Methodology for the fuel axial design for BWR type reactors using evolutive computation

A methodology for the fuel assembly axial design of Boiling Water Reactors (BWR) using an optimization process based on evolutionary computing is presented. The methodology was automated in a computational system. The main characteristics of the system are presented, showing how the Genetic Algorithms (G A) optimization process is linked to the three dimensional, neutronics and thermal-hydraulics reactor simulator Core Master Presto (CM-Presto), which is used to evaluate the fuel axial design performance. Each fuel design evaluation is performed through the simulation of a reload pattern where the fresh fuel has the characteristics of the fuel axial design to be evaluated. The objective function of the process searches the design with the minimum average enrichment which allows to meet the cycle target energy, under the thermal limits and reactivity constraints, which are used as safety criteria of the nuclear fuel design. The methodology is validated searching an optimal axial design for Laguna Verde (L V) cycle 5 Unit 1. The optimization system shows very good results concerning the convergence and the performance to search optimal fuel designs. (Author)