Skip to main content
Infrastructure Resilience Conference 2018

Full Program »

The Robustness of Hierarchical Critical Spatial Infrastructure Networks

The robustness of critical infrastructure networks is pivotal as these are the building blocks for the economic state and wellbeing of societies (Rinaldi, 2004; HM Treasury, 2010); disruption to critical infrastructure networks can result in significant financial as well as human consequences. Infrastructure networks such as those for communications, energy and water, are increasingly exposed to a complex set of hazards, with climate change increasing the occurrence of weather-related events including flooding and lightning storms (Royal Academy of Engineering, 2011). Therefore, improving the robustness of infrastructure networks and thus reducing the impact of such hazards upon them can create a more dependable set of networks aiding economic security and national competitiveness.

Modelling of spatial infrastructure network robustness has traditionally focused on their local topological properties, with networks often characterised via their degree distribution (Newman, 2003). However, it has been suggested that this simple measure misses many of the critical properties of network robustness, particularly in relation to networks that may be hierarchically organised (Pastor-Satorras et al., 2004). Therefore, in this paper we use a suite of higher-level graph metrics; namely, betweenness centrality, cycle basis and assortativity coefficient to evaluate the robustness of 42 critical spatial infrastructure networks to targeted and random failure models. Moreover, we statistically compare and associate the structure of these infrastructure networks with a family of 8 graph-theoretic models ranging from random through to hierarchical tree’s in order to associate failure characteristics with the inherent structure of the networks.

The analysis revealed that 39 of the 42 infrastructure networks analysed exhibited failure characteristics similar to hierarchical graph-models, including networks such as the national electricity transmission and distribution network for the UK, and the Tube network in London. These hierarchical networks experienced total failure on average 27% quicker than the non-hierarchical networks and were also shown to fragment into multiple components after <10% of node assets were removed; thus showing little positive redundancy (multiple cycles) in terms of structural characteristics compared to non-hierarchical networks. The results highlight that many critical infrastructure networks which have previously been considered relatively robust on the basis of degree distribution are in-fact highly vulnerable to targeted failures when considered in terms of the their overall structural connectivity and accessibility.

References HM Treasury (2010) Strategy for National Infrastructure. UK: HM Treasury,. Newman, M.E.J. (2003) 'The Structure and function of complex networks', Physics, pp. 1-58. Pastor-Satorras, R., Alexei, V. and Vespignani, A. 440 (2004) 'Topology, Hierarchy, and Correlations in Internet Graphs' Lecture Notes in Physics. pp. 425-440. Rinaldi, S.M. (2004) 'Modeling and Simulating Critical Infrastructures and Their Interdependencies', International Conference on System Sciences. Hawaii. Royal Academy of Engineering (2011) Engineering the Future. London, UK: Royal Academy of Engineering.

Craig Robson
Newcastle University
United Kingdom

Stuart Barr
Newcastle University
United Kingdom


Powered by OpenConf®
Copyright ©2002-2016 Zakon Group LLC