Guidelines for Identification of Hazardous Highway Curves

Abstract

Curve-warning treatment can be extremely cost-effective because of the low cost of treatments, such as curve warning signs or markings, and the potentially large number of target crashes. However, measures are needed to identify and prioritize treatment of potentially hazardous rural curves. An empirical Bayes-based procedure is presented for prioritizing potential treatment sites on the basis of crashes that may be classified as occurring because of the presence of curves. The attractiveness of the procedure is enhanced by the fact that the data and calculations are also a part of the evaluation of treatment that may be applied to sites identified. Alternative levels of the procedure can be selected, depending on the data available. Ontario data were used to calibrate supporting models. It is hoped that experience gained from using the guidelines can lead to the development of procedures that can be incorporated into the U.S. Department of Transportation’s Manual on Uniform Traffic Control Devices.