Traffic Safety: Warrant Modeling
Unsignalized intersections near railroad at-grade crossings were of concern in the profession due to safety concerns. NCHRP commissioned a study, which was done by the Transportation Program at what is now NYU Tandon School of Engineering, with KLD as the other member of the team.
At the time, no traffic signal warrant existed that would consider this situation on its own merit, and because of the locations of such intersections, none of the then- existing warrants were often triggered. Such intersections were often controlled by a stop sign on the minor approach, the one crossing the tracks. Yet situations shown in the sidebar existed with some regularity.
The team focused on the risk of fatal accidents in two ways: (a) fatality rates can be derived from the literature for signalized intersections that meet the volume-based warrants of the MUTCD, and (b) it can be estimated when the train-vehicle fatality rate exceeds this level near the contiguous unsignalized intersection. The case was made that a indicates an accepted risk level when one installs an intersection, and that there are many cases in which b exceeds this acceptable risk, so that it is logical to consider a signal installation to reduce the total level of fatalities.
The recommended approach essentially assigned the train-vehicle fatalities to the absence of the signal, and sought to reduce the risk to the level commonly accepted at signalized intersections.
This work was backed by field work to update the gap-acceptance behavior at such intersections and an extensive set of simulation runs that considered many combinations of volumes, traffic composition, resultant queues, distance between the intersection and the track, and probability of violators sitting on the tracks.
The work resulted in the addition of Traffic Signal Warrant 9 to the 2009 MUTCD, after extensive review by the NCHRP Panel, the NCUTCD, the FHWA reviewers for the MUTCD, and other peer reviews.