On-Line Critical Control Rod Pattern Prediction Algorithm for BWR Plant Startup.

Abstract

This paper describes an on-line algorithm for predicting the critical control rod pattern, which has been developed to reduce the mental strain on operators while withdrawing control rods in the BWR plant startup operation. The proposed algorithm estimates a target eigenvalue (eigenvalue bias) for a three-dimensional neutron kinetics model with a neutron source incorporating actual neutron detector readings. The critical control rod pattern is then predicted based on the estimated eigenvalue bias. The algorithm has been verified using data obtained from an actual startup operation on a BWR model-5 plant, and the estimated eigenvalue bias agreed well with the effective multiplication factor at the criticality actually determined from the operator's judgement.