OCR error (C:\Conversion\JobRoot\00000ANP\tiff\2000W7GK.tif): Unspecified error

-------

-------
                                               IV-^jitti Canter for EnviTvwticr.t£l

                                 —            i^r./'.-Tl'str^ot (3PM52)
                          EDWIN B.  ERICKSOM    f*-^-}'-^- PA. 19103
 Citizens Concerned With  the  Environment:

      The Mid-Atlantic  area that encompasses EPA's  Region III has
 a diverse  environment.   In our region, we  are  challenged to find
 innovative ways  to  protect the public health  and the environment
 we  share  with various  species  of flora  and  fauna.   This  is  a
 challenge we gladly accept.

      This  report summarizes the  current status  of environmental
 problems within the region in terms of the relative risks that are
 posed by eighteen defined problem areas.  This is our first attempt
 to  use  the comparative  risk evaluation  methodology, which  is an
 innovative approach to ranking  environmental problems.

      We  at EPA  share  everyone's  concerns about  the  continued
 presence of contaminants  and pollutants  which  threaten our health
 and damage our  environment.   Often,  new  problems  appear  before
 existing problems can  be brought under  control  or  eliminated.
 Restraints on  the Federal budget  require agencies  like EPA to
 direct resources  to those areas where they can do  the  most  good,
 that is to achieve the greatest  reduction  in environmental risks.

      EPA Region  III  Comparative Risk  Project  is our  response to
 the question, "Which environmental problems pose  the highest risk
 to public health and  the environment?"  We are currently working
 on the next  step, determining how to  address  those highest risk
 problems.

      Our understanding of the  complexities of the  environment in
 which we live continues to improve, but demands  further examination
 in our approach to solving environmental problems.   Government at
 all  levels must work  together  to find  solutions  that manage
 environmental risks.  Therefore,  we are releasing  this  report to
 help identify the relative risk of our environmental problems and
 focus our attention to reducing those  risks.

      I hope  this report will  help us  improve  as  environmental
 managers and help the public to better understand our  activities.
 We would be pleased to have your reactions  to this  report.
Edwin B. Erickson
Regional Administrator
                                             Re£ion
-------
                        EXECUTIVE SUMMARY

     Environmental  problems  which  have  only  recently  "come  to
light"  or for  which  existing  control  programs  have not  been
completely successful present high  risks to  human  health and the
environment in the area that encompasses  EPA's Region III.  While
the technology to address these problems  is  generally available,
the associated costs will be prohibitive.  Region III does not have
the necessary  resources  in terms of  personnel, funds,  or  legal
authority, to  resolve all  of  these  problems  in  the immediate
future.

     Region III conducted the Comparative  Risk Project to determine
which of  the  Region's environmental problems posed  the greatest
risks to human health, ecology,  and societal welfare.

     Region Ill's Comparative Risk Project is consistent with the
national report,  Unfinished  Business, and with similar projects in
Regions I and X.   This summary discusses the major findings of the
project, the methodology used to  obtain  those findings, and some
of the additional benefits derived from undertaking this project.

     FINDINGS

     The findings of the project are priority rankings of eighteen
environmental problem areas  in terms of human health, ecology, and
welfare risks.  The  project  focused  on the uncontrolled portion of
an environmental  problem,  known as residual risk.  Risks that are
currently  abated were  not  considered.    The  findings were  as
follows:

     *  The  highest  ranking  human  health  risks are  indoor  air,
indoor radon,  pesticides  in food,  radiation  (other  than radon),
nonpoint sources of pollution,  and  contamination associated with
the operation and maintenance of public water supplies.

     * The problem areas that rank highest for ecological risks are
the physical modification of terrestrial  and aquatic habitats and
nonpoint source discharges of pollution.

     * In the welfare  category,  the  problem areas with the highest
ranking are criteria air pollutants and acid deposition.

-------
     *  Assisted regional  management  in planning  and  priority
setting in Region  III.   This analytical tool may be used  by the
senior level managers in the  decision-making  process  employed in
establishing resource allocations.

     * Determined the importance of explaining the limitations of
this  project.     The   project  does  not  indicate   trends  in
environmental  programs,  or  the  consequences  of  reducing  or
eliminating base programs.

     * Identified the need to put more effort  into developing and
improving data management systems.   Regional and state data bases
were often inaccessible or lacking information.

-------
METHODOLOGY

     The method used to compare environmental problems is closely
aligned with the procedure used  for  risk assessment.   The Region
approached  the   project  systematically,  utilizing   the  best
professional   judgment   among   Regional  experts.     The  staff
represented all regional environmental programs with a variety of
scientific expertise.  Three work groups were formed to address the
problem  areas  based on human  health,  ecology,  and  welfare.
Existing data  on pollutants,  exposure,  effects,  and  costs  were
collected and  analyzed.   Where gaps existed,  work  groups relied
heavily on  professional  judgment.   In  some areas,  the  project
represents the Region's  best  opinion, given the limited knowledge,
rather  than   quantitative   analysis.     The  participants  felt
relatively confident iri their final rankings.

     The regional  staff  analyzed  problem areas  which generally
correspond with  existing EPA programs  or statutes.    The Region
excluded certain problem  areas that  were beyond  the  influence of
EPA Region  III,  such as  global  warming and ocean dumping.   The
final list consisted of eighteen problem areas that were analyzed
for this report.

     For each  problem area,  three types  of  risk  were evaluated;
public  health  risk,  which included  cancer and non-caner risks;
ecological risks; and welfare  risks, or social costs.   Each type
of risk was  evaluated separately.  No attempt was made to determine
whether one category was more important  than another or to combine
risks across problem areas.
BENEFITS

     The Region  benefitted  from the Comparative Risk  Project in
several ways:

     * Identified  that  high-ranking public health  problems have
little  or  no  ecological  impacts,  except for  nc  ^oint  source
discharges of pollutants.  Regional attention needs to be focused
in both areas to reduce risks.

     * Identified several problem areas which did  not pose the high
risk  that  were  expected,   including  the  hazardous  waste  and
underground storage tank programs.

     *  Increased the  level  of communication  among the  various
environmental programs.  Provided  a greater  understanding of the
relative  risks  among   the   problem  areas.     The  work  group
participants were especially  enlightened  to  the  risks  associated
with other environmental problems.

-------
                                TABLE OF CONTENTS
I. Introduction	    1
       Themes in Environmental Management	    1
       Origin and Focus of the Comparative Risk Project	    2
       Purpose of This Report	    3

II. Overview of the Project's First Year	    4
       The Process	    4
       Roles and Responsibilities of Participants  	    6
       Problem Definitions	    8
       Ground Rules 	   10

III.  Ranking Environmental Risks in Region III  	   14
       Health Risk	   14
              Methods	   14
              Ranking  Results  	   17
       Ecological Risk  	   19
              Methods	   19
              Ranking  Results  	   21
       Welfare Risk	   24
              Methods	   24
              Ranking  Results  	   27
       Comparison of Rankings Across Types of Risk	   30
       The Effect of Problem Definitions on the Rankings	   34

IV.  Observations	   37
       Comparison with Unfinished Business Findings	   37
       Level of Confidence in Ranking Results  	   41
       Information Gaps That Might Be Filled in  the Future	   42
       Benefits of the Project	   43

V. Next Steps	   45

-------
                                        I.  Introduction
Themes in  Environmental Management

        The early 1970's  marked the beginning of a new national commitment to restore and protect the
environment.  The Environmental Protection Agency (EPA) was created to clean up the gross and highly
visible pollution that choked our nation's air and fouled our rivers and streams.  Using newly developed
environmental laws,  regulations, and technologies,  EPA and  state environmental agencies  have made
significant progress  over  the  past  two  decades in controlling air  pollution and restoring  water  quality.
During this time, several  major  themes  in environmental management have emerged to make this project
timely.

Setting Priorities

        One important theme in the evolution of environmental protection has been an increase in the
complexity and number of issues without an accompanying increase in  resources to address them.  With
advances in scientific knowledge and measurement techniques, previously unrecognized problems have been
detected.  DDT, asbestos,  PCBs, dioxin,  radon,  indoor air pollution, the ozone layer, and  many other
environmental problems have come to  the forefront of public concern.  As scientific understanding and
public awareness increase, the list of issues continues to grow.

        Increased governmental responsibilities for managing these issues, particularly at the state and local
levels, have  not always  been matched  with sufficient funds.  In  an  era of tight budgets,  government
environmental programs need  to set priorities carefully to get the  greatest possible benefit out of available
resources.

Risk Analysis

       Government  has  traditionally faced  environmental challenges in a reactive mode  --  setting and
enforcing environmental standards. With increased understanding of the problems comes the realization that
a reactive approach must be supplemented with new strategies that better anticipate and plan for future
environmental concerns.  A second  important theme in environmental management is the development of
risk analysis  as a tool to help  sort through environmental problems.  Risk analysis can provide a common
denominator across disparate  environmental areas to indicate the seriousness  of the problem.  To avoid

-------
reactive policy-making, government officials can use risk analysis to evaluate and compare the multitude of
environmental problems.

Regional Differences

        The third theme providing important background for the Region III Comparative Risk Project is
the growing importance of regional differences in environmental management.  National pollution control
programs have effectively provided most citizens with a minimum level of environmental protection. Many
of the remaining problems vary from area to area and require site-specific, tailored controls at the regional,
state, or local level for effective  mitigation.   Nationwide requirements cannot always  address unique local
conditions, such as exposure patterns, hydrogeology, or meteorology, which demand unique local solutions.

        These three themes underscore the need  to set environmental priorities carefully.   Region III
believes the Comparative Risk Project will provide insight to help develop a  new environmental planning
process.
Origin  and Focus  of the  Comparative Risk  Project

        EPA Headquarters earmarks nearly all the resources it provides  to the Regions  to accomplish
specified levels of work output.  The Regions have limited influence over  national priorities or specified
work levels.  For several years,  Region III has sought more flexibility from  EPA Headquarters to address
unique regional problems.  Nearly all of the resources provided by EPA Headquarters to the Region are
earmarked by program.  Region III has attempted  to improve this situation by  systematically determining
regional priorities, how they differ from national priorities, and how to obtain increased flexibility to address
them.

        Region III developed its MERITS (Measurable Environmental Results Initiatives) program to help
address high-risk, high-priority regional problems. Every year the Region solicits MERITs proposals from
the staff.   It encourages the staff to develop proposals  that address emerging  high-risk  environmental
problems that are inadequately addressed by existing programs.  A committee  of regional  managers and
senior staff evaluates the proposals in terms  of their feasibility and risk reduction benefits. The Region has
completed over 50 projects that would otherwise not have occurred.

       Although Region Ill's MERITS have been nationally recognized as an innovative method of focusing
resources on high-priority environmental problems,  the  MERITs process has limited scope. The Region

-------
allocates only a small fraction of its  resources to MERITs projects.   The Region wanted to conduct a
comparative risk analysis  to expand on the MERITs process by addressing the entire scope of regional
programs.  The analysis would produce a relative ranking of environmental programs and a starting point
for developing comprehensive regional resource allocation priorities.  In a narrow sense, the Region wanted
to use comparative risk to allocate its  own discretionary resources by identifying  problems for development
of MERITs proposals.  In a  broader sense, the Region planned  to  use the assessment to  increase its
flexibility  in  allocating  non-discretionary resources  by  highlighting what is  unique about  Region  Ill's
environmental problems.

        Region  Ill's  Comparative Risk Project is designed to improve the Region's process for setting
management priorities among environmental problems. The central premise is that environmental protection
priorities should be placed on reducing demonstrated risks.   The project has three specific objectives:

        o  To  develop a relative  ranking  of environmental  problems  in Region  III  based  upon the
           seriousness of the risks they pose.

        o  To establish a  set of effective initiatives to work toward  mitigating these problems.

        o  To enhance the Region's management system using the rankings of problems and  initiatives.

        The Comparative  Risk Project has  two phases corresponding generally to the first two of these
objectives:  a risk assessment phase and a risk management phase.  In the first phase,  the regional  staff
analyzed and ranked the major environmental problems  facing the Region. They ranked the  problems in
terms of the risks they pose to  human  health, to natural ecosystems, and to economic welfare in the Region.
The project covered  all of the environmental media  - air,  surface water, land and ground water.  In the
second phase, now under way, the staff is evaluating initiatives to mitigate these problems.  The Region will
consider the ranking results and initiatives in the  regional budget and planning processes.
Purpose  of This Report

       This report focuses on the activities and results of the first phase of the project  The report briefly
describes the general process  the Region used in  conducting the comparative risk analysis and  provides
details on the project's major  steps.  The report describes in more detail the  analytical methods used to
rank the  environmental problems, and the ranking results.

-------
                         II. Overview  of the  Project's First Year
The  Process
       The main objective in the first phase of the project was to rank the risks posed by the major
environmental problems in Region III.  The basic approach included defining the list of environmental
problems, developing methods to analyze the risks they posed, collecting and analyzing data, and ranking
the problems using the data and best professional judgment.  This section briefly describes the process the
Region used in the first phase of the Comparative Risk Project.
                      STEPS IN THE  FIRST PHASE OF THE COMPARATIVE RISK PROJECT
Beginning the Project

       In the spring of 1987, staff in the Environmental Services Division (ESD) drafted a work plan for
the Comparative Risk Project.  The  Environmental Management Committee (EMC), a standing multi-

-------
media committee in Region HI, and a committee of senior managers reviewed and modified the work plan.
On July 9, 1987, the Regional Administrator approved the work plan and sent it to EPA Headquarters.
This marked the formal beginning of the project.

Designating Roles and Responsibilities of Participants

       Once the work plan had been approved, the Region organized personnel to perform specific project
tasks.  The Region designated project  management staff, a Steering Committee to provide direction and
oversight,  the EMC to coordinate and resolve issues, and three work groups (one work group for each of
health, ecological, and welfare risk) to perform the analysis and rankings. EPA Headquarters provided staff
assistance and contractor support.

Defining the Problem Areas

       The project management staff based Region Ill's problem list on the list from Unfinished Business.
The Steering Committee made several major changes to the original list by requiring the problem areas to
be mutually exclusive and by dropping problems that were national in scope and unlikely to be  addressed
at the regional level.

Establishing Ground Rules

       To structure the ranking of environmental problems, the Region adopted several important ground
rules  identical to those used in Unfinished Business.  The rules focus on risk as  the best way of comparing
the relative seriousness of environmental problems.  In particular, the  project examines risks that remain
after  assuming current levels of environmental controls as the base.

Developing Analytical Methods for Comparing and Ranking Problem Areas

       Each of the three work groups  developed  methods to provide a structure for determining the data
to collect and analysis to perform for each problem area.  These methods imposed consistency within each
work  group across the  problem areas.

       Lack of standard risk assessment methods  in several areas, unfamiliarity with risk analyses, and the
novelty of the  project posed  a problem for the work  groups in developing  analytical risk assessment
methods.  Each work group with the assistance of EPA Headquarters staff and contractors basically designed
its own new method of analysis after considering available options.

-------
Analyzing the Problems

        Each work group assigned a problem area to a single individual in the group. The lead individual
for a problem area analyzed the problem in accordance with the work group's methodology and reported
the results to the entire work group.  As a first step, the lead work group member directed contractors to
prepare plans that outlined the data to collect and analysis to be performed for the problem area. These
plans typically relied  on  the results of national  analyses, modified to reflect  regional data  sources.
Contractors  provided summaries of the results  to the lead individuals.

Ranking the Problems

        Each work group  used a similar procedure to develop its final  rankings.  The lead work group
member for a problem area reviewed the data  and  analysis  provided by the contractors and developed a
tentative ranking or score  for the problem.   The lead described the analysis and  the proposed ranking to
the entire work group. Intensive discussions  followed, often involving questioning the analysis  or data,
developing group judgments, and directing work group members or contractors to do further analysis.  The
discussions continued until the group reached a consensus on the problem's  rank.  After  the work groups
discussed each problem area, they reviewed  the entire list of rank-ordered problems, making adjustments
as necessary.  Each work group concluded by characterizing  its  level of confidence  in the final ranking.

Reviewing and Documenting the Rankings

        Representatives of the  work groups presented their risk rankings to  the EMC and the  Steering
Committee.   Neither  group made any changes to the rankings.  After  approval of the rankings by the
Steering Committee, the project staff wrote  this report summarizing the results of the first phase of the
project.
Roles  and Responsibilities of Participants

        More than 50 regional employees representing all EPA's environmental programs participated in
the Comparative Risk Project. They included senior professional staff and all levels of management The
participants are listed in Appendix 2.   The Region organized these participants to perform the following
functions:

-------
Project Management Staff

       The Region designated as the project manager an individual in ESD who was familiar with risk
assessment and  had managed the Region's MERITS  process.  The Geographic Studies Branch at EPA
Headquarters  provided a  full-tune staff person to support Region Ill's project.  EPA Headquarters also
provided other staff and contractor assistance as needed.

Steering  Committee

       The Steering  Committee  was formed to provide general guidance and oversight for the project;
assign personnel; review and approve the work plan, list of problem areas, project ground rules, analytical
methods, and final rankings; evaluate regional activities to address project results; and integrate the project
with the  Region's planning and resource  allocation  processes.    Chaired  by the  Deputy  Regional
Administrator, the  Steering Committee consists of the  Region's Division Directors and Office Heads.  The
Committee has final approval on  all  significant project activities.

Environmental Management Committee (EMC)

       The EMC is a standing committee that consists of mid-level managers and senior  staff  across all
programs in the Region. The Branch Chief of the Environmental Assessment Branch in the  Environmental
Services Division chairs the EMC. For purposes of the Comparative Risk Project, the EMC  assists with
project design and  implementation and provides recommendations to the Steering Committee on such issues
as the  list of problem areas, analytical methods, and development of the final rankings.  The EMC  also
coordinates and resolves issues  raised by the three work groups.

Work Groups

       Three work groups (Health Work Group, Ecological Work Group, and Welfare Work Group) were
formed to carry out the bulk of the analysis needed to rank the problem areas.  Chaired by a Branch Chief,
each work group consists of approximately ten individuals, ranging from technical staff to regional managers,
who represent all  major  program areas.  Although  the  Region does not have many individuals with
methodological expertise in health, ecological, or welfare risk assessment, the project organizers attempted
to assign  individuals with appropriate  experience to the work groups.  Three lexicologists sit on the Health
Work Group; two  ecologists are on the Ecological Work Group; and one economist sits on the Welfare
Work Group.

-------
Problem Definitions

        Region III had many options to consider in defining the problem areas. . Environmental problems
may be divided by pollutants (e.g., benzene, microbials, cadmium), by sources of pollution (e.g., automobiles,
power plants,  underground storage tanks), by media  (e.g.,  air, surface water), by geographic region (e.g.,
Chesapeake Bay,  the Kanawha Valley, Virginia), or by other factors.  The project management staff decided
to define the problem areas to correspond with areas addressed  by major EPA programs, so that it would
be easier to translate the results of a risk ranking into implications for program priorities.  This produced
a mixture of problem areas defined as pollutants, sources and media.

        To allow  a comparison of regional and national environmental risks, the project management staff
decided to modify the list from Unfinished Business, which included 31 problem areas roughly corresponding
with EPA national program areas. The Region modified the list in two ways - by defining the problems
to be mutually exclusive and by excluding problem areas outside regional control.

        Defining  the problems to be  mutually exclusive was a major divergence from  the  Unfinished
Business list  Where a risk might plausibly be included in any of several problem areas, the Region assigned
it explicitly to one problem area,  to avoid double-counting the risk.  For  example,  if a  Superfund  site
contaminates ground water that is used for  drinking, the resulting  health  risk  could be counted as  a
Superfund site problem, a ground-water problem, a drinking water  problem, or all three.  Multiple EPA
programs  may address  this problem.  The Region  assigned this risk to  the Superfund  problem area.
Regional staff devoted substantial effort to identifying areas of potential program overlap and assigning them
uniquely to a problem area.

        Although the Region wanted  a comprehensive list  of problems,  it did not want  to include
environmental problems  of a  national or global scope which  could  not be solved  by regional action.
Consequently, the project organizers excluded problem areas that  were beyond the Region's level of control,
such as global warming or ozone depletion.  The final list of problem areas covered all environmental risks
addressed by Region III.

        Exhibit 1 shows the resulting list of 18 environmental problems.  Appendix 1 provides full definitions
of the problems.  The definitions identify what is included and excluded from each problem.  The definitions
play an important role in explaining the rankings.

-------
                                         EXHIBIT  1
                                   List of Problem Areas
1.   Criteria Air Pollutants
2.   Hazardous/Toxic Air Pollutants
3.   Radon - Indoor
4.   Indoor Air Pollutants Other Than Radon
5.   Radiation Other  Than Indoor Radon
6.   Industrial Point Source Discharges to Surface Water and Air
7.   POTW Discharges to Surface Water and Air
8.   Nonpoint Source Discharges to Surface Water and Air
9.   Management of Hazardous Waste at RCRA Facilities
10.  Hazardous Substances at CERCLA  Sites
11.  Solid Waste Management
12.  Releases from Underground Storage Tanks
13.  Other Ground-Water Contamination
14.  Other Pesticide Contamination
15.  Physical Modification of Aquatic Habitats
16.  Physical Modification of Environmentally Sensitive Terrestrial  Habitats
17.  Acid Deposition
18.  Operation and Maintenance of Water Supply Facilities

-------
Ground Rules


        Region III adopted important  ground rules that affect the interpretation  of the results  of the
comparative risk analysis.


Focus on Risks


        To measure the relative  seriousness of environmental problems, the project  organizers focused on
risks.  Risk measures capture the  ultimate  impacts of the  problems.  For example, scrubbers are not
installed on electric power plants to reduce emissions to meet  air quality standards; rather they are installed
to protect people  from getting sick or lakes  from acidification.  Pollution control actions are designed to
eliminate the ultimate impact of environmental problems on humans or the ecology of an area. Region Ill's
comparative risk  analysis focused on three  basic types of risks: human health  effects, ecological effects,
and welfare  effects.
                                           THREE TYPES OF RISK


   Health risk - Cases of human disease or injury caused by the environmental problem.

        The health  effects considered ranged from cancer (e.g.,  lung cancer from indoor radon)  to  learning
   disabilities (from airborne lead) to gastrointestinal disease (from pathogens in drinking water) to angina
   pain  (from carbon monoxide)  to numerous other non-cancer effects.

   Ecological risk  -  Damage  to the structure and function of natural ecosysteas fro* the enviroraaental
   problem.

        Some examples of  the effects considered include: eutrophication of water bodies from nutrients in
   nonpoint  source  runoff;  loss  of  species'  range,  breeding grounds,  and  other  effects  from  physical
   modification of habitat; and reduced growth rate of trees and increased susceptibility to pests in forests
   exposed to high levels of  ozone.

   Welfare risk - Economic losses to hunan activities caused by the environmental problem.

        Examples  include increased maintenance expenses  for  paints  and other materials  exposed  to acid
   deposition;  reduced  recreational  use of water  bodies  polluted  by  industrial  dischargers; costs of
   replacing or treating drinking water  supplies contaminated by hazardous waste site leachate; and costs
   of  treatment and lost productivity for  individuals suffering adverse health effects from environmental
   pollutants.  Also  included under welfare risks are intangible damages, such  as the adverse effects of
   odors or reduced visibility  associated with  air pollution.  Only  the primary or direct welfare damages
   were  counted.  Market manifestations  of welfare  damages (e.g., reduced  real  estate prices  in  polluted
   areas) were thought to provide possible measures for welfare damages, but were not considered themselves
   to be damages.
                                                   10

-------
 Consider Only Residual Risks

        The project assessed the "residual risks" associated with each problem area.  Residual risks are the
 impacts that remain, given the current levels of controls and non-compliance with regulatory requirements.
 The project did not assess risks in the absence of control actions or after current requirements are met.
 This ground rule of considering only residual risks forced the Region to evaluate the existing environmental
 situation. It also affects the policy implications one might draw from the risk ranking:

        o  A  problem area might  rank low because the risks it presents are: (a)  inherently low, or (b)
           inherently high, but a control program has successfully reduced them.

        o  The risk rankings provide guidance for allocation of new incremental resources among programs,
           but no guidance for reallocating existing resources.  Residual risks  provide a guide to  problem
           areas most in  need of further efforts.  They do not provide an indication of how much risks
           would increase in a problem area if present programs were dismantled or reduced.

        o  A problem area can appear to present high residual risks currently, even though compliance with
           existing laws and regulations will lower the risks  considerably.

 Separate Risk Assessment from Risk Management

        The third project ground rule was  to consider only the risks associated with the problem areas.
 Other attributes of the problems that are important in determining how to address them  -- such as the cost
 or technical feasibility of dealing with them,  or public  opinion about them, for  example  -- were  not
 considered in the ranking  process and were reserved for consideration later in  the project. Assessment of
 the seriousness of each environmental problem (risk assessment) was carefully separated from evaluation
 of what could  be done about each problem (risk management).  The risk rankings alone do not represent
 the Region's priorities for action.   Setting  priorities  involves balancing the severity  of each problem with
judgments about how effectively it can be addressed.

 Use Both Quantitative Data and Expert Judgment

        The project organizers wanted the risk rankings to be as objective as possible.   There had been
several other instances, for other purposes, in which  regional management had listed which environmental
problems they  considered most serious in  the Region.   These lists, though, were  based upon  opinion,
judgment, and  little analysis. The Comparative Risk Project was unique in that it sought  to generate a risk

                                                11

-------
ranking in a systematic, objective, and data-driven way. The project participants aimed to generate and use
as much quantitative data on risks as possible.  But there were inevitable constraints:

        o  The work groups had limited time, staff, and budget for the analysis.  Full risk analysis, in which
           data on emissions and ambient concentrations are collected, exposures are modeled, and ultimate
           impacts are projected, is  very costly and time-consuming.

        o  Even if resources had been unlimited, risk analysis is an inherently uncertain process.  Subjective
           interpretation of the results of any risk analysis is always necessary, weighing the strength of the
           data base used and the validity of the assumptions made.

In short, the project organizers realized that the  rankings could riot be completely objective.  The  project
staff used their own  expert judgment in instances when they could not obtain data or perform analyses.
Their approach was to quantify results to the extent the available data and time allowed, to recognize the
universal need to supplement the data with judgment, to make the judgments in an objective manner, and
to address significant data gaps through future refinement of  the estimates and assessments.
        To stretch regional
resources, the project staff
relied  extensively on risk
analyses that had  already
been   done   for   other
purposes   prior  to  this
project, and adjusted them
to   suit    their    needs.
Unfinished  Business  was
An example occurred  in  estimating the health risks associated  with  the
toxic air  pollutants problem area. A  contractor modeled the health risks
in Region  III from 22 toxic air pollutants.   But  there are far more than
22 toxic air pollutants in total.  The health work group had to decide
whether  the 22 modeled pollutants constituted a large or a small fraction
of the total risks associated with the  problem area.  The health work group
relied  on  the   judgment  of  the  regional  air  program  staff  and  the
contractor,  and  information  in Unfinished Business (deriving  from OAQPS
staff judgments)  to make  this decision.   The work group  decided  that
although a few  important  air toxics were omitted from the analysis  and
although synergistic  effects  had not been considered,  the  risks  were
estimated  conservatively for the  22 chemicals,  and  they probably did not
drastically understate the risks for the entire problem area.
particularly  helpful.    In
many cases, the project staff extrapolated from incomplete analyses.

        At times, a work group relied  on analyses that  had been done for olher geographic  areas  and
adjusted them to fit Region III.  Health risks from industrial plants discharging to surface waters had been
estimated for the State  of Pennsylvania; the work group  scaled them up to represent the entire  Region.
In other cases, a  work group  qualitatively adjusted  estimates of different quality in order to make them
more comparable (e.g., where one  analysis was based on highly conservative, worst-case assumptions, while
another analysis was based on more realistic assumptions). And in many cases, there were simply data gaps
that a work group had no choice but to fill by using judgment
                                                  12

-------
        In sum,  the rankings should be viewed  more  as the informed judgment of the  Region  III
professional staff, based on quantitative data to the extent possible, than as the results of a scientific risk
assessment.  No scientific group has  peer reviewed  the  results  or analyses,  because the  results are
fundamentally not scientific. But they are not simply opinion either.  The project participants collected  large
amounts of data and conducted extensive analyses.  They took care  to make their judgments systematically
and objectively.  They developed the rankings by using carefully developed methods for each of the types
of risk.
                                                13

-------
                     III. Ranking Environmental Risks in Region III
        Each work group designed analytical methods, conducted analyses, and ranked the 18 environmental
problems.  In many instances, the analytical method a work group chose substantially influenced the group's
specific results.  For example, the way in which the Health Work Group chose to weigh cancer relative to
non-cancer health effects influenced the way they ranked risks from criteria air pollutants (which are not
thought to be carcinogenic) relative to those from radon (which causes cancer, but few non-cancer effects).
This chapter describes each work group's ranking results and the methods for developing them.
Health Risk Methods

        The health work group's methodology involved estimating four categories of health effects - risks
to highly exposed individuals and aggregate risks to the entire Region III population, for cancer and for non-
cancer effects - for particular chemicals chosen to represent each problem area.  The work group estimated
risks in each of these categories by developing exposure scenarios typical of each problem area.
        An  exposure  scenario  summarized  one
instance  in which humans  could be  exposed to         some exposure scenarios for different problem
                                                    areas included:
pollutants  from  a  problem area.    For  each
                                                     o The  ambient concentrations of 22 toxic air
exposure  scenario,  the  key data  included  the        pollutants  in each Region  in county.
concentrations of contaminants to which individuals      0 The concentrations of contaminants  in drinking
were  exposed,  the duration  of  the exposure, the        water intakes and f ish downstream of each of  12
                                                       sampled industrial dischargers.
number of people exposed,  and an indication of
                                                     o The indoor concentrations of radon, chlordane,
the degree to which the scenario represented the        formaldehyde and other substances in residences
                                                       where sampling  has been conducted.
entire  problem area.  The  representativeness of
each scenario was  part^arty  tapomn,, as ft.      «
work  group and contractor had  to  decide how to  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
scale up from the particular exposure scenarios for
which data were available to  the entire problem area.  The scaling up processes differed substantially  from
one problem area to the next. Sometimes an exposure scenario represented an average case, and most of
the residents in the Region were expected to be exposed similarly to those in the scenario. Sometimes the
scenario  was an extreme case,  and relatively few individuals in the  Region  were likely to be similarly
                                                14

-------
 exposed.  The work group used its judgment in making such projections.

        Once the work group estimated regional exposure, it projected cancer risks (for both the regional
 population and  the maximum exposed individual) by using standard EPA  methods.   The work group
 converted exposure to dose by making standard assumptions about such quantities as  the amount of air
 people  breathe  and the amount of water they drink each day.   The work group used  the latest EPA
 potency estimates for each carcinogen, and multiplied dose by potency to estimate risk.

        Because standard methods do not exist for estimating non-cancer risks, the work group developed
 its own approach, as follows:

        o  For pollutants for which appropriate data were available (for example, criteria air pollutants and
           lead), the work group used epidemiologically derived estimates of potency in combination with
           Region III  data on ambient concentrations and exposed populations in order  to estimate  the
           number  of cases of adverse health effects directly.

        o  For some other effects (illnesses from  microbes, and diseases transmitted by animals such as
           rabies), the number of cases was estimated directly by adjusting upward or downward the number
           of reported cases of such effects.

        o  The  work group estimated  the number  of  cases of non-cancer effects from  the remaining
           pollutants with the  following equation:

            (number of people  exposed at hazard index > 1)  x  (hazard index) x 0.01,

           where hazard index = dose/RfD1.

           This  equation was calculated for each pollutant having a hazard index greater than one for  the
           problem area.  Population risk was obtained by summing across pollutants.

        o  A proxy  for non-cancer individual risk was calculated as the  highest hazard index (dose/RfD)  for
           the problem area.
     l. The RfD, or Reference Dose, for a chemical is an estimate (with uncertainty spanning perhaps an
order of magnitude) of a daily exposure to the human population (including sensitive subgroups) that is
likely to be without appreciable risk of deleterious effects during a lifetime.
                                                15

-------
The work group was aware that both the hazard index approach and  the  formula  for estimating cases
constituted gross oversimplifications  of actual lexicological relationships.  Nevertheless, the work group
believed that such "short cut" proxies for non-cancer risks were necessary,  given the project's time and
budget constraints.

        Work group members and contractors estimated risks for each problem area in this manner.  The
problem-leads described the results of these analyses to the other work group members. The work group
debated and modified these estimates, in many cases adjusting the  estimated risks  for a problem area
upward or downward, based upon professional judgment.  The work group eventually created four separate
rankings of the 18 problems: for population and individual risks, and for cancer and  non-cancer effects.

        The work group then employed various mathematical techniques  to quantitatively combine the four
risk rankings and obtain an aggregate health  risk ranking. The mathematical techniques yielded different
aggregate ranking results.  For example, using one technique where cancer cases were weighted  heavily
relative to  cases of  non-cancer effects resulted  in Indoor Air Pollution ranking as the highest risk and
Criteria Air Pollutants as the fifth highest.  But  using a different technique where cancer  and  non-cancer
cases were weighted equally resulted in Criteria Air Pollutants ranking as the highest risk  and  Indoor Air
Pollution as the  seventh highest.

        The work group members used  these various results and their judgment to reach a  consensus
aggregate ranking of the problem areas.  The consensus ranking did not  exactly match any  of the rankings
obtained by the  various mathematical techniques.

        The work group decided that cancer effects were more important than non-cancer  effects because
non-cancer effects have a wider range of  severity than cancer. Also, the  work group had more confidence
in then" procedures for estimating the number of cancer cases than the number of non-cancer cases.  The
work group members  then decided  to weight population risks much more  heavily than  individual risks
because they judged  population risks more  important in broad, Region-wide  priority-setting decisions.

        The consensus ranking by the work group reflected these  decisions: it most closely resembles a
ranking based on cancer population risks.
                                                16

-------
Health Risk  Ranking Results

        Exhibit 2 below shows the work group's ranking of health risks.  The highest ranked problem areas
all caused substantial numbers of cancers, which were judged by  the work group to be generally the  most
severe of all the environmentally-induced adverse  health effects.

        o  Indoor air pollution was  ranked as  the highest  risk.   Large numbers of excess cancers  were
           projected from chlordane, passive smoking, and formaldehyde, as well as a substantial number
           of non-cancer effects.  Only a few of the multitude  of indoor air pollutants were explicitly
           considered.  The other pollutants would raise the risk estimates far above what was calculated,
           which was already so high that there was  no question about the ranking.

        o  Radon was ranked second.  High lung cancer risks for both populations and individuals  were
           the primary  hazard.

        o  Other pesticide contamination was ranked  third.  This  was largely due to substantial residues of
           carcinogenic pesticides on food and the universal human exposure to these contaminants.

        o  Radiation other than radon was ranked fourth.  Excess cancers caused by natural background
           radiation (about which little can be done) were responsible.

Problem areas involving little cancer risk  but substantial non-cancer risks were ranked below these four
problems with high cancer risks.

        o  Acid  deposition was ranked  seventh. The  health effects of airborne sulfates were defined to be
           included in this problem area.  They have been linked epidemiologically to a range of adverse
           effects, including premature mortality.

        o  Criteria air pollutants were ranked ninth.  This was due to the work group's emphasis on cancer
           and to further discounting most of the criteria  air pollutant effects by a factor of 100 to reflect
           their  particularly low severity.

        o  Problem areas  with exposure primarily through contaminated ground water (Other Ground-
           Water. RCRA Sites. CERCLA Sites. Solid Waste. UST) were ranked low primarily because few
          people are exposed for  long periods to high levels of  contaminants in  ground water from such
          sources.  The work group  estimated  population risks  as low,  though individual risks can be  high.

                                                17

-------
                                        EXHIBIT 2


                                   Health  Risk Ranking

1.  Indoor Air Pollutants Other Than Radon
2.  Indoor Radon
3.  Other Pesticide Contamination
4.  Radiation Other Than Indoor Radon
5.  Nonpoint Source Discharges to  Surface Water and Air
6.  Operation and Maintenance of  Water Supply Facilities
7.  Acid Deposition
8.  POTW Discharges to Surface Water and Air
9.  Criteria Air Pollutants
10. Other Ground-Water Contamination
11. Industrial Point Source Discharges to Surface Water and Air
12. Management  of Hazardous Waste at RCRA Facilities
13. Hazardous Substances at CERCLA (Superfund) Sites
14. Hazardous/Toxic Air Pollutants
15. Solid Waste Management
16. Physical Modification of Environmentally Sensitive Terrestrial Habitats
17. Releases from Underground Storage Tanks (USTs)
18. Physical Modification of Aquatic Habitats
                                              18

-------
        Several major uncertainties are inherent in these health risk rankings.  Perhaps most important, as
previously discussed, are the relative weights in the aggregate ranking given to cancer vs. non-cancer and
population vs. individual risks.  If the work group had chosen different weights, a different  ranking would
have resulted.

        Another uncertainty involves the procedure chosen to estimate non-cancer risks in situations where
exposures were  compared with RfDs.  EPA has  no accepted procedure for risk  estimation  in such cases,
and the work group developed its own mathematical formula.  The work group was aware that its formula
substantially oversimplified the complex nature of dose-response functions for non-cancer effects.   The
formula may have overestimated the number of cases of non-cancer effects for problem areas  (nonpoint
sources, POTWs, and  industrial point sources) where nearly all  the exposures were at levels only slightly
above RfDs.  This was because  substantial safety factors are built into RfDs, and, contrary to the work
group's formula, there is probably only a low risk from exposures only slightly above RfDs.

        Additional uncertainties derive from the data introduced into the work group's methodology.  Data
on the health effects of chemicals, on potencies,  and on epidemiological relationships are uncertain.  The
data on pollutant concentrations  and exposed populations ranged from very good (e.g., a huge number of
measurements of the indoor radon concentrations to which Region III residents are exposed) to poor (e.g.,
extremely uncertain modeling of the concentrations of pollutants to which individuals downgradient of leaking
waste sites or USTs might be exposed).

        Despite these  uncertainties,  the differences in numbers of health effects estimated for the various
problems  were  so substantial that the work group felt confident in its  relative ranking.
Ecological Risk  Methods

        Risk assessment was more difficult for the ecological work group than for the health work group.
Whereas substantial  portions of the methods for assessing health risks have been standardized, there is no
consensus on how to assess  ecological risks.   Ecological systems consist of multiple species, each perhaps
reacting differently to pollution stresses.  Over time, the behavior of the entire system can be quite different
from that of any single portion of it.  Similarly, the  data available  for implementing any ecological risk
assessment method are  more limited than the data relating to health risks.

        The ecological work group began by considering the effects of individual stressors (e.g., acids, metals,
sediment) from problem areas on different types of ecosystems (e.g., coastal wetlands, forests, lakes).  This

                                                19

-------
effort was abandoned because of the difficulty in isolating the effects of a single stressor in environment.
where multiple stressors are present simultaneously.  The work group also devoted substantial effort to
formulating a useful ecosystem typology, with the aim of assessing  the  risks from a  problem area to each
type of ecosystem individually, and then adding across ecosystems.   However, this approach proved to be
too complex.

        Eventually, the ecological  work group devised a  qualitative risk assessment method based on the
experience and opinions of the work group members. Although scored quantitatively, the method relied
fundamentally on qualitative judgments about the intensity and geographic extent of ecological damages from
each problem.  Contractors provided data  to help form  many of the judgments.  The  method was based
on developing scores for four criteria, combined through a mathematical formula:

      Problem area score = ('Severity score + Reversibility score) x (Source score  + Target score)
        o  The severity score characterized the frequency with which the ecosystem was exposed to stressors
           at durations resulting in acute and/or chronic toxicity.

        o  The reversibility score characterized the length of time required for an ecosystem to recover from
           the impacts of the stressors.

        o  The source score depended on whether the sources of contamination associated with the problem
           area  were widespread or local, and whether the loadings were predictable or not.

        o  The  target score  depended on whether impacts were  local or widespread in one or  several
           ecosystems, and whether environmentally sensitive/unique ecosystems were involved.

        Prior to  the ranking meeting, each ecological work group issue lead reviewed the available data and
developed tentative scores for the four criteria. The entire work group discussed the data and the proposed
scores for each of the four criteria, and reached a consensus on scores for each problem.  The total score
for a  problem was the sum of its criteria scores, and the problems were ordinally ranked on the basis of
their total scores. The work group members reviewed the ordinal ranking they generated by relying directly
on the scores, and decided not to change  it.  They considered collapsing the numerical ranking of problems
into risk groupings instead (putting problems into high, medium, or  low ecological risk groups), but decided
to leave their 1  - 18 ranking of the problem areas.   Several problem areas were  left tied in the ranking.
                                                20

-------
Ecological Risk Ranking Results

        The resulting ecological risk ranking is shown below in Exhibit 3.   Some of the highlights of the
ranking follow:

        o Physical modification of terrestrial habitats and physical modification  of aquatic habitats were
          ranked as tied  for  causing  the  greatest ecological risk.  They cause extremely widespread,
          devastating, and typically irreversible impacts.  They can damage not only the specific location
          where they occur, but also nearby areas by changing water flow patterns, disrupting migration
          pathways, altering food webs, etc.

        o Nonpoint source discharges  to surface waters were ranked as causing the  third highest risks.
          These discharges are extremely widespread and can render some areas virtually lifeless (as in
          some streams affected by abandoned mine drainage). The states' 305(b) water quality assessment
          reports show nonpoint sources as by far the leading cause of degradation in  streams, lakes, and
          estuaries.

        o Acid  deposition was ranked  as causing  the fourth greatest risk.  The work group  based this
          ranking largely on aquatic effects, judging effects on forests to  be unproven  in Region  HI.
          Adverse impacts on lakes, headwaters streams, wetlands and even Chesapeake Bay (substantial
          air deposition of nitrates) were identified.  Acid deposition could be ranked  as even higher risk
          if effects studies become more conclusive, as precipitation in Region III is the most acidic in the
          nation.

        o CERCLA  sites  were  ranked fifth.  Ecological  investigations  for  Superfund  sites revealed
          substantial  impacts  occurring off-site,  frequently extending into  receiving  water  bodies.
          Extrapolating from several case studies to  the entire universe  of potential NPL  sites, the work
          group estimated significant ecological damage in sensitive environments.

       o Criteria air pollutants and air toxics were each ranked sixth, with great uncertainty attached to
          this ranking.  Data on the actual ecological effects associated with these problems were minimal.
          The problems were ranked moderately high because they involved widespread pollution and there
          was a presumption of adverse impacts.
                                                21

-------
o  Five problem areas were ranked at the bottom as causing no or minimal ecological risks: indoor
   radon, indoor air pollution, other ground-water contamination, other pesticide contamination, and
   operation and maintenance of water supply facilities.  Note that ground-water contamination and
   pesticides may cause significant ecological damage.  However,  the definitions of these problem
   areas  adopted for the purposes of the Comparative Risk Project excluded the routes by which
   they cause such damage.  Ecological impacts from ground-water contamination were limited by
   definition because the impacts of contaminated ground water on surface water were assigned to
   the nonpoint sources problem area.  The pesticides problem was  defined to include only risks
   through residues on food and  risks  to applicators.   Risks to  non-target species through non-
   food routes (e.g., surface runoff, dermal absorption) were excluded.
                                        22

-------
                                         EXHIBIT 3
                                  Ecological Risk Ranking
1.  Physical Modification of Environmentally Sensitive Terrestrial Habitats
1.  Physical Modification of Aquatic Habitats
3.  Nonpoint Source Discharges to Surface Water and Air
4.  Acid Deposition
5.  Hazardous Substances at CERCLA (Superfund) Sites
6.  Criteria Air Pollutants
6.  Hazardous/Toxic Air Pollutants
8.  Releases from Underground Storage Tanks
9.  Industrial Point Source Discharges to Surface Water and Air
9.  Radiation Other than Indoor Radon
9.  POTW Discharges to Surface  Water and Air
12. Management of Hazardous Waste at RCRA Facilities
13. Solid Waste Management
14. Indoor Radon
14. Indoor Air Pollutants Other than Radon
14. Other Ground-Water Contamination
14. Other Pesticide Contamination
14. Operation and Maintenance of Water Supply Facilities

       Note - the  final five problems were believed to  cause no ecological risk and were ranked as tied
for lowest risk.
                                              23

-------
        The work group noted substantial uncertainty in its ranking results. Ecological risk assessment, in
the absence of standard methods, was a very difficult task.  The  work group had particular difficulty in
deciding how to weigh disparate impacts in arriving at a single judgment about ecological risk - for example
in comparing a problem causing substantial impacts in relatively few locations (e.g., CERCLA sites) with
one causing moderate impacts to large areas (e.g., criteria air pollutants).  Another difficulty came from the
apparent need to weigh  different ecosystem types against each other ~ for example, to decide whether a
particular impact in a wetland caused more damage than a different impact in a forest.

        The work group also noted that the project's ground rules  seemed to facilitate health risk analysis
more than ecological risk analysis.  The directive to assess current residual risks seemed to ignore the long-
term, cumulative ecological impacts that many environmental problems will cause. The list of problem areas
did not incorporate ecologically meaningful distinctions. Several problem areas had no ecological significance
(e.g.,  indoor air pollution, drinking water), several were defined in an artificial way that excluded ecologically
important effects (e.g., the pesticides problem area  excluded the routes by which most ecological damage
occurs from pesticides), and some lumped together extremely different ecological impacts (e.g., the nonpoinl
sources problem area  included both  acid  mine drainage and eutrophication of  lakes  from excessive
agricultural nutrient runoff).

        In sum, the ecological work group members did not feel very confident in their results.  They
believed that the  three  highest ranked problem areas (physical  modification of aquatic and  terrestrial
habitats, and nonpoint sources) clearly caused  more damage than the other problems, and that  the  five
problem areas ranked at the bottom caused lesser risks than the others.  However, they had little confidence
in the accuracy of the rankings for  the other ten problems.
Welfare Risk Methods

        The welfare work  group had a major advantage over the other work groups: there was broad
agreement on using the dollar value of damages as the common denominator for measuring welfare risks.
The work group agreed to  estimate the annual dollar losses caused by each environmental problem as the
key factor on which they would base their ranking. The major methodological issues the work group dealt
with were the  definition of  the boundaries of welfare risk, and secondary factors other than dollar damages
that should enter into the ranking process.

        The work group decided  to consider nine categories of welfare effects.
                                                24

-------
                                CATEGORIES OF WELFARE EFFECTS AND EXAMPLES
   Effect on agriculture and livestock:  Reduced crop yields due to vegetation damages from ozone.
   Effect on forestry:  Reduced commercial forestry yields due to reduced growth in trees affected by ozone.
   Effect on water suitability:   The costs of treating or replacing contaminated water supplies.
   Effect on coMercial fishing:  Reduced profits to the commercial  fishing industry from decreases in
   finfish and shellfish populations due to environmental pollution.
   Effect on re-use of land:  Economic losses resulting  from  reduced options in use of  land  following
   environmental contamination, such as at a CERCLA site with contaminated soil.
   Effect on Materials (Manufacturing,  residential,  commercial):   Accelerated  deterioration of  outdoor
   masonry,  statues, etc,, exposed  to airborne acids.
   Effect on recreational opportunities:  Reduced fishing, boating, and swimming opportunities  in  surface
   waters affected by pollution.
   Effect on aesthetic value: Offensive odors from diesel  exhaust.
   Effect on health:  Cost of treating lung cancer due to exposure  to indoor  radon.
        The economic costs of environmentally induced diseases were included as welfare damages.  These
included both the medical costs of treating the disease, and the loss of economic productivity incurred by
someone with the disease.   There was debate within the work group over this  approach.  Some felt that
health effects would be double-counted if they were considered by both the health and the welfare work
groups.  The results reveal that inclusion of health costs ultimately made some, but not a major, difference
in the welfare  rankings.  Health costs contributed less to total  welfare damages than did  some  other
categories of damages. The work group estimated health costs by attaching standard "cost per case" values
for different sorts of health effects to the health work group's estimates of the  numbers of health effects
for each problem area.

        There was also some debate over including aesthetic effects as  a  category of welfare  damages.
There was agreement that effects such as reduced visibility, odors, and poor tasting water were legitimate
damages, but there  was great concern about how they could accurately be measured.  Studies monetizing
such damages were  extremely rare, and often came  to debatable conclusions. The work group decided to
include aesthetic damage estimates, but to perform sensitivity analysis on the ranking for any problem areas
where they were significant.

        The  work group took care to distinguish welfare damages from control costs, and to exclude control
costs from the analysis. In general, the work group considered control costs to be costs incurred to prevent

                                                  25

-------
 generation of pollution, while welfare damages were costs that occurred subsequent to the pollution having
 been generated.  The distinction became difficult in such areas as the costs of mitigating household radon
 and the costs of remediating CERCLA sites.  In such cases the work group adopted another rule.  To the
 extent a polluter was legally required to mitigate his  pollution, his expenses of doing so were counted as
 a control cost To the extent the mitigation expenses were incurred voluntarily by a victim of the pollution
 to reduce his damage, they were  counted as a welfare damage.  Radon mitigation expenditures were thus
 counted as a welfare damage (but to the extent they were assumed to occur, they reduced the medical costs
 due to radon's health effects).  The costs of remedial action at CERCLA sites were split:  source control
 and ground-water  remediation were  counted  as  control  costs, while  costs of treating  or  replacing
 contaminated drinking water supplies were counted as welfare damages.

        The work group also decided to reflect in its ranking scheme several factors in addition to the total
 annual dollar estimate for  the welfare damages.  These factors were:

        o  The geographic extent of the welfare damage;

        o  The impact to individuals from the welfare  damage (in effect, considering individual welfare risk
           in addition to population welfare risk);  and

        o  The reversibility of the welfare damage.

 The work group developed a mathematical formula to  combine the dollar magnitude of damages with these
 qualitative adjustment factors.  The formula was designed to weight the dollar losses more heavily than the
 qualitative factors.

       At the work group's ranking meeting, the problem-leads reviewed the results of contractor analyses
 with the  work group and proposed scores for  the  problems.  The work group then developed consensus
 scores for each category of welfare damage caused  by each problem area  (e.g., damage to materials,
 recreation, aesthetics).  Using the formula, the  group then summed the scores across damage categories to
 obtain a  total score for each problem area.

       Based upon the scores, the welfare work  group ranked the 18 problems ordinally, and also grouped
 the 18 problems into high, medium and low risk categories.  The  work group members believed that there
were major  uncertainties in monetizing welfare damages, and that the full ordinal ranking might overstate
 the degree  of confidence in their results.  They  felt more comfortable in portraying  their findings in less
 precise terms.  The scores for  the problem areas fell into three sharply different ranges,  and the three

                                                26

-------
ranges provided the basis for the high, medium and low risk categories.  The work group felt very confident
that there were substantial differences in risk between problems in different groups, but much less confident
in risk differences within a group.

        After developing the two rankings, the work group considered modifying them.   They performed
several rough sensitivity analyses, investigating what would happen if they:

        o  Discounted some of the more speculative damage estimates for air pollution effects (for example,
           visibility damages, and acid deposition damages to forests); and

        o  Combined the damages  from all problem areas affecting ground water into  a single problem
           covering all ground-water contamination.
Welfare  Risk Ranking Results

        Exhibit 4 shows the resulting welfare risk ranking.  The following are some of the welfare risk
ranking highlights.

        Criteria air pollutants and acid deposition comprised the high welfare risk category.

        o  Criteria air pollutants were ranked as the highest welfare risk.  The estimated annual damages
           were about $700 million to materials, about $200 million to crops, and about $100 million to
           forests.

        o  Acid deposition was ranked as the second highest welfare risk.  Annually it caused health care
           costs of $220 million $190  million in damage to materials.  Speculative estimates of annual
           visibility losses and forestry damage were put at $550 million and $105 million, respectively. Acid
           deposition ranked solidly as the second-highest welfare risk, even if visibility and forestry damages
           were completely discounted.

Each  of seven problems appeared to cause  moderately high welfare damages;  clearly  lower than those
caused by the first two ranked problems but still substantial  The work group placed these problems in the
medium welfare risk category.
                                                27

-------
        o  Nonpoint source discharges ranked third. Total annual damages were estimated at $265 million,
           consisting largely of recreational losses ($150 million), aesthetic damages ($70 million) and effects
           on water suitability ($30 million).

        o  Operation and maintenance of water supply facilities ranked fourth, with annual damages of $150
           million from the effects of corrosive water on materials, and $65 million from the health effects
           of lead leached from pipes and trihalomethanes from the process of disinfecting drinking water.

        o  Indoor air pollutants other than radon ranked fifth.  Annual damages were  estimated at $130
           million from health effects, and $20 million in mitigation costs.  Assumptions were made about
           how many households would be likely to mitigate indoor air pollution, thereby reducing health
           costs but incurring mitigation costs.

        o  Radon was ranked sixth.  It was analyzed similarly as indoor air pollutants, and similar results
           were obtained: annual health damages of $130 million and mitigation costs  of $20 million.  It
           was ranked lower than indoor air pollution because of the qualitative factors used by the work
           group to adjust the dollar damage estimates.

        o  Other problem areas  with moderately high annual welfare damages included other pesticide
           contamination f$96 million annually in health costs),  POTW  discharges ($88 million annually.
           mostly in recreational damages), and radiation other than radon ($77 million annually in health
           costs, nearly all from natural background radiation').

The remaining problem areas were  ranked as causing low welfare  risks.  Each was estimated to cause
annual damages of $40 million or less.

        Five problem areas (USTs, RCRA sites, CERCLA sites, solid waste management and other ground-
water contamination) caused damages primarily through ground water, primarily involving contamination of
current and  potential future water  supplies.   The  work  group  believed  that splitting  ground-water
contamination into five smaller pieces biased its ranking downward.  If all the  ground-water damages were
combined into a single ground-water contamination problem  area, it would have ranked about seventh in
terms of welfare  risk,  with annual damages of $116 million.  USTs would have been the  largest single
contributor.  Damages via private wells would substantially exceed  those in public water supplies.   In
addition, damages to  currently  used ground-water supplies would slightly  exceed the estimated value of
damages to currently unused  supplies that might be needed in the future (option demand).
                                                28

-------
                                        EXHIBIT 4

                                  Welfare  Risk Ranking
       HIGHER RISK PROBLEMS:

1.  Criteria Air Pollutants
2.  Acid Deposition


       MEDIUM RISK PROBLEMS:

3.  Nonpoint Source Discharges to Surface Water and Air
4.  Operation and Maintenance of Water Supply Facilities
5.  Indoor Air Pollutants Other Than Radon
6.  Indoor Radon
7.  Other Pesticide Contamination
8.  POTW Discharges to Surface Water and Air
9.  Radiation  Other Than Indoor Radon


       LOWER  RISK PROBLEMS:

10. Releases From Underground Storage Tanks
11. Industrial Point Source Discharges to Surface Water and Air
12. Other Ground-Water Contamination
13. Management of Hazardous Waste at RCRA Facilities
14. Physical Modification of Environmentally Sensitive Terrestrial Habitats
15. Solid Waste Management
16. Physical Modification of Aquatic Habitats
17. Hazardous Substances at CERCLA (Superfund) Sites
17. Hazardous/Toxic Air Pollutants
                                             29

-------
       The welfare work group  members felt  that they were on  solid ground with their  ranking
methodology. However, they were uncertain about estimating the dollar magnitude of the different welfare
effects because of the  limited  data and studies available.  Monetization of environmental damages is a
specialized field of economics.  The work group and the supporting contractors were very dependent on the
few existing studies in this area. In many instances, the work group had  to base its estimate of a particular
welfare damage on a single study that had only limited applicability.  In sum, the major uncertainty in the
welfare work group's ranking derived  primarily  from the lack of data, rather than from assumptions made
in the methodology.
Comparison of Rankings Across Types of Risk

       Exhibit 5 compares the three work groups' rankings.  Exhibit 6 displays the same information slightly
differently; it shows for each problem area the rankings given to it by the three work groups.
                                               30

-------
             a?
            45

            1

                5 I I S I ! S S
                o oo o oooooooooooooo
                            1
I
I
1
8
o
18
                      I
                o o o o oooooooooooooo
                     I
                88l8*°3efi



                ooo o oooooooooooooo
               HIGHER RISK       LOWER RISK
                       31

-------
                                       EXHIBIT 6
                             Ranks for  Each Problem Area
        Problem Areas
1.   Criteria Air Pollutants
2.   Air Toxics
3.   Indoor Radon
4.   Indoor Air
5.   Radiation
6.   Industrial Point Sources
7.   POTWs
8.   Nonpoint Sources
9.   RCRA Sites
10.  CERCLA Sites
11.  Solid Waste Management
12.  USTs
13.  Other Ground-Water Contamination
14.  Other Pesticide Contamination
15.  Aquatic Habitat Modification
16.  Terrestrial Habitat Modification
17.  Acid Deposition
18.  O&M of  Water Supply Facilities
Health
Rank
9
14
2
1
4
11
8
5
12
13
15
17
10
3
18
16
7
6
Ecological
Rank
6
6
14
14
9
9
9
3
12
5
13
8
14
14
1
1
7
14
Welfare
Rank
1
17
6
5
9
11
8
3
13
17
15
10
12
7
16
14
2
4
                                             32

-------
 The Steering Committee decided not to combine these rankings to produce an overall ranking aggregating
 the three types of risk.  Nevertheless,  looking across the three separate rankings, several conclusions are
 evident.

        The degree  of risk posed by an environmental problem seems to depend critically on the type  of
 risk.  None of the problem areas ranked first or second by one of the  three work groups is ranked first  or
 second by another work group.  None of the problems ranked last or next-to-last by one of the work groups
 is ranked similarly by  another work group.

        However, some problem areas do rank consistently above or belo^ the  median  across the  three
 types of risk.  The consistently higher-risk problems  include:

        o  Nonpoint source discharges (ranked #5, #3 and #3 for health, ecological and welfare risks);

        o  Acid deposition (ranked #1, #4 and #2); and

        o  Criteria air pollutants (ranked  #9, #6 and #1).

 Problem areas that rank consistently below the median for each type  of risk include:

        o  Solid waste management (ranked  #15, #13 and #15);

        o  RCRA sites (ranked #12, #12 and #13); and

        o  Other ground-water contamination (ranked #10, #14 and #12).

        Several problem areas rank very high in health risks, but very low in ecological risks.  These include
 indoor air  pollutants other than  radon, indoor radon, other pesticide contamination, and operation and
 maintenance of water supply facilities.  (But note that the low ecological risk  ranking for other pesticide
 contamination is a function of its restricted definition.) Other problem areas rank very high in ecological
 risks, but very low in health risks.  They include aquatic and terrestrial habitat modification, CERCLA sites,
 and air toxics.

        The ranking  disparities between welfare risks  and health or ecological risks do not seem as extreme
 as between health and ecological risks. Thirteen of the 18 problem areas received welfare risk rankings that
were tied with or between their health and ecological risk  rankings.  In effect, welfare risk seems to be

                                                33

-------
combining or mediating between health and ecological risks.  This should not be surprising, since welfare
damages include monetized estimates of both human health and ecological effects.
The Effect of Problem Definitions on the Rankings

        Several of the problem areas received rankings that differ from what might be expected.  This was
at least partly because the Region defined these problems in a mutually exclusive fashion, which contrasts
with the popular conception of what their titles imply.  However, EPA programs are not mutually exclusive.
They overlap; several of them can address the same risk.  Particular attention should  be  paid to  the
definitions of the following problem areas.

        1. Criteria Air Pollutants.  Criteria Air  Pollutants were defined to exclude ambient sulfates, which
were included instead in Acid Deposition. Because sulfates cause substantial health and welfare (visibility)
damages, Criteria Air Pollutants would rank higher if they were  included.

        2. Acid Deposition. The health  and visibility effects of airborne sulfates were  responsible for all
of Acid Deposition's health ranking.  Excluding visibility from Acid Deposition would still leave it ranked
as very high for welfare damages.

        3. Other  Radiation.  Natural background radiation was virtually completely responsible for  this
problem's health and welfare rankings. Presumably, the Region can do nothing about natural background
radiation.

        4. Nonpoint Sources. The Region defined this problem to include releases from contaminated in-
place sediments.  It thus included health  risks from eating fish that have  bioaccumulated PCBs and heavy
metals.  The Region might instead have attributed these risks to  the problem areas that contaminated the
sediments in the first place -- generally either industrial or municipal point sources.  But the Region judged
that  most of the  current sediment contamination is due to point source discharges long ago, and that
defining  risks  from sediments  as  within the scope  of the point source problem  areas  would  thus be
inconsistent with the objective of assessing the risks associated with current  levels of pollution from each
problem area.  Releases from sediments were probably responsible for roughly half of the estimated health
risks from Nonpoint Source Discharges.

        5. Industrial Point Source and  POTW Discharges.  Health risks from these  problems were
correspondingly lower because sediment releases were  included in Nonpoint Sources.

                                                34

-------
        6. Operation and Maintenance of Water Supply Facilities. Because this problem area included only
drinking water  risks attributable  to  contaminants not covered under other problem areas,  it differed
substantially from the purview  of the  drinking water program.  It effectively  included  only  risks from
trihalomethanes from  disinfection of water supplies and lead  leached  from water distribution systems.
Perhaps surprisingly, though, these pollutants were  found to cause the bulk of the health  risks from
contaminants in drinking water.  An expanded definition of the problem area to cover the entire drinking
water area would  not increase its  health or welfare rankings very much.

        7. Other Ground-Water Contamination.  Like the above problem area, this problem included only
ground-water risks attributable to  contaminants not covered under other problem areas.   It,  too, differed
substantially from  the purview of the ground-water program.   This problem effectively included only risks
from nitrates and  pesticides (both from agricultural leaching) and microbial contaminants in private wells
(from septic tanks).  Several aspects of this definition are notable:

        o  The  contaminants  included in this problem area  were  responsible for a moderate fraction --
           probably more than half -  of all health risks arising from contaminated ground water.

        o  In the  ecological risk area,  surface discharge of contaminated ground water was counted under
           Nonpoint Sources.  While this did not significantly raise  the  Nonpoint Sources ecological risk
           ranking,  it  did  exclude from the definition of Other Ground-Water Contamination  the only
           significant means by which ground-water contamination could result in ecological damage.

        o  The welfare work group did a separate analysis, placing the annual damages  from all ground-
           water contamination at $116 million.   This would probably rank about seventh in terms of
           welfare risk if it were  defined as an entire problem area.  As it was defined, Other  Ground-
           Water  Contamination accounted for only  $30 million of this total.

        o  When estimating risks from Other Ground-Water Contamination, both the health and the welfare
           work groups made efforts to include damages occurring in private wells in addition to those in
           public water supplies.

        8. Other Pesticide Contamination.  This problem area included only risks to pesticide applicators,
farm workers, and humans from residues on  food.  It thus did not cover all varieties of pesticide risks.
Most important, its ecological risks were minimal, because most pesticide risks  to plant and animal life,
which are probably quite substantial, were covered under Nonpoint Sources  and  Air  Toxics.  Health risks

                                                35

-------
via  consumption of fish that have  bioaccumulated pesticides were covered under Nonpoint Sources, and
those via consumption of drinking water were covered also under Nonpoint Sources or under Other Ground-
Water Contamination.
                                              36

-------
                                       IV. Observations
Comparison with Unfinished  Business Findings

        The problem areas that were ranked in the Region III Comparative Risk Project and the methods
for and results of doing so were generally similar to those  in the EPA Headquarters Unfinished Business
project.  Comparing the results is straightforward, with the  exception of two items:  the Region III  project
used slightly different definitions for  some problem areas  than did Unfinished Business, and Unfinished
Business did not combine  their separate health rankings for cancer and non-cancer effects.

Health Risk

        Both projects rated radon, indoor air pollution, and  pesticides among the highest health risks.
Unfinished Business also ranked worker exposures and exposures to consumer products as causing very high
health risks, but the Region III project did not consider these problem areas. Drinking water contamination
ranked quite high in health risks for both studies.

        At the low end of the  health  risk ranking, both projects listed USTs, non-hazardous waste,  RCRA
sites, and, somewhat higher in Unfinished Business. CERCLA sites.  It is evident that the important factor
causing these problem areas to be ranked as  low risk in both studies was the limited population exposures
typically associated with contaminated ground water.   Ground water as a medium differs from air or  surface
water.   Pollution of ground water tends to be slow-moving and  localized, and exposure to  contaminated
ground  water can  usually be avoided at modest cost by obtaining  alternate  or treated water supplies.
Although the health risk for a person exposed to contaminated ground water may  be as high as that for
someone exposed to polluted air or surface water, far fewer people are exposed to contaminated  ground
water.  The health ranking methods used in  both the Region III  project and  Unfinished Business  tended
to weight population risk far more than they did individual risk.

        A few health risk  rankings differed between the Region  III  project and Unfinished  Business:

        o   Region III ranked the three surface water problem areas (nonpoint source discharges, industrial
           point sources, and  POTWs) all as somewhat  higher  risk than  did  Unfinished Business.  The
           discrepancy was particularly great for nonpoint sources, which the Region ranked as  #5 for
                                               37

-------
           health risk. Headquarters ranked nonpoint sources as #20 of 31 for cancer risks and as medium
           for non-cancer risks.

               This difference in ranking may represent a real difference in risk between Region III and
           the entire  nation.  Region III is more densely populated, and has a large concentration of older
           industrial and POTW dischargers, and a unique abandoned mine drainage problem.   Or,  the
           ranking difference may be because the Region III health work group obtained data on toxic
           chemicals  in edible fish tissue, and  performed several  modeling analyses to  project human
           exposures to contaminants in surface water via fish consumption and drinking water.  Unfinished
           Business used professional judgment, with little data.  However, the Region III work group's
           analyses in this area used highly conservative assumptions, and may thus have overstated risks.

        o  The air toxics ranking in Unfinished Business  showed substantially higher health risks than  the
           Region HI ranking.  The disparity turns on  different judgments about the degree to which  the
           specific air toxics analyzed represent  the entire  universe  of  air toxics, and the extent to which
           air toxics cause non-cancer effects.

Ecological Risk

        The ecological rankings were also quite similar across the studies.  Physical modification of habitats,
nonpoint sources, and criteria air pollutants/acid deposition were ranked as posing high risks.   Unfinished
Business ranked as even higher risks global warming and ozone depletion, but the Region did not consider
them.  Both studies found radon, indoor air pollution, drinking water, solid waste disposal, other  ground-
water contamination, and RCRA sites  to present the lowest  ecological risks.  Several differences in  the
ecological rankings included  the  following:

        o  Unfinished  Business found the risks posed by point sources (both industrial  and POTW) and
           nonpoint sources  to be similar. In contrast,  the  Region ranked nonpoint sources clearly higher
           and industrial point sources and POTWs medium to low.   The  ecological work group used data
           that Unfinished Business paid less attention to on the  number of stream miles and lake acres
           degraded by the different sources of water pollution. The data showed nonpoint sources to be
           by far the greatest source of degradation.

        o  The Region ranked CERCLA sites as high risk, while Unfinished Business ranked them as low
           risk.  The  Region ranked USTs and other radiation as medium risk while Unfinished Business
           found  them  to be low risk.   These  ranking  differences  probably reflect the high density of

                                                38

-------
           CERCLA sites and USTs in Region III, and differences in ecological scoring methods between
           the two projects.  The Region III method tended to attribute higher  risks to source types that
           are particularly widespread,  and that release highly toxic substances  unpredictably.  In effect,
           there was a  large element of evaluating potential risks in  the Region's approach.  Unfinished
           Business, by contrast, based  its ecological rankings more on the actual record  of observed
           ecological damages attributable to each problem. In any case, such differences in judgment might
           be expected  on problems like these for which data on ecological effects are extremely limited.

Welfare Risk

       At first glance,  Region Ill's welfare risk rankings match those from Unfinished Business less well
than do the health and ecological risk rankings.  Several problem areas that the Region ranks high in terms
of welfare risks (indoor air pollution, indoor radon, other pesticide contamination, and radiation) are  ranked
very low  by Unfinished  Business. This is because Region III counted health care costs in its welfare risk
estimation  and Unfinished Business did not, and each  of these problems  causes high health care costs.
Otherwise, the welfare rankings are generally similar.  In both studies the greatest damages were attributed
to criteria air pollutants, with the next highest damages caused by  nonpoint sources.  The rankings diverge
somewhat in  two respects:

       o  As in the ecological rankings, Unfinished Business places industrial point sources and POTWs
           in  the high risk category, while  the Region categorizes them as  medium.  Again, the probable
           reason  is that the Region attributed a lower share of the damages occurring via surface water
           to point sources  (a  higher share to nonpoint sources)  than did  Unfinished  Business based on
           Regional 305(b)  report data.

       o  The Region's relative ranking of the ground-water and waste problem areas was exactly opposite
           to that of Unfinished Business.  The Region ranked USTs and other ground-water contamination
           as medium risks, with RCRA  sites, solid waste  disposal, and CERCLA sites as  low risks.
           Unfinished Business ranked them in an inverse order. The Region III welfare work group's data
           and procedures for considering ground-water damages were substantially more sophisticated than
           those of Unfinished Business. The work group quantitatively considered impacts on private wells
           and impacts  to currently unused ground water, and used a  large regional data base to allocate
           damages among categories of sources responsible for the damages.  Unfinished Business did not
           use such data.
                                                39

-------
Reasons For Similarities To National Risk Rankings

        There is a substantial similarity between the Region III results and those from Unfinished Business.
The rankings agree far more often than they disagree, and many of the divergences seem explainable by
definitional or methodological variations.  Ranking differences that reflect real risk differences -- where a
problem is substantially more  or less risky in Region  III  than in the nation as a whole --  seem rare.

        There are several cautions against drawing such conclusions, however. The first is  that the  Region
III and Unfinished Business projects used substantially similar analytical methods and data. Some individuals
from Headquarters and among the consultants participated in both projects.  It is possible  that the similar
rankings resulted from the common methods, assumptions, and judgments,  and not necessarily  from the
similarity of environmental risks.

        Another  point to note is that the Region and Headquarters both  ranked risks  over large
geographical areas that are each large aggregations of disparate elements of risk.   If the Region III project
and  Unfinished Business were to have ranked  finer elements, the projects might have found much more
extensive evidence of geographic differences in risk.   Following are two examples of this point:

        o  The nation as a whole  and Region III both encompass diverse land uses.  Each has major
           cities,  industrial areas,  agricultural areas,  forest lands, etc..   If such  studies were to  analyze
           problems in smaller and more homogeneously defined geographic areas, they would undoubtedly
           uncover larger distinctions in risk.  A ranking of environmental risks in an agricultural area (e.g.,
           Sussex County, DE) would be much different from one for an urbanized area (e.g., Philadelphia).

        o  Similarly, risk rankings would begin  to diverge if the problem areas were more finery defined.
           For example, while nonpoint sources may cause similarly high ecological risks in Region III  and
           the entire nation, the rankings would differ if the component portions of nonpoint sources were
           ranked individually.  Nonpoint source impacts from mining (e.g., abandoned  mine drainage)
           would rank very high in Region III,  but would not be very important in New England.  Similar
           differences  across Regions might begin to emerge if other problem  areas were subdivided --
           perhaps by splitting criteria air pollutants  into those  from stationary vs. mobile sources, or by
           splitting non-hazardous waste sites into landfills vs. incinerators.

        The Region's conclusion is that a  risk  ranking of environmental problems, when  conducted at a
broad level, will show  little geographic variation.  What is true for one Region will be generally true  for
another.  There is a level of  risk  associated with each environmental problem  that does not vary much
                                                40

-------
geographically.  This supports national priority-setting when very broad questions are at issue.  But as the
ranking or  priority-setting becomes finer, focusing on smaller  geographic areas or smaller portions of
programs, geographic distinctions become much more important.
Level of Confidence  in Ranking Results

        How accurate are the  rankings produced in the Region  III Comparative Risk Project?  This
question is difficult to answer.  Very little of the analysis underlying the rankings is sufficiently quantitative
to be subjected to  traditional sensitivity analysis.   Most of the conclusions rely on a mixture of facts and
judgments, and there is very little basis for knowing how much the judgments might be in error.

        Arguing against the accuracy of the rankings are several factors:  the limited resources for analysis
and data acquisition, the unavailability of data on many key questions, and the  novelty of the risk analysis
methods used for  some areas.   On the other hand, the project  participants who made  the  necessary
judgments  formed  a very  high quality  group.   They  were experienced  Regional professional  staff,
representing  all program  areas, and  Headquarters  staff familiar with  Unfinished Business  and  other
Headquarters comparative risk  analyses.   The contractors had national reputations for expertise in risk
analysis in  their particular subject areas.

        The best way to judge the quality of the ranking results is to rely  on the assessment of the
individuals  who generated them.  For  the  most part, the participants  feel comfortable with their relative
rankings of the problem areas.  They believe that there are  very substantial differences in risk between
problem areas  that have been  captured accurately  in  the rankings,  despite missing data and  unproven
methods.

        When work group participants were asked whether they would recommend redoing the ranking using
better data and methods, they argued against doing so for at least several years.  This was primarily because
they did not believe the rankings would change much.

        This belief  is supported by noting the similarity in the risk  rankings with  the Unfinished Business
project The similar results for the two projects, arrived at generally independently, suggest either that both
projects are generally capturing  the truth, or that both have made some  consistent  methodological error.
                                                41

-------
        In sum, the participants in the Comparative Risk Project appreciate the imperfections in what they
have  done,  but believe  that  the  rankings of the  problem areas reflect their relative risks reasonably
accurately.
Information Gaps That Might Be Filled in the Future

        The  participants in  the  project noted several areas where their conclusions were  particularly
uncertain.  These areas were not necessarily those for which data were the most limited or of the poorest
quality. In several cases, the work groups noted problem areas  for which data were poor but to which a
ranking could nevertheless be assigned confidently.  For some other problems, much high-quality data were
available, but the problem area still could not be ranked confidently because of the absence of a single key
piece of information.  The work groups noted several areas where additional information and analysis could
do much to improve the confidence of the rankings, and where the gaps in information and analysis could
probably be filled at reasonable cost.  These areas  may be the focus of some future efforts to improve the
reliability of the ranking results.

        The following data elements were available  at the time the comparative risk analysis was performed,
but were not  obtained for the project.  In some cases the reason was that limited project resources explicitly
prevented staff from acquiring further data on certain topics; in other cases the need for the data did not
become apparent until too  late in the process.

        o   A better understanding of the health effects of lead, adapted in a manner to fit the work group's
           scoring approach.

        o   A critical review of the epidemiological evidence for  health effects  from airborne sulfates.

        o   A full compilation of typical  concentrations of indoor air pollutants.

        o   An exhaustive search for studies  relating to the ecological effects of toxic air  pollutants.

        o   Data on concentrations at which toxic  air pollutants  are  found in the Region, relative  to the
           lexicological thresholds at which they may begin to cause non-cancer human health effects.

        o  Data on actual, as opposed to modeled, concentrations of pesticide residues on foods as they
          are consumed.

                                                42

-------
        o  A study of the 305(b) reports from the Region III states, including: an assessment of the validity
           of aggregating data across the states, developing some judgment about conditions  in  the
           unassessed stream miles, and  extracting from them  data about the relative  effects  of waste
           problem areas (e.g.,  solid waste, UST, CERCLA sites) vs. water problem areas.

        o  A canvas for new studies monetizing welfare damages from environmental problems.

In addition, some other data  elements were not available at the time  of the  Comparative Risk Project
analysis because they are  only now being generated.  The following new pieces of data should be  available
within the  next year or so.

        o  Results  of acid deposition research from Pennsylvania State University, especially on terrestrial
           effects.

        o  Better  data  on toxic chemicals  in the effluents from point sources, particularly for sources
           discharging to  priority water bodies.

        o  Greatly  expanded data on chemical contaminants in larger public water systems, as they meet
           the new monitoring requirements.

        o  Information from the national survey of pesticides in  ground  water.

        o  Improved techniques for ecological risk assessment, as a result of the efforts of EPA's  Office of
           Research and Development.

        o  Ambient air quality  data for 1988, for which meteorological conditions were probably worse
           than in other recent  years.
Benefits  of the Project

        The Comparative Risk Project is not completed.  Eventually it will contribute to better resource
allocation in the Region, with efforts targeted where they can provide the greatest environmental benefits.
The risk assessment findings of the first year of the project provide a clear and useful description of the
environmental problems remaining in the Region. The primary benefits of the project will come when these

                                                43

-------
problems are addressed.  In the meantime, though, before the risk management portions of the project
are completed, the project has already provided some important benefits.

       o  The Region has an understanding of not only the relative risks from the environmental problems,
          but also the  causes or "anatomy" of risk for each problem.  The more detailed understanding
          of which pollutants, pathways, source types, or geographic areas ("hot spots") cause most of the
          risk in each problem area makes it easier to design efficient, targeted programs to reduce risks
          in each area.

       o  The Regional staff participating in the project have received practical training in risk assessment,
          and have developed a better cross-program perspective.  Many project participants say they have
          profited from the opportunity to learn from their colleagues about environmental problems and
          programs other than their own. As EPA continues to move toward a cross-media approach to
          environmental management, this broader view by the staff will become more important.

       o  Senior Regional managers have gained a vision of environmental problems and directions.  The
          comprehensive  purview of the analysis and rankings gives managers a  good  perspective for
          strategic planning.

       o  The project has focused all Regional participants on reduction of health, ecological, and welfare
          risks as the ultimate reason for environmental protection.
                                               44

-------
                                         V. Next Steps
        In its next phase, the Region III Comparative Risk Project will move generally from risk assessment
into risk management.  The Region will use the rankings to allocate its resources more efficiently and better
manage the risks that  have been studied.  The Region's  senior managers feel comfortable with the  risk
rankings and are developing specific  plans for how to use them.  The Region's major objectives will be to:

        o  Work toward systemic changes.  The Region will work to foster management processes in EPA
          and states that  reconcile  the priorities suggested by  the comparative  risk rankings and those
          mandated by other concerns.  The Region will pursue two themes: greater attention in resource
          allocation to risks  and the opportunity to reduce them,  and greater  flexibility to respond to
          unique local and regional conditions.  The Region will pursue the themes at three  levels: it will
          participate in Headquarters initiatives to institutionalize risk-based decision-making; it will develop
          and discuss  options for addressing high-priority problems more effectively within the Region; and
          it will work with the states to promote the use of the Comparative Risk Project  in upcoming
          grants negotiations.

        o  Initiate MERITS projects.  The Region will seek to implement MERITs for problems identified
          as particularly high risk.  MERITs will be targeted at  the particular elements of high-risk issues
          that contribute to their high ranking.

        o  Involve the  Region III states.  The Region will develop a strategy to  increase its involvement
          with the states so that both can participate effectively  in each other's planning and  management
          processes.

        o  Analyze and strengthen the project's results. The Region will document and further analyze the
          results of the first phase of the project. This will include comparing the results with those from
          other comparative risk projects.

        o  Communicate the project's results.   The Region will develop and implement a strategy to
          communicate the project results to the public and relevant government agencies in Region III.

        o  Evaluate Regional  resource allocation relative  to residual risks.  An initial look suggests that
          the Regional Office may  not be directing its resources where the highest risks lie.  The  five

                                                45

-------
           highest health risk problem areas are each the subject of small Regional programs  - indoor
           air pollution, radon, pesticides, radiation, and nonpoint source discharges.  The same is true of
           the four highest ecological risk areas -- modification of terrestrial and aquatic habitats, nonpoint
           sources, and acid deposition.  The same is true again of two of the three highest welfare risk
           areas  - acid deposition  and  nonpoint sources.  At the  other extreme,  the Region devotes
           substantial resources to problems that rank much lower in terms of residual risks:  CERCLA and
           RCRA sites, USTs, air toxics, industrial point sources, and POTWs. There are several reasons
           why resource allocation should not necessarily match residual risks, but the Region will evaluate
           this  issue.

        The ultimate test of the utility of  the Region III Comparative Risk Project is whether it contributes
to improved allocation of the Region's resources for protecting the environment.   The first phase of the
project has provided a comprehensive picture of the relative seriousness of the environmental problems
facing the Region.  The task  of building  on these findings lies  ahead.
                                                46

-------
APPENDIX 1 - DEFINITIONS OF REGION III PROBLEM AREAS
            ISSUE
          INCLUDES
         EXCLUDES
 1. Criteria Air Pollutants
Ambient Sulfur Dioxide, PM 10
(TSP prior to approved PM 10
SIP), CO, NOx, ozone & related
VOCs, and lead.
Acid deposition.
2. Hazardous/Toxic Air Pollutants
NESHAPs substances (approved
and   pending),  Acutely   Toxic
Chemicals List, pesticides, routine
& accidental releases.
Toxics from wastewater treatment
plants,    CERCLA    sites,
radionuclides  NESHAPs,  solid
waste disposal, RCRA TSD, air
deposition impacts.
3. Radon - Indoor
Indoor radon exposures from any
source.
Occupational exposure, outdoor
exposure.
4. Indoor Air Pollutants Other
Than Radon
All  indoor  exposures  to  air
pollutants for example:  asbestos,
formaldehyde, tobacco, CO, NOx,
and pesticides.
Occupational exposure
5.  Radiation Other Than Indoor
Radon
Naturally    occurring,
manufacturing, radioactive waste
disposal,  indoor  radiation other
than radon, non-ionized activities
(microwaves,  high-tension lines,
etc.).
Medical x-rays, CERCLA sites,
cosmic rays exposure in aircraft,
exposure from ozone  depletion,
occupational  exposure,  nuclear
power plant accidents.
6.  Impacts  of  Industrial  Point
Source  Direct  Discharge  of
Wastewater  on Surface  Waters
and Air
Pollutants   in   wastewater
generated by all privately-owned
sources    that    are    directly
discharged   to  surface  waters
(including   wetlands)    through
discrete    conveyances    or
volatilized to air.
Discharges  to  or from  publicly
owned   treatment   facilities,
treatment  sludges,  ground-water
impacts    from    wastewater
treatment, and  physical  impacts
from  discharges of dredge and
fill material.

-------
            ISSUE
          INCLUDES
         EXCLUDES
7. Impacts of POTW Discharges
on Surface Water and Air
Pollutants    in    wastewater
generated by all publicly owned
sources   that   are   directly
discharged  to  surface  waters
(including   wetlands)   through
discrete    conveyances    or
volatilized   to   air,   indirect
industrial   discharges,   and
combined sewer overflows.
Discharges to or from  privately
owned    treatment   facilities,
treatment sludges,  groundwater
impacts    from   wastewater
treatment, and physical impacts
from  discharges of dredge and
fill material.
8. Non-point Source Discharges
to Surface Waters
Discharges  from  non-discrete
conveyances including agricultural
runoff,   industrial    runoff,
silvicultural   runoff,   pesticide
runoff, surface discharge of septic
tanks,  stormwater runoff,  mine
drainage,  contaminated in-place
sediments, air deposition, oil and
gas  operations,  and  chemical
discharges   from  disposal  of
dredge and fill materials.
Acid deposition impacts, discrete
discharges    of    contaminated
ground   water,    solid   waste
disposal, hazardous  waste sites
(RCRA   &   CERCLA),   and
physical impacts  from discharges
of dredge and fill material.
9. Management  of Hazardous
Waste at RCRA Facilities
All discharges to air, soil, surface
water  and ground water  from
active   and    closed    RCRA
facilities,  waste  transportation,
and  illegal   disposal/lack  of
capacity.
Discharges   to   wastewater
treatment plants and criteria air
pollutants.
10.  Hazardous   Substances  at
CERCLA Sites
NPL  sites and potential  NPL
sites.    Illegal  disposal/lack  of
capacity.
Discharges    to   wastewater
treatment plants and criteria air
pollutants.
11. Solid Waste Management
Multi-media  discharges to  air,
soil,  surface  water, and ground
water   from   all   household,
municipal,  and industrial  waste
not  regulated by  RCRA as  a
hazardous    waste,   treatment
sludges, waste transportation, and
illegal disposal/lack of capacity.
Discharges    to   wastewater
treatment plants and criteria air
pollutants.
12.  Releases  from Underground
Storage Tanks
All  substances  released   from
underground storage tanks, such
as gasoline, pesticides, solvents,
and oil.
RCRA-regulated
CERCLA sites.
tanks   and

-------
            ISSUE
                          INCLUDES
                                          EXCLUDES
13.    Other
Contamination
Ground-Water
Pollutants contaminating ground
water  from  such  sources  as
agriculture;  industry,  municipal
activities; silviculture; oil, gas &
mining  operations;  pesticides;
UlC-related discharges; road salt;
urban    runoff;   underground
discharges  from  septic  tanks;
saltwater intrusion; and naturally
occurring fluorides.
CERCLA  -   and  RCRA  -
regulated   sites,  underground
storage  tanks,  and solid  waste
disposal.
14.   Other
Contamination
  Pesticide   Residues on  and  in  food and
                 applicator exposure.
                                Surface water runoff, aerial drift,
                                ground-water    contamination,
                                manufacturing,  disposal,  non-
                                commercial and non-agricultural
                                applicators.
15.  Physical   Modification
Aquatic Habitats
            of   All physical changes to aquatic
                 habitats such  as  dredging and
                 filling  of  wetlands, dams, and
                 channelization.
                                 Chemical impacts from disposal
                                 of dredge and fill materials.
16.  Physical   Modification  of   All physical changes to sensitive   Chemical impacts  from disposal
Environmentally   Sensitive   terrestrial  habitats  such  as dam   of dredge and fill materials.
Terrestrial Habitats               building,   strip   mining,   and
                                 highway construction.
17.  Acid Deposition
                 All  damages  caused by wet or
                 dry    deposition    of   acidic
                 compounds from the atmosphere.
                                Primary impacts of sulfur oxides,
                                NOx, and VOCs.
18.  Operation and Maintenance
of Water Supply Facilities
                 All water treatment facilities and
                 distribution networks.
                                Contamination in the raw water.

-------
APPENDIX 2 - LIST OF PROJECT PARTICIPANTS
Region 3:

Robert Allen
Sheila Briggs
Henry Brubaker
Maryann Bucknavage
Jeffrey Burke *
Jon Capacasa
Gail Caron
Robin Cole
Barbara D'Angelo
Joe Davis
Roger Devin
Bruce Diamond
Richard Fetzer *
Alyce Fritz
James Harper
Kim Hummel
Greene Jones
Fred Knapp
Robert Kramer
Stanley Laskowslti
Dennis Lohman
Paula Luborsky
Thomas Maslany
Andy McErlean
Tony Meadows
Israel Milner
Bruce Molholt
Alvin Morris
Randy Pomponio
Ron Preston
Robert Reed
Charles Rhodes
Robert Runowski
Charles Sapp
Roland Schrecongost
Dianne Sims
Roy Smith
Bruce Smith
Henry Sokolowski
Rebecca Taggart
Lawrence Teller
Jean Thompson
Virginia Thompson
James Topsale
Michael Vaccaro
Janet Viniski
Randy Waite
Stephen Wassersug
Carey Widman
Patricia Wilbur *
William Wisniewski
Sidney Worthington
Denis Zielinski
         Current and former project managers
EPA Headquarters:

James Hemby
Steve Horn
Eva Ring
Catherine Tunis

-------
I] 'I111  I11  " ™ 111 r  '  [	""If"    [
                                                                       f"

-------
T"! 1
      ! Il
   I  I
      i iill


               •
                j::
                         . M
                          •;  • i"
                          i  :!
                                                             3


                                                             !

-------