High-tensioned cable barriers have been introduced in the United States as a median barrier option, primarily as a result of state agencies looking for a cost-effective means by which to mitigate or eliminate cross-median crashes. This paper studies the impact of cable barriers on crash frequency and severity in a way that can be utilized in the benefit-cost estimation for new cable barrier installations where other cost components have to be included, such as right-of-way and bridge structure widening. A before-and-after analysis with negative binomial and logistic regression was applied to data obtained from eight states to investigate the impact of cable barriers on the frequency and severity of crashes on a roadway segment. Three types of crashes were investigated for both the frequency and severity of crashes: single vehicle, multiple vehicle—same direction, and multiple vehicle—opposite direction (also called cross-median or crossover crashes). These three crash categories were considered separately with the assumption, confirmed with the results, that they are all affected by the presence of a median barrier. The result indicates that an installation of cable barriers, both low- and high-tensioned, eliminates 94% of multiple vehicle—opposite direction crashes. The before-and-after analysis showed that installing cable barriers results in an increase in reported single-vehicle crashes on wide, depressed medians by about 70%. In contrast, on narrow, depressed medians, the installation of a low-tensioned cable barrier essentially results in no change in safety. For multiple vehicle—same direction crashes, there is a weak, if any, effect on safety. Considering severity, the proportion of severe crashes decreases by more than 8% when adding a lane in each direction and installing a low-tensioned cable barrier on a wide, depressed median. The reduction in the proportion of severe crashes is less than 6% when the width is reduced in addition to the widening of the roadway and installation of a low-tensioned cable barrier. The narrower median is also the only median treatment that shows a considerable change in the proportion of crashes in response to an increase in average annual daily traffic. Finally, the effect of installing high-tensioned cable barriers on severity measured with the proportion of injury and fatal crashes is small if not negligible. (Author/publisher)
Abstract