RELAP5/Mod3.3 and MARS3.0a Modeling of a Siphon Break Experiment

Pool water plays a very important role as a final heat sink for most pool-type research reactors following postulated events. Therefore, one of design criteria for the reactors is that the water level of reactor pool must not decrease below a predefined elevation even against the most severe accident due to ruptures of coolant boundary of connecting systems to the reactor pool. In order to accomplish the design criterion, all the connecting systems are usually arranged to be above the elevation of reactor core. However, some research reactors with a downward flow in the reactor core have a primary cooling system located below the elevation of reactor core because of meeting an available net positive suction head of pumps in the system. These reactors have a provision consisting of pipes penetrating a reactor pool wall at a higher elevation than that of reactor core and siphon break devices to meet the design criterion. A series of experiments was carried out to figure out thermal hydraulic characteristics during siphon is blocked and establish design requirements for siphon breaker. The experimental study provided a lot of data and observations to the process of siphon break, but it does not provide a sufficient theoretical analysis and present practical design requirements applicable to industry. The experimental range is not also sufficient to cover operating conditions of siphon breakers for research reactors. A series of numerical simulations on the experimental data has been tried by using thermal hydraulic system analysis codes, RELAP5/Mod3.3 and MARS3.0a. This paper includes a part of the numerical simulations. First output from this study shows an importance of an adequate use of thermal hydraulic models in the codes and a big different prediction between the two codes especially in relation to the use of choked flow option. From this study, it seems that RELAP5/Mod3.3 has some problems on the control of a choked flow option-flag or the prediction of a choked flow criterion.