This chapter builds on the lessons learned in the earlier chapters of this report and focuses on recommendations that encourage better decision making during the post-remedy-selection phase of remediation at complex groundwater sites (as opposed to Chapter 3, which focuses primarily on the original remedy selection process). This focus is critical because the Committee has concluded that regardless of the remedial technologies applied at complex sites, removal of sufficient mass to reduce contaminant concentrations in groundwater to levels that allow for unlimited use and unrestricted exposure is unlikely for many decades. Furthermore, no transformational remedial technology or combination of technologies appears capable of overcoming the inherent technical challenges to restoration at these complex sites. Rather, the nation’s cleanup programs are transitioning from remedy selection into remedy operation and long-term management (LTM), potentially over long timeframes. The implications of the limitations of existing technologies to attain unlimited use and unrestricted exposure levels throughout the impacted aquifer should be more fully reflected in the decision-making process used in existing cleanup programs and should be recognized earlier in the regulatory cycle of these complex sites.
Better decision making is needed at key points in the life cycle of a complex groundwater contamination site to address issues that frequently arise, including (1) what is a “reasonable time frame;” (2) what is the definition of contaminant removal “to the maximum extent practicable;” (3) when should active remediation at a complex groundwater site be transitioned to a passive remedy, such as monitored natural attenuation (MNA) or natural attenuation (NA); and (4) can consensus be reached on a “dimin-
ishing returns” concept applied to the performance of active remedies for groundwater cleanup at complex sites?
This chapter provides a series of recommendations that, in the judgment of the Committee, will accelerate the transition of a site to one of three possible “end states,” where this term simply means a state where long-term management will be implemented if required. These “end states” are (1) closure in which unlimited use and unrestricted exposure levels have been attained (presumably no long-term management will be required at such sites, although even these sites can be subject to reopeners should conditions change); (2) long-term passive management (e.g., using MNA, NA, physical containment, PRBs, institutional controls, or some combination thereof), and (3) long-term active management (e.g., indefinite hydraulic containment using pump and treat or other active remedies requiring continuous operation). Complex sites under both passive and active long-term management could eventually transition to the closure end state, but the time frame extends many decades into the future.
The acceleration of this transition to one of three end states is premised on using remedies that are fully protective of human health and the environment in combination with more rapid acceptance of alternative end states other than clean closure, taking risk reduction, life-cycle costs, and technical feasibility into account. The transition of a site to either passive or active long-term management must be accomplished in a manner that is transparent, reduces long-term risks to an acceptable level, and is practical and cost effective, among other goals.
All complex contaminated groundwater sites will ultimately transition from investigation, through remedy selection, implementation, and operations, to long-term management and ultimately (without any time frame constraint) to attainment of unlimited use and unrestricted exposure goals. If these concentration levels have not been achieved after some reasonable time period, a site will require long-term monitoring and management under either passive remedies such as MNA or an active remedy such as pump and treat. The number of complex sites where achieving unlimited use and unrestricted exposure goals within a reasonable timeframe is unlikely is not known precisely, but as discussed in Chapter 2 is estimated to be in the range of 12,000 sites and may be much higher.
Over more than 30 years of remediation experience at contaminated groundwater sites following the passage of federal statutes (CERCLA and RCRA) and state regulations have shown that the duration of the cleanup process—from initial discovery of contamination to installation of “final” remedies (as opposed to interim actions designed to eliminate imminent
threats)—can take 10 to 25 years at complex sites. If during this time interim actions have not halted or substantially slowed the migration of contamination, contaminants can spread both laterally and vertically, making the site even more difficult and costly to address. A desire to transition more quickly from remedy operation to some final end state was expressed frequently to the Committee over the last two years. In particular, many of the DoD personnel expressed frustration over the long timeframes and financial demands resulting from continued delays in decision making at key points in the remedial process. A reduction in these timeframes is desired during many phases of cleanup, not just between remedy operation and the transition to long-term management or actual closure, but also between initial site characterization and remedy selection and in the period subsequent to remedy implementation during which the effectiveness of the remedy is assessed.
Accelerating decisions throughout the cleanup process is difficult for a number of reasons, as many initiatives undertaken in the past have made clear (Clean Sites, 1990; DOE, 1998; EPA, 2011a; various ITRC documents). First, the investigatory process is inherently difficult, expensive, and may not result in an accurate conceptual model of the site, at least initially. There are uncertainties and subjectivity in selecting the appropriate technologies for site characterization and remediation, including differing perspectives over the intensity of data collection and its timing and cost effectiveness. Second, there can be disagreements among potentially responsible parties, the states and even within the ten EPA Regional offices over the appropriate risk management approach to take at an individual site (i.e., how much aggressive source removal is warranted when considering the time to reach groundwater cleanup goals and the high costs of such actions). Third, when residual contamination is left in place, stakeholders have increasingly argued for remedy modification to accelerate risk reduction, to consider more recently developed sustainability metrics, or to address previously unknown risks such as exposure to volatile chemicals via vapor intrusion (see Chapters 5 and 6).
In addition to reducing the time to reach cleanup goals, potentially responsible parties (PRPs) such as the DoD strive to minimize life-cycle costs at these complex sites. Some estimates of future costs for site cleanup exceed $300 billion (in 2004 U.S. dollars) (EPA, 2004). As discussed in Chapter 2, the Committee’s upper estimate of $127 billion (see Chapter 2) is likely an underestimate of future liabilities. At the same time, severe budget constraints at the federal and state levels have led to the need for prioritization of resource expenditures to ensure that the greatest risks at contaminated sites are mitigated and that long-term containment of contamination is achieved to ensure no unacceptable risks to human health and the environment. In this context, the DoD in particular has established
aggressive goals to reduce costs at the majority of their sites (Conger, 2011; Yonkers, 2011). Based on the experience of some Committee members, private sector PRPs with large portfolios of complex groundwater sites have also established strategies to accelerate cleanup, with a goal of reducing life-cycle costs. Strategies have included setting up separate companies responsible for remediation of legacy sites, or outsourcing the management of complex sites through the use of experienced program management companies that oversee remediation.
Innovation would be particularly helpful in three areas of decision making. The first is making decisions in a more prompt manner, as delays result in higher transaction costs and increase the risks of ineffective interim remedies. Second, better decision making requires better metrics for demonstrating progress—based not upon regulatory milestones, but upon quantifiable, transparent metrics of remedial performance and human health risk reduction. Third, the decision on when to transition to long-term management should be formally recognized as the point where further active remediation results in little or no decreases in contaminant concentration, and the unit cost of the remedy increases much faster than the reduction in contaminant concentrations. The benefits of improved decision making at complex sites may include (1) reduction in the duration between decisions in the cleanup process; (2) cost savings at particular sites; (3) more rapid restoration of impaired groundwater resources, thereby allowing unrestricted use of at least some portions of a site; (4) more rapid mitigation of exposure pathways while long-term strategies are being considered; and (5) minimizing the long-term risks and financial burdens, to the public and PRPs, associated with groundwater sites where residual contamination is likely to persist for long timeframes. Of course, not all of these benefits can be simultaneously realized.
EPA’s Existing Site Remediation Process for Groundwater
EPA recently summarized its existing guidance for restoring contaminated groundwater (EPA, 2011b). While its groundwater restoration Road Map focuses on CERCLA, EPA has stated elsewhere that the methodology also applies to groundwater remediation under the RCRA Corrective Action program (see Chapter 3). Figure 7-1 shows the current EPA decision framework, which includes five key decision points (diamonds in the figure): (1) determine if selected remedy is viable, (2) post operations, determine if operational data are sufficient to evaluate performance of the remedy, (3) determine if achievement of the remedial action objectives (RAOs) in the record of decision (ROD)—usually defined as restoration where groundwater is a potential source of drinking water—can be achieved or not, (4) if restoration is likely, determine whether in fact RAOs have been achieved,
FIGURE 7-1 EPA recommended process for restoring contaminated groundwater at Superfund sites.
SOURCE: EPA (2011b).
and (5) if restoration is unlikely, evaluate whether alternative technologies can overcome the restoration limitations of the technologies included in the ROD. In the event that available and demonstrated technologies cannot meet RAOs, a Technical Impracticability (TI) waiver can be granted. The transition to site completion thus requires modification of the RAOs, possibly modification to the remedy, and continued evaluation and assessment of an active remedy. There is, however, no discussion of a transition to passive or active long-term management. In the opinion of the Committee, lack of guidance on this transition is unfortunate given the likelihood that many contaminated groundwater sites pose technical challenges that will not allow for achievement of unlimited use and unrestricted exposure throughout the entire contaminated aquifer for many decades.
The EPA Road Map includes a process to monitor performance of the remedy and conduct regular (every five years) reviews of the adequacy and protectiveness of the remedy, considering new information if available. Similarly, under the RCRA corrective action program continued operation of the industrial facility requires approval of a permit that will include necessary monitoring and reporting requirements where contamination remains on site, as well as verification that remedies are meeting such requirements as eliminating off-site contaminant migration.1 Many Superfund facilities have now been through multiple five-year reviews, although statistics on the average number of five-year reviews or the number of non-Superfund sites that have received similar regulatory reviews are not readily available. Given the lengthy life cycle of these complex sites (and associated costs), the development of an alternative process for addressing groundwater contamination is warranted that accounts more explicitly for the likelihood of residual contamination and provides sufficient engineering and legal controls to ensure that exposure risks are below acceptable limits. An alternative decision process is shown in Figure 7-2.
Initial Steps of the Alternative Decision Process
Figure 7-2 includes the processes currently followed at all sites regulated under CERCLA and at many complex sites regulated under other federal or state programs (RCRA Corrective Action or state Superfund sites), but it provides more detailed guidance for sites where recalcitrant contamination remains in place at levels above those allowing for unlimited use and unrestricted exposure. The first two decisions in Figure 7-2, shown as diamonds, come after a remedy is in place as directed in the ROD. These diamonds represent two key questions that must be answered (usually during the five-year review): (1) is the remedy protective of human health and the environment (considering all exposure pathways, including vapor intrusion), and (2) have maximum contaminant levels (MCLs) or other RAOs been met at the designated point(s) of compliance (POCs).
With respect to remedy protectiveness, EPA’s five-year review guidance (EPA, 2001a) states that “if the risk associated with the cleanup levels currently being achieved by the remedy are within EPA’s acceptable risk range, the remedy generally should be considered protective.” Protectiveness may be demonstrated through a variety of means including a human health risk assessment that demonstrates site risks for all exposure pathways fall
FIGURE 7-2 Key decisions for complex sites with groundwater contamination.
within EPA’s acceptable risk range of 10–6 to 10–4, interruption of exposure pathways, and other measures. Based on the experience of the Committee, designation of a CERCLA remedy as “not protective” during the five-year review is relatively infrequent and when it does occur, action is eventually taken. The MEW Superfund facility in Mountain View, California, is an example of a site that was considered “protective” through one five-year review cycle until 2009, when the remedy was no longer considered protective because of newly identified human health risks from vapor intrusion. Similar cases may arise elsewhere as the potential risks from vapor intrusion become better understood.
With respect to the second decision, there are certainly instances where sites regulated under CERCLA have reached the level of unlimited use/unrestricted exposure at the points of compliance, but as discussed in Chapter 2 many of these sites were listed early in the program’s history and likely did not have significant contamination to begin with. A larger number of these delisted sites have persistent groundwater contamination (see Chapter 2). For complex DoD sites, few have achieved unlimited use and unrestricted exposure levels in groundwater. Thus, at many groundwater sites the answer to this second question will be “no,” leading to the third decision.
Has an Asymptote Been Reached?
Figure 7-2 begins to diverge from the EPA Road Map for remediating groundwater at complex sites at the third decision point, which asks whether an asymptote, or a point of diminishing returns, has been reached in the benefits gained by applying the current remedy, considering both time and cost. The benefits metrics considered here are derived from the cleanup objectives for the site and might include reductions in groundwater or indoor air concentrations, shrinking of the dissolved plume footprint, reductions in source zone mass, or an increase in the potential beneficial uses of the site and any affected resources (e.g., aquifers and surface water bodies).
Remedy performance is typically reported on a quarterly schedule and may include data collected at daily or weekly intervals. This provides sufficient data to adjust and optimize systems on monthly to quarterly timeframes and to assess, on annual timeframes, if operation of the remedy continues to provide significant benefits. Given this context of typical data monitoring and reporting cycles, the question of whether an asymptote has been reached can be rewritten as “given their current trends, will there be significant changes in the benefits metrics (e.g., groundwater concentrations) over the next 12 months and after another five years of remedy operation?” The 12-month and five-year periods are relevant and practical because they represent a reasonable minimum future projection period given typical data sets and the mandated CERCLA five-year review cycle,
respectively. The Committee’s experience with complex sites suggests that there will be few cases where optimal remedy implementation reaches asymptotic conditions in less than 12 months.
Answering the question of “will there be significant changes in the benefits metrics over the next 12 months and after another five years” will involve professional technical judgment, input from all stakeholders, and consideration of the legal framework. Discussion will likely center on the definition of “significant changes” for the benefits metrics. It should be noted that this question is only relevant in cases where there is groundwater contamination above levels allowing for unlimited use and unrestricted exposure and only in cases for which remedy optimization has already occurred.
Some of the monitoring tools mentioned in Chapter 6 could be particularly useful for determining whether a site has reached an asymptote, such as compound specific isotope analysis. For example, if contaminant concentrations were found to be steady (or even decreasing), and isotopic signatures were not changing along the treatment gradient, it would suggest that no further degradation is occurring (i.e., concentrations are too low for biodegradation to operate or the transformation capacity of the aquifer has been exhausted). In that case, it would suggest that the remedy is no longer having an effect and that the point of diminishing returns had been reached. Other examples discussed in Chapter 6 include the application of diagnostic tools to measure reductions in mass flux/mass discharge from source areas, which could also be used to assess whether asymptotic limits have been reached.
It should be noted that while a clear, purely technical trigger for determining when the asymptote has been reached would be desirable, this decision must be made in the context of the legal framework within which the site is managed and the sometimes competing stakeholder interests. Regulators and impacted communities are often unwilling to accept that a remedy will not reach its objectives in a certain timeframe. However, it is generally the experience of the Committee that stakeholders will understand the technical limitations of the remedy provided that such discussions are conducted in a transparent manner (see further discussion of stakeholder interactions later in this chapter). Furthermore, the Committee has observed that members of affected communities do not support the indefinite expenditure of resources on activities that do little or nothing to reduce risk. Thus, the asymptote analysis is not just a valuable decision-making tool, it is also a vital part of communicating the challenge of groundwater remediation to the public.
The paradigm embodied in Figure 7-2 strikes a balance between deciding not to undertake any risk reduction efforts as part of the original remedy because of the impracticability of attaining drinking water standards
at complex sites versus continuing to expend extensive resources when further remedial work provides no substantial incremental risk reduction (as measured by decreases in contaminant concentration or other appropriate benefits metrics).
The Transition Assessment
For sites that have reached an asymptote, Figure 7-2 leads to a transition assessment. The transition assessment is envisioned as an analysis similar to a focused feasibility study that considers alternatives for site management—choosing a new remedy or transitioning to long-term management (such as monitored natural attenuation) or the other alternative approaches outlined in Chapter 3 and ESTCP (2011). For each of the possible alternatives, the transition assessment considers the nine remedy selection criteria of CERCLA or similar criteria established under other regulatory regimes, particularly the risk of any residual contamination expected to remain in aquifers that are not likely to be restored; costs such as life-cycle costs and the marginal costs of remediation compared to the level of risk reduction achieved; and state and community acceptance. Although the actual decision making occurs at the same time through comparison of the alternatives, for simplicity we begin first with a discussion of whether a new remedy is warranted, followed by the transition to long-term management. Risk and cost considerations are also detailed here, while stakeholder concerns are presented in a subsequent section.
The transition assessment is partly a response to the ongoing debates regarding the cost effectiveness of remedies at complex sites that have been operating for several years. Although some additional costs will be incurred to conduct the transition assessment, it is anticipated that if the paradigm in Figure 7-2 is followed, overall life-cycle costs will be reduced while maintaining the goal of protecting human health.
Is a New Remedy Warranted?
When asymptotic conditions have occurred and concentrations remain above MCLs or other restoration goals, an evaluation is required of whether modifications to or replacement of the existing remedy are warranted. This decision must be consistent with the legal requirements in existing consent orders or settlements (as well as existing remedy selection criteria). Where a non-federal PRP has entered into a legally binding agreement, this agreement generally provides a legal release from future cleanup liability if the PRP implements the original remedy selected for the site and the remedy remains protective. Thus, at such sites the decision would
proceed to the question about passive management without consideration of whether a new remedy is warranted.
Regulators and PRPs may agree to modify the mandates in existing regulatory instruments (e.g., an administrative order or settlement) to include consideration of alternative remedies. This may be desirable where a private company is trying to follow a new internal policy (e.g., a sustainability policy), or where the application of a new remedy is likely to attain drinking water standards and the incremental cost is less than the existing remedy. Similarly, if attaining drinking water standards will contribute to making the area more likely to be redeveloped, a company may agree to perform a new remedy even if the cost of implementing the new remedy exceeds the cost of the existing remedy. There is no legal release provided in the federal facilities agreements that govern cleanup at DoD sites. Governmental PRPs may decide as a matter of agency policy, or be ordered by states, to perform additional remedial work in the appropriate circumstances.
In cases where the site does not have a legal release, one question to be asked is, “is a new remedy warranted based on the original CERCLA or State remedy selection criteria?” The answer to this question could be “yes” if a new technology has been recently developed that could address the contamination more effectively (i.e., significantly reduces the timeframe to achieve restoration compared to the remedies considered during the initial feasibility study), and is cost effective and practical (see Chapter 3 for a more complete discussion of remedy selection criteria). The Committee expects that a new remedy might be warranted where its implementation would achieve drinking water standards in a significantly shorter period of time, the technology is clearly feasible, and the total incremental life-cycle costs of a new remedy are less than the continued costs of the existing remedy.
The answer to the question could be “no” if it is determined, based on balancing all the remedy selection criteria, that aquifer restoration using an active remedy is unlikely within a certain timeframe (say 100 years). For example, if a new remedy were predicted to reduce contaminant concentrations in groundwater to drinking water standards in 1,000 years versus 10,000 years, the answer to the question would be “no.” Predictive numerical models of the type described in Chapter 6 are critical to answering such questions. Furthermore, this approach must be tailored by the relevant regulatory agencies to conform to existing statutes and regulations. If no new remedy is warranted, the transition assessment for such a site would then determine which of the two long-term management end states would be preferable.
The Southeast Industrial Area (SIA) on the Anniston Army Depot is an example of a site where the questions in Figure 7-2 have been variously tackled, but the physical complexities in the subsurface at the SIA
Anniston Army Depot, Anniston, Alabama
The Anniston Army Depot (ANAD) is an active military facility that occupies 15,200 acres in northeastern Alabama. The storage, maintenance, and industrial functions of ANAD have generated solid and liquid wastes including metals, pesticides and herbicides, chlorinated and petroleum hydrocarbons, and solvents, among others. From the 1940s through the late 1970s, wastes were disposed of on-site in trenches, lagoons, landfills, and other surface impoundments. Since 1976, investigations have focused on contamination of shallow groundwater at the facility and off-site groundwater by chlorinated solvents and metals. The contaminant source areas are the Landfill Area, Trench Area, Northeast Area, and Southeast Industrial Area (OU1), which are underlain by a residuum layer and bedrock. The total mass of TCE in groundwater is estimated to range from 3.6 to 27.1 million pounds, 87 percent of which is in the residuum lithologic unit (Malcolm Pirnie, 2006; SAIC, 2005). In addition, over 99 percent of the total TCE mass is present as DNAPL. TCE and other contaminants continue to migrate vertically and horizontally from the source areas, impacting groundwater. The primary receptor is Coldwater Springs, a prolific natural spring supplying potable water to the City of Anniston, Alabama.
The 1991 interim ROD required a groundwater extraction and treatment system known as the Groundwater Interceptor System (GWIS). An air stripping system is removing TCE to levels below drinking water standards (i.e., less than 5 μg/L) at the springs. However, in the most recent five-year review (EPA, 2010a), EPA determined that the interim remedy at OU1 is not protective because the on-site pump and treat system is not significantly reducing the extent or mobility of TCE contamination in the groundwater. Possible exposure to contaminants in Coldwater Springs water is unlikely if the treatment system is operating, and regular monitoring continues to show that TCE is effectively removed by the air stripper (EPA, 2010a).
Given the large amount of TCE DNAPL present at this site, the inability to
and disagreements over the applicability of a TI waiver have resulted in lengthy delays in agreeing on a final remedy for the site. As discussed in Box 7-1, human health exposure due to TCE releases from the site has been eliminated by the use of wellhead treatment at the point of exposure (water supply source), and yet migration of the TCE in groundwater is not under control and thus the remedy is not fully protective as noted in a recent five-year review (EPA, 2010a). A focused feasibility study was recently conducted to compare alternative remedies, including combinations of enhanced groundwater extraction, additional use of in situ oxidation technologies, enhanced bioremediation, and land use controls, among oth-
locate the contaminant transport pathways between OU1 and Coldwater Springs, the technical difficulties in restoring the groundwater quality beneath the site, and the complex lithology in the subsurface, the Army applied for a TI waiver to replace the RAO of achieving the TCE MCL throughout the aquifer system. However, because the current interim remedy is considered “not protective,” the waiver was denied and additional efforts have been made by the Army, under direction from EPA and the State of Alabama, to remove greater quantities of TCE and other compounds from the subsurface.
The Army’s 2011 Focused Feasibility Study (Tetra Tech, 2011) evaluated enhanced pump and treat and bioremediation for their estimated time to achieve RAOs in the groundwater compared to the present remedy, using solute flow and transport models (Modflow, Remchlor, and MT3D). The predicted time for the TCE to reach the MCL ranged from 1,233 years to over 10,000 years for the various alternatives. Estimated net present values costs for applying the two alternative strategies are approximately $17 million and $21 million (in 2011 dollars), respectively. Another interim ROD is expected soon.
ANAD exhibits many characteristics of complex sites—complex hydrogeology, large quantities of DNAPL, continued uncertainties in establishing the exact pathways of contaminant transport, and very long timeframes to achieve restoration. In addition, the site is affecting a critical water supply, but these impacts have been mitigated through the installation of air stripping to meet drinking water standards. The continued use of the air stripping system does pose risks to the consumers, but these risks can be mitigated through proper monitoring and operation of the treatment system. Finally, because the interim remedy was not protective, alternatives were evaluated that appear to be capable of reducing the mass discharge of TCE to the major receptors. However, the estimated time of remediation greatly exceeded 100 years, and it is thus difficult to determine the benefits of choosing any alternative compared to continued operation of the pump-and-treat system at Coldwater Springs. As long as the interim remedy is deemed “not protective,” additional remedial efforts will be required to meet revised RAOs for the site.
ers. Modeling of the potential performance of the alternatives predicted timeframes for reaching groundwater cleanup goals ranging from 1,233 to 10,000 years (neither end of which is likely to be considered “reasonable”). Furthermore, attempts to negotiate a TI waiver for TCE in a portion of the site were rejected by regulators. The current status of the site following completion of the focused feasibility study appears to be preparation of an interim ROD, which will require implementation of an in situ bioremediation remedy to achieve revised remedial action objectives for the site. The path to a final ROD is uncertain at this time (Laurie Haines-Eklund, AEC, personal communication, 2012).
Consideration of Costs Curing the Transition Assessment
The issues discussed in Chapter 3 concerning the consideration of cost in the remedy selection phase are likely to arise during the transition assessment, particularly the question of whether to use a broader range of discount rates in calculating costs. For example, the current present value of costs could be recomputed using the opportunity cost of capital discount rate that is appropriate at that time. (As in Chapter 3, the Committee is not advocating for any particular discount rate; the choice of discount rate is based on balancing various policy considerations including statutory requirements). Costs for the various future alternatives—MNA, other passive management, continued or new active management—could then be compared to one another on a common economic basis.
With respect to the methods used to estimate the costs of the various alternatives, probabilistic cost models could be considered for estimating remedial and operation/maintenance costs and potential environmental liabilities (e.g., see Hayes et al., 1996). This approach identifies a range of statistical probabilities to address the uncertainty of critical issues faced during remediation, including not only the extent of contamination, but also construction, operation and maintenance, and legal/regulatory risks that may occur. The approach generates ranges for the median costs, fair value, and other parameters as a probability-weighted average for several possible scenarios.
This reevaluation of costs during the transition assessment is relevant for several reasons. First, existing costs represent a sunk investment and the relevant questions deal with the best path forward from the current point in time. Second, initial remedy selection is based on projected risk reductions and projected costs, both of which could be wrong. Third, economic conditions change and might make remedy implementation more or less costly. Finally, failing to do this assumes that the present value of costs is fixed at the time of the ROD. Indeed, by allowing the cost analyses to adapt through time, one can better inform any remedial actions that might be undertaken. Such updating of costs would occur during a transition assessment or during the five-year review if the remedy was found not to be protective.
Risk Assessment as Part of the Transition Assessment
The transition assessment provides an opportunity to reevaluate site risks above and beyond what would normally occur during a five-year review. When new information becomes available that would indicate the remedy may no longer be protective, such as new toxicity information that indicates greater chemical toxicity, the development of new health-based
criteria (e.g., MCLs), or the identification of new exposure pathways (see Chapter 5 and the MEW case study), this is supposed to be reflected in five-year review reports. Under RCRA corrective action, such changes might trigger modifications to the existing Part B permit for operating industrial facilities. Such considerations can indicate the need for additional actions if a remedy is no longer protective. However, because the five-year review focuses primarily (if not solely) on lack of protectiveness, the risk analyses that occur during a typical five-year review provide an incomplete understanding of risks posed at a site presently and in the future. A transition assessment would instead include more comprehensive risk assessment, including
• Better understanding of risk reductions as predicted in the ROD compared with actual risks reduced from ongoing remedial activities. Evaluation of both individual risks and population risks could be conducted (see Chapter 3).
• Risks associated with various future alternatives—passive management, MNA, continued or new active management. These risks could be compared to one another, in terms of the magnitude of the relative risk reduction and the time scale to achieve the projected risk reduction.
• More explicit consideration of uncertainty that would reveal not only the confidence in existing risk estimates, but also the potential impact of other factors (e.g., new dose-response studies, future groundwater and land uses) on risk estimates in the future.
Is Passive Long-Term Management Acceptable?
The transition assessment considers whether or not a passive long-term management option (such as MNA, NA, a permeable reactive barrier, or physical containment) is appropriate. The current regulatory approach to the use of MNA is to require that the timeframe for restoration using MNA is “reasonable,” usually considered to be less than 100 years (EPA, 1999, 2007, 2011c; ITRC, 2010; USGS, 2007). However, for most non-petroleum-based contaminants, MNA will often require longer timeframes for restoration, and there have been instances where MNA has been allowed even though restoration will likely not occur in 100 years. Where site-specific risk analysis indicates that residual contamination will not pose unacceptable future risks, the use of a longer timeframe criterion for transition to MNA is critical to avoid inefficient use of limited resources. This may require the definition of a zone within the aquifer where alternative RAOs would be established (equivalent to a “TI Zone” in the current EPA Road Map).
An example of the transition to passive long-term management is the MEW case study, where regional pump-and-treat systems have been operating for about 15 years. The site will likely be transitioning to MNA once a certain target concentration has been met. Another example is the Solvents Recovery Service of New England Superfund facility in Southington, Connecticut, that had solvents in bedrock (EPA Region 1, 2010). Modeling indicated that the solvent plumes would not reach MCLs for 500 years under baseline conditions. Even if 95 to 99 percent of the overburden mass could be removed, it would take approximately 250 years to clean up the bedrock. Thus, the chosen approach is to use pump-and-treat, MNA, and institutional controls.
The end state of long-term passive management (see Figure 7-2) requires that an acceptable and cost-effective monitoring program be established to ensure that the passive management alternative is protective of human health and the environment. The necessary engineering and legal controls would be established as part of the management plan. Finally, transparent communication with community representatives would be an essential component of the plan for any site in this “end state.” It should be stressed that sites in this end state may eventually achieve restoration throughout the aquifer, but it is expected that the timeframe for this to occur would exceed many decades (see the dashed lines in Figure 7-2).
If MNA or passive containment is not appropriate based on a balancing of all of the remedy selection factors, a long-term active management program would be developed. This would include a comprehensive outreach program to the affected community, a thorough plan for monitoring contaminant concentrations in groundwater, five-year reviews for Superfund facilities, and maintenance of institutional and engineering controls.
The Role of the Five-Year Review in More Effective Long-Term Management
Many of the above sections have mentioned doing certain tasks as part of the five-year review process. Indeed, although the five-year review is shown as a single box in Figure 7-2, it would occur at regular intervals at all sites with residual contamination, under either active or passive long-term management. Thus, the questions posed in Figure 7-2, including those in the transition assessment, could theoretically all occur under the umbrella of an improved five-review process. This suggests the need to comment on the current status of five-year reviews.
In 2011, EPA’s Office of the Inspector General (OIG) found that 84 percent of the five-year reviews performed since 2006 were overdue as of April 28, 2009 (EPA, 2011d). Also, the EPA OIG noted that EPA regional staff did not consistently follow CERCLA policies for updating the Com-
prehensive Environmental Response, Compensation, and Liability Information System (CERCLIS) database on the status and actions taken to address prior review issues. In addition, recommendations from prior reviews were not always followed through. The report concluded that “EPA does not have effective management controls to monitor the completion of five-year review recommendations at federal government Superfund sites.”
In response to the Inspector General’s report, EPA recently issued priorities for the five-year review program that would require more stringent management controls to ensure that recommendations in the five-year review reports are tracked, monitored, and implemented by EPA and the federal agencies (EPA, 2011e). The Inspector General’s report specifies the actions that remedial project managers must undertake to ensure that the reviews are completed on time and are performed in an independent manner, particularly with respect to the protectiveness determination. Most importantly for federal facilities, given the historic disagreements between EPA and DoD, it confirms that five-year reviews are enforceable under the Federal Facility Agreements (EPA, 2011e).
This Committee did not perform a review of the adequacy of five-year review reports, although it read many such reports as part of its evaluation of groundwater sites removed from the National Priority List (see Appendix C). The Committee also considered other evaluations of five-year reviews (EPA, 2011e; GAO, 2009; Pham, 2010; Probst, 1999; Schiller, 2009). In general, these evaluations indicate a need to improve the quality and comprehensiveness of these reviews, particularly because of the importance of the “protectiveness” decision. Given the opportunity to improve decision making as outlined above, both revision of the five-year review guidance and issuance of separate guidance for non-Superfund sites to incorporate and institutionalize the transition assessment is the recommended path forward for implementing the process shown in Figure 7-2. For each new component, the five-year review reports should have enough documentation for stakeholders to replicate the calculations and understand the support for the conclusions. This suggests that five-year review reports will need to become much more comprehensive than they have been in the past.
Because EPA guidance argues that cleanups at similar types of sites should follow the same process and generally should result in the same or very similar remedial actions (see Chapter 3), the transition assessment should apply to RCRA sites undergoing corrective action for groundwater cleanup and other cleanup sites. However, the RCRA corrective action program is implemented through corrective action permits while a facility is still in operation. As a practical matter, while a plant is operating the primary focus of the corrective action plan is to protect human health and prevent expansion of the plume. As a result, typically RCRA corrective actions early in the process do not involve the selection of a “final” remedy.
Once the final RCRA corrective action remedy is selected and implemented, then the transition assessment applies and should be used. It should also be noted that Figure 7-2 applies primarily to complex sites. A transition assessment is not necessary at sites where the remedy is likely to attain drinking water standards within a reasonable period of time.
At complex groundwater sites that are the focus of Figure 7-2 where achievement of long-term remedial objectives is difficult, complex community engagement strategies are essential because the potential risk from the contaminated groundwater is likely to persist for decades and in some cases many human generations. Informed community input can improve cleanup decisions at these complex sites, and communities that have had the opportunity to participate in the decision-making process are more likely to support, rather than challenge, key decisions. Furthermore, tools such as asymptote analysis that give communities a clear sense of the options for long-term site management can enable community members to both understand and participate in the making of highly technical decisions.
Three decades of experience at hazardous waste cleanup sites have demonstrated that the affected public responds positively when it is viewed as a partner with a stake in the outcome of cleanup activities. This partnership goes beyond simply communicating to the public what decision-makers want to do, but it must not go as far as always doing what site neighbors propose. Rather, it involves listening to stakeholder viewpoints in such a way that they feel empowered to provide constructive input into the decision-making process. The trust inherent in this interaction is the primary reason the public will support cleanup decisions (EPA, 2001b; Siegel, 2007). At complex groundwater sites where future predictions are inherently uncertain, the public tends to trust decision makers when they believe that a serious effort has been made to achieve cleanup goals and that systems are in place to ensure that any residual contamination can be effectively and safely managed.
Most regulatory programs provide for some level of community involvement, but CERCLA is the most explicit, with community acceptance being the ninth criterion for remedy selection under the National Contingency Plan. It does not give members of affected communities—even if they all agree—direct authority over the selection of remedies, but it legally recognizes that they have a significant role to play. At Superfund facilities and at projects addressed under other federal and state authorities, community members are routinely invited to comment on proposed plans, in writing and sometimes at public meetings, or to indicate their preferences
among remedial alternatives. Federal agencies—notably EPA, DoD, and the Department of Energy—have gone further and promoted the formation of site-specific advisory boards. The Defense Department currently sponsors 265 Restoration Advisory Boards at active, closed, and former installations, while DOE hosts local boards at eight major cleanup sites (DoD, 2011; DOE, 2011). Similarly, 68 EPA Community Advisory Groups facilitate information exchange among the local community, EPA, state regulators, and other federal agencies regarding the remediation of Superfund facilities (EPA, 2011f). Other similar advisory groups have been formed under state and local auspices. These community-based organizations are advisory and do not hold statutory decision-making authority. Yet where they function properly, they provide site neighbors with the opportunity to exert real influence over cleanup decisions.
Under the classic model of public input, agencies develop a small number of remedial options, selected a preferred alternative, and ask the public what it prefers. In most cases, key decisions, such as the establishment of remedial objectives, have already been made. Judgments on future land uses, which often influence remedial options, have also been developed. It is difficult for the public to do anything more than push for the expenditure of more money. On the other hand, with advisory groups engaged community members can help make the early decisions—about remedial objectives or anticipated land use—that influence future decisions (EPA, 2011g). EPA has long-standing guidance (EPA, 2011g) to encourage community involvement at the site-characterization stage of the process even though the statute does not include explicit requirements for community involvement activities at this stage.
For community involvement to be successful, a great deal of effort needs to go into building trust among the parties, and community members need to be educated on the technical and process aspects of cleanup. Ideally, they have access to their own, independent technical consultants. The most effective of these consultants not only translate official documents and the underlying science into a form more easily understood by the public, but they are in a position to focus public concerns on decisions where they can make a difference. A number of programs provide independent technical assistance to community groups, the largest and oldest of which is EPA’s Technical Assistance Grant (TAG) program. Since 1988, 323 TAGs have been awarded (205 providing $50,000 or less and 15 providing more than $250,000) (Catalogue of Federal Domestic Assistance, 2011).
Once initial remedial decisions are made, advisory board members are likely to show less interest in meetings, and agencies may be unwilling to support regular meetings with no clear agenda. There are few examples of effective public participation in the reopening of remedial decisions. All this may suggest a reduced role for the public once a site has reached the
stages found in Figure 7-2. The Committee believes that just the opposite is needed, because at these complex sites if remedies fail to achieve objectives or become unprotective many important decisions will have to be made. Reconvening dormant advisory groups just prior to the initial five-year review would allow the public to gather and digest new information about the remedy, understand site complexities that are preventing the attainment of remedial objectives, and frame sophisticated responses to the questions posed in Figure 7-2.
This reengagement of the public is supported by CERCLA, which states that EPA must provide an opportunity for a public meeting prior to the “adoption of any plan for remedial action” (underlining added) (42 U.S.C. § 9617(a)). EPA policy supports “going beyond the letter of the law” and recommends the implementation of additional community involvement activities not required by the NCP (EPA, 2005). In light of this policy and the benefits, it is reasonable to interpret the statute as allowing, if not requiring, a public meeting whenever there is a major change to the remedial plan, e.g., a decision that additional work is needed to ensure the protectiveness of the remedy as a result of a five-year review, or a decision to transition to more passive long-term management.
One example of where a community advisory body has been engaged over the long term is at the Moffett-MEW Regional Groundwater Plume in Mountain View, California (see Appendix B). At this site, EPA is leading a focused feasibility study to consider ways to accelerate groundwater remediation in response to a five-year review that found a long-standing pump-and-treat remedy to be unprotective. The local Community Advisory Board developed a list of remedial priorities and offered its own screening of remedial technologies, and EPA has incorporated the community’s ideas into its study. One of the reasons that community involvement was sustained long after the 1998 installation of the regional groundwater remedy is that other issues became pressing, such as wetlands cleanup, the emergence of vapor intrusion from the regional plume into an adjacent residential area, and the preservation of historic Hangar One at Moffett Field.
Another very important aspect of community involvement in the process found in Figure 7-2 is that communities should have a role to play in establishing and overseeing institutional controls. Because institutional controls may impede a community’s future groundwater, building, and land uses, EPA Regions are already directed to “ensure communities have meaningful opportunity to review proposals for site remedies and provide adequate information to allow informed public comment regarding the choices between cleanup alternatives that either achieves levels that allow for unrestricted use, or leave levels that lead to restricted uses and rely on institutional controls” (EPA, 2010b, p. 9). Once engineering and institutional controls are in place, informed neighbors can be in a position to
provide valuable information on possible failures that might otherwise go unnoticed by regulators who seldom visit the site (EPA, 2010b). In fact, EPA already supports a larger role for local residents, community associations, and interested organizations (representing those who live or work near the site and thus have a vested interest in ensuring compliance with institutional controls) (EPA, 2010b).
EPA’s five-year review guidance (EPA, 2001a) recognizes the need for some level of community notification at the beginning and end of the five-year review process, with optional additional engagement at high-profile sites. The guidance suggests contacting local health and other government officials, community organizations, members of advisory groups, and technical assistance grant recipients at these stages. Historically, there have been few instances where the public has played an active role in the five-year review process, particularly at sites or facilities where remedies are all in place. The challenge is to devise mechanisms for long-term public involvement at sites where people think they are “done.” First, it is essential that PRPs or regulators retain and regularly update lists of interested members of the public, and local media should be encouraged to publish annual site updates. Where advisory boards have been meeting, it may be useful to convene annual reunions to review site progress. At sites where most engaged community members frequently use the Internet or social networks, discussion forums should be encouraged. At some sites, it may prove helpful to move the locus of public involvement to permanent local institutions such as public health departments. All of these vehicles could be “re-activated” at the beginning of each five-year review cycle.
At many complex sites, contaminant concentrations in the plume remain stalled at levels above cleanup goals despite continued operation of remedial systems. At such sites, which may number in the tens of thousands, there is no clear path forward to a final end state, such that money continues to be spent with no concomitant reduction in the risks posed by the sites. If the effectiveness of site remediation reaches a point of diminishing returns prior to reaching cleanup goals and optimization has been exhausted, the transition to MNA or some other active or passive management should be considered using a formal evaluation. This transition assessment, which is akin to a focused feasibility study, would determine whether a new remedy is warranted at the site or whether long-term management is appropriate (see Figure 7-2). This analysis, particularly the graphical demonstration that contaminant concentrations at a site have reached an asymptote, is likely to enhance public involvement in long-term decision making and thus strengthen public confidence in the decisions made.
As part of the transition assessment, the costs of remedial options should be updated, ideally using probabilistic methods, to reflect the current present value of costs in order to more effectively compare the costs of various future alternatives. To the extent possible, the marginal cost of remediation through time should be quantified. Also as part of the transition assessment, formal post-remedy risk assessment should be conducted to compare alternatives, including long-term active or passive management. This risk analysis should explicitly consider uncertainty and take into account both individual and population risks, where appropriate, such that the relative risk reduction of the alternatives can be determined.
Five-year reviews are an extremely valuable source of field data for evaluating the performance of remedial strategies that have been implemented at CERCLA facilities and could be improved. To increase transparency and allow EPA, the public, and other researchers to assess lessons learned, more should be done, on a national basis, to analyze the results of five-year reviews in order to evaluate the current performance of implemented technologies. EPA’s technical guidance for five-year reviews should be updated to provide a uniform protocol for analyzing the data collected during the reviews, reporting their results, and improving their quality.
Informed public involvement characterized by two-way communications and that builds trust is a key element of good remedial decision making, particularly at complex sites. Federal agencies generally go beyond the minimum level of public involvement by forming or recognizing advisory groups and providing technical assistance, such that members of affected communities have the opportunity to influence remedial decision making. Nonetheless, public involvement tends to diminish once remedies at a site or facility are in place. No agency has a clear policy for sustaining public involvement during long-term management.
Regulators and federal responsible parties should work with members of existing advisory groups and technical assistance recipients to devise models for ongoing public oversight once remedies are in place. Such mechanisms may include annual meetings, Internet communications, or the shifting of the locus of public involvement to permanent local institutions such as public health departments. Such involvement is critical to the success of the processes found in Figure 7-2.
The cost of new remedial actions may decrease at complex sites if more of them undergo a transition to passive long-term management. However, long-term management at complex sites will still demand substantial long-term funding obligations. Failure to fund adequately the long-term manage-
ment of complex sites may result in unacceptable risks to the public due to unintended exposure to site contaminants.
Catalogue of Federal Domestic Assistance. 2011. Superfund Technical Assistance Grants (TAG) for Community Groups at National Priority List (NPL) Sites. https://www.cfda.gov/index?s=program&Mode=form&ab=step1&id=a83b10849f1dc74453b499c5d9e0370f.
Clean Sites. 1990. Improving Remedy Selection: An Explicit and Interactive Process for the Superfund Program.
Conger, J. 2011. Memo: New goals for DERP. July 18, 2011.
DoD (Department of Defense). 2011. Defense Environmental Programs Annual Report to Congress, FY 2010, p. 54. http://www.denix.osd.mil/arc/upload/FY10DEPARC_06_Restoration_Final_Report-2.pdf.
DOE (Department of Energy). 1998. Accelerating Cleanup: Paths to Closure. DOE/EM-0362.
DOE. 2011. About the EM Site-Specific Advisory Board. http://www.em.doe.gov/stakepages/ssababout.aspx.
EPA (U.S. Environmental Protection Agency). 1999. Use of Monitored Natural Attenuation at Superfund, RCRA Corrective Action, and Underground Storage Tank Sites. EPA/540/R-99/009. Washington, DC: EPA Office of Solid Waste and Emergency Response.
EPA. 2001a. Comprehensive Five-Year Review Guidance. Washington, DC: EPA Office of Solid Waste and Emergency Response.
EPA. 2001b. Stakeholder Involvement & Public Participation at the U.S. EPA: Lessons Learned, Barriers, & Innovative Approaches. EPA-100-R-00-040. http://www.epa.gov/publicinvolvement/pdf/sipp.pdf.
EPA. 2004. Cleaning up the nation’s waste sites: Markets and technology trends. Washington, DC: EPA Office of Solid Waste and Emergency Response.
EPA. 2005. Superfund Community Involvement Handbook. EPA 540-K-05-003. Washington, DC: EPA Office of Solid Waste and Emergency Response. http://www.epa.gov/superfund/community/cag/pdfs/ci_handbook.pdf.
EPA. 2007. Monitored Natural Attenuation of Inorganic Contaminants in Ground Water Volume 1—Technical Basis for Assessment. EPA/600/R-04/027. Washington, DC: EPA Office of Solid Waste and Emergency Response.
EPA. 2010a. Five Year Review for Anniston Army Depot, SIA OU1.
EPA. 2010b. Institutional Controls: A Guide to Planning, Implementing, Maintaining, and Enforcing Institutional Controls at Contaminated Sites. EPA-540-R-09-001. Washington, DC: EPA Office of Solid Waste and Emergency Response. http://www.epa.gov/superfund/policy/ic/pdfs/PIME-IC-Guidance-Interim.pdf.
EPA. 2011a. Presumptive Remedy for CERCLA Municipal Landfill Sites. http://www.epa.gov/superfund/policy/remedy/presump/clms.htm.
EPA. 2011b. Groundwater Road Map Recommended Process for Restoring Contaminated Groundwater at Superfund Sites. OSWER 9283.1-34. Washington, DC: EPA OSWER.
EPA. 2011c. An Approach for Evaluating the Progress of Natural Attenuation in Groundwater. EPA 600/R-11/204. Ada, OK: EPA Office of Research and Development.
EPA. 2011d. Stronger Management Controls Will Improve EPA Five-Year Reviews of Superfund Sites Report No. 12-P-0251. Washington, DC: EPA Office of Inspector General.
EPA. 2011e. Memorandum from Mathy Stanislaus, Assistant Administrator for Solid Waste and Emergency Response to Carolyn Copper, Director for Program Evaluation, EPA Office of Inspector General, Re: Response to the Office of Inspector General Final Evaluation Report, “Stronger Management Controls Will Improve EPA Five Year Reviews of Superfund Sites (Report No. 12-P-0251) (April 18, 2012).
EPA. 2011f. Technical Assistance Services for Communities. EPA 540-K-08-001. http://www.epa.gov/superfund/community/tasc/pdfs/tasc_brochure_english.pdf.
EPA. 2011g. Community Involvement Initiative, Action 1: Draft Work Plans, Superfund Remedial Process Improvements at 4. http://www.epa.gov/oswer/engagementinitiative/superfund.pdf.
EPA Region 1. 2010. First Five-Year Review Report for the Solvents Recovery Service of New England, Inc. Superfund Site. Boston, MA: EPA Region 1.
ESTCP (Environmental Security Technology Certification Program). 2011. Alternative Endpoints and Approaches Selected for the Remediation of Contaminated Groundwater.
GAO (Government Accountability Office). 2009. Hazardous Waste Sites: Improved Effectiveness of Controls at Sites Could Better Protect the Public. http://www.gao.gov/products/GAO-05-163.
Hayes, T. D., D. G. Linz, D. V. Nakles, and A. P. Leuschner, eds. 1996. Management of Manufactured Gas Plant Sites, Vol. 2. Amherst, MA: Amherst Scientific Publishers.
ITRC (Interstate Technology and Regulatory Council). 2010. Technical/Regulatory Guidance: A Decision Framework for Applying Monitored Natural Attenuation Processes to Metals and Radionuclides in Groundwater. December 2010.
Malcolm Pirnie. 2006. Technical Impracticability Evaluation (Draft), Southeast Industrial Area, Anniston, Army Depot, Anniston, Alabama.
Pham, T. 2010. Assessment of Failures in Long-Term Management of Contaminated Sites. Worchester Polytechnical Institute. March 10, 2010.
Probst, K. 1999. Institutional Controls: The Next Frontier. Washington, DC: Resources for the Future. http://www.wpi.edu/Pubs/E-project/Available/E-project-032410-230908/unrestricted/Final_-_Failures_at_Remediated_Sites.pdf.
SAIC (Science Applications International Corporation). 2005. Comprehensive Groundwater Remedial Investigation, Southeast Industrial Area, Phase III, at Anniston Army Depot, Anniston, Alabama (Draft).
Siegel, L. 2007. Communities and Chemical Warfare Material Disposal. Mountain View, CA: Center for Public Environmental Oversight. http://www.cpeo.org/pubs/C&CWMD.pdf.
Tetra Tech. 2011. Focused Feasibility Study for Southeast Industrial Area OU1 of ANAD, Draft, May 2011.
USGS (U.S. Geological Survey). 2007. A Framework for Assessing the Sustainability of Monitored Natural Attenuation. Circular 1303. 2007.
Yonkers, T. 2011. Policy for Refocusing the Air Force Environmental Restoration Program. February 24, 2011.