Sensitivity of Travel Reliability Measures to Major System Changes

The performance-based planning and programming approach established by the Moving Ahead for Progress in the 21st Century Act (MAP-21) has necessitated the process of estimating congestion (delay) and reliability benefits from transportation system improvements. Agencies are experiencing a shift in policy focus from the average travel time to the reliability of travel time. By examining empirical data on before-and-after archived travel times at identified project locations to find evidence of observed reliability benefits, this paper addresses a gap in the literature about the overall role that reliability benefit estimations play in the decision-making process during transportation project prioritization, and the relative effects of different strategies on different reliability measures. In the process, the study investigated why and how the way the reliability measures are defined can affect their ability to capture the impacts (and benefits) of major system changes. Data for four different broad improvement categories and three different reliability performance measures (planning time index-95 [PTI95], planning time index-80 [PTI80], and level of travel time reliability [LOTTR]) revealed varying impacts of different strategies on different measures. PTI80 was better able to capture ground-level benefits experienced by most travelers and showed higher sensitivity to system improvements than PTI95. Another key finding was that LOTTR’s threshold-based definition limited its ability to capture observed reliability benefits accurately. The results provide a picture of magnitude and relative levels of observed benefits from different system improvements, which could help planners and practitioners better understand, anticipate, and incorporate the benefits expected from planned changes.