Uncertainty Study in the Sectional Method-Based Aerosol Behavior Analysis Model for Severe Accidents in Nuclear Power Plants

Under severe accident conditions in nuclear power plants (NPPs), aerosols are one of the primary forms of fission product release, making the study of aerosol behavior crucial for assessing severe accident source terms. In severe accident analysis programs, the description of aerosol behavior is predominantly based on solving the General Dynamic Equation (GDE) for aerosols. Since the GDE is a partial differential equation, analytical solutions are often unavailable. Aerosol behavior is influenced by a combination of factors, including their physicochemical properties, severe accident phenomena, and engineered safety features, leading to significant uncertainties in the models describing their transport behavior. This study employs the DAKOTA program to conduct uncertainty and sensitivity analyses on the parameters related to the coagulation and deposition models in the Containment Aerosol Behavior Simulation and Analysis (CABSA) program. Using the Wilks non-parametric method based on Latin Hypercube Sampling (LHS), the uncertainty propagation of input parameters is quantified, and the impact of parameters on the model is evaluated through correlation coefficients. The results indicate that the median particle size of aerosols, the thermal diffusion boundary layer thickness, the thermal regulation coefficient, and the dynamic shape factor significantly influence aerosol behavior. Additionally, the Brownian diffusion boundary layer thickness, the lower limit of aerosol particle size, and particle density also exhibit moderate effects on aerosol behavior. This research provides valuable insights for the subsequent use and optimization of aerosol behavior analysis models, contributing to the improvement of accuracy and reliability in severe accident source term analysis for NPPs.