New Ways for Optimizing Equilibrium Cycle Reactor Reload Patterns

To judge the performance of a selected reload pattern over more than one operation cycle, considering the equilibrium cycle behavior is a suitable and well-defined instrument. As it takes many iterations in successive cycles to approach the equilibrium cycle sufficiently, this puts an additional burden to the already very demanding problem of finding the optimum reload pattern for a next operation cycle only. At the Delft University of Technology two different ways are investigated to tackle the problem, while avoiding most or all of the normally necessary iterations in core cycle behavior for an equilibrium cycle. One way is to treat the optimization problem by a mixed-integer nonlinear programming approach. This involves the formulation of the problem, including the neutronics behavior of the core, in a set of algebraic equations. Then the equilibrium cycle behavior can be included simply by the addition of one equation, relating the nuclide densities at each reactor node at begin of a cycle to the nuclide densities at the end of the cycle. This approach is demonstrated using a rather simple neutronics model. The other way of dealing with the problem is to apply perturbation theory to evaluate the core behavior for a certain reload pattern by considering this as a perturbation to a pattern with known characteristics. The theory for deriving the change in nuclide densities due to a change in reload pattern for an equilibrium cycle is summarized and essential characteristics are discussed. A numerical example demonstrates some of the applications for reload pattern optimization. (author)