W. Kanning

Cost-optimal design of roiver dikes using probalilistic methods

Given the large investments in dike construction and reinforcements, optimisation of these interventions could contribute to significant cost savings. In this study, a probabilistic method is developed for obtaining the geometry of a river dike cross-section that fulfils a specified target reliability at a minimal cost. The method takes into account multiple relevant failure mechanisms and various geometric parameters of the dike. Large numbers of cross-sections are generated and the reliability for each section is computed using FORM, Monte Carlo sampling, and numerical integration. The method is applied to several cases representing typical river dike profiles that are found in the Netherlands. Results show that for the optimal geometry the largest fraction of the target failure probability is assigned to those mechanisms with a relatively high cost of reinforcement. Also, comparison with the current semi-probabilistic design guidelines shows that a decrease in costs of up to 15% could be reached with the new optimisation method.