Model for resilient water distribution systems
FRS researchers have developed a model that supports tailored strategies to make water distribution systems more resilient, based on the location and criticality of the disruption.
The resilience of water distribution systems (WDSs) has gained increasing attention in recent years. Analysing the system’s performance loss when disruptions occur and its subsequent recovery is one way to assess the system’s resilience. However, there is currently no model for characterising the performance loss and recovery behaviour of WDSs, making it a challenge to translate observations to insights to assess and enhance resilience.
In a paper published in external page Journal of Water Resources Planning and Management, Dr Beatrice Cassottana, Dr Nazli Yonca Aydin and Prof. Loon Ching Tang present a recovery function to model the performance of post-disruption WDSs over time. The model enables a time-continuous quantification of the WDS response throughout the duration of disruption, including periods of loss and restoration of service. Specifically, the recovery function is meant to model the water delivery service of a WDS disrupted by water leakages, with parameters to characterize the absorptive, adaptive, and recovery capabilities of a WDS.
The model proposed in “Download Quantitative Assessment of System Response during Disruptions: An Application to Water Distribution Systems (PDF, 1.7 MB)” was applied to two benchmark networks, Net3 and C-Town, under different disruption scenarios and recovery strategies. Analysis showed that the model supports the development of tailored strategies based on the location of disruption and the criticality of the disrupted node, and therefore, allows resources to be allocated more efficiently.
Compared to other metrics that quantified WDS resilience based on performance observed at specific points of time, recovery functions allow comparisons of system responses under different disruption and recovery scenarios. This makes it possible to identify areas for improvement to enhance resilience. In addition, the model is effective in identifying disruptions scenarios with unusual recovery behaviors, such as slow performance loss followed by a sudden recovery, which would have not been detected using existing hybrid metrics.
The model therefore supports the design of resilient WDSs by addressing system criticality with effective strategies. Moreover, resources can be efficiently allocated by developing tailored resilience strategies according to the location of disruption. More importantly, the methodology presented in the paper is general and can be applied to different water networks.
Further analysis could also support the development of resilience strategies to enhance the absorptive, adaptive, and recovery capabilities of systems based on topology and unique attributes. These strategies might include buffer capacity to help a WDS to absorb and adapt to disruptions, or emergency routines such as disconnecting failed components through isolation valves to prevent cascading failures during disruptions.
Dr Beatrice Cassottana and Prof. Loon Ching Tang are in the Cyber-Physical Systems Resilience module of the Future Resilient Systems (FRS). Dr Cassottana is also the principal investigator of the Disaster REsilience Assessment, Modelling, and INnovation Singapore project, of which FRS alumni, external page Asst Prof. Nazli Yonca Aydin from TU Delft, is a collaborator.