Distribution of Value of Time and Ways to Model Value of Time in Long-Range Planning Models
As managed lanes (ML) become more integrated in regional urban networks with existing general purpose (GP) lanes, the distribution of travelers’ value of time (VOT) is becoming more important for transportation planning agencies to quantify in order to accurately predict future travel patterns. Since travelers’ VOT varies depending on a multitude of factors, this study investigates ways that we can determine the VOT distribution of a region from existing travel data as well as effective ways that we can model VOT using traffic assignment algorithms. In networks with available link volumes and toll data on segments where travelers have the option of choosing to stay on the GP lanes or entering a ML facility, a VOT distribution can be inferred assuming that travelers who enter the ML choose to do so based on a certain “threshold” VOT. When modeling these VOT distributions, errors are observed in the traffic assignment results when both the continuous nature of VOT distributions are discretized, and when varying toll values are assumed to be constant. Specifically in the context of TransCAD software, link travel time errors appear to be much less significant than flow errors when tested on a nine node network. Additional experimentation on larger regional networks is needed to verify the significance of these errors and their impact on predicted travel patterns.