I would like to measure the length of queue that accumulates at an at grade railway crossing. I am wondering if I can do this by simply imputing the same amount of red time that a vehicle would experience at the given railway crossing into a regular intersections red time, then measuring the distance of the queue. Would this work ok? Is there a better way to model this?

sounds good to me. We've been wanting railroad crossing simulation (mostly in the form of signal pre-emtion) for a while now.