Empirical Evaluation of J-Turn Intersection Performance

Abstract

Most crashes on high-speed rural expressways occur at intersections with minor roads. Through and left-turn movements on minor roads are especially susceptible. One low-cost alternative for improving safety of at-grade intersections on such expressways is the J-turn. This study evaluated the effectiveness of unsignalized J-turn intersection design in Missouri with field studies, crash analysis, and traffic conflict analysis. Field studies used detailed video-derived data at a J-turn site and a control site. Crash analysis included an empirical Bayes before–after safety evaluation of five J-turn sites. The empirical Bayes analysis found that the J-turn produced a 31.2% reduction in crash frequency for all crashes and a 63.8% reduction in crash frequency for injury and fatal crashes. Annual disabling injury crashes decreased by 91.6%, and minor injury crashes decreased by 67.9%. Annual right-angle crashes decreased from 8.6 to 0.8, a 90.2% reduction. No left-turn, right-angle crashes were reported after installation of the J-turn at any study site. The average time-to-collision conflict measure was four times higher at the J-turn site as compared with the two-way stop-controlled (TWSC) site of minor road turning vehicles; this result indicated greater safety at the J-turn site. The average wait time at the J-turn site (5 s) was half that at the control site (11 s), and the average travel time at the J-turn site was approximately 1 min more than at the TWSC site. Several study findings are transferable to J-turn installations in other states.