Results of Nuclear Design Accuracy Evaluation on BN-600 Hybrid Core

Shono, Akira; Sato, Wakaei*; Hazama, Taira ; Iwai, Takehiko*; Ishikawa, Makoto

Nuclear design accuracy on the BN-600 hybrid core has been evaluated using the JNC's nuclear analysis system for FBR cores, by utilizing the critical experiment analysis results on BFS-62 configuration that had been obtained under JNC's efforts for Russian surplus weapons plutonium disposition. In the BN-600 hybrid core design, a part of the current UO2 fuel region is replaced by MOX fue1, and the Peripheral blanket region by stainless steel reflectors, respectively. These changes were simulated in a series of critical experiment configurations (BFS-62-1 to 4). Based on the analysis results on both BFS-62 configurations and other fast reactor cores, nuclear design accuracy on the BN-600 hybrid core has been evaluated by applying both the group constant adjustment method and the bias method. Evaluated nuclear parameters include, the criticality, fission rate distribution, sodium void reactivity, control rod worth, burn-up reactivity loss, etc. It is concluded, by applying the group constant adjustment method, that the evaluated accuracy (uncertainty) of most of the nuclear parameters can be decreased to less than half of those based on the basic nuclear constant without reflecting any experimental data. The improvement was mainly achieved by reducing the covariance of the iron elastjc cross section. This significant effect results from the feature of the BN-600 hybrid core, which has relatively larger power density, adopts U235 as the main fissile nucljde, and has the stainless steel reflector surrounding the fuel region. In addition, good consistency of analysis results between the BFS and other fast reactor cores is confirmed. Information obtained by BFS-62 experiment show significant contribution to the accuracy improvement. It is also found that the bias method shows less significant effects on the accuracy improvement than the group constant adjustment method. Furthermore, the bias method may degrade the accuracy for certain nuclear parameters that have large e