Modeling Impact of and Mitigation Measures for Recurring Freeway Bottlenecks

Abstract

Recurrent congestion is a continually growing problem on urban freeways. Facility expansions cannot keep pace with the growing vehicle demand. Low-cost mitigation measures are one way to alleviate the congestion at recurring bottleneck locations. Low-cost measures typically have a life of approximately 10 years and costs ranging from $8,000 to $2.45 million. While benefits have been realized in field applications, there hasn't been a lot of study regarding the performance of these measures in terms of added capacity. While modeling has long been a tool for planning and analyzing freeway networks, there has been little reported regarding its use for estimating the benefits of low-cost freeway improvements. In this study, the author tested proposed treatments at two sites using both a macroscopic and microscopic model. Because empirical performance information of these measures is not available, a quantitative analysis would not be reasonable since confidence in the values reported would be low. Current bottleneck identification methods typically either predict breakdown in real-time, or analyze detector data off-line. In order to identify bottlenecks from recorded aggregated data in an off-line model, criteria were generated to identify active bottlenecks and analyze the models' performance in an empirical and qualitative manner. Application of the criteria has been shown to provide reliable bottleneck identification to the calibrated pre-treatment case and expected results in the post-treatment cases.