This paper discusses the results of a literature review of studies relating intersection geometry to accident rates. It also discusses a research project currently being undertaken at the Queensland University of Technology to determine the effect of unsignalised intersection geometry, traffic volumes and other parameters on the rates and various types of accidents occurring for Australian conditions. The literature review revealed that most studies have found traffic flow to be by far the most important independent variable, with geometric parameters having only a small effect. The majority of studies have identified that the minor road flow affects accident rates more than the major road flow. Several statistically significant geometric and other parameters have been found across the various studies. However, there is little consistency between the results of these studies. If a particular study identified an important parameter (other than traffic volume), then this parameter often was not considered by other studies, was not found to be important, or was found to have the opposite effect. The only results found consistent across two or more independent studies are, that T-intersections are safer than cross-intersections, lit intersections are safer than unlit intersections, and larger stop signs on the minor legs result in a lower accident rate. .This paper also discusses the proposed approach to be undertaken for the research project, given consideration of the results of the literature review. (a).
Abstract