2019 CSCE Annual Conference - Laval (Greater Montreal)

2019 CSCE Annual Conference - Laval (Greater Montreal) Conference

Multi-attribute Metric for Assessing Resilience of Water Distribution Networks

Mr. Ahmed Assad, Concordia University (Presenter)
Dr. Osama Moselhi, Concordia University
Dr. Tarek Zayed, The Hong Kong Polytechnic University

Water distribution networks (WDNs) are critical infrastructure systems responsible for securing adequate quantities of safe, high-quality water to the public. Ensuring a proper function of water systems has always been a major concern for utilities and municipalities because of their direct impact on public health and safety. Resilience assessment of these networks is emerging as an important requirement in planning and management of WDNs. In this context, it is desirable for water networks to be strong enough to withstand disruptions with least impact on their performance and to enable fast recovery in case of service loss.Several models have been developed to consider resilience in design of WDNs, but much less in their operation and maintenance. Those that targeted operation and maintenance were limited to one source of hazards like earthquakes.

The ultimate objective of a current research is to develop a holistic resilience-based management method for water distribution networks. This paper presents a newly proposed metric to assess resilience of WDNs considering multi-hazard events.

A detailed framework and algorithm are developed to estimate loss in resilience arising from a given source of hazard.  The metric is based on two components robustness and redundancy. Robustness of WDNs is modeled by integrating reliability and criticality of its water mains. Graph theory is employed to quantify the connectivity and redundancy in the network. The metric is then formulated as a weighted sum of the two components. Several codes were developed to capture a scenario-based assessment of various hazard events. Data from City of London, ON was fetched to implement the developed model. The results will identify the critical components of the network which are responsible for a total service loss. This type of output can be of help to decision makers in setting priorities for maintenance of their WDNs.