Development of methodology of probabilistic safety assessment for radioactive waste disposal in consideration of epistemic uncertainty and aleatory uncertainty

ABSTRACT Probabilistic safety assessment has not been performed for radioactive waste disposal owing to the difficulty of dealing with the probability distributions of the parameters included in the long-term safety assessment of radioactive waste disposal. In this study, we develop a methodology of probabilistic safety assessment in consideration of both epistemic uncertainty and aleatory uncertainty, in which the probability density function (PDF) and cumulative distribution function (CDF) of the maximum annual dose can be calculated. We also propose an approach to demonstrating dose assessment results in compliance with the stepwise target annual doses of likely and less-likely scenarios according to the occurrence probability of the scenario without classifying the probabilities of parameters involved prior to the safety assessment. For the likely scenario, we can employ the larger of the modal value of the PDF and the 50th percentile of the CDF to meet the target annual dose (10 µSv y-1). For the less-likely scenario, we can adopt the 95th percentile of the CDF as the assessment result for comparison with the target annual dose (300 µSv y-1).