Probabilistic Risk Assessment Study for Uncertainty Quantification of Spent Pebble-Bed Dry Cask Storage System Operation

Abstract

Abstract The development of on-site interim dry storage systems has become an important issue, not only for existing nuclear power plant (NPP) operations but also for next-generation NPP designs, due to space limitations in the spent fuel pool. To ensure the safety of the spent fuel dry storage system operation, probabilistic risk assessment (PRA) elements are utilized throughout the non-light water reactor (LWR) lifecycle. Uncertainty quantification is an essential aspect to evaluate the risk through PRA performance, particularly in assessing off-site risks associated with advanced NPPs and related facilities. This study implements PRA technical requirements for uncertainty quantification for the dry cask system storing spent fuel pebbles containing Tri-Structure Isotropic particles in stainless canisters on-site. The standardized PRA elements in this paper are demonstrated sequentially through 1) initiating event selection, 2) event sequence analysis through event tree, 3) data analysis from the PRA-related materials, 4) event sequence quantification including uncertainty quantification using Phoenix Architect, 5) mechanistic source term analysis by employing ORIGEN 2.2, 6) radiological consequence analysis through MicroShield simulation, and 7) risk integration for quantifying the risk using the Frequency-Consequence (F-C) target. The case study assumes a canister drop as an initiating event that can lead to the release of fission products from the spent fuel. For analyzing and quantifying the event sequences, damage states of pebbles and canisters, and operation failure of the heating, ventilation, and air conditioning system are considered. To quantify the radiological dose at 5 km of the exclusion area boundary, three release categories are established, including direct exposure, noble gas release, and radionuclide release. The uncertainties, including epistemic uncertainty and aleatory uncertainty, are considered for event sequence and consequence quantification aspects. Due to the lack of information for both the initiating event frequency and failure probabilities of events, the constrained noninformative distribution is utilized to reduce the impact of epistemic uncertainty. Also, assumptions are established for the damaged pebbles’ condition, such as number of pebbles and release fraction uncertainty. The uncertainties are quantified using UNCERT, which is a module in the Phoenix Architect. The integrated risk, considering uncertainties regarding the frequencies and consequences, is shown in the F-C curve, including licensing basis event (LBE) target line, for the non-LWR design as the risk-informed conceptual storage system operation. The safety cases for the release categories, along with their associated event frequencies, are validated by ensuring that their uncertainty ranges fall below both the F-C target lines and the LBE risk-significant criterion. The suggested methodologies can be extended to improve the analysis of LBE for dry cask storage operation for pebble-bed filled canisters by examining the additional topics, including uncertainty due to pebbles’ movement during conveying or dropping or damaged pebbles’ distribution in a canister, and so on.