Coupled AEM-PSO Model for the Optimum Location and Design of a Radial Collector Well in an Alluvial Riverbed

The radial collector wells provide large quantity of filtered water when constructed in the alluvial riverbeds getting low flow from upstream but their construction cost is very high. Simulation/optimization model AEMPSO-OD is developed to find the optimum location of the well within the river bed for a synthetic problem to minimize the capital cost by using analytical element method and particle swarm optimization. The AEM model generates line-sinks for well laterals by using values obtained from each position vector generated by PSO model and also creates other elements such as aquifer, resistance specified line-sink for river channel and in-homogeneity to represent no-flow boundary. The objective is to minimize the total cost subject to constraints for minimum discharge, residence time, maximum discharge intensity in radial drain and maximum induced recharge from river. The optimum is solution obtained by model in terms of location, length of each lateral and water level to be maintained in the well. The result shows that well discharge is highly increased by proximity of stream. As the well is located away from stream to increase residence time for better filtration, the induced recharge rate is also decreased which helps controlling sedimentation. The model is successful in finding optimum location and design. AEM is preferred as other mesh based method involves change in mesh with different locations populated by optimization model.