-------
Inspector General Divisions
 Conducting the Audit:
Western Audit Division,
Lead Division

Northern Audit Division

Southern Audit Division
Regions Covered:
Region  2

Region  4

Region  5

Region  6

Region  7

Region  9

Region 10
Program Office Involved:
Office of Solid Waste and Emergency Response

-------
               UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                          WASHINGTON, D.C. 20460
                                                             OFFICE OF
                                                        THE INSPECTOR GENERAL
                          August  6,  1996
MEMORANDUM
SUBJECT:   Consolidated Report on EPA's Leaking  Underground
           Storage Tank Program, Report No. E1LLF5-10-0021-6100264
FROM:     Michael Simmons
          Deputy Assistant Inspector General  for
             Internal and Performance Audits

TO:       Elliott Laws
          Assistant Administrator for
             Solid Waste and Emergency Response

     Attached  is our final report titled Consolidated Report on
EPA ' s Leaking  Underground Storage Tank Program .   The  purpose of
this consolidated report is to identify performance issues that
need attention so that Agency managers can target resources to
improve the  overall program.

     This report identifies corrective actions the Office of
Inspector General (OIG)  recommends involving  EPA's Leaking
Underground  Storage Tank (LUST)  program.  As  such, it represents
the opinion  of the OIG.   Final determinations on  the  matters in
the report will be made by EPA managers in accordance with
established  EPA audit resolution procedures.  Accordingly, the
findings described in this report do not necessarily  represent
the final EPA  position and are not binding upon EPA in any
enforcement  proceeding brought by EPA or the  Department of
Justice .

     In accordance with EPA Order 2750, we have designated the
Assistant Administrator for Solid Waste and Emergency Response as
the Action Official for this  report.  The Action  Official is
required to  provide this office with a written response to the
audit report within 90 days of the final report date.   The
                                                      Printed on Recycled Paper
                                                        Recycled/Recyclable
                                                        Printed with Soy/Canola Ink on paper that
                                                        contain* at laast 50% recyded fiber

-------
response should address all recommendations.  For corrective
actions planned but not completed by the response date, reference
to the specific milestone dates will assist us in deciding
whether to close this report.  We have no objection to the
release of this report to the public.

     Should you or your staff have any questions about this
report, please call Truman R. Beeler, Divisional Inspector
General for Audit,  Western Division at (415) 744-2445, or Charles
Reisig, Team Leader at (206)  553-4032.

Attachment

-------
                        TABLE OF CONTENTS




                                                             Page




EXECUTIVE SUMMARY 	  1




CHAPTERS




     I -  INTRODUCTION	  1




            PURPOSE	  1




            BACKGROUND	  1




            SCOPE AND METHODOLOGY	  4




            EVIDENCE EXAMINED	  5




            INTERNAL CONTROLS	  6




     2 -  SOME HIGH RISK SITES NOT CLEANED UP 	  7




     3 -  COST RECOVERY PROGRAMS NEED IMPROVEMENTS	 23




     4 -  ACTIVITY REPORTS MISLEADING	 33




APPENDIX




     APPENDIX A - ACRONYMS	 45




     APPENDIX B - OSWER RESPONSE TO DRAFT REPORT 	 47




     APPENDIX C - REPORT DISTRIBUTION 	 61

-------
                     EXECUTIVE SUMMARY
PURPOSE
BACKGROUND
Congress
Created the
LUST Trust Fund
The purpose of this consolidated report is to
identify issues that need attention so that
Agency managers can target resources to
improve the overall program.  This report
summarizes the issues relating to the Leaking
Underground Storage Tank (LUST) Program which
have been identified by Office of Inspector
General (OIG) audits.  Several OIG audits
identified issues involving States' oversight
and enforcement of the clean-up of
significantly contaminated sites, cost
recovery of LUST Trust Fund expenditures, and
activity reporting.

The LUST Program's current focus is on
program results and not on adherence to
prescribed procedures.  We have identified
three areas where program results should be
improved.
Leaking underground storage tanks pose
significant environmental risks throughout
the country.  Because of national problems
with leaking underground tanks, the Congress
amended the Resource Conservation and
Recovery Act (RCRA) in 1984 to regulate
underground storage tanks.  In 1988, the
Federal rules for regulating tanks took
effect.  The rules, promulgated in the Code
of Federal Regulations (40 CFR 280 to 281),
require underground storage tank systems to
be designed and upgraded to protect against
leaks.

To further address the need to clean-up
contamination from leaking tanks, the
Congress passed the Superfund Amendments and
Reauthorization Act (SARA) in 1986.  This Act
created a $500 million LUST Trust Fund,
financed by one-tenth of a cent tax on each
gallon of gasoline, diesel, and aviation
fuels sold.  The Fund is to be used to
finance the clean-up of petroleum releases
from underground storage tanks.  The Congress

-------
States operate
the LUST
Program with
EPA Oversight
RESULTS IN
BRIEF
subsequently removed the $500 million ceiling
and by 1995 the Fund had grown to over $1.5
billion.

SARA authorized EPA to provide LUST Trust
Funds to the States to operate the program
and clean-up the contaminated sites.  States
are expected to play the primary or lead role
in managing the clean-up process because they
are better able to respond quickly and
effectively to leaks due to their proximity
to, and experience with, local site
conditions.

The EPA's role is to establish the regulatory
criteria, aid the States in establishing and
implementing their programs, and evaluate the
States' implementation of their programs
according to the work plan requirements in
the cooperative agreements.
The results of OIG audits are consolidated in
this report.  Those audits reviewed nine
States' LUST programs which received 26
percent of the total LUST Trust Fund
allocation for fiscals 1987 - 1995.  We
focused on three areas:  (i) oversight and
enforcement of high risk sites; (ii) cost
recovery from responsible owners and
operators; and (iii) States' reporting of
program results.

We found that the States' LUST programs can
be improved.  Specifically, we found that
States could improve:

•    Oversight and enforcement on some high
     risk sites.

•    Cost recovery from owners and operators
     who are responsible.

•    Reporting of program results.

Although States were making notable progress
in cleaning-up many LUST sites, our reviews
of 249 high risk sites disclosed that 126
(one half)  had not received appropriate
corrective action.  The States did not focus
sufficient resources to remediate potentially
                               11

-------
PRINCIPAL
FINDINGS
Oversight and
Enforcement
Needed for Some
High Risk Sites
Some Cost
Recovery
                    significant threats to human health and the
                    environment.

                    There were major differences in results
                    between the States' cost recovery programs.
                    Some States had effective programs that
                    recovered significant amounts from owners and
                    operators, other States achieved partial
                    recoveries, and two States did not have cost
                    recovery programs.  One result of ineffective
                    cost recovery programs is that owners and
                    operators of leaking tanks were not held
                    liable for the costs of cleaning-up their own
                    sites.

                    We found that the States' activity reports
                    (reporting of program results) were generally
                    unreliable and potentially misleading.
Our findings are summarized below and
discussed in detail in CHAPTERS 2 through 4
of this report.

We found that five States could improve
oversight and enforcement to remediate some
high risk sites.  We reviewed 249 high risk
sites and found that 126 sites did not get
adequate oversight or enforcement to ensure
protection of human health and the
environment.

Overall, States were making notable progress
in cleaning-up LUST sites, but some high risk
sites needed more oversight and enforcement
actions.  Under the LUST Program, States are
responsible for establishing clean-up
programs that protect human health, and the
environment.  Some States have more effective
programs than other States.  For example,
Ohio did a good job of managing LUST
activities and directing its resources toward
the clean-up of high priority sites.
However, California, Idaho, Kansas, New York,
and Oregon could have directed more attention
toward oversight and enforcement of
responsible party clean-ups for some serious
sites.

For the States in our review, there were
major differences among the cost recovery
                               111

-------
Programs Need
Improvement
Activity
Reports Were
Not Accurate
RECOMMENDATIONS
programs and their success in achieving
intended results.  We found that: (i) two
States maintained effective cost recovery
programs; (ii) three States needed
improvements in their cost recovery programs;
and (iii) two States did not have cost
recovery programs.

The primary purpose of a cost recovery
program is to provide incentives for owners
and operators to clean-up releases from their
own tanks.  When States do not recover clean-
up costs from those owners and operators who
are responsible for paying, the States is:
(i) relieving them of their financial
obligation; (ii) removing the financial
incentive to clean-up contamination; and
(iii)  shifting the burden of clean-up costs
to the Federal government.  Ineffective cost
recovery programs hinder the efforts of the
LUST Program to protect human health and the
environment.

We found that the quarterly activity reports
submitted by States to EPA did not accurately
report LUST Program results.  For the States
reviewed, results were generally overstated.

The activity reports are used as a management
tool to track the most important program
activities.  Therefore,  it is necessary that
the information be accurate, complete, and
comparable for effective management of the
LUST Program.
Specific recommendations follow the findings
in CHAPTERS 2 through 4.  In summary, we are
recommending that the Assistant Administrator
for Solid Waste and Emergency Response
implement management controls to ensure that:

1.   States provide adequate oversight and
     enforcement on high risk sites.

2.   States develop and implement cost
     recovery programs.

3.   Reliable and timely information is
     obtained, maintained,  reported and used
     for decision making.
                               IV

-------
AGENCY COMMENTS
AND OZ6
EVALUATION
OI6 Evaluation
A draft report was provided to the Office of
Solid Waste and Emergency Response (OSWER) on
May 16, 1996.  The OSWER responded to the
draft report on June 27, 1996.  The response
is summarized below and comments on specific
audit findings are described at the end of
the chapters.  The response generally did not
acknowledge the problems identified in our
report, but it either agreed with the
recommendations or presented alternative
actions that meet the intent of our
recommendations.

The response said that the draft report did
not capture the tremendous success of its
State partners in completing over 140,000
LUST clean-ups, nor the magnitude of the
States' workload.  As of March 31, 1996,
there were over 314,000 confirmed releases
from underground storage tanks.  According to
OSWER these numbers are "astronomical and
unprecedented in other environmental
programs."

OSWER stated that the audit report
conclusions propose a greater level of
oversight and a more intrusive role for EPA
vis-a-vis state programs.  It agreed that
management and accountability are important,
but those requirements present a difficult
quandary about asking States to do more with
less.  However, OSWER did describe some
actions that have been,  or will be,  taken to
address the recommendations in our report.

We believe that OSWER's response does not
appropriately address the issues described in
our report.

We believe that OSWER7s emphasis on "the
magnitude of reported releases" circumvents
the issues.  We specifically focused our
audits on those sites that posed the greatest
threat to human health and the environment.
OSWER's claim of 314,000 confirmed releases
does not appropriately distinguish between
sites where health and/or environmental risks
are serious (a small portion of confirmed
releases)  and those sites where the risk is
relatively insignificant (a majority of the

-------
sites).  As described in CHAPTER 2, our
conclusion is based on the testing we
completed for the high risk sites we sampled.
We found that 125 of 249 high risk sites did
not receive sufficient oversight or
enforcement.

Additionally, the 314,000 confirmed releases
and 140,000 clean-ups cited by OSWER are
suspect.  As described in CHAPTER 4, we found
that the numbers were generally overstated.

We believe that we have given the States
credit for making significant progress.  When
we found that a State had an effective
program in one or more of the three areas, we
gave recognition to that State in the report.

Finally, we believe that our recommendations
do not require an unreasonable level of
oversight and a more intrusive role for
OSWER.  Our recommendations are in accordance
with the Government Performance and Results
Act of 1993 (GPRA) which support managing for
results.  Also, the recommendations are not
in conflict with the performance partnership
initiatives which have requirements for
accountability.  The recommendations in
CHAPTER 2 relate to the more efficient use of
scarce resources by focusing them on the
higher risk sites.  The recommendations in
CHAPTER 3 relate to cost recoveries from
responsible parties which would provide
incentives to clean-up releases from their
own sites.  The recommendations in CHAPTER 4
relate to the establishment of program
accountabi1ity.
           VI

-------
                         CHAPTER 1
                        INTRODUCTION
PURPOSE
BACKGROUND
Health Risks
The purpose of this consolidated report is to
identify performance issues that need
attention so that Agency managers can target
resources to improve the overall program.
This report summarizes the issues relating to
the LUST Program which have been identified
by OIG audits.  Several DIG audits identified
issues involving States' oversight and
enforcement of the clean-up of significantly
contaminated sites, cost recovery of LUST
Trust Fund expenditures, and activity
reporting.

The LUST Program's current focus is on
program results and not on adherence to
prescribed procedures.  'We have identified
three areas where program results should be
improved.
Leaking underground storage tanks pose
significant environmental risks throughout
the country.  Because of national problems
with leaking underground tanks,  the Congress
amended the RCRA in 1984 to regulate
underground storage tanks.   In 1988 the
Federal rules for regulating tanks took
effect.  The rules, promulgated in the Code
of Federal Regulations (40  CFR 280 to 281),
required underground storage tank systems to
be designed and upgraded to protect against
leaks.

Leaking tanks pose a long term health threat
because they usually contain petroleum
products, which contain many harmful
substances.  For example, benzene is a known
carcinogen which is often found in fuels and
related fuel vapors.  Contact with petroleum
products can cause adverse  effects including
nausea; irritation of the skin,  eyes,  and
throat; liver and kidney damage; and loss of
reflexes.

-------
Congress
Created The
LUST Trust Fund
States Receive
LUST Funds
Through
Cooperative
Agreements
To further address the need to clean-up
contamination from leaking tanks, the
Congress passed the SARA in 1986.  This Act
created a $500 million LUST Trust Fund,
financed by one-tenth of a cent tax on each
gallon of gasoline, diesel, and aviation
fuels sold.  The Fund is to be used to
finance the clean-up of petroleum releases
from underground storage tanks.  The Congress
subsequently removed the $500 million ceiling
and by 1995 the Fund had grown to over $1.5
billion.

SARA authorized EPA to provide LUST Trust
Funds to the States to operate the program
and clean-up the contaminated sites.  States
are expected to play the primary or lead role
in managing the clean-up process because they
are better able to respond quickly and
effectively to leaks due to their proximity
to, and experience with, local site
conditions.

The EPA regional offices provide LUST Trust
Funds to the States for implementing the
program through cooperative agreements (CAs).
States use this money to finance:  (i) the
cost of site corrective actions;
(ii) enforcement action undertaken to get
owners and operators (responsible parties) to
take corrective actions; (iii) oversight of
responsible-party lead clean-ups; (iv) cost
recovery efforts of Trust Fund expenditures;
and (v) reasonable and necessary
administrative expenses directly related to
these activities.

Each LUST Trust Fund CA includes a work plan
that establishes requirements for the State's
clean-up program.  In general, the States
oversee the clean-ups undertaken by the
responsible parties.  This oversight involves
monitoring the clean-up plans, clean-up
progress, and disposal of contaminants.  The
States also undertake the clean-up of sites
where a responsible party cannot be
identified or where the responsible party is
insolvent.  The EPA's role is to establish
the regulatory criteria, aid the States in
establishing and implementing their programs,
and evaluate the States' implementation of

-------
                    their programs  according to the work plan
                    requirements.
Allocation of
Trust Funds to
the States
   Leaking underground tank removal and clean-up

A total of 55 States and territories  have
CAs.   From fiscals 1987 to  1995  over  $545
million in Trust Funds were appropriated to
EPA,  and EPA awarded almost $473  million to
the States.  States reported cumulative Trust
Fund expenditures totaling  over  $361  million
as of the end of fiscal 1995.

-------
                    LUST Trust Funds allocated to the States
                    through CAs for fiscal 1995 totaled $61.2
                    million.  The following graph illustrates how
                    the fiscal 1995 LUST Trust Fund allocation
                    was distributed among the 10 EPA regions.


                      FISCAL 1995 LUST TRUST FUND ALLOCATION
                                   Dollar Amount by EPA Region
                            Millions
                                      4567

                                        EPA Region
                               8  9
10
                    To supplement LUST Trust Funds, 46 States
                    have established State financial assurance
                    funds to help owners and operators satisfy
                    the Federal statutory requirement for
                    evidence of ability to pay the costs of
                    corrective action.  While these funds
                    collected more than $1 billion annually, many
                    are beginning to face solvency issues as
                    reimbursement requests increase.
SCOPE AND
METHODOLOGY
We performed our audit in accordance with
Government Auditing Standards -issued by the
Comptroller General.  This consolidated
report represents a summation of findings
from nine DIG audits of States' LUST
programs.  The audits covered the periods
between August 1987 to October 1995.  LUST
Trust Funds allocated to the nine States
totaled $124 million out of $473 million  (26
percent) for fiscals 1987 - 1995.

The scope of this consolidated report was
limited to the adequacy of the States'
procedures in three areas:  (i) oversight and

-------
                    enforcement of corrective actions for high
                    risk sites  (those sites that had contaminated
                    drinking water or otherwise had been
                    determined to be a significant threat to
                    human health and the environment);  (ii) cost
                    recovery of Trust Fund expenditures; and
                    (iii) activity reporting.

                    Audit findings in three areas were reported
                    in seven audit reports and two sets of
                    position papers to the individual States.
                    The seven reports were issued to California,
                    Idaho, New Mexico, Ohio, Oklahoma, Oregon,
                    and Tennessee.  The position papers were
                    issued to Kansas and New York.  All the
                    audits were performed by the OIG staff except
                    for Oregon.  Oregon's audit was performed by
                    independent public accountants under contract
                    with the OIG.

                    Six audits had been completed for California,
                    Idaho, New Mexico, Oklahoma, Oregon, and
                    Tennessee with varying scopes prior to the
                    decision to summarize common findings in a
                    capping report.  To provide expanded coverage
                    of the LUST Program, we judgmentally selected
                    the States of Kansas, New York, and Ohio to
                    audit for fiscals 1994 and 1995.   We also
                    wanted to determine whether problems
                    identified in earlier reports continued to
                    exist in more recent time periods.  Our field
                    work in those States was conducted between
                    January 1995 and November 1995.
EVIDENCE            To accomplish our objectives we:
EXAMINED
                    •    Reviewed applicable laws, regulations,
                         and directives.

                    •    Interviewed responsible EPA and State
                         officials.

                    •    Examined records maintained by the
                         States.

                    The records reviewed included:  CAs, States'
                    statutes and policies, management site
                    listings, site files including oversight and
                    enforcement documents, and States'
                    priority/assessment systems.

-------
INTERNAL            We reviewed the internal controls associated
CONTROLS            with the States7 LUST programs to the extent
                    we considered necessary to accomplish our
                    objectives.  The controls we evaluated
                    included priority or assessment systems/
                    management site listings, document controls,
                    and information systems.

                    The internal control weaknesses noted in our
                    review are discussed in detail in CHAPTERS 2
                    through 4 of this report.  Our
                    recommendations, if implemented,  should
                    improve the internal controls in those areas.

                    Due to inherent limitations in any system of
                    internal control,  errors  or irregularities
                    may occur and not be detected.   However,
                    based on our review,  nothing else came to our
                    attention that caused us  to believe that
                    either the program office or the regions were
                    not in compliance with any other terms and
                    conditions of applicable  agreements,  laws,
                    and regulations for the three areas of the
                    LUST programs tested and  summarized in this
                    report.

-------
                         CHAPTER 2

           SOME HIGH RISK SITES NOT CLEANED-UP
OVERSIGHT AND
ENFORCEMENT
NEEDED FOR SOME
HIGH RISK SITES
We found that one State did a good job of
directing its resources toward clean-up of
high priority sites.   However, five States
could improve oversight and enforcement to
remediate some high risk sites.   We reviewed
249 high risk sites and found that 126 sites
did not get adequate oversight or enforcement
to ensure protection of human health and the
environment.

Overall, States were making notable progress
in cleaning-up LUST sites,  but some high risk
sites needed more oversight and enforcement
actions.  Under the LUST Program,  States are
responsible for establishing clean-up
programs that protect human health and the
environment.  Some States have more effective
programs than other States.  For example,
Ohio did a good job of managing LUST
activities and directing its resources toward
clean-up of high priority sites.   However,
California, Idaho, Kansas,  New York,  and
Oregon could have directed more attention
toward oversight and enforcement of
responsible party clean-ups for some serious
sites.

We found that these States did not either:
(i) assess some LUST sites to determine the
risk to human -health and the environment; or
(ii) assign the appropriate level of
oversight or enforcement relative to the
potential risk.  As a result, States did not
always initiate clean-up efforts on some
sites that were the most hazardous and
threatening to human health and the
environment, including those that posed a
threat to drinking water.  In addition, we
found that some States' management reports
were not accurate and as a result did not
identify some serious sites.

-------
Management
Reports Were
Inaccurate
BACKGROUND
LUST site status reports were not accurate  in
Idaho, Kansas, New York, and Oregon.  As a
result, these States did not have reliable
information to effectively monitor LUST site
remediation progress.  Without a current,
accurate, and complete site status report,
management could not assure that significant
sites were cleaned-up timely.
As discussed previously, SARA established the
LUST Trust Fund to assure rapid and effective
response to releases from leaking underground
storage tanks that pose a threat to human
health and the environment.

Subtitle I of the Resource Conservation and
Recovery Act authorized EPA to provide LUST
Trust Funds to States for the clean-up of
LUSTs through CAs.  LUST CAs are agreements
between EPA and States that provide the basis
which EPA oversees and manages the use of
LUST Trust Funds.  CAs identify the amount of
funds allocated to each State and establish
LUST Program performance requirements.

OUST gives the States latitude to tailor
their own programs and does not specify
clean-up levels or administrative procedures.
OUST has encouraged States to develop
programs where clean-up efforts are based on
factors that influence human health and the
environment.

OSWER Directive 9650.10, dated
February 8, 1989, provides guidance on LUST
CAs and for prioritizing LUST site corrective
actions.  The Directive states:

The State will ensure that a priority system
for addressing underground storage tank
petroleum release sites is established and
maintained.  The system will incorporate the
two priorities set in Section 9003(h)  of the
Resource Conservation and Recovery Act.
These priorities are:

•    Releases which pose the greatest threat
     to human health and the environment.

-------
•    Sites where the State cannot identify a
     solvent owner or operator of the tank
     who will undertake action properly.

In a memorandum, dated March 11, 1993, the
Director of the OUST, wrote:

     ... even in streamlined programs, all
     sites cannot and should not receive the
     same degree of oversight.  States will
     likely need to focus their limited
     technical resources on more complex or
     hazardous sites.  Therefore, we
     encourage the use of risk-based priority
     systems as a way to determine the
     appropriate degree of oversight that
     should be given to specific sites.

OSWER Directive 9610.17 dated March 1, 1995,
states that EPA's regulations dealing with
Underground Storage Tank (UST) corrective
actions are aimed at protecting human health
and the environment.  Under the regulations,
UST implementing agencies, including EPA, are
expected to establish goals for the clean-up
of UST releases based on consideration of
factors that could influence human and
environmental exposure to contamination.

The Directive recommends that States use a
risk-based decision-making approach as an
integral part of the corrective action
process at sites where leaking underground
storage tank systems have released petroleum
products into the environment and thus
created risks to human health and the
environment.

The EPA's role is to establish the regulatory
criteria, aid the States in establishing and
implementing their programs, and evaluate the
States' implementation of their programs
according to the requirements.  OMB Circular
A-123 provides guidance to Federal managers
as they develop and implement strategies for
reengineering agency programs and operations.

-------
OMB policy for
management
accountability
SOME HIGH RISK
SITES NEEDED
MORE ATTENTION
OMB Circular A-123 defines management
accountability as:  "the expectation that
managers are responsible for the quality and
timeliness of program performance, increasing
productivity, controlling costs and
mitigating adverse aspects of agency
operations, and assuring that programs are
managed with integrity and in compliance with
applicable law."
We selected judgmental samples of high risk
sites to determine if the levels of oversight
and enforcement were appropriate in
California, Idaho, Kansas, New York, Ohio,
and Oregon.  These six States received $97
million of the $473 million (21 percent) LUST
Trust Funds awarded for fiscals 1987 - 1995.

Of the 249 high risk sites that we reviewed,
we found that 126 (one half)  did not get
adequate oversight or enforcement to ensure
protection of human health and the
environment.  We limited our scope to
oversight and enforcement of high risk sites
because they should receive priority in times
of scarce resources.   We selected sites
designated by the States as either high
priority or as having petroleum contaminated
groundwater.
                               10

-------
                                 ,-  ;
CALIFORNIA
                      Site where soil and groundwater were contaminated

                    In our reviews of high risk sites, we did  not
                    take  exception to States' priority or
                    risk-based  systems, except in the State  of
                    California  where the system was not adequate
                    to identify the most serious LUST sites.   We
                    issued findings when States failed to
                    identify or to provide oversight or
                    enforcement of high risk sites in accordance
                    with  their  own policies.

                    At each State, we reviewed the State's
                    priority policy/procedures for addressing
                    LUST  site corrective actions.  We found  that
                    each  State  had a policy in place but that  the
                    policy was  not always followed.  We reviewed
                    site  files  and held discussions with program
                    officials for the 249 high risk sites.   Of
                    the 249 sites reviewed, we determined that
                    for 126 sites oversight and enforcement  could
                    be improved as described below.
We reported that the priority system for LUST
clean-up was unacceptable and that the
State's LUST enforcement program was
ineffective.
                                11

-------
Priority System
For Clean-ups
Was
Unacceptable
California's
Enforcement
Program was
Ineffective
We concluded that the State's priority system
did not identify some sites most
environmentally threatening to groundwater.
Further, we found that the priority system
established by the State was not being
followed.

As a result, some leaking tank sites
affecting drinking water were not being
cleaned-up.  We found that 48 of 69 leaking
tank sites, identified as affecting drinking
water, were not being cleaned-up timely.  The
leaks at these 48 tank sites had been known
for 3 to 14 years.

For example, a site in Vacaville, California,
where a leaking tank at a gas station was
discovered in March 1988, was found to be
polluting a domestic well used for drinking
water.  The owner had not responded to a
two-year-old request for a final clean-up
plan.  In this case, the regional board
(State) had requested a final clean-up plan
from the owner of the site by May 1990.
However, as of February 1992, nearly two
years later, there was no communication to
explain why the plan had not been received.
Since leaking tank sites which affect
drinking water were the highest priority
within the State, action should have been
taken on this site within two years.

At another site in Sonora, California, a gas
station had leaking tanks that polluted two
drinking water wells.  The leaking tanks were
discovered in June 1985.  No action had been
taken on this case since it was referred to
the regional board  (State) in January 1989.

We judgmentally selected 38 leaking tank
sites to review the status of enforcement
actions.  We concluded that enforcement
action appeared appropriate at only one site.
For the other 37 sites, the enforcement
actions were determined to be inadequate as
follows.

•    For 13 sites, the State took no
     enforcement action although the leaks
     had been reported for 2 to 11 years.
                                12

-------
                         For 19 sites, where clean-up and
                         abatement orders were issued, the owner
                         or operator had not complied with the
                         terms of the order and no enforcement
                         actions were taken to assess penalties
                         for noncompliance.

                         For 5 sites, where clean-up-and-
                         abatement orders were issued including
                         $226,540 in assessed penalties, the
                         penalties were only a small fraction of
                         the amounts that could have been
                         assessed, ranging from a minimum of
                         $1 million to a maximum of $10.4
                         million.
IDAHO               We reported that greater emphasis on
                    corrective action was needed for some higher
                    priority LUST sites and that the State's
                    management reports were inaccurate.

Oversight           Although Idaho had made substantial progress
Needed              in addressing the remediation of LUST sites,
Improvement         sufficient attention had not been given to
                    some LUST sites that posed the greatest
                    threat to human health and the environment.
                    We found that for 9 sites, out of a sample of
                    23, oversight of responsible party
                    remediation efforts had not been conducted on
                    some significantly contaminated LUST sites.

                    For example, the Steve's Auto Repair site had
                    been assigned no priority ranking.  The State
                    became aware of the LUST site during
                    October 1991.  Significant petroleum
                    contamination of the soil and groundwater had
                    been identified at the site.  A laboratory
                    report showed that the contaminant benzene
                    was identified at a concentration of 23,000
                    parts per billion.  The State's clean-up
                    standards required remediation action when
                    benzene levels exceed 5 parts per billion.
                    As of the end of fiscal 1993, the responsible
                    party had not initiated remediation action at
                    the site and the State had not conducted any
                    followup actions.

                    In another example, the Pioneer Square site
                    was identified as a LUST site during July
                    1992 and was assigned a priority ranking of


                                13

-------
Management
Reports were
Inaccurate
eight  (very high).  Significant petroleum
contamination of the soil and groundwater had
been identified at the site.  We found that
the State attempted to get the responsible
party to initiate the remediation process,
but were unsuccessful as of the end of fiscal
1993.

Our review of the site status report covering
the fourth quarter fiscal 1993 and site files
for 23 judgmentally sampled sites disclosed
that the status report included incorrect
information on 14 (61 percent) of the sites.
Specifically we found that the report
included:  (i) two sites that were not
confirmed LUST sites; (ii) seven sites with
incorrect confirmed release dates; (iii) two
sites with incorrect clean-up initiated
dates; and (iv) three sites with incorrect
confirmed release and clean-up initiated
dates.
KANSAS
Insufficient
Oversight on
Some LUST Sites
We reported that oversight and enforcement
could be improved for some sites which posed
a significant threat to human health or the
environment and that data bases were
inaccurate and not effectively used.

We selected a judgmental sample of 24 sites
which were reported as having contaminated
ground water and found that 6 sites did not
have adequate oversight of responsible party
remediation.  Groundwater contamination,
which is an indication of a serious
condition, was discovered several years
earlier at those sites and sufficient
remediation activities had not been
performed.

For example, the Home Oil site was one of the
six sites with significant contamination and
was an example where oversight of responsible
party remediation needs to be improved.

The State was aware of petroleum
contamination at this site for 12 years, yet
the contamination remained uncorrected.
During July 1983, the district office
verified the existence of petroleum
contamination in a garden well at a private
                               14

-------
                    home.  The staff suspected that the Home Oil
                    Service Station's underground storage tanks,
                    located about a half block away, were the
                    source of the release.  However, the district
                    office waited until March of 1990 to send a
                    corrective action letter to the owner and
                    operator of the service station.  The letter
                    requested a site assessment to identify the
                    size of the ground water contamination plume
                    and to develop and implement a remedial plan.
                    The letter did not establish deadlines for
                    the requested remedial activities.

                    The district office did not followup on the
                    site until February 1995 when it took ground
                    water samples at the service station.  The
                    samples confirmed that a LUST existed at the
                    site, showing a level of 13,080 parts per
                    billion for the primary gasoline constituents
                    of benzene, toluene, ethyl benzene, and
                    xylene (BTEX).  Staff stated that the site
                    was considered a "hot" site because of the
                    very high BTEX level.  The site was recently
                    assigned a priority ranking of 129 (very
                    high) because it was near a public water
                    supply and posed a significant threat to the
                    water.

Enforcement was     Kansas did not effectively pursue enforcement
not Effective       of responsible party corrective actions for
                    some LUST sites.  As a result,  Kansas could
                    not be sure that the most significantly
                    contaminated sites were cleaned-up timely.

                    As of the end of fiscal 1994, Kansas had
                    issued administrative orders to responsible
                    parties for 11 sites.  We reviewed
                    administrative orders and site files for 6 of
                    the 11 sites and interviewed district office
                    personnel to determine if the State
                    effectively pursued enforcement of the
                    responsible parties' corrective actions.

                    Kansas had effectively pursued enforcement
                    for two of the six sites but had not pursued
                    enforcement for the other four sites.  At one
                    site, the State did not pursue enforcement
                    actions after the responsible party refused
                    to accept an administrative order.  At two
                    other sites, the administrative order, which
                    directed remediation activities for both
                    sites,  was dismissed because there may have

                                15

-------
Data Bases Were
Inaccurate and
Not Effectively
Used
been additional responsible parties for the
contamination.  However, as of March 1995,
additional responsible parties had not been
identified and clean-up activities had not
been initiated.

As an example of some sites where enforcement
was not effectively pursued, the district
office confirmed that a LUST existed at the
St. Mary's Oil site during November 1990.  A
soil sample indicated that the release had
caused significant petroleum contamination.
Sample results showed a total hydrocarbons
level of 2,000 parts per million.  The State
requires remediation when the total petroleum
hydrocarbons level in the soil exceeds 100
parts per million.

An administrative order was issued to the
responsible party during March 1991 because
the responsible party had refused to conduct
a site investigation to identify the extent
of contamination at the site.  The order
directed the responsible party to perform a
site investigation, prepare a remediation
plan, and implement the plan.  The order also
established a schedule for the directed
activities.

Four years later (March 1995) the responsible
party had not complied with the order and the
district office had not taken any additional
followup actions.  The State recently
assigned a priority ranking of 105 which
means the site posed a significant risk to
human health and the environment.

Kansas maintained LUST site information on
two data bases.  We found that these two data
bases contained inaccurate site information
for 8 of 24 sites.  In addition,  site
information in the data bases had not been
used effectively by the State managers to
monitor the remediation of high risk sites.
                               16

-------
NEW YORK            We reported that  additional  oversight  of
                    responsible party corrective actions could
                    strengthen the program,  some errors in
                    priority  classifications were noted, and  the
                    data  base contained  inaccuracies.

Oversight Could     We selected a judgement  sample of  58 high
Be Improved         priority  releases which  were located within
                    the jurisdictions of Regions 1 and 6 of the
                    State's 9 regional offices.   Our review
                    disclosed that 42 releases had appropriate
                    oversight, 5 releases were not active  LUST
                    sites, and 11 releases did not have
                    sufficient oversight.

                    We found  that the regional offices had not
                    provided  sufficient  oversight of responsible
                    party remediation efforts for 11 of the
                    releases  during fiscal 1994.   Oversight
                    weaknesses at each region are discussed
                    below.

                    Region 1

                    We found  that sufficient oversight had not
                    been  provided for 8  of 30 releases which  were
                    reported  as high  priority during fiscal 1994.
                    For example, we found the following for the
                    Mobil Service Station site.

                    In October 1991 petroleum contaminated soil
                    was found during  the removal of an
                    underground storage  tank at  the service
                    station.  Regional spills staff reported  that
                    groundwater contamination was possible and
                    assigned  a high priority classification to
                    the release.  They sent  a corrective action
                    letter during November 1991  requesting the
                    responsible party to install five  groundwater
                    monitoring wells  at  the  site.  The purpose of
                    the wells was to  determine  if groundwater
                    contamination existed.

                    The responsible party failed to conduct any
                    groundwater  investigation during  1992. The
                    regional  office notified the responsible
                    party during November 1992  that the  State
                    would retain a private contractor  to  install
                    the requested monitoring wells and would  seek
                    reimbursement  for all costs  plus  penalties.


                                17

-------
During December 1992 the responsible party's
attorney responded by notifying the regional
office that they would decide within two
weeks whether the responsible party would
undertake the requested groundwater
investigation.

As of the end of June 1995 the file contained
no evidence that the responsible party
conducted any groundwater investigation or
that the regional office followed up on the
site after December 1992.

Region 6

We found that sufficient oversight had not
been provided on 3 of 28 releases which were
reported as high priority during fiscal 1994.
For example, we found the following for the
Burkhart Brothers site.

Petroleum was found floating on the
groundwater in two UST excavations during May
1990.  A groundwater investigation report,
dated July 1991, showed a level of 27,000
parts per billion in the groundwater for the
petroleum compounds of BTEX.  The State's
acceptable clean-up level for BTEX is 30.7
parts per billion.

During August 1991, the regional office
issued a corrective action letter to the
responsible party requesting additional
investigation work, quarterly groundwater
sampling, and a soil venting system proposal.
The file contained no evidence that the
responsible party completed the required
activities.

The regional office last followed up on the
site in March 1992.  This followup was a
telephone call to the responsible party
requesting an update on the status of
activities at the site.  During the telephone
conversation,  the responsible party informed
the regional office staff that he was not
sure he would be financially able to continue
the investigation, and that he would get back
to the Region shortly with the status of the
site.  There was no evidence in the file that
the responsible party provided any additional
updates on the status of the site.

            18

-------
Data Base
Contains
Inaccuracies
Region 6 staff said that the large spill
response workload prevented them from
addressing corrective actions on all high
priority LUST releases and that some higher
priority releases were overlooked.

Site priority classifications were not
accurate for 9 of 58 releases.  Specifically,
our review showed that some releases assigned
high clean-up priorities should have been
assigned lower priorities,  in addition, some
numeric class codes did not identify whether
a responsible party was willing to undertake
corrective actions.  Also, the State did not
accurately identify LUST releases or
erroneously classified releases as cleaned-up
for 13 of 68 sites.  The State used the data
base to perform various trend analyses and to
obtain status information on specific sites.
OHIO
The State did a good job of managing LUST
activities and directing its resources to
clean-up high priority sites.  We reviewed
case files for 20 high priority sites and
found that the State identified and responded
to the most significant sites where there was
an imminent threat to human health and the
environment.  We concluded that the State
effectively used a ranking system to direct
its limited resources to the most significant
sites.
OREGON
Enforcement
Actions Were
Not Timely
Contract auditors reported that enforcement
actions were not performed in a timely
manner.

The auditors determined that an excessive
time lag occurred between the time a release
was reported and an enforcement action was
initiated for 11 out of the 11 sites
reviewed.

For example, Leather's Oil had a reported
release on November 3, 1988.  On February 9,
1989, the respondent agreed to a unilateral
order.  However, on December 24, 1990 a
notice of violation with intent to assess
penalties was issued by the State.  This
resulted in a time lag of two years from the
                                19

-------
                    release to when enforcement action was taken.
Site List Was
Not current/
Accurate/ or
Complete
The State's priority system was not updated
on a regular basis to reflect significant
changes, such as the immediate danger being
eliminated or the corrective action
completed.
MORE EMPHASIS
IS NEEDED FOR
SOME HIGH RISK
SITES
RECOMMENDATIONS
We found that some potentially serious sites
were not assessed as to the risk they posed
or some sites that were assessed, were not
assigned appropriate attention for the level
of risk.  In our opinion, these conditions
occurred because the States had not always
followed their own priority or risk-based
decision-making approach.  As a result,
States did not oversee or enforce remediation
efforts on some significantly contaminated
LUST sites.

Since the primary purpose of the LUST program
is to focus corrective actions on sites that
pose the most serious threats to human health
and the environment, we believe that EPA
needs to give more emphasis to clean-ups of
high risk sites.  It needs to implement
controls to promote effective management and
accountability for the LUST program.

The enactment of the Government Performance
and Results Act of 1993 (GPRA) shifted the
focus of Federal management and
accountability away from inputs, such as
personnel levels and adherence to prescribed
processes, to a greater focus on achieving
desired results.  The establishment of new
partnership agreements between EPA and the
States is an opportunity for EPA to reassess
management controls and accountability over
the LUST Program to ensure that all high risk
sites receive appropriate attention.
We recommend that the Assistant Administrator
for Solid Waste and Emergency Response:

1.   Implement management controls, as
     required by OMB Circular A-123, to
     ensure that States provide adequate
                               20

-------
AGENCY COMMENTS
AND OIG
EVALUATION
OIG Evaluation
                         oversight and enforcement on high  risk
                         sites.

                         Use the cooperative agreements between
                         EPA and the States as a tool to achieve
                         the expected results.

                         Measure the States performance to
                         determine if they are meeting the
                         expected results.
OSWER generally disagreed with our finding
and conclusions, but concurred with the
recommendations.  It stated that for the
sites where we took exception to the level of
oversight or enforcement we:  (i) provided no
evidence that there were any negative human
health or environmental consequences; (ii)
focused on the very small minority of sites;
and (iii) did not give credit when things
went well.  It interpreted the statement in
the report "We believe that EPA needs to give
more emphasis to clean-ups of high risk
sites" to mean that States should numerically
rank LUST sites and spend money (for
corrective action or oversight)  only on the
highest ranked sites.

However, OSWER stated that all nine States
audited are now in the process of
implementing risk-based corrective action
(RBCA) thus making substantial improvements.
"RBCA represents a paradigm shift in the way
LUST regulating agencies do business," from
following predetermined and standardized
clean-up levels to setting site-specific
levels based on risk.  OSWER believes that
increased implementation of RBCA is the best
approach to address the concerns outlined in
the report.

We believe that our report contains adequate
evidence that there were a significant number
of sites that were not cleaned-up which had
negative human health and environmental
consequences.  The sites included in our
report were identified by each State as ones
that pose the most serious health and
environmental risks.  The States identified
serious or high risk sites as those which had
                                21

-------
high levels of contamination found in
drinking water, domestic water wells,
groundwater, or near a public water supply.

Groundwater is the source of drinking water
for half the United States.  It is also used
for irrigation, livestock watering, and
industrial uses.  Petroleum is a known
carcinogen and other adverse effects of
contact are nausea; skin, eye, and throat
irritation; loss of reflexes; and liver and
kidney damage.

As previously stated, we limited our audit
scope to high risk sites.  While fewer in
number than the low risk sites, the high risk
sites require more immediate clean-up actions
because they pose the most significant threat
to human health and the environment.  Our
conclusion is based on the testing we
completed for those high risk sites where we
found that 125 sites out of 249 did not get
adequate oversight or enforcement.

OSWER's interpretation of the report that we
are recommending a specific, inflexible,
numerical ranking system is incorrect.  This
issue was previously discussed with OUST's
staff in July 1995.  At that time we clearly
explained that we were not recommending a
numerical ranking system that would limit the
expenditure of LUST funds to high risk sites.
We said that our audits had identified a
number of high risk sites that had not
received adequate oversight or enforcement.
We were told about RBCA and that_ OUST
requires States to have a system in place to
focus limited resources on the sites that are
a threat to human health and the environment.

OSWER's proposed actions to implement RBCA
meet the intended purpose of the
recommendations.  The effectiveness of RBCA,
however, depends upon whether the States
fully implement an assessment of all sites
and assign oversight on all sites according
to risk.  As stated in the report, all nine
states had assessment procedures.  However,
some sites were excluded from the assessment
or were never assigned oversight when risk
was assessed as high.
            22

-------
                         CHAPTER 3

      COST RECOVERY PROGRAMS NEED IMPROVEMENTS
                    For the States in our  review, there were
                    major differences in results between the
                    States' cost recovery  programs.  We found
                    that:  (i)  two States  maintained effective
                    cost recovery programs;  (ii) three States
                    needed improvements  in their cost recovery
                    programs;  and (iii)  two  States did not have
                    cost recovery programs.

                    In our opinion,  two  States did not implement
                    cost recovery programs because they either
                    did not believe in recovery of clean-up costs
                    from owners and operators, or never devoted
                    time to develop a cost recovery policy.
                    Three States did not maintain adequate site-
                    specific records to  support all recoverable
                    costs because the States did not follow EPA
                    requirements for various reasons.

                    The primary purpose  of a cost recovery
                    program is to provide  incentives for owners
                    and operators to clean-up releases from their
                    own tanks.   Ineffective  cost recovery
                    programs hinder the  efforts of the LUST
                    program to protect human health and the
                    environment.

                    When States do not recover clean-up costs
                    from those owners and  operators who are
                    responsible for paying,  the State is:
                    (i) relieving them of  their financial
                    obligation; (ii)  removing the financial
                    incentive to clean-up  contamination; and
                    (iii) shifting the burden of clean-up costs
                    to the Federal government.
BACKGROUND          Subsection 9003(h)(6),  Subtitle I of RCRA
                    states that whenever costs have been incurred
                    by a State for undertaking corrective action
                    or enforcement action,  the owner or operator
                    shall be liable  for such  costs.
                               23

-------
Office of Solid Waste and Emergency Response
(OSWER) Directive 9610.IDA outlines EPA's
LUST Trust Fund cost recovery policy.  The
policy states that each State is to make
reasonable efforts to contact owners and
operators who are liable for LUST sites,
notify them of their liability for
enforcement and corrective action costs, and
demand payment.
      Clean-ups are costly, averaging $150,000

The OSWER Directive states that the policy
provides States with the autonomy and the
incentive necessary to pursue recoveries
aggressively and' efficiently.  States with
CAs will litigate and settle recovery claims
without the routine involvement or
concurrence of EPA or the Department of
Justice.  Also, States may retain any Trust
Fund monies they recover for use on
            24

-------
STATES' COST
RECOVERY
PROGRAMS RANGE
FROM EFFECTIVE
TO NONE
                    additional Fund-eligible clean-ups and
                    activities.

                    The OSWER Directive states responsible
                    parties are liable for all costs of
                    corrective action and enforcement, including
                    interest, indirect and "management and
                    support" costs associated with these
                    activities that are paid for by the Trust
                    Fund.  Responsible parties are also liable
                    for Trust Fund expenditures made by States in
                    overseeing responsible party clean-ups.

                    The OSWER Directive requires States to
                    establish a cost accounting system to support
                    cost recovery claims in a judicial action and
                    provide evidence that costs claimed are
                    reasonable and necessary.  States must
                    document site-specific costs where they have:
                    (i) performed an emergency response;
                    (ii) begun a detailed site investigation; or
                    (iii) determined that an owner or operator is
                    or is likely to be recalcitrant.
For the States in our review, there were
major differences between the cost recovery
programs.  We found that: (i) California and
New York maintained effective cost recovery
programs; (ii) Idaho, Oklahoma, and Tennessee
needed improvements in their cost recovery
programs; and (i.ii) Kansas and Ohio did not
have cost recovery programs.

We reviewed cost recovery programs for seven
States: California, Idaho, Kansas, New York,
Ohio, Oklahoma,  and Tennessee.  The seven
States received $105 million out of $473
million (22 percent) Trust Funds allocated
for fiscals 1987 through 1995.  In addition
to Federal requirements, we reviewed each
State's policy and site files to determine if
the State was following its own policy.  For
example, Idaho's policy states that interest
is assessed on Trust Fund expenditures used
for corrective action and enforcement
activities.   We reported a finding when we
found that the State did not assess interest
according to its own policy..
                                25

-------
EFFECTIVE COST
RECOVERY
PROGRAMS
California
New York
                    States are required to have a cost recovery
                    program that is consistent with OSWER
                    Directive 9610.10A.  This directive provides
                    States with flexibility and discretion in
                    determining under what circumstances they
                    will pursue such costs, which costs to
                    pursue, and how much effort to devote to
                    pursuit of these costs.  States are not
                    required to recover costs against a
                    responsible party where the LUST Trust Fund
                    had been used only for oversight of
                    responsible party corrective actions.
California and New York had effective cost
recovery programs.  California had a recovery
rate of about 90 percent for oversight costs
of responsible party clean-ups.  New York had
a recovery rate of 83 percent for direct site
clean-up costs.

We found that California's cost recovery
program effectively recovers the costs of
oversight of responsible party corrective
actions.  The state's cost recovery program
for oversight of responsible party corrective
actions was started in July 1989.  It had
recovered about $30 million from responsible
parties since the beginning of the program.
The $30 million represents both State and
LUST Trust Funds recovered.

States are not required to recover costs for
oversight of responsible party clean-ups
because the potential benefits might not
exceed the costs to recover.   It is at the
discretion of the State to recover oversight
costs.  California demonstrated that recovery
of oversight costs can be worthwhile.

The State of New York was effective in
recovering costs from responsible parties
when the State performed the clean-ups with
LUST Trust Funds.  Through June 1995, New
York spent about $7.3 million for direct site
clean-ups on 24 sites.  We reviewed cost
recovery files for 11 of the 24 sites and
found that the State had efficiently and
aggressively pursued cost recovery.  The
State recovered $1.5 million out of $1.8
million spent on direct site clean-ups (an 83
                               26

-------
COST RECOVERY
PROGRAMS NEED
IMPROVEMENTS
Idaho
Oklahoma
Tennessee
                    percent success rate)  for  six  completed  cost
                    recovery cases.
We found three States where cost recovery
programs needed improvements to enhance the
incentives for owners and operators to
clean-up releases from their own tanks.  The
absence of strong programs also relieved the
responsible parties from the financial
liability for some clean-up costs.

The State completed four cost recovery
settlements during the period fiscal 1993 and
through the first quarter fiscal 1994.  It
recovered 72 percent of identified costs on
the four settlements.  Although the State had
incurred Trust Fund expenditures totaling
$198,000 for the sites, only $189,000 was
identified for cost recovery purposes.  In
addition, the State did not document why it
accepted reduced cost recovery settlements
totaling $144,000 and did not attempt to
charge interest.

The State maintained incomplete records of
site-specific costs paid with LUST Trust
Funds and therefore, could not recover all
the costs.  The State's site-specific records
documented the clean-up contractor's costs
but did not include allocated portions of
indirect costs, management and support costs,
and interest.  To ensure that the State could
effectively attempt to recover clean-up costs
from responsible owners and operators, the
LUST CA required the State to establish and
maintain cost records on a site-specific
basis.

The State did not maintain complete records
of site-specific costs paid with LUST Trust
Funds and therefore, could not recover all
the costs.  The site-specific records
included clean-up costs but not enforcement
costs, indirect costs, management and support
costs, and interest.  To ensure recovery of
appropriate costs from responsible owners and
operators, the State must establish and
maintain cost records on a site-specific
basis.
                                27

-------
NO-COST
RECOVERY
PROGRAMS
Kansas
Ohio
Ohio and Kansas had not implemented cost
recovery programs.  This resulted in
responsible parties for 138 sites not being
held liable for any clean-up costs.  The
States had used LUST Trust Funds for direct
site clean-ups and cost recoveries should
have been pursued.

The State of Kansas did not implement cost
recovery procedures and, therefore, did not
pursue cost recovery from responsible
parties.  Since the inception of the program
the State pursued only one cost recovery
settlement and did not demand payment for 17
other sites.  As a result, the State had
recovered only $21,748 out of $303,001 Trust
Fund expenditures on direct site clean-ups.

Kansas officials stated that the Federal
mandate for cost recovery runs in opposition
to the State's user friendly approach.

The State of Ohio had not recovered Federal
and State funds spent to clean-up 121 LUST
sites and did not keep complete cost records.
Since 1988, the State had only recovered
$93,000 in costs and penalties from two
responsible parties.  We were unable to
estimate the costs associated with 121 sites
because the State did not maintain
site-specific cost documentation as required.

The State did not implement a cost recovery
program because it never devoted time to
develop a cost recovery policy.  However, in
late 1995 the State prepared draft cost
recovery procedures and had submitted them to
EPA Region 5 for review and approval.
MORE EMPHASIS
IS NEEDED FOR
SOME COST
RECOVERY
PROGRAMS
We believe that EPA needs to place more
emphasis on the improvement of some States'
cost recovery programs.  An important purpose
of cost recovery is to provide incentives for
owners and operators to clean-up releases
from their own tanks.  An aggressive cost
recovery policy could be used as an effective
tool to stimulate site clean-ups.
Specifically, responsible parties should be
held liable to clean-up releases from their
                               28

-------
RECOMMENDATIONS
AGENCY COMMENTS
AND OIG
EVALUATION
                    own tanks, or reimburse the State for all
                    costs to clean-up their site.

                    The EPA's policy recognizes that there are
                    situations where cost recovery will not be
                    possible or feasible.  In these situations,
                    the State is required to fully document its
                    decisions and to formally close out all
                    cases.  In fulfillment of its
                    responsibilities under OMB Circular A-123, we
                    believe that EPA should re-evaluate its
                    policies and procedures to ensure that States
                    are accountable for establishing effective
                    cost recovery programs.
We recommend that the Assistant Administrator
for Solid Waste and Emergency Response:

1.   Implement management controls, as
     required by OMB Circular A-123, to
     ensure that States develop and implement
     cost recovery programs in accordance
     with the statutory requirements and EPA
     policy.

2.   Use the cooperative agreement between
     EPA and the States as a tool to achieve
     the expected results.

3.   Monitor and measure the States' cost
     recovery programs to determine if they
     are meeting the expected results.
OSWER acknowledged that improvements would be
beneficial to some State cost recovery
programs.  However, it did not believe that
the problems in a few states were indicative
of systemic, national problems.  Its primary
concern is that this chapter generally fails
to recognize the State flexibility and
discretion that intentionally has been built
into the cost recovery policy.

OSWER generally concurred with the
recommendations and said that it will
implement management controls to ensure that
States develop and implement cost recovery
programs in accordance with EPA's policy.
                                29

-------
                    The response states that contrary to several
                    of the statements made in the report States
                    are not required to both document and pursue
                    all costs for which UST owners could be held
                    liable under Subtitle I.   Also,  it stated the
                    cost recovery policy encourages but does not
                    require States to either document or pursue
                    these costs.

                    OSWER stated that the draft report does not
                    provide any evidence that harm has occurred
                    to human health or the environment due to
                    ineffective state cost recovery programs.  It
                    also said that the draft report improperly
                    equated lack of documentation of cost
                    recovery decisions with the absence of a cost
                    recovery program.   For example,  according to
                    EPA,  Region 7,  and the State,  Kansas did have
                    a cost recovery policy but that cost recovery
                    was not pursued for various reasons.

OIG Evaluation      We believe that flexibility and discretion
                    can be desirable attributes in a national
                    policy.   Some benefits are identified in
                    OSWER Directive 9610.10A which states that
                    the two most innovative aspects  of EPA's cost
                    recovery policy for the LUST Trust Fund
                    should provide States with the autonomy and
                    the incentive necessary to pursue recoveries
                    aggressively and efficiently (emphasis
                    added).   Unfortunately,  flexibility and
                    discretion can also be excuses for States'
                    failure to manage effective cost recovery
                    programs.   Ohio did not implement a cost
                    recovery program because  it never devoted
                    time to develop a recovery policy.   Kansas
                    did not pursue cost recovery from responsible
                    parties because cost recovery runs in
                    opposition to the State's user friendly
                    approach.

                    The response  states that  under Subtitle I of
                    RCRA,  UST owners and operators are liable for
                    costs incurred by EPA or  a State in
                    performing enforcement or corrective action
                    at LUST sites.   However,  if a State fails to
                    have  a  cost recovery program under OSWER's
                    policy,  then  those owners and operators are
                    not held liable for the costs incurred.

                    We believe that our report identifies
                    sufficient weaknesses in  States'  cost

                               30

-------
recovery programs to demonstrate a need for
corrective action.  In addition, we selected
the state of Kansas for review based on a
recommendation from OUST.  Its staff believed
that Kansas' LUST program was typical and
representative of the other States.  We found
that Kansas did not have a cost recovery
program.

OSWER's response states that our report does
not provide a demonstrated connection between
a State's cost recovery program and
protection of human health and the
environment.  We believe this connection is
made in OSWER Directive 9610.10A which states
that EPA expects this policy will help cost
recovery to become a practical and effective
tool that States will use to both stimulate
and fund more clean-ups of releases from
underground storage tanks.  The connection is
that by stimulating and funding more clean-
ups of releases, cost recovery helps to
protect human health and the environment.
            31

-------
(This page left intentionally blank.)

-------
                         CHAPTER 4

                ACTIVITY REPORTS MISLEADING
ACTIVITY
REPORTS WERE
HOT ACCURATE
BACKGROUND
Quarterly activity reports submitted by
States to EPA did not accurately report LUST
Program results.  For the States where we
reviewed activity reporting,  we found that
States generally overstated their results.
The overstatements ranged from 7 to 47
percent.  We believe that accurate
performance information is needed to make
sound management or program decisions.   Sound
decisions ensure that the program:  (i)  meets
intended goals, (ii) assesses the efficiency
of processes, and (iii) promotes continuous
improvement.

In our opinion, States did not accurately
report program accomplishments because they:
(i) did not correctly use the OUST
definitions for confirmed releases,  clean-ups
initiated, and clean-ups completed;
(ii) improperly reported unregulated tanks;
(iii) had reporting systems that did not
count actual activities; and (iv)  placed too
little emphasis on reporting requirements.

The activity reports were used as a
management tool to track the most important
program activities.   Therefore,  it is
necessary that the information be accurate,
complete, and comparable for EPA to
effectively manage the LUST Program.
States with LUST CAs are required to submit
quarterly activity reports to EPA,  disclosing
data on various LUST Program activities.
This reporting mechanism helps EPA program
officials monitor and oversee program
progress by State and EPA region.  EPA
Headquarters uses the information to make
nationwide funding decisions, to report
program progress to Congress, and to report
the results of the program in the Agency's
financial statements.
                               33

-------
Activity Data
Reported
Through Fiscal
1995
Through fiscal 1995, more  than 303,000
releases had been reported in activity
reports nationwide.  In  1995  alone,
new releases were confirmed and reported.
The upcoming 1998 deadline for upgrading,
replacing, or closing underground storage
tank systems will probably increase that
number.  Owners and operators often discover
contamination not previously  identified as
they evaluate their tank systems to decide
whether to upgrade, replace,  or close them
before the deadline.

All States and territories, as well as some
local governments, have  corrective action
programs employing a total of about 1,500
technical staff.  Most corrective actions are
undertaken by responsible  parties with State
and local oversight.  Through fiscal 1995,
OUST reported that clean-ups  have been
initiated at more than 238,000 sites and
completed at more than 131,000 of them.   In
spite of this progress,  more  than 172,000
sites still need to be cleaned-up.   The graph
below illustrates the reported numbers of
confirmed releases, clean-ups initiated,  and
clean-ups completed for  each  of the 10 EPA
regions as of the end of fiscal 1995.
                       CUMULATIVE LUST SITES THROUGH FISCAL 1995
                                       By EPA Raglan
                            Thousands
                         too
                             123456789   10

                                        EPA -Region

                      Source: Fiscal 1995 Fourth Quarter Activity Reports (unaudited)
                                34

-------
Management          OMB Circular  No. A-123, Management
Accountability      Accountability  and  Control,  states  that
and Control         management  controls are the  organization,
                    policies, and procedures used  to  reasonably
                    ensure that ... reliable and timely information
                    is obtained,  maintained, reported and  used
                    for decision  making.
ACTIVITY            Quarterly  activity  reports  submitted  by
REPORTS ARE NOT     States to  EPA  did not  accurately  report LUST
RELIABLE            Program results.  For  the States  where we
                    reviewed activity reporting, we found that
                    States generally overstated their results.
                    The overstatements  ranged from 7  to 47
                    percent for the various  activities and time
                    periods.   We believe that accurate
                    performance information  is  needed to  make
                    sound management or program decisions.  Sound
                    decisions  ensure that  the program meets
                    intended goals, assesses the efficiency of
                    processes, and promotes  continuous
                    improvement.

                    We evaluated the accuracy of the  reported
                    number of  confirmed releases, clean-ups
                    initiated, and clean-ups completed for six
                    States for various  time  periods.  These six
                    States (California, Idaho,  Kansas, New
                    Mexico, New York and Ohio)  represent  22
                    percent of the reported  confirmed releases,
                    24 percent of  the reported  clean-ups
                    initiated, and 21 percent of the  clean-ups
                    completed.

OUST's              The following  are OUST's definitions  for the
Definitions         three major LUST program activities as
                    described  in the Agency's fiscal  1995
                    Financial  Statements and in a Memorandum,
                    dated December 21,  1990, to the Regional
                    Program Managers.

                    •    Confirmed releases  are the number of
                         sites where the owner/operator has
                         identified a release from a  petroleum
                         underground storage tank, reported the
                         release to the state/local agency, and
                         the implementing  agency has  verified the
                         release.
                                35

-------
                    •    Clean-ups initiated are the total number
                         of confirmed releases which the State or
                         responsible party (under State
                         supervision)  has initiated management of
                         petroleum contaminated soil, removal of
                         free product, or management or treatment
                         of dissolved petroleum contamination.

                    •    clean-ups completed are the total number
                         of confirmed releases where clean-up has
                         been initiated and where the State has
                         determined that no further clean-up
                         actions are necessary to protect human
                         health and the environment.

                    We reviewed the controls for data
                    compilation,  supporting documentation,  and
                    the reporting process for the three program
                    activities listed above.   The cumulative
                    totals for the above listed three activities
                    were used to determine Regional Trust Fund
                    allocations and were identified as
                    performance measures in the Agency's
                    financial statements.

                    Each State had its own system for data
                    compilation and reporting of the three
                    program activities to EPA regions.   Of the
                    six States,  we had reported findings on the
                    accuracy of the activity data as described
                    below.

California          The activity report dated,  December 31,  1991,
                    significantly overstated the number of
                    confirmed releases and clean-ups initiated.
                    These overstatements occurred because some
                    sites were counted more than once and some
                    information in the data base was inaccurate.
                    The number of leaking tank sites was also
                    overstated because unregulated tank sites
                    (above ground tank sites)  and unlocatable
                    sites were included in the activity reports.
                    Unregulated heating oil tank sites,  farm tank
                    sites,  and above ground tank sites are
                    ineligible and should not have been included
                    in the activity reports.

                    Clean-ups Initiated.   We found that the
                    number of reported site clean-ups initiated
                    in the State  was inflated by about one-third.
                    The activity  report stated that 5,860 site
                    clean-ups were initiated.   However,  the

                               36

-------
Idaho
Kansas
supporting LUST information system showed
that only 4,459 clean-ups were initiated.
The difference of 1,401  (31 percent)
represented a significant overstatement of
the reported number of clean-ups initiated.

Confirmed Releases.. Leaking tank sites that
had not been confirmed were included in the
report as confirmed leaking sites.  We found
the State assumed that all sites coded in the
LUST data base as "No Action" were reported
as confirmed leaks.  However, according to
regional board personnel, some unconfirmed
sites were included in the "no action"
classification, and therefore, were
improperly reported as confirmed leaks in
activity reports.

The quarterly reports did not accurately
report LUST Program activity during fiscal
1993.  The number of confirmed releases were
understated by 16 percent, clean-ups
initiated were overstated by 18 percent, and
clean-ups completed were understated by 57
percent.  In addition, the quarterly reports
included LUST Program activity that had
occurred prior to the reporting periods.

The inaccuracies occurred because not all
performance data had been correctly compiled
by the computer system used to maintain LUST
site information, information in the
quarterly reports was not reconciled to the
supporting records, and some regions reported
LUST site activity late.

Fiscal 1994 accomplishments were accurately
reported for confirmed releases,  but
clean-ups initiated and clean-ups completed
were overstated by 7 percent and 13 percent,
respectively.  In addition, the number of
confirmed releases, clean-ups initiated, and
clean-ups completed, reported in the first
quarter fiscal 1995, were overstated by
14 percent, 23 percent, and 47 percent,
respectively.

These inaccuracies occurred for various
reasons, including:  (i) multiple leaking
underground storage tanks located at the same
location were incorrectly classified as
separate confirmed releases; (ii) leaking

           37

-------
                    underground heating oil tanks were improperly
                    classified as confirmed releases; and
                    (iii)  source records were not reconciled with
                    the data base used to prepare activity
                    reports.  We also found staff that compiled
                    activity report data for the first quarter of
                    fiscal 1995 had not received adequate
                    training.

New Mexico          The State submitted activity reports that
                    contained inaccurate data for confirmed
                    releases.   We found that the number of
                    confirmed releases in the reports did not
                    match the number in the State's LUST data
                    base for fiscals 1990 to 1992.   For fiscal
                    1992,  the State reported 701 confirmed
                    releases but the data base supported only
                    616, a 14  percent difference which represents
                    a significant overstatement.

                    The reason this occurred is that reports were
                    not reconciled to the data base.

New York            At the time of our review,  the  State did not
                    report actual activities for some program
                    accomplishments and reported incorrect
                    information on some high priority sites.   For
                    confirmed  releases and clean-ups initiated,
                    New York reported leaks before  they were
                    confirmed  and clean-ups initiated prior to
                    knowledge  that physical clean-up activity had
                    begun.   In addition,  we tested  a sample of 68
                    active sites during fiscal 1994.   We found
                    that the State's data base contained
                    incorrect  information for 13 LUST releases,  a
                    19 percent error rate for our sample.

                    New York counts confirmed releases by
                    querying the spills data base each quarter
                    and reporting all LUST notifications of leaks
                    or suspected leaks.   The State  adjusts the
                    total  notifications for a small number of
                    sites  where it has been confirmed that no
                    leak occurred.   While the State adjusted for
                    a few sites where it confirmed  no leak
                    occurred,  the majority of the activities
                    reported were based on a notification and not
                    on a confirmed release.

                    New York assumes that clean-ups will be
                    initiated  for 100 percent of the confirmed
                    releases as computed in the preceding

                               38

-------
Ohio
paragraph,  it reports the total confirmed
releases, adjusted for an estimated number of
sites where the responsible party is unknown,
as clean-ups initiated, even though the State
does not know if an actual clean-up has been
initiated.

The State significantly overstated its
accomplishments for the number of confirmed
releases, clean-ups initiated, and clean-ups
completed.  These overstatements occurred
because the State included sites with no
contamination.

The State reported all tank removals as
completed clean-ups, even the ones that did
not have a release of petroleum.  By
including clean tank removals, the State
overstated the number of clean-ups completed
by 35 percent for the period January through
March 1995.

We found the State's data base supported
significantly fewer accomplishments than it
reported for January through March 1995.  The
State's data tracking system did not identify
the specific sites reported for these
accomplishments.  The State also could not
provide an explanation of the methodology it
used to compute the reported accomplishments.
MORE EMPHASIS
IS NEEDED ON
REPORTING
REQUIREMENTS
Based on our findings, we believe that the
activity reports were inaccurate.  Management
or program decisions based upon those reports
may be at risk because the activity data was
used to make nationwide funding decisions, to
report program results to the Congress, and
to report performance measures in the
Agency's financial statements.

We cannot statistically evaluate the extent
of the error rate in the activity reports
because methodologies and time periods varied
among the audits we are summarizing.  For
example, we tested 100 of the three
performance measures for fiscal 1993 in Idaho
but it was not practical to test 100 percent
in Kansas.  We had planned to audit 100
percent of fiscal 1994 but chose to randomly
                                39

-------
select a sample because Kansas had integrated
the support into the entire LUST file system.

However, based on the results of our audits
of six States that comprise 21 to 24 percent
of the cumulative totals for the three
reported activities, we believe the problems
and errors are indicative of the extent of
the misstatement of the nationwide reports.
Even if no other State had errors, these six
States comprise a significant amount of the
totals and their combined errors would
significantly affect the nationwide reports.
Therefore, we believe the misstatements
represent a significant reporting problem.

OUST needs to take action to improve the
management controls over activity reporting
for performance measures.  OUST should
implement controls to ensure all the States
use the OUST definitions, report only
regulated tanks, improve systems to count
actual activities, and place more emphasis on
reporting requirements.

We recognize that some States calculate the
reported numbers from data base information
and because of large volumes of activities,
as in the State of New York, it is not
feasible to physically count and view
supporting documentation for each activity
reported.  However, we believe that those
States should sample test their methodology
or otherwise ensure that reported information
is accurate.

We believe that OUST should implement
controls to ensure that all States report
comparable data.  For example, Ohio which
counts every tank pull has significantly
overstated its three activities.   Strict
adherence to the definitions would provide
results that can be compared among the
States.

OUST requires the States to report ten LUST
activities, including the three activities
identified as performance measures.  Accurate
reporting would provide a sound basis to make
decisions and to focus their efforts and
improve performance.
           40

-------
Validation and
Verification
RECOMMENDATIONS
AGENCY COMMENTS
AND 016
EVALUATION
The enactment of the GPRA shifted the focus
of Federal management and accountability away
from inputs, such as personnel levels and
adherence to prescribed processes to a
greater focus on achieving desired results.
The GPRA calls for validation and
verification of the results and follow-up for
exceptions.

OUST reported in the Agency's fiscal 1995
financial statements that the data was
validated and verified by EPA regional
offices.  However, based upon the
deficiencies that we found, we conclude that
the Regions had not identified the reporting
errors and had not taken steps to ensure that
the States were reporting accurate
information.
We recommend that the Assistant Administrator
for Solid Waste and Emergency Response:

1.   Implement management controls, as
     required by OMB Circular A-123, to
     ensure that reliable and timely
     information is obtained, maintained,
     reported and used for decision making.

2.   Specifically, direct EPA regions to use
     the cooperative agreements between EPA
     and the States to require them to report
     the actual status of their LUST programs
     by:  (i). using the EPA definitions; (ii)
     excluding unregulated tanks;  (iii)
     improving reporting systems to count
     actual activities; and  (iv) placing more
     emphasis on reporting requirements.
OSWER disagreed that the misstatements in
activity reports represent a significant
reporting problem.  It said our conclusion
that activity reports were generally
unreliable was overstated and unsupported
based on the limited size of the databases
examined.  For example, New York reported
over 12,000 confirmed releases of which we
investigated 70, a small percentage.
                                41

-------
                    OSWER generally concurred with the
                    recommendations and said that it will
                    implement management controls to ensure that
                    reliable and timely information is obtained,
                    maintained, reported, and used.

                    OSWER disagreed that strict adherence to the
                    definitions would provide results that can be
                    compared among the States.  It stated that by
                    design the program results varied between
                    individual States.  It also stated that
                    accurate, complete and comparable information
                    is not necessary to manage the LUST program.

                    It stated that the performance measures were
                    never intended to be used to manage the LUST
                    program.  The measures were originally
                    developed to determine the extent of the LUST
                    problem, to track nationwide remediation
                    efforts, and to establish a baseline of
                    performance for individual States.

                    OSWER disagreed that the activity reports
                    were used to determine State allocations^  It
                    said the data was used to determine Regional
                    allocations and that the measures used in the
                    allocation formula serve as a surrogate for
                    State performance.  It recognized that the
                    measures are not perfect, but believe they
                    are appropriately used to determine EPA
                    Regional allocations.

                    OSWER stated that activity report data is
                    only one of many tools used for management
                    decisions.   It also stated that the Regions
                    evaluate State program progress,  identify
                    weaknesses and strengths in State programs,
                    and work with the States to identify
                    continuous improvement activities.
OIG Evaluation      We believe that OSWER's response does not
                    appropriately address the issues described in
                    our report.   The Agency's response failed to
                    address our concern that the States are not
                    reporting their activities accurately.

                    As explained in the report,  our conclusions
                    were based on our findings that six States
                    did not accurately report their activities.
                    We found a high degree of misstatement where

                               42

-------
States generally overstated their
accomplishments.  The States that we reviewed
represented about 23 percent of the reported
activities.  Overstatements reported by those
States ranged from 7 to 47 percent.  That is
a significant reporting problem.

OSWER's response stated that New York,
reported over 12,000 confirmed releases, and
we agree.  However, what it failed to address
was that New York did not actually confirm
these releases.  Therefore, to report that
the State confirmed over 12,000 releases is
incorrect.  OUST's definition of a confirmed
release requires that the implementing agency
(State) has to verify that the release did
occur.  We took exception to the State
reporting that these releases were confirmed
when in fact they were not.

OSWER's comments about information gathering
and its use of the data is in direct conflict
with GPRA.  In fiscal 1995, the LUST program
continued its participation in a GPRA
performance pilot.  OUST used the three
performance activities that we reviewed to
evaluate progress in meeting its performance
goal under the pilot.

The GPRA requires that agencies determine if
their programs have produced real results.
The GAO guidelines to implement the GPRA
state that "Managing better requires that
agencies have, and rely upon, sound financial
and program information.  It also states that
to be successful in measuring performance,
agencies must ... link responsibility to
establish accountability for results; and use
data that was sufficiently complete,
accurate, and consistent to be useful in
decision-making."

We believe that OSWER did use activity
reports to manage the program because it
responded that the original intent of the
measures was to determine the extent of the
LUST problem, track remediation efforts, and
to establish a baseline of performance.

We do agree with OSWER that the Regions meet
with the States for mid-year and end-of-year
meetings.  Information is gathered informally

            43

-------
and anecdotally at these meetings.  However,
the Regions did not monitor the States'
performance to the extent that significant
reporting errors were identified and
corrected.
           44

-------
                      APPENDIX A
                         ACRONYMS
Acronym                        Name
 BTEX      benzene, toluene, ethylbenze, and xylene
 CAs       Cooperative Agreements
 CFR       Code of Federal Regulations
 EPA       Environmental Protection Agency
 GPRA      Government Performance and Results Act of 1993
 LUST      Leaking Underground Storage Tank
 OIG       Office of Inspector General
 OMB       Office of Management and Budget
 OSWER     Office of Solid Waste and Emergency Response
 OUST      Office of Underground Storage Tanks
 RBCA      risk-based corrective action
 RCRA      Resource Conservation and Recovery Act
 SARA      Superfund Amendments and Reauthorization Act
 UST       Underground Storage Tank
                            45

-------
(This page left intentionally blank.)
                 46

-------
                          APPENDIX B

                 OSWER RESPONSE TO DRAFT REPORT
Attach are the Office of Solid Waste and  Emergency Response's
comments to the draft report.   The specific comments reference
page numbers of the draft report.   The  page numbers have changed
in the final report.
                               47

-------
(This page left intentionally blank.)
                 48

-------
             UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                        WASHINGTON, D.C. 20460
                                                          OFFICE OF
                                                     SOLID WASTE AND EMERGENCY
                           JUN 2 7 !996                     RESPONSE
 MEMORANDUM

 SUBJECT:   Comments on "Consolidated Report on EPA's LUST Program,
           Draft Report No.

 FROM:      Elliott Laws
           Assistant Admf

 TO:        Michael Simmons
           Deputy Assistant IG for Internal & Performance Audits

     Thank you for the opportunity to review the draft capping
 report  on the audits of state leaking underground storage tank
 (LUST)  programs.   The results of nine state audits were utilized
 to develop this capping report.   Those audits began in August
 1987 — nearly nine years ago.   This is a very long period of
 time from which to consolidate audit results.

     The  early audits conducted were of nascent state LUST
 programs.   In fact,  the LUST Trust Fund program was authorized by
 Congress  in late  1986,  and EPA awarded the first LUST Trust  Fund
 Cooperative Agreement to a state in 1987.  The audited state
 programs,  and their peers not audited,  have made substantial
 improvements since then.   Unfortunately,  the draft report does
 not capture nor adequately reflect the continued development of
 these state programs,  the tremendous success of our state
 partners  in completing over 140,000 LUST cleanups,  nor the
 magnitude  of the  states'  workload.   As of March 31,  1996,  there
 were over  314,000 confirmed releases from underground storage
 tanks (USTs).   This  means that,  on average,  each state has had to
 respond to over 6,000  releases during the last nine years.   These
 numbers are astronomical  and unprecedented in other environmental
 programs.

     As you are aware,  the Agency has moved aggressively towards
 establishing a  new,  more  cooperative,  and less directive
 relationship with our  state partners,  epitomized in the oversight
 reform and  performance  partnership initiatives.   In addition,
 because of  our  strong  reliance on state programs the Agency  has
 begun planning  to reduce  over time the federal role in the
 UST/LUST program  except on tribal lands.   This is also at a  time
when there  has  been  a  35%  reduction in our FY 1996  LUST Trust
Fund appropriation.  Over  80% of  LUST Trust Funds have been
dedicated to  cooperative  agreements with  states.   This funding

                            49

-------
 shortfall will  result in reductions in the number of state LUST
 staff.  While the President's FY97 budget requested that the FY95
 appropriations  level  be reinstated, the House Appropriations
 Subcommittee's  recent markup indicated a continuation of the 35%
 reduction during  FY97.   As  we continue to work with our state
 partners in implementing the LUST program,  we must balance the
 draft report's  conclusions  that  propose a greater level of
 oversight and a more  intrusive role for EPA vis-a-vis state
 programs.  While  we certainly agree that fiscal management and
 accountability  are important,  it presents a difficult quandary
 when we ask the states  to do more with significantly less funding
 and at a time when they are  forced to  reduce staff.

     Because a  number of  the state audits referenced in the draft
 report are dated,  and since  the  Regions  previously commented on
 the individual  state audits, we  have not commented extensively on
 state-specific  information or  conclusions, nor  commented on their
 veracity or applicability to the present day.   Instead,  we  have
 focused on national issues and the recommendations.

     Please contact Lisa  Lund or Josh Baylson in the Office of
 Underground Storage Tanks at 703-603-9900 if you have any
questions.

Attachment

cc: Lisa Lund,  OUST
                           50

-------
         OUST'8  COMMENTS ON OZ6 DRAFT CONSOLIDATED REPORT
       EPA'8  LEAKING UNDERGROUND STORAGE TANK (LUST) PROGRAM


 COMMENTS ON  EXECUTIVE SUMMARY AND INTRODUCTION

      The second paragraph on pages ill & 1 is awkward and
 inadvertently attributes a negative purpose to the capping
 report.   The paragraph should be recast as follows:

      The LUST Program's current focus is on program results and
      not on  adherence to prescribed procedures.   This approach is
      consistent with the Government Performance and Results Act
      of  1993 (GPRA),  which has shifted the focus of federal
      management and accountability away from inputs,  such as
      personnel  levels,  and adherence to prescribed processes, to
      a greater  focus on achieving desired program results.   We
      have identified three areas where LUST Program results can
      and should be  improved.

      Pages iv & 2 state,  in describing the statute,  "...LUST
 Trust Fund,  financed by one-tenth of a cent tax on each gallon of
 gasoline sold.1*  This statement is factually incorrect.   The tax
 is on motor  fuels;  it is not limited to gasoline.

      Page 2  (under  States Receive LUST Funds...)  describes
 allowable state uses  of LUST Trust Funds.   The  list is incomplete
 and should be expanded to include "oversight of  responsible-party
 lead  cleanups."  We  suggest adding this as iii,  making the
 current  iii  ->  iv, and  the current iv -> v.


 COMMENTS  ON  CHAPTER  2

      The  draft  report  states  (pages 7  & 10)  "...125 sites did not
 get adequate oversight  or enforcement  to ensure  protection  of
 human health and the  environment."   Yet the draft  report  fails to
 demonstrate that these  sites  caused any proximate  harm to human
 health or the environment.  As noted in the transmittal
 memorandum,  this conclusion was  based  in part on information that
 is over five years old.   In  addition,  this  alleged deficiency is
 reported to have created management and oversight  problems  for
 these  sites.   No evidence  is presented that there  were any
 negative human health or  environmental consequences at these or
 anv other sites.  We also  are disappointed  that  the draft report
 focused on the very small minority  of  sites where  there appeared
 to be problems and ignored the vast majority of  sites where
 things went well.   Most of the states  have  made  significant
progress since the audits.  In fact, all  nine states audited are
now in the process of implementing  risk-based corrective  action
 (RBCA) .

     RBCA allows regulatory agencies to focus limited resources
on sites that pose the highest risks,  protecting human health and

                            51

-------
 the environment while ensuring that all sites move towards
 closure.  RBCA enables implementing agencies to focus site
 assessment data gathering, categorize or classify sites,
 determine what (if any) further action is necessary, better
 establish cleanup goals, and decide on the level of oversight
 necessary.

      RBCA represents a paradigm shift in the way LUST regulatory
 agencies do business, from following predetermined and
 standardized cleanup levels to setting site-specific levels based
 on risk.  It also allows regulatory agencies to identify high
 risk sites based on actual risk — proximity to pathways and
 receptors — instead of simplistically classifying all
 groundwater sites as high risk,  and allows other important
 factors to be considered.   For example,  environmental justice
 concerns and the need to address Brownfields,  which according to
 traditional ranking systems would often be rated low risk because
 they have not contaminated drinking water sources,  can be better
 addressed using a site specific  decision-making approach.  We
 believe increased implementation of RBCA is the best approach to
 address the concerns outlined in the draft report.

      The column at the top of page 8 states "Management Reports
 Were Inaccurate."  A more  accurate statement is:  Management
 Reports were Sometimes Inaccurate.   Likewise,  the associated text
 should read "LUST site status reports were not always accurate in
 Idaho,  Kansas,  New York, and Oregon.   As a result,  these States
 did not always  have reliable information to effectively monitor
 LUST site remediation progress."

      The last sentence in  the same paragraph makes  an unwarranted
 leap in judgement by stating "Without a  current,  accurate,  and
 complete site status report,  management  could  not assure that
 significant sites were cleaned up  timely."   This  statement
 discounts,  without justification,  the extensive knowledge of the
 state technical staff overseeing sites and  of  their management.
 Just because information is  not recorded in  writing,  does not
 mean that it does not exist.   A more  accurate  statement  would be:
 ...management may not be able to assure  that significant sites
 were cleaned up in a timely  fashion.

      Later  on page 8;  OSWER  Directive 9650.10  was revised and
 reissued  on May 24,  1994,  as  OSWER  Directive 9650.10A.   The next
 paragraph quotes  the directive as  stating  "The State  will ensure
 that  a  priority system for addressing underground storage tank
 petroleum release  sites is established and maintained."   The
 appropriate context  for this  statement is as it applies  to  state-
 lead  corrective and/or enforcement  actions.  This statement does
 not  apply to  oversight of responsible party-lead  corrective
 actions.  The role of  priority systems is discussed again (page
 20) when  the  report  states "We believe that  EPA needs to give
more emphasis to  clean-ups of high  risk  sites."   Implicit in this
                           52

-------
 statement and others in the draft report is the view that states
 should numerically rank LUST sites and spend money (for
 corrective action ojr oversight) only on the highest ranked sites.

      Given that 97% of LUST sites are cleaned up by responsible
 parties,  our goal has been to move all sites through the  cleanup
 process concurrently, regardless of their rank.   We agree that
 the focus of state oversight activities should be on those sites
 with the greatest risk, but not to the exclusion of low risk
 sites,  at which corrective actions may stall without state
 review, approval, and oversight activities.   Lack of attention  to
 lower risk sites creates many problems, including Brownfields,
 and environmental justice concerns.  We believe that this is
 generally a more protective approach than using a priority system
 to  decide which cleanups may move forward and which will  stall
 due to the limited capacity of most existing oversight  programs.
 This policy was explained in detail in an April  25,  1994,
 memorandum from David Ziegele to Charles Reisig  of the  OIG.

      Page 11 states,  in reference to the audit of California's
 LUST program,  that "...the priority system for LUST clean-up was
 unacceptable and that the State's LUST enforcement program was
 ineffective."   We continue to agree with Region  9's position in
 response  to the California audit,  i.e.,  that we  do not  believe
 that their priority system was unacceptable,  but needed
 refinement,  and that we do not believe that  their LUST
 enforcement program is ineffective;  California has taken over
 9,000 LUST enforcement actions.

               Recommendations 1. 2 & 3  (page 21)

      Since RBCA allows regulatory agencies to focus  limited
 resources on sites that pose the highest risks,  protecting human
 health  and the environment while ensuring that all  sites move
 towards closure,  we believe increased  implementation of RBCA is
 the  best  approach to  address the concerns outlined in the draft
 report.

      We will implement the following management  controls to
 monitor state  progress in  addressing oversight and enforcement of
 high  risk sites through the use  of RBCA,  utilizing the
 cooperative  agreements as  tools  to assist in implementing RBCA,
 and measuring  the progress of  state  implementation of RBCA.

      We will continue to actively  participate in the Partnership
 in RBCA Implementation (PIRI)  with state governments, the
American  Society  for  Testing and Materials,  Amoco Oil Corp.,
British Petroleum Oil Co.,  Chevron USA,  Exxon Co.  USA,  Mobil Oil
Corp.,  and Shell  Oil  Co. to reduce the  risk  from LUSTs  by
targeting those releases for cleanup which pose  the greatest risk
to public health  and  the environment.   Other groups that will
contribute funding, technical  expertise,  or  other support  to this


                            53

-------
 effort include Unocal,  the American Petroleum Institute, the
 Petroleum Marketers Association of America,  the Environmental
 Bankers Association, and Environmental and Commercial Insurance
 Inc.   Under a Memorandum of Understanding,  each of the partners
 contributes funds towards the development and delivery of a
 series of training sessions for states to teach the new risk-
 based approaches to cleanup.   The three module training is
 followed by an implementation (policy)  design phase and actual
 implementation of RBCA.   We will track states'  progress in
 receiving the training  modules and implementing RBCA.   We will
 monitor this progress through PIRI and as part of  our annual,
 Regional Strategic Overview planning process with  the Regions.
 In addition,  we will educate Regional  managers  about RBCA,  PIRI,
 use of the cooperative  agreements,  and our efforts to track state
 progress at the next UST/LUST Regional  Managers meeting
 (currently planned for  September 1996).


 COMMENTS ON CHAPTER 3

      We acknowledge that  improvements would  be  beneficial to some
 state cost recovery programs.  We  do not  believe,  however,  that
 the problems  in a few states  as  described in the draft report  are
 indicative of systemic, national problems.   For example,  although
 two of the seven state cost recovery programs were judged
 ineffective in the draft  report, two states'  programs  were
 considered very effective, while three  others were .identified  as
 needing some  degree of improvement.

      States are required  to have a* cost recovery program  that  is
 consistent with EPA's LUST Trust Fund Cost Recovery  Policy  (OSWER
 Directive  9610.10A).  Under Subtitle I of RCRA, UST  owners  and
 operators  are liable  for  costs incurred by EPA  or  a  state in
 performing enforcement or corrective action  at  LUST  sites.   On
 page  22, the  draft report incorrectly states that  Subtitle  I
 "requires  that  the parties responsible for contaminating  sites
 clean  them up or  reimburse the State for doing  so.H  While  tank
 owners  and operators  are  liable  for such costs, the  Cost  Recovery
 Policy  provides states with significant flexibility  and
 discretion in determining under what circumstances they will
 pursue  such costs, which costs to pursue, and how much effort  to
 devote  to  pursuit  of  these costs.  OUST further clarified the
 Cost Recovery Policy  in a memorandum from Lisa  Lund  to the  UST
 Regional Program Managers, dated November 8,  1994,  which  we have
 provided previously to the DIG.

     Our primary concern with chapter 3 is that  it generally
 fails  to recognize the state flexibility and discretion that
 intentionally has been built into the Cost Recovery  Policy.  For
 instance, the OIG  is critical of states that failed  to both
document and pursue all  costs for which UST owners could  legally
be held liable under Subtitle I.  This position is inconsistent

                           54

-------
with the  Cost Recovery Policy.  It also is internally
inconsistent with other portions of chapter 3.  For  example,  on
page 24,  the draft report is accurate when it asserts that states
are only  required to document site-specific costs in a  limited
set of circumstances.   The draft report is accurate  again  on  page
25, when  it asserts that "States are not required to [document
or] recover costs ...  for oversight, of responsible party
corrective  actions."   These provisions of EPA's Cost Recovery
Policy•constitute major exceptions to the draft report's
contention  that  states must document and pursue all  costs  for
which UST owners could be held liable.  In addition, the draft
report criticizes states that failed to allocate portions  of
indirect  costs,  management and support costs, and to assess
interest  on cost recovery claims.  The Cost Recovery Policy
encourages  but does not require states to either document  or
pursue these costs as  part of their cost recovery program.  Thus,
contrary  to several of the statements made in this chapter,
states are  not required to both document and pursue  all costs  for
which UST owners could be held liable under Subtitle I.

     We also are concerned by the statements: "Ineffective cost
recovery  programs hinder the efforts of the LUST program to
protect human health and the environment" (pages vi and 22) and
"An aggressive cost recovery policy could be used as an effective
tool to stimulate site  clean-ups" (page 28).   These and similar
statements  throughout  the chapter are unsupported by any findings
in the draft report.  At no point does the draft report provide
any evidence that harm  has occurred to human health or  the
environment due  to ineffective state cost recovery programs.  The
fact that chapter 3 criticizes the state of Ohio for lack of a
cost recovery policy while chapter 2 singles out Ohio for praise
in "managing LUST activities and directing its resources to
clean-up high priority  sites" (page 19)  indicates the lack of a
demonstrated connection between a state's cost recovery program
and protection of  human health and the environment.   In our
experience  managing the national LUST program,  the presence of
more than forty  state fund reimbursement programs for UST'
cleanups is  a  far  greater influence in stimulating responsible
party-lead  cleanups than a state's cost recovery program for LUST
Trust Fund monies.

     When considering potential cost recovery actions,  the Cost
Recovery Policy encourages states to make decisions that are in
the best interests of their  overall LUST program and that reflect
efficient use of Trust Fund  monies.   Regardless  of the  action
taken by a state in exercising its discretion in cost recovery
cases,  the state is required to fully document its decision and
to formally close out all  cases.   The draft  report acknowledges
this point on page 28 and  we agree.   We believe  in some
instances, however, that the draft report improperly equated  lack
of documentation of cost recovery decisions  with the absence of a
cost recovery program.   For  example,  page 24  includes the


                           55

-------
 statement tha- Kansas  does  not have a cost recovery program.  EPA
 Region 7 and the state, however,  indicate that Kansas does have a
 cost recovery policy,  but that cost recovery was not pursued for
 various reasons.  Both the  Region and state agree that the
 problem was that these decisions  were not documented.
                               1.  2   and 3  facre 28)
      We will  implement management controls to ensure that states
 develop and implement cost recovery programs in accordance with
 Subtitle I requirements and EPA's Cost Recovery Policy for the
 LUST Trust Fund.  We will use the cooperative agreements as a
 tool to assist  in achieving these results.   We will monitor
 states'  progress as part of our annual, Regional Strategic
 Overview planning process with the Regions and states.  In
 addition, we  will discuss these expectations at the next UST/LUST
 Regional Managers meeting (currently planned for September 1996) .
 COMMENTS ON CHAPTER 4

     The draft report states (page v) "We found  that the States'
 activity reports were generally unreliable and potentially
 misleading."  This statement is overstated and unsupported based
 on the limited size of the database examined, the  actual findings
 associated with the data, and EPA's use of the data.  A more
 appropriate statement is: the states' activity reports  were not
 always accurate and therefore potentially misleading.   Similarly
 (pages vi, 29 & 31) there are several references that "activity
 reports were not accurate," or "EPA did not accurately  report
 LUST program results."  In all these cases the description should
 be modified to read:  activity reports were not always accurate
 and EPA did not always accurately report LUST program results.
 On page 37, the draft report notes "we conclude  that  the Regions
 had not identified the reporting errors and had  not taken steps
 to ensure that the States were reporting accurate  information."
 To be accurate,  this  statement needs to be modified to  read:  the
 Regions had not always identified...and had not  taken steps to
 ensure that the states were always reporting accurate
 information.

     The draft report states (page 36)  that "... we believe the
 misstatements represent a significant reporting  problem."   OUST
 strongly disagrees with this statement.   A more  accurate
 statement would  be that the misstatements have caused reporting
 problems.   We are concerned that the draft report  draws
 conclusions about all state data based on very few data points.
 For example,  in  New York,  it appears that data for 70 sites were
 investigated.  New York reports over 12,000 confirmed releases,
70 sites  constitute significantly less than one  percent of the
total  release population.   The  draft report recognizes  this by


                            56

-------
stating  "we a nnot statistically evaluate the error rate in the
activity reports....H

     The statements concerning California (page 32)  that "The
activity report...significantly overstated the number of
confirmed releases and cleanups initiated," and (page 33)
"leaking tank sites that had not been confirmed were included in
the report as confirmed leaking sites" are misleading.   In  fact,
EPA Region 9 believes that the state has been under-reporting due
to the large number of local databases.   California  was  initially
reporting state-regulated tanks.   However,  after a training
session  was conducted for the state,  the data was modified  and
the reporting problem fixed.   Furthermore,  it is appropriate  for
states to report multiple confirmed releases at the  same site if
the new  release  occurs after work on the original release is  well
underway and the second release requires a separate  remedial
response.   Finally,  in California the term "no action" means  that
the state or local implementing agency has not had a chance to
either enter the data into the system or has not reviewed the
site assessment  or corrective action plan.   It does  NOT  indicate
whether  or not the site is a confirmed release.  In  other words,
a release can be confirmed and documented without additional  data
being entered into the database.

     OUST disagrees with the statement (page 36) that "Strict
adherence to the definitions would provide results that  can be
compared among the States."   As with any data base,  we
acknowledge that the  numbers are  not always perfect, people
occasionally make  mistakes,  and there is not always  complete
consistency in the way states interpret  the definitions.
However,  it is more important to  assure  consistent interpretation
of the measures  within individual states rather  than between  the
states.   We have encouraged  constant improvement within  state
programs  and not competition  between  programs.   This concept  is
further  reinforced by the  design  of  the  national UST program.  We
adopted  a  decentralized model that empowers state and local
programs.   This  arrangement  allows states to run programs that
are tailored to  the needs  and demands of their  local environments
and regulated communities, and by design leads  to variations
between  individual state programs.   These variations make it
invalid to  compare state programs to  each other  using only  the
reporting data.

     The draft report  states  (pages  vii  & 29)  that "it is
necessary that the information be accurate,  complete, and
comparable to better manage the LUST  program."   This  sentence is
unsubstantiated  and should be  deleted.   The STARS measures  were
never intended to be used  to  manage  the  LUST program.  The  data
is only one  of many tools we  use  to make management  and  program
decisions  in  the LUST  program.  The  STARS measures were
originally developed to  determine the extent of  the  LUST problem,
to track nationwide remediation efforts,  and to  establish a


                           57

-------
 baseline of performance  for  individual states.  The STARS
 measures provide us with a general  indication of national program
 progress.  The measures  were not intended to compare states to
 each other, but instead  to serve as a relative measure and as a
 way to identify state progress and areas for improvement.  For
 example, we have asked the Regions to investigate the reasons why
 individual states fall below the national average for completing
 cleanups and to determine if this is a program management problem
 or merely an indication  of a preponderance of complicated
 groundwater sites.  Regions  and states also use the data to
 monitor state progress and to continuously improve state
 procedures using total quality management tools.

      The draft report states (pages 29 & 35)  that "EPA
 Headquarters uses the information to make nationwide funding
 decisions, to report program results to Congress,  and to report
 the results of the program in the Agency's financial statements."
 Since the STARS data is only one source of information,  it would
 be more accurate to state:  EPA Headquarters uses the information
 to help make nationwide funding decisions,  to help report program
 progress to Congress,  and to help report the results of  the
 program in the Agency's financial statements.

      The draft report notes that "the cumulative totals  for the
 above listed activities were used to determine State Trust Fund
 allocations...."  This is an incorrect statement.   The data is
 used to determine Regional  allocations,  not state  allocations.
 Furthermore,  OUST disagrees and no evidence is presented that
 substantiates the bold statement (page 37)  "that accurate
 reporting would provide a more  equitable  distribution of LUST
 Trust Funds among the  States."   This statement should be deleted.

      The STARS measures used in the national  allocation  formula
 serve as a surrogate for state  performance.   We  recognize that
 the  measures  are not perfect, but believe they are  appropriately
 used to determine EPA  Regional  allocations,  i.e.,  the amount of
 LUST Trust Fund money  each  Region receives  per year.  We work
 closely with  representatives  from the  ten Regions,  in addition to
 consulting with states,  to  determine the  components  of the
 Regional  Allocation  Formula.  Regional Program Managers  actually
 determine the  amount of individual  state  awards  based on a
 variety of  factors,  including additional  state data  (such as
 spending  patterns),  state reviews,  and their day-to-day  knowledge
 of the  state programs.

     As noted  above, the  STARS  data  is only one  of many  tools
used for  management  decisions.   Each Region develops an  annual
Regional  Strategic Overview  (RSO) to assess the  status of state
progress  in implementing  the  program and  identifying key
improvement areas for  the coming  year.  This process is  based on
strategic discussions  that OUST has  with  Regional management and
the Regions have with  the states. In mid-year  and end-of-year


                           58    :

-------
 state evaluations, the Regions evaluate state program progress,
 identify weaknesses and strengths in state programs,  and work
 with the states to identify .continuous improvement activities.
 In addition, within each Region, state project officers/program
 managers have day-to-day contact with each of the states.

      We use information generated from all of these activities  to
 report program results within the Agency and to Congress.   We
 believe we are accurately portraying program results  and making
 sound management decisions.   In fact,  OUST's judicious  use  of
 data,  collegial management approach,  and innovative strategic
 planning method have long been praised by states and  the
 regulated community,  and similar concepts are being diffused
 throughout the Agency under the direction of the White  House
 under the rubric of Total Quality Management,  state empowerment
 (performance partnerships),  and oversight reform.

                  Recommendations l &  2  foaae 37)

     We will implement the following management  controls to
 ensure that reliable  and timely information  is obtained,
 maintained,  reported,  and used.

       OUST will revise (or issue revision memorandums)  the UST
 Grant  Guidance and LUST Trust  Fund Cooperative Agreement
 Guidelines for FY98 and beyond.   (It is  too  late  to revise the
 guidances  for FY97 since negotiations are well underway.)  The
 revisions  to both  guidances will address regional and state
 verification of state  reporting data, clarify the definitions and
 clearly  delineate  reporting requirements  (i.e.,, report  only
 federally  regulated Subtitle I  USTs).

     In  the  interim, OUST will  address this  issue at our next
 UST/LUST Regional  Managers meeting  (currently scheduled for
 September  1996).   The  Regions will work with the  states in mid-
 year and end-of-year state evaluations to assess  state  data.
 Headquarters will  monitor the progress of Regional verification
 of state data  as part  of our annual Regional Strategic  Overview
 planning process with  the Regions.

     OUST also  is  in the process  of converting approximately
 twenty states' data tracking systems to a new windows-based
 system know as UST-Access.  This  will provide these state
programs with a state-of-the-art  tracking system  and additional
capacity for data verification.
                          59

-------
(This page left intentionally blank.)
                 60

-------
                          APPENDIX C

                       REPORT DISTRIBUTION
Office of Inspector General

          Inspector General  (2410)

Headquarters Office

          Assistant Administrator for
            Solid Waste and Emergency Response (5101)

          Director, Office of Underground Storage Tanks (5401W)

          Agency Followup Official, (3101), Attn:  Assistant
            Administrator for Administration and Resource
            Management

          Agency Followup Coordinator (3304), Attn:  Director,
            Resource Management Division

          Audit Followup Coordinator (5101)

          Associate Administrator for Regional Operations and
            State/Local Relations  (1501)

          Associate Administrator for Congressional and
            Legislative Affairs  (1301)

          Associate Administrator for Communications,  Education
            and Public Affairs  (1701)

          Headquarters Library  (3304)

Ten Regional Offices

          Regional Administrator

          Regional LUST Program Director

          Audit Followup Coordinator

          Public Affairs Office

          Regional Library
                                61

-------
External
          General Accounting Office
                               62

-------