Logical expansion of the recursive decomposition algorithm for infrastructure interdependency analysis

This article deals with the network reliability analysis of interdependent lifeline networks whose survival is defined by complex general system events with multiple source and demand node pairs. A new method, the Logical Expansion of the Recursive Decomposition Algorithm (LE-RDA), is introduced to handle complex events by successively applying a generalized decomposition process for two logical functions (intersection and union). The method is applied to an electric power grid and a water network in Memphis/Shelby County to estimate the seismic performance, functional degradation, interaction effect, and the contribution of network component groups to water flow distributions, all of which are defined by a complex mixture of correlated events with shared components. The significance of the interdependency effect is addressed and the network component groups' priorities are identified for various earthquake magnitudes and interdependency levels.