Characterisation of resilience metrics in full-scale applications to interdependent infrastructure systems

Critical infrastructure (CI) plays a crucial role in our daily life and CI resilience is nowadays a priority for policymakers and a strategic objective for operators. This trend prompted a considerable research effort on measuring resilience, which proposed several different metrics aiming to capture resilience properties of a CI system, both generic and problem-specific. Selecting the appropriate metric for a given application, based on properties, major strengths, or weaknesses of existing metrics, is not trivial. To this end, this paper first performs a systematic literature review of resilience metrics suitable for application to networked infrastructure systems, where a representative subset of different types of metrics is selected. We advance the state-of-the-art by defining a few new metrics to partly cover the gaps in the extant literature. The final set of resilience metrics is rigorously compared by looking at their behaviour when the properties of the system vary. In particular, we evaluate the ability of each metric to assess the influence of the system's structural and operational conditions on its resilience profile. The proposed and implemented evaluation procedure uses the DMCI tool (Dynamic Functional Modelling of Vulnerability and Interdependencies of CIs) to simulate the dynamics of an interdependent networked infrastructure system. A large portion of the transportation network connecting Italy and Switzerland is used as the empirical setting.