A synthetic water distribution network model for urban resilience
Nasir Ahmad, Mikhail Chester, Emily Bondank, Mazdak Arabi, Nathan G. Johnson, Benjamin L. Ruddell
Abstract
Water distribution networks (WDN) are one of the most critical infrastructures, providing water for essential needs. However, the dearth of information on WDNs due to weak historical records, limited willingness to share data, and security concerns limit a researcher’s understanding of the criticality, adaptability, vulnerability, and interdependencies of WDNs. To address this challenge, we develop a model entitled SyNF (Synthetic Infrastructure) for synthetic WDN generation. SyNF uses a roadway network, water demand, and water source locations to synthesize topology, diameter, and service year of pipes, and location and power requirements of pumps. To show SyNF’s capabilities, we start with the City of Tempe and scale the model to Phoenix metro’s seven major cities. We find a 6% average dissimilarity on pipe size distribution between the original and synthesized WDN in validating SyNF. We also discuss how SyNF can advance our understanding of the criticality, vulnerability, and resilience of WDNs.