[My apologies for the temporary absence of this item. I have been updating it.]
A supplementary report on ridership projections and other impacts on the Yonge Subway is now available on the TTC’s report website.
While I do not agree with all of the report’s conclusions, this is a refreshing attempt to look at the growth and development of the transit system on a wholistic, networked basis, rather than as a single line.
The TTC persisted in using subway train capacities that do not match their own service design criteria, ignoring current over capacity problems and downplaying future growth. This has changed between the December and January staff reports, and with that change come important new concerns about current and future available capacity.
At this point we have no idea of the feasibility of the proposed Bloor-Yonge platform reconstruction project. The TTC alternately treats this as something for the indefinite future or as a co-requisite of the extension’s construction, depending on which report one reads. Cost, constructability and operational impact during the conversion are all unknowns.
I remain seriously concerned that the TTC is playing a dangerous game with capacity of the subway system and views the downtown relief line as a far-distant, last resort fix. This will push more and more passengers into a single route and make the system even more vulnerable to delays and disruptions than it is today.
Today’s TTC meeting produced the expected result, the endorsement by the Commission of the project, but it is still subject to a long list of caveats. The definitive list is in the recommendations of the Toronto Executive Committee from January 5, 2009 (item EX28.1).
However, after presentations by me, Karl Junkin and David Fisher, there was considerable debate. Vice-Chair Mihevc wound up as the sole dissenting vote in the approval motion, but it was clear that the complexity of the issues related to the future of the Yonge Subway is now grasped by the Commission.
Well, almost all of them. Commissioner Perruzza seems to think that York Region should be free to build whatever it wants, connect it to Toronto’s system, and let us worry about how to deal with the aftereffects. Unfortunately, nobody has stepped forward, certainly not from York, offering to actually pay for it.
How Many People Will Fit On a Train?
Today’s supplementary report (at page 5) quotes the previous Rapid Transit Expansion Study (2001) citing a capacity per train of 1,200 in even older documents. The RTES goes on to acknowledge that the TTC had adopted a lower standard of 1,100 in recognition of the limitations of passenger handling when the space per person reaches the theoretical maximum. In fact, as I noted above, the TTC’s current loading standard is 1,000 per train. This has s significant impact on discussions of line capacity and demand.
Current service runs at 25.5 trains/hour, for a loading standard of 25,500 passengers. Depending on which report one reads, the actual current demand at Wellesley southbound in the AM peak is roughly 28-30,000 per hour and service is already overcrowded. Any regular rider knows this.
The new Toronto Rocket trains will increase the capacity by about 10 percent to 1,100 per train. However, with a projected future demand of up to 42,000, this still requires 38 trains/hour or a headway of 95 seconds. This is below the 105-second lower limit that even the TTC thinks is feasible.
Additional capacity could be provided by a seventh car, but this would have pervasive effects at carhouses, on signal systems and on platform doors should the TTC opt to install them before the seventh car is added to an existing fleet. The extra car would be shorter and this would change the standard spacing of train doors along the platform. Co-existence of 6-car and 7-car trains on the same line would be challenging.
If train capacity could be expanded to 1,200 with a 7th car, the 42,000 per hour demand would still require 35 trains/hour or a 103-second headway. This would push the line right to its new theoretical limit with no room to spare for surge loads or absorbing minor delays.
It is troubling that this analysis appears now in the January report, but did not show up in the original December staff presentation.
How Has Demand Changed on the Yonge Line?
The supplementary report includes charts of ridership on both the Yonge subway and on GO’s northern routes. Although the subway has recovered from its lower demand in the 1990s, its share of the market for trips into the core remains low at 46.2%, well below the 54.1% achieved in 1987. The reason for this is that the market is bigger, and GO has taken all of the growth.
A chart of GO’s ridership is a bit confusing because it ends in 1999 when GO’s modal split was 16%. In the text of the report, however, we learm that it has grown to 19% by 2006.
The TTC’s own fleet plan projects future growth at 1.35% per year on the Yonge line, and projections of new demand in decades to come show that over 4/5 of the additional demand will originate within the 416. The remainder will come into the city from the 905 on the Richmond Hill extension. 10% comes from Transit City routes feeding additional riders into the subway. 70% comes from population and employment growth plus other, unspecified network improvements. Only 20% comes from the Richmond Hill extension.
20%? This is truly odd. The forecast growth in riding south of bloor ranges from 7-12,000 in the peak hour. This means that the Richmond Hill extension is expected to contribute somewhere between 1,400 and 2,400 passengers to the peak load (the total contribution will be higher because some riders are bound for stops north of Wellesley).
The growth projections are in stark contrast to claims by some that stagnant employement would limit the room for growth in demand to the core (and therefore for new lines). If that is so, why is GO Transit planning to double capacity for its operations at Union Station?
In December, TTC staff claimed that peak hour demand on Yonge would be reduced by 2,300 riders shifted to the Spadina Subway extension. In the January report, it turns out that this diversion is for the peak period, and that the peak hour only accounts for 1,300. This is a trivial diversion of demand, probably within the margin of error of the demand model.
Also in December, TTC staff claimed that the current headway on the Yonge line (141 seconds) would be adequate to handle opening day demand in 2017. This is utter rubbish. The current headway is not handling the 2009 demand, let alone growth over 8 years or new riders attracted by the extension.
Again, the contrast between the December and January reports is troubling.
What is the Relationship Between GO and TTC Ridership?
This is a complex question, one that has been systematically ignored in decades of planning. The issue is complicated by the fact that GO has a completely separate fare system and doesn’t really want much to do with trips internal to the 416. However, as GO plans major service expansion, this artificial segregation of routes no longer makes sense.
By 2017, we will have TTC subway service roughly every 4 minutes at Richmond Hill Centre providing a trip to Union in about 40 minutes (possibly less depending on train speed). Meanwhile, nearby, GO will be running at least 6 trains per hour taking 35 minutes or so to get to Union (less if they figure out how to speed up the line). Depending on where you start and where you’re going, either service may be the more attractive. Metrolinx projections suggest that GO riders will outnumber TTC by over 2:1, but this presumes very frequent GO service.
An important issue in all of this is the question of fares and fare integration. GO charges more than the TTC, although there will be the question of access costs to Richmond Hill Centre. Will GO subsidise YRT and VIVA passengers? What will the TTC’s “Zone 2” fare be for riders north of Steeles (their own financial and demand projections assume a second fare)?
On a more regional note, with many frequent, all-day rail services, will riders’ perception of GO evolve into that of a “regional rapid transit network” rather than a commuter operation?
The TTC January report acknowledges the role of GO and notes that improvement of GO service is an important co-requisite of the subway extension. This is another important change — the acknowledgement that GO exists and will play a major role in regional travel.
The Downtown Relief Line
The TTC notes the Metrolinx projections that substantial load will be diverted from the Yonge/Bloor interchange by the Downtown Relief Line, but suggests that these projections may be optimistic. If the TTC has a better model, it is long overdue that differences between TTC and Metrolinx projections be resolved. We cannot have one agency publishing plans whose projections are dismissed, when it suits them, by another.
As I have already written, the DRL itself may not appropriately end at the Danforth Subway but could extend north to Eglinton. Much of the south end of the Don Mills LRT will require some form of dedicated infrastructure. The TTC still claims that they might cook up some form of surface alignment, but this ignores severe problems of neighbourhood impact, not to mention questions about how all of the transferring passengers would be handled. A junction of a Don Mills LRT, DRL and Danforth Subway at Pape Station is unthinkable.
If the DRL continues north to Eglinton, it will become a major route into the core for riders in the Don Mills corridor and on the eastern part of the Eglinton LRT.
The TTC regards the DRL as a complete waste of money because it diverts riders at considerable cost, riders who would be on the system anyhow. This is another specious argument. Any new line attracts riders because it brings more people in range of faster transit. Moreover, this line relieves congestion on Yonge and avoids very large capital and operating costs to increase that line’s capacity. That relief has a value both in terms of more attractive trips for riders and in keeping the Yonge line well below its maximum theoretical capacity.
As previously reported, the TTC hangs all its hopes for additional capacity on the combined effects of:
- TR trains (about 10% additional internal capacity)
- 7-car trains (another 10%)
- Shorter headways (105 vs 141 seconds, about 35%)
- Bloor-Yonge changes to handle the combined added capacity and pedestrian traffic
A consulting company will be retained soon to perform a detailed analysis of the Bloor-Yonge project including evaluation of a 6-platform configuration with additional capacity on the BD subway. This is discussed in some detail elsewhere on this site with the original 1988 TTC study of the scheme.
The report is due in the fall of 2009. What will happen if the BY project proves far more complex than anticipated? How will the line actually operate during the period when half of the existing platform capacity has been removed but the new centra platform on the Yonge line has not been added? What is the operational effect of concentrating so much traffic through one node on the system? How will the BD subway (regardless of the number of platforms) absorb PM peak traffic arriving at BY and at St. George at a rate much higher than today?
If the BY project is infeasible either due to cost or problematic construction, the TTC’s plan falls apart. Without that station project, they cannot substantially add to demand on the Yonge line, and they must find another way to handle future growth in riding.
The professional approach appears to be that everyone will cross their fingers and toes in hopes that somehow the station can be rebuilt and all of the other problems will just go away.
How Fast Will the Trains Run?
Reports about the coming automatic train control system state that they will permit a 10% reduction in trip times and hence fleet requirements. Some of these reports have been rather vague, if not downright inaccurate, about how this would be done. Today I confirmed that what is actually intended is to run the YUS in “high rate”, or its equivalent for the TR trains.
For those with very, very long memories, the H (and T) series of cars have two performance settings, low rate and high rate. The BD line ran in high rate for a time, and the trips were quite sprightly. Faster acceleration. Hills? What hills? Indeed, the occasional lucky rider would find a high rate train on Yonge where the trip from Eglinton to Finch would be cut from roughly 12 to 9-10 minutes. (The effect is less pronounced where stations are closer together.)
However, the H-1 cars had very serious hunting problems at 45 mph. At this speed, the trucks would sway violently, and operation at high rate was considered unsafe. Over the years, a series of excuses, including questions of track maintenance costs and a desire to maximize the car orders going to Thunder Bay, left both subway routes operating in low rate. One issue with moving to high rate is that timing signals must be reset to allow for the faster speed profile of the trains.
With the move to ATC, speed control is a question of software configuration within the train control system, and the concept of signal blocks and timing for those blocks disappear. Enter high rate operation and a saving of many trains. Note that running faster trains does not add to line capacity, only closer headways do that. If you run the same number of trains at higher average speeds, the headways get shorter and the capacity goes up. If you save on equipment by running fewer, faster trains on the same headway as before, the capacity is unchanged.
Planning a Network
The supplementary report states in several places that network planning is essential to addressing complex questions of system growth such as these. Indeed, the report notes that the new streamlined approval process creates problems because it is project focussed and does not allow for detailed evaluation of alternative schemes and network effects. We need to know which option we want to build before we start the EA because we won’t have time for discussions later.
This is foolish, and already we have seen both in some Transit City projects and in the just-announced Metrolinx Weston Corridor study a process to involve communities and discuss options before the EA even formally starts. In effect, we have put back part of the old, onerous EA process that was ditched in the name of efficiency.
The TTC and by extension City Council is coming to appreciate the need for network planning. That was supposed to be Metrolinx’ job, but unfortunately, they were more interested in building an “end state” 2031 plan that is at best a guideline, not a definitive network. Metrolinx is also paranoid and argumentative about any suggestions that their plan might change even though their own plan states that it is an example of what might be done, not cast in stone.
The real challenge for Metrolinx (and through them Queen’s Park) will come with the hard reality of budgets. They have a $50-billion plan, but only $11-billion or so in the bank. It’s easy to draw lines on maps when you don’t have to pay for them, and your only concern is that every part of the GTAH gets something to keep it happy.
Sadly, the underlying planning by Metrolinx gives little guidance on how we get from today’s network and demand to the wonderful new world where Metrolinx logos will be more common that Tim’s, Starbucks and the Second Cup put together.
This is a time for thorough, public debate about the scope of transit improvements needed in the GTAH. The TTC, prodded by criticism including my own, has at least recast their own subway plans with a broader view of issues and the options we must consider.