On April 7 and 10 two major outages on Line 2 Bloor-Danforth were caused by hydraulic fluid spills from work cars. The TTC has sidelined its entire fleet of work cars pending inspection and repair. This has placed much subway infrastructure work on hold at a time when there is already a backlog.
A previous leak incident in 2024 led to TTC Board reports, and the findings were not impressive. Some work cars were beyond their useful lives, and there were ongoing issues with inspection and maintenance of equipment.
- Hydraulic Oil Spills on the Subway (May 2024)
- Subway Hydraulic Fluid Spills – Investigative Report (December 2024)
A common thread in these delays has been a car leaking fluid onto the rails over an extended distance before this problem is discovered. That leads to extensive, manual cleanup work to ensure that revenue service trains can accelerate and brake without slipping.
Problems with these cars are not the only source of major disruptions, and failing signal systems requiring service suspensions are common.
The TTC has not published a list of signal failure incidents to give a sense of their frequency and severity, let alone any plan to improve reliability over the decade before a new ATC system can take over.
It is rather comical that TTC Board meetings can include extensive discussion of a new enterprise asset management system, but little info on actual condition, maintenance and plans for what we have. We should not have to wait for a large IT project to know what the issues are today.
Behind both the fleet and signals problems lie bad capital budget planning by TTC management in response to political pressure to trim spending. In past years, key interrelated projects were downplayed or sidelined including:
- The need to refresh and expand the subway work car fleet
- The need to convert Line 2 to Automatic Train Control
- The need for a new Line 2 fleet
Under former CEO Rick Leary, in his early days, there was a sense that existing infrastructure and fleet could be stretched out to about 2040, fully ten years longer than the then-current target of 2030. In time, the ATC and new trains projects were restarted, but with much later delivery dates. According to the December 31, 2025 Major Projects Update Report, the delivery of 55 new trains will not complete until 2035, and the final cutover of ATC will not occur until 2037. Existing systems will have to last another decade.
Another factor is the timing of the Scarborough and North Yonge subway extensions which are planned to open in the early 2030s and will need new trains before the existing Line 2 fleet is retired. This is further complicated by demand projections showing the need for growth trains in the mid 2030s. Trains for new lines and extra service require production capacity that would otherwise go to a replacement fleet.
With ATC for Line 2 a decade away, the Scarborough extension will have to be built with conventional signals to tide it over until an ATC-capable fleet is running.
The TTC has not published a fleet plan showing how the various proposed deliveries of new trains will be staged. There is also the small matter of storage and maintenance space for the expanded fleets, and projects for new yards are not yet funded. The Province is happy to announce new subway lines, but conveniently omits the very large cost of the maintenance facilities.
This ties back to the work car reliability and fleet size issues because the volume of subway work will not decline, but will actually increase both thanks to aging infrastructure and system expansion. There are already severe scheduling problems for the work plans due to conflicting requirements for this fleet.
The focus at an upcoming TTC Board meeting will no doubt be on recent failures, but there are much larger issues affecting subway reliability for the coming decade and more. 2027 might seem a long time (and an election) away, but planning for that budget is already underway. The Board should demand a detailed review and plan to address the situation, and this should not be another “we’ll get back to you next year” report.
Steve Munro 
On Thursday I was on one of the first trains to get past Ossington that morning and I was shocked to see it was one of the newer trains you can walk the full length of. Not sure why it was on the line.
Steve: Likely borrowed from Line 1 due to inability to get trains from Greenwood yard.
LikeLiked by 1 person
Would
everywork car have a hydraulic system? If not, why ground them all?I am amused that according to the CPTDB wiki, RT-3 which is a crane car dates from the 1920s. Which implies it’s got Witt running gear.
Steve: If you look at the photo of RT-3 on the Transit Toronto website, you will see that it has subway trucks, not Witt trucks, and the body is clearly newer than a 20’s era Peter Witt. BTW what is your source that this is the culprit vehicle, and if so for only one or for both incident?
No, all work cars do not have hydraulic systems. This was covered in the 2024 reports on the subject. However, the issue is whether there is a generic problem with maintenance and pre-service inspection regardless of the technology on each car.
LikeLike
Maybe it’s just me but haven’t all these hydraulic leaks been in the same location?
Steve: No they have not, but a common factor is that the effects (trailing leaking fluid onto tracks) generally occurs as cars return to home base at Greenwood shops. The first incident was near Kipling and fouled rail east to Keele, The second was near Broadview and fouled rail from there to Greenwood Wye.
If you look at my article from December 2024, you will see a table of previous incidents and their locations.
LikeLike
Hopefully the delivery of 55 new trains WILL be completed BEFORE 2035, though. 2035 is excessive with a 2030 start date (deliveries of the previous order were completed faster than that despite delays in the process). If anything they should aim for an early 2030s completion of 55 trains and around 2035 for extensions (hopefully what ends up happening), not vice versa.
To be clear, a 100% new fleet is required before even partial ATC conversion can begin (which would predate the final cutover by probably a few years)?
Steve: That 2035 date looks suspicious to me, but as I said in the article I suspect that at least the YNSE and SSE trains are coming ahead of at least part of the order so that these extensions can open in the early 2030s. This would push the actual replacement trains back in the queue. ATC cannot be enabled while any of the T-1 trains remain in service because they cannot operate under ATC control and for various non-trivial reasons cannot be retrofitted.
The whole issue of a delivery schedule needs to be examined publicly. One might even suspect that it is artificially long to keep a minimal level of production at Thunder Bay active for as long as possible regardless of how this would affect TTC fleet planning.
Another issue I did not mention is that the TTC still entertains the option of installing a different vendor’s system on Line 2 from Line 1. A big effect here would be the work car fleet which requires ATC to operate across the network. These cars are not limited to one line like the TRs on Line 1. I think this is a foolish decision, and if the Line 2 contract goes to a system other than Alstom’s Urbalis, I expect they will run into issues with fleet and operational flexibilty. It is possible to have multiple systems on one network, but why introduce that complexity unless there is some fatal flaw in the Alstom system on Line 1? And if there is such a flaw, why haven’t we heard about it?
LikeLike
47 of 75 have hydraulics. Only 17 have an alarm when the fluid gets low. According to the Hatch Report.
According to this week’s capital report, it has been determined that 2 cars (27 and 40) are best to be replaced vs overhauled…
Steve: For readers the report is here and the reference is at page 10.
The Hatch report on the previous hydraulic failures unfortunately did not go into what each car uses hydraulics for…and whether the equipment is part of the vehicle design or a piece of equipment that could be replaced…
It also did not investigate whether there are alternatives to the existing hydraulic use cases. In many cases there likely is.
Electric motors, and batteries/super capacitors can get a lot more done than they used to. There may also be situations where hydraulics are still needed but there are safer designs, or the design uses less fluid.
If I was the board this is one significant avenue for investigation, especially with the vehicles at end-of-life.
LikeLike
Are their construction stages (especially YNSE) shaping up for an early 2030s opening, though?
Steve: What might happen given Metrolinx incompetence matters less than the fact that the political dates are soon after 2030. Any fleet plan that says “extension xxx won’t actually open until 2036” would not see the light of day, and planning has to be based on political dates. TTC has been saved from a fleet crisis only by the covid-era ridership loss and then the delays in opening dates for extensions.
Jheez, making it artificially long is the last thing we need after all the egregious delays that already happened. Here’s to hoping it can & will instead be accelerated as much as possible to be as close as possible to the original target deadline (pre-Leary). According to a CBC post, “the contract will likely provide work at the plant for at least 7–10 years”, so 7 years from now for 55 trains out of 70 sounds like a reasonable target.
Steve: Provided that we get the first 55 for T1 replacements. It’s worth noting that Kawasaki will deliver 378 new cars to New York City that were ordered in November 2025 in 2028 through 2030. This is possible because they have the production capacity.
Why are they still entertaining that idea, and reluctant to implement a uniform system across the network? And when do they expect to make the decision and award the ATC contract?
Steve: I suspect heavy lobbying by competitors to have an open bid coupled with TTC hoping it can get a better price through competition. How much they factor in extra costs of co-existence particularly in the long run remains to be seen. According to the Major Projects Update the award is expected in Q2 2026.
LikeLike
Why does it always seem like there’s no one at the helm?! Ah, I guess that should be, there’s no one driving the train? Why do you know all these things and the people in charge do not? THAT is a very good question!! Even when you report it, they don’t seem to get the memo. Now I know things that those in charge don’t. How crazy is that?!!!!
Thanks for the info, Steve!!!
LikeLike
Fair enough, as long as they’ll be able & willing to change fleet plans to accelerate replacement to the early 2030s if & when it becomes clear the extensions won’t open until later in the decade.
In that case, it’s unfortunate Kawasaki didn’t get this contract (though there might’ve been even more delays if CRRC or Hyundai got it instead). Based on the TR order, Thunder Bay certainly has the capacity to deliver the first cars 3–4 years after getting the contract, and deliver enough cars to replace an entire fleet in another 3–4 years.
Steve: It’s all about CanCon and keeping Thunder Bay working, if at less than full capacity.
Hopefully they’ll choose the same ATC for L2, that would also solve the problem of fleet planning by ensuring the SSE/YNSE trains are identical to, and thus interchangeable with, the 55 replacement trains (though at least the SSE trains would be identical to the replacement trains regardless).
LikeLike
I was not suggesting that RT-3 was the leaky culprit. I was just amused that i theory there’s a work car that dates back to the Witt era. Even if it is great-grandfather’s axe by now.
It does point to the age of the work cars. As the wiki lists, there’s a large contingent of cars from 2000 or earlier. Those would be 25 years or older by now.
Since they mostly go out overnight, I wonder if their maintenance is following the “out of sight, out of mind” principle.
LikeLike
It would be so fun to blame this on #DoFo and his conservative ilk, but it’s not any better in any other part of Canada. Worse, the same incompetent governance is also applied to (formerly universal) health care and education!
And if you think having the feds take any of this over recall that the Carney is in that same conservative ilk regardless of his party’s name.
LikeLike
If they really want to keep Thunder Bay busy as long as possible, perhaps they could exercise part of the add-on option order of up to 150 additional trains, as offered by Alstom. Though not sure why they even bother offering such a large option order when the TTC would never need that many, even half of that option order would be enough to allow early replacement of the entire TR fleet.
Steve: The RFP for the car order makes it clear that replacement of the TRs is an optional add-on. In effect, Alstom becomes the vendor of record for subway cars for the next few decades without a re-tender.
LikeLike
The detail about work cars leaking fluid for “an extended distance before this problem is discovered” really stood out—it’s wild that these critical vehicles aren’t fitted with basic sensors to alert operators instantly. At my old job in a warehouse, we had a similar issue with aging forklifts leaking hydraulic oil; management ignored it until a near-accident forced a full fleet audit. Feels like the TTC is heading for the same painful lesson with their backlog while the board debates software instead.
LikeLike
It’s striking how the work car leaks keep cropping up as a recurring failure mode — you’d think after the 2024 incident there’d be a faster inspection cycle. Living near the Bloor line, I’ve definitely noticed more unexplained delays this month, and it’s frustrating to think a hydraulic leak could cascade into hours of cleanup while signal issues pile up unrecorded. Makes you wonder what an enterprise asset management system will actually do if basic inspection records aren’t even being tracked today.
LikeLike