4:05 PM - 4:20 PM
[2G14] Development of a gas-liquid two-phase flow experiment database and a simulator-based machine learning method
Keywords:Gas-liquid two-phase flow, Wire-mesh sensor, Pool scrubbing, Machine learning, Real-time simulator
During a severe accident (SA) at a nuclear power plant, pool scrubbing is expected to be performed, whereby gases containing fission products (FP) are passed through water. This will remove FP from the gas to the water, thus reducing the amount of FP released into the environment. This paper reports on the development of a detailed experimental database on gas-liquid two-phase flow and a simulation method which reflects the experimental data using machine learning, which are important for evaluating the effectiveness of FP removal by pool scrubbing.