This study will develop crash modification factors/functions to provide information on the safety impacts of signal installation with and without turn lanes in two lane roads in rural and suburban areas. Crash modification factors/functions will be developed for two situations at intersections in two lane roads in rual and suburban areas: (1) Signal installation without left turn lanes, and (2) Signal installation with left turn lanes. Different methods including empirical Bayes before-after and cross-sectional regression models will be considered for the evaluation.
Data from 117 intersections on two lane roads in rural and suburban areas in North Carolina were used to determine the safety effect of signalization with and without left turn lanes. This was a before-after study that was conducted using the empirical Bayes method. Before signalization, all the 117 intersections were controlled by stop signs on the minor legs. As part of implementing the empirical Bayes method, safety performance functions were estimated using data from a reference group of minor road stop controlled intersections. Results have been provided for three and four leg intersections separately. Five types of crashes were investigated: total, injury and fatal, rear end, frontal impact (type 1), and frontal impact (type 2).
It is clear that the introduction of signals without the addition of left turn lanes resulted in a reduction in total crashes, injury and fatal crashes, and frontal impact crashes (both types), and an increase in rear end crashes. When left turn lanes were added, rear end crashes decreased as well. Injury and fatal crashes and rear end crashes benefited the most from the addition of left turn lanes. Overall, frontal impact crashes did not benefit from the addition of the left turn lanes. These results along with information about the cost of adding left and right turn lanes could be used by NCDOT to determine the locations where these turn lanes would be most cost-effective.
