A modular gas-cooled reactor design with a thermal output of 600 MWt and a core exit temperature of 950 °C has been designed by the Korea Atomic Energy Research Institute based on the GT-MHR reactor concept which adopts a prismatic core. A sensitivity study on the transient plant behavior during a postulated depressurized LOFC accident concurrent with the failure of the RCCS was performed. In the transient analysis, the GAMMA+ code which can handle multi-dimensional, multicomponent problems was used. The RCCS is a passive system which is very reliable and supplies a significant heat removal mechanism during abnormal conditions in a GCR. To investigate the safety characteristics of a GCR under the one of the worst accidental scenarios, a simultaneous failure of the RCCS with a depressurized LOFC was assumed. The thermal behavior of the reactor system was analyzed in various conditions. It is found that the maximum temperature of the reactor fuel compact could exceed 1600 °C at about 50 h at the condition of a depressurized LOFC with a failure of the RCCS. A problem with the structural integrity of the reactor pressure vessel could also be a critical factor. The insulation of a reactor cavity wall serves as a dominant obstacle against a heat transfer from the reactor vessel to the surrounding ground when the RCCS fails to operate. Without insulation material on the reactor cavity wall, the gradients of the increasing rate of the maximum temperature diminish and the peak values decrease. The maximum temperatures of the fuel compact and the reactor vessel are less sensitive to the concrete and surrounding soil properties, those are the thermal conductivity and volumetric heat capacity, when the insulation material is used. The uncertainties in the properties of the concrete and the surrounding soil become significant without an insulation material in the cavity. To improve the safety of a modular GCR, more effective and feasible heat removal mechanism need to be devised based on the comprehensions on the heat transfer characteristics.
Read full abstract