Full-scale dryout/rewet instability tests in a BWR rod bundle and TRACE assessment

•BWR Instability rests were conducted in a full-scale BWR rod bundle.•Dryout/rewet cycles were observed until failure to rewet.•The effect of the minimum film boiling temperature in the failure to rewet is investigated.•2D inverse heat transfer analysis were conducted.•TRACE simulations are compared with the experimental data. The regulatory purpose of the current work is to study fuel damage mechanisms and processes that were predicted to occur under certain unstable conditions that could arise during postulated anticipated transients without SCRAM (ATWS) for boiling water reactors (BWRs) operating in the maximum extended load line limit analysis plus (MELLLA+) expanded operating domain. The Nuclear Regulatory Commission (NRC) conducted experiments at the high pressure Karlstein Thermal-Hydraulic Experimental Facility (KATHY) using a full-scale simulated BWR fuel assembly. The simulated assembly includes representative features of modern BWR fuel bundles, including commercial spacers, part length fuel rods and water rods. KATHY includes two features that allowed the NRC to conduct experiments that subject the simulated fuel assembly to unstable conditions similar to those expected during postulated ATWS events. These features are the downcomer, which allows the loop to be operated under natural circulation flow, and the proprietary SINAN control program, which allows the heater rod voltage to vary in response to changing thermal-hydraulic conditions, which in turn, simulates reactivity feedback. The NRC conducted several tests which produced growing flow and power oscillations and these oscillations were allowed to proceed up to the point where the heater rod surface failed to rewet and went into the early stages of heat-up. The experimental data were analyzed to determine the temperature threshold for failure to rewet and the experimental results have been compared to various models of the minimum stable film boiling temperature (Tmin). The results show that the homogeneous nucleation temperature provides a reasonable estimate of this temperature threshold. Further, the NRC's TRAC-RELAP Advanced Computational Engine (TRACE) code was used to simulate the experiments to assess the TRACE performance for simulating failure to rewet during postulated ATWS-I events. When compared to the experimental data, TRACE produces slightly conservative estimates of the cladding temperature when an appropriately low Tmin value is used (i.e., one consistent with homogeneou