Assessment of pressure vessel failure due to pressurized thermal shock with consideration of probabilistic-deterministic screening and uncertainty analysis

The research presented in this dissertation is intended to address open issues encountered in a U.S. Nuclear Regulatory Commission (NRC)-sponsored project aimed to re-evaluate pressurized thermal shock (PTS) risk in pressurized water reactors. The objectives of the research were to (1) provide a framework including uncertainty characterization and treatment for PTS studies; (2) develop a set of screening criteria for scenario selection; (3) develop a simplified and reliable computation tool for PTS analyses performed for other than re-licensing or regulatory purposes; and (4) propose to Oak Ridge National Laboratory (ORNL), developer of a computation code for PTS re-evaluation project under contract to the NRC, a procedure on probabilistic fracture mechanics modeling and uncertainty treatments.; The PTS evaluation consists of three disciplines: probabilistic risk assessment (PRA), thermal hydraulics, and probabilistic fracture mechanics. The major contribution of this research was to provide a framework for the PTS analysis including uncertainty characterization and treatment, which is consistent with PRA principles. The uncertainty analysis proposed in this research adopted the concept of distinction between aleatory uncertainty and epistemic uncertainty for practical purposes.; To ease the computational burden incurred by introducing thermal hydraulics uncertainties, two approaches were proposed: (1) to develop a set of scenario screening criteria, and (2) to develop a simplified and reasonably accurate fracture mechanics computing package that performs fracture mechanics calculations rapidly and conservatively. MATH-PFM is a package that can fulfill the goal of the second approach. It was implemented in MATHMATICA, a popular commercial computation software. MATH-PFM was verified against the ORNL code and proved to be slightly conservative to assure accurate screening. The screening criteria developed in this research by using MATH-PFM were able to exhaustively screen out scenarios that are insignificant to PTS risk.; In support of the ORNL effort in developing a PTS computation code, this research proposed a new methodology for PTS studies and devised a detailed procedure to be implemented in the code.