Estimation of The error of boron carbide oxidation model in SOCRAT code using The BOX experiments

In the code validation process, a standard approach to quantify the capability of an individual model to simulate a relevant process or phenomenon that occurs during an accident at NPP implies calculation of the bias, or a difference between the calculated and measured values for parameters of interest. The bias values do not always provide information about the accuracy of the mathematical model since the uncertainties that arise in numerical modeling and experiments are ignored. The paper discusses the modeling error, associated with approximations and simplifications in the boron carbide oxidation model (part of the best estimate severe accident code SOCRAT). Isothermal and transient BOX tests in the temperature range of 800–1500 °C are used as a reference basis for validation. The modeling error is estimated by an advanced approach based on the ASME V&V 20 standard with account for uncertainties associated with the input data and the measured values.