Safety Evaluation of Signalized Restricted Crossing U-Turn Intersections

Chapter 2. Literature Review

Kramer first developed RCUT intersections in the mid-1980s.⁽¹⁾ A similar concept, called a J-turn intersection, was developed independently in Maryland in the late 1980s. Since the development of the concept, there has been a steady stream of published literature on the design. Recently, FHWA commissioned an exhaustive review of this literature for its Restricted Crossing U-Turn Intersection Informational Guide.⁽¹⁾ The Guide summarized the literature in nine chapters, including the following:

• Multimodal considerations.
• Safety.
• Operational characteristics.
• Operational analysis.
• Geometric design.
• Traffic control devices.
• Construction and maintenance.

The Guide posits that, while there is still much unknown about RCUTs, significant knowledge has been gained in the past 25 years that highway agencies can use to design and operate RCUTs.

The Guide found no previous papers making substantial contributions on the safety of signalized RCUTs. However, and as previously noted, the Guide did review three papers that made substantial contributions to the knowledgebase on the topic of the safety of unsignalized RCUT intersections.⁽¹⁾ The first of these three major studies was completed in North Carolina in 2010.⁽¹⁾ The authors examined 13 rural sites where a two-lane minor road met a four-lane divided major road. Before the conversion to an RCUT, the intersection used two-way stop control. STOP signs controlled the RCUT crossovers. The authors used comparison groups to account for other changes from the before to the after periods and used an empirical Bayesian (EB) method to account for potential regression to the mean. Total crashes decreased between 27 and 74 percent, depending on the analysis method employed. The number of fatal and injury, angle, and left-turn crashes decreased substantially with conversion to the RCUT design, while the number of sideswipe, rear-end, and other crashes decreased slightly or increased.

The second major study to examine the safety of unsignalized RCUTs was completed in Maryland in 2012.⁽³⁾ The nine sites were like those in the North Carolina study, except the U-turn crossovers and minor street right turns had merges, while the left-turn crossovers had YIELD signs. The authors reported results from several types of analyses, but they considered an EB analysis with a calibrated safety performance function that produced a 44-percent decrease in total crashes from the before to the after period to be the most reliable.

Edara et al. produced the third major study to examine the safety of unsignalized RCUT intersections.⁽⁴⁾ The five rural sites had STOP signs on the minor street approaches and crossovers before and after conversion to the RCUT design. The EB analysis procedure accounted for regression-to-the-mean bias using the crash prediction model from the Highway Safety Manual. The results from the Missouri study showed a 35-percent reduction in total crashes and a 54-percent reduction in injury crashes. Similar to the research conducted in North Carolina, researchers in Missouri observed a large reduction in angle crashes with RCUT installation.