Design Principles and Criteria for an Effective Long-Term Institutional Management System: Findings and Recommendations
The foregoing chapters have demonstrated several key points:
Only a small number of U.S. Department of Energy (DOE) sites can now be remediated to a level that permits unrestricted use. For most of the remaining sites, only an intermediate level of cleanup and safety is currently possible, enough to permit some uses, but not enough to permit unrestricted use.
Everything that will need to be considered in making decisions for the long-term disposition and management of residually contaminated sites cannot be precisely known. Instead, decisions often will have to be made under conditions of irreducible uncertainty.
Sustained vigilance will be required, yet there is reason to be skeptical of our collective societal ability to sustain such vigilance.
To return to the prescient observation of Alvin Weinberg noted at the beginning of this report, management of nuclear wastes and contaminated sites will require “both a vigilance and a longevity of our social institutions that we are quite unaccustomed to.” Given the scope of the challenges and current limits in scientific understanding and technical capability, these problems cannot simply be made to “go away” by application of existing scientific and technical know-how. Nor is it prudent to adopt the position that the influences that can erode management systems and organizations will be successfully held at bay over time by existing institutions, including but not limited to DOE. Under the circumstances, there are no existing formulas or “cookbook” solutions that can simply be pulled off the shelf and put into place. Instead, the best guidance that can be offered involves two components: first, the need for sober recognition of the magnitude of the challenge that needs to be faced, and second, the need to favor institutional and technical systems that will have high probabilities of being able to anticipate and correct problems, and more broadly, to minimize future regrets.
DESIGN PRINCIPLES AND CRITERIA
The challenge is to do the best one can in dealing with a problem that cannot be fully known in advance. This appears to come down to placing emphasis on prudence and precaution. Specifically, there are nine characteristics of institutional design that need to be emphasized:
defense in depth;
complementarity and consistency;
stability through time;
follow-through and flexibility.
Defense in Depth: Layering and Redundancy
In Chapter 5, the characteristics of layering and redundancy were discussed. “Layering,” as used in this study, refers to using more than one element to accomplish basically the same purpose; “redundancy” refers to having more than one organization responsible for basically the same task. While the concept was discussed in Chapter 5 in terms of stewardship activities, it applies more broadly to the total system of institutional management for the site as well. For example, layering could occur by having both contaminant isolation measures (such as engineered barriers) and stewardship measures (such as deed restrictions and zoning), all intended to accomplish the same purpose of preventing undue exposure to residual contaminants. Similarly, organizations such as the site owner, local citizens groups, and regulatory agencies might all have the right and responsibility to oversee the ongoing management of a site and to ensure that safety measures are performed. Redundancy requires careful coordination and mutual trust to avoid chaos.
Complementarity and Consistency
“Complementarity and consistency” as the phrase is used here, refers to having contaminant reduction, contaminant isolation, and stewardship measures that support and enhance each other, rather than hobbling or detracting from each other. If, for example, a contaminant isolation measure (such as waste entombment) makes it difficult to carry out a stewardship activity (such as monitoring and oversight), the components of the site's institutional management system will not be complementary and well integrated. It is also particularly important to guard against cases where the day-to-day incentives for agency management personnel and contractors run counter to official agency policy. For example, if official policy is to secure the highest possible levels of public safety, it is important to have the agency's official reward system (e.g., salary increases, promotions, contractor bonuses) reinforce that policy rather than being based on other factors that are politically popular or easy to quantify (e.g., number of acres or sites transferred to private hands).
“Foresight” refers to anticipating how the components of the system will work, individually and together, and making preparations in a timely fashion. The committee has observed in the past the tendency of DOE to make commitments in such documents as Tri-Party Agreements to remediation actions that are not technically feasible, often resulting in delays and loss of trust in DOE. For example, if a contaminant isolation measure such as a cap requires monitoring, but the monitoring capability is not designed into the barrier system, retrofitting may be difficult and expensive. On the other hand, a barrier system of multiple layers, coupled with monitoring equipment that could detect a failure in the first line of defense before the second is threatened, might reflect better foresight and reduce long-term costs in accordance with the old adage that “an ounce of prevention is worth a pound of cure.” Similarly, preparations for legal use restrictions and appropriate systems of enforcement need to be made well before property transfers are conducted. Foresight sometimes may be constrained by uncertainty, but it should be employed to the greatest extent possible.
Accountability: Ability to be Monitored and Enforced
“Accountability” means both “answerable” and “capable of being explained.” As used here, it refers to the ability of both the human and the technical components of the site's management system to be monitored and, if necessary, corrected through renewed remediation activities, enforcement, or other means. If people responsible for various site management and oversight tasks cannot be held answerable to the interested and affected public for their actions or non-actions, or if the site 's remedial technologies and its physical environment do not perform as expected, yet that deviation goes unnoticed and unexplained, the efficacy of the site's protective system is likely to erode over time.
“Transparency/visibility” as used here refers to having site disposition decisions that are not only rational, but also clearly articulated and readily accessible to public scrutiny. People need to understand both the site disposition decision and its rationale to be able to evaluate effectively whether the site's protective system is working as anticipated. Without this transparency, the public still may be able to evaluate whether the system is failing, but only after the failure has become evident or more serious.
One of the characteristics of what appear to be relatively reliable organizations seems to be a high degree of openness and visibility such that errors can readily be seen and understood to be problematic by a wide range of people, and those people are reasonably free from what Martin (1999) terms “the suppression of dissent.” Under such circumstances there can be a significant increase in the probability that an error will be detected and corrected. This approach runs counter to the tendency to favor organizational secrecy or to solve problems by putting them “out of sight, out of mind.” Not only are transparency and visibility needed for an open analytic-deliberative process involving citizens as well as regulators and management personnel (National Research Council, 1996a), but transparency and visibility can improve system safety and lay the groundwork for accountability.
“Feasibility” refers to having an institutional management system that is technically, economically, and institutionally possible to implement within a specified time period. If, for example, the disposition decision calls for a remedial technology that has not yet been fully developed or tested, the system will not be feasible unless this limitation is overcome. Similarly, unwarranted institutional expectations (e.g., expecting local governments to be impervious to development pressures, or expecting DOE to continue to receive high levels of funding for oversight into the indefinite future [see Probst and Lowe, 2000]) may lead to infeasible assumptions about site management.
Stability Through Time
“Stability” refers to the likelihood, based on reasonable estimates, that the components of the site management system and the system as whole will continue to perform as expected. A continued, stable investment in resources must be assumed to accomplish this stability. Stability may be much more likely with some elements, specifically those requiring a minimum of upkeep, monitoring, oversight, and enforcement. In some cases, measures that increase stability may lead to decreased flexibility, particularly in terms of institutional performance. However, analysts such as LaPorte and Keller (1996) have argued that nuclear waste management appears to create greater need for “institutional constancy ” than is possible from the typical approach in many policy institutions, namely “muddling through.”
Iteration: Revisiting Site Disposition Decisions
“Iteration” refers to the concept that, when a site's uses must be restricted because of residual contaminants, it is desirable to periodically reconsider both how well the site's protective system is working and whether it can
be improved. Iterability is thus motivated by both caution (i.e., a recognition of the need to plan for fallibility) and optimism (i.e., a recognition that better technologies or institutional arrangements may become available in the future). While pressure for expanded use of the site and nearby resources (e.g., water) may be the greatest impetus to revisit site disposition decisions, iterability should be routinized. In other words, it should be integrated into the total site disposition decision process.
Follow-Through and Flexibility
Given the impossibility of establishing a “best” solution in advance for the vast majority of DOE sites, it is obvious that there will also be a need for more systematic iteration or reconsideration on a periodic basis. The need for iteration coupled closely with follow-through mechanisms is thus a matter of both caution (i.e., recognition of the need to plan for fallibility) and of optimism (i.e., a recognition that better science and technologies or institutional arrangements may become available in the future, particularly if institutions of the present and near future make the needed investment in research and development). The challenge is that, even though experience to date has provided some evidence of success, there is little evidence that present-day institutions can be counted on to provide follow-through to act on new information with the degree of long-term reliability and rigor that current contamination problems require. Under the circumstances, perhaps the best that can be done is to plan more carefully about the kinds of approaches and institutions that could be developed to do a better job, as noted earlier, keeping in mind at all times the importance and preferability of minimizing regret. By this, we mean developing a decision strategy overall that avoids foreclosing future options where sensible, takes contingencies into account wherever possible, and takes seriously the prospects that failures of institutional controls or other stewardship measures in the future could have ramifications that a good steward would want to avoid triggering through inappropriate action in the present. In practice, as noted earlier, this may mean deciding not to tear down certain structures that may prove to have historical value as future museums or interpretative sites, or choosing to leave sites in federal control as environmental preserves or wildlife refuges rather that starting a process of private commercial development that could well prove to be self-intensifying and difficult to control or redirect. There is also a need to recognize, first, that some options will entail limiting future options but may still, on balance, be preferred; and second, that some options have already been foreclosed by past actions.
In summary, a “one-size-fits-all” formula for institutional management is not advocated in this study. Instead, institutional management should be tailored to the needs and conditions of each site. Nevertheless, the general precepts spelled out in this chapter—the system design criteria, iterative decision process, and requisite attributes of institutional mechanisms—should be considered whenever a DOE waste site will, after remediation, have residual contamination necessitating restrictions on the uses of the site or nearby resources.
The following discussion pertains generally to sites across the U.S. Department of Energy (DOE) legacy waste complex, rather than addressing particular sites and facilities. In addition to information and perceptions gained in the course of this study by presentations from DOE personnel, contractors, representatives of other agencies, other experts, and interested public citizens, as well as from tours of a number of sites, the members of the committee also drew on a wide spectrum of individual background experience and knowledge of many other relevant activities and studies, in particular, current studies by the National Research Council.
Almost All Sites Will Require Future Oversight. Although considerable progress has been made over the past decade in treating and stabilizing wastes at DOE sites, numerous contaminated units within these sites cannot be made safe for unrestricted release. Moreover, at many of these sites radiological and hazardous contaminants are already migrating, or can be expected to migrate, beyond site boundaries.
The challenge for long-term site institutional management is therefore to assure that risks posed by such migration are successfully managed, not only within site boundaries, but in nearby areas where site managers will likely have less effective control.
Engineered Barriers Have Limited Lives. Engineered barriers have limited design lives compared with the time periods over which wastes will remain hazardous, and hence, will require ongoing surveillance and maintenance, and in some cases periodic replacement, to assure their continued ability to isolate wastes.
Designing and maintaining long-lasting engineered barriers pose significant challenges for long-term institutional management. Such a system must direct attention to research and development aimed at improving the performance of both the physical systems that isolate wastes from the environment and the human institutions upon which the long-term effectiveness and monitoring of engineered barriers depends.
Institutional Controls Will Fail. Institutional controls and other stewardship activities are being heavily relied upon in the DOE planning for long-term management of sites where hazards will remain. Past experience with such measures suggests, however, that failures are likely to occur, possibly in the near term, and that humans and environmental resources will be put at risk as a result. The circumstances under which stewardship measures fail need to be better understood than they are at present, as must the risks associated with such failures.
There is a need to carry out more systematic research on the types of institutional forms and incentive structures that have shown greater reliability to date, and to develop and put into place those forms and structures that appear to have the greatest promise for at least reasonably vigilant oversight and stewardship over a period of decades to a century or more. Some of the problems can be expected to last for many thousands of years. Although one might conclude that stewardship might be an alternative to cleanup or isolation for periods of tens to a few hundred years, it may not be an acceptable approach for multi-thousand-year problems.
Conduct “Institutional Performance Assessments.” While risk assessments have been used extensively to guide cleanup decisions at DOE sites, they appear less well suited to the quantitative assessment of alternative long-term disposition than strategies that rely on stewardship measures.
There is a need to develop the techniques and data so that what might be called “institutional systems performance assessments” can be conducted, using the same conceptual approach as for technical performance assessments; the means to do so do not exist at present. As in the scientific and technical arena, where lack of data, understanding, and analytical capability often limit the utility of risk assessment, the risk of failure of management strategies involving long-term stewardship measures is likely dominated by the contribution of “unknown unknowns”—that is, flaws in aspects of management systems that are unknown or unrecognized by analysts at the time risk assessments are conducted. It might be argued that “institutional system performance assessments” should be subjected to the same standards applied to analysis of the performance of technical systems, particularly if institutional control is viewed as an alternative to a technical remedy. For example, a performance assessment of a high-level waste repository must consider any feature, event, or process having a probability of greater than 1 in 10,000 of occurring in a 10,000-year period. If institutional performance assessments are required to meet less stringent standards, it could tend to make institutional options appear to be less “risky” than technical remediation options simply because the scope of potential risks considered is narrower in the institutional case.
Remediation Efforts Do Not Always Account for Long-Term Institutional Management Needs. Many remediation efforts at DOE sites, though oriented toward future or end states, in fact aim to achieve interim or temporary cleanup goals.
Remediation planning at individual DOE waste sites is not currently occurring in a way that explicitly takes into account the needs and limitations of long-term stewardship. In particular, site planning now occurring across the DOE complex does not adequately or realistically consider the weaknesses of some stewardship measures. It is very likely that, for at least some sites where DOE is currently “completing ” cleanups in the sense described at the beginning of this report, future remediation planning to revisit contamination problems will prove necessary to assure the degree of protection originally intended.
Present Remediation Should Aim to Facilitate Possible Re-Remediation. Actions taken today should aim to maximize chances that future generations have the capacity to identify and attend to unanticipated problems should they emerge at sites requiring long-term institutional management. Necessary steps include (a) assuring that the scope and severity of waste-site problems that might emerge in the future are as minimal as we can reasonably make them today, (b) assuring that future waste-site stewards have adequate resources (scientific, technical, organizational, financial, and informational) to take action when deemed necessary, and (c) assuring that monitoring and surveillance systems put in place today have (and retain) the capacity to detect unfavorable changes in site conditions at the earliest possible time.
Models Used in Remediation Decisions Are Inadequate. Remediation decisions at many DOE sites are relying on modeled estimates of long-term contaminant transport. The modeling approaches currently in use are often not “state of the art,” nor have they been systematically reviewed to ensure the appropriateness of the assumptions used in generating these estimates. Moreover, the current “state of the art” (in environmental modeling and related computational science and technology) does not adequately capture the complicated reality that must be dealt with at many DOE sites.
Basic Research is Needed to Improve Long-Term Remediation Effectiveness. Whether modeling is relied upon or not, greater emphasis should be placed on developing long-term decision strategies that will be robust and adaptive in the face of actual results turning out differently than originally intended. This will not occur without concerted attention to basic research that improves understanding of the actual rather than the modeled environments at DOE sites, particularly the subsurface environments, and the dynamics of contaminant transport within them.
Assessment of Long-Term Impacts of Private-Sector Reindustrialization is Needed. The use of private-sector reindustrialization at DOE sites, while in many ways laudable, needs to be examined from a long-term institutional management perspective that takes into account the inherent fallibility of stewardship measures and other limitations in society's ability to manage contaminated sites over the long term.
Early and careful planning is necessary in several arenas to begin the process of developing an institutional management program for long-term disposition of radioactive and hazardous waste sites that ensures the protection of the public and the environment. The committee recommends that DOE commit the time and funding needed to develop and implement effective plans devoted to five key principles: 1) plan for uncertainty, 2) plan for fallibility, 3) develop appropriate and substantive incentive structures, 4) undertake scientific, technical, and social research and development, and 5) plan to maximize follow-through on phased, iterative, and adaptive long-term approaches.
Plan for Uncertainty
Cleanup strategies and “post-remediation” site management planning strategies must be able to adapt to a wide range of variation in possible outcomes. It is far more sensible to anticipate a wide range of possible outcomes than to focus first on defining a “most likely” outcome and then to add uncertainty by applying uncertainty ranges. Remediation strategies that avoid foreclosing options for dealing with possible outcomes often should deserve preference over those that do not.
Plan for Fallibility
Other things being equal, contaminant reduction and removal should be preferred over contaminant isolation, and either is preferable to the imposition of stewardship measures that have a high risk of failure. Stated more
simply, a precautionary approach, that is, one that is self-consciously risk averse and therefore takes remedial actions even when harm is not clearly demonstrated, argues for erring on the side of contaminant reduction and removal to safer locations. Strategies that maximize the visibility of sites with residual contamination (e.g., nuclear historic parks) are preferred over those that, by making sites less apparent to the public, increase the chances that knowledge of potential risks will eventually be lost. For land containing potentially hazardous residual contamination, it is best to encourage uses that are likely to be relatively constant through time (e.g., ecological reserves) over those likely to be subject to frequent change (e.g., commercial development).
Far greater effort needs to be made to assure that information about contaminated sites is preserved and communicated effectively to future site users. It is important that relevant records on residual contaminants, remedial actions taken, and technologies used be preserved in forms that will remain accessible and readily understood by future generations. At present, it is not possible to assure that this will happen, but the committee recommends devoting greater energy to the task. Disseminate information broadly to all interested parties, and rely where possible on institutions with proven ability and positive incentives to preserve records over the long term.
Much has been published in the academic literature in recent years of the notion of “government failure.” Government may work best when serving as a referee between parties of equal power (government's adjudicative function), but less well when it serves as a partisan “player.” An agency's “organizational culture ” is an additional consideration. Agencies that develop effective mechanisms for accomplishing one mission frequently find it very difficult to shift to other tasks. The novel demands of long-term institutional management suggest that many existing agencies, including DOE, can be expected to have difficulties adjusting to the expectations embodied in the long-term institutional management construct. This places a premium on assuring that at least some elements of future stewardship are developed from the ground up, rather than simply creating a long-term institutional management task within the structure of an existing agency.
Develop Appropriate and Substantive Incentive Structures
One of the reasons for institutional failure over time may well be that future institutions and their employees will face pressures or respond to incentives that were not adequately understood or anticipated when the institutions were first put in place. Efforts should be put forth to identify and examine the nature of incentive systems within whatever types of institutions that seem to have long-term successes at maintaining their missions, and to ask whether similar incentive structures can be built into stewardship organizations and systems.
In virtually all cases, however, stable long-term funding mechanisms and access to other needed resources appear to be necessary components of incentive structures that maintain institutional focus and effectiveness through time. Trust funds are one mechanism of this type worthy of exploration in that they could reduce susceptibility to future budget cuts from Congress or other governing bodies. Active citizen oversight of long-term management should be likewise encouraged, with stable funding or financial rewards for detection of lapses in the stewardship system, although it should never be relied upon as the lynchpin of a long-term institutional management program, since citizen groups also suffer “atrophy of vigilance.”
Undertake Scientific, Technical, and Social Research and Development
Effective long-term institutional management of waste sites requires attention to limits in knowledge wherever they inhibit our ability to apply the three sets of institutional management measures discussed in the report. This requires attention to both the basic science and technology needs and the organizational and human performance aspects of long-term planning systems. Continuing efforts already underway to improve or demonstrate the short-term performance of engineered barriers are of obvious importance for extrapolating performance to a long-term institutional management system. The likelihood that such barriers will eventually fail suggests a need for greater emphasis on the specific elements of research that improve our ability to detect and correct failures of barriers once in place. The likelihood that engineered barriers will fail also highlights the need for basic research aimed at improving understanding of the surface and subsurface environments at waste sites and the dynamics of residual contaminants within them. Where transport modeling is relied upon to estimate the extent of plume
migration, the errors in prediction associated with simplifying assumptions and lack of fundamental knowledge take on greater importance when very long-term planning is required. This in turn highlights the need for more attention to basic science aimed at improved understanding of the surface and subsurface environments. Attention must be given to ways to make planning more robust to accommodating “surprises” that change our understanding of residual contaminant behavior.
Plan to Maximize Follow-Through on Phased, Iterative, and Adaptive Long-Term Approaches
A long-term institutional management framework should not be static. Adaptation to changing conditions or unexpected outcomes can only be possible if the overall strategy is iterative, but iteration works best if plans have been laid for follow through on successive phases. In essence, long-term institutional management needs to be oriented toward collaborative, adaptive learning by systematically and actively seeking opportunities that cause learning and rethinking to occur.