Simulation-based capacity estimation and speed-drop analysis of on-ramp merging sections in a corridor with mixed traffic stream

Understanding on-ramp merging process is important for effective planning, design and operations of the on-ramp merging area. Mixed traffic streams with varying vehicle sizes and maneuverability, and lack of lane-discipline leading to lateral interactions are common in developing countries. Available studies on on-ramp merging operations in mixed traffic streams did not explore the effect of traffic composition and merging section geometry on the capacity and speed drops in the merging area. In this study, field data collected from the on-ramp merging section of an urban highway was used to develop a calibrated and validated microsimulation model. The model calibration involved optimizing conflict area and car-following parameters, and validation included comparing the model performance with another section on the same corridor. The calibrated model was used to evaluate the capacity and speed drops in the merging area for 600 scenarios representing various combinations of vehicle compositions, flow rates, merging section lengths, and mainline and on-ramp widths. Scenarios simulated for three hours with and without on-ramp traffic revealed merge area length, truck composition in mainline, merging traffic flows, and width of mainline and on-ramp as the statistically significant factors. The capacity dropped about 15% for mainlines with two, three, and four lanes. A second-order polynomial speed drop model developed from the simulation data indicated optimum merging section length of 100 m for mainline with two lanes and 150 m or more for mainlines with three and four lanes. The speed drop is maximum at about 1400 PCU/h/lane mainline flow and monotonically increase with ramp flow.