Spatial optimality in power distribution networks

Critical infrastructures can be characterized as networks where nodes and edges are embedded in space. Transportation networks, the internet, and power grids, are examples of networks where spatial constrains are relevant. An important consequence of space is that there is a cost associated with the length of edges which in turn has important effects on the topological structure of and on the dynamical processes which take place on these spatial networks. In this paper we investigate the effect of the wiring cost in the spatial organization of a sample of power distribution networks by means of shuffling the networks in systematic ways. We show that although they share similar topologies, suboptimal networks (i.e., those with topologies not organized to reduce the wiring cost) seem to accumulate more failures. Consequences of these results and further work are finally discussed and outlined.