Safety and Operational Effectiveness of Protected Only Versus Protected/Permitted Left-Turn Signal Phase

Abstract

Roadway intersection fatalities account for around 24% of total traffic fatalities each year, with left-turning vehicles associated with most of these crashes. To manage left-turning movements, various left-turn signal phases like protected only (PO), protected/permitted left-turn (PPLT), and flashing yellow arrow (FYA) are currently in use. However, their overall effectiveness has not been well established. The paper evaluates the safety and operational effectiveness of PO over PPLT left-turn signal phase using sample intersections from Louisiana. A total of 151 signalized intersections with uniform left-turn signal phases at all of their approaches were selected. Crash data recorded over five years at those intersections comprising a total of 13,278 crashes, including 1,275 left-turn crashes, were collected. A crash modification factor (CMF) was developed using the negative binomial model (NBM) to assess the safety benefits of PO over PPLT. The Highway Capacity Manual 6th Edition (HCM 6) control delay field measurement technique was used for the operational analysis. The study first developed CMFs for total crashes and only left-turn crashes at various severity levels. CMF of PO over PPLT revealed that PO significantly reduced all severity levels of left-turn crashes by more than 50%, however, it was only able to reduce injury and fatal total crashes. Second, the study found average control delay at PO (50.69 s/vehicle) significantly higher than at PPLT (46.04 s/vehicle). The difference in delay was only significant during the morning peak hour. Overall, from both safety and operation perspectives, the results indicate that PO performs better than PPLT.