Researchers apply travel time reliability measures in modeling and project evaluation

By Chris McCahill

Transportation planners and traffic analysts who typically measure road performance in terms of delay are beginning to incorporate measures of travel time reliability, which describe the hourly and daily variation in travel times due to congestion and transit delay. These variations are considered an added cost incurred by travelers, which affects their mode choice and other travel decisions. Two new reports prepared for the Strategic Highway Research Program (SHRP 2) provide a better understanding of the value of travel time reliability, and also insight into what might be the most appropriate uses for such measures.

In Maryland, researchers incorporated travel time reliability measures into benefit-cost analyses for highway projects by including both travel time reductions and reliability improvements, noting that a failure to do so might result in undervaluing project benefits. Management from the Maryland State Highway Administration generally approved of the reliability measures, which the agency already uses, but raised concerns about focusing too heavily on congestion relief. During a presentation by the research team, management suggested that reliability should be viewed as a better goal than “fixing congestion,” which they considered “not necessarily feasible in today’s world of financial constraint and other competing issues,” according to the report.

Researchers in Oregon improved their travel demand modeling and dynamic route assignment capabilities by developing a “travel time equivalent,” which accounts for the reliability of network segments throughout the day. They simulated 15 scenarios by incorporating bus rapid transit and variable message signs on selected corridors in the Portland area. Transit ridership increased and overall corridor performance improved when automobile and transit reliability were both accounted for. The benefit of variable message signs, which help balance automobile traffic between corridors, was also more pronounced when reliability was accounted for, despite causing a slight decrease in bus rapid transit ridership.

Both groups of researchers estimated the value of travel time reliability to be roughly 75% of the value of overall travel time—which is consistent with the literature—but also recognized that these values vary by mode, trip purpose, time of day, and other local conditions, meaning that context-specific estimates are generally preferred. The Oregon team estimated the value of travel time reliability using stated-preference surveys administered online, while the Maryland team based their estimates on a year’s worth of probe-based travel data.

Chris McCahill is a Senior Associate at SSTI.