Abstract
High-occupancy toll (HOT) lanes are in operation, under construction, and planned for in several major metropolitan areas. The premise behind HOT lanes is to allow single occupant vehicles (SOVs) to access high occupancy vehicle (HOV) lanes (and theoretically, a higher level of service) if they are willing to pay a toll. To maintain a high level of service in the HOT lanes, the toll rate is set dynamically to restrict the number of SOVs which access the facility as it nears capacity. Thus, HOT facilities provide operators of transportation systems with an additional tool: pricing. In order to effectively use pricing, it is critical to understand driver behavior when faced with a set of traffic conditions and toll levels. This thesis presents the results of an empirical investigation into the relationship between toll rate, traffic conditions, and SOV driver behavior, based on data from the dynamically-tolled I-394 HOT facility in Minneapolis, Minnesota. Analysis of the empirical data indicated that of the SOVs using the HOT lanes, 87.5% use the HOT lanes at predictable rates throughout the AM peak period, even when there is no clear travel time advantage. After accounting for these “regular” users, the remaining “price-sensitive” SOV drivers utilize the HOT lanes at greater rates when the cost per hour of commute time saved is lowest. A model was developed that incorporates both of these findings, predicting HOT lane usage rates based on time savings, time of day, and toll rates with an R<sup>2</sup> value of 0.684, <i>n</i> = 27831. When compared to the historic HOT utilization rates only, i.e. assuming all HOT lane SOVs are “every day” drivers, the resulting model has an R<sup>2</sup> value of 0.675, <i>n</i> = 27831. Thus, the pricing structure as in place at this facility, appears to have a negligible influence on behavior. This may indicate serious implications, as many HOT facilities are under consideration partly for their potential as traffic management tools.
