Risk assessment for hydraulic design associated with the uncertainty of rainfall

The purpose of this paper is to propose a method that incorporates the uncertainties of the depth and the duration of rainfall into the risk assessment for hydraulic design. In this paper, the risk of a hydraulic system is defined as the probability of failure of the hydrological system. The Hasofer-Lind reliability index (HLRI), which is a popular index for risk assessment, is used to improve the computation efficiency. The evaluation of the HLRI can be transformed into a constrained optimization problem. To solve the constrained optimization problem, modified simple genetic algorithms (SGA), which combine the penalty function and the constraint handling technique proposed by Deb (2000), is developed. The proposed method can produce the relation of the probability of failure versus the central safety factor. Based on the relation, a comprehensive benefit-cost analysis can be performed. The optimal alternative can be selected according to the result of the benefit-cost analysis. The proposed method provides an aid for performing the benefit-cost analysis.