Capacity Estimation on Turboroundabouts with Gap Acceptance and Flow Level Methods

In the literature, linear models and exponential models based on gap acceptance theory are distinguished. Parameters for the linear models can be estimated only at the level of traffic flow, whereas the gap acceptance theory assumes that behavioral parameters (critical gap, follow-on time, and minimum headway) can be estimated at the vehicle level, and then capacity can be determined. However, in the latter method, measurements must also be made under saturated conditions because of so-called pseudoconflict (caused by vehicles leaving the roundabout in the opposite leg direction). An analysis of data sources (vehicle level and traffic flow level) indicates that the parameters estimated at the vehicle level correspond with those estimated at the level of traffic flow for single-lane roundabouts, but not for more complex situations. Despite this deficiency, better results can be gained with a model based on the present gap acceptance theory than with a linear model, provided that the parameters are adjusted to consider traffic flow measurements under saturated conditions. Additional research will be necessary to determine which underlying assumptions of the gap acceptance models cause these deficiencies. In Germany, capacity is higher in turboroundabouts than in compact two-lane roundabouts because of better use of the inner lane in the turboroundabout.