Impact of Operational and Restoration Interdependencies on Cost and Disruptive Effect in Multilayered Infrastructure Networks

Abstract

Network-based modeling and simulation of interdependent infrastructure systems has traditionally modeled 1 to 2 interdependency subtypes within a given model. Typically, different interdependency subtypes are modeled with special sets, variables, and parameters rather than integrated into the model as a standard feature. The most thorough current modeling efforts have modeled only four of the nine interdependency subtypes identified throughout academic literature. This paper presents the first model to incorporate all nine of the identified interdependency subtypes in a multiobjective mixed-integer program. The model is tested using a simulated flood event and a modified interdependent infrastructure dataset representing a medium- to large-scale military base. Model results identified a satisfactory short-term recovery level avoiding an additional $6 million of temporary repairs for a minimal increase in system operability. Inclusion of all nine interdependency subtypes developed accurate repair and cost schedules, whereas exclusion of any of the nine interdependency subtypes resulted in over or underestimations in cost and repair time. This model provides a more accurate repair schema when recovering from a system disruption.