What Does Subway Construction Cost?

An interesting article on the site Pedestrian Observations was recently linked on Twitter by Taras Grescoe (@grescoe), the author of Straphanger.  In Comparative Subway Construction Costs, Alon Levy attempts to bring together projects worldwide, adjust for currencies, inflation and other factors to derive comparable US$ values for subway construction.  The numbers are interesting in light of complaints about overpriced construction in Toronto.

The underlying problem, of course, is that no two projects are the same.  Varying proportions of underground construction, different soil/rock/water conditions, variations in station numbers/size/depth, not to mention rolling stock procurement and yard/maintenance facilities all affect the total cost and hence the cost/km.  Stir in political differences and the ease or difficulty with which projects are approved, and the number of variables is quite high.

All that said, Toronto’s costs are not wildly out of line.  This is not to say that they may not be excessive, but the cause could be our extended design and approval process and a preference for deep bore tunneling that drives up construction costs compared to shallower cut-and-cover.

The argument for LRT has always turned on the availability of a surface option where it is practical.  Whether we choose to exercise this in every possible case is another matter.

There are cases where underground construction is the only practical way to build a line, but that should not condemn Toronto to building only subways without looking at alternatives.

60 thoughts on “What Does Subway Construction Cost?

  1. I saw that article too – I thought it was very odd that Vancouver’s tunnelling costs seem so much lower than Toronto. Anyone know why?

    Steve: Cut and cover, not deep bore. Also the stations are smaller.

    Like

  2. Your post says

    “but the cause could be our extended design and approval process and a preference for deep bore tunneling that drives up construction costs compared to shallower cut-and-cover”.

    If these are the things that drive up costs, then maybe they should be changed. The approval process was changed for wind turbines – so why can’t things be changed for projects such as transit and transportation that has great benefit to the City.

    Steve: Actually, the EA process was shortened from about two years to six months, at least on paper. However, it still takes an extra six months’ work just to get to the point where there is enough info to conduct a reasonable “Transit Project Assessment”. Deep tunnels is a harder nut to crack because there is strong opposition to the upheaval that cut-and-cover construction would bring.

    Like

  3. To expand a bit on Steve’s comment about Vancouver’s recent Canada Line costs and the smaller stations, its trains are only 2 car trains, with platform lengths to match. Astonishingly short to see in person, when most other subways/metros world wide have much longer lengths to maximize the carrying capacity. As stations are the most expensive part of any subway’s construction, Vancouver’s managed to minimize this.

    Steve: There is already concern in Vancouver that they will hit capacity limits thanks to the station size even with frequent train service.

    Like

  4. I never understood the praise for the Canada Line to be honest. Low cost but everything that entails. It has station platforms that are 40 to 50 metres in length (which are pathetically small) and single track terminals (low frequencies there) with the consequences that follow: low capacity. Huzzah! The platforms are expandable to 50 metres!

    Like

  5. Steve,

    I wonder how much more disruption a cut-and-cover construction creates. Consider that stations are every kilometre and platforms are 150m – so the open excavation would be about 20% of the total length. Thus, shallow tunnel has a greater extent of disruption, but for a shorter amount of time since all excavations (by definition) are shallower. Deep tunnel, with deeper stations, has less length of disruption, but longer disruption at the major intersections – and the major intersections are the main cause of traffic delays and inconvenience.

    Like

  6. I think it is questionable why Toronto wants to keep Eglinton and Sheppard as LRT when it is planning to implement $2 billion per year transit taxes in the GTA. The taxes are supposed to fund a $34 billion transit expansion plan, although this number is questionable (it is unclear how long the taxes will be in place, possibly much longer than 17 years, and revenues will likely go up each year due to population growth and inflation though costs will go up as well). This is in addition to the $8.6 billion for Transit City, or the $16 billion for transit as a whole in the GTA (not sure where this number comes from) that is already allocated.

    This would be an unprecedented increase in the budget for transit expansion, and is comparable to the enormous budgets in cities like Paris, London, Seoul, etc. Although clearly a lot of this money needs to be put toward GO Transit expansion, the downtown relief line and improving local bus service, there is also an urgent need for relieving the extremely busy and severely congested Highway 401, and serving densely populated areas outside the downtown core. The Eglinton line should become an elevated subway in the outer sections, and be extended west to Mississauga (the 2nd largest employment area after downtown) and the Sheppard subway should be extended east (where there are numerous significant condo developments) and the downtown relief line needs to be extended further north along Don Mills to Finch.

    If we are willing to tolerate elevated construction on the outer portions of Eglinton and cut and cover construction on Sheppard and the northern part of the DRL, these extensions will be a lot less expensive. Tunnel boring is needed to reduce disruption in areas where the road is narrow (like the central portions of Eglinton or the downtown portion of the DRL) but is it really necessary on 6 lane suburban arterials where it would probably be possible to keep 2 lanes each way open during construction with cut and cover. LRT construction is much more disruptive than deep bore subway construction.

    The numbers are all available on the Big Move website and have also been discussed in other articles here. The current projects (the $16b) include:

    the Spadina extension ($2.6b),
    the Eglinton line ($4.9b),
    the SRT replacement and extension ($1.8b),
    the Sheppard and Finch LRTs ($1b each),
    the Georgetown South project, UPX and Union Station Train Shed upgrade ($1.9b),
    the York and Mississauga busways ($1.7b)

    These are quoted in uninflated dollars on various base years, mainly 2010.

    The “Next Wave” totals $22-23b worth of projects (2014$) of which the major components are:

    Two subway lines (DRL and Yonge extension) ($10.8b)
    LRT lines in Hamilton and Mississauga ($2.6b)
    Various GO improvements ($5.8b)

    That’s 15 years’ worth of the 75% of $2b/year that Metrolinx would keep, and a few years more would go to inflation over that period. The remaining 25% would go to a variety of highway and municipal projects. This still leaves part of The Big Move unfinished.

    The Eglinton extension to the airport is likely to be bumped up in priority as part of the 2016 review of The Big Move, and I can think of a few other additions. The last thing we need is to add several billions in additional cost for subway or elevated construction on Eglinton beyond the current bounds, or in replacements for the Sheppard/Finch LRT lines. There is also the small matter of higher operating costs for this type of infrastructure.

    As to condos on Sheppard, as I have written here before, subway lines in Toronto are NOT supported primarily by residential density on the lines but by the wider area feeding into major stations by surface routes. The density of condos needed to support Sheppard in their own right is vastly higher than anything that will be built along that corridor.

    LRT is NOT more disruptive than deep bore construction especially at stations where an excavation is required and at the tunnel launch sites (eg Black Creek on Eglinton).

    Like

  7. I do not like comparing transit project costs. Like Steve said, the cost of the project is based on many variables. I have seen simple streetcar projects that cost $100M/mile(!), and when you delve deeper into the costs, you see the project involved utility relocation, streetscape improvement, consultant fees, etc. I have seen LRT projects that are pretty cheap. The final cost for the Hampton Roads LRT project came in at $318M for a 7.4 mile project. Not bad for LRT.

    On the costs of tunneling, I was under the impression Cut and Cover was actually more expensive than using a TBM?

    Steve: As with all project costs, “it depends”. Cut and cover can be more expensive if there are a lot of utilities in the way, something deep bore can avoid if it goes down far enough. (Some underground work in Madrid avoided this problem by choosing an alignment that were comparatively utility-free. This is not always possible.) However, station costs can be affected by depth and by the complexity of existing underground infrastructure. Then there is the question of ground conditions: soil, rock, water each of which brings its challenges and limitations.

    Like

  8. To expand upon what others have already noted, Vancouver is seeing the downsides of the much-vaunted PPPs in the construction of the Canada Line. Future expansion will be extraordinarily expensive because costs were cut heavily by the private partner in the initial phase, shortening the underground stations to a measly 50 m and single tracking the terminus stations and their approaches.

    Steve: To which the obvious question is: why were the specifications for the line so sloppily written that this downsizing was possible? Were the project sponsors more interested in getting an “affordable” project and deliberately skimped on provision for future growth?

    Like

  9. Andrew: You constantly bring up the issue of 401 congestion. It’s been stated many times on here, and other message boards that a subway on along Sheppard will do nothing to reduced congestion on the 401. Sheppard is NOT a viable alternative to the 401. The majority of drivers on the 401 are heading to destinations outside the 416. There is a reason the 410 was originally called the Toronto Bypass.

    Regarding Eglinton, I work in Mississauga, and a lot of people commute out of Toronto to work in the Skymark area, but it’s no where close to requiring a subway. The BRT being built and the future ECLRT connection to the BRT should be able to handle future demand easily.

    Mississauga should consider adding a limited stop service along Eglinton Avenue to Ninth Line too.

    Like

  10. For Vancouver, I agree that the question is how they allowed the specifications to be so sloppily written. I doubt the cost would have been exorbitant if the stations were longer (100m).

    If minimizing cost was so important, they could have kept the short stations for all elevated stations (which would be relatively easy to expand later) and/or for all lightly travelled stations (with less passenger traffic at a station, having only some of the cars line up at the station would still provide adequate service).

    These short stations really provide a good excuse to not use the Vancouver model, even though it seems there were a lot of positives that could be taken from it.

    Like

  11. L Wall said:

    “I never understood the praise for the Canada Line to be honest.”

    From my perspective the praise comes from the fact that is was actually built, and built in its entirety. Yes, the project has some serious compromises, yes they will cost the city later, but they got a complete rapid transit line, providing quite good service for at least a decade or two, and they did it without the decades of back and forth with no real progress that have become so standard.

    Steve: And they had a hard date to hit of a major international event and had the support of senior governments, both of which are missing in Toronto. There was an attempt to use the Pan Am Games as a transit “stimulus”, but almost none of the “Pan Am” transit lines will open (or even be on a plan) by the time the games are only a memory. Indeed, the Spadina extension (which serves York U, a games venue) was on the drawing boards long before Toronto bid for and won the games, ergo no causal relationship although one was claimed until the project fell behind schedule due to a construction accident (among other things). The SRT won’t have been rebuilt, and the Sheppard LRT won’t be there either. Its proposed extension to UTSC (another venue) died with Rob Ford’s election.

    Total federal contribution to this? $670m for the Spadina line, and a $300m “commitment” for Sheppard as and when we get around to it. Even the Vancouver Airport Authority chipped in to the cost, something notable by its absence for the UPX in Toronto.

    Like

  12. “Why didn’t Vancouver built stations for their ‘normal’ length of trains (four cars as I recall)? This seems a it ridiculous in my opinion.”

    Sorry for the double post, but this jumped out at me. The short version is that the P3 model they used allowed ANY design, specifying only capacity and automation. The winning bidder came up with very short but subway sized trains instead of the ICTS style long and short, and that is what the stations were built for.

    I’m honestly not convinced the short platforms are the real problem, the automation is designed for the same kind of 90ish second frequencies ICTS/Skytrain can handle, but the single track terminals on each branch most definitely are not, and neither will be especially easy to expand (the only credible option in Richmond seems to be creating a one way loop in fact). That said, even as is it is realistically going to be quite a few decades before the platform extensions that can be done won’t be able to handle demand on the maximum frequencies allowed by terminals.

    The real scandal is that business with cut and cover; deep bore had been promised to everyone along Cambie but then it turned out that nothing in the contracts specified deep bore so naturally the contractor saved what they could and went with cut and cover. A perfect example of the sort of thing that get dismissed as a detail and then causes huge problems with badly run P3s, and quite a cautionary tale for Metrolinx. That said, I’m all in favor of cut and cover, but we can’t go promising deep bore then ripping everything up either.

    Like

  13. Not sure if you’ll want to publish another comment by me, but just dropping in with an expansion on the Vancouver P3.

    Steve:

    “To which the obvious question is: why were the specifications for the line so sloppily written that this downsizing was possible? Were the project sponsors more interested in getting an “affordable” project and deliberately skimped on provision for future growth?”

    Basically what it boils down to is that according to the documents the line wasn’t downsized. It was all specified in terms of passenger capacity, and the design built meets those specifications, which are (theoretically) on the same order as the existing skytrain lines.

    Steve: This begs an obvious question: why is there already handwringing in Vancouver about the capacity limitations of the line, specifically its stations? Translink replied in June 2010 about this issue saying that the line was (at that time) carrying a peak load of about 5k, but was capable of 15k with more trains. On a 90 second headway (40 trains/hour), that’s 375 passengers per train, on average, although individual trips will vary through the peak.

    Stephen Rees has a rejoinder to Translink’s claims, and the comment thread contains a lot of interesting info. What is quite clear is that Vancouver is constrained as much by the terms of its P3 contract as by physical limitations of the fleet and infrastructure.

    As you say, a cautionary tale for Metrolinx, not to mention for Ontario and Toronto as we will almost never know the details of whatever arrangements are included in the P3 contracts here thanks to “commercial confidentiality”.

    Like

  14. The Canada line has one glaring deficiency, lack of super elevation on the curves! How can you build a metro line that can’t negotiate curves without significantly slowing down?

    Like

  15. Not Metrolynx. Infrastructure Ontario. They advised on the P3 for the Ottawa LRT, and are advising on the K-W LRT. By the time it hits T.O., the kinks will be worked out. Just like with Presto.

    Steve: I hate to sound biased, but I have always viewed IO as a Ministry looking for a Mission. As the bloom fades from the P3 rose, I hope that this Ministry will fade from the prominence it was given under Dalton McGuinty.

    Like

  16. Andrew said:

    The Sheppard subway should be extended east (where there are numerous significant condo developments)

    An interesting little story in the news related to this surfaced recently which suggests that not only are the citizens of the inner suburbs against the idea of the intensification needed to justify a Sheppard subway, but that Rob Ford would also side with them in fighting against it. As a result, it’s becoming obvious that the Sheppard subway is just a vote buying scheme rather than a solution to our transit problems.

    Also, I’m still waiting for someone to explain to me why someone would park their car at STC, ride the subway to Downsview, and then rent a car to continue their journey westwards since that’s the only way that the Sheppard subway will make a dent with the congestion currently on the 401.

    Like

  17. Toronto Streetcars said:

    Why didn’t Vancouver built stations for their ‘normal’ length of trains (four cars as I recall)? This seems a it ridiculous in my opinion.

    Another factor is that the Canada Line is isolated from the rest of the network, so the proposal to use shorter, non-LIM trains with higher frequencies was acceptable under the expectations produced for the line.

    I suppose that the Eglinton Crosstown LRT being LRT is a good example too … it is best as LRT because the corridor has a specific set of needs…and it does not need to be a part of the Subway network in order to accomplish that.

    The DRL, on the other hand, needs to be a part of the network.

    As for the proposal for elevated subway lines in the outskirts of the city … an elevated line is costly too. Moreover, the city has no growth plans that justify full subway capacity along the entire corridor.

    I’m now thinking that if the TTC had gone with LRT for Sheppard, (and specifically not tried to connect it to the subway network) all the money used to build the track connections could have been put to extending the LRT to Scarborough Town Centre and Downsview, and things would be entirely different now.

    Cheers, Moaz

    Steve: Sadly, at the time the Sheppard line was proposed, the TTC didn’t want to have anything to do with LRT, and Mel Lastman (whose pride and joy that line was) was on record as saying “real cities don’t use streetcars”. It has taken over two decades to get beyond that blinkered view of how Toronto’s transit could develop, and then we got that cracking-great Mayor Ford.

    Going back two more decades takes us to the early 70s when, had Queen’s Park been willing to look at transit as a transportation system rather than as a tech toy, we might have actually built a suburban rail network long before we had what are now called “inner suburbs”, but then were empty fields.

    Like

  18. For $2.1 billion and a five year build span, Ottawa is getting a 12.5 km LRT, equipped with 34 Alstom CITIDIS LRVs.

    Alstom’s press release in February 2013.

    Interesting that this project will take five years to complete, while the the recently opened systems in France, of similar length. took only took two/three years.

    Steve: How much tunneling was involved in the comparative French systems you cite?

    Like

  19. French tram systems usually involve little to no tunneling. It’s rare that there’s significant demand/obstacles for a tunnel to be needed.

    I really have no way to gauge the price for Ottawa’s LRT (which really, has the capacity and design of a metro system, at least on the line as-is), but it seems awfully expensive when you consider that the majority of the line was already grade-separated. It certainly punctures the idea of building BRT first and then converting as a cost-saving measure.

    But hopefully Ottawa can use their two grade-separated corridors to make an effective system. One can imagine a sort-of Sbahn system where tram lines converge on the central, grade-separated corridors, and then diverge again at the other end. One wonders if Gatineau will ever want to get in on the action, or whether the two provincial governments will ever think for a second that things might be better if they worked together.

    Like

  20. And to highlight a hilarious comparison between the two, Ottawa’s LRT line will have 180 m platforms (120 m for outdoor stations, to start). Three-and-a-half times the length of the Canada Line.

    Really, the Confederation Line probably isn’t a good comparable to other LRT projects in terms of cost, simply because it’s being built to the standard of a metro. It’s simply being called LRT because future expansion plans envision at-grade corridors.

    Like

  21. How much would it cost to extend Sheppard to Downsview and add north-to-east and south-to-west turns so that trains can be serviced from the Wilson Yard?

    Steve: Here is my back-of-the-envelope calculation. The distance from the end of the existing structure to Downsview is about 4.3km, and so at a minimum of $300m/km, we are looking at $1.3b for openers.

    The Downsview extension was built to allow for a west to south curve from the Sheppard Subway, but not for a west to north. That connection, by the way, was for a service track to allow Sheppard trains to access Wilson Yard, not for an integrated service between the two lines which would require grade separation. It is physically impossible to have a “Dufferin West” station on the Sheppard line and run trains through onto the Spadina line. The connection would probably work something like the junction at Sheppard-Yonge where a train from the Yonge line can turn east onto Sheppard, but misses Sheppard-Yonge station in the process.

    Metrolinx estimated the cost of the Sheppard West line at $1.48b (2010$) and that’s without a west-to-north junction. For some reason, their length estimate is greater than mine, but they probably go well west of Dufferin with a station structure and tail tracks. Note also that they have only one other station (Bathurst). Add $100m or so per station if you want more.

    Some have talked about a west-to-north curve (and its corresponding south-to-east mate) to provide through-routed trains to York U and beyond, and that will require grade separation which has not been included in the design north of Downsview Station. If such a structure came in at under $250m, I would be very surprised. As with the west-to-south quadrant, it would not be possible for through Sheppard-to-York trains to stop at Downsview.

    I think we are looking at $1.7b to $2b in 2013$, more including inflation, for your proposal.

    Like

  22. Didn’t I read somewhere (here?) that a long section of the Canada Line going through a wealthy suburb was built underground rather than elevated when the residents raised unholy hell? The cost for that stretch then roughly doubled. Shades of the Ontario gas plant switcheroo.

    It seems to me that elevated is the way to go here, for the DRL and perhaps some other outlying areas. Not only is it less expensive but it’s also quicker to build, increasing the likelihood that some of us will actually get to use it while we’re still able. And the view would be better too.

    Between Rob Ford and the traffic/transit inaction, Toronto is fast becoming an international laughing stock.

    Steve: And that stretch was originally to be built with deep bore tunnel, but to save money (and because the contract didn’t specify construction technique), they changed to cut-and-cover causing a real stink.

    More generally, anyone who thinks that they are going to build elevated trains any place there is existing or planned residential land use is in for a huge and probably futile fight. Networks costed on the basis of elevated structures are fantasies.

    I look forward to your proposed elevated alignment of the DRL.

    Like

  23. I don’t think it’s entirely fair to criticize Ottawa’s high cost on the converted BRT grounds. Ultimately the cost is not coming from the already grade separated sections but from the central tunnel that has yet to be built precisely because every design ever envisioned has either been similar to the current at grade arrangement or massively expensive. We’re talking about a very deep, quite long tunnel through hard rock in a major urban centre; it’s going to be expensive no matter what. The real question is how expensive further lines are going to be, and that has yet to be seen.

    As for converting BRT in general I think the real point that needs to be made is what the timeline has been in Ottawa, and how many other systems have proceeded with it in any form; this idea of BRT as a rail precursor may have it’s place, but in very specific circumstances that need to be detailed from the beginning. My view has always been that for the most part if BRT isn’t an acceptable permanent solution you really need to be looking at a different technology.

    Like

  24. As Gauephat stated, there is next-to-nothing tunneling involved with the French systems, instead, most of these systems somehow are able to manage to convert streets to pedestrian/tram only zones.

    Clip of Bordeaux.

    Like

  25. I suppose that the TTC could run trains in revenue service from Don Mills to Downsview and back (bypassing Sheppard-Yonge station). If they used six carriage trains they could lock out the doors on two carriages … and once those Sheppard-line trains got onto the Yonge-University-Spadina line they would have space for passengers south of York Mills, who would be facing full trains otherwise.

    The challenge would be figuring out how many trains to shift from one line to the other, and do it in a way that does not disrupt or delay service.

    Steve:

    The Downsview extension was built to allow for a west to south curve from the Sheppard Subway, but not for a west to north. That connection, by the way, was for a service track to allow Sheppard trains to access Wilson Yard, not for an integrated service between the two lines which would require grade separation.

    If the Sheppard subway line was extended to Downsview, would the Sheppard-Yonge connection (which cost a ton of money…and if I recall correctly, forced the line to be cut back from Victoria Park to Don Mills) stop being used for train movements?

    Steve: Probably.

    Also, you say the provisional connection is to the Wilson Yard and doesn’t allow integrated service. If the extension were built, could Sheppard line trains access bypass Downsview but still access Wilson station?

    Steve: No. The point I was making is that this is not designed as a grade separated double track connection. West to south movements would have to cross the northbound track.

    If that were the case then TTC could potentially offer a Don Mills-Don Mills loop that would share Yonge-University-Spadina line tracks (bypassing Sheppard-Yonge and Downsview stations but transfers could be made at York Mills and Wilson).

    Again, the only real benefit is to get Sheppard line trains onto the Yonge-University-Spadina line to offer passenger relief by feeding less-full Sheppard line trains into Y-U-S.

    Cheers, Moaz

    Steve: This whole discussion is getting rather silly. Real relief for the subway problems will only come with the introduction of new services both as a new NS subway line and with very substantial increase in GO to bleed off the 905 to downtown demand.

    Like

  26. Roccinante said:

    It seems to me that elevated is the way to go here, for the DRL and perhaps some other outlying areas. Not only is it less expensive but it’s also quicker to build, increasing the likelihood that some of us will actually get to use it while we’re still able. And the view would be better too.

    Steve said:

    I look forward to your proposed elevated alignment of the DRL.

    The DRL would be underground up to Bloor and certainly an extension would be underground up to the Don Valley … but once the line gets there, an elevated line could cross the Don Valley and run along Don Mills to Eglinton, presumably without upsetting too many people since there aren’t many apartments right on the street.

    I suppose the two schools on the east side of Don Mills Road at Overlea might have some people protesting to ‘think of the children’ but I’m sure that the children would be happy to have ‘subway’ access.

    If the DRL was extended further from Eglinton then it would have to go underground north of Wynford.

    Cheers, Moaz

    Steve: That’s not very much of an elevated. Obviously the river crossing would be on a bridge, but then there’s the question of running through Thorncliffe Park rather than coming straight up Don Mills. I’m not sure an elevated station would be too popular at the Overlea intersection, let alone at Don Mills & Eglinton where it would have to connect with the underground LRT station.

    By the way, the schools are on the west side of Don Mills.

    Like

  27. “I think we are looking at $1.7b to $2b in 2013$, more including inflation, for your proposal.”

    I was looking at just extending the subway west and allowing the trains to be serviced from Wilson, nothing about integration to York U which I believe everyone and their mother used to ask about.

    So to build the extension and then the connection to travel to and from the Wilson Yard to the Sheppard Line extension, that would be $1.5b (2010$) for the line including the Downsview/Dufferin West station & Bathurst Station?

    There is no point in interlining Sheppard to any line because they run 4-car trainsets anyway. To interline you might as well make it a 6-car trainset and then tear down the walls on the existing Sheppard Line stations that cover up part of the length as that part of the platform is not used (but built in advance).

    For that $1.5b anyway, is that part of The Big Move next generation priority list along with Yonge extension and DRL?

    Steve: No. It is not in the Next Wave due to low projected demand.

    Like

  28. A rough observation leads me to think that cut and cover may work best in Toronto. It seems that this method provides the most advantages where stops are relatively close together and operates primarily under established roads. Deep tunnel looks like it could work best where the line does not follow the road above, and where station spacing is wider.

    When it comes time to build the DRL, we should really perform some studies to figure out which method would be best. While deep tunnel may be less disruptive, the cost savings of cut and cover may make this mega-project more acceptable.

    Like

  29. WRT the Canada Line – the Request for Proposals (it is a 35 year Public Private Partnership, based on a design, build, partially finance, operate and maintain model) specified certain core requirements for the system – capacity was if I recall 14,000 ppdph and it needed to travel to the airport within 24 minutes. It was up to the bidders to design a system that accomplished those parameters.

    The vehicles selected are Hyundai Rotem vehicles (3.0m) which are wider than Bombardier MKII SkyTrain cars (2.6m), and are comparable in width to other airport lines (such as JFK Airtrain (wider Bombardier MKII cars)). The wider cars provide for higher capacity than the MKII cars.

    Capacity of a line is a function of not only physical characteristics (train width, and length) but also frequency – something that Jarrett Walker has noted. (i.e. don’t get fixated on length, because if those trains only come every 10 minutes in rush hour, that’s not very good service).

    Copenhagen’s Metro has similar physical characteristics to the Canada Line (short trains, high frequency), so it is not out of line with other fully automated light metro systems.

    The RFP left it up to the bidder to propose the exact alignment and the construction method (the RFP provided it to run down Cambie, but it was up to the bidder to place the tunnels (or trench) anywhere within the corridor. The overall cost of the various bids would have been obviously a function of the vehicles, stations, tunnelling method, alignment, etc. selected. IntransitBC selected an alignment under the northbound lanes of Cambie (not under the median, because of the desire to preserve trees in the median and there is a large BC Hydro conduit under the median that would require relocation). They also designed the line with a variety of side-by-side and stacked tunnel sections depending on the section and available RoW widths. Cut and cover construction allowed stations close to the surface, reducing passenger access time and the number of escalators (ie future maintenance costs). The short platform lengths allowed cut and cover construction to remain close to the surface on the steep hilly sections (longer platforms would have required a deeper bored tunnel to keep grades between station within tolerance).

    In essence, IntransitBC “thought outside the box” on a number of items in order to design and construct a system that met the required parameters.

    For the new Evergreen Line (Bombardier MKII), the tunnel will apparently be a single bore housing both inbound and outbound tracks – that would reduce costs and construction time versus twin-bored tunnels.

    In Toronto, there doesn’t seem to be nearly as flexible a process for the design and construction of lines. (i.e. if you leave the design to the project bidders, there may be a lot more talent among engineering firms bidding on transit project than in-house within TTC).

    Steve: A few points about Toronto practices. The choice of deep bore tunnel is as much political as it is practical, although specifics for certain lines make the decision easier. It is, for example, hard to understand why the Spadina extension travels north of York University through areas where a cut-and-cover would have been possible. Crossing a few important corridors (the CNR, Hydro and the 407) would have been an issue, but not an insurmountable one.

    Eglinton is a different situation. The street has many changes of grade from east to west, and a tunnel (no matter how it is built) could not follow this topology. The result is that in some places it would be too deep for cut-and-cover construction. Eglinton is also a major utility corridor making close-to-the-surface construction more difficult than if the line were under a minor road or a field.

    There are still unhappy memories here of some of the cut-and-cover work for the Sheppard subway, and these colour all discussions of that option.

    .

    Like

  30. Steve,

    Sorry, I posted before proof reading. Please replace with this:

    If you want a bit more “silly” discussion, another idea I had was to have a Don Mills-Downsview-Union-Finch line (8′ freq), a Don Mills-Downsview-VCC line (8′ freq) and a Finch-Union-VCC line (4′ freq.)m and a Finch-Union-Downsview (short turn) line (8′ freq). These are rough peak frequencies and would change with ATC. All trains would be 6 car (eventually 7) as described by Moaz.

    To achieve this, the new Sheppard line Downsview Station would not be located under the current station, but would be located farther east of Dufferin and share the bus station (no transfers needed between these two stations due to the interlining). The line would have to go straight under the Spadina line at current Downsview Station and then an ‘S’ curve is needed to join in to the Spadina line just north of Wilson. The other ramps would make an ‘S’ curve to join the Spadina line just east of the new the Downsview Station. Only the main line would have to go under the Spadina line, all other ramps would be constructed in the open field, away from traffic, and would be built above existing lines.

    For 4km of Sheppard Construction, you actually create two Sheppard lines that are 18km and 40km in length.

    As for the distant unhappy memories of cut-and-cover construction, it is probably better than the more recent memory of nothing being built. Given the choice between cut-and-cover and TBM, they will choose TBM. Given a choice between cut-and-cover and nothing, they would love cut-and cover.

    Like

  31. Small cars and 90 second headways to meet the throughput requirements is nice in theory but as discussed here previously, dwell times have a thing or two to say about your theoretical system.

    Like

  32. Sjors:

    Leaving aside the issue of demand a would-be subway line on Sheppard West would generate, there is no operational benefit from linking the Sheppard line to Wilson yard and having the Sheppard trains being serviced from there. On the contrary, it would be a pain in the neck, operationally speaking.

    Wilson yard is already constrained in its ability to feed revenue trains onto the main line in the morning. In the early AM service build-up period, there are daily conflicts between the revenue trains that are dispatched to the mainline to build up service and the work trains returning to the yard at the end of the night shift. This is not an issue of yard capacity (which will be expanded), but rather of track geometry at the yard entrance, that limits the number of train movements per hour. On a typical weekday, the first revenue train is supposed to leave the yard in deadhead mode at around 5:25 AM and service build-up continues – on paper- for about 2 hours, until about 7:30 AM. In practice, the revenue trains are often late leaving the yard (due to said conflicts), which obviously impacts the ability to build-up service in time for the height of rush hour. As a side note, I find it particularly amusing hearing people talking about 90 sec or 100 sec headways, given that we are unable to even operate the current schedule effectively.

    In the evening, all revenue trains are supposed to “run in” by 2:15 AM on weekdays and about 2:45 AM on the weekends (when Davisville yard is closed). By the time the YUS will operate solely with the new TR trains, there will be a reduction in a number of service trains being stored at Davisville overnight (because the existing yard is designed for 2-car married pairs, not 6-car TR train consists), with a corresponding increase in the number of trains being stored at Wilson overnight and dispatched on the mainline in the morning. This will reduce the night time maintenance window (with a corresponding impact on overnight work) and will lengthen the service build-up period in the morning. This problem is likely to get even worse after the Spadina extension opens for service, especially since the track layout at the yard entrance is designed in such a way that revenue trains can only be effectively be fed southbound on the Spadina line (a northern connection from Wilson yard to the mainline is under construction as part of the Spadina project. We’ll see how effective it will be and whether it will foul up service in the morning even more than it is now).

    So given these operational constraints, the last thing we need is to feed extra trains from Wilson on the Sheppard line. Sheppard operates quite effectively now – essentially as a completely separate line. There is no need to return the trains to the yard every evening to be serviced. Basic maintenance is done overnight in the tailtracks at Sheppard Yonge. The same can be done if the line is ever extended further west to Downsview or wherever. No need to also foul up service on Sheppard every day (it’s fouled up enough on the YUS already). When trains are due for pit inspections/wheel grinding etc. they can be sent down to Davisville which is 10 minutes away. And that’s not a frequent occurrence anyway.

    Like

  33. Ok, this idea is obsolete now that the Spadina line is being extended to Vaughan, but how about this:

    Extend Bloor-Danforth north along the RT alignment, then West along Sheppard to Don Mills. Extend Sheppard West to Bathurst, then curve north and connect in to Downsview station.

    Voila! The Yonge-University-Spadina-Sheppard-Danforth-Bloor line!

    I call it “The Pretzel”.

    How much does it cost? Who cares! It’s a pretzel!

    Steve: There is a bizarre fascination with loops and, er, pretzels, that is very hard to understand. If we’re really serious, given recent announcements, maybe we should have a loop that includes downtown, Pearson Airport, Vaughan (everything has to go through Vaughan) and Pickering Airport.

    Like

  34. “I look forward to your proposed elevated alignment of the DRL.”

    Very good of you.

    The existing rail line running northeast through Riverdale would be ideal. Commuter rail and the El share ROWs in Chicago, a city with which we have some things in common. Why not here? Strictly speaking it wouldn’t be elevated, other than on bridges passing over Queen, Dundas, etc., but it would be grade or otherwise separated from street traffic, which would benefit everyone.

    It could turn north and run through the Greenwood yards to link with the Danforth line and perhaps continue northwards underground from there. Or it could continue along the rail line towards Scarborough where there are a variety of route options. Using one of the less commercial east-west downtown streets such as Richmond or Adelaide would be a good option there.

    Steve: The route you describe is more or less the one proposed 50 years ago and which I generally have supported. However, what you describe would not be elevated, certainly not through downtown. On one hand you are extolling elevateds, but then you propose an alignment that is mostly not.

    Like

  35. Ben Smith said:

    While deep tunnel may be less disruptive, the cost savings of cut and cover may make this mega-project more acceptable.

    The question with that is whether cut and cover is really cheaper than using a TBM when dealing with bedrock; especially when you are unable to use explosives.

    Like

  36. Steve:

    There is a bizarre fascination with loops and, er, pretzels, that is very hard to understand. If we’re really serious, given recent announcements, maybe we should have a loop that includes downtown, Pearson Airport, Vaughan (everything has to go through Vaughan) and Pickering Airport.

    While we’re on the strange and the terrible, want to put anything on how long it will be before we start hearing about “extending” the ARL to Pickering Airport? Every site plan I’ve seen definitely has provision for a rail spur to the terminal site.

    Steve: I cannot help thinking that one reason Jim Flaherty is trying to revive Pickering Airport is to assist in getting his proposed GO service to Peterborough “off the ground”, so to speak. A rail link across the top of the city will not be easy, but before we start building it, let’s see if the current government and Minister last long enough to launch the project.

    Like

  37. Re: Timur Urakov on Sheppard West

    What you say about the need for Sheppard trains to access Wilson is entirely true, but my understanding has always been that the really attractive thing is not Sheppard itself, but in the ability it creates to get trains from Wilson quickly over to the northern end of the Yong line. Particularly if whatever yard we build on the Richmond Hill extension is limited in size (pretty likely seeing as it looks like they want to put it underground) this could be quite a major improvement.

    Steve: The problem of logistics at Wilson would remain even with the Sheppard link and would be compounded by the need for Yonge trains to cross over the Spadina line to reach the Sheppard connection. The proposed yard in Richmond Hill has space for twelve trains.

    Like

  38. Steve:

    This whole discussion is getting rather silly. Real relief for the subway problems will only come with the introduction of new services both as a new NS subway line and with very substantial increase in GO to bleed off the 905 to downtown demand.

    Which the city and province and GO have not managed to do in 30 years … 50 if you just look at DRL-ish proposals.

    I’d say that I’m in good Toronto company.

    Steve:

    I cannot help thinking that one reason Jim Flaherty is trying to revive Pickering Airport is to assist in getting his proposed GO service to Peterborough “off the ground”, so to speak. A rail link across the top of the city will not be easy, but before we start building it, let’s see if the current government and Minister last long enough to launch the project.

    I’m curious about Flaherty’s outright opposition to the revenue tools for the Big Move and Transit expansion. Initially it looks like he is being ideologically-driven, but when we realize that he is MP for Whitby-Oshawa even more questions are raised.

    Politicians love to be able to cut ribbons … does the Big Move not offer enough ribbons to cut in Whitby-Oshawa? The Durham Pulse brt-lite begins operations on June 29 … maybe an invitation to the ribbon-cutting would help a little?

    Cheers, Moaz

    Like

  39. That’s the first time I’ve seen a diagram for the storage facility in Richmond Hill … the dotted lines beyond the storage facility for “future extension” is interesting. I’m still deciding if I should laugh or cry. I always thought Steve was joking about the extension to Barrie.

    Sadly it is completely plausible we will see a “Barrie Relief Line” long before a DRL, Eglinton extension to the Airport, or SMLRT (sigh). I have to thank Pasternak for starting this whole “insert suburban community with barely any local bus service” Relief Line crap.

    Steve: There have been rumblings about an extension to Major Mac and construction of a new yard there. I don’t know whether this has fallen off of the map, and a big issue would be whether York region wants to foot the bill for 1/3 of the extension.

    Like

Comments are closed.