An application of an evolutionary algorithm in the study of a boiling water reactor fuel lattices with minor actinides

In this work, a method for carrying out the neutronic design of a boiling water reactor fuel lattices is shown. Generated designs include a vector of minor actinides that was added in one of the positions of a typical 10x10 fuel assembly. This vector is composed of neptunium, americium and curium concentrations. The decision variables used in this design process were the average enrichment of U-235 (w/o), the concentration of tri-oxide of gadolinium Gd2O3 (w/o), neutronic multiplication factor in an infinite medium (k-infinity), and the local power peaking factor of the fuel lattice. In addition to, a fixed average void fraction (%), not presence of control rod, and the temperature of fuel, among other are assumed as search conditions. The considered restrictions on decision variables, established that k-infinity must be into the interval of ±100 pcm with respect to reference value; the average enrichment of fuel lattice must be under the reference fuel enrichment but not less than 1.0% with respect to the reference lattice; a similar restriction is also assumed for neutronic burnable poison concentration, it must be inside the range defined by ± 1.0% around reference value; local power peaking factor must not be greater than assumed reference. CASMO4 code was used for carrying out the simulation of fuel lattices through the designing process. Obtained fuel lattices fulfilled each one of the restrictions applied in this solution approach to fuel lattice design problem, when some of positions inside the array may selectively include concentrations of minor actinides.