The DRL East as LRT

This post has been created to hold a comment thread that began under the “Poor Frozen Streetcars” about whether the DRL should be built with subway or LRT technology.

My position is clear and, as far as I am concern, this issue is beyond debate.  I am moving comments already received into this thread.

19 thoughts on “The DRL East as LRT

  1. I have often wondered why the Eastern portion of the DRL could not be just an LRT, simply the southern end of a Don Mills LRT. The Cross town has made it clear that LRTs can be underground where required, and there are railway lands etc to come west in as required. Would this not save billions, and provide better service (smaller trains higher frequency). Speed is a matter of stops.

    Steve: I am publishing here the text of the email I sent to your private query on the same topic.

    Malcolm wrote:

    Steve, I am somewhat confused as to why the Downtown Relief line must be a “subway” as opposed to an LRT. While I appreciate the desire for the capacity, I would have thought that a 4 car, 90 second headway LRT would be enough to meet the need in question into the foreseable future. This would also allow for a single LRT coming south on Don Mills, intercepting both the Sheppard and Eglinton LRTs, as well as the Danforth Subway.

    It could Cross the Don at Millwood Road, follow Broadview to Eastern and come West into Downtown. I would suggest that a fair amount of this route could be done at surface, could also go down Pape, then follow the Railroad tracks.

    Any notions. Wanted to just ask the question, as I have read many times the DRL “must” be subway, and it strikes me that 20k per hour per direction of additional capacity would go a long ways.

    I replied:

    The basic answer to your proposal is that on street operations are incapable of supporting the kind of service you propose. Even the Bloor streetcar in its heyday just before the subway opened only had a capacity of about 8-9k/hour, and it certainly was not an express service. This involved very frequent 2-car PCC trains (roughly the equivalent of one new LFLRV) on headways below 2 minutes. The streetcars more or less “owned” the streets. You are proposing trains four times as long to achieve a service design capacity of just under 20k/hour.

    The projected demand south of Danforth is well above 10k with a good potential for growth, and this is definitely not in on-street LRT range. If we had a surface right-of-way available, I would be the first to suggest that we use it (in the same manner as the Scarborough RT/LRT). However, such a corridor does not exist. Surface operations into downtown could not possibly handle this demand.

    Another big problem is passenger handling at stops. If a service carrying 10k/hour pulls up to a stop and drops or picks up a substantial load, those pedestrians go to or come from somewhere. A protected pathway for the pedestrians is as important as one for transit vehicles. (The Bloor car had a loading island in the middle of the street with direct access to the subway.)

    I think it makes far more sense to engineer the DRL as a subway between downtown and Eglinton (as described in articles I have written) because it will have to be grade separated over almost the entire route and might as well be a subway for the capacity it will require. At Eglinton would be the transition to an LRT further north on Don Mills. If the station is properly designed (not like that disaster at Kennedy for the RT), the transfer between the Don Mills LRT, the Eglinton LRT and the DRL should be a straightforward connection. This is a location with lots of room, and the design would not be constrained by fitting into a confined space with existing buildings.

    Malcolm then replied:

    The question becomes, with a dedicated Right of Way on surface, and signal priority, could you not get close. Take over centre of Broadview or Pape. Would mean reducing lanes of traffic. Perhaps, you could even skirt the valley, beside the DVP.

    To which I replied:

    The short answer is no. You would still have very large trains operating on what are very local streets where a takeover of a street simply is not possible. Pape and Broadview are only 4 lanes wide, and they are largely residential.

    Then there is the problem of what to do when the line gets downtown to the eastern waterfront and the core.

    This is simply not an LRT line.

    To finish up the thread, there are two phrases in the comment: “LRTs can be underground where required” and “smaller trains higher frequency”. The whole issue with the DRL East is that a great deal of the line will have to be grade separated and this reaches a point where 100% grade separation is probably easier than a partial arrangement. Don’t forget that the Crosstown as originally designed only had 1/3 of the mileage underground. As for smaller trains, Malcolm’s own proposal is to run a 2 minute headway of 4 car LRV trains. That’s not “smaller trains” by a long shot, and the shorter the headway, the less practical on street operation becomes without severe traffic management issues.

    We cannot just keep wishing away the difficulty of the DRL by saying “let’s make it an LRT” and then trying to shoehorn a partly surface alignment into the map. That was the mistake the TTC made with the Transit City EA that looked at some rather laughably inappropriate surface routings for the section of the Don Mills line south of O’Connor.


  2. Steve:

    “The projected demand south of Danforth is well above 10k with a good potential for growth”

    Metrolinx’s modelling of the 25-year Big Move network in 2008, put the peak hour ridership at 17,500 compared to 16,400 for the Bloor-Danforth subway! While these seem high, they do include the Don Mills LRT from Danforth to Highway 7, and unlike the DRL modelling, don’t accidently siphon off much of the demand to the GO Richmond Hill line.

    If the demand in 40-years time is anywhere close to these numbers, LRT would be woefully inadequate. The same study only put the demand on the Don Mills LRT north of Danforth at 5,000.

    Clearly subway isn’t going to go to Highway 7 up Don Mills/Leslie anytime soon. Nor is LRT going to make any sense in a tunnel downtown.

    Surely the only decision to be made is do you change between subway and LRT at Danforth, or Eglinton.

    Steve: My feeling is that the “LRT” north of Danforth will have to be grade separated at least as far north as the intersection at Overlea/Gateway, and an underground link to the Eglinton line would probably not hurt either. Building interchange capacity at Pape Station for the transfer traffic will be difficult, and the more of this that can be shifted further north, the better.


  3. The concern I would have with using a “subway” as opposed to grade separated / tunnelled LRT is that with projections of 17.5K per hour peak seems more in keeping with an LRT, which can be easily engineered to handle 20K, and subways only really make sense with projections north of this. Calgary is making similar accommodations on its C-Train by adopting a 4 car 2 minute headway. 3 Car 2 minute headway seems to be good for up to 15K per hour. So start with all stations being 4 car capable. In crush they could handle well above 20k per hour (per Bombardier 30K). The money not spent could go to extending the line West as well to the Exhibition grounds and beyond.

    Steve: The problem here is that 17.5k/hr may be within “LRT” capability, but it is very high end LRT with an almost completely segregated right-of-way. When people (including you) commenting here talk about LRT for the DRL, they propose surface alignments on streets that are only 4 lanes wide, and from which the traffic cannot be diverted to nearby roads as Calgary did with its downtown transit mall. The line would be “LRT” in name only, with no hope of surface operation.

    Just as a check on your math, for service design purposes, a 30m LRV has a capacity of about 150 (this is average load, not crush), and so a 3-car train is 450. 30 trains/hour gives 13,500/hour as a design capacity. 4-car trains would get you to 18,000/hour.

    Bombardier cites higher values by using a lower space/passenger and assuming that this would be sustained over the peak hour. We all know what that sort of loading looks like by visiting Bloor Station during any AM peak. I have had no end of people here talking about how we should build for future growth, but suddenly, the wind has shifted and I see people trying to argue that we should build a line that would have little room for this even though it is in a busy and obviously growing demand corridor. (These are not necessarily the same people, but it is fascinating to hear contrary arguments that support or oppose LRT depending on the writer.)

    Finally, with surface LRT, I come back to the issue that people have to get on and off this thing. The stations take a lot of space, and the pedestrian traffic to/from platforms is a major concern for other traffic, not to mention simply having platform space to store them. Grade separation is not simply to keep the LRVs and other road traffic separated, but also to deal with pedestrian circulation at busy stops.


  4. One of the reasons I resist subway, on a route like this is 2 fold.

    1- means forcing more transfers, and this will make rides harder for people. I think through trips are generally preferable.

    2- means a likely much lower level of service joining the Don Mills/Eglinton and Sheppard routes in future. Partly due to it being seen as a low priority, partly due to high costs with this subway segment. Also I prefer we avoid building another 40k/day subway route like the current Sheppard line and then splicing something on the end. Taking 10k/hour out of Bloor would buy a lot comfort there for quite some time.

    If there is a reasonable forecast that shows near term use in excess of 20k/hour, then subway it should be. I especially have no issues, should this be extended as subway to an alignment that meets the Crosstown LRT, as I suspect growth along this LRT will be higher than numbers I have seen suggest.

    Steve: Your concern is for the transfer between a subway and LRT wherever the junction between the two would be. If the subway ends at Eglinton, then only those riders originating north of Eglinton face a transfer compared to a through LRT line. This is a far lower number than would face crowding on the lower end of of the route whose capacity would be strained. LRT, especially if it is not completely grade separated, cannot handle the numbers projected for the lower end of the line. The comparison with Sheppard is inappropriate because that subway is operating at a capacity of roughly 7,000 per hour and the TTC has no plans to improve service.

    Please note that I believe we have exhausted this debate and I will not post further exchanges.


  5. While I know that Steve is getting, if not already completely, tired of the “DRL as LRT” nonsense, I’d just like to quickly point out that those who think it can be done on the surface along Pape should look at what happens to Pape when it reaches Riverdale Avenue as you travel south from Danforth.


  6. I find it amusing that people would propose running 400-foot long trains down old city streets just to save a transfer. There are portions of the old city where traffic signals are 200 feet apart. There are many more portions where side streets are less than 200 feet apart. That doesn’t seem to stop people from suggesting these trains could run on the surface.


  7. Malcolm Newell:

    Also I prefer we avoid building another 40k/day subway route like the current Sheppard line and then splicing something on the end. Taking 10k/hour out of Bloor would buy a lot comfort there for quite some time.

    The thing about that, Simon, is that the Sheppard Stubway would not have had to be one if Mel ‘Bad Boy’ Lastman knew and could’ve done the research on LRT; it would have been LRT for most of the route of the current Sheppard Stubway and for the rest of Sheppard Avenue. Unfortunately we know how people (especially the people who voted for him) can be, so we got what we got. That doesn’t mean that you are right in this instance, though.

    L Wall:

    I find it amusing that people would propose running 400-foot long trains down old city streets just to save a transfer. There are portions of the old city where traffic signals are 200 feet apart. There are many more portions where side streets are less than 200 feet apart. That doesn’t seem to stop people from suggesting these trains could run on the surface.

    People are tired of buses and subways both, and I can see why:

    Buses pollute, are too loud, are too small for what they offer (even with the articulated buses coming back in service this year) and just don’t feel ‘right’ to them (a better version of the ‘ghetto’ feeling that our cousins to the south have about buses), plus they need lots and lots of maintenance (which adds up over the long run even if a transit system is fully funded by provincial/state and federal governments.)
    Subways are ‘troglodyte transportation’ (as Gord Perks once said back before he was a city councillor) and the reason for that feeling is because people have to go down a flight of steps to a cave to catch a train traveling from one cave to another, instead of being able to see where they are going (the reason the TTC and other transit systems that have subways have got trains with all of the systems that tell people where and when to exit.) People want to be able to do that, and LRT/streetcars give them that feeling.


  8. I hope Steve will indulge me this comment. I find it amusing that many places Steve is accused of being an LRT fanatic who hates subways and wants LRTs absolutely everywhere. Then for the Don Mills line where he presents a strong case for a subway, many go out of their way to try to find ways to build LRT, and not necessarily even grade-separated LRT.

    As usual with this sort of thing it’s probably different people, but I still find it funny.


  9. I guess that people in the suburbs are too good for LRT and deserve subways, even if there is no financial justification for the service while latte sipping downtown pinkos don’t deserve a subway and can do with LRT. But wait, the DRL will help people from the suburbs get to their downtown jobs and won’t really help those who live and work downtown.

    These are people whose minds are made up and do not want to be confused by facts. If we could only get 1 true LRT line built then people could see what it really is and it is not the St. Clair car.


  10. I find your comment interesting

    “The whole issue with the DRL East is that a great deal of the line will have to be grade separated and this reaches a point where 100% grade separation is probably easier than a partial arrangement. Don’t forget that the Crosstown as originally designed only had 1/3 of the mileage underground.”

    I tried to make the same argument with the Eglinton/SRT line. It had been apparent for quite some time that the transfer at Kennedy was a big issue, and that the combined Eglinton/SRT was a much lower cost than the B-D subway extension. For phase 1, almost 80% would have been grade separated and for the ultimate destination of Pearson, it would still be over 60% grade separated. The 1/3 underground should be meaningless because of course Metrolinx would be smart enough to re-evaluate things (technology, grade-separation, etc.) as the underground portions began to increase.

    Also, if I look at the Don Mills LRT combined with DRL, well less than 50% would be underground (Seneca to Gateway vs. Gateway to Spadina).

    Steve: There are two differences you are ignoring here. In the case of the original Eglinton-Crosstown-Scarborough line, a chunk of that “grade separation” is the SRT right-of-way which is an LRT, not a subway corridor. Also, the projected peak demand on the route is well within LRT capabilities. By contrast, the “grade separation” for a DRL East line would be mostly underground except for the Don Valley crossing north of O’Connor. The peak demand on the DRL is beyond the capability of on-street LRT.


  11. The only way I can see a downtown line being LRT, or an extension of the Don Mills line to be more specific (one could argue that subways ARE LRTs, but let’s not make this complicated*) is if the downtown portions were grade separated and every other train short turned at Pape. Depending on the routing, you might be able to run at grade along Lake Shore, though at high frequencies it would likely create some form of bunching and/or cross traffic problems due to signalling.

    *For those who want a better understanding of what I am trying to say, read this article by James Bow.

    Steve: The problem with Lake Shore is that it is south of the rail corridor. Connection to Union and points west is much more difficult than if the line runs north of the corridor.


  12. “(one could argue that subways ARE LRTs, but let’s not make this complicated*)”

    I do wonder sometimes if anyone from Toronto has ever ridden the Boston Green Line, the Cleveland Blue & Green Lines, the Pittsburgh PAT lines, the Philaelphia “Surface-Subway” lines, LA’s Blue & Expo lines, Chicago’s Brown, Pink & Yellow lines, or even Toronto’s own “diveunders” of streetcar lines at various intersections, and on the approach to Union Station. There seems to be a madness afoot in Toronto which makes sharp distinctions between subways and streetcars — there’s a continuum here, and they’re all trains. Grade-separate where appropriate, exclusive right-of-way where appropriate, street-running where not appropriate.

    Steve: I first rode the Pittsburgh, Philadelphia and Boston systems in the mid-1960s and have always had a very strong sense of what “streetcars” could do given the chance to operate as “LRT”. The Riverside line in Boston could have been the template for Scarborough, but Queen’s Park was more interested in throwing millions at dubious high technology.


  13. Steve. This is my first time visiting your website and there is some great discussion going on. I am looking for your comments/post relating to the DRL and why it must be a subway to Eglinton at Don Mills. Can you direct me to the post?

    I trust it answers the question why we wouldn’t use LRT for full route? Why not build a surface LRT along Don Mills from Sheppard to O’Connor and a grade separated LRT from O’Connor along Pape and into downtown across King or Queen?


    Steve: This has been covered in various places, but the short answer is that the demand south of Danforth is projected to be at the high end of LRT capabilities. Conversely, demand on the north end of the route would be much lighter. With the need for grade separation at least as far north as the Don Valley crossing, and the fact that a DRL East will be built much sooner than a Don Mills LRT north from Eglinton, then Eglinton is a good place for the northern end of the subway.

    The route into downtown will not be along King or Queen for various reasons. Queen is too far north for what is now the centre of development in the core, and Front/Wellington would make a better route in than King for various reasons related mainly to where everything underground is located.


  14. Nathanael. The problem with your comment is that Toronto is not getting LRT like the ones you mentioned above. And when LRT approaches the kind as the Riverside branch in Boston, then you basically have a grade separated aboveground subway.

    This obsession with discrediting subways in Toronto is really getting ridiculous.

    Toronto is not planning modern LRT (which functions more like than a subway than any of you want to admit). And subways do not always have to be underground.

    The key here is ensuring fully grade separation and relatively far spaced stations, regardless of the technology used. And by grade separation I do not mean St. Clair style, which has no full priority, and cannot operate at high speeds.

    If you think Toronto is ever going to allow Calgary style LRT to be built along our streets, you are dreaming. Can you think of Sheppard, Eglinton, or Finch with huge medians with chain link fences, crossing arms, massive stations, etc. No, because from an urban design perspective it only works on highway style streets like Calgary has.

    Whether we like it or not, we have to accept that a lot of Toronto’s rapid transit expansion will have to be underground, aboveground (elevated), or built in railway corridors. Not going down the middle of a street and stopping at red lights.


  15. Railway corridors are the obvious ones. It can be done relatively cheap and fast. Underground lines especially suburbs cost too much for the amount of taxes people are willing to pay and the benefits are rather dubious.

    Yes I’m looking at you Vaughan and Scarborough. Once you start talking about the actual amounts of money needed to build, operate, and maintain the gold tunnels, support always drops.

    I also wonder why polls show the people screaming loudest for the underground subways that you’re talking about are drivers that don’t take transit to begin with and I surmise wouldn’t do it after construction either; Who would want to take a smelly bus to the station?

    The question is what is the purpose? Do you think everyone “deserves” the same level of service irrespective of cost/benefit? Is it a right? Is the benefit of a few in the interests of “fairness” justification enough? There certainly isn’t any economic rationale for it.

    Steve: I’m not sure that it is fair to argue that the loudest screamers for subways are drivers, and that this may somehow diminish the validity of their arguments. Yes, drivers tend to think of travel in terms of first class, relatively unimpeded service, but the only examples they have been given in Toronto are subways. Every time someone argues that “economic rationale” dictates cuts to surface transit service because we can’t afford it, or that we have to relent on transit priority because other road users, particularly trucks, might be disadvantaged at some “economic cost”, we drive people more and more to believing that the only way their travel will get serious attention is with the expenditure of billions of dollars.


  16. L. Wall. Subways don’t need to be underground in suburban areas unless required due to situations like Sheppard, where there was no reasonable aboveground option.

    The subway to Vaughan should have been elevated after York University. For Scarborough, many portions should be elevated as well.

    And don’t discredit car drivers. Yes you do need their opinion. Because if you are going to just build transit to carry people who don’t have a choice, then don’t even bother spending the money.

    Transit is supposed to be for everyone, and attract people out of their cars. It can only do that by listening to what people want. And people want FAST transit.

    A Canadian Urban Transit Association survey just found that biggest issue Canadians have with transit is travel times being too long compared to driving.

    With about 80% of TTC riders having a choice to hop in a car tomorrow, I think you should be paying attention to them.


  17. Steve I don’t believe anyone who would only accept fully grade separated transit as their only option would be too interested in the cheap and easy stuff like improved bus service frequencies, an “E” express route here or there, or cheaper at grade BRT/LRT. It’s not going to convert those people and in my opinion it costs too much for too little return to make them more than an afterthought in planning.

    The subway to Vaughan should have been elevated after York University. For Scarborough, many portions should be elevated as well.

    Building any new grade separated right of way (even if elevated) is still going to cost an order of magnitude more than at grade. One of the primary benefits of going separated is increased service capacity but most of that would be wasted in most areas of the city.

    I would rather spend money improving bus services (especially off peak) and building BRT and LRT to serve the customers in the here and now and call that bluff. Go ahead go out and drive and see how far you get.

    Steve: The people for whom subways are the only option are not the target market, but transit planning cannot simply stop all improvement waiting for a massive subway network that will never arrive. Even if we build a few subway lines, they will not handle anywhere near all of the travel demand, and bus improvements are essential to the network as a whole.


  18. I don’t think that you can look at only one option. I would need to see the cost comparison between the DRL options (realistic ones). One of the reasons for the initial question was my understanding of relative costs. I had been under the impression that even underground an LRT was notably less expensive than an Hrt type subway.

    I was looking for a way to get something past the politics of high cost projects for downtown. However Steve seems to be saying you won’t save enough to make it work. Also my basic proposal to keep it cost effective would mean using railway lands which lands you at Union. It appears that this will create a lot of future issues as a bottleneck. If you can tunnel an LRT for a lot less than a tunnel a subway (enough to give you another kick at the cat later) then it needs to be looked at.

    However Metrolinx in effect already says they will look at what is likely the most extreme version, LRT in existing rail alignments in the Don Valley if this is feasible I suspect they will push it as it will likely be politically easier.

    Based on Steve’s reaction our question would need to be what equipment can we run on that choice later to expand capacity, and what additional alignments can we access for future growth. For myself I worry that we need to look at getting the basic problem solved soon and make sure the solution is either cheap enough to be replaced or be expandable.

    Steve: There are a few problems here. First off, the TTC has consistently downplayed the importance of the DRL, and even now still talks of it only as a line north to the Danforth for “relief” in the most limited sense, not to provide an alternate route to Thorncliffe Park and Don Mills. Second, the Metrolinx demand projections for the DRL are at the high end of LRT, well above the levels projected for the Eglinton line of even for the Scarborough Subway.

    On Eglinton, a substantial portion of the line can be above grade (thinking in the context of the future Airport extension and other LRT lines in Scarborough) and the demand is easily within LRT range. Yes, there’s the central tunnel and that’s unavoidable, but the situation on the DRL is quite different with very limited opportunity for surface running especially where the demand is heaviest.

    The shopping list of routes shown by Metrolinx includes every scheme that has been floated in the past few years, including LRT on the rail corridor. This is a complete non-starter, as are a few others, but if Metrolinx just dismisses the idea without doing some analysis, they will endure floods of complaints from each proposal’s advocates that “the fix is in” for a preferred solution.

    At today’s Metrolinx board meeting and the press scrum that followed, Metrolinx was quite firm in talking about there being no one solution, and that various alternatives would likely be implemented over time. Some would be short term to provide whatever relief might be easily available, some medium term and some (notably the DRL) longer term simply because of the lead time for construction. They are very strong on GO rail taking up the challenge at least in the medium term and this means having a serious look at the capabilities of an upgraded service in the north-south corridors. We don’t even know what the maximum capacity of such a service might be yet, or how it would be fed out in the 905. That’s why we need a broad-based study.


Comments are closed.