U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number:  FHWA-HRT-11-067    Date:  June 2012
Publication Number: FHWA-HRT-11-067
Date: June 2012


Field Evaluation of A Restricted Crossing U-Turn Intersection




Given the much lower volume of traffic emerging from North Franklin Road compared to that from U.S. 15 Business, the number and severity of conflicts at the conventional intersection suggests that the RCUT is a safer design. At the RCUT, only 2 low-severity conflicts were observed among 242 vehicles making a right-turn maneuver during 1 h of observation. At the conventional intersection, 2 low-severity conflicts were observed among 79 vehicles that turned right during nearly 7 h of observation. However, these right-turn conflicts are not directly comparable, as the RCUT was yield-controlled and provided an acceleration lane, whereas the conventional intersection was stop-controlled and lacked an acceleration lane, perhaps because the bridge immediately south of the intersection was too narrow to accommodate another lane. Neither the yield control nor acceleration lanes are necessary attributes of an RCUT design. The two conflicts with right-turning vehicles at the conventional intersection might have been avoided if an acceleration lane was available and used.

At the RCUT U-turn, 1 low-severity conflict was observed during observation of 42 vehicles over 5 h, whereas at the conventional intersection, 4 conflicts were observed among 36 left-turn and through movements. Furthermore, two of the conflicts at the conventional intersection that were judged to be of moderate severity could not have occurred at an RCUT where direct left turns from the minor road are not possible. Four of the conflicts at the conventional intersection involved vehicles making direct left turns or crossing movements from the minor road.

If traffic conflicts as defined here are predictive of crash rates, then the RCUT design, which eliminates direct left turns and through movements from the minor road, is safer than the conventional design.


In the case of right turns from the crossroad, the percentage of vehicles changing lanes to accommodate merging vehicles at the conventional intersection (13 percent) was about the same as that at the RCUT (15 percent). In the case of left turns from the crossroad, the percentage of vehicles changing lanes to accommodate merging vehicles was 14 percent, which was considerably higher than the 5 percent that changed lanes in response to vehicles merging from the RCUT U-turn. However, the frequency of lane changes was too small to allow tests of the significance of the differences.

Induced weaving appears to be similar at RCUT and conventional intersections. However, it should be noted that because all minor road traffic must turn right at an RCUT, the number of right-turn movements at the main intersection will increase, and thus the amount of induced weaving between the main intersection and the U-turn is likely be greater than at crossroads where direct left-turn and through movements are allowed.

Travel Time

Comparison of travel times suggest that through and left-turn movements take about a minute more at the RCUT than at the conventional intersection. Travel times at both intersections included time spent waiting for gaps in the through traffic. Thus, for the observed intersections, the provision of acceleration lanes at the RCUT did not compensate for the extra travel distance that was required. However, were the traffic volumes on the four-lane divided highway higher than those observed here, the wait times for an acceptable lag at the conventional intersection would likely have been longer. Thus, the travel time advantage of the conventional intersection would likely be less.

Acceleration Lanes

When traffic was present in the through lanes, drivers at the RCUT utilized the acceleration lanes for both right turns and U-turns most of the time. Eliminating U-turn acceleration lanes because they are not used in three out of four cases would either increase travel times when through traffic is present or cause the rate of conflicts with through vehicles to increase. From the limited observation made in this study, it appears that both right-turn and U-turn acceleration lanes are a valuable part of the RCUT design and should be implemented in future RCUT deployments.


Three approaches were used to estimate the effect of an RCUT conversion on crashes. All three approaches led to the same conclusion: the RCUT design reduces crashes. A simple 3-year before and 3-year after analysis suggested a 30 percent decrease in the average number of crashes per year. An analysis that adjusted the observed crash rate at RCUT locations for the observed crash rate at nearby conventional intersections on the same corridors suggested a 28 percent decrease in the average annual number of crashes. An EB analysis that adjusts for, among other things, the expected number of crashes at similar intersections and AADT suggested a 44 percent decrease in crashes.

Not only did the expected number of crashes decrease between 28 and 44 percent, but the crash data also suggest that the overall severity of crashes that occur is lower with the RCUT design than at conventional stop-controlled intersections. A 9 percent reduction in the proportion of crashes that result in injuries or fatality was observed.


The RCUT design should be considered for minor road intersections with four-lane divided highways where there is a sufficient volume on the minor road. The RCUT design greatly reduces the probability of angle crashes at the cost of a minimal increase in travel time. As volume on the divided highway increases, the travel time penalty is likely to decline and the safety benefit is likely to increase.


Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101