Toronto Eglinton Line 5 | ?m | ?s | Metrolinx | Arcadis

[rbt]

I'd think that the outcome of such comparison will vary greatly from one route to another.

No doubt but over a hundred routes some kind of correlation would be exposed (or not). Of course, correlation != causation, but it's hard to devise a test to find causation if you don't know what you're starting with.

Many of the examples of stops which should be eliminated are on routes with some of the highest ridership in the city; the same city with one of the highest per capita riderships in North America. It just feels wrong to assume those same routes are misconfigured by default as clearly something has been done right for the culture where those routes exist.

 

[crs1026]

In reality, given the number of lights and, and the increase in dwell times per STOP, the speed upgrades you're talking about would be lucky to increase average vehicle travel speed by a few km/h.

The math and logic behind this is solid, but I wonder if there is a qualitative factor that is being overlooked.

If you have a series of lightly used stops where each vehicle may or may not need to stop, you introduce variability in the trip time of each vehicle. The schedule has to be timed on the assumption that every vehicle will make every stop. But in practice the vehicle will blow by the lightly used stops some of the time....and then have to run slower to burn up the time saved.

By limiting the stops to main intersections where there are likely to be passengers waiting for each vehicle, the timing becomes more regular and reliable, because the duration of stops is consistent.

I make these comments having used the grade separated portion of the 501 for most of my life. In theory it ought to be a lRT quality zone, but it has always been characterised by extremes of hold-on-tight racing and casual lallygagging as some operators struggle to catch up with schedule while others struggle not to get ahead of their timing. This is operationally undesirable, but it's also hugely frustrating to passengers. And it builds a culture of plodding instead of zooming.

I don't have a position on what the stop spacing should be on Crosstown, but I believe that every LRT we build should exhibit the stop-rush-stop cadence of our subways. There are places where the character of a LRT changes en route (Boston's Green Line, for instance, changes mid-route from subway to surface street car) but this is not what we are trying to build on Eglinton. In transit, you don't sell the trip timing, you sell the zoom factor.

- Paul

 

[smallspy]

The math and logic behind this is solid, but I wonder if there is a qualitative factor that is being overlooked.

If you have a series of lightly used stops where each vehicle may or may not need to stop, you introduce variability in the trip time of each vehicle. The schedule has to be timed on the assumption that every vehicle will make every stop. But in practice the vehicle will blow by the lightly used stops some of the time....and then have to run slower to burn up the time saved.

By limiting the stops to main intersections where there are likely to be passengers waiting for each vehicle, the timing becomes more regular and reliable, because the duration of stops is consistent.

- Paul

If that was the case, then shouldn't he subway's running time not be variable? Since, you know, there are no intersections with lights to worry about, and it has to stop at every station.

But in fact, that is not the case. There is over a 10% difference in timing from the busiest (and therefore slowest) time of the day to the quietest (and therefore fastest).

There are far, far more factors at play than just traffic lights and the number of stops. The TTC knows this, and takes this into account - that's why virtually every transit route in the City runs faster at 10pm than it does at 8am - and that includes the subways.

Dan
Toronto, Ont.

 

[crs1026]

If that was the case, then shouldn't he subway's running time not be variable? Since, you know, there are no intersections with lights to worry about, and it has to stop at every station.

That's what makes it consistent. Every train makes every stop. Some stops may only be a few seconds at some times of day, but that's a lot less variability than not decelerating, stopping, and starting.

Line 2 Islington westbound to Kipling is no picnic at some times of day... it can be a long wait for one's train to make it to the platform at Kipling if things are bunched up ahead. But the subway train doesn't dog it all the way from Runnymede, the operators go full speed until they catch up with the ones ahead. I don't see that on 501..... travel time from Roncy to Humber was anyone's guess (and the new shuttle buses are proving much better).

- Paul

 

[TheTigerMaster]

I make these comments having used the grade separated portion of the 501 for most of my life. In theory it ought to be a lRT quality zone, but it has always been characterised by extremes of hold-on-tight racing and casual lallygagging as some operators struggle to catch up with schedule while others struggle not to get ahead of their timing. This is operationally undesirable, but it's also hugely frustrating to passengers. And it builds a culture of plodding instead of zooming.

The LRTs, I believe, will be running on headway-based scheduling. So I don't expect that we'll see situations where the trains will sit and wait for several minutes, while they try to catch up with their schedule.

Also, the subway also experiences delays for scheduling adjustments. For example, when the train sits in stations for a long time, or travels slowly between stations. When taking the subway on low-demand periods, such as early Saturday mornings, the train move noticably quicker than any other time of day.

 

[Hopkins123]

The Relief Line stop spacing along Queen is roughly the same as Line 2. We have stations on, University, Yonge, Sherbourne, Sumach, Broadview and Pape (6 Stations)

On the same stretch of Bloor, there are stations on Bay, Yonge, Sherbourne, Castle Frank, Broadview, Chester and Pape (7 Stations)

In particular, distance between Sherbourne and Yonge on Queen is exactly the same as the distance on Bloor.

The implication I'm trying to make here is that the stop spacing should be closer to reflect the greater population density in the core versus up by Bloor Street where density to the north is restricted by the natural barrier of the Rosedale Valley Ravine.

So it should also be 7 stations from University to Pape - University, Yonge, Jarvis, Parliament, Sumach, Broadview and Carlaw.

One station at Sherbourne to serve the whole area from Jarvis to Parliament is unreasonable, it should be two.

 

[Steve X]

The LRTs, I believe, will be running on headway-based scheduling. So I don't expect that we'll see situations where the trains will sit and wait for several minutes, while they try to catch up with their schedule.

Also, the subway also experiences delays for scheduling adjustments. For example, when the train sits in stations for a long time, or travels slowly between stations. When taking the subway on low-demand periods, such as early Saturday mornings, the train move noticably quicker than any other time of day.

In the tunnels, ATO will just do it's thing. Once east of Laird, the operator drives it like a streetcar. It's prone to the same problems as St Clair where lights, intersection blocked by cars, difference in acceleration/braking, operators playing games and etc. Transit priority may fail at an intersection adding to additional travel time along the line. TTC can still restrict intersection crossing to 10 km/h like the current streetcar ROWs.

 

[muller877]

In the tunnels, ATO will just do it's thing. Once east of Laird, the operator drives it like a streetcar. It's prone to the same problems as St Clair where lights, intersection blocked by cars, difference in acceleration/braking, operators playing games and etc. Transit priority may fail at an intersection adding to additional travel time along the line. TTC can still restrict intersection crossing to 10 km/h like the current streetcar ROWs.

Will they have an operator from Laird to Mt Dennis? Would save a bunch of money if the operator left the train and the computerized system took over.

With ATO central control could manage the speeds/stops to even out the headways as it travels in the tunnels.

 

[EastYorkTTCFan]

Will they have an operator from Laird to Mt Dennis? Would save a bunch of money if the operator left the train and the computerized system took over.

With ATO central control could manage the speeds/stops to even out the headways as it travels in the tunnels.

I think even with ATO they will still have an operator on board just incase something happens so they can override it.

 

[Monarch Butterfly]

Will they have an operator from Laird to Mt Dennis? Would save a bunch of money if the operator left the train and the computerized system took over.

With ATO central control could manage the speeds/stops to even out the headways as it travels in the tunnels.

The SRT is automated already. So between Laird and Mt. Dennis it'll be the same. The operator acts as a guard, or pair up with a another employee to act as a fare inspector.

 

[Steve X]

The SRT is automated already. So between Laird and Mt. Dennis it'll be the same. The operator acts as a guard, or pair up with a another employee to act as a fare inspector.

I highly doubt the operator/guard would be a fare inspector. The operator would sit in the cab all the time watching for potential hazards and be the person closing doors. Fare inspection is better off with 3-4 officer team in most of the day. Alternatively they could just check everyone's fare at Eglinton Station in rush hour. TTC said they will change it's bylaw to allow inspection on any part of the system.

Having an operator on board really helps when things go wrong. On the SkyTrains, sometimes Translink think the trains were block by debris till their employees arrive and realize the train actual ran over someone. TTC handles these situation much better than Translink does. In the last board meeting, some of the board members were question if it's absolutely necessary (by law) to have the emergency alarm on board. Andy Byford dodged the question but did say it have saved lives in the past and is a good thing to have despite all the delays from misuses. An operator sitting on board in the cab is absolutely necessary to be able to request for proper help.

 

[smallspy]

That's what makes it consistent. Every train makes every stop. Some stops may only be a few seconds at some times of day, but that's a lot less variability than not decelerating, stopping, and starting.

Line 2 Islington westbound to Kipling is no picnic at some times of day... it can be a long wait for one's train to make it to the platform at Kipling if things are bunched up ahead. But the subway train doesn't dog it all the way from Runnymede, the operators go full speed until they catch up with the ones ahead. I don't see that on 501..... travel time from Roncy to Humber was anyone's guess (and the new shuttle buses are proving much better).

- Paul

But it isn't consistant. Why does the B-D need 110 minutes to make a round trip in the middle of the day, but only 97 at night?

The signals haven't changed. There aren't fewer stations being served. The route isn't shorter.

And thus my point. The number of stops isn't the final bearing on how long a route will take to run. It does have a bearing, yes, but it's most certainly not the only one.

Dan
Toronto, Ont.

 

[crs1026]

But it isn't consistant. Why does the B-D need 110 minutes to make a round trip in the middle of the day, but only 97 at night?

The signals haven't changed. There aren't fewer stations being served. The route isn't shorter.

And thus my point. The number of stops isn't the final bearing on how long a route will take to run. It does have a bearing, yes, but it's most certainly not the only one.

Dan
Toronto, Ont.

I think we are saying the same thing, sort of. On the subway, the running time between stops doesn't change over the day. Once you know how many stops there are, and how far apart they are, you have a constant running time.

What changes is dwell time in stations. Trains take longer to load/unload in peak periods, yes. That accumulates over the length of the run, so you schedule longer running times at peak. However, on leaving the platform, a train takes exactly as long to reach the next station as they would at any other time of the day. (OK, if trains bunch up, they may run on yellow lights at slower speed, but even at rush hour the schedule tries to avoid this.)

I agree- If you assume that LRT always stops at every stop, then it's the same for LRT. The number of stops doesn't matter, dwell time is what influences trip time. I'm questioning if this is the scenario being designed. In the extremities, if we design with the intermediate local stops included, these stops may have no one boarding or requesting a stop. Some LRT's will blow right through the stop, just as the 501 frequently does at Colborne Park and South Kingsway. That results in a much more variable running time than if they are designed to stop everywhere.

You can schedule the same way as the subway - allow more time at peak, on the assumption that more LRT's will stop more often at those times, and for longer. But there is a limit to how often this will be true. You will have a greater number of cases where the LRT gets ahead of schedule.

If you limit the stops to the main intersections where there is heavy transfer business, then the actual operation most closely approaches subway conditions....the LRT stops at every stop. We're back to dwell time as the variable factor.

The issue is not "more" stops versus "less" stops or "longer" trips versus "shorter" trips - it's consistency. Variability will play havoc with traffic priority signalling or timed signalling. The solution is inevitably operators electing to coast....there is an art to streetcars coasting towards a stale green when the operator wants to hit the yellow. They use the forced stop to kill time through one light cycle.

The point I am beating to death is, build in whatever number of stops you want....but don't allow coasting to creep into the design. Low-use stops influence schedule which influences coasting.

- Paul

 

[am29]

tbh they should've just dug under the intersections in the above ground part so they avoid the light. Don't think it would've added that much to the project. Would probably save a lot of time and cause less traffic disruption.

 

[Zach6668]

tbh they should've just dug under the intersections in the above ground part so they avoid the light. Don't think it would've added that much to the project. Would probably save a lot of time and cause less traffic disruption.

There are currently 15 signalized intersections between Laird and Kennedy that you're suggesting we should have tunneled under. That would most certainly add a ton to the project. Utility relocation being a huge expense nobody ever thinks about, plus the actual act of tunneling, mining, or however you expect to do it, and somehow with *less* surface disruption? Not a chance.

Would it speed up trip times? Maybe. Probably vehicle trip times, but where are you putting the stops/stations? Under the intersections? So now you're adding the cost of elevators, escalators, stairs, all sorts of additional concrete, etc. No? You'll put them before the portals? Now you're adding a bunch of walking time, and maybe roundabout walking depending on how you expect users to get to the middle of the road stops. Signalized crosswalks? That'll slow auto-traffic down.

Can we just accept that you can't get across a large, bustling city at absurdly high speeds? Movement takes time. You could build a two-stop, 20-km tunnel to whisk you across the city at the highest speeds, but it'll be useful for a tiny amount of people. Stop spacing, running speeds, intersection treatments, grade separations, etc - these are all compromises. There's no clear "better way", it all just depends on your specific needs. Not everyone needs the same things as you, so compromises are made, and compromises are made when limited capital dollars are at play, as well.

Crosstown will be a useful transit line. You'll be able to get across the city in a reasonable time, and to dozens of points in between.