Book Description

Congestion pricing and managed lanes (ML) have been recently gaining interest around the country as a congestion management tool and as a means of revenue generation for facility maintenance and expansion as well repayment of highway construction debts. Congestion pricing in MLs entails one of several strategies, including time of day pricing, dynamic pricing based on predicted/anticipated traffic conditions, and real-time dynamic pricing based on actual traffic conditions. The overall goal of this study has been to develop a Decision Support System (DSS) tool based on drivers' revealed willingness to pay (WTP) values. This should determine more effective dynamic toll pricing that achieves the ML corridors' operational goals. A key challenge has been estimating drivers' revealed WTP as influenced by their perceived values of time or by other factors such as enhanced safety and more reliable travel times. While there are significant advances made in the available methods to estimate WTP, research still lacks in the area of dynamic pricing. Indeed, for dynamic toll pricing systems, setting real-time toll prices based only on drivers' average WTP values appears ineffective. However, the WTP values estimated through existing methods represent the average value of travel time saving and/or reliability (VOT and/or VOR) for the total population. This makes the current approaches more compatible with static networks, which cannot efficiently address the nature of dynamic corridors. Another major drawback associated with current methods is that the travelers WTP values are measured in terms of price paid to save one unit of travel time (VOT). However, the travelers' WTP to use MLs has been shown to be for a number of intertwined reasons and not just for time savings. This study suggests a number of unique approaches in estimating WTP values. These include a new revealed data source as well as an alternative analysis method for estimating WTP. To obtain more accurate results, the study was limited to the North Tarrant Expressway (NTE) drivers in North Texas and was conducted for different time periods. For this study, traffic count data were reduced from the camera images for different vehicle categories and for five different time periods, including AM and PM peaks, AM and PM inter-peaks, and off-peak periods. In addition, real-time toll prices associated with the study segment and the day and time of the data collection were obtained from the NTE website. The data analysis method involved an existing toll pricing model (TPM) developed in a former Texas Department of Transportation study for setting tolls for MLs. The model was modified and calibrated based on actual ML shares and associated toll prices for the NTE ML corridor. The modified version of TPM (version 5.0) can be employed as a DSS tool to estimate the WTP values for drivers of any vehicle class and for any time of day. Values of about $119, $101, $71, $75, and $59 per hours were estimated as the revealed average WTP for the NTE SOV drivers during AM peak, PM peak, AM inter-peak, PM inter-peak, and off-peak periods, respectively. In addition, a value of $85 per hour was estimated for the mean revealed WTP (all periods inclusive) for the NTE SOV drivers. The results of this study showed relatively high WTP values and ML share percentages for the NTE drivers, indicating a high level of acceptance of MLs in the region. Finally, this study suggested applying a new paradigm in WTP estimation studies. The employed data collection and analysis methods were two components of the new paradigm. Besides, the new paradigm recommended evaluating real-time WTP by time of day instead of average WTP values for dynamic pricing schemes. The last component was a recommendation to attribute the WTP values to the travelers' willingness to pay to drive one unit distance on toll lanes instead of to save one unit of travel time.The DSS tool developed in this study for the NTE ML has the potential to be used by ML operators to measure the real-time WTP values for the ML users. The results of this new methodology may not directly address the questions about travelers' behavior in terms of their reasons to choose between the MLs and GPLs. However, these results can significantly contribute to decision making about transportation policies, in particular, the policies associated with dynamic congestion pricing for ML corridors.