Throughput during peak-period traffic at signalized intersections with high left-turn demand commonly creates a bottleneck. Two intersection configurations designed to increase the capacity at locations without adding lanes are the dynamic reversible left-turn (DRLT) lane at diamond interchanges and contraflow left-turn (CLT) lane at signalized intersections. The DRLT design replaces back-to-back left-turn bays with reversible lanes that span the distance between interchange nodes. The direction of flow alternates within a signal cycle. Signals and changeable-message signs control the direction of flow. The DRLT configuration may result in drivers intending to make a left turn at the downstream ramp viewing opposing traffic in the lane they are to use in a subsequent signal phase. 

In a CLT intersection, a gap in the median allows drivers turning left to queue in a lane that is normally used by opposing through traffic. The CLT design allows for an additional left-turn lane without widening the roadway. Two experiments, conducted in a highway driving simulator, evaluated driver comprehension and lane usage in DRLT and CLT intersections compared to conventional intersections, which were also simulated. Data were obtained from 96 participants in each experiment. The results of the DRLT interchange experiment showed that up to 25 percent of participants made inappropriate lane changes when first encountering the DRLT lane and observing opposing traffic in their lane prior to receiving a green signal. 

Early signs of confusion dissipated in later trials, such that only 4 percent of participants made more than 1 inappropriate lane change in the DRLT interchange. Participants in the CLT-intersection experiment used the CLT lane more frequently after the first encounter, with probability of use rising to 58 percent during the second exposure when there was a long queue in the nonreversible left-turn lane. Lane changes in the CLT intersection were uncommon and occurred in only 5 percent of trials. Under some circumstances, however, acceleration- and brake-release delays were longer in the CLT intersection than in the conventional intersection configuration. These delays did not improve during the second exposure. Findings show that signs and markings used in both studies will work well in a real-world environment. Recommendations are provided for additional and modified signs and markings on the approach sides of DRLT and CLT intersections to improve comprehension and comfort of drivers who may see opposing traffic in their lane while waiting for a green signal.

• DTFH61-13-D-01024\

• Operations

• FY 2002-2022 / Operations

• Intersections