ANALYSIS OF MANAGED LANES ON FREEWAY FACILITIES

Abstract

Roadway agencies face growing challenges to expand freeway capacity. Under consideration of rising construction costs, right-of-way limitations, and environmental constraints, transportation agencies are seeking solutions to efficiently manage the demand on existing freeway facilities and to provide options for travelers. The concept of Managed Lanes (MLs) is an increasingly popular countermeasure that aims to make the most efficient use of freeway facilities, by restricting access to one or more lanes, and to certain vehicle classes on a facility that is parallel to existing General Purpose (GP) lanes. With limited work being documented for analytical evaluation of ML facilities, transportation agencies often rely heavily on time-consuming and cost-ineffective simulation experiments for ML performance analysis. Therefore, there is an urging need for a framework to guide the practitioners for a systematic evaluation of ML system. This research aims at filling this gap by developing a methodological framework for analyzing freeway facilities with ML and GP lanes operated in parallel. The framework acknowledges that the composition and behavior characteristics of the ML traffic stream are expected to be quite different from those for the GP lanes in terms of traffic volume, free-flow speed, capacity, vehicles type, etc. The framework further considers that there may still be certain levels of interactions between these two lane groups, especially for those facilities do not have physical (barrier) separations, either en route or at access points, between them. Within that framework, different modules were developed based on sensor-measured or simulation-generated data, including the characterization of ML speed-flow relationship, the frictional effect of adjacent lane traffic speed, the adjustment for cross-weave effects, and the development of side-by-side facility-wide ML and GP performance measures. Thus, the proposed methodology is sensitive to different GP and ML segment types (basic, weaving, etc.) and separation styles (stripe, buffer, barrier), and is capable of analyzing extended facilities across multiple time periods. The methodology is further implemented in a computational engine for testing and application. Furthermore, a mesoscopic simulation model, entitled twin-cell modeling approach, on the basis of Cell Transmission Model (CTM), is developed in this research to describe the traffic evolution of parallel facilities. The twin-cell modeling approach quantifies the frictional effect, which is unique to the parallel facilities, via data mining techniques, and also incorporates a weaving component that extends the original CTM model to the grid-level parallel facility modeling. In summary, this research attempts to develop an analytical methodology for evaluating freeways with parallel GP lanes and MLs. The methodology is supported by extensive field data collected at ML facilities in several states of the U.S. The developed methodological framework presents an important, new approach for the performance analysis of ML-enabled freeway segments, which is valuable in providing guidance for analysts in evaluating freeway segments in the presence of concurrent GP lanes and MLs.