Evaluation of the robustness of critical infrastructures by Hierarchical Graph representation, clustering and Monte Carlo simulation

•We look at the robustness of networked CIs under a system-of-systems viewpoint.•We propose a Hierarchical Graph for CIs representation.•We adopt a multistate model to describe different degrees of degradation of the components.•We account for different priorities in the partitioning of the flow in the network.•We resort to clustering techniques for large-scale systems and we evaluate the robustness by Monte Carlo simulation.

In this paper, we present a methodological work that adopts a system-of-systems (SoS) viewpoint for the evaluation of the robustness of interdependent critical infrastructures (CIs). We propose a Hierarchical Graph representation, where the product flow is dispatched to the demand nodes in consideration of different priorities. We use a multi-state model to describe different degrees of degradation of the individual components, where the transitions between the different states of degradation occur stochastically. The quantitative evaluation of the CIs robustness is performed by Monte Carlo simulation. The methodological approach proposed is illustrated by way of two case studies: the first one concerns small-sized gas and electricity networks and a supervisory control and data acquisition (SCADA) system; the second one considers a moderately large power distribution network, adapted from the IEEE 123 node test feeders. The large size of the second case study requires hierarchical clustering for performing the robustness analysis.