A recent exchange on Twitter piqued my curiosity with the question “Is the Spadina car slower than it used to be”. A quick review of my archived tracking data for this route gave a simple answer “yes”, but there is more going on that just the speed of vehicles.
A related question dates back to a 2005 Globe article by Stephen Wickens comparing travel times on the 511 Bathurst streetcar which operates in mixed traffic to times for 510 Spadina which operates with “transit priority”. The Bathurst car won, much to the TTC’s displeasure.
Travel Times on Spadina: 2006 to 2021
Regular readers of this site will be familiar with the travel time comparisons I have published for many routes, most recently on RapidTO corridors. On parts of the streetcar network, I have been collecting data for many years, and this allows a review of route behaviour over an extended timeline.
The following charts show travel times between Bloor and Front Streets on Spadina subdivided by hour of the day at various times from late 2006 to mid 2021. The screenlines for these measurements are in the middle of intersections, and so a time “at King” is the time at which a streetcar crossed the intersection. Similar, “at Bloor” is the time a car passed under Bloor Street in the entryway to Spadina Station.
Each block of values contains monthly averages within one hour. For example, the leftmost block below contains data for the hour from 6 to 7 am. The bright red column within each block at January 2016 corresponds to the point where Flexity streetcars officially replaced the older CLRVs.
The bright yellow and bright pink columns (May 2015 and 2021 respectively) are months used for the CLRV vs Flexity comparison in the next section.
Because there is so much data, I have split each direction into two panels with data before 3 pm in the first, and after in the second. (Click to see an expanded version.)
Data for weekends are similar to weekdays. The charts are included in the full chart sets at the end of the article.
There is a consistent pattern over the years, although there is a more sustained rise in travel times northbound than southbound particularly later in the day and evening.
CLRV vs FLexity Speeds
January 2016 brought the schedule change where headways and running times were based on Flexity operation rather than CLRVs. Charts in this section compare operating speeds for mid-May 2021 with those for mid-May 2015 when the Flexity fleet was very small and almost all vehicles on Spadina were CLRVs.
The charts below are a selection of times through the day from the AM peak, midday, PM peak and mid-evening on weekdays. In each case the red line shows the May 11-22, 2015 data (mainly CLRVs) while the green line shows May 10-21, 2021 (Flexity) speeds. (For the full hour-by-hour charts, download the files linked at the end of the article.)
For southbound charts, the direction of travel is right to left. There is a regular sawtooth pattern corresponding to the location of stops. Speeds are high as cars accelerate away from a stop and then fall as they approach the next one. The effect of double stopping with nearside stops for signals and farside stops to serve passengers is clear.
The dotted lines show overall trends in values, but they are not moving averages (which would be very jagged due to the constant change in values). For the most part, the red 2015/CLRV speeds are higher than the green/Flexity speeds with a few exceptions.
A consistent characteristic of the CLRV data is that speeds are faster on departure from stops and particularly leaving Spadina Station and climbing the ramp up to Sussex. An important distinction here is that this could be due to inherently better acceleration and driving practices on the CLRVs, or to the pervasive slow orders now in place on all intersections with other streetcar routes. With most stops being farside, the better CLRV acceleration probably dominates this effect.
Note that these charts do not show dwell times.
On the northbound charts, the direction of travel is left to right. As with the southbound data, the CLRV trend lines (red) tend to be higher than for Flexity data (green), but not by as much.
Bathurst vs Spadina
This section compares weekday travel speeds for the period May 10-21, 2021 on routes 510 Spadina and 511 Bathurst. As in the section above, four hours are shown as examples and the full sets of hourly charts covering 6 am to midnight are linked at the end of the article. Both routes operate entirely with Flexitys, the only streetcars remaining in service, and so any differences are related to route characteristics, not to vehicle type.
In the charts below, the 511 Bathurst data are shown in green, while the 510 Spadina data are in red. In almost all cases, the Bathurst speeds are higher than on Spadina.
An important issue about minor stops is that if a streetcar does not have to drop out of the regular traffic flow, it will probably stay in the “green wave” even without transit priority at signals associated with those stops. On Bathurst this can make for speedier trips than on Spadina where farside stops guarantee that streetcars will fall out of the green wave. The absence of transit priority can then prevent streetcars from getting through the next intersection without being held.
The Bathurst cars have several advantages over those on Spadina:
- Bathurst stops are nearside, and when a streetcar crosses an intersection, it does not have to slow down again until the next stop. Many Spadina stops are farside, and streetcars have the double penalty of stopping twice and waiting for left turn traffic which has priority.
- Between Bloor and Harbord, both routes have one stop at Lennox/Bathurst and Sussex/Spadina. This can be a double stop for southbound Spadina cars, but a single stop for Bathurst cars.
- Although both streets have stops between Harbord and College at Willcocks and Ulster respectively, Spadina cars must operate more slowly around Spadina Crescent while Bathurst cars have a straight run. Also, Bathurst cars do not always stop at Ulster because of low demand.
- Between College and Queen, both streets have a stop between College and Dundas at Nassau, and they each have a stop between Dundas and Queen at Sullivan/Spadina and Robinson/Bathurst. Overall, the Bathurst cars win out, but not by as much as further north.
- At Adelaide Street, Spadina cars can be held by south-to-east left turning traffic that has priority over streetcars even though there is no streetcar stop here. Spadina cars no longer serve the Charlotte Street loop regularly, and most southbound streetcars are headed straight through the intersection.
- Bathurst is a narrower street than Spadina with much simpler intersections.
The direction of travel in the southbound charts is right to left. In most cases, except where noted, the legend on the charts refers to the intersection with Bathurst. King and Bloor Streets are used as references because these are at the same relative location crossing both streets. Other intersections are slightly offset due to the wandering nature of some streets notably Front, Richmond, Dundas and College.
Northbound charts are read from left to right, the direction of travel. Bathurst cars do well compared to Spadina cars in sections north of King, between Queen and Dundas, and in much of the distance from College to Bloor. Some of this is due to lighter demand at intermediate stops and some to intersection design.
Full Chart Sets
Travel times for 510 Spadina 2006 to 2021:
Speed comparison for 510 Spadina May 2015 vs May 2021:
Speed comparison for 510 Spadina vs 511 Bathurst May 2021:
I’m curious what you might do to improve service on Spadina. Is the problem structural, operational or both?
Steve: Nothing can be done about the acceleration characteristics of the Flexitys. As for the slow orders at intersections, TTC has reacted to a problem at specific locations by imposing system-wide rules that slow down streetcar operations. This will be very hard to change because “safety” is a motherhood issue. Streetcar systems all over the world operate through intersections at speed, but somehow they cannot do this in Toronto.
Getting true transit priority at signals will be like pulling teeth while moving cars takes precedence. One big problem inherent in the signal and stop spacing is that the advance notice a signal can receive that a streetcar is coming is fairly short in many cases. That said, I suspect there is a strong aspect of a collective shoulder shrug and an attitude that nothing can/will be done.
There will always be an advantage for the Bathurst car because it is a simpler street with less traffic overall.
I plan to do an updated review of 512 St. Clair as well. An important difference there is that in theory there are some transit priority signal locations, and double stopping should be less common.
Thanks Steve, great analysis and finally gives me data to justify the many times I’ve decided not to take the Spadina car because I felt it would be too slow. I had no idea Bathurst was actually faster however, I am not surprised. The state of Spadina should be an embarrassment to the TTC because it could be so much better given the advantages of ROW operation.
There are 2 simple improvements to Spadina I’ve always thought they should implement:
1. Straight through movements on Spadina should happen first and left-turning vehicles should come at the end of the cycle.
2. Minor street intersections should have a sensor right after the previous stop to detect approaching streetcars and ensure they get the green (prevent the light from changing).
The first is something I’ve seen done in suburban arterials in the US. Dedicated lefts sometimes come at the end of the cycle because conditions at the intersection warrant it. Since Spadina is mostly far-side stops this would speed the streetcar getting to the stop to service it.
The second just seems really simple to do since unlike at major streets where there are other streetcar routes fighting for priority, the minor streets are likely running on a sensor already and adding another along with a sensible intersection program would lead to fewer holdups.
A much bigger challenge is the Spadina station loop. The fact that only one Flexity can serve the platform at a time is a drastic capacity reduction from the days when 3 CLRVs could serve that platform. There needs to be some plan to expand it, though I realize moving walls or pillars isn’t trivial, it isn’t as extensive a change as at the Union loop. However, if that can’t be accommodated maybe they could modify the Flexities to have a door program that allows operators to only open 3 of the 4 doors and allow two cars to serve the platform, having one with diminished entry or egress capacity.
Steve: Spadina Station is an embarrassment. The TTC knows the platform can be expanded, but at some cost because of structural issues. As for the doors, “broken as designed”. The operator can open the front door independently of others, and the second door can operate independently for ramp deployment. But selective opening of each door set under operator control? No. Bad design, and who knows how much it would take to fix this. At the very least this is the kind of change that should go into the spec for the 60 additional cars, and be implemented as a retrofit during major overhauls for the rest of the fleet.
I didn’t include the loop in my analysis because I wanted to concentrate on on-street operations. But, yes, it’s as if every possible opportunity is taken to slow down the Spadina car and make service worse for riders.
The section between Front and Queen can be infuriatingly slow. The lights are clearly timed to get cars onto the Gardiner and the streetcars end up waiting for a long time at red lights.
Steve: Yes. It is a textbook example of what “transit priority” actually means in Toronto. And if we ask for better, the answer always is that if streetcars got more priority, traffic downtown would be snarled beyond belief with auto congestion that would … wait for it … slow down transit.
Forgive me if this has been addressed elsewhere. I was under the impression that city council is (still) waiting for a report about turning on “real and useful” signal priority for the Spadina streetcars. I believe it was a report requested by a councillor several years ago. Am I mistaken? Has there been any progress regarding this issue?
Steve: Like so much that Council might request, this has disappeared into the mists thanks both to covid and a general attitude that things cannot be fixed.
Steve, as a transit operator I can attest to the accuracy of your post. Spadina is a mess for many reasons, not just due to Flexity vs CLRV differences.
The biggest issue is pedestrians crossing the ROW, often without looking for oncoming vehicles, causing pedestrian-streetcar contacts. There have been too many injuries, including deaths, so a slow order has been imposed between Sullivan and Dundas on Spadina forcing streetcars to crawl along at 20kph between those stops. The simple solution would be to erect a fence to prevent pedestrians from crossing mid-block, but city councillors’ interference stopped that from happening.
Another issue is the pantographs on the Flexities. There have been numerous issues with panto-drops due to poor design, hence slow orders in and out of Spadina tunnel and under the Gardiner overpass.
Due to passengers interfering with door operations on the Flexities, it’s often difficult for speedy operation, especially on Spadina, which is particularly harder hit with this issue than other lines. People would hold the doors for other people who weren’t at the stop, and who were in no hurry to get to the vehicle, thus making everyone on board wait, further delaying service.
Diversions that can occur on other routes, like the 506, 505, 501, & 504, cause havoc as they utilise Spadina, because it has the ability to interconnect with other lines via ‘Grand Unions’, whereas other north-south lines do not. These diverting routes severely impact Spadina times not just due to increased Flexity traffic, but because switches that should be automatic are either OOS and set to manual operation, or because transit turn-priority signals are missing/non-operational, making it difficult for Flexities to turn into oncoming traffic.
The biggest difference as you’ve alluded to is that the shorter length of the CLRVs permitted double-platforming not just at Spadina station, but along the route as well. The length of the Flexities means that only one car can occupy a platform at a time, even at Bremner with its exceptionally long platform. If diverting cars could occupy the same platform as a Spadina car, this would permit passengers to transfer between lines more easily, as well as permit faster vehicle operation.
Spadina is a lost cause of poor transit planning, improperly designed infrastructure, badly designed equipment, and interference from the public and city council, all in one caldron. I’m sure sometime, somewhere, it’ll be used as a teachable moment in a classroom on how not to design public transit.
LikeLiked by 2 people
Something which may be missing from the data is the extremely long dwell times where passengers are trapped in the tunnel at Spadina waiting to exit during rush hour. Technically you’re at Spadina, but here are 3 bunched streetcars ahead of you and it can take up to 10 minutes to actually get to the platform. This is because of the much larger design of the new streetcars and the platform which was not expanded to accommodate them. I’m not sure if the TTC’s travel time data accounts for this.
The 510 is weird in that its travel time is highly sensitive to overcrowding, despite the dedicated ROW. At night, it races, but is genuinely approaching walking speed during rush hour due to the factors mentioned in this article and my comment. Makes for an extremely frustrating commute. I used to take the 510 daily pre-covid but walking over to Bathurst may not be the worst idea going forwards!
Steve: I will publish a follow up article on the time taken by cars queued at the terminal. I should mention that crew breaks/changes at Spadina Station don’t help either. This was a problem with shorter cars, but with the Flexitys the single car platform compounds the problem.
No streetcar coming, a “leading left turn – A sequence where the oncoming thru green is next after the green left turn arrow.” is fine.
Streetcar coming, a “lagging left turn – A sequence where the green left turn arrow is next after the oncoming thru green.” is fine.
Will the city actually use up-to-date technology to get true transit priority? Unlikely with the current crops of politicians and bureaucrats.
510 at a low of 8.5 km/hr to a high of 11.5 km/hr, shows the the whole thing is ridiculous. It’s not much faster than walking (5 – 6.5 km/hr). The point is that by avoiding reporting absolute speed over the distance of the whole route, the TTC avoids the embarrassment of just how awful the service is while being able to do nothing to improve the ability to get from A to B.
Problems with lagging left turns on Spadina would be caused by the fact that the left turn lanes are short and if you ended up with too many cars waiting and it would back into the left hand through lane and if there are cars turning right it would block the right hand through lane. This would cause the transportation gurus (read autophiles) to go crazy.
There is a discussion of this on the Canadian Public Transit Discussion Board, based on these observations. Some of the commenters may be actual TTC operators. The discussion starts down towards the end of the page.
It sounds like the root cause of the problems is a City Council with no interest in transit. Do *any* of the city council members make transit their top priority? Even one or two? Accomplishing anything might require an actual majority, but are there even a few?
Call it a crazy idea, but if the length of the Flexity’s is such a problem on the Spadina route, what about using shorter cars, about the length of the CLRV’s?
Likewise, using a Wheel-Trans type bus on low volume routes such as MAPLE LEAF?
You know, using the appropriately-sized vehicle for the job? Like, using real subway cars on the Ontario Line?
Just a fantasy of mine ….
Steve: Considering you would wait at least four years for a CLRV-sized car, assuming an existing design could be adapted easily, I’m not sure that is a solution. Also, you don’t buy a small fleet for one line. We’ve been down that route before.
WT vehicles on Maple Leaf? You are aware that pre-pandemic it carried 5,000 riders a day? That’s a LOT of WT buses and operators.
Hi Steve, I checked out Alstom and Skoda, and the shortest streetcars they make are 30m length vs. 15m length for a CLRV. Talk about suppliers not meeting what the market needs. However, Alstom might come through, they are “flexible”.
A 15m car does not have to be dedicated to one line only. There are other routes, such as KINGSTON RD or Lakeshore or night service where a smaller car is appropriate.
Steve: Both Kingston Road and Lakeshore cars supplement service on routes with full-sized vehicles, primarily King. That is not a place to send small cars. As for the night service, there is no point in changing off full-sized daytime cars for shorter night cars. The service will be half-hourly either way.
As to “coming through”, unless there is an off-the-shelf car the TTC could buy, the development cost for a small fleet would make the unit price very high
Regarding the MAPLE LEAF route, 5000 riders a day, that might be 50 riders on 100 buses? Neither number is realistic, as pre-pandemic there were an average 5 passengers per bus, including me, and I would wait 30+ minutes for the bus. No difference today.
Steve: The peak vehicle requirement for Maple Leaf is six buses. The higher riding count comes from the fact that 59 Maple Leaf is in effect a branch of 52 Lawrence West and many of those riders do not board on the unique segment of the 59. Those riders still have to be carried somehow.
I found this blast from the past regarding Skoda cars for the TTC.
Steve: The importance of keeping any new car order in Ontario (i.e. in Thunder Bay) was so strong that even though Skoda offered to bring a demonstrator to Toronto, this was rejected by the TTC. Who knows how we would have fared with another vendor’s car, or for that matter whether Bombardier would have built better vehicles if their management had not been so pre-occupied by an ultimately failed attempt to get into the aircraft market.
What, is the 59 MAPLE LEAF bus back in play??
I have not ridden the 59 bus lately (nor any bus, for that matter). However, I work at the corner of Ingram and Kincourt. When I see a 59 go by, it generally has more than five people aboard. up to a standing load. Note that the unique part of the 59 route passes through an industrial area with shift workers (similar to, say, the 123 Sherway), and then through a collection of midrise apartment buildings on Gulliver.
My experience, when I did ride the 59 westbound, is that yes, there wasn’t much ridership west of Jane. That doesn’t mean that the first section, from Lawrence and Benson to Lawrence and Culford, has buses running nearly empty.
Steve: It is in play in the mind of one of our readers who sees it as a gold mine of savings for the TTC.
Steve says that the 59 MAPLE LEAF might be seen as a gold mine of savings for the TTC (presumably by removing buses from the route).
Years ago, on another forum, there was a long discussion about changing around the 169 HUNTINGWOOD route. It was debated with such detail and fervor, it was as if changing this one route would make the TTC a shining beacon of efficiency combined with public service.
That didn’t quite work out, maybe 59’s the charm. 🙂
Steve: Over the years I have been amazed at how many people think that the “savings” from a bus here or there will actually make a huge difference to TTC finances. They talk about empty buses, but forget that the pre-pandemic service standard was that 10 boardings per vehicle hour was the minimum needed to justify a route. During the off peak period, the round-trip time on Maple Leaf is one hour, and therefore it meets the standard if only 10 people ride each bus at some point in its round trip. They will not all be on the buses at the same time.
This is a fundamental point about transit. Some portions, directions and times of day are very busy on some routes, and some are not. It would be easy to start nibbling away at the network, not just for “poor performing” routes, but also route segments/branches that were not seen as “pulling their weight”. It is, for example, more difficult for a bus route with strongly directional demand (e.g. outbound in the evening) to generate the sort of utilization numbers of a busy route with lots of ons-and-offs and bi-directional demand. But are the riders who use these services any less deserving than those who happen to use busy corridors not to mention portions of the subway network that are almost deserted at times?