Posts in this thread have examined the general design proposed for the Richmond Hill subway and the many demand estimates for this line. Now I will turn to the impact of this line on the larger network.
As many have pointed out in comments to the previous items, the Spadina/VCC extension was supposed to offload the Yonge subway. We now know, according to the TTC’s estimates, that the effect will be a reduction of less than 10% of the existing demand southbound at the peak point, Wellesley Station. Meanwhile, the availability of a competing subway line in the established Yonge Street corridor will attract many more riders.
The TTC manages a rabbit-in-the-hat trick by claiming that demand relative to capacity on the subway in 2017 will be the same as it is today thanks to Spadina diversion and more commodious trains. That’s a very big, very fat rabbit, and I suspect it’s more of a canard.
Development will continue in York Region, and if anything the availability of frequent transit service, both on GO and on the TTC, will offset any effect that long-term increases in energy costs and commuting might have on travel demand and the decision to live far out of the core area. Demand will grow on the subway both from the 905 and from within the 416.
Notwithstanding the 2017 claim, the TTC has many options to increase capacity including:
- Addition of a 7th 50-foot long car to expand train capacity by about 10%
- Operation of a headway as low as 105 seconds using automatic train control
- Reconstruction of Bloor-Yonge Station to separate boarding and alighting traffic streams
- Diversion of traffic from the lower Yonge line to a Downtown Relief line
The addition of a short 7th car to the TR trainsets brings added capacity, but there is a problem with timing if this were included in the base car order now at Bombardier. The longer trains will require more precise stopping at the platforms, and the TTC feels that it is impractical to expect this of their operators who are now used to having a cushion because trains are shorter than platforms. Only after ATC is running on the full YUS in 2015 could 7-car trains be operated. The TR cars will be here long before 2015.
Another issue is that the TR fleet now on order is not large enough to displace all of the T-1 trains, and a supplementary order will be needed to fully convert the YUS. The TTC may be unwilling to operate trains of varying lengths on the same line as passengers will either wait where no train will stop, or they will avoid the ends of the platforms in anticipation of a short consist.
Any scheme to add platform doors to the YUS also requires trains of uniform door arrangement. Mixing trains of different layouts will require a lot of extra doors that would eventually be obsolete. As I have already written, I believe that platform doors are much more a make-work project than a vital addition to the subway system. Nice-to-haves, not must-haves.
Automatic Train Control
The current schedule for ATC implementation is:
- 2013 Eglinton to Union (replacing the oldest signalling equipment on the line)
- 2014 Finch to Eglinton
- 2015 Union to Downsview
The TTC claimed that ATC was an absolute pre-requisite for any expansion of subway service, but this runs headlong into the claim that the same headway will do nicely, thank you, for 2017 as for today. This is not to say that upgrading the signalling system was a waste of money, but the way the project was presented created an artificial crisis to encourage funding.
An important side-effect of ATC is the ability to run in either direction on either track. This permits bidirectional operation at widened headways between points where trains can switch from one track to the other. The TTC plans to reinstall the crossovers formerly at King, College and St. Clair to add flexibility in operation on the line during service disruptions.
There have been statements (although inconsistent) about the ability to run 24-hour service because trains can be diverted around work sites. This is not quite as simple as it seems because the power feeds are not, as I understand things, set up to allow selective shutdown of one direction’s power.
A 105-second headway can be operated provided that trains are allowed to approach more closely to each other than the current system permits. The two clear block rule on today’s system prevents trains from creeping right up behind their leaders at busy stations, and potential dwell time is lost getting a following train to the platform. ATC allows this close spacing without the safety exposures of doing the same thing manually (as is technically possible with the existing signals at many locations).
Short headways also demand that some trains short turn because the geometry of existing terminals makes it physically impossible to cycle trains in and out that quickly. Leaving aside the less than split-second timing of crew changes, the length of crossovers imposes lower bounds on the cycle.
In this context, the TTC’s plan to send all PM peak service to Richmond Hill is dubious unless they are going to radically change the track layout compared with current designs. An “opening day” headway equal to current operations will fit in a standard terminal, but a reduction to 105 seconds will not.
The Bloor-Yonge Station changes do not increase capacity per se, but reduce dwell times by separating the passenger streams. This is a pre-requisite for shorter headways because the sum of the dwell time and the time to cycle the next train onto the platform must be no more than the headway, preferably less for flexibility.
The TTC has resurrected a 20-year old scheme to restructure the station with a new centre platform on the upper (Yonge line) level, and a possible pair of side platforms on the lower (Bloor line) level. See diagrams at pages 33-34 of the online version of the staff presentation.
I have already published the report on which this design is based (browse “Part 1 of an 8-part review on my site). That report claimed that the construction would require an extended closing of Bloor Station, but I am told by TTC staff that a subsequent design changed this. Essentially, the plan would be to cut away the existing platforms in short sections replacing them with temporary structures to maintain access to the trains. New tracks would be laid under the temporary platforms, and operations would switch over to the relocated tracks and narrowed side platforms once an entire side of the station was completed.
What has not yet been addressed is the issue of construction at the north end of the station where it is physically inside the structure of the Bay, and in the approach area north of the platform where the tracks must be moved further apart. Foundations of adjacent buildings are close behind the existing tunnel walls including the Asquith Bell Canada switching centre.
Connection to the Bloor line’s existing platform requires excavating a new passageway under the existing station, and side platforms on the Bloor line itself bring their own challenges. Some of the structures required to serve new side platforms on the Bloor level would probably best be built as part of an initial construction for the new third Yonge platform, and this will add to the cost and complexity of such a project.
The TTC will undertake a feasibility study to update the 1988 report, and staff (at least some of the staff who are also gung-ho on the Richmond Hill subway itself) seem really enthusiastic about this project. The implications for service quality during a multi-year construction project are daunting, and the cost will be substantial. Both of these must be offset against alternative ways to maintain traffic through this critical junction at or below current levels.
The Downtown Relief Line (East)
Metrolinx includes a DRL running from Dundas West to Pape Station via the Weston Corridor, an indeterminate east-west alignment, possibly Queen Street, and the Danforth subway in its 25-year plan. I believe that this is too far in the future even assuming that this phase of the plan could begin in “year 16” after the 15-year plan is completed.
For its part, TTC staff are less than enthusiastic about this option, an odd situation considering that it could add another subway project comparable to the Richmond Hill line to their future workload. The problem, of course, is that if we make the DRL a co-requisite for the RH extension, this will trigger a funding crisis (as if we don’t already have one) in the overall scheme of transit projects.
[Note to my many commentators: Please don’t set off a discussion of which alignment at DRL might take in response to this post. The issue is whether we have a DRL at all. Technology and alignment will affect the total cost, but in the context of the RH subway I would like to keep that discussion for another thread.]
The staff presentation includes a sketch map at page 36 of possible DRL alignments. All of them proceed south from Pape to at least Queen, and the three options shown enter downtown via (1) Queen, (2) Eastern, Front and Wellington, or (3) Eastern, Rail Corridor, Front.
These trace their origins to the DRL plan that was based on ICTS/RT technology and which went down to Eastern Avenue to access a possible carhouse site. An alternative alignment via the Rail Corridor from Pape (or further east) is not included, and this is a considerable oversight.
At its north end, the DRL plans go to Pape because it is a major bus terminal (25 Don Mills, 81 Thorncliffe Park), and because with an ICTS yard on Eastern, no connection to the existing subway network is required.
However, the larger context for a DRL today is the Don Mills LRT study. The south end of this Transit City line suffers from a hangover of the Don Valley Transportation study which included BRT operations to Pape, Broadview or Castle Frank Stations. None of these makes sense at the capacities expected on a Don Mills line, but this is an example of “the tyranny of old studies” I have written about before.
The original scheme to run the Don Mills LRT across the Leaside Bridge has encountered problems both of bridge capacity (possible, but difficult) and curve radius (turns both end of the bridge are difficult). Moreover, designs for on-street operation on Broadview or Pape belong in a fantasy world where we can convert busy four-lane streets into transit malls. This sort of “design” gives LRT a very bad name, and if I were cynical, I might think it was a deliberate plot to demolish support for the proposal. [Cynical? Me? You jest!]
A complete rethink of this section is long overdue including a recognition that the line will have to be underground at least to the point where it reaches the Don Valley. The problem then becomes crossing that valley, entering Thorncliffe Park and continuing north.
My own view [I can hear the wild laughter now] is that the line should pass under Thorncliffe Park and continue up Don Mills to a major junction at Eglinton. This would provide a continuous rapid transit link from Eglinton to Downtown in the Don Mills corridor and eliminate much transfer activity at the Danforth Subway. A Don Mills LRT, operationally separate from the DRL itself, would run north from Eglinton. [Yes, there are design alternatives here including the question of a link to the CPR corridor and GO service to Agincourt and beyond.]
Nobody wants to talk about this sort of scheme because it is very expensive. However, it must be remembered that whatever will be done with the Don Mills LRT, the south end of the line is going to cost a lot and probably be at least partly underground. The question is one of the marginal cost of integrating this part of the line with the DRL.
Moreover, a DRL will divert much demand from the Yonge line and Bloor-Yonge station thereby avoiding the cost of additional station capacity not to mention the greatly increased fleet needed to handle all the traffic on the Yonge line.
At this point, readers may think I am just about as certifiable as the TTC engineers who want to dismember Bloor-Yonge station. My point is that we have to look at the alternatives both to understand what might be done and how the network would benefit (if at all) from various alternative schemes. The greatest challenge, politically more than technically, is that the pressure to build new lines has the construction sequence all wrong, and the possibility of at least temporarily overloading the existing system too high.
The Downtown Relief Line (West)
While we are on the subject, I may as well add comments about the west branch of the DRL. I believe that it is not as critically required because demand from the west is, or will be, served by a richer collection of services on GO Transit, and the University Subway (to the degree that it has reserve capacity).
A related project is the Weston/Airport corridor which, if it were detached from the stupidity of Blue 22, could be regarded not just as a regional GO project, but also as a new local service connecting northern Etobicoke and Weston to downtown. I am not convinced that this should be through-routed with the DRL east, and doing so could create technology issues by forcing both lines to be either LRT or full subway.
The Sheppard West Subway and Wilson Yard
The TTC has included the Sheppard West subway in its grab-bag because it is one of the options for the Sheppard/Finch corridor issues. They have concocted a story that goes roughly like this:
- If trains must enter service from Richmond Hill southbound, they will have to leave the carhouse earlier.
- This will cut down on the time available for line maintenance and increase the cost of dead mileage.
- Therefore we need a Sheppard connection, possibly with a station at Bathurst, to grant access from Wilson Yard to the NorthYonge subway.
This is a badly skewed argument. First off, the really early morning trains should originate at Davisville, not at Wilson. Second, a link across Sheppard West would cost at least $600-million at current construction costs. The interest alone on that could pay for a lot of extra dead-heading.
If storage is provided somewhere on the extended Yonge line, it does not have to be in an open air “yard” like Wilson or Davisville, but could be in underground tracks adjacent to the running structure. We’re not talking about housing a dozen or more trains, only enough to prime the start of day service.
(An operational point: carhouse moves from Wilson Yard to Richmond Hill could only occur at the beginning and ending of service because there is no east-to-north or south-to-west connection from the Sheppard Line to the Yonge line. The moves would require trains to reverse through the existing links which face south, not north.)
The TTC talks of the ability to enlarge Wilson Yard, but they need to put this in context. How many trains can it hold? What is the effect on fleet size (and space needs) of dropping the headway to 105 seconds? What is the realistic limit on the number of cars Wilson Carhouse can service?
The purpose of these three posts has been to ask many questions, not to answer them definitively. I have some ideas, and so do many others. The vital point is that we do not fixate on a single “solution” before we really understand the problem and the options for dealing with it.
As I said in Part 2, the first thing we need is a detailed model of origin-destination patterns for 2017, the year the RH subway would open. Where does everyone want to go? Which services serve which demand? What is the growth potential for both ridership and service in various corridors? What are the comparative costs of alternative ways to produce capacity in the Yonge corridor? What are the costs and risks of concentrating all riding in a single corridor?
If the riding estimates that have been published are correct, there will be a large and growing demand for capacity into Toronto and within Toronto for service. We have to accommodate this. The real problem may be that the demand verges on outstripping the amount of service various governments are willing to pay for. Building a single subway extension may keep the folks in Richmond Hill happy, but at the expense of the greater network.
Metrolinx produced an integrated Regional Plan, but did very little to examine the relationship between its components. This is happening, belatedly, now that the Board has recognized the importance of looking at bundles of lines, at collections of alternatives. Projects like the Richmond Hill subway threaten to end-run the integrated planning Metrolinx claims is so important.
In the current economic climate, there maybe pressure to approve and fund this route without critically looking at alternatives. The fact that nobody could actually start building it until well beyond the end of the current recession will be lost in the political hoopla. Vital money that could have been spent sooner on works we really need (not necessarily transit) will be sequestered in a future subway line’s trust account.
I am not unalterably opposed to subway construction if the need can be demonstrated, and the need is not the result of a gerrymandered demand model. However, I also want more transit in more places, more options for people to travel around the region. Burning up every dollar we have on a few lines won’t accomplish that.