The Secret Sheppard Subway Report

On February 15, 2012, the Star’s Royson James wrote about a TTC report prepared in March 2011 for Mayor Rob Ford on the Sheppard Subway.  The article included a photo of the report’s summary.

Royson James graciously provided me with a copy of the document, and it is available here for those who want to see the whole thing.  I suspect that it is only part of an even larger report because this material only covers one big question: why are the assumptions from the Network 2011 study done back in 1986 no longer valid?  There is no discussion of construction costs, project financing, or any comparison of alternative schemes.

2011.03 Transit Technology Summary and Background

2011.03 Transit Technology Table

Note:  These files were prepared by scanning the copy I received, which itself was a previous generation copy including a lot of marginalia.  The text was imported into and formatted as a new Word document with approximately the same layout (and typography) as the original.  This allowed it to be “printed” in PDF format (the files linked above) rather than a much larger set of images of the scanned sheets.

The report contains a few rather intriguing comments that won’t sound new to regular readers of this site, but which raise questions about the planning assumptions underneath decades of work by the TTC, City Planning and other agencies.

Planners and politicians make grand statements about how policies, official plans and zoning will focus development in locations and patterns of their choice.  In practice, this does not actually happen because the best intentions are inevitably diluted by political reality.  Developers build where there is a real market, not where a plan tells them they should build.  Jobs move around in complete ignorance of city, regional and provincial goals.  Do you own some land that doesn’t fit the plan?  Just sit on it until a friendly government comes to power and get a brand new, as-of-right zoning upgrade.

The idea that transit will shape development is demonstrably false because so many parts of the city with subway stations have not, in fact, developed at all.  This may be due to neighbourhood pressure, or to a policy of preserving the “old” parts of the city because that character has a value greater than massive redevelopment.  A neighbourhood may simply not be ready for development, or may have the wrong character.

This is particularly striking for residential development where local amenities and the “feel” of a neighbourhood are more important than with an industrial/commercial/office development.  People may work in office towers surrounded by pedestrian-hostile roads and parking, but they want to go home to something friendlier.

Because the market for commercial real estate and the jobs it brings has shifted to the 905, much of the development in nodes originally intended for employment has been residential.  This completely changes the transit demand pattern.  Instead of many commuters travelling “in” to a few nodes, we have residential areas that spawn outward trips all over the GTAH.  Subway plans presumed the concentrated trip making that nodes full of employees would create, and these have not materialized.

We are now seeing this pattern even in downtown Toronto with the growth of the condo market.  Many residents live and work downtown, but a considerable number are “reverse commutes” out to the 905, trips for which both the local and regional systems are very badly equipped.

The idea of “downtown North York” or “downtown Scarborough” has simply not materialized in the form expected three decades ago.  Actual employment levels at these two centres are about 1/3 (North York) or 1/5 (Scarborough) of the 1986 projections.  This should be a lesson for today’s planners and politicians who think they can forecast and direct future growth patterns with the aid of a few maps and regulations.

The employment growth projected back in 1986 for “Metropolitan Toronto” (now the City of Toronto) was a rise from 1.23-million to 1.9m.  In fact, employment grew only to 1.30m by 2011 with the lion’s share of the jobs going instead to the 905.  With the absence of strong nodes for new jobs, there was little chance of improving the modal split to whatever commercial development did occur.  Combining lower than predicted growth and a failure to achieve the projected transit modal split leaves us with demand projections that are completely meaningless.

Far too often, there is a political imperative to make the future look better than it might be, or at least to do a proper sensitivity analysis, a “what if” scenario for conditions that don’t match what we would like to see.  Any subway financing scheme that depends on future ridership must answer basic questions:  will those riders actually arrive, and will land development occur in a manner that will generate trips the subway will serve?

We have already seen development in the Sheppard corridor, but it is unclear whether this attracts buyers because it is near the 401 and DVP (and thus to a wide set of GTA destinations), or because it is near the subway.  That development is generating many car trips because, for most destinations, auto travel is the only real option.  The market share for transit at the North York and Scarborough centres is barely half what was projected in 1986, and the compound effect of much lower employment means that transit demand to these centres is a trivial fraction of what Network 2011 was intended to serve.

One item caught my eye in the section of “Public’s travel patterns and behaviour”.  Not only were the employment and mode share values used to model demand considerably above what actually happened, assumptions were  made about the way the Sheppard subway would get its passengers.  Regional and local bus services would be gerrymandered to force riders onto the Sheppard line (at least in the model), but riders actually preferred to go to Finch Station where there was a chance of getting a comfortable spot on a train.

Another assumption in the demand model was that the cost of driving would rise substantially both through higher gas prices and the cost of parking.  Neither of these materialized, although based on typical motoring behaviour, without a very  good network of transit alternatives, the pricing of auto trips does not discourage much travel.

This begs a vital question for all regional planning — can we trust the models?  What assumptions went into the model for our new transit  network, and have these been tested against actual patterns of development and of the regional economy?

The projected demands on new transit lines made back in 1986 were substantially higher than today’s expectations:

  • The Sheppard subway was expected to have 15,400 peak riders by 2011, but the actual number on the existing line is 4,500.  The projected peak demand for the full line in 2011 is now 6-10,000.
  • The Eglinton subway was expected to have 17,600 peak riders by 2011, but the LRT projection is now reduced to 5,200 (based on having the central section underground).
  • The Downtown Relief line was projected to have 11,700 peak riders by 2011, and the demand projection today is 12,000.  This is no surprise given that the DRL would serve a demand that actually existed 25 years ago, rather than a notional demand in a regional plan.

In previous articles, I have discussed the matter of the TTC’s Capital Budget and the mounting cost of simply keeping the subway system running.  Nothing lasts forever, and many systems are wearing out.  We are now on the third major generation of vehicles, there are problems everywhere with station finishes and equipment, water penetration and damage is an ongoing headache, and the signal system must be completely replaced.  Contrary to statements by some subway advocates, subways do not last for 100 years without major investments in rehabilitation.

Back in 1986, the TTC had not yet reached the point where the subway had started to wear out.  The oldest line (Yonge from Eglinton to Union) was only 32 years old, and much of its first generation equipment was still functional.  The TTC now knows that the subway system has an ongoing cost of $230m operating (routine maintenance) and $275m capital (major systems replacement) every year.  Looked at another way, simply maintaining the subway system consumes about 1/6 of the annual operating budget, and a substantial slice of the non-expansion related capital budget.

There is a large backlog of needed capital repairs with a shortfall of $2.3-billion in the 10-year capital budget thanks to provincial cutbacks in capital funding.  Building more subway lines will only add to this set of maintenance costs a few decades in the future.

Finally, we have a bit of creative history writing.  Why, the TTC asks, was LRT not embraced as an option back in 1986?  They claim that at the time it was a poorly understood mode with only limited use, particularly in North America.  What we now think of as “modern” LRT had not yet evolved.  This statement ignores the LRT renaissance in Europe and suggests that despite new LRT systems in North America (notably Edmonton’s and Calgary’s), it was too soon for the TTC to embrace the mode.

I will not dwell on the fact that the Scarborough ICTS system was brand new, and the idea that an “intermediate capacity” system between buses and subways already might exist was simply not in accord with provincial policy.

In fact, the TTC’s love for LRT is a very recent phenomenon.  When the Ridership Growth Strategy was first proposed in 2003 for “short term” service improvements, TTC subway planners were terrified that their pet projects had fallen off of the map.  The RGS was hastily amended to include a commitment to the Spadina and Sheppard extensions, and this move has been cited ever since as “proof” that the TTC supports the Sheppard line.  It would be another four years before the Transit City scheme was launched.

LRT was well-established around the world before the Transit City plan was announced, but it took a major rethink of Toronto’s transit network at the political level, combined with the economic constraints against subway building, for LRT to get the consideration it deserved.  Transit City was not perfect, but it got Toronto thinking about what might be built.

This report is a year old, and its existence shows that the pro-subway forces in Toronto, notably in the Mayor’s office, did not want an informed, public discussion of subway plans to occur.  Observations about the changing growth patterns in Toronto raise important questions about the future role of transit, indeed of the ability of transit to serve the region as we have actually built it.  Far too much effort is concentrated on the subway-vs-LRT battle in a few corridors when the real challenge lies “out there” in the growing and very car-oriented 905.

116 thoughts on “The Secret Sheppard Subway Report

  1. DavidC said: “It’s on City website but, of course, no details to date.”

    Other than a subtle reference to a Roman Emperor who also had a ‘career altering experience’ on the Ides of March.


  2. Regarding the LRT/Subway between Don Mills Sta. and say Victoria Park.

    A tunnel has to be bored under the Don Valley. That’s a fixed cost, LRT or Subway. Very little cost savings in running the LRT to Don Mills. It’s time to fix a mistake.

    There are four TBM’s on the Spadina Extension. They even have names : Holey, Moley, Yorkie & Torkie. Surely two of them will be in running order by 2013. Not a big cost saving but what did they do in Madrid?

    Steve: The LRT would surface west of Consumers Road, while the subway will go all the way to Vic Park and will almost certainly have tail tracks beyond the station. That’s two more underground stations costing at least $125-million each, plus the longer tunnel. As Karl Junkin has pointed out in an earlier comment, the tunnel (regardless of technology) will be too shallow to build with a boring machine and will have to be cut-and-cover.


  3. I thought everyone agreed it made sense to take the subway to Vic. Park and start the LRT there, since it’s going to be underground anyway.

    Steve: The extra cost of subway stations at Consumers and Victoria Park, plus the extra tunnel (and no doubt tail tracks beyond) could not be justified by the ridership.


  4. As Karl Junkin has pointed out in an earlier comment, the tunnel (regardless of technology) will be too shallow to build with a boring machine and will have to be cut-and-cover.

    Oh, I have confused corridors in the previous comment (I automatically think Eglinton when you say DVP because I’m used to instinctively thinking of Sheppard as crossing the 404, even though they’re the same highway except with different ownership). I thought you were talking about the reported design change (the status of which I remain unclear on) to stay under at the DVP along Eglinton and what that tunnel would entail – sorry, my mistake (that’s what I get for commenting late at night). I haven’t seen any profile drawings for Sheppard’s LRT tunnel as no vertical alignment drawings were published in the EA, only horizontal alignment, except at the GO Uxbridge crossing. I’ve always found that kind of curious since it is in stark contrast to other Transit City EAs that all had vertical alignments published.


  5. Cut and cover across 404/DVP should be interesting and expensive.

    For cut and cover you need a work zone on top for vehicle access. The work zone has to be more than 20 ft long (2 lanes).

    The standard decking for cut and cover (12″ x 12″ timbers and a wear surface) is OK for vehicles travelling 50 kph. For cars travelling 120 kph something else will be needed.

    At the Airport they built a tunnel under a runway (cut and cover, no TBM). The runway was shut down for the work.

    Anyway let the designers figure it out.


  6. Dear Steve:

    General-Specific re: modes & application:

    How viable would a monorail or gondola system or Wuppertal or mag lev solution be for GTA usage?

    Would LRT be best system for Highway 7 in York Region?

    Steve: All elevated systems share a common problem: their effect on the streets over which they run. If a route can go “cross country” and interact with streets only where it crosses them, then the structure is less detrimental. However, many routes people want to travel align with streets, not with the few available rights-of-way. Elevated systems require stations that occupy a considerable amount of space. Where this can be integrated with new development, the effect can be reduced, but this presumes new or redeveloped land along a route.

    Gondola systems do not have high capacity on a par with the sort of rapid transit lines we need in the GTA, and their application is far more appropriate to locations with geographic constraints notably absent in our area.

    I note that you don’t mention Skytrain technology which, at least, has the advantage of being a developed mode used in rapid transit configurations.

    The common issue through all of this is that we can put transit on the surface with BRT or LRT, or we can move it above or below grade. That move has a big effect on the cost of implementation and operation. LRT has the advantage that it can be at grade where room is available, but can run in tunnel where space is tight.

    I cannot help thinking of Yonge Street in Richmond Hill where even VIVA cannot get a right-of-way to avoid road widening. How would an elevated look in the middle of that street, or in any of the other former small towns around York Region?


  7. Being a Richmond Hill resident, I can’t see anyway for LRT to run up Yonge without tunneling between Harding and Crosby unless they were to remove the parking lane which seems even less likely than a Yonge LRT.

    Originally York considered turning the Highway 7 BRT into a LRT once demand justified it, but they seem to have dropped all mention of this from their VivaNext proposals, with only future LRT appearing as a possible extensions of the Don Mills and Jane TC routes up to Highway 7. They do have a disclaimer saying BRT routes are being designed with ability for conversion to LRT as ridership increases.

    Steve: I never had much faith in York’s commitment to LRT because it would require a recognition that street space should be taken on a wider basis than just those locations where it was already easy to drop in a BRT lane and platforms.


  8. My impression is that the plan for Viva is to build what’s currently planned and then have sell itself to the public so that people clamour for expansion.

    One nice thing about tunnelling between Harding and Crosby basically involves going straight into the side of a steep hill, so the portal would be very small.

    I’m more worried about the Highway 7 section under the 404. Right now, the plan seems to be to build a single-lane median, so eastbound buses and westbound buses would have to take turns.


  9. Peter said, “I can’t see anyway for LRT to run up Yonge without tunneling between Harding and Crosby unless they were to remove the parking lane which seems even less likely than a Yonge LRT. “

    On my York Region Options page on my site, I outline what could be built with the $1.4 billion expected cost of a subway extension north of Steeles if LRT were to be used instead. LRT on Yonge would have to be in a tunnel from about half way between Harding and Major Mac to a couple of blocks north of Crosby.

    Peter also said, “Originally York considered turning the Highway 7 BRT into a LRT once demand justified it, but they seem to have dropped all mention of this from their VivaNext proposals.”

    While not mentioned on the VIVAnext site, I believe that when the EAs were done on the busways, the phrase “protect for future conversion to LRT” was used frequently. With rapidways now under construction, it is not necessary to make mention of this as the focus is on the current work. That does not mean that it is off the table.


  10. Leo said about Highway 7 at the 404, “Right now, the plan seems to be to build a single-lane median, so eastbound buses and westbound buses would have to take turns.”

    That is true, however it is a far cry better than what a number of the display boards at the open houses showed: the rapidway lane for each direction would merge into traffic just on the 404-side of the off-ramp traffic lights on each side of the 404. I knew at that time that there was only space to add one lane through the underpass and wondered why they were not even considering doing that. At least that one lane will be added.

    There has been no indication if they may have transit signals to help guide the buses through this gauntlet. Another idea might be to use the rapidway lane for only one direction during peak times, if it weren’t for the fact that Highway 7 tends to be almost equally busy in both directions at both peak times. Perhaps that may change a little when the Beaver Creek-Allstate Parkway mid-block crossing over the 404 is complete.


  11. Calvin Henry-Cotnam said: LRT on Yonge would have to be in a tunnel from about half way between Harding and Major Mac to a couple of blocks north of Crosby.

    I’m intrigued by this. What forces the portal south of Major Mac and north of Crosby (as opposed to portals immediately north of Major Mac and immediately south of Crosby)?


  12. “David Gunn had a very practical approach. Put buses where needed. When the road can’t accommodate more, put in light rail. When light rail can’t meet the demand, then build a subway. In short, ridership determines the mode, in contrast to “build it and they will come.”

    And if we followed David Gunn’s view, our transit ridership would be dismal. David Gunn’s management style is good for one thing, and that is operating transit under a tight budget, and with the goal of only carrying people who have no choice, by cramming them into the most crowded buses you can put out.

    I find this really concerning that transit advocates in this day and age don’t understand we are not in 1950’s Toronto anymore. If you wait for demand to build on a bus or LRT train, you will never get it today. Because people today have options (cars). They are not going to crowd onto buses, and then onto LRT, and then wait for subways to be built.

    If the bus or LRT service is slow and can’t compete withe car, then they will just drive, and the demand will never be there.

    We see this in a number of places.

    Demand for Vancouver rapid transit projects were cut in half when Transit City LRT was explored, instead of Skytrain technology. Vancouver could have built LRT, but the additional ridership would never have come, because the LRT was too slow to attract car drivers in large numbers.

    The GO Train here is a great example of not creating demand. Many would say to run the trains once a hour until demand warrants more service. But this ignores the fact that people are not riding the GO Train now, because they don’t want to wait an hour for a train if they get out of a show early, or something.

    Transit must compete. It can not rely on just having riders like it did in the past.

    Toronto’s transit system competed with the car well into the 1980’s, and this showed with massive ridership gains.

    But waiting for demand to rise is not going to happen in the modern society. You have to create the demand to an extent by building an attractive service. I don’t care if it is busways, lrt, or subways. But the service has to be fast and compete with the car, as well as offer frequent service, and service at most times of the day and night seven days a week.

    Steve: The massive gains of the 80s came from population and job growth within the 416. You are confusing “competing with the car” with “good transit”. They are not the same thing.

    I could believe some of the pro-subway rants if only I had a sense that the Ford Faction believed in providing good surface transit on routes that don’t have their precious subways. Instead, they cut service because it isn’t “efficient” enough. We can’t decry the bean-counters when talking about LRT, but act just like them when looking at day-to-day operations.

    Transit competes when a bus I need to get to a subway station actually runs often enough that I don’t need a timetable, and with a reasonable expectation that I will be able to get on.


  13. If they go this route, I honestly hope they spend the extra bit and just convert Sheppard LRT. What most Scarborough riders are tired off is transferring. You can sugarcoat it all you want, with same platform transfers and the like. But it just plain sucks.

    If they go LRT, convert Sheppard and build LRT from Weston to the zoo.


  14. Honestly, this historical report kinda feels like failure. So they planned and built a subway and do everything else that was required to get ridership on the subway. No serious matching increases in density. No promotion of Scarborough and North York City Centres as centres of employment (and only limited effort at residential density in STC over the last few years). Given this history, would even LRT get the success that Miller and his ilk imagined? Seems like development plans never support the transit plans.


  15. Steve, given that everyone (except Rob Ford) says a Sheppard subway is unfeasible, why not try to integrate the Sheppard LRT rather than forcing yet another transfer on the TTC ridership. I was in Pittsburgh last summer and they run an LRT/Subway lite in their city. The vehicles are very similar to the proposed LRT vehicles. However they have one unique difference, they have two sets of doors. One set of high floor doors for stations that are underground in the core (subway portion) and other major stations along the route. And one set of low floor doors for stations that are located beside or in the median of the street.

    Using this type of vehicle we could build the Sheppard East LRT run the vehicles through the Sheppard subway tunnels and all would be, hopefully, happy no?

    Steve: Pittsburgh’s door arrangement is left over from an era when low floor cars were uncommon, and the old fleet consisted of cars for loading at street-level platforms. The door arrangement is primarily for high-platform stations and this constrains capacity when loading through the “low” door of which there is only a single narrow one per side per car. Another possible arrangement is the high/low scheme used in San Francisco with movable steps in the stairwells. In either case, this would require a special fleet of non-standard cars just for the Sheppard line. All of the new maintenance facilities are designed for low floor cars with equipment on the roof, but a high floor car has underfloor gear for which shops must have pits for access.

    It’s worth noting that the height of subway platforms is considerably greater than the existing “high floor” of a CLRV/ALRV streetcar and, by implication, of any “high floor” LRV we might procure. There would still be a mismatch between a “high floor” LRV and the subway platforms.

    I really think we should concentrate our efforts and our available capital on getting the Sheppard LRT built rather than spending a lot of time thinking about conversion of the subway. That may be a project for another day, but it is quite some distance in the future.


  16. I do concur with John about Sheppard. If Sheppard is to be an LRT, then make it all an LRT. Otherwise you are talking about least two transfers to get downtown: one at Yonge and one at Don Mills. And if it were all an LRT line, it would make it easier if the TTC ever wanted to head west along Sheppard with an LRT.

    I would like to see Transit City built. The only new subway line we need at the moment is the DRL. But the issue I see with transit is that people will use their cars if it is more practical for them, so yes a good transit system has to be able to compete with the car.

    Steve: Making conversion of the existing subway to LRT part of the package almost guarantees that an LRT won’t be built. Double transfers are commonplace for all who get downtown by taking a bus to the BD subway and then transfer at Yonge or St. George. The idea that we should eliminate a transfer at Don Mills as a high priority in the network plans begs the question of whether this is more important than other improvements including the DRL.


Comments are closed.