On a Best-Estimate Approach to the Calculation of Dryout Probability During BWR Transients

Abstract

A method is proposed whereby uncertainty of any dryout margin measure (figure of merit) may be quantified when the only experimental information available for validation is whether dryout has occurred or not. The method does not involve the heater temperature, except as a discrete dryout indicator. This is an advantage when analysing anticipated operational occurrences for which the acceptance criterion refers exclusively to the probability of dryout occurrence. The derived uncertainty provides a direct relation between the simulated dryout margin and the aforementioned probability. Furthermore, the method, which is based on logistic regression, has been designed to be consistent with more common parametric methods of uncertainty analysis that are likely to be used for other parts of a thermal hydraulic model. One example is provided where the method is utilized to assess statistical properties, which would have been difficult to quantify by other means