Resumen
A detention reservoir is one of the most effective engineered solutions for flood damage mitigation in urban areas. Detention facilities are constructed to temporarily store storm water and then slowly drain when the peak period has passed. This delayed drainage may coincide with upstream floods and aggravate the flood risk downstream. Optimal operation and design are needed to improve the performance of detention reservoirs for flood reduction. This study couples hydrologic simulation software (EPA-SWMM) with an evolutional optimizer (extraordinary particle swarm optimization, EPSO) to minimize flood damage downstream while considering the inundation risk at the detention reservoir. The optimum design and operation are applied to an urban case study in Seoul, Korea, for historical severe flooding events and designed rainfall scenarios. The optimal facilities outperform the present facilities in terms of flood damage reduction both downstream and in the detention reservoir area. Specifically, the peak water level at the detention pond under optimal conditions is significantly smaller than that of the current conditions. The comparison of the total flooded volume in the whole watershed shows a dramatic reduction of 79% in a severe flooding event in 2010 and around 20% in 2011 and in 180 min designed rainfall scenarios.