The high-speed statistical criticality evaluation method based on the multidimensional interpolation for on-demand criticality evaluation

In the decommissioning of Fukushima Daiichi Nuclear Power Station (NPS), the removal of fuel debris cannot be avoided. The fuel debris is currently considered to be sub-critical. Volume ratio of debris/water has strong sensitivity to the criticality. Thus, the criticality estimation is required in the debris removal process. While, the best estimation of the criticality is difficult due to the unclear properties and/or geometries of debris, therefore, random sampling with Monte Carlo calculation was used in current evaluation. However, the Monte Carlo calculation spend three days to one week for the evaluation due to its high calculation cost. Therefore, it is difficult that Monte Carlo apply to on-demand statistical criticality evaluation. For these background, multidimensional interpolation was applied to the statistical criticality evaluation. In this study, fundamental validation was examined by comparing with Monte Carlo results. The proposed method was 614,400 times faster than Monte Carlo calculation, additionally, difference of mean value was 0.9 % δk. As a result, proposed multidimensional interpolation showed good agreement with direct Monte Carlo calculation. In the future work, simplification of the model, evaluation of applicable limit and operation procedure will be examined. (authors)