Mechanism study of siphon break using RELAP5/MOD3

The phenomenon of siphon break is a passive physical process and it is important for the safety design of pool type Liquid Metal Fast Breeder Reactor (LMFBR) and spent fuel pool of the nuclear power plant adopted Light Water Reactors (LWRs). The phenomenon can affectively mitigate the Loss of Coolant Accidents of LMFBR and spent fuel pool of LWRs. In the present study, the system thermal hydraulic code RELAP5/MOD3 is used to simulate the experimental results of siphon break apparatus of Beijing Tsinghua University. The experimental system had an upper tank, a main pipe including a riser and a downcomer, a siphon break pipe and a lower tank. This device is roughly 3.5 m high. Initially, the upper tank, the main pipe and lower tank are filled with water. The water level in the upper tank is high enough to cover the inlet of the siphon break pipe. At the beginning of the experiment, a valve on the main pipe is switched open. The water in the upper tank drains to lower tank through the riser and the downcomer of the main pipe due to the effect of siphon. When the water level of the upper tank drops below the air inlet of siphon break, the air enters the main pipe through the siphon break pipe. The flow in the downcomer transits from single phase to two phase flow. The water velocity decreased to zero and the loss of water from upper tank is terminated. In the present study, a RELAP5/MOD3 input deck of the test facility is built to simulate the siphon break phenomenon. The simulated results of flowrate in the main pipe, pressure drop in downcomer and water level in the upper tank over time are compared with the experimental results. The results have demonstrated that the general trends of the experimental results can be caught by RELAP5 code. Nevertheless, the time duration of siphon break phenomena as predicted by the code is significantly less than the experimental results. A detail analysis of the code simulation results has shown that the two phase multipliers as predicted by the code is about an order of magnitude smaller than the experimental results. It is believed that the difference is due to the assumption of fully developed flow in the code calculation. Sensitivity studies of the different downcomer pipe diameter and outlet valve opening show that the time duration and water loss are positively related to the downcomer diameters, but not related to the outlet openings. The results of RELAP5/MOD3 simulations are also compared with the results of CFD (Computational Flow Dynamics) simulation. (author)