HIGHLIGHTS FROM
A Study of the Implementation of the
RCRA Corrective Action Program
Office of Solid Waste
U.S. Environmental Protection Agency
April 9, 2002
-------
Highlights Report
A Study of the Implementation of the RCRA Corrective Action Program, published on April 9,
2002, provides a complete discussion of the findings from EPA's data collection on facilities which have a
final and/or stabilization remedy in place.1 This highlights report presents background on the data collection
effort, information on the characteristics of the facilities examined in the study, and an overview of the main
findings from the report related to facility contamination, remedy selection process, remedies selected, and
remedy implementation. For more information on the full study, or to obtain a copy of the study, please
check the Office of Solid Waste website http://www.epa.gov/osw/ under "corrective action," or contact Paul
Balserak (703-308-0486).
I. Background
A. Introduction
The Permits and State Programs Division (PSPD) of EPA's Office of Solid Waste (OSW)
developed the RCRA Corrective Action Questionnaire in the Spring of 1997. This survey was designed to
provide a national level understanding of how the Corrective Action program has been implemented since its
promulgation. The questionnaire collected site-specific information on the implementation of Corrective
Action, including the nature of contamination, the Corrective Action decision-making process, and the
remedies selected. This information will enable EPA headquarters to more effectively support states and
EPA Regions in implementing the Corrective Action program.
As the purpose of the project was to collect information on Corrective Action implementation, EPA
queried the Resource Conservation and Recovery Information System (RCRIS) database for those
hazardous waste management facilities subject to Corrective Action that had selected a final remedy and/or
implemented stabilization measures. This search generated a list of 889 facilities. From this list, EPA
selected a statistically representative sample of facilities from which to gather information. EPA surveyed
the EPA Regional and/or state regulator responsible for overseeing Corrective Action at each sample facility.
The questionnaire included 53 questions that applied to the entire facility. These questions covered
topics such as facility background, status of Corrective Action, contamination at the facility, remedy
selection, institutional control usage and public participation. In addition, each respondent was requested to
answer a set of 18 questions for up to four specific waste management units at the facility. This report
presents the results of the facility-wide questions. It does not present results from the 18 unit-specific as
these results are not statistically representative of the larger sample.
Regulators responsible for 65 facilities completed the survey. EPA entered the information from
these facilities into the RCRA Corrective Action Implementation Database (RCAID). EPA then
extrapolated the RCAID data to represent all 889 Corrective Action facilities with a final remedy selected
and/or stabilization measures implemented. The results presented in this report are based on the
1 EPA published a version of the complete report on September 25, 2000, and a version of the Highlights
Report on October 18, 2000. Based on a subsequent review of the groundwater point of compliance (POC) data
(in particular, the response rates to the POC questions) and the statistical methods used in analyzing those data,
EPA has reevaluated the POC findings. The new POC findings are discussed in a new April 9, 2002 version of the
complete report (see page 4-7). The Highlights Report no longer presents the POC results. These changes
regarding the POC findings represent the only changes made in producing the April 9, 2002 version of each report.
Highlights Report 2 April 9, 2002
-------
extrapolated data. Also, unless otherwise noted, the results reflect extrapolated facilities for which there
was a response, excluding "unknown" responses and cases where no response was given.
The facilities surveyed are not representative of all facilities subject to Corrective Action. Facilities
that had selected final remedies and/or implemented stabilization measures were the focus of the survey.
These facilities were further along in the Corrective Action process than other Corrective Action facilities,
and therefore may differ significantly from other facilities subject to Corrective Action. The findings
presented in this report are subject to three primary limitations:
• The design of the survey does not allow for analysis of the unit-specific data. These data
could not be extrapolated because respondents did not answer the unit-specific questions
for all units subject to Corrective Action at the facility.
• The accuracy of the survey responses determines the quality of the analysis. While internal
inconsistencies within an individual facility survey were resolved where possible, other
errors or mistakes made by survey respondents, such as those resulting from the lack of
readily available information on the respondents part, could not be detected and corrected.
• The small sample size results in relatively large confidence levels associated with
extrapolated variables. Thus, many results are not statistically significant.
B. Facility Description
Thousands of facilities that generate and manage hazardous waste have releases that are subject to
RCRA Subtitle C Corrective Action requirements. The facilities that were the focus of this analysis all had
selected a final remedy and/or implemented interim measures as of 1996 according to RCRIS.
Number of Units and Facility Size. The extrapolated RCAID facilities ranged from sites with one small
contaminated solid waste management unit to facilities with extensive contamination spread over more than
25 units. Most RCAID facilities had at least five solid waste management units that were potentially subject
to Corrective Action (see Exhibit 1). Over 90 percent of all RCAID facilities were smaller than 1,000 acres;
about a third were smaller than 10 acres. There was no clear relationship between number of units and
facility size.
Exhibit 1
Number of Units Per Facility
Number of Units
Less than 5 units
5 to 9 units
10 to 25 units
More than 25 units
Total*
Percent of Responses
30%
35%
25%
15%
100%
* Total does not sum due to rounding.
Location. The survey was designed to be statistically representative at the national level, not the state or
EPA Regional level. Thus, the results should not be used for state-by-state or Region-by-Region
comparisons. Whether the lead regulator is the state or EPA Region is affected by the state in which a
Highlights Report
April 9, 2002
-------
facility is located. The lead regulatory authority was generally divided evenly between EPA Regions and
states.
Priority. RCAID facilities that had implemented stabilization measures and/or selected a final remedy
tended to be high priority facilities. Almost 80 percent of them were high priority under the National
Corrective Action Priority System (NCAPS) ranking. The NCAPS rankings were similar for facilities that
had and those that had not selected a final remedy. RCAID facilities also had NCAPS rankings similar to
the 1,714 facilities identified by the Agency for purposes of tracking Government Performance Results Act
(GPRA) goals.
Land Use and Industry Classification. The types of land use at or near facilities often influence RCRA
Corrective Action decisions. As expected, most RCAID facilities, 87 percent, were used primarily for
industrial purposes. Only 6 percent of facilities had an on-site residential land use. However, 60 percent of
facilities had residential land use within one-half mile of the facility boundary in the direction of contaminant
migration, and 30 percent had recreational land use within one-half mile.
The majority of all RCAID facilities were in manufacturing industries. Of the facilities reporting a SIC:
• 35 percent were in the chemical manufacturing sector;
20 percent were in wood preserving, petroleum refining, or other manufacturing;
• 18 percent were in a services sector; and
• 25 percent were scattered among other classifications.
II. Study Highlights
Below, EPA presents abbreviated discussions on the main findings from the RCAID study. More
complete presentations of these findings can be found in the main report.
A. Nature of Contamination and Exposure
The nature and extent of contamination at RCRA corrective action facilities are diverse.
Contamination ranges from small spill areas requiring soil cleanup to extensive on-site and off-site soil,
sediment, and groundwater contamination, including polluted drinking water sources.
Major Sources of Contamination. Survey respondents were asked to name one or two areas of concern
that contributed most significantly to contamination at the facility (see Exhibit 2). Spills were a major source
of concern at over half of the facilities. Landfills, surface impoundments, and underground storage tanks
also contributed significantly to facility contamination.
Types of Contaminated Media. Almost all RCAID facilities had both soil and groundwater contamination,
while almost one-third had sediment contamination. At roughly half of all RCAID facilities, contamination
had migrated beyond the facility boundary (see Exhibit 3). About half of the facilities with contaminated
ground water reported off-site plume migration.
Highlights Report 4 April 9, 2002
-------
Exhibit 2
Primary Areas of Concern for RCAID Facilities
Area of Concern
Spill area
Landfill
Surface impoundment
Underground tank
Above ground tank
Percent of
Facilities*
55%
35%
25%
25%
10%
Area of Concern
Waste pile
Sewer
Container/container storage area
Land treatment unit
Other type of area
Percent of
Facilities*
5%
3%
3%
< 1%
20%
Facilities could provide more than one answer to this question.
Exhibit 3
Off-Site Contamination by Media
Medium
Soil
Ground Water
Sediments
Surface Water
Air
Total
Facilities with
Contaminated Media*
99%
90%
30%
20%
5%
100% (of 889 facilities)
Facilities Where Contamination
Migrated Off-Site
5%
45%
15%
20%
5%
48% (of 889 facilities)
* Facilities could provide more than one answer to this question.
Types of Contaminants. At over half of the RCAID facilities, the contamination included only one type of
contaminant. Less than one-third of all RCAID facilities had three or more types of contaminants. The
most common types of contaminants found at RCAID facilities were:
Volatile organic compounds (VOCs): 85 percent of facilities;
Semi-volatile organic compounds (SVOCs): 40 percent;
• Metals: 25 percent; and
• Polychlorinated biphenyl (PCBs): 10 percent.
Dense non-aqueous phase liquids (DNAPLs) and light non-aqueous phase liquids (LNAPLs) are classes of
contaminants that are difficult to remove from ground water. Approximately 55 percent of RCAID facilities
had known or suspected NAPLs on-site. DNAPLs were slightly more common than light non-aqueous
phase liquids LNAPLs. About one-fourth of facilities were contaminated with both DNAPLs and LNAPLs.
Groundwater and Surface Water Use and Contamination. Almost half of the extrapolated RCAID facilities
were located above an actual or potential source of on-site drinking water. One-fourth of these facilities had
been using the on-site aquifer as a source of drinking water. Contamination had been detected in an actual
or potential source of drinking water at approximately 90 percent of facilities.
Over half of the RCAID facilities were located within two miles, in the direction of contamination, of an
aquifer that was designated as an actual or potential source of drinking water. At half of these facilities, the
off-site ground water was used for drinking water. Also, at 80 percent of these facilities, contamination had
been detected in an actual or potential source of drinking water.
Highlights Report
April 9, 2002
-------
Bodies of surface water were found within a quarter mile of over half of all RCAID facilities.
• Nearly all of the facilities were located within two miles of a body of water;
The surface water body closest to the facility was used as a source of drinking water for 15
percent of RCAID facilities; and
• Of this 15 percent, over half detected contamination in a medium associated with the
surface water body.
B.
Remedy Selection Process
EPA has long recognized the need for the Corrective Action program to be procedurally flexible,
with many different approaches to remedy selection. This section presents the data on several aspects of
the remedy selection process for RCAID facilities.
Innovative Site Characterization and Phased Corrective Action. Almost 30 percent of all RCAID facilities
used some type of innovative characterization approach. Low priority facilities were more likely than
medium and high priority facilities to employ these approaches.
Approximately 70 percent of all RCAID facilities used a phased approach for at least one step in the
Corrective Action process. The phased approach was used more often by facilities with medium and low
NCAPS rankings than by facilities with high rankings.
Action Levels and Risk Assessments. About 60 percent of all RCAID facilities used one or more types of
action levels. Action levels were developed on a facility-specific basis at 15 percent of facilities. More
commonly, facilities took action levels from standardized lists such as state action levels, EPA Regional
action levels, and proposed Subpart S levels. Use of action levels did not appear to depend on whether the
remedial alternative was based on residential or non-residential land use assumptions. Action levels were
used most often for soil, as shown in Exhibit 4.
Exhibit 4
Use of Action Levels Across Media Types
Medium
Soil
Ground Water
Surface Water
Sediment
Air
Percent of RCAID facilities
55%
40%
15%
10%
8%
Approximately half of all RCAID facilities conducted some form of facility-specific risk assessment as part
of the remedy selection process:
• 15 percent of facilities conducted both a human health and an ecological risk assessment;
• 35 percent conducted only a human health risk assessment; and
1 percent conducted only an ecological risk assessment.
Highlights Report
April 9, 2002
-------
RCAID facilities with residential, recreational, or agricultural use on-site were more likely to have both
human health and ecological risk assessments than facilities with industrial and commercial land uses. About
80 percent of the human health risk assessments were conducted by the owner or operator rather than by
the lead regulatory agency. Most of the human health risk assessments used the Risk Assessment Guidance
for Superfund (RAGS), either alone or in conjunction with other guidance. Almost all the RCAID facilities
performing an ecological risk assessment employed EPA ecological risk assessment guidance.
Public Participation. Over 80 percent of all RCAID facilities engaged in some form of public participation.
As shown in Exhibit 5, the most common forms were announcements in newspapers, magazines, and
journals. According to the data, the use of public participation mechanisms did not appear to depend on the
oversight mechanism used, the lead regulatory agency, or the NCAPS ranking of the facility. Public
participation was reported to have rarely influenced the selection of the remedial alternatives. No public
comments were received at over 40 percent of facilities.
Exhibit 5
Methods of Public Participation at RCAID Facilities
Method of Public Participation
Announcements
Fact Sheets
Information Repository
Mailings to Facility Mailing List
Informal Public Meetings
Formal Public Meetings
Door-to-Door Communication
Multilingual Communication
Percent of Facilities that Used Method
55%
40%
40%
40%
25%
20%
3%
3%
Voluntary Corrective Action. Over half of all RCAID facilities were reported to have conducted some
remedial action at the facility as a voluntary action. Voluntary remedial actions are those actions that are
designed to meet RCRA requirements, but are not specifically required in a permit or order. High priority
facilities, as well as facilities with EPA as the lead regulatory agency, were slightly more likely to undertake
voluntary Corrective Actions than were lower ranked facilities or facilities with state leads.
Of the facilities with a voluntary action subject to oversight, the majority conducted "routine
communications" with the lead regulatory agency, had regulator "review of proposed work," or had
regulator "review of completed work." Of the facilities that conducted voluntary remedial actions, over half
were provided some level of assurance that the cleanup should satisfy regulatory requirements, primarily
through the lead regulatory agency's approval of proposed work plans.
C.
Characteristics of Selected Remedies
This section presents the main findings from the RCAID work on remedy selection.
Institutional Controls. Slightly more than half of all RCAID facilities employed some form of institutional
controls. Generally, when a facility used institutional controls as a part of the remedy, it employed multiple
institutional controls. The most common institutional controls were compliance monitoring and restrictive
Highlights Report
April 9, 2002
-------
covenants (see Exhibit 6). Approximately 70 percent of the facilities using institutional controls selected
remedies that were based on a non-residential land use scenario.
Exhibit 6
Institutional Controls at RCAID Facilities
Compliance Monitoring
Restrictive Covenants
On-Site Use Restrictions
Access Restrictions
Notices
Off-Site Use Restrictions
% of Facilities
with Institutional
Controls
70%
70%
60%
50%
45%
15%
% of all
RCAID
Facilities
40%
40%
30%
28%
25%
10%
Institutional controls were used less frequently at facilities with low Corrective Action costs than at facilities
with high costs. Whether the facility had employed interim measures or implemented a final remedy did not
appear to significantly affect the use of institutional controls.
Treatment in Final Remedies. Approximately 70 percent of facilities anticipate some on-site treatment of
contaminated media.
• Almost 60 percent of all RCAID facilities anticipate on-site treatment of over half of the
contaminated media at the facility;
Roughly 50 percent of RCAID facilities anticipate on-site treatment of more than 75
percent of the contaminated media; and,
• Approximately 25 percent of facilities anticipate on-site treatment of 100 percent of the
contaminated media at the facility.
Some additional amount of media could be treated off-site. These figures represent expectations about
future treatment at facilities, rather than actual treatment performed to date.
Natural Attenuation. Approximately 20 percent of all RCAID facilities selected natural attenuation as one
component of the remedial alternative. Ten percent of facilities employed natural attenuation as a part of
the remedy for addressing off-site contamination. Natural attenuation was rarely used as the sole remedy
for contamination at the facility. At almost all the facilities using natural attenuation for ground water, there
was no known or suspected DNAPL contamination. The primary classes of contaminants for which natural
attenuation was employed as a remedy were VOCs and organics. None of the sites that used natural
attenuation had metal or pesticide contamination.
Financial Responsibility. The survey indicated that about one-fourth of all RCAID facilities had already
demonstrated financial responsibility. Financial assurance is commonly not required until a final remedy is
selected. The type of financial assurance mechanism used was identified for 15 percent of all RCAID
facilities. Of these respondents, 70 percent used the financial test or corporate guarantee and 20 percent
used a letter of credit.
Highlights Report
April 9, 2002
-------
The Corrective Actions that were subject to financial responsibility took longer to complete on average than
other Corrective Actions. While it may be expected that financial assurance would be required more
frequently for higher cost cleanups, the data did not clearly indicate such a difference.
D. Remedy Implementation
Facilities with more extensive contamination were likely to require more costly cleanup activities, as
well as more effort by EPA or the state to oversee the remedial activities. The time and costs estimated by
respondents varied widely.
Cost. The cost of cleanup at RCAID facilities varied widely, with estimates ranging from under $1 million
to over $50 million. Of facilities with reported total costs, over half estimated total cleanup costs of under
$5 million. Less than 10 percent had estimated total costs of over $50 million. Other cost findings included
the following:
Facilities with only one contaminated medium tended to be less costly to remediate than
facilities with several contaminated media (see Exhibit 7).
• The percentage of facilities with off-site migration of contaminated ground water had higher
cleanup costs on average than facilities without off-site migration.
• Facilities with commercial or industrial land use on-site tended to have significantly lower
costs than facilities with residential, recreational, or agricultural land use on-site.
Time Required to Complete Corrective Actions. As would be expected, the cost of Corrective Action and
the time required to complete Corrective Action were positively correlated:
85 percent of facilities with estimated costs under $1 million were expected to complete all
remedial activities in less than 5 years; and
80 percent of facilities with estimated costs over $50 million were expected to take at least
10 years to complete remedial activities.
Exhibit 7
Corrective Action Costs and Number of Contaminant Types
(Cumulative Percentages)
Number of
Contaminants Types
1 type
2 types
3 types
4 types
5 types
Total
<$1
million
65%
10%
0%
0%
0%
30%
<$5
million
80%
60%
30%
0%
0%
55%
<$10
million
85%
65%
65%
2%
0%
70%
<$25
million
95%
90%
90%
10%
85%
90%
<$50
million
95%
90%
90%
45%
85%
90%
< or>
$50 million
100%
100%
100%
100%
100%
100%
Multiple contaminant types (e.g., VOCs, SVOCs, etc.) at a facility were positively correlated with the length
of time required to complete Corrective Action (see Exhibit 8). In addition, facilities with off-site
Highlights Report
April 9, 2002
-------
contaminant migration on average were expected to spend more time to complete remedial actions than
facilities without off-site contaminant migration.
Exhibit 8
Time to Complete Remedial Activities by Number of Contaminant Types
(Cumulative Percentages)
Types of
Contaminants
5 types (1%)
4 types (5%)
3 types (27%)
2 types (17%)
1 type (50%)
Total (100%)
Time to Complete Remedial Activities
< 5 years
0%
0%
0%
10%
45%
25%
< 11 years
0%
50%
2%
70%
80%
55%
< 30 years
90%
60%
75%
90%
85%
80%
All years
100%
100%
100%
100%
100%
100%
Highlights Report
10
April 9, 2002
------- |