To investigate the micro-natural circulation phenomenon within the pressure vessel of a small nuclear reactor during cold shutdown maintenance, a comprehensive reactor pressure vessel model was developed utilizing the RELAP5 3D program. A sensitivity analysis was performed on the node division of critical control components, including the reactor core, descent section, chamber, and bypass flow channel, to assess the efficacy of the improved modeling scheme. The thermohydraulic characteristics of micro-natural circulation at varying power levels were examined. Furthermore, the Collier formula was employed to refine the low-flow heat-exchange model. The results indicate that the node division in core areas, such as the reactor core and the descending section, is highly sensitive to simulation outcomes. The optimized model effectively addresses the issues of overestimating circulation flow and underestimating the temperature difference in heat exchange present in the original model. Within the pressure vessel, three parallel flow paths exist: the main circulation, local short circulation, and bypass circulation, with the main circulation flow rate comprising nearly 69%. Power significantly influences micro-natural circulation. At a power level of 25 kW, the steady-state flow rate of natural circulation reaches 0.95 kg/s, while the steady-state temperature difference between the inlet and outlet of the reactor core is 3.7 K. Conversely, at a power level of 5 kW, the microcirculation flow rate is approximately 0.53 kg/s, and the temperature difference between the inlet and outlet of the reactor core is about 1.3 K. Compared to the original heat-exchange model of the RELAP5 program, the Collier formula more accurately characterizes the heat-exchange behavior under low-flow micro-natural circulation conditions. These research findings provide a theoretical basis and modeling reference for the thermal simulation of cold shutdown conditions in pressurized water reactors, as well as for nuclear safety assessments during maintenance.
Fulong et al. (Mon,) studied this question.