Control Blade Pattern Optimization Using Evolutionary Algorithm for the Canadian SCWR

Abstract

The Canadian SCWR has been simulated with the diffusion code PARCS, including proposed reactivity control devices. Transport calculations of the lattice cell have been performed with SCALE/TRITON to determine the few-group neutronic parameters as functions of burnup for input in PARCS. 89 cruciform control blades have been implemented in the model to remove part of the excess reactivity and to balance radial core power. The goal of this study is to minimize the channel power ripple by adjusting the control blades throughout the cycle. This was achieved by coupling PARCS to an optimization software (DAKOTA). An automated process has been developed to find a control blade sequence that holds each channel power as close to constant as possible while keeping the core reactivity around the desired value throughout the cycle, while also maintaining acceptable maximum nodal powers.