Comparative Analysis Of Roundabout Capacity Analysis Methods

According to the Washington State Department of Transportation (WSDOT), the use of roundabouts has been increasing due to its numerous benefits. Compared to other types of intersection (conventional) controls, a roundabout improves traffic safety, reduces delays, lessens the severity of crashes, improves traffic flow, and minimizes maintenance costs. The Federal Highway Administration (FHWA) report that in some previous study findings the injury and fatality crashes were reduced by 78-82% after converting signalized intersections to roundabouts. This study used data from the City of Fairborn, Ohio, for a project where they were preparing to replace a 50-years old signalized intersection (at Kauffman Avenue and Colonel Glenn Highway) with a modern roundabout. This signalized intersection experienced 18 crashes between 2013 and 2015. The modern roundabout was scheduled to be constructed in 2019 by the City of Fairborn, and it was designed with three legs (similar to the existing intersection layout), with three entries, three exits, and one circulating lane. Two software programs were used in this study, which are RODEL and Highway Capacity Software (HCS7), and two manual methods were also used, which are 2010 Federal Highway Administration method (Roundabouts: An Informational Guide) and 2016 Highway Capacity Manual (HCM 6). The main aim of this study was to compare capacity analysis results to the same data inputs among all methods. The results from this study are helpful in explaining the performance of each method and provide an understanding of analyzing roundabout capacity. The comparative results for each method include capacity, volume-to-capacity ratio, control delay, level of service (LOS), 95th percentile queue length for each approach/lane, and the entire intersection control delay and LOS. The results show that RODEL (geometric model) predicted the largest capacity, the smallest volume-to-capacity, the lowest delay, the best LOS, the shortest 95% queue length, and the lowest delay and best LOS for entire intersection. This supports the conclusions of previous studies. The FHWA 2010 results were quite the opposite; estimated the lowest capacity for the same traffic and geometric inputs, largest queues, delays and volume-to-capacity ratio. The results from HCS7, 2016 HCM (HCM 6), 2010 FHWA and RODEL (HCM 2016 model) were similar because they use the same methodology. HCS7 software automates the HCM6 methodologies.