Estimating Queue Lengths at Signalized Intersections Near Highway-Railroad Grade Crossings

Abstract

Traffic signals at intersections located in closed proximity to a roadway-rail grade crossing often benefit from coordination with nearby railroad signals in order to alleviate both safety and operational issues associated with such locations. Analysis of traffic queues at these locations is of primary importance for determining if the normal signal operation needs to be preempted for railroad operations. Such queuing analysis becomes even more critical when direct observations of traffic queues are not possible or where the assessment is needed for a future location. Underestimation of queue lengths may lead to an unsafe design while significantly overestimated queues may cause unnecessary traffic delays consequently leading to violations of the active traffic control devices at the crossing. This research evaluated and compared commonly used micro-simulation models (i.e. Sim-Traffic and VISSIM) for their adequacy in estimating the queue spill-back lengths from traffic signals in the proximity of railroad crossings. Several additional comparisons were made between the queue estimation from the simulation-based and other deterministic analytical methods including Highway Capacity Software, Synchro, and a Railroad Assessment Tool that was developed as a part of this research. These comparisons helped identify the differences and specific limitations of each method in estimating adequate queue lengths. Recommendations were developed based on each model’s ability to capture the impact of various significant traffic factors on queue estimation, while minimizing the risk of underestimated queues.