Deploying bus bypasses in dense, congested cities: Practical findings from simulations and abstractions

Abstract

Simulations of idealized urban arterials reveal how a bus bypass, or queue jump, can worsen travel conditions for cars and buses. Problems occur when bypass operations affect a critical bottleneck, where queues form and grow long. To expedite bus movements through an intersection, a bypass’ dedicated traffic signal periodically detains cars. This creates voids in car flow that propagate forward. When a bypass was placed at a bus stop immediately upstream of a critical bottleneck, the voids diminished bottleneck discharge flow. Cars therefore encountered higher delays. Buses were also penalized, albeit unintentionally, because they shared lanes with cars.

This damaging cross-modal effect was eliminated by replacing the bypass with a bus pullout, sans dedicated signal. Although buses were delayed when re-entering traffic, they benefited more from the pullout than from the bypass. The pullout was also better for buses than was a do-nothing bus-stop treatment. The latter had dwelling buses impede car flows, which triggered the cross-modal effect. These findings point to changes needed in the design standards used for bypasses.

The bypass performed better, as intended, when placed on a congested arterial link further upstream of the critical bottleneck. This time, the forward-moving voids in car flow were compressed upon colliding with residual queues downstream. The voids were thus eliminated before they could do damage by propagating through the critical bottleneck. Further simulations unveil other conditions for which a bypass is well suited. Simulations also indicate that present findings hold for wide ranging bus flows and congestion levels.