CONTAIN Code Calculations for the LA-4 Experiment

Abstract

The thermal hydraulic and aerosol post-test predictions which have recently been completed for the LA-4 experiment using the latest version of the CONTAIN code are discussed. Modeling aerosols in nuclear reactor accident scenarios poses formidable computational challenges. The governing equations for aerosol particle coagulation, deposition, and condensation are difficult to solve over the many orders of magnitude of concentration expected in such scenarios. The modeling of particle growth by vapor condensation is particularly challenging because time scales of interest are much shorter than the time scales of interest in containment analysis. Furthermore, aerosol condensational growth is intimately coupled to the thermal hydraulics, and the two processes should not be solved independently. These difficulties have motivated the development of a moving-grid numerical technique for the containment severe accident code, CONTAIN. This technique is used to incorporate hygroscopic effects and the Kelvin effect in the CONTAIN aerosol condensation model. In this work applications of the moving-grid technique are reported. In particular, the technique is used to analyze the LACE LA-4 experiment, which simulated processes that could occur in light water reactor containments under severe accident conditions. In this experiment, measurements were made of the aerosol concentrations of hygroscopic CsOH and nonhygroscopic MnO in a steam-filled vessel. Rapid aerosol removal was observed during vessel depressurization and was attributed to condensation on aerosols enhanced by the hygroscopic nature of CsOH. The CONTAIN post-test blind calculations of the LA-4 thermal hydraulics predicted slightly higher pressures and gas temperatures than were reported. A review of CONTAIN input for completeness and uncertainty ranges for modeling parameters has culminated in a revised input deck for the thermal hydraulic portion of the experiment. Using this input, a best-estimate thermal hydraulic calculation is made and the results are in good agreement with data. The original CONTAIN post-test blind calculations without the hygroscopic effect did not reproduce the reported aerosol removal. Including the hygroscopic effect in the aerosol model increases the deposition rate significantly, and brings the rate in good agreement with the experiment as shown in a best-estimate aerosol calculation. The sensitivity of aerosol deposition to thermal hydraulic conditions is discussed as well as the effects of solubility on the thermal hydraulics.