Superfund: Improved Reviews and Guidance Could Reduce Inconsistencies in
Risk Assessments (Letter Report, 08/10/94, GAO/RCED-94-220).

The Environmental Protection Agency (EPA) is required to assess the
risks to human health posed by each of the Superfund program's hundreds
of hazardous waste sites. Because EPA uses these risk assessments to
determine whether and how sites should be cleaned up, they are a sources
of considerable scrutiny and controversy. For example, both industry and
environmental groups have criticized risk assessment for inconsistently
estimating the amount of contamination that persons living near the
Superfund sites may have been exposed to. This report reviews 20 of the
approximately 70 risk assessments EPA did in 1992. GAO provides
information on (1) whether the risk assessments adhered to EPA guidance,
(2) whether they varied among the regions and sites, and (3) how EPA
monitors the quality and consistency of risk assessments. GAO focused
particularly on how risk assessments measured human exposure to
hazardous contaminants and calculated the resulting risk. GAO also
provides data on the sources of contamination described in these risk
assessments and the risks associated with these sources.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-94-220
     TITLE:  Superfund: Improved Reviews and Guidance Could Reduce 
             Inconsistencies in Risk Assessments
      DATE:  08/10/94
   SUBJECT:  Environmental monitoring
             Hazardous substances
             Waste disposal
             Quality assurance
             Risk management
             Evaluation methods
             Health hazards
             Waste management
             Toxic substances
             Information gathering operations
IDENTIFIER:  Superfund Program
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

August 1994

SUPERFUND - IMPROVED REVIEWS AND
GUIDANCE COULD REDUCE
INCONSISTENCIES IN RISK
ASSESSMENTS

GAO/RCED-94-220

Improving Superfund Risk Assessments


Abbreviations
=============================================================== ABBREV

  ARAR - applicable or relevant and appropriate requirement
  CERCLA - Comprehensive Environmental Response, Compensation, and
     Liability Act
  EPA - Environmental Protection Agency
  NPL - National Priorities List
  OSWER - Office of Solid Waste and Emergency Response
  VOC - volatile organic compound

Letter
=============================================================== LETTER


B-257513

August 10, 1994

The Honorable John D.  Dingell
Chairman, Committee on Energy and Commerce
House of Representatives

The Honorable Al Swift
Chairman, Subcommittee on Transportation
 and Hazardous Materials
Committee on Energy and Commerce
House of Representatives

EPA's regulations implementing the Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA) require an
assessment of the risks to human health posed by each of the
Superfund program's hundreds of hazardous waste sites.  Because these
risk assessments are a tool, among several, that the Environmental
Protection Agency (EPA) uses to determine whether and how sites
should be cleaned up, they have been a source of considerable
scrutiny and controversy.  For example, both industry and
environmental groups have criticized risk assessments for
inconsistently estimating the amount of contamination to which those
living near Superfund sites may be exposed.\1

Partly because of these concerns, the administration's proposal for
reauthorizing Superfund calls for writing a new risk assessment
regulation that will establish more specifically how risk assessments
are to be conducted. 

In response to your questions about the way EPA assesses the health
risks posed by Superfund sites, we reviewed 20 (2 from each of EPA's
10 regions) of the approximately 70 risk assessments conducted in
1992.  In this report, we provide information on (1) whether the risk
assessments adhered to EPA's guidance, (2) whether they varied among
regions and sites, and (3) how EPA monitors the quality and
consistency of risk assessments.  As you requested, we focused
particularly on how risk assessments measured human exposure to
hazardous contaminants and calculated the resulting risk.  You also
asked for data on the sources of contamination described in these
risk assessments and the risks associated with these sources.  This
information appears in appendixes II through VI. 


--------------------
\1 We discussed some of these issues in Superfund:  Risk Assessment
Process and Issues (GAO/T-RCED-93-74, Sept.  30, 1993). 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

In general, the 20 risk assessments we reviewed adhered to EPA's
guidance and were prepared consistently across EPA's 10 regions. 
Most of the assessments followed the guidance in how they identified
which contaminants were present and used similar assumptions in
measuring human exposure to hazardous material in the environment. 
However, some of the assessments we reviewed did not follow EPA's
guidance for (1) estimating the level of contamination, (2)
adequately describing the assumptions and uncertainty inherent in the
assessments, and (3) calculating the total risk from all means of
exposure.  As a result, risks were estimated to be either higher or
lower than they would have been if the risk assessment team had
followed the guidance. 

Although the risk assessments we reviewed generally followed
consistent approaches, they sometimes used different assumptions in
estimating exposure, specifically, in (1) judging how sites might be
used in the future and how much contamination would remain there and
(2) determining how people absorb contaminants through their skin. 
When such inconsistencies occurred, risks were estimated differently
even though the sites had similar characteristics. 

To monitor the quality and consistency of risk assessments, EPA has
conducted annual quality assurance reviews--collecting extensive data
for about half of the assessments conducted each year.  But these
reviews produce only summary descriptive information on how risk
assessments were done and do not analyze for inconsistencies among
risk assessments.  EPA staff agreed that the reviews could be
improved by analyzing for consistency and noted that this would not
require significantly more resources because the reviews already
gather most of the necessary information. 


   BACKGROUND
------------------------------------------------------------ Letter :2

With the enactment of CERCLA in 1980, the Congress created the
Superfund program, authorizing a trust fund to clean up the nation's
most severely contaminated hazardous waste sites.  The program was
extended in 1986 and in 1990 and is now being considered for
reauthorization.  Under CERCLA, EPA reviews contaminated areas and
then places the nation's most highly contaminated sites on a priority
list for investigation and cleanup.  Since 1980, EPA has included
over 1,200 sites on the list. 

EPA begins work at each listed site by conducting a "remedial
investigation" to determine whether the nature and extent of
contamination warrant cleanup.  One element of this investigation is
the baseline risk assessment--a scientific evaluation of any current
and potential threats to human health the site would pose if no
cleanup occurred.  The risk assessment looks at the critical
components of risk:  the toxicity of the contaminants present and the
likely exposure to these contaminants.  Along with other information,
the assessment may be used to determine whether cleanup is warranted
and to describe the potential risks for those who must decide how to
clean up the site and for the local community. 

The risk assessment has four steps:  (1) collecting and analyzing
data from site samples to determine the types and levels of chemicals
present; (2) estimating the extent to which populations living at or
near the site might be exposed to these chemicals through various
"pathways"--for example, by drinking contaminated water or touching
contaminated soil; (3) assessing the toxicity of the chemicals; and
(4) calculating the risks from exposure to this contamination. 
(Figs.  I.1 and I.2, in app.  I, illustrate the steps of a risk
assessment and the ways people can be exposed to hazardous waste.) A
team of EPA regional staff that generally includes the site's project
manager, a toxicologist, and other technical staff perform these risk
assessment steps. 

The risk of developing cancer from the chemicals is expressed as a
probability, such as 1 in 10,000.  The risk of developing other
health problems is determined by calculating a numerical "hazard
index" that evaluates whether health problems could occur from
exposure to the hazardous chemicals.  A value greater than 1 means
that a risk of health effects exists. 

EPA's policy states that risks greater than 1 in 10,000 for
carcinogens and greater than 1 for noncarcinogens are considered
serious enough to require cleanup action.  However, other factors in
addition to the risk assessment also influence whether and how a
Superfund site is cleaned up.  CERCLA states that cleanups must meet
"legally applicable" or "relevant and appropriate" requirements
(ARARs), including applicable or relevant federal and state
environmental regulations.  ARARs in some instances require cleanup
when the risk assessment's results would not.  For example, CERCLA
requires that surface water, like ponds or rivers, near hazardous
waste sites meet the legally applicable federal Clean Water Act's
standards, independently of risk assessment results.  EPA must also
consider factors such as the cost, feasibility, and ecological risk
in deciding how to clean up Superfund sites.\2

To improve the quality of risk assessments and promote consistency,
EPA issued Risk Assessment Guidance for Superfund in 1989.  This
document lays out guidelines for each step of evaluating the health
risks at Superfund sites, identifies sources of data, and provides
values that can be used in calculating exposure and the resulting
risk.  Because of scientific uncertainty in the data used in
calculating risk, as well as the variable conditions at each site,
the guidance grants leeway for professional judgment.  Nevertheless,
the guidance recommends that estimates of risks be conservative--that
the estimates include a "margin of safety." In calculating human
exposure to site contaminants, EPA recommends using a "reasonable
maximum," or higher than average but still realistic number.  For
example, to measure how long someone might live in a home near a
waste site, EPA looked at statistics on how long Americans live in
their homes and then selected a value (30 years) that represents the
experience of the upper 10 percent of the population.  EPA developed
this conservative approach in order to ensure that site cleanups done
under CERCLA protect human health. 

Both the uncertainty and the assumptions about exposure to
contamination that are involved in Superfund risk assessments have
drawn considerable criticism.  Private parties responsible for
cleaning up sites have argued that risk assessments use excessively
conservative assumptions and that EPA does not use consistent
assumptions from site to site.  Both industry and environmental
advocacy groups have criticized risk assessments for not adequately
accounting for scientific uncertainty and variability. 


--------------------
\2 We are currently evaluating the role of risk assessments and other
factors in site cleanup decisions. 


   RISK ASSESSMENTS ADHERED TO
   MOST GUIDANCE, EXCEPT FOR THREE
   AREAS
------------------------------------------------------------ Letter :3

For the most part, the 20 risk assessments we reviewed adhered to
EPA's guidance.  Specifically, the assessments used the prescribed
equations in calculating people's daily intake of contaminants;
justified excluding detected contaminants from further evaluation;
combined risks from separate carcinogenic and noncarcinogenic
contaminants; used the specified values for the toxicity of
contaminants given people's length of exposure; and adequately
justified excluding specific types of exposure.  However, of the risk
assessments we reviewed,

  3 did not follow EPA's guidance for estimating the level of
     contamination, thereby overstating the level of contamination;

  19 did not adequately explain the uncertainty and variability in
     the data used and the assumptions made; and

  7 did not include proper calculations of the total risk to people
     who could come into contact with several sources of
     contamination, thereby understating risk. 

EPA officials said these deviations from the guidance occurred
primarily either because the risk assessment team followed different
guidance developed in one of EPA's regions or because the team found
it difficult to obtain information needed to follow the general
guidance EPA provided. 


      EPA OVERSTATED THE LEVEL OF
      CONTAMINATION IN THREE CASES
---------------------------------------------------------- Letter :3.1

EPA's guidance recommends that the risk assessment team use data from
site samples to calculate the level of contamination that people
might actually encounter at the site.  In particular, the guidance
states that the measure of contamination (such as the milligrams of a
contaminant per kilogram of soil) should be an estimate of the
average of the site samples.\3 This estimate is important because it
helps determine how much contamination people could come in contact
with each day. 

Three of the 20 risk assessments we reviewed did not follow EPA's
guidance on this matter.  Instead, they used a higher estimate,
resulting in an overstatement of how much contamination people could
encounter at the site.  Two used the highest level of contamination
found, while the other used an estimate much higher than the average
but not the highest level found at the site.  For example, the risk
assessment team for the PSC Resources site, a solvent-recycling
facility in Massachusetts, used the sample with the highest amount of
contamination to determine how much someone could come in contact
with each day.\4 Using the highest contamination level implies that
exposed people would spend all their time at the most contaminated
part of the site, which is likely to overestimate exposure. 

The risk assessment for PSC Resources followed guidance by EPA Region
I (Boston) that conflicted with EPA's national guidance. 
Specifically, guidance by Region I recommended that its risk
assessment staff use the highest level of contamination found, rather
than the average level.  EPA regional and headquarters staff
explained that this conflict between headquarters and Region I
stemmed from a disagreement they had over the best way to measure the
level of contamination.  Region I officials said they are now
changing the region's policy to make it conform to the national
guidance. 


--------------------
\3 The risk assessment team is supposed to use a standard statistical
procedure, with the result that in only 5 out of 100 instances will
the actual average be greater than the value EPA uses to describe the
level of contamination. 

\4 EPA's guidance states that when an adequate number of samples is
not available (generally 20 to 30 samples are sufficient), the
highest level of contamination found at the site may be used. 


      MOST RISK ASSESSMENTS DID
      NOT ADEQUATELY DISCUSS
      UNCERTAINTY AND VARIABILITY
---------------------------------------------------------- Letter :3.2

Precisely estimating risk is difficult because the data used to
derive the estimates often are uncertain and variable.  Such
uncertainty occurs when EPA cannot determine an exact value to use in
estimating risk.  For example, the risk assessment team would have no
way of knowing the exact number of days each year that people living
at a given site would be exposed to waste.  Data also may vary
considerably.  For example, because the level of contamination is
different at various locations at a site, the potential for
exposure--and therefore the risk--will vary from spot to spot.  In
cases in which precise data are not available, EPA often uses
assumptions to help approximate values.  Given such uncertainty and
variability, it is important to disclose and attempt to measure how
the assumptions used in the risk assessment affect its outcome.  In
so doing, risk assessment teams are supposed to demonstrate that a
range of potential risk exists at the site, which is more realistic
than a single estimate of risk.  Identifying the sources of
uncertainty and variability also allows the risk assessment team to
decide where better information might improve the risk estimate. 

EPA's guidance recommends that a risk assessment communicate the
precision of its estimates by explaining the limitations of the data
in all of its steps.  Such disclosures should include a description
of any uncertainty, variability, or assumptions used in the risk
assessment.  Specifically, EPA states that the risk assessment should
provide the ranges of possible values in the data used throughout the
assessment.  EPA also directs that the assessment explain both the
reasons for the values or assumptions used and their impact on the
calculated level of risk.  For example, when determining how
frequently people are exposed to waste when working at a site, a risk
assessment would be expected to (1) provide a realistic range for the
number of days people typically spend at work, (2) explain how a
given assumption (for example, 250 days per year) was selected, and
(3) evaluate whether that value would tend to overstate or understate
the risk. 

Nineteen of the 20 risk assessments we reviewed did not follow EPA's
guidance established to encourage full disclosure of the limitations
of the data.  Specifically,

  18 did not include any information on the ranges of possible values
     to measure exposure (for example, the number of days people
     might typically work at the site);

  7 did not explain how they arrived at the values or assumptions
     used in calculating risk (for example, why a risk assessment
     included an assumption that people would work 250 days per year
     at the site); and

  10 did not explain how the values or assumptions affected the risk
     estimate (for example, whether using 250 working days per year
     would tend to overstate or understate the risk). 

When risk assessments did communicate information about uncertainty
and variability, it was easier to understand how the data and
assumptions they used affected the risk estimates.  For example, the
risk assessment for the Revere Textile Prints site in Connecticut
included the range of possible values for determining human exposure,
such as how many days during the year workers, residents, and
trespassers could come in contact with contaminated areas, and how
much contaminated material, such as soil or groundwater, they could
take in each day.  These values often varied by several orders of
magnitude.  The risk assessment included a table explaining how the
risk assessment team selected values from the ranges.  (For example,
for trespassers, who could be on the site from 1 to 365 days per
year, the assessment assumed 52 days on the basis of one or two
visits a week during spring, summer, and fall and none in winter.) It
also explained that in many cases the assumed values overstated the
typical exposure and could result in a high estimate of risk. 

EPA officials said that obtaining the information necessary to
discuss uncertainty and variability in data is difficult.  In
particular, the agency only has limited data on the ranges of
exposure to contaminants.  The officials also told us that their
staff had varying levels of training and experience in the
statistical analysis tools needed to describe limitations in data and
the resultant effect on risk estimates.  They noted, however, that
staff in Region III (Philadelphia) have become leaders in this area
by developing guidance on how to include this uncertainty and
variability in the overall assessment of risk. 

EPA officials acknowledged that the agency could improve its handling
of uncertainty and variability.  They noted that although the
agency's guidance states that risk assessments should fully disclose
such limitations, the guidance is silent on how to statistically
measure these limitations and their impact on the overall estimate of
risk, as well as how to use the information when selecting a remedy. 
They added that Region III's experience could provide a basis for
more specific guidance on the issue.  However, they cautioned that
progress will be slow until further research provides better data on
how people are exposed to hazardous waste sites. 


      A THIRD OF THE ASSESSMENTS
      UNDERSTATED RISK BY NOT
      CONSIDERING MULTIPLE SOURCES
      OF CONTAMINATION
---------------------------------------------------------- Letter :3.3

As part of estimating the human health risk at a Superfund site,
EPA's guidance directs risk assessment staff to calculate the total
risk faced by people who could be exposed to more than one source of
contamination while living or working at the site.  According to the
guidance, risk from different contaminants and sources can be added
together--people could drink contaminated water and come in contact
with contaminated soil, for example, over the same time period.  At
18 of the 20 sites in our review there was the potential for people
to be exposed to more than one source of contamination, but in 7
instances the risk assessment did not calculate the total risk from
being exposed to contamination from more than one source. 

People living at the Revere Textile Prints site in Connecticut, for
example, could be exposed to contaminants in both the groundwater and
soil.  However, the risk assessment calculated the risks from the
groundwater and soil separately without combining them to show the
total risk to someone living at the site.  The assessment estimated
that the individual risks of contracting cancer from drinking
groundwater and from coming into contact with the soil were 5 in
10,000 and 4 in 10,000, respectively.  But the risk assessment for
Revere Textile did not include a calculation of total risk even
though people exposed to both risks would face, under EPA's guidance,
a total risk of 9 in 10,000, or about twice as high as the risks from
groundwater or soil considered separately. 

EPA headquarters officials acknowledged that the agency's guidance
calls for each risk assessment to include a calculation of total risk
when appropriate but said that they believed that individually
calculating a risk for each medium (groundwater, soil, etc.) was
often sufficient to determine whether the site required cleaning up. 
However, the officials agreed that in order to compare the risks at
various Superfund sites, it was necessary to calculate the combined
risk posed by all sources of contamination at each site.  They said
that calculating the total risk posed little technical difficulty and
added that the administration's proposed reauthorization bill would
require EPA to develop comprehensive risk assessment regulations that
would include rules on combining risks from different sources. 


   RISK ASSESSMENTS USED GENERALLY
   CONSISTENT APPROACHES, BUT SOME
   STEPS VARIED
------------------------------------------------------------ Letter :4

The 20 risk assessments in our review were generally consistent for
most aspects of the risk assessment process.  For example, the risk
assessments generally used consistent assumptions about (1) the
quantities of groundwater and soil people consume or ingest each day,
(2) the number of days per year people live or work at the sites, (3)
the nature of the population exposed to the contamination (i.e.,
adults, children, or sensitive populations), and (4) the
characteristics of exposed populations (i.e., people's weight and
number of years of exposure).  (See apps.  II and III for more
information on these values.)

However, in two areas, many of the 20 risk assessments used
inconsistent assumptions even when sites had similar contamination. 
This variation occurred in assumptions about

  how the site would be used in the future, what the routes of
     exposure would be, and how much contamination would remain at
     the site and

  how much contamination people absorb through their skin. 

These inconsistencies occurred because EPA left to the risk
assessment team many of the decisions about how to measure exposure. 
As a result, different estimates of risk were developed for sites
even with similar characteristics.  To the extent that these risk
assessments are used to make cleanup decisions, this could mean that
one site might be cleaned up and another similar site might be left
alone only because of arbitrary differences in assumptions.  EPA
officials said that risk assessment staff were granted such latitude
because the assessments must take into account the unique conditions
at each site and because scientific uncertainty abounds in certain
aspects of the process.  However, the inconsistencies we found were
not related to differences in sites' conditions.  The EPA officials
also noted that in several instances, the agency is working on needed
clarification to its guidance. 


      RISK ASSESSMENTS MADE
      INCONSISTENT JUDGMENTS ABOUT
      FUTURE LAND USE AND LEVELS
      OF CONTAMINATION
---------------------------------------------------------- Letter :4.1

EPA's guidance calls for the risk assessment team to determine (1)
how a Superfund site will be used in the future, (2) how people will
be exposed to the sources of contamination, and (3) what level of
exposure this will generate.  The guidance then directs the
assessment team to factor these judgments into an evaluation of
future exposure to site contamination.



         HOW LAND WILL BE USED IN
         THE FUTURE
-------------------------------------------------------- Letter :4.1.1

Judgments about how land will be used are crucial to determining the
potential for human exposure to hazardous waste.  In fact, EPA
officials noted that assumptions about the future use of land were a
major determinant of many cleanup decisions.  For example, if the
land is used for homes, people are exposed over longer periods than
if the land is used for business, and a more stringent cleanup could
be required.  The risk assessment guidance gives general directions
on determining how land might be used in the future, such as looking
at population trends and zoning plans, but leaves the decision to the
risk assessment team.  In addition, EPA encourages the risk
assessment team to take a relatively conservative approach in
deciding the future use of the land.  Reasoning that EPA cannot
control local zoning or other land use restrictions, the guidance
suggests that the risk assessment assume that in the future, the land
will be residential even if no one lives there now. 

Consequently, most risk assessments in our review assumed that the
land at hazardous waste sites would be used differently in the future
than it is now.  The risk assessments assumed that residential
development would eventually occur on hazardous waste sites, even
though few sites had residences directly on them now.  We found that
under similar circumstances, risk assessments forecast different land
uses for their sites.  Specifically, of the 17 sites in our review
that were either abandoned or used for industrial or recreational
purposes, risk assessments assumed that 12 would have homes built on
them in the future but that 5 would never be used for residential
purposes.  (See app.  IV for more information on risk assessments'
assumptions about the uses of land and app.  V for the risk
associated with these uses.)

Three landfill sites we reviewed demonstrate the variation in risk
assessment teams' judgments about future land use.  All three sites
had similar conditions:  inadequate coverings over the landfill,
nearby residences, and contaminated groundwater affecting the
residents' drinking water.  Although landfills seem unlikely sites
for residential development, at the Hercules 009 Landfill in Georgia
and the Woodstock Landfill in Illinois, the risk assessments
concluded that people would build homes on them in the
future--exposing residents to contaminated soil and water every day. 
In contrast, at the Strasburg Landfill in Pennsylvania, the risk
assessment concluded that the site would not be developed but that
occasional trespassers would come in contact with the contamination
at the site.  The risks measured at the Hercules 009 and Woodstock
landfills indicated the need for cleanup, but the risk at the
Strasburg site did not exceed the criteria for cleanup.\5

We also found that risk assessments were inconsistent in their
evaluation of how the use of adjacent land affected the future use of
the site.  Ten of the sites in our review had nearby homes:  In eight
of these cases, the risk assessment used this fact to decide that the
site itself could be developed for residences in the future; two
other assessments concluded, however, that the sites would not be
developed for residential use. 

EPA officials acknowledged that estimating how land might be used is
a somewhat subjective and often contentious step in the risk
assessment, but, they pointed out, the agency is developing new
guidance on forecasting land use that directs the risk assessment
team to consult with the local community on such issues as zoning and
the use of adjacent land in making the decision.  This would offer a
more predictable and systematic approach than assuming that
residential use would prevail in almost all cases.  However, the
officials cautioned that some seemingly inconsistent decisions would
still occur because it is difficult to craft specific guidance that
covers every potential situation concerning land use. 


--------------------
\5 Even though the risk for the Strasburg Landfill was below the
level at which action is required, EPA is replacing the site's eroded
cap because contamination was expected to increase as the cap
continued to erode.  In addition, EPA will continue to treat the
contaminated water for several nearby residences, according to the
site's manager. 


         WHAT FUTURE EXPOSURE
         ROUTES WILL BE
-------------------------------------------------------- Letter :4.1.2

Exposure to contaminants also depends on any restrictions that may be
placed on how a site will be used in the future.  Such restrictions
would include the local government's requiring residents to use water
from an uncontaminated source, such as the municipal water supply,
rather than private wells--effectively preventing exposure to any
contaminated groundwater. 

The risk assessments we reviewed were not consistent in selecting
future exposure routes, especially when actions to reduce exposure
were already in place.  For nine of the sites in our review, local
governments provided residents with water from an uncontaminated
source, rather than a private well.  For two of these sites--Idaho
Pole Company in Montana\6 and Commencement Bay/Nearshore Tideflats in
Washington--the risk assessment did not include future exposure to
contaminated groundwater for on-site residents.  For the other seven
sites, including Atlantic Wood Industries in Virginia, Hercules 009
Landfill in Georgia, and Farmers' Mutual Cooperative in Iowa, the
risk assessment did include exposure to groundwater.  When exposure
to groundwater was evaluated, it often resulted in estimates of
substantial risk--residential exposure to contaminated groundwater
was the most common source of risk in the risk assessments we
reviewed.  (See app.  VI.)

EPA headquarters officials explained that when in doubt, risk
assessments should include exposure to contaminated groundwater,
disregarding the alternate water supply.  While this is a
conservative approach, EPA officials explained that it is consistent
with the agency's policy that assessments measure risk before any
actions are taken to reduce exposure.  The officials said that as
part of their revision to the guidance on assessing future land use,
they would clarify how to treat actions already taken to control
exposure to waste. 


--------------------
\6 For the Idaho Pole Company site, exposure to groundwater from a
well at currently unoccupied residences off the plant property was
evaluated, although groundwater that could supply future residences
on the plant property was not.  The site's manager said that the Safe
Drinking Water Act's standards would require cleaning up the
groundwater even though the risk assessment did not specifically
evaluate future on-site residents' exposure to groundwater. 


         HOW MUCH CONTAMINATION
         WILL REMAIN
-------------------------------------------------------- Letter :4.1.3

Hand in hand with determining how land will be used, risk assessments
also must estimate how much contamination will remain on that land in
the future.  EPA's guidance acknowledges that many models exist for
predicting future levels of contamination and therefore leaves the
decision of how to calculate this contamination up to the risk
assessment team.  As a result, the risk assessments we reviewed
calculated the level of future exposure differently even for sites
with similar types of contamination. 

Although most of the risk assessments in our review took the
conservative approach of assuming that people who lived or worked on
the site in the future would encounter the same levels of
contamination as today, two assumed that the level of contamination
would decrease over time.  For example, the risk assessment for the
Idaho Pole Company site in Montana, a wood treating facility, assumed
that contaminants in soil and water would degrade over time to
between 13 percent and 50 percent of their current concentration. 
The site's project manager told us that he consulted with several
experts in the agency about the model used to determine how much the
chemicals would degrade.  In contrast, the risk assessment for the
Atlantic Wood Industries site in Virginia, another wood treating
facility with many of the same contaminants, assumed that
contamination would remain at its current level. 

EPA headquarters officials said that risk assessment teams are
uncertain about how to calculate future levels of contaminants
because they increase or degrade into another hazardous chemical. 
Given this uncertainty, EPA said that many risk assessment teams make
the assumption that the contamination level will not change rather
than take on the complex task of estimating degradation, especially
if the degradation will not have a significant impact on the selected
cleanup.  The officials said that because EPA currently lacks the
resources to model the degradation of a range of contaminants under
various conditions, the agency is unlikely to provide specific
guidance in this matter in the foreseeable future.  As a result, some
risk assessments are likely to develop different estimates of future
risk under similar conditions. 


      RISK ASSESSMENTS
      INCONSISTENTLY DETERMINED
      HOW PEOPLE ABSORB
      CONTAMINANTS THROUGH THEIR
      SKIN
---------------------------------------------------------- Letter :4.2

EPA's guidance tells risk assessment teams to determine ways that
people are likely to be exposed to contamination through their
activities at a site.  Such exposure routes include dermal (skin)
contact, ingestion (swallowing), or inhalation (breathing).  Many of
the risk assessments we reviewed assumed that dermal exposure to
chemicals in soil occurred differently.  Specifically, we found
inconsistencies in

  whether risk assessments evaluated dermal absorption of chemicals
     in soil and

  how the assessments estimated the level of contamination people
     absorb through their skin. 

EPA officials explained that such differences arose because good
information for how people absorb contaminants from soil was
unavailable.  Consequently, some risk assessments did not include
this type of exposure, and some used inconsistent approaches to
measuring exposure from touching contaminated soil. 


      SOME RISK ASSESSMENTS DID
      NOT EVALUATE DERMAL EXPOSURE
      TO SOIL
---------------------------------------------------------- Letter :4.3

Because data on absorption rates (the percentage of a contaminant
that is absorbed into the bloodstream) often are lacking for many
chemicals found in soil that may be absorbed through the skin, EPA
has left to the risk assessment team the decision of whether to
include this route of exposure.  As a result, of the 13 risk
assessments in our review that addressed contaminated soil, 3
excluded dermal contact with the soil as a means of exposure--even
when other types of exposure to contaminated soil were included.  EPA
staff told us they often were uncomfortable evaluating dermal contact
with soil because known absorption rates for chemicals in soil exist
for only three of the hundreds of contaminants that can be found at
hazardous waste sites.  For example, risk assessment staff in Region
II (New York) believed that inadequate absorption data made the
calculations too uncertain.  Therefore, they adopted a policy of
never calculating dermal exposure for contaminants without absorption
factors.  In contrast, risk assessment staff in other regions
calculated dermal exposure at their sites by substituting the known
absorption rates for similar contaminants. 

In those cases in which dermal contact with soil was included, it
resulted in calculations of substantial risk.  For six of the risk
assessments we reviewed, the dermal exposure route alone resulted in
a cancer risk in excess of 1 in 10,000, EPA's threshold for requiring
cleanup action. 


      RISK ASSESSMENTS USED
      DIFFERENT DERMAL ABSORPTION
      RATES
---------------------------------------------------------- Letter :4.4

When risk assessment teams decided to evaluate dermal contact with
soil, they used different approaches to estimating the amount of
exposure.  The risk assessment team must determine whether people
exposed to the chemicals in soil will actually absorb them into their
bodies.  However, EPA's guidance is largely silent on how to estimate
absorption rates and suggests that risk assessment staff consult
scientific literature on the topic.  Consequently, we found that risk
assessments frequently used a default absorption rate of 100 percent,
and those that assumed lower absorption often used different rates
for the same chemicals.  This problem was most prevalent in risk
assessments that considered the risk of dermal exposure to
cancer-causing contaminants known as volatile organic compounds (VOC)
in soil. 

Nine of the risk assessments we reviewed evaluated dermal exposure to
VOCs.  Two of the assessments assumed that the absorption rate for
these chemicals was 10 percent; two assumed it was 25 percent; three
assumed it was 50 percent; and two assumed it was 100 percent.  In
these cases, the ten-fold difference in absorption rates between the
lowest and the highest would be enough to change the estimated chance
of contracting cancer in a lifetime from 1 in 100,000 to 1 in 10,000. 

EPA headquarters and regional officials acknowledged that considering
dermal contact is important to measuring exposure to contaminated
soil.  The officials also acknowledged that risk assessments should
not come to different conclusions about the level of exposure and
risk when people are exposed to similar contamination.  Superfund
officials said they recently asked EPA's Office of Research and
Development to review its policy on dermal contact with soil. 
Specifically, the review will focus on whether it is appropriate to
calculate risk without the proper absorption factors for specific
site contaminants.  Accordingly, EPA expects to establish a more
consistent policy this year on the basis of this review. 


   HEADQUARTERS' REVIEWS DO NOT
   IDENTIFY INCONSISTENCIES IN
   RISK ASSESSMENTS
------------------------------------------------------------ Letter :5

EPA regions are responsible for conducting risk assessments in
accordance with national guidance.  Since 1990, the Office of Solid
Waste and Emergency Response (OSWER) in EPA headquarters has
conducted annual quality assurance reviews intended to monitor
whether regions were following EPA's guidance for risk assessments. 
This effort has included (1) extracting from site documents
information about how the assessment was conducted, (2) compiling and
analyzing the information, and (3) reporting the results (based on
about half the risk assessments completed each year) to regional
staff.  We reviewed both the worksheets OSWER staff used to record
the results of their risk assessment reviews and the summary reports
based on these reviews. 

We found that OSWER had collected extensive, generally accurate data
during its quality assurance reviews.  We also found that OSWER's
reviews checked for compliance with important risk assessment
guidance and revealed some problems.  For example, the reviews showed
that some risk assessments were not calculating the level of site
contamination as recommended.  OSWER officials used this information
to identify which areas of guidance risk assessments generally were
not following. 

OSWER gathered information on all four steps of the risk assessment
process:  collecting and analyzing data, estimating the extent of
exposure, assessing the toxicity of contaminants, and calculating the
resultant risk.  For each risk assessment reviewed, extensive data on
how the above steps were conducted were entered into a worksheet. 
OSWER staff then rolled up the data from each worksheet into a
package of summary information describing the characteristics of each
region's risk assessments for the year.  For example, the packages
included tables identifying which sources of contamination and land
uses were evaluated in the past year's risk assessments. 

However, our review of the summary reports and our discussions with
regional officials indicated that OSWER neither analyzed the
information nor communicated the results in ways that help risk
assessment teams improve the consistency of their work.  For example,
OSWER did not seek to determine whether different absorption rates
were used for the same contaminants even though its quality assurance
review worksheets contained information on the absorption rates used
in risk assessments.  Instead, most of the data generated were simply
frequencies, expressing, for instance, how often risk assessments
evaluated contaminants absorbed through the skin or selected
residential as the land use for the future. 

EPA regional officials told us that while the descriptive information
in the quality assurance reports was interesting, they would find it
more useful to receive critical feedback on individual risk
assessments--specifically, whether their individual risk assessments
were consistent with other risk assessments.  OSWER officials told us
that the quality assurance reviews were originally designed to
describe the results of the regions' risk assessments and to help the
regions to follow the agency's guidance more closely.  However, they
noted that since most of the data needed to analyze risk assessments
for consistency is already gathered annually, it would not be
difficult to add this type of analysis as part of the quality
assurance reviews.  OSWER officials acknowledged that if they had
analyzed the assessments in this way, their quality assurance reviews
could have uncovered the kind of inconsistent approaches to risk
assessments we identified in our review.  In addition, OSWER
officials told us that by including measures of consistency in
headquarters' reports to the regions, they could have emphasized the
need for such consistency and pointed out specific places where it
was lacking. 

OSWER officials pointed out that in addition to disseminating the
descriptive information in the quality assurance reviews, they have
developed some ad hoc procedures for sharing information on risk
assessments.  For example, regions participate in a monthly telephone
conference, in which they discuss specific policies on risk
assessment and their implementation.  EPA also organizes work groups
on particular problems with risk assessments, such as the one formed
to consider dermal exposure to contaminated soil.  OSWER officials
acknowledged, however, that they could promote better compliance and
consistency by supplementing these discussions with improved
information from the annual quality assurance reviews. 


   CONCLUSIONS
------------------------------------------------------------ Letter :6

On the whole, the 20 risk assessments we reviewed adhered to EPA's
guidance and were relatively consistent with one another--both in
following the guidance and showing a high degree of consistency in
most areas we reviewed.  However, our review uncovered several areas
in which clearer guidance could improve the consistency of risk
assessments.  For example, we found that EPA guidance permitted--and
risk assessment teams used--different approaches to determine how
people could absorb hazardous contaminants through their skin and
describe the level of precision in risk estimates.  To the extent
that risk assessments are used to make cleanup decisions, this could
mean that one site might be cleaned up and another similar site might
be left alone only because of differences in how the assessments were
done rather than actual differences between the sites. 

EPA has acknowledged many of the gaps we found in its guidance and
should be commended for its ongoing effort to give more specific
direction rather than requiring risk assessment teams to improvise
when data are lacking.  In particular, EPA's plans to provide better
guidance on estimating future land use, dermal exposure to
contaminated soil, and the rates at which contaminants are absorbed
into the bloodstream will help to produce more consistent risk
assessment methods. 

Despite these improvements, EPA headquarters staff acknowledged that
the agency could make better use of its reviews of how its regions
conduct risk assessments.  Although EPA has in place an extensive and
detailed quality assurance and monitoring system, that system does
not analyze risk assessments to determine whether they are consistent
with one another.  Analyzing for consistency would not require
significantly more resources given that the annual reviews already
gather most of the necessary data.  Furthermore, such analysis would
help EPA identify additional areas in which guidance or research is
needed to increase the consistency of risk assessments.  It would
also serve as the basis for feedback to regions that would better
meet regional staff's needs. 


   RECOMMENDATION
------------------------------------------------------------ Letter :7

We recommend that the Administrator of EPA ensure that the agency's
quality assurance reviews be used to identify and reduce
inconsistencies in risk assessments by

  analyzing individual risk assessments to determine whether they are
     consistent with national practices and

  providing the results of this analysis to the regions. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :8

As requested, we did not provide a draft of this report to EPA
officials for written comment.  However, we met with EPA officials,
including the Deputy Director of the Hazardous Site Evaluation
Division, representatives from OSWER and the Office of Research and
Development, and regional staff and discussed with them the facts in
this report.  They agreed with the information provided, and we have
incorporated their comments, suggestions, and responses where
appropriate throughout the report.  We also have noted various
improvements the agency plans to make, including providing better
guidance on determining land use and evaluating dermal exposure to
contaminants, and the agency's acknowledgment that it could use the
results of quality assurance reviews to promote greater consistency
among risk assessments. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :9

To obtain information on each of the three issues raised in the
request, we reviewed 20 risk assessments approved in fiscal year
1992.  (See app.  VII for a list of the sites.) These selections
covered not only each of EPA's 10 regions, but also the major types
of Superfund sites (e.g., mining, landfill, and industrial).  We also
reviewed headquarters' and regions' risk assessment guidance.  Our
review focused primarily on the way EPA assesses human exposure to
contamination, and calculates and characterizes the risks sites pose. 
In these areas, we identified nine major requirements and evaluated
whether risk assessments followed them.  We also summarized and
compared the information risk assessments used in estimating exposure
to hazardous contaminants to determine whether they used similar
values and methods.  In addition to reviewing risk assessments, we
interviewed EPA officials, such as project managers and
toxicologists, and private parties responsible for cleaning up the
sites we reviewed.  We evaluated EPA's quality assurance process by
reviewing the available worksheets for our risk assessment case
studies and the summary reports and by interviewing the OSWER
officials who designed the reviews and the regional staff who use
them.  We conducted our review from September 1993 to April 1994 in
accordance with generally accepted government auditing standards. 


---------------------------------------------------------- Letter :9.1

As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days
from the date of this letter.  At that time, we will send copies of
the report to the Administrator of EPA.  We will also make copies
available to others on request. 


Please contact me on (202) 512-6112 if you or your staff have any
questions.  Major contributors to this report are listed in appendix
VIII. 

Peter F.  Guerrero
Director, Environmental
Protection Issues


STEPS IN A RISK ASSESSMENT AND
WAYS PEOPLE CAN BE EXPOSED TO
HAZARDOUS CONTAMINANTS
=========================================================== Appendix I

   Figure I.1:  Steps in a Risk
   Assessment

   (See figure in printed
   edition.)

\a "Media" refers to specific parts of the environment, including
groundwater, soil, and air. 

   Source:  The Environmental
   Protection Agency's (EPA) risk
   assessment guidance.

   (See figure in printed
   edition.)

   Figure I.2:  Ways People Can Be
   Exposed to Hazardous
   Contaminants

   (See figure in printed
   edition.)

   Source:  EPA's risk assessment
   guidance.

   (See figure in printed
   edition.)


INTAKE RATES FOR DRINKING
GROUNDWATER AND SWALLOWING SOIL
========================================================== Appendix II

In calculating exposure to hazardous waste, risk assessments estimate
the amount of contamination people take in each day (the intake
rate).  Part of this calculation requires determining how much soil
or groundwater or air people consume daily.  To analyze the specific
data the Environmental Protection Agency (EPA) used to measure
exposure to contaminants and resulting risk, we reviewed each
exposure pathway in our 20 risk assessment case studies.  "Exposure
pathway" refers to the specific way someone contacts contaminants.  A
pathway is defined by

  how the land is used (whether the use is residential, industrial,
     etc.);

  what the source of the contamination is (groundwater, soil, etc.);

  whether the exposure is now or in the future;

  how old the exposed person is (whether the person is an adult or
     child); and

  how the contact occurs (touching, swallowing, etc.). 

One pathway might be defined by an adult resident of the site
drinking groundwater in the future.  We identified 357 separate
pathways in the 20 risk assessments we reviewed.  Risk assessments
varied greatly in the number of pathways they evaluated--ranging from
3 to 48 pathways for a single site. 

EPA recommends that risk assessment teams assume that residents (both
adults and children) on or near a Superfund site consume 2 liters of
groundwater per day.  EPA's guidance also recognizes that people
inadvertently consume a small amount of soil during the course of the
day and recommends that the risk assessment teams assume that adults
swallow about 100 milligrams (mg) of contaminated soil per day.  (The
typical aspirin tablet contains 325 mg of aspirin.) Because
children's activities bring them into contact with soil more
frequently, EPA recommends a higher value of 200 milligrams of soil
per day for them.  Table II.1 shows that although risk assessment
teams may develop their own estimates on the basis of information
about the site, most of the risk assessments in our review used the
recommendations in EPA's guidance. 



                          Table II.1
           
            Intake Rates Used in Risk Assessments
            Compared With EPA's Recommended Rates


                                          Higher       Lower
Exposure pathway,         Recommende        than        than
population, and                    d  recommende  recommende
recommended level              level     d level     d level
------------------------  ----------  ----------  ----------
Drinking groundwater
Adult (2 liters)                  23           0           0
Swallowing soil
Adult (100 mg)                    11           1           1
Child (200 mg)                     8           0           4
============================================================
Total                             42           1           5
------------------------------------------------------------

EXPOSURE FREQUENCY
========================================================= Appendix III

To help the risk assessment teams measure exposure accurately, EPA
provides recommendations on the number of days per year that someone
could be exposed to waste.  According to EPA, residents should be
expected to spend 350 days at home each year, and workers 250 days
per year at their jobs.  Although risk assessment teams may develop
their own estimates on the basis of information about the site, we
found that more than half of the assessments used the recommendations
in EPA's guidance. 



                         Table III.1
           
           Number of Days Used to Estimate Exposure
           in Risk Assessments Compared With EPA's
                      Recommended Number


                                     More than    Fewer than
Land use and         Recommended   recommended   recommended
recommended number     number of     number of     number of
of days                     days          days          days
------------------  ------------  ------------  ------------
Residential                  113            33            52
 (350 days)
Industrial                    40             0            14
 (250 days)
============================================================
Total                        153            33            66
------------------------------------------------------------

LAND USES
========================================================== Appendix IV

EPA's guidance directs the risk assessment team to determine how land
at and near a hazardous waste can be used currently and in the
future.  Land use largely determines how much people are exposed to
waste--residential use of land generally leads to the highest
exposure.  Figure IV.1 illustrates that future residential use
predominated in the exposure pathways.  Of all pathways evaluated in
the 20 risk assessments we reviewed, 56 percent were for future
residential use of the site. 

   Figure IV.1:  Number of
   Exposure Pathways Evaluated in
   Risk Assessments, by Current
   and Future Land Uses

   (See figure in printed
   edition.)

Note:  In some cases, risk assessments assumed land would be used in
the same way now as in the future.  These pathways are categorized as
"both" current and future exposures. 


LAND USE AND ASSOCIATED RISKS
=========================================================== Appendix V

We found that residential and industrial settings produced the
highest proportion of exposure pathways with risk estimates exceeding
the level at which EPA requires cleanup action.  This was true for
both cancer and non-cancer risks.  Fifty-three of 121 residential
exposure pathways had cancer risks in excess of 1 in 10,000, and 51
of 126 pathways had hazard indexes greater than 1.  (See figs.  V.1
and V.2).  Residential and industrial land uses assume more frequent
exposure to waste than other kinds of land use. 

   Figure V.1:  Number of Exposure
   Pathways With Cancer Risks
   Exceeding EPA's Criterion, by
   Land Use

   (See figure in printed
   edition.)

Note:  This figure includes both current and future land uses. 

   Figure V.2:  Number of Exposure
   Pathways With Non-Cancer Risks
   Exceeding EPA's Criterion, by
   Land Use

   (See figure in printed
   edition.)

Note:  This figure includes both current and future land uses. 


RISKS ASSOCIATED WITH CONTAMINATED
MEDIA
========================================================== Appendix VI

EPA requires cleanup if the chance of developing cancer exceeds 1 in
10,000 for a pathway or combination of pathways.  For non-cancer
health problems, EPA requires cleanup if the calculated "hazard
index" exceeds a value of 1.  In the 20 risk assessments we reviewed,
of the exposure pathways that included an estimated risk, 41 percent
had cancer risks and 42 percent had non-cancer risks requiring
cleanup under EPA's policy.  Risks exceeding EPA's criteria most
commonly resulted from groundwater and soil ingestion, particularly
for future exposures.  (See tables VI.1 and VI.2.)



                          Table VI.1
           
            Cancer Risks Associated With Selected
                      Exposure Pathways



Media and exposure route         Yes      No     Yes      No
----------------------------  ------  ------  ------  ------
Groundwater
------------------------------------------------------------
Ingestion                          4       4      18       6
Dermal                             2       4       5       6
Inhalation                         0       5       3       7

Soil
------------------------------------------------------------
Ingestion                          4       9      13      10
Dermal                             4      17      12      12
Inhalation                         0       3       2       5

Air
------------------------------------------------------------
Inhalation                         9      15       5       9
============================================================
Total\b                           23      57      58      55
------------------------------------------------------------
Note:  Tables VI.1 and VI.2 include some of the same pathways.  When
site contaminants were thought to cause both cancer and other health
problems, exposure pathways included both types of risks. 

\a In some cases, risk assessments assumed that land would be used in
the same way in the future as it is now.  We included these in the
"current" column. 

\b This table does not include all 357 pathways in the risk
assessments we reviewed because (1) some pathways did not evaluate
cancer risk and (2) we did not include infrequent pathways, such as
eating produce grown in contaminated soil. 



                          Table VI.2
           
             Hazard Index Levels Associated With
                  Selected Exposure Pathways



Media and exposure route         Yes      No     Yes      No
----------------------------  ------  ------  ------  ------
Groundwater
Ingestion                          6       2      23       2
Dermal                             1       5       3       8
Inhalation                         1       4       3       4

Soil
------------------------------------------------------------
Ingestion                          4       9       9      14
Dermal                             4      15       4      18
Inhalation                         1       2       1       1

Air
------------------------------------------------------------
Inhalation                         6      12       7       5
============================================================
Total\b                           23      49      50      52
------------------------------------------------------------
Note:  Tables VI.1 and VI.2 include some of the same pathways.  When
site contaminants were thought to cause both cancer and other health
problems, exposure pathways included both types of risks. 

\a In some cases, risk assessments assumed that land would be used in
the same way in the future as it is now.  We included these in the
"current" column. 

\b This table does not include all 357 pathways in the risk
assessments we reviewed because (1) some pathways did not evaluate
non-cancer health risk and (2) we did not include infrequent
pathways, such as eating produce grown in contaminated soil. 


LIST OF CASE STUDY SITES
========================================================= Appendix VII


      REGION I
----------------------------------------------------- Appendix VII:0.1

PSC Resources, Massachusetts
Revere Textile Prints, Connecticut


      REGION II
----------------------------------------------------- Appendix VII:0.2

Endicott Village Wellfield, New York
Witco Chemical Corp., New Jersey


      REGION III
----------------------------------------------------- Appendix VII:0.3

Atlantic Wood Industries, Virginia
Strasburg Landfill, Pennsylvania


      REGION IV
----------------------------------------------------- Appendix VII:0.4

Chem-form Inc., Florida
Hercules, Inc.  009 Landfill, Georgia


      REGION V
----------------------------------------------------- Appendix VII:0.5

Torch Lake, Michigan
Woodstock Municipal Landfill, Illinois


      REGION VI
----------------------------------------------------- Appendix VII:0.6

American Creosote Works, Louisiana
Fourth Street Refinery, Oklahoma


      REGION VII
----------------------------------------------------- Appendix VII:0.7

Farmer's Mutual Cooperative, Iowa
Red Oak City Landfill, Iowa


      REGION VIII
----------------------------------------------------- Appendix VII:0.8

Eagle Mine, Colorado
Idaho Pole Company, Montana


      REGION IX
----------------------------------------------------- Appendix VII:0.9

Pacific Coast Pipeline, California
San Gabriel Valley, California


      REGION X
---------------------------------------------------- Appendix VII:0.10

Commencement Bay/Nearshore Tideflats, Washington
Joseph Forest Products, Oregon


MAJOR CONTRIBUTORS TO THIS REPORT
======================================================== Appendix VIII

RESOURCES, COMMUNITY, AND ECONOMIC
DEVELOPMENT DIVISION, WASHINGTON,
D.C. 

Bernice Steinhardt, Associate Director
Stanley J.  Czerwinski, Assistant Director
Sharon Butler, Senior Evaluator
Stephen M.  Cleary, Staff Evaluator
Fran Featherston, Senior Social Science Analyst
John H.  Skeen, III, Managing Editor

CHICAGO REGIONAL OFFICE

Jim Musial, Regional Management Representative
Katherine Siggerud, Evaluator-in-Charge
Paul J.  Schmidt, Senior Evaluator
Melvin Rodriguez, Staff Evaluator
John Zarem, Computer Programmer Analyst
LaKale Williams, Technical Adviser
Rimini Butler, Support Services Technician

