Risk assessment of K Basin twelve-inch and four-inch drain valve failure from a postulated seismic initiating event

摘要

The Spent Nuclear Fuel (SNF) Project will transfer metallic SNF from the Hanford 105 K-East and 105 K-West Basins to safe interim storage in the Canister Storage Building in the 200 Area. The initial basis for design, fabrication, installation, and operation of the fuel removal systems was that the basin leak rate which could result from a postulated accident condition would not be excessive relative to reasonable recovery operations. However, an additional potential K Basin water leak path is through the K Basin drain valves. Three twelve-inch drain valves are located in the main basin bays along the north wall. Five four-inch drain valves are located in the north and south loadout pits (NLOP and SLOP), the weasel pit, the technical viewing pit, and the discharge chute pit. The sumps containing the valves are filled with concrete which covers the drain valve body. Visual observations indicate that only the valve's bonnet and stem are exposed above the basin concrete floor for the twelve-inch drain valve and that much less of the valve's bonnet and stem are exposed above the basin concrete floor for the five four-inch drain valves. It was recognized, however, that damage of the drain valve bonnet or stem during a seismic initiating event could provide a potential K Basin water leak path. The objectives of this analysis are to: (1) evaluate the likelihood of damaging the three twelve-inch drain valves located along the north wall of the main basin and the five four-inch drain valves located in the pits from a seismic initiating event, and (2) determine the likelihood of exceeding a specific consequence (initial leak rate) from a damaged valve. The analysis process is a risk-based uncertainty analysis where each variable is modeled using available information and engineering judgement. The uncertainty associated with each variable is represented by a probability distribution (probability density function). Uncertainty exists because of the inherent randomness associated with the distribution of values that a variable may assume, and because of a lack of knowledge concerning a variable. Engineering judgement and technical information are used to develop the variable probability density functions, but the bounds of the probability density function are based on physical limitations. The uncertainty, described by probability distributions, is propagated through the analysis by Monte Carlo convolution techniques. The corresponding results are developed as a probability distribution and expressed in terms of the corresponding complementary cumulative distribution function (''risk curve'').