pp 2667–2672
Superfund and Mining Megasites
Legacy contamination at large, complex, industrial hazardous waste sites creates management and financial challenges for EPA’s Superfund program.
What should a program established to clean up relatively contained industrial hazardous waste sites do when faced by thousands of tons of contaminants spread throughout a 150-mile-long corridor? This is essentially the challenge faced by the U.S. EPA at the Coeur d’Alene River Basin Superfund site in northern Idaho. Superfund sites are commonly perceived to be former landfills or industrial disposal areas. However, these sites range tremendously in size and can encompass hundreds of square miles and a range of natural environments and human communities. Addressing the contamination at these large complex sites creates tremendous management and financial issues for the Superfund program (1–3).
Mining and contamination in the Coeur d’Alene River Basin
Mining began in the Coeur d’Alene River Basin in the late 1800s; the area contains some of the most historically productive lead and silver mines in the U.S. as well as a few currently active mines. However, historical mining and ore processing have left contamination across the northern panhandle of Idaho and into eastern Washington state. As described and referenced in great detail in the NRC report (6), mine waste, particularly tailings from milling activities, was discharged into rivers of the region and transported downstream throughout the floodplain of the Coeur d’Alene River, into the 50-mi2 Lake Coeur d’Alene, and down the Spokane River into eastern Washington. An estimated 121 million tons (t) of metals-enriched mine tailings were produced by the mines and mills that operated in the Coeur d’Alene River Basin, with ~62 million t discharged to the basin’s streams (7). The topography and habitats of the area are extremely diverse, including high-gradient mountain streams, slow-moving meandering rivers, wetlands, and large and small lakes. At this site, the riverbed, riverbanks, and thousands of acres of wetlands and floodplains are contaminated with lead-containing sediments. Surface waters throughout the basin have elevated levels of dissolved metals, particularly zinc. Lead concentrations in wetland sediments in some areas of the basin have proven acutely toxic to waterfowl, with >18,000 acres exceeding lead levels estimated to cause adverse effects to waterfowl. Zinc in surface waters in areas throughout the basin can be thousands of times greater than EPA’s ambient water quality criteria. In the 1970s, blood-lead levels in children living adjacent to the Bunker Hill Mine and Smelting Complex were severely elevated. The smelting complex has closed, and cleanups in the area surrounding the smelter have been completed. Blood-lead levels in children have declined markedly since the 1970s and now appear to be approaching those of same-age children in the U.S. general population.
EPA is charged with protecting public health and the environment. In pursuit of this goal, EPA administers the Superfund program to identify contaminated sites, determine responsible parties, and finance cleanups when liable parties are unknown or are unable. The Superfund program is authorized under the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (P.L. 96–510); it is implemented under the National Contingency Plan (Code of Federal Regulations, Title 40, Part 300). Contaminated sites possessing sufficient documented hazards to human and environmental health can be placed on Superfund’s National Priorities List (NPL) to guide further investigation and remediation. The most expensive of the NPL sites are “megasites”, defined as sites where cleanup is anticipated to cost >$50 million.
Legacy contamination at hard-rock mining sites (particularly abandoned mine lands) poses serious management issues for Superfund (e.g., see 4). A 2004 EPA Inspector General report identified 156 hard-rock mining sites in the U.S., 40% of which were on the NPL, that could cost as much as $24 billion to clean up. At least 19 of these are likely to be megasites (5). Mining megasites are the largest and most complex of these sites; they involve multiple contaminants and contaminant sources and large volumes of waste material that have accumulated over many years of mining activity and are often dispersed over wide areas.
Coeur d’Alene River Basin case study
One such mining megasite is the Coeur d’Alene River Basin in the northern panhandle of Idaho (see box to the left) (6). After the site was listed on the NPL in 1983, cleanup efforts, including residential soil replacements and floodplain soil and sediment removals, focused on the most contaminated areas in a 21-mi2 “box” surrounding the smelting complex. In 1998, EPA began applying Superfund requirements beyond the box to contaminated areas throughout the 1500-mi2 Coeur d’Alene River Basin project area. A 2002 record of decision (ROD) (8) addressed the entire project area, excluding the box (which was the subject of earlier RODs). This ROD contained a “final remedy” to address contamination-related human-health risks and an “interim remedy” to begin to address ecological risks at a cost of $359 million over 30 years. The selected remedial strategy was highly controversial partly because of its long time frame, its high costs, and its perceived uncertain scientific and technical basis.
Cleaning up the Coeur d’Alene River Basin is a major challenge for EPA’s Superfund program. The amount and wide distribution of waste materials preclude complete remediation with traditional cleanup approaches such as removal and capping. Portions of the communities and transportation infrastructure are built on top of and out of mining waste. Periodic floods continue to redistribute these wastes throughout the basin, and the contaminants undergo changes in chemical speciation and bioavailability as they are transported through the river basin. Thousands of people and their properties in multiple political jurisdictions are involved. It is an extensive, expensive, complex, and controversial site, with private parties unable or unwilling to accept responsibility and finance most of the remaining cleanup.
The Coeur d’Alene River Basin is one of the largest mining-related Superfund sites in the U.S., but it is not unique. For example, just east of the Coeur d’Alene River Basin, across the Idaho–Montana border, the Clark Fork Operable Unit of the Milltown Reservoir–Clark Fork River Superfund site includes 120 river miles contaminated with metals stemming from mining activities in upstream reaches (9). These upstream areas, the Anaconda Smelter and the Silver Creek Butte Superfund sites, are also megasites and together encompass nearly 400 mi2.
NRC review of Superfund decision making
In 2003, the U.S. Congress instructed EPA to arrange for an independent evaluation of the scientific and technical decision making within the Coeur d’Alene Basin Superfund site by the National Academy of Sciences (NAS). In response, the National Research Council (NRC) convened the Committee on Superfund Site Assessment and Remediation in the Coeur d’Alene River Basin. The committee’s report, released at the end of 2005, contains a range of conclusions and recommendations related to EPA’s remedial investigation, human and ecological assessment, remedial planning, and decision making (6). The final charge to the committee was to assess “lessons from the Coeur d’Alene case that may be applicable to other similar Superfund sites.” As a result, the report contains a series of findings and recommendations for considering mining megasites at the national level.
As it reviewed the remedy identification and selection in the Coeur d’Alene River Basin and cleanups at other mining areas, the committee observed several problems in applying Superfund requirements to mining megasites. This article summarizes several of the national-level findings, particularly in relation to the management of large, complex mining megasites in the Superfund framework. We use the Coeur d’Alene River Basin to illustrate the issues and include information from relevant sections of the report. However, only that report represents the committee’s consensus conclusions and recommendations.
No final remedy
Superfund regulations focus on identifying and implementing a final cleanup remedy. However, this implies that a final solution can be clearly defined before remediation begins. In the case of large mining sites, where remediation may involve many decades of sequential remedial actions, a final remedy may never be achieved (10). The most obvious problem with cleaning up megasites such as the Coeur d’Alene River Basin is the massive quantities of contaminated waste materials (including waste rock, tailings, and tailings-contaminated sediments) that cover a large geographic area in a variety of upland, wetland, and aquatic environments. The complexity and volume of contaminated material practically eliminate the potential to completely remove, cap, and treat the contaminated materials; and they make practical and effective remedies very difficult to design and implement. Indeed, the volume of mining wastes present in the Coeur d’Alene River Basin is so large that it is doubtful that complete remediaton could ever be attained. It is estimated that >100 million yd3 of contaminated materials are present in the basin. EPA’s models indicated that even after aggressive cleanup, dissolved metal concentrations will remain above the recommended levels for hundreds of years (8).
The NRC committee found that even the limited removals proposed for the Coeur d’Alene River Basin will be costly and difficult and, in some cases, may not even be feasible. The extent to which proposed remedial measures would reduce dissolved metals concentrations in surface waters is unclear. Engineering solutions to a contaminated region as large and geographically complex as the Coeur d’Alene River Basin are not obvious. Remediation must be viewed as a long-term process involving numerous individual remediation projects, only some of which can be specified at the beginning. Given the inevitably high uncertainty about the design and ultimate success of the proposed remedies, any estimates of the duration and cost of the remediation are necessarily crude approximations.
A long-term process
Because of the difficulty of implementing a final remedy, the cleanup of a site like the Coeur d’Alene River Basin will require a long-term commitment to implement and maintain the cleanup actions undertaken. Remedies for environmental and human health will require continued efforts to control land use, protect the integrity of the remedies, and deal with flood-related recontamination, which is inevitable in a watershed such as this one. Indeed, the $359 million remedial estimate is only a first step in achieving environmental protection and, if nothing more is done, it is still estimated to take hundreds of years to achieve water-quality standards, and significant quantities of contaminated sediments will remain uncontrolled in the environment. The Coeur d’Alene River Basin site is not alone; EPA believes that more than half of the mining sites currently listed on the NPL will require operation and maintenance in perpetuity (5).
Cleanup of mining megasites necessarily must be viewed as a long-term process with an uncertain outcome. Management of these sites over the many decades needed to complete the remediation process requires the development of institutions with the capability to oversee engineering operations, minimize the impact of remediation on local communities, and maintain the institutional controls needed to control human exposures. It also requires the implementation of a long-term monitoring strategy that will provide more specific information on the causes of human-health and environmental risks and the sources of contamination causing these risks, evaluate the effectiveness of remediation efforts, and monitor changes in human and environmental health.
A bureaucratic process
To many observers, cleaning up a site under Superfund appears to be a bureaucratic, cumbersome, and inefficient process. Millions of dollars and many years can be spent undertaking studies, producing massive reports, and attempting to come to agreement on a remedy that will adequately protect human health and the environment while complying with the other Superfund requirements. This process was established initially to address more limited industrial waste sites, and it may not be appropriate for cleanup at a large, geographically complex mining megasite like the Coeur d’Alene River Basin.
The Superfund process calls for EPA to first gather all the necessary information (the remedial-investigation [RI] phase), then evaluate alternatives for addressing all the human-health and environmental risks identified in the information-gathering stage (the feasibility-study [FS] stage), and then decide on the best remedies for reducing these risks to acceptable levels (documented in the ROD). The Coeur d’Alene River Basin FS goes to great lengths to select, document, cost, and compare five alternative strategies. However, none of these remedies was selected. The ROD is based on a massive effort and extensive documentation through the Superfund site decision documents (including the RI, human and ecological risk assessments, FS, and proposed plan) as well as a series of public meetings and responses to written comments. However, in reality, the actions documented in the ROD may or may not be implemented on the basis of financial, societal, or on-the-ground considerations. This process can be confusing and frustrating to an uninitiated public seeking to genuinely contribute to a final decision by reviewing and commenting on decision documents.
As noted by the NRC committee, complexities inherent in an ecosystem as multifaceted as the Coeur d’Alene River Basin do not mesh well with the rigidity of the Superfund process. As an area increases in complexity, the certainty of cost, volume, and remedial efficacy estimates decreases, as does the certainty that selected remedies will be implemented. In reality, large geographically complex sites such as the Coeur d’Alene River Basin cannot be remediated in a short time frame, and efforts to describe the entirety of the problem and chart a path to completion (as attempted in the Superfund process) become less realistic with increasing complexity of the site. These decision documents—even when based on best understanding and engineering practices and considering the uncertainty involved—open the agency to criticism that the decisions are not being followed or are incorrect. Although the current Superfund requirement to develop these decision documents may be unavoidable, it is unfortunate, because mining megasite environments are not conducive to these types of encompassing descriptions and predictions.
A path forward
Given these problems, the NRC committee concluded that an effective program for mining megasites should emphasize long-term management of sites and include a stable management structure, with federal, state, and local representation, recognizing that the remediation process inevitably will take decades to complete. The objectives of the program would be to protect human health and the environment using a combination of institutional controls, active remediation, and habitat restoration. From the beginning, data collection, evaluation, and the decision-making process should be designed so that it is focused on establishing a durable process for long-term management of mining megasites rather than selecting “final” remedies that cannot truly be final. Because of the long-term commitment required, active involvement by the affected states and local communities is essential. Long-term management requires long-term management structures.
Implementing adaptive management
The NRC report describes adaptive management as a six-step process for defining and implementing environmental management actions under conditions of high uncertainty about the outcome of those actions. The process, based on previously described principles (11–13), contains the following interactive steps:
- Assessing the problem. This begins by defining the scope of the problem, defining measurable management objectives and potential management actions, and specifying key indicators for each management objective in the relevant spatial scales and time frames. Explicit forecasts are then made concerning the responses of the indicators to alternative management actions. Finally, key uncertainties are identified, and the implications of these uncertainties with respect to the effects of alternative management actions are described.
- Designing a management plan. The management alternatives identified in step 1 are compared on the basis of the likelihood of meeting the management objectives, cost, risk of undesired consequences, and other relevant criteria. In addition, a monitoring protocol is designed that specifies the types and quantities of baseline data; the frequency, timing, and duration of monitoring; the necessary indicators; and the appropriate spatial scales. Most importantly, indicator values that trigger a change in the management actions or objectives must be specified.
- Implementing the plan. The management plan should be implemented as specified. Circumstances requiring deviations from the original plan should be identified in advance and should be understood and agreed to by all stakeholders. Any such deviations must be clearly documented.
- Monitoring. Implementation monitoring should include three components: monitoring for implementation or compliance (were the actions taken as planned?), monitoring for effectiveness (did the plan meet objectives?), and monitoring to validate the model parameters and relationships (which hypothesis is correct?).
- Evaluating results obtained from monitoring. Results obtained from monitoring are compared with the forecasts from step 1. The evaluation should explain why the results occurred and should include recommendations for future action.
- Adjusting the management plan. The models used to make the initial forecasts should be updated, and the objectives of management should be reviewed and possibly adjusted. New forecasts are made, and management actions are revised as necessary.
Among several recommendations, the committee concluded that at those sites where it is unlikely that final remedies can be identified and implemented, a rigorous adaptive-management process with well-defined performance milestones, monitoring strategies, and evaluation criteria should be established (see box on the next page). The adaptive approach recognizes that the information needed to design a remedy that will meet all performance goals may not be available when remediation begins. This approach involves establishing goals and developing a monitoring program that measures progress toward them and provides data needed to adjust the remedy to meet the goals. It also emphasizes continuous real-time evaluation of remediation success and replacement of ineffective or inefficient approaches.
Current and future role of adaptive management
Adaptive management appears to be endorsed, if not implemented, by EPA. For example, recent guidance on the management of contaminated sediments advocated adaptive management, “especially at complex sediment sites to provide additional certainty of information to support decisions” (14). But the question remains whether EPA can or will implement adaptive management as a remedial strategy. Superfund regulations do not preclude the use of adaptive management, but EPA guidance supports the development of Superfund decision making as a comprehensive process moving from the RI through the FS to defining a remedial plan.
At the Coeur d’Alene River Basin site, the ROD made only cursory mention of adaptive management. Instead, an interim 30-year set of prioritized remedial actions was established. After the ROD, EPA issued an environmental management plan that proposed an adaptive-management approach for the basin; however, the committee concluded that the plan did not adequately adhere to accepted procedures for adaptive management. For example, many of the proposed monitoring indicators were only loosely connected to the remediation goals specified in the ROD, and no explicit links between monitoring results and changes to the remedy were provided. To some extent, the required 5-year reviews of sites and requirement to adjust unsuccessful remedies may be considered to be adaptive management; however, the process could be made much more effective if a more rigorous approach were adopted.
The authors recognize that adaptive management appears to provide less certainty about the path and outcome of remediation than does a conventional remedial plan of defined scope, duration, and cost. This uncertainty can present a real burden to affected parties, including industry and the public. The stigma of Superfund designation has negative financial and psychological consequences, and an open-ended adaptive-management remedial process may reasonably be unacceptable to stakeholders. Responsible parties may perceive the situation as too open-ended, with possible expenses and time frames not defined. Individuals or organizations with little tolerance for leaving contaminants in place may view adaptive management as much less ambitious and stringent than complete removal. The authors believe these objections are outweighed at sites like the Coeur d’Alene River Basin, because the size and complexity of the environment imply that high uncertainty concerning the cost, duration, and success of the remedy is unavoidable regardless of the management process followed.
Any decision-making process about Superfund megasites will remain difficult and contentious. Environmental remediation is expensive, and levels defining “clean” and the path to achieve those levels are open to wide debate and interpretation. Recognition by all parties of the long-term nature of the remediation process at these sites, together with the development of appropriate management structures, stable financing, and the use of adaptive management, would provide a significant improvement over current approaches to achieving cleanup of mining megasites such as the Coeur d’Alene River Basin.
Acknowledgments
The authors wish to acknowledge the contributions of the entire committee that authored the NRC report Superfund and Mining Megasites—Lessons from the Coeur d’Alene River Basin: David Tollerud (chair), University of Louisville; Herbert Allen, University of Delaware; Lawrence Barnthouse, LWB Environmental Services; Corale Brierley, Brierley Consultancy; Edwin Clark, Earth Policy Institute; Thomas Clarkson, University of Rochester; Edmund A. C. Crouch, Cambridge Environmental, Inc.; Alison Cullen, University of Washington; Joseph Graziano, Columbia University; David Johnson, State University of New York; Ronald Kendall, Texas Tech University; John Kissel, University of Washington; Thomas Lapoint, University of North Texas; David Layton, Lawrence Livermore National Laboratory (retired); C. Herb Ward, Rice University; Spencer Wood, Boise State University; and Robert Wright, Children’s Hospital Boston.
Karl E. Gustavson is a senior program officer at the NRC, NAS, Board on Environmental Studies and Toxicology. Lawrence W. Barnthouse is president of LWB Environmental Services, Inc., and an adjunct associate professor of zoology at Miami University. Corale L. Brierley is president of Brierley Consultancy. Edwin H. Clark, II, is a senior fellow at the Earth Policy Institute. C. Herb Ward is a professor of civil and environmental engineering and ecology and evolutionary biology at Rice University. Address correspondence about this article to Gustavson at kgustavson@nas.edu.
References
- Final Report of the Superfund Subcommittee of the National Advisory Council for Environmental Policy and Technology, 2004, http://www.epa.gov/oswer/docs/naceptdocs/NACEPTsuperfund-Final-Report.pdf.
- Probst, K. N.; et al. Superfund’s Future: What Will it Cost? Resources for the Future: Washington, DC, 2001.
- Schmidt, C. W. Not-So-Superfund: Growing Needs vs. Declining Dollars. Environ. Health Perspect. 2003, 111, A162-A165.
- Fields, S. The Earth’s Open Wounds: Abandoned and Orphaned Mines. Environ. Health Perspect. 2003, 111, A154-A161.
- EPA. Nationwide Identification of Hardrock Mining Sites; Report No. 2004-P-00005; OIG: Washington, DC, 2004.
- NRC. Superfund and Mining Megasites-Lessons from the Coeur d’Alene River Basin; National Academies Press: Washington, DC, 2005.
- Long, K. R. Production and Disposal of Mill Tailings in the Coeur d’Alene Mining Region, Shoshone County, Idaho: Preliminary Estimates; Open-File Report 98–595; U.S. Geological Survey: Tucson, AZ, 1998.
- EPA. Bunker Hill Mining and Metallurgical Complex: Operable Unit 3, Record of Decision; EPA Region 10: Seattle, WA, 2002.
- EPA. Clark Fork River Operable Unit Record of Decision; EPA Region 8 Superfund: Helena, MT, 2004.
- Moore, J. N.; Luoma, S. N. Hazardous Wastes from Large-Scale Metal Extraction: A Case Study. Environ. Toxicol. Chem. 1990, 24, 1278–1285.
- British Columbia Forest Service. An Introductory Guide to Adaptive Management for Project Leaders and Participants; Ministry of Forests: Victoria, Canada, 1999.
- British Columbia Forest Service. Definition of Adaptive Management. Adaptive Management Initiatives in the BC Forest Service; Ministry of Forests: Victoria, Canada, 2000.
- U.S. Department of Energy. Environmental Response Design and Implementation Guidance; DOE/EH-413–9915; Office of Environmental Policy and Guidance, Office of Program Integration, and National Environmental Training Office: Washington, DC, 1999.
- EPA. Contaminated Sediment Remediation Guidance for Hazardous Waste Sites; EPA-540-R-05–012; Office of Solid Waste and Emergency Response: Washington, DC, 2005.
