Integrated Stochastic Approach for Risk and Service Estimation: Passing Sight Distance Application

Abstract

The geometric design of highway elements is based on a set of standards that do not consider the risk levels associated with these designs. Safety elements including risk measures need to be identified and incorporated into the design process. The focus of this paper is a procedure that can be used by practitioners as a framework to conduct trade-off analyses between risk and service measures of specific design elements. The procedure builds on the randomness of parameters affecting the design attributes of any highway element. By accounting for the inherent variability of the determining parameters, random distributions of the design element can be derived. Then, using stochastic simulation, each of the design values can be tested to assess its risk measures. As a result, a risk index can then be attached to every design value in the random distribution. Finally, level of service (LOS) measures can be estimated and a trade-off analysis between LOS and safety could be conducted. The application focused on the design of the passing sight distance (PSD), which is an example of a highway element traditionally analyzed with no risk measures. Within the analysis, the risks and LOS measures associated with the use of the current PSD standards were assessed. The American Association of State Highways and Transportation Officials (AASHTO) Green Book values were shown to be conservative. The Manual on Uniform Traffic Control Devices (MUTCD) values fall within the acceptable risk and level of service (LOS) levels.