Coupled decay heat and thermal hydraulic capability for loss-of-coolant accident simulations

The ANSI/ANS-5.1 standard “Decay Heat Power in Light Water Reactors” is a well-known way to predict the decay heat as a function of time for lightwater reactors. The standard calculates the decay heat as a function of the reactor power history and time after shutdown and can be applied for reactor safety analysis to ensure core coolability. Typically, the standard is applied for the entire core, using the history of the total power production to calculate the total decay heat in the core. Industry has some interest in modifying pressurized water reactor (PWR) cycles for operation at higher burnups by extending their cycle lengths. This would improve the economics of the current reactor fleet, but it is unknown if the high-burnup fuel rods would maintain their integrity, especially during accident scenarios. It is also unknown whether the ANS standard applied to the whole core is sufficient for these high-burnup rods or whether a more detailed approach should be pursued. This paper presents advancements in the Virtual Environment for Reactor Applications (VERA) to be able to calculate region-by-region decay heat everywhere in the reactor core. This is done by calling ORIGEN decay heat interfaces, which are well validated and used extensively in reactor and fuel cycle analysis. This capability was implemented for steady-state and transient. Several calculations were performed. First, decay heat curves were calculated during a reactor SCRAM of Watts Bar Unit 1 cycle 3 at the end of cycle using both the core-averaged ANS standard and the region by region ORIGEN calculations. This represents a conventional PWR core design, and the results show that the ANS standard is on the conservative side compared with ORIGEN, meaning that ORIGEN predicted lower overall heat during and immediately following the SCRAM. Next, decay heat curves were calculated using both methods for a hypothetical equilibrium high-burnup PWR core. In this case, ORIGEN predicted more decay heat during the SCRAM and beginning around a minute after the SCRAM completed. This shows that for high-burnup cores, the ANS standard may not be conservative. Finally, loss of coolant accident (LOCA) simulations were conducted with the TRAC/RELAP Advanced Computational Engine (TRACE) for the high burnup core using four decay heat treatments: (1) fixed-shape ANS treatment (TRACE model), (2) fixed-shape VERA-ORIGEN treatment (TRACE model with core power scaling factor taken from VERA-ORIGEN calculation), (3) spatially dependent VERA-ORIGEN, and (4) VERA-ANS (VERA time-dependent power with core-averaged ANS decay heat). The results of these TRACE simulations showed that there was little difference between the ANS and ORIGEN power calculations when using a fixed shape, with the limiting rod being one which was high power at steady-state. However, when using the region by region ORIGEN calculations the results were quite different; there was a much larger spread in the peak cladding temperatures and a high-burnup rod became limiting instead of a high-power rod. This indicates that for high-burnup cores, it may be necessary to resolve the spatial dependence of the decay heat to accurately predict the conditions of high-burnup rods during LOCAs. This is an important result that should be taken into account when developing high-burnup core designs.