A Few Questions About Eglinton-Crosstown (Update 3)

Updated July 4, 2011 at 11:10 am:

Recently, the Toronto Star reported that Metrolinx had claimed that the travel time by underground LRT from Kennedy Station to Jane Street would be 25 minutes.  They have now confirmed that the correct figure should be 35 minutes.

Also, the full presentation given at the recent board meeting regarding the Eglinton line is now available online.  Only the station design portion had been posted originally.

Updated June 26, 2011 at 10:00 am:

The June 23 presentation on the Eglinton LRT line contained a great deal of interest.  In addition to the presentation on station design available from the Metrolinx site, staff gave us a new ridership projection based on the “all underground” alignment of the route.

Total ridership in 2031 is projected to be 30% higher than with the surface/subway LRT design primarily because of the higher speed and the availability of a through route from the Scarborough leg to the Eglinton line. Most of the increase comes from shifted riding from the Danforth subway to Eglinton, with a minor contribution from auto-to-transit migration. Riding on the Scarborough leg drops in the projection due to the line’s termination at McCowan and the absence of a through connection to Sheppard East.

Peak point ridership doubles from about 6,000 to 12,000 passengers per hour westbound to Yonge-Eglinton station.

This chart also plots flows on other parts of the network, although no numeric values were supplied in the presentation. I have requested this info so that the full implications of the chart can be better understood.

The chart asks more questions than it answers:

  • The demand westbound from Kennedy is about 75% of the demand arriving at Yonge, although there may be some turnover along the route. Details of this would indicate the degree to which the new network is simply funneling traffic from northeastern Scarborough to the Yonge Subway as opposed to providing service to many points in between.
  • Placing a scale on the chart to estimate the values on other network segments, notably the Yonge Subway southbound, the values appear to be in the same range as 2011 demand, or possibly lower. A reasonableness check is needed on the projection as a whole, and these values also feed into the debate about the need for added YUS capacity and/or a Downtown Relief Line.
  • The projection leaves the Sheppard Subway in its current configuration, and we need to know what the flows would look like with the proposed extensions to STC and Downsview.
  • It is unclear where the Yonge subway ends in the model and whether the projection includes the effect of a Richmond Hill extension.
  • The role of GO Transit as part of the service carrying riders from the outer part of the 416 to downtown needs to be explored. Metrolinx plans include service on the CPR through northeastern Scarborough, and this could be an important alternate route for riders travelling to downtown.

Metrolinx is also considering the implications of the Eglinton corridor for major GO Transit interchanges, aka “Mobility Hubs”

Although the text on this page acknowledges the difficulty of a connection to the Richmond Hill GO service in the Don Valley, it completely misses the opportunity for a hub at Leslie Street where the CPR crosses Eglinton Avenue. The planners appear to be considering the network and service levels as they now exist on GO rather than the likely status by 2020 when the Eglinton line will open. This is echoed by the lack of “bumps” in the demand projections at locations where transfer traffic to/from GO might be expected.

The report on station design provoked some discussion.

The proposed locations were chosen based on ridership potential, spacing, access and development potential. Although the “Mobility Hubs” map shown earlier implies a station at Black Creek, this is actually replaced by a station at Weston linking with the rail corridor.

Stations will be 125-130m long of which 97m will be the platform and the balance for utility functions, notably fan shafts. Building codes now require considerably more fire safety provisions in underground stations and this adds to the size and cost of the structures. Work on preliminary plans for five stations is underway. These will be presented at public consultations in fall 2011, Feb/Mar 2012 and May/June 2012 as they are refined.

The tunnel depth will range down to 30m. Although a preliminary design for the line to pass under the Don River has been prepared, Metrolinx is looking at an alternative crossing on a bridge between Brentcliffe and Leslie. This is expected to save at least $100m in construction cost, and avoids a need to change the location and depth of Laird Station to accommodate the approach to an underground crossing. The bridge would lie to one side of Eglinton rather than the originally proposed street-running in the middle of a widened roadway.

There was no mention of how the east branch of the Don will be handled between Wynford and Bermondsey Stations. This would also be the segment where any interchange with the Richmond Hill GO service would have to fit into the plans.

With the change to fewer stations, the issue of surface bus service came up again. Metrolinx is playing rather coy on this subject saying that this is a TTC decision, and the original Environmental Assessment for the line foresaw no local bus service.

This may seem to be a relatively minor issue, but it is an example of Metrolinx’ abdication of responsibility for local transit service. Many people will be affected if their existing short walks to a bus stop are replaced with long, possibly hilly walks to a rapid transit station. This comes up at every public meeting, and ducking the issue simply provokes resistance to the project.

[I will declare an interest in this matter as I regularly use a local stop at Banff and Eglinton, halfway between Mt. Pleasant and Bayview Stations, near a family home.]

Operation of the line will be fully automated with an onboard crew (like the SRT) as monitors, although the cars will have manual controls for use in any open sections such as yards and future surface LRT construction, not to mention during failures of the control system.

The projected 12K peak demand (in 2031) can be handled with a two minute headway of three car trains, and this includes some headroom for surges. The automated control system will allow headways down to 90 seconds, and this provides room for growth. Metrolinx and TTC are working on operating cost estimates, but these are not yet available.

Finally, the project now has a new official “branding” as the “Eglinton-Scarborough Crosstown” route with emphasis on “Crosstown”.

The typeface chosen for this is claimed/intended to be similar to the existing “Toronto Subway”. The resemblance is only superficial, and saying that it “looks like” the original does not make it so. Note particularly the difference in the “R”.

Update:  Here’s what the logo looks like in Toronto Subway.  (Thanks to Jonathan Chen)  Note to Metrolinx:  Don’t claim you are trying to match the old typeface, use it.

The name provoked some amused questions about how this would fit into a world including Etobicoke, home of Mayor Ford who is responsible for tearing apart the Transit City plan. Presumably by the time we actually have to worry about this, the words “Eglinton-Scarborough”, already a light gray, will have faded from view.

Original post from June 12:

Two major Metrolinx projects — the Eglinton-Crosstown LRT and the Georgetown South rail corridor — have had their share of road shows in past weeks, and this will continue through early summer.  Metrolinx does feel-good publicity almost too well, but that’s not the sort of thing community groups particularly want to hear, especially when they’ve heard it all before.

Some of the knottier questions have a political edge, the ones the hapless staffers at public meetings cannot possibly answer, the problems that even the professional politicians and Metrolinx board and management will duck.  Some are technical questions that should have an answer, but they are either evaded or answered inconsistently.

Here’s a grab bag from recent events along Eglinton.

Where Will the Stops Be Located?

Metrolinx dodges this question every time it comes up because they are still working on the design for the outer parts of the line (west of Keele, east of Laird).  However, they also repeat that there will be up to 26 stations.  It doesn’t take a planning degree to make a good guess where these will be.

First, we need to look at the original proposed route map for the Eglinton LRT.  Between Jane Street and Kennedy Station, there are 27 stations including the terminals.  Tacking on the SRT adds 5 more for a total of 32.  This means that getting down to 26 will require the elimination of 6 stops.  Here’s an attempt at picking them from west to east.

  • Oakwood.  This stop is notable by its absence from a list of stations where design work is already in progress, and it is also fairly close to both Dufferin and Eglinton West.
  • Chaplin.  This was a proposed site for extraction of the tunnel boring equipment (one set coming east from Black Creek, the other west from Laird), but it is unlikely to survive as a station location without major redevelopment.
  • Leslie.  There is little at Leslie and Eglinton to warrant a station unless GO begins operation of considerable service on the CPR line to Agincourt and beyond.  “Considerable” is not two trains each way Monday to Friday.  Will things be different in 2020 when the Eglinton line finally opens?
  • There were 9 surface LRT stops between Don Mills and Kennedy, of which 4 are at major arterials.  The Ferrand/Wynford stops will likely be consolidated, probably on the Wynford side of the DVP.  A consolidated Victoria Park and Pharmacy stop with an off-street bus loop could also make sense, as part of redevelopment of the “Golden Mile”.  Further east, there will be some debate about the “in between” stops given that the N-S arterials are about 1km apart, but most will not survive financial and ridership review.
  • Ellesmere.  This is the most lightly used station on the SRT because it has an extremely difficult walking transfer connection to the 95 York Mills bus.  Parachutes and trampolines would be required to speed up this process.

This gives us more than 6 candidates, and I suspect that Metrolinx will attempt to keep the cost of this project under control by trimming as much as possible.  This issue needs to be discussed openly sooner rather than later, and relates to the next question.

What Replacement Bus Service Will Operate?

When a good chunk of the LRT was to operate at grade, this question only affected the central, tunnelled portion from Laird to Keele.  However, the issue is more complex now that the line will be underground.  Walking distances between stations will be longer than to existing bus stops, and Eglinton is not the flattest street in Toronto.  Some stations will be quite deep, and the vertical access adds to the travel time to or from a vehicle.

East from Yonge, there are many routes, but the most frequent services are 34 Eglinton East, 54 Lawrence East and 100 Flemingdon Park.  Others run infrequently if at all.  To the west, there is 32 Eglinton West briefly supplemented by the two Avenue Road services, when they run.  The TTC would save much grief at public meetings by simply stating that there will be a surface bus that would operate (say) on no worse than a 15 minute headway.  This would be comparable to the 97 Yonge bus in the “old” section of the city south of York Mills.  (I would argue that all of the parallel-to-subway routes should be improved as an accessibility measure, but that’s a topic for a different post.)

This is an example of a jurisdictional split between Metrolinx (responsible for building the rapid transit line) and the TTC (responsible for operations).  As Metrolinx/GO expands out from the core, they will produce new demand patterns that local transit systems may be ill-equipped (and funded) to address.

How Will the Line Cross the Don River?

Although the Metrolinx/Ford Memorandum of Understanding has an escape clause for surface construction at the Don River, various project staff have claimed that the line will go under the river.

From detailed drawings in the EA for the LRT proposal, we can see how much different the grades and elevation (depth underground) the line would have to be to get under the river.  Five pages from the EA are discussed below.

  • Plate 68 shows the proposed portal east of Brentcliffe.  This shows the tunnel alignment (near the bottom of the drawing) and the declining elevation of Eglinton Avenue as it drops into the valley.  Note that the vertical and horizontal scales are not equal, and the grade looks much steeper than it actually is.  A 5% grade is the maximum that would be provided for LRT, and for full subway 4% is preferred (the grade from Summerhill to St. Clair is 4%).
  • Plate 69 shows the bridge over the west branch of the Don at Sunnybrook Park.  Note the additional depth of the river valley itself below the elevation of the bridge deck.
  • Plate 70 shows Eglinton rising out of the valley and passing below the CPR viaduct.
  • Plate 76 shows the Wynford station at the west side of the east branch of the Don.  The plan does not show the elevation of the river itself, but the drop off into the valley is visible.
  • Plate 77 shows the east side of the Don.

There is no question that a tunnel is possible, but going under two rivers adds substantially to the cost, and it affects the depth of nearby stations which may have to be further underground to keep grades down to the river crossings within allowable limits.

Going under both branches of the Don is an expensive decision, and at the very least Metrolinx should keep the options open until the cost tradeoffs are understood.  Such a cost saving on Eglinton presents Mayor Ford with a dilemma:  making Eglinton cheaper to build frees up money that could underwrite the Sheppard subway.

Underground Through Weston

During the Eglinton LRT EA, the TTC and Metrolinx were immovable on the subject of an underground line through Mt. Dennis, the section on either side of Weston Road.  This produced local resentment in the vein of “North Toronto gets a tunnel, but we poor folks get stuck with surface operations and community disruption”.

The TTC produced plans for an underground alignment, but rejected this option to keep costs down.  Here are excerpts from a January 2010 presentation.

  • Recommended Weston Stop:  This shows the arrangement for a surface stop on Eglinton west of Weston Road (the location is dictated by space constraints of the underpass to the east).  A common centre platform is shared by both directions of travel.
  • Recommended Alignment — Property Impacts:  Because Eglinton is narrow where the platform would be located, there are extensive property effects notably for a block of houses on the north side of Eglinton that would be demolished to allow for road widening.  To say that this is unpopular would be an understatement.
  • Underground Alignment:  This shows a variation with the line passing under Black Creek, the rail corridor and Weston Road underground.  Whether both Black Creek and Weston would actually get stations given the cost and likely demands at each point is hard to say, but this illustrates the general placement of the line.
  • Underground Alignment — Property Impacts:  Like the surface scheme, this design requires demolition of houses on the north side of Eglinton.  However, much of this is due to the presence of a storage track west of the station which is intended for use as an emergency turnback.  This was a clear example of the TTC adopting a design that would skew the choice to their preference by making the demolition of homes common to both schemes.  The station has a crossover east of the platform.  This is quite adequate for emergency operations, or it could be dropped completely as there will be a crossover at Keele which also has a surface bus loop, and there would also be a crossover at the terminal, Jane Street.

The alignment through Mt. Dennis has been a delicate issue for some time, and Metrolinx would be wise to begin addressing this as soon as possible.  Delay only feeds a suspicion that the line will never get past Black Creek or, at best, a connection to the Weston rail corridor.

Service Level

In response to a question about the level of service riders might see, a Metrolinx official replied that there would likely be 3-car LRV trains running every 6 minutes at peak, 12 at off peak.  Metrolinx needs to understand that “rapid transit” is not the same as a GO train every hour.  On a route where the average journey will likely be under 20 minutes (few will make the 45 minute trip from STC to Jane), long headways will contribute substantially to the trip length and to a perception that service is less than adequate.  There’s a reason why the TTC runs subway trains every five minutes whether they are needed or not.

If this was an off-hand remark, it needs to be rethought, and quickly.  Odd, indeed, that Metrolinx can muse on service levels in the Eglinton LRT subway, but not for what is almost an afterthought, a parallel surface bus.

Future Airport Service

After spending $8b to keep Mayor Ford happy by burying the entire line, we will still only be able to ride to Jane Street at best.  Extending the line to the airport is a task for the next decade, presumably when this can be done on the surface without interference from the Mayor’s office.  There will probably be a variation on Kipling Station’s Airport Rocket running west on Eglinton for many years.

Metrolinx still speaks of both the Eglinton and a future Finch line reaching the airport, but this is not going to happen using LRT in the current political climate.  Meanwhile, we will have the Air Rail Link starting in 2015, but this is still to be a premium fare service and will be of little value to the most important group of potential riders, the people who work at the airport.

It’s worth noting that the ARL will provide about 18,000 daily seat-trips both ways from Union to Pearson (with a stop at Dundas West and possibly a station somewhere in the Weston area).  They will not all be full — that is the nature of all-day transit service.  (That’s 126 seats/train, 4 trains/hour, 18 hours/day.)  The Airport Rocket already carries over 4,000 passengers daily to and from the airport, and more will eventually ride along Eglinton.

Metrolinx needs to start thinking about the airport beyond the scope of the ARL, and for riders whose trips do not originate downtown.  Service may be provided by the TTC and any number of 905 transit agencies, but the “mobility hub” at Pearson should not grow haphazardly depending on the local funding and service decisions of each municipality.

119 thoughts on “A Few Questions About Eglinton-Crosstown (Update 3)

  1. Why does everyone thinks the Eglinton line will be cancelled like it was in the 90’s. Its a different era folks, it would be political suicide for Hudak to cancel the line.

    Steve: If Hudak plans to get elected in the 905, he won’t care what he does with Eglinton. Moreover, he could run on one platform and then “discover” that he couldn’t afford to build the line afterward. When you have a majority, you can ignore a lot of people, even if only 40% of them actually voted for your party.


  2. Regarding the recurring idea that Hudak will “Harris” the Eglinton Crosstown line: We must ensure that this is an election issue.


  3. Why does everyone think the Eglinton line will be cancelled like it was in the 90′s?

    Because it does not make sense. The demand is not there for a subway, LRT will not benefit shoppers in the area east of Keele and the uproar from the construction mess will drown it. Remember St.Clair!

    A better choice is a relief line into downtown along with LRT lines above ground on wider streets such as Lawrence or Steeles. Eglinton is slow speed, frequent stop route with lots of residential and retail along much of it.


  4. @Ray Kennedy

    “The demand is not there for a subway”

    …and therefore, we shouldn’t make any upgrades to transit? Even though the TTC’s numbers destroy the argument for subways, the TTC’s numbers also clearly state that future ridership would be too much for any Bus-related technology to handle.

    Too little for subways. Too much for buses. That’s why they also recommended LRT.


  5. Jacob Louy said:

    Regarding the potential for an LRT network for Richmond Hill, wouldn’t this be possible without the presence of the Yonge Subway? Couldn’t a Richmond Hill LRT network connect to the subway within the City of Toronto instead?

    Too many people think that the best way to develop is to extend the subway. That is what probably would have happened if Bill Davis had decided to extend “Metro” out into the suburbs to include Port Credit, Cooksville, Malton, Woodbridge, Thornhill, Richmond Hill, Markham and Pickering – and I can imagine that the development situation would be terrible and the TTC would be totally bankrupt.

    We did not end up with extended subways and a regional ART system (or regional busway) running along highway and hydro corridors. Instead, we ended up with “mini-Metros” (not the trains, but the regional municipalities) surrounding Toronto, wide roads to support cars, and potential for independent LRT / BRT systems along these roads (if the politicians can open their eyes to the possibilities and face the potential backlash).

    YRT has developed the VIVA system along the main corridors in York, which might be LRT one day. Brampton has introduced Zum buses and Mississauga has introduced the MiExpress service.

    Finally, Mississauga and Brampton have decided that the Hurontario corridor needs to be LRT from downtown Brampton to Port Credit, and that it they should ask the Ontario government to push forward with the project.

    We do not need one massive GTA transport authority that will bleed our funds dry. What the Ontario Government needs to do is facilitate these local improvements and expand GO to connect the inter-regional connections. They are doing some of these things, but they need to do it faster.

    Jacob Louy said:

    Too little for subways. Too much for buses. That’s why they also recommended LRT.

    Eglinton was originally supposed to be a busway extending from Eglinton West out to Mississauga to meet the Transitway and connect to the airport. Then it somehow became a subway. Then someone realized that Eglinton had developed and it would make a great cross-town route (from, say, Keele to Laird). But someone else realized that the cost of a subway would be astronomical and an LRT would save enough money to pay for a full cross-town route.

    So LRT made for a good compromise on the passenger capacity side, but more importantly, it allows the line to be built longer and faster than if it is just underground or if it is a subway.

    The needs and challenges are different for the suburbs and for Toronto but the message is pretty clear – LRT is the best choice whether you are looking for a compromise in capacity or extended flexibility or more km of rapid transit for your buck.


  6. I think we need to clarify something here when the word “rapid transit” is used.

    People love using the word “rapid transit” when describing VIVA, ZUM, and in the middle of the road streetcar service like Spadina, and the Transit City lines.

    These services are not “rapid transit”. They are upgraded local transit service, or “limited stop” services. But they do not have full grade separation that is required to truly make something rapid (aside from a point to point express bus).

    My friend brought this up the other day and I thought it made total sense.

    So your comment that LRT is good because you get more km of rapid transit for your buck, does not address that the LRT we are getting for that buck is not true rapid transit.

    Now if it was LRT like in Calgary, then it would be a totally different story and classified as rapid transit.

    Anyway in terms of capacity, I don’t know how many riders you guys need to see that Eglinton is justified as an underground LRT for the whole route. 300,000 daily riders is a lot, and that actually signals that this line should probably use subway trains, as the LRT trains are going to be full from day one.

    Like everything in Toronto, ridership will be vastly higher than projected.

    At 300,000 daily riders, the Eglinton Crosstown will be busier than most subway lines in North America, aside from NYC subway lines. It will be busiest than the entire MARTA system, and will have over half the ridership of the Chicago “L” system. It will also carry over half the ridership of the Boston subway, and just under the ridership of the Philly subway.

    Not bad for one line huh?

    Steve: Yes, especially impressive for a line with a peak point demand of 12k/hour. The projected annual trips are 100-million, and this would usually be converted, for busy urban routes, to a daily figure by dividing by 300 (weekends count as one day). That gets us to the 300k/day figure (for the benefit of those who might wonder about the math).

    Eglinton is a busy route with many over lapping, relatively short (compared to the route in total) trip patterns. It’s not a commuter line with only inbound peak traffic, and the trips don’t all want to travel the full length of the route. For a more conventional route, we would take that 12k, double it for the AM peak, double it again for the PM peak, and double it a third time for all-day riding. That would barely get us to 100k riders, but the higher number comes because we are linking (at least) three routes together — the SRT, Eglinton East and Eglinton West — not to mention bidirectional flow.

    But what happens if the Eglinton demand could be intercepted by a DRL at Don Mills? What would the peak demand look like intercepted at Don Mills? That’s why I want to see network projections, not just one line’s version. Where are those riders supposed to fit on the Yonge line at Eglinton? If they are currently BD riders, the only saving grace is that we will shift the transfer problem north from Bloor to Eglinton, but we won’t get rid of it.


  7. Steve said …

    “That gets us to the 300k/day figure (for the benefit of those who might wonder about the math).”

    When the Keele-Woodbine Bloor-Danforth subway opened in 1966, its daily ridership was only 131,000 passengers per day. That’s right. A full six-car 500 foot platform subway line for 131k per day.

    Now we are bulding Eglinton as a three-car LRT subway (with 300 foot platforms) for a projected demand of 300,000 passengers per day?

    I want to go back to the 60s. At least the planners in charge back then knew what they were doing.


  8. If I understand the political shell games that have followed this line for the past couple of years correctly, what has essentially transpired is that:

    1) A crosstown rapid transit line was originally envisioned for the Eglinton corridor with Metrolinx favouring ICTS technology so that it could be integrated with a rebuilt Scarborough RT (supposedly justified by ridership projections at the time. The books may have been cooked though, it’s debatable)

    2) The Miller team behind Transit City worked out a compromise deal that featured a Stadtbahn-like line with both sub-surface and surface operations at different parts of the route, to free up funding for other Transit City projects. This line attracted fewer riders but also represented a lower capital investment.

    3) With the political changing of the guard, we’re basically back to square one, with all the other Transit City lines dumped. And we have a mad scheme to funnel more money into the black hole that is the Sheppard line, supposedly based on gullible private sector ‘investors’ and fairy dust.

    Essentially this project is now going to be what most people would consider a “subway” (i.e. a fully grade separated, mostly subterranean line) albeit with capacity less than what we’re used to on the YUS/BD lines in Toronto, what might be called a “Light Metro” in other places. I wonder how much different the costs really would be between LRT and HRT given the same amount of tunneling (representing the vast majority of the capital cost) will be required? If these numbers are to be believed (and I like you Steve await more details), they seem to be well within the realm of HRT.

    More importantly than what ridership technology this line uses, however, is the fact that with all this focus on Eglinton there seems to be ZERO discussion ongoing about the important DRL East except in internet forums like this one! Steve, do you know of any updates on the TTC’s Downtown Rapid Expansion Study? The TTC page for it hasn’t changed in months and I have heard little mention of it in the news media.

    Steve: The TTC’s 2012 capital budget presentation explicitly states that a DRL isn’t even in the cards until all of the options for added capacity on YUS have been exhausted. I believe this is extremely short-sighted as it assumes that the capacity gains the TTC claims can actually be achieved on a reliable basis. Bluntly, I don’t believe them, but then I am not an engineer and am not “qualified” to comment. The TTC completely misses the fact that a DRL east would add a completely new north-south corridor to the transit system and would benefit travellers generally to the near eastern part of the city. It’s not just a “relief” for Bloor-Yonge. However, their insistence on looking only at the southern part of the line more or less guarantees that its potential benefits are never examined.

    Politically they can get away with this because any spending on a DRL would be seen as a “downtown” project and it would take a back seat to things like Sheppard which will carry far fewer riders.


  9. Even with comparable station spacing (400-600 metres), the Bloor Danforth Line travels at an average speed of 29 km/hr, 4-7 km/hr more than a surface line with the same station spacing.

    Is this difference in speed all due to red-light delays? Considering how the Bloor Danforth Subway barely makes it past 50 km/hr between stations, a surface line could travel almost as fast (provided that the road speed limit is indeed 50 or 60 km/hr).

    Is it therefore possible to implement very aggressive transit signalling priority so that LRV’s will rarely encounter red light delays (thereby making a surface route almost as fast as a subway of similar station spacing)? Has this aggressive implementation been done elsewhere in the world, and what are the impacts on traffic?

    Steve: Toronto’s signal priority system is designed from the premise that transit gets fitted in when possible, not that it has absolute right of pre-empting other traffic. None of our streetcar lines has stops far enough apart to be comparable with an LRT implementation, although over the past decade or so, the growth of signals at intersections between stops has produced additional points of delay. The detection system used in Toronto treats each intersection as independent, and with close signal spacing, an arriving streetcar is detected so close to some intersections that they cannot react in time. Moreover, nearside stops can foul up the design because the streetcar burns up green time serving customers.

    When people compare systems and signalling schemes, it’s important to be aware of the differences in implementation (mixed traffic vs right-of-way), stop spacing and the design policy on giving transit a clear run through cross-streets between stops.


  10. “The projected annual trips are 100-million.”

    Assuming the annaul trips on the entire system will be 500 million, that’s one out of five trips on the TTC involving a leg on the Eglinton/Crosstown/SRT.

    I must say that this seems high. Even counting ridership on the ex-SRT section of Eglinton, I have a hard time imagining one out of five (or one out of six, or one out of seven) TTC trips involving Eglinton. These kinds of numbers just don’t make sense to me.


  11. M. Briganti says, “When the Keele-Woodbine Bloor-Danforth subway opened in 1966, its daily ridership was only 131,000 passengers per day. That’s right. A full six-car 500 foot platform subway line for 131k per day.

    What is the source of this number? Is it shortly after opening? What was the ridership in 1967? 1968? 1969?

    I can certainly see ridership being light in 1966, as the line was pretty short and people were do doubt used to riding parallel services, of which there were plenty. (More than parallel services on Eglinton.)

    What was the planned ridership of Bloor-Danforth? You’re comparing an opening-day ridership with a planned end-state ridership for Eglinton.

    Furthermore, I don’t understand why people are trotting out the “300,000” figure as if it was gospel truth. One out of five trips on the TTC involving a stretch on Eglinton?

    As far as I’m concerned, the 300,000 is a fantasy number. Let’s look at these 2009 ridership numbers:

    Eglinton East: 26,300
    Eglinton West: 41,100
    Lawrence East: 33,800
    Lawrence West: 22,200
    Wilson: 23,500
    York Mills: 23,600
    Finch East: 44,600
    Finch West: 42,600
    Steeles East: 23,700
    Steeles West: 27,500
    That adds up to 308,900.

    Where are the predicted 300,000 riders on Eglinton going to come from exactly?

    Steve: One problem with a lot of modelling is that it works fine for regional trips, but breaks down for local trips where access time to and from a line is a significant part of the total. Also, the grid within the models tends to be coarse and a corridor like Eglinton can take in a wide swath for modelling purposes.

    The flip side of your point — the existing riding on parallel routes — is how much of that riding is modelled to move to Eglinton itself? We know from earlier subway openings that, unlike motorists and expressways, transit riders will not travel miles out of their way to ride a few stops on “rapid” transit.

    As and when the TTC comes out with a projected service plan for the remaining surface network, and it shows widespread cuts because of presumed migration of trips to the Eglinton line, then I will at least believe that the numbers have some credibility. Meanwhile, I will remember the history of the King car whose service was cut from 1’40” to 4’00” when the Bloor line opened in 1966, only to be quickly restored. It now runs every 2’00” including trippers. Also, one can look at Finch east of Yonge for an example of riders who choose to remain on their very frequent bus services.


  12. A recent Star article said “The original [Eginton] LRT plan called for only an 11-kilometre tunnel in the middle. With the switch, the trains are now expected to travel at subway speeds, averaging 34 km/h versus 22 km/h. The trip between Jane and Kennedy [20 km] will take about 25 minutes, rather than 45.” That is a 20 minute saving!

    However, is the Star in error? I did the calculations and got that the trip between Jane and Kennedy will take about 35 minutes (all underground) , rather than 44 – a 9 minute saving.

    Steve: Yes, the Star is in error. Either Jack Collins from Metrolinx botched the numbers, or the Star goofed. My money is on Jack Collins overstating the saving.

    It is 19.5km from Kennedy to Jane, and at 34km/h, that’s about 35 minutes, not 25. I have sent notes both to the Star and to Metrolinx asking for a clarification.


  13. Regarding transit priority now…

    Is the St Clair Route really Transit Priority (where traffic lights truly respond to an approaching streetcar)? Or is St Clair simply a bunch of well-coordinated traffic signals for transit?

    Oakwood Eastbound is the only location where I thought I witnessed a traffic signal actually responding to an approaching streetcar by holding a green for a few more seconds after the pedestrian countdown has finished. I have yet to see this anywhere else on St Clair.

    Are the traffic signals supposed to hold a green for a few more seconds for an oncoming streetcar for farside stops? Is that what we’re supposed to see on a regular basis?

    Steve: It depends, and a big part of the problem relates to the way Toronto has implemented “priority”. As I mentioned in a previous comment, the zone of control for any traffic signal can extend at most to the next signal down the line in the Toronto implementation. Rather than sensing an oncoming car and giving it a wave of greens, the signal system handles each intersection individually. When they are very close together, it is possible for a car to be detected “too late” for its presence to affect the next signal.

    Also, at some intersections (major ones like Dufferin or Bathurst), there is little if any transit priority visible to an observer. This happens when the importance of green time for a cross street takes precedence, but also because it is possible for a streetcar to get out of sync with the traffic wave around which the green signals for the street are timed. This happens on Spadina, particularly northbound, where traffic and streetcar leave on the same green, but the streetcar stops again for a large passenger volume (e.g. Dundas). By the time it leaves, the traffic wave is well north of the streetcar, and it can hit a red at the next signal.

    The mixture of nearside and farside stops, plus turn phases, adds even more complexity to the situation.

    To be fair to the traffic engineers, there is only so much green time to go around, and at some point, attempting to tweak the setup will produce minimal overall benefit. It then becomes a policy issue of whether to take green time away from cars to benefit transit vehicles.


  14. The way Metrolinx seems to be envisioning the origin-destination patterns, ie pretty much everyone is headed for the Yonge line and then Downtown in a long-haul pattern, doesn’t an SRT-Eglinton route from the east almost completely negate the need for a Sheppard Subway extension to the same starting point? How much would both of these lines rely on feeder routes in Scarborough to justify their operation?

    Steve: See my previous comment about modelling regional vs local travel. I would love to see how many people their model thinks will move from bus routes to the “LRT subway”, and what the OD pattern is for these riders. We are investing huge sums of money based as much on political as technical/planning principles.


  15. If you want full signal priority, than you need crossing arms.

    Calgary does this, and their trains get up to around 80 km/h on stretches of the road like the one you see in this video. I would fully support grade separated surface rail like this, as this is true rapid transit. But it just would not work well in Toronto where going underground makes more sense.


  16. How much would it cost to make provisions for a potential future expansion of the Eglinton Crosstown stations to 5-car length, while retaining the LRT rolling stock?

    I can imagine a situation when Eglinton (with 3-car trains) becomes a bottleneck while Yonge does not. That can happen if DRL is built according to the Metrolinx 2007 plan, from downtown to Bloor / Danforth only. Such configuration can divert enough riders from the southern segment of Yonge to keep it in a manageable state, but will not relief Eglinton.

    The idea to extend DRL East up to Eglinton / Don Mills makes a lot of sense and deserves complete support. However, such extension is not yet a part of any official plan, and there is no guarantee that it gets bundled with DRL.

    Ditto for a frequent GO service, or a parallel LRT on Lawrence. Those enhancements can potentially divert riders from Eglinton, but currently there is no commitment to implement them.

    Steve: The biggest problem with making this type of provision is the station design. First, the length of level track (no greater than 0.3% gradient) has to be extended by two car lengths. This can affect grades nearby and possibly even the elevation of the station. Also, the emergency exit and vent shaft get shifted to the end of the “longer” platform so that if the station is enlarged, they don’t tie up a lot of platform space. That affects the placement of these shafts. Finally, the length of the hole that has to be dug grows and this adds to the station cost and construction effects. The result is that you do most of the work without getting the full length station. This is fundamentally different from leaving provision for a future station (such as at North York Centre) where a completely new station was added to an existing line.


  17. Steve said: “The biggest problem with making this type of provision is the station design. [balance of quote snipped]”

    Is this not what they did with the Sheppard line though? As far as I know provision was made to extend the platforms using false walls and things of that nature.. perhaps the same thing can be done along Eglinton? I mean if it works for a full blown subway line why can it not work for a smaller LRT line?

    Steve: I didn’t say it can’t be done, simply that the changes are not trivial. Given the hilly nature of Eglinton, it’s not just a question of digging a bigger hole.


  18. @Michael

    Along the Eglinton East portion between Leslie and Kennedy, there are only 3 locations where an LRV may be slowed down unnecessarily: Swift Drive, Thermos/Sinott Road, and Rosemount Drive (Lebovic uses that downstream median U-turn treatment). I find it hard to believe that if full transit priority is implemented at these locations, that would slow an LRV down to 22-25 km/hr, compared to 29km/hr for subways.

    Sure, without crossarms, an LRV may have to slow its speed for caution, but I can’t see how this accounts for the big drop in speed.


  19. To answer Ed’s question, the 131,000 all-day figure I quoted for Bloor-Danforth was from October, 1966 (several months after the line opened). Ridership rose to 170,000 passengers per day by 1969 with the extensions to Islington and Warden.

    Bloor-Danforth is currently at 480,000 per day, so how can they justify building Eglinton as a 3-car LRT subway with a 300,000 daily ridership estimate? There’s no margin for growth later on, especially with an extension to Pearson. If these numbers prove to be correct, the line should be built as 6-car HRT now for uniformity with the rest of our subway network, and for future growth.

    As for the TTC’s own ridership predictions from the past, let me say that their own projection for BD in 1960 (which indicated that during rush hours it would see roughly 2/3 the ridership that Yonge experienced) was fairly accurate with the actual passenger counts taken six years later when the line finally did open. Their estimates for southbound passenger loads on University with the Spadina subway (south of St. Clair W.) were also remarkably accurate ten years before they actually happened. These predictions were made in the mid-1960s because there were fears that a Spadina subway would overload University south of Bloor.

    Steve: Thanks for this historical background. The question remains of whether Metrolinx demand projections are accurate given factors already discussed here about the model which generated them.


  20. Happy Dominion Day Steve!

    Thanx again for your efforts. Many items that you highlight are topics that I wouldn’t necessarily have given a whole lot of thought to before you post them, but when here I truly enjoy reading about. Your thoughts, research and passion give me pause and help to widen my sphere of knowledge. Much appreciated!

    Dennis Rankin

    Steve: You’re welcome!


  21. Demand projections for Toronto and Canadian systems in general are very very conservative, compared to our American counterparts.

    There is no doubt that ridership on Eglinton will be higher than projected, as a whole segment of the population is getting rapid transit.

    I actually find that the TTC under estimates all the time. They do this a lot with express bus routes (I do not know what it is, but the TTC seems to hate anything that does not stop every block). I remember when the Airport Rocket started carrying something like three times the ridership the TTC projected. And the 190 Scarborough Centre Rocket also started carrying something like double what the TTC projected.


  22. Concerning the estimated 20 minute time saving from burying the eastern portion of the Eglinton Crosstown, Steve replied :

    Yes, the Star is in error. Either Jack Collins from Metrolinx botched the numbers, or the Star goofed. My money is on Jack Collins overstating the saving.

    I don’t think you can blame this one on The Star. I had heard that time reduction could be as low as 2 minutes, so at the May 31st meeting at Leaside Memorial I asked just how much of a time saving we would get for our extra 2 billion dollars, and the reply at that time was also 20 minutes.

    For this reason, I believe that this is the official TTC/Metrolinx position. Whether it is accurate is a separate question.

    Steve: Do the math. 34kph will take you only 14.1km in 25 minutes, but 19.8km in 35 minutes. From Kennedy to Jane is 19.5km. For Metrolinx to claim a 25 minute trip would require an average speed of 46.8 km/h, and unless they are planning to stop at only a handful of stations, this is physically impossible. Metrolinx is wrong.


  23. My namesake, Michael, suggests that the TTC has been grudging in its implementation of the Airport Rocket. I have been taking this service ever since it was every 45 minutes in the morning rush hour time. I only travel sporadically – when I fly – and my observations are by their nature anecdotal. However, I have always been favourably impressed with the way that the service has grown along with ridership. In the early days the passenger make-up included an awful lot of airport/airline employees – people with uniforms and badges. If they were not in the majority, they were at least approaching half. As members of the flying public discovered this excellent facility, service levels have improved and the airport workers/flight attendants/pilots have been dispersed over a larger pool of passengers. Now with frequent service throughout the day, most Rocket passengers are going on a flight.

    Crowding has always been an issue – but more related to suitcases than the actual volume of passengers. A standard city bus is not the best at accommodating a near capacity load of passengers many of whom have two or three bags each.

    As you have pointed out in previous posts – the Airport Rocket – as a point to point service – is one of the lower cost recovery routes. I see this commitment by the TTC as a shining example of wise planning to meet social need, rather than providing the minimum service.

    On a trivial note unrelated to the larger issues at hand – I once took the 3:15 a.m. Ossington bus and the 3:30 a.m. Bloor Night bus to an early morning flight. Some of the Bloor Night Buses terminate at the airport. I didn’t appreciate being up so early, but it was as quick as a cab and $50 cheaper.


  24. @Michael

    The reason why I even bother to respond to your comments is because I am genuinely curious to know which projects/lines have had ridership underestimated in Toronto.

    So which specific projects/lines have had ridership underestimated in Toronto?

    And how can you be so sure that the TTC underestimated the ridership for Eglinton? Is it because they took into account all regional transit network changes, which you seem to be so against?


  25. After riding various systems in the US that have 10 to 15 minute base service for LRT and HRT and carry 30 000 to 50 000 passengers per day the TTC should convert most of their bus and street car lines t0 HRT or LRT at least to go by systems that are getting built in the states. The Americans seem to love the idea of a good transit system as it will get other people off the road so that they can drive faster.

    The Baltimore subway has huge and half empty parking lots at it surface stations. The west end of the line runs through cheap corridors that have almost no one living within walking distance. I think that Baltimore gets down to three or two 2 car trains on the line in the evening. They want one man operation to save money then have all sorts of people sitting or standing around in stations.

    I don’t think that most of them could contemplate running the Spadina car line or many of the heavy bus lines.


  26. “Oakwood Eastbound is the only location where I thought I witnessed a traffic signal actually responding to an approaching streetcar by holding a green for a few more seconds after the pedestrian countdown has finished. I have yet to see this anywhere else on St Clair.”

    St Clair Avenue at Yonge Street definitely has transit priority implemented. Normally the pedestrian signal on the north side does not last more than 7 seconds in the walk phase. However, whenever there is a streetcar that is leaving the station heading westbound (or possibly the other direction too, I’m not sure, it hasn’t been nearly as obvious), this phase is extended significantly.


  27. This line has been plagued with the ongoing debate as to the type of technology that should run on it. Now that we know the line will be fully grade separate the debate rages onwards, which bring to my question which may sound a bit juvenile.

    Considering HRT, conventional LRT and MARK II are not able to negotiate the tight curves or the current SRT line, why isn’t the TTC and Metrolinx looking at re-launching the MARK I cars via Bombardier. They can be brand new MARK I that address all the problems that occurred in the original cars. Is that an option, has the TTC even study the feasibility of going along that route?

    Who know maybe the cost of forcing Bombardier to manufacture new MARK I trains is cheaper than trying to fit HRT, LRT, or MARK II trains on the existing SRT alignment.

    Steve: Bombardier has no interest in building Mark I trains as there would be no market for them outside of this specific Toronto application. Given the length of the Crosstown line, now and possibly in the future, there is no point in building one special order of cars for what is a limited problem, and in the process locking in a need to buy even more in the future if the line is extended.

    There are only two curves — one at Kennedy and the other at Ellesmere. The one at Kennedy will be abandoned when the new through underground connection is built between the Eglinton and SRT lines. The one at Ellesmere will be replaced. While that won’t be cheap, it’s not an area where they have to worry about disturbing utilities and buildings. Also GO/Metrolinx owns the Uxbridge subdivision now, and so negotiations about work under the existing rail corridor will be a matter of offices within one agency talking to each other.


  28. While it’s nice that Metrolinx has corrected the error, unfortunately I expect it’s likely to already have done damage to any chances of resurrecting the lower-cost, at-grade alignment. It’s already in print and can be referenced again and again, whereas we have a quiet correction by Metrolinx (and one that may not be officially recognized by the Star, because it was Metrolinx’s error, not the Star’s).

    Steve: I’m watching to see whether it is repeated in some official material as opposed to the hearsay that passes for analysis in some quarters.


  29. Leslie stop could be eliminated since there is almost nothing but that would make Laird stop and Don Mills stop too far apart. Moving Laird to Brentcliffe could be an alternative.


  30. A while back I posted ridership for major crosstown bus routes, trying to figure out how on earth the Eglinton LRT could conceivably approach a ridership of 300,000 that heavy-rail advocates are taking as gospel truth:

    Eglinton East: 26,300
    Eglinton West: 41,100
    Lawrence East: 33,800
    Lawrence West: 22,200
    Wilson: 23,500
    York Mills: 23,600
    Finch East: 44,600
    Finch West: 42,600
    Steeles East: 23,700
    Steeles West: 27,500
    Which adds up to 308,900

    So thinking further….

    If you truly believe that Eglinton will run up 300,000 riders, it’s obviously functioning as a regional line, pulling in riders from the northern 2/3rds of Toronto; only a small fraction of this will be local Eglinton riders.

    And if that is the case, the line shouldn’t go along Eglinton, which is a mere 4 km north of the Bloor-Danforth line and 10 km south of Steeles. In fact, the best bet is to have a crosstown subway on Sheppard, which is 10 km north of Bloor-Danforth and only 4 km south of Steeles. In other words, we should learn for our experience placing the Spadina line less than 4 km west of the Yonge line (although it will make it to about 8 km west by its Vaughan terminus).

    If the ridership for a full-length Sheppard line isn’t projected to be 300,000, I don’t see how Eglinton can possibly approach that. Yes, there is more local ridership on Eglinton, but not more than a few tens of thousands.

    Ah, but the problem is that Eglinton needs more capacity than buses can deliver. So the LRT is the answer. But you have to tunnel the LRT. So here come the ALRT/HRT advocates saying “while you’re at it, you might as well make it an ALRT/subway/maglev/My Magic Pony.” And here we are.

    Will a crosstown Sheppard subway negate the need for more buses on Eglinton? My opinion is no, not at all. Look at the ridership on Finch East, which parallels the Sheppard subway. People pile on Finch East buses, even west of Don Mills; they don’t go south to the Sheppard subway, even though the buses are packed, and are stuck in long backups of traffic in the peak directions. Yet….they stay on the bus.

    In fact, Eglinton won’t get 300,000 riders, nor will a cross-town Sheppard subway put much of a dent in Finch ridership, let alone Eglinton’s. (Look at the crowds on the Dufferin bus, even though the Spadina subway is, for a large section, “conveniently” close.)

    Therefore, build a grid of LRT as the bus services are overwhelmed….or build subways and STILL have to build LRT on parallel lines as they get overloaded. You can build a lot of surface LRT for the cost of a subway. Of course Eglinton has to be tunneled anyway, and Ford refuses to imagine transit vehicles on the surface of the earth under the sunshine and in the daylight air. (Perhaps Robbie thinks that only vampires ride TTC and is loath to expose these taxpaying creatures to sunlight.)

    So the tunnel-where-necessary Eglinton line becomes an all-tunneled (except where elevated) extravaganza, meaning no more money to upgrade other routes, and the Sheppard subway extension remains a mirage.

    In other words, Transit City was a good plan to deal with distributed demand, and Ford has backed us into a hole (complete with TBM) from which every “escape” such as ALRT just means digging the hole deeper.

    Alas, a subway to China has certain technical difficulties….


  31. @Ed:
    For the sake of argument, assume Eglinton takes over half of Bloor-Danforth’s ridership (495,280 * .5 = 247,640) and add the Eglinton bus routes. That gets you 315,040.

    A lot of Bloor-Danforth ridership is driven by the bus routes it intercepts. Eglinton will intercept the same bus routes a bit earlier, and transferring from a mixed traffic bus to a rapid transit line early might prove appealing.

    I suspect that riders will figure out the whole capacity thing out and not transfer early if there’s no capacity for them, but Eglinton will inevitably poach some from Bloor-Danforth.


  32. It looks like the location of the fan rooms prevents any future expansion of the platform, that would be necessary to accommodate longer train sets. This, I believe is too short sighted. The entire line reminds me of the Canada Line in Vancouver, which was pretty at capacity as soon as it opened.

    @Felix There has been a fair bit of development at the North-East corner of Leslie and Eglinton in the past couple years, and I imagine these residents (as well as Sunnybrook park users) will want access to transit.


  33. Leo says, “For the sake of argument, assume Eglinton takes over half of Bloor-Danforth’s ridership (495,280 * .5 = 247,640) and add the Eglinton bus routes. That gets you 315,040.”

    Eglinton will likely pull some passengers from Bloor-Danforth, but half? Making Eglinton busier than Bloor-Danforth? I don’t think so.


    Other than the subway transfer stations (Bloor/Yonge, St. George) some of the busiest stations include Broadview, Spadina, Islington and Kipling. Very little of the patronage here will go to Eglinton, because the main service is to the south for the first two, and the services are well west of the end of the Eglinton line for the other two. (There’s a link to station usage somewhere in recent postings, but I can’t find it.)

    Steve: Actually, Broadview gets a lot of traffic from the bus routes pouring in passengers from the north, but of these routes, only Flemingdon Park goes to Eglinton. The station stats are here.

    Many other lesser-used stations such as Runnymede, Dundas West, Lansdowne, Ossington, Bay, Sherbourne, Castle Frank, Coxwell, and Main have feeder services whose main business is in the Lakeshore-St. Clair band and again Eglinton won’t draw these away in any quantity.

    I can see Weston Road, Keele, Dufferin, Don Mills, and Victoria Park riders heading to Eglinton in some numbers, but certainly not all. Dufferin for example is busier south of Eglinton than north of it.

    If Eglinton pulls even 100,000 riders from Bloor-Danforth I’d be quite surprised.

    Of course if your argument is valid, we can transfer T1s to Eglinton and convert Bloor-Danforth to LRT like it was always meant to be!


  34. When people claim that the capacity of LRT is around 10,000 pphpd-12,000 pphpd (roughly), does that include how much passenger traffic the LRT stations themselves can take?

    Does that mean a typical LRT station can handle twice this number (for each direction)? Can they take 20,000 passenger/hour, assuming that half would go in one direction and the other half goes in the other direction?

    Steve: Much depends on the design. No station has 20K/hour right at that station. The people passing through on the trains don’t count for purposes of platform and access capacity. For example, close to 30K may pass through Wellesley at the peak, but they don’t all get off at that station. Problems arise with far fewer people trying to use the stations further south, especially when an escalator is out of service (notably at College where there is no alternative).


  35. How many passengers, theoretically, can use a surface streetcar/LRT platform? For example, does the westbound platform of Deer Park Crescent have enough room to handle 12,000 boarding passengers in one hour?

    Steve: Absolutely not. As I said before, not even a subway station has to handle that sort of volume in most cases. Another constraint for surface platforms is that people must get to and from them. That’s why the Bloor streetcar had a transfer station in the middle of Bloor Street east of Yonge for its connection to the Yonge subway.


  36. Eliminating the Leslie stop leaves a very wide gap between Don Mills and Laird. I live in some condos on the north east corner of that intersection (three towers with a significant amount of residents), not to mention plans for a huge seniors residence in the old Inn On the Park building, the current construction of the 3 phased “scenic” condos on the south west corner and the large amount of new townhomes on the same corner. Leslie is a major feeder street that connects the north to Eglinton, which is currently served by the 51 and 54 bus lines. The walking time from Laird to Don Mills is at least 25 mins at a good pace in good weather.


  37. Hi Steve!

    Looks like the TTC and Metrolinx launched a new site for the LRT.

    Guess they took a few tips from the logo I made, haha! 🙂

    Steve: And if you look closely, you will see that the station animations show high floor platforms.


  38. I can’t understand why, if they’re building the whole thing underground, why they don’t make it a subway so it’s interchangeable with the other rapid transit lines? Or so that the surface routes can feed directly into the Eglinton line? As David Gunn says, the Eglinton line as is will cost more and carry less.

    Steve: The 100% undergrounding is simply Queen’s Park rolling over for Rob Ford while wasting a lot of money. The whole idea of value for money planning at Metrolinx is dead in the water.


Comments are closed.