Empirical Framework for Characterizing Infrastructure Failure Interdependencies

This paper develops an analytical framework with empirical applications to characterize infrastructure failure interdependencies (IFIs). It uses major electrical power outages as the context for understanding how extreme events (within or external to the power system) lead to failures of other infrastructure systems, given a major electrical power outage. The paper takes an empirical approach by examining the patterns of IFIs that occurred in three kinds of events: the August 2003 northeastern North American blackout, the 1998 Quebec ice storm, and a set of three 2004 Florida hurricanes. Data sources include media reports and official ex post assessments of the events. The results characterize IFIs in terms of the sectors affected, and the consequences for society. We developed scales to characterize the consequences of IFIs in terms of impact and extent indices. A comparison is provided of IFIs arising in all five events discussed in the paper, as a basis for considering priorities for risk mitigation. The most significant IFIs in all five events included effects on HVAC in buildings; effects on water systems; effects on health systems, including hospitals and public health efforts, and effects on road transportation systems.