Appendix A
Summary of Scenario-Based Studies of the Remediation of High-Level Waste in Tanks
This committee is by no means the first group to adopt a scenario-based approach to the cleanup of wastes in the Department of Energy (DOE) complex in general, or Hanford tank wastes in particular. Therefore, to provide some historical perspective, in this appendix we review several reports dealing with the Hanford tank wastes from a risk-based or scenario-based approach. These reports were chosen for their direct or indirect relevance to the issue of identifying technology needs. The reports discussed here use methods that are similar to the end state approach that we propose, and could be used (or adapted) to identify technology development needs; moreover, some studies contain explicit recommendations for needed technology development. The bibliography at the end of this appendix lists all risk-based reports that were provided to this committee by DOE and its contractors. Note that this list may not be comprehensive, as there may well be additional risk-based applications of which the committee has not yet become aware.
Evaluating Alternative Strategies for Tank Waste Treatment
Two studies apply a decision analysis or risk-based approach to evaluating alternative strategies for tank waste treatment. In particular, Johnson et al., (1993) deals with the decision of whether to pursue extensive separations, in order to reduce the volume of the ultimate high-level waste (HLW) form or whether to pursue high-capacity vitrification. The motivation for this study was the growing concern in the early 1990s about the ability of the then-planned Hanford Waste Vitrification Plant to successfully vitrify all single- and double-shelled tank waste. Therefore, this study adopted a decision analysis approach to evaluate the performance of various alternative strategies for processing the tank wastes.
As a result of technical uncertainties at the time of that study, Johnson et al., (1993) recommended that plans and technology development for two alternative bounding technologies for the Hanford tanks waste disposal program be carried forward in parallel:
- extensive separations followed by vitrification at a relatively small plant (possibly even at another DOE site); and
- high-capacity vitrification with minimal pretreatment.
Even though the option of minimal pretreatment followed by high-capacity vitrification was desirable from the point of view of several criteria (e.g., the use of relatively simple and mature
technologies), it was recommended that the final selection between these two technology paths be deferred to preserve the flexibility to choose the best technology in the future, once the uncertainties of each were more fully understood.
The flexible or contingent nature of this recommendation is in keeping with decision analysis theory about the value of information. In situations where the cost of pursuing two strategies in parallel is not exorbitant and a final decision is not absolutely necessary, deferring the final decision until key uncertainties have been resolved can facilitate a better result.
The two-track strategy recommended by Johnson et al., (1993) was never fully implemented at Hanford. In particular, an integrated technology development program focused on extensive separations was never funded. More recently, following completion of the March 1996 Tri-Party Agreement milestone to evaluate enhanced sludge washing, it was determined that enhanced sludge washing was likely to lead to an acceptably small volume of high-level waste, and advanced separations were not required. Moreover, high-capacity vitrification sufficient to permit "minimal pretreatment" is also not being seriously pursued at this time. Thus, Hanford seems to be pursuing a middle course, rather than the two bounding approaches recommended by this study.
The report by Hesser et al., (1995), undertaken in response to a request by the then DOE Assistant Secretary for Environmental Management, is somewhat more far-reaching. In particular, it lays out risk-based cleanup strategies consistent with an expected annual Hanford site funding profile of $1.05 billion and addresses not only tank waste but also nuclear materials, solid waste, environmental contamination to groundwater and soils, and major facilities. The report notes that at the hypothesized funding level, "Cost reduction efforts alone cannot achieve the necessary savings. Changes in schedules and/or scope of cleanup are necessary."
The report considers four general strategies: the existing baseline strategy; an extended version of the baseline strategy; a risk-based strategy; and a composite strategy that addresses land use and mortgage reduction in addition to risk reduction. With regard to tank waste, both the risk-based strategy and the composite strategy involve in situ disposal of liquid waste. In particular, the report states that under a risk-based strategy, "In situ disposal of tank waste would be the preferred option. Technology for in situ disposal should be developed." Hesser et al., (1995) note that this approach would be less acceptable to local stakeholders than the existing baseline strategy, but would reduce worker risk (by nearly an order of magnitude), expected cost (by more than $10 billion), and the time required to complete the cleanup (by more than 10 years); there were no significant differences between the baseline and risk-based strategies with regard to public risk.
Scenario-Based Approaches to Characterization of Hanford Tanks Waste and Technology Development
The next two reports discussed are concerned primarily with developing rather than applying scenario-based methodologies. In particular, Colson et al., (1997) develop a risk-based approach specifically intended to help in making decisions regarding the characterization of tank wastes. This effort was undertaken in response to a request by the DOE Assistant Secretary for Environmental Management, who noted:
"The potential impact of a technically flawed strategy [for waste characterization] is significant. Underestimated risks can result in unnecessary exposure of workers or the public to hazardous materials. Overestimated or ill-defined risks can unnecessarily constrain processing actions and greatly increase the costs of tank waste remediation.
Delays and increased costs from improperly structured efforts can erode public confidence as well as support."
A variant of the methodology developed and applied in this report may well be usable to aid in determining technology development needs instead of waste characterization needs, since both types of activities are aimed at uncertainty reduction. However, the report in some places adopts fairly sophisticated terminology and mathematical methods. More illustrations of how the method can be adapted to situations of varying degrees of complexity might help to ensure that the method is readily applicable in practice.
It is worth noting, by the way, that this report specifically addresses methods of managing potential conflicts of interest, and also recommends "early and open sharing of information and ideas among the project teams, DOE, regulators, stakeholders, tribes, and the public." These issues are clearly critical in today's stakeholder environment.
The draft report by Franklin et al., (1996) lays out a possible method specifically intended for prioritizing some technology development efforts-particularly those efforts undertaken for contingency planning purposes. As such, it is perhaps more directly related to the charge of this committee than the other reports reviewed in this appendix.
The report considers a number of discrete scenarios (ranging from the existing baseline scenario to in situ remediation) based on externalities such as the possible lack of a repository, significant budget reductions, or regulatory changes. Therefore, the methodology described in the report, or a similar approach for addressing the same issues, could be one way of implementing the process described in Chapter 2 of this report. In particular, the report considers a much broader range of scenarios than many of the other studies we have reviewed (including in situ disposal), which is appropriate since the proposed method is intended as a contingency planning tool. To be fully useful and defensible, applications of this method must be based on models that are technically as sound as possible and on parameter estimates that reflect the best available data.
Supporting Analyses of Safety and Environmental Risk
The reports reviewed in this appendix provide information that can be useful in future risk-based decisions among alternative remediation scenarios, potentially providing a basis for achieving reduced risk at reasonable cost. Several probabilistic safety assessments of the Hanford tanks (MacFarlane et al., 1994, 1995a, and 1995b) are described below. The work described in these reports constitutes a comprehensive risk analysis of all 18 tank farms at the Hanford Site.
The effort involved in this analysis was extensive, and hence could not be reviewed in detail by this committee. However, the information they provide can be useful in a number of technology development decisions using a scenario-based approach. For example, MacFarlane et al., (1995a) conclude that only 35 of the 177 waste tanks contribute more than 1 percent each to the total risk prior to waste retrieval. This information can be helpful in identifying plausible scenarios for use in an end state approach. For example, the indication that many tanks pose a relatively low risk could lead to the inclusion of the apparently less-expensive in situ disposal scenario as one of the alternatives to be considered in identifying technology development needs. Similarly, MacFarlane et al., (1995b) conclude that sluicing is associated with substantially lower risks than mechanical or pneumatic retrieval of tank wastes, even in tanks that are known to be leaking. This again can be useful input in focusing further technology development activities, either by encouraging greater emphasis on technologies for sluicing or by ensuring that future development
activities in support of mechanical and pneumatic retrieval technologies focus on minimizing retrieval risks.
The committee also recognizes that these reports were intentionally limited in scope. In particular, the extent to which they consider the effects of human error is unclear, and they cover only retrieval, not such post-retrieval processing steps as vitrification. Therefore, it may well be worthwhile to perform additional risk analyses in the future to address these issues.
Finally, the letter report by Jacobs Engineering (Nelson, 1995) determines how many tanks would need to be retrieved (and how many could be handled by in situ disposal) to meet a 10-4 risk criterion on site. The report does not make any recommendations or assumptions about whether in situ disposal is reasonable or desirable, but rather it examines what the implication would be for disposal requirements.
Based on the criterion of 10-4 or less risk of contracting cancer by a hypothesized future farmer on the site (presumably per lifetime), the report concludes that this criterion could be achieved by retrieving 99 percent of the waste from roughly 40 percent of the tanks and remediating the remainder of the tanks on-site with a cap-and-fill strategy. The report also estimates that this approach would yield substantial cost savings over the current baseline. The cost savings are anticipated to be less than proportional to the reduction in the number of tanks to be retrieved, largely because the cost of the melter needed to vitrify the retrieved wastes would probably not scale linearly with the required throughput.
The report does not discuss the political acceptability of the proposed risk-based approach. Since the tanks assumed to be remediated in situ generally contain only a small fraction of the high-risk radionuclides, capping and filling is assumed to be an acceptable method of stabilizing the nonretrieved wastes. Therefore, the report does not discuss the technology development that would be necessary for more substantial forms of in situ stabilization, such as in situ vitrification.
Summary
The studies summarized here (on average about three years old) all use end state approaches, including at least some consideration of alternative scenarios. The committee believes that these studies contribute significantly to our overall understanding of the health, environmental, and programmatic risk from the Hanford tanks, and that the general approach adopted therein can be a useful way of identifying technology development needs. However, any effect these studies may have had on actual decisions regarding either tank waste remediation in general or technology development programs in particular was not evident from the documentation and other information available to the committee.
References
Colson, S.D., R.E. Gephart, V.L. Hunter, J. Janata, and L.G. Morgan. 1997 (April). A Risk-Based Focused Decision-Management Approach for Justifying Characterization of Hanford Tank Waste: Revision 2. Pacific Northwest National Laboratory PNNL-11231, Richland, WA.
Franklin, A.L., J.L. Stoops, M.J. Quadrel, J.S. Dukelow, Jr., and N. Mahasenan. 1996. Tanks Focus Area Risk-Based, Scenario Based Approach to Technology Development Methods Report: Draft. Pacific Northwest National Laboratory, Richland, WA.
Hesser, W.A., P.A. Baynes, P.M. Daling, T.F. Demmitt, R.D. Jensen, L.E. Johnson, L.D. Muhlstein, S.M. O'Toole, A.L. Pajunen, M.B. Triplett, J.L. Waite, and T.M. Wintczak. 1995 (May). Development of a Risk-Based Approach to Hanford Site Cleanup. Pacific Northwest Laboratory, Richland, WA.
Johnson, M.E., M.L. Grygiel, P.A. Barnes, J.P. Bekemeier, B.D. Zimmerman, and M.B. Triplett. 1993 (March 31). Tank Waste Decision Analysis Report. Westinghouse Hanford Company WHC-EP-0617 Draft, Richland, WA.
MacFarlane, D.R., T.F. Bott, L.F. Brown, D.W. Stack, J. Kindinger, R.K. Deremer, S.R. Medhekar, and T.J. Mikschl. 1994 (May). Probabilistic Safety Assessment for Hanford High-Level Waste Tank 241-SY-101. Los Alamos National Laboratory, Los Alamos, NM, and PLG, Inc.
MacFarlane, D.R., D.W. Stack, J. Kindinger, R.K. Deremer, S.R. Medhekar, and T.J. Mikschl. 1995a (September). Probabilistic Safety Assessment for Hanford High-Level Waste Tanks. Los Alamos National Laboratory and PLG, Inc., LA-UR-95-1900, TSA-11-94-R110.
MacFarlane, D.R., D.W. Stack, J. Kindinger, R.K. Deremer, S.R. Medhekar, and T.J. Mikschl. 1995b (July). Probabilistic Safety Assessment for Retrieval of Hanford Single-Shell Tank Wastes. Los Alamos National Laboratory, Los Alamos, NM, and PLG, Inc.
Nelson, M.E. 1995 (May 10). Selective Retrieval. Letter report to C. Haass, U.S. Department of Energy, Richland Operations Office, Richland, WA. Jacobs Engineering Group, Inc., Kennewick, WA.