The evaluation of modern roundabouts as an alternative to signalized and two-way stop controlled intersections in a urban and rural environment

Abstract

The comparison of roundabouts with other intersection forms of traffic control is becoming an increasingly common occurrence. With little overall experience within the United States, the data for roundabout comparison is somewhat varied. This report includes the results of five models created using VISSIM traffic modeling software in the comparison of a two-way stop, signal and roundabout in an urban environment, and a two-way stop and roundabout in a rural environment. A large amount of field data in the form of traffic volumes, vehicle types, gap timing and headway distances were used in the calibration and validation of the various models. Multiple runs of the models were conducted to attain an expansive data pool from which to evaluate the different modes of traffic control. Measures of effectiveness included crash reduction, delay, and queue length. Due that crash data could only be obtained for the before period of the base case studies, a detailed crash reduction analysis could not be conducted. A limited analysis was conducted to review the possible reduction in the possibility of accidents and their severity. Delay values indicated the roundabout provided the best performance, followed by the signal, and lastly the two-way stop. Average queue length data indicated that the roundabouts functioned with no notable queuing experienced. The results of the signal indicated improved performance on that of the two-way stop. The results indicate that roundabouts are a viable alternative to a two-way stop and signal. For a more accurate, long-term evaluation, growth projections should be applied to existing conditions to determine the operational effectiveness of the intersection traffic control type at some point in the future. Depending on the purpose and need of any intersection being evaluated, additional focus can also be concentrated on obtaining additional field information for driver behavior, gap timing, headway distances, delay times and average queue lengths.