Sediment Dredging at Superfund Megasites: Assessing the Effectiveness (2007)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

6 Dredging at Superfund Megasites: Improving Future Decision-Making INTRODUCTION The preceding chapters discussed sediment management and dredging at Superfund megasites and included sections on assessing the effectiveness of dredging for removing contaminated sediment to attain remedial-action objectives and achieve specified cleanup levels. The as- sessment included the review of 26 projects from which general conclu- sions were developed with respect to the appropriate use and limitations of dredging in meeting risk-based goals. From those conclusions, the committee developed guidelines with respect to favorable site conditions under which dredging should be more likely to achieve long-term reme- dial-action objectives. The committee also offered recommendations for monitoring to facilitate scientifically based and timely decision- making to improve dredging effectiveness. In this final chapter, the committee addresses the charge in the statement of task to consider “how conclusions about completed and current operations can inform future remedial decision-making” and to “develop recommendations that will facilitate scientifically based and timely decision making for megasites in the future.” Specifically, we seek to identify how lessons learned from experience may inform future prac- 240 

Improving Future Decision-Making 241 tices and management of contaminated sediment at megasites. This in- cludes the expected role of dredging in the future and the issues and fac- tors that need to be addressed to ensure the effective use of dredging as a component of contaminated sediment remediation. Most of the commit- tee’s earlier recommendations focus on these issues at the site-specific level, but this chapter focuses on the national level. MANAGING SEDIMENT MEGASITES IN THE FUTURE With the establishment of Superfund in 1980, we now have the op- portunity for retrospective analysis at dozens of sediment sites to evalu- ate decision-making, field experience, and remedial effectiveness where dredging has been selected for sediment cleanup. In the past, a rigorous evaluation of whether site remediation achieved risk reduction goals and what factors contributed to or limited the achievement of those goals was often just not done. Although information is available from various sites with respect to volume of bulk material removed or sediment con- centration achieved, that information does not permit determination of the degree to which remedial objectives for risk-reduction were achieved. Thus, it is not easy to determine which approaches resulted in risk reduction under various site conditions. The difficulty stems partly from the lack of comprehensive post-dredging monitoring data and from the fact that followup assessments typically do not quantify uncertainty in both risk measurements and predictions. In hindsight, it is clear that there are limitations to dredging effec- tiveness. With this historical perspective comes the opportunity to learn and improve how we think about and implement environmental dredg- ing. Perhaps nothing is more important than to step back and derive common lessons from experience, as was done in Chapter 4. This type of review needs to be continuous and needs to part of a shared experience among regulators, practitioners, and the public. As described in Chapter 2, sediment megasites are among the most challenging and costly sites on the National Priorities List (NPL). Megasites are conventionally defined as sites with remedial activities costing at least $50 million; the U.S. Environmental Protection Agency (EPA) has defined contaminated sediment megasites as sites for which the sediment component of remedial activities will cost at least $50 million. 

242 Sediment Dredging at Superfund Megasites The charge to this committee focused on megasites, but the dearth of such sites with completed dredging remedies and with good pre- dredging and post-dredging data has meant that the committee re- viewed smaller sites, or individual projects at megasites. The projects evaluated did not include any of the magnitude (that is, tens of miles of river stretches and thousands of hectares) and time frames (for example, decades in the Hudson River [TAMS Consultants, 2000]) that can be an- ticipated at the largest of the current and future megasites. Megasites with a broad spatial area and large volume of contaminated sediments will likely require multiple seasons of dredging.1 The larger scale and time frames will increase the chemical exposures and residual produc- tion related to operations, and make it more difficult to fully characterize contaminant distribution, sources of contamination, and conditions un- favorable for remedial operations. At these sites, risk based goals may not be achievable in the foreseeable future due to the long time frame, complexity of the sites, and the limitations of available technologies. The committee recognizes that experience with remediation at larger sites might reveal challenges not faced at the smaller sites and has attempted to anticipate such challenges to the extent possible in making its recom- mendations. Cleanup of contaminated sediment megasites incorporates large temporal and spatial scales that create two distinct issues: the human health and ecosystem risk-reduction benefits achieved by isolated reme- diation in a large-scale watershed are difficult to predict and quantify; and the large spatial scales and long time lines, coupled with the com- plexity and heterogeneity of large-scale megasites, suggest that varied and combined remedial approaches will be appropriate. We can do a better job of addressing those issues by taking a broader, basin-wide view in contaminated sediment management and by embracing more flexible approaches. Those issues are discussed further below. The Need for Regional-Scale Perspectives Because contaminated sediment megasites are influenced by re- gional-scale phenomena, watershed and airshed contributions to sedi- For example, at current levels of operation, it will take more than 25 years to 1 complete dredging at New Bedford (Dickerson and Brown 2006). 

Improving Future Decision-Making 243 ment contamination at any site must be viewed in a larger framework to permit valid predictions about cleanup and risk reduction. Sediment megasites can span an entire waterbody (such as the Lower Fox River in Wisconsin) or be located in a watershed amongst other contaminated sediment sites. Bridges et al (2006) comment that several watersheds in the United States contain multiple contaminated sediment sites in close proximity to one another and that effective sediment management will require a more holistic approach to understanding multiple sources of contaminants (sediments, outfalls, and non-point sources) and their cu- mulative impacts in a waterway. The public has a right to know what benefits will be achieved for particular investments. For example, if the risk being addressed is asso- ciated with the consumption of fish, a valid question is, How much will contaminants in fish decrease as a result of this action? Some organisms travel great distances and there is need to understand their movement and variable exposure to the Superfund sites. At the same time, the long- distance movement of toxics from the site throughout the larger water body needs to be understood. Finally, there needs to be an understand- ing of secular changes in basin-wide conditions and how they might re- late to cleanup at a specific site. (For example, whether basin-wide con- centrations of contaminants are declining because of point source reduction or whether contaminants are migrating in from other contami- nated areas in the wider basin.) These factors contribute heavily in evaluating site specific data on concentrations in fish species and the broader water body before and after a cleanup. As such, a regional approach to modeling and analysis of contami- nants at megasites is needed to better understand their effect on resident and migratory fish and on the flux of contaminants within the wider ba- sin (e.g. Linkov et al. 2002; von Stackelberg et al. 2002). Because of the difficulty in accurately estimating several of the necessary parameters and inputs for these types of models (particularly fish exposure to con- taminated sediments, differences in movement of various fish popula- tions or life stages, contaminant concentrations in prey, and uptake and loss kinetics of mobile species occupying areas of high and low expo- sure), their uncertainty will remain a concern. However, because these issues are of particular importance at megasites and where multiple Superfund sites exist in close proximity, the development of these mod- els is essential. 

244 Sediment Dredging at Superfund Megasites A related issue is the lack of essential tools for understanding how reductions in sediment toxicity or biologic exposure will enhance ecosys- tem response and benefit ecosystem recovery. Much of our understand- ing of these topics is wholly observational or is derived from ancillary measures (such as sediment chemistry). In itself, such information pro- vides little capability for predicting community or organism response after remediation. Understanding the ecosystem dynamics that affect recovery entails larger regional-scale phenomena, such as larval recruit- ment, food-web interactions, and fate and transport processes.2 The Use of Adaptive Management at Megasites Given the difficulty in predicting dredging effectiveness, and the limited number of available alternative technologies, what changes can be made to improve the remedy selection and implementation process to ensure more effective and cost-effective remedies? A major challenge to decision-makers is the uncertainty about whether—and how well—a remedy will work at a site. Experience has shown the wisdom of well-designed pilot field tests and experimentation prior to committing to a specific final cleanup remedy. Pilot testing, in- cluding monitoring of appropriate environmental variables, for example as part of the feasibility study, is used to test the performance of a tech- nology or approach, understand the factors affecting its performance, and to provide information on how, if necessary, the remedy should be adapted to achieve desired goals. In this way, the information generated in the pilot tests and monitoring becomes a key component of the rem- edy selection, design, and implementation process. The use of a structured process of selecting a management action, monitoring the effects of the action, and applying those lessons to opti- mize a management action is generally referred to as adaptive manage- ment (e.g., NRC 2003, 2004, 2005; Bridges et al. 2006; Linkov et al 2006a, b). As described in NRC 2004, “There is no prototype for its implementa- tion, and no ‘cookbook’-type set of steps or building blocks that will im- mediately constitute an adaptive-management program. It is context- specific, it involves feedback and learning between scientists, managers, 2 As discussed in Chapter 5, biodynamic approaches that are linked with prin- ciples of functional ecology can help to bridge this knowledge gap. 

Improving Future Decision-Making 245 and stakeholders.” NRC (2005) recommends an adaptive-management approach at Superfund megasites where it is unlikely that final remedies can be identified and implemented. That report describes adaptive man- agement as a six-step interactive process for defining and implementing management policies under conditions of high uncertainty regarding results of remedial actions (see Box 6-1). Bridges et al. (2006) described the need for greater flexibility in sediment management processes because “the more strictly linear deci- sion making process characterized as the ‘decide and defend’ approach to remedial decision making does not contain sufficient flexibility” and is unable to accommodate or benefit from other approaches such as adap- tive management. In the current Superfund process, the ROD is the end result of a long and often difficult and contentious process of conducting studies and receiving and responding to input from stakeholders often with divergent and impassioned views of the type and extent of reme- diation that is required (see Chapter 2 for greater detail on the remedy selection process). A ROD often selects a specific remedy and predicts its ability to achieve cleanup levels and remedial action objectives at the site. Because the scale of megasites is so large, a variety of unanticipated conditions can greatly influence the results of a remediation. When remedies are selected without the benefit of actual, on-the-ground feed- back on the effect of the remediation, there is a greater chance that un- foreseen conditions and events will hinder progress or limit the effec- tiveness of the remediation. The ROD process can be reopened to amend or modify a ROD on the basis of information gathered after implementa- tion begins, however, instituting an adaptive-management process from the outset recognizes the uncertainty inherent in predicting remedial results and allows adaptation of the remedy based on site experience to optimize progress toward attaining remedial goals. In this process, the primary goals of Superfund, the protection of human health and the environment, remain paramount. As such, adap- tive management is not a means to permit or sanction less rigorous cleanups, or to avoid public input or scrutiny of the decision making process. The principles of transparency and public notification remain essential and the adaptive-management process at a site needs to be de- veloped in concert with stakeholders and insights from monitoring and testing need to be shared with them so that they can contribute to 

246 Sediment Dredging at Superfund Megasites BOX 6-1 Six-Step Adaptive-management Process 1. Assessing the problem, including establishing measurable management objectives, key indicators of those objectives, quantitative or conceptual models to predict effects of remedial alternatives on the indicators, and forecasts of re- sponses of indicators to remedial actions. 2. Designing a management plan, including comparing and selecting re- medial actions and, importantly, selecting indicator values that will trigger a change in management actions. 3. Implementing the plan, including documenting and agreeing with stakeholders on those circumstances that might require deviations from the plan. 4. Monitoring for effectiveness and for verifying and updating the concep- tual model. 5. Evaluating results obtained from monitoring, including comparing re- sults with forecasts from earlier modeling, seeking to explain why results oc- curred, and provide recommendations for future action. 6. Adjusting the management plan in response to the monitoring results, including implementing recommendations, reviewing and updating models, and developing new forecasts, management objectives, and management actions as necessary. Source: Adapted from NRC 2005. adapting the remedy, if necessary. It is expected that adaptive manage- ment could be implemented in the current legislative framework because CERCLA and the NCP have great flexibility and do not preclude adap- tive management (NRC 2005). That implementation would need to be reviewed by EPA to best fit CERCLA requirements. There is progress toward implementing adaptive management at contaminated sediment sites. Recent EPA guidance (EPA 2005) endorses the general concept, stating Project managers are encouraged to use an adaptive management approach, especially at complex sediment sites to provide addi- tional certainty of information to support decisions. In general, this means testing of hypotheses and conclusions and reevaluating site assumptions as new information is gathered. 

Improving Future Decision-Making 247 There are also examples of sites where adaptive-management prin- ciples, if not an explicit, rigorous adaptive-management process, have been applied in remediating contaminated sediment sites (see Box 6-2). In sum, the desired outcome of this more flexible approach is to al- low, indeed to encourage, adaptation to realities on the ground in an ef- fort to achieve remedial goals in as efficient and cost effective manner as possible. These suggested changes reflect the need to make decisions in the face of uncertainty while allowing managers and stakeholders to re- spond to, and take advantage of, unanticipated events and a variety of possible future outcomes through the design of a flexible, iterative learn- ing process. BOX 6-2 Examples of the Application of Adaptive-management Principles in Sediment Remediation In the Fox River, WI, two demonstration projects were conducted during the Remedial Investigation/Feasibility Study at Sediment Management Units 56/57 and Deposit N (Foth and Van Dyke, 2000; Montgomery Watson, 2001). The projects provided useful information on implementability, effectiveness, and expense of large-scale dredging at the site and were used to inform future deci- sion-making. In Operable Unit (OU) 1 of the Fox River, the ROD permits flexibility in achieving cleanup levels and stipulates additional actions (further dredging or capping) if dredging doesn’t achieve desired results. Following dredging, if sam- pling shows that the 1 ppm action level has not been achieved, a surface- weighted average concentration (SWAC) of 0.25 ppm may be used to assess the effectiveness of PCB removal. If that SWAC of 0.25 ppm has not been achieved for OU 1, the first option is that additional dredging may be undertaken to en- sure that all sediments with PCB concentrations greater than the 1 ppm action level are removed throughout the particular deposit. A second option is placing a sand cover on dredged areas to reduce surficial concentrations to achieve a SWAC of 0.25 ppm for OU 1 (WI DNR/EPA 2002). Finally, in the case of the Grasse River, several large-scale dredging and capping projects have revealed site-specific conditions that limited the effective- ness of the remediation (including dredging and capping). This site-specific in- formation can then be used in development of a revised Analysis of Alternatives Report (Alcoa Inc. 2005). 

248 Sediment Dredging at Superfund Megasites The Future of Dredging While some improvements have been made to dredge design and operation (for example, precision positioning systems or dredge head or bucket modifications to reduce resuspension), in many respects dredging as a technology has not changed dramatically in the last few decades. What has changed is how dredging is applied. Devices designed and proven for navigational or maintenance dredging are now pressed into service for specific and precise contaminant-mass removal or to attain specific sediment contaminant concentrations in what are often complex settings and difficult conditions. In addition, it is often difficult to accu- rately characterize the sites, and define the degree of uncertainty about the effectiveness of different remedial approaches. The committee found that most of the sites that it examined exhib- ited one or more conditions unfavorable for dredging and concluded that dredging alone is unlikely to be effective in achieving both short- term and long-term cleanup levels at many sites. However, its effective- ness as a contaminant-mass removal technology will ensure its use at most sites where mass removal is necessary (such as where navigational, source reduction, or sediment stability concerns are present). Where un- favorable conditions exist, it is likely to be implemented—in conjunction with capping, in situ treatment, or monitored natural recovery—as part of a combined remedy. In the future, dredging will continue to play an important role in the management of Superfund megasites and should be viewed as one of several approaches that may be necessary for their cleanup. CONCLUSIONS AND RECOMMENDATIONS The committee envisions that some combination of dredging, cap- ping or covering, and natural recovery will be involved at all megasites. In situ treatments may also be required at many sites. Thus, all remedial approaches should be considered in the site evaluation, and the interac- tions among the various approaches should be well understood. Dredg- ing for mass removal itself may be attractive from the viewpoint of the public, but it alone does not necessarily produce risk reduction. A better appreciation of the existing risks before dredging and what is required to 

Improving Future Decision-Making 249 achieve desired risk reduction, both in the short term and in the long term, is needed. The challenge to this committee was twofold: to make pertinent technical recommendations (contained in Chapters 4 and 5) and to rec- ommend changes in the management of the Superfund program that will ensure that the technical recommendations are implemented. The com- mittee believes that three kinds of changes are critical to improve deci- sion-making and increase dredging-remedy effectiveness at contami- nated sediment megasites. First, owing to the complexity, large spatial scale, and long time frame involved, the management of contaminated megasites should em- brace a more flexible and adaptive approach to accommodate unex- pected conditions and events, new knowledge, technology changes, and results of field pilot tests. Second, improved risk assessment should specifically consider the full range and real-world limitations of remedial alternatives to allow valid comparisons of technologies and uncertainties. Third, EPA needs a centralized focal point for coordinated assess- ment of contaminated sediment megasites for better consistency in site evaluations, remedy selection, and for increased focus and communica- tion among EPA management and technical staff on what works and why. Similar recommendations have been discussed and developed by other groups3, but it is hoped that in the aggregate the committee’s rec- 3The notion that large, complex sites need a more adaptive approach to rem- edy selection and implementation was the topic of much discussion at the meet- ings of the EPA National Advisory Council for Environmental Policy and Tech- nology Superfund Subcommittee (NACEPT 2004). The need for adaptive- management approaches at complex contaminated sites has also been discussed in the academic community for some years (for example, Cannon 2005). Various National Research Council committees and other independent reviews have ad- vocated similar approaches. For example, the National Research Council advo- cated the use of flexible phased implementation and adaptive management in environmental remediation (NRC 2001, 2003, 2005), recommended that the wide array of risks associated with implementing a remedy be explicitly considered (NRC 2001, 2005), and recommended that the limitations associated with dredg- ing and the potential for production of residual contaminated sediment be con- 

250 Sediment Dredging at Superfund Megasites ommendations add more specificity than past efforts regarding the effec- tive remediation and management of contaminated sediment. The three recommendations, which are described in more detail below, will in some cases require additional resources and, equally challenging in large organizations, new ways of doing business. The committee cannot stress enough that because of the potentially huge cost and the complexity of sediment megasites, the costs and efforts required to change standard operating procedures are worth the up-front investment that will be re- quired. As noted, many times cleaning up sediment megasites sites may take decades from investigation to cleanup and cost hundreds of millions of dollars. The cost of implementing the recommendations in this report should be viewed in that context. In fact, the committee is concerned that if its recommendations are not implemented, many hundreds of millions of dollars of government and private funds will be wasted on ineffective remedies for contaminated sediment megasites. 1. An adaptive management approach is essential to the selection and implementation of remedies at contaminated sediment megasites where there is a high degree of uncertainty about the effectiveness of dredging. If there is one fact on which all would agree, it is that the selection and implementation of remedies at contaminated sediment sites are complicated. Many large and complex contaminated sediment sites will take years or even decades to remediate and the technical challenges and uncertainties of remediating aquatic environments are a major obstacle to cost-effective cleanup. Because of site-specific conditions—including hydrodynamic set- ting, bathymetry, bottom structure, distribution of contaminant concen- trations and types, geographic scale, and remediation time frames—the remediation of contaminated sediment is neither simple nor quick, and the notion of a straightforward “remedial pipeline” that is typically used to describe the decision-making process for Superfund sites is likely to be at best not useful and at worst counterproductive. sidered (NRC 1997, 2001). Yet, at the time of the present review, little progress has been made in implementing those recommendations. 

Improving Future Decision-Making 251 The typical Superfund remedy-selection approach, in which site studies in the remedial investigation and feasibility study establish a sin- gle path to remediation in the record of decision, is not the best approach to remedy selection and implementation at these sites owing to the in- herent uncertainties in remedy effectiveness. At the largest sites, the time frames and scales are in many ways unprecedented. Given that remedies are estimated to take years or decades to implement and even longer to achieve cleanup goals, there is the potential—indeed almost a certainty— that there will be a need for changes, whether in response to new knowl- edge about site conditions, to changes in site conditions from extreme storms or flooding, or to advances in technology (such as improved dredge or cap design or in situ treatments). Regulators and others will need to adapt continually to evolving conditions and environmental re- sponses that cannot be foreseen. These possibilities reiterate the importance of phased, adaptive ap- proaches for sediment management at megasites. As described previ- ously, adaptive management does not postpone action, but rather sup- ports action in the face of limited scientific knowledge and the complexities and unpredictable behavior of large ecosystems [NRC, 2004]. 2. EPA should compare the net risk reduction associated with the various remedial alternatives, taking into account the limitations of each approach in selecting site remedies, such as residuals and resus- pension. One subject of great interest and concern at contaminated sediment Superfund sites is the risk-based comparison of remedial alternatives to support selection of a remedy (Bridges et al. 2006; Wenning et al. 2006). The committee was charged only with evaluating the effectiveness of dredging and not with comparing the effectiveness of remedial alterna- tives. However, the committee recognizes that the effectiveness of a dredging remedy depends on good planning, and good planning in- cludes an evaluation of net risk reduction associated with each remedial alternative. Therefore, the committee recommends evaluating the net risk reduction of remedial alternatives to facilitate scientifically based decision making at megasites. 

252 Sediment Dredging at Superfund Megasites Baseline risk is quantified in the remedial investigation for all NPL sites, but the feasibility study may or may not include a quantitative es- timate of the risks posed by alternative remedies. EPA (2005, p. 2-14) in- dicates that “although significant attention has been paid to evaluating baseline risks, traditionally less emphasis has been placed on evaluating risks from remedial alternatives, in part because these risks may be diffi- cult to quantify.” Even if such quantitative comparative risk assessment is provided, it might be limited in scope. For example, the feasibility study for the Upper Hudson River (TAMS Consultants 2000) included a quantitative comparison of human health and ecologic risk reduction for the fish-consumption exposure pathway, the pathway associated with the highest risk estimates. However, the analysis did not quantify short- term effects on the local community or workers or other effects that might occur during dredging; it concluded that “there is no reliable means of quantifying potential short-term impacts from activities such as sediment resuspension, habitat loss, or other transient effects.” Each remedial alternative offers its own set of risk-reduction bene- fits and possibly the creation of new exposure pathways and associated risks. A confounding issue is that site conditions can change in ways that help to reduce risk. That would be the case, for example, with deposition of cleaner material over residual contamination. Site-specific measure- ments and models need to incorporate an understanding of such site fea- tures both spatially and temporally to support valid comparisons of re- medial alternatives. Environmental responses to remediation, including sediment and biota concentration changes, are complex and difficult to predict. During remedy selection, the uncertainty around estimates of responses to remediation should be recognized and quantified to the extent war- ranted to optimize decision-making. For example, EPA established a tiered approach to probabilistic risk assessment in the Superfund pro- gram, as shown in Figure 6-1 (EPA, 2001). Using that approach, one pro- ceeds from a less expensive point estimate sensitivity analysis to more expensive and time-consuming quantitative uncertainty-analysis meth- ods. The question is: When are the more advanced methods useful or necessary? Box 6-3 illustrates a situation in which additional quantitative analysis might be warranted. It presents an idealized comparison of risk estimates, including inherent uncertainty, for two remedial alternatives. 

Improving Future Decision-Making 253 Tier 3 Advanced PRA 2-D MCA Probabilistic Sensitivity Analysis (Microexposure Modeling, Bayesian Statistics, Geostatistics)a Characterization of Variability and/or Uncertainty Increasing Complexity/Resource Requirements Complete RI/FS Process Tier 2 PRA 1-D MCA Probabilistic Sensitivity Analysis Tier 1 Point Estimate Risk Assessment Point Estimate Sensitivity Analysis Problem Formulation/Scoping/Work Planning/Data Collection = Decision Making Cycle: Evaluation, Deliberation, Data Collection, Work Planning, Communication At each tier, a decision may be to exit the tiered process aExamples of advanced methods for quantifying temporal variability, spatial variability, and uncertainty. FIGURE 6-1 EPA’s tiered approach to the use of probabilistic risk assessment (PRA). DMCA: decision-making cycle analysis. Source: EPA 2001. Some potential effects will remain difficult to accurately quantify and compare (for example, the impact of a large dredging project on quality of life issues such as noise or light pollution) or potential psycho- logical consequences from not implementing a removal remedy (for ex- ample, if community members perceive that an unmitigated threat to human health exists in their environment). Other “implementation risks” (risks potentially imposed by the implementation of a remediation strat- egy) such as worker and community health and safety, equipment fail- ures, and accident rates associated with an active remediation are given 

254 Sediment Dredging at Superfund Megasites little consideration in EPA’s feasibility studies at Superfund sites (Wen- ning et al. 2006). Cura et al. (2004) identify several challenges associated with comparative risk assessment, given data limitations and the un- avoidably subjective nature of quantifying some risks associated with dredged-material management decisions. However, ignoring those types of risk in comparisons of remedial options is not the solution and may have undesirable consequences, particularly when the cost of being wrong is high (Bridges et al. 2006). BOX 6-3 Importance of Quantifying Uncertainty in Risk Estimates In the hypothetical case outlined in the figure below, remedial alternative 1 appears to result in lower risk than alternative 2. If the uncertainty in these risk estimates is not quantified in some way, a risk manager might proceed with al- ternative 1. However, the uncertainty in this estimate is sufficiently high that one cannot be certain about this conclusion, and in fact a higher risk might result from implementing alternative 1. Such an outcome is obviously not desirable and even more problematic if alternative 1 is the more costly remedial alterna- tive. With the benefit of the quantitative uncertainty analysis, and depending on the magnitudes of risk estimates and remedy costs, a risk manager might elect to gather more data (for example, with a pilot field test) to reduce uncertainty in the risk estimate for remedial alternative 1 before making a selection. A quantita- tive uncertainty analysis can reveal significant contributors to uncertainty in risk estimates, which are the most useful subjects of further study and data collec- tion. Error bars reflect Best uncertainty Estimate about best of Risk estimate of risk Remedial Remedial Alternative Alternative #1 #2 Hypothetical comparison of risk predicted for two remedial alternatives, including quanti- fication of uncertainty associated with the risk estimates. 

Improving Future Decision-Making 255 3. There is a great need for centralized EPA resources, responsi- bility, and authority at the national level to ensure that necessary im- provements are made so that contaminated sediment megasites are remediated as effectively as possible. As discussed in Chapter 2 and elsewhere in this report, it became abundantly clear during the committee’s work that EPA has not devoted adequate resources and senior management attention to the issue of con- taminated sediment, given the scope of the problem and the huge costs incurred by the federal government, the private sector, and others. If the recommendations in this report (and the many good reports that have gone before) are to be successfully implemented, some group in the Superfund program should be given the resources and responsibility to make needed changes and should then be held accountable. EPA is in the best position to gather and evaluate relevant data on a national level, so it is natural for EPA to lead the effort in monitoring the progress and sharing experiences on dredging at megasites. Because every EPA region has on-the-ground experience with dredging at some megasites, regular review and shared experience can inform decision- making and raise the overall level of technical expertise. Whether by a more robust Contaminated Sediments Technical Advisory Group or some other mechanism, a consistent set of design and monitoring princi- ples should emerge and grow from such efforts. Such information should be publicly available. The goal is generating a greater under- standing of sound remediation principles and best practices and their uniform application among sites. The difficulty that the committee had in obtaining information about Superfund contaminated sediment sites and the lack of consistent data on those sites point to a need for a much stronger national program that has the authority and responsibility for overseeing and evaluating EPA’s Superfund contaminated sediment ef- forts. The recommendations made here and by many earlier independent evaluations are unlikely to be implemented by the current patchwork approach to managing contaminated sediment sites. Resources, author- ity, and strong leadership are needed to ensure that the recommenda- tions in this and prior reports are implemented in a timely manner. It is impossible to identify a focal point for contaminated sediment sites in the current Superfund office organizational structure. Yet those sites are among the most challenging and expensive sites on the NPL. 

256 Sediment Dredging at Superfund Megasites Years of experience suggests that to garner the needed resources, focus, and management attention for a problem of this magnitude, it is neces- sary not only to create a “critical mass” of personnel and expertise but to clearly identify those responsible and accountable for implementing needed changes and policies. The committee strongly recommends that this gap be addressed. Specific responsibilities include the following: • Gather data to define the scope of the contaminated sediment problem. • Track current and likely future contaminated sediment megasites that are on the NPL and in other EPA programs. • Review site studies, remedies, and monitoring approaches at contaminated sediment megasites to assess whether best practices are being implemented, including whether regions are complying with na- tional sediment and other program guidance. • Ensure that adaptive-management approaches are applied at contaminated sediment megasites where there is substantial uncertainty about the effectiveness of dredging and other remedial approaches. As part of this effort, it is critical that EPA staff communicate clearly to local citizens and other stakeholders objective information about what dredg- ing and other remedial options can and can not accomplish, as well as inform them about the inherent uncertainties of remedy effectiveness at sediment sites. For an adaptive-management approach to be successful, public involvement should occur “early and often.” • Ensure adequate pre- and post-remediation monitoring at com- plex contaminated sediment sites. • Evaluate the effectiveness of sediment remediation in near “real time” at major sediment cleanup projects to determine whether selected remedies are achieving their intended goals and to develop lessons learned. • Create a centralized, easily accessible, and up-to-date repository of relevant data and lessons learned regarding sediment remedies, in- cluding dredging and other approaches to facilitate information transfer among regional and headquarters staff working on these sites and the public. • Develop and implement a research strategy for evaluating ways to improve the assessment, monitoring, and cleanup of contaminated 

Improving Future Decision-Making 257 sediment sites, including the development and testing of new technolo- gies. • Serve as a focal point for coordination and communication among the many EPA programs and federal agencies who are involved in the cleanup of contaminated sediment sites. For these functions to be implemented, EPA Headquarters program staff will need to review and provide input to regional decisions, and present senior program managers with data and information about rem- edy effectiveness and the approaches tried at different sites on an on- going (and real time) basis. This will require the commitment of senior mangers in EPA Headquarters and the 10 EPA regional offices to work together. Some of the specific tasks that will be needed to implement the above functions are described in more detail below. EPA should define the scope of the problem. One of the necessary tasks will be to define the scale of the megasite-sediment problem. As noted in Chapter 1 of this report, EPA has attempted to define the extent of contaminated sediment in the United States since at least the 1970s. The latest report (EPA 2004), based on the National Sediment Inventory (NSI) database, surveys about 9% of the water-body segments in the United States and classifies 43% of this nonrandom sample as having probable associated adverse effects. How- ever, EPA’s efforts have fallen short of the systematic assessment needed to define the scope of contaminated sediment that may require remedial action.4 Even at Superfund sites, EPA’s efforts to determine the geo- graphic extent and volume of contaminated sediment appears episodic, 4 Similarly, a recent EPA Office of the Inspector General report (EPA 2006, P.19) concluded that “EPA’s 2004 National Sediment Quality Survey report did not provide a complete assessment of the extent and severity of sediment con- tamination across the Nation, nor did it fully meet the requirements of the Water Resources Development Act. . . . As a result, EPA cannot accurately estimate the volume and risks posed by contaminated sediments on a national scale. Such a national assessment would better enable EPA to ensure that it devotes resources to contaminated sediment issues that pose the greatest risks to human health and the environment.” 

258 Sediment Dredging at Superfund Megasites and, as described in Chapter 2, there is no current list of contaminated sediment sites, nor does the Agency evaluate new NPL sites when they are listed to develop a “watch list” of those sites that are likely to be fu- ture megasites. In this regard, conclusions of one of EPA’s earliest re- ports on the subject (EPA 1987) still holds true: “Although it is reason- able to say that there is significant in-place contamination in U.S. waters, it is not possible with the current level of knowledge to quantify the problem. We do not know and cannot even begin to estimate, for exam- ple, the river miles affected or the cubic yards of sediment involved.” Of course EPA can not and should not wait until it has compiled a definitive picture of the contaminated sediments problem in the United States to move forward with site cleanups. Defining the scope of the sediment problem is important for two reasons. First, it will help to place the magnitude of the problem in proper perspective to help in understanding how much of the problem has already been addressed and how much remains to be done. A con- crete goal for EPA should be to have an on-going process of evaluating newly listed NPL sites, as well as major contaminated sediment sites ad- dressed by other programs, to understand the magnitude and severity of contaminated sediments. From those evaluations, the agency should produce a report that describes the number of past, active, and probable future contaminated sediment Superfund sites and the number of likely megasites. The report should describe the types of contaminants and the volumes of contaminated sediment and lay out an estimate of likely fu- ture costs of cleanup and long-term monitoring. This kind of information was not available from EPA, which made it difficult to understand the scope of the problem. Second, documenting how much work remains to be done and at what cost should help senior EPA management and other officials to identify the most pressing program and research needs. For example, if many more site remedies remain to be executed or listed on the NPL, it makes sense to invest in developing new technologies for remediation. If few remedies remain to be chosen, then it may be that developing moni- toring tools is of greater importance. All the recommendations in this report will take staff time and money to implement. By clearly defining the scope of the problem, EPA management will have the information it needs to identify the most im- 

Improving Future Decision-Making 259 portant tasks to accomplish the goal of improving the scientific basis of selecting the most effective remedies for contaminated sediment sites. EPA should ensure that adequate monitoring strategies are implemented. As highlighted in Chapters 4 and 5, one difficulty in understanding the effectiveness of dredging is that statistically valid baseline pre- dredging condition assessments are generally not done. Without ade- quate pre-dredging and post-dredging monitoring, it is impossible to make the valid comparisons that are necessary to support definitive statements about the degree to which remedial objectives have been at- tained as a result of dredging. Much greater attention should be given to sufficient monitoring to allow valid statistical comparisons of conditions before and after dredging. That will require considerable forethought in sampling design, sampling methods, and analytic techniques. The long period from site investigation through remedial action to required 5-year reviews compounds the problem. All that points to the need for EPA (and hopefully other federal agencies with a stake in this arena, for example, the Army Corps and the Navy) to invest in better and more consistent measuring tools to monitor conditions in the field more reliably and efficiently. The committee rec- ommends greater efforts to develop better methods to measure sediment stability and transport processes, biogeochemical processes, and pore water concentrations and fluxes. EPA should develop and implement a contaminated sediment research and evaluation strategy. One of the key elements of an improved sediment-cleanup pro- gram is to establish a coherent research and evaluation strategy to fill critical information gaps. In Chapter 4, the committee reached some spe- cific conclusions regarding factors that contribute to or limit dredging effectiveness. The EPA research and evaluation strategy should build on the work of the committee to ensure that experience gained in dredging in a variety of combinations and situations is translated into useful guid- ance to EPA regions and communicated to the full panoply of external stakeholders in a timely and transparent fashion. 

260 Sediment Dredging at Superfund Megasites The objective of the research and evaluation program should be to answer key questions as to what risk reduction will be achieved by dif- ferent technical approaches, under what site conditions, and with what certainty. The agency needs to answer those questions through pilot studies and data collection efforts that monitor baseline to long-term conditions and that stress robust sampling and statistical analysis. To this end, EPA should undertake or commission real-time independent evaluations of the effectiveness of dredging and other remedies at con- taminated sediment sites, especially megasites. The reviews would build on the committee’s analyses and assess the effectiveness of dredging and other remedies at all major sediment cleanup projects and seek to under- stand the factors that contributed to or limited the effectiveness of the cleanup approach. This kind of study should either be conducted by a neutral external organization (either academic or non-profit) or, if con- ducted by EPA, be made subject to external peer review. It should be clear at the outset that an external organization conducting the review will have full control of the results, and that the final report will be made publicly available. This type of systematic evaluation will require EPA’s Superfund of- fice and Office of Research and Development to work together to fill the critical gaps in guidance and standard protocols. This effort should also involve other agencies, such as the U.S. Geological Survey, the Army Corps of Engineers, the U.S. Navy, and the National Oceanic and At- mospheric Administration, who work on contaminated sediment sites. To implement a unified research and evaluation strategy successfully, EPA will need to ensure that appropriate resources are applied and that the various EPA and other government offices involved are held ac- countable for timely implementation of the strategy once it has been de- veloped. While there are only a few general approaches to sediment reme- diation, there is room for improvement in their performance. Improving and optimizing remediation systems has long been a cornerstone in en- vironmental engineering and remediation. The refinement, modification, and development of sediment remediation approaches and technologies can overcome limitations to remedial performance and improve effec- tiveness. Therefore, research to improve and develop new remediation technologies, site-characterization techniques, and monitoring tools is essential to advance sediment remediation and should be supported. 

Improving Future Decision-Making 261 Efforts to understand and promote those practices and operations that improve remediation effectiveness, and the training of decision makers and practitioners in those operations, is also critical to advance the field and improve the performance of remedial operations. In sum, EPA’s efforts should focus on moving forward with reme- dies at sites and, at the same time on investing the effort needed to make sure that each new pilot test or remedy implemented increases our col- lective knowledge of what works and what does not work and why. Be- cause many of contaminated sediment sites are vast and remediating them will be expensive, it is worth investing time and resources now to try to ensure more cost-effective remedies in the future. Such a focus is needed if the country is to make the best possible use of the billions of dollars that will be spent on site remediation. REFERENCES Alcoa Inc. 2005. Remedial Options Pilot Study Work Plan: Grasse River Study Area Massena, New York. Alcoa Inc., Massena, New York.. February 11, 2005. Bridges, T.S., S.E. Apitz, L. Evison, K. Keckler, M. Logan, S. Nadeau, and R.J. Wenning. 2006. Risk-based decision making to manage contaminated sediments. Integr. Environ. Assess. Manage. 2(1):51-58. Cannon, J.Z. 2005. Adaptive management in superfund: Thinking like a contami- nated site. New York U. Environ. Law J. 13(3):561-612. Cura, J.J., T.S. Bridges, and M.E. McArdle. 2004. Comparative risk assessment methods and their applicability to dredged material management decision- making. Hum. Ecol. Risk Assess. 10(3):485-503. Dickerson, D., and J. Brown. 2006. Case Study on New Bedford Harbor. Presenta- tion at the First Meeting on Sediment Dredging at Superfund Megasites, March 22, 2006, Washington DC. EPA (U.S. Environmental Protection Agency). 1987. An Overview of Sediment Quality in the United States. EPA-905/9-88-002. Office of Water Regula- tions and Standards, U.S. Environmental Protection Agency, Washington, DC, and Region 5, Chicago, IL. June 1987. EPA (U.S. Environmental Protection Agency). 2001. Risk Assessment Guidance for Superfund: Vol. III - Part A, Process for Conducting Probabilistic Risk Assessment (RAGS 3A). EPA 540-R-02-002. Office of Emergency and Re- medial Response, U.S. Environmental Protection Agency, Washington, DC. December 2001 [online]. Available: http://www.epa.gov/oswer/riskassess ment/rags3adt/index.htm [accessed May 14, 2007]. 

262 Sediment Dredging at Superfund Megasites EPA (U.S. Environmental Protection Agency). 2004. The Incidence and Severity of Sediment Contamination in Surface Waters of the United States: Na- tional Sediment Quality Survey, 2nd Ed. EPA-823-R-04-007. Office of Sci- ence and Technology, U.S. Environmental Protection Agency, Washington, DC. November 2004 [online]. Available: http://www.epa.gov/water science/cs/report/2004/nsqs2ed-complete.pdf [accessed Dec. 26, 2006]. EPA (U.S. Environmental Protection Agency). 2005. Contaminated Sediment Remediation Guidance for Hazardous Waste Sites. EPA-540-R-05-012. OSWER 9355.0-85. Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. December 2005 [online]. Available: http://www.epa.gov/superfund/resources/sediment/pdfs/guidance.pdf [ac- cessed Dec. 26, 2006]. EPA (U. S. Environmental Protection Agency). 2006. EPA Can Better Implement Its Strategy for Managing Contaminated Sediments. Evaluation Report No. 2006-P-00016. Office of Inspector General, U. S. Environmental Protection Agency, Washington, DC. March 15, 2006 [online]. Available: http://www. epa.gov/oig/reports/2006/20060315-2006-P-00016.pdf [accessed Oct. 22, 2007]. Foth & Van Dyke. 2000. Summary Report: Fox River Deposit N. Wisconsin De- partment of Administration, Wisconsin Department of Natural Resources. April 2000 [online]. Available: http://dnr.wi.gov/org/water/wm/foxriver/ documents/sediment/depositn_report.pdf [accessed Jan. 10, 2007]. Linkov, I., D. Burmistrov, J. Cura, and T.S. Bridges. 2002. Risk-based manage- ment of contaminated sediments: Consideration of spatial and temporal patterns in exposure modeling. Environ. Sci. Technol. 36(2):238-246. Linkov, I., F.K. Satterstrom, G. Kiker, C. Batchelor, T. Bridges, and E. Ferguson. 2006a. From comparative risk assessment to multi-criteria decision analysis and adaptive management: Recent developments and applications. Envi- ron. Int. 32(8):1072-1093. Linkov, I., F.K. Satterstrom, G.A. Kiker, T.S. Bridges, S.L. Benjamin, and D.A. Belluck. 2006b. From optimization to adaptation: Shifting paradigms in en- vironmental management and their application to remedial decisions. In- tegr. Environ. Assess. Manage. 2(1):92-98. Montgomery Watson. 2001. Final Summary Report: Sediment Management Unit 56/57 Demonstration Project, Fox River, Green Bay, Wisconsin. Prepared for Fox River Group and Wisconsin Department of Natural Resources, By Montgomery Watson. September 2001 [online]. Available: http://dnr.wi. gov/org/water/wm/foxriver/documents/finalreport/final_summary_report. pdf [accessed Jan. 10, 2007]. NACEPT (National Advisory Council for Environmental Policy and Technol- ogy). 2004. Final Report. Superfund Subcommittee of National Advisory Council for Environmental Policy and Technology. April 12, 2004. 264pp 

Improving Future Decision-Making 263 [online]. Available: http://www.epa.gov/oswer/docs/naceptdocs/NACEPT superfund-Final-Report.pdf [accessed May 14, 2007]. NRC (National Research Council). 1997. Contaminated Sediments in Ports and Waterways. Washington, DC: National Academy Press. NRC (National Research Council). 2001. A Risk-Management Strategy for PCB- Contaminated Sediments. Washington, DC: National Academy Press. NRC (National Research Council). 2003. Environmental Cleanup at Navy Facili- ties: Adaptive Site Management. Washington DC: National Academies Press. NRC (National Research Council). 2004. Adaptive Management for Water Re- sources Project Planning. Washington, DC: National Academies Press. NRC (National Research Council). 2005. Superfund and Mining Megasites – Les- sons from the Coeur d’Alene River Basin. Washington DC: National Acad- emies Press. TAMS Consultants, Inc. 2000. Hudson River PCBs Reassessment RI/FS Phase 3 Report: Feasibility Study. Prepared for U.S. Environmental Protection Agency, Region 2 and U.S. Army Corps of Engineers, Kansas City District. December 2000 [online]. Available http://www.epa.gov/hudson/fs000001. pdf [accessed May 14, 2007]. von Stackelberg, K.E., D. Burmistrov, D.J. Vorhees, T.S. Bridges, and I. Linkov. 2002. Importance of uncertainty and variability to predicted risks from tro- phic transfer of PCBs in dredged sediments. Risk Anal. 22(3):499-512. Wenning, R.J., M. Sorensen, and V.S. Magar. 2006. Importance of implementation and residual risk analyses in sediment remediation. Integr. Environ. As- sess. Manag. 2(1):59-65. WI DNR/EPA (Wisconsin Department of Natural Resources and the U.S. Envi- ronmental Protection Agency). 2002. Record of Decision, Operable Unit 1 and Operable Unit 2, Lower Fox River and Green Bay. Wisconsin Depart- ment of Natural Resources, Madison, WI, and the U.S. Environmental Pro- tection Agency, Region 5, Chicago, IL. December 2002 [online]. Available: http://dnr.wi.gov/org/water/wm/foxriver/documents/whitepapers/Final_R ecord_of_Decision_OUs_1-2.pdf [accessed Jan. 11, 2007].