A generalized optimization methodology for isotope management

Abstract

This research, funded by the Department of Energy's Advanced Fuel Cycle Initiative Fellowship, was focused on developing a new approach to studying the nuclear fuel cycle: instead of using the trial and error approach currently used in actinide management studies in which reactors are designed and then their performance is evaluated, the methodology developed here first identified relevant fuel cycle objectives like minimizing decay heat production in a repository, minimizing Pu-239 content in used fuel, etc. and then used optimization to determine the best way to reach these goals. The first half of this research was devoted to identifying optimal flux spectra for irradiating used nuclear fuel from light water reactors to meet fuel cycle objectives like those mentioned above. This was accomplished by applying the simulated annealing optimization methodology to a simple matrix exponential depletion code written in Fortran using cross sections generated from the SCALE code system. Since flux spectra cannot be shaped arbitrarily, the second half of this research applied the same methodology to material composition of fast reactor target assemblies to find optimal designs for minimizing the integrated decay heat production over various timescales. The neutronics calculations were performed using modules from SCALE and ERANOS, a French fast reactor transport code. The results of this project showed that a thermal flux spectrum is much more effective for transmuting used nuclear fuel. In the spectral optimization study, it was found that a thermal flux spectrum is approximately five times more effective at reducing long-term decay heat production than a fast flux spectrum. This conclusion was reinforced by the results of the target assembly material optimization study, which found that by adding an efficient moderator to a target assembly designed for minor actinide transmutation, the amount of decay heat generated over 10,000 years of cooling can be reduced by over 50% through a single pass in a fast reactor without exceeding standard cladding fluence limits.