Addressing epistemic uncertainty in a radioactive waste disposal facility siting: examples of applications

摘要

The siting of a radioactive waste disposal facility involves site survey, site selection and site confirmation. Throughout the siting process a series of calculations and data analyses are performed with increasing complexity as increasing amounts and more detailed data becomes available. Each one of these stages requires a multidisciplinary approach to evaluate the most important aspects that affect the performance and safety of the disposal facility. The evaluation of these aspects has uncertainties that are related to the complex interactions amongst a large number of parameters, lack of data, and ignorance as to natural processes. Simplifications are used to reduce the size and complexity of the computational model and this also causes uncertainty in the validity of the predicted performance. Due to lack of data and ignorance of the exact physical processes that are involved, linguistic expressions are used to describe natural processes. Expressions such as "strongly fractured rock" or "low U content" are very powerful in the sense that they can carry robust information that has an important impact on decision-making. Therefore, a mathematical representation of these expressions would be an important tool for the analyst to effectively use this linguistic information. This paper presents some applications of the fuzzy logic tools to address the epistemic uncertainty throughout the stages of the disposal facility implementation.%National Nuclear Energy Commission, CDTN/CNEN, Belo Horizonte, MG, Brazil ;Brookhaven National Laboratory BNL, Upton, NY, USA;