Evaluation of Methodologies for Analyzing Freeway Ramp Weaving

Weaving sections are common design elements on freeway facilities such as near ramps and freeway-to-freeway connectors. Traffic operational problems often exist at weaving areas even when traffic demands are less than capacity because of the complexity of vehicle interactions, resulting in poor level of service and potential safety problems. The objective of this research was to evaluate the methodology for analyzing freeway weaving in the Highway Capacity Manual 2010 (HCM 2010) and other methodologies for evaluating weaving, such as Level D and Leisch, that are used for design and analysis in California. The evaluation was based on the comparison of methods' predictions with field data. Emphasis was placed on the evaluation of ramp weaves consisting of a one lane on-ramp followed by an off-ramp with a continuous auxiliary lane. The findings from the evaluation of the HCM 2010 method on existing California weaving sites (93 data points) indicated that the method overpredicted the traffic densities and associated loss at ramp weaves by 24% on average. Additional data were collected at three ramp weaving sites in San Diego, California, by using Bluetooth readers and video recordings. The HCM 2010 method was found to overpredict density by an average of 13.4%. The Level D and Leisch methods overpredicted density by about 3% on average compared with observed values. The HCM 2010 method under predicted weaving section capacity when the weaving volume was above 30% of total weaving volume.