Future Demand on the Downtown Subway Network

Recent discussion about the Downtown Relief Line study and its Terms of Reference sent me back to the TTC’s Downtown Rapid Transit Expansion Study (DTRES) published last year for a look at the demand projections.

What I found there was rather troubling.

The TTC looked at three scenarios to model future shortfall in network capacity by 2031:

  • The existing TTC and GO networks
  • An enhanced “reference network” with improved subway and GO service
  • The reference network plus the Yonge extension north from Finch to Richmond Hill

The demand model outputs appear in three separate tables within that study, but it is not until we consolidate the information that some anomalies really jump out.


There are four sets of numbers in this table with columns corresponding to the three model networks.

  • Capacity:  This gives the capacity of each route based on service levels and train lengths.
  • Inbound demand:  This is the modeled demand on the network.
  • V/C:  This is the ratio of demand to capacity.  A value near to or greater than 1.0 indicates that the line will be over capacity during at least part of the peak period.
  • Inbound deficiency:  Where the capacity is lower than the demand, this is the magnitude of the shortfall.

The capacity of the reference network is about 50% greater than the existing one.  Note that for the northern GO services, ten-car trains are assumed although 20% could be added to the capacity with 12-car trains on the same presumed schedules.  (The model also considered the east-west GO services and their effect in draining trips off of the BD subway that would otherwise contribute to demand south of Bloor Station.)

The modeled demand is also about 50% greater than the demand that the model assigns to the “existing” network configuration.  This shows the modeled effect of increased transit service on network demand.  However, this also begs the question of where those trips would be if the TTC and GO improvements did not take place.  An obvious useful addition to the discussion would be the added road trips, or the trips simply not taken because there was no network capacity to handle them.

The big surprise is that there is almost no difference between the total demand with or without the Richmond Hill extension.  Indeed, most changes are re-assignments of trips from GO lines and the University subway in the “reference” network to the Yonge subway in the “reference + YSE” network.

Route                    Without YSE     With YSE
University Subway           25,100        23,500
Yonge S of Bloor            35,800        39,400
Barrie GO                    7,500         7,400
Richmond Hill GO             2,500         2,200
Stouffville GO               8,600         8,000
Total                       79,500        80,500

Why would we spend billions of dollars building a subway to Richmond Hill to carry no more total riders on the network than we do without it?

There are two obvious responses to this question:

  • Some of the new trips have destinations at or north of Bloor Street and therefore they do not contribute to the count of riders into the core area.
  • In the model’s world, the subway extension does not attract any net new trips beyond what would occur simply with better service on the subway to Finch and enhanced GO services (i.e. with the reference network).

This is a rather strange situation considering that the holdup on building the Richmond Hill extension arose from the claim that it would overload the Yonge line.  However, in the model, it does this primarily by attracting trips that would otherwise have been on GO or on the extended University subway.

(At this point, I have to wonder whether a similar methodology produced the inflated ridership projections for the Scarborough Subway, but that is another matter.)

The model shows very low ridership on the Richmond Hill line.  Indeed, the greatest number of riders (2,900) is obtained with the “existing” network and the value falls even though GO service is improved in the “reference” and “reference + YSE” networks.  This implies that the model prefers to assign trips to the “faster” Yonge subway especially when it goes all the way north to Richmond Hill.

On the BD line, although an increased capacity is included in the model (about 27%), ridership only goes up in the section east of Yonge.  This implies either that demand from the west is static (difficult to believe) or that it is going somewhere else in the model.  Where?  Is growth assigned mainly to GO because it competes well with the subway for traffic in Mississauga while to the east Scarborough is poorly served by GO?

There is no question that Toronto needs more capacity into the core area, but the modeled numbers in the DTRES are suspect.  If anything, they may understate the problem and the potential benefits of alternatives to stuffing more riders onto the Yonge subway.

The TTC has a long history of downplaying the need for anything beyond Yonge subway capacity expansion (more trains, new signals, bigger stations) to the detriment of long-term planning for better GO service and new TTC subway or LRT services.  For many years, all we heard about from TTC was the need for a Richmond Hill subway.  Any other project was cold-shouldered because it threatened that favoured scheme.  Only when capacity problems could not be ignored did the TTC turn to the “DRL” as a possible solution.

Toronto has been ill-served by this blinkered planning, and coming studies on the future of the transit network (without regard to the paint scheme on the vehicles) must be based on a fair and accurate assessment of how new and improved services will contribute to moving passengers and limiting the growth of congestion in Toronto.

Inching Ahead on Downtown Relief (Updated)

Updated December 4, 2013 at 7:00 pm:

Toronto’s Planning & Growth Management Committee considered the proposed consultation process from City staff regarding the Downtown Relief Line and approved it with only one minor amendment, that a contest be set up to name the new line.

Those of us who remember the last such contest will know that it produced, as a moniker for the line now serving Scarborough Town Centre, that heartwarming name, “RT”.

Two presentations were made by Metrolinx and by City staff.

The Metrolinx presentation gives an overview of the Yonge Relief Network Study which will consider capacity problems and approaches to relief from a network point of view including the possibility that some of the modeled demand can be shifted from the subway to GO Transit.  Included in this is a map (page 4) showing the projected locations and degrees of capacity shortfall on the 2031 network.

What is quite striking about this map is that while there may be a severe problem south of Bloor Station, the situation north to Sheppard is not exactly rosy with demands ranging up to 100% of capacity.  The need to divert demand north of Bloor is quite evident, although it is rarely mentioned in discussions of “downtown relief”.

The City presentation gives a précis of the report, but shows more clearly (page 10) that at this stage all that is happening is consultation on the Terms of Reference and the Public Consultation that would occur in a future study.  This will come back to Council in the spring of 2014 for approval of the full study.

This is a hybrid version of the Environmental Assessment process.  A Transit Project Assessment is intended to be fairly brief (120 days) based on a predetermined project.  However, the DRL is so far-reaching and expensive a proposal that the City wants to ensure a valid review of all options has taken place before locking in to the formal TPAP review which offers little opportunity for amendment.

An important factor will be the co-ordination and integration of the network-wide study by Metrolinx and the local route and station planning by the City and TTC.  Metrolinx will have a good sense of the network options by mid-2014 just at the point the City process enters its “alternatives” phase and the creation of a “long list” of options.

The greatest challenge in the short term will be to ensure that the Terms of Reference do not preclude options that should be considered.  In this regard, limiting the study area to the initial phase of a DRL from downtown to Danforth may be valid for detailed planning of alignments and stations, but not from the larger view of how a second phase north to Eglinton or other extension options might affect the route selection on the first phase.

The City plans extensive consultation on the ToR including public input from community and advocacy groups.

Another challenge will be the credibility of demand and capacity figures used to model the future network.  For reasons that I will discuss in a separate article, some of the modeling numbers cited by the TTC’s Downtown Rapid Transit Expansion Study are suspect on two counts:

  • The projected service levels and demands on the northern GO lines (a) do not necessarily match with Metrolinx’ plans as stated in The Big Move, and in particular demand assigned to the Richmond Hill GO line is trivial.
  • The projected increase in total demand flowing into the core from the north with the Richmond Hill subway is only 1,000 per peak hour.  Much of the extra subway demand comes from trips reassigned from the GO rail lines.

If the numbers for the Yonge subway and the GO network are dubious, then what of the potential demand from a DRL, especially the component north of Danforth to which the TTC assigns a low demand and priority?  Earlier at the same P&GM meeting, City staff presented information about the “Feeling Congested” consultations that have been underway.  This included a preliminary evaluation of various additions to the transit network in which the Don Mills corridor, including the proposed LRT to Steeles, ranked highly (see map, page 19).

A future line cannot both be highly ranked and of little value, and this suggests at a minimum that the criteria by which it was evaluated are not the same in each case.  The challenge for the DRL’s Terms of Reference process is to determine which evaluation is correct.

The original text of this article from December 2, 2013, follows the break.

Continue reading