Impact of Traffic Incidents on Reliability of Freeway Travel Times

Abstract

Travel time reliability is a key measure of a freeway system's performance. Traffic incidents are one of the more important factors affecting travel time reliability because they reduce the capacity of a freeway segment and generate a temporary bottleneck. In this study an empirical travel time reliability analysis was conducted with the use of 4 years of travel time and incident data collected on Interstate 5 and Interstate 405 in the Seattle, Washington, metropolitan area. The incident data used for this study are notable for their highly detailed information for each incident. Three incident types (i.e., shoulder, single-lane, and multiple-lane incidents) were considered in the data analysis. Travel time reliability was analyzed through several measures, including travel time variability, buffer index, and probability of freeway segment traffic breakdown. The results show that incidents result in higher values for all these measures. To be more specific, multiple-lane incidents induce the highest buffer index and variability in freeway route travel times, whereas shoulder incidents induce the lowest buffer index and variability. Travel time variability and buffer index during multiple-lane incidents increase by an average of 205% and 237%, respectively, compared with normal conditions (conditions with no incidents). Compared with normal conditions, shoulder incidents can significantly increase the probability of freeway segment traffic breakdown. In general, incidents can significantly reduce travel time reliability. The findings in this study can help shape incident mitigation and management policies for different incident types, especially when the aim is to improve freeway travel time reliability.