Subway Fleet Planning(?)

The TTC’s ongoing inability to correctly provide for its fleet requirements gets embarrassing at times.  A big one comes during consideration of capital budgets when there always seem to be brand new projects that appear out of thin air even though they should have been foreseen.

The questions of subway capacity and fleet size are not small change, something to be fixed up with the underspending in a few minor accounts.  If you get it wrong in one direction, nobody can fit on the trains, and the lead time to fix this condemns riders to horrendous service.  If you get it wrong in the other direction, you have a bloated fleet, an investment in trains and yard space that might have gone to more deserving projects, and a padding factor that works against efficiency in maintenance because there’s always another spare train in the yard. 

The TTC’s surrent subway fleet plans do not include cars for the subway extensions beyond Steeles Avenue, nor for the allegedly closer headways we require to accommodate all of the riding growth the Richmond Hill line will bring.  Cynics might be tempted to say that the whole question of subway capacity has been manipulated to the TTC’s budgetary advantage, but I have always taken the view that much that appears to be Machiavellian can be explained by bad planning.

This post and the much longer paper attached to it will examine TTC fleet planning and the degree to which we are not seeing the full story of options for future subway services.

Current Fleet Plans

The TTC plans to consolidate into two fleets by the middle of the next decade.  This will give us a fleet of T-1 cars for use on Bloor-Danforth and Sheppard, and a fleet of “Toronto Rockets” or “TR” cars for the Yonge line.  However, current orders for TR cars will only replace the existing “H” series fleets with a modest provision for expansion.  TTC’s current plans for additional service include:

  • Seven additional trains in 2010 for the current shortfall in service.
  • Three additional trains in 2010 to extend short-turn operations on the Spadina line to Glencairn.
  • Sixteen additional trains (7 for BD and 9 for YUS) over the period up to 2025 to accommodate growth at 1.35% per year.

The 2010 plans may be delayed until the fleet is actually big enough to support the additional services.

The ability of the BD line to actually handle so many additional trains is dubious because of limitations on turanaround times at Kennedy and Islington (roughly 2’10” minimum as discussed elsewhere in this blog).

The Yonge line will not be able to handle more frequent service until two projects are completed.

  • Automatic train control allowing for closer train spacing especially at stations like Bloor with long dwell times.
  • Provision of turnback points (now planned for Finch and Downsview) where half of the service can pull into pocket tracks and return south.  The VCC and Langstaff terminals would become much more complex if all service ran to the ends of the line, and demand is nowhere requiring this level of operation.

The Yonge line’s projected requirements could be offset by new and improved services such as GO service to Richmond Hill and the Downtown Relief Line.  Unfortunately, it is dubious whether a DRL would be in place (even if it were moved up in the Metrolinx priority list) in time for the Langstaff opening.  The TTC will be forced to provide very frequent service over much of the Yonge line even though the real congestion is only downtown.  This will keep the carbuilders happy, but it will concentrate both demand and the impact of any service problems on one line.

The current fleet plan and capital budget do not contain any provision for extra trains for improved headways on Yonge.  The budget for the two extensions does include a provision for additional cars, although this is buried inside a larger item and we don’t know whether this is adequate for our needs.

A bigger fleet needs a larger carhouse and more ongoing maintenance work.  This is only partly provided for in the budget.

The Spare Ratio

A big issue with the current subway fleet (not to mention the entire surface fleet) is the spare ratio, the number of vehicles in addition to scheduled service that are needed for ongoing maintenance and repairs.  Within my lifetime, spare ratios over ten percent were considered excessive in the transit industry, but modern vehicles are more complex and less reliable, and the common values have drifted up well into the 15% range or more.  (The streetcar fleet now stands at about 33%.)

The TTC expects its new fleet to reduce spare requirements through improved reliability.  Three cheers and all that, but I can’t help wondering how the situation was allowed to become so bad in the first place.  One way or another, the elimination of the last of the “H” trains will allow for a lower spare ratio once the entire fleet is made up of T-1 and TR cars.  This offsets but does not eliminate the pressure from riding growth and route extensions.

Capacity Options for the Yonge Line

There are various ways to increase capacity on the Yonge-University Subway.

  • The TR trainsets have about 10% more capacity because of the gangways between cars and the ability of passengers to distribute themselves through the trains.
  • Headways could be reduced from the current level of about 2’20” to roughly 1’50” (roughly a 27% improvement) through automatic train control.  Although 1’30” headways have been discussed, they are impractical given the actual route layout and station spacing.  The 1988 headway study (presented elsewhere on this blog in great detail) observed that below a 1’50” headway, scheduled speed actually goes down because the trains are too close together to move at track speed.
  • A seventh 50-foot-long car could be added to the TR sets.  This would add roughly 10% to the capacity of each train.

Station Capacity

Several of the Yonge line stations have constrained capacity for passenger movements from the platform to the mezzanine with current service levels, never mind an improvment by 40% (the combined effect of the first two bullets above).  Congestion is particularly bad at stations with on stair and one escalator when the escalator is under maintenance.

Although the TTC is working to add second exits at various stations, a concerted program would be required to bring critical Yonge line stations such as Wellesley, College and Dundas up to scratch.  There may also be fire code issues with University line stations where there is only one pathway to the surface through a common mezzanine level.

Bloor-Yonge and St. George Stations

These stations handle transfer traffic between the two subway lines and they are constrained for expansion of capacity by nearby underground structures and watercourses.  The situation at Bloor-Yonge was discussed in the 1988 study.

A big problem exists with the BD line during the PM peak because of limitations on the extra service that can be provided there.  If passengers arriving northbound at Bloor (or St. George) are 40% more numerous than today, the service on BD will not be able to take them away fast enough to clear the platform between trains.

Schemes to increase station capacity have been proposed, but I cannot help thinking that this is one more example of how subway planners compound the problem of overcommiting the capacity of one corridor rather than looking at alternatives.

A Few Concluding Thoughts

Subway planning and the need for additional capacity into the core are complex issues, and Toronto has been poorly served by years of jurisdictional isolation (GO, TTC) and underfunding (build pet subways like Sheppard while ignoring long term problems in areas of high demand).

Recently, Metrolinx has at least thrown some of the options for downtown-oriented capacity back into the hopper, but many of these are years away in the back end of the 15-year plan if not the 25-year map.  The problems, however, will face us within the next decade.

A much more extensive review of these issues is contained in Subway Fleet and Route Planning, a discussion paper I completed a few weeks ago.  Originally I had intended that it find its way to the staff of various agencies, but have decided to publish it here for maximum exposure.

18 thoughts on “Subway Fleet Planning(?)

  1. It only gets worse if fares get harmonized. I talked to Charles Wheeler about the impacts of fare integration on Yonge subway capacity, and he said that a report on that is on the to-do list for the TTC. However, I don’t think it is being treated with the urgency that it really needs, since fare integration is a high priority on Metrolinx’s agenda. The results should be predictable: Yonge can’t handle fare-integration with YRT/Viva. The Downtown Relief Line should be a prerequesite for fare-integration, and the TTC needs to put its foot down and demand the province put the money into that project before fare-integration is allowed to proceed. So long as the city isn’t on the hook anymore for funding capital expansion, there’s little excuse for the city to be hesitant about supporting investment in the DRL.


  2. The Yonge Extension and the Transit City LRT line will certainly put significant pressure on the downtown subway network. I know it has been debated before, but if they are as successful as they are hoped, it will result in more riders system-wide, but with many more headed downtown. There’s three new mid and highrise office towers well underway, and we’re not seeing the job dispersion that some downtown politicians were hoping for when compromising on suburban subways.

    I am glad to see that the utility of the DRL is increasingly recognized by many now, though it has been pushed back into the 25 year plan, with TC and Yonge North ready first. I think it’s almost backwards.

    The utility of the DRL is apparent. What I don’t get is why the Richmond Hill line is advanced as an “express rail” line before Stouffville, which would at least be able to take passengers coming from Scarborough and the north east off the Danforth Line, and runs in a better corridor. The Bala Sub is the most isolated and winding of all of the routes heading into downtown Toronto, and misses the points where it could pick up passengers in logical areas in the city. It would pick-off many passengers coming straight in from Richmond Hill and perhaps Thornhill (if there’s a stop at Bayview and John), but apart from Finch (Old Cummer), it isn’t that useful. Oriole could work if they moved the station north, but it misses Don Mills, it winds through the Don Valley (making stations at Eglinton, Flemingdon/Thorncliffe and Bloor very difficult) and is really suited for commuter, not transit, service. I can’t see it making enough of an impact on the Yonge Line’s crowding, especially if we maintain the fare disco-ordination status quo. There would probably need to be a station at Eglinton/Wynford, a difficult location to build.

    The DRL has the potential, especially in feeding passengers from the north at Pape (or beyond) via the Don Mills corridor, and intercepting passengers before they get to Bloor-Yonge.


  3. Many of these problems are the same problems faced by other cities. Older subway stations just cannot handle newer crowds.

    As for the university line and the single-exits, I find it sadly ironic that stations like St.Patrick and Queens Park had their second exists closed for safety reasons (I think this occurred after a number of rapes in the area) only to possibly have them re-opened for different safety reasons. Seems the options are either you are un-safe, or you are un-safe.

    Steve: Actually, those are not exits, only crossovers between the two platforms. Because they didn’t actually go anywhere, they were walled off for security reasons. The issue was a murder at St. Andrew Station which, oddly enough, does not have any such nooks and crannies. The “jail” at Museum was a result of the same event. The space had to be left open as it leads to an air shaft.


  4. The seventh car is interesting. I wonder how they intend to do that? From what I understand, the car can only be added once the Automatic Train Control is in place, or if the train enters the station at a reduced speed, so the driver can pull up precisely at the right spot. And wouldn’t adding a seventh car require modifications to a train-set? or maybe they can simply add it as a trailer with hook-ups?

    Does the TTC stable trains on the along the line, like Paris?

    Steve: The seventh car would probably be a trailer inserted in the middle of a six-car set with appropriate connections for electrical passthroughs of control wiring. Yes, ATC is a pre-requisite for precise stopping on the platform.

    The TTC does store some trains online to simplify startup at the beginning of service, but not many.


  5. The TTC is probably working on the assumption that all new YUS ridership will be going strictly north-south into York region and not making any transfer moves to or from the Bloor St. subway.

    It really doesn’t matter — the TTC is, and always has been, thoroughly incompetent at planning and running the subway system properly. They could have expanded St. George and Bloor-Yonge in the 1960s and 70s when there were no adjacent buildings in the way.

    There were other quickie / cheapie measures they could have examined … for example:

    They could have installed a crossover west of Bay Lower and alternately stubbed trains to that platform, instead of letting it rot to hell — a scheme that would have given us a 3rd transfer point and reduced platform crowding at St. George and B-Y.

    They could also have included a provision for extending the Spadina subway south of Bloor along Spadina Avenue for additional downtown capacity. Instead, they come up with a DRL that will probably never be built.

    Steve: To be fair to the TTC, the various plans for additional capacity into the core area were sabotaged by an alliance of suburban politicians who actually believed that places like “Downtown North York” could complete with the financial district, and by city politicians who were trying to strangle growth downtown by limiting transportation capacity. They didn’t figure in GO Transit which has provided the lion’s share of additional commuting traffic into the core, and any ideas of stopping growth are long behind us. Meanwhile the transit planning exercises focus on the suburbs. They deserve transit too, but pretending downtown office spaces and commuters don’t exist is pure folly.


  6. I’m not at all sure anyone should ever hold their breath for the DRL. I support it just as much as any interested party ever possibility could but when a project, no matter how worthy it is, gets put off for 25 years I find it questionable at best that it’ll ever get built.


  7. “The ability of the BD line to actually handle so many additional trains is dubious because of limitations on turanaround times at Kennedy and Islington (roughly 2′10″ minimum as discussed elsewhere in this blog).”

    Hmm, that should be Kipling.

    Having recently ridden gangwayed subways in Santiago and Buenos Aires, I’m not sure if the “self-redistributing passengers” notion will work well in practice–as soon as the crowd increases, everyone sticks where they are like limpets. Same as people not moving to the back of ALRVs.

    Steve: Ouch!! Considering that I was at Kipling this morning, I should have remembered where the end of the line actually is! I tend to concur also about the gangways from a distribution standpoint, but we should pick up some space from the elimination of the cabs and the elimination of the inter-car space.


  8. I plan to redistribute myself in gangway-cars, and there are a few others that will too. I cant go one day without seeing someone run between the cars while a train is at a stop. Even if only 10% of people do this, it means 10% less crowds at the “crowded end”


  9. The DRL, it only makes sense if a very large majority of jobs are in the Toronto core, but lots of companies are moving out of the core, Look at all the industrial and commercial building going on in places like Richmond Hill and Vaughan. The DRL probably made more sense 20 years ago, then it does now. Now I think extending the Sheppard to run from Markham road to Kipling would make more sense. This in fact makes more sense then another subway line to a traffic corner in Vaughan.

    This would mean that bus service from York Region could run into about 10 or 15 stations, instead of 2. Alter Transit city to operate Streetcars between the Bloor and Sheppard subways, although loops at Steeles and the Lakeshore probably make most sense. This would allow better balance between the Yonge and Spadina lines, meaning you get twice the utilization of each added train.

    The main problem for the TTC though, is that once a decade or two, they come up with a megaproject, rather then lots of little projects along the way., For example, if the plan for Sheppard had been to open it, then start work on extending it to Victoria Park, they would be opening that station now, and starting work on Warden.

    Steve: Notwithstanding new jobs in the 905, there is continued growth in demand into the core, but there isn’t room for it in the future on the Yonge subway.


  10. Also worth noting that the enormous residential and, for lack of a better word, ‘entertainment’ growth in the places that would be served by the DRL (especially if done along the rail corridor, as is sensible. Liberty, the Queen West triangle, the area around Bathurst and Front and Bathurst and Spadina, the Distillery etc.–these are now major destinations for living, working, and playing, and they have zero rapid transit. If the DRL made sense 25 years ago, the case for it is much, much stronger now.


  11. A key point in regards to jobs growth downtown is that adding a DRL would spur office building development in the downtown – in other words, development and improved transportation infrastructure can be symbiotic.


  12. Karl Junkin wrote, “It only gets worse if fares get harmonized.”

    Likely yes, but full harmonization should include GO’s network as well, political will notwithstanding.

    I have posted elsewhere that I believe in a zone system (no new boundaries, just the current ones made 2km wide so commuters from either side into the boundary area don’t have to pay extra) that charges a supplement (perhaps $1) instead of a full second fare and transfers would be equally accepted by both connecting systems.

    This would put added strain on the Yonge subway for sure, but what if this was extended to GO? If the fare from Yonge and Hwy 7 to downtown is $4 ($3 base fare plus $1 zone supplement) using the subway, that should also be the fare for using GO. If I need to take a bus from Yonge and Major Mac to the subway and take the subway to Queen station and it costs me the $4, I should be able to take the bus to RHC and get on a GO train to Union and take the subway from there north to Queen for the same fare.

    Traveling from Pickering to Mississauga should be the same fare whether by GO or using DRT/TTC/MT (notwithstanding the absence of a transfer point currently between DRT and TTC, nor the fact it would take ages).

    Dallas and Fort Worth have this sort of fare integration between DART and The T along with Trinity Rail Express, and the commuter rail lines in Oslo Norway have full integration with local transit operators (your zone-based commuter rail fare gives you access to the transit system in each zone it covers).

    I know this has to be paid by someone, by are we not already paying this and likely more through the cost of congestion and environmental impact? The task is to change mindsets to understand the true costs of these things.


  13. The RHC-Union GO Train service has problems that must be resolved.

    I’ve posted this elsewhere, including in Metrolinx submissions, that the line needs to be re-routed south of the CPR to use the CPR Belleville Don Branch (which GO would be free to purchase, that branch has been disconnected (meaning only the switch was removed) altogether from the CPR network). Someone once compared the travel times from Union to Sheppard-Yonge on the subway (making all stops, of course) and the from Union to Oriole (non-stop, of course)… and there’s only 3 minutes difference. That’s unacceptable. The CPR would offer significant time savings. There’s also the very strong potential of a hub at Don Mills on the north side of Eglinton.

    Oriole and Langstaff need platform shifts to the north (half-a-km, and 125m, respectively).

    At least half the stations along the line require renaming, too… potentially all stations along the RH line should be renamed, for a variety of reasons and existing problems.


  14. Way back when in first year Economics we learned the term “ceteris paribus” or “other things being equal” meaning that you look at one variable in isolation.

    It seems that is what the TTC is doing with planning for the Yonge Subway.

    Transit City has two north-south routes: Don Mills and Jane. In addition there would be East-West routes along Finch (west of Yonge) and Sheppard (east of Yonge-subway & LRT) as well as Eglinton further south. York Region plans improvements to the VIVA service. All these measures would seem to provide some alternatives for using the Yonge Subway (depending on where you start from and where your destination is.)

    Improved GO service could funnel people into the downtown core.

    Have any of these Transit improvements been considered in considering whether to extend the subway line especially in light of all the bottlenecks that exist now with the current level of demand?

    Steve: The short answer is “no”. To be fair to the TTC, the Richmond Hill subway is not their idea, but the brainchild of politicians and citizens in York Region for whom one underutilized subway line isn’t enough. Having said that, the Yonge extension will actually be better used than the Spadina line, but that was in a different riding and served a politically well-connected university.

    At the other end of the scale, Metrolinx has done full network simulations 25 years out that show demands on both major subways at levels lower than today because of riding diverted mainly to GO and the DRT. The problem here is that the entire network is unlikely to be built, and we are not seeing projections for intermediate stages to determine which additions would have the most benefit in the short term.

    Finally, the TTC has somewhat dishonestly, I think, made the ATC conversion of the YUS line a pre-requisite for the Richmond Hill extension even though they have made no provision for the fleet needed to provide the level of service required for the alleged future demand. This is much a question of the politics of shaking funding out of senior governments as it is good planning, and without the Richmond Hill subway was a convenient stick, the TTC would have a weaker argument for full funding of the ATC project.


  15. I believe that the Downtown Relief Line should not be HRT, LRT or GO transit but more like City Rail in Sydney Australia or the electrified suburban service in Melbourne. In the west end I would use the Weston Sub corridor with stations at the major cross streets, Finch, Sheppard, Wilson, Lawrence, Eglinton, St. Clair, Bloor, Dundas-College-Lansdowne, Queen, King, Exhibition, Bathurst Spadina, and Union Station. In the East end I would follow the CN to Pape or Green wood and then follow a subway up to the CP line and then follow it to North Scarborough with stations at Parliament, Queen, Pape-Gerrard, Danforth, O’Connor, Eglinton, Lawrence, Warden-Ellesmere, Kennedy-Sheppard, Finch-Markham. I would use electrified GO transit type vehicles with high platform loading and doors in the end sections above the wheels. For travel of this distance you need a higher comfort vehicle than the standard HRT cars. While it might not provide a large transfer from the Bloor-Danforth line at Dundas West and Pape or Greenwood it would intercept riders from Bloor-Danforth at take them downtown at a faster speed than standard HRT. You could probably run more standard HRT trains on the inner end to provide additional capacity between Dundas West and Pape-Greenwood.

    While everyone is looking at HRT, LRT, BRT, ICTS, the RT and standard GO I strongly believe from my travels that there is a heavy rail electrified service between HRT and GO that has a place in Toronto. Electrification and ATC would allow a higher speed and frequency service than GO with closer station spacing. It would have a wide enough spacing to have a significant speed advantage over the TTC’s HRT and with a higher proportion of seats it would provide a more appealing ride for someone coming in from North York or Scarborough than the TTC and a more frequent and more convenient service than GO. In Summary the vehicles would be 85’ long GO transit style bi-levels with wider doors in the end sections above the wheels that would service a high platform station. Power collection would probably be pantograph and overhead. Sydney uses 65 foot vehicles like this in their suburban service that work like a cross between standard subway and GO style system. They also mix 85’ inter city trains on part of this line but they cannot do the downtown loop because of curve restrictions.

    A branch that continued up North and west from Pape could join the Bala Sub and provide service to Richmond Hill. The existing right of way is too convoluted and isolated to be of much use. Service could also be sent out the CP’s Galt Sub, GO’s Uxbridge Sub and GO’s Newmarket Sub, though the latter is close to the Spadina subway. I do not think it is wise to run any of these into Summerhill Station via CP’s old North Toronto station as it would overload the Yonge Subway where it could not afford it. I think that we need to look outside of the conventional systems and try something new. There is nothing wrong with a different style service and its vehicles would still have a lot in common with the existing GO bi-levels.


  16. Again, we need a moat with fire breathing swans. Keep those wanks in the 905 area from polluting our transit system, especially when they don’t pay for it with their taxes.


  17. Hi Steve:-

    In regards to M. Briganti’s Dec. 8th comment “They could have installed a crossover west of Bay Lower and alternately stubbed trains to that platform, instead of letting it rot….”. I’m not sure that would be physically possible given the lay of the land and existing structure. Indeed an intriguing idea. East of Bay lower might be achievable if the money would allow for the layout of all of the physical plant at this end is better, but even here, there may not be the space to have a full trainset sit clear of the new x-over and still have its backside be in the clear of the WB switch at the west end of Yonge.

    It might work if it were only a trailng x-over instead of a full double x-over, but even then, we’d be looking at inches to spare. Given a delay on the low level platform, for any reason, the incoming train may be holding up service if it cannot be in the clear while awaiting the platform, therefore then the west end being the more desirable but next to impossible to shoehorn into place and still make it work.


    Steve: There is definitely no room west of lower Bay for a crossover and associated tail tracks. The curves leaving the station begin immediately west of the platform, and they are tight to dodge around building foundations. East of lower Bay would also be challenging because, as Dennis mentions, a westbound train waiting for clearance through the crossover would likely not be clear of the switch west of Yonge Station and would therefore hold westbound service.


Comments are closed.