The effects of guardrail placement at signalized railroad-highway at-grade crossings in Iowa

Cyr, Justin
Major Professor
Peter T. Savolainen
Chris M. Day
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Journal Issue
Civil, Construction, and Environmental Engineering

Railroad-highway at-grade crossings have historically been the subject of substantive research given the heightened risk of fatal and severe injury when traffic crashes occur at these locations. To improve safety, active warning devices, such as flashing lights and gates, have been installed to warn drivers of oncoming trains. Various studies have been conducted on the effectiveness of these devices, but research has been limited with respect to the occurrence of crashes involving these warning devices. The supports for such devices are generally rigid and non-breakaway, posing as a fixed-object obstacle for motorists. A few states, including Idaho, Iowa, and Washington, include guardrail designs to protect the signal. However, guidance is generally limited in this regard.

To that end, this study involves an in-service evaluation on the safety performance of signalized railroad crossings within Iowa. The study leverages data from 2007 to 2016 to examine how the number and severity of crashes varies depending upon whether guardrail is installed at these locations. Over 1800 signalized crossings were analyzed and results show that crashes are more frequent in the presence of guardrail; however, such crashes tend to result in reduced severity compared to collisions directly involving the roadside hardware (i.e. without a barrier). Simulation analyses using the Roadside Safety Analysis Program suggest that moving the railroad signal further from the roadway provided the best alternative in both the presence and absence of guardrail.

Improving the guardrail design and ensuring the proper construction could make this system more beneficial. Ultimately, few crossings include such a design in which only a limited crash sample is available to compare to a base-case scenario. Handbooks suggest using a crash cushion at these locations, but no designs were found and even fewer crossings nationwide provide such a barrier.