Comparison of Cascading Effects in Critical Infrastructure Networks based on Simulation

Abstract

Natural hazards such as floodings affect critical infrastructures (CIs) with increasing frequency and intensity. The effects are hard to predict due to the change of the hazards over time (e.g., due to climate change), but also due to interdependencies between CIs This uncertainty about the impact of a flooding can be captured in a probabilistic model that describes the effects in a network of CIs. To this end, the functionality or availability of a CI is described through a state, measured on a 5-tier scale. The state changes with a certain probability depending on the infrastructure and on the threat. These transition probabilities may be influenced by factors such as the resilience of a CI against the considered threat.

This probabilistic model is the basis for a Cascading Effects Simulation (CES) that allows to mimic the spreading of the effects of the flood through the CI network by probabilistic changes of the sates of the nodes. It allows to investigate different types of floods, e.g., heavy rainfall for a short time or long periods of rain. The simulation also allows to investigate different setups of the CI network, i.e., it is possible to investigate how changes affect the spreading process. Possible changes are:

* Adaption of the network topology, e.g., adding backup components such as a second supplier of power * Adaption of the reaction of one or more components to the threat, e.g., increasing protection measures against flooding

The simulation can be re-run for different options and comparison of the results provide insights into the effectiveness of the considered changes. This comparison is supported by a graph visualization of the CI network, where nodes are coloured depending on their state to demonstrate how much they are affected.