INVESTIGATION OF THE BEST ESTIMATE AND ANALYSIS OF UNCERTAINTY METHODOLOGY TO LOSS OF COOLANT ACCIDENTS IN THE CANDU REACTOR

Abstract

Large loss of coolant accidents in the CANDU reactor are simulated using the Modal Approximation Code suite, the results of which are substantiated using industry codes. Results for the integrated hot bundle energy are used to perform a comparison of the Best Estimate and Analysis of Uncertainty (BEAU) safety analysis methodology with the Code Scaling, Applicability and Uncertainty (CSAU) and the Automated Statistical Treatment of Uncertainty (ASTRUM) methodologies. The three safety analysis methods are applied to the code results to determine the 95/95 value of the hot bundle energy associated with the loss of coolant power pulse. The code results are also used to examine the response robustness of each method. The 95/95 values determined using the three safety analysis methods are consistent to within 0.59%. The CSAU results represent the lowest value of bundle energy at 5.763 MJ (over the power pulse duration). The ASTRUM value is calculated to be 5.772 MJ and is established as the least costly method to implement. The BEAU 95/95 result is determined to be the highest at 5.797 MJ indicating lesser margin for the safety evaluation criteria The BEAU methodology is the most complex to implement, the resultant response function however is found to be the most robust to perturbations in the reactor physics parameters associated with the loss of coolant accident. In analyses were the parent distributions of these parameters are not well defined or where significant reanalysis of results may be required, the BEAU methodology is the most suitable safety analysis method.