News   Apr 19, 2024
 669     0 
News   Apr 19, 2024
 635     2 
News   Apr 19, 2024
 1K     3 

Improvements in Median Transit Design

reaperexpress

Senior Member
Member Bio
Joined
Nov 20, 2009
Messages
3,265
Reaction score
6,458
Location
National Capital Région de la Capitale Nationale
There is general agreement that the St. Clair and Spadina "LRT" lines are not shining examples of LRT design. They are notoriously slow, largely due to poor transit priority and inherently obstructive vehicular left turns.

We were told that Transit City would be a large improvement upon that design, but personally, I don't see it. Based on the following animation from VivaNext, the future rapidways don't offer any improvements either:

[video=youtube;ko4s0JpQPRM]http://www.youtube.com/watch?v=ko4s0JpQPRM[/video]

Given that so much of the planned transit in the GTA is situated in the median of roads, it is imperative that we improve on the design used on Spadina and St. Clair, rather than rehashing it in larger iterations.

I've heard numerous suggestions to improve the design of median LRT or BRT, ranging from modified light cycles all the way to grade separation. This is the thread for those ideas.
 
One thought I had is: Since left-turning cars conflict with median transit, why not move the conflict point away from the intersection where we can have a simple intersection with 100% transit priority?

Here's a diagram of a possible layout:

attachment.php


To turn left, cars cross the ROW at a traffic light which will always be green unless there is a transit vehicle. It's basically a level crossing at an angle. Once the cars have reached the left side of the ROW, they are free to turn left whenever there is a gap, just like at any other intersection, and unlike on St. Clair, where they must wait for the light cycle.

Main Advantages:
- It vastly increases the percentage of a light cycle that is green for transit.
- Because it inherently has nearside stops, it would additionally speed up transit that doesn't have transit priority by only stopping once at each intersection.
- It would reduce the average amount of time cars wait to turn left, by allowing left turns whenever there is room.

The main disadvantages are that it takes up a lot of space, and that the nearside stops make transit priority a bit more difficult.

Here is a diagram illustrating the 3 possible phases for the through street. Phases for the perpendicular street are not shown, since they would be unchanged. "1 phase left" is the current LRT design (think St. Clair, VivaNext or Sheppard). "2 phase left" is the idea with staged lefts.

attachment.php


Black lines: Vehicular traffic can proceed.
Grey lines: Vehicular traffic can proceed when it's safe to do so.
Red lines: Transit can proceed.

The only place I can think of where this would be possible on our current streetcar network is on The Queensway at Windermere and Ellis Avenues. Maybe with this road layout the TTC would remove the 7km/h speed limit currently in those intersections.

Any thoughts, improvements, concerns or criticisms of this idea?
 
Last edited:
Of course some drivers will start crossing the ROW before there is space in the turn lane for their car, and be forced to sit in the ROW blocking oncoming trains. This applies for people not paying attention, people who don't know how to drive properly, and jerks.
 
Of course some drivers will start crossing the ROW before there is space in the turn lane for their car, and be forced to sit in the ROW blocking oncoming trains. This applies for people not paying attention, people who don't know how to drive properly, and jerks.

True. I'd thought of that, but I think it should be OK as long as there is a long left turn lane, to minimize the occurance of such an event.
 
This would never work in Toronto. Too many selfish drivers that would simply block up the ROW while waiting to make their left turn in this configuration. The best approach is a simple one: When a bus approaches an intersection, right before the left turn signal, let the bus through via a short maybe 5 second green light for the bus. Then continue on with left turn cycle and general green. The city is too afraid to experiment with such things because engineers say it will delay left turning cars. In fact, it is the cars that are delaying the buses.
 
This would never work in Toronto. Too many selfish drivers that would simply block up the ROW while waiting to make their left turn in this configuration. The best approach is a simple one: When a bus approaches an intersection, right before the left turn signal, let the bus through via a short maybe 5 second green light for the bus. Then continue on with left turn cycle and general green. The city is too afraid to experiment with such things because engineers say it will delay left turning cars. In fact, it is the cars that are delaying the buses.

I admit that it does pose a risk to selfish and/or stupid drivers, but I expect that people will eventually learn that if they stop on the tracks, a transit vehicle will come and sit there honking at them. If that is not sufficient, it is quite simple to hand out hefty fines for anyone blocking the ROW.

The idea of having a bus-only phase is interesting, though it would inevitably add to the total light cycle, which makes it more likely that the bus will get a red light in the first place. As well, that would decrease the amount of green time for the cross street (which could be a concern if there is surface transit on it).

While I do appreciate simple solutions, I must ask: Is this design really so much more complicated than the default LRT design that it would be rejected? Which part is so complicated?

I don't really get what you're saying about left turning cars. Are you referring to the road design or your light idea? This road design benefits left turning cars as well as transit, so it's exactly the type of thing that the roads department might implement.
 
Last edited:
Do Spadina and St Clair even have signal priority?

And what's the point of designing median ROWs for cars instead of for transit with signal priority? Far-side stops are the best for transit riders, and when you build a major transit project you should build it for riders, not for drivers.

If left-turns obstruct median transit, then get rid of the left turns, it's as simple as that.

The real question is why so much planned transit is on the street. The Eglinton LRT for example should not be on-street at all.
 
Do Spadina and St Clair even have signal priority?

The lines have all the infrastructure for transit priority, but the roads department refuses to turn it on.

And what's the point of designing median ROWs for cars instead of for transit with signal priority? Far-side stops are the best for transit riders, and when you build a major transit project you should build it for riders, not for drivers.

This is not designed for cars, it's designed for transit. The benefits to left-turning drivers are simply side effects, but decreasing the wait time for them allows us to reduce the length of the left-turn phase. That benefits everyone equally, since during left turn phases, no one can cross the intersection, whether they be transit, cyclists, pedestrians or cars.

Obviously if you want good surface transit you use transit priority. This design doesn't change that. However, since transit can proceed in 2 out of 4 phases (as opposed to 1 out of 4 in the current design), it reduces the probability that transit will get a red light (it is very difficult to guarantee a green light for on-street transit, even with transit priority).

Would you care to elaborate on why far-side stops are best for transit riders? Because in my opinion, near-side stops are best for transit riders.
The reason near-side stops are more difficult for transit priority is that the signals don't know the length of time it will take the transit vehicle get to the intersection, so it must hold the green until the transit vehicle has crossed it. This holds up cars on the cross-street, but you don't seem to mind holding up cars, so I doubt that's a downside in your eyes.
Nearside stops have the additional advantage that when transit gets a red light, it can put that time to use loading passengers, instead of just sitting there doing nothing.
You may think that with transit priority, transit should get never get a red light, but from what I can tell, it is difficult to do that on median-based transit. This is because it takes a while before the lights can change from red to green, due to long pedestrian countdown and many required light phases.
If you take a look at the transit priority algorithm used by the TTC (page 12 of this presentation), all it does is reduce the wait for a green, and extend the green. It doesn't eliminate reds entirely.

If left-turns obstruct median transit, then get rid of the left turns, it's as simple as that.

Quite frankly, that's what I would like us to do too, but unfortunately if we tried to do that, people would make such a fuss that any project proposing that would likely get canceled outright.

The real question is why so much planned transit is on the street. The Eglinton LRT for example should not be on-street at all.

Absolutely. Off-street transit is almost always faster, even at grade, because intersections are far simpler so transit priority is more effective. However, it is unrealistic to expect all transit to be off-street. That would mean canceling most of Transit City, all of VivaNext, and most of the Mississauga LRT.
 
Last edited:
Jughandles and signalized Michigan Lefts would work where there's room, but in such event, off street running (with railway gates like in Edmonton) is even better for places like Eglinton in the Richview corridor. Phoenix has "service lane" style left turn offsets on Washington and Adams along the LRT but they can work only as they are one-way streets. The only jughandle left turn I can think of for regular traffic in Toronto is eastbound Kingston Road at Midland.
 
Last edited:
Transit City plans for Eglinton along the Richview corridor called for some right turn, left turn lanes, where cars would exit Eglinton through a lane on the right side, that would curve and provide an opportunity to make a left onto the cross-road. This would eliminate the need for left turns at these intersections.
 
Transit City plans for Eglinton along the Richview corridor called for some right turn, left turn lanes, where cars would exit Eglinton through a lane on the right side, that would curve and provide an opportunity to make a left onto the cross-road. This would eliminate the need for left turns at these intersections.

Yes. I think that various left turn schemes were proposed on Eglinton. At some point (such as pages 17/18 of this presentation), U-turns were going to be on cross streets, but later (such as page 15 of this presentation) they moved the U-turn locations onto the median of Eglinton to increase green time for transit and pedestrians.
 
Last edited:
Would you care to elaborate on why far-side stops are best for transit riders? Because in my opinion, near-side stops are best for transit riders.
The reason near-side stops are more difficult for transit priority is that the signals don't know the length of time it will take the transit vehicle get to the intersection, so it must hold the green until the transit vehicle has crossed it. This holds up cars on the cross-street, but you don't seem to mind holding up cars, so I doubt that's a downside in your eyes.
Nearside stops have the additional advantage that when transit gets a red light, it can put that time to use loading passengers, instead of just sitting there doing nothing.

I think that with a certain detection / control system in place, far-side stops can be better for transit. If you can detect a transit vehicle that has just passed the previous traffic light, you can then decide to either hold the green longer (so that the transit vehicle passes during the current cycle), or actually hold the red longer (so it turns green just before the transit vehicle arrives). In theory, this could completely exclude any red / waiting time for transit.
 
This would never work in Toronto. Too many selfish drivers that would simply block up the ROW while waiting to make their left turn in this configuration. The best approach is a simple one: When a bus approaches an intersection, right before the left turn signal, let the bus through via a short maybe 5 second green light for the bus. Then continue on with left turn cycle and general green. The city is too afraid to experiment with such things because engineers say it will delay left turning cars. In fact, it is the cars that are delaying the buses.

I'd go for this scheme. A significant practical advantage of it is that no capital investment is needed (except maybe to upgrade the software that controls traffic lights).
 
Do Spadina and St Clair even have signal priority?

The lines have all the infrastructure for transit priority, but the roads department refuses to turn it on.

Not right. St. Clair DOES have priority and it is switched on. Just shows that signal priority does not specifically mean "guaranteed green light".

In fact, ALL streetcar routes have some form of signal priority. (Believe it or not!) The system on St, Clair is more intelligent that that on other lines.

Check out this (old-ish) presentation on TTC signal priority:
http://www.signalsystems.org.vt.edu/documents/July2004AnnualMeeting/attach/m8_Sinikas.pdf
 
Good lord. Did the VIVA people even watch that video? How can they think that video sells the idea? The bus waits for turning cars, then by the time it pulls up to the bus stop a whole lot of cars have already passed, and before the bus can even leave the stop all cars have passed it and the light cycle has restarted. I watch that video and I feel like heading out to a car dealership.
 

Back
Top