Application of Mathematical Optimization Techniques to Nuclear Reactor Reload Pattern Design.

摘要

Nuclear power plants are an important source of electricity in many countries. The uranium that is used in these plants is very costly, and in order to reduce both fuel cost and nuclear waste, it should be utilized efficiently. Nuclear fuel management is concerned with the question of how to utilize the uranium in a safe and efficient way. Our research dealt with one important aspect of fuel management, namely the placement of fuel elements in the reactor core. A nuclear reactor core generally consists of a number of nuclear fuel elements, which can be divided into groups of different ages. At the end of a core operation cycle (which usually lasts from 12 up to 18 months), the oldest group is discarded from the core, the remaining fuel elements may be reshuffled in the core grid, and the empty positions are filled with new, unburned fuel elements. A scheme that tells where each element should be moved to is known as a reload pattern. The positions of the elements in the reload pattern can have big influences on the efficiency of the reactor. In current practice, such patterns are usually designed with the use of techniques that are mathematically known as direct search heuristics.