Application of Bayes Method in Evaluation of ROP/NOP Trip Setpoint
Abstract of the technical paper presented at:
37th Annual Conference of the Canadian Nuclear Society
Dumitru Serghiuta, John Tholammakkil, Canadian Nuclear Safety Commission, Ottawa, Canada
David Alain Stephens, McGill University, Montréal, Canada
Anthony O’Hagan, Emeritus Professor, Sheffield University, UK
Independent Regulatory evaluation of the ROP/NOP trip setpoint under heat transport system aging conditions is necessary for verification and confirmation of the adequacy of licensees’ proposed values of the installed trip setpoints. CNSC staff have undertaken a multi-phase research project to develop a Bayesian statistical framework (SF) and software to support regulatory independent verification and enhance the input information for a risk-informed decision making (RIDM) process.
The Bayesian SF has been designed to estimate the functional failure distributions for a ROP/NOP system with fixed design control variables (number of detectors, location of detectors and value of trip setpoint). Specifically, the Bayesian SF should answer the question: "What is the estimated frequency of NOP system (functional) failure and the uncertainty around that estimate?"
The Bayesian approach has been selected because the design ROP/NOP methodology is based on a probabilistic model. By design, Bayesian methods natively consider the uncertainty associated with the parameters of a probability model (even if those uncertain parameters are believed to be fixed numbers). They are often used, as proven practice, in the areas of reliability and probabilistic assessment as the proper way to make formal use of subjective information.
The paper describes the theory, technical basis and software requirements, which were developed during the first phase of the project and completed in March 2016; the approach taken in developing the algorithms coded in the software; and some of the criteria for verification and manufacturing of benchmark problems for qualification purposes.
To obtain a copy of the abstract’s document, please contact us at email@example.com or call 613.995.5894 or 1.800.668.5284 (in Canada). When contacting us, please provide the title and date of the abstract.
- Date modified: