Risk-based decision making to manage contaminated sediments

Authors


Abstract

This paper summarizes discussion among the 7 authors who served on an expert panel at the Third Battelle International Conference on Remediation of Contaminated Sediments held in New Orleans, Louisiana, USA, in January 2005. In this article, the authors review how sediment management decisions are currently made and address the question of how management decisions should be made in the future. It is arguably the case that sediment remediation presents greater challenges and more complexity than traditional land-based clean-ups. Although understanding of these challenges and complexities has grown over the last 25 y, there has been, until recently, relatively little innovation in the approaches used to manage the environmental risks posed by contaminated sediments. New methods that facilitate a more rigorous analysis of the multiple criteria considered in decision making have been developed. These methods, collectively known as multicriteria decision analysis (MCDA), coupled with the use of comparative-risk assessment and cost/benefit analysis, are proposed as an effective, efficient, and credible foundation for evaluating remedy alternatives at contaminated sediment sites.

EDITOR'S NOTE:

This paper is among 9 peer-reviewed papers published as part of a special series, Finding Achievable Risk Reduction Solutions for Contaminated Sediments. Portions of this paper were presented by the author at the Third International Conference on Remediation of Contaminated Sediments held in New Orleans, Louisiana, USA in January 2005.

INTRODUCTION

The magnitude of the challenge posed by contaminated sediment is large. The US Environmental Protection Agency (USEPA) has estimated that, in the United States, approximately 1 billion m3 of surficial sediment are sufficiently contaminated so as to pose a potential risk to human health and the environment (USEPA 1997). While relatively little quantitative information exists on the extent of the global scope of the problem, there is little reason to think that the situation in the United States is markedly different in other industrialized countries. Given the role sediments play as both a sink and a source of contaminants in aquatic systems, numerous government and private organizations are working to develop the policies and the technical means to address the challenge of how best to manage contaminated sediments.

Addressing sediments is a 2-step process involving assessment, then management. Sediment assessment is generally viewed as the process used to characterize sediment within the context of a given activity or purpose (e.g., habitat use, dredged material disposal, land farming, habitat construction, and so on). Sediment management generally refers to the process of making decisions and taking actions to either preserve or expand recreational and economic services provided by a waterway or to limit the negative impacts of sediment upon a system. Sediment management strategies or options can include a large set of actions, ranging from no action, the imposition of engineering controls, or the use of more aggressive, intrusive activities related to removing, containing, or treating sediments.

Over the last 30 y, several countries have proposed or promulgated environmental laws and regulations that address a broad range of stresses imposed by human activity on the environment. For example, sediment quality is addressed in the United States by several state and federal programs and regulations, including, most prominently, the National Environmental Policy Act; the Clean Water Act; the Marine Protection, Research, and Sanctuaries Act; and the Comprehensive Environmental Resource, Compensation, and Liability Act (NRC 1997). Within the European Union, the Water Framework Directive (WFD) contains far-reaching provisions intended to secure and manage water resources, and largely by implication, sediments, at the river basin scale (Brils 2003; Brooke 2004a, 2004b). Sediment management is also directly or indirectly included in different European directives, such as the Waste Directive and Habitats Directive (Köthe 2003). Guidelines for management of coastal dredged material have been developed by several international maritime conventions, including the London Convention, Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR), and the Helsinki Convention (den Besten et al. 2003; Bergmann and Maass, forthcoming).

While improvements in environmental quality were realized in the United States and several countries following implementation of environmental regulatory programs, the current focus these programs have on a specific medium, (e.g., water, soil, and air), environmental system (e.g., rivers, estuaries, and marine systems), contaminant (e.g., PCBs, pesticides, and Hg), or activity (e.g., navigation dredging, wastewater discharge, and environmental cleanup) could, in some cases, impede achieving efficient long-term solutions to risks posed by sediments (Apitz and Power 2002). As our understanding of sediment in aquatic systems has evolved, it has become increasingly evident that effective and sustainable management strategies for aquatic environments should be focused on larger scales, such as watersheds and coastal zones. While conceptual approaches for preserving or restoring environmental quality in aquatic environments are being developed (e.g., Apitz and White 2003), significant institutional barriers remain at present.

Against this backdrop, the Third Battelle International Conference on Remediation of Contaminated Sediments held in New Orleans, Louisiana, USA, in January 2005 hosted a panel of experts, who were asked to review how sediment management decisions are currently made and answer the question of how management decisions should be made in the future. The panel was challenged to address a series of questions listed in Table 1. This article summarizes the discussions and recommendations prepared by the panel and presented at the meeting.

WHAT IS RISK-BASED DECISION MAKING?

While it is entirely feasible to develop decision frameworks based on goals other than risk management, within an environmental context, risk-based decision making is widely considered the most appropriate foundation for managing contaminated sediments (Bridges, Berry, et al. 2005; Bridges, Kiker, et al. 2005). Effective risk-based decision-making is informed by 3 fundamental principles, (1) adverse conditions are driven by site-specific factors, (2) uncertainty is always present, and (3) risks are managed, and only rarely eliminated.

Site-specific assessment and management

The nature and magnitude of risks related to sediment contamination are largely dependent on site-specific conditions. Such conditions include the nature and extent of the contamination, the concentrations and bioavailability of those contaminants, and the broad range of processes (pathways) by which those contaminants contact humans and wildlife. Recognition of the fundamental role of site-specific factors and processes is evident in the degree of emphasis given to developing comprehensive conceptual site models in the conduct of risk assessment and the execution of risk management (ASTM 1995). Risk assessment should inform risk management by providing decision makers with sufficient process-level understanding about how risks are being produced at a site such that informed decisions can be made about managing risk through interdiction within the relevant processes. Risk assessment should inform decision makers about the pathways of exposure, the human and wildlife populations at risk, and the risks associated with implementing different engineering or remedy options. The information developed from a comprehensive risk assessment should aid in balancing trade offs among costs, short- and long-term effectiveness, implementability, and environmental hazard (Apitz et al. 2005b; Wenning et al. 2006). These issues, and others, are discussed in US Superfund guidance concemed with evaluating remediation options at contaminated sediment sites (USEPA 2005).

Among the key elements of an effective sediment assessment, based on biological and chemical measures (USEPA 2001), are a well-designed and site-specific conceptual site model; transparent and well-thought-out data collection and analysis; explicit discussion of heterogeneity, uncertainty, and scaling issues; and carefully selected reference sites and decision criteria (Apitz et al. 2005a, 2005b). If an assessment concludes that unacceptable risks exist, risk-management strategies must be evaluated, selected, implemented, and their success evaluated. Evaluating the proposed remedial alternatives in terms of risk reduction, to inform the selection process, will involve assessing risks under the diverse set of conditions that would result from implementing the various management alternatives (Reible et al. 2003; Apitz et al. 2005b; Bridges, Berry, et al. 2005; Wenning et al. 2006).

The role of uncertainty

Heterogeneity and complexity are intrinsic components of all natural systems, whether anthropogenically impacted or not. Thus, uncertainty, whether due to incomplete knowledge or inherent variability in processes or parameters, is an everpresent component of environmental assessment and decision making (von Stackelberg et al. 2002; Vorhees et al. 2002). The existence of uncertainty necessitates that risk assessment and risk management are based in likelihoods and probabilities. In complex systems such as aquatic environments, multiple, competing hypotheses and management options will be considered; therefore, sediment management decision making must balance knowledge, uncertainty, and policy to support decisions based on the information available (Apitz et al. 2005a). The degree to which uncertainty complicates or interferes with this decision making will be determined by the consequences of making the wrong decision. If the costs of making the wrong decision are high, either for the environment or for the organization providing the capital to execute the decision, then there will be less tolerance for uncertainty, in comparison with cases when the costs of being wrong are lower.

Managing risk

In all but the most trivial cases, the practical objective of risk-management strategies is to reduce risks rather than eliminate them. The reason for this constrained objective is that most environmental risks cannot be eliminated. While this statement may strike some as being unnecessarily pessimistic, we would maintain that it is nevertheless true, as complete containment or removal of contaminants from a heterogeneous and complex environment is neither economically nor technologically feasible.

The challenges facing those confronted with managing the risks posed by contaminated sediments have been described in detail elsewhere (e.g., NRC 2001) and amply demonstrate the nature of the problem. The risks posed by contaminated sediments result from a complex set of physical, chemical, and biological processes. The physics of sediment-transport processes, the chemistry of contaminant partitioning and bioavailability, and the biology of the hundreds of potentially exposed species at a contaminated sediment site complicate the assessment of baseline risks as well as the evaluation of remedial alternatives.

Table Table 1.. Critical questions for sediment management decision making
What are the major impediments to effective decision making in managing contaminated sediment?
How do we better define our objectives in sediment management? For example, are we making decisions to meet regulatory criteria, reduce certain contaminant loads, protect humans, manage watersheds, protect specific sensitive species, etc.?
What role should risk assessment play in remedial decision making at contaminated sediment sites?
How do we more securely link the risk assessment and decision processes being applied at contaminated sediment sites to the management objectives developed for those sites?
What are the key elements of effective comparison-based decision making that gives more comprehensive consideration to the benefits and costs, the advantages and disadvantages, of each management alternative under consideration?
Given the diversity of environments, receptors, risks, and time and spatial scales that are relevant to the broad range of management or remedial options currently available, how does one make a fair comparison among the options?
What are our conceptual models for linking risk assessment to management?
How do we develop frameworks that can embrace emerging and currently unaddressed toxicants and risk processes? How do we develop consistent but adaptable risk frameworks?
What structuring approaches, methods, or tools are available for facilitating effective and efficient decision making? What gaps are there in the current suite of decision-making tools?
When and how should stakeholder values be elicited and included in the decision-making process?
How can the presence of uncertainty in risk estimates be more effectively addressed or managed within the decision-making process?
What improvements can be made in the way uncertainty associated with the performance of remedial technologies is addressed or managed in the decision-making process?
What information should be collected to validate the decision process?
What changes could be made in existing decision models for contaminated sediment sites that would encourage the development and use of innovative solutions to sediment management or cleanup?
How can the principles of adaptive management be applied at contaminated sediment sites?

The physical complexity of sediments and the aquatic environments that contain them provide for no easy remedial options. Sediment resuspension during dredging and the operational constraints that result in the presence of residual contaminated sediments after dredging disabuse all but the casually engaged that dredging is a surgical solution to the problem. The dredging option is further complicated by the fact that, once removed, the sediments must be transported and either treated or stored, which involves transferring and transforming the risks to other locations and receptor groups. In situ capping of sediments has proven to be a useful alternative to removal via dredging; however, concern about the long-term performance of caps is a subject of continual discussion. In situ treatment technologies offer some promise, although they are presently in the early stages of technology development; however, it is unrealistic to expect such technologies to resolve all the technical and logistical challenges posed by contaminated sediments.

All of the management options thus far discussed involve the use of invasive procedures, in which the environment is either damaged or altered by the actions taken. The rate and extent to which ecosystems recover from the actions taken will depend on the magnitude of the disturbance and any number of site-specific factors. The viability of Monitored Natural Recovery (MNR), which does not rely on the use of invasive techniques but on natural sedimentation and degradation processes, is vulnerable to concerns about process rates and uncertainties, the amount of time necessary to achieve acceptable risk reduction, and the risk of contaminant or sediment dispersal during the recovery process.

These considerations, and others, have led those who have given careful consideration to the problem of contaminated sediments to conclude that there are no zero-risk options for managing contaminated sediments and that a “systems engineering approach” should be used to manage risks (e.g., NRC 2001). Such an approach seeks to identify optimal management strategies through an explicit analysis of tradeoffs among the risks, costs, and benefits associated with the remedial options for achieving management objectives for a site. In this way, consideration will be given to the temporal scales for risks, costs, and benefits, including risks related to consumption/generation of energy/carbon dioxide during transport, accidents, and worker safety, and contaminant risks at the containment or treatment site (NRC 1997; Sorensen et al. 2004).

THE INGREDIENTS FOR RISK-BASED DECISION MAKING

If the premise holds that decision making to date has generally been performed in a less than holistic fashion and without the full benefit of risk-benefit analysis as discussed above, then how should decision making with regard to sediments evolve in the future? Seven recommendations are made here that should help foster more effective, risk-based decisions at contaminated sediment sites.

Develop a watershed-scale perspective

Decisions concerning sediment remediation should be informed by a thorough understanding of the biological, chemical, and hydrodynamic processes ongoing in the watershed. Source control should be a leading priority of an overall strategy to improve the quality of surface water and sediments. Sediments are a reservoir for contaminants in watersheds, as well as a source of contaminants (though not in the original sense). At the scale of an individual site or sediment project, knowledge about sources and fate and transport process within the watershed will ensure attention is given to controlling upstream hotspots before pursuing management action at downstream locations. Non-point-source inputs and combined sewer overflows are just 2 sources within watersheds that can make significant contributions to degraded sediment conditions. More effective sediment management will require a more holistic approach to source control and management.

The potential for cumulative impacts represents another issue that can be addressed through a watershed-scale approach. Several watersheds in the United States and elsewhere contain multiple, individual contaminated sediment sites located in close proximity to each other (e.g., the Passaic River, Newark Bay, Hackensack River, Berry's Creek, and the Meadowlands in northern New Jersey, USA). Addressing ecological risks in situations where fish and fish-eating birds may migrate across site boundaries, or otherwise experience exposures from several sites, pose significant challenges to existing regulatory structures. Developing watershed-based management approaches for improving sediment quality will require relevant agencies and stake-holders to formulate specific legislative, regulatory, and policy actions to accomplish more integrated management. These approaches are being developed in Europe, with the WFD mandating the management of all waters, and by association, sediments and contaminant sources, at the river basin scale (Apitz and White 2003; Brils 2003; Brooke 2004a, 2004b; Aptiz, Carlon, et al., forthcoming; Apitz, Elliott, and Fountain, forthcoming; Heise and Apitz, forthcoming).

Pursue comparison-based decision making

The premise of comparison-based decision making is that a remedy decision should be based on projections of the performance of the remedial alternatives under consideration (e.g., Sorensen et al. 2004). Performing detailed projections will require using the approaches and information collected during a risk assessment to compare and contrast the environmental risks and benefits of each of the remedy options. Information about chemical behavior and transfers between sediment, water, and biological compartments should fold into engineering considerations about how each remedy alternative interrupts contaminant exposure over the timescales of interest. Given that decisions, under such a framework, are made on the basis of modeled predictions of remedy performance, care should be taken to consider and address the role of uncertainties in those predictions. As depicted in Figure 1, selecting the best remedy alternative based on realistic estimates of performance that consider uncertainty can both inform and complicate decision making. However, open recognition and analysis of uncertainty in performance projections will emphasize the need for using a comprehensive suite of decision criteria (that consider both benefits and costs) given that, within the predictive power of available data and models, it may be difficult to distinguish alternatives on the basis of any single performance metric.

Figure Figure 1..

Risk-reduction comparison for three remedial alternative scenarios for a hypothetical site. Solid lines represent the best estimate for the risk-reduction trajectory and dashed lines represent uncertainty about the best estimate.

Give more attention to the role of spatial scales at sediment cleanup sites

One largely unexplored issue in sediment management concerns how the effectiveness of remedial technologies varies as a function of spatial scales. Whether the technology under discussion involves dredging, capping, in situ treatment, or monitored natural recovery, it is relevant to ask how the effectiveness of that technology might change as a function of the size of the contaminated site. Would, for example, dredging be as effective a remedy at a 0.1-ha site as at a 10-ha site, given the same contaminants and concentrations? Greater variation in physical parameters, including hydrodynamic conditions and sediment geotechnical properties will, on average, be encountered at larger sites. Numerous logistical elements of a remediation, including the size and number of equipment pieces required, access and feed rates for consumable materials, how long it will take to complete the action, etc., will be affected by spatial scale. The collective experience to date in sediment cleanup has been gained from relatively small sites (i.e., < 0.1 ha); however, several sites in the United States (e.g., the Hudson River, NY; Fox River, WI; Passaic River, NJ; Portland Harbor, OR; the Palos Verdes Shelf, CA; and Upper Columbia River, WA; all USA) and elsewhere span large distances (20–200 km). With regard to making comparisons among proposed remedial options, it would be particularly important to develop an understanding of where the breakpoints in effectiveness might fall and how the remedial alternatives might differ with respect to the spatial scale of the proposed remediation.

Seek to use a collaborative decision-making process

The legal, regulatory, and administrative procedures that must be followed to reach a decision at a contaminated site will differ as a function of which authorities and agencies are at play. The degree of cooperation and collaboration, or conversely, the degree of competition and antagonism, that exists among the parties engaged in the process will affect decision making in numerous ways more significant than the general ambiance of project meetings. Anecdotal observation of sediment projects currently underway provides evidence that progress in a project is more efficiently achieved (measured in terms of time and monetary costs) when cooperation and collaboration dominate. The perspectives of stakeholders will differ; however, working together in a cooperative and collaborative manner does not require that stakeholders surrender their perspectives, only that the parties place equivalent value on conserving resources while making progress toward completing the project.

According to the National Research Council (NRC 1997), the most successful sediment management projects involve a broad range of stakeholders early and often. Stakeholder involvement at all levels of the decision process is mandated in Europe by the WFD, and case studies for stakeholder involvement are being developed (Gerritts and Edelenbos 2004). Stakeholder groups can include representatives from local communities and governments; fishermen, industries, ports, environmental and public-interest groups; and regulatory and trustee organizations from local, state, and national organizations and indigenous peoples. This broad representation allows all interested parties to be involved with and understand the problems and their investigation and resolution, fostering trust and mutual understanding (NRC 2001; USEPA 2002).

Be realistic, flexible, and innovative

Risk assessments that form the basis for remedial decisions are designed and conducted to result in protective conclusions, i.e., to err on the side of overestimating environmental risk. While in many respects such a bias is appropriate, one of the consequences of such a bias toward greater estimated (and thus perceived) risk for remedial decision making is that more aggressive remedies will tend to be selected than would likely be the case in the absence of this bias. Although conservative risk assumptions should always be used early in the assessment process, final decisions should be supported by refined, realistic estimates of risk provided by site-specific data and sound analytical approaches. It is increasingly evident that such an approach is best supported by a well-designed, site-specific, and tiered assessment process (Bridges, Berry, et al. 2005).

Decision making based on overprotective risk estimates is further compromised by overly optimistic (or poorly supported) estimates of remedy performance and cost, increasing the likelihood for a less than optimal outcome. Selecting a remedy for which costs are underestimated (or the size of the project's funding stream is overestimated or uncertain) can result in substantial project delays (increasing the duration of exposures and risks). More realistic estimates of the risks and benefits, as well as of the costs, of proposed remedial actions will provide for more informed and credible decision making (Wenning et al. 2006). According to Wenning et al. (2006), 2 aspects of risk are necessary to consider: implementation risks that address the benefits, costs, and hazards associated with remedy engineering; and residual risks that address the shortand long-term biological and environmental benefits and hazards posed by a remedy. In either case, basing decisions on realistic estimates (as opposed to optimistic estimates under ideal conditions) of remedy performance will help build credibility in the overall process.

Although it can be argued that excessive interest in retaining flexibility in remedial decision making can hinder progress and slow project completion, there appears to be ample room for injecting more flexibility into current management processes before reaching such an extreme position. There is a need for greater flexibility at several levels within the process. The more strictly linear decision process characterized as the “decide-and-defend” approach to remedial decision making does not contain sufficient flexibility to accommodate or benefit from increasingly favored comparison-based and adaptive management approaches, nor does it support a watershed-scale management approach, which requires that data flow freely between the investigation, assessment, and remediation aspects of the decision process (Apitz, Carlon, et al., forthcoming).

It is arguably the case that sediment remediation presents greater challenges and complexities than traditional landbased cleanups. While it is true that our collective appreciation and understanding for these challenges and complexities has grown over the last 25 y, there has been relatively little innovation in the technologies used to manage the risks posed by contaminated sediments over this period. Recently, there has been some encouraging evidence that focused energy is being directed to researching and developing new approaches and techniques for managing contaminated sediments, including in situ approaches. However, guidance for organizing how and where these approaches would be used at actual sites is lacking. Clearly, the innovative approach would have to offer some advantage in terms of greater risk reduction, lower costs, conserved habitat, etc., compared with an existing technology. But choosing the innovative option will require having an approach for addressing and managing the uncertainties attendant to using a new approach. To address the large volumes of contaminated sediment worldwide that require some form of management, innovative, cost-efficient approaches will be required. Comprehensive guidance for guiding the application of innovative approaches is needed.

Make use of adaptive management principles

Recognition of the fact that uncertainty is a component of all assessment and management activities has exposed some weaknesses in the traditional, linear regulatory processes, wherein data are collected, data are analyzed, and then a decision is made. The dominating role of uncertainty in many environmental problems, including contaminated-sediment management, argues forcefully for the use of adaptive management as an approach to managing uncertainty within the decision-making process (Linkov, Satterstrom, Kiker, Batchelor, et al. forthcoming; Linkov, Satterstrom, Kiker, Bridges, et al. forthcoming). Adaptive management emphasizes the role of performance monitoring and uses the results of such monitoring to make necessary adjustments in management actions over time. The iterative approach used in adaptive management reduces the unrealistic burden placed on risk assessors to provide 1 definitive prediction of environmental risks. The principles of adaptive management would encourage the use of pilot studies to test the efficacy of technologies being considered for broad use at a site. Management strategies that begin with considering or implementing less invasive management options (e.g., MNR and capping) while reserving more invasive measures (e.g., dredging) to cases where the required risk reduction cannot be achieved through less aggressive means, will conserve habitat, financial resources, and decision flexibility (it would be difficult, if not impossible, to undo an invasive option like dredging once implemented). For these reasons, an adaptive approach to sediment management will also be more receptive to the use of innovative technologies.

Build a decision structure with people, a process, and tools

Successful environmental decision making in complex settings will depend on the extent to which 3 key ingredients are integrated within an organizing structure: people, process, and tools (Kiker et al. 2005). Having the correct combination of people is the 1st essential element to the overall decision process. The 3 groups of people involved in the decision-making process, although in different manners and degrees, are (1) decision makers, (2) scientists/engineers, and (3) stakeholders. Each group has its own perspectives, motivations, methods of envisioning solutions, and societal responsibilities.

While risk assessment is a largely scientific process, risk-management decisions must, ideally, balance the outcomes of the scientific-risk assessment process with social, economic, technological, cultural, political, and legal considerations in order to achieve society's goals for a site and the watershed as a whole (Heise and Apitz, forthcoming). Decision makers spend most of their effort in defining the problem context and the overall constraints to the decision. Much of what might be understood as context and constraint will be derived from the relevant laws and regulations to which decision makers must be responsive. In addition, they may have primary responsibility for selecting the final remedy and its means of implementation.

Stakeholders may also provide input to problem definition, but have the highest degree of interaction in helping to formulate success criteria and contributing value judgments for weighting the various success criteria. Depending on the problem and regulatory context, stakeholders may have some responsibility in ranking and selecting the final option (Gerrits and Edelenbos 2004). Scientists and engineers provide the data and analysis necessary to understand the magnitude of the problem and to parameterize the criteria for selecting among the management alternatives. Although they may also take a secondary role as stakeholders or decision makers, their primary role is to provide the technical details as requested by the decision process. There is growing recognition that the successful interaction between these parties requires clear communication and careful planning (Gerrits and Edelenbos 2004).

It is reasonable to expect that the decision-making process may differ in specific details among regulatory programs/project types. Comprehensive decision processes, as described by Kiker et al. (2005), are composed of 2 parts. The 1st part of the decision-making process generates a list of potential management alternatives, success criteria for evaluating the alternatives, and value judgments for weighting the criteria. The 2nd part of decision making involves the ranking of the alternatives by the application of criteria levels and value weights.

There are 4 broad and somewhat overlapping groups of tools that are relevant to environmental decision making (Kiker et al. 2005): (1) modeling/monitoring, (2) risk analysis, (3) cost-benefit analysis, and (4) stakeholder involvement. While those involved in contaminated-sediment assessment and management will be well acquainted with examples of tools within the first 3 categories, most will be less familiar with the use of multicriteria decision analysis (MCDA) techniques to organize and integrate stakeholder views within the decision-making process (see, e.g., Kiker et al. 2005; Linkov, Sahay, et al. 2005; Linkov, Varghese, et al. 2004).

The use of MCDA techniques offers significant benefits in complex decision-making environments (e.g., managing contaminated sediments). Multicriteria decision analysis provides a structure and process for generating and mapping preferences of stakeholder groups in a way that can be linked with the other technical tools, e.g., modeling/monitoring, cost benefit estimation, and risk analysis. Comparative risk assessment (CRA), in particular, provides a powerful approach for analyzing risk-reduction scenarios for the range of management alternatives under consideration at a contaminated-sediment site (Kane Driscoll et al. 2002; Cura et al. 2004; Bridges, Berry, et al. 2005; Bridges, Kiker, et al. 2005). When used in combination, CRA and MCDA provide the means to accomplish true comparison-based decision making (NRC 2001).

Multicriteria decision analysis offers at least 4 primary benefits. First, the structure provided by MCDA calls for the establishment, in an explicit form, of project objectives at the earliest phases of a project. These objectives will inform what risks are evaluated during the risk assessment and serve as the basis for establishing decision criteria for the evaluation of potential remedies (Kiker et al. 2005). Second, MCDA provides the means to not only document differences between stakeholders, but to also quantitatively explore whether those differences are consequential, i.e., whether the differences are large enough to affect the outcome of the decision. There is little benefit to spending time and money on resolving differences that don't really matter. Third, the structure and analytical qualities of MCDA also lend themselves to formally evaluating the value additional information may contribute to a decision, e.g., through the resolution of specific uncertainties in an important decision criterion. In this way, the formal structure and rules provided by MCDA make it possible to prioritize, in a quantitative fashion, the information needed for a decision based on a set of pre-established criteria. Fourth, MCDA offers an approach for facilitating the practical and consistent application of decision criteria across a program (e.g., the 9 National Contingency Plan criteria in the US Superfund program) and for documenting how the criteria were used to reach a decision.

Reticence to using MCDA to structure and facilitate decision making could be motivated by concerns regarding its compatibility with existing regulatory structures; the costs of applying MCDA to sediment projects, including the resources necessary to familiarize team members with the approaches and tools; and discomfort on the part of some stakeholders with a more interactive and open decision-making process.

CONCLUSION

This article summarized discussions among the 7 authors who served on an expert panel at the Third Battelle International Conference on Remediation of Contaminated Sediments held in New Orleans, Louisiana, in January 2005. The authors were asked to review how sediment management decisions are currently made and answer the question of how management decisions should be made in the future. In this article, the authors described the need for innovative and cost-efficient approaches to address the large volumes of contaminated sediment that require management worldwide. It is encouraging that greater focus and energies are being directed toward research and development of new approaches for assessment and management of contaminated sediments, including in situ approaches. A complementary investment is needed to refine what can be an overly linear and narrowly focused environmental decision-making process to incorporate broader consideration of multiple objectives and criteria within a watershed-scale context.

Environmental decision making involves complex trade offs between divergent criteria. The traditional, linear approach to environmental decision making involves the valuation of multiple criteria into a common unit, usually monetary, and, thereafter, performing standard engineering optimization procedures. Extensive scientific research in the area of decision analysis has exposed many weaknesses in this approach (Belton and Stewart 2001). At the same time, new methods that facilitate a more rigorous analysis of multiple criteria have been developed. These methods, collectively known as MCDA methods, coupled with the use of comparative-risk assessment and cost/benefit analysis, would promote more effective, efficient, and credible decision making at contaminated sediment sites.

Acknowledgements

The authors wish to thank Battelle for providing the opportunity for us to meet and discuss these issues. We gratefully acknowledge the financial support of our respective organizations and sponsoring programs. This manuscript has been reviewed in accordance with US Army Corps of Engineers policy and permission was granted by the Chief of Engineers to publish this material.

Disclaimer—The views presented within this article are those of the authors alone and do not necessarily represent the policies of our respective organizations. The peer-review process for this article was managed by the Editorial Board without the involvement of Board members T. Bridges and R. Wenning, both of whom appear as authors in this article.

Ancillary