Yesterday’s launch by York Region of their Yonge Subway Now website brought to the fore the question of just how much room remains on the Yonge Subway for additional riders. Over many years, claims about capabilities of new subway technologies together with changing projections for future demand have left Toronto in a position where its subway is badly overloaded with little relief in sight.
This article traces the evolution of those claims and the reality of what can actually be provided to show that building a Relief Line is not a project for a future decade but one that must begin now.
The origins of this problem can be traced back two decades to the ridership plummet brought on by the mid-1990s recession when the TTC lost roughly 20% of its demand. The Yonge line, which had been packed full in the late 1980s, suddenly had room to spare and the idea of northern extensions took root then. An eastern relief line had been on the books for years as a Queen Street subway swinging north to Thorncliffe Park and Don Mills, but it was beaten out through a combination of suburban political pressure and an overt distaste for the line in TTC management. Gary Webster, former Chief General Manager, said on more than one occasion, in effect, why build a new line when you can fit the riders on the one we already have.
The TTC expected to implement a new signal system eventally allowing them to operate much more frequent service and hence greater capacity, and that was pretty much the end of the debate. Issues such as station capacity (notably at Bloor-Yonge where a billion-dollar project would add transfer capacity between the Yonge and Bloor lines) were treated as marginal considerations, not as a central, necessary part of any capacity growth.
The DRL was also hijacked for a time by Queen’s Park’s love for new technology with a proposal for the Scarborough RT/ICTS technology that would have placed an elevated structure on Front Street directly in front of Union Station. Historic sensitivity was never their strong suit.
The subway signalling system became part of the expansion dreams in 2008 when the TTC authorized the first steps in replacing the aging technology on the original 1954 Yonge line. It was a period of great uncertainty and cutbacks in funding on many fronts including transit, and the TTC feared it would not get funding for a simple “state of good repair” signal project. This became a project as much about enabling growth in subway capacity as one of badly needed maintenance, and it perpetuated the idea that lots of room would be available with the new system.
2.47 The ability to improve capacity on the line has in the past been limited by the current train control system, which can only handle the turn-around of 25 trains per hour. As part of the Metrolinx Quick Wins package of rapid transit initiatives in 2008, the Yonge subway line capacity improvements were funded in order to quickly achieve customer and system improvements. The Quick Wins investment of $293 million provided funding to the TTC for the installation of Automatic Train Operation (ATO) and a one-third funding contribution towards the acquisition of higher-capacity subway trains – ‘Rocket’ trains. [Metrolinx Yonge North Benefits Case Analysis, p 17]
Anyone who has been following the signalling project knows that $293m is small change for everything the TTC needs to install. This was only an early stage in a complex project, not a full scale solution to provision of better service.
Various capacities have been claimed for a subway line, and the Metrolinx BCA actually says that 1,320 for a train is a reasonable number. The report attempts to skate around crowding problems by overstating the design capacity and effectively committing the TTC to running packed trains.
4.27 Demand levels forecast on the line are higher than the subway planning capacity estimates (38,000 passengers per hour per direction at 180 passengers per subway car). However, it is acknowledged that maximum allowable capacity is likely to be higher. Assuming 220 passengers per subway car, the forecast level of demand is below the allowable capacity for all options after ATO and Rocket train roll out.
4.28 The majority of demand on the line will come from transit users who will transfer from other services (bus or GO), but it is estimated that 1,909 new transit trips (i.e. modal transfer from auto) will be generated in the AM peak period for Option 1. [p 35]
There is a compound problem of an assumed headway that the TTC now feels is unachievable, and the use of a load factor high enough to guarantee substantial crowding and loading delays.
The TTC designs service on a lower basis, although for the Yonge line they have no choice but to overload the cars because they cannot run more trains.
- The scheduled service on Yonge is a train every 141 seconds, or 25.53 trains/hour. At the design capacity of 1,100 per train, this is a capacity of about 28k passengers/hour.
- They are actually carrying about 31k/hour at peak, and this works out to 1,347/train, roughly the level the report suggests is reasonable. Anyone who rides the subway knows what this looks like, and the problems it brings with crowding and delays at major stations. This is a crowding level that is acceptable only in an emergency, not as a matter of daily operation.
- The problem is made even worse by the fact that the TTC rarely achieves its scheduled service level because of crowding and long dwell times. This is compounded by door problems and medical emergencies flowing from the overcrowded trains.
In some of the earlier studies of the new signal system, and in the Yonge North BCA, the TTC and Metrolinx have stated a target headway of 105 seconds, equivalent to a capacity of 38k at 1,100/train. However, the TTC has since reduced the target service to 110 seconds, or 36k, based on constraints that would prevent actual achievement of the shorter headway. Even 110 seconds will be challenging given the way TTC terminals operate today.
The chart below is taken from the July 2016 CEO’s Report showing the actual throughput of trains southbound at Bloor Station in the AM peak.
The following chart shows the ups and downs of demand and capacity changes projected by Metrolinx. This material was originally in the Technical Summary of the Yonge Relief Network Study and reproduced in a Metrolinx Board presentation in June 2015. It shows:
- The existing demand of 31k/hour which exceeds the design capacity of 28k/hour,
- The projected growth in demand to 2031,
- The projected diversion of riders to the Vaughan subway (TYSSE) and GO/RER,
- The projected riding for the “base case” in the Network Relief study, and
- The marginal increase in demand caused by extending the subway to Richmond Hill.
The GO/RER diversion is affected by at least three factors:
- Improvement of service from Richmond Hill to Union on GO from every 30 to every 15 minutes,
- Improvement of service on the Stouffville corridor to 21 trains over the am peak,
- A fare differential between GO and the subway that limits the attractiveness of GO as an alternative.
The service improvement to Richmond Hill is already in GO’s plans, but the service for Stouffville is now proposed as 4 trains/hour from Unionville and 3 trains/hour from Lincolnville. Without seeing the details, it is difficult to say how the effect of GO/RER would raise or lower the diverted passenger count, but it is important to note the importance of this effect in creating headroom for projected new riders.
The technical report includes an option dubbed “RER Plus” that would have provided even more frequent service, notably to Richmond Hill, which was dropped from consideration because it had minimal effect on ridership. The exact details should be explored because it implies that there is a finite new demand that may have been sated (in the model) by the 15 minute headway in the “base case”. Moreover, the limited growth with RER could be a function of fare levels in the model, a factor which the SmartTrack and Scarborough Subway studies show affects where trips are assigned. Would the GO network do a better job of “relief” if its fares were comparable to those paid for subway travel?
The projected demand on the Yonge North extension is quite substantial, and much larger than numbers for either the Vaughan or Scarborough subways. Most of this demand comes from existing transit trips that simply feed into a subway station further north than they do today. In the chart below, “Option 1” in blue is a line built in a single project all the way to Richmond Hill as opposed to variations on an interim operation to Steeles, “Option 2”. The cumulative load southbound at NY Centre Station is 26.7k/hour, almost the entire design capacity of the subway before planned improvements through new signalling and automatic control. This does not leave a lot for riders from Sheppard southward.
(The figure below is taken from the Yonge North BCA.)
There is a big problem with this chart in that it conflicts with demand projections in the Conceptual Design Study [45MB pdf].
Projected ridership at Richmond Hill Centre Station increased 25 percent relative to the TPAP
The projected transit ridership – developed during the TPAP – was updated during the Conceptual Design Study to reflect revised 2031 land use projections from York Region and the City of Toronto. Results suggest that more people (about a 25% increase) will board the subway at Richmond Hill Centre Station during the morning peak hour in 2031 than previously projected during the TPAP. The resultant increase reflects the implications of new secondary plans and development applications received since the completion of the TPAP in 2009. Based on current analysis, it is projected that ridership volumes in the Yonge corridor will warrant a subway within the next 10 years (i.e. before the year 2021). The minimum ridership required to warrant a subway is approximately 10,000 passengers during the peak hour. Within the next 10 years, over 12,000 peak hour passengers are expected to board or depart at RHC station alone. By 2031, ridership at RHC is expected to increase to 14,000 peak hour passengers. [p ii]
Additional demand, not included in the studies, is expected to arise from increased residential density within Toronto south of Steeles. [Conceptual Study p 2]
The options for subway relief show substantially reduced demand at Bloor-Yonge station both for on-train loads and for transfer movements.
The 32,300 riders under “do nothing” are the same riders as in the “YRNS Base Case” above and they represent 89% of the future subway capacity before the effect of a Richmond Hill extension. The projected peak hour demand at North York Centre station of 27,626 implies either that there will be comparatively few riders boarding south from there to Bloor, or that there is very substantial turnover at major intermediate destinations to free up capacity for “in town” riders. Such turnover, however, would come at the cost of additional dwell time as one set of passengers leaves and another boards.
Although the modelling was done, according to the reports, with the Crosstown LRT included, we do not know how many riders it contributes and especially what the difference will be between existing bus feeder loads and the LRT line at Eglinton Station. The developing land use pattern at Yonge-Eglinton is at a higher level than originally planned, and this too needs to be factored into the model. What is missing is a recognition that subway congestion analysis is not only a matter for south of Bloor, but for the entire line north to Finch.
The study looks at accumulated demand, not at the operational side of the subway including a basic question: how frequent would trains have to be (or how low must demand be kept) to preserve an environment where the line does not bog down at stations and lose capacity through an inability to run its scheduled service. Moreover, there are capacity issues not just at Bloor but at stations further south on Yonge where clearing the platform between more frequent trains would be difficult.
On the Bloor line in the PM peak, transfer traffic will arrive at Bloor-Yonge northbound faster than it does today, and today’s rate already strains the lower level station. Overcoming this constraint would not be easy, and the construction project alone would bring substantial disruption for years.
The cost of expanding capacity in existing stations has never been included as a cost of the “do nothing” option even while the pricetag for a DRL usually brings much hand-wringing over affordability.
The Yonge Relief Network Study is quite clear: the “long” version of the DRL will be needed and alternatives to it will provide only superficial, short-term relief.
8.2 Analysis of the base case indicated that the development of RER provides significant relief to Line 1 prior to 2031. Of the anticipated growth in ridership of 6,600 passengers in the peak hour, RER will abstract 4,200. However, by 2031 it is anticipated that ridership on Line 1 will be approaching its theoretical capacity and further transit development will be required.
8.3 The overarching conclusion of the YRNS is that a new rapid transit option will be required post 2031 to abstract growing demand on Line 1 and accommodate future growth. [p 49]
8.11 This analysis should determine a preferred project concept for Line 1 relief that builds upon the surface and tunnelled options included in the YRNS evaluation. Phasing of potential subway components is a critical consideration for this work. Based on the YRNS evaluation, it is likely that a short subway option may provide relief post 2031, with an extension to Line 4 at Don Mills providing longer term relief and a complete transit network.
8.12 The development of this project should consider key factors influencing crowding on Line 1 in determining a timeline for implementation and project development:
• Rate of population and employment growth
• Impact of fare integration on RER’s abstraction from Line 1
• Impact of new transit projects above those included in the YRNS base case
• Crowding impacts at Bloor-Yonge Station [p 50]
Future demand growth will be a critical issue in the analysis given that the Yonge line is projected to be effectively “full” by 2031. Except for a recent levelling-off of demand (which itself could be a factor of capacity constraints and unattractive service), an annual growth rate of 2% is not out of the question especially with strong growth of office space in the core area. Only a slight change in that rate would eat through any residual capacity very quickly.
There is a major problem underlying this debate with competing ridership projections and service plans. From York Region’s point of view, any model showing that there is room for growth from a subway extension is ideal. From Toronto’s point of view, subway capacity is already overcommitted even with expected improvements. From the TTC’s point of view, there is an inconsistency in presentation of options, particularly the cost and effect of “do nothing” for the existing downtown network. A consolidated, consistent view of the issues is long overdue.
Toronto talks boldly about investing in its future, but this ideal is short-changed when it comes to transit capacity. Subway relief will not be simple or inexpensive, but this is an investment that must be made.