\
UK,1
\
Office of Inspector General
AUDIT REPORT
PROGRAM ENHANCEMENTS WOULD
ACCELERATE SUPERFUND SITE
ASSESSMENT AND CLEANUP
E1SFF3-08-0021 -4100180
JANUARY 31,1994
-------
Inspector General Division
Conducting the Audit:
Regions Covered:
Central Audit Division
Kansas City, Kansas
Regions 1 through 10
Headquarters
Offices Reviewed:
Office of Emergency and
Remedial Response (Superfund)
-------
**
\ UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. O.C. 20460
January 31, 1994 OFFICE OF
THE INSPECTOR GENERAL
MEMORANDUM
SUBJECT: Program Enhancements Would Accelerate Superfund Site
Assessment and Cleanup
Report Number E1SFF3-08-0021-4100180
FROM: Michael Simmons (\»\l<**>«».«\ SOT^^-T^^CV^-—
Associate Assistant Inspector General
for Internal and Performance Audits
TO: Elliott P. Laws
Assistant Administrator for Solid Waste
and Emergency Response
Attached is our report entitled "Program Enhancements Would
Accelerate the Superfund Site Assessment Process and Cleanup".
The report contains recommendations on site inspection
prioritization, Congressional disclosure, and site deferral.
Your staff has implemented some improvements and plans more
changes to address the issues we identified. The objectives of
our audit were to:
(1) evaluate EPA's policies and procedures for evaluating
sites,
t
(2) determine the current status of sites in the site
assessment process, and
(3) evaluate proposed deferral policies.
Action Required
We have designated you as the Action Official for this
report. In accordance with EPA Order 2750, you are reguested to
provide this office a written response to the audit report within
90 days of the report date. Your response should address all
recommendations. For corrective actions planned but not
completed by the report date, reference to specific milestone
dates will assist us in deciding whether to close the report.
This audit report contains findings that describe problems
the Office of Inspector General has identified and corrective
actions it recommends. This audit report represents the opinion
of OIG. Final determinations on matters in this audit report
Recycted/B*cyaab«o
Pnmta win SovC^nc** "•• >• —aw :na:
contain* «l MJM bO>. ••CYCMC '>cw
-------
will be made by EPA managers in accordance with established EPA
audit resolution procedures. Accordingly, the findings described
in this audit report do not necessarily represent the final EPA
position.
We appreciate the positive attitude exhibited by your staff
during the audit. Their willingness to implement changes to the
site assessment process and to accept input on the deferral
process should positively impact the Superfund program.
We have no objections to the release of this report to the
public.
If you or your staff have any questions, please contact
Nikki Tinsley, Divisional Inspector General, at (913) 551-7824 or
Jeffrey S. Hart, Audit Manager, at (303) 294-7520.
Attachment
-------
EXECUTIVE SUMMARY
PURPOSE
Between 1980 and June 1993, the Superfund site assessment program
evaluated over 23,000 potential Superfund sites. In fiscal 1993,
Environmental Protection Agency (EPA) site assessment officials
budgeted over $63.9 million to evaluate sites to determine which
sites qualify for listing on the National Priorities List (NPL)--
a list of the nation's worst hazardous waste sites. Yet, prior
audits have found that EPA was not meeting legally mandated site
assessment deadlines. As a result, some sites that qualify for
listing on the NPL are still not fully evaluated. Through site
assessment activities, EPA implements an important Superfund
program goal, to ensure that the worst sites are cleaned up
first. Superfund reauthorization in fiscal 1994 presents an
opportunity to seek congressional approval for program
improvements. Our audit objectives were to: (1) evaluate EPA's
policies and procedures for evaluating sites, (2) determine the
current status of sites in the site assessment process, and (3)
evaluate proposed deferral policies.
BACKGROUND
The Comprehensive Environmental Response, Compensation, and
Liability Act of 1980 (CERCLA) created EPA's Superfund program to
respond to releases of and threatened releases of hazardous
substances. CERCLA Section 105 required that EPA, through the
National Contingency Plan (NCP), identify methods and procedures
to carry out the provisions of the Act. Among other provisions,
the NCP contains methods for discovering and investigating sites
that may be a threat to human health and the environment and
requires that EPA identify which sites are a priority for Federal
Superfund cleanup. Since 1980, Congress has provided $15.2
billion in spending authority to carry out the provisions of
CERCLA, as amended.
The site assessment program is the primary mechanism for
determining which sites qualify for the NPL. NPL sites gualify
for Federal participation in cleanup activities. Accurate,
timely site assessment is the foundation of the remaining steps
in the Superfund cleanup process. As a site moves through the
site assessment process, EPA refines its knowledge of the extent
and nature of contamination at the site. The final step in site
i
Report No. E1SFF3-08-0021-4100180
-------
assessment is preparing Hazardous Ranking System (HRS) packages
necessary to list a site on the NPL, deferring the site to
another Federal authority, or determining that the site does not
gualify for further Federal involvement.
The Office of Emergency and Remedial Response (OERR), through its
Hazardous Site Evaluation Division (HSED), implemented Site
Inspection Prioritization (SIP) in October 1991 to reevaluate a
backlog of 6,467 sites that had been evaluated, but NPL listing
decisions had not been made. SIP is estimated to cost over $70
million. The backlog existed primarily because site assessment
managers felt it would be preferable to make final listing
decisions after revising the HRS model. The Superfund Amendments
and Reauthorization Act of 1986 (SARA) significantly changed the
site assessment criteria. From 1986 through 1991 EPA evaluated
sites but was reluctant to recommend listing sites on the NPL.
RESULTS IN BRIEF
Since October 1991, regions reevaluated 942 sites in the 6,467
SIP backlog. About 56 percent of the 942 sites were eliminated
from the Superfund program, 16 percent were classified as a high
priority, and 27 percent were classified as a low priority for
further Federal involvement. The next step for high and low
priority sites is evaluation for listing on the NPL.
Better SIP implementation and management could have improved
program effectiveness and efficiency. HSED managers implemented
SIP before they developed written national program guidance.
Without guidance, regions approached SIP differently. As a
result, information about program results was incomplete and
accounted for inconsistently and HSED cannot use program results
information to assess national accomplishments.
Upon SIP completion, EPA could have a backlog of over 3,100 sites
to list on the NPL. At the current rate, it would take EPA 27
years to list these sites. Although the Superfund Accelerated
Cleanup Model may accelerate the cleanup process, it will have
little impact on the site listing process. Implementing a well
designed deferral policy could significantly reduce the backlog
of sites needing cleanup and help EPA meet CERCLA site assessment
goals. Deferral allows States, potentially responsible parties,
and other Federal authorities to cleanup potential Superfund
sites and effectively removes the site from the Federal
inventory.
ii
Report No. E1SFF3-08-0021-4100180
-------
PRINCIPAL FINDINGS
Regions conducted SIPs for nearly 2 years without guidance.
Without guidance, regions implemented SIP inconsistently and HSED
cannot be confident that an important site assessment goal, to
assess the "worst sites first," was accomplished. SIP guidance
issued in August 1993 focuses on worst sites first, but does not
provide a planned approach with measurable program goals.
Regions inconsistently recorded completed SIPs in the
Comprehensive Environmental Response, Compensation, and Liability
Information System (CERCLIS) which caused management information
to be unreliable. HSED's management control documentation did
not include adequate control techniques to ensure national
consistency in program implementation. According to Headquarters
program management officials, other higher priorities also
adversely impacted SIP. Better SIP planning could have ensured
more consistent regional program implementation and more
meaningful program results information.
Despite its efforts to accelerate site assessment and cleanup,
EPA has not met CERCLA goals to list sites on the NPL within 4
years of site discovery. When EPA completes the SIP program it
is expected to have identified over 3,100 sites, of the 6,467 in
the SIP backlog, that will require HRS packages. At EPA's
current budget rate, it will take over 27 years and about $105
million to make NPL decisions on these sites. EPA may have to
reevaluate sites again that remain in the backlog for an extended
time period. Based on past EPA per-site cleanup costs, we
estimated it will cost about $74.5 billion to cleanup the 3,100
sites that will likely be listed. EPA does not currently include
this cost estimate in its annual report to Congress. Not knowing
the magnitude of the hazardous waste site cleanup problem
adversely impacts the public, Congress, and parties responsible
for site cleanups. Although EPA has developed initiatives to
improve the Superfund process, those initiatives will have
minimal impact on the potential NPL backlog.
Adopting an expanded site deferral policy could reduce the
potential NPL backlog, conserve site assessment resources,
accelerate site cleanups, and help EPA meet the CERCLA mandated
site assessment timeliness goals. EPA considered adopting a
broad deferral policy in the past, but did not do so primarily
because of congressional concern. Comments on EPA's prior
deferral proposal were mixed. EPA began a State deferral pilot
program in September 1993. An even broader deferral policy could
iii
Report No. E1SFF3-08-0021-4100180
-------
permit other Federal authorities and/or responsible parties to
perform CERCLA-quality hazardous waste site cleanups on sites
awaiting NPL listing. As a result, the NPL could serve as a
better management tool in setting priorities for addressing sites
that require Federal involvement.
RECOMMENDATIONS
OERR should ensure that regions reevaluate high priority sites
first. It should ensure that regions consistently account for
accomplishments, conduct SIPs only on sites that need them, and
implement national programs consistently. OERR should fully
disclose to Congress the estimated cost to cleanup sites that,
upon completion of SIP, will likely be proposed for NPL listing.
EPA should respond to concerns expressed by various parties about
deferral and take steps to implement a carefully designed
deferral policy that would involve governments, industry, and
affected parties in the cleanup process.
EPA COMMENTS AND OFFICE OF INSPECTOR
GENERAL EVALUATION
The Office of Solid Waste and Emergency Response (OSWER)
generally agreed with the results of our audit and has
implemented or plans to implement policies and procedures to
address the issues we raised. The only recommendation OSWER
questioned concerned, disclosing to Congress the estimated cost of
cleaning up sites. We changed our recommendation to include only
those sites with completed SIPs that are likely to be listed on
the NPL. We resolved some issues through discussions with OSWER
officials and added new information to the report. The positive
attitude OSWER managers showed during our audit and their
willingness to explore ways to improve the site assessment
process should have a positive impact on site assessment and
improve management in general.
IV
Report No. E1SFF3-08-0021-4100180
-------
TABLE OF CONTENTS
Page
EXECUTIVE SUMMARY i
CHAPTERS
1 INTRODUCTION 1
PURPOSE 1
BACKGROUND 1
SCOPE AND METHODOLOGY 7
PRIOR AUDIT COVERAGE 8
2 SIP CAN BE IMPROVED 9
LACK OF NATIONAL SIP GUIDANCE RESULTED IN
INCONSISTENT SIP IMPLEMENTATION .... 9
SIP GUIDANCE COULD BE IMPROVED 14
ADDITIONAL CONTROL TECHNIQUES MAY ENSURE
NATIONAL CONSISTENCY 14
CONCLUSIONS 15
RECOMMENDATIONS , 15
EPA COMMENTS AND OIG EVALUATION 16
3 UNREPORTED COSTS TO CLEANUP POTENTIAL NPL SITES MAY
EXCEED $74.5 BILLION 17
POTENTIAL NPL LISTING WILL TAKE $105 MILLION AND 27
YEARS TO ELIMINATE 17
NOT MEETING CERCLA TIMELINESS GOALS ADVERSELY IMPACTS
THE SUPERFUND PROGRAM 19
EPA'S REPORTED COSTS TO CLEANUP POTENTIAL NPL
SITES IS UNDERSTATED 20
v
Report No. E1SFF3-08-0021-4100180
-------
POTENTIAL NPL BACKLOG IMPACTS THE PUBLIC, CONGRESS,
AND RESPONSIBLE PARTIES 21
NEW INITIATIVES WILL NOT ELIMINATE NPL BACKLOG . . 22
CONCLUSIONS 23
RECOMMENDATIONS 23
EPA COMMENTS AND OIG EVALUATION 23
4 AN EXPANDED DEFERRAL POLICY MAY HELP REDUCE THE POTENTIAL
NPL BACKLOG 25
EPA'S PRIOR ATTEMPT TO ADOPT AN EXPANDED DEFERRAL
POLICY WAS UNSUCCESSFUL 25
EPA PROPOSES LIMITED DEFERRAL 27
A WELL DESIGNED AND IMPLEMENTED DEFERRAL POLICY WILL
HELP ENSURE PROGRAM SUCCESS 28
CONCLUSIONS 29
RECOMMENDATIONS 30
EPA COMMENTS AND OIG EVALUATION 30
EXHIBITS
EXHIBIT 1.1: SITE ASSESSMENT PROCESS 3
EXHIBIT 1.2: SIP BACKLOG AS OF JUNE 30, 1993 5
EXHIBIT 1.3: SIP PROGRAM RESULTS (OCTOBER 1991 -
APRIL 1993) 6
EXHIBIT 2.1: NUMBER OF SIPS NOT IDENTIFIED BY HSED
(OCTOBER 1991 - APRIL 1993) 13
EXHIBIT 3.1: ESTIMATE OF POTENTIAL NPL SITES IN THE SIP
BACKLOG 19
APPENDIXES
APPENDIX I: EPA COMMENTS 32
vi
Report No. E1SFF3-08-0021-4100180
-------
APPENDIX II: ABBREVIATIONS 45
APPENDIX III: DISTRIBUTION 46
Vll
Report NO. E1SFF3-08-0021-4100180
-------
CHAPTER 1
INTRODUCTION
PURPOSE
Between the enactment of the Comprehensive Environmental
Response, Compensation, and Liability Act of 1980 (CERCLA)
and June 30, 1993, over 23,000 sites were evaluated through
the Superfund site assessment program. The fiscal 1993 site
assessment budget totaled over $63.9 million. The site
assessment program is the Environmental Protection Agency's
(EPA) primary mechanism for determining which sites qualify
for the National Priorities List (NPL), the list of the
nation's most serious hazardous waste sites. Prior audits
found that EPA had not met site assessment goals and, as a
result, potential NPL sites were not fully evaluated.
EPA's policy is to cleanup the worst NPL sites first. Timely
identification of the worst sites is critical to the
Superfund process. Superfund program expiration on
September 30, 1994, and its subsequent reauthorization
presents an opportunity to increase the effectiveness of the
site assessment process. Accordingly, our objectives were
to:
— evaluate EPA's policy and procedures for
evaluating sites for inclusion on the NPL,
— determine the current status of EPA site
assessment activities, and
— evaluate proposed policies regarding site
deferral to States.
BACKGROUND
Congress enacted CERCLA establishing the Superfund program to
respond to releases and threatened releases of hazardous
substances. CERCLA required EPA to establish criteria for
determining priorities among releases of hazardous
substances. CERCLA authorized a 5-year, $1.6 billion trust
fund to pay for Federal cleanup of sites. CERCLA was revised
and expanded by the Superfund Amendments and Reauthorization
1
Report No. E1SFF3-08-0021-4100180
-------
Act of 1986 (SARA). SARA reauthorized Superfund for another
5 years and provided an additional $8.5 billion. The
reauthorization required a more accurate assessment of the
risk to human health and the environment sites pose through a
revised Hazardous Ranking System (HRS) model. More recently,
the Omnibus Budget Reconciliation Act of 1990 extended the
Superfund program through September 1994 and increased
funding by $5.1 billion to a total of $15.2 billion in
spending authority.
CERCLA Section 105 required a National Contingency Plan (NCP)
to establish procedures and standards for responding to the
release of hazardous substances. SARA required the NCP to be
revised to include methods for discovering and investigating
sites where hazardous substances are present and criteria for
determining priorities among hazardous substance releases for
taking removal and remedial actions.
EPA can cleanup sites through removal and remedial actions.
Removal actions are short-term actions which stabilize or
cleanup a hazardous site that poses a threat to human health
and the environment. Remedial actions are longer term and
usually more expensive actions at sites listed on the NPL.
EPA identifies NPL sites through the site assessment process.
Primary objectives of the site assessment process are to
evaluate sites to determine if they qualify for listing on
the NPL and ensure that EPA carries out its policy to cleanup
worst sites first. ( sites are discovered by regional EPA
offices, State agencies, and citizens who file a petition.
Once discovered, sites are entered into the Comprehensive
Environmental Response, Compensation, and Liability
Information System (CERCLIS), EPA's computerized inventory of
potential hazardous waste sites. CERCLIS contains a
historical record of all events that have occurred on a site
under the Superfund program. EPA then evaluates the
potential for release from a site through a series of steps.
At the end of each step, EPA reviews the results to determine
whether to further evaluate the site. Those sites that
require further evaluation are prioritized as either an "H" -
high priority for listing on the NPL or "L" - low priority
for listing on the NPL. Sites that do not require further
Superfund evaluation are classified as "N" - no further
remedial action planned/site evaluation accomplished, or "D"
- deferred. Exhibit 1.1 depicts the site assessment process;
a discussion of the steps follows:
2
Report No. E1SFF3-08-0021-4100180
-------
EXHIBIT 1.1
SITE ASSESSMENT PROCESS
— Preliminary Assessment
(PA); PA is a limited-scope
investigation performed by
States and/or EPA at every
CERCLIS site. Readily
available information is
obtained from the site and
existing site information is
reviewed. PA is designed to
distinguish between sites
that pose little or no
threat to human health and
the environment, and sites
that require further
investigation. PA also
identifies sites requiring
assessment for possible
emergency response actions.
— Site Inspection (SI); SI
is a more detailed site
investigation, typically
including the collection of
waste and environmental
samples, to identify sites
with a high probability of
qualifying for the NPL and
to identify sites posing
immediate health or
environmental threats which
require emergency response.
— Site Inspection
Prioritization (SIP); SIP
established a "temporary"
intermediate step in the
site assessment process to
update Sis that were
conducted under the pre-SARA
HRS model. SIP examines
backlogged sites to gather
additional information and
set priorities among sites
Report No. E1SFF3-08-0021-4100180
-------
for further investigation or to eliminate sites from further
investigation.
-- Expanded Site Inspection (ESI): ESI includes additional
field activities such as monitoring well installation, air
sampling, geophysical studies, drum or tank sampling, and
complex background sampling studies to collect all data
necessary to prepare the HRS scoring package.
— HRS Package Preparation: The HRS process calculates a
score based on information collected during the PA, SI, SIP,
and ESI which is used to prepare an HRS package. The score
is based on several individual factors for each site that are
combined mathematically. Any site with an HRS score of 28.50
or greater is eligible for the NPL.
— NPL Listing; EPA places sites on the NPL through proposed
rulemaking in the Federal Register. EPA reviews public
comments and modifies site scores, when appropriate, if new
information is made available. Sites with scores remaining
above 28.50 are published as a final rule in the Federal
Register and added to the NPL. EPA does not disclose
Superfund cleanup cost estimates to Congress and the public
until sites are listed on the NPL.
In some cases, EPA may defer a site during the site
assessment process. EPA may defer sites to other Federal
authorities where that authority has the means to ensure the
site is cleaned up. Deferring a site permits cleanup without
listing the site on the NPL and without the expenditure of
Superfund money. Currently, EPA defers sites to the Resource
Conservation and Recovery Act program and to the Nuclear
Regulatory Commission.
EPA has developed new Superfund initiatives: the Superfund
Regional Pilots project and the Superfund Administrative
Improvements initiative. Under the Superfund Regional Pilots
project, regions have implemented pilot projects to
streamline and accelerate Superfund response activities. For
example, the Superfund Accelerated Cleanup Model (SACM) is
one initiative being tested under the Superfund Regional
Pilots project. The Superfund Administrative Improvements
initiative considers deferral of sites to States, which would
help eliminate sites from the potential NPL backlog.
CERCLIS indicates that EPA's Superfund program has identified
4
Report No. E1SFF3-08-0021-4100180
-------
EXHIBIT 1.2
SIP BACKLOG
AS OF JUNE 30, 1993
over 37,000 sites and determined that about 22,000 sites are
not eligible for further Federal funding. Over 1,200 have
been listed on the NPL—indicating those sites are among the
nation's worst hazardous waste sites and require Federal
participation in remedial cleanup activities. The remaining
15,400 sites have not been fully evaluated; about 6,400 have
had Sis completed and need NPL listing decisions. These
sites make up the SIP program backlog.
SIP Responds To HRS Revisions
In October 1991, EPA's Office of Emergency and Remedial
Response (OERR), Hazardous Site Evaluation Division (HSED)
implemented SIP as a temporary step in the site assessment
process to ensure site assessment decisions were based on the
revised HRS model and to make screening decisions on sites
using minimal resources. HSED identified a backlog of 6,467
sites that were evaluated prior to September 1992—the date
it issued final SI
guidance. These
sites were
evaluated through
the SI stage under
the original HRS
model. EPA
recognized that to
ensure these sites
complied with the
revised HRS model,,
some of the 6,467
sites would require
additional data
collection. Using
HSED's estimates,
SIP will cost over
$70 million (6,467
sites x 190
hours/site x
$57/hour).
The backlog grew
primarily because
EPA Site Assessment
officials felt it
would be preferable
to make final site
NUMBER OF SITES
567
REGIONS
10
Report No. E1SFF3-08-0021-4100180
-------
decisions after revising the HRS model. HSED allocated
approximately 20 percent of its fiscal 1992, and 24 percent
of its fiscal 1993 site assessment budget for SIP. Exhibit
1.2 on page 5 illustrates the regional SIP backlog. '
SIP Reevaluated And Prioritized Sites
Between October 1991 and April 1993, site assessment
officials reevaluated 942 sites using SIP. Of the 942 sites,
526 (56 percent) were eliminated from further Superfund
program involvement, 258 (27 percent) were classified as a
low priority for further investigation, 153 (16 percent) were
classified as a high priority for NPL listing, and 5 sites (1
percent) were not prioritized. Exhibit 1.3 illustrates the
results of SIP to date.
EXHIBIT 1.3
SIP RESULTS
(OCTOBER 1991-APRIL 1993)
No Further Action
526 56%
Low Priority
258 27%
Not Prioritized
5 1%
High Priority
153 16%
1 Information for this exhibit was taken from an HSED
report dated June 30, 1993, and totals 6,397 sites. According to
an HSED analysis of sites with Sis completed prior to July 31,
1992, there are a total of 6,467 sites in the SIP universe.
Report NO. E1SFF3-08-0021-4100180
-------
SCOPE AND METHODOLOGY
We performed our fieldwork from October 1992 through
September 1993. We interviewed site assessment program
representatives in HSED and all 10 EPA regions to obtain
information regarding the site assessment process. We
reviewed general policies and procedures used by HSED and the
regions for evaluating sites for inclusion on the NPL. In
addition, we judgmentally selected 10 States to obtain
information regarding State site assessment programs. We
conducted telephone interviews with representatives in
Massachusetts, New Jersey, Pennsylvania, Florida, Michigan,
Texas, Missouri, California, and Washington. With the
exception of Colorado, we selected the State in each region
with the largest number of sites in CERCLIS. We visited
Colorado because we have an office in Denver.
We reviewed CERCLIS site assessment accomplishments for all
942 sites with SIPs completed between October 1991 and April
1993. We obtained SIP completion information from HSED, the
Region 8 CERCLIS coordinator, and representatives of three
regions. We sought regional clarification of SIP
accomplishments for those regions where information obtained
from HSED and the Region 8 CERCLIS coordinator differed
significantly. We obtained CERCLIS site assessment
accomplishments from a CERCLIS "List 8" report (April 21,
1993), obtained from the Region 8 CERCLIS coordinator.
We did not assess controls within the CERCLIS or WasteLan
systems, or assess,controls over the information recorded in
CERCLIS.
We reviewed HSED's Federal Managers' Financial Integrity Act
(FMFIA) documentation to evaluate control objectives and
techniques related to site assessment. We evaluated HSED's
fiscal 1992 and 1993 management control plans and its fiscal
1993 event cycle documentation. Additionally, we reviewed
OSWER's fiscal 1992 assurance letter.
We performed our audit in accordance with Government Auditing
Standards (1988 Revision) issued by the Comptroller General
of the United States. No other significant issues came to
our attention that warranted expanding the scope of our
review.
7
Report No. E1SFF3-08-0021-4100180
-------
PRIOR AUDIT COVERAGE
Prior Office of Inspector General (OIG) audits have not
evaluated EPA's site assessment program. However, prior
General Accounting Office (GAO) reports indicated that EPA
was experiencing difficulty in meeting CERCLA mandated site
assessment deadlines and identifying all potential hazardous
waste sites and costs during the site assessment process. A
November 1988 GAO report entitled "Missed Statutory Deadlines
Slow Progress in Environmental Programs" reported that EPA
was having difficulty meeting site assessment deadlines
established by SARA. GAO recommended that managers report
the reasons why SARA deadlines were missed. A July 1992 GAO
report entitled "EPA Cost Estimates Are Not Reliable or
Timely," criticized EPA for not including cleanup costs for
sites it intended to add to the NPL. The report recommended
that EPA improve the usefulness of reports to Congress by
refining budget estimates to include costs associated with
cleanup of sites expected to be included on the NPL.
In a December 1987, report entitled "Extent of Nation's
Potential Hazardous Waste Problem Still Unknown," GAO found
that CERCLIS was not an accurate picture of the nation's
hazardous waste problem because EPA policies did not require
States to include all potential hazardous waste sites in
CERCLIS. GAO concluded that the public and Congress were not
fully informed about the amount of work facing EPA and the
States, and the level of resources that should be allocated
to the Superfund program. Although GAO recommended that EPA
issue a formal CERCLIS reporting policy to be followed by the
regions and States, EPA had not issued guidance requiring
States or regions to report potential hazardous waste sites
at the time of our review.
Also, GAO reported on site deferral. An August 1989, GAO
report entitled "State Cleanup Status and Its Implications
for Federal Policy" recommended that if a deferral policy was
implemented, cleanup of deferred sites should be consistent
with the NCP; States' eligibility for deferrals should be
based on a State's ability to meet specified standards; and
EPA should have the right to monitor State cleanup
performance on deferred NPL sites.
8
Report NO. E1SFF3-08-0021-4100180
-------
CHAPTER 2
THE SITE INSPECTION PRIORITIZATION (SIP) CAN BE IMPROVED
Hazardous Site Evaluation Division (HSED) representatives
conducted the SIP program between October 1991 and August
1993, without written program guidance. Without guidance,
regions implemented SIP inconsistently and HSED could not
ensure that an important SIP goal, to assess the worst sites
first, was accomplished. Regions recorded completed SIPs in
CERCLIS inconsistently which caused management information to
be unreliable. HSED's management controls did not include
adequate control techniques to ensure national consistency in
program implementation. Recently issued program guidance
reiterates EPA's December 26, 1991, policy to propose worst
sites first for listing on the NPL. Although this is a
positive step, the guidance does not provide a planned
implementation approach with measurable program goals.
According to HSED, other higher priorities prevented it from
timely issuing guidance. Better planning could ensure more
consistent regional program implementation and more
meaningful program information.
LACK OF NATIONAL SIP GUIDANCE RESULTED
IN INCONSISTENT SIP IMPLEMENTATION
Regions began implementing the SIP program in October 1991,
almost 2 years before receiving guidance on how SIP should
operate. HSED did not ensure that regions directed SIP
toward EPA's site assessment goal to identify worst sites
first; and in fact, regions performed SIPs on sites that
should not have needed them. Regions implemented SIP
inconsistently and recorded program results in CERCLIS
inconsistently which caused management information to be
unreliable.
In 1986 Congress amended CERCLA by passing SARA. SARA
required that revisions be made to the HRS to ensure to the
maximum extent feasible the HRS accurately assessed the
relative risk to human health and/or the environment posed by
each site. As a result, site assessment managers developed
new guidance for conducting PAs, Sis, and published revised
HRS rules. The new rules changed how sites were scored and
changed how EPA determined if a site scored high enough to be
proposed for the NPL. Inclusion of a site on the NPL
9
Report NO. E1SFF3-08-0021-4100180
-------
qualifies the site for further Superfund evaluation and
cleanup. Between the passage of SARA in 1986 and August
1993, HSED told us it:
1) Began evaluating sites based on new HRS rules
(March 1991).
2) Issued revised PA guidance (OSWER Directive 9345.0-
01A—September 1991).
3) Began the SIP program to reevaluate sites
previously evaluated under the old HRS (October
1991) .
4) Issued revised SI guidance (OSWER Directive 9345.1-
05—September 1992) .
5) Issued memo guidance to Regions for setting NPL
priorities (OSWER Directive 9203.1-06—October
1992).
6) Published new HRS rules (OSWER Directive 9345.1-07-
-November 1992).
7) Issued SIP guidance to regions (August 1993).
Regions Implemented SIP Without Guidance
HSED implemented SIP in October 1991, but did not implement
program guidance until August 1993. HSED delayed issuing SIP
guidance because of resource constraints and other
priorities. Regions did not consistently implement SIP or
site assessment practices. Regions performed SIPs on sites
that should not have needed them.
HSED stated that higher priority needs precluded its issuing
timely SIP guidance and said it made some hard decisions in
the context of competing workload and priorities that caused
it to implement SIP the way it did. HSED representatives
told us that they developed guidance documents and training
programs for PAs, Sis, revised HRS scoring, and SACM
implementation from December 1990 to March 1993, and this,
along with other priorities, precluded issuing SIP guidance
until August 1993. HSED felt it was better to implement the
program without guidance than to delay the implementation for
2 years. Although expediting site prioritization was an
10
Report NO. E1SFF3-08-0021-4100180
-------
important goal, implementing SIP without national program
guidance resulted in regions taking different approaches to
reevaluating their site backlogs.
We discussed program implementation with EPA regions. Most
regions had not evaluated the nature of their SIP backlogs
and had not developed a worst sites first approach. For
example, one region reviewed SIP candidate sites and divided
them into groups according to the level of effort needed to
evaluate each site. This approach was beneficial for
workload allocation, but did not ensure worst sites were
addressed first. Another region identified levels of effort
needed to evaluate its sites, but did not review its backlog
and categorize sites before it began performing SIPs. Again,
the region did not take steps to ensure worst sites were
evaluated first. Other regions had not evaluated their
workloads or established a methodology for reevaluating
sites.
Some regions conducted SIPs for nearly 2 years without
national, written guidance and on sites that should not have
needed reevaluation. HSED officials stated that many regions
began using the revised HRS model when it became available in
March 1991, several months prior to SIP implementation in
October 1991. Sites that were reevaluated using the revised
HRS model should not need SIPs. In addition, some regions
performed SIPs on sites with Sis performed after SIP
implementation. Any site that had an SI performed after SIP
implementation should have used the revised HRS guidance.
Making timely site assessment guidance available to regions
could have prevented these unnecessary reevaluations.
One region said that there is another category of sites in
its "SIP" backlog that do not require SIPs. Region 7's
Information Management Coordinator and Superfund Branch Chief
told us that Region 7 may have a significant number of sites
that require no further Federal action but appear as SIP
candidates. These sites were classified as either high or
low priority sites upon the completion of PAs or Sis.
Subsequent to that evaluation some action occurred that
changed the status of the site and made further federal
involvement unnecessary. For example, Region 7 conducted a
removal at a site previously identified as a high priority
and as a result the site requires no further Federal
involvement. CERCLIS does not permit recording this
situation in its current preliminary assessment data fields.
11
Report NO. E1SFF3-08-0021-4100180
-------
As a result, some sites in the SIP backlog are sites that do
not need further evaluation; they are in the backlog only
because the CERCLIS database has no means to capture the
changed site status. Providing guidance that permitted
regions to identify and eliminate these sites prior to
beginning the SIP program would have provided better
management information about the nature and extent of the SIP
backlog and permitted better program management.
We reviewed the 942 SIPs performed from October 1991 through
April 1993 and found that 83 were on sites with Sis after the
March 1991 revised HRS guidance was available and 27 were on
sites with Sis after SIP was implemented. EPA officials
offered two reasons that SIPs could have been necessary: (1)
because a site was previously classified as needing no
further Superfund evaluation or deferred to another Federal
authority and new information indicated the site should be
considered for the NPL, or (2) contractors and State
representatives that performed Sis had not been trained to
use the revised HRS. One region accounted for most of the 27
sites that had both an SI and a SIP after October 1991.
Regional officials explained that a contractor had begun the
evaluations using the pre-SARA HRS model; the region directed
the contractor to gather data needed for the revised model,
but the contractor scored the sites using the pre-SARA HRS
model. SIPs were needed to prioritize these sites and
rescore the site using the new HRS.
Of the 6,467 sites,in the SIP backlog, 1,263 (almost 20
percent) had Sis completed after March 1991. Sites with Sis
performed after the revised HRS model was available should
not need a SIP. Likewise, sites with Sis performed after SIP
was implemented should not need SIPs because the revised HRS
model was available to be used when performing Sis.
Some Regions Did Not Implement
CERCLIS Programming Changes
Inconsistently counting SIP accomplishments in CERCLIS
provided HSED managers inaccurate information about SIP
results and hindered effective program management. We
obtained program results information from HSED but could not
reconcile it with information we obtained from the regions
and from another CERCLIS report. Without accurate,
consistent program information, HSED and regions could not
accurately determine the number or results of completed SIPs,
nor could they evaluate accomplishments against program
goals.
12
Report No. E1SFF3-08-0021-4100180
-------
HSED revised CERCLIS programming guidelines after
implementing SIP; however, not all regions implemented the
revisions. The revised CERCLIS programming guidance required
regions to change the way SIPs were tracked in CERCLIS.
Specifically, the revised guidance required regions to
replace the result of the last SI with the result of SIP.
EXHIBIT 2.1
NUMBER OF SIPS NOT
IDENTIFIED BY HSED
(•OCTOBER 1991-APRIL 19931
NUMBER OF SIPS
We reviewed SIP
accomplishments for all
10 EPA regions and found
that 4 regions did not
record SIP accomplish-
ments in accordance with
revised CERCLIS
programming guidance.
When HSED changed the
method of recording SIPs,
these regions did not
recede SIPs recorded
prior to the revised
guidance. Because
regions did not recode
previously completed
SIPs, HSED could not
accurately identify SIP
program accomplishments.
For example, one region
conducted 46 SIPs, none
were recorded correctly
and none were identified on the HSED list of completed SIPs.
Regions should have receded prior SIP accomplishments so that
all SIPs in CERCLIS were recorded consistently. Exhibit 2.1
shows 115 SIP accomplishments we identified which were not
identified by HSED. In a similar example, one region
conducted 67 SIPs; 55 were not coded correctly but were
included on the list of completed SIPs we obtained from HSED.
Conflicting information about program accomplishments makes
it difficult for HSED to manage SIP. Inconsistent and
inaccurate program status and resulting information misleads
the information users and precludes managers from accurately
measuring accomplishments or determining remaining workload.
10
13
Report NO. E1SFF3-08-0021-4100180
-------
SIP GUIDANCE COULD BE IMPROVED
During our review, HSED issued SIP guidance that clarified
many aspects of SIP. It explains the program, identifies the
program's goal and gives some advice on program activities.
More detailed SIP guidance could ensure consistent regional
program implementation, more meaningful program information,
and more effective and efficient use of program resources.
Through SIP, EPA plans to make decisions on backlogged sites
to enable it to address worst sites first in the future. The
guidance states that regions must set site priorities on a
worst site first basis and ensure that Superfund cleanups are
timely and efficient, but only suggests one method to
identify worst sites. The guidance does not suggest ranking
backlogged SIP sites based on prior site assessment
information as a method to move worst sites to the NPL first.
Additionally, it does not provide techniques to ensure that
SIPs are conducted only on those sites evaluated under the
pre-SARA HRS or that require reevaluation because additional
information becomes available. The SIP approach is not well
defined and there are no goals to evaluate program results
against.
HSED's guidance included sites in its universe which may not
require SIPs. It states that SIPs should be conducted on
sites that were evaluated before the implementation of the
revised HRS model and have not received a final NPL decision/-
but, it identifies' August 1992 as the revised HRS model
implementation date and says sites evaluated prior to that
date are eligible for SIPs. HSED officials told us that some
regions began using the revised HRS model in March 1991.
Sites with Sis performed using the revised HRS model should
not need SIPs. We identified 1,263 sites with Sis completed
between March 1991 and August 1992; a portion of these sites
should not need SIPs. Additionally, the guidance states that
SIPs may be conducted on sites where new information is
necessary to determine whether a site should be screened out
or investigated further for probable inclusion on the NPL.
14
Report No. E1SFF3-08-0021-4100180
-------
ADDITIONAL CONTROL TECHNIQUES MAY
ENSURE NATIONAL CONSISTENCY
Although HSED identified promoting and ensuring consistent
application of guidance and training as a primary control
objective, its event cycle documentation did not include
adequate control techniques to ensure the objective was met.
HSED identified the need to promote and ensure national
consistency in the use of guidance and training among regions
and States. To meet this control objective, HSED listed
seven control techniques. However, in order to ensure
consistency in the use of guidance and training, additional
control techniques are necessary. For example, evaluating
program results for consistency is a necessary management
control technique.
HSED's control techniques to ensure national program
consistency included procedures to promote regional and State
involvement in guidance development. However, HSED did not
identify control techniques to ensure the consistent
application of guidance. For example, regions did not
consistently record SIP results. Including a control
technique to review regional implementation of guidance would
have identified inconsistencies.
CONCLUSIONS
HSED implemented SIP in October 1991, but did not issue
program guidance until almost 2 years later. Without
guidance, regions implemented SIP inconsistently and
accounted for SIP accomplishments in CERCLIS inconsistently.
The SIP backlog includes sites with Sis performed after
regions implemented the revised HRS model. Some sites in the
SIP backlog appear only because EPA has no means to code the
site as complete. Program guidance clarifies some aspects of
the program, but does not ensure that high priority sites in
the backlog are evaluated and moved to the NPL first.
RECOMMENDATIONS
We recommend that the Assistant Administrator for Solid Waste
and Emergency Response require the Director, OERR to:
1. Ensure that SIPs are conducted on high priority
sites first and that SIPs are not conducted on
15
Report NO. E1SFF3-08-0021-4100180
-------
sites that do not need them.
2. Ensure that regions consistently account for SIPs
in CERCLIS.
3. Develop a method to record a non-NPL site
completion for sites that are not eliminated as a
result of site assessment work but, because of some
later action such as a removal action, will not
require further EPA work.
4. Revise management controls to include techniques
that would provide reasonable assurance that
regions and States consistently apply guidance.
EPA COMMENTS AND PIG EVALUATION
OSWER agreed with our recommendations and has completed,
begun, or plans to take corrective action for each
recommendation. (See Appendix I.) HSED officials told us
they intend to discuss methods to ensure worst sites are
reevaluated first at the National Site Assessment Meeting in
April 1994. OSWER also initiated a quarterly, formal
management report to track SIPs and to ensure SIPs are being
recorded consistently in CERCLIS.
In our draft report we recommended that EPA not destroy
historical SI completion data in CERCLIS. Because HSED
managers told us that they have access to historical data
from EPA's annual "frozen database", we eliminated that
portion of our recommendation. In addition, HSED managers
told us that they plan to analyze SIP program results this
fiscal year. OSWER agreed to implement changes to CERCLIS
that will permit recording the completion of a non-NPL site
even though the site was not eliminated as a result of site
assessment work.
An HSED manager told us that HSED intends to incorporate
control techniques in its FMFIA documentation to ensure
national program consistency. OSWER's response stated that
it intends to review current control techniques with regions
at the time of the next FMFIA update.
16
Report No. E1SFF3-08-0021-4100180
-------
CHAPTER 3
UNREPORTED COSTS TO CLEANUP POTENTIAL
NPL SITES MAY EXCEED $74.5 BILLION
Despite its efforts to accelerate site assessment and
cleanup, EPA has not met CERCLA time goals for listing sites
on the NPL and has not reported an estimated $74.5 billion in
future cleanup costs. SIP is expected to identify about
3,336 sites that will require HRS packages. At the current
rate and current cost of preparing HRS listing packages, it
will take EPA over 27 years and an estimated $105 million to
make NPL decisions on these sites. EPA may have to
reevaluate sites again that remain in the backlog for an
extended time period. EPA cannot ensure that the most
hazardous sites are being cleaned up first without fully
evaluating the sites in the backlog. Based on past EPA
cleanup costs per site ws estimated the cost to cleanup sites
that will likely be listed on the NPL at over $74.5 billion.
Not knowing the magnitude of the hazardous waste site cleanup
problem adversely impacts the public, Congress, and parties
responsible for site cleanups. Although EPA has developed
initiatives to improve the Superfund process, those
initiatives will have a minimal impact on the potential NPL
backlog.
POTENTIAL NPL LISTINGS WILL TAKE $105
MILLION AND 27 YEARS TO ELIMINATE
SIP will identify a potential NPL backlog; some sites will
require additional investigation and all will require HRS
package preparation to determine if they qualify for the NPL.
We estimated that it will take EPA 27 years and about $105
million to eliminate the potential NPL backlog. If EPA
continues to take as long as it has in the past to list
sites, some sites will have to be reevaluated again,
requiring additional resources. In addition, EPA must
continue to assess and prepare HRS packages for new sites
that are discovered each year.
We reviewed SIP accomplishments for each region and
determined that many sites will require additional Federal
expenditures upon SIP completion. As of April 1993, regions
had conducted 942 SIPs, 14.6 percent of the 6,467 site
backlog. SIP eliminated 526 from further participation in
17
Report NO. E1SFF3-08-0021-4100180
-------
the Superfund process. The remaining 416 sites,
approximately 44 percent, will require HRS listing packages.
We estimated the cost of conducting additional Superfund
evaluations on sites that are not screened out during SIP.
HRS packages must be prepared for all sites proposed to the
NPL. EPA budgets $31,350 (550 hours at $57/hour) to prepare
an HRS package. Using EPA's budget estimates, we estimated
that it will cost approximately $105 million to prepare HRS
packages on the 3,336 site backlog. EPA budgets for 122 HRS
packages per year; so, it will take over 27 years to complete
HRS packages in the potential NPL backlog.
Exhibit 3.1 on page 19 provides an estimate, by region, of
sites in the current backlog that will require HRS package
preparation and may become NPL sites. We used SIP results
from October 1991 through April 1993 to estimate sites that
may require HRS packages. We used the results of all sites
proposed for listing through June 30, 1993, (historically 94
percent of all sites proposed for listing are listed) to
estimate sites that may be listed on the NPL.
Additional resources may be needed to reevaluate sites that
remain in the backlog for an extended time period. EPA's HRS
Guidance Manual (November 1992) states that the HRS score
should reliably reflect the site's eligibility for the NPL.
If the site score is greater than or equal to 28.50, the
scorer should be confident that the score will remain at or
above 28.50 after the Headquarter's Quality Assurance/Quality
Control Review and public comments. As information becomes
older and site conditions change, EPA will need to reevaluate
or collect additional information on sites to ensure that HRS
packages are adequately supported.
18
Report No. E1SFF3-08-0021-4100180
-------
EXHIBIT 3.1
ESTIMATE OF POTENTIAL NPL
SITES IN THE SIP BACKLOG
REGION
1
2
3
4
5
6
7
8
9
10
TOTAL
SITES IN
THE BACKLOG
871
733
419
1,307
1,398
838
261
58
331
181
6,397
SITES REQUIRING
HRS PACKAGES
679
264
105
588
839
670
131
12
17
31
3,336
POTENTIAL
NPL SITES
638
248
99
553
789
630
123
11
16
29
3,136
NOT MEETING CERCLA'TIMELINESS GOALS
ADVERSELY IMPACTS THE SUPERFUND PROGRAM
Despite efforts to .accelerate site assessment and cleanup,
EPA has not met time goals for listing sites on the NPL and
cannot ensure that the most hazardous sites are being cleaned
up first. When enacted, SARA required that all old sites and
any new sites in CERCLIS be screened from further Superfund
evaluation or be listed on the NPL within 4 years. If EPA
does not meet the timeframes, it must disclose the reason to
Congress. SIP Guidance requires that the site assessment
process screen out less serious problems and expedite action
at sites that require additional Superfund response. The
guidance also requires that site priorities be set on a worst
site first basis to ensure that cleanups are timely and
efficient. EPA issued guidance for setting NPL candidate
site priorities in October 1992 which required regions to
divide NPL candidate sites into high or low priorities to
19
Report NO. E1SFF3-08-0021-4100180
-------
ensure the worst sites are addressed first. However, SIP
guidance does not require that regions reevaluate high
priority sites first. At the current rate EPA is evaluating
and listing sites it cannot meet SARA NPL listing goals. EPA
will not be able to ensure it is addressing worst sites first
until the backlog is eliminated and all sites are listed.
EPA'5 REPORTED COSTS TO CLEANUP POTENTIAL
NPL SITES IS UNDERSTATED
EPA's estimate of the cost of cleaning up Superfund sites is
significantly understated. EPA's interpretation of CERCLA
reporting requirements leaves Congress unaware of the total
costs associated with listing and cleaning up hazardous
sites. If EPA does not fully disclose all costs, Congress
does not have complete information to determine future
Superfund funding levels and priorities, establish meaningful
goals, or legislate other needed changes in the Superfund
program.
CERCLA requires EPA to report annually to Congress an
estimate of the resources needed to complete Superfund
implementation. CERCLA Section 301 (h)(I)(G), requires EPA
to submit an annual report to Congress which includes an
estimate of the resources, including the number of work years
or personnel necessary, to conduct activities associated with
CERCLA implementation. EPA interpreted this requirement as
the cost of completing cleanups of existing NPL sites.
EPA developed the Outyear Liability Model (OLM) to project
costs, activity levels, and resource needs associated with
the Superfund program. The OLM combines historical trends
and program activities with expected program conditions to
develop a comprehensive analysis of Superfund program
resource and funding needs. The OLM estimate included in
EPA's 1990 annual report to Congress, addressed only the cost
of cleaning up existing NPL sites (proposed and final). EPA
estimated $16.4 billion to complete cleanup on existing
sites.
EPA's estimate does not disclose the costs of evaluating or
cleaning up sites in the potential NPL backlog. EPA will
spend about $105 million to prepare HRS packages for the
3,336 sites in the backlog. We estimated the cost to
evaluate and cleanup the 3,136 sites that will likely be
placed on the NPL. Based on EPA per site cost estimates, the
20
Report No. E1SFF3-08-0021-4100180
-------
cost of conducting cleanups on the 3,136 sites would be over
$74.5 billion (3,136 sites at $23.76 million per site).
Additionally, based on historical trends of the site
assessment process, approximately 5 percent of the new sites
entered in CERCLIS (i.e., new sites discovered) will be
listed on the NPL.
POTENTIAL BACKLOG IMPACTS THE PUBLIC.
CONGRESS, AND RESPONSIBLE PARTIES
Not knowing the magnitude of the hazardous waste site cleanup
problem adversely impacts the public, Congress, and those
responsible for site cleanups. The potential NPL backlog
hinders EPA's ability to ensure that the worst sites are
addressed first. Although EPA assigns sites high and low
priorities, neither the site assessment process nor the MRS
model identifies which sites within those broad categories
are the worst sites. EPA notifies the public of the extent
and nature of contamination at a site during proposal to the
NPL. Additionally, EPA does not estimate the cost of
cleaning up sites until listed on the NPL. Therefore, until
that time the public is unaware of the location and
seriousness of those hazardous waste sites that EPA
determines are the worst until the sites are added to the
NPL. Congress is unaware of the potential cost and magnitude
of the Superfund program and EPA has not fully reported
problems in listing sites on the NPL or the potential cost
and magnitude of the hazardous waste site problem. Unless
EPA accurately estimates the costs associated with cleaning
up potential NPL sites and provides this information to
Congress, Congress will not have accurate, complete, and up-
to-date information to make informed decisions during
Superfund reauthorization.
NPL listing and the accompanying potential financial
liability can have an adverse impact on responsible parties.
CERCLA requires responsible parties to pay for cleaning up
hazardous waste sites. The time between site discovery and
listing leaves responsible parties in doubt about their
CERCLA liability because selecting a remedy and estimating
cleanup costs is done after sites are listed on the NPL.
Until liability is established, investors may be reluctant to
invest money, banks may be reluctant to make loans, and
corporate managers may be reluctant to expand operations.
NEW INITIATIVES WILL NOT ELIMINATE
THE POTENTIAL NPL BACKLOG
21
Report No. E1SFF3-08-0021-4100180
-------
In response to public and congressional criticism about the
slow pace of the Superfund process, EPA has developed new
Superfund initiatives; Superfund Regional Pilots and the
Superfund Administrative Improvements initiative among them.
EPA developed SACM, an initiative under the Superfund
Regional Pilots program, to establish major initiatives to
streamline and accelerate Superfund response activities. EPA
developed the Superfund Administrative Improvements
initiative to determine what improvements it could make under
the current law and identify improvements that require
legislative approval. An important part of this initiative
involves site deferral to States. While the SACM initiative
will have little impact on the potential NPL backlog, the
Superfund Administrative Improvements initiative could help
to reduce the potential NPL backlog.
The Superfund initiatives set a framework for streamlining
assessment and cleanup activities involving an integrated
assessment function, early actions, long-term remediation,
regional decision teams, and appropriate integration of
enforcement, community relations, and public participation
throughout the process. Superfund assessment under SACM
integrates previously separate removal and remedial
assessments into a single process. Under SACM, the
assessment processes operate concurrently. EPA projects that
integrating assessment functions will cut several years from
the site assessment and cleanup process. HSED stated that
EPA instructed the Regions to use information developed in
the SIP program to identify sites for early actions under
SACM.
A single assessment function combines elements of traditional
removal assessments, PAs, Sis and Remedial Investigations
(RI). After the PA or SI stage, sites found to have serious
contamination problems can proceed directly to more detailed
Rl-level data collection and risk assessment. SACM may
accelerate the PA and SI steps in the site assessment
process, but it will have minimal impact on the listing
process for sites in the SIP backlog. Sites in the potential
NPL backlog that will be identified by the SIP program
already have completed PAs and Sis and the opportunity to
integrate removal and remedial assessment is decreased.
One element of the Superfund Administrative Improvements
initiative is to expand State roles to expedite cleanup of
Superfund sites through deferral. Deferring NPL-caliber
sites to States could reduce EPA's potential NPL backlog. We
22
Report NO. E1SFF3-08-0021-4100180
-------
discuss EPA's deferral policy in greater detail in Chapter 4.
CONCLUSIONS
Completion of SIP may identify as many as 3,136 additional
NPL sites. EPA has not met, and in the near future likely
will not meet, CERCLA site evaluation timeliness goals. If
EPA continues to budget for the preparation of 122 HRS
listing packages per year, it could take over 27 years to
prepare listing packages for all the sites now in the
potential NPL backlog. Although recent initiatives, such as
the SACM and the Superfund Administrative Improvements
initiative, may help the Superfund program in general, other
changes are necessary to meet CERCLA goals for listing sites.
Until sites are fully evaluated and listing decisions are
made, EPA cannot determine which sites are among the worst
and which should be cleaned up first. EPA has not fully
informed the public, Congress, and potentially responsible
parties (PRP) of the total cost and magnitude of the
Superfund problem.
RECOMMENDATIONS
We recommend that the Assistant Administrator for Solid Waste
and Emergency Response require the Director, OERR to:
1. Include an analysis of the resources required for
EPA to meet its SARA site assessment goals in
HSED's fiscal 1996 budget request.
2. Disclose to Congress the estimated cost of cleaning
up those sites that upon completion of the SIP
program will likely be proposed for listing on the
NPL.
EPA COMMENTS AND PIG EVALUATION
OSWER officials generally agreed with our recommendations.
(See Appendix I.) In our draft report, we recommend that EPA
advise Congress why it is unable to meet site evaluation
timeliness goals. OSWER said that it had reported why it did
not meet goals but had not advised Congress of the resources
required to meet goals. OSWER managers agreed to include an
analysis of the resources required to meet SARA site
23
Report No. E1SFF3-08-0021-4100180
-------
assessment goals in HSED's fiscal 1996 budget request. In
our draft report we recommended that EPA advise Congress of
the estimated cost to assess and cleanup potential NPL sites
currently in the backlog. OSWER questioned disclosing to
Congress the cost of cleaning up all potential NPL sites.
OSWER does not believe its information regarding the number
and type of sites in the SIP backlog is accurate enough to
estimate the cleanup cost of sites not yet listed on the NPL.
We believe that it would be appropriate, however, to estimate
cleanup costs for those sites that have had SIPs completed.
Since the purpose of the SIP program is to characterize the
nature of a site and identify which sites will be proposed
for NPL listing, EPA should have a reasonable basis to
estimate cleanup costs at that time. Accordingly, we changed
our recommendation to include only those sites with completed
SIPs that are likely to be listed on the NPL. Congress
should be fully informed of the potential Federal liability
during annual appropriation deliberations.
24
Report No. E1SFF3-08-0021-4100180
-------
CHAPTER 4
AN EXPANDED DEFERRAL POLICY
MAY HELP REDUCE THE POTENTIAL NPL BACKLOG
Adopting an expanded site deferral policy could reduce the
potential NPL backlog, conserve site assessment resources,
accelerate site cleanups, and help EPA meet CERCLA site
assessment timeframes. EPA considered adopting a broad
deferral policy in the past, but did not do so in response to
congressional concern. Comments on EPA's prior deferral
proposal were mixed. EPA has proposed using pilot projects
to evaluate deferring sites to States. An even broader
deferral policy could permit other Federal authorities and/or
responsible parties to perform CERCLA quality hazardous waste
site cleanups on sites awaiting NPL listing. The NPL could
serve as a better management tool in setting priorities for
addressing sites.
EPA'S PRIOR ATTEMPT TO ADOPT AN EXPANDED
DEFERRAL POLICY WAS UNSUCCESSFUL
In 1988, EPA sought comments on a proposal to defer sites to
States, additional Federal authorities, and responsible
parties. EPA sought and received comments from citizens,
States, and interested environmental groups on its proposal.
Many comments were positive and identified benefits that
could be gained from adopting a deferral policy. Other
comments were cautiously optimistic and some opposed
deferral. Congress and staff voiced concerns about the
legality of deferral, about cleanup and enforcement
authorities, and about funding levels of other entities.
HSED managers told us that some of the opposition to deferral
expressed in 1989 still exists today and "...would have to be
dealt with during the process of setting up such a program."
In December 1988, EPA sought public comments on deferral
(e.g., it held public hearings) and it agreed to consider
comments prior to implementing a deferral policy. Industry
and State representatives generally supported deferral. Both
noted that deferral would permit cleaning up more sites in a
timely fashion and eliminate duplication of State and Federal
effort. State representatives believed that deferring sites
to States would provide more leverage to negotiate agreements
with PRPs because the State could assure the PRP that the
25
Report No. E1SFF3-08-0021-4100180
-------
site would stay out of the Federal process. State experience
with PRPs has shown that the PRPs want to stay off the NPL.
One State representative noted the positive effect deferral
would have on State/EPA relationships and that Superfund cost
savings could be redirected to State programs. Other State
participants argued that State deferral would eliminate
confusion over site responsibility. Industry representatives
said that it would be more efficient to deal with one State
bureaucracy and that overlapping State and EPA
responsibilities slowed past cleanup efforts.
Industry/trade association representatives, State and local
government officials, and individuals supported deferring
sites to responsible parties. They noted that deferring
sites to responsible parties, in connection with an
enforcement agreement, would provide a strong incentive to
"voluntarily" cleanup sites, cooperate with State and Federal
authorities, and avoid the allegedly bureaucratic listing
process. They added that deferring cleanup to responsible
parties would allow scarce Superfund money to be used only
for those sites where no other resources are available.
Critics of an expanded deferral policy believed EPA was using
the policy to limit the scope of the Superfund program by
excluding certain categories of hazardous waste sites and
shifting the burden of cleanups to State and other Federal
authorities. Congressional committee staff suggested that
deferral was inconsistent with the legislative history of
SARA, was illegal, and had to be authorized by Congress.
Environmental group participants questioned the legality of
deferring sites. In its December 21, 1988, Federal Register
notice, EPA stated that it would keep the current deferral
policies in effect and not implement a general deferred
listing policy until it considered comments on its draft
policy.
During 1989 confirmation hearings, EPA's Administrator told
Congress that EPA would consider congressional concerns
before it adopted a deferral policy. Congress was concerned
that other Federal laws did not have cleanup authority and
were underfunded. Congress also noted that some State solid
waste laws did not provide enforcement authority like
Superfund.
26
Report No. E1SFF3-08-0021-4100180
-------
EPA PROPOSES LIMITED DEFERRAL
As part of its Superfund Administrative Improvements
initiative, EPA proposed a State deferral policy in June
1993. Its objective is to increase States' roles in the
Superfund process and take advantage of State capabilities.
According to the initiative, candidate sites for State
deferral are NPL-caliber sites—sites that will likely
achieve an HRS score of 28.50 or above. As discussed in
Chapter 3, there may be over 3,000 NPL-caliber sites in the
SIP backlog.
The proposed State deferral initiative focuses on NPL-caliber
sites that have not been proposed for listing. It would give
States the responsibility and authority to cleanup sites
without listing the sites on the NPL. Under the deferral
proposal:
EPA and States will openly negotiate a division of
responsibility for NPL-caliber sites to determine
which agency should address a given site.
States will have the initial responsibility to
cleanup NPL-caliber sites that were classified as
low priority sites during initial EPA evaluations.
— EPA will defer sites as long as satisfactory
cleanup progress is made.
i
EPA will retain the right to list sites on the NPL
as leverage to motivate reluctant responsible
parties to conduct cleanups under State
supervision.
EPA initiated pilot projects in September 1993 and
anticipated that interim guidance establishing criteria for
State deferral would be effective in March 1994. EPA
attorneys stated that EPA has the legal discretion to exclude
sites from NPL listing through deferral. They also said EPA
has the authority to implement a deferral policy without
congressional approval.
In its June 23, 1993, Superfund Administrative Improvements
report EPA noted that, at current budget levels, it will be
unable to address the environmental threats at some sites for
27
Report NO. E1SFF3-08-0021-4100180
-------
many years. EPA also noted that by identifying sites and
postponing cleanup, PRPs are left in doubt about their
financial liability. In addition, local communities remain
at risk from unremediated sites and without the productive
use of the related land.
A WELL DESIGNED AND IMPLEMENTED DEFERRAL
POLICY WILL HELP ENSURE PROGRAM SUCCESS
A carefully designed and implemented deferral policy with
public and congressional input could significantly improve
the efficiency and effectiveness of the Superfund program.
As part of implementation, EPA should respond to previously
expressed concerns. A broader deferral policy may reduce the
time and cost of cleaning up sites that are currently
awaiting listing on the NPL by encouraging greater
participation of all parties involved in hazardous waste
cleanup.
When it last considered an expanded deferral policy, EPA
sought comments through the Federal Register and through
public meetings. It discussed deferral during congressional
hearings. EPA representatives stated that they would
consider and respond to the comments. To maintain its
credibility and assure public confidence in an expanded
deferral policy, EPA should review and analyze prior comments
and respond to them.
Most EPA officials we talked to supported an expanded
deferral policy. They identified two elements that they
thought should be specifically addressed in the deferral
policy: (1) site cleanups should meet NCP standards,
including community involvement in the Superfund process;
and, (2) EPA and States should develop guidelines, controls,
and oversight procedures to ensure consistent deferral
implementation. In addition, we believe that EPA should
establish cleanup timeframes during site deferral
negotiations and establish measurable deferral program goals.
EPA plans to have a regional/Headquarters work group address
key State pilot implementation questions, and assess early
State-lead experiences. Building upon the knowledge gained
by assessing State led deferred sites, a regional/Head-
quarters work group could also provide valuable information
regarding deferring sites to other Federal authorities and
responsible parties. The same advantages that derive from
28
Report No. E1SFF3-08-0021-4100180
-------
deferring sites to States could derive from deferring sites
to other Federal authorities and responsible parties. For
example, deferring more sites would reduce the potential NPL
backlog, conserve site assessment resources, accelerate site
cleanups, and help EPA meet CERCLA site assessment
timeframes. EPA should design its deferral program to ensure
maximum participation in hazardous waste cleanups that meet
established standards and goals.
CONCLUSIONS
When completed, we estimate SIP will have identified over
3,000 NPL-eligible sites. Unless USED establishes other
procedures to get these sites cleaned up, they will remain in
the potential NPL backlog for years. By carefully
implementing a broader deferral policy, EPA can solicit the
help of willing participants to increase the number of
cleanups underway at one time.
EPA recognizes the benefits of site deferral, and plans to
implement a limited deferral policy on a pilot project basis.
A few years ago, EPA considered but did not implement a
broader deferral policy. Although EPA told Congress and the
public it would not implement broader deferral without
answering public comments, it began testing the use of
deferral to States without responding to concerns previously
expressed by Congress, environmental groups, States, and
others. However, EPA representatives told us that they
intend to consider and respond to these comments. Not
involving affected parties in an expanded deferral policy may
bring unnecessary public and congressional criticism and
decrease confidence in EPA's ability to cleanup hazardous
waste sites.
A well designed and implemented deferral policy would reduce
the potential NPL backlog and help EPA meet its CERCLA
timeframe for listing sites. Deferring site cleanup avoids
NPL listing and promotes a cooperative rather than an
adversarial relationship with responsible parties.
Establishing a mechanism to use non-Federal resources in
managing some site cleanups would allow EPA to use Superfund
resources for additional cleanups.
29
Report No. E1SFF3-08-0021-4100180
-------
RECOMMENDATIONS
We recommend that the Assistant Administrator for Solid Waste
and Emergency Response:
1. Consider elements identified by EPA and State
officials as essential to a successful deferral
policy and determine the proper Federal role when
establishing oversight procedures, rules, and
responsibilities before implementing a broader
deferral policy.
2. Consider deferring sites to PRPs as well as States
and other Federal authorities.
3. Respond to prior deferral proposal comments and
involve those affected in deferral decision making
activities.
EPA COMMENTS AND PIG EVALUATION
OSWER agreed with our recommendations. (See Appendix I.)
OSWER is currently working on a deferral policy under EPA's
Administrative Improvements Initiative. It indicated that it
is discussing the essential elements of a successful deferral
policy, as part of Superfund reauthorization, with the
public, States, environmental groups, industry, and Congress.
OSWER's response did not address the necessity of
establishing oversight procedures prior to implementing a
broader deferral policy. An HSED representative told us
however, that as EPA develops deferral policy it will address
implementation issues.
In its comments, OSWER also stated that it is considering the
issue of "voluntary cleanups" by PRPs (what we refer to as
PRP deferral) but believes it should move cautiously. It
indicated that decisions about "voluntary cleanups" will be
made by Congress. We agree that a major policy change such
as having PRPs conduct voluntary cleanups should be entered
into cautiously.
OSWER agreed that it should respond to prior deferral
comments and involve affected parties in deferral decision
making activities. EPA has already taken steps to ensure all
30
Report No. E1SFF3-08-0021-4100180
-------
of the major stakeholders are involved in deciding how to
implement deferral.
31
Report No. E1SFF3- 08-0021-4100180
-------
APPENDIX I
.S8&
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
OFFICE CF
SCUD WASTE AND EMESG
RESPONSE
MEMORANDUM
SUBJECT: DIG Draft Report: EPA's Superfund Site Assessment
Process E1SFF#-08-0021-xxxx
FROM: Elliott P. Laws
Assistant Administratef
TO: Kenneth A. Konz
Assistant Inspector General
for Audit
This memorandum responds to the subject draft audit report
per your November 15, 1993 request. We have implemented several
of the recommendations made in this report. We appreciate the
work that you and your staff have done in your review of this
process. The following is OSWER's response to your
recommendations:
Recommendation to the Assistant Administrator, O3WER,page 16, #1
Refine Sight Inspection Prioritization (SIP) Guidance
to establish procedures to ensure high priority sites
are reevaluated first and to ensure SIPs are not
conducted unnecessarily.
OSWER Response
The Final SIP Guidance, which was issued in August 1993,
contains procedures to ensure that high priority sites are
reevaluated first, as you recommended (copy attached). We wish
to reiterate that a substantial amount of discussion took place
before the guidance was issued. SIP implementation was discussed
over a two year period at the regular Regional/Headquarters
Section Chiefs meetings (attended by Headquarters and Regional
Section Chiefs, and the Site Assessment Chief), and during the
regular Section Chief conference calls. In addition, a formal
SIP presentation was given at the March 1992 Annual Site
Assessment National Meeting, attended by over 200 site assessment
personnel from Headquarters, regions, states, contractors and
Federal agencies.
32
Recycled/Recyclable
Printed with Soy Canoia InK on paper that
contains at least SON recycled tlbw
-------
Recommendation to the Assistant Administrator, OSWER, page 17, #2
Ensure that regions consistently account for SIPs in
CERCLIS and ensure recording SIP accomplishments does
not destroy other, important historical data.
OSWER Response
Your recommendations have been incorporated. In FY 94, HSED
is initiating a more formal, quarterly management report to track
the progress toward eliminating the SIP backlog, and to ensure
that SIP accomplishments are being recorded correctly in CERCLIS.
In addition, SIP accomplishments are being incorporated into the
FY 94 Superfund Consolidated Accomplishments Plan (SCAP) Reports,
and will be included in the FY 94 Superfund Quarterly Management
Reports.
We also believe that replacing the result of the last Site
Inspection (SI) with the new SIP recommendation will not destroy
the historical data. Historical CERCLIS data is maintained by
Headquarters in "frozen" databases. Although we do not
anticipate the need to revisit the old SI qualifiers, we do have
the capability of pulling these qualifiers from past frozen
databases if the need arises.
Recommendation to the Assistant Administrator, OSWER, page 17, #3
Provide a manner to record a non-NPL site completion
for sites that are not eliminated as a result of site
assessment work but because of some later action will
not require further EPA work. Possible solutions
include providing an "NPL type" work complete indicator
for non-NPL sites, expanding event codes in the site
assessment field, or adding a new event qualifier.
OSWER Response
We agree that CERCLIS should have such capability, and will
implement your recommendation in FY 94. We are currently
considering various options, including adding new Preliminary
Assessment (PA) and Site Inspection (SI) qualifiers to identify
sites that are referred to the Removal program for cleanup,
indicating whether further remedial assessment is necessary.
Recommendation to the Assistant Administrator, OSWER, page 17, #4
Revise FMFIA control documentation to include
additional control techniques that would provide
reasonable assurance that Regions and States
consistently apply guidance.
OSWER Response
33
-------
The Headquarters FMFIA event cycle documentation report
contains the event cycle "develop site assessment guidance to
promote national consistency" with the control objectives to
"promote national consistency among the states and Regions in
conducting site assessments." Numerous control techniques to
achieve this objective are outlined in the documentation
guidance. We will share this information with the Regions at the
time of the next FMFIA update.
Recommendation to the Assistant Administrator, OSWER,, pacre 24, #1
Develop realistic budget requirements for site
assessment, and reevaluate the allocation of resources
necessary to prepare Hazardous Ranking System (HRS)
packages in accordance with SARA goals.
OSWER Response
In formulating the annual Superfund budget, the Office of
Emergency and Remedial Response (OERR) first identifies their
total funding need by specific program area, i.e. pre-remedial or
remedial. These needs are then prioritized across the Divisions
based on Program priorities outlined in the Superfund Program
Management Manual. In FY 93, site assessment activities received
approximately $100M of the $1,050M of extramural resources
managed by OERR.
EPA has established measures such as the Site Screening and
Assessment Decisions SCAP measure to monitor how the priorities
are met. This measure monitors regional progress towards
recommending sites for Hazardous Ranking System (HRS) scoring.
For example, as expanded site inspections are completed, the
measure monitors how many sites require no further action and how
many are recommended for HRS scoring. Through this measure, EPA
can verify the rate at which sites are being assessed and scored.
To ensure resource distribution is optimal, the site assessment
budget requirements are reevaluated on an annual basis.
Recommendation to th® Assistant Administrator, OSWER. page 24, /2
Advise Congress of the full cost of the Superfund
program by including estimated costs of assessing and
cleaning up potential NPL sites currently in the
backlog.
OSWER Response
The recommendation advises including potential NPL sites in
outyear liability information. While this is a reasonable
request and the Agency has researched the possibility of
34
-------
icluding these sites, analyses have shown that due to
.actuating NPL listing rates and varying site characteristics of
. w sites, we cannot accurately predict the number and type of
: tes that will be listed in any given fiscal year. For these
j jsons, outyear estimates consist of the current NPL only.
Recommendation to the Assistant Administrator, OSWER, page 24, #3
Advise Congress, as required by CERCLA section 116, why
EPA is unable to meet CERCLA site evaluation timeliness
goals.
OSWER Response
Through our annual report, as well as the annual budget
debates, we inform Congress. We are open to any suggestions
which you might have on how we can make this process more
effective.
Recommendation to the Assistant Administrator, OSWER, page 30, #1
Consider elements identified by EPA and State officials
as essential to a successful deferral policy and
establish oversight procedures, rules, and
responsibilities before implementing a broader deferral
policy.
OSWER Response
Under the current reauthorization process, EPA is looking at
all aspects of the Superfund program in an effort to design a
much more efficient and effective program. The roles of the
states and Potentially Responsible Parties (PRPs) are critical in
this and are being discussed with members of the public, States,
environmental groups, industry and Congress.
Recommendation to th® Assistant Administrator, OSWER, page 30, #2
Consider deferring sites to PRPs as well as States and
other Federal authorities.
OSWER Response
We are currently exploring State Deferral under the
Superfund Administrative Improvements initiative. However, the
Agency feels the need to move cautiously on the deferral issue
since it is part of the larger reauthorization debate. In
addition, Congress placed language in the FY 1994 Appropriations
Conference Report stating the following:
35
-------
"Under the Agency's Superfund Administrative
Improvements initiative, several procedures have been
designed to expedite the current process. While the
conferees are supportive of such an initiative, these
efforts should in no way hinder the listing of sites on
the Superfund National Priority List. The Agency
should continue its normal procedures under Superfund
rules and regulations, including the listing of
Superfund sites if warranted by the administrative
record."
The issue of "voluntary clean-ups" by PRP's brings with it
issues in addition to those being worked out under the State
deferral pilots. This is also part of the larger debate for
Reauthorization and involves the debate over remedy selection and
clean-up standards versus clean-up levels. Congress will make
the final decision over the next year on issues such as these.
Until then, we will continue to work with all parties to
stimulate the discussion.
Recommendation to the Assistant Administrator, OSWER, page 30, #3
Respond to prior deferral proposal comments and involve
those affected in deferral decision making activities.
OSWER Response
We agree, and the Agency is already committed to an open
process during the Superfund reauthorization debate and has
involved all of the major stakeholders in this debate. With the
use of public forums,' the involvement of the National Advisory
Council on Environmental Policy and Technology (NACEPT), States
and other outreach efforts to industry and the public, those
affected by Superfund have been active players in discussions on
how a reauthorized Superfund program should be structured.
We appreciate the opportunity to share with your office our
response concerning the Superfund Site Assessment Process. If
you have any questions, please call Charlene Dunn, OSWER Audit
Follow-up Coordinator, at 202-260-9466.
Attachment
cc: Sharon Hallinan
Johnsie Webster
Sandra Lee
36
-------
United States
Environmental Protection
Agency
Office of
Solid Waste and
Emergency Response
Directive 9345.1-15F5
August 1993
Office of Emergency and Remedial Response
Hazardous Site Evaluation Division
EPA/540/F-93/037
Quick Reference Fact Sheet
The purpose of this fact sheet is to provide guidance to Environmental Protection Agency (EPA), State, and
contractor staff responsible for conducting Site Inspection Prioritizations (SIPs). Of the 36,000 sites currently in
the Comprehensive Environmental Response, Compensation, and Liability Act Information System (CERCLIS)
inventory, approximately 16,700 have undergone Site Inspections (Sis). Of those, however, over 6,000 sites still
require final site disposition decisions (Figure 1). This backlog has made it difficult for EPA to evaluate sites
efficiently on a worst sites first basis. Consequently, EPA established the SEP to address this backlog and to make
decisions on these sites.
Figure 1
SITE INSPECTION PHJORmZATJON BACKLOG
Regional
0
D3IEP
V »
REOIOM
Osia m at Sff/BS
BACKGROUND
The Superfund Amendments and Reauthorization Act
(SARA) of 1986 required that EPA revise the Hazard
Ranking System (HRS), me primary mechanism used
to list sites on the National Priorities List (NPL). In
December 1990, EPA promulgated and published the
revised HRS in the Federal Register (55 FR 51532),
superseding the original HRS. During a period of
transition to the revised HRS, sites were evaluated
through the SI stage under the original HRS;
however, EPA felt it would be preferable to make
final site disposition decisions on these sites after
revising me HRS. Information for these sites needs
to be updated to evaluate the site using the revised
HRS. A final decision may be to list the site on the
NPL, make a Site Evaluation Accomplished (SEA)
determination, or defer the site to another Federal
authority (e.g. Resource Conservation and Recovery
Act (RCRA) or Nuclear Regulatory Commission
(NRC)). An SEA decision means that, based on
currently available information, the site does not meet
the criteria for inclusion on the NPL and Federally
funded remediation. Sites designated SEA are
subsequently referred to the appropriate State agency
for any further action.
The goal of the SEP is to gather any additional
information necessary, following the completion of
the SL to help set priorities among these sites for
NPL listing or to screen mem from further Superfund
attention. At a minimum, this would generally
require gathering data to update the site evaluation
and determining whether the HRS score Ls greater
than 28.5. Typical SEP data gathering efforts may
37
-------
include collecting additional site information (e.g.
historical use) and "target" information (e.g., wells
within 4 miles, surface water intakes, fisheries and
sensitive environments within 15 miles downstream).
SI?a may also entail collecting limited samples if this
is required to make a screening decision. The
number of samples for an SIP should range from a
few up to the normal number typically collected for
an SI. An SIP should rarely result in the need for
further investigation through the Expanded Site
Inspection (ESI) stage. ESIs should be reserved for
those sites clearly headed for the NPL and where
significant fieldwork (e.g., well installation or
extensive air monitoring) or other non-routine data
collection activities are necessary.
The SEP is a temporary, intermediate step in the Site
Assessment program to update old Sis and make
screening decisions on a discrete universe of sites
using minimal resources. Therefore funding for SEP
activities is expected to be available for the next two
to four years. The SIP backlog should not continue
to grow. Current Preliminary Assessments (PAs) and
Sis are being completed according to guidance
developed for the revised HRS (see Guidance for
Performing Preliminary Assessments Under CERCLA,
OSWER Directive 9345.1-01A, September 1991 and
Guidance for Performing Site Inspections Under
CEROLA, OSWER Directive 9345.1-05, September
1992). The updated guidance documents recommend
the use of intermediate scoring tools (PA Scoresheets,
PA-Score, and SI Worksheets) to make screening
decisions using site information normally available at
the PA and SI stages. These scoring tools typically
require less site information and effort to make a
screening decision than using PREscore. At the SIP
stage the majority of sites should be scored using SI
Worksheets at a minimum; however, the decision of
which scoring tool (PA Scoresheets, PA-Score, SI
Worksheets, or PREscore) to use for SIPs will be
made on a site by site basis.
REVISED HRS REQUIREMENTS
EPA revised the HRS to comply with the
requirements set forth by SARA. To better assess
the relative degree of risk to human health and me
environment, EPA modified the approach for
evaluating the ground water, surface water, and air
migration pathways that were addressed in the
original HRS and incorporated a direct exposure
pathway (soil exposure) into the composite score used
to evaluate sites. In general, the HRS score reflects
the risk associated with each pathway by estimating.
(1) the likelihood of a release of hazardous
substances; (2) the quantity and toxiciry or other
harmful characteristics of on-site wastes; and (3) the
risk to both human and environmental targets.
The revised HRS requires more data than the original
HRS to evaluate a site. The revised HRS evaluates
ground water discharge to surface water, human food
chain exposure, soil exposure, and the potential for
air release. In addition, the evaluation of risk to
ecosystems or environmental targets is broader in the
revised HRS than the original HRS, and the
calculation of waste quantity is more comprehensive.
However, the most crucial information that will need
updating for an SIP is the target data. Table 1
presents a comparison of target needs between the
original and the revised HRS models and identifies
what information might be missing in old Sis.
SEP UNIVERSE
In general, an SIP should be assigned for non-Federal
facility sites which had Sis completed prior to th
implementation of the revised HRS and have nc
received a final NPL decision. SIPs are appropriate
where more information Ls necessary to determine
whether a site should be screened out (designated
SEA or deferred to another Federal authority), or
investigated further for probable inclusion on the
NPL. The extent of additional information required
to make this final decision and the probability of NPL
listing are both important in determining whether an
SIP or an ESI is appropriate. SIP candidates are sites
with an SI completion date entered into CERCLIS
prior to the implementation of the revised HRS (i.e.,
August 1, 1992) and may include: (1) sites without
an event qualifier (i.e., high priority, low priority,
SEA, or deferred), or (2) sites with an event qualifier
of high or low priority.
Sites may not have event qualifiers due to CERCLIS
coding errors. The Region should review tile
information for these sites tc determine whether a
decision is possible. Sufficient information may be
available to screen out the site from further CERCL J
investigation (SEA or defer to anodier Fede^b
authority). The appropriate event qualifier should
then be entered into CERCLIS. If additional
38
-------
Table 1: Original vs. Revised HRS Target Data Requirements
PATHWAY
ORIGINAL HRS
TARGETS
REVISED HRS
TARGETS
Ground Water Migration
Pathway
Ground water use
Distance to the nearest
well/population served within 3
miles
Distance from a source to the nearest drinking water "-veil
Population served by drinking water wells within four
miles
Apportioned population for blended systems
Resources
Wellhead protection areas
Surface Water Migration
Pathway
Overland/Flood
Migration Component
Surface water use
Distance to the nearest intake/
population served
Distance to sensitive environment
Drinking Water Threat;
• Distance to nearest drinking water intake
• Average flow (cubic feet per second)
• Population served by drinking water drawn by intakes
along the surikcc water migration pathway within 15
downstream miks
• Apportioned population for blended systems
" Resources
Human Food Chain Threat:
• Location of fisheries
• Annual harvest (in pounds) of human food chain organisms
Environmental Threat:
« Location of sensitive environments
• Wetlands frontage length (in miles)
Ground Water to Surface
Water Migration
Component
NA
Same as above
New component to the surface water pathway. If both the
overland/flood and the ground water to surface water
components are scored, the greater of the two component
scores ii selected.
Soil Exposure Pathway
NA
The original HRS included a direct
contact pathway, but that pathway
was not i*lr*d*trd in the overall
HRS migration score.
Resident Population Threat:
• Number of individuals who live, work, attend school or
day care within property boundaries and within 200 feet of
observed contamination
« Location of terrestrial sensitive environments within the
area of observed contamination
a Resources
Nearby Population Threat:
« Number of individual* who live or attend school within a
one-mile travel distance from any source with observed
contamination
» Attractivcaesa/accesaibility of sources
Air Migration Pathway
Land use
Population wkhin four miles
Distance to sensitive
environments
a Distance from an emission source to the nearest individual
a Population within a four-mile radius of sources
• Resources within one-half mile of sources
« Distance from sources to sensitive environments within
four miles of sources
• Total wetland acreage within four miles of sources
39
-------
information is required to make a final decision,
either a high priority or low priority recommendation
should be entered as the event qualifier (Regional
guidance should be consulted to distinguish between
higb and low priority recommendations). These sites
would then fall into the second category of sites.
Sites with high or low priority event qualifiers may
still require additional data, eidier desktop or
analytical, to support a revised HRS score. Since die
average level of effort per SIP is 190 hours, sites
which require significantly more effort to support a
final decision (> 550 hours) may not be appropriate
for an SIP assignment. If a greater level of effort is
required to fill SI data gaps and to support a final
decision, particularly if the site is likely to be listed
on the NPL, then an ESI may be more appropriate.
In general, an ESI should be assigned if extensive
fieldwork. or unusual data collection activities are
required prior to preparing an NPL package. Factors
mat should be examined to help determine whether an
SIP or an ESI should be assigned are provided in
Table 2. In addition to sites with and without SI
event qualifiers, there may be other sites evaluated
under the original HRS which need further evaluation
via the SIP. This includes sites where an incorrect
site decision was made or new file information is
made available which might significantly alter the
HRS evaluation of the site. The decision as to
whether an SIP is appropriate for these sites will be
made on a case by case basis. Note, however, that
these sites are not included as part of ,me SEP bacMog
(Figure 1) because site decisions have already been
entered into CERCLIS.
SETTING PRIORITIES
Figure 1 shows chat some Regions have a very lar^
SIP backlog and may need to incorporate additional
measures to set priorities among their SEP candidates.
One mernod may be to identify sites located in
vulnerable geographic areas. Features that may
characterize an area as vulnerable include: population
density, geologic and hydrogeologic features, surface
water intakes, fisheries, municipal drinking water
wells, wetlands, and other considerations. Much of
mis information is available in existing databases that
can be incorporated into a Regional geographic
information system (GIS). By plotting rnese features,
sites located in vulnerable areas could then be
identified.
The issue of setting priorities is not limited to SIPs.
If further investigation is warranted at die conclusion
of an SEP, these sites, which are now ESI candidates,
must also be prioritized. It is critical to set priorities
for ESI candidates, not only to comply with EPA's
policy of assessing the worst sites first, but also to
allow for the efficient use of limited resources. ESIs
consume an average of 1,000 hours; therefore
Regions must set priorities for ESI candidates whl
keeping in mind their budgetary constraints. Region
may use more detailed prioritization schemes to
further identify ESI candidates.
Regions are encouraged to investigate their Regional
GIS capabilities as well as other prioritization
methods, not only to address the SEP backlog, but
also to help direct other environmental protection
efforts on a worst sites first basis.
Table 2: Factors to Determine SIP or ESI Assignment
SIP
Limited data are necessary to determine whether or
not the site will attain & score greater than the cutoff
score for NPL eligibility
NPL eligibility is uncertain
SI completed but no HRS score calculated
SI completed but preliminary HRS scoring assumes
primary targets without sample results
ESI
Substantial data collection is necessary to
HRS package (>550 hours)
prepare NPL quality
Probable NPL site
SI score (completed wi& SI Worksheets)
score for NPL eligibility
is greater than the cutoff
SI sampling has verified contamination at primary targets; site
score is greater than the cutoff score for NPL eligibility
40
-------
SEPs and SACM
The Superfund Accelerated Cleanup Model (SACM)
requires better integration of all Superfund program
components to make cleanups more timely and
efficient. During an SEP, activities should be
coordinated to ensure that data collected support
assessment, enforcement, and response activities.
The Regional Decision Team (RDT), which is
responsible for making site decisions to ensure early
risk reduction, will establish the strategy for
addressing sites. SIP data collection efforts should be
consistent with these strategy decisions.
The basic principles of SACM assessment are built
upon the need to eliminate redundancy and expedite
the Superfund process. SEPs will help identify
priority sites so that EPA resources are expended on
sites that require prompt risk reduction. For further
information, refer to Assessing Sites Under SACM •
Interim Guidance OSWER Directive 9203.1-051,
December 1992.
SIP ACTIVITIES
Activities to be conducted for an SIP will depend on
the additional information necessary to update the old
SI in accordance with current guidance and the
revised HRS. In all cases, however, site information
must be reexamined to update current site conditions,
satisfy revised HRS requirements, and identify me
potential need for removal actions (see the next
section for further information on identifying
potential removal actions during SIPs).
Because activities necessary for SEPs will vary due to
the quality and comprehensiveness of site information
that is available, three levels have been identified to
meet the goal of an SEP as illustrated in Table 3.
It is estimated that each SIP will require at least
updated HRS scoring; two-thirds will require desktop
data collection and updated scoring; and one-third
will require desktop data collection, updated scoring,
and limited sampling. It is estimated that an SEP will
average 190 hours per site. EPA Regional staff will
help determine the level of effort necessary for each
SEP assigned. The choice of which HRS scoring tool
to use (PA Scoresheets, PA-Score, SI Worksheets, or
PREscore) will depend on the amount of information
available for the site. At a minimum, revised HRS
scores must support each SEP decision. Sites with
revised HRS scores below the cutoff (28.5) are
screened out, and will receive an SEA decision in
CERCLIS. Sites with revised HRS scores above the
cutoff will be recommended for either an ESI (if
extensive information/data collection is still required)
or for a full HRS package.
LeveJ A: The first step in conducting an SIP is to
generate a revised HRS score. This typically
requires collecting new target information for the
revised HRS target distance limits (see Table 1}.
This first activity is necessary for ajl SEPs to generate
a site score according to the revised HRS, identify
data gaps, and determine whether additional SEP
activities (desktop data collection and/or sampling)
are necessary to make a final site disposition
decision. If desktop data collection and sampling are
not necessary, the site's revised HRS score should be
documented with these new target data incorporated,
and a final site disposition decision should be entered
into CERCLIS.
Level B: Most SEPs will likely require the collection
of additional site specific desktop data beyond
Table 3: SIP Levels versus SIP Activities
SB? LEVEL
LEVEL A
LEVELS
LEVEL C
SEP ACTIVITIES
« Updated (revised HRS) scoring
• Updated (revised HRS) scoring
« Desktop data collection
• Updated (revised HRS) scoring
« Desktop data collection
« Limited Sampling
APPROXIMATE
TOTAL HOURS
40-60
80- 100
350 - 550
41
-------
updating a site's target information. Level B SIPs
typically include researching and updating site
information because site conditions may have changed
significantly since the completion of the old SI. All
-appropriate data sources (EPA, State, municipal, etc.)
should be researched to ensure that information is
updated for the SIP. The SI Data Summary
(Appendix B of the Guidance for Performing Sue
Inspections Under CERCLA) and the Site Assessment
Information Directory (SAID) may be useful data
collection tools for this task. After this new
information is collected, the site's revised HRS score
should be documented incorporating this new data,
and a final site disposition decision should be entered
into CERCLJS.
Level C: It is estimated that approximately one-third
of SIPs will require sampling activities hi addition to
the activities described for Level A or B SEPs. EPA
Regional staff will determine the appropriate
sampling strategies necessary for Level C SIPs.
After sampling activities are conducted, the site's
revised HRS score should be documented
incorporating the new analytical data, and a final she
disposition decision should be entered into CERCLIS.
Site visits may be necessary for Level A, B, or C
SIPs to verify and update site conditions, evaluate the
need for a potential removal action, identify target
information for BCRS scoring, and/or conduct
sampling activities. As a cost savings measure,
Regions should consider scheduling SIPs hi
geographic clusters so that site,' visits can be
combined. SIP field activities may also be combined
with other Site Assessment and integrated assessment
field activities.
SIP products will depend on what activities are
conducted and will be determined by EPA Regional
staff. For example, the final product for an SIP
requiring only Level A activities may consist of a
brief memo updating site and target information along
with completed revised HRS scoresheets. Products
for an SIP requiring Level B or C activities may
consist of a full report, similar to the SI reporting
format, along with completed HRS scoresheets.
SEPs will be tracked in CERCLIS as a subevent of
the SI. Refer to the sidebar for the appropriate
CERCLIS entry protocols.
CERCLIS Data Entry for SIPs
8 Enter the SIP as a subevent (SP) to the
last SI.
8 The completion date is the date the report
is accepted by the Region and a
disposition (event qualifier) is made on
the report.
9 Replace the last SI event qualifier with
the new SIP event qualifier (Higher,
Lower, Deferred, or SEA).
Only sites where the last SI completion date is
prior to August 1, 1992 are eligible for SIPs.
Sites with SI qualifier "N" or "D" are not
eligible for SEPs unless new information
relevant to the decision becomes available.
IDENTIFYING POTENTIAL REMOVAL
ACTIONS DURING SIPs
SEP field activities can be very useful in identifying
sites where a potential removal action may b
necessary. Removals are relatively short-ters
actions, compared to the long-term remedial solutions
that the NPL addresses. Removal actions are
designed to respond to situations that require
immediate action to diminate a present threat or to
avoid a more serious future problem (for example,
containing leaking drums of hazardous substances to
prevent ground water contamination). Removal
actions can be of an emergency, time-critical, or non-
time-critical nature and can include, but are not
limited to, any of the following (see Superfund
Removal Procedures, OSWER Directive 9360.3-01,
February 1988):
* Fencing the site
• Providing 24 hour security to restrict public
access
e Stabilizing waste sources such as leaking
drums or overflowing surface
impoundments
8 Physical removal of hazardous substances
6 Capping areas of obvious contamination
8 Assessing me need to temporarily relocate
populations
0 Providing alternative drinking water
supplies
42
-------
Table 4 outlines the factors that EPA considers in
determining the appropriateness of a removal action
pursuant to section 300.415(b) of the National Oil
and Hazardous Substances Pollution Contingency
Plan, commonly known as the NCP (40 CFR Part
300).
Under the revised HRS, waste removals may be
considered for HRS scoring purposes under certain
circumstances. For more information concerning the
requirements for considering removal actions, refer
to The Revised Hazard Ranking System: Evaluating
Sites After Waste Removals, OSWER Directive
9345.1-03FS, October 1991.
SIP SAMPLING
Because SIPs that include sampling will require
significantly more hours, some analysis should be
conducted to determine if sampling is critical for
making a final decision. For example, if a
preliminary site score is 28.5 or greater, all targets
for which actual contamination (level I and II) is
suspected should be identified. By examining various
scoring scenarios, the site score should be tested to
determine whether the site score will fell below the
28.5 cutoff for NPL eligibility if the targets with
suspected actual contamination are scored as
potentially contaminated. If the site score falls below
28.5 with this modification, sampling is necessary to
verify the suspected contamination of the target
receptors. If the site score remains 28.5 or greater.
sampling may not be necessary. On the rare occasion
where an SIP results in the need for further
investigation via an ESI, new data obtained from
samples collected for me SIP may help set priorities
for sites needing further work.
For additional information concerning sampling
guidance for the revised HRS requirements, refer to
Guidance for Performing Sice Inspections Under
CERCLA, OSWER Directive 9345.1-05 and the
Hazard Ranking System Guidance Manual, OSWZR
Directive 9345.1-07.
In summary, the goal of the SIP is to gather any
additional information necessary, following the
completion of the SI, to make decisions on this
discrete universe of sites. Activities conducted for an
SEP should be consistent widi current guidance,
including SACM, and should result in sites being
either removed from further Superfund attention or
recommended for NPL package preparation.
EPA developed the SIP as a cost effective,
intermediate step in the Site Assessment process to
screen out less serious problems and expedite action
at sites that require additional Superfund response.
Site priorities must continue to be set on a worst first
basis to ensure that Superfund cleanups are timely
and efficient.
Table 4: Remoyai Action Criteria
1. Actual or potential exposure to nearby human populations, animals, or the food chain from hazardous substances
or pollutants
2. Actual or potential contamination of drinking water supplies or sensitive ecosystems;
3. Hazardous substances or pollutants or contaminants in drums, barrels, tanks, or other bulk storage containers,
that may pose a threat of release;
4. High levels of hazardous substances or pollutants or contaminants in soils largely at or near the surface, that
may migrate;
5. Weather conditions that may cauje hazardous substances or pollutants or contaminants to migrate or be released;
6. Threat of fire or explosion;
7. The availability of other appropriate Federal or State response mechanisms to respond to the release; and
8. Other situations or factors that may pose threats to public health or welfare or the environment.
43
-------
REFERENCES
U.S. Environmental Protection Agency, 1988. Superfund Removal Procedures, OSWER Directive 9360.3-01.
U.S. Environmental Protection Agency, 1990. Hazard Ranking System, Final Rule, 55 FR 51532, December 14,
1990.
U.S. Environmental Protection Agency, 1990. National Oil and Hazardous Substances Pollution Contingency Plan.
40 CFR Part 300. 55 FR 8666, March 8, 1990.
U.S. Environmental Protection Agency, 1991. Guidance for Performing Preliminary Assessments Under
CERCLA, OSWER Directive 9345.1-01 A.
U.S. Environmental Protection Agency, 1991. Revised Hazard Ranking System: Evaluating Sites After Waste
Removals, OSWER Directive 9345.1-03FS.
U.S. Environmental Protection Agency, 1991. Site Assessment Information Directory (SAID), October 31, 1991.
U.S. Environmental Protection Agency, 1992. Assessing Sites Under SACM—Interim Guidance, OSWER
Directive 9203.1-051.
U.S. Environmental Protection Agency, 1992. Guidance for Performing Site Inspections Under CERCLA,
OSWER Directive 9345.1-05.
U.S. Environmental Protection Agency, 1992. Hazard Ranking System Guidance Manual, OSWER Directive
9345.1-07.
Additional copies can be obtained from:
National Technical Information Service (NTIS)
U.S. Department of Commerce
5285 Port Royal Road
Springfield, VA 22161
(703) 487-4600 or (800) 3364700
Order*: PB93-963340
or
EPA Employees
Superfund Documents Center (SDQ
U.S. Environmental Protection Agency
401 M Street, SW (OS-245)
Washington, DC 20460
(202) 260-9760 or (202) 260-2596 (FAX)
44
-------
APPENDIX II
ABBREVIATIONS
CERCLA
CERCLIS
EPA
ESI
FMFIA
GAO
HRS
HSED
NCP
NPL
OERR
OSWER
DIG
OLM
PA
PRP
RI
SACM
SARA
SI
SIP
Comprehensive Environmental Response,
Compensation, and Liability Act of 1980
Comprehensive Environmental Response,
Compensation, and Liability Information
System
Environmental Protection Agency
Expanded Site Inspection
Federal Managers' Financial Integrity Act
General Accounting Office
Hazard Ranking System
Hazardous Site Evaluation Division
National Contingency Plan
National Priorities List
Office of Emergency and Remedial Response
Office of Solid Waste and Emergency Response
Office of Inspector General
Outyear Liability Model
Preliminary Assessment
Potentially Responsible Parties
Remedial Investigation
Superfund Accelerated Cleanup Model
Superfund Amendments and Reauthorization Act
of 1986
Site Inspection
Site Inspection Prioritization
45
Report No.' E1SFF3-08-0021-4100180
-------
APPENDIX III
DISTRIBUTION
Headquarters Office
for Regional Operations
(1501)
and
Inspector General (2410)
Assistant Administrator for Solid Waste and Emergency
Response (5101)
Assistant Administrator for Administration and Resources
Management (3101)
Associate Administrator
State/Local Relations
Associate Administrator for Congressional and
Legislative Affairs (1301)
Associate Administrator for Communications,
and Public Affairs (1701)
Director, Resource Management Division (3304)
Headquarters Audit Follow-up Coordinator (3304)
Headquarters Library (3404)
Education,
Regional Offices
Regional Administrator, Regions 1 - 10
46
Report No. E1SFF3-08-0021-4100180
------- |