Developing Capacity Models for Local Roundabouts

Abstract

Modern roundabouts are gaining popularity in many parts of the United States. However, because of a lack of well-established capacity models, many transportation professionals rely on international experience when analyzing the capacity of U.S. roundabouts. From a comprehensive evaluation of U.S. roundabouts, the authors of NCHRP Report 572 developed capacity models for single-lane and multilane roundabouts. NCHRP Report 572 suggests that these models can be calibrated for local conditions by adjustment of two key parameters: critical headway and follow-up headway. However, there is little direct connection between such calibrated models and the actual flow rates achieved in the field. For communities with congested roundabouts, this paper presents a streamlined process to develop capacity models. The process consists of video data collection, data processing and verification, and model development. The process was applied to a case study of three roundabouts in Carmel, Indiana. The single-lane roundabout capacity model developed in the case study is substantially different than the model in NCHRP Report 572. Overall, the streamlined process to develop capacity models for local roundabouts proved valuable, and it can be scalable according to available resources. The capacity models developed by this process are based on actual flow rates at local congested roundabouts and will provide an accurate capacity estimation for planning new roundabouts or capacity improvements. The difference between the case study's capacity model and the NCHRP Report 572's capacity model may be partially attributed to driver familiarity. The study underscores the importance of using capacity models that are applicable to local conditions.