A Mixed-Integer Programming Approach to Optimal Operations of Drinking Water Treatment Plant with Verification on an Industrial Site

A constant treatment rate is expected in a water treatment plant for a cost-effective water supply. But this is not always possible due to demand variations. Therefore, the operator decides the treatment stages based on the plant capacity and switches from one stage to another when demand changes. This study proposes a mixed-integer programming approach to minimize such switches while ensuring water quality and satisfying drinking water tank capacity constraints. Considering daily operations, time-dependent binary variables are used to describe the hourly treatment quantity, resulting in a mixed-integer linear programming (MILP) problem. The proposed approach is verified in an industrial treatment plant. Compared with conventional operations, the number of switches can be reduced by about 30% through the optimization.