Analysis of the effect of directional traffic volume and mix on road traffic crashes at three-legged unsignalized intersections

Abstract

Traffic volume is the most common factor associated with road traffic crash frequency. Use of aggregate volume measures may lead to biased prediction of crash frequencies. Disaggregate analysis that involve hourly volumes can better explain this effect. In low and middle-income countries traffic mix may vary with respect to direction of travel on a road and such variation in proportion leads to crashes and conflicts among crossing vehicles at unsignalized intersections. This paper aims to explore the above relationship in the context of three-leg unsignalized intersections in Malaysia. First exploratory analysis of geometric and traffic variables was performed and their relationship with the crash frequency and type of conflicts was examined. Then probability of crashes was calculated with respect to the percentage of motorcycles that moved in a certain direction and proportion between volume in each direction. Segregation of the data with respect to traffic direction and traffic mix indicated that sites with less percentage of motorcycles moving in the far side direction experienced more crashes as compared to sites with greater percentage of motorcycles moving in the far side direction, given the condition that the volume ratio between each direction was larger than one. Results showed that the chances of crashes and serious conflicts increased upto 25% and 67% respectively; as the percentage of motorcycles that moved in the far side direction decreased from 31%, subject to the condition that the ratio of volume between each direction remained greater than one. It was concluded that traffic mix and traffic volume together influence the crash frequency at unsignalized intersections.

Keywords

• Unsignalized intersection
• Volume
• Crash
• Conflict
• Conditional probability