Optimal operations and scheduling of water pump stations considering energy consumption

The operations of pump stations in water distribution system consist of scheduling each pump's status in different time periods, regulating water storage levels, meeting demands with satisfactory water quality at required flow rate and pressure, and also minimizing energy consumption. In this paper, we divide each day into several peak/off-peak intervals depending on the water demands, to incorporate various levels of energy costs. For pump scheduling, the on/off status will be decided with a limitation on maximum number of daily switches. We also consider the storage levels in different tanks, flow rates and amounts on pipelines with corresponding capacities, water pressure requirements at demand and junction nodes. The objective is to minimize the total energy consumption cost. To solve the optimization problems with both integer and continuous decision variables, we present reformulation techniques to linearize the nonlinear terms and iteratively apply the EPANet as a gray-box to decide the operations of water system. Additionally, we select a random group of solutions as initial ones applied in genetic algorithm to find near-optimum solutions. The outcomes will be tested on a water distribution system, and results for operations of pumps can be presented to the manager of a water system to facilitate daily optimal decisions.