At its board meeting on June 25, 2015, Metrolinx will consider an update on the study of capacity relief for the Yonge Street Corridor in Toronto.
The report states that projected demand on the Yonge line can be handled for the next 15 years:
1.a. Significant relief to the Yonge Subway will be achieved with currently committed transit improvements underway including:
i. TTC’s automatic train control and new subway trains;
ii. The Toronto-York Spadina Subway Extension; and
iii. GO Regional Express Rail
1.b. Based on [the] above, more rapid transit service and capacity that is currently funded and being implemented will meet the future 15 year demand, assuming current forecasts on the growth rate of downtown employment and the implementation of TTC automatic train control on the Yonge Subway.
Continued work is recommended:
2. Direct the Metrolinx CEO to work with the City of Toronto City Manager and the TTC CEO to develop an integrated approach to advance the Relief Line project planning and development, incorporating further business case analysis and the findings of the Yonge Relief Network Study to:
- further assess the extension north to Sheppard Avenue East to identify a preferred project concept,
- inform the planning underway by the City of Toronto and TTC to identify stations and an alignment for the Relief Line from Danforth to the Downtown area
- continue to engage the public in this work as it develops
3. Direct staff to work in consultation with York Region, City of Toronto and the TTC to advance the project development of the Yonge North Subway Extension to 15% preliminary design and engineering.
The emergence of a variation on the Relief Line that would operate north to Sheppard is quite a change from days when even getting discussion of a line north of Eglinton was a challenge. The context for this emerges by looking at the alternatives for “relief” that were considered and how they performed.
The next report to the Metrolinx Board will be in Spring 2016. The challenge will be to keep planning for a Relief Line “on track” in the face of the excitement and political pressures for GO RER, SmartTrack and a Richmond Hill Subway.
Politically, this study exists in a challenging environment where just about every other transit proposal on the planet (and possibly several neighbouring galaxies) takes precedence because almost nobody wants to grasp the nettle of advocacy for more subway capacity into the core area. The idea of a line east then north from downtown has been around for a very, very long time including incarnations as the Queen Street subway and as a Skytrain/RT line when this technology was expected to blanket the city with low-cost rapid transit. Later, the TTC had grandiose dreams that it could carry every passenger who might arrive on the Yonge line with shorter headways and heroic reconstruction of the Bloor-Yonge Station.
There was a period in the mid-1990s when the ridership losses of a then-recent recession drove subway demand well below its historic highs of the 1980s, a level to which it has now returned. This gave a false sense that there was plenty of room to spare. By the time an extension of the Yonge Subway north to Richmond Hill was on the table, Toronto Council said “enough is enough” and demanded that capacity relief be guaranteed before such an extension would be built. That spawned the current round of Relief Line reviews.
Additional historical background is available in a January 2009 TTC report.
When relief first became an issue for Metrolinx in September 2013, the RER plan did not yet exist, although a lot of the background work to it was already underway. Many proposed mechanisms for capacity relief did not consider the possibility of substantially increased GO capacity as a network-wide project. The update notes:
With RER – which will provide medium term relief – the focus is now on when additional relief is needed for Yonge Line and then on the intervention. [p. 10]
Meanwhile, the City of Toronto is undertaking a study and consultation about the lower end of a Relief Line with corridor options broadly speaking via Queen or King through downtown, and via Broadview or Pape through Riverdale. The clear favourite already is a Pape/Wellington alignment, and an important consideration is the ability to extend north from Danforth.
Downtown Toronto has received the bulk of growth for the past two decades, and this trend is accelerating as the desire of firms to locate offices downtown accelerates. Downtown now has 35% of the jobs, and the number has grown by 51% over 1993-2013. In the decade from 2003-2013, growth has been 80% reflecting the acceleration out of the 1990s recession. [For greater detail on growth within Toronto, please see the May 2015 edition of How Does The City Grow? from Toronto’s Planning Department.]
A series of charts ending on page 20 show the anticipated evolution of demand and capacity.
Yonge Line Capacity
The scheduled service today on the Yonge Subway (Line 1) is 141 seconds, or 25.5 trains per hour, during the AM peak period. At a design load of 1,080 passengers per train for the TR fleet which serves this line, the peak capacity is roughly 28,000 per hour. Line capacity is projected to increase to 36,000 per hour through the implementation of automatic train control (ATC) in 2020.
This will require that peak headways be reduced from 141 seconds to about 110 seconds, or close to 33 trains/hour. ATC is technically capable of handling this subject to geometric constraints such as terminal configuration, and expedited flows through congested areas such as Bloor-Yonge made possible with better regulation of train spacing.
Over the years, the TTC has made various claims for the ATC-related capacity increases, but these have gradually declined. As recently as fall 2012, they claimed a 35% potential bump in capacity to 38,000 per hour, but this has been scaled back to 36,000. Achieving this will stress the capability of Finch terminal, and it is uncertain whether a 110 second headway can be operated without change in the terminal geometry (for example, use of a split terminal involving multiple tail tracks, or operation of partial service to a point further north such as a Steeles Station). Much is riding on the claims of what ATC can achieve with the existing network, something we cannot verify until the new signal system is in place and crew management at the terminal becomes razor-sharp.
Another way to increase capacity is for the TTC to buy longer trains for YUS, but that is not in the cards until the mid 2020s. The existing TR sets would move to BD and the new trains would operate on Yonge with roughly 10% more capacity. This would either add to line capacity, or drop the required headway back to about 120 seconds which is not quite as tight as 110. (30 trains/hour with a capacity of about 1,200 gives 36,000 passengers/hour.) This is still one of the possible options according to a conversation I had with TTC CEO Andy Byford at the Board meeting on June 21.
The Evolution of Demand
Current subway loads exceed the design level, and the line is actually carrying 31,200 per peak hour. Of this, about 1,300 (4%) of the riders are expected to shift to the Spadina extension (TYSSE) when it opens at the end of 2017. (Total riding on Spadina will go up somewhat more thanks to new riders attracted by parking, and increased riding from YRT/VIVA feeder services.)
In the absence of other changes, growth will add 6,600 riders on Yonge by 2031, and a Richmond Hill extension would add a further 2,400. Without relief, this would bring the total demand to 38,900 passengers/hour, well above the line’s capacity. A related problem is that Bloor-Yonge Station cannot accommodate the traffic likely at such a level, and a $1-billion plan exists to very substantially increase passenger flow capabilities there. Any scheme that avoids this pressure also avoids the cost, time and disruption for the construction project.
GO RER (including SmartTrack, it would appear) is expected to divert 4,200 riders/hour from the Yonge line. This is actually not a very large number compared with the total demand. This has major implications for the “selling” of SmartTrack as the line that will solve every problem. GO’s expansion may well carry more riders, but they will mainly be commuters who would not otherwise be on the Yonge line. No demand estimates for the GO RER services are included in the Metrolinx report.
This leaves us with a net projected demand in 2031 of 34,700, slightly below the planned 36,000 capacity, but with little headroom for growth or error. Under this scenario, the Relief Line might not be needed “today”, but there is no time for dallying with the thought that serious planning for it can be put on the back burner.
Options For Relief
Metrolinx planners modelled various options to see how they would affect demand on the Yonge Subway:
- Increase GO service beyond the planned RER levels, a so-called “RER Plus”, on the Richmond Hill and Stouffville corridors
- Build a Relief line either:
- Downtown to Danforth (“Relief Line Short”)
- Downtown to Sheppard (“Relief Line Long”)
- Danforth to Bloor via Downtown (“Relief Line U”)
- Build a surface LRT corridor from Downtown to Don Mills
The surface LRT option differs from the subway in that it has fewer stops (8 vs 14) and therefore is intended more as a Don Mills to Downtown service than one that also picks up demand along the way. The anticipated running times are comparable, but this would depend on the availability of a totally protected surface alignment. It is hard to believe that a surface line running on streets could achieve a 20 minute trip time as this would give an average speed in excess of 45km/h.
In the chart above, the “Do Nothing” option includes RER and SmartTrack.
The RER+ network shows almost no effect on subway demand. I asked Metrolinx about this.
In the chart on page 31, RER+ is shown to have a minimal effect versus “Do Nothing”. Is it correct that the “base case” here is actually GO RER, not the current network configuration? In other words, is it a situation that RER itself will bleed off some amount of riding and there isn’t much left over for enhanced RER to pick up in this time frame?
Metrolinx (Leslie Woo, Chief Planning Officer, via email): You are correct. The base case includes RER and all other committed/funded projects. “Do Nothing” means the base case or nothing beyond current commitments. The low attraction of RER Plus is an indication that RER really optimizes the diversion from the Yonge Subway.
In other words, RER picks up almost all of the riding available for diversion from the Yonge line, and further enhancement of service has little effect in the model.
The projected demand for the “Long” option to Sheppard is very impressive, well into the territory where subway technology is the only reasonable choice, especially considering the areas through which the line must pass. The offsetting benefits on the Yonge and Bloor-Danforth subways are very large to the point where expansion of Bloor-Yonge capacity would not be necessary. Moreover, the reduced demand on both existing subway lines would have effects on fleet size and operating costs. All of these are offsets against the cost of a “Long” Relief Line.
Indeed, the Long Relief line would cut demand on the Yonge Subway by about 1/3 (11,600 riders out of 32,300), an impressive contribution to the network. It is hard to imagine a Yonge line with barely more than 20,000 riders on it in the peak hour.
On the Bloor line east, demand falls by about 1/4.
The comparative cost of the Long Relief Line at $7.8b and SmartTrack at $8b shows an obviously much greater benefit for the Long Relief Line. An obvious question here is the need to reconcile claims for SmartTrack with the service levels and ridership that come out of the demand model. SmartTrack may well have riders, just not the ones we seek to divert from the core of the subway system.
Both the Relief Line and LRT options will receive further study, but RER+ has been dropped from the mix. This is important if only because it turns the focus for additional core-oriented service to a line serving the inner part of Toronto that can have a major effect in redistribution of demand, rather than on the outer ends of the network. This is not to minimize the importance of GO, but to recognize that there are two separate markets and one “solution” cannot serve both of them.
A technical report with details of the demand projections will be available in July 2015.