Quantitative assessment methods for railroad preempted signalized intersections

Thomas M Brennan, Purdue University


On October 25, 1995, seven fatalities occurred during a train collision with a school bus crossing railroad tracks adjacent to a signalized intersection in Fox River Grove, IL. In 1996 the National Transportation Safety Board (NTSB): Highway/Railroad Accident Report # PB96-916202 made several recommendations to the the United States Secretary of Transportation, one of which was “… the use and maintenance of railroad and highway traffic signal recording devices…….to readily determine that the highway signals and railroad-activated warning devices are coordinated and operating properly.” To date, this particular recommendation had not been addressed.^ This dissertation proposes (1) screening and prioritizing procedure for signalized intersection adjacent to highway-railroad crossings, (2) performance measures to evaluate track clearing effectiveness, and (3) direct measurement of highway-railroad synchronization operation. The proposed screening and prioritization procedures were applied to 595 state maintained intersections, which identified 20 intersections for future field investigation. Changes in track clearing effectiveness were used to evaluate 2802 preempt events to compare a pre/post steerable signal head installation that was installed based upon the Manual of Uniform Traffic Control Devices (MUTCD) Section 8D.07 guidance. This analysis concluded there was no impact on the effectiveness of track clearance based on the signal head change. In regards to the NTSB recommendations, this dissertation defined a procedure for assessing railroad synchronization applied it to over 7,648 preempt events recorded on a railroad interconnected traffic signal controller. Improvements to the synchronization and transition were made based the incremental analysis of performance measures. As a result of this analysis, the fixed track clear green phase was first extended by 5 seconds, followed by a subsequent refinement implementing a variable track clear green phase that terminated when the gate was down. As a result of those improvements the frequency of gates going down after the end of track clear green was reduced from 42% to 10% to 0% respectively. The effectiveness of track clear green based on post track vehicle presence was also improved reducing vehicle presence from 33 to 3 to 0 vehicles respectively. The dissertation concludes with specific recommendations regarding performance measures, new traffic signal timing techniques, and potential geometric improvements that can be used to reduce train/vehicle conflicts.^




Darcy M. Bullock, Purdue University.

Subject Area


Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server