Probabilistic flood hazard assessment method considering local intense precipitation at NPP sites

摘要

In recent years, the flooding risk of major facilities has increased significantly due to heavy rainfall events. These facilities must evaluate both external and internal flooding risk such as from heavy rainfall, flash floods, watershed flooding, river flooding, and coastal flooding. This study first estimates flooding from extreme rainfall with local intense precipitation and analyzes the resulting impacts on buildings and roads at a specific nuclear power plant (NPP) site, with which a roughness coefficient according to the landuse condition is estimated. A two-dimensional external flooding hazard analysis is then carried out with tidal levels as the external boundary conditions, and based on the results, new hazard curves for the inundation depth with frequency and duration are developed for the NPP site to show the relationships among rainfall, flood depth, and annual exceedance probability. To match the proper probability distribution types to the flood hazard curves, the fit was analyzed through Akaike's information criterion verification. After analyzing the correlation between the flood depths, the mode values were calculated through Monte Carlo simulation with the verified probability distribution types. Finally, probabilistic flood hazard curves for the NPP site were obtained based on the calculated mode values. Representatively, when a 10(6)-year return period rainfall occurs at the nuclear power plant site, the mode flood depth was found to be 1.07 m for power plant 1 and 0.61 m for power plant 2. In this way, the approach of this study is expected to support the waterproof design of critical facilities, flood prevention function design, the advancement of flood prevention measures and procedures, and the evaluation of flood mitigation functions.