CASMO-5 Criticality Analysis

Sammanfattning

This project evaluated the bias and uncertainty of keff of samples of spent nuclear fuel from various PWRs, using measured isotope concentrations. This project initially covered 260 samples from sixteen reactors. These samples had different fuel types (UO2 or MOX), with various enrichments (0.231% - 4.657%) and burnup ranges (3.4 - 75.0 GWd/tUi). The majority of the operating history information and reactor design specifications used for the CASMO-5 simulations were obtained from the database SFCOMPO 2.0, with additional data sourced from various technical reports. The simulated predicted isotopic concentrations were compared with the measured isotopic concentrations, which were also obtained from the database. For the isotopic concentration assessment, the predicted concentrations were adjusted by tuning the power density using 148Nd or burnup as a calibration point for the power density, achieving an uncertainty of ± 1%. Before tuning in, the average isotopic concentration uncertainty was -0.44%, over all reactors, with a range of -74.65% to 53.27%, after excluding outlying samples. A larger number of samples were ultimately excluded from the the final criticality analysis, due to missing information on design parameters, or because of unreasonably large differences between measured and calculated isotopic concentrations from the CASMO-5 simulations. After the isotopic concentration assessment, 128 samples were evaluated for the criticality analysis. The criticality analysis was performed using KENO-Va simulations to evaluate the effective multiplication factor, keff . Four simulation variants were used to evaluate the impact of different isotopic concentration inputs on KENO-Va results: one with predicted isotopic concentrations, and three with measured isotopic concentrations where missing or zero values were replaced with predicted values adjusted using the median, 10th percentile, or 90th percentile of the relative differences between predicted and measured concentrations. The result from this analysis showed the average bias across all samples was 862 pcm, and the average uncertainty was 2173 pcm. The total ∆ki in this project was determined to be -3921 pcm. This value is significantly different than the ∼ -1841 pcm reference value from the technical reportof ISG-8R3, which cites NUREG/CR-7108 which is a similar analysis on PWRs. It is important to minimize these biases and uncertainties to ensure accurate and reliable criticality analyses of spent nuclear fuel, which can improve the reliability of storage and disposal safety margins.