Risk based optimal design of detention dams considering uncertain inflows

This study deals with the optimal design of detention dams under flood discharge uncertainties using a simulation-based optimization approach. An extended methodology is represented by integrating the ant colony optimization (ACO), artificial neural networks, and various risk measures including: expected flood discharge, value at risk, and conditional value at risk (CVaR). For this purpose, first, the neural network is trained by the results of a hydrodynamic model and then it is used to measure different risk indices under flood uncertainties. The proposed approach is then applied to a real case and optimal designs are determined by the search algorithm-i.e. ACO. Different optimal designs are obtained for the storage detention dams when different risk concepts are implemented as the objective function in the system modeling. Particularly, when the expected value measure is combined with the CVaR, the cost of optimal design is nearly two times smaller than those obtained by the formulation with independent objective functions whereas the obtained solution could efficiently minimize both E(Qd) and CVaR. The optimal solutions have especial capabilities in terms of performance and cost levels and this gives the stakeholders and decision makers to construct a framework to choose the final design when there are different attitudes and interests for flood risk management.