The Optimization of Cobalt Adjusters with OPTEX-3

Abstract

A numerical approach is taken in OPTEX to optimize in-core devices and fuel management in CANDU reactors at equilibrium refuelling. The approach consists in linearizing the system characteristics about points in the feasible domain of the state variables, and then using mathematical programming techniques to direct the search towards an improved state. Variational techniques, found in Generalized Perturbation Theory (GPT), are used to define the true gradients of the system characteristics. This paper will present recent additions and improvements to the code, dealing with the introduction of a variable load of cobalt in the adjuster rods. The power shaping capabilities of OPTEX-3 will also be discussed.