CANDU-6 fuel optimization for advanced cycles

Abstract We implement a selection process based on DRAGON and DONJON simulations to identify interesting thorium fuel cycles driven by low-enriched uranium or DUPIC dioxide fuels for CANDU-6 reactors. We also develop a fuel management optimization method based on the physics of discrete on-power refueling and the time-average approach to maximize the economical advantages of the candidates that have been pre-selected using a corrected infinite lattice model. Credible instantaneous states are also defined using a channel age model and simulated to quantify the hot spots amplitude and the departure from criticality with fixed reactivity devices. For the most promising fuels identified using coarse models, optimized 2D cell and 3D reactivity device supercell DRAGON models are then used to generate accurate reactor databases at low computational cost. The application of the selection process to different cycles demonstrates the efficiency of our procedure in identifying the most interesting fuel compositions and refueling options for a CANDU reactor. The results show that using our optimization method one can obtain fuels that achieve a high average exit burnup while respecting the reference cycle safety limits.