Resilience assessment for interdependent urban infrastructure systems using dynamic network flow models

Critical infrastructure systems are becoming increasingly interdependent, which can exacerbate the impacts of disruptive events through cascading failures, hindered asset repairs and network congestion. Current resilience assessment methods fall short of fully capturing such interdependency effects as they tend to model asset reliability and network flows separately and often rely on static flow assignment methods. In this paper, we develop an integrated, dynamic modelling and simulation framework that combines network and asset representations of infrastructure systems and models the optimal response to disruptions using a rolling planning horizon. The framework considers dependencies pertaining to failure propagation, system-of-systems architecture and resources required for operating and repairing assets. Stochastic asset failure is captured by a scenario tree generation algorithm whereas the redistribution of network flows and the optimal deployment of repair resources are modelled using a minimum cost flow approach. A case study on London's metro and electric power networks shows how the proposed methodology can be used to assess the resilience of city-scale infrastructure systems to a local flooding incident and estimate the value of the resilience loss triangle for different levels of hazard exposure and repair capabilities.