Validation of a proposed MUF evaluation method based on the Monte Carlo technique

Since 2015, the Republic of Korea (ROK) has conducted material balance evaluations (MBEs) through national inspections under the IAEA's State Level Approach. MBE models based on analysis of variance (ANOVA) and expression of uncertainty in measurement (GUM) have been applied to assess the measurement uncertainty of material unaccounted for (MUF). However, these models may overestimate or underestimate propagated the systematic error, potentially leading to biased results. To decrease bias, a Monte-Carlo (MC) method with random numbers (RNs) is proposed. RNs, based on a normal distribution with the International Target Value’s uncertainty, were applied to the U-235 weights of virtual nuclear material items. Simulations generating RNs were conducted for up to 1,000 iterations and adjusted MUF values were calculated 30 times per iteration. Through iterative repeated calculation, an of 37.2167 kg was calculated under the optimized iteration number of 700. This result was similar with the previous result for MUF uncertainties yielded by ANOVA and GUM based MBE models. The proposed MC-based MBE model provides another option to evaluate MUF and supplement the IAEA and GUM based MBE models. It enables reliable assessments with limited uncertainty data and can support facility-specific measurement guidelines in national inspections.