Dispersion of traffic queues

Vehicles move away from signalized intersections in queues and start to accelerate, causing the vehicles to naturally disperse from the queues. It is important to specify the distance across which the traffic queue has fully dissipated as this distance will determine the positions at which signalized intersections are to be established. The objective of a recent study was to determine this queue dispersion distance on two peri-urban roads around the Stellenbosch district, namely the R44 and R310. The selection of a time headway distribution model is a fundamental part in determining the dispersion distance of a traffic queue. Research has shown that the model that provided the most accurate fit is the shifted negative exponential distribution. This model was formulated by shifting the negative exponential distribution model with the minimum time headway. The distance from the intersection at which point the queues have dissipated fully, was determined by using a time headway of 3.5 sec as being the cut-off between followers and non-followers. The dispersion distance for the R310 was determined to equal 595 m and for the R44 at 690m. It must however be noted that the R310 is susceptible to the ability of vehicles to cross lanes which can influence the accuracy of the dispersion distance. This paper describes the background theory, the data collection and analyses and the results.