| Summary: | Critical infrastructure and similar lifeline networks rarely operate in isolation. Whether a subway car depend on electricity, or the control of a power station on the telecommunications network, their ability to operate and cooperate is keenly dependent upon the state of the connected network. Understanding how the nature of these couplings affect the networks in the face of a significant perturbation and throughout their respective recovery processes is a key element to
building robust and resilient infrastructures. Previous research has revolved around building models for specific types of networks usually with bidirectional couplings and a percolation approach to failure. This paper presents a simple framework for modelling the dynamic behavior of pairwise coupled, spatially embedded infrastructure networks. The proposed model is based on directed links connecting pairs of networks and allows for fractional nodal functionality, as mutual dependences
between nodes and a binary approach to operability is not the general rule in infrastructure networks. This allows for examination of how the structure of inter-network couplings affects the magnitude of sustained damage post perturbation and the subsequent repair required to fully restore functionality. We find that a wider geographical spread between the points connecting coupled networks is associated with an increased magnitude of sustained damage, as well as a higher fraction of
damaged links which must be repaired in order to fully restore functionality. Additionally, the relative location of coupled nodes in relation to one another can increase the probability of sustaining significant damage, i.e. damage that cannot be restored by repairing initially broken edges. Surprisingly, there is also evidence suggesting that properties of the nodes that couple the networks do not have a statistically significant effect on the magnitude of damage each network
sustains.
|
|---|
