First-order perturbation for uncertainty quantification in the deterministic truncation of Monte Carlo method

Abstract

A new approach to estimate SD from uncertainties of CMFD parameters was tested. The method was able to conservatively predict RESD of the simulation based on several tens of inactive cycles. For this estimation, the first-order perturbation formula was used to replace power iteration step, which accounts for significant computation time. The first-order perturbation formula for cross sections was able to result almost identical multiplication factors from the samples, and SD was very similar. In the future, additional effort to precisely estimate RESD based on a few inactive cycles will be further studied. Also, SD estimation of pin-power distribution will be done shortly.