Wye Didn’t Subway Integration Work? A Look Back at Subway Operations

The recent 40th anniversary of the Bloor-Danforth Subway brings up another piece of TTC history:  the integrated operation of the Bloor-Danforth and Yonge-University lines.  For those of you who weren’t here to see it, there were three interlaced routes:

  • Keele to Woodbine
  • Eglinton to Keele via Downtown
  • Eglinton to Woodbine via Downtown

This operation is possible because there are connection tracks west of St. George Station from the upper (University-Spadina) level to the lower (Bloor-Danforth) level merging just east of Spadina Station.  North of Museum, there is a junction whose eastern tracks lead to the lower level of Bay Station and thence to the eastbound Bloor subway with the junction just west of Yonge Station.  You can read about this in some detail on Transit Toronto’s site here.

I recently received a note from Mimmo Briganti (who contributed to the article linked above):

I was wondering if you know anything about the politics surrounding the discontinuation of *integrated* service on the Yonge-University and Bloor-Danforth segments in 1966.

I understand the TTC was opposed to Norman Wilson’s Bloor-University wye during the design/construction phase and intentionally misoperated and induced delays on the integrated system so that they could justify moving to the current operating format.  Is this true?

As a daily Bloor-University subway rider for the past 20 years, this issue always crosses my mind when changing trains at St. George or Bloor-Yonge.  I don’t think this is what Norman Wilson had in mind when he designed the system.

I’ve read numerous reports from the 60s on the subject and I always walk away with the impression the TTC created artificial delays at the wye by forcing all trains to run in proper sequence, and by using a rigid schedule-based switching system when better alternatives were available.  Chicago, for instance, used the same Identra coils we had to control platform signs *and* the switches at their junctions (instead of relying on trains to follow a strict schedule).

The issue here is one of operational pig-headedness, not of politics, although it’s hard to tell them apart in any large organization.  TTC staff fought the full wye design, but were forced by the Commission (then an appointed citizen body rather than a committee of Council) to build and operate it.  The way that this was done guaranteed it would fail.

Trains on each of the three routes were forced to stay in order at the junction, and inevitably, things got screwed up when one train was slightly delayed.  A minor holdup with a sticking door, a late crew change, or an ill passenger would throw everything into chaos.  Imagine if we tried to run surface routes like this with buses and streetcars waiting at junctions so that they could leave in the correct, pre-ordained sequence.

Why did trains have to wait for each other?  Scheduling.  Rather than treating them as three separate routes that happened to share tracks (the way any sensible system with interlined service would), the TTC built a totally integrated schedule where trains switched from route to route as the day progressed.  This scheme preserved a rigid alternation of destinations as headways changed, but immensely complicated operations in the process.  The trains had to stay in the correct sequence for all of those route swaps to work properly.  There were also crewing implications of this arrangement, but I won’t go into that here.

Even after they changed from rigid alternation of trains at the wye, they still had to resequence things at terminals and this produced further delays as trains were held for their place.

As the Transit Toronto article points out, the TTC went on to use the Origin-Destination Survey of riders to buttress its case against integration.  In summary, about two-thirds of all trips were unaffected by the integrated service because they didn’t travel through the wye junction.  Of the remainder, roughly half obtained a direct ride they wouldn’t have otherwise while the other half might have to let a train pass by because it didn’t go to their destination. 

The TTC interpreted these figures to indicate that the integrated service was more of an inconvenience than a convenience to its riders (everyone was affected by the ragged service brought on largely by their mis-handling of the operation, but that was never mentioned) with only one-sixth of the riders deriving a benefit.

We can play with figures a lot, but the important idea was to avoid the transfer moves, congestion and delays at Bloor-Yonge and St. George stations.  Anyone coming from the west whose destination was from about Queen Station south to Union and back up University Avenue would choose a “downtown” train.  From the east, the break-even point was further south, around King Street, because of the longer trip via University.  Every rider who avoided a transfer also lessened the crunch on platform, stairway and escalator capacity at the two interchanges.  Moreover, riders from the Danforth line who might otherwise change at Yonge could be persuaded to use the University line because of the transfer-free trip.

Alas, it was not to be.  Lower Bay Station has been closed to all but film shoots, occasional fantrips, training courses and other odd bits since September 1966.  The inbalance in demand on the two lines, one of the original reaons advanced for breaking them apart, has vanished and headways on B-D and Y-U-S are almost identical. 

Of course, the extension of the University line up Spadina makes integration impractical today.  Possible, sort of, but the complications are not worth the effort because too little capacity remains on University to absorb many trains merging from Bloor-Danforth.


I have a copy of the integrated subway schedule, and it is thing to marvel at for the amount of work that must have gone into working it all out.  Trains are scheduled to the nearest six seconds all day long.  Trains meander from route to route, sometimes making only half or one trip on, say, Eglinton-Keele before becoming a Keele-Woodbine and later a Woodbine-Eglinton train.

The TTC went from running one simple subway line between Eglinton and St. George to the complexity of an integrated three-branch operation.  It’s no wonder it didn’t work, but clearly it was designed to fail.

2 thoughts on “Wye Didn’t Subway Integration Work? A Look Back at Subway Operations

  1. I have been very interested in subway integration and believe with today’s technology, sequencing delays could be avoided.  If there is one thing those who run transit systems miss, is riders do not like to lose time and be inconvienced by transferring.

    Columnist Ed Drass has pointed out Bloor-Danforth trains delay for sequencing at Ossington, Keele, Chester, and the major interchanges such as St.George, Bloor-Yonge and Kipling.

    Steve:  These delays are not for train sequencing.  The intent is to keep trains on the line regularly spaced.  Some crews used to make very short station stops and creep up on the preceding train.  Their intent was to get to the terminal early and have a longer layover. The resulting irregular service would annoy riders.  The tradeoff here is many short delays for more reliable service overall.

    Right after the wildcat strike an emergency occured on a Bloor-Danforth train at the subway interchange delaying all trains including Yonge-University-Spadina.  You have mentioned the congestion at Yonge/St.George interchanges during rush hour.  Does the TTC have to wait for a tragedy to address this?

    Steve:  The only way an emergency could block both lines would be one that required a major power cut.  If a line is held up, an integrated subway operation will not necessarily continue to operate.  It would not, by itself, resolve crowd backups at St. George and Bloor-Yonge.

    On the cost side what business would spend money on something that does not generate revenue or is value added?  The TTC recently upgraded the tracks and switches at St.George/Spadina interchange as trains still use the wye for service adjustments. 

    Steve:  The work at St. George/Spadina was for routine maintenance. 

    I have written counciller Saundercook who took an interest in Lower Bay to relieve commuter congestion but like everything else the matter seems to have gone to sleep.  You would think the TTC realizes increasing numbers of passengers are coming from Kennedy and Kipling and going south of Bloor.


  2. Thanks for clearing up a couple of things.  I am curious to visit Lower Bay station and had tried emailing the TTC regarding a tour to no avail.  All tours are for prospective filmakers.

    I visited a couple of websites from subway infiltraters (No I was not seriously thinking of attempting this!) who commented that Lower Bay is dangerous due to frequency of trains switching lines.

    Any active line requires routine maintenance which costs money.  I’m sure there must be some routine maintenance to prevent Lower Bay from falling into complete disrepair.  Again money being spent on something that could generate revenue and alievate congestion at St.George, Spadina & Bloor/Yonge.

    It seems most businesses would only spend money as an investment to maximize service and generate potential revenue.  Add to this passengers, some maybe tourists who have to carry bundles, laptops, and children or the handicapped having to get off one train go up or down lugging their belongings to change trains.  I believe The TTC should integrate.

    Steve:  Although trains do go through lower Bay station from time to time, they are hardly “frequent”.  This track is mainly used for moving cars between the University and Bloor lines to get them to Greenwood Shops (accessed via the wye between Donlands and Greenwood stations).  Neither Wilson nor Davisville carhouses have the equipment needed for major repair work.  The track is also used by work trains, usually in the middle of the night.  Most special work cars are based at Greenwood.

    As for transferring passengers, the issue is whether an integrated service (including the Spadina subway) could be operated feasibly with three lines merging southbound on University, as opposed to the original two.  Although theoretically possible, the frequency of “downtown” trains from Bloor-Danforth would at best be one-in-three, maybe one-in-four at peak, and most people won’t wait that long.  It would also limit the number of trains that could be sent up the Spadina line to the point where it would be overloaded.  A very large expenditure to change the signalling system allowing more trains on the Yonge-University line could deal with some of the capacity problems, but with the very close station spacing, I do not think that even newer signals can squeeze many more trains onto the line.


Comments are closed.