dysfunction (e.g., significant increases in serum or urine indicators
                 of hepatotoxicity or nephirotoxicity).  Any neuropathy with gross
                 changes in behavior, sensory or motor performance.

         14      Pronounced pathologic changes with severe organ dysfunction (e.g.,
                 very substantial increases in serum or urine indicators of
                 hepatotoxicity or nephrotoxicity). Any neuropathy with loss of
                 behavioral or motor control  or loss of sensory ability.

         15      Death or pronounced life-shortening.
        Source:  Ross, 1994 (equivalent to DeRosa et al., 1984), with minor adaptations.
Working Draft                               C-12                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      5.    Assign a potency or "dose rating value" (RVd) to the MED according to the following
           truncated logarithmic relationship (see DeRosa et al. [1984] for background on the
           theoretical and empirical bases for this scheme):

           RVd=10(iflogMED<-3)
           RVd = -1.5 log MED + 5.5 (if -3 < log MED < 3)
           RVd=l (if log MED > 3)

      6.    Assign an "effect rating value" (RVe ) to the effect associated with the MED based on
           the 0 to 15 scale illustrated in Table 4.

      7.    Calculate the CS as follows: CS = RVd x RVe

      8.    If there are multiple MEDs for a given route of exposure: (1) choose adequate chronic
           MEDs over subchronic MEDs, (2) of the remaining MEDs, choose the one based on the
           best data, and (3) if multiple MEDs remain, choose the one that results in the highest
           CS.

      9.    If CSs have been derived for multiple routes of exposure, evaluate the chemical for
           chronic/subchronic toxicity according to the highest CS, regardless of the route upon
           which it  is  based.   This  represents an adjustment to the original  Mammalian
           Chronic/Subchronic Toxicity approach (USEPA,  1995d), which advocated using an
           inhalation-based CS  only in the absence of any based upon oral or dermal data. The
           revision reflects Agency commentary  (USEPA, 1996a, 1996b) that emphasized the
           importance of the inhalation route for human exposure to many chemicals that are, or
           could be, found in drinking water.

      The reviewed literature  does not clearly indicate  a recommended scoring approach for
situations in v/hich the only available dose value is a NOAEL - by definition, no  scorable effect
occurs at this close. This would seem equivalent to an RV£ of 0 (as currently indicated in Table 4),
thus resulting in a  CS also of 0, regardless  of the NOAEL-based RVd.  As another default
alternative, the following approach could be adopted during data entry: double the NOAEL value
and use this adjusted dose to calculate an RVd in the normal way, then multiply by an assumed RVg
of 1.  In other words,  the relatively conservative assumption is made that doubling the reported
NOAEL value would cause detectable, but minimal, toxicity. The resulting CS range would be from
1 to 10, would be of low data quality and could be modified by SAR/QS AR considerations. While
probably not of overriding importance to the final  ranking  in most instances,  it would help
distinguish between those chemicals having high or low chronic/subchronic NOAELS in the absence
of other quantitative chronic/subchronic data.

      Mammalian Carcinogenicity

      Carcinogenicity is another major toxicity characteristic, the scoring of which is patterned after
the RQ scheme, as adapted by EPA in its Chemical Use Clusters approach (USEPA, 1993). It was
originally developed by the Agency's Carcinogen  Assessment Group (CAG)  and considers the
Working Draft                               C-13                             November 1996
-------
                        Conceptual Approach for Contaminant Identification
weight-of-evidence (WOE) that the substance is a human carcinogen, as well as the substance's
carcinogenic potency (USEPA, 1987).

      Chemicals that are categorized by USEPA's WOE methodology as belonging to either Group
A (known human carcinogens), Group B (probable human carcinogens, either Bl or B2) or Group
C (possible human carcinogens) are subsequently categorized as having "high", "medium" or "low"
potency.  This is done by estimating the dose associated with a life-time excess cancer risk of 10%
(EDjQ, which is a less uncertain value than  the cancer slope factor,  "qj*"), then calculating a
potency factor "F" equal to 1/EDjQ .  From this is derived a human "RQ Potency Factor", which can
be used to determine the carcinogenicity score.  If no RQ Potency Factor has been derived, a qj*
slope factor may be used.  If neither is available, agency SAR/QSAR estimates can provide a low
quality estimate of potency.  Examples of this type  of derivative potency data, along with other
features of the method, are summarized in Table 5.

      Group D "not classifiable for human carcinogenicity" chemicals can be ranked by agency
SAR/QSAR approaches, while Group E "human noncarcinogens" are given a score of 0 for this
characteristic.  Professional judgment must be exercised (perhaps strengthened by  independent
SAR/QSAR analysis) to in effect estimate both WOE and P for chemicals having in vivo/in vitro
carcinogenicity data, but which have not yet been evaluated and categorized in some manner by
EPA. In such situations, the methodology allows the data enterer to select a "response strength" level
indicator  (e.g., high, medium, low or negative) upon which the carcinogenicity characteristic score
will be based. The methodology ranking output format provides an indicator symbol (see Data
Sufficiency Indicators Section) that such derived scores were based on nonagency-reviewed,
professional- judgement-based assessment of the available animal and short-term in-vitro data.

      It is recognized that new, more flexible guidelines for carcinogen risk assessment have been
proposed (USEPA, 1996c), and that they are likely to be substantially adopted. However, until they
are officially adopted in final form and until chemical evaluations performed under them are actually
available, significant modifications to this toxicity ranking methodology's approach, descriptive text
and database application are considered to be premature. Nevertheless, it is anticipated that the new
approach can be functionally integrated into the current, more  rigid WOE-based procedure that
serves as the basis for this toxicity ranking methodology's carcinogenicity-ranking process. For
example, one possible scheme for establishing approximate equivalency between old and new "WOE
categories" could be the following:

           Current WOE            Proposed WOE Descriptor/Subdescriptor
           Category

           A or B                 Known/Likely (all)

           C                      Cannot Be Determined/Suggestive Evidence

           D or Unclassified       Cannot Be Determined/Conflicting,
                                  Inadequate or No Data

           E                      Not Likely/Two Negative Animal Studies, etc.
Working Draft                              C-14                             November 1996
-------
    >n
^x  S
 03  >,  S
*O  ed "3
"ob5 13
        Me

        f
        £

^S |

O 
            a  a  ±
           5  5  a
            ob ob .3
               o  6
           O  CN)  1>  00

           d  d S  S
                                     M a>


                                     W) BO
                                    o  o
                                    7  °P
                                     A  A
                                                              s
                                                              •^H

                                                              "8
                                                              O
                                                              •a

                                                   00
                                                  s
W


•4-^
S
Q.
S
1
Q
t:
o
Oi
1
U


^
g
S
£
U
U
C3
U
*
I?


S
<


RQ Potency Factors




G
S
Structure Activity Tee
Rank




1
3
Chemical Category H
Toxicity Estimate






q^* (Cancer Potency)



U


£
3
%
PH
>>
O
1
Of
&




a
S
Structure Activity Tee
Rank




c
cd
3
Chemical Category H
Toxicity Estimate




g
S
Structure Activity Tee
5 Rank
*w
S
'«3
C/3
ca
°1
Q P
1

Chemical Category H
Toxicity Estimate



T3


0
II
§
o
C/3



w
-------
                        Conceptual Approach for Contaminant Identification
      Genotoxicity

      The approach for scoring genotoxicity is adapted from the OTS scheme reported by O'Bryan
and Ross (1988), is largely based on WOE considerations and is presented in Table 6.  This revision
of the scheme has increased the number of defined scoring categories in an effort to reduce the
amount and variability impact of any needed professional judgment. In effect, the scheme attempts
to largely delineate scoring decisions based upon WOE considerations, while still permitting the
exercise of some professional judgment in order to increase the WOE-based category score when
data is available that indicates high-to-very-high chemical potency. The revised scheme's nonzero
scores range from  15 to 150,  with the higher scores reflecting the degree of seriousness accorded
to evidence suggestive of potential human gerni cell genotoxicity.  Mid- and low-range scores reflect
the quantitative or qualitative types of data that are somewhat less suggestive of hazard to the human
gene  pool and,  at the  same time, somewhat more  typical of  that often  used to  support the
determination of a chemical's carcinogenic potential.

      Reproductive and Developmental Toxicity

      The final toxicity characteristic utilized for scoring in this methodology is that of Reproductive
and Developmental Toxicity. The current revised approach arrives at a final characteristic score by
multiplying a potency score times what is primarily a WOE score. In deference to Agency comments
on the original scoring scheme (USEPA, 1996a, 1996b)  and guidance for developmental toxicity risk
assessment  (USEPA, 1991), SOE considerations have largely  been eliminated.  The potency
categories and revised scoring values (0 to  15) are the same as those presented in Table 2 for
Mammalian Nonlethal Acute Toxicity, with the addition that in the absence of quantitative data,
qualitative evaluations of potency can also serve as the basis of scoring (e.g., high = 15, medium-
high = 12, medium = 9, medium-low = 4.5, low = 1.5, no effect or negative = 0).  The WOE
categories and scores (0 to 10) are summarized in Table 7.  The latter were partially adapted from
the Great Lakes Basin Sunsetting scheme reported by Foran and Glenn (1993) and, following this
revision, are fewer in number, simpler and virtually free of SOE parameters.  As with the other
principal toxicity characteristics, final scores for this characteristic can range up to a maximum of
150 points.

      RO Category

      While not strictly speaking a toxicity characteristic, the RQ  category of a chemical may be a
readily available data element that could be utilized in this ranking process. Therefore, to increase
the methodology's versatility, the five RQ categories provisionally have been assigned the following
revised score values:

         1 Ib   =100 points
       10 Ib   =  80 points
      100 Ib   =  40 points
     l,0001b   =  20 points
     5,000 Ib   =  5 points
Working Draft                              C-16                             November 1996
-------
                         Conceptual Approach for Contaminant Identification
                     TABLE 6: GENOTOXICITY CHARACTERISTIC6
  Category
   Score       	Criteria7
   150      Convincing/strong epidemiological data suggesting genotoxicity in human
             populations, OR suggestive epidemiological data combined with evidence of
             chemical interaction with mammalian germ cell DNA or with evidence of
             rnutagenicity/clastogenicity in a single mammalian germ cell assay, OR
             evidence of rnutagenicity/clastogenicity from multiple or replicated mammalian
             g;erm cell assays
   140      Suggestive epidemiological evidence of genotoxicity in humans, OR chemical
             interaction with mammalian germ cell DNA, OR evidence of
             mutagenicity/clastogenicity in a single, unreplicated mammalian germ cell assay
   130      Evidence of genotoxicity in multiple/replicated nonmammalian germ cell
             assay(s) or mammalian dominant lethality assays
   120      Evidence of genotoxicity in a single, unreplicated nonmammalian germ cell
             assay or mammalian dominant lethality assay
   110      Evidence of genotoxicity from replicated assays of three test systems (in vivo/in
             vitro) other than those listed above, OR evidence from unreplicated assays of
             four or more such test systems
   100      Evidence of genotoxicity from replicated assays of two, OR from unreplicated
             assays of three, test systems (in vivo/ in vitro) other than those listed above
    80      Evidence of genotoxicity from unreplicated assays of two test systems (in
             vivo/in vitro) other than those listed above
    65      limited evidence of genotoxicity: replicated positive assays of a single test
             system (in vivo in vitro) other than those listed above, OR unreplicated positive
             results in two such test systems in conjunction with negative results in one or
             two others
    50      Limited evidence of genotoxicity: unreplicated positive assay of a single test
             system (in vivo in vitro) other than those listed above
    30      Limited evidence of genotoxicity: equivocal results in one or more test systems
             (in vivo/in vitro) other than those listed above, OR mixed positive and negative
             test results in a single such test system
    15      Limited evidence of nongenotoxicity:  negative results in one or two test
             systems
     0      Only negative results in three or more test systems
       6 Adapted from O'Bryan and Ross (1988).

        Genotoxicity: mutagenicity (in vivo and in vitro), including dominant lethality; clastogenicity (in vivo and in
vitro); DNA interactions (binding/damage/repair). In the absence of data, scores can be assigned on the basis of chemical
analog data or structure-activity relationships (SAR). If response magnitudes are known to be high to very high, the
category score may be increased by one or two levels.
Working Draft                               C-17                               November 1996
-------
                       Conceptual Approach for Contaminant Identification
           TABLE 7:  REPRODUCTIVE AND DEVELOPMENTAL TOXICITY:
                          WOE CATEGORIES AND SCORES8
    Category
     Score                      	 Criteria
       10      Sufficient epidemiological evidence in humans, OR compelling but limited
               epidemiological evidence plus observed effects in one: mammalian test
               species, OR observed effects in two or more mammalian test species

        8      Compelling but limited epidemiological evidence, OR observed effects in
               one mammalian test species

        6      Positive results in multiple: short-term in vivo/in vitro tests

        4      Equivocal epidemiological or mammalian test species data, OR positive
               results in one, or equivocal/mixed results in more than one, short-term in
               vivo/in vitro test

        2      Suggestive negative epidemiological evidence, OR no observed effects in
               one mammalian test species, OR negative results in multiple short-term in-
               vivo/in vitro tests

        0      Sufficient negative epidemiological evidence, OR compelling but limited
               negative epidemiological evidence plus no  observed effects in one
               mammalian test species, OR no observed effects in two or more
               mammalian test species
        Adapted from Foran and Glenn (1993).
Working Draft                              C-18                             November 1996
-------
                        Conceptual Approach for Contaminant Identification
      These revised scores are based on the maximum chronic toxicity points achievable by a
chemical in each of the RQ categories according to the RQ chronic toxicity "RV^ x RVe" scheme
(DeRosa et al., 1984;  USEPA,  1985),  which essentially has been  adopted in this  ranking
methodology (see above).  For ranking purposes, RQ category points are initially treated and
designated as derived rather than real data (see Data Sufficiency Indicators Section), since generally
it will not be known whether the RQ category was based on a chemical-physical parameter or a
toxicity parameter  (and even if the later, which one will not be known).   Furthermore, in
consideration of these uncertainty concerns as expressed in the original methodology documentation
(USEPA, 199:5c, 1995d)and subsequent Agency review commentary (USEPA, 1996a, 1996b), the
revised relative weight of the RQ characteristic has been reduced by no longer scaling its maximum
value up to the 150 points of the principal toxicity characteristics.

      Data Gaps and SAR/OSAR

      As has been noted previously, SAR/QSAR approaches may be utilized to supply scores for
those toxicity characteristics lacking available experimental and epidemiological data. The particular
approaches and the characteristics for which they can be  used will depend upon the  in-house
capabilities of EPA. Fitting S AR/QSAR output within the scoring schemes for each characteristic
will rely upon professional judgment.  If no "real" or SAR/QS AR data can be generated for a given
chemical, it will be designated currently unrankable.

      Evaluation Output

      This section briefly describes the types of evaluation output that can be derived after chemicals
are put through the toxicity characteristic scoring process.

      Ranking Score: Weighting. Aggregation and Normalization

      The primary output number to be generated by the methodology is the ranking score  based on
available real or derived toxicity information. As already indicated above, this score will  take into
account the relative importance of the toxic effects evaluated.  This endpoint weighting is  achieved
through assigning different maximum obtainable scores to those toxicity characteristics that are
judged to be of higher or lower significance or reliability than the "typical"  toxicity characteristic.
In light of  the; methodology's design goals of identifying drinking  water contaminants  that may
require regulation or the generation/acquisition of additional important toxicity data,  different
maximum scores are currently assigned to only two characteristics. RQ category receives somewhat
less weight primarily because it is a derivative value that, by itself, leaves significant uncertainty
regarding its basis.  Mammalian Skin/Eye Irritation and Sensitization receives substantially lower
weight because; such data, while frequently available, are deemed to be of significantly less relevance
to ordinary drinking water hazard scenarios.  As revised, the remaining six characteristics all have
the same maximum score potential (i.e., 150 points), which reflects a judgment of their approximate
equivalency in terms of maximum endpoint SOE and/or perceived societal concern.

      Because the proposed methodology does not force the evaluator to determine a score  for every
toxicity characteristic, the individual scores for each evaluated characteristic cannot be aggregated
to derive the final ranking without biasing the results towards chemicals with more input data and
Working Draft                               C-19                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
jeopardizing the characteristic weighting scheme. For example, chemical A with high scores in acute
lethal, acute nonlethal and skin/eye irritation effects, plus a moderate score in chronic/subchronic
toxicity, would outrank chemical B having the highest possible score for carcinogenicity and no
other data - a result probably misleading with respect to the actual levels of regulatory concern that
would be anticipated to attach to each of the chemicals,

      Furthermore, score normalization procedures also appear to permit undesirable results. For
example, normalizing characteristic scores by the maximum obtainable scores for the evaluated
characteristics could result in a 150 point maximum-value carcinogen (no other data) being ranked
the same as a 10 point maximum-value eye irritant (no other data), each having a normalized score
of 1.0. Normalizing by the number of characteristics evaluated also presents potential difficulties,
as a maximum carcinogen and eye irritant would score well below just a maximum carcinogen ([150
+ 10]/2 = 80 versus 150/1  = 150).  In effect, score normalization is not  compatible with toxicity
endpoint weighting.

      Therefore, it is recommended that score aggregation and normalization procedures not be
incorporated into the ranking calculation.  The original toxicity ranking methodology (USEPA,
1995a, 1995d) did propose one additional weighting procedure for consideration - that of exposure
pathway weighting.  Because the human exposure of concern is that of chemical contaminants in
water, it was proposed that toxicity characteristic scores derived from inhalation exposure (versus
oral or dermal) be multiplied by a factor of 0.75 to reflect a presumed lower relevance of such data
for the majority of drinking water contaminants.  Because the Agency called this operational
assumption into question (USEPA, 1996a, 1996b), especially for volatile compounds, this revision
of the methodology has eliminated the inhalation route adjustment factor of 0.75.  However, the
ability to incorporate route-specific adjustment  factors  into the ranking  process via simple
programming changes remains a feature of the methodology's application software.

      For equivalent rankings, those based upon real data will be placed higher than those based
upon derived data (see Data Sufficiency Indicators Section). If equivalent rankings still exist, the
revised methodology database application will now order them  according to the scores obtained in
the next highest scored characteristic (e.g., a 150-point carcinogen with a 50 point genotoxicity score
would rank above a 150-point carcinogen with a 10 point  eye irritation score). Any  substances
having tied rankings based on the two highest  scored characteristics will  simply  be ranked
alphabetically.

      Professional Judgment and Data Sufficiency Indicators

      Previous sections have suggested that this toxicity ranking methodology can minimize and
simplify, but not eliminate, the exercise of professional judgment. When substantial, "non-routine"
use of professional judgment is  required during a particular ranking exercise, it is anticipated that
explanatory text will be included in an accompanying summary report.  Indications of more routine
applications of professional judgment can be gleaned from examining the Toxicity Ranking Input
data forms (Table 8), i.e., from whether an identified study was selected for possible use as  the basis
for the characteristic's score, from the indicated evaluation of the study's quality (for mammalian
chronic/subchronic toxicity studies; "high", "medium" and "low" can be more formally defined in
the report for each ranking exercise), and from the descriptions and comments found in the "Effect"
Working Draft                              C-20                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
column for several of the characteristics. These data "input" forms are included as "output" in the
ranking exercise's final summary report.   Additionally, ranking scores based principally on
professional judgment or derived data are identified in the Ranking Summary of Chemicals output
form (Table 9) with data qualifiers ("d" for USEPA-supplied SAR/QSAR data, interpretations, etc.,
and "dp" for score quantitation based upon professional judgment supplied by the ranking exercise's
evaluating toxicologist or other non-agency source). Note that when a ranking is based upon the RQ
Category score, an "r" qualifier is displayed in this output form.

      The Toxicity Ranking Summary of Chemicals output form also includes two "data sufficiency
indices" (DSI1 and DSI2) in order to provide a readily accessible indication of how much unqualified
and total data were available and utilized to evaluate each chemical for ranking. DSI1 represents the
weighted fraction of the eight toxicity characteristics for which "real", unqualified data were located.
It is calculated by summing the maximum category scores for those characteristics having such data,
then dividing this total by the sum of the maximum scores obtainable for all eight toxicity
characteristics (i.e., 1010 points).  DSI1 values may range from 0.00 (i.e., all available data were
qualified) to a maximum of 0.90 (910/1010, i.e., unqualified data were available for all seven non-
RQ toxicity characteristics; RQ data is by definition qualified).  Similarly, DSI2 represents the
weighted fraciion of toxicity characteristics for  which any data  (qualified or unqualified) were
located for each chemical. Consequently, its value may range from 0.00 (i.e., no useable data of any
kind were located for any toxicity characteristic and the chemical is unrankable) to 1.00 (i.e., useable
data were located for eight of the toxicity characteristics).  In a general sense, the higher these
indicators, the more complete and reliable is the chemical's ranking score.

      To better present a picture of data sufficiency and toxicity profiles for the ranked chemicals,
an additional summary table (Toxicity Characteristics Summary of Chemicals, Table 10) is provided
in the ranking report. It is similar to the current Toxicity Ranking Summary of Chemicals output
form (Table 9). The table's entries include: Rank Number, Chemical Name, and the scores (with
any appended data qualifiers) and  the data sufficiency indices (DSI1 and DSI2).

METHODOLOGY LIMITATIONS AND UNCERTAINTIES

      Like all methodologies devised in part to simplify an inherently complex task, this Toxicity
Ranking Methodology is  limited by the simplifying restrictions that  are intended to promote
uniformity and general user-friendliness. Toxicity characteristic breadth and scoring weight are of
necessity somewhat arbitrary, and  like any such system, the quality of the ranking output is highly
dependent on the extent and quality of the input data.  Meaningful toxicology cannot be performed
without the exercise of at least a minimal amount of professional judgment, and to that extent the
methodology remains subject to a certain amount of variability, bias and even outright error. When
used as intended,  i.e., as  a screening methodology to establish a preliminary prioritization of
chemicals' overall toxicity and an identification of their apparent data gaps, the method is anticipated
to serve adequately.  Typically, though, this use will involve the use of limited time and information
resources, and will thus require that the ranking undergo a reasonable level of professional scrutiny
and adjustment prior to its final acceptance for regulatory or research-prioritization activities. The
more time and information that are made available at the data review and input selection stage, the
more  reliable will be the resultant toxicity profiles and rankings.  Through continued use, more
Working Draft                              C-21                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
specific limitations can be expected to become apparent that will require some degree of professional
accommodation and/or methodology revision.
Working Draft                               C-22                               November 1996
-------
s
$
fc
u
HH
X
o
H

do
PQ
    CO

    8
    S
    a
    m
          iu
£
          V)
 B
                           I
                  s
                           *:
                           CO
                               IN 
-------
      £
      H
      8
      on
      I
u
o
si
S

o
z
      g
      ON
      3
      3
      H
             s
             1
          o?
       in
       Cil
              «
              I
              o
              "5
              w
              |
              z
               8
             I


                   rsS
                     I
n

s
ci
O
8
                       o o i- i- o T-
                                  o o o o e> o o o
                       oooodoooooooooi
                       S588SSS5?

                       Q.
                       •a
                   111
iii
                                             >0> O T- fM f> •» 
                                             t- 
-------
s
5
&
P
o
t>o
^
u
PH
O

>
at
CO
CO
U
h— I
H
CO

s
w
H
U
ffi
U
>

I
 a
 PQ
|
            O
            •5

si
                       o
°
  U
                       V)
  i
                     =1
                     £i
                         OO^-T-Oi-OOOOOOOOO
                                                        in
                         '"TMf-jt^wi^U)^;'*';'!-*^
                         o  o o o  o o o  d o d o o o  o o
         is
         R JS   !?S
                                £ 2 o» o
           •» S> o
                                                        o o
                           Ul
                                                 8 N 8 J5 «5 N
-------
                       Conceptual Approach for Contaminant Identification
REFERENCES

ARET. 1993. Accelerated Reduction/Elimination of Toxics. ARET Criteria Sub-committee report.
Revision date September 27,1993.

CLC. 1992.  Canadian Labor Congress. CLC proposal for the October meeting of the ARETS
(Accelerated Reduction/Elimination of Toxics Scoring) Criteria Working Group. Department of
Field Services. Ottawa, Ontario: Canadian Labor Congress Environment Bureau.

Davis GA, Swanson M, Jones S. 1994.  Comparative evaluation of chemical ranking and scoring
methodologies.  EPA Order No. 3N-3545-NAEX.  University of Tennessee, Center for Clean
Products and Clean Technologies, Knoxville, TN.

DeRosa CT, Stara JF, Durkin PR.  1984. Ranking chemicals based upon chronic toxicity data.
Cincinnati, OH:  U.S. Environmental Protection Agency, Office of Research and Development.
Syracuse, NY: Syracuse Research Corporation.

Environ Corporation.  1986. Examination of the severity of toxic effects and recommendation of
a systematic approach to rank adverse effects.  Prepared for U.S. Environmental Protection Agency,
Office of Environmental Criteria and Assessment, Cincinnati, OH.

Foran JA, Glenn BS. 1993. Final report. Criteria to identify chemical candidates for sunsetting in
the Great Lakes Basin.  The George Washington University, Environmental Health and Policy
Program, Department of Health Care Sciences, Washington, DC.

Hushon JM, Kornreich MR. 1984. Scoring systems for hazard assessment. In: Hazard assessment
of chemicals: current developments. Vol.3. Academic Press, 63-109.

Klein W, Kordel W, Klein AW, et al. 1988. Systematic approach for environmental hazard ranking
of new chemicals. Chemosphere 17:1445-1462.

MDNR. 1987. Michigan Department of Natural Resources. Critical Materials Register (Criteria
and Support Documents).  Michigan Department of Natural Resources.

O'Bryan TR, Ross, RH.  1988. Chemical scoring system for hazard and exposure identification. J
Toxicol Environ Health 1:119-134.

OME. 1988. Ontario Ministry of the Environment. The effluent monitoring priority pollutants list
(1987). Hazardous Contaminants Coordination Branch.  Toronto, Ontario: Ontario Ministry of the
Environment.

OME. 1989. Ontario Ministry of the Environment. The effluent monitoring priority pollutants list
(1988 update). Hazardous Contaminants Coordination Branch. Toronto, Ontario: Ontario Ministry
of the Environment.
Working Draft                             C-26                            November 1996
-------
                       Conceptual Approach for Contaminant Identification
OME.  1992. Ontario Ministry of the Environment. Candidate substances list for bans or phase-
outs. Hazardous Contaminants Branch and Water Resources Branch. Toronto, Ontario: Ontario
Ministry of the Environment.

Ross RH. 1994.  Personal communication.  Unreleased draft report.  Workshop on weight of
evidence - hazard identification for noncancer health effects. August 26-28, 1987.  Sponsored by
the Science  Policy Integration Branch, Office of Policy,  Planning  and Evaluation.   U.S.
Environmental Protection Agency.  July 1988.

Sampaolo A, Binetti R.  1986.  Elaboration of a practical method for priority selections and risk
assessment among existing chemicals.  Regul Toxicol Pharmacol 6:129-154.

Sampaolo A, Binetti R. 1989. Improvement of a practical method for priority selections and risk
assessment among existing chemicals.  Regul Toxicol Pharmacol 10:183-195.

USEPA. 1980.  U.S. Environmental Protection Agency. Guidelines and methodology used in the
preparation of health effects assessment chapters of the consent decree water quality criteria.  Fed
Regist 45:79318-79379.

USEPA.  1985.  U.S. Environmental Protection Agency.  Notification requirements; reportable
quantity adjustments; final rule and proposed rule. U.S. Environmental Protection Agency. Part n.
Fed Regist 50:13455-13522.

USEPA. 1987.  U.S. Environmental Protection Agency. Hazardous substances; reportable quantity
adjustments; proposed rules.  U.S. Environmental Protection Agency.  Part n. Fed Regist 52:8139-
8171.

USEPA. 1988.  U.S. Environmental Protection Agency. Recommendations for and documentation
of biological values for use in risk assessment.  EPA/600/6-87/008. Cincinnati, OH: U.S. EPA
Office  of Research  and Development, Office of Health  and  Environmental  Assessment,
Environmental Criteria and Assessment Office.

USEPA. 1989a. U.S. Environmental Protection Agency. Guidelines for authors of EPA Office of
Water Health Advisories for drinking  water contaminants.  Final.  Washington, DC: U.S. EPA
Office of Water, Office of Drinking Water.

USEPA. 1989b. U.S. Environmental Protection Agency. Exposure factors handbook.  EPA/600/8-
89/043. Washington, DC: U.S. EPA Office of Research and Development, Office of Health and
Environmental Assessment, Exposure Assessment Group.

USEPA. 1992  U.S. Environmental Protection Agency. Guidelines for developmental toxicity risk
assessment.  Fed Regist 56(234):63797-63826.

USEPA. 1992. U.S. Environmental Protection Agency. Draft report:  A cross-species scaling factor
for carcinogen risk assessment based on equivalence of mg/kg3/4/day. Fed Regist 57(109):24151-
24173.
Working Draft                              C-27                             November 1996
-------
                       Conceptual Approach for Contaminant Identification
USEPA. 1993. U.S. Environmental Protection Agency. Draft report. Chemical use clusters scoring
methodology. Office of Pollution Prevention and Toxics.  Washington, DC: U.S. Environmental
Pollution Agency.

USEPA.  1994. Toxicity Ranking Methodology Options. TR-1690-5-2.  Cleveland, OH: Life
Systems, Inc.

USEPA.  1995a.  Toxicity Ranking Methodology. Draft Report.  TR-1690-5-3. Cleveland, OH:
Life Systems, Inc.

USEPA. 1995b. Toxicity Ranking Methodology Test-Run  Results. TR-1690-5-5.  Cleveland, OH:
Life Systems, Inc.

USEPA. 1995c. Toxicity Ranking Methodology Validation Report. TR-1690-5-6.  Cleveland, OH:
Life Systems, Inc.

USEPA. 1995d. Toxicity Ranking Methodology. Final Report. TR-1690-5-3A. Cleveland, OH:
Life Systems, Inc.

USEPA. 1996a. U.S. Environmental Protection Agency. Letter: EPA Deliverables for Revisions
to Toxicity Ranking Methodology (technical; direction). From Bruce S. Mintz, Work Assignment
Manager, Health  and Ecological Criteria Division, Office of Science and Technology, Office of
Water, USEPA. Washington, DC: to Mr. Jeffrey J. Mosher, Project Manager, The Cadmus Group.
 Dated June 27, 1996.

USEPA.  1996b.  U.S. Environmental Protection Agency.  Memorandum: Review of Toxicity
Ranking System.  From  Gail Robarge, Director, Water  Staff, Office;  of Research and Science
Integration and Elaine Z. Francis, Director, Toxics/Pesticides Staff, Office of Research and Science
Integration, to Margaret J. Stasikowski, Director, Health and Ecological Criteria Division, Office of
Science and Technology,  Office of Water. Washington, DC:  USEPA.  Dated May 22, 1996.

USEPA. 1996c. U.S. Environmental Protection Agency. Proposed guidelines for carcinogen risk
assessment.   EPA/600/P-92/003C.   Washington,  DC:   U.S.  EPA  Office of Research and
Development.

Weiss M,  Kordel W, Kuhnen-Clausen D, et al.  1988.  Priority setting of existing chemicals.
Chemosphere 17:1419-1443.
Working Draft                              C-28                             November 1996
-------
                    Conceptual Approach for Contaminant Identification
                            Appendix D:

                      Exposure Ranking System
                        for the Contaminant
                        Identification Method
Working Draft
November 1996
-------
                          Conceptual Approach for Contaminant Identification
                               [This page intentionally left blank.]
Working Draft                                                                      November 1996
-------
                       Conceptual Approach for Contaminant Identification
                                   APPENDIX D:

                       EXPOSURE RANKING SYSTEM FOR
                 THE CONTAMINANT IDENTIFICATION METHOD
INTRODUCTION

     The U.S. Environmental Protection Agency's (EPA) Office of Water (OW) is in the process
of developing a Contaminant Identification Method to support regulation development under the
Safe Drinking Water Act (SDWA).  One part of the Contaminant Identification Method involves the
use of an exposure  ranking system for drinking water contaminants.   As a first step in the
development of an exposure ranking system, a review of currently available ranking systems and
potential scoring criteria is  warranted.  This document provides an initial summary of available
ranking systems and a brief description of potential scoring criteria or attributes. The remainder of
this section describes  the purpose of this study, relevant background information, and the technical
approach employed.  The other sections in this document include: Section 2 - Currently Available
Ranking Systems; Section 3 - Criteria for Scoring; and Section 4 - Next Steps.

     Purpose

     The purpose of this document is to summarize and present currently available information
regarding contaminant  ranking systems for drinking water exposure.   The   Contaminant
Identification Method requires the use  of an exposure/toxicity ranking  system  to support the
regulation development process. The toxicity ranking methodology is described in Appendix C.

     The information summarized in this document will serve as the basis for the development of
the exposure ranking system to support the Contaminant Identification Method. In addition, this
summary will be part of an information package to support an EPA stakeholder meeting on the
Contaminant Identification Method to be held in December 1996.

     Background

     This document summarizes the currently available information on the drinking water exposure
ranking component of the toxicology/exposure ranking method, including describing the technical
approach, presenting the findings, and providing recommended next steps.  This information will
be  the  basis for developing the exposure ranking  methodology, including the exposure
characteristics, scoring criteria, and weightings.

CURRENTLY AVAILABLE RANKING SYSTEMS

     The four sources of hazard ranking systems identified for summarization in this document
were:
           OST's Toxicity Ranking Methodology (1996)
           OGWDW's Contaminant Ranking Scheme (1992)
Working Draft                             D-l                             November 1996
-------
                        Conceptual Approach for Contaminant Identification
           OSW's Waste Minimization Prioritization (1996)
           Superfund Hazard Ranking System (1990)

      OST's Toxicitv Ranking Methodology (1996)

      This Toxicity Ranking Methodology is described in detail in Appendix C.

      OGWDW's Contaminant Ranking Scheme (1992)

      In  1992, OGWDW developed a Priority Pollutants List for E>rinking Water based on a
exposure/toxicity ranking system.  The objective of this ranking system was to identify and prioritize
contaminants that pose a potential threat to drinking water. The method developed was based on a
health risk-based approach for identifying and ranking drinking water pollutants (USEPA, 1992).

      The developers of this system conducted a literature search for existing ranking schemes.  Six
scoring systems that involved exposure ranking schemes were identified for review.  The citations
for these six ranking systems are:

           Brown,  S.L, et al., 1984 "Ranking Algorithm for the EEC Water Pollutants;" Final Report,
           Contract No. ENV/223/74-EN, Stanford Research Institute, Menlo Park, CA.

      •     Fiksel J and  M Segal,  1982 "An approach to Prioritization of Environmental Pollutants: The
           Action Alert System;" Final Draft Report, EPA  Contract 68-01-3857, Auther D. Little,
           Cambridge, MA.

      •     Brown, S.L.  et al., 1976 "Systems for Rapid Ranking of Environmental Pollutants - Selection
           of Subjects for Scientific and Technical Assessment Reports;" EPA Contract No. 68-01-2940,
           Stanford Research Institute, Menlo Park, CA.

      •     Ross, R.H. and J. Welch, 1980 "Proceedings of the EPA Workshop on the environmental
           Scoring of Chemicals (August 13-15,1979);" ORNL/EIS-158,EPA-560/11-80/010, Oak Ridge
           National Laboratory, Oak Ridge, TN.

           USEPA, 40 CFR Part 300, December 14, 1990, Hazard Ranking System: Final Rule. Volume
           55, No., 241.

           ATSDR, 1991 "Support Document for the CERCLA 104 Revised Priority List of Hazardous
           Substances that will be the Subject of Toxicological Profiles;" Agency for Toxic Substances and
           Disease  Registry, Prepared in cooperation with the U.S. environmental Protection Agency.

      The rationale for identifying chemicals was based  solely on  EPA's  Integration Risk
Information System (IRIS), Storage and Retrieval System (STORET), and Pesticide in Ground Water
Data Base (PGWDB). A list of chemicals was identified by screening IRIS for the existence of
toxicity criteria measured by quantitative health risk information. This resulting group of chemicals
was cross-referenced with STORET and PGWDB for  occurrence in ambient water measured by
frequency of detection (USEPA, 1992).
Working Draft                               D-2                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
      In this method, the developers ranked the selected chemicals according to various hazard
potential characteristics. The developers labeled the ranking score as the Risk Index.  This Risk
Index was based on the following four factors:

      •     Production quantity of the chemical (PQ)
      •     Exposure quantity of the chemical in water (EQ)
      •     Occurrence of the chemical in water (OW)
           H uman health risk (HR)

      Some of these factors included additional subfactors. The exposure quantity of the chemical
in water was based on the quantity of chemical that is released to water (in tons) and the fate of the
chemical in water based on a persistence factor. The persistence factor was derived from the EPA's
Hazard  Ranking System Method (i.e., the half-life of a chemical in water or, if lacking,  the
octanol/water partition coefficient) (USEPA, 1990). The occurrence of the chemical in water score
was derived from a frequency of detection score and a concentration score. The human health risk
factor was based on a carcinogenicity score and a non-carcinogenicity score (USEPA, 1992).

      Each factor and subf actor was scored from 1 to  10 on the basis of the available data.  The
scores were normalized, weighted, and aggregated to obtain the final Risk Index for each chemical.
The overall Risk Index was computed using the following equation:

      Risk Index = [(Wj*PQ) + (W2*EQ) + (W3*OW)] * (W4*HR)

           \\Tiere,     Wj = 1
                       W2 = l
                       W3=3
                       W4 = 5

      In this system, the data source for the production quantity of the chemical was the Hazardous
Substances Database. The  sources used to derive  the exposure quantity of the chemical scores
included the following: (1) the Toxic Release Inventory (TRI) system for release data; (2) the Permit
Compliance System (PCS) for release data; (3) the Superfund Chemical Data Matrix (SCDM) for
persistence data. For the occurrence of the chemical in water, the Agency for Toxic Substances and
Disease  Registry's (ATSDR) HAZDAT system was used. IRIS data was employed in the derivation
of scores for human health risk (USEPA,  1992).

      In this ranking system, each factor or subfactor was  scored on the basis of the following
criteria:

      •     The range of scores for each factor or subfactor was developed by examining the actual
           data to preserve the distribution of the data.
      •     All the ranges developed for scoring each factor and subfactor were set from  1 to 10.
      •     For each  factor, the score was normalized to account for multiple subf actors.
      •     Risk Index scores were calculated based on the equation presented above.
Working Draft                               D-3                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      The developers of this ranking system acknowledged that the quality and availability of the
underlying data is a major issue in the scoring scheme and that this scheme does not take into
account issues of uncertainty. The developers suggested that a uncertainty analysis of each database
could be conducted and the  uncertainties across the various databases; could be compared.  The
developers also acknowledged that lack of data was a problem.  When data were not available, the
developers used best professional judgement for assigning scores (USEPA, 1992).

      OSW's Waste Minimization Prioritization (1996)

      EPA developed  the  Waste Minimization National Plan9, with  extensive  stakeholder
involvement, to promote waste minimization. To develop this plan, stakeholders expressed a need
for prioritization tools to identify source reductions  and recycling priorities, including a flexible
screening tool (i.e., prioritization tool) and a list of constituents (i.e., chemicals) of concern (USEPA,
1996b).

      The Office of Solid Waste (OSW) established the Waste Minimization Prioritization Team,
consisting of representatives from EPA headquarters, EPA regions, and several states to select a risk-
based prioritization tool.  The objectives of this team were to: assess  the needs of different
stakeholder groups for risk-based prioritization tools; evaluate available tools against these needs;
and provide recommendations to EPA management for selection of a prioritization tool that could
be used to create a list of prioritized chemicals (USEPA, 1996b).

      The needs identified by the stakeholder groups included: a risk-based prioritization tool; a
multi-media viewpoint (i.e., not just focus, on waste but consider EPA air, water, and waste
programs); and flexibility in establishing goals and implementing waste minimization.  In addition,
the stakeholders decided that rather than developing a new prioritization  tool, they would use a
current tool, with only slight modifications, if one was available that would  meet all of their
requirements.  The core criteria for their prioritization tool, as characterized by the stakeholders,
included persistence, bioaccumulation potential, and tox icily as indicators of risk.  The stakeholders
noted that the commonly used measures of persistence are: medium-specific half-life; hydrolysis
half-life,  biodegradation half-life, volatilization half-life; and overall half-life. Common criterion
for bioaccumulation potential were noted as: fish bioconcentration factor; unspecified aquatic
bioconcentration factor; and octonal/water partition coefficient (K^,) (USEPA, 1996b).

      The Waste Minimization Prioritization Team identified a number of available prioritization
tools and evaluated these options against set criteria. A tool refers to any system that can be used
to rank, score, or set priorities based on the chemical constituents or properties of a contaminant or
waste. In addition to a comprehensive literature  search the team used! the following three major
sources of information to provide summaries of candidate screening tools:

      •     University of Tennessee, 1994 "Comparative Evaluation of Chemical Ranking and Scoring
           Methodologies;" Prepared by University of Tennessee Center for Clean Products and clean
           Technologies. The authors reviewed 14 ranking and scoring tools and the report summarizes
           52 of the 147 tools.
       9 U.S. EPA, November 1994, EPA 530-R-94-045
Working Draft                               D-4                               November 1996
-------
                        Conceptual Approach for Contaminant Identification
     ' •     USEPA, 1994 "Environmental Targeting Systems;" Prepared by the EPA Ad Hoc Regional
           Workgroup on Targeting Systems. Summarizes and compares 35 screening methods developed
           by EPA and state organizations.

           USEPA,  1994  "Chemical and Waste Prioritization Systems Used  in OECD Countries:
           Overview, Issues, and Research Needs;" Prepared  by ICR, Inc. for EPA.  Summarizes and
           compares seven prioritization tools, focusing on international tools.

      Based on the literature search and these three major sources of information, the Waste
Minimization Prioritization Team identified a preliminary list of 57 prioritization tools. These tools
are listed in Exhibit 2-1. Using this list as a starting point, a three-tier evaluation was conducted to
limit further evaluation to only those tools that met the team's criteria.  In general, the three tier
evaluation focused on the ability of these tools to address the following:

      •     quantitatively evaluate persistence, human and ecological bioaccumulation potential,
           and human and ecological toxicity;
      •     address different exposure pathways or media;
      •     evaluate  the databases  of information available for persistence, bioaccumulation;
           potential, and toxicity;
      •     review the ability to score categories chemical/wastes;
      •     evaluate factors used for prioritization;
      •     evaluate the software requirements, or the potential to be put in a software package;
           review system maintenance and user support requirements;
           address the ability to overlay with other regulatory lists (e.g., RCRA and TRI);
           address any peer review performed; and
           evaluate the use within the past six years.
      After the three tier evaluation process, only four of the candidate tools  met the team's
requirements  lor a prioritization tool (Identification  numbers 2, 5, 33, and 53).  The  Waste
Minimization Prioritization Team is currently in the process of finalizing their recommendations
regarding these four tools (USEPA, 1996b).

      Superfund Hazard Ranking System (1990)

      In 1980, Congress enacted the Comprehensive Environmental Response, Compensation, and
Liability Act  (CERCLA), commonly called Superfund, in response to the dangers posed by
uncontrolled releases of hazardous substances, contaminants, and pollutants.  Section 105(a)(8)(A)
of CERCLA required EPA to establish criteria for determining priorities among release or threatened
releases of hazardous substances for the purpose of taking remedial action. Criteria and priorities
are to be based on relative risk or danger to public health and take into account the population at risk,
the hazard potential of the hazardous substance, the potential for contamination of drinking water
supplies, the  potential for direct human contact,  and the potential for destruction of sensitive
ecosystems (USEPA, 1990).

                                       Exhibit 2-1
Working Draft                               D-5                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
                          OSW Initial List of Prioritization Tools
 Identification
    Number
                           Name of Tool (Date)
       1
       2
       3

       4
       5
OPPTS Persistent Bioaccumulative Chemicals Screening (1994)
University of Tennessee Chemical Ranking System (1994)
OAQPS Hazardous Air Pollutant Ranking System - Section 112(g) of the Clean Air
  Act Amendments (1994)
OW Sediment Contaminant Ranking (1994 - Draft)
Use Clusters Scoring System Methodology (1993)
       6
       7
       8
       9
       10
OPPTS TRI Listing Guidelines (1991- Draft)
OPPTS TSCA New Chemicals Health and Environmental Review Process (1992)
OPPTS Existing Chemicals Screening Program (1991)
OAQPS Focus Chemicals for Great Waters Study (1991- Draft)
CERCLA Section 102 Reportable Quantities Adjustment Process (1989)
       11
       12
       13
       14
       15
OPPTS Chemical Scoring System for Hazard and Exposure Identification (no data)
OPPTS TSCA TRI Chemical Risk Assessment Pre-Screeitiing Methodology (no date)
Water Quality Guidance for the Great Lakes System-bioaccumulative chemicals
(1995)
Region VI Human Health Risk Index (under development)
Interagency Testing Committee Scoring Methodology for Ranking Chemical Testing
  Candidates (1993)
       16
       17

       18
       19
       20
ATSDR CERCLA 104 Priority List of Hazardous Substances (1992)
Minnesota Indexing System for comparing Toxic Air Pollutants Based upon Potential
  Environmental Impacts (1993)
Oregon Cross-Media comparative Risk Assessment Model (no date)
Canad's toxic Substances management Policy (1994)
Canadian Accelerated Reduction/Elimination of Toxics Scoring Protocol (1994)
       21
       22
       23

       24
       25
Uniform System for the Evaluation of Substances (USES), The Netherlands (1994)
UN ECE Task Force on Persistent Organic Pollutants (1993)
European Community Informal Working Group on Priority Setting Proposal for
  Priority Setting of Existing Chemical Substances (1992)
Critique of the Ontario HazEird Assessment System-Canadian Labor Congress (1992)
Candidate Substance List for Bans or Phase-Outs-Ontario (1991)
       26
       27

       28
       29
       30
Substances and Preparations, Dangerous to the Environment-Denmark (1990)
Great Lakes Water Quality Agreement Standard Methods-International Joint
  Commission (1989)
Existing Chemicals of Environmental Relevance-Society of German Chemists (1989)
Second Priority List-Society of German Chemists (1989)
Criteria to Identify Chemical Candidates for Sunsetting in the Great Lakes (1993)
Working Draft
                           D-6
November 1996
-------
                         Conceptual Approach for Contaminant Identification
  Identification!
    Number
                           Name of Tool (Date)
       31
       32

       33

       34
       35
Systematic Data Collection and Handling Priority Settings (1989)
Multipathway Risk Assessment (Limiting Pathways Analysis) for the Hazardous
Waste
  Identification Rule-HWIR (1994)
WMB Setting Priorities for Hazardous Waste Minimization-Modified Hazard
Ranking
  System (1994 - Draft)
OAQPS Source Category Ranking System (1993)
Ranking the Relative Hazards of Industrial Discharges to POTWs and Surface Waters
  (1991)
       36
       37
       38

       39
       40
Region m Chemical Indexing for the TRI, Part I: Chronic Index (1993)
North Carolina Waste Reduction Management System (1993)
Minnesota Non-Hazardous Industrial Waste Targeting and Pollution Prevention
Project
  (no date)
Classification System for Hazardous Chemical Wastes (1990)
Risk-based Enforcement Strategy or Multi-Media Ranking System (1993 - Draft)
       41
       42
       43

       44
       45
TRI Environmental Indicators Methodology, (1992 - draft)
OPPTS Graphical Exposure Modeling System (1992)
Continuous Release (Priority Assessment Module)-Emergency Response Notification
  System (1991)
OSWER Superfund Hazard Ranking System (1990)
OPPTS Targeting Pollution Prevention Opportunities Using the 1988 TRI (1990)
       46
       47
       48
       49
       50
OPPTS TRI Risk Screening Guide (1989)
Region IV Environmental Targeting System (1994)
National Corrective Action Prioritization System (1991)
Region X Risk Based Multi-media Targeting System (no date)
U.S. Department of Defense Priority Model (1993)
       51
       52
       53
       54
       55
Decision-Support System for Prioritization of Multimedia Dischargers (1993)
Priority List of Environmental Chemical Hazards in Poland (1991)
Ranking Great Lakes Persistent Toxics-General Motors System (1992)
Identify, Screen, and Rank Toxic Chemicals (1992)
Carnegie Mellon Facility Ranking Methodology (1995)
       56
       57
Subtitle D Industrial Wastes Ranking Methodology (1992)
Environmental Fate and Risk Assessment Tool (1995)
Working Draft
                           D-7
November 1996
-------
                        Conceptual Approach for Contaminant Identification
      EPA developed the Hazard Ranking System (HRS) to meet these requirements. The HRS is
a scoring system used to assess the relative threat associated with actual or potential releases of
hazardous substances at sites.  The HRS is the primary tool for determining whether a site is to be
included on the National Priorities List (NPL), the EPA's list of sites that are priorities for long-term
evaluation and remedial response.  The HRS is a crucial part of EPA's program to address the
identification of actual and potential releases. Under the original HRS, a score was determined for
a site by evaluating three migration pathways: ground water, surface water, and air (USEPA, 1990).

      In 1986, Congress enacted  the Superfund Amendments  and Reauthorization Act (SARA).
Section 105(c)(l) required EPA to amend HRS to assure to the maximum extent feasible, that the
hazard ranking system accurately assesses the relative degree of risk to human health and the
environment posed by sites and facilities subject to review.  The subsequently revised HRS assesses
the human health risks associated with actual or potential contamination of surface water used for
recreation or drinking water and take into account the potential migration of any hazardous substance
through surface water to downstream sources of drinking water. Two  criterion were added for
evaluating sites: actual or potential contamination of the ambient air and threats through the human
food chain.  In addition, high priority  was assigned to facilities where  the release of hazardous
substances resulted in the closing  of drinking; water wells or contamination of a principal drinking
water supply (USEPA, 1990).

      The revised HRS considers each of the following site-specific characteristics:

      •     The quantity, toxicity, and concentrations of hazardous constituents that are present in
           waste;
      •     The extent of,  and potential fDr, release  of such hazardous  constituents  into the
           environment; and
      •     The degree of risk to  human heeilth and the environment.

      The HRS includes a scoring system based on factors grouped into three categories.  The factor
categories are multiplied and then normalized to 100 points to obtain a pathway score.  The four
pathways considered under HRS are: the ground water migration pathway; surface water migration
pathway; soil exposure pathway;  and air migration pathway. The factors for these pathways are:
observed or potential release (net precipitation, depth to aquifer, travel time, overland flow, and
water by flood), waste characteristics  (toxicity, mobility, persistence., bioaccumulation, area of
contamination, and waste quantity), and targets (food chain individual, sensitive environments,
nearest well, population, workers,  resources, and wellhead protection area).The final HRS score is
obtained by combining the pathway scores using a root-mean-square method (USEPA,  1990).

CRITERIA FOR SCORING

      Scoring under most hazard ranking systems involves the assignment of scores to several
criteria and combining these individual scores in some manner for an overall hazard ranking score.
This section provides a list of potential scoring criteria for a hazard ranking system on drinking water
exposure.
Working Draft                              D-8                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      The hazard ranking for the assessment of exposure to drinking water contaminants involves
several criteria related to the population exposed, including the frequency and duration of exposure
and the concentrations or quantities of the chemicals.  These criteria can be characterized by direct
information (e.g. monitoring chemical  levels in drinking  water or its sources) and  indirect
information (e.g., production, release,  and fate and transformation data).  For the hazard ranking
systems, criteria characterized by direct information is considered more important than criteria
characterized by indirect information.  For many potential drinking water contaminants, however,
data sources for criteria based on direct information is lacking. Therefore, in many cases, hazard
ranking systems must rely primarily on criteria based on indirect information.

      The purpose of this section is to list and describe potential scoring criteria relevant to hazard
ranking for the assessment of drinking water exposure. Criteria based on both direct and indirect
information are listed.  As  many criteria as  possible are described at  this early stage in the
development of a drinking water exposure ranking system.  The final exposure ranking system will
include some subset of these scoring criteria and might include other additional scoring criteria
identified as the ranking process is developed. In the actual hazard ranking system, the scoring
criteria will be aggregated based on some data weighting factors. Another possibility in combining
the scoring criteria involves the normalization of scores to compensate for the lack of data.

      The remainder of this section summaries the potential scoring criteria for an exposure ranking
system. The list of attributes was developed from available literature, previous ranking systems, and
discussions with EPA staff.

      Production

      A commonly  used indicator of the potential of contaminant exposure is annual production.
Since this information is quantitative and is readily available from compiled sources, scoring for
ranking purposes is rather straightforward. Annual production values provide an indication of the
quantity of chemicals available for release into the environment. Production amounts, however, only
provide information  on the amounts intended for entry into commerce and do not provide insight as
to the amounts released to the environment. Therefore, production volumes are not an index of
exposure for chemicals in waste streams, such as emissions and effluents, which directly pollute the
environment.  In addition, production figures do not address exposure to impurities, captive
intermediates, naturally occurring chemicals, or degradation products (Hushon and Kornreich, 1984
and USEPA, 1992).

      Use/Application

      Use and application patterns can provide information on the type, frequency, and potential for
human contact  Information on application rates for pesticides on crops, for example, is readily
available. Use and application information coupled with leachability (i.e., persistence and mobility)
data provide a strong indication of  potential groundwater contaminants.   For non-pesticide
chemicals, however, accurate and current data on use information is not as complete.

      Since different uses/applications result in varying human exposure, a quantitative evaluation
of uses/applications would be useful  for scoring or ranking purposes.   In some cases, general
Working Draft                               D-9                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
categories of use/applications could be determined based on the structure and properties for similar
chemicals. Different uses and applications could be used to differentiate between environmental and
consumer exposures and occupational sources (Hushon and Kornreich, 1984).

      Release

      Under EPA's Toxic Release Inventory, a release is defined as an on-site discharge of a toxic
chemical  to the environment.  Releases of chemicals to the environment occur as a result of
production losses, uses, and disposal. Emissions to the air, discharges to bodies of water, disposal
to land, and contained disposal into underground injection wells are considered releases under TRI
(USEPA,  1994).

      Releases to air include stack (i.e., through air streams such as staicks, vents, ducts, or pipes)
and fugitive (i.e., equipment leaks, evaporative losses, and building  ventilation system losses)
emissions. Releases to water include discharges to streams, rivers, lakes, oceans,  and other bodies
of water.  Releases due to runoff, including stormwater runoff, are considered releases to water.
Underground injection is a contained release  of a fluid into a subsurface well for the purpose of
waste disposal. Under TRI, this release includes Class I (i.e., municipal and industrial hazardous
wastes) and Class V (i.e., non-hazardous wastes) wells. Releases to land include  disposal of toxic
chemicals in wastes to a landfill, to land treatment/application farming, to a surface impoundment,
and other land disposal such  as spills, leaks, or waste piles (USEPA, 1994).

      In  scoring or ranking systems, release information is often used in conjunction with
production, use, or disposal volumes. Release rates can be used as multipliers to modify production
amounts based on the release mechanisms (USEPA, 1992).

      Chemical/Physical Properties

      Chemical  and physical properties of potential contaminants can provide  insight as to the
persistence, distribution, and transformation of chemicals  in various environmental media. Also,
chemical and physical properties can be used to score or rank similar contaminants as a surrogate
for limited information on  other scoring criteria.

      Water Solubility

      The water solubility of a chemical provides insight as to the fate and transport of a chemical
in the environment.  It has been found that chemicals that are highly soluble in water tend to remain
dissolved in the water column and are less  likely to adsorb onto sediments, suspended solids, or
soils.  As a result, they tend to be highly mobile in soil, more easily biodegraded by microorganisms,
and are less likely to volatilize from water. For chemicals with a low solubility, the opposite is true.
These chemicals are more likely to bind strongly to soils, suspended solids, and sediments, and
volatilize readily from water. They are less  likely to be biodegradable in surface  waters and soils.
All organic chemicals are soluble in water to some extent. Chemical solubilities  as low as 1 Aig/L
have been measured, while other  chemicals are considered to be infinitely soluble in water (i.e.,
soluble in all proportions)  (Howard,  1990).
Working Draft                              D-10                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      Vapor Pressure

      The vapor pressure of a chemical provides insight into the transport of a chemical in the
environment. The vapor pressure of a chemical influences its volatility from soil and surface waters.
Chemicals with high vapor pressures tend to evaporate readily from dry soil surfaces. In water,
however, both  the chemical's vapor pressure and solubility determine its evaporation rate from
surface waters  and wet soil  surfaces.  Vapor pressure with water solubility are used to calculate
Henry's Law constant (see below) (Howard, 1990).

      Henry's Law Constant

      For many chemicals, volatilization can be an important removal process, with half-lives on
the order of several hours. Henry's Law constant, the air/water partition coefficient for a chemical,
provides a qualitative indication of volatilization.   This non-dimensional constant relates the
chemical concentration in the gas phase to its concentration in the water phase.  By dividing the
vapor pressure in atmospheres (atm) by the  water solubility in mmole/m3,  however, it can be
expressed in terms of atm-m3/mole. Henry's Law constant provides a qualitative indication of a
chemicals's evaporation rate from water and moist soils.  These constants can be measured directly,
calculated from the water solubility and vapor pressure, or estimated from the structure of the
chemicals (Howard, 1990).

      Environmental Fate and Transformation

      Fate and transformation data are useful in demonstrating how chemicals behave in soil, water,
and air, and how human exposure is likely to occur and should be considered under an exposure
ranking system. Factors relevant to fate and transformation that might be considered under a ranking
system include data on the chemicals persistence, distribution, reactivity, and transformation through
such processes  as biodegradation, photolysis, and hydrolysis.  Distribution factors that are often
considered include bioaccumulation and bioconcentration potentials. Fate and transformation data
can be determined by field data, laboratory data, and by surrogate data.  Some of these factors have
been used in exposure scoring as weighting factors or multipliers instead of additive terms because
of their broad impact on exposure (Howard, 1990 and Hushon and Kornreich, 1984).

      Persistence

      Persistence of a chemical in the environment can be defined as the tendency of the chemical
to persist, or survive, in the environment without transformation into  another chemical form
(USEPA, 1996b).  Persistence provides an indication of the ability of a  chemical to exist in the
environment and, therefore, be available for exposure  to people.

      In hazard ranking systems, the persistence of chemicals is reflected in scoring against those
chemicals that ;ire easily degraded or transformed.  Persistent chemicals are those believed to spread
through the environment.  The possibility of exposure to these chemicals is  increased with increased
environmental  residence time.   It  should be noted that if a  chemical is  easily degraded or
transformed, the resulting degradation  products might be more toxic than the chemical initially
Working Draft                              D-11                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
released.  In these cases, the secondary pollutants could be considered under the hazard ranking
system (Howard, 1990 and Hushon and Konireich, 1984).

      In EPA's Hazard Ranking System, persistence is measured by the half-life of a chemical in
water (T05) (USEPA, 1990). As defined uncier the Hazard Ranking System, the T05 of a chemical
is calculated from the hydrolysis half-life, biodegradation half-life, volatilization half-life, and the
photolysis half-life as follows:

           T05 = l/(l/h + 1/b + 1/p + 1/v)

           where:          T05 = overall half-life of a chemical
                            h =  hydrolysis half-life
                            b =  biodegradation half-life
                            v =  volatilization half-life
                            p =  photolysis half-life

      Mobility

      Mobility of a contaminant in the environment was measured in EPA's Hazard Ranking System
using the octonal/water partition coefficient (K^).  This coefficient is an indicator of the fate of
organic chemicals in the environment and is related to water solubility and soil/sediment adsorption
coefficients.  Generally,  chemicals  with  low K^ tend to have high water solubilities and low
adsorption coefficients.

      Mobility of a contaminant in  the environment also is closely related to its soil adsorption
characteristics. For many chemicals, especially pesticides, experimental soil or sediment partition
coefficients are available.  This data is used to derive an adsorption coefficient (Kd).  The Kd values
for individual soils or sediments  can then be normalized to the organic carbon content of the soil or
sediment by dividing by the organic content (K^.).  Of the numerous soil properties, such as organic
carbon content, particle size, clay mineral composition and cation-exchange capacity,  organic carbon
is the most important for undissociated organic chemicals.  Koc can be determined experimentally
or calculated using the water solubility or K^ and various regression equations. The adsorption
values are used to determine the likelihood of mobility through soil or adsorbing onto sediments.
      Leachability

      Leachability is usually employed to characterize pesticides.  Leachability includes both
mobility and persistence factors. Leaching refers to the downward movement of pesticides into the
soil.

      Bioaccumulation/Bioconcentration

      Bioaccumulation and bioconcentration are considered measures of the distribution of
contaminants in the environment.  Bioaccumulation is the uptake and retention of chemicals by
living organisms. Organisms are exposed to pollutants through contaminated water, sediment, and
Working Draft                               D-12                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
food. A pollutant bioaccumulates if the rate of intake into the living organism is greater than the rate
of excretion or metabolism. Bioacculumation results in an increase in the tissue concentration
relative to the exposure concentration in the ambient environment. The presence of contaminants
in animal tissue reveals the presence of pollutants in water bodies that routine monitoring of water
alone can miss.

      Monitoring Data

      Drinking water and  source water concentrations are direct  measures of drinking water
exposure to chemical contamination. Lacking precise exposure data, occurrence data serve as a
surrogate for dose.   Aspects of monitoring data that are important include: analytical methods
employed; number of samples; percent of samples positive; the minimum detection limit; the range
of concentrations; and statistical characteristics such as the mean, median, and standard distributions.

      Monitoring data reflect the prevalence of chemical contamination by the frequency  of
detection of contaminants and also the magnitude of the problem as  determined by the maximum
chemical concentration detected. The maximum detected concentration is a useful parameter since
it reflects the worst-case scenario.

      Occurrence in Drinking Water

      Monitoring data taken after water treatment comprises  occurrence in drinking water. Sample
locations include in the clearwell of a water treatment system, the entry to the distribution system,
and at the tap. Drinking water occurrence data can be found in federally-sponsored national surveys,
state and regional (including watersheds) studies,  and local evaluations. Of particular importance
is data by system size (i.e., small, medium, and large by population and flow) and source water (i.e.,
ground and surface), and to a lesser extent location (i.e., geologic conditions) and ownership (i.e,
public versus private).

      Occurrence in Source Waters

      Monitor] ng data for source waters provides an indication of contaminant distribution. Source
water occurrence data indicate the prevalence of contaminants in the  potential influent of drinking
water supplies (i.e, pre-treatment levels).

      Exposure

      Exposure information includes the characterization of the size and type of populations exposed
as a factor in ranking chemicals. For drinking water exposure, the size of the population is the
number of people served by public drinking water systems.  A subset of this population, however,
may be affected significantly by a contaminant and this population subset would be identified as a
vulnerable population. For most contaminants considered for drinking water hazard ranking, precise
exposure data will be lacking.

      Vulnerable Populations
Working Draft                               D-13                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      Vulnerable populations are those people that are highly susceptible to specific health effects.
These populations can include infants, children, pregnant women, elderly, and those with serious
illnesses including the immunocompromised. Scoring for this attribute would reflect the existence
of potential vulnerable populations. The 1996 Amendments of the Safe Drinking Water Act requires
the Agency to prioritize new regulations of contaminants that present the greatest public health
concern, including their effects on vulnerable populations.

      Exposure Data

      The frequency and duration of exposure must be known to determine the actual quantity of
a substance taken in by humans over a specified time and route.  This dose information is not readily
available for most chemicals and is rarely used in scoring.

NEXT STEPS

      The EPA's Contaminant Identification Method Workgroup will continue the development of
a contaminant ranking system for drinking water exposure. One of the next steps for the Workgroup
is to prepare for the  scheduled December  1996 stakeholder meeting  on the Contaminant
Identification Method.  An exposure ranking system for drinking water contaminants will be a
primary topic at this meeting. This document on exposure ranking systems and scoring criteria will
be part of a support package provided to the stakeholders during the meeting.

      This document identifies some examples and sources for ranking systems and associated
scoring. No literature search was conducted to identify additional published ranking systems. A
revised draft of this document should be  prepared to support the December 1996 stakeholder
meeting.  Specifically, a literature search should be performed to identify available ranking systems.
In addition, several  ranking systems were  identified in this document that need to be evaluated
further. Also, this document should be expanded to cover additional sections on: (1) ranking system
issues; and (2) potential data sources for the scoring criteria.

      A section on ranking system issues would involve a discussion on the following topics:

      •     data selection, availability, prioritization;
      •     exposure criteria;
      •     weight or strength-of-evidence;
      •     score aggregation, weighting and normalization;
      •     chemical categorization versus individual ranking;
      •     professional/expert judgement;
      •     complexity, transparency, and overall ease of use;
      •     flexibility of ranking system;
      •     computerized, or capable of being computerized; and
      •     validation of the ranking system.
      A section on potential data sources would provide an evaluation of available databases for the
following topics:
Working Draft                              D-14                              November 1996
-------
                        Conceptual Approach for Contaminant Identification
      •     production data;
      •     release data;
      •     use and application data;
      •     chemical and physical properties;
      •     fate and transformation data;
      •     occurrence data (drinking water and source water); and
      •     vulnerable populations.

      A draft exposure ranking system will be developed after the December 1996 stakeholder
meeting.  To develop a draft ranking system, several issues would have to be addressed, including:
specific scoring criteria; the use of weightings and/or normalization procedures; the selection of data
sources for the individual scoring criteria; and the procedure for aggregating scores for the various
scoring criteria.  In addition to these technical questions, software considerations would have to
addressed. It is possible only the framework of a draft exposure ranking system could be used for
the purpose of soliciting feedback in the form of comments and suggestions from the stakeholders
at the December 1996 meeting.
Working Draft                               D-15                              November 1996
-------
                       Conceptual Approach for Contaminant Identification
REFERENCES

Howard, P.H., 1990 Handbook of Environmental Fate and Exposure Data for Organic Chemicals:
     Lewis Publishers, Chelsea, MI, pp xi-xxi.

Hushon, J.M. and M.R.  Kornreich, 1984 "Scoring System for Hazard Assessment;" In: Hazard
     Assessment of Chemicals: Current Developments. Volume 3, Ed. Jitendra Saxena, Academic
     Press, Orlando, FL.

USEPA, 1996a "Toxicity Ranking Methodology;" Final Report, Prepared for U.S. Environmental
     Protection Agency, Office of Science and Technology under EPA Contract 68-C3-0342, July
     24, 1996.

USEPA, 1996b "Waste Minimization: Where to Begin?;" U.S. Environmental Protection Agency,
     Office of Solid Waste and Emergency Response, Draft Document, Washington, DC.

USEPA,  1994 "1992 Toxics Release Inventory:  Public Data  Release;" U.S. Environmental
     Protection Agency, EPA 745-R-94-001, Washington, DC.

USEPA, 1992 "Development of a Priority Pollutants List for Drinking Water;" U.S. Environmental
     Protection Agency, Office of Ground Water and Drinking Water, prepared under EPA
     Contract 68-C9-0040, Washington, DC.

USEPA, 1990 "Hazard Ranking System; Final Rule;" 40 CFR Part 300, 55 FR 241, December 14,
     1990.
Working Draft                             D-16                            November 1996
-------