Assessment of sensitivity methodologies for a pressurized and depressurized loss of forced cooling in the high temperature gas-cooled test reactor

The goal of this study is to perform a preliminary assessment of loss of forced cooling events in the High Temperature Gas-cooled Test Reactor (HTGR-TR). This assessment is a demonstration of a sensitivity analysis framework that can be readily built upon. The framework employed two different sensitivity methodologies to highlight the necessity and benefits of including multiple metrics by comparing the predicted coefficients. An analysis will be performed for both a pressurized loss of forced cooling (PLOFC) and a depressurized loss of forced cooling (DLOFC). In both cases, the study will include the subsequent conduction cooldown in the reactor core. In particular, we will focus on a small, non-exhaustive, group of thermal properties within the HTGR-TR. The sensitivity study was used to exemplify how the framework can help determine which thermal properties have the most impact on the transient peak fuel temperature and 8-day (“late stage”) coolant outlet temperature. This study will include the calculation of both non-variance-based sensitivity coefficients and moment-independent Delta indices through the use of Monte Carlo sampling to compare the findings of both methodologies. An envelope of predicted fuel temperatures was generated through the sampling of the parameters during the sensitivity study. A peak fuel temperature of approximately 1500 K was predicted for both the PLOFC and DLOFC. The peak temperature was dominated by steady-state conditions, resulting in the same peak temperature for both transients. After the onset of both transients, a subsequent peak temperature was observed to be 1300 K for the PLOFC and 1400 K for the DLOFC. The fuel temperature did not exceed the steady-state temperatures during either transient. In both the PLOFC and DLOFC, the coolant inlet temperature and thermal conductivity were identified as most important for the transient fuel temperature. The Delta indices also identified these parameters as important with respect to the transient fuel temperature distribution. For example, the thermal conductivity had a Delta index of 0.28 and 0.5 for the PLOFC and DLOFC, respectively. In addition, an analysis of the 8-day coolant outlet temperature indicated a shift in parameter importance between the PLOFC and DLOFC with respect to nominal coolant pressure and moderator thermal conductivity. For example, a vastly differing Delta index of 0.1 and 0.7 was computed for the thermal conductivity during PLOFC and DLOFC, respectively. Discrepancies between the sensitivity studies highlight the potential strengths and weaknesses of different sensitivity approaches, emphasizing the value of including different methodologies during sensitivity analyses.