Analyzing the Vulnerability of the Intersections in Large Networks during a Crisis Occurrence (Case study: Tehran)

The intersections in the network are considered as one of the network vulnerability during a crisis occurrence. This study aimed to evaluate the role and importance of intersections in network performance and accessibility damages during a crisis occurrence through which the prioritization of intersections for optimal allocation of resources to ensure the appropriateness of network performance is made possible. Accordingly, a method was proposed to identify vulnerable intersections during a crisis based on comparing their delays in two situations 1) without lights and any kind of control equivalent to the maximum turbulence at the intersection and 2) lighted with optimal timing schedule as a situation equivalent to the control of the intersection by the police. In such a method, the intersection delay in both conditions is obtained from the mean delay of the links leading to it based on the calibrated delay functions. Also the basis of the optimal timing schedule and its parameters to apply the delay function in the lighted intersection mode is Webster method. In addition, it is assumed that the pattern of trips in pre-crisis situation based on the daily trips is maintained to determine the volumes reached at the specific intersection from each link leading to it during crisis occurrence, which this pattern is obtained from analyzing named “critical link analysis” as one of the results of trip assignment models. It is worth noting that the results of this analysis for a given link are a matrix, indicating the contribution of the origin-destination pairs passing through the volumes observed in that link. Subsequently, the passing volume from a specific link related to each origin-destination pair in crisis situation is obtained through multiplying the ratio of crisis demand of that origin-destination pair to that of the pre-crisis situation by the corresponding array in the matrix obtained from the critical analysis of the that link based on pre-crisis situation trip demand assignment. The sum of results for all pairs of origin-destination is equal to the total volumes passing through the link in crisis. Then, the proposed method was implemented in Tehran based on the mean estimated evacuation demand due to the earthquake caused by the existing main faults to evaluate its efficiency. In this study, value of 46% was considered as the limit of acceptable changes in delay at intersections to identify vulnerable intersections. This value is the minimum possible value for this threshold, provided that it is possible to construct a matrix of the shortest free travel time from the all damaged zones to all evacuation destinations in the modified network due to the removal of all vulnerable intersections. It was obtained from a repetition-based process. According to the results of the project implementation in the case of study, 12.7% of network intersections are known as vulnerable intersections. Further, the roads with arterial functional category have the most vulnerable intersections per unit length of roads, indicating their low reliability of travel time in the whole network in the case of this research, which is considered as a result of the inappropriate structure of the network in separating the duties of the roads, especially providing access.