Determining Louisiana\u2019s Roundabout Capacity

US Transportation Collection2021PDFTech ReportCodjoe, JuliusThapa, RajuAyernor, Sam-Mark AnsahLoker, MatthewLouisiana Transportation Research CenterLouisiana. Department of Transportation and DevelopmentUnited States. Department of Transportation. Federal Highway AdministrationLouisiana. Department of Transportation and DevelopmentLouisianaUnited StatesEstimatingHighway capacityRoundaboutsTraffic delaysAutomatic headway controlCalibrationfollow-up timecritical headwayLOSSIDRAFHWA/LA.17/640Final Report, 01/2019 \u2013 09/2020This study calibrated the HCM roundabout capacity models to incorporate local Louisiana driver behavior at single-lane roundabouts, using gap acceptance parameters measured from field data at forty-one approaches from seventeen roundabouts throughout the state. The correlation between these gap acceptance parameters and certain geometric features were explored. Delays, queue lengths, and levels of service (LOS) were determined using appropriate equations from the HCM 6th Edition (HCM 6). Furthermore, a SIDRA analysis was undertaken to determine whether the default environment factor (EF) of 1.2 currently being used by DOTD for the Build Year design was appropriate. Using a maximum likelihood estimation analysis to determine the gap acceptance parameters, it was deduced that the average follow-up time and critical headway for Louisiana drivers is 3.36 seconds and 4.76 seconds respectively. The national averages from the HCM 6 were 2.60 seconds and 4.70 seconds respectively. No notable difference in gap acceptance parameters existed across DOTD districts/parishes, nor different functional classes of roads, although arterials had the least follow-up times and critical headways. To determine whether geometric features contribute to the observed follow-up times, inscribed circle diameter (ICD), splitter width, central island diameter, and approach lane width were compared to the follow-up times. Even though follow-up times most correlated with ICD, none of the relationships proved strong enough to be included in the local model calibration. The local model was developed by recalibrating the HCM 6 model to a fixed intercept based on the average observed follow-up time, and a slope parameter based on regression of the data. The local model results showed that it performed closer to the HCM 2010 model than the HCM 6 model. The SIDRA analysis showed that an environment factor (EF) of 1.06 generated approach capacities with the least root mean squared error (RMSE) when compared to approach volumes from the local model for given circulatory volumes. This value corresponds to an average age of eight years for the 17 roundabouts evaluated. Comparing delays, queue lengths, and LOS theoretically determined for the roundabout approaches, to the SIDRA outputs generated mixed results. Future studies should find innovative ways to determine these parameters directly from the field to compare with SIDRA outputs.121