Water distribution networks as flexible loads: A chance-constrained programming approach

Abstract

There is a greater need for flexibility in the power distribution network (PDN) due to increasing levels of renewable energy resources. Here, we consider using the water distribution network (WDN) as a flexible load. We formulate a chance-constrained multiperiod optimization problem to schedule water distribution pumps subject to WDN and PDN constraints while managing power demand forecast uncertainty. To do that, we develop a control policy that adjusts the WDN’s operation when a PDN constraint violation is present. Since the resulting problem is nonconvex, we utilize approximation and relaxation techniques to transform the problem into a convex program and solve via the scenario approach. Through detailed case studies, we verify the performance of the control policy to ensure network constraints are satisfied despite uncertainty. We find that we can successfully schedule and control the WDN to provide flexibility to the PDN for many realistic water and power demand scenarios.