Artificial neural network modeling for fission gas release in LWR UO2 fuel under RIA conditions

A fission gas release (FGR) model was developed by using an artificial neural network method to predict fission gas release in UO2 fuel under reactivity initiated accident (RIA) conditions. Based on the test data obtained in the CABRI test reactor and nuclear safety research reactor, the model takes into account the effect of the five parameters: pellet average burnup, peak fuel enthalpy, the ratio of peak fuel enthalpy to pulse width, fission gas release during base-irradiation, and grain size of a fuel pellet. The parametric study of the model, producing a physically reasonable trend of FGR for each parameter, shows that the pellet average burnup and the ratio of peak fuel enthalpy to pulse width are two of the most important parameters. Depending on the combination of input values for the five parameters, the application of the model to a fuel rod under typical RIA conditions of light water reactor produces 1.7-14.0% of FGR for the pellet average burnup ranging from 20 to 70 MW d/kg U. (C) 2010 Elsevier B.V. All rights reserved.