Roadway departure crashes (RDCs) occur when a vehicle crosses a centerline, edgeline,or otherwise leaves the intended path of travel. The Virginia Department of Transportation (VDOT) has devoted significant efforts to reducing RDCs. As part of the effort to consider countermeasures, the current project identified best practices in low-cost delineation and data availability through a comprehensive literature review, VDOT district interviews of the current state of practice, and compilation of relevant crash modification factors (CMFs) and effectiveness information. Based on these results, treatments were selected and their effectiveness was evaluated based on Virginia crash sites. Additionally, a human factors study was performed to develop treatment effectiveness and configuration on the rural portion of the Virginia Smart Roads at the Virginia Tech Transportation Institute. The treatments evaluated included curve warning signs (CWS), edgelines, centerlines, post-mounted delineators (PMDs) at 20- and 40-foot spacing, and plastic inlaid markers (PIMs). Finally, the project team also developed a Microsoft Excel-based tool for evaluating benefit-cost ratios(BCRs) based on a range of input variables such as type of facility, annual average daily traffic (AADT), and treatment considerations.
The review of available CMFs showed that, in general, most of the low-cost delineation treatments have a potentially positive effect on reducing crashes, with a few studies showing some potential increase in crash counts for a few countermeasures. The aggregated CMF values suggest that delineators and signs have the highest potential for crash reductions, followed by rumble strips, pavement markings, and raised pavement markers (RPMs).
In the human factors study, participants overwhelmingly selected the PMDs as the most effective treatment. Participant ratings in the human factors study show statistically significant differences between PMDs and other treatments for curve visibility, perceived treatment effectiveness, perceived curve sharpness, and curve awareness. There was no statistical difference among ratings for CWS, edgelines, or PIMs. The two PMD spacings showed more improved performance indicators than the other treatments on many of the measured factors, but the difference was not typically statistically significant.
The benefit-cost analysis toolset developed as part of this project allows engineers to compute BCR not only for all crashes but also for specific crashes depending on the availability of CMFs. The benefit-cost analysis toolset shows significant benefits for all the treatments. As an example, using edgelines on two-lane rural roads with an AADT of 5,000 produces an estimated BCR of between6.8 and 12.1, assuming a CMF between 0.8 and 0.9. Similarly, a BCR is estimated between 26 and 67 for using PMDs on a 1,000-foot curve with just one RDC each year on average, assuming a CMF between 0.85 and 0.95.