PNNL

Abstract

The FRAPCON-3.3[1] fission gas release (FGR) model has been modified to predict fission gas release from mixed oxide (MOX) fuel. This is accomplished by increasing the diffusion coefficient used for UO2 fuel by a small factor to better predict measured FGR data from full length commercially irradiated PWR fuel rods and from Halden instrumented assemblies with MOX fuel. The MOX used in the commercial and Halden test rods was fabricated using the MIcronized MASter blend (MIMAS) process developed by Belgonucleaire and used for MOX fabrication in French reactors. The FGR data from Halden tests consisted of refabricated rod segments of these commercially irradiated fuel rods that were further irradiated at steady-state powers and FGR data from power ramped tests IFAs 629.1, 629.3 and 606. The code has also been compared to IFA-633 that has been irradiated from the start in Halden at steady-state low power operation and increased rod powers in a slow stepwise manner to measure the onset of FGR. Recently the code has been compared to MOX FGR data from the Advanced Test Reactor in the US. The ATR tests involved short pin segments with MOX irradiated under steady-state power operation up to three different burnup levels. The ATR tested MOX fuel was fabricated using weapons grade plutonium and a fabrication process similar to MIMAS. These comparisons to FGR data demonstrate that the FRAPCON-3.3 code provides a good prediction of FGR for commercial PWR and Halden MOX irradiations at low and high burnup. The code overpredicts FGR in the low burnup ATR tests at low burnup but provides a good prediction of the ATR tests at high burnup. These results also demonstrate that the weapons grade MOX results in similar or lower FGR to that from commercially reprocessed MOX fuel using the MIMAS process, for a given temperature-burnup-time history.