Characterization of Vulnerability of Road Networks to Random and Nonrandom Disruptions Using Network Percolation Approach

Abstract

This paper examines the vulnerability of road networks to two types of disruptions by modeling the percolation dynamics in road networks under different disruption scenarios. The objective of this paper is threefold: (1) to examine if the theoretical network robustness measure proposed in the literature is applicable for measuring the integrity of road networks during disruptions; (2) to unveil the impacts of network size on the overall vulnerability of road networks; and (3) to compare the performance profile of road networks to random and nonrandom types of disruptions. To that end, this study first modeled the road system in a community as a planar graph. Then, the percolation dynamic in the road network during the flood is captured by assigning different removal probabilities to nodes in the road network according to Bayes’ rule that take floodplain types, node-elevation, and street-grade as inputs. In the end, an overall road network robustness measure and its temporal changes were obtained and for random and nonrandom scenarios, using road networks with different sizes. The results were compared in order to characterize the vulnerability of road networks under different scenarios. The proposed method was applied to the road network in central Houston during Hurricane Harvey. The results show that: (1) the theoretical network robustness measure is applicable to assess the road network robustness; (2) compared to the random percolation model, the probability (Bayes’ rule) based percolation could lead to a greater decrease in the network robustness; and (3) the percolation profiles of the road networks with different sizes are not significantly different. The findings of this study could not only inform resilience enhancing decisions by the stakeholders but also could serve as a foundation for future vulnerability related research.