Highway geometric design in mountainous areas has been a typical challenge. The combination of short horizontal curves and restricted right-of-way is a common ground for contemplating design exception in British Columbia, Canada. In practice, collision modification factors (CMFs) are advocated as quantitative measures of changes in road features on safety. However, in many situations, there are no CMFs in the literature to predict the safety impact of changing particular road features. An important example of these road features is sight distance restriction on horizontal curves. A mechanism for risk measurement has been proposed in earlier work to assist designers in comparing the safety impact of different deviations from sight distance requirements. This paper attempts to answer the questions as to whether it is possible to reduce overall risk and achieve consistency in such reduction without demanding wider right-of-way. This problem was formulated in a multi-objective optimization framework. Following this methodology, it was possible to achieve an average reduction in risk of 25% on the nine critical cross-sections. This reduction in risk was achieved without demanding wider right-of-way and without creating measurable increase in expected collision frequency due to independent re-dimensioning of different geometric elements. On theoretical grounds, this paper represents another step into the direction of developing fully probabilistic geometric design standards. On practical grounds, this paper provides an important decision mechanism that enables the efficient use of available right-of-way for new highway construction. Case studies in this paper have been applied on a major highway development in British Columba, Canada.

Additional Metadata
Keywords Cross-section elements, Highway geometric design, Optimization, Reliability analysis
Persistent URL dx.doi.org/10.1016/j.ssci.2012.02.001
Journal Safety Science
Ismail, K, & Sayed, T. (Tarek). (2012). Risk-optimal highway design: Methodology and case studies. Safety Science, 50(7), 1513–1521. doi:10.1016/j.ssci.2012.02.001