Hybrid SWMM and particle swarm optimization model for urban runoff water quality control by using green infrastructures (LID-BMPs)

In the past decades, climate change, population growth, urbanization, changes in land-use, and outdated runoff collecting networks have affected the quantity and quality of urban runoff in many countries. In this research, a new methodology for planning green infrastructures (GI) for runoff water quality improvement in urban areas was proposed. This framework aims in optimized designing of the type and area of low impact development-best management practices (LID-BMPs) for urban areas. Three management practices, including infiltration trenches, bio-retention basins, and permeable pavements, were considered together with an urban drainage network. The stormwater management model (SWMM) was used for the rainfall-runoff simulation, and the multi-objective particle swarm optimization (MOPSO) algorithm was utilized for LID-BMP optimization. The proposed SWMM-MOPSO model was applied to an urban area in Northwestern Tehran, Iran, and an optimized combination of the BMPs was determined. To evaluate the performance of the optimized BMPs, its results were compared with conditions where a single type of BMPs (i.e., infiltration trenches, bio-retention basins, and permeable pavements) was implemented to each of the sub-basins. For this purpose, a single type of BMPs was allocated by different percentages of each sub-basins area. Results showed that by the application of these single BMP types, the whole basin's concentrations of total suspended solids (TSS), total phosphorous (TP), and total nitrogen (TN) were reduced by 97%, 68%, and 72%, respectively. It was seen that the bio-retention basin was the most effective single BMP in water quality improvement. In the case of determining an optimum combination of the three BMPs, the SWMM-MOPSO model was applied to minimize the TSS concentration. Results showed that comparing the single bio-retention basins, the optimum combination of BMPs reduced the TSS concentrations by 10-12%. The proposed hybrid SWMM-MOPSO simulation-optimization model was instrumental in the optimal designing of the LID-BMPs and controlling runoff water quality.