Why “Professionals” Didn’t Design Transit City

Now and then, word reaches me that some of the professional transit folk in these parts have their noses out of joint because Transit City wasn’t designed by their elite brotherhood.

Some of them want more subways.

Some of them want more Bus Rapid Transit.

Some of them think the lines are in the wrong place.

Well, I hate to disappoint them, but Transit City didn’t get designed to their liking for some very good reasons.

First off, there’s the small matter of Toronto’s Official Plan.  The original idea was to include a transit plan, but the then-head of TTC management would not countenance the City planning where transit lines would go.  Therefore, what we got instead was a reproduction of the then-official TTC plan consisting of some surface priority routes and a few subway extensions.

The surface priority routes were not chosen because they made sense from a planning point of view, only because they were routes that already had frequent service.  Planning for future development only by looking at past service is a bad idea, and the Official Plan was looking at major changes with the Avenues proposal.

The two subway lines (Sheppard East and York University) were on the TTC’s map because they had been railroaded into the “Ridership Growth Strategy” when some TTC staff and other subway advocates got miffed that RGS might mean the end of a focus on subway construction.  The Commission went along with adding these rapid transit lines — whose very premise was contrary to the RGS goal to produce low-cost short term improvements —  as a fourth RGS category. 

Former TTC chair Howard Moscoe told me “don’t worry, they will never get built because we won’t have the money”.  Needless to say, the fact that the TTC endorsed these lines as their next subway projects gave them “legs”.  Sadly, these lines did nothing to improve service to the city as a whole.

Some have argued for new transit corridors to be built in available rights of way.  The Official Plan wisely designates streets where real people actually will live as areas for intensification, not hydro corridors, rail lines and expressways.  Transit should go where people are and if this means taking road space to do so, then so be it.

Some have argued for increased use of BRT over LRT forgetting that the capacity of BRT, in a two-lane, middle-of-the-road operation cannot handle the projected demand for the Transit City corridors.  High capacities claimed for BRT are only achieveable with multi-lane stations and infrequent crossings with intersecting streets.

Some have argued for more subways versus LRT forgetting that in the corridors we are considering, an LRT subway can handle the high demand in the central sections while surface operations are quite viable elsewhere.

Some fret that people will not be able to travel the length of the GTA at great speed on Transit City forgetting that it is designed for comparatively shorter trips, not to move large crowds from Malvern to Long Branch.

For my part, I have been waiting since 1972 to see a transit plan that really looks at the system as a network, that doesn’t try to spend a fortune on a subway-in-every-ward approach, that recognizes that there are many ways to provide transit service of which an important component is LRT.

Transit City was a political proposal, but one grounded in reasonable planning principles — put transit where the people are and where we plan for them to be, and scale the infrastructure to the needs of the corridors.  Some changes will inevitably be made as the network is fine-tuned, but the basics are sound.  If anything, Transit City suffers from not being ambitious enough.  It was scaled to be financially within reach (a stretch, but not impossible) and with a construction timeframe where people might actually get to ride the network in their lifetimes.

It fundamentally altered the discussion about transit from “what line will we build next” to “what should the network look like” before Metrolinx drew its first map or Queen’s Park announced MoveOntario.  Moreover, it broke the pattern of “more of the same” planning that gave us nothing but ruinously expensive subway proposals.

Operationally, we will have major debates about how streets should be designed, what real “transit priority” will look like at intersections, how far apart stops will be, and how road capacity will have to be sacrificed to transit to make this scheme work.  But for once we will look hard at what we can do with something other than subways rather than dismissing this out of hand. 

The “professionals” have a long history of throttling debate on alternatives and Toronto, nominally a “transit city”, continued to grow with an auto mentality because the transit schemes were too expensive and workable options to improve the network never saw the light of day.

I as a transit advocate and Toronto as a city waited 35 years to see a transit plan we could actually believe in.  We don’t have another 35 years to haggle about what we should build, and those  who try to undermine Transit City do us all a disservice.  Discussion and debate, yes.  But don’t dismiss it just because the “wrong” people drew the map on a napkin over coffee.

46 thoughts on “Why “Professionals” Didn’t Design Transit City

  1. Steve said: “A two minute headway of 60m trains (two-car units) would give a capacity of at least 8,000 per hour depending on assumptions about the design capacity of vehicles.”

    But 8,000 pphpd, or 16,000 counting both wings of the DRL, is a weak relief for the two downtown subway lines that together carry at least 60,000 pphpd during the peak. The system is overloaded already; those extra 16,000 would be a substantial improvement for a constant passenger flow, but leave little room for further growth. Moreover, even if we assume a higher LRT capacity (with 3-car vehicles and 1.5 min headways) and speak of 30,000 or 35,000 pphpd (both DRL wings combined), even that might not be enough if the downtown employment grows substantially. Meanwhile, the unique rail corridor would be used up.

    Of course, the equation changes dramatically if the subways get a serious help from GO; but that looks somewhat uncertain. OK, perhaps Union station will not be a bottleneck after the renovation; but how many 2-nd tracks have to be installed for the frequent service? are the corridors wide enough to add those tracks? how much the upgrades of multiple GO stations in 416 will cost? and, some rail lines go through residential areas, will the residents accept 15-min or 10-min train headways instead of the present 1 or 2 h?

    Steve: No tracks are added to Union to get more capacity, but the rather generous layovers of trains on platforms will disappear. These are not sleeping cars to Montreal.

    Speaking of costs, the equation here is (light rail DRL + massive GO enhancements) versus (heavy rail DRL + some GO enhancements). It is not immediately obvious which combo is more expensive, given that the DRL will be fully grade-separate in either case.

    Steve: The point is that we have to look at the alternatives, not make assumptions that the only way to handle demand into the core is with yet another subway line. If GO had been integrated with TTC years ago, the way we think about planning rapid transit would be much different today.

    Steve said: “Actually it won’t intercept as much traffic as you might think because the only point where it would be fed by an east-west route is at Eglinton. The critical catchment area for traffic diversion from the subway network is further north and east, right where GO could provide a direct link to downtown.”

    Well, not only Eglinton, but Lawrence East as well, that route runs via Eglinton anyway. Plus, Thorncliffe Park, Overlea, Flemmington Park areas, all rather dense. Plus, passengers from the Don Mills LRT, and perhaps even Sheppard LRT routed down Don Mills. Plus, it is possible to route Leslie and some of Victoria Park buses to the subway terminus at Don Mills / Eglinton.

    Steve: The Lawrence East bus does not run on Eglinton until Leslie, not Don Mills, and the Thorncliffe/Flemington areas already feed the Bloor Subway via the bus routes. They do not represent a net addition of demand in the corridor, only a transfer of demand from a bus feeder to the Don Mills line. The Leslie bus does not run often enough to make any difference to anything, and the Victoria Park bus will connect with the Eglinton LRT. It’s hard to say how the passenger flows will redistribute themselves, but don’t forget the Eglinton line will be in place first.

    Regardless, it will be nice to have reasonably frequent GO service on the Stouffville and Agincourt / Seaton lines, and good integration with surface routes, Sheppard E LRT in particular. That will be an extra option for the riders from North Scarborough, while also serving the 905. I am just not sure if those GO enhancements can shave enough of peak demand to make a noticeable difference for subways, in case the overall ridership to downtown keeps growing.


  2. @Calvin Henry-Cotnam:
    Here’s a lot info that may be of interest (the TTC does not seem to advertise this info, I must admit, you really have to be looking for it to find it):

    Download Chapter 7 and on page 23 is a cost breakdown. Something else to also weigh in the comparisons is that the Spadina extension includes Wilson Yard expansion, but Transit City does not include costs for new carhouses.

    Calvin:The comparison with the Prince Edward Viaduct is poor and perhaps disingenuous for one simple reason: the provisions on the bridge were not to be used immediately, with expected increase in use to a point where all use would have to be halted for a period of time for conversion to take place. Think of the Bloor-Yonge station. Forward thinking in 1954 would have been to build a then-unneeded centre platform.

    I don’t understand why you say it is disingenuous, the conditions seem to be the same. The Viaduct was constructed with provision for a rapid transit line to run on the lower level (originally not for the TTC since the TTC didn’t exist at the time, as Steve has kindly pointed out, but the argument still holds since it was ultimately used… eventually) that was not to be used immediately. Eglinton is being given subway-sized tunnels in width and LRT-sized tunnels in height, even though the extra width will not be used immediately; just like the Prince Edward Viaduct.

    The argument about Yonge simply does not work at all, because Bloor-University were designed as one project, and Bloor should never have needed a centre platform if it worked the way they designed it to. They built a wye to carry people through the core without transfers, and that was forward thinking, but although it was forward thinking, it wasn’t well-managed, and in some aspects not well-designed either (that wye has some inherent design flaws that cripple it no matter how the service is managed). So they were forward thinking in 1954, they just weren’t as good at it as they were in 1914 ;). As to whether or not we can simply start using the wye again today to deal with the capacity problems, the short answer is no due to the Spadina Line.

    Calvin: That said, this city needs to play 40 years of catch-up in improving the overall network, and Transit City fits the need to fill this priority.

    Well, yes, if you treat the north-of-Bloor part of the city in isolation, that statement is absolutely true. The biggest concern I have about Transit City, as I’ve stated here several times, is how it will make Bloor-Yonge and southern Yonge in general worse than it already is today. What would be most prudent is to have the DRL built first to ensure Transit City can grow the network without pushing it over its limit. However, I agree with the notion that we can’t afford to wait anymore for some real suburban improvements to the network on a large and cost-effective scale. That said though, they could have been more strategic in picking which lines (or parts of lines) get built first so that the new pressure is mostly taken on the Spadina and Bloor West lines instead of Yonge.


  3. Several points to comment or follow-up on…

    Karl Junkin wrote, “GO trains are by-and-large packed by the time they cross into the 416.”

    While this is very true, and plays a significant impact on the idea of using GO as a way to improve intra-416 travel, are there any figures available on use of 416 stations as an exit point for 905 commuters? I suspect that the York U stop on the Barrie line is mainly an exit in the mornings and an entrance in the afternoons, but what about others? If Sheppard West were implemented, would it add more to a crowded train going downtown in the mornings, or siphon off some of the crowd that would benefit from a southbound subway instead of having to double-back from Union?

    My thanks to Karl for pointing out the specifics of the Spadina budget. As also pointed out, the Wilson Yard expansion costs will also be removed from the comparison (at the same time, I will highlight where some of the LRT cost examples actually include storage and maintenance facilities).

    On the issue of building the Eglinton line for future upgrade to full subway, Karl said, “I don’t understand why you say it is disingenuous.”

    The comparison was made to the King Edward Viaduct being forward thinking. In the case of the Viaduct, the lower deck was for future use and did not see active service until it was needed for the Bloor Danforth subway. This is not comparable to building in provisions on a line that will be used in revenue service. Conversion to full subway will not be possible by closing it for a couple of Sundays or done during overnight periods. The line will have to be totally shut down, I suspect for at least 2 months, though 6 months is probably more realistic. If the rider ship of the line grows to the level that this will be needed, the impact of shutting down the line will be so great that it might make it impossible to do (hence, my reference to Bloor station).

    Care should be taken to look at this really closely. If there is a hint of needing full subway, then build it now. I suspect that this will not be justifiable. I do suspect that building the underground LRT line with platforms long enough for 5 or 6 cars is justifiable. While the cost of building a subway-up gradable LRT tunnel will be less than a full $250 million per kilometre subway, I strongly believe that it will be more than the cost of building a basic LRT tunnel, even with 5-6 car length stations ($130-160 million/km).

    Edmonton built its underground stations long enough for 5 cars (their cars are 25 metres long, so the stations are only four cars long for the 30 metre cars we are talking about). The Grandin extension opened in 1989 added 800 metres and one underground station for a cost of $67.1 million, or $117.7 million in today’s dollars (using a 3% inflation rate for 19 years). That puts it at $147.1 million per kilometre.

    Steve: Don’t forget that we are going to be running 30m long LRVs and there is nothing to prevent the “subway” part of the Eglinton line from (a) being extended and (b) operated with longer trains. You don’t have to use “subway” cars to run “subway” service especially since the LRT operation will have floor-level loading just like a subway train. It’s a lot easier to figure out how to handle carhouse moves from an LRT subway (simple things like breaking trains into shorter pieces if they have to traverse part of the surface network) than to make a fully operational connection from an Eglinton subway into the existing rapid transit network.

    This is conceivable only if the Eglinton line makes it all the way to Kennedy. Connections at Yonge or Spadina are geometrically impossible. Please don’t waste everyone’s time with discourses on how we might shoehorn one in.


  4. Steve wrote, “It’s a lot easier to figure out how to handle carhouse moves from an LRT subway (simple things like breaking trains into shorter pieces if they have to traverse part of the surface network)”

    Even this may be simpler, after all, will it really be necessary to break such moves? The 2-car (or perhaps 3-car) limit for surface operations will primarily be imposed by the need to keep platforms at stops within limits usually determined by city blocks. A movement of an out-of-service train that is longer would not impose the problems that an in-service train would.


  5. Steve: I am letting these comments through in the interest of debate, but believe that this has been done to death and will not post any further installments.

    For those who have been here before, I have added to comments at the end of the section about the proposed Keele junction.


    Greenwood Wye; I understand what you are saying, but fundamentally, taking two main differences into consideration, there is little difference between the level crossing at Greenwood and the double-crossovers at any terminus, where every second train crosses over the opposite direction at-grade on the way out of the terminus. The only differences at Greenwood is that this crossing is between stations rather than right in front of one, and that a 90-degree turn is involved (both affect line of sight, but that should be unimportant, we have signal systems for that).

    Steve: The big difference is that at a terminal you have one outbound train whose departure gets priority (the route is claimed about one minute before departure time to avoid conflict with an incoming train), and an inbound train that has layover time built into its schedule for terminal operations. At Greenwood, you would have two mainline services for which delays could not be tolerated.

    Keele; I propose this location not because it would be light engineering work. It is heavy engineering work, make no mistake. There are other more important issues that make this site attractive:

    I propose this location because, first, it is probably the only location where the change can be made without massive destruction to nearby properties (or streets, for that matter). Expropriation of existing structures would be minimized at this location, though not avoided altogether. Second, very important, it is the only location I can identify where Bloor-Danforth services can be maintained at all times (with perhaps a one-off Sunday being exception) in operation over the entire construction period.

    Turnback ops require at least one (preferably two to stay afloat in the rush hours) tail track(s) immediately west of Keele to work. From what I gather from Google Maps, the large developments are all out of the way for this. Whether or not the new track(s) can go under the existing houses or not with some underpinning, I can’t say (but it’s been done before).

    Keele must be converted to a three-track two-platform arrangement since trains would be entering the station from both Dundas West (due to alignment geometry, it is impossible for an interlined DRL to stop at Dundas West) and coming off the Weston Sub on the east side (otherwise, we’ll end up with the same problem Museum inherently had). The existing Keele south track would remain as is (ideally at least), expansion would take place on the north side (we’d lose the bus loop, but…)

    Is it tricky? That’s one word to describe it. However, compared to other locations (where “tricky” would certainly be an understatement) and the network benefits it can carry, it has its merits.

    Steve: There is a fundamental problem with using the existing Keele Yard trackage as a lead into a western branch of a DRL. Those tracks extend into Dundas West Station, and there is a door about half-way down the stairs to platform level that leads into the eastern end of that structure. If you project those tracks further east, you will be immediately above the existing station building and will collide with the basement of the Crossways building on the east side of Dundas. Don’t even think about shifting the tracks because other structures (now or planned) are in the way no matter where you turn.

    Alas, this reminds me too much of the TTC Engineering Department’s scheme for the reconstruction of Bloor-Yonge Station. Anything is possible with the application of enough money, but the underlying question is why do it in the first place. Your proposal seems to be bent on a U-shaped integration of a DRL and the BD subway, and you refuse to consider alternatives.

    GO Transit; The options here are quite limited, but are talking about it intercepting feeders, or alleviating the subway at subway/GO transfer points? If the latter, there are many problems with this, but I imagine you are well aware of that, so I suspect it is the former, but in that case the feeders that it can work with wouldn’t make much of a dent anyway.

    GO trains are by-and-large packed by the time they cross into the 416. Heck, Bradford didn’t even stop in the 416 at all prior to York U station being added. It will be interesting to see how the new Sheppard West station works (I’m pessimistic on this one, actually).

    GO can’t do that much if you think about the restrictions GO is forced to work with. It certainly won’t be enough to really address the problems at Bloor-Yonge/St.George and southern Yonge. What space is made would get filled up again anyway by new riders and we’re back to where we started. A DRL can make a much bigger dent, so that even when new riders inevitably start to fill up the space created by the alleviator, it should still be a big improvement.

    Furthermore, GO is currently in a position where demand for GO service is growing faster than GO can increase supply. Given that this is the conundrum that GO is in, which has both pros and cons for GO, talking about GO alleviating subway capacity is very premature and GO would indeed scoff at the idea given the problems they’ve already got on their plate.

    However, the biggest problem with integrating fares here, is that if GO and the TTC are going to be on the same fare, that would involve the TTC being uploaded by the province, and taken out of Toronto’s control to be at the mercy of Queen’s Park. I don’t think any of us here want that.

    Steve: I am talking about substantial increases in GO capacity, some of which are already on the drawing boards including the Lakeshore electrification, improved service on the Georegtown line and new service on the Agincourt/Seaton line.

    As for integration, Metrolinx is going to take over GO, and there is strong pressure for regional integration generally. We need to stop planning as if these are separate systems.


  6. Don’t you think it’s going to be rather expensive to convert the Sheppard subway to LRT? the fact that Transit City is all LRT is all well and good but it seems to me that the money that would be spent on that would be better spent just getting the LRT lines built. Convrting the Sheppard subway to LRT would only help defeat the purpose of building LRT rather than subways.

    Steve: This depends on what your goal is. If you cringe at the thought of discontinuity between a Sheppard East LRT, a Sheppard Subway and a Sheppard West bus, and dream of a single line end to end, then the choices are fairly basic. Either you build a much longer subway and waste a small fortune, or you bite the bullt and use the existing subway for LRVs.

    Yes, a surface operation wouldn’t have been a bad idea if that’s what we had built at the outset, but the subway is sitting there. The question is whether it can be converted reasonably quickly at a cost competitive with an all-surface LRT. Don’t forget that surface operations face challenges both at the Don Valley Parkway crossing and at Yonge Street.


Comments are closed.