Prediction plays a prominent role in both the linear methodology and adaptive staging scheme. Both approaches use prediction as tool for the acquisition of pertinent information about elements of repository performance. The distinction between the ways each approach uses prediction is one of emphasis.
In the linear methodology, prediction is heavily used and often becomes the goal of some public policies instead of a means to study repository performance. For instance in the Yucca Mountain Project, the goal of providing evidence that the repository will safely contain radioactive waste has been substituted by a goal of predicting the repository performance using computer simulations. While the DOE does not state this in their literature, the goal substitution is evident in the fact that the TSPA is the relied upon method for arguing that the repository is safe. Moreover, as problems with the geology of Yucca Mountain have surfaced over the years, the DOE has responded by suggesting more heavily engineered waste canisters and drip shields to prolong the amount of time that passes before the waste seeps into the surrounding geological strata. Prediction is used to demonstrate that these engineered barriers will perform as expected. Furthermore, the use of novel alloy materials in waste canisters only increases DOE reliance on prediction because these alloys don’t exist in nature where their response to corrosion over long periods of time could be studied.
In contrast, adaptive management uses prediction where it is applicable but acknowledges that prediction has limits and that social forces play a primary role in making policy decisions. Whereas the DOE’s answer to the revelation that the unsaturated zone in Yucca Mountain could be much wetter than previously thought was to introduce engineered barriers, adaptive management might seriously consider another disposal site in light of this information and increasing local opposition to the project. In principle, geologic repositories are effective at containing waste for extended periods of time due to the particular geology of the site. Whenever new data surface that reveal the geology might different than expected, it is questionable whether the site still can be classified as a geologic repository. Some have asked this very question of Yucca Mountain.
Adaptive staging also emphasizes a level of outside participation in the repository design process that a linear methodology does not. By relying almost exclusively on prediction, a linear framework excludes the voices of those who don’t contribute to the production of predictions. In a flexible, stepwise process, the technical information in the site investigation is peer-reviewed by outside scientists and presented to local communities, local governments, and other organizations for comment and evaluation. Thus, any decisions are the result of a synthesis of information from scientific studies, socio-political investigations, and public opinion, rather than being the product of only technical and bureaucratic considerations. By giving prediction an important but not exclusive role in policymaking, adaptive staging can make scientifically informed and socially sensitive decisions. If prediction is the only barometer for repository viability, public opposition is inevitable.