Have they bunched yet? An exploratory study of the impacts of bus bunching on dwell and running times

If transit agencies wish to retain and attract riders, they need to provide reliable and efficient services. Transit agencies tend to run high-frequency bus routes during peak hours, and in many cities, different routes can also overlap along major corridors. In some instances, consecutive buses can arrive at a shared stop simultaneously or one bus may arrive while another bus is currently servicing the stop. This phenomenon, known as bus bunching, can delay buses and passengers, and is usually inefficient. In this study, we attempt to understand how bus bunching from the same or different routes can impact bus operations, specifically dwell and running times. This research uses stop-level records obtained from automatic vehicle location (AVL) and automatic passenger counter (APC) systems from TriMet, Portland, OR. Using linear modeling, we find that bus bunching increases both dwell and running times. Specifically, when different routes bunch or are scheduled to arrive at a bus stop within a short time frame, or when buses from the same route arrive with a short time frame, dwell times increase by ~10 s. Similarly, bus bunching from the same route or different route prolongs running times by ~40 s. Our findings suggest that bus schedulers and operators should consider adding more time between consecutive buses from different routes at shared stops to minimize the negative impacts that we observed from bus bunching.