SRT Conversion to LRT Not a Dead Issue

The possibility of converting the Scarborough RT to Scarborough RTtrue LRT when its current fleet reaches the end of its life, instead of conversion to ICTS Mark-II, is still being considered by the TTC.

In the supplementary agenda for the October 23 TTC meeting, the status update on Transit City includes a section on the planned upgrading and extensions of the Scarborough RT. The following paragraph appears in the document:

The project team is currently re-visiting the option of converting the Scarborough RT from its current vehicle technology to light rail technology, when the current fleet of vehicles reaches the end of its service life. Such a conversion is being investigated as a means of providing more routing flexibility, and in order to take advantage of possible benefits of a vehicle technology which would be common to the other Transit City lines. In support of this option, a structural analysis is underway of the existing Scarborough RT infrastructure. The project team is continuing its work on the development of conceptual designs for a new maintenance facility.

In my opinion, this conversion would be a wise choice. When looking at converting the existing SRT to either ICTS Mark-II or LRT, the LRT option is slightly more expensive. However, when one considers extending the line, the cost per kilometre is significantly less for the LRT option. Furthermore, the benefits of using the same technology as the rest of the Transit City network is immense.

Though actual use traffic patterns will dictate what routes will operate, I strongly suspect that at least during rush hours it will be convenient to have split operations on both the Sheppard route and the SRT route. What I mean by this is that the Sheppard route could see some running all the way along Sheppard to Meadowvale while some would branch to Malvern Town Centre, and also the SRT could run with these two same destinations in the northeast. This would provide no transfer service from Malvern Town Centre to Scarborough Town Centre and Kennedy as well as to points west along Sheppard and the same thing from points east along Sheppard.

In actual use, if the first LRT arriving was not the one you need for a transfer-less journey, you could either wait for the one you want, or you could take it to the transfer point. It is very likely that you might get to your destination earlier by opting for the transfer choice. I have experienced this very type of service as it currently exists in Denver. Take a look at my Denver page to get an idea of how the routes work with this sort of interlining. There are five routes forming an “X” between the southwest suburbs, southeast suburbs, and western and eastern downtown areas. They all interline for three stations where one can transfer if the train they are on is not going to their destination. There is a similar interlining arrangement in the southeast suburbs where a branch provides routes to downtown and a suburb-to-suburb route.

8 Responses to “SRT Conversion to LRT Not a Dead Issue”

  1. W. K. Lis Says:

    The Scarborough RT was supposed to have been originally to use CLRV’s running as multi-unit trains on their right-of-way. However, the government of the day forced the TTC to use ICTS instead.

    Now the TTC wants to replace the ICTS with LRV’s, but will they be Low-Floor LRV’s, so that the same LFLV’s from the Transit City lines can be used on the RT (in multi-unit trains of course)?

    Cal’s comment: Sort of, maybe yes, or no. Clear? I am not certain just where the TTC actually stands on this…

    Back in 2005/6, the TTC ran a study on the future of the SRT, lead by Richard Soberman (an article on the final report is on Steve Munro’s website). From accounts I have read, Soberman sounded like he was leaning towards an LRT conversion, but at the time the TTC had a very anti-LRT attitude. Added to this, the TTC already had a completed EA for the extension of the line, at least to Sheppard and Markham Road. When the report came out, there was a full 180 from what Soberman had been leaning towards throughout the study. The report did compare conversion costs of the current line to either Mark-II ICTS or LRT, but did not compare extension costs.

    At the last round of EA open houses earlier this year, it appeared that the decision was final that the SRT would be converted to Mark-II ICTS and extended. One of the display boards outlined different options that had been considered, including LRT conversion. Despite it appearing to be a done deal, I took the time to submit comments that LRT conversion should be given more serious consideration, with justifications. There must have been a significant amount of feedback suggesting this, because it appears to back on the table, maybe. The cynical side of me suggests that it is only in this current report to appease people such as myself. On the other hand, the TTC may very likely want to do away with ICTS all together, but have to do so in small steps as there may be pressure to keep a showcase for ICTS in Ontario.

  2. Scarborough Guy Says:

    Scarborough will need to have at least one subway line going into it. Right now there is none as sheppard stops well short of it and bloor line goes 2 blocks in to kennedy. Scarborough rt is the only one that spans pretty much the whole thing with the malvern extension.

    Cal’s comment: I am curious, please cite the anticipated ridership levels for justifying a subway line in Scarborough. I would be really interested to know where exactly a line can be drawn in Scarborough where the need to move at least 10,000 people per hour per direction exists in Scarborough.

    I used to be a “we need a network of many subway lines all over the city” person, until I saw for myself just what can be done with LRT for a fraction of the cost. My fear with Transit City is that it will not be as rapid as LRT can be, and this is most unfortunate for Scarborough, where there are a number of corridors that could be used to provide a very rapid implementation of LRT along the lines of Calgary’s C-Train or The Ride in Denver. These types of LRT implementations can rival subway for speed, while providing a more suitable level of capacity at a comparatively proportionate level of cost.

  3. Michael Forest Says:

    Scarborough is the most remote part of 416, so perhaps subway extensions into / towards Scarborough are desirable for the sake of speed and connectivity, even though they won’t carry 10,000 pphpd.

    I would think of the following scheme:

    1) Extending Danforth subway to STC, and adding LRT line STC - McCowan - Sheppard - Neilson - Malvern Centre.

    It is possible, of course, to place LRT on the present SRT alignment and thus avoid the subway extension. However, subway extention would cut a lot of transfers (almost all of Scarborough will have a one-sit ride to a subway station).

    Cal’s comment: While this sounds like a great idea, one must consider that other speed and connectivity possibilities may be more feasible for a much lower cost, namely the role that GO Transit will play in the future. Granted, a better and equitable fare integration will be needed.

    2) Eglinton LRT gets extended east of Kennedy, along Eglinton and Kingston Rd. (areas with high intensification potential), up to Morningside Mall. In other wards, it absorbs a portion of “Scarborough-Malvern” LRT. That will connect Eglinton to Lakeshore East GO line, as well as several Scarborough’s N-S bus routes. There will be no need in LRT north of Morningside Mall / Lawrence, as those areas have a limited intensification potential, and will have more direct connections to subway.

    While Transit City plans show distinct routes, once the infrastructure is in place there is no reason not to have interlining routes or run-through routes that will provide riders with a one-seat ride through areas that warrant it, even if only during peak hours (similar to VIVA’s Pink route that provides this during peak times).

    I would disagree with the lack of need for LRT on Morningside north of Lawrence. It is not simply for intensification purposes that lines such as this are proposed, but for network connectivity. A true network provides alternate paths of movement.

    3) Extending Sheppard subway west to Downsview and east to Kennedy, and adding LRT line along Sheppard east of Kennedy. Then, a lot of E-W trips in the north will use a combination of Sheppard and Spadina subways (for example Finch & Warden - bus to Sheppard - subway to Downsview - subway to Wilson - bus to Keele and Wilson). Buses serving eastern parts of Finch and Ellesmere could then operate off the Kennedy terminus.

    Extending Sheppard west to Downsview is a network enhancing project, which I am generally in favour of. Getting funding support is another issue, though. I have argued that if this could be arranged, and a subway to York U and beyond is really necessary, then it should be a western extension of Sheppard that should go to York U and VCC.

    Speaking of costs: the subway extentions would cost a lot more than the portion of LRT they are replacing, but the relevant measure is the total cost of each plan. Let’s estimate all new lines east or north of Kennedy Stn, and east of Don Mills Stn.

    Plan currently on books:
    - SRT upgrade and extension to Malvern: 1.75 B
    - Sheppard E LRT: 800 M
    - “Scarborough-Malvern” LRT: 600 M
    - Total: 3.15 B

    LRT-only plan:
    - SRT to LRT conversion: 500 M
    - LRT extension to Malvern: 500 M
    - Sheppard E LRT (less the section shared with STC-Malvern) : 700 M
    - “Scarborough-Malvern” LRT (less the section shared with STC-Malvern): 550 M
    - Total: 2.25 B

    Subway / LRT combo plan:
    - Danforth subway to STC: 1.2 B
    - Sheppard subway to Kennedy: 1.2 B
    - LRT extension to Malvern: 500 M
    - Sheppard E LRT (off Kennedy, less the shared section with STC-Malvern) : 400 M
    - Eglinton / Kingston Rd. extension of Eglinton LRT: 400 M
    - Total: 3.7 B

    So, the subway / LRT combo would cost more than the LRT-only plan, but not x5, and not even x2. If it results in a much better connectivity / travel times, then perhaps this is the way to go.

    Your numbers are somewhat reasonable, though I believe the Bloor-Danforth extension to STC would be closer to $1.5B because the distance of the required alignment is closer to 6 km. The line is east-west at Kennedy, so an as-the-crow-flies alignment to STC is not possible. The number I get for a Sheppard extension to Kennedy is $1.25B, so I won’t quibble over $50M.

    The bigger issue is the gap between Sheppard/Kennedy and STC. Extending Sheppard only to Kennedy would extend its current Pre-Metro status. Why go to the expense of extending the Bloor-Danforth line to STC, but not connect it with a Sheppard extension to make both extensions network enhancing? Of course, with these two extensions in place it would make logical sense to build the connecting link, which would be another $733M.

    There is one flaw in this cost comparison, and it comes from your final comments about the subway/LRT plan not even being twice the cost of LRT only. This is partly because you are including additional kilometres of LRT construction in both summaries which are separate from the comparison (the costs of LRT construction that appear in each summary, such as the Sheppard LRT east of Kennedy not adjusted for shared costs) should be removed from the comparison in order to look at the cost ratio. The comparison is further complicated by the choices for the SRT. While a kilometre of underground subway will on average cost about five times the cost of a kilometre of above ground LRT, we are not simply swapping things kilometre for kilometre, mainly due to the SRT options.

  4. Michael Forest Says:

    Cal, a few comments:

    1) Regarding the LRT on Morningside north of Lawrence. It should be taken into account that TTC’s light rail routes are planned for relatively high capacity. So, if there is no local density around the route, and it does not connect other major trip generators, that route will be underused.

    My understanding is that UTSC is the only significant trip generator near Morningside north of Lawrence. If the subway terminus stays at Kennedy, then UTSC is the logical end of the Eglinton - Kingston Road LRT. But if the subway terminus is moved up to STC, then the connection from UTSC via Ellesmere (whether LRT or bus) will be more attractive.

    I am not sure about the chances of intensification along Morningside further north; and the Malvern to Kennedy trip via that route would be to long anyway.

    The passenger network connectivity can be maintained by buses. The rail maintenance connectivity can be achieved by installing mixed-traffic non-revenue tracks on Morningside.

    Cal’s comment: There are a number of factors that work into the model that determines of LRT capacity is warranted. UTSC is the main trip generator at this point in time, and the travel pattern for that destination would be altered by extending the subway to STC, but a number of other things will take shape before this line comes online, including many of the proposed plans for GO Transit, which would alter the choice of getting to STC for a subway connection versus getting to Guildwood station for a GO connection for downtown-based trips.

    2) Regarding the cost of Danforth subway extension to SRT: The 1.2 B mark comes from the well-known report of Dr. Soberman. If the real cost is going to be considerably higher than that, then perhaps we should stick with LRT solution instead of subway. LRT would be slightly less convenient, but still provide the needed capacity and a pretty decent speed (being on the fully separate ROW). Obviously, the cost factor cannot be ignored.

    That figure must be inflation-adjusted to compare 2008 dollars with 2008 dollars.

    3) The eastward extension of Sheppard subway to Kennedy can be viewed as network-enhancing, if we count the Stouffville GO line. I am quite surprised that the RTP does not list that line as one of frequent Express services, even within the 25-year framework. That would be a great connection between the Toronto’s north-east and downtown, the speed is pretty good there (23 min from Agincourt to Union).

    Adding Sheppard subway to Agincourt transit hub would further enhance its potential. Quite a few bus routes in the north and east could operate off that hub, plus the LRT along the rest of Sheppard. Plus, if new condos mushroom near Sheppard subway between Don Mills and Kennedy, their dwellers could use Stouffville GO line as a shortcut to downtown. Actually, that route would be faster than the one via Yonge (provided that the destination is close to Union).

    Now you are extending the line to GO Agincourt. This puts it’s cost at $1.37B for a proper analysis. I am not convinced your proposal is a bad idea, but I do believe it would be far more useful if the two subway lines made a connection, and that puts its cost way up. One day, there will be more justification and need for subway construction, but our major problem currently is that public transit needs to play catch-up for several decades of neglect. There is only so much money to go around and the wider it can be spread the closer we will be to having an overall network that is useful to all.

    4) I don’t think that the Kennedy - STC chunk of subway should be bundled with the Don Mills - Kennedy chunk. We are not in the situation of having to build either both or none.

    The Don Mills - Kennedy section (which will see higher usage, and be more conductive to the network connectivity) can be built now, while the Kennedy - STC section can be deferred until after the 25-year plan.

  5. Michael Forest Says:

    And Cal, there is no flaw in my cost comparison. I intentionally formulated it that way.

    I do not dispute the fact that, say, a 6 km subway extension will cost several times more than 6 km of LRT in a surface ROW.

    Nevertheless, I believe it is useful to compare the costs and benefits at the whole network level, rather than just a particular route. From that viewpoint, the cost advantage of LRT over subway gets naturally diluted, because both network options (all-LRT versus subway-and-LRT) share a lot of LRT parts.

    Cal’s comment: It is true that we must look at “the big picture” to see the various options. However, to take “big picture” costing figures to make a claim like “subway is less than twice the cost of LRT” distorts the purpose of such analysis. In our situation with limited funds to be used to make as much of the system more useful to everyone, that extra cost may seem not too bad when looking at Scarborough alone, but could provide several kilometres of LRT construction elsewhere in the city that could place 200,000 more people within a five minute walk of a higher order transit line, and another 600,000 within a five minute bus connection.

  6. Michael Forest Says:

    RE: Kingston / Morningside LRT and its connection to Guildwood GO

    You are right, this is a point that I missed. If many people are expected to use that connection, then the LRT at least to UTSC is justified regardles to other routes. Not so sure about Malvern, as it is quite far from Guildwood.

    Cal’s comment: Though less useful, the connection from UTSC to Malvern has three reasons that I can think of. First, there is Malvern Town Centre as a destination for this line, which I know from personal experience having lived in the Morningside/Ellesmere area for over a decade. Second, some badly needed connections (as well as fare integration) between the TTC and DRT will add sources and destinations for trips involving this line. Third, a connection up to Sheppard for access to the carhouse that will likely be on Sheppard near Meadowvale will be necessary anyways, so why not make this track useful for revenue service.

    As for a connection from Sheppard up to Malvern Town Centre, regardless of what happens with the SRT, it would likely serve not only the Eglinton/Malvern line, but also a branch of the Sheppard line, at least during rush hours. I can see the possibility of Sheppard being run with a branch straight out to Meadowvale and another to Malvern Town Centre. If the SRT were to be converted to LRT, the portion from Markham and Sheppard to Malvern Town Centre would be built under the Malvern line’s construction and I can also see this service being operated with one destination at MTC and the other at Meadowvale.

  7. Michael Forest Says:

    I have a concern regarding the Sheppard E route, and the concept of “almost everyone within 2 km from a rapid transit line” in general.

    The problem is that TTC is not designing Transit City lines as true rapid transit. Exceptions are the tunneled portion of Eglinton, and the replacement of SRT running on its guideway. The rest of light rail, in dedicated lanes but with frequent stops, will run just 10 - 15% faster than ordinary buses (according to the Metrolinx modeling for Sheppard E line).

    Cal’s comment: I am on side with that concern. The increase in speed over ordinary buses does not sound great, but could be increased further if we have true signal priority. This does not simply mean that LRVs get a green signal phase before anyone else, but that they get the phase when they need it. In creating all the different city pages on this site, I have seen with my own eyes that this is possible and makes a great difference. Toronto’s roads and traffic department has not been very co-operative, given that the signal priority hardware installed on Spadina has never been turned on. Though, I suspect part of this may be due to that hardware not being true priority hardware, given that while Fleet Street was closed for track reconstruction, the transit signals at Fleet and Strachan continued to get their phase.

    The TTC’s design is not necessarily bad for all lines, in fact it might have advantages in both accessibility and construction costs. Finch W line, for example, should do well, since many users will transfer at Finch W subway (almost midpoint) and head either towards downtown or to York U. If the LRT portion of trip is within 15-20 min, people will just like its reliability and comfort, and not care that another design could make it a few minutes faster. Eglinton Crosstown should be good as well: it connects to subways at three points, plus the tunneled section will accelerate the whole line. That line might not be a great tool to travel from Kennedy to Pearson, but many people will find it useful for medium-length trips, including cross-Yonge.

    The emphasis on Transit City has been on “faster local service”, which is not exactly “rapid transit”. Even given that a full subway qualifies as “rapid transit”, would you want to take one from Kenendy to Pearson?

    But what about Sheppard E LRT? It will be heading towards nothern Scarborough, which is the most remote part of Toronto due to the Lake’s geometry. Yonge to Kipling is 12 km in the south, 14 km in the north. Yonge to Eglinton / Kingston is about 15 km. But Yonge to Morningside is about 20 km.

    So, how beneficial is the planned LRT?
    - For those living near Sheppard and not far from subway (say up to Midland), light rail will be very beneficial. They will get a reliable and comfortable ride.
    - For those living near Sheppard but further east, light rail will be somewhat beneficial. But they’d rather have something faster.
    - But those living “wihin 2 km” from Sheppard, say along Finch or Ellesmere, probably won’t bother transferring to LRT, as it will be only slightly faster than their regular bus.

    For the most part, this is correct, though I suspect perception might make some change their travel patterns when the benefit may be marginal.

  8. Maury Markowitz Says:

    I’m confused: the original conception of the ICTS was that the capital costs for an at-grade system were lower than a LRT because they didn’t need the tracks to be built below grade. After watching the more-than-a-year long process to replace the 1 km or so of the Bathurst line from Bloor to Dupont, I see their point. St. Clair is an absolute disaster, and from what I can see, they’re progressing backwards. And for what? There’s zero performance improvement, and the claims of better service frequency are clearly countermanded by the extensive experience we have on Spadina.

    The ICTS, on the other hand, uses fairly conventional roadbeds, laid on gravel. They go in quick, with minimum prep. I find it very suspect that the cost-per-k for ICTS is greater than LRT. Yes, it needs its own right-of-way, but critically, it already has that. Sure, laying new line at the end, especially aboveground, would be expensive, but there’s no parallel at all with LRT.

    Cal’s comment: ICTS does NOT use fairly conventional roadbeds - quite the contrary. It must be on a concrete roadbed in order to provide a stable foundation for the reaction rail which must remain at a spacing from the bottom of the train within a limited tolerance. It is LRT that can use ballasted tie construction.

    In addition to this infrastructure, ICTS is a mini-metro which means that it must be in a 100% separated right of way, including station infrastructure. When not elevated or underground, that means it must be totally fenced in with no level crossings with any roads. Even when elevated, any close structure has to have fencing installed - something I hadn’t realized until my recent travels on Vancouver’s SkyTrain.

    Are there numbers on this? Trustworthy ones? The rest of the world seems perfectly happy with ICTS: Vancouver and Kuala Lumpur are expanding their systems continually, and the public reaction to the systems in other cities seems generally positive (especially Beijing). The only places that still think the ICTS isn’t “da bomb” are Toronto and Detroit, the two places where it was forced down their throats by government bodies higher up the scale. One might conclude that this is a political issue primarily.

    No, these are the two places where there is a significant amount of that thing called snow - significant amounts of wet snow in particular. Vancouver had more than its usual experience with snow this past winter, but each year they have some shutdowns, but not as much as the Scarborough RT. Mark-II is no different than Mark-I in this respect. ICTS works well in the other locations, and not overly bad in Vancouver, but Toronto and Detroit have proven it is not suited to significant snow.