Quantitative Analysis for Attaining and Maintaining Continuity of Knowledge

摘要

The current American administration is markedly pro-nuclear and currently expending significant time and resources to make the American nuclear industry competitive internationally. As such, a long term solution for storage of spent nuclear fuel (SNF) is required, and the Yucca Mountain nuclear waste repository is again being prepared for licensing. As the specifics for what containment and surveillance (C/S) measures need to be implemented to assure that nuclear material accountancy (NMA) allows the inspectors to draw a safeguards conclusion, a method is required for quantifying the level of confidence that the SNF stored in perpetuity is in fact still where it has been concluded to be. Elaborate probability models are developed and utilized by financial analysts to forecast economic behavior and by insurance companies to predict catastrophic events, and those models that accurately represent the real world result in measured financial success for those that develop them. As a corollary, a similar statistical model needs to be developed to inform the C/S practices employed at the Yucca Mountain site as well as at other long term storage facilities in order to quantitatively attain and maintain continuity of knowledge (CoK). A statistical model based on existing data such as NMA violations, standing C/S methods, etc. and based on how proposed C/S methods will impact CoK uncertainty is developed and presented in this paper. This proposed statistical model can serve as a "way ahead" for future statistical modelling in order to provide actionable and reliable data to decision-makers.