Modeling of Aerosol Fission Product Scrubbing in Experiments and in Integral Severe Accident Scenarios

Aerosol scrubbing experiments are recalculated for selected POSEIDON-II series tests and TEPCO-TOSHIBA-HITACHI tests with the MELCOR/SPARC code and with the BUSCA code. The major uncertainties in the calculations are identified and the results of the analyses are used in sensitivity simulations for long-term station blackout accident sequence in a boiling water reactor (BWR) with MELCOR/SPARC. The accident sequence is similar in nature to what happened in Fukushima Unit-3.

In the analyses of experiments, the basic characteristics of the thermal-hydraulic behavior were captured very well by MELCOR/SPARC. The trends in the calculated values of decontamination factors (DFs) for aerosols were in good agreement with the data, predicting the dependence of decontamination on the effect of the submergence depth, of steam content in the gas, and of aerosol particle size. However, the absolute values of DFs as calculated by the codes were rather sensitive to changes in the default input options and the agreement with experiments was not convincing.

In the integral BWR sequence simulations, the predicted DFs for aerosols in the wetwell (WW) was sensitive in the same way as for the experiments. For this type of a scenario, the fission product (FP) releases to containment—represented by Cs and I compounds—were mostly discrete events of short duration. The release path was from the reactor vessel to WW during the operation of the safety relief valves and subsequent venting of the containment from the WW gas space. Even though the passage through the WW water in these simulations was the only way for fission products to reach the environment, the sole scrubbing potential of the wetwell was not the determining contributor to the FP retention. At least of the same importance were the details of the sequence progression, timing of events, FP speciation, and other factors.