A MINLP-based approach for the design-for-control of resilient water supply systems

Abstract

The improvement in resilience of water supply systems by increasing their redundancy, either in energy or in connectivity, is a common priority when doing rehabilitation and expansion. This however can come at the cost of other aspects of network performance, such as leakage management. In this work, we consider the designfor-control problem of adding new connections (from a predefined set of candidate pipes) to water supply systems to improve their resilience to failure events while minimizing the impact on leakage management under normal operating conditions. We present a mixed-integer non-linear programming formulation of the problem of optimal link addition for the minimization of average zone pressure, a surrogate measure of pressure dependent leakage. We implement a method based on spatial branch-and-bound to solve the problem on a case study network from the literature and an operational network part of an urban water system in the UK. Finally, we validate the improvement in network resilience resulting from the addition of new connections by performing an a posteriori critical link analysis, using the hydraulic resilience measure of reserve capacity.