A Framework for Estimating Future Traffic Operation and Safety Performance of Restricted Crossing U-Turn (RCUT) Intersections

Abstract

Despite numerous studies demonstrating the effectiveness of Restricted Crossing U-Turn (RCUT) intersection design, its implementation remains uneven and close to zero in some large states, including California. This paper provides a comprehensive framework to estimate the operational and safety performance of future RCUT designs. The framework is demonstrated for a geometrically constrained intersection located on a high-speed rural expressway. The operational evaluation relies on microscopic simulation models of existing TWSC and alternate RCUT designs used to estimate network-wide performance measures. Two approaches are demonstrated for future safety estimation; first, an HSM-prescribed Empirical Bayes (EB) approach that uses Safety Performance Function (SPF) predictions combined with the crash history of the site. For typical applications, EB estimates may be combined with CMFs for RCUT found in the literature. This approach remains the preferred option for safety estimation. However, for geometrically constrained locations where atypical RCUT designs need to be evaluated, a surrogate measure-based approach that uses trajectory data from the simulation model is described. Surrogate measure-based safety analysis revelated that the RCUT design with no-left turn from mainline would be the most appropriate design for this location. The framework demonstrated here may be used by agencies to estimate the future benefits of the first-time application of treatments that have been successful elsewhere.