Prediction of void fraction uncertainty bands in BWR fuel bundles using the CIAU methodology

One of the tasks of the Boiling Water Reactor (BWR) Full-size-fine-mesh Bundle Test (BFBT) is to examine the derivation of uncertainty bounds on the void fraction of a BWR fuel bundle. A Code (with the capability of) Internal Assessment of Uncertainty (CIAU) was originally proposed by the Univ. of Pisa as a means of deriving uncertainty bounds in Nuclear Power Plant (NPP) calculations based on the difference between simulated and experimental data. This paper demonstrates how the core of the CIAU methodology May be adapted to generate uncertainty bands around subchannel void fraction predictions made by RELAP5-3D. In three of the four cases tested, the experimental subchannel void fraction fell within the predicted uncertainty bounds over 98% of the time. The fourth case consisted of low pressure, power and flow boundary conditions which led to poor subchannel void accuracy. An accuracy analysis indicates that this particular case has a sample averaged bias error of 0.191, which is about one order of magnitude higher than the other three cases examined. A planar average taken over all of the subchannels shows that the differences between simulated and experimental voids are less than 0.05 (in absolute terms) in all four of the cases examined. (authors)