Sectorization of Water Distribution Systems for Optimal Observability

Resilient water distribution systems have the ability to anticipate, monitor, react to and learn from critical events. While numerous resilience metrics have been proposed for the reacting and anticipating functions, monitoring metrics are understudied. This work proposes an a-priori monitoring metric based on structural observability theory, the minimum number of sensors, and demonstrates how it can be integrated into problem formulations for optimal sectorization. The dividing stage of WDS sectorization is modelled as a MINLP, while the minimum sensor placement problem is solved using zero forcing formulated as an IP. To provide an example, the methodology is applied to the benchmark network Apulia.