Update 4 October 21, 2012 at 8:30 pm:

It’s intriguing to look back at coverage of the DRL the last time this was a major issue.  Mike Filey passed along a clipping from the Star from December 2, 1982 that makes interesting reading.  My comments are at the end in Postscript 2.

Update 3 October 20, 2012 at 3:20 pm:

A postscript has been added discussing the various demand simulations as a group rather than individually.  Charts of total demand southbound from Bloor Station as well as pedestrian activity at Bloor-Yonge are provided to consolidate information from several exhibits in the background paper.

Update 2 October 19, 2012 at 11:00 am:

This article has been reformatted to merge additional information from the background study as well as illustrations into the text.

At its meeting on October 24, 2012, the TTC will consider a report on the Downtown Rapid Transit Expansion Study.  The full background paper is also available on the TTC’s website.

A study by the City of Toronto and TTC, including consultations with Metrolinx, concludes that transit demand to the core by 2031 will grow at a rate that exceeds the capacity of all of the current and planned transit facilities.  Ridership will be 51% higher than today.  The residential population south of College from Bathurst to Parliament will grow by 83%, and employment by 28%.

Capacity is an issue today as Table A-1 in the background paper shows.  Several corridors into downtown are already operating over their design capacity.  This is particularly the case on GO where the target is to have few standees, and there is more room for additional passengers in the design capacity than on the TTC subway services.

Table A-2 shows the projections for 2031.  All of the shortfalls are on GO, but the TTC lines are close to saturation.  This presumes a considerable increase in the capacity of various lines.  For example, the YUS goes from a design capacity of 26,000 to 38,000 passengers per hour (pphpd), an increase of 46% which may not actually be achievable.  Similarly, the BD line goes to 33,000 pphpd, an increase of 27%.

Exhibit 1-10 shows the components of projected capacity increase including 36% from running trains closer together.  As discussed at some length on this site previously, the constraints on headways arise at terminal stations.  A 36% increase in trains/hour implies a headway of about 100 seconds as compared with 140 today.  This cannot be achieved with existing terminal track geometry, not to mention the leisurely crew practices at terminals.

On the GO lines, the projected capacity on Lakeshore West doubles, and smaller increases are seen on other routes.  It is worth noting that the projected capacity of the north-south corridors to Stouffville, Richmond Hill and Barrie are nowhere near the level of service implied by The Big Move, probably because these lines are not targets for early electrification.  This contributes to the capacity shortfall in the northern sector.  Recommendation 1 of the study includes encouragement that Metrolinx review the possibility of increased capacity in those three corridors.

The full list of lines included in the modelled network can be found in the background study at section 1.2.1.

The streetcar system (plus sidewalks and bike lanes) will absorb the growth in short-distance travel to and from the core.  This has implications for the future of the streetcar fleet once the new LFLRVs come into service — the immediate retirement of older equipment may have to wait while the backlog of growing demand is addressed.  Cycling, pedestrians and transit will require dedication of an increased amount of road space including transit priorities and aggressive controls on abuse by motorists and delivery vehicles.  With growth numbers like those, the idea that people will breeze through a congestion-free downtown by car is laughable.

Exhibit 1-8 in the background study gives a revealing breakdown of trips arriving in the core area during the AM peak.  GO has 88% of the trips from Peel Region,  95% of the trips from Durham (on a much smaller base than Peel), but only 55% of the trips from York Region.  This shows the degree to which GO has not picked up demand from the north, and this demand is now filling up the Toronto subway system.

This problem persists in the 2031 projections in Exhibit 1-9.  The TTC’s share of trips from York Region to downtown falls only from 45% to 41%.  GO may be adding capacity, but only slightly faster than what is needed to handle greater demand, not as an incentive to shift riding from the TTC network.

Also revealing is the breakdown of trips within the 416.  Of the trips from east and west of downtown entering the core, the surface routes carry about 30%.  The outside-416 trips swamp the local 416 trips, and surface trips only make up 8% of the overall total.  However, the surface system is an important part of local travel within the city, and this role should increase as the population south of Bloor rises.

Exhibit 1-9 implies a fall in the absolute number of trips handled by the surface network outside the core area boundaries.  The percentage share for surface routes falls from 30% to 23% while the total number of trips does not rise much.  This implies a diversion of trips to other modes without the addition of any east-west subway capacity other than better service on the BD line.  I am not sure whether this represents a real shift or if it is an artifact of the modelling process.

Table 1-5 gives comparative streetcar capacity figures for the downtown routes with three sets of figures: existing, the effect of a 1-for-1 replacement by new LFLRVs, and the capacity of a nominal 3-minute headway.  What is not shown is any indication of the latent demand on streetcar routes whose service has been frozen for years, nor of the effect of current and planned development that will increase demand along the streetcar lines.

Elsewhere in the report, the TTC claims that 2,600 is the practical capacity of a streetcar line based on a 3-minute headway of LFLRVs.  This statement must be challenged on two counts.  First, it is already agreed that more road space and time must be given to transit, not less.  Second, if we presume that 2,600 is the upper bound for a surface line in mixed traffic, we face huge costs for demands that are far below those where subway operation is affordable.

The subway network’s function will be to handle medium-distance trips, although “medium” is a bit of a stretch when the subway will eventually extend into territory once the preserve of GO transit’s rail and bus network.  Indeed, one reason for crowding on the subway is the limited growth of GO thanks to constraints at Queen’s Park.  Metrolinx talks a good game, but it’s always “subject to funding”, and little of that materializes.

Other recommendations in the background study include:

• Do not proceed with the Yonge Subway Extension in advance of the provision of additional rapid transit capacity into the downtown.
• TTC and the City of Toronto undertake the studies and actions needed to protect for a possible future expansion of Bloor-Yonge station and develop a plan for improvements that will be needed in the future.
• City of Toronto continue to study means of reducing congestion in the downtown area via the optimization of existing infrastructure in its ongoing “Downtown Transportation Operations Study.”
• Maintain, and where possible enhance policies in the City’s Official Plan that will help to minimise the need for future investments in rapid transit facilities.
• TTC conduct further investigation into the future demands and transfers expected at subway stations in the downtown and identify those stations that should be given priority in TTC’s station modernization program. King Station in particular will see high passenger demand and operational issues regardless of the presence of a DRL.

These recommendations emphasize that the DRL is not something about which decisions can be made in a vacuum.  It is a pre-requisite for expansion of the existing Yonge line, and other parts of the network, both transit and roads, will have to adapt to provide greater capacity in the system.  This is not a matter of drawing one line on a map, finding some money, and building it.

Two families of network improvements are included in the TTC study:

• The Downtown Relief Line (DRL) and variations
• A Lakeshore Subway (or equivalent) in the GO corridor

Because more pressure on network capacity lies east of Yonge, each group of options includes an “east only” version as well as an “east + west” option.  The DRL options include versions ending at Pape Station or extending north to Don Mills and Eglinton.  This gives six configurations.

[Note: The TTC study shows the DRL going under King and up Roncesvalles, but the alignment is only for the purpose of illustration.  Commenters should note that I will not entertain yet another battle among competing alignments here.  We have done this issue to death already.  What is clear is that a “DRL” cannot hit everyone’s pet project on its way across downtown, and Toronto needs to decide just what function that line would provide.]

An east-only DRL ending at Pape Station or at Eglinton:

An east-west DRL from Dundas West Station to Pape Station or to Eglinton:

A east Lakeshore line from Union to Rouge Hill

An east-west Lakeshore line from Long Branch to Rouge Hill

Leaving aside the choice of inside-416 stations for the termini of a Lakeshore route, the basic problem it has is that it competes with the Bloor-Danforth line but does not shift much riding off of Yonge-University.  Attention, therefore, focuses on the DRL options.

The TTC notes that some benefits could be obtained by improvements to GO’s north-south services, but these are outside the scope of their study.  Some work has already been done looking at a network adapted from the one used in the electrification study.  This is precisely the sort of detail that should be made public to inform discussion of options rather than remaining hidden within Metrolinx.  Any debate about future funding streams needs to know what is possible and how soon we might see it.

What we are now seeing is the cumulative effect of decades of deferred investment in transit across the region at a time when transit demand is taking off.  A few more trains here or there, and a subway extension every few decades, and piles of glossy reports just are not enough.

The background study contains a short history of the evolution of downtown rapid transit plans (section 1.1).  This includes an odd remark about the retention of streetcars, now celebrating its 40th anniversary:

a citizen’s group in support of streetcars lobbied to retain the streetcar lines in favour of a new subway, [the Queen line]

Well, no, actually.  What was happening was that the TTC planned to gradually eliminate streetcars and replace them with inferior service, and with no guarantee that a Queen subway (which would not have served large areas of the streetcar network anyhow) would ever be built.  A tiny bit of revisionist history.

The paragraph goes on to note that the Queen subway was cancelled to concentrate efforts in the suburbs.

Band-aid efforts on the existing network are not enough either, although the demand projections include the high end of the TTC’s projected capacity range on YUS.  Only one month ago, in its 2013 Capital Budget report, the TTC downplayed the need for an additional rapid transit line into downtown to relieve the Yonge-University subway:

… all of these projects could conceivably improve carry capacity on that line by 40% or more over time. Funding and completing them could put off the need for the $10 billion or so Downtown Relief Subway Line by 10 or 20 years at a fraction of the cost.

The projects in question include:

• Automatic Train Control
• Fleet expansion to allow ATC
• A shift to 500-foot long trains (essentially replacing the TRs which have not all been delivered yet so that they, in turn, would replace the T1s in the Bloor-Danforth line) together with carhouse and storage expansion for the larger fleet (in length and numbers of trains)
• Installation of platform doors (not funded, and described in private conversation by Andy Byford as not necessarily a viable solution)
• Capacity expansion at Bloor-Yonge Station

The TTC has not provided a consolidated costing of these improvements nor have they indicated which of them might not be required if a DRL were implemented.  Almost all of the list above is not funded, and cost estimates, such as they exist, are scattered through multiple projects in the Capital Budget.  The comment cited above uses a DRL cost ($10b) which is larger than the price quoted even for a full DRL from Dundas West to Eglinton ($8.3b).  At this point, the TTC is only talking about a $3.2b route from downtown to Pape Station.  One wonders just how deliberately misleading the anti-DRL statements from TTC management are and have been for past decades.

A further problem is that the TTC mixes improvements that actually provide net new capacity with those intended to improve reliability.  If, for example, there are fewer track fires because of platform doors, this improves reliability, but it does not increase capacity unless we assume that such fires are so common that they will always interrupt peak service.  Fires and suicides are not the only source of service delays, and an external review of the TTC has already reported that until they can get their reliability in much better shape, achieving much higher capacity is impossible.

Pape is not the place to stop, and the line should go north to Eglinton.  This will intercept more traffic and will eliminate some of the need for passenger transfer capacity at Pape and Danforth.  Such a line would provide much improved transit to major neighbourhoods at Thorncliffe and Flemingdon Parks, and would be more than a diversionary route for traffic on the Danforth subway.

Exhibit 1-11 gives an overview of existing and projected transfer movements at Bloor-Yonge.  These will increase substantially by 2031, and the station as now configured cannot handle the added pedestrian activity.  The TTC has a separate study underway of capacity at Bloor-Yonge, but this has not yet been published.  The question then will be whether the proposed improvement to station capacity is physically possible, affordable and acceptable to nearby landowners who may be affected.  If we cannot handle more transfer movements at Bloor-Yonge, a discussion of increase line capacity runs aground fairly quickly.

Detailed projections for the DRL show that it will have considerable effects in reducing demand at Bloor-Yonge.  When this is combined with the possible benefit of increase north-south capacity on GO, heroic work at Bloor-Yonge to accommodate new transfer capacity may not be required.  The TTC, however, seems to have the attitude that it is needed no matter what.  The validity of this position should be challenged as the relative benefit, costs and effects of various combinations of options are reviewed.

Exhibits 1-12 and 1-13 examine projected increase in station demands on the downtown “U” and shows very large changes in passengers entering the subway from Queen Street south.  Most affected is Union where boardings would triple from 5,700 to 17,300.  Other stations have comparable ratios (3X) but many fewer passengers.  The effect of higher subway ridership shows up in all stations’ departure numbers.  The combined effect of these changes shows the need for more station capacity along the Yonge-University line, not just at the two key stations.

The diagrams below are only those showing overall changes in demand flow for each option.  Many more details are available in section 4 of the background paper.

DRL east to Danforth:  The projected peak demand westbound to Yonge is 11,700 although fewer riders are diverted from the Danforth (5,600), Yonge (4,700) and University (1,000) subways.  The model presumes a GO connection at Gerrard that would attract about 3,000 transferees per hour at peak.  The number of boardings at Pape is 9,400, considerably larger than the number of trips diverted from the Danforth subway, and this likely shows the volume of traffic attracted from the north (something the study’s authors appear to have missed).  This option also reduces the west-to-south transfer demand at Bloor-Yonge to below 2001 levels.

DRL from Dundas West to Danforth:  When the western leg of the DRL is added, demand at the peak point east of Yonge goes up to 13,600 because destinations west of University have been added as options for riders using the DRL.  Anyone who has seen the substantial counterpeak demand on the King car going to offices west of downtown will understand this pattern.  The biggest change, understandably, is on the Bloor line west of St. George from which traffic would divert to the DRL west, and on the University line.  Projected peak demand on the DRL west is 12,900, but this depends on substantial transfer traffic at Queen and Dufferin.

DRL with Eglinton Option:  The Eglinton option does not have much effect on the peak point projected demand for the DRL east or for transfer traffic at Bloor-Yonge as compared with a line ending at Pape Station.  This implies that the model is not finding much “new” demand to assign to the corridor with a faster trip between Eglinton and Danforth than on the existing bus routes.

The Lakeshore options primarily affect the BD subway by diverting east-west trips away from it.  The effect on the Yonge line and on Bloor-Yonge station is considerably less than with the DRL options.  A Lakeshore service also adds even more demand to Union Station which will be straining to handle growth in GO Transit and waterfront-based trips.

As a general observation, the demand model appears to be “force feeding” the DRL by assumed GO-to-DRL transfers that require additional stops in locations where they may not be physically or operationally practical as well as the station capacity to handle the passenger movements.  Whether riders would actually make these transfer moves requires more detailed analysis.  This behaviour may be particularly hard to obtain in the PM peak where passengers would be trying to get on outbound GO trains already packed with riders from Union Station.

This report begs a much more complex question — there is a pressing need for a detailed analysis of the transit system, all of it, not just the parts each agency cherry-picks for its own purposes.  GO Transit must stop treating travel inside the 416 as something that is not its job.  Transit studies need to look at surface route growth, not just at subways.  The whole debate about LRT will remain a sideshow while the Ford brothers have influence in Toronto, but the larger problem of surface capacity on bus and streetcar routes will not go away.

Some bus routes logically will transform to LRT, some to BRT, some to simply a mix of local and express buses.  Road space for auto users will come under attack, and subways cannot possibly fulfill the needs in every corridor.  All of our transit lines will need more money for fleet growth, garaging, operations and maintenance.  In the 905, we need to know the implications of greatly improved GO service on local transit networks, especially for off peak and counterpeak travel.

With this report, the TTC has seized the initiative in the political debate about transit expansion and could leave Metrolinx in the dust thanks to that agency’s dependence on Queen’s Park to make policy.  The focus is on a subway line, something that will keep the Fords happy, but the continued importance of the streetcar system for “in town” trips is part of the overall plan.

At long last, we have the TTC openly talking about the need for much-increased rapid transit capacity downtown, a debate that was sidelined decades ago when a proposed DRL was dumped in favour of the Sheppard Subway.

Postscript (October 20, 2012)

A comment on this article sent me looking at the information from the many simulations in the background study to see how the component flows changed, or not, for each configuration of the network.  The information is much more easily digested in this consolidated format rather than flipping back between many pages in the source document.

Charts of Demand at Bloor Yonge

Table of Demand At Bloor Yonge

The “charts” file linked above contains three pages.  The first shows the modelled demand at Bloor-Yonge under various scenarios.  Note that “walk in” traffic at Bloor-Yonge is not included in the published data, and this may be important as we will see later.

• The leftmost column shows the current demand and its source from the three legs of the subway.
• The next two columns show the projected demand in 2031 without and with the Yonge Subway Extension to Richmond Hill.
• The next six columns show the projected demand in 2031 with the YSE and with each of the six alternatives studied as a “relief” line.

The most notable aspect of this chart is that the demand from the north in 2031 with the YSE added in equals all of the existing demand.  Growth and new ridership attracted by the extension completely replace the capacity now used by transfer traffic from the BD subway.

Various relief schemes reduce the BD transfer demand, but the total travel south from Bloor remains well above current levels in all cases.

The second chart shows the various forms of activity on the Bloor southbound platform.  This includes passengers leaving the station from southbound trains, transfers from southbound trains to each direction of the BD line, and transfers from the BD line onto the southbound service.  There is a very substantial rise in total activity for 2031, but this arises from various sources:

• There are many more transfers from the eastern leg of the subway to southbound service, and a smaller increase from the western leg.
• There are many more transfers from the northern leg of the subway to westbound service.
• There are many more passengers arriving from the north and leaving the station.

A “relief” line will only address the first of these three groups.  The effect can be seen in the lower platform activity for various configurations notably those with a full DRL from Dundas West that intercepts traffic from both directions, not just from the east.

The “walk out” traffic is interesting because of the big jump compared to current figures.  This is shown on the third chart where the projected number of passengers for whom Bloor-Yonge is a destination (i.e. where they will leave the system) will more than triple by 2031.  This begs two questions.

First, where are they going?  New construction at Bloor-Yonge is overwhelmingly residential, not commercial, and yet a massive rise in “walk out” traffic implies a growth in destinations like offices or schools.  Is the land-use information in the demand model valid?

Second, if these numbers are accurate, the pedestrian flows through the station exits will also increase substantially.  Can the station actually handle this level of activity?

The published figures do not include “walk ins”, something that should grow substantially given the amount of residential construction planned for the Yonge-Bloor area.

The crucial missing part of the study is the “out of scope” examination of how much demand from the north can be diverted onto GO Transit.  Considering the cost of handling the much higher demand on Yonge south, the absence of this information is a glaring omission, and shows the problem inherent with the TTC studying only the “inside 416” effects.  It would be nice to say this should be Metrolinx’ job, but that agency is notorious for doing very little in public where the work can be examined and debated.

When we are looking at multi-billion dollar options for subway expansion we need to be sure we are not ignoring possible alternatives.  For decades, the TTC has downplayed the importance of the DRL as a potential relief valve or the Yonge-University subway.  The possible contribution of GO should not be ignored as part of an integrated, dare I say, regional solution.

Postscript 2 (October 21, 2012)

Back on December 2, 1982, the Star ran an article by Rick Brennan.  I am not reproducing it in total in respect of copyright, but here are the main points.

TTC Chief General Manager warned that downtown could grind to a halt in as little as seven years without a DRL.  “With the office development that is coming online in seven to ten years, we can’t service it, period,” Savage said.  “We can’t do it with the existing system”

The forecast for new jobs in the core was 40,000 in the the decade to come, and as many as 90,000 by the year 2000.  As we know, the world did not end.  What happened?

First, the lion’s share of the growth was handled not by in-town trips on the subway, but by the expansion of GO Transit and the rise of commuting from what would become the “905”.  Also, actual growth over the two decades did not match predictions due to a severe recession in the early 90s.  Ridership fell by 20%, and the TTC obtained unexpected “relief” for its capacity problems.  Toronto has now passed its old record ridership set before that crash, but the subway’s capacity has not grown to match.

In 1982, the line would have run from Donlands to Union and was predicted to cost $400-million, rather less than the $3.2-billion foreseen for a line from Pape to St. Andrew.

Savage noted that surface transit options could not help the problem because of traffic downtown.  Nothing much has changed on that count, but more to the point, this is not just a question of drawing lines on a map.  If riders are not originating where a surface line can help them, then they won’t divert from their current travel pattern.

Another important difference over the two decades is that “downtown” and the residential community serving it are “fatter” than they used to be.  At one time, a Queen West subway would have primarily served residential areas to the north and feeder routes from the west.  Only a small pocket of high-density development around King and Jameson was well south of the Queen Street corridor.  Similarly, a subway to the east would not even consider development south of Queen to the lake.

Today, development is pushing south into former industrial lands both east and west of downtown, and the dense area which will be too far for convenient subway service is becoming much “wider”, further away for walking access.  At the same time, the job market downtown is spreading away from the core at Bay and King both to the south, and east-west.  No new rapid transit line can serve all of the growth areas, especially those south of the rail corridor.

As they are today, Bloor-Yonge and Wellesley Stations were the hot spots of problems for the subway in 1982.  Savage the “relief” subway as a decade at most away, but as the Star noted:

“Ironically, the biggest opponent to the relief line has been the City of Toronto.”

This was the “old” pre-amalgamation City which was attempting to throttle development downtown by choking off expanded transportation routes.  Politically, this put Toronto in the odd position of supporting new subways that would replace pressure on downtown with suburban job growth.  Little new suburban rapid transit was built, and the sprawl of homes and jobs through the outer 416 and 905 developed around auto travel.

There is an important lesson here for the Official Plan review now in progress.  Whatever form we hope to see Toronto take, we must be prepared to see it through with a transportation system that will support the plan rather than sabotage it.  Decades of inaction leave us with the need to simultaneously catch up with growth we ignored and to build for growth happening now and in the future.  This will not be cheap, and we won’t be able to afford everything we might like to see.  The missing DRL is a reminder of what happens when planning ideology trumps actual experience on the ground.