Large-scale network systems involve a large number of variables, making the design of real-time controllers challenging. A distributed controller design allows to reduce computational requirements since tasks may be divided into different subsystems, making possible to guarantee real-time processing. This paper proposes a constrained optimization based distributed control design by applying a novel population and masses dynamics approach. The distributed controller design is applied to the Barcelona Drinking Water Network (DWN) in order to illustrate its effectiveness in the control performance.