To start the workshop, an introductory presentation by William (Bill) Levitan, Associate Deputy Assistant Secretary for Site Restoration for Department of Energy’s (DOE’s) Office of Environmental Management (EM), provided insights on the motivations for and expectations of the workshop series. The first session of the workshop focused on promoting effective and efficient risk-informed decision making. To address this topic, federal and academic practitioners provided background and examples of effective use of risk in complex decisions related to environmental remediation decisions. The session was moderated by Michael Kavanaugh, Principal, Geosyntec Consultants and planning committee member, who provided a short introduction to the session by highlighting findings and recommendations from past National Research Council (NRC) reports. Next, Bernard (Bernie) Goldstein, Professor Emeritus, Department of Environmental and Occupational Health, University of Pittsburgh, presented the keynote address on the historical use of risk and sustainability within U.S. government decisions.
This set of briefings was followed by speakers who provided case studies and examples of decision making for complex remediation sites:
- Reggie Cheatham, Director of the Federal Facility Restoration and Reuse Office, Environmental Protection Agency, and
- Paul Black, Ph.D., Principal, Neptune and Company, Inc.
This chapter provides summaries of the key points made by each of these individuals and by participants in the subsequent discussion sessions.
These statements reflect the viewpoints of the individual speakers, not the consensus views of the workshop participants or of the National Academy of Sciences.
2.1 INTRODUCTORY PRESENTATION: DOE EM’S PERSPECTIVES ON THE WORKSHOPS
William (Bill) Levitan
EM has multiple considerations when making remediation decisions: agreements with the sites and site stewards, the environment, and the taxpayers. DOE is not alone in reconsidering how remediation decisions should be made. Several reports and recent policy documents support this statement. For example, Alternatives for Managing the Nation’s Complex Contaminated Groundwater Sites (NRC 2013), the recent release of the Environmental Protection Agency’s (EPA’s) groundwater strategy,1 and the U.S. Nuclear Regulatory Commission’s (USNRC’s) white paper on policy strategies for risk management provide evidence of the need for a national conversation on decision making for contaminated site cleanup.
There are several challenges to establishing a decision-making process for site remediation. The process must balance many concerns and impacts including
- Short-term and long-term impacts
- Worker and community impacts
- Local and global impacts
- Cost and risk mitigation
- End states and future use
It is becoming increasingly apparent that some wastes will remain on site after active cleanup activities are completed (recognized by a potential shift from the basic question of “How clean is clean?” to “How much waste might be left behind?”). It is not yet clear how to incorporate risk and the sustainability process into decisions addressing residual contamination.
DOE is seeking advice on how to complete remediation and make decisions related to long-term stewardship of its sites. These topics are highlighted in the statement of task (Appendix A) and are shown in Figure 2-1.
Projected cost profiles of existing DOE remediation programs indicate peak spending at about $8 billion per year (see Figure 2-2). Current budgets are not at this level nor are they expected to increase significantly.2 To ad-
2 The funding level for EM in fiscal year 2014 was $5.6 billion (see http://energy.gov/sites/prod/files/2013/04/f0/FY2014_EM_Congressional_Budget_Request.pdf).
FIGURE 2-1 Intended main topics of the National Academy of Sciences’ Workshop series.
SOURCE: Levitan 2014.
dress this shortfall, remediation end states for the sites may be reconsidered or further defined; remediation activities may be sequenced; and selection of remedies consistent with site end states may be considered.
Finally, Mr. Levitan provided the following eight topics as DOE’s main reflections from the first workshop:
- The determination of end states is critical to the decision-making process.
- EM and the U.S. government are “not going away” and are committed to addressing the legacy wastes.
- Flexibility allows remediation goals to be accomplished over time (both a blessing and a curse).
- The intrinsic value of groundwater or environmental resources is ill defined.
- Adaptive management approaches are needed to incorporate new technologies and practices.
FIGURE 2-2 Projected spending for DOE’s Office of Environmental Management (EM) site cleanup. The spending peaks at $8 billion in the coming years. Current funding levels for EM are ~$5 billion to $6 billion per year and are not expected to increase.
SOURCE: Levitan 2014.
- Sustainability should be considered as part of the remediation process because it can better accommodate multiple concerns including societal and economic values.
- The sequencing of remediation work should be considered.
- Communication is critical—both for risk assessments and throughout decision making.
2.2 SUMMARY OF DISCUSSION SESSION
Workshop 1 Highlights. The planning committee chair Paul Gilman (Covanta Energy) commented that EM’s reflections and overview of the first workshop were fair. He suggested adding two additional topics to Mr. Levitan’s list (see Vol. I, Ch. 3):
- the Savannah River Site’s decision-making process as a model for risk-informed decision making with stakeholder support,3 and
- David (Dave) Maloney’s example of flexibility in contracting to allow for innovative solutions for cleanup as another highlight.
Other participants added to comments from Workshop 1. A workshop participant highlighted Michael Truex’s presentation on the use of monitoring and modeling for remedy assessment. Dave Maloney noted that Workshop 1 included discussions about a movement away from prescriptive standards, which define how to reach a target value to performance-based standards, which in turn allow for a variety of methods to reach target values without relaxing the environmental standards themselves.
Workshop participants offered several other comments.
Challenges to Decision Making. Planning committee member Patricia Culligan (Columbia University) mentioned that an additional challenge to EM’s decision making is that the stakeholders will change over the life cycle of the remediation. Mr. Levitan agreed that the stakeholders change over time, but added that at any given moment stakeholders are diverse because there are so many of them. Rateb (Boby) Abu-Eid (USNRC) emphasized two critical aspects of the decision-making process: future land use or end use is very important to determine and uncertainties can be large when considering long timeframes.
Waste Types of Future Reactors. Willie Preacher (Shoshone-Bannock Tribes) expressed concern about the types of waste the next generation of nuclear reactors (i.e., Gen IV)4 will produce and the implications for future waste management. Mr. Levitan responded that poor waste practices existed in the past but are no longer followed. Current laws and regulations such as the National Environmental Policy Act (NEPA) forbid dumping or burial (previously accepted practices). Further, Gen IV development—taking place at the Idaho National Laboratory (INL)—and the associated waste practices are currently monitored to confirm that compliance is maintained.
3 For more details on the Core Team Process, see the summary of Mary Flora’s presentation titled, “The Core Team Process: Making Risk-Informed Decisions for On-Site Monitoring,” in Chapter 4 of this report.
4 Nuclear reactors have been grouped into four generations. Generation I (or Gen I) were developed in 1950-60s. Most reactors in operation today are Gen II reactors. Gen III (or 3+) reactors are considered advanced reactors; they are operational in Japan or under construction elsewhere. Gen IV designs are still being developed.
2.3 HIGHLIGHTS FROM NATIONAL RESEARCH COUNCIL REPORTS
Michael (Mike) Kavanaugh
Planning committee member and the moderator for the first technical session of the workshop, Dr. Kavanaugh (Geosyntec), introduced the session with a short overview of relevant NRC reports and their findings:
Wastes will remain at many sites. Dr. Kavanaugh chaired the committee that wrote the recent groundwater contamination report, Alternatives for Managing the Nations Complex Contaminated Groundwater Sites (NRC 2013). The committee estimated that the United States has at least 126,000 contaminated groundwater sites with residual contamination levels that exceed cleanup goals. This number includes 3,650 DOE sites. DOE has stated that the majority of its sites will require long-term stewardship of residual and stored wastes (DOE 1999).
Long-term stewardship is difficult. Long-Term Institutional Management of U.S. Department of Energy Legacy Waste Sites (NRC 2000) concludes that effective long-term stewardship will be difficult to achieve.
Analysis results need to be technically credible. Finding 8 of Risks & Decisions about Disposition of TRU and High-Level Radioactive Waste (NRC 2005) enumerated a list of characteristics of a credible decision-making process, highlighting the importance of credible and believable results (emphasis added): “An effective and credible risk-informed, decision-making process has several characteristics. It is (1) participatory; (2) logical; (3) consistent with current scientific knowledge and practice; (4) transparent and traceable; (5) structured with reasonable independence of the decision authority from the petitioner; (6) subjected to thorough, independent peer review; (7) technically credible, with believable results; and (8) framed to address the needs of the decision process” (NRC 2005, p. 7).
Containment systems will eventually fail. Long-Term Institutional Management of U.S. Department of Energy Legacy Waste Sites (NRC 2000) states that DOE planners should safely assume that any containment system for long-lived radioactive wastes will ultimately fail over the wastes’ lifetimes.
With this as the backdrop to the first session, Dr. Kavanaugh introduced the speakers.
2.4 HISTORY OF RISK AND SUSTAINABILITY IN DECISION MAKING FOR COMPLEX SITES
Bernard (Bernie) Goldstein
Over the past several years, the U.S. government’s thinking about risk and sustainability has evolved to the point that, currently, we are at a tipping point between the two approaches to decision making. Twenty years from now, people will recognize the present moment as the cusp of adoption of sustainability practices into remediation decisions.
To provide credibility and historical context to this statement, one must consider the progression of decision making within environmental management:
- Early years of site remediation assumed a command-and-control approach—“It’s dirty, go clean it up.”
- The current approach to site remediation includes risk assessment and management, which recognizes that not all risks are obvious and visible. Exposure is used to calculate risk—“If you understand exposure, then you can understand the risk.”
- The near-term future of site remediation is moving toward inclusion of sustainability frameworks, which account for value judgments and other considerations during the decision-making process.
The U.S. government’s adoption of risk assessment to inform decisions took many years. Released in 1983, the NRC’s Red Book5 (NRC 1983) outlined a risk assessment framework to guide government decisions. However, EPA took a full decade to develop the tools and processes necessary to measure and assess risk (i.e., risk was defined in terms of exposure). Integration of sustainability principles into decision making is likely to follow a similar timeframe and pattern. Factors contributing to the current movement toward the use of sustainability principles to inform decisions are as follows:
- Current and emerging problems are more complex and challenging (the easy sites have been cleaned up, the challenging sites remain),
- Sophisticated tools to address these complex problems are becoming available,
- Sustainability is becoming a common approach to address broader economic, social, and environmental issues, and
- A sustainability approach that allows consideration of economic
5 The full title of the report commonly referred to as the “Red Book” is Risk Assessment in the Federal Government: Managing the Process.
and social issues in environmental protection is of potential value to the competitiveness of the United States.
Three NRC reports track the progression from risk to sustainability approaches: the Red Book (NRC 1983) defined a risk assessment framework for EPA; Science and Decisions (NRC 2009) defined a risk-based decision framework; and the “Green Book”6 defined a framework for a sustainability process for EPA (NRC 2010). The NRC committee that authored the Red Book discovered that many of the regional EPA workers were incorporating risk into their decisions before headquarters directed them to do so. A similar finding was made by the committee that authored the Green Book; it saw that sustainability principles were already being employed by practitioners to guide decisions at local levels.
The Green Book purposefully did not define “sustainability.” Rather, concepts compatible with a sustainability approach were referenced in other existing government documents including the National Environmental Policy Act (42 U.S.C. § 4331(a)):
“[T]o create and maintain conditions under which man and nature can exist in productive harmony, and fulfill the social, economic, and other requirements of present and future generations of Americans,” and “[T]he continuing responsibility of the Federal Government” is to, among other things, “fulfill the responsibilities of each generation as trustee of the environment for succeeding generations.”
Although not defining sustainability, the Green Book described how it could be used to guide decisions. A sustainability framework has two levels (see Vol. I, Sec.1.6):
- Sustainability Framework Level 1: Components that define the agency-wide process (an agencies’ high-level framework); and
- Sustainability Framework Level 2: Elements of Sustainability Assessment and Management (practitioner-level framework).
The Green Book contains several recommendations to support the incorporation of sustainability principles into EPA’s culture (NRC 2011, p. 5). In particular
Recommendation 3.1: The committee recommends EPA adopt the proposed Sustainability Framework. The proposed Sustainability Framework requires a comprehensive approach including specific processes for incorporating sustainability into decisions and actions. As part of the framework, EPA should incorporate into its decision making upfront
6 Dr. Goldstein chaired the committee that issued Sustainability and the U.S. EPA.
consideration of sustainability options and analyses that cover the three sustainability domains (social, environmental, and economic), as well as trade-off considerations. The framework was developed with the intent that EPA could apply it to any decision to which a need arose.
In the last sentence of this recommendation, the 2011 NRC committee recognized that there may be no need to add processes to some U.S. government decisions. However, for other decisions that require simultaneous consideration of social, environmental, and economic issues, a sustainability framework should be used. To implement the framework, more tools are needed.7
How does an organization move from consideration of risk to sustainability to guide its decisions? Some risk assessment practitioners are concerned that risk assessments are not needed in sustainability frameworks. Risk is part of the consideration of sustainability, but the sustainability framework goes beyond risk assessment. Recommendation 5.1 from the Green Book makes this clear: “The committee recommends EPA include risk assessment as a tool, when appropriate, as a key input in its sustainability decision making” (NRC 2011, p. 6). Practically speaking, a sustainability framework adds the factor of time into decision making—a factor missing in the risk assessment approach—making it valuable for DOE decision makers, who must consider long timeframes in site remediation decisions. A sustainability framework considers not only how to minimize risk but also how to maximize benefit.
The Science and Decisions (NRC 2009) report, with some modifications to the nomenclature, outlines how to incorporate sustainability into decisions. The decision-making framework is presented as three phases (see Figure 2-3). Phases 1, 2, and 3 in the framework show how risk assessment is part of a larger decision-making process.
This approach can be mapped to Sustainability Framework Level 2 of the recently proposed sustainability framework (NRC 2011). This shows that risk assessment has naturally been moving toward sustainability for a number of years.
However, sustainability and risk assessment are not the same thing. A public health analogy can highlight an important difference between sustainability and risk. In public health there are three levels of care: primary, secondary, and tertiary. Primary care is aimed at preventing the contraction of a disease or condition. Secondary care is focused on treating a contracted disease before adverse symptoms have occurred. Tertiary care treats the symptoms of the disease, attempting to prevent pain or further damage. In the analogy, primary care corresponds to a sustainability approach. It strives to avoid conditions that would be more costly and impactful in the
FIGURE 2-3 The risk-based, decision-making framework proposed by the Science and Decisions report (NRC 2009). Although not specifically mentioned, many components of a sustainability framework are represented.
SOURCE: Goldstein 2014.
future. Tertiary treatment is analogous to risk assessment. Its goal is to treat an existing problem that was created by failure of other preventative measures.
The current practice of risk assessment has several problems, which could carry over to future sustainability frameworks if not addressed. Box 2.1 provides an example of over-reliance on risk assessment without using common sense to balance the results.
The short analysis of this scenario shows that one can expect a single excess cancer death over the time period of 1.75 million years, roughly the time that humans have existed on the planet. Although sustainability frameworks require consideration of future generations, one should be careful about how the results of risk assessments—such as this example—are used to guide decisions. Risk to future generations currently depends on radiation exposures over periods of millennia, which implies that risks of cancer to humans remain unchanged over time. But this is not likely to be the case, as suggested by the progression of our understanding of other diseases and conditions. For example, in the early 20th century, Yellow
Stealth Farmer Scenario for Risk Assessment
Risk assessment for exposure risk is calculated for a remote DOE legacy site that is contaminated but secured through institutional controls. This includes a “stealth farmer” scenario.
Unbeknownst to anyone, a family of four climbs the fence and lives off the land. The family secretly develops subsistence life style on DOE land contaminated to a level of 1/100,000 risk. The family is replaced every 70 years with another family of four.
The calculated risk of one cancer death from residual radiation is 1/100,000 lifetime. Using these numbers, one adverse event (excess cancer death) will occur every 1.75 million years.
Humans have been on the planet for approximately 1.75 million years.
SOURCE: Modified from Goldstein 2009, 2014.
Fever was potentially deadly and feared by many, but now the disease is understood and treatable (we can vaccinate against it). Isn’t it possible that our understanding of and ability to treat cancer will significantly change over the next several thousand if not millions of years?
In 1994, DOE established a Comprehensive Environmental Response, Compensation, and Liability Act Disposal Facility (CERCLA)/NEPA Policy that states if DOE relies on and follows the CERCLA process (see CERCLA’s nine criteria in Box 2.2), then no separate NEPA document or process would be ordinarily required.8 Under this policy, DOE ensures that it will take steps to involve the public as early as possible in the decision-making process. NEPA requires an environmental impact statement (EIS) and a clear definition of the project goal with options considered (including the “do nothing” option). Although CERCLA does not require an EIS, for some of DOE’s biggest sites, such as Hanford, an EIS for the full site might have been a beneficial exercise.
In summary, sustainable decisions for environmental management involve maximizing benefits while minimizing risks. It may be better to show maximum benefit now than to focus on reducing risks to individuals millions of years from now.
CERCLA’s Nine Criteria
The analysis of alternatives under review reflects the scope and complexity of the site problems and the alternatives being evaluated and considers the relative significance of the factors within each criterion. The nine criteria are part of the National Contingency Plan (40CFR300.430(e)(9)).
The nine evaluation criteria are as follows:
1. Overall protection of human health and the environment
2. Compliance with ARARs (applicable or relevant and appropriate standards)
Primary Balancing Criteria
3. Long-term effectiveness and permanence
4. Reduction of toxicity, mobility or volume
5. Short-term effectiveness
8. State acceptance
9. Community acceptance
2.5 SUMMARY OF DISCUSSION SESSION
Maximizing Benefit, Minimizing Risk. Participants raised several questions about maximizing benefit. Mr. Levitan (EM) asked if the term “optimization” instead of “maximization” might be an alternative way of thinking about sustainability. Optimization allows for more factors to be maximized (instead of the implied single variable for maximization). Richard Mach (U.S. Navy) suggested that optimization is concerned with maximizing benefit, minimizing risk, and finding the right balance between options. Dr. Goldstein agreed with both comments.
Craig Benson (University of Wisconsin) pointed out that the boundaries of the optimization need to be carefully considered. There are consequences of restricting optimization objectives due to system boundaries, such as lines of funding. As the United States becomes more resource constrained, we should consider expanding those boundaries. Dr. Goldstein responded that the separation of federal and state systems may present fundamental challenges to the proposed modification to barriers such as lines of funding.
Planning committee member, Patricia Culligan (Columbia University)
asked for further examples of maximizing benefits. Dr. Goldstein provided the following example from the Green Book (see NRC 2011, p. 117): The lighting systems in New York (NY) City Public Schools were leaking polychlorinated biphenyls (PCBs) into the school buildings, and EPA required their replacement. Under the initially proposed lighting-fixture replacement schedule, replacement costs coupled with a tough budget situation would have necessitated teacher layoffs. By balancing the cost savings from the new, more energy-efficient lighting systems and staggering replacement of the lights, the NY public school system and EPA worked out a solution that reduced the risks of PCB exposure while retaining all of the teachers. This example also highlights how time is introduced into a sustainability-based decision. Mr. Levitan provided another real-life example from a large site in which a small quantity of asbestos fibers was released during DOE cleanup. Because the release violated the Clean Air Act, EPA fined DOE. To offset the cost of the fine, DOE laid off cleanup workers.
CERCLA, NEPA, and Sustainability. Charles (Chuck) Powers (Consortium for Risk Evaluation with Stakeholder Participation [CRESP]) provided historical background on the policy decision to equate CERCLA’s nine steps to NEPA. Lois Schiffer, then assistant attorney general at the Department of Justice (DOJ), voiced at least two concerns about this decision: (1) provisions within NEPA that would have allowed the stopping of cleanup actions and (2) increased participation in the decision-making process.9 These concerns were to be alleviated by full adherence to all of CERCLA’s nine criteria. Dr. Powers also noted that sustainability enters into the CERCLA process through the Remedial Investigation and Feasibility Study (RI/FS).
Planning committee chair Paul Gilman (Covanta Energy) responded to Dr. Power’s comment on sustainability and CERCLA. The statement was made during the first workshop that CERCLA prevented EPA from incorporating sustainability. Yet the point has been made throughout both workshops that creative work with the stipulated agreements allows for sustainability principles to be part of the decisions (and while economic issues cannot be part of a rule, they can be included in the discussions with stakeholders that lead to decisions). Robin Anderson (EPA, Superfund program) responded by saying that if “sustainability” could be better defined, then it may be more easily incorporated into CERCLA. Questions remain about the number of options to consider for optimization and how to prioritize those options. She recognized that there currently are ways to incorporate sustainability principles into CERCLA’s nine criteria.
9 See memorandum dated January 23, 1995, from Lois J. Schiffer, assistant attorney general, Department of Justice (http://energy.gov/sites/prod/files/G-DOJ_nepa_cercla_cleanups.pdf).
2.6 FEDERAL FACILITY CLEANUP DIALOGUE SERIES AND BROWNFIELDS DISCUSSIONS
As a regulator, EPA is in a unique position to offer perspectives on the decision-making process for contaminated sites. A few myths about CERCLA and Federal Facility Cleanup (see Figure 2-4) have arisen and must be dispelled, based on EPA’s experience from many site closures.
“CERCLA is not flexible,” is a common myth.
CERCLA allows for a variety of remediation decisions at different sites that offer cost and time savings while protecting human health and the environment. It is very flexible.
Second, the idea that cleanup goals cannot be achieved is a myth.
In early 2000, Rocky Flats and Mound were extremely challenging sites and were unlikely to be cleaned up for decades. Both of these ventures have been successful and have completed cleanup.
Third, “Federal Facility Agreement (FFA) milestones are not negotiable” is also myth.
In reality, these agreements are renegotiated frequently because of changing funding levels and changing conditions on the sites (e.g., new contamination identified).
Finally, it is a myth that groundwater cleanup cannot be achieved.
Groundwater is a valuable resource, and EPA has worked to ensure its protection. The vast majority (90 percent) of CERCLA/Superfund sites have a selected groundwater remedy.
Risk is a part of CERCLA. It is used to guide a variety of decisions including, policy directives from other agencies. CERCLA’s intention, purpose, and design drive decisions at the sites to be risk-based. Risk is considered throughout the CERCLA process to develop defensible decisions by Remediation Project Managers (RPMs) and others.
In the mid-1990s, federal agencies and stakeholders involved with federal facility environmental cleanup met to discuss how to “do more with less.” The Federal Facilities Environmental Restoration Dialogue Committee (FFERDC) produced a report (FFERDC 1996), outlining principles that are still relevant and important today. Specifically, EPA and EM regularly use Chapter 5, Funding and Priority Setting, as a resource. The report suggests that discussions leading to decisions should focus on accomplishments and review next steps based on available funding. This is especially relevant during times of tight budgets. The prioritization process should be transparent and should include factors beyond risk exposure.
A sustainability framework offers a way to incorporate these additional
FIGURE 2-4 Federal Facility Cleanup Myths and Facts, from the perspective of the Environmental Protection Agency, Federal Facility Restoration and Reuse Office. NCP=National Contingency Plan, RAO=remedial action objectives.
SOURCE: Cheatham 2014.
factors. EPA’s Brownfields program is modeled after the Department of Defense’s Base Realignment and Closure (BRAC) program. The focus of Brownfields is on economic development. Sustainability approaches allow the integration of societal and environmental values into Brownfields’ decisions. Brownfield programs—and environmental cleanup activities, in general—need to focus on accomplishments and take more credit for the jobs created by cleanup activities. More than $7 billion/year is put into the economy by cleanup (more if private industry is considered). Cleanup, jobs, and economy are all parts of the pillars of a sustainability framework.
Protecting the nation’s water is one of EPA Secretary McCarthy’s top priorities.10 EPA strongly supports inclusion of the use, value, and vulnerability of groundwater in decision making. But determining the value of groundwater is challenging. Because the primary responsibility of groundwater rests with the states, the public works within the states are best equipped to assess its value. At the same time, cost and technical limitations can prevent some sites with contaminated groundwater from being
remediated to pre-contamination or maximum contaminant level (MCL) standards. The recent NRC report (NRC 2013) on groundwater remediation provides good examples and background.
“Technical impracticability” or TI waivers have been created to accommodate challenging cases. Although TI waivers are difficult to obtain, they have been granted. Whether or not a TI is in place, each site needs to consider a realistic approach to restoration based on actual and reasonable planned end uses and other sustainability factors.
There are no quick fixes for complex contaminated sites. Often remediation activities uncover additional contamination and waste problems. EPA supports incorporation of new technologies to reduce cost and time. Federal facility sites are a great test bed for these innovations because the U.S. government “holds all the marbles”—if the technology does not perform as expected, then the government can recover. Finally there are two points of caution regarding cost savings:
- cheaper cleanups based only on cost savings for today will ultimately cost more for future generations, and
- EPA will not allow the “writing off” of groundwater to shorten cleanups.
2.7 SUMMARY OF DISCUSSION SESSION
CERCLA Reviews and Timeframes for Analysis. Planning committee member Michael Kavanaugh (Geosyntec) noted that the focus of the recent NRC study (NRC 2013) was complex sites in which cleanup cannot be expected in a reasonable timeframe.11 The study did not conclude that groundwater should be “written-off.” He asked how “timeframe” enters into EPA’s decisions on groundwater. Mr. Cheatham responded that when CERCLA sites have undergone four or five 5-year reviews and have reached steady-state conditions, discussion between EPA and the Primary Responsible Party (PRP) have often spurred on changes to the site remedies. He noted that the NRC report is helping EPA with these discussions, but EPA is realizing that the discussions need to occur with the right people (e.g., public water works and utilities).
FFERDC Background. Marolyn Parson (Savannah River Site’s Citizens Advisory Board) asked for more details on the FFERDC. Mr. Cheatham succinctly described it as “an effort that came and went but left valuable insight.” The FFERDC was reconvened in 2010 to continue the federal
11 Dr. Kavanaugh was the chair of Alternatives for Managing the Nation’s Complex Contaminated Groundwater Sites (NRC 2013).
dialogue. The main difference between the 1996 and 2010 groups is that the discussions have shifted to long-term stewardship, which is encouraging. Current efforts of a group that sprung from the 2010 meetings focus on improving the 5-year review process. The group is holding monthly meetings. Bill Levitan (EM) added that the original FFERDC membership included the Site-Specific Advisory Boards (SSABs).
TI Waivers and Sustainability. A workshop attendee asked about technical impracticability and the need to move forward with sustainability approaches, which might be in conflict. Mr. Cheatham responded that the TI waiver exists, and a few have been granted. How this relates to a sustainability framework is difficult and something he will consider in the future. Bill Levitan (EM) noted that this question targets the essence of the workshops—how do we make risk-informed decisions that recognize the large number of issues simultaneously? Richard Mach (U.S. Navy) noted that for complex sites—especially one that would be the subject of his presentation day (see Mach 2014)—the Department of Defense’s (DOD’s) approach is to consider multiple options for each site. A TI is not necessarily applied for even though the problem is technically impracticable, because it is still possible to come up with a sustainable, long-term solution (see Chapter 3, Bethpage case study example).
2.8 RISK-BASED APPROACHES FOR REMEDIATION DECISIONS
Optimizing the decision-making process requires the balancing of risk against benefit. One challenge to this approach is that risks of exposure can be measured and quantified whereas the “costs” of benefits based on value judgments are not easily quantified. Decision analysis is needed to make defensible, reproducible, and logical decisions that incorporate risk and value judgments for complex sites.
Models are an important tool in decision analysis. Most practitioners of decision analysis understand that—at some level—all models are wrong. Yet, if appropriately applied, then models can be useful. “Useful models” are as simple as they need to be—but no simpler—and they are based on reality. Conservatism in models results in outcomes that are unreliable because their assumptions are not based on reality. Conservative decisions can be made using realistic models, but good decisions cannot be made using conservative models.
Waste and cleanup regulations have impact beyond the environmental remediation of contaminated sites; the regulations can also impact the nuclear energy industry. Overly conservative cleanup standards result in
“the waste management tail wagging the nuclear energy industry’s dog.” Thirty years ago, regulations and guidance—including Data Quality Objectives (DQOs)—were developed as part of nuclear cleanup standards. At the time, simplification made sense. The models, processes, and guidance should be revisited to take advantage of the technologies and experience that have been developed over the decades. The regulations and guidance are currently being updated, but the proposed changes do not go far enough. Meaningful and significant change is understandably difficult but in this case needed.
Environmental remediation is site specific, but aspects of the risk assessment for a given site are often not site specific. For example, the population and threat models (e.g., resident and intruder models) developed for more highly populated sites are also commonly used for remote sites. Probabilities of the likelihood of encounter should be assigned for each of the sites. Another example is the importance of value judgments of the stakeholders for each site. WIPP had strong state and local support, whereas Yucca Mountain did not. Risk assessments normally do not include stakeholder values, but it is possible to account for them using decision analysis tools.
What is decision analysis exactly? Decision analysis combines risk analysis and statistics with an aim to use optimization to guide decisions. One can think of this as “formalized common sense.” Humans are not good at considering more than a few aspects or components of a problem at a given time. A decision analysis framework that incorporates value judgments allows for simultaneous consideration of all aspects of a complex problem (e.g., a Bayesian DQO). When faced with technical complexity, uncertainty, cost and value judgments, and multiple competing objectives, decision analysis provides an approach for making logical, reproducible, and defendable decisions. The analysis is easier to understand even if it is more complicated and time consuming to explain. Ultimately, it is easier to communicate and explain the results because the results are based on realistic models. This approach also avoids “re-dos” because it is more difficult to disagree with the results.
EPA has recently incorporated decision analysis into its decisions. Examples include the following:
- SMARTe (Sustainable Management Approaches and Revitalization Tools)—Brownfields project, the focus was on economy, the decision framework brought in environmental and societal issues to allow for balancing
- Re-Imaging Cleveland—original focus was on revitalization of economy, needed to bring in and balance other values
- DASEES (Decision Analysis for a Sustainable Environment, Economy, and Society)—an EPA model driven by stakeholder involvement, accounts for value judgments
Nuclear legacy sites have an inherently long timescale associated with cleanup. One approach that utilizes decision analysis for sites with long-lived radioactive wastes is to make a decision now but plan to revisit the analysis at least once per generation (e.g., every 25 years). The decision context will change as the remedy’s effect on contaminants evolves and changes to populations occur. Now is the time to adopt decision analysis that can support sustainability frameworks into the decision-making process.
2.9 SUMMARY OF DISCUSSION SESSION
Models. Tom Nicholson (USNRC) asked how abstraction and realism should be balanced in models. Dr. Black responded that it depends on the stage of the problem at which the abstraction is happening. There are examples of models for very complex sites that are exceedingly detailed (i.e., the site has been segmented into one cubic meter sections). These models are parameter rich but data poor. It is better to focus on the broader problem—including uncertainties and sensitivities—and to set up an experimental design across the model space.
Bayesian Statistics. Planning committee member, Mike Kavanaugh (Geosyntec), asked for an example of a Bayesian DQO. Dr. Black explained the differences between traditional and Bayesian DQOs. Traditional guidance suggests the use of classical statistical methods to support implementation of the quantitative steps of the DQO process. However, classical statistical methods have several limitations that make them ill-suited for environmental remediation decision making, as follows:
- They begin with two hypotheses (null and alternative), which limits the number of options to two,
- The statistical evaluation of null and alternative hypotheses is asymmetric, and
- Classical statistics cannot be used to evaluate the probability or uncertainties that contamination may exist at a site, or to estimate the quantities of contamination that may exist in the future.
However, Bayesian statistics can easily accommodate multiple decision options (more than a null and an alternative hypothesis), are mathematically symmetric in addressing each decision option, and can estimate probabilities and uncertainties associated with each option. Bayesian DQOs can balance risks associated with the environmental system against costs and value judgments to arrive at an optimal decision given the information available.
High Consequence/Low Probability Events. Dr. Kavanaugh asked about low-probability, high-consequence events. Dr. Black acknowledged that this problem is difficult but addressable. Models can provide insights for these events but they must include the event’s building blocks. This approach does not model the low-probability event itself but provides insight into how it might occur.
Qualitative versus Quantitative Information. Rateb (Boby) Abu-Eid (USNRC) commented that this discussion is partly about quantitative and qualitative probabilities and uncertainties. Quantitative uncertainties can be related to many unquantifiable value judgments. One needs to account for these contributions. Dr. Black admitted that addressing qualitative uncertainties is challenging. However, it is better to quantify value judgments in the models than to leave them out altogether. Needed are tools of the type suggested by Dr. Goldstein. For example, models provide insight, and sensitivity analysis provides information to guide decisions.