In December 1984, the Department of Energy (DOE) published draft environmental assessments (EAs) to support the proposed nomination of five sites and the recommendation of three sites for characterization for the first radioactive-waste repository. A chapter common to all the draft EAs (Chapter 7) presented rankings of the five sites against the postclosure and the preclosure technical siting guidelines. To determine which three sites appeared most favorable for recommendation for characterization, three simple quantitative methods were used to aggregate the rankings assigned to each site for the various technical guidelines. In response to numerous comments on the methods, the DOE has undertaken a formal application of one of them (hereafter referred to as the decision-aiding methodology) for the purpose of obtaining a more rigorous evaluation of the nominated sites.

The application of the revised methodology is described in this report. The method of analysis is known as multiattribute utility analysis; it is a tool for providing insights as to which sites are preferable and why. The decision-aiding methodology accounts for all the fundamental considerations specified by the siting guidelines and uses as source information the data and evaluations reported or referenced in the EAs. It explicitly addresses the uncertainties and value judgments that are part of all siting problems. Furthermore, all scientific and value judgments are made explicit for the reviewer. An independent review of the application of the decision-aiding methodology has been conducted by the Board on Radioactive Waste Management of the National Academy of Sciences; the comments of the Board are included as an appendix to this report.

In spite of its advantages, the formal analysis cannot address every aspect of the site-recommendation decision and thus its results will not form the sole basis for that decision. The site-recommendation decision is analogous to a portfolio-selection problem because the DOE is not choosing a single site for repository development; rather, the DOE must choose, from a suite of five well-qualified sites, three sites for site characterization. Combinations of three sites possess properties that cannot be attributed to individual sites, such as diversity of geohydrologic settings and rock types. Thus, the three sites indicated as moat preferable by the multiattribute utility analysis reported here do not necessarily constitute the most preferred combination when these portfolio effects are taken into account. The relative advantages of other combinations of three sites as portfolios together with other information the Secretary of Energy believes is important to making the decision are examined in a separate report.