Void fraction distribution in BWR fuel assembly and evaluation of subchannel code

Void fraction measurement tests for BWR fuel assemblies have been conducted as part of a Japanese national project. The aim was to verify the current BWR void fraction prediction method. Void fraction was measured using an X-ray CT scanner. This paper describes typical results of void fraction distribution measurements and compares subchannel-averaged void fraction data with current subchannel analysis codes. The subchannel analysis codes COBRA/BWR and THERMIT-2 were used in this comparison. The agreement between data for an actual BWR fuel assembly with two unheated rods was good, but in the case of many unheated rods, the codes were unable to predict well large void fraction gradient in the radial direction observed in the measured data over the unheated rod region. The prediction errors of COBRA/BWR and THERMIT-2 for the subchannel-averaged void fraction were 〈A〉 (average of difference between measurement and calculation) = -1.1%, σ (standard deviation) = 5.3% and 〈A〉 = -2.2%, σ = 6.3%, respectively.