Modeling s-t Path Availability to Support Disaster Vulnerability Assessment of Network Infrastructure

Abstract

The maintenance of system flow is critical for effective network operation. Any type of disruption to network facilities (arcs/nodes) potentially risks loss of service, leaving users without access to important resources. It is therefore an important goal of planners to assess infrastructures for vulnerabilities, identifying those vital nodes/arcs whose debilitation would compromise the most source-sink (s-t) interaction or system flow. Due to the budgetary limitations of disaster management agencies, protection/fortification and planning for the recovery of these vital infrastructure facilities is a logical and efficient proactive approach to reducing worst-case risk of service disruption. Given damage to a network, evaluating the potential for flow between s-t pairs requires assessing the availability of an operational s-t path. Recent models proposed for identifying infrastructure vital to system flow have relied on enumeration of all s-t paths to support this task. This paper proposes an alternative model constraint structure that does not require complete enumeration of s-t paths, providing computational benefits over existing models. To illustrate the model, an application to a practical infrastructure planning problem is presented.