Accidents are a complex process involving many contributory factors. The understanding of the accident process has often been sought by the use of accident data. Although accident data provide a direct relationship to estimating accident risk, there are many drawbacks associated with the use of these data. The major drawback with the use of accident data is the very fact that traffic engineers have to wait for accidents to occur before any interventions can be made. This alone is significant as the time span required to collect a sample size is often a three-year period. The many deficiencies with accident data have led to alternative measures such as traffic conflict techniques (TCT's) to estimate accident risk. In this investigation. traffic conflict techniques were used to estimate accident risk. There are four basic traffic conflict concepts and the development of these techniques was based on the accident process. The aim of this investigation was to highlight the differences between these concepts and to assess the applicability of these concepts to vehicle-pedestrian conflicts. The investigation was based on applying the various conflict techniques to data obtained at three intersections in the Durban CBD. In order to record the data an innovative method of using digital imaging was employed. This led to the development of a computer program to analyse conflict events. Analysis of the intersections based on the conflict techniques indicates that the intersections of Pine-Field and Commercial-Grey have a high probability of road users being involved in a "serious event" once there is an interaction between them. However, the probability for Commercial-Albert intersection is low thus indicating a safe intersection for vehicle-pedestrian interactions. The number of "serious events" at these locations was found to be related to the interacting traffic volumes - the conflict rate increases with increasing traffic volume. The use of conflict-volume models and accident models together with the conflict concepts agree that the accident risk is related to the conflicting traffic volumes and speed of the road users.

Thesis (M.Sc.Eng.)-University of Natal, Durban, 2003.