This question is related to both Synchro and SimTraffic, so I'm going to post it to both boards for the benefit of anyone searching for info later.

I have a 5-signal coordinated system that runs north-south along a two-lane street with occasional turning pockets. At two of the intersections I have north or southbound approaches where two lanes are provided but are not explicitly designated by the marking or signing for particular movements. Both are 2-phase signals with permitted left turns.

The left lane in each case operates as a through lane in the absence of left turning vehicles, which are a fairly minor movement (10-20 vph), and the right lane operates as a right-turn-only lane. However, when a vehicle is waiting to turn left, most through vehicles (500-600 vph) move to the right-side lane to cross the intersection, exiting on a wide, single lane in one case and on two lanes that soon merge to one lane in the other case. They can do this without much interference from right turning vehicles, which are about 50 vph.

In SimTraffic, I can simulate this behavior fairly well by assuming that both lanes are shared with the adjacent turn movement. Then, I can lower the positioning distances from their default values so that the high end of the range is just longer than the length of the accel lane after the signal. This way, only the most conservative drivers will not attempt to pass around a waiting left-turner, since this is what I've observed in the field. (It would be nice to be able to set override values for positioning distance on a link-by-link basis, but since there are few other opportunities for optional lane changes in this network, it works OK).

The problem that remains in SimTraffic is that with the lowered positioning distance, the through vehicles will now queue adjacent to each other on a red light, something that has not been observed in the field. In real life, drivers seem to perceive the left turning vehicles ahead as contributing more points toward the positioning advantage than the through vehicles stopped on red. Is there any way to have my cake and eat it too on this one?

In Synchro, the dilemma I face is with coming up with the best set of timing plans while accounting for this situation. I could model these approaches with a short left-turn only bay capable of storing just a single vehicle in addition to the dedicated through and right turn lanes, but that would underestimate the impact of the left-turners. Would an adjustment to the sat flow rate be appropriate here?

On the other hand, if for the sake of optimization I assume in Synchro that these approaches have a shared left turn lane and a right turn-only lane, then I overestimate the true impact of the left turners because they would always block all of the through traffic. If I use the two-shared lanes solution that I use for the SimTraffic model, would a change in the lane utilization factor be most appropriate to account for the non-simultaneous use of the two lanes by through vehicles?

Any ideas you might have would be appreciated, or if this just serves as fodder for future software improvements, that would be great, too.

Hi Josh,

You asked: "Is there any way to have my cake and eat it too on this one?"

You have handled this correctly. However, this behavior is global so no way to really get it all (sorry). We are looking at these types of issues for future models.

Regarding the Synchro analysis, you ask about the saturation flow. I would say that yes, an adjustment to the saturation flow would be in order. You might still treat this as a single lane group. As you suggest, you'd want to make an adjustment to the lane utilization factor to account for the unbalanced flow. Synchro, like the HCM, handles this type of macro analysis with lane groups.

Thanks for the comments,

John.

John Albeck
Trafficware