Abstract As one of the “next generation” nuclear power systems, the Gas Turbine Modular Helium Reactor (GT-MHR) achieves improved safety by relying upon inherent thermal storage characteristics and passive cooling features to maintain fuel and reactor vessel integrity if all its active cooling systems fail to operate in the unlikely event referred to as conduction cooldown. A helium-cooled reactor core, an uninsulated reactor vessel, and a set of surrounding natural convection air cooling panels facilitate removal of decay heat from the core. Due to high temperatures and large thermal gradients, the prevailing heat transfer mechanisms during conduction cooldown include natural convection of pressurized helium combined with thermal radiation inside the vessel and natural convection of air combined with thermal radiation outside the vessel. Because of the unique geometry and thermal conditions of the GT-MHR and limited available test data, the applicability of empirical-based heat transfer correlations is limited. In designing the passive natural draft cooling system for the GT-MHR, computational fluid dynamic (CFD) models have been developed to model the passive heat transfer processes and predict the resultant reactor fuel and component temperatures. To initiate validation of the analytical methods used for this critical safety application, a benchmark heat transfer experiment performed at the Japanese Atomic Energy Research Institute (JAERI), which partially models the in-vessel and ex-vessel decay heat removal heat transfer of the GT-MHR, was simulated using the finite element-based FIDAP CFD analysis package. The steady state model incorporated conjugate heat transfer, diffuse grey-body surface radiation, and a Boussinesq model for buoyancy driven convection. Since Rayleigh numbers for the experiment ranged from 106 to 107 inside the vessel and from 108 to 109 outside the vessel, both laminar and turbulent flow models were evaluated. The analysis results show good agreement with the experiment test data. Although radiation accounted for more than two-thirds of the heat rejected from the vessel, the strength of the natural convection flow caused the highest temperatures both inside and outside the vessel to occur around the upper vessel head, far removed from the area of highest heat generation. The laminar flow model shows better agreement with the test data than the turbulent model, which seemed to over-predict the vessel thermal boundary-layer thickness and under-predict the axial temperature gradient resulting from natural convection. This investigation provided an excellent starting point towards the validation of the CFD models used to predict Decay Heat removal conditions in the GT-MHR. Further experiments performed at higher temperature conditions are needed to determine the extent of model complexity required to accurately simulate the passive decay heat removal mechanisms in the GT-MHR.
Jonathan Berkoe (Sun,) studied this question.