Experimental analysis and uncertainty quantification using random sampling technique for ADS experiments at KUCA

Nuclear data-induced uncertainties of neutronics parameters (neutron multiplication factor k_eff, one-point kinetics parameters and prompt neutron decay constant α) are quantified for lead-bismuth zoned accelerator-driven system experiments at the Kyoto University Critical Assembly, in order to contribute validation for subcritical core analysis. The random sampling technique using SCALE6.2.1/Sampler/NEWT/PARTISN is utilized for the validation and the uncertainty quantification, because the random sampling technique is applicable for a problem which is not easy to apply the perturbation theory. Consequently, it is confirmed that the numerical results of α reasonably agree with the experimental ones, compared with the nuclear data-induced uncertainties. In addition, it is clarified that the nuclear data-induced correlations between α and k_eff and between α and neutron generation time Λ are strongly negative and positive, respectively. This fact implies that the numerical predictions of k_eff and Λ can be improved by the data assimilation technique using subcritical experimental results of α, which can be directly measured even for a deep subcritical system. (author)