This is the accessible text file for GAO report number GAO-13-252 
entitled 'Superfund: EPA Should Take Steps to Improve Its Management 
of Alternatives to Placing Sites on the National Priorities List' 
which was released on May 10, 2013. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as 
part of a longer term project to improve GAO products' accessibility. 
Every attempt has been made to maintain the structural and data 
integrity of the original printed product. Accessibility features, 
such as text descriptions of tables, consecutively numbered footnotes 
placed at the end of the file, and the text of agency comment letters, 
are provided but may not exactly duplicate the presentation or format 
of the printed version. The portable document format (PDF) file is an 
exact electronic replica of the printed version. We welcome your 
feedback. Please E-mail your comments regarding the contents or 
accessibility features of this document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

United States Government Accountability Office: 
GAO: 

Report to Congressional Requesters: 

April 2013: 

Superfund: 

EPA Should Take Steps to Improve Its Management of Alternatives to 
Placing Sites on the National Priorities List: 

GAO-13-252: 

GAO Highlights: 

Highlights of GAO-13-252, a report to congressional requesters. 

Why GAO Did This Study: 

Under the Superfund program, EPA may address the long-term cleanup of 
certain hazardous waste sites by placing them on the NPL and 
overseeing the cleanup. To be eligible for the NPL, a site must be 
sufficiently contaminated, among other things. EPA regions have 
discretion to choose among several other approaches to address sites 
eligible for the NPL. For example, under the Superfund program, EPA 
regions may enter into agreements with PRPs using the SA approach. EPA 
may also defer the oversight of cleanup at eligible sites to 
approaches outside of the Superfund program. GAO was asked to review EPA
’s implementation of the SA approach and how it compares with the NPL 
approach. This report examines (1) how EPA addresses the cleanup of 
sites it has identified as eligible for the NPL, (2) how the processes 
for implementing the SA and NPL approaches compare, and (3) how SA 
agreement sites compare with similar NPL sites in completing the 
cleanup process. GAO reviewed applicable laws, regulations, and 
guidance; analyzed program data as of December 2012; interviewed EPA 
officials; and compared SA agreement sites with 74 NPL sites selected 
based on their similarity to SA agreement sites. 

What GAO Found: 

The Environmental Protection Agency (EPA) most commonly addresses the 
cleanup of sites it has identified as eligible for the National 
Priorities List (NPL) by deferring oversight of the cleanup to 
approaches outside of the Superfund program. As of December 2012, of 
the 3,402 sites EPA identified as potentially eligible, EPA has 
deferred oversight of 1,984 sites to approaches outside the Superfund 
program, including 1,766 Other Cleanup Activity (OCA) deferrals to 
states and other entities. However, EPA has not issued guidance for 
OCA deferrals as it has for the other cleanup approaches. Moreover, EPA’
s program guidance does not clearly define each type of OCA deferral 
or specify in detail the documentation EPA regions should have to 
support their decisions on OCA deferrals. Without clearer guidance on 
OCA deferrals, EPA cannot be reasonably assured that its regions are 
consistently tracking these sites or that their documentation will be 
appropriate or sufficient to verify that these sites have been 
deferred or have completed cleanup. Under the Superfund program, EPA 
oversees the cleanup of 1,313 sites on the NPL, 67 sites under the 
Superfund Alternative (SA) approach, and at least 38 sites under 
another undefined approach. 

The processes for implementing the SA and NPL approaches, while 
similar in many ways, have several differences. EPA has accounted for 
some of these differences in its SA guidance by listing specific 
provisions for SA agreements with potentially responsible parties 
(PRP), such as owners and operators of a site. One such provision 
helps ensure cleanups are not delayed by a loss of funding if the PRP 
stops cleaning up the site. However, some EPA regions have entered 
into agreements with PRPs at sites that officials said were likely 
eligible for the SA approach without following the SA guidance. Such 
agreements may not benefit from EPA’s provisions for SA agreements. 
EPA headquarters officials said the agency prefers regions to use the 
SA approach at such sites, but EPA has not stated this preference 
explicitly in its guidance. In addition, EPA’s tracking and reporting 
of certain aspects of the process under the SA approach differs from 
that under the NPL approach. As a result, EPA’s tracking of SA 
agreement sites in its Superfund database is incomplete; the standards 
for documenting the NPL eligibility of SA agreement sites are less 
clear than those for NPL sites; and EPA is not publicly reporting a 
full picture of SA agreement sites. Unless EPA makes improvements in 
these areas, its management of the process at SA agreement sites may 
be hampered. 

The SA agreement sites showed mixed results in completing the cleanup 
process when compared with 74 similar NPL sites GAO analyzed. 
Specifically, SA agreement and NPL sites in GAO’s analysis showed 
mixed results in the average time to complete negotiations with PRPs 
and for specific cleanup activities, such as remedial investigation 
and feasibility studies, remedial designs, and remedial actions. In 
addition, a lower proportion of SA agreement sites have completed 
cleanup compared with similar NPL sites. SA agreement sites tend to be 
in earlier phases of the cleanup process because the SA approach began 
more recently than the NPL approach. Given the limited number of 
activities for both NPL and SA agreement sites in GAO’s analysis, 
these differences cannot be attributed entirely to the type of 
approach used at each site. 

What GAO Recommends: 

GAO recommends, among other things, that EPA issue guidance to define 
and clarify documentation requirements for OCA deferrals and clarify 
its policies on SA agreement sites. EPA agreed with the report’s 
recommendations. 

View [hyperlink, http://www.gao.gov/products/GAO-13-252]. For more 
information, contact David C. Trimble at (202) 512-3841 or 
trimbled@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

EPA Defers Oversight of a Majority of Sites Eligible for the NPL to 
Approaches Outside of the Superfund Program: 

Processes for Implementing the SA and NPL Approaches Differ in a Few 
Significant Ways: 

SA Agreement Sites Showed Mixed Results in Completing the Cleanup 
Process When Compared with Similar NPL Sites: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: Long-Term Cleanup Approaches by EPA Region: 

Appendix III: Data Analysis of SA Agreement Sites and Similar NPL 
Sites: 

Appendix IV: Comments from the Environmental Protection Agency: 

Appendix V: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: Number of Sites by Region and Long-Term Cleanup Approach, as 
of December 2012: 

Table 2: Percentage of Sites by Region and Long-Term Cleanup Approach, 
as of December 2012: 

Table 3: Median Length of Negotiations in Months for Agreements 
between EPA and PRPs at 66 SA Agreement Sites and 74 Similar NPL Sites 
for Agreements Finalized from June 2002 through December 2012: 

Table 4: Median Length of Cleanup Activities for 66 SA Agreement Sites 
and 74 Similar NPL Sites Completed from June 2002 through December 
2012: 

Figures: 

Figure 1: EPA's 10 Regions: 

Figure 2: Approaches Identified by EPA to Clean Up Sites Eligible for 
the NPL: 

Figure 3: Long-Term Cleanup Process at Sites under the Superfund 
Program: 

Figure 4: Current Sites Identified as Eligible for the NPL Using 
Different Long-Term Cleanup Approaches, as of December 2012: 

Figure 5: Number of SA Agreement Sites by EPA Region, as of December 
2012: 

Figure 6: Construction Completions at SA Sites and Similar NPL Sites: 

Abbreviations: 

CERCLA: Comprehensive Environmental Response, Compensation, and 
Liability Act: 

CERCLIS: Comprehensive Environmental Response, Compensation, and 
Liability Information System: 

EPA: Environmental Protection Agency: 

GPRA: Government Performance and Results Act of 1993: 

IG: Inspector General: 

NPL: National Priorities List: 

NRC: Nuclear Regulatory Commission: 

NRD: natural resource damages: 

OCA: Other Cleanup Activity: 

PRP: potentially responsible party: 

RCRA: Resource Conservation and Recovery Act: 

SA: Superfund Alternative: 

[End of section] 

United States Government Accountability Office: 
Washington, DC 20548: 

April 9, 2013: 

The Honorable Henry A. Waxman: 
Ranking Member: 
Committee on Energy and Commerce: 
House of Representatives: 

The Honorable John D. Dingell: 
House of Representatives: 

The Environmental Protection Agency (EPA) estimates that one in four 
Americans lives within 3 miles of a hazardous waste site. Many 
hazardous waste sites pose serious risks to human health and the 
environment, and their cleanup can be expensive and take many years to 
complete. While these sites may not necessarily be subject to a 
federal cleanup requirement, several approaches exist to address such 
long-term cleanups. EPA manages the Superfund program--the federal 
government's principal program to clean up hazardous waste sites--
under the Comprehensive Environmental Response, Compensation, and 
Liability Act (CERCLA) of 1980.[Footnote 1] Under this program, EPA 
can place sites with contamination that is sufficiently severe on the 
National Priorities List (NPL), a list of sites for attention under 
the federal Superfund program that includes sites among the nation's 
most seriously contaminated.[Footnote 2] At sites on the NPL, EPA 
oversees the cleanup, which may be performed by potentially 
responsible parties (PRP)[Footnote 3] or by EPA itself. Aside from 
placing an eligible site on the NPL, EPA can also ensure cleanup 
through the Superfund Alternative (SA) approach.[Footnote 4] Under 
this approach, PRPs agree to clean up sites (hereafter referred to as 
SA agreement sites), and EPA does not list the sites on the NPL at 
that time, which allows PRPs to avoid the perceived stigma of having a 
site on the NPL. Since EPA first issued guidance on the SA approach in 
2002, EPA and its Office of Inspector General (IG) have evaluated the 
approach and made various recommendations to improve its 
implementation. In 2010, EPA reported to the IG that it had 
implemented the recommended actions.[Footnote 5] Where EPA decides not 
to address the site under the Superfund program (i.e., list the site 
on the NPL, use the SA approach, or otherwise retain oversight), EPA 
may defer[Footnote 6] sites whose contamination makes them eligible 
for the NPL to other federal and state cleanup approaches.[Footnote 7] 

EPA's regional offices may discover potential hazardous waste sites, 
or such sites may come to EPA's attention through reports from other 
federal agencies, state agencies, or citizens. EPA then reviews 
available information about each site to decide whether to add it to 
the Superfund program database--the Comprehensive Environmental 
Response, Compensation, and Liability Information System (CERCLIS). At 
some point after a site is added to CERCLIS, EPA assesses the severity 
of the contamination at the site to determine whether the site is 
eligible to be placed on the NPL. If the site has some contamination 
but is not eligible for the NPL, EPA will not pursue a long-term 
cleanup under the Superfund program; state cleanup programs or other 
programs may still address the contamination at the site. Long-term 
cleanups at sites eligible for the NPL that are conducted under the 
Superfund program generally follow an established process. This 
process consists of several phases, including studying site 
conditions, selecting a method to clean up the site, and conducting 
the actual cleanup. In some cases, cleanup of a site is divided into 
smaller parts, known as operable units, and cleanup may proceed at 
different rates at each of these operable units.[Footnote 8] To 
accomplish the cleanup, EPA may, among other options, negotiate and 
enter into agreements with PRPs for them to address contamination at 
the site. These agreements may cover one or more phases of the cleanup 
process and may address one or more operable units. Furthermore, EPA 
may enter into more than one agreement with a PRP at a site. 
Alternately, for sites listed on the NPL, EPA may conduct the cleanup 
itself and pursue costs from PRPs through administrative or judicial 
actions. 

In addition to conducting long-term cleanups, EPA may use its 
Superfund emergency response authorities to conduct removal actions at 
sites. Removal actions are usually short-term cleanups at sites that 
pose immediate threats to human health or the environment. Under the 
removal program, EPA has conducted thousands of cleanup actions 
instead of or in combination with long-term cleanups. 

In this context, you asked us to review the implementation of the SA 
approach and how this approach compares to listing sites on the NPL. 
Our objectives were to examine (1) how EPA addresses the cleanup of 
sites it has identified as eligible for the NPL, (2) how the processes 
for implementing the SA and NPL approaches compare, and (3) how SA 
agreement sites compare with similar NPL sites in completing the 
cleanup process. 

To conduct this work, we analyzed applicable federal laws and EPA 
regulations and guidance to understand the available approaches to 
address hazardous waste sites that are reported to the Superfund 
program and have a level of contamination that makes them eligible for 
the NPL. We also conducted interviews with officials in all 10 EPA 
regions. To determine how EPA addresses the cleanup of sites eligible 
for the NPL, we analyzed EPA data from the CERCLIS database to 
establish the number of such sites that are being addressed through 
each approach as of December 2012. To assess the reliability of these 
data, we analyzed related documentation, examined the data for errors 
or inconsistencies, and interviewed agency officials about any known 
data problems and to learn more about their procedures for maintaining 
the data. We determined the data to be sufficiently reliable for 
calculating durations for completing different cleanup activities, 
including negotiations, at SA agreement and NPL sites. We also 
interviewed a nonprobability, convenience sample of officials from 13 
state cleanup programs who were familiar with available cleanup 
approaches. The sample consisted of representatives from state 
environmental departments taking part in an Association of State and 
Territorial Solid Waste Management Officials conference call who 
agreed to speak with us. Because this was a nonprobability sample, the 
results of our analysis cannot be generalized to all states; however, 
these officials provided important information about the cleanup 
process. To compare the processes for implementing the SA and NPL 
approaches, including the cleanup process and EPA's oversight, we 
discussed the processes with EPA headquarters officials and officials 
from all 10 of EPA's regions, obtained relevant supporting 
documentation, and analyzed previous reviews of the SA approach. To 
determine how SA agreement sites compare with similar NPL sites in 
completing the cleanup process, we constructed a comparison group of 
74 NPL sites with agreements between EPA and PRPs similar to 
agreements at 66 SA agreement sites.[Footnote 9] For example, all 
sites had agreements that were entered into from June 2002 through 
December 2012 that involved cleanup actions at the site. Additionally, 
to match the characteristics of most SA agreements, we included only 
NPL agreements involving relatively few PRPs and agreements whose 
estimated costs were similar to costs at SA agreement sites. Once we 
identified the group of NPL sites from these agreements, we compared 
the duration of relevant cleanup actions, including negotiations, with 
those at SA agreement sites. The 74 NPL sites that were selected for 
comparison with SA agreement sites are not representative of the 
universe of all NPL sites because they were selected based on specific 
criteria. For example, we selected for comparison only NPL sites at 
which a PRP agreed to conduct at least some part of the cleanup. The 
results of our analysis cannot be generalized to all NPL sites; 
however, these sites can provide important information about the 
cleanup process. Appendix I provides a more detailed description of 
our objectives, scope, and methodology. 

We conducted this performance audit from December 2011 to April 2013 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

Background: 

This section discusses EPA's process for assessing sites under the 
Superfund program and the approaches identified by EPA for conducting 
long-term cleanups at sites eligible for the NPL under the Superfund 
program and under other available approaches. 

EPA's Process for Assessing Sites under the Superfund Program: 

Under the Superfund program, EPA assesses hazardous waste sites for 
long-term cleanups through a specific process. At some point after a 
potential hazardous waste site is reported to the Superfund program 
and entered into CERCLIS, EPA regional officials, their contractors, 
or states acting under cooperative agreements with EPA evaluate the 
relative potential for a site to pose a threat to human health and the 
environment. EPA's 10 regional offices each are responsible for 
implementing Superfund within several states and, in some cases, 
territories. Under CERCLA, EPA may only pay for a remedial action at a 
site if the relevant state agrees, among other things, to pay a 
portion of the cleanup expenses, as well as all operations and 
maintenance costs. In addition, under a cooperative agreement with 
EPA, a state may assume the lead oversight role at a site in the 
Superfund program. Figure 1 shows the states included in each of the 
10 EPA regions. 

Figure 1: EPA's 10 Regions: 

[Refer to PDF for image: illustrated U.S. map] 

Region 1: 
Connecticut: 
Maine: 
Massachusetts: 
New Hampshire: 
Rhode Island: 
Vermont: 

Region 2: 
New Jersey: 
New York: 

Region 3: 
Delaware: 
District of Columbia: 
Maryland: 
Pennsylvania: 
Virginia: 
West Virginia: 

Region 4: 
Alabama: 
Florida: 
Georgia: 
Kentucky: 
Mississippi: 
North Carolina: 
South Carolina: 
Tennessee: 

Region 5: 
Illinois: 
Indiana: 
Michigan: 
Minnesota: 
Ohio: 
Wisconsin: 

Region 6: 
Arkansas: 
Louisiana: 
New Mexico: 
Oklahoma: 
Texas: 

Region 7: 
Iowa: 
Kansas: 
Missouri: 
Nebraska: 

Region 8: 
Colorado: 
Montana: 
North Dakota: 
South Dakota: 
Utah: 
Wyoming: 

Region 9: 
Arizona: 
California: 
Hawaii: 
Nevada: 

Regin 10: 
Alaska: 
Idaho: 
Oregon: 
Washington: 

Source: GAO analysis of EPA information; Map Resources (map). 

[End of figure] 

During the initial phases of the long-term cleanup process--known as 
preliminary assessment and site inspection--EPA regional officials or 
their counterparts evaluate the potential need for additional 
investigation or action in connection with a release of hazardous 
substances from a site. Specifically, the preliminary assessment phase 
involves an evaluation of readily available information about a site 
and its surrounding area to determine if the release or potential 
release poses enough of a threat to human health and the environment 
that further investigation is needed. If further investigation is 
needed, a site inspection is performed. During this phase, 
investigators typically collect environmental and waste samples to 
determine what hazardous substances are present. Information collected 
during the preliminary assessment and site inspection is used to 
calculate and document a site's preliminary Hazard Ranking System 
score, which indicates a site's relative threat to human health and 
the environment based on potential pathways of contamination.[Footnote 
10] Sites with a Hazard Ranking System score of 28.50 or greater are 
eligible for listing on the NPL. Information collected from the 
initial assessment phases to develop Hazard Ranking System scores is 
not intended to be sufficient to determine either the extent of 
contamination or how to clean up a particular site. After a site is 
determined to be eligible for the NPL, EPA chooses which long-term 
cleanup approach is best suited to the site. In some cases, EPA may 
conduct a short-term cleanup known as a removal action or otherwise 
delay selection of a long-term cleanup approach. 

Approaches Identified by EPA for Conducting Long-Term Cleanups at 
Sites Eligible for the NPL: 

EPA may choose among several approaches to address sites with a 
relative threat to human health and the environment that is 
sufficiently severe to make them eligible for listing on the NPL. For 
long-term cleanups, EPA can retain oversight of sites under the 
Superfund program or defer the oversight of sites to other approaches, 
as shown in figure 2. 

Figure 2: Approaches Identified by EPA to Clean Up Sites Eligible for 
the NPL: 

[Refer to PDF for image: flow-chart] 

Sites eligible for the National Priorities List (NPL) proceed to a 
cleanup approach: 

Superfund program approaches: 
* NPL; 
* Superfund Alternative; 
* Other. 

Approaches outside the Superfund Program: 
* Deferral to EPA's Resource Conservation and Recovery Act program; 
* Deferral to the Nuclear Regulatory Commission; 
* Formal state deferral; 
* Other Cleanup Activity (OCA) deferral to one of the entities below: 
- States; 
- Federal agencies; 
- Private parties; 
- Tribes. 

Source: GAO analysis of EPA information. 

[End of figure] 

Superfund Program Approaches for Conducting Long-Term Cleanups: 

Under its Superfund program, EPA conducts long-term cleanups using 
three approaches. The first and most common approach under the 
Superfund program involves listing a site on the NPL. To do so, EPA 
first proposes the site for listing on the NPL in the Federal 
Register. EPA then accepts public comments on the proposal and 
responds to the comments in a second and final Federal Register 
listing announcement of the site; then the agency may place on the NPL 
those sites that continue to meet the requirements for listing. The 
second approach that EPA may use under the Superfund program is the SA 
approach, which began informally in the 1990s whereby some EPA regions 
negotiated site cleanup agreements with PRPs for sites that PRPs, 
states, or local government officials and communities did not want to 
have listed on the NPL. To promote consistency across regions, EPA 
issued guidance in 2002 formalizing the SA approach, which it 
subsequently updated in 2004 and 2012.[Footnote 11] According to EPA's 
guidance, to qualify for the SA approach, (1) a site's contamination 
must make it eligible for listing on the NPL; (2) EPA must anticipate 
a long-term cleanup at the site; and (3) there must be a willing, 
capable PRP who will negotiate and sign an agreement with EPA to 
perform the investigation or cleanup. The third approach EPA can use 
for long-term cleanup of sites is to address sites under the Superfund 
program but not list them on the NPL or address them through the SA 
approach.[Footnote 12] These "Other" sites under the Superfund program 
can vary widely and include, among others, some sites with cleanup 
agreements that preceded the SA approach. Under these older 
agreements, for which there was no guidance at the time they were 
negotiated, EPA agreed not to list the site on the NPL, and the PRPs 
agreed to conduct the cleanup, according to EPA officials. 

Irrespective of the approach chosen, all sites under the Superfund 
program approaches follow the same general phases for long-term 
cleanup, as shown in figure 3, and EPA officials oversee the cleanup 
at all of these Superfund sites. 

Figure 3: Long-Term Cleanup Process at Sites under the Superfund 
Program: 

[Refer to PDF for image: process illustration] 

Phase: Preliminary assessment. 

Phase: Site inspection. 

Phase: Remedial investigation. 

Phase: Feasibility study; 
Milestone: Record of Decision. 

Phase: remedial design. 

Phase: Remedial action; 
Milestone: Construction complete. 

Phase: Postconstruction completion[A]. 

Source: GAO analysis of EPA information. 

Note: Phases of the long-term cleanup process may overlap, and 
multiple phases may be concurrently under way at a site. 

[A] Postconstruction completion includes activities such as operation 
and maintenance, long-term response actions, and 5-year reviews, which 
ensure that Superfund cleanup actions provide for the long-term 
protection of human health and the environment. 

[End of figure] 

After the initial phases of the long-term cleanup process, EPA or a 
PRP conducts a two-part study of the site: (1) a remedial 
investigation to further characterize site conditions and assess the 
risks to human health and the environment, among other actions, and 
(2) a feasibility study to evaluate various cleanup options to address 
the problems identified in the remedial investigation. At the 
conclusion of these studies, EPA selects a remedy for addressing the 
site's contamination and develops a cost estimate for implementing the 
remedy; both of these are included in a record of decision. According 
to EPA officials, the level of cleanup depends on site-specific 
conditions, not the particular approach selected. EPA or a PRP then 
develops the method of implementation for the selected remedy during 
the remedial design phase and implements it during the remedial action 
phase, when actual cleanup of the site occurs. Multiple cleanup 
activities can occur within a given phase at the same or different 
operable units at one site. For example, one remedial action at an 
operable unit may address soil contamination, while another remedial 
action at the same operable unit may address groundwater 
contamination. When EPA or a PRP finishes the cleanup remedy at a 
site, all immediate threats have been addressed, and all long-term 
threats are under control, EPA generally considers the site to be 
"construction complete." For sites listed on the NPL, when EPA, in 
consultation with the state, determines that no further site response 
is appropriate, the agency may delete the site from the NPL. EPA 
reports achievements at NPL sites, including completion of some phases 
of the cleanup process, as part of the agency's implementation of 
provisions under the Government Performance and Results Act of 1993 
(GPRA).[Footnote 13] The act requires federal agencies to develop 
strategic plans with outcome oriented agency goals and objectives, 
performance measures to track the progress made toward achieving 
goals, annual goals linked to achieving the long-term goals, and 
annual reports on the results achieved. EPA does not report publicly 
the same achievements for sites that are not on the NPL. 

CERCLA provides EPA with authority to enter into agreements with PRPs 
to conduct cleanup actions at sites;[Footnote 14] this authority is 
relevant to all three approaches where EPA maintains oversight under 
the Superfund program. Model agreements for different phases of 
cleanup with standard provisions are to serve as the basis for 
negotiations and for the agency's legal documents. According to EPA 
officials, the agency typically uses legal documents known as 
"administrative orders on consent"--which do not require court 
approval--to record the agreements between EPA and PRPs for conducting 
remedial investigation and feasibility studies. An EPA agreement with 
a private party for conducting a remedial action generally takes the 
form of a "consent decree," which must be approved by a court. 
[Footnote 15] Although agreements under the SA approach generally 
follow these model agreements, EPA guidance states that SA agreements 
are to include specific provisions, depending on the phase of cleanup 
to which the agreement applies. These provisions are intended to 
ensure equivalency between the SA and NPL approaches. In addition to 
negotiating agreements, EPA has authority to issue enforcement orders, 
such as "unilateral administrative orders," or to coordinate with the 
Department of Justice in seeking an injunction to require PRPs to 
conduct cleanup.[Footnote 16] 

Approaches Outside of the Superfund Program for Conducting Long-Term 
Cleanups: 

As an alternative to addressing a site under the Superfund program, 
EPA may defer oversight of the cleanup of a site eligible for the NPL 
to other cleanup approaches, including federal and state cleanup 
programs. For example, EPA may defer a site from its Superfund program 
to its Resource Conservation and Recovery Act (RCRA) program. Congress 
passed RCRA in 1976, establishing requirements, as well as giving EPA 
regulatory authority, for the generation, transportation, treatment, 
storage, and disposal of hazardous waste.[Footnote 17] While CERCLA 
focuses on cleanup of sites where hazardous substances have been 
released, including inactive and abandoned hazardous waste sites, RCRA 
generally focuses on facilities currently generating, treating, 
storing, and disposing of hazardous waste--hazardous materials that 
are destined for disposal or recycling. RCRA authorizes EPA to issue 
administrative cleanup orders where an imminent and substantial danger 
to health and the environment may exist. At a given site, certain 
authorities of CERCLA and RCRA may be applicable to a cleanup. EPA 
also may defer certain sites to the Nuclear Regulatory Commission 
(NRC). The NRC licenses commercial nuclear facilities, including power 
reactors, and regulates and oversees their safe operation, including 
the decommissioning and decontamination associated with shutting down 
a licensed reactor. In what are known as "formal state deferrals," EPA 
may also defer sites to states or other entities, such as federally 
recognized tribes where applicable conditions are met. According to 
EPA guidance on this approach, the EPA region and the state cleanup 
program should enter into a memorandum of agreement certifying that 
the state has the necessary authority and capability to adequately 
supervise the PRP's cleanup actions, among other things. The state 
will then oversee cleanup actions conducted and funded by the PRPs at 
the site. The quality of these cleanup actions should be substantially 
similar to a cleanup required under CERCLA authorities, according to 
EPA guidance on this approach. Under this formal state deferral 
approach, the EPA region still negotiates the level of oversight 
appropriate for the particular site. 

EPA may also defer oversight of the long-term cleanup of a site 
eligible for the NPL through the Other Cleanup Activity (OCA) 
approach. OCA deferrals go to one of four types of entities (described 
below): states, federal agencies, tribes, or private parties. 

* OCA deferral to a state places a site under that particular state's 
environmental regulations, as opposed to CERCLA authorities. In 
contrast to formal state deferrals, the OCA deferral to a state 
involves no formal EPA oversight other than periodic discussions 
between EPA regional officials and state officials. Since 2012, EPA 
guidance has indicated regions should have these discussions. 

* OCA deferral to federal agencies places a site under that particular 
federal agency's oversight and authorities, according to EPA. Certain 
federal agencies, such as the Department of Defense, have 
responsibility and authority for some or all cleanups at their 
facilities.[Footnote 18] EPA assigns a status of "Other Cleanup 
Activity: Federal Facility Lead" to federal facilities that EPA tracks 
in its CERCLIS database and are being cleaned up outside of the NPL 
approach[Footnote 19] (these sites are eligible for listing but are 
not listed on the NPL). EPA periodically checks in with other federal 
agencies on the status of cleanup work at these sites. 

* OCA deferral to a tribe places the site under that tribe's 
environmental regulations. EPA periodically checks in with tribal 
regulators on the status of cleanup work at these sites. 

* OCA deferral to private parties applies to certain sites where the 
cleanup is conducted by a private party. 

EPA Defers Oversight of a Majority of Sites Eligible for the NPL to 
Approaches Outside of the Superfund Program: 

EPA most commonly addresses the cleanup of sites eligible for the NPL 
by "deferring" oversight to approaches outside of the Superfund 
program. EPA regions select the cleanup approach and defer oversight 
of more than half the sites eligible for the NPL to approaches outside 
of the Superfund program, primarily through OCA deferrals. Though OCA 
deferrals include the majority of NPL eligible sites, EPA's guidance 
on this approach is less detailed than guidance on other approaches. 

EPA Regions Exercise Discretion in Selecting the Cleanup Approach for 
Sites: 

EPA provides regions with discretion in selecting the cleanup approach 
for a given hazardous waste site. According to the Superfund Program 
Implementation Manual--which lists EPA's Superfund program management 
priorities, procedures, and practices--each region is to select an 
appropriate cleanup approach after determining a site is eligible for 
the NPL. Officials in all 10 regions said that when they select 
cleanup approaches they attempt to use the most appropriate cleanup 
approach for a given site. For example, complex sites, such as 
contaminated waterways, may be more suited to the NPL approach than to 
deferral to a state cleanup program because EPA typically has more 
resources to oversee and manage such complex cleanups. Officials in 
four regions noted that states will sometimes request the NPL approach 
for large or complex sites. 

EPA regions can establish their own processes for selecting a cleanup 
approach for a given site. Three of the 10 regions have some type of 
regional guidance related to their decision-making process. For 
example, Region 7 has guidance for its regional decision team that 
outlines the stakeholders within the region who will participate, when 
the team will meet, and how decisions are to be made at the meeting. 
Region 10 has guidance that focuses on how the region will prioritize 
sites that are eligible for the NPL. All of the regions may consult 
with relevant stakeholders across EPA programs about a given site, 
whether the regions have written guidance or not. These stakeholders 
might include staff from the office of regional counsel or the removal 
program. Five of the 10 regions use regional decision teams to 
evaluate sites that have been found eligible for the NPL and select 
which approach should be used to clean up the site. The other 5 
regions do not use regional decision teams, opting instead for more 
informal decision-making processes or meetings on an as-needed basis. 
For example, in Region 5 there is a practice of coordinating between 
the region's long-term cleanup and removal programs on sites that may 
be of interest to both programs. 

EPA officials said the regions consider many potentially relevant 
factors to select the appropriate approach for each site. The major 
factors influencing regional officials' choice of cleanup approach at 
a site include the preferences of the state regarding how the site 
will be addressed and the existence of a PRP that is willing to and 
capable of addressing the site. Specifically, officials in all 10 
regions highlighted state preference as a factor they consider. State 
preference can be particularly important because EPA has a policy of 
obtaining state concurrence before listing a site on the NPL. 
According to Region 5 officials, if a state opposes an NPL listing, 
they will typically give preference to other approaches, such as an 
OCA deferral to the state or the SA approach. In addition to state 
preference, officials in 9 of 10 regions said that the existence of a 
willing and capable PRP can be a factor in determining the cleanup 
approach. For example, a willing and capable PRP is necessary for the 
SA approach, which requires the PRP to conduct the cleanup under an 
agreement with EPA. EPA officials in one region said that the 
existence of a PRP can also be important for cleanups under state 
cleanup programs because states can have very limited funding to 
conduct cleanups on their own. State environmental officials from four 
states we contacted confirmed that they had limited or, in some cases, 
no state funding to conduct their own long-term cleanups. 

Regional officials also identified other factors that can sometimes 
influence what cleanup approach the region will select. For example, 
officials in Region 5 noted that if the contamination presents an 
immediate threat to health and safety, they may use the Superfund 
removal program, which is more suited to a quick response than long-
term cleanup approaches. Depending on the circumstances at a site, the 
removal program may be sufficient to deal with all of the 
contamination, or the site may need to be referred to a long-term 
cleanup approach for further work. Regional officials can also 
consider other relevant legal authorities that could apply to a site, 
such as RCRA. When a site is eligible for cleanup under both RCRA and 
Superfund, EPA policy provides that the agency generally will defer 
the site to the RCRA program for cleanup.[Footnote 20] 

EPA Defers Oversight of More Than Half the Sites Eligible for the NPL 
to Approaches Outside of the Superfund Program: 

Among the 3,402 sites reported to the Superfund program in CERCLIS 
that EPA has identified as having contamination making them eligible 
for the NPL,[Footnote 21] EPA deferred 1,984 sites to cleanup 
approaches outside of the Superfund program (see figure 4). Sites 
under the Superfund program make up the 1,418 sites that remain, with 
the vast majority of those sites being addressed through the NPL. 

Figure 4: Current Sites Identified as Eligible for the NPL Using 
Different Long-Term Cleanup Approaches, as of December 2012: 

[Refer to PDF for image: illustrated pie-chart] 

Sites under Superfund program approaches: 42%: 
National Priorities List (1,313); 
Superfund Alternatives (67); 
Other (38). 

Sites under approaches outside of the Superfund program: 58%: 
Other Cleanup Activity deferrals (1,766): 
* States (1,585); 
* Tribes (1); 
* Private parties (77); 
* Federal agencies (103); 
Deferrals to EPA's Resource Conservation and Recovery Act program 
(193); 
Formal state deferrals (21); 
Deferrals to the Nuclear Regulatory Commission (4). 

Source: GAO analysis of EPA CERCLIS data. 

Notes: Sites listed above only include sites that are currently active 
in CERCLIS and do not include sites that have been archived by EPA or 
sites where no further site response is required, such as sites 
deleted from the NPL. 

[End of figure] 

EPA addresses more sites eligible for the NPL through the OCA deferral 
approach than any other cleanup approach: 1,766 of the 3,402 sites (52 
percent). Moreover, because EPA deferred most of these 1,766 sites to 
states, OCA deferrals to states account for about 47 percent of all 
identified eligible sites.[Footnote 22] EPA regions' use of OCA 
deferrals to states ranges widely, from 7 sites in each of three 
regions (6, 7, and 8) to 470 sites in Region 1 (see appendix II for a 
breakdown of cleanup approaches by region). According to officials in 
Region 1, states in the region have mature environmental programs 
willing and capable of overseeing many sites, which makes the OCA 
deferral to states well suited to that region. In contrast, officials 
we spoke to in some regions noted that they needed to consider states' 
capacity to oversee a site before using the OCA deferral to states. 
Nine states have no OCA deferrals, and other states oversee hundreds 
of these sites, with the most in Massachusetts (247 sites), New Jersey 
(221), and California (180). Environmental officials in several states 
we contacted confirmed that states' use of and experience with OCA 
deferrals can differ substantially. One state official noted that 
these differences are likely related to how industrialized a state may 
be and the extent of cleanup programs in a given state. OCA deferrals 
to federal agencies, private parties, and American Indian tribes 
account for an additional 181 sites. OCA deferrals to federal agencies 
primarily involve military sites: 76 percent of these deferrals were 
to the Army, Navy, or Air Force. In addition, a majority of OCA 
deferrals to private parties come from Florida where, on the basis of 
a state law, PRPs can conduct cleanup without any formal agreement or 
order from the state, according to Region 4 officials. PRPs conducting 
such cleanups must submit regular reports to the state on their 
progress, and the state reserves the right to take the PRP to court, 
if necessary. 

EPA currently addresses 1,418 sites (42 percent of those identified as 
eligible for listing on the NPL) through approaches under the 
Superfund program--most commonly, through listing the sites on the 
NPL. Specifically, sites listed on the NPL account for 1,313 sites, 
over 90 percent of sites under the Superfund program.[Footnote 23] 
According to officials in one region, EPA has access to more resources 
than states and typically addresses sites that require greater or more 
specialized resources through the NPL approach. For example, regional 
officials noted, states face different limitations that can prevent 
them from pursuing cleanup under their programs including: technical 
capacity, legal resources, and financial resources. In addition, EPA 
officials in four regions noted examples where a state environmental 
program requested that the Superfund program pursue NPL listing 
because the state was having trouble getting a PRP to cooperate or the 
PRP went bankrupt. 

In addition to listing sites on the NPL, EPA also oversees the long-
term cleanup of sites through two other approaches under the Superfund 
program. First, the Superfund program currently oversees 67 sites 
under the SA approach.[Footnote 24] Second, EPA oversees at least 38 
other sites with long-term cleanups under the Superfund program for 
which EPA has no documented definition and no consistently applied 
method of counting. EPA officials provided different estimates of the 
number of such sites. One EPA official provided a method to identify 
these sites based on a code in CERCLIS, which resulted in the 38 sites 
listed above. However, another EPA official provided us a list of 35 
such sites that had reached the remedial action phase. Of these 35 
sites, 12 matched the 38 sites identified by the code in CERCLIS. In 
addition, 16 of the sites on the list of 35 had a code of "status 
undetermined" or had no code at all. EPA regional officials also 
identified other specific sites under the Superfund program, but some 
of those sites could not be identified by the code in CERCLIS, were 
not on the list of 35, or had no code at all. As of December 2012, 270 
sites had the "status undetermined" code, and 101 had no such code in 
CERCLIS, making it impossible to determine the exact number of sites 
that EPA oversees under the Superfund program that are not being 
addressed under either the NPL or SA approaches. Tracking of these 
sites is discussed later in this report. 

EPA addressed the remaining sites eligible for the NPL through 
different deferral approaches, primarily through deferrals to the RCRA 
program. Specifically, deferrals to the RCRA program account for 193 
of these remaining sites (89 percent).[Footnote 25] Aside from these 
deferrals to the RCRA program and a few deferrals to NRC, EPA deferred 
21 sites to state programs using the formal state deferral approach in 
4 of its 10 regions. EPA officials said that the OCA deferral to 
states approach has largely replaced the formal state deferral 
approach, and EPA does not anticipate using the formal deferral 
approach much in the future. 

EPA Has Less Detailed Guidance for OCA Deferrals Than for Other Less 
Commonly Used Deferral Approaches: 

As discussed above, EPA addresses more sites through OCA deferrals 
than any other approach but has less guidance to define this approach 
or how deferral decisions should be documented than for its other 
deferral approaches. Unlike OCA deferrals, EPA has guidance or other 
documents outlining the process for deferrals to the RCRA program, 
deferrals to the NRC, and formal state deferrals. These documents 
clearly define or provide mechanisms to define the roles of the 
Superfund program and the entity that will conduct oversight at the 
site. For example, guidance for the formal state deferral approach 
specifies that the EPA region and the state should enter into a 
memorandum of agreement in which they clarify mutual expectations for 
their interaction and each party's responsibilities at deferred sites. 
[Footnote 26] After the deferral, the region continues to review the 
state's progress and conduct any other activities required by its 
individual agreements with the state in each case.[Footnote 27] 

In contrast, EPA has not issued guidance focused on OCA deferrals that 
clearly defines the different types of OCA deferrals or what detail 
would be sufficient or appropriate to support its decisions at these 
sites. Instead, EPA describes OCA deferrals in the Superfund Program 
Implementation Manual (which is updated annually). EPA recently added 
to its instructions in the manual regarding sites with OCA deferrals. 
Specifically, in its 2012 version of the manual, EPA added more 
language explaining that there is to be no continuous and substantive 
involvement on EPA's part while cleanup work is ongoing at OCA 
deferral sites. In addition, in this version of the manual, EPA added 
an instruction for regions to check on the status of OCA sites 
periodically. Officials in EPA regions noted that they use different 
approaches for tracking OCA sites; for example, for an OCA deferral to 
states, EPA regions' tracking activities range from checking state 
websites to meeting with states to receive status updates every 3 
months. Officials in some regions noted that they will need to modify 
their processes to meet this new instruction. 

Even with EPA's additions to the manual, the available instruction 
does not clearly define each type of OCA deferral, particularly OCA 
deferrals to private parties, which has resulted in inconsistent 
identification of those OCA deferrals by different regions. While the 
manual defines OCA deferrals generally, it does not define each type 
of OCA deferral. When asked to define OCA deferrals to private 
parties, Superfund program officials in headquarters referred us to 
EPA regional officials for more information, and officials in 6 of 10 
EPA regions were unsure about how to define OCA deferrals to private 
parties or how they should be used.[Footnote 28] Moreover, officials 
in another 6 regions confirmed that some sites identified as OCA 
deferrals to private parties in CERCLIS should have been identified as 
OCA deferrals to states. Without clearer guidance defining the 
different OCA deferrals, EPA cannot be reasonably assured that it is 
consistently tracking its OCA deferral sites in CERCLIS, which can 
make it difficult to identify what entity is responsible for 
conducting oversight at the site. 

In addition, the manual instructs regions to track OCA deferrals and 
completion of cleanups at these sites but does not clearly specify the 
documentation required to support these actions. The manual provides 
overarching program management priorities, procedures, and practices 
for the Superfund program. For OCA deferrals, the manual explains how 
EPA regions should identify the deferral date of an OCA site and the 
date of completion of cleanup at that site,[Footnote 29] but it 
provides little detail on what type of documentation would be 
acceptable to support these determinations. For example, according to 
the manual, the deferral date of a site entered into CERCLIS is 
supposed to be "supported by existing documentation," described as 
"documentation between EPA and the non-EPA party leading the cleanup," 
with no further detail of what documentation would be appropriate or 
sufficient. In addition, the instruction for entering the date for 
completion of the cleanup refers to required documentation, without 
further clarification about what documentation is needed from the 
entities conducting oversight. Regional officials told us that, in 
practice, the amount and type of documentation regions collect to 
support OCA deferrals covers a broad range, including no written 
documentation, an e-mail from a state official, letters from state 
officials attesting to the cleanup, or a copy of the legal order or 
agreement between the state and PRP. Similarly, regions relied on 
different forms of documentation, including various e-mails, letters, 
or reports from state officials to document the completion of cleanup 
at OCA deferral sites. Officials in three regions reported that there 
was no consistent standard for documentation within their region. 
Moreover, Region 9 officials noted that the region had not tracked the 
completion of cleanups at OCA deferrals in CERCLIS in the past and may 
have no documentation for some of its older OCA deferral sites. 
Without guidance that details the documentation needed to support 
regions' OCA deferral decisions, EPA cannot be reasonably assured that 
its regions' documentation will be appropriate or sufficient to verify 
that these sites have been deferred or have completed cleanup. EPA 
officials noted they were working on additional guidance for OCA 
deferrals. However, these officials said that development of the 
guidance was in the planning stage; therefore, a draft of this 
guidance, detailed information on what will be included in the 
guidance, or a planned issuance date for the guidance, were not yet 
available. 

EPA provides the least detailed guidance for the small number of sites 
that are undergoing long-term cleanup under the Superfund program 
outside of the SA and NPL approaches. Such sites do not have specific 
guidance at the program level, regional level, or a section in the 
Superfund Program Implementation Manual describing how they should be 
defined or tracked. In contrast, EPA has developed instructions in the 
manual for how to track sites cleaned up under the SA approach. EPA 
also has guidance for the NPL approach, such as how the agency should 
propose, list, and delete sites from the NPL. EPA officials noted that 
sites that are cleaned up under the other Superfund program approach 
often involve unique situations, making it difficult to establish any 
guidance that would cover all possible situations. For example, one of 
these sites is using a hybrid approach under both RCRA and CERCLA 
authorities, according to an EPA official. However, in 10 cases, 
regional officials described these sites as standard cleanups under 
CERCLA authority that used standard procedures. While there are unique 
and standard cases among sites being cleaned up under the other 
Superfund program approach (i.e., outside of the SA and NPL 
approaches), EPA officials could not provide a reliable estimate of 
these other sites because the agency has no consistently applied 
method for counting them. Without a method to identify and track such 
sites, EPA headquarters has no way to determine the extent to which 
regions use this approach or evaluate regions' use of this approach. 
As a result, it will be difficult for EPA headquarters to hold regions 
accountable for using the approach. 

Processes for Implementing the SA and NPL Approaches Differ in a Few 
Significant Ways: 

The processes for implementing the SA and NPL approaches have 
similarities, but also several differences, some of which EPA has 
accounted for through specific provisions in its agreements with PRPs 
at SA agreement sites. However, some sites may not benefit from EPA's 
efforts to account for these differences. Furthermore, the agency's 
tracking and reporting of SA agreement sites differs significantly 
from its tracking and reporting of NPL sites. Using the SA approach at 
sites has certain potential advantages for EPA and some PRPs and 
states, but communities' views on this approach are mixed. 

The SA and NPL Processes Are Similar in Many Ways and EPA Has 
Accounted for Some Differences Between Them: 

The processes for implementing the SA and NPL approaches have many 
similarities. According to the agency's SA guidance, at its SA 
agreement sites, EPA is to generally act in accordance with the 
practices normally followed at sites listed on the NPL. For example, 
according to EPA guidance, SA agreement and NPL sites should follow 
the same investigation and cleanup processes, including the phases and 
milestones of long-term cleanups shown earlier in figure 3. EPA 
regions should also use the same response techniques, standards, and 
guidance for SA agreement sites as they do for NPL sites. According to 
EPA's guidance, SA agreements should eventually achieve cleanup levels 
that are comparable to those required at NPL sites. EPA regions should 
also take steps to ensure equivalency between the SA and NPL 
approaches in the absence of NPL listing. 

Despite these similarities, there are certain differences in the 
overall processes and EPA's authority under the NPL and SA approaches. 
Through specific provisions in its SA agreements with PRPs, EPA has 
sought to make the two approaches comparable by accounting for the 
following four key differences: 

* First, EPA has the authority to pay for remedial actions only at 
sites listed on the NPL.[Footnote 30] To account for this difference, 
SA agreements include a provision to help ensure cleanups are not 
delayed by a loss of funding if the PRP ceases work during the 
remedial action phase of cleanup. Specifically, this provision 
requires the PRP to obtain a readily available source of funds that 
the agency can use if it needs to take over the cleanup work. EPA can 
use those funds to continue the work while the agency lists the site 
on the NPL, if necessary. 

* Second, EPA is authorized to provide technical assistance grants 
that help communities participate in decision making only at sites 
that are listed or proposed for listing on the NPL. An initial grant 
of up to $50,000 is available to qualified community groups so they 
can contract with independent advisors to help the community 
understand technical information about the site. EPA includes a 
provision in SA agreements to help ensure that a community's 
opportunity to receive technical assistance at an SA agreement site is 
comparable to that at an NPL site. This provision requires the PRPs, 
with EPA oversight, to administer and fund a technical assistance 
plan, under which a qualified community group can receive up to 
$50,000 for the same purposes as a technical assistance grant from EPA. 

* Third, if a PRP were to clean up an SA agreement site to the extent 
that it no longer scored at least 28.50 on the Hazard Ranking System, 
according to EPA, it might lose the option of listing the site on the 
NPL, a concern that is not present when a site is listed on the NPL. 
To prevent this, SA agreements state that the PRP will not challenge 
listing the site on the NPL if a partial cleanup of the site results 
in changed site conditions. EPA officials noted that this provision 
gives the agency assurance that it can step in and clean up the site 
under the NPL approach if the PRP were to default on the SA agreement. 

* Fourth, CERCLA states that an action for natural resource damages 
(NRD) at NPL sites must start within 3 years after completion of the 
remedial action.[Footnote 31] This period is longer than the general 
statute of limitations for NRD claims, which states that an action 
must start within 3 years after the discovery of the loss and its 
connection with the contamination. SA agreements contain a provision 
that clarifies that the longer statute of limitations for NPL sites 
also applies to SA agreement sites. 

Even with EPA's efforts to achieve equivalence of SA agreement and NPL 
sites through these provisions, some sites may not benefit from these 
efforts because EPA regions have entered into agreements with PRPs at 
sites that they said were likely eligible for the SA approach without 
following the SA guidance. Agreements at such sites may not, for 
example, ensure that a community has access to funds to pay for 
technical assistance or that remedial action can continue if a PRP 
stops cooperating. Officials from some EPA regions told us they have 
continued to enter into agreements with PRPs since 2002 without 
following the SA guidance. We identified six sites where this has 
occurred as follows: 

* In Region 7, officials entered into an agreement with a PRP to 
conduct remedial design and remedial action at a site. Regional 
officials stated that the SA approach, which can be suggested for a 
site by the PRP or the region, never came up during their discussions 
with the PRP. 

* In Region 10, officials stated that the agreements they had entered 
into with PRPs at five sites might qualify for the SA approach but 
that, at the time they entered into the agreements, the officials had 
not focused on whether the agreements met the SA criteria; rather, 
they were focused on obtaining enforceable agreements. 

According to EPA headquarters officials, if regions are going to 
conduct a long-term cleanup under the Superfund program at a site, but 
not list it on the NPL, the agency prefers regions to use the SA 
approach. EPA headquarters officials said that they believed this 
preference was implicit in the agency's SA guidance and stated they 
discussed this preference with regional officials at periodic 
meetings; however, they also acknowledged that this preference is not 
stated explicitly anywhere in guidance for the regions. If regions 
continue to enter into agreements for some sites without following the 
SA guidance, these sites may be denied some of the advantages built 
into the SA agreements to ensure that the cleanups will be comparable 
to those under the NPL approach. 

EPA's Tracking and Reporting of Certain Aspects of the Process under 
the SA Approach Differ Significantly from That under the NPL Approach: 

Some differences remain between the way EPA tracks sites under the SA 
and NPL approaches. In CERCLIS, EPA tracks sites' status in relation 
to the NPL regardless of any changes in cleanup approach. 
Specifically, sites that have been proposed for listing on the NPL, 
are currently on the NPL, have been deleted from the NPL, or have been 
removed from proposal can always be identified as such in CERCLIS, 
which allows EPA to accurately identify sites that are or have been on 
the NPL. In contrast, EPA cannot similarly track an SA agreement site 
as such if it is subsequently listed on the NPL.[Footnote 32] 
Specifically, EPA currently tracks SA agreement sites through a single 
database code identifying only that a site has an SA agreement, and 
the identifying code is not maintained in the database if the site is 
later added to the NPL. The agency has not clarified in its guidance 
when to leave this SA identifying code in place, and when to remove 
it, even though the EPA IG recommended in a 2007 report that EPA 
develop specific instructions on when to use the SA designation and 
update the Superfund Program Implementation Manual (which is updated 
annually) to incorporate these instructions.[Footnote 33] According to 
the IG report, these instructions should specify that the SA code 
should not be removed even if the site is cleaned up or proposed for 
the NPL, so that controls over documentation of sites with SA 
agreements can be maintained. As the EPA IG pointed out, absence of 
guidance can result in poor quality data on the SA universe. While EPA 
indicated in 2010 that it would implement this recommendation, 
[Footnote 34] the 2012 manual does not include any instructions about 
maintaining the SA code. Because EPA has not implemented the IG's 
recommendation, the manner in which the agency tracks the identity of 
SA agreement sites in CERCLIS is incomplete. For example, while an EPA 
website identifies all sites that have or have had SA agreements, 
three sites that had SA agreements and were later added to the NPL 
cannot be identified in CERCLIS as having had SA agreements. As a 
result, all sites that have had SA agreements are not identifiable in 
CERCLIS, which may hamper EPA's ability to effectively manage long-
term cleanups and track outcomes at SA agreement sites. 

Furthermore, the standards for specifying what documentation is 
sufficient to support the Hazard Ranking System score of SA agreement 
sites are less clear than those for NPL sites. When sites are proposed 
for listing on the NPL, EPA procedure requires they have a Hazard 
Ranking System documentation record--a specific document that includes 
detailed justification for the Hazard Ranking System score. In 
contrast, both the 2004 and 2012 SA guidance state that EPA should 
have "adequate documentation" supporting a Hazard Ranking System score 
of 28.50 or higher but do not define what is meant by "adequate" 
documentation or provide criteria for assessing adequacy. The guidance 
documents specify that regions may rely on a draft Hazard Ranking 
System documentation record or "other adequate documentation," but do 
not provide an explanation of what other documentation might be 
adequate. EPA headquarters officials told us that documentation of a 
preliminary calculation of the Hazard Ranking System score during the 
initial assessment phases would qualify as adequate, and said that 
this has been discussed with regional officials during periodic 
meetings. EPA officials acknowledged, however, that this 
interpretation of the guidance has not been included in any written 
guidance to the regions. As the EPA IG pointed out in its 2007 report, 
consistent and reliable documentation of Hazard Ranking System scores 
at SA agreement sites is an internal control to ensure compliance with 
the SA guidance and approach.[Footnote 35] Under the federal standards 
of internal control, agencies are to clearly put in writing (i.e., in 
management directives, administrative policies, or operating manuals) 
internal controls, such as this interpretation of the guidance, and 
have them readily available for examination.[Footnote 36] Without more 
specific written guidance, EPA regional officials may not develop 
adequate documentation of Hazard Ranking System scores at SA agreement 
sites. 

In addition to the differences in its tracking, EPA has not reported 
the agency's performance on the progress of cleanup at SA agreement 
sites as it has for NPL sites. EPA reports achievements at NPL sites, 
including completion of some phases of the cleanup process, as part of 
the agency's implementation of provisions under GPRA, which generally 
aims to hold federal agencies accountable for using resources wisely 
and achieving program results. Two of the Superfund program's three 
GPRA performance measures--sites where human exposure is under control 
and sites that are ready for their anticipated use--refer only to NPL 
sites. One additional performance measure tracks the completion of the 
initial assessment phases, which generally precede EPA's decision 
about which cleanup approach to use at a site, including the SA or NPL 
approach. EPA's Office of Solid Waste and Emergency Response, which 
manages the Superfund program, reports these performance measures for 
NPL sites in several annual reports available on EPA's 
website.[Footnote 37] However, EPA does not include in these reports 
the cleanup milestones reached at SA agreement sites, such as how many 
SA sites have human exposure under control. The EPA IG recommended in 
2007 that EPA track and report the same GPRA performance measures at 
SA agreement sites as it does at NPL sites.[Footnote 38] As the IG 
reported, by measuring and tracking all performance measures at SA 
agreement sites, EPA could demonstrate the outcomes of the Superfund 
program's work and provide an incentive to regions by more thoroughly 
accounting for their performance. In 2010, EPA indicated that it would 
implement the IG's 2007 recommendation to track and report all 
Superfund GPRA performance measures at SA agreement sites using an 
annual report.[Footnote 39] EPA officials noted that the agency has 
begun tracking Superfund performance measures for SA agreement sites, 
but they acknowledged that EPA is not reporting these results 
publicly. Until the agency reports performance information on the 
progress of cleanup at SA agreement sites as it does for NPL sites, 
EPA is not providing the public and Congress with a full picture of SA 
agreement sites. Without such information, Congress lacks complete 
information on the progress of the Superfund program to inform its 
legislative actions, including appropriations. 

The SA Approach Has Potential Advantages for EPA, Some PRPs, and 
States, but Communities' Views on its Benefits Are Mixed: 

Using the SA or the NPL approach can have advantages or disadvantages 
for the parties involved, including EPA, PRPs, states, and 
communities. Specifically, using the SA approach generally allows EPA 
to avoid at least some of the cost and time associated with listing a 
site on the NPL. For example, NPL listing requires preparation of a 
Hazard Ranking System documentation record, which is not required for 
sites with SA agreements.[Footnote 40] EPA officials estimated each 
such record costs an average of about $65,000 to prepare. In addition, 
when EPA decides to propose a site for listing on the NPL, the agency 
sometimes conducts an expanded site inspection if further information 
is necessary to document a Hazard Ranking System score. EPA officials 
estimated this step costs about $92,000 on average. In addition, to 
list a site on the NPL, EPA has to work through the formal listing 
process, including issuing notices in the Federal Register with a 
public comment period. This process takes time to complete, which may 
affect the progress at the site. In Region 3, EPA officials stated 
that the volume of comments received on a particular site proposed for 
the NPL, in addition to the likelihood of litigation from one or more 
parties if the site were finalized on the NPL, led the region to 
address the site through the SA approach. 

Some EPA regions have seen the advantages of using the SA approach 
more than others. As shown in figure 5, of the 67 SA agreement sites, 
57 sites, or 85 percent, are in EPA Regions 4 and 5. 

Figure 5: Number of SA Agreement Sites by EPA Region, as of December 
2012: 

[Refer to PDF for image: illustrated U.S. map] 

Region 1: 0; Did not use the SA approach; 
Region 2: 0; Did not use the SA approach; 
Region 3: 2; Used the SA approach infrequently; 
Region 4: 23; Used the Superfund Alternation (SA) approach frequently; 
Region 5: 34; Used the Superfund Alternation (SA) approach frequently; 
Region 6: 1; Used the SA approach infrequently; 
Region 7: 1; Used the SA approach infrequently; 
Region 8: 1; Used the SA approach infrequently; 
Region 9: 3; Used the SA approach infrequently; 
Region 10: 2. Used the SA approach infrequently. 

Source: GAO analysis of EPA data; Map Resources (map). 

[End of figure] 

Differences in usage of the SA approach among regions relate to a 
region's specific circumstances and preferences. According to EPA 
headquarters officials, Regions 4 and 5 had early experience with SA 
agreements and may have been more comfortable in starting new ones as 
a result. These two regions have also listed many sites to the NPL 
since the SA approach was formalized in 2002. Region 4 officials told 
us that they have found that the SA approach is best suited to sites 
with one or two PRPs and no questions about the PRPs' liability or 
ability to conduct the investigation or cleanup. They said they have 
also found that it is helpful when a PRP has a financial interest in 
finishing the cleanup quickly, as in the case of a potential 
redevelopment project at a site. Other regions have used the SA 
approach in limited circumstances. For example, officials in Region 9 
described one case in which they pursued the SA approach because the 
state did not want a particular site listed on the NPL. Two regions--
Regions 1 and 2--have never used the SA approach. Officials in Region 
1 explained that few sites that have willing and capable PRPs and are 
eligible for the NPL come to the region's attention because state 
programs prefer to take on oversight of such sites. Region 2 officials 
said they did not see a reason to use the SA approach--if a site's 
contamination is severe enough, the region will propose the site to 
the NPL, unless the state is addressing the site. 

Using the SA approach allows PRPs to avoid the perceived stigma 
associated with an NPL site, according to EPA officials. Sites with SA 
agreements have to meet all of the qualifications of NPL sites, and 
thus may have contamination that is just as severe, but the potential 
stigma of NPL listing appears to influence PRPs. Officials in 7 of 10 
regions mentioned the stigma of an NPL site as a concern for PRPs. 
Concerns about this stigma may also arise when a company is to be sold 
and does not want to list an NPL site as part of its liabilities, 
according to EPA regional officials. Related to this stigma, EPA 
officials said they believed that avoiding listing on the NPL may help 
local government officials and PRPs in some cases, such as 
facilitating a site's redevelopment or its financing. Previous reports 
have also pointed to the potential stigma of NPL listing as motivation 
for pursuing a different cleanup approach.[Footnote 41] For example, 
an assessment of the effectiveness of the SA approach in Region 4 
(hereafter referred to as the Region 4 study) found that sites using 
the SA approach may have a higher potential for redevelopment than 
comparable NPL sites if avoiding this stigma increases PRPs' financing 
options and their willingness to redevelop.[Footnote 42] 

In addition, some states generally prefer that EPA not list sites on 
the NPL, according to EPA officials, which makes the SA approach more 
appealing. According to EPA policy, EPA typically obtains a state's 
concurrence before listing a site on the NPL. Officials in all 10 EPA 
regions mentioned the states' views as one of the factors they used to 
determine whether to pursue an NPL listing or other approaches. 
Moreover, officials in 4 of 10 regions said there were states in their 
region that were generally reluctant to have EPA list sites on the 
NPL. For example, Region 9 officials said two of their states 
generally do not want EPA to list sites on the NPL; specifically, one 
of these states wanted to avoid the associated stigma of having NPL 
sites in the state. 

The SA approach also has advantages and disadvantages for communities. 
According to an EPA official, it may be easier for communities to 
obtain technical assistance funds from PRPs at SA agreement sites than 
to obtain the equivalent funds from EPA at NPL sites. This official 
said obtaining funds from PRPs at SA agreement sites often involved 
the absence of a "match" requirement as well as fewer paperwork 
requirements for the communities because the technical assistance 
plans do not have to follow federal grant requirements. However, under 
the SA approach, communities have no opportunity for a formal comment 
process on EPA's selection of the SA approach itself, as they do under 
the NPL approach. Specifically, when EPA proposes a site for the NPL 
in the Federal Register, the public has 60 days to comment on the 
proposed listing. EPA then responds in writing to significant public 
comments in conjunction with the final Federal Register listing 
announcement of the site. No such opportunity exists when EPA decides 
to enter into an SA agreement at a site, although EPA provides 
numerous opportunities under the SA approach for communities to 
comment on the cleanup process. 

Communities also may have mixed reactions to the SA approach for other 
reasons as well. According to EPA officials, communities may have 
concerns about the SA approach and may require outreach from the 
agency to explain the approach. For example, at one site in Region 5, 
the region expanded its outreach efforts after some community members 
protested the use of the SA approach at the site. A regional official 
explained that some individuals in the community believed the site 
would not follow the same cleanup process as an NPL site. Some 
community members may support listing on the NPL over the SA approach 
to bring increased attention to a site, helping to ensure its cleanup. 
Other regional officials said other community members may be more open 
to the SA approach and oppose listing on the NPL for fear of its 
effect on property values. The Region 4 study confirmed that the SA 
approach is often considered advantageous by community members and 
leaders concerned about property values and stigma.[Footnote 43] 
However, this report also found that other community members require 
confirmation that the process will not result in more limited 
resources or reduced remediation compared to listing on the NPL. 

SA Agreement Sites Showed Mixed Results in Completing the Cleanup 
Process When Compared with Similar NPL Sites: 

For sites with agreements from June 2002 through December 2012, SA 
agreement sites and similar NPL sites we selected showed mixed results 
in the time needed to complete negotiations for agreements, specific 
cleanup activities, and achieving the construction completion 
milestone (see appendix I for more details on our objectives, scope, 
and methodology and appendix III for more information on our results). 
Specifically, SA agreement and NPL sites in our analysis showed mixed 
results in the average time to complete negotiations with PRPs and for 
specific cleanup activities, such as remedial investigation and 
feasibility studies, remedial designs, and remedial actions. In 
addition, a lower proportion of SA agreement sites have reached 
construction completion compared with similar NPL sites. SA agreement 
sites tend to be in earlier phases of the cleanup process because the 
SA approach began more recently than the NPL approach. 

For agreements finalized from June 2002 through December 2012 at sites 
in our analysis, SA agreement and similar NPL sites showed mixed 
results in the length of time to complete negotiations, with SA 
agreement sites taking about as long as similar NPL sites for remedial 
investigation and feasibility study negotiations and less time for 
remedial design and remedial action negotiations. EPA regional 
officials confirmed that negotiations can be faster at SA agreement 
sites because the PRPs are more cooperative. For example, Region 4 
officials highlighted one SA agreement site where the PRP pushed for a 
quicker negotiation process by turning in documents ahead of 
deadlines, unlike many other PRPs. In another case, Region 5 officials 
said they negotiated three SA agreements for remedial investigations 
and feasibility studies covering 19 sites of a similar nature with the 
same PRP. Region 5 officials noted that these negotiations were 
particularly smooth and cooperative. Moreover, the Region 4 study also 
found, based on interviews with PRPs and EPA officials, that the tone 
of SA negotiations is more productive than at NPL sites.[Footnote 44] 
However, given the relatively limited number of negotiations for both 
NPL and SA agreement sites in our analysis, the differences in the 
average length of negotiations cannot be attributed entirely to the 
type of approach used at each site. 

The SA agreement and similar NPL sites in our analysis showed mixed 
results in the length of time it took to complete specific cleanup 
activities,[Footnote 45] with SA agreement sites taking substantially 
longer for remedial investigations and feasibility studies on average 
and about the same time for remedial designs and remedial actions. 
While SA agreement sites took substantially longer on average than NPL 
sites to complete remedial investigations and feasibility studies, 
these differences do not appear to be exclusively attributable to the 
SA and NPL approaches. For example, several remedial investigations 
and feasibility studies at SA agreement sites took a long time to 
complete due to individual circumstances at the site, such as dealing 
with a proposal to sell on-site materials to a manufacturing company, 
late participation from PRPs in the process, or coordination with 
other cleanup efforts. SA agreement sites and NPL sites in our 
analysis took about the same time on average to complete remedial 
designs and remedial actions. 

A lower proportion of SA agreement sites have reached construction 
completion compared with similar NPL sites in our analysis (see figure 
6). 

Figure 6: Construction Completions at SA Sites and Similar NPL Sites: 

[Refer to PDF for image: stacked vertical bar graph] 

SA sites: 
Sites that have not achieved construction completion: 63; 
Sites that achieved construction completion: 3. 

Similar NPL sites: 
Sites that have not achieved construction completion: 47; 
Sites that achieved construction completion: 27. 

legend: 

SA sites: Sites with Superfund Alternative (SA) agreements. 

Similar NPL sites: Sites listed on the National Priorities List (NPL) 
with agreements similar to SA agreements. 

Source: GAL analysis of EPA data. 

[End of figure] 

We selected NPL sites for our analysis that had agreements put in 
place between EPA and PRPs from June 2002 through December 2012, as SA 
agreement sites do. According to EPA officials, however, because the 
SA approach began in 2002, and the NPL approach was initially 
authorized in 1980, more SA agreement sites began cleanup later than 
NPL sites and, therefore, are in earlier phases of the cleanup 
process. For example, 23 percent of the sites in our NPL comparison 
group have ongoing remedial investigations and feasibility studies--a 
phase that precedes selection of the remedy and remedial action--while 
almost 60 percent of SA agreement sites are in this phase. Since EPA 
began implementing the NPL approach over 30 years ago, there were more 
sites ready to negotiate agreements during the period of our analysis 
for later phases in the cleanup process, which lead to construction 
completions. 

Conclusions: 

EPA regional officials are responsible for choosing the appropriate 
long-term cleanup approach for sites with contamination that makes 
them eligible for the NPL. To do so, they select from among several 
approaches, including deferring responsibility for the oversight of 
site cleanup outside of the Superfund program. Of these sites deferred 
outside of the Superfund program, EPA has deferred about 1,800 sites 
through OCA deferrals--more sites than any other approach--but the 
agency has not issued guidance focused on this long-term cleanup 
approach. Instead, EPA describes OCA deferrals in the Superfund 
Program Implementation Manual, which does not clearly define each type 
of OCA deferral, particularly OCA deferrals to private parties. This 
has led to inconsistent coding of OCA deferrals in CERCLIS by 
different regions. Moreover, EPA's guidance does not specify in detail 
the documentation regions should have to support their decisions on 
OCA deferrals or completion of cleanup at these sites. As a result, 
EPA regions collected varying types and amounts of documentation--
including, in some cases, no documentation--to support OCA deferrals. 
EPA officials noted they were currently working on additional guidance 
for OCA deferrals, but they had not set an issuance date for this 
guidance. Without clearer guidance on OCA deferrals, EPA does not have 
reasonable assurance that it can consistently track its OCA deferral 
sites in CERCLIS or that its regions' documentation will be 
appropriate or sufficient to verify that these sites have been 
deferred or have completed cleanup. In addition, EPA officials could 
not provide a reliable estimate regarding the number of sites with 
long-term cleanups under the Superfund program that are being cleaned 
up through approaches other than the NPL and SA approaches--the 
"other" Superfund program sites--because there is no consistently 
applied method for tracking them. While the agency's estimates of the 
number of such sites is relatively small, without a method to identify 
and track such sites, it is difficult for EPA headquarters to 
determine the extent to which regions use this other approach under 
the Superfund program, evaluate regions' use of this approach, or hold 
regions accountable for using this approach. 

Furthermore, EPA guidance has made clear since 2002 that the agency 
should try to make SA agreement sites equivalent to NPL sites in terms 
of the level of cleanup achieved, among other things. EPA has largely 
accomplished this through adherence to the Superfund cleanup process 
and by adding certain provisions to SA agreements to address key 
differences between the NPL and SA approaches. The agency has not 
clarified to regions in its guidance that the SA approach is the 
preferred approach for long-term cleanup of sites under the Superfund 
program not listed on the NPL. Without clear guidance, agreements at 
such sites may be denied some of the advantages built into the SA 
agreements to ensure that the cleanups will be comparable to those 
under the NPL approach. Also, while EPA accurately identifies NPL 
sites in CERCLIS, the agency cannot do the same for SA agreement sites 
because it has not clarified in writing when the database code that 
identifies sites with SA agreements should remain in place and when it 
should be removed. In addition, EPA's standards for specifying what 
documentation is sufficient to support the Hazard Ranking System score 
at SA agreement sites are less clear than those for NPL sites. Unless 
EPA improves its tracking of SA agreement sites and clarifies its 
policies, its ability to effectively track outcomes of the SA approach 
at these sites and manage long-term cleanups at sites under the 
Superfund program may be hampered. Finally, while EPA reports 
performance information for NPL sites under GPRA, it does not report 
performance information on the progress of cleanup at SA agreement 
sites in an equivalent manner. Without such information on SA 
agreement sites, Congress lacks complete information on the progress 
of the Superfund program to inform its legislative actions, including 
appropriations. 

Recommendations for Executive Action: 

To improve the Superfund program's management of sites with 
contamination that makes them eligible for the NPL, including 
management of the SA approach and deferrals of cleanup oversight to 
other entities, we recommend that the Administrator of EPA take the 
following four actions: 

* Provide guidance to EPA regions that defines each type of OCA 
deferral and what constitutes adequate documentation for OCA deferral 
and completion of cleanup. 

* Develop a method for EPA headquarters to identify and track other 
sites with long-term cleanups under the Superfund program (i.e., those 
that are outside of the NPL and SA approaches). 

* Update EPA's written policies on SA agreement sites, including 
taking steps such as clarifying whether the SA approach is EPA's 
preferred approach for long-term cleanup of sites under the Superfund 
program and outside of the NPL, specifying what documentation is 
sufficient to support the Hazard Ranking System score at SA agreement 
sites, and defining when the database code that identifies sites with 
SA agreements should remain in place. 

* Report performance information on the progress of cleanup at SA 
agreement sites in a manner that is equivalent to such reporting for 
NPL sites. 

Agency Comments and Our Evaluation: 

We provided a draft of this report to EPA for review and comment. In 
written comments, which are included in appendix IV, EPA agreed with 
the report's recommendations and stated that it believes the report 
contains substantial useful information. Regarding the first 
recommendation, EPA stated that it added more detail on OCA tracking 
in its fiscal year 2012 Superfund Program Implementation Manual, but 
it acknowledged that more guidance is needed. Regarding the second 
recommendation, EPA stated that it agreed with the recommendation 
without further comment. Regarding the third recommendation, EPA said 
that it will clarify that the SA approach is generally the agency's 
preferred enforcement approach for CERCLA non-NPL sites that are "NPL-
caliber," where feasible and appropriate. Finally, regarding the 
fourth recommendation, EPA stated that it agrees with this 
recommendation as it pertains to reporting under GPRA and provided 
further information on how EPA reports measures at SA agreement sites. 
EPA also provided technical comments on the draft report, which we 
incorporated as appropriate. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies to the 
Administrator of the Environmental Protection Agency, the appropriate 
congressional committees, and other interested parties. In addition, 
the report will also be available at no charge on the GAO website at 
[hyperlink, http://www.gao.gov]. 

If you or your staff members have any questions about this report, 
please contact me at (202) 512-3841 or trimbled@gao.gov. Contact 
points for our offices of Congressional Relations and Public Affairs 
may be found on the last page of this report. GAO staff who made major 
contributions to this report are listed in appendix V. 

Signed by: 

David C. Trimble: 
Director, Natural Resources and Environment: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

This appendix provides information on the scope of the work and the 
methodology used to examine (1) how the Environmental Protection 
Agency (EPA) addresses the cleanup of sites it has identified as 
eligible for the National Priorities List (NPL), (2) how the processes 
for implementing the Superfund Alternative (SA) and NPL approaches 
compare, and (3) how SA agreement sites compare with similar NPL sites 
in completing the cleanup process. 

To examine how EPA addresses the cleanup of hazardous waste sites with 
a level of contamination that makes them eligible for the NPL, we 
analyzed applicable federal statutes and EPA regulations and guidance 
to determine the available approaches to address sites that are 
reported to the Superfund program. We then obtained and analyzed data 
from EPA's Comprehensive Environmental Response, Compensation, and 
Liability Information System (CERCLIS), the Superfund program's 
database, as of December 2012. Specifically, we analyzed EPA's CERCLIS 
database to determine how many sites EPA currently classified as 
undergoing long-term cleanup under each approach, both nationally and 
by each of EPA's 10 regions.[Footnote 46] According to EPA officials, 
sites under each of the long-term cleanup approaches would have a 
Hazard Ranking System score of at least 28.50, otherwise the site 
would have been classified as "no further remedial action planned." 
Thus, all sites identified as being under a long-term cleanup approach 
were considered to have contamination making them eligible for the 
NPL. This analysis involved the review of EPA's non-NPL status code, 
NPL status code, and SA code. In addition, we conducted semistructured 
interviews with officials in all 10 EPA regions to understand each 
region's processes for selecting among long-term cleanup approaches 
and why regions used the various approaches. We also obtained relevant 
supporting documentation from these regional officials. In addition, 
we interviewed EPA headquarters officials about the assessment process 
and cleanup approaches. Finally, we interviewed a nonprobability, 
convenience sample of officials from 13 state cleanup programs who 
were familiar with available cleanup approaches. The convenience 
sample consisted of representatives from state environmental 
departments taking part in an Association of State and Territorial 
Solid Waste Management Officials conference call who agreed to speak 
with us. Because this was a nonprobability sample, the results of our 
analysis cannot be generalized to all states; however, these officials 
provided important information about the cleanup process. 

To compare the processes for implementing the SA and NPL approaches, 
including the cleanup process and EPA's oversight, we analyzed 
available documentation on the two approaches, including guidance and 
prior reviews. These reviews included an EPA Inspector General (IG) 
report on the SA approach, as well as several reports on the approach 
by EPA. We reviewed key findings and recommendations from the IG's 
report, as well as the evidence provided by EPA to demonstrate its 
implementation of the report's recommendations. We found the evidence 
to be sufficient to assess whether EPA had implemented these 
recommendations. In addition, we interviewed officials in all 10 EPA 
regions to determine how each region implemented the SA approach and 
obtained relevant supporting documentation. Finally, we interviewed 
EPA headquarters officials knowledgeable about the SA approach. 

To compare how SA agreement sites and similar NPL sites complete the 
cleanup process, we identified SA agreement sites and constructed a 
comparison group of 74 NPL sites with agreements between EPA and 
potentially responsible parties (PRP) similar to those at SA agreement 
sites as follows: 

* We identified 67 SA agreement sites using the SA code and added to 
that 3 SA agreement sites with their SA code removed after the site 
was listed on the NPL for a total of 70 SA agreement sites; we 
identified these three sites through our interviews with EPA 
officials. We then obtained data on the legal actions taken at these 
sites from EPA officials in the Office of Site Remediation 
Enforcement, which included all agreements at these sites. Based on 
discussions with EPA officials and the SA guidance, we isolated 
agreements at SA agreement sites by selecting: (1) agreements entered 
into between June 2002 (the date of the issuance of the first SA 
guidance) and December 2012; (2) administrative orders on consent or 
consent decrees; and (3) agreements involving a PRP-led combined 
remedial investigation and feasibility study, remedial design, or 
remedial action. After excluding four sites with SA codes that had SA 
agreements that were not relevant to our study, we had 66 SA agreement 
sites for our analysis.[Footnote 47] 

* We constructed our comparison group of 74 NPL sites starting with 
the approximately 1,300 sites on the NPL. Specifically, we identified 
the 702 sites with (1) a combined remedial investigation and 
feasibility study, (2) remedial design, or (3) remedial action led by 
a PRP. We requested data on the legal actions taken at these sites 
from EPA officials and identified agreements similar to SA agreements 
based on the date the agreement was entered into, the type of 
agreement, and whether it included PRP-led long-term cleanup actions. 
In addition, we dropped any NPL sites from Regions 1 and 2 from the 
analysis because neither region has used the SA approach. 

To more precisely align the NPL comparison group with SA agreement 
sites, we analyzed, for SA agreements, the number of PRPs involved and 
estimated costs for PRP-led actions. According to EPA officials, SA 
agreement sites generally tend to have fewer PRPs. Based on this 
analysis and EPA's comments, we established thresholds for different 
variables that agreements in our NPL comparison group could not 
exceed. Specifically, NPL agreements could have: (1) no more than 
seven PRPs involved and (2) administrative orders on consent with 
estimated values between $100,000 and $5,000,000 or consent decrees 
with estimated values between $125,000 and $30,000,000.[Footnote 48] 
These ranges covered the vast majority of SA agreements. 

After we identified the NPL sites with agreements similar to SA 
agreement sites, we merged the data on the legal actions with cleanup 
action data for NPL and SA agreement sites. We kept (1) combined 
remedial investigation and feasibility studies, (2) remedial designs, 
and (3) remedial actions at sites if the action was explicitly listed 
as a remedy in an SA agreement or an SA-similar agreement (for NPL 
sites). We identified negotiations related to cleanup actions of 
interest by comparing the completion date of the negotiation with the 
completion date of the agreement in EPA's legal action data. For 
remedial investigation and feasibility study negotiations, we kept any 
negotiation with a completion date up to 180 days before the date of 
an administrative order on consent for that site. For remedial design 
and remedial action negotiations, we kept any negotiation with a 
completion date up to 2 years before the date of a consent 
decree.[Footnote 49] After keeping these cleanup actions of interest, 
we computed the durations of specific cleanup activities by 
calculating the difference in months between the start and completion 
dates of identified actions included in CERCLIS. We then calculated 
the mean and median durations for the SA and NPL groups, as well as 
related ranges. We compared the means and medians of the durations to 
assess whether reported results are affected by a possible skewed 
distribution. We decided to report the median because it is less 
sensitive to extreme values and provides a better estimate of the 
"average" duration for this analysis. Because only three SA agreement 
sites had reached the construction completion milestone, we were 
unable to compare the groups across the entire cleanup process; 
instead, we compared completion of specific activities, such as 
remedial designs. The results of our analysis cannot be generalized to 
all NPL sites because the 74 sites were a subset of all NPL sites 
selected to be as similar as possible to SA agreement sites based on 
key characteristics related to cleanup durations such as having a PRP 
that agreed to conduct at least some part of the cleanup. The 
comparison group was created for purposes of assessing whether 
alternative approaches for addressing the long-term cleanup of 
hazardous waste sites under the Superfund program can make a 
difference in cleanup durations and not for making generalizations 
about the larger universe of all NPL sites. 

We conducted additional analyses on our SA and NPL groups to determine 
if there were any unaccounted distributional differences within each 
group that would materially affect our results. Specifically, we 
examined the sensitivity of our results to differences in regional 
distribution because the SA approach has different regional usage 
patterns than the NPL approach. While 85 percent of SA agreement sites 
are in Regions 4 and 5, only 34 percent of the similar NPL sites are 
in Regions 4 and 5. In one analysis, we restricted SA agreement and 
similar NPL sites to Regions 4 and 5, and the results were generally 
similar to the analysis using the full set of SA agreement and similar 
NPL sites.[Footnote 50] In addition, we examined the sensitivity of 
our results to differences in the complexity of SA agreement sites and 
similar NPL sites measured through the distribution of megasites 
[Footnote 51] and single operable unit sites in each group. The 
results for length of negotiations were not sensitive to differences 
between SA agreement sites and similar NPL sites in the distribution 
of megasites, though the results for the length of cleanup activities 
were somewhat sensitive to distributional differences.[Footnote 52] 
The results for length of negotiation and cleanup durations were, in 
general, not sensitive to differences in the distribution of sites 
with one or more operable units. 

To assess the reliability of the data from EPA's CERCLIS database used 
in this report, we analyzed related documentation, examined the data 
for errors or inconsistencies, and interviewed agency officials about 
any known data problems and to learn more about their procedures for 
maintaining the data. Where there were discrepancies in the data, we 
worked with EPA officials to clarify. For example, we identified 
certain SA agreement sites that did not appear to have agreements with 
long-term cleanup actions and reviewed these with EPA officials. 
Miscoded data were corrected, and EPA officials provided explanations 
for unique circumstances with certain agreements. We determined the 
data to be sufficiently reliable for calculating durations for 
completing different cleanup activities, including negotiations, at SA 
and NPL sites. 

We conducted this performance audit from November 2011 to April 2013 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

[End of section] 

Appendix II: Long-Term Cleanup Approaches by EPA Region: 

Tables 1 and 2 provide a breakdown of cleanup approaches by region. 
Table 1 shows the number of sites within each region that are being 
cleaned up under the various cleanup approaches. Table 2 shows each 
region's percentage of the total number of sites cleaned up under each 
approach. 

Table 1: Number of Sites by Region and Long-Term Cleanup Approach, as 
of December 2012: 

Region: 1; 
NPL: 100; 
SA: 0; 
Other sites under Superfund program: 0; 
OCA deferral to states: 470; 
OCA deferral to federal agencies: 4; 
OCA deferral to private parties: 1; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 2; 
Formal state deferrals: 0; 
Deferral to the NRC: 0; 
Total: 577. 

Region: 2; 
NPL: 215; 
SA: 0; 
Other sites under Superfund program: 8; 
OCA deferral to states: 247; 
OCA deferral to federal agencies: 5; 
OCA deferral to private parties: 0; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 5; 
Formal state deferrals: 0; 
Deferral to the NRC: 0; 
Total: 480. 

Region: 3; 
NPL: 170; 
SA: 2; 
Other sites under Superfund program: 1; 
OCA deferral to states: 119; 
OCA deferral to federal agencies: 22; 
OCA deferral to private parties: 5; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 23; 
Formal state deferrals: 0; 
Deferral to the NRC: 2; 
Total: 344. 

Region: 4; 
NPL: 185; 
SA: 23; 
Other sites under Superfund program: 11; 
OCA deferral to states: 289; 
OCA deferral to federal agencies: 7; 
OCA deferral to private parties: 55; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 71; 
Formal state deferrals: 5; 
Deferral to the NRC: 0; 
Total: 646. 

Region: 5; 
NPL: 244; 
SA: 34; 
Other sites under Superfund program: 3; 
OCA deferral to states: 111; 
OCA deferral to federal agencies: 22; 
OCA deferral to private parties: 5; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 13; 
Formal state deferrals: 0; 
Deferral to the NRC: 0; 
Total: 432. 

Region: 6; 
NPL: 90; 
SA: 1; 
Other sites under Superfund program: 0; 
OCA deferral to states: 7; 
OCA deferral to federal agencies: 0; 
OCA deferral to private parties: 3; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 20; 
Formal state deferrals: 0; 
Deferral to the NRC: 0; 
Total: 121. 

Region: 7; 
NPL: 68; 
SA: 1; 
Other sites under Superfund program: 2; 
OCA deferral to states: 7; 
OCA deferral to federal agencies: 0; 
OCA deferral to private parties: 0; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 7; 
Formal state deferrals: 8; 
Deferral to the NRC: 1; 
Total: 94. 

Region: 8; 
NPL: 53; 
SA: 1; 
Other sites under Superfund program: 1; 
OCA deferral to states: 7; 
OCA deferral to federal agencies: 1; 
OCA deferral to private parties: 6; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 11; 
Formal state deferrals: 1; 
Deferral to the NRC: 1; 
Total: 82. 

Region: 9; 
NPL: 113; 
SA: 3; 
Other sites under Superfund program: 1; 
OCA deferral to states: 222; 
OCA deferral to federal agencies: 37; 
OCA deferral to private parties: 1; 
OCA deferral to tribes: 1; 
Deferral to EPA's RCRA program: 40; 
Formal state deferrals: 0; 
Deferral to the NRC: 0; 
Total: 418. 

Region: 10; 
NPL: 75; 
SA: 2; 
Other sites under Superfund program: 11; 
OCA deferral to states: 106; 
OCA deferral to federal agencies: 5; 
OCA deferral to private parties: 1; 
OCA deferral to tribes: 0; 
Deferral to EPA's RCRA program: 1; 
Formal state deferrals: 7; 
Deferral to the NRC: 0; 
Total: 208. 

Region: Total; 
NPL: 1,313; 
SA: 67; 
Other sites under Superfund program: 38; 
OCA deferral to states: 1,585; 
OCA deferral to federal agencies: 103; 
OCA deferral to private parties: 77; 
OCA deferral to tribes: 1; 
Deferral to EPA's RCRA program: 193; 
Formal state deferrals: 21; 
Deferral to the NRC: 4; 
Total: 3,402. 

Source: GAO analysis of EPA data. 

Notes: Sites listed above only include sites that are currently active 
in CERCLIS and do not include sites that have been archived by EPA or 
sites where no further site response is required, such as sites 
deleted from the NPL. 

[End of table] 

Table 2: Percentage of Sites by Region and Long-Term Cleanup Approach, 
as of December 2012: 

Region: 1; 
NPL: 8%; 
SA: 0%; 
Other sites under Superfund program: 0%; 
OCA deferral to states: 30%; 
OCA deferral to federal agencies: 4%; 
OCA deferral to private parties: 1%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 1%; 
Formal state deferrals: 0%; 
Deferral to the NRC: 0%; 
Total: 17%. 

Region: 2; 
NPL: 16%; 
SA: 0%; 
Other sites under Superfund program: 21%; 
OCA deferral to states: 16%; 
OCA deferral to federal agencies: 5%; 
OCA deferral to private parties: 0%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 3%; 
Formal state deferrals: 0%; 
Deferral to the NRC: 0%; 
Total: 14%. 

Region: 3; 
NPL: 13%; 
SA: 3%; 
Other sites under Superfund program: 3%; 
OCA deferral to states: 8%; 
OCA deferral to federal agencies: 21%; 
OCA deferral to private parties: 6%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 12%; 
Formal state deferrals: 0%; 
Deferral to the NRC: 50%; 
Total: 10%. 

Region: 4; 
NPL: 14%; 
SA: 34%; 
Other sites under Superfund program: 29%; 
OCA deferral to states: 18%; 
OCA deferral to federal agencies: 7%; 
OCA deferral to private parties: 71%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 37%; 
Formal state deferrals: 24%; 
Deferral to the NRC: 0%; 
Total: 19%. 

Region: 5; 
NPL: 19%; 
SA: 51%; 
Other sites under Superfund program: 8%; 
OCA deferral to states: 7%; 
OCA deferral to federal agencies: 21%; 
OCA deferral to private parties: 6%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 7%; 
Formal state deferrals: 0%; 
Deferral to the NRC: 0%; 
Total: 13%. 

Region: 6; 
NPL: 7%; 
SA: 1%; 
Other sites under Superfund program: 0%; 
OCA deferral to states: 0%; 
OCA deferral to federal agencies: 0%; 
OCA deferral to private parties: 4%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 10%; 
Formal state deferrals: 0%; 
Deferral to the NRC: 0%; 
Total: 4%. 

Region: 7; 
NPL: 5%; 
SA: 1%; 
Other sites under Superfund program: 5%; 
OCA deferral to states: 0%; 
OCA deferral to federal agencies: 0%; 
OCA deferral to private parties: 0%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 4%; 
Formal state deferrals: 38%; 
Deferral to the NRC: 25%; 
Total: 3%. 

Region: 8; 
NPL: 4%; 
SA: 1%; 
Other sites under Superfund program: 3%; 
OCA deferral to states: 0%; 
OCA deferral to federal agencies: 1%; 
OCA deferral to private parties: 8%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 6%; 
Formal state deferrals: 5%; 
Deferral to the NRC: 25%; 
Total: 2%. 

Region: 9; 
NPL: 9%; 
SA: 4%; 
Other sites under Superfund program: 3%; 
OCA deferral to states: 14%; 
OCA deferral to federal agencies: 36%; 
OCA deferral to private parties: 1%; 
OCA deferral to tribes: 100%; 
Deferral to EPA's RCRA program: 21%; 
Formal state deferrals: 0%; 
Deferral to the NRC: 0%; 
Total: 12%. 

Region: 10; 
NPL: 6%; 
SA: 3%; 
Other sites under Superfund program: 29%; 
OCA deferral to states: 7%; 
OCA deferral to federal agencies: 5%; 
OCA deferral to private parties: 1%; 
OCA deferral to tribes: 0%; 
Deferral to EPA's RCRA program: 1%; 
Formal state deferrals: 33%; 
Deferral to the NRC: 0%; 
Total: 6%. 

Region: Total; 
NPL: 100%; 
SA: 100%; 
Other sites under Superfund program: 100%; 
OCA deferral to states: 100%; 
OCA deferral to federal agencies: 100%; 
OCA deferral to private parties: 100%; 
OCA deferral to tribes: 100%; 
Deferral to EPA's RCRA program: 100%; 
Formal state deferrals: 100%; 
Deferral to the NRC: 100%; 
Total: 100%. 

Source: GAO analysis of EPA data. 

Notes: Percentages may not add up to 100 because of rounding. Sites 
listed above only include sites that are currently active in CERCLIS 
and do not include sites that have been archived by EPA or sites where 
no further site response is required, such as sites deleted from the 
NPL. 

[End of table] 

[End of section] 

Appendix III: Data Analysis of SA Agreement Sites and Similar NPL 
Sites: 

In this appendix, we discuss the results of our analysis of the median 
length of negotiations and the median length of cleanup activities at 
SA agreement sites and similar NPL sites, which consisted of NPL sites 
with agreements similar to SA agreements. Appendix I includes more 
information on our methodology. 

Median Length of Negotiations at SA Agreement Sites and Similar NPL 
Sites: 

As shown in table 3, for agreements with PRPs finalized from June 2002 
through December 2012, SA agreement sites and similar NPL sites in our 
analysis showed mixed results in the length of time to complete 
negotiations, with SA agreement sites taking about as long as similar 
NPL sites for remedial investigation and feasibility study 
negotiations and less time for remedial design and remedial action 
negotiations. 

Table 3: Median Length of Negotiations in Months for Agreements 
between EPA and PRPs at 66 SA Agreement Sites and 74 Similar NPL Sites 
for Agreements Finalized from June 2002 through December 2012: 

Cleanup activities involved in negotiations[B]: Remedial investigation 
and feasibility study; 
Median length of negotiations in months: SA: 7[D]; 
Median length of negotiations in months: NPL: 8; 
Number of negotiations for agreements between EPA and PRPs[A]: SA[C]: 
29; 
Number of negotiations for agreements between EPA and PRPs[A]: NPL: 13. 

Cleanup activities involved in negotiations[B]: Remedial design and 
remedial action: 
Median length of negotiations in months: SA: 9; 
Median length of negotiations in months: NPL: 14; 
Number of negotiations for agreements between EPA and PRPs[A]: SA: 21; 
Number of negotiations for agreements between EPA and PRPs[A]: NPL: 47. 

Source: GAO analysis of EPA data. 

Note: Our analysis included 66 SA agreement sites that had relevant 
cleanup actions. Some other SA agreement sites were excluded from the 
analysis because of having agreements that were not relevant to our 
work. 

[A] Sites can have more than one of each type of negotiation so the 
numbers of negotiations in the table do not always reflect the number 
of sites. 

[B] EPA tracks negotiations for combined remedial investigation and 
feasibility study, as well as combined remedial design and remedial 
action in its CERCLIS database as they are defined in the Superfund 
Program Implementation Manual. 

[C] In some cases at SA agreement sites, EPA did not record 
negotiations in CERCLIS, such as when a PRP approached EPA to begin 
negotiations. This occurred with two different negotiations in the SA 
group: one negotiation for a multisite agreement covering 11 sites and 
another negotiation for a multisite agreement covering 2 sites. 

[D] At six SA agreement sites, EPA and the PRP negotiated a single 
multisite agreement that took about 3 months. However, EPA recorded a 
3-month negotiation at each of the six sites in CERCLIS, which may 
overstate the median for the SA agreement sites in our analysis. These 
six sites are the only case of bundled negotiations with recorded 
negotiations in CERCLIS for the sites in our analysis. 

[End of table] 

Given the relatively limited number of negotiations for both NPL and 
SA agreement sites in our analysis and the effect of unique sites, the 
differences in the median length of negotiations cannot be attributed 
entirely to the type of approach used at each site. Unique conditions 
at each site have the potential to affect negotiations between EPA and 
the PRP beyond the cleanup approach selected. 

Median Length of Cleanup Activities at SA Agreement Sites and Similar 
NPL Sites: 

As shown in table 4, the SA agreement sites and similar NPL sites in 
our analysis showed mixed results in the length of time it took to 
complete specific cleanup activities,[Footnote 53] with SA agreement 
sites taking longer for remedial investigations and feasibility 
studies on average and about the same time for remedial designs and 
remedial actions on average. 

Table 4: Median Length of Cleanup Activities for 66 SA Agreement Sites 
and 74 Similar NPL Sites Completed from June 2002 through December 
2012: 

Cleanup activities[A]: Remedial investigation and feasibility study; 
Median number of months to complete a cleanup activity: SA: 69; 
Median number of months to complete a cleanup activity: NPL: 50; 
Number of cleanup activities: SA: 14; 
Number of cleanup activities: NPL: 5. 

Cleanup activities[A]: Remedial design; 
Median number of months to complete a cleanup activity: SA: 20; 
Median number of months to complete a cleanup activity: NPL: 22; 
Number of cleanup activities: SA: 9; 
Number of cleanup activities: NPL: 20. 

Cleanup activities[A]: Remedial action; 
[Empty]; 
Median number of months to complete a cleanup activity: SA: 31; 
Median number of months to complete a cleanup activity: NPL: 30; 
Number of cleanup activities: SA: 6; 
Number of cleanup activities: NPL: 35. 

Source: GAO analysis of EPA data. 

Note: Our analysis included 66 SA agreement sites that had relevant 
cleanup actions. Some other SA agreement sites were excluded from the 
analysis because of having agreements that were not relevant to our 
work. 

[A] Sites can experience more than one cleanup activity in a given 
phase, so the number of activities listed under each approach in the 
table does not always reflect the number of sites in our analysis. 

[End of table] 

Twelve of the 14 remedial investigations and feasibility studies at SA 
sites took longer than 50 months to complete, which is greater than 
the median for NPL sites in our analysis, as well as the median of 51 
estimated by EPA for PRP-led remedial investigation and feasibility 
studies that began after June 2002. However, given the relatively 
small number of cleanup activities for both NPL and SA agreement sites 
in our analysis and differences at the site level, the differences in 
the median length of cleanup activities cannot be attributed entirely 
to the type of approach used at each site. For example, several 
remedial investigations and feasibility studies at SA sites took a 
long time to complete due to individual circumstances at the site, 
such as dealing with a proposal to sell on-site materials to a 
manufacturing company, late participation from PRPs in the process, or 
coordination with other cleanup efforts. SA agreement sites and NPL 
sites in our analysis took slightly less than 2 years on average to 
complete remedial designs and slightly less than 3 years on average to 
complete remedial actions. 

[End of section] 

Appendix IV: Comments from the Environmental Protection Agency: 

United States Environmental Protection Agency: 
Office of Solid Waste and Emergency Response: 
Washington, D.C. 20460: 

March 18, 2013: 

Mr. Alfredo Gomez: 
Acting Director: 
Natural Resources and Environment: 
U.S. Government Accountability Office: 
Washington, DC 20548: 

Dear Mr. Gomez: 

Thank you for the opportunity to review and comment on GAO's draft 
report, "Superfund: EPA Should Take Steps to Improve Its Management of 
Alternatives to Placing Sites on the National Priorities List" (GA0-13-
252). The purpose of this letter is to provide the EPA's comments on 
this draft report. The EPA appreciates the GAO's work on this subject 
area and your collegial working relationship and dialogue with our 
staff. The EPA generally agrees with the GAO's findings, conclusions, 
and recommendations and is providing technical comments we believe 
will improve accuracy and clarity of the final report. 

The draft report examines current non-National Priorities List (NPL) 
approaches being used to address sites determined by the Superfund 
remedial site assessment program to need long-term remedial cleanup 
attention. The draft report focuses on sites addressed through the 
EPA's Superfund Alternative Agreement Approach, sites deferred or 
referred to other cleanup programs such as state programs, and other 
approaches. The GAO recommends the EPA improve guidance on selecting 
and tracking non-NPL cleanup approaches and report progress on sites 
using the Superfund Alternative Approach in a manner that is 
equivalent to how the EPA reports progress at NPL sites. 

GAO Recommendation #1: 

Provide guidance to EPA regions that defines each type of Other 
Cleanup Activity (OCA) deferral and what constitutes adequate 
documentation for OCA deferral and completion of cleanup. 

EPA Response: 

EPA agrees with this recommendation. EPA added more detail on OCA 
tracking in its FY2012 Superfund Program Implementation Manual but 
acknowledges more guidance is needed. 

GAO Recommendation #2: 

Develop a method for EPA headquarters to identify and track other 
sites with long-term cleanups under the Superfund program (i.e., those 
that are outside of the NPL and SA approaches). 

EPA Response: 

EPA agrees with this recommendation. 

GAO Recommendation #3: 

Update EPA's written policies on SA agreement sites, including taking 
steps such as clarifying whether the SA approach is EPA's preferred 
approach for long-term cleanup of sites under the Superfund program 
and outside of the NPL, specifying what documentation is sufficient to 
support the Hazard Ranking System score at SA agreement sites, and 
defining when the database code that identifies sites with SA 
agreements should remain in place. 

EPA Response: 

EPA agrees with this recommendation and will clarify that the SAA is 
generally the. Agency's preferred enforcement approach for CERCLA non-
NPL sites that are NPL-caliber, where feasible and appropriate. 

GAO Recommendation #4: 

Report performance information on the progress of cleanup at SA 
agreement sites in a manner that is equivalent to such reporting for 
NPL sites. 

EPA Response: 
EPA agrees with this recommendation as it pertains to reporting under 
the Government Performance Results Act (GPRA). Please see EPA's 
technical comments for recommended revisions to the draft report that 
describes how EPA reports measures at these sites. 

In closing, we believe that there is substantial useful information in 
this report and appreciate this opportunity to comment. Please feel 
free to contact me or your staff may contact Randy Hippen in EPA's 
Office of Superfund Remediation and Technology Innovation at 703-603-
8829 with additional questions. 

Sincerely, 

Signed by: 

Mathy Stanislaus: 
Assistant Administrator: 

Enclosure: 

cc: 
Cynthia Giles, OECA; 
Barry Breen, OSWER; 
Lisa Feldt, OSWER; 
Elliott Gilberg, OECA/OSRE; 
David Kling, OECA/FFEO; 
James Woolford, OSWER/OSRTI; 
Reggie Cheatham, OSWER/FERRO; 
Barnes Johnson, OSWER/OSRTI; 
Becki Clark, OSWER/OSRTI; 
Lisa Price, Region 6; 
Venessa Bowie, OCFO; 
Mark Howard, OSWER; 
Johnsie Webster, OSWER; 
Nancy Browne, OECA/OSRE; 
Doug Ammon, OSWER/OSRTI; 
Art Flaks, OSWER/OSRTI; 
Art Flalcs, OSWER/OSRTI; 
Kevin Hollingsworth, OSWER/OSRTI. 

[End of section] 

Appendix V: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

David C. Trimble, (202) 512-3841, trimbled@gao.gov: 

Staff Acknowledgments: 

In addition to the individual named above, Vincent P. Price, Assistant 
Director; Elizabeth Beardsley; Eric Charles; Pamela Davidson; Armetha 
Liles; Cynthia Norris; and Nico Sloss made key contributions to this 
report. 

[End of section] 

Footnotes: 

[1] CERCLA, Pub. L. No. 96-510, 94 Stat. 2767 (1980) (codified as 
amended at 42 U.S.C. §§ 9601-9675 (2012)). Hereinafter, references to 
CERCLA sections are as amended. 

[2] There is no legal requirement that EPA clean up a site on the NPL 
or that it do so under a particular time frame. As we have previously 
reported, EPA's future costs to conduct remedial construction at 
nonfederal NPL sites will likely exceed recent funding levels. The 
limited funding, coupled with increasing costs of cleanup, has forced 
EPA to choose between cleaning up a greater number of sites in a less 
time and cost-efficient manner or cleaning up fewer sites more 
efficiently. See GAO, Superfund: EPA's Estimated Costs to Remediate 
Existing Sites Exceed Current Funding Levels, and More Sites Are 
Expected to Be Added to the National Priorities List, [hyperlink, 
http://www.gao.gov/products/GAO-10-380] (Washington, D.C.: May 6, 
2010). 

[3] PRPs generally include current or former owners and operators of a 
site or the generators or transporters of the hazardous substances. 

[4] Where hazardous waste sites are owned or controlled by a federal 
agency, that agency may also have a significant role in cleanup. 
Processes and provisions specific to these federal sites are generally 
not discussed in this report and, according to EPA, the SA approach 
has not been used at a federal site. 

[5] We did not assess the extent to which EPA implemented all of the 
IG's recommendations because it was beyond the scope of our review. 

[6] For purposes of this report, deferral refers to sites where EPA 
elects not to use its Superfund authorities for a long-term cleanup 
because another program will provide oversight of the site's cleanup. 

[7] Forty-seven states have followed the federal government's lead and 
established their own version of the Superfund program to identify and 
clean up sites not covered by the federal program. 

[8] Operable units may address geographical portions of a site, 
specific site problems, or initial phases of a cleanup action. 

[9] Our analysis of how SA agreement sites compare with similar NPL 
sites in completing the cleanup process did not include all SA 
agreement sites because some of these sites did not have relevant 
cleanup activities. 

[10] These potential pathways of contamination include groundwater 
migration, surface water migration, soil exposure, and air migration. 

[11] EPA, Response Selection and Enforcement Approach for Superfund 
Alternative Sites (Washington, D.C.: June 24, 2002); EPA, Revised 
Response Selection and Settlement Approach for Superfund Alternative 
Sites (Washington, D.C.: June 17, 2004); and EPA, Updated Superfund 
Response and Settlement Approach for Sites Using the Superfund 
Alternative Approach (SAA) (Washington, D.C.: Sept. 28, 2012). 

[12] For non-NPL sites, EPA does not have the option to perform a long-
term cleanup itself and seek reimbursement from PRPs; however, EPA may 
enter into settlements with PRPs where applicable conditions for 
enforcement actions are met. 

[13] The Government Performance and Results Act of 1993, Pub. L. No. 
103-62, 107 Stat. 285 (1993), amended by the Government Performance 
and Results Act Modernization Act of 2010, Pub. L. No. 111-352, 124 
Stat. 3866 (2011). 

[14] CERCLA § 122(a), 42 U.S.C. § 9622(a) (2012). 

[15] CERCLA § 122(d)(1)(A), 42 U.S.C. § 9622(d)(1)(A) (2012). EPA 
tracks separately the duration of negotiations for (1) remedial 
investigation and feasibility study agreements and (2) remedial design 
and remedial action agreements. 

[16] CERCLA § 106(a), 42 U.S.C. § 9606(a) (2012). 

[17] Pub. L. No. 94-580, 90 Stat. 2795 (1976) (codified as amended at 
42 U.S.C. §§6901-6992k (2012)). Although RCRA amended the Solid Waste 
Disposal Act, Pub. L. No. 89-272, Title II, 79 Stat. 997 (1965), the 
amended law is nonetheless sometimes referred to as RCRA, a convention 
we follow here. Subtitle C of RCRA, 42 U.S.C. ch. 82, subch. III (§§ 
6921-6939f), governs hazardous waste management. Hereinafter, 
references are to RCRA as amended. 

[18] See, e.g., 10 U.S.C. § 2701 (2012), giving DOD responsibility for 
environmental restoration at its facilities, among other things. See 
also Executive Order 12580, Superfund Implementation, 52 Fed. Reg. 
2923 (Jan. 23, 1987). The executive order was issued in 1987 to 
respond to the Superfund Amendments and Reauthorization Act and 
delegates to EPA certain regulatory authorities that the statute 
assigns to the President, while delegating to the Departments of 
Defense and Energy authority for removal and remedial actions at their 
facilities, subject to certain provisions of CERCLA. This executive 
order generally gives other federal agencies authority for cleanups at 
non-NPL listed sites. 

[19] Federal facility sites on the NPL are subject to certain 
additional requirements, see 42 U.S.C. § 9620(e) (2013). 

[20] See 48 Fed. Reg. 40,658 (Sept. 8, 1983) (discussing original 
deferral policy), 60 Fed. Reg. 14,641 (Mar. 20, 1995) (providing 
explanation of deferral policy and revisions to RCRA policy), 54 Fed. 
Reg. 41,000 (Oct. 4, 1989) (discussing EPA development of its approach 
to RCRA deferral), 53 Fed. Reg. 23,978 (June 24, 1988) (discussing 
revisions to RCRA deferral). See also EPA Memorandum, Coordination 
between RCRA Corrective Action and Closure and CERCLA Site Activities 
(1996); and EPA, "NPL Deletion/Deferral Policy and RCRA Subtitle C 
Corrective Action," EPA: 540-R-95-002g (1995). 

[21] We identified the approximately 3,400 sites based on whether the 
site was undergoing long-term cleanup. According to EPA officials, 
sites under each of the long-term cleanup approaches would have a 
Hazard Ranking System score of at least 28.50, otherwise the site 
would have been classified as "no further remedial action planned." 
These sites do not include sites that EPA has archived in CERCLIS. 

[22] States also oversee cleanup at an undetermined additional number 
of sites that may be eligible for the NPL but are not listed in 
CERCLIS. States are not obligated to report all potentially eligible 
sites to the Superfund program, and environmental officials in several 
states confirmed that they have conducted or overseen cleanups at 
sites not listed in CERCLIS that may have been eligible for the NPL. 

[23] NPL sites do not include sites that have been deleted from the 
NPL. 

[24] EPA has had SA agreements at 70 sites in total; 3 of these sites 
have switched from the SA approach to the NPL approach. 

[25] EPA officials said the Superfund program had deferred 10 sites to 
the RCRA program in the last 5 years, and that deferrals from the RCRA 
program to the Superfund program probably have been more common 
recently, because of business bankruptcies. Active hazardous waste 
treatment, storage, or disposal facilities at which contamination has 
occurred may be eligible to be cleaned up as corrective actions in the 
RCRA program. These cleanups are generally the responsibility of the 
party owning or operating the permitted facility. If the party goes 
bankrupt, the party may be unable to complete the RCRA corrective 
action. 

[26] EPA Office of Solid Waste and Emergency Response, Guidance on 
Deferral of NPL Listing Determinations While States Oversee Response 
Actions, EPAl540/F-95/002 (Washington, D.C.: May 1995), 4. 

[27] CERCLA also establishes requirements related to formal state 
deferrals; see 42 U.S.C. § 9605(h) (2012). 

[28] All six of these regions had at least one site identified as an 
OCA deferral to private parties. 

[29] The completion of cleanup date at an OCA deferral site is the 
date of the determination that cleanup was successfully completed, 
that cleanup was not necessary, or that the other entity will not 
complete cleanup and the site will be referred back to the Superfund 
program. 

[30] Specifically, CERCLA established a trust fund from which EPA 
receives annual appropriations for Superfund program activities. 
Superfund trust funds are available for long-term cleanups only at 
sites on the NPL. EPA may use these funds for investigation and 
removal actions at any nonfederal site. EPA also can seek 
reimbursement from the PRPs after incurring these costs. 

[31] NRD claims are made for injury to, destruction of, or loss of 
natural resources. CERCLA defines natural resources broadly to include 
land, fish, wildlife, groundwater, and other resources belonging to or 
managed by federal or other governmental entities. 

[32] Under certain circumstances--for example, if EPA determines that 
a PRP that has entered into an SA agreement for a site is not 
adequately fulfilling the requirements of the agreement--EPA may 
decide to list the site on the NPL. 

[33] EPA IG, EPA Needs to Take More Action in Implementing Alternative 
Approaches to Superfund Cleanups, 2007-P-00026 (Washington, D.C.: June 
6, 2007). 

[34] EPA Office of Enforcement and Compliance Assurance, 
"Certification to Close OIG Audit Report, 'EPA Needs to Take More 
Action in Implementing Alternative Approaches to Superfund Cleanups'" 
(Washington, D.C.: Aug. 30, 2010). 

[35] EPA IG, EPA Needs to Take More Action, 6. 

[36] GAO, Standards for Internal Control in the Federal Government, 
[hyperlink, http://www.gao.gov/products/GAO/AIMD-00-21.3.1] 
(Washington, D.C.: November 1999). 

[37] See, for example, EPA, The Office of Solid Waste and Emergency 
Response, Fiscal Year 2010 End of the Year Report (Washington, D.C.). 
Available at [hyperlink, 
http://www.epa.gov/superfund/accomplishments.htm]. Accessed March 22, 
2013. 

[38] EPA IG, EPA Needs to Take More Action, 10. 

[39] EPA Office of Enforcement and Compliance Assurance, 
"Certification to Close OIG Audit Report," 2. 

[40] Region 4 prepares a draft Hazard Ranking System documentation 
record for SA agreement sites because the region wants the site to be 
ready for listing on the NPL, if necessary. In addition, in some 
cases, sites were already proposed to the NPL when the SA agreement 
was entered, negating any savings from avoiding this step. 

[41] See, for example, EPA, "Superfund Alternative Approach Baseline 
Assessment" (Washington, D.C.: Apr. 14, 2011); GAO, Superfund: Greater 
EPA Enforcement and Reporting Are Needed to Enhance Cleanup at DOD 
Sites, [hyperlink, http://www.gao.gov/products/GAO-09-278] 
(Washington, D.C.: Mar. 13, 2009); Industrial Economics Incorporated, 
Effectiveness Assessment of the Region 4 Superfund Alternative 
Approach, a report prepared for the EPA, November 2010. 

[42] Industrial Economics Incorporated, Effectiveness Assessment, ES-7. 

[43] Industrial Economics Incorporated, Effectiveness Assessment, ES-8. 

[44] Industrial Economics Incorporated, Effectiveness Assessment, ES-9. 

[45] Multiple cleanup activities can occur within a given phase at the 
same or different operable units at one site. Completion of one 
cleanup activity, such as a remedial investigation and feasibility 
study, does not necessarily mean all work in that phase has been 
completed. Our analysis looks at individual activities within given 
phases. 

[46] We only reported information from CERCLIS's active inventory of 
sites, as opposed to the archived inventory. EPA archives sites when 
no further interest exists at the site under the federal Superfund 
program based on available information. 

[47] We excluded four SA agreement sites that did not have relevant 
agreements involving a PRP-led combined remedial investigation and 
feasibility study, remedial design, or remedial action. First, one 
agreement involved EPA's use of a special account where the PRP funded 
the initial remedial investigation work based on the understanding 
that the PRP would conduct remedial action. Second, another site's SA 
agreement involved a PRP-led removal action and no SA agreement for 
remedial investigation and feasibility study, remedial design, or 
remedial action. Third, one SA agreement site had no legal action 
data. Finally, at one SA agreement site, the PRP began conducting 
cleanup activity before establishing an SA agreement with EPA. 

[48] We analyzed administrative orders on consent and consent decrees 
separately because consent decrees are used for remedial actions which 
can have higher costs than the activities under an administrative 
order on consent. 

[49] Consent decrees, which are used for remedial actions, must be 
approved by a court. Due to this requirement, there can be delays 
between completion of negotiations and the decree becoming official, 
so we included agreements with completed consent decrees within 2 
years of the end of the negotiation. 

[50] The one exception involved remedial actions, with NPL comparison 
sites taking 20 months longer, on average, to complete remedial 
actions than their SA agreement site counterparts in our regional 
analysis restricted to regions 4 and 5. However, this comparison 
involved only a small number of NPL sites since Regions 4 and 5 only 
contain around one-third of the complete set of NPL comparison sites 
used in our analysis. 

[51] Generally, a site is considered to be a megasite if the combined 
extramural, actual and planned, removal, and remedial action costs 
incurred by Superfund or by PRPs are greater than $50 million. The 
megasite designation may be applied to any federal or non-federal 
facility NPL or non-NPL site. 

[52] While the duration of cleanup activities was somewhat sensitive 
to the distribution of megasites, it was not always in the expected 
direction. For example, SA agreement sites in which megasites were 
excluded took longer on average to complete remedial action activities 
than the entire SA group. In contrast, NPL comparison sites in which 
megasites were not included took less time, on average, to complete 
remedial investigation and feasibility study activities than the 
entire set of NPL comparison sites. 

[53] Multiple cleanup activities can occur within a given phase at the 
same or different operable units at one site. Completion of a cleanup 
activity, such as a remedial action, does not necessarily mean the 
site has completed the entire phase. 

[End of section] 

GAO’s Mission: 

The Government Accountability Office, the audit, evaluation, and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the 
performance and accountability of the federal government for the 
American people. GAO examines the use of public funds; evaluates 
federal programs and policies; and provides analyses, recommendations, 
and other assistance to help Congress make informed oversight, policy, 
and funding decisions. GAO’s commitment to good government is 
reflected in its core values of accountability, integrity, and 
reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO’s website [hyperlink, http://www.gao.gov]. Each 
weekday afternoon, GAO posts on its website newly released reports, 
testimony, and correspondence. To have GAO e-mail you a list of newly 
posted products, go to [hyperlink, http://www.gao.gov] and select 
“E-mail Updates.” 

Order by Phone: 

The price of each GAO publication reflects GAO’s actual cost of 
production and distribution and depends on the number of pages in the 
publication and whether the publication is printed in color or black 
and white. Pricing and ordering information is posted on GAO’s 
website, [hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or 
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card, 
MasterCard, Visa, check, or money order. Call for additional 
information. 

Connect with GAO: 

Connect with GAO on facebook, flickr, twitter, and YouTube.
Subscribe to our RSS Feeds or E mail Updates. Listen to our Podcasts.
Visit GAO on the web at [hyperlink, http://www.gao.gov]. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 
Website: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]; 
E-mail: fraudnet@gao.gov; 
Automated answering system: (800) 424-5454 or (202) 512-7470. 

Congressional Relations: 

Katherine Siggerud, Managing Director, siggerudk@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, DC 20548. 

Public Affairs: 
Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, DC 20548. 

[End of document]