Toronto’s Transit Capacity Crisis

In recent days, Mayor Tory has announced, twice, a ten point program to address crowding on the TTC. The effectiveness of this program is limited by years of bad political decisions, and the hole Toronto has dug itself into is not one from which it will quickly escape.

This article is a compendium of information about the three major portions of the “conventional” (non-Wheel-Trans) system: subway, bus and streetcar. Some of this material has appeared in other articles, but the intent here is to pull current information for the entire system together.

Amendment February 15, 2018 at 5:30 pm: This article has been modified in respect to SmartTrack costs to reflect the fact that over half of the cost shown as “SmartTrack” in the City Manager’s budget presentation is actually due to the Eglinton West LRT extension which replaced the proposed ST service to the commercial district south of the airport. A report on SmartTrack station costs will come to City Council in April 2018. Eglinton LRT costs will take a bit longer because Council has asked staff to look at other options for this route, notably undergrounding some or all of it.

Subway Capacity

The capacity of a subway is dictated by a few basic factors:

  • The number of trains per hour that can be operated through the peak point on a line.
  • The capacity of a train.
  • The ability of stations to accommodate passengers to and from trains, and to provide buffer space for extra passengers when service is delayed.
  • Geometric and other constraints at pinch points, notably terminals, that could limit the trains/hour throughput.

The peak point trains/hour is constrained both by the signal system (which sets a lower bound on train spacing) and by passenger handling at stations (dwell time for passengers to leave and board the trains). Dwell times can rise substantially under crowded conditions when there are many passengers to handle and little space for them to move around. A subway could have a theoretical capacity, including “crush” rather than “service design” loading, but this works against dwell time by adding to congestion on board trains.

It is possible for a line’s capacity to go down when crowding goes up because the reduced throughput of trains at stations more than offsets the added number of riders per train. This affects not just a peak point station like Bloor-Yonge, but many adjacent stations where crowding on trains slows the exchange of passengers between trains and platforms.

This problem also applies to surface routes. Those who argue that there’s “room for a few more” can do more harm than good in their penny-pinching pursuits. “Efficiency” is a combination of several factors, and jammed full vehicles are counterproductive.

The Yonge-University-Spadina subway, Line 1, is scheduled to have a train every 2’21”, or 25.5 trains/hour. At a service design capacity of 1,100/train, this translates to about 28,000/hour of line capacity in theory. However, the actual service operated hovers in the 90% range of the schedule, and can fall lower when delays cause gaps in service. This shaves off about 2,800/hour reducing the line’s effective capacity to about 25k/hour. It actually carries more because more riders are packed into trains than the service design level. There is a similar problem on the Bloor-Danforth subway, Line 2, but its overcrowding is less severe looked at over the entire peak period.

The most important change in line capacity will be the introduction of Automatic Train Control (ATC) and an increase in the scheduled service to at least 30 trains/hour. This would take the design capacity up to 33k/hour. A further increase to 33 trains/hour would get the line up to about 36k. This is the TTC’s current target. In the past, claims were made for a possible 40 trains/hour service with a capacity in the 45k/hour range, but this is not physically possible on the Toronto system.

Fleet and Carhouse Capacity

The TTC now operates 63 trains in the AM peak period on Line 1 YUS, of which 2 are the recently restored “gap” trains. Half of the service runs from Finch to Glencairn and half runs through from Finch to Vaughan. The scheduled headway on each branch is 4’42” combining to provide a 2’21” service south from Glencairn (25.6 trains/hour). The PM peak uses 63 trains with all service running from Finch to Vaughan on a 2’36” headway (23 trains/hour).

Moving to a schedule of 30 trains/hour would require about 71 trains based simply on scaling existing service up proportionate to the higher frequency of service. With the benefit of ATC and higher possible speed on Line 1, this can be reduced to 68 according to the TTC [email from Brad Ross, February 14, 2018]. It is unclear whether this would include gap trains. The Toronto Rocket (TR) fleet serving Line 1 includes only 76 trains, and so a 68-train peak service would require a low spare ratio of only 11.7%, well below the TTC’s accustomed level. (Originally there were to be 80 TR trains, but 4 trains’ worth of cars were repurposed to become six 4-car sets on Line 4 Sheppard.)

Beyond the 76-train fleet, an additional 10 are required to get to 33 trains/hour, and there is no provision for these in the capital plans. Management will bring a proposal to the Board for procurement of these trains once they are sure how many are needed and when. [email from Brad Ross]

A related issue will be the storage needed for even more trains.

Line and Station Capacity

Various schemes have been proposed to deal with Line 1 crowding. In addition to ATC and more trains/hour, these include diversion of traffic to the TYSSE (Vaughan extension) and to GO/RER/SmartTrack. Demand will rise from population and employment growth, from new LRT service feeding in on Eglinton, and further if the subway is extended north to Richmond Hill. The net effect of all this is to take the projected demand to only slightly below the design capacity with ATC and 33 trains/hour. However, we know that the TTC does not actually operate all of its scheduled service and that missing ten percent is equivalent to 3 trains/hour.

A further problem with the projections (contained in a 2015 report on the effect of a Richmond Hill extension) is that any additional capacity provided on the subway will immediately be swamped by latent demand that is constrained only by the existing level of service. It is far from clear that the anticipated relief (more YUS service, demand diversion to the Vaughan extension, GO or SmartTrack) pending the construction of a parallel Relief Line will be adequate to reduce crowding below current levels especially when the growing population is factored in.

In order for dwell times at Bloor-Yonge to be reduced to a level that would permit much more frequent service, passengers must be able to move freely both between trains and platform, and within the station itself. A full train meeting a full platform leads to gridlock. To avoid this there must be both room on the trains and in the station to avoid dwell times that are so long that the planned train frequency cannot be operated. ATC will help somewhat by allowing trains to bunch closer together, and less time will be wasted between trains “A” departing from the platform and train “B” arriving, but the problem of station capacity remains especially when there is a delay.

In a report on the TTC’s supplementary board agenda for February 15, 2018, management examines the many issues raised by the overcrowding at St. George and Bloor-Yonge on January 30, 2018. They note that a gap of seven minutes in service at Bloor-Yonge “creates crowding and difficult passenger circulation conditions”. The situation at St. George is only slightly better with a nine minute gap having the same result. To address this, at least for Bloor-Yonge, the TTC has reintroduced “gap trains” which are dispatched from Davisville southbound to fill wide gaps from Finch and clear the platform at Bloor before the next packed train from Finch arrives. This applies only to the AM peak southbound on Yonge. There is no provision for gap trains elsewhere nor anywhere in the PM peak.

As service builds up in the future with more frequent Line 1 YUS trains, this will also affect crowding on the Line 2 BD platforms where the same service operated today will be fed by more frequent connecting demand from Line 1. This is of particular concern during the PM peak when the primary flow is outbound from Yonge-University to Bloor-Danforth, not the other way around. Eventually, in the mid-2020s, the BD line itself will have more frequent service and this will in turn affect the rate of transfer traffic onto the YUS in the AM peak.

These concerns have spawned plans to expand platform capacity at Bloor-Yonge, but this is extremely difficult with much of the station structure hemmed in by (and occasionally inside of) buildings. A further challenge is that any construction could reduce capacity while it is underway.

The ability to construct any modifications is severely constrained as the station lies largely under buildings at Yonge and Bloor Streets. Specifically, these building foundations limit the space available to enlarge the station. Inside the station, current customer congestion also constrains construction. Careful staging of any improvement would be required to ensure that impact to station performance is minimized, but this would increase the duration of the construction activity.

The preferred option will be developed to 10% design to address the requirements to the year 2041. Included in this analysis is the impact of all the expansion projects, sequencing and other endeavours that effect Line 1 and Line 2 performance (ATC signalling, new subway trains, etc.). Fire ventilation, emergency egress and other impacted infrastructure associated with Line 1 and Line 2 capacity improvements will be considered in the design development for the station.

The early identification of potential concepts is expected to be available for TTC staff review by mid-year 2018 with final feasibility reporting in early 2019. [p 8]

Station capacity is not only an issue for Bloor-Yonge, but it also affects other major stations downtown. If trains are going to arrive more frequently, they will deliver passengers to platforms at a higher rate and the crowd from one train must be able to clear before the crowd from the next train arrives. Stations from College to King pose problems, especially if some of the exit capacity is blocked for maintenance (e.g. an out of service escalator).

Platform Edge Doors (PEDs)

Platform doors have been proposed as a way to improve safety at crowded stations, and this became a talking point after the crush on January 30 at Bloor-Yonge. The current cost estimate for this type of installation is about $24 million/station. [Globe & Mail, February 12, 2018]

Automatic Train Control is considered to be essential for the operation of PEDs through precise stopping of trains. This limits the applicability to Line 1 YUS and Line 4 Sheppard in the immediate future.

An important question is the actual purpose of the project. If it is intended as a suicide barrier or to limit garbage blowing onto the tracks, then the scope reaches across the system even to lightly used stations. If the goal is crowd control and safety, then only the busiest stations need to have doors fitted.

Platform doors will not add to station capacity, but will only provide a barrier and safety measure between riders on the platform and the trains. The goal for such a project must be clearly understood up front so that costs and expectations are scaled accordingly.

Terminal Capacity

Terminals present a special challenge for operating short headways.

  • The physical length of a crossover combined with the radius of the turnouts at switches sets a lower bound on the time a train requires to move from being stationary (either departing or on approach) to being clear of the shared trackage.
  • There are constraints on the entry speed to stations when there is a train stored on the tail track beyond the platform.
  • The nature of break times for train crews, the need for operators to walk the train length (an issue for one person operation), and the punctuality of crews being ready to depart as soon as their route is cleared all affect the ability operate a service at a terminal’s maximum capacity.

The TTC has yet to attempt a trial to determine whether they can actually operate at a rate of 30 or 33 trains/hour (corresponding to headways of 2’00” and 1’49” respectively) on a sustained basis and under the most restrictive (tail tracks occupied) conditions.

Bloor-Danforth Line 2

There is no plan to add capacity to the BD line until 2026 when the Scarborough Subway Extension (SSE) opens and the existing BD line converts to ATC operation. The fleet plans provide for extra trains for the extension and for more frequent service, but only in 2026.

There are conflicts between the schedules for related projects:

  • SSE opens 2026 with ATC-only operation.
  • BD Line 2 begins cutover from conventional to ATC signals in 2026 completing in 2029.
  • Additional service, which requires ATC, is planned for 2026.
  • T1 trains, which are not ATC capable, are replaced from 2026 to 2030.
  • A new carhouse and yard at Kipling must be available to receive the T1 replacement trains, but there is no construction budget in the 10 year capital plan.

These timings simply do not work, and resolving the conflicts will require both the new fleet procurement and construction of the new carhouse at Kipling to be brought forward.

The T1 Subway Fleet

The TTC has two fleets of subway cars.

  • The “T1” trains are used on Line 2 BD, and they will reach their end-of-life in the late 2020s. Current plans call for their replacement beginning in 2026. They are not capable of conversion to ATC at a reasonable cost considering their remaining lifespan.
  • The “TR” trains are used on Line 1 YUS and Line 4 Sheppard and they all have ATC controls. Most trains are configured in six-car sets for YUS, but those for Sheppard are four-car sets. The original 80 six-car trainset TR order was changed to 76 six-car sets and 6 four-car sets to provide equipment for Sheppard.

Problems in the fleet plans arose from decisions that changed how the T1 fleet could be used.

  • The TYSSE to Vaughan was originally designed to have conventional signals, not ATC. The T1 fleet was sized to provide trains for this extension.
  • The move to ATC over the entire Line 1 YUS, with the TYSSE built as ATC-only, meant that additional TRs were required for Line 1 and the surplus T1s did not have a home.
  • The move to ATC for Line 4 Sheppard displaced T1s from that line increasing the surplus of T1 trains.
  • The need to reactivate Keele Yard was a direct result of shifting all of the T1 fleet to Line 2 BD.
  • Planning for the SSE assumed that surplus trains in the existing fleet would be available for the extension to STC, but that extension will require ATC-equipped trains.

Line 3 SRT

When the rebuilding of SRT trains for an extended lifespan completes in 2019, the peak scheduled service will be increased from 5 to 6 trains.

Line 4 Sheppard

There is no plan for additional service on Sheppard in the fleet plan.

Parallel Routes for Subway Relief

Express Buses and Streetcar Routes

Operation of parallel services to the subway offers limited relief given that surface vehicles are much smaller than subway trains, they must operate over city streets already plugged with traffic, and the existing stations are not set up to support large volumes of passengers transferring to a parallel service.

Moreover, the TTC’s standard response to any calls for service improvements has been “we have no buses and no garage capacity”. Their ability to create a major new “relief” service is constrained by the belt-tightening that has limited growth of bus services generally over the past two terms of Council.

TTC staff will study the various options.

The measures to be investigated will include, but will not be limited to:

• Express buses on Yonge Street, together with transit priority measures
• Express buses from other stations on Line 2 Bloor-Danforth direct to downtown
• Improvements to streetcar service and streetcar terminal capacity improvements at Broadview, Spadina, and Dundas West stations, which are served by streetcar routes to downtown

The work will be done with input from TTC planning staff, TTC Operations, TTC Engineering, City of Toronto Transportation, and City of Toronto planning staff. The work will be expedited, with an initial report to the TTC Board in May 2018.

The work will include:

• Review of past work on express bus relief services
• Identification of service options, including routings, corridors, and express stops
• Consultation with key stakeholders
• Projection of ridership and of the degree of subway relief to be provided
• Resource requirements for buses and Operators
• Required transit priority measures to deliver the necessary services
• Evaluation of value for money of the relief services, including comparison of the relief services to other services that could be operated with the same resources [p 7]

This proposal is grasping at straws, hoping that there is some way to bleed off demand from the subway. If this were attempted, a small number of frequent point-to-point services with strong transit priority measures have a better chance of success than many less frequent operations that could raise expectations they cannot meet. The idea should not be to give riders over a wide part of the subway an alternate route to downtown, but to locate a major demand (for example at Finch terminal), riders who all want to go to the core area. Any attempt to provide service that stops “like the subway” at many locations enroute would be limited by the need to follow the subway route and by the overhead of serving those stops. Riders hoping to board enroute would have to decide whether to stay inside the subway or take their chances with the surface bus.

The level of transit priority to make services like this work is considerable, and experience on King Street shows that taking away road capacity from general use is an uphill battle. This hill grows steeper the further one moves from downtown where transit dominates travel on the streets.

Another important consideration will be the operation of a downtown pickup point for outbound service in the PM peak. The volume of would-be riders will be quite substantial and they will require considerable space wherever they are going to board. This terminal area will be more complicated with many routes heading to a variety of destinations than with a few trunk services.

To put a bus service in context, the service design capacity of an articulated bus is 77 of whom 31 would be standees. To divert 10% of the subway’s capacity, say 2,500 riders/hour, would require about 32.5 buses/hour or a headway of just under two minutes. That is a very frequent service, and it would take over a good deal of the road capacity wherever it operates especially at stops.

As for the streetcar routes, let us pretend for the moment we actually have vehicles that could be used for a parallel service. The main route into downtown is 504 King, and it already suffers from overcrowding. If anything it requires more service for riders who live and work along the route today, not as a subway relief valve. There is also a question of platform capacity at Broadview and Dundas West Stations which have chronic problems of streetcars backing up onto the street.

Spadina Station suffers from the way the streetcar platform is operated with, usually, only one Flexity on a platform that is almost long enough to hold two. Streetcars regularly back up into the approach tunnel, and the longest part of a rider’s journey on the 510 car can be the glacial approach to the platform. Two changes at this location are required:

  • A way must be found operationally for streetcars to have dedicated unloading and loading zones that can be occupied at the same time.
  • Crew management must, like a busy subway terminal, operate with much less chin-wagging between staff and leisurely departures blocking service. Terminal layovers are simply not possible here and streetcars should move through the station as quickly as possible.

Both of these would address problems for the 510 Spadina car, but there is no capacity for an additional “downtown” service overlaid on the existing route.

Streetcar routes are quite obviously local services, and riders who might be diverted from the subway to them must be prepared to trade off a faster, if much more crowded, trip via the subway which they are already riding for a local service that, will take considerably longer to reach downtown, especially from the west.

GO Transit and SmartTrack

GO Transit does not plan significant improvement in train frequencies in the near future pending completion of infrastructure for its “RER” (Regional Express Rail) services in the mid 2020s. Current service from Richmond Hill to Union operate half-hourly inbound from 6:25 to 8:25 am, and outbound from 4:30 to 6:00 pm with additional trains on longer headways on both shoulders of the peak. This service will improve to every 15 minutes in the peak direction as part of RER.

A doubling of GO service adds 2,000 passengers per train, or 4,000 per hour, assuming ten car trains and full loads. Achieving this depends on local bus feeder services and parking to bring riders to the trains.

Politically, one incentive for subway service to Richmond Hill is the idea that riders can come into Toronto for one fare as they now do at Vaughan and Highway 407 stations. However the problem of the separate local fare remains, and there is a limit on how many commuters can be accommodated in ever-larger parking structures.

Many riders originating in York Region are not going to Union Station, and they will remain customers for the subway which serves many other nodes enroute to downtown. That said, the diversion of traffic onto the GO corridor should occur sooner rather than later to address existing congestion.

Correction: The following paragraph has been corrected to reflect the split between the SmartTrack costs related to GO stations, and the Eglinton West LRT extension which replaces the originally proposed branch of SmartTrack to the airport. The modified text is in italics below.

SmartTrack, for all its branding, is simply an increase in service frequency and new stations within Toronto. What is actually planned is minuscule compared to the claims made for this service in the 2014 election, but the service has a huge pricetag, about $1.5 billion. (There is $3.7 billion in the City of Toronto capital budget for 2018-27, but this includes about $2b for the Eglinton West LRT which is the replacement for the western leg of “SmartTrack”.) Operating costs including the subsidy of free transfers from TTC services to GO/ST trains are extra.

Both SmartTrack and the SSE address demand coming in from Scarborough, and their competition for the same market was (and remains) an issue within debates over the appropriate transit solution for Scarborough’s travel. To the west, SmartTrack will add stations at Mount Dennis, St. Clair and Liberty Village, as well as serving an improved subway connection at Dundas West/Bloor. Demand projections show little effect at Bloor-Yonge from SmartTrack. It will skim off some growth and latent demand, but will not actually reduce traffic at Bloor-Yonge from existing levels.

With all the focus on subway relief, it is important to remember that GO transit has other routes that are not part of SmartTrack. Any service and fare offerings should be available on all of these routes to divert travel that might otherwise use the subway.

The Relief Line (aka the Don Mills Subway)

The Relief Line has been written about at great length elsewhere and I am not going to duplicate that work here. In any event, any “relief” is a decade or more away, and the RL still competes in the minds of many who insist on viewing this as a “downtown” transit project even though the principle benefits will flow to suburban riders who need more capacity to reach their destination.

The situation has not been helped by the TTC who downplayed the need for this route for decades, and shortened the proposal limiting its reach as an alternative travel corridor. Back in the 1960s, the “Queen Subway” would have run east from downtown then north to Eglinton and Don Mills. When the proposal was scaled back to end at Danforth, its political attractiveness waned along with the amount of actual relief the route could provide. The high cost was also seen as diverting funding from suburban expansion including, now, the Richmond Hill and Scarborough subways.

Demand projections show that the major relief comes from taking the Relief Line well north of Danforth all the way to Sheppard and Don Mills. This would provide a truly parallel route into downtown as well as new rapid transit service to the entire corridor. Development sites, current and proposed, along the RL make it much more than a sleepy route that would only see demand during the height of the peak with little reason to ride other than as a bypass around Bloor-Yonge.

The political challenge is to move beyond detailed planning now underway to actual construction and service. Although John Tory argued four years ago that SmartTrack would solve every problem one might name, he is now forced to acknowledge the importance of the Relief Line. Neither of these will open while he is still in office, but the decisions and delays will haunt Toronto long after he has retired. The situation is even more complex at the provincial level where the commitment to build an expensive new line into downtown Toronto is not met with unbridled enthusiasm by any party.

Off Peak Fares

The concept of using lower off-peak fares to attract ridership away from the peak period will be part of an update on the Ridership Growth Strategy in July 2018.

Many questions surround such as scheme:

  • Would off-peak fares be subsidized while leaving peak fares as they are, or would peak fares go up in an attempt to recover lost revenue?
  • Crowding is an issue only for parts of the network at specific times and locations. Should a major fare change be implemented across the board affecting all peak riders, and benefiting all off-peak riders?
  • What are the equity effects for riders who have long trips and for those who cannot shift their travel times?
  • Would the fare differential be available to all riders, or only to those paying “full” adult fares? How would this plan interact with various fare discounts and passes?
  • Is an investment in a new fare structure a productive way to “buy” peak capacity compared to other options including better service on surface routes?

Off Peak Service

Another effect of lower off peak fares will be to encourage demand that could overload existing off peak services both on the shoulders of the peak and more generally through the midday, evening and weekend. Better service at these times does not require more vehicles, but it does require more operators to drive the buses, streetcars and subway trains.

A proposal to improve off-peak service on Line 1 YUS failed thanks to budget limits a few years ago. It was part of a package that included improved early Sunday service, but only part of the scheme survived the budget knives.

Part of the $2 million in added transit funding approved by Council will go to adding four trains to the early evening service on Line 1.

On the surface network, the quality of off peak service is under attack by TTC Board members who feel that the Service Standards are too generous. They argue that there is nothing wrong with having a few standees, but this diverts attention from the routes where off peak crowding is more than a few riders for short periods.

Squeezing the more lightly loaded routes offers limited savings as we found during the Ford era, and it reduces or eliminates capacity and attractiveness for more riders. The TTC has been preoccupied with static “ridership” numbers for so long that it ignores the very real fact that crowding exists beyond its own approved standards. The focus for budgets is not on how to run more attractive service, but on finding ways to make do with existing vehicles and staffing levels.


The bus fleet suffers from a shortage of both vehicles and garage space. These are related in that one cannot have buses without somewhere to store and maintain them. The problem goes back to the transition between Mayors Miller and Ford when the Transit City Plan was killed. At that time, bus fleet and garage planning assumed two things:

  • More buses and garage space would be needed for riding growth and improved crowding standards.
  • Transit City LRT lines would, when they opened, replace major bus routes that would cap the need for more garages.

What actually happened was that Transit City was killed, and the crowding standards were rolled back to pre-Ford levels. This, in turn, allowed provision for vehicles and a new garage to be deferred in the capital budget. This also benefitted the operating subsidy pressures by reducing the amount of service that could be provided.

Construction of McNicoll Garage eventually started, but it will not open until 2020. A small additional space has been purchased adjacent to Malvern Garage that will house 50 more buses starting later in 2018. No additional garages have reached the planning and budget stages, although the TTC is at least looking for potential sites.

A few dozen buses will come free in fall 2018 as streetcar routes convert back to rail operation, but that is the extent of spare resources available for better peak service on the bus network overall. The next big jump in bus availability will come with the opening of the Eglinton and Finch LRT lines in 2020-21.

Plans for express buses have been half off until 2019 pending both fleet growth and budgetary approval.

There is no service plan that takes into account possible growth in ridership across the bus system, and certainly not for any parallel services to supplement the subway.


Streetcar delivery issues are well-known with the new Bombardier fleet, and crystal ball gazing to produce an implementation and service plan is a cottage industry not unlike the creation of fantasy transit maps. Many configurations of streetcar service can be sketched, but until cars are actually arrive on a dependable basis, any fleet and service planning is hamstrung by uncertainty of the actual fleet that will be available. This is compounded by reliability issues with the old streetcars which, despite the investment of millions in reconstruction, are still vulnerable to very cold weather.

As of February 18, 2018, there will be 91 buses operating on streetcar routes in the AM peak period. Allowing for spares, there are over 100 buses that could be serving riders on bus routes, but instead are operating on streetcar lines to make up for Bombardier’s non-performance.

This raises a budgetary issue because as and when these vehicles are available, they would represent a 6% increase in peak service on the bus network for which there is no available operating funding.

Another problem for the streetcar fleet is that when it is underutilized because buses run on streetcar routes, there is a temptation to retire cars and keep only the most reliable cars on the street. This process can get ahead of new car deliveries throttling the ability to operate more streetcar service.

Very roughly speaking, if all of the Flexitys were in Toronto by the end of 2019, it would be possible to operate the entire streetcar system with them, including an increase in capacity of about 25% on all routes that do not already have them, and a more substantial increase on King with a 1:1 replacement of old cars by new. This would bring a long, long overdue improvement on the streetcar routes which have been starved for service for over two decades.

The actual timing of these changes, and their degree, will depend both on Bombardier’s deliveries, and on City Council’s willingness to pay for improved service.

Further increases in the streetcar fleet to address future growth lie in two potential orders:

  • A 60-car add-on order to the current Flexity contract with Bombardier
  • A further order for new cars for potential Waterfront lines west to the Humber and east to the Don.

If all of these were actually purchased, the TTC will be tight for streetcar storage, especially considering that Leslie Barns will do double-duty as both a carhouse and as an overhaul facility when the new fleet is old enough to reach its first major overhaul cycle.


Where Does the Money Come From?

There are two problems with paying for all of this and they are at the heart of all debates on transit growth.

  • Funding: This term refers to the actual source of money that will pay the bills. Will it be the farebox, the municipal tax base, provincial or federal programs?
  • Financing: This refers to how the money will be obtained. Will it be on a pay-as-you-play basis with current revenues paying the bills (typical for the operating budget), or will it be some combination of current revenues and borrowing, including P3 schemes where borrowing is done by a private company which passes on the cost through an operating lease?

The TTC’s capital program has been artificially limited in scope by the City’s self-imposed cap on the ratio of borrowing costs to property tax revenue. There is a provincial mandate that this stay below 25%, but Toronto has its own cap of 15% as a management strategy to (a) leave headroom and (b) provide an attractive, well-managed city to attract good debt ratings and lower borrowing costs. Recently, this cap has proved impossible to meet in an absolute sense, and it is now averaged over a ten-year period to allow for a coming crunch in borrowing requirements (mainly due to the SSE, the Gardiner and SmartTrack).

The TTC is forced to set its capital budget based on a debt target from the City, and this results in about 1/3 of the approved capital needs having no funding. A further (and growing) list of projects is not even on the “approved” list, and these can only be added with Council’s approval. There is much that the transit system needs and wants, but paying for it is quite another matter.

Among the factors now affecting capital planning are:

  • 10 additional trains to allow Line 1 YUS to operate at 33 trains/hour
  • Timing of the Line 2 BD fleet replacement & carhouse
  • Subway station expansion
  • Platform doors
  • One or two additional bus garages beyond McNicoll including a rolling facilities migration to electric vehicles
  • Bus fleet expansion
  • 60 or more new streetcars for ridership growth and new Waterfront services

Underlying all of this is the problem between spending that is in the operating budget (more service, fare subsidies) versus the capital budget (more trains and buses, new lines). The nature of funding schemes from other governments pushes Toronto away from any new operating costs toward spending on buying and building things, even though they must then find the funds with which to operate the new assets. Capital spending also has the political advantage that it is “other people’s money” and the hit on operating costs lies years on the future.

Fleet Plans

In a future update, I will add here the detailed fleet plans from the TTC’s 2018-27 capital budget.

37 thoughts on “Toronto’s Transit Capacity Crisis

  1. How many riders could the TRs handle if half the seats were removed? If all but the blue seats were removed? (Say, on those gap trains, not every train set.) Drastic, nasty, but short-term relief?

    Steve: Yes, drastic and nasty, and utterly unsympathetic to the many riders who would have no place to sit down especially during much of the day when overcrowding is not an issue.

    Liked by 1 person

  2. Little mention about using the 6 BAY bus as an alternative for the Line 1 from south of Bloor Street West. Before the Yonge Subway opened, the BAY streetcar was an alternative (from the Earlscourt loop at St. Clair West and Caledonia, along St. Clair, south on Avenue Road to Davenport, then south on Bay from Davenport). With the opening of the Yonge, the route was replaced, over time, with trolley buses and now buses.

    With the streetcars and trolley buses, the BAY route was more heavily used than with diesel buses. With rush-hour headways of 5 minutes, there is little draw for users to use the BAY bus instead of the subway. I’m sure that includes Mayor Tory.

    Steve: Mayor Tory lives close to an entrance of St. George Station, and I doubt he would schlepp over to Bay Street. In any event, he goes in to work early and subway crowding would not be a problem for him.

    If the Bay bus ran every 2 minutes, that would add 18 buses per hour or roughly 900 passengers’ capacity to the route, less than one subway train. It would help, but it’s no panacea. FYI I use the Bay bus sometimes enroute to or from City Hall, and am familiar with the vagaries of the “service” it provides. During the off peak, I’m sure to check how it is running when I am a few stations further east before making the choice of getting off at Yonge or Bay.

    Liked by 1 person

  3. Steve would this not require?

    1) acknowledging and getting temporary storage and additional buses.

    2) looking seriously at a spot where you could actually divert serious ridership to core

    3) being willing to actually spend some serious money

    So even if magically found a route, to give meaningful relief would mean moving thousands of riders per hour, like 3 or 4 thousand or something like 60-100 buses per hour. Today that would mean creating a route (ie one that was able to be dedicated to bus) – serious work, and well, having the buses, serious money – and drivers – also serious money. Has council shown any willingness to spend money, or grab space from roads or parks to make this work?

    Steve: Are you living in some sort of fantasy world?


  4. How much money was spent to cancel the Scarborough LRT. I am asking about sunk costs only.

    How much money will have been spent on the SSE by the end of this year (all previous years included)?

    Please state in fixed year dollars (say 2017). The more the answer to those two questions, the better as that would make the much needed Scarborough subway more likely.

    Thank you!

    Steve: At year end 2017, spending on these projects is:

    • SLRT design (sunk costs): About $75 million
    • SSE design: $90 million
    • SRT life extension: $31 million

    This is approximately 6% of the current estimate of the subway’s cost. We are nowhere near the point of no return in a subway/LRT decision from a financial point of view.

    Life extension of the SRT is ongoing and includes not just vehicle overhauls, but also infrastructure work (track, power, structural repairs, elevators, escalators, signals).

    Figures are in cumulative as spent dollars. I am not going to delve into past budgets to pick out each year and inflate it to 2017$. See the detail pages for the 2018 Capital Budget at page 23.


  5. TTC fares for GO and UPX will solve the problem.

    Steve: If only GO/UPX had sufficient service to both to attract riders (long headways are a real pain in the ass if you miss your connection), and capacity to carry the resulting demand. Also, at locations where a TTC-to-GO transfer might occur, the walk is considerably longer than the off-complained of transfer between the RT and subway at Kennedy.


  6. When ATC goes live, and if the TTC goes with driverless trains, would it be possible to run more trains (especially off-peak) without too much extra cost? Presumably there’d be cost in electricity, and depreciation / wear-and-tear, but would there be any disadvantage otherwise?

    Steve: The TTC does not plan to run driverless trains, although this would technically be possible. A common misconception about the subway is the portion of costs attributable to on-train staff. There are far more people looking after stations and on standby as maintenance crews to deal with problems than there are on the trains. If we went to driverless trains, there would be some offset for an increase in roaming security staff. We have already seen a similar effect where Presto was supposed to “eliminate” collectors. It will, but they will be transformed into Customer Service Agents, and my suspicion is that to perform this job effectively (i.e. to have a visible presence in the larger stations) the TTC will need more CSAs than the existing complement of collectors.

    For un/loading trains and dwell times, how much does the door size and quantity affect things? Would having wider doors and/or more of them be helpful? This would reduce seating, but might be worth it. We probably can’t retrofit larger doors on the TRs, but since the T1s are going to be retired “soon-ish”, perhaps (a) move the TRs to Line 2, and then (b) buy new wider-doored trains for Line 1? (This would have to be considered in an PED decisions.)

    Steve: The doors on the T1s and TRs are already wider than on older trains. There is a point where making doors wider is counterproductive in terms of lost seating and the size of the door machinery, not to mention the structural issues of having an even wider part of the car without a sidewall structure. This was an issue in the T1 design. The same argument applies to having more doors.


  7. Another thought came to mind: driverless trains have been discussed (as I mention above), but have you heard any research into driverless trams? Autonomous cars are in the news a lot, and I’m curious to know if anything has been done with streetcars. Doing a quick search, there are some news articles on the topic, but how pie-in-the-sky is the idea from your knowledge of the industry?

    Steve: A driverless tram on a private right of way would be effectively the same as a subway train on the RT technology (which is driverless in Vancouver). On street operation is quite another matter. Pie in the sky does not begin to describe the issues here. Again, in an attempt to eliminate the driver, one creates much greater complexity in the vehicle including higher maintenance staffing requirements (and that’s assuming they don’t hit anything).


  8. It amazes me in the past the TTC was once able to operate the King car and the Dundas car in two directions from a single track at Dundas West, but today cannot manage to avoid backups at Spadina.

    Steve: The old schedules did not include provision for recovery time at the terminals, and padding to reduce short turns enroute.


  9. I think what you are saying is that we are in a bit of a pickle. The neglect of (what I maintain is) the last 40 years has caught up and there are no simple solutions.

    However, I am very bothered by the suggestion (not by you) that ridership is not growing and therefore service increases are no longer necessary. This is not true. Overcapacity has been the rule for decades and a period of no or slow growth is an opportunity to match the needed service to the demand. Increases in all service at the “shoulders” of rush hour is an immediate need and while that cannot be “instant” the only delay is “months” while the appropriate operators get trained. (Or overtime.)

    After that, the commission of new bus garages and bus deliveries is only 3 or 4 years away – not the same as building a subway. As you have noted – transit riders are going to lots of nodes in Toronto. Let’s get going on serving all of those nodes asap.

    The same for streetcars. Bombardier’s delivery schedule has been abysmal, but let’s get going on the additional 60 (or 100) streetcars in addition to the current contract. If that needs to be another supplier, let’s get that specification settled now. If not, hold Bombardier to a delivery schedule with penalties.

    These are all things that can be done relatively cheaply and relatively quickly. Why aren’t our politicians acting.

    Steve: Because they only care about subways, subways, subways, and don’t want to spend another penny actually paying for improved service.


  10. The question we all should be asking is who are the people causing the crowd and where they come from. Here is an example to illustrate the problem. John Doe lives at Neilson and Finch. He has two choice in how to get to work at Queen Station using the TTC. He can take the 199C to Finch and use Line 1. This person will cause the capacity issue by crowding out passengers from boarding say at Summerhill. The other option would be to take the 133 bus and transfer using Line 3, 2 and 1. He now causes the issue of overcrowding at Bloor Yonge Station.

    John Doe pays only about $150 per month for the Metropass to go and return from work. He also gets the benefit of unlimited travel. If this person is to take the 199 to Smart Track Finch East, he would be diverted from Bloor Yonge. He will also have a faster trip even though he would backtrack from Union. It will also cost him at least $200 more per month. Diverting this person is easy. All the city has to do is subsidize his GO/Smart Track fare and since he lives in Toronto, it would be easy to make the case as John Doe is a taxpayer too.

    Someone coming in from Richmond Hill to use Line 1 is a completely different scenario. How do you convince someone to take the Richmond Hill Line when it is not frequent, not available during midday and weekends, costs more and local transit do not work well connecting to it. Taking out the regional demand from Line 1 will go a long way in easing the congestion at Bloor Yonge.

    The Richmond Hill Line is really difficult to expand with the flooding issue at the Don River. Even if a billion dollars is found to fix the issue, the curvy nature of this route is very problematic. Tilting trains also change the angle in which the pantograph contacts the overhead wire. Electric tilting trains are more complicated than tilting DMUs. If the Richmond Line ever goes to 2 way all day frequent service, it will require its own trainset and an expensive upgrade on the line. Via Rail do not run on that line often enough to contribute to line upgrades.

    Steve: I think your analysis shows a big problem with the perception of subway vs GO in the Richmond Hill context, not to mention other corridors. The question of comparative fares is one of public policy. If we want to encourage a shift from one mode/route to another, this can be done through pricing (a comment you have often made in other contexts). The current service on the RH GO line is not my point of reference, and as I noted in the article, there are plans to move to a 15 minute peak direction service.

    GO’s biggest problem with this line is not flooding (which is comparatively rare), but the fact they don’t own all of it, and prefer to invest in corridors they control. The section that does have water problems is from the bottom end of the line at the Union Station Rail Corridor to, roughly, Leaside. This is the section that must be raised above the typical flood levels. I’m quite familiar with what that looks like as I can see the line at Bloor Street simply by looking out my window. Before rejecting this idea out of hand, I would prefer to see an engineering study from GO/Metrolinx, not vague claims that it’s too expensive. “Expense” is relative to benefits, and as we well know, Metrolinx is all about Business Case Analysis to show the tradeoffs.

    The relief line is nice, but it takes a long time to build. It still does not divert regional travel from Line 1. The fastest way to build would be use monorail technlogy. Prefabricated guideways are much quicker to build than make a tunnel. The crisis is here now, planning something that will take 10 years to complete is not prudent.

    Steve: A monorail is unlikely given the area the Relief Line would traverse. Do you really expect to put a monorail down the middle of Queen Street? This reminds me of the UTDC’s original plan for an ICTS Relief line that would have run along Front Street, including an elevated station, right in front of Union Station.

    The RL is not intended to relieve regional travel, only to intercept traffic that is within Toronto and now flows west to Line 1. Please remember that there are riders within the 416 too, and “regional” includes a lot of suburban Toronto, not just the 905. Each component (GO, TTC) has a role to play. You want quick relief with a monorail, but downplay what could be done with GO Richmond Hill. No single project will solve the entire problem.

    Mr. Tory needs to do something quick before the elections. Imagine what someone like Doug Ford can do. Can Mr. Tory and Mr. Verster do a pilot program? How about a sticker where TTC pass holders can travel unlimited on the Kitchener and Stouffville for an added fee? This can work today and without the wait from construction. If it is successful, the TTC can start to route buses to those GO/Smart Track stations. Yes, the relief line needs to be built and the Richmond Hill Line needs to be upgraded. However, placing hopes on a future solution when people might get injured from overcrowding is not right.

    Steve: A sticker for free transfers to the Kitchener or Stouffville line is not going to solve Bloor-Yonge problems. Remember also that there are other GO lines to which inside-416 passengers could transfer, not just the two “SmartTrack” corridors.


  11. In a previous post, you discussed the garage crowding issue in more detail. You reproduced a chart showing that the garage capacity now is 1,630 buses, but that the actual current fleet is 1,881 buses, which is 251 buses more than the garages will hold — or, put another way, the garages are at 115% of capacity. Further, while 2017 was the worst year for crowding, the fleet has been over capacity since 2012, and 2015-2016 were almost as crowded as 2017.

    A logical question might be what they are doing with those extra 251 buses. The public understands what crowding looks like on the subway or on a bus or streetcar, and you can understand how you can cram in more than the stated capacity if you accept that people will be closer together. It’s harder for a lay person to understand how that would work with buses. Obviously they’re not parking them out on the street, or running 251 extra buses on Blue Night routes [note: potential service enhancement and garage space mitigation technique]. If you have 1,881 cars trying to squeeze into an 1,630-space parking lot, the only way to do that would be to park 251 cars in the aisles. The follow-up question would be, if they have a way of squeezing in 15% more buses now, why is 4.5% over capacity a problem in 2027?

    I say this partly to play devil’s advocate, but also because I suspect garage operations are the hidden backstage components really only known to a subset of transit enthusiasts and advocates, let alone the general public.

    Steve: Remember that except for the middle of the night, there are 500 or more buses on the road. The capacity crunch comes when most of the fleet is “at home” and has to be serviced and set up for the following day. Buses do park in the aisles because the garage staff know that they will depart at set times, unlike a parking lot for cars. Think of the capacity we could achieve simply by converting the expressways to parking lots (formally) and giving everyone assigned arrival and departure times! Now imagine the chaos when one of those parked cars won’t start, and becomes an obstacle to every other vehicle trying to leave.

    Another change in recent years is the increase in the ratio of spare vehicles to scheduled service. This change has resulted in better vehicle performance because there is more preventative maintenance, and the TTC has now reached the point that there are on occasion a few more buses available for service than are actually needed. The result is that buses of dubious reliability don’t go out on the street to make up the service, and the fleet is more reliable. However, those extra spares mean that the fleet is bigger even if there is no change in service levels, and that adds to garage congestion.

    When John Tory trumpeted special funding for the purchase of 99 buses a few years ago, half of them went to backfill the streetcar fleet, and the other half went to increasing the spare pool.


  12. Notable from its absence on the list is reducing the amount of capacity that is wasted through bunching. When you have three buses in a row — one packed, one comfortably full, and one virtually empty — that’s wasted capacity that could be actually used if the service was better spread out. (I am preaching to the choir here.) Better still, not only is it making better use of capacity, it makes all the riders’ experience better. We know that some level of bunching is unavoidable, but I think your analyses have shown that there are a lot of things under TTC’s control that they could use as tools to at least reduce the magnitude of bunching.

    A more subtle example of this is routes that have branches with incompatible headways. Your Feb. 3 post on crowding shows the example of the 91 Woodbine having standees south of O’Connor in the midday peak. That is the part of the route that, in theory, has extra capacity from the route formerly known as 91A, but it only runs every half hour and its schedule doesn’t necessarily align with the main route… if it runs NB two minutes after a 91, it only picks up a handful of riders that might have just missed the 91 and are not going beyond O’Connor, and the odd rider that might be going to Parkview Hills and is actually specifically waiting for it. In those cases, it’s not really contributing at all to capacity on the main route.

    Steve: Yes, the 91/93 route is an example of the lack of co-ordination of two “blended” services. Mind you, the old Parkview Hills branch of the 91 ran on its own schedule separate from rest of the line. All that has happened is a change in route number for reasons only the TTC boffins will understand. That said, there are also problems with branching services even when they have co-ordinated headways. The absence of line management means that the “A” and “B” services are not meshed properly, and it is not uncommon to see the combined part of a route with ragged service. The TTC claims that they will be able to better manage this with their new vehicle monitoring system (aka “VISION”), but the laissez-faire culture regarding service management is deeply entrenched at TTC. I am not hopeful about this.

    Finally, your comment about capacity going down as crowding goes up is a useful reminder, especially for bus routes. I can point to an opposite example of this, on one Canada Day a few years ago when 92 Woodbine South had maximum service pretty much all day, not just in the lead-up to the fireworks — basically running buses as frequently as Woodbine station could handle them. It meant they weren’t stuffed during the day, could load and unload faster, and could run faster and more frequently, but the big part was that it delayed the start of the real serious crowding before the start of the evening fireworks.

    Steve: I have been on the 92 when it is running wall to wall service for an event at the beach, and yes, it’s impressive what can be done when huge amounts of time are not wasted packing every last passenger into a bus before it departs.


  13. My gut feeling is that construction of the Eglinton LRT station is running late. Every time they send out a construction notice of some work that they need to do, that work actually ends up happening a few months after they said it would in the notice. If Eglinton station is late, then the whole Eglinton LRT line won’t open as scheduled, meaning that we won’t have extra buses available in 2021.


  14. Steve said: Are you living in some sort of fantasy world?

    No I was rather asking – what sort of political fantasy was being spun? I mean it is all doable, if you are willing to really commit, and take heat elsewhere. If you were willing to take road space, or run through parkland. I just cannot imaging – actually achieving any let alone all, these days in Toronto. It would merely be political suicide.

    Steve: My tongue was, of course, planted firmly in my cheek with that remark!


  15. If they made the premium express buses regular fare and rerouted them on King Street to increase their reliability that would divert a small number of riders at a low cost and maximize bus utilization during peak periods.

    Steve: Actually, those buses would just get in the way, and they would consume green time at intersections that should be used for more local streetcar service. Also, to make any substantial dent in subway demand would require vastly more service on the 14x routes than is now operated.


  16. Steve: Because they only care about subways, subways, subways, and don’t want to spend another penny actually paying for improved service.

    The obsession with limiting property tax increases to “inflation” (even though the true rate is a little higher) is very damaging to our City. It is not only transit that is suffering – a whole lot of services are badly underfunded and crumbling infrastructure is not repaired or upgraded.

    On a more humorous note – and I only paraphrase a bit for effect – the TTC has announced that they are upgrading the retail in the subway. It will allow us, they say, to shop for our milk and eggs during the delays.

    Steve: That whole new retail strategy is such a pile of BS that I could not bring myself to write about it. The Board spent an hour discussing the report which suggests revenue might rise by a whole $3 million/year over the next six years. This for an organization with a $2 billion annual budget. They get enthusiastic about this penny ante stuff, but improve service? Oh that would cost money.

    By the way, don’t hold your breath for picking up milk and eggs as that requires power and refrigeration (among other things). There are considerable constraints on what can be added to subway stations based on available space, building codes (some existing shops are grandfathered and could not open as new today), power and plumbing needs. The report mentions this, but the discussion more or less ignored these issues.


  17. Me: How much money (sunk cost) was spent to cancel the Scarborough LRT?

    Steve said: About $75 million (sunk cost).

    Can we sue the Ford family to recover that? Can we sue Bombardier? Because there is no way that Bombardier could have produced the vehicles in time.

    Steve: Remember that if Transit City had proceeded as planned, we would likely have had the 70% low floor cars like the ones they built for Minneapolis. Our delivery problems stem, in part, from discarding that original proposal and going with what was then a brand new design of a 100% low floor car which Bombardier proved they could not produce competently. Also, the streetcar manufacture would have started before Bombardier was preoccupied with its aircraft project to the detriment of management and capital for the rail side of their operations.


  18. At this point, the situation in the Toronto area has been allowed to deteriorate so extensively for so long, I really think all three levels of government would have to admit that there’s a very serious problem and commit to sinking serious money into building the large U shaped relief line, a number of badly needed LRT lines, expansion of the existing streetcar network where it makes sense and increasing bus service across the whole city, and aggressively expanding the capacity of the commuter rail system on an aggressive time scale to make any serious progress on the problem.

    Politically, this would be a tough if not impossible sell to do a 180 degree turn after decades of baiting voters with the cheap, nasty, miserly “don’t spend any money” shortsighted routine that’s been going on for so long. Additionally, if somehow the funding taps do get opened up a bit, the last, absolute worst thing that could be done is blow the money on some kind of pie-in-the-sky scheme and build a short distance of something dubious like a hydrogen powered Maglev monorail or other similar piece of idiotic industrial policy item like that instead of delivering something serious that would count against mitigating some of the problem on our hands.


  19. I am not as limber as I once was, though I still can climb stairs – albeit slowly and sometime one foot at a time. However, after climbing the three steps at the front, CLRVs/ALRVs are much easier to walk around than LFLRVs.

    I infer that you would agree that 70% low floor – here and in service – would be better than the slogan inspired 100% low floor that is not here and not in service. Is that correct?

    Steve: It’s a question of timing. Toronto should have been ordering replacement “legacy” cars quite a long time ago, probably before Transit City was announced, but the political situation would have worked against it without the momentum of the suburban LRT expansion. The first proposal we got was for 70% low floors, but then along came the shiny new 100% low floors. At the time, it appeared that they would show up “tomorrow” but things didn’t quite work out like that.

    The current cars do not have friendly seating. The seats over the trucks are not as accessible as seats with a raised floor, steps and even no step access. (Such as on the buses.) The floors, while ostensibly low floor, are not even. There are humps (over the trucks) that are not marked with yellow caution tape, that are a tripping hazard for those of us with reduced mobility. There are dips – also not marked – at at least some of the doors.

    Compromises have been made to meet a theoretical accessibility standard. I am not convinced that the finished product really is “accessible”. Low floor access with clearly marked steps to 30% might actually provide better and safer accessibility.

    Steve: That “dip” at one of the doors is extra depth to reduce the height the ramp has to handle when deployed on street.


  20. Steve: Because they only care about subways, subways, subways, and don’t want to spend another penny actually paying for improved service.

    Yes, buses and streetcars are a quicker fix, but they have limited capacity. And if they do work to fix the current crowding on the bus and streetcar routes, all they will do is drop more riders into the overcrowded subway.

    Given the timeline issues with SSE and B/D expansion/upgrading would it make sense to put the SSE on hold and divert the money to the LRT east extension and the DRL?

    Is there enough room on the DVP and the political will to designate dedicated BRT lanes from Don Mills to Downtown?

    Steve: The challenge would be in finding room on downtown streets for a high capacity bus service.

    Would surveying the ridership moving through Bloor/Yonge station give us hints on origin and destinations that would help find diversion fixes that would stretch capacity?

    Steve: There are other ways to get this information, especially once Presto is rolled out, by using trip tracking to pair up inbound and outbound origins during the two peaks.

    There is pent up demand for better operated and more transit in Toronto as evidenced by the “we need to do something” attitude in the air. Take the bull by the horns and budget some real money (0.5-0.8 B$) to transit operations.


  21. Steve – could the TRs now on Sheppard be moved back to YUS and replaced with T1s? Thanks.

    Steve: No. They are built as a four-car consist. Also, once the YUS converts to ATC, the T1s would not be able to run over it to and from Davisville Yard without some modification to allow them to “talk” to the new signal system and tell nearby ATC trains where they are.


  22. Steve said: “In order for dwell times at Bloor-Yonge to be reduced to a level that would permit much more frequent service, passengers must be able to move freely both between trains and platform, and within the station itself. A full train meeting a full platform leads to gridlock. To avoid this there must be both room on the trains and in the station to avoid dwell times that are so long that the planned train frequency cannot be operated. ATC will help somewhat by allowing trains to bunch closer together, and less time will be wasted between trains “A” departing from the platform and train “B” arriving, but the problem of station capacity remains especially when there is a delay.”

    It would greatly increase efficiency in handling passenger load at stations if the passengers exited from one set of train doors, and entered from the other – more flow through, less friction, less jostling.

    However, that requires three platforms. Such a rebuild could be looked at for the busiest stations in the system, for example, Yonge-Bloor, Queen, and Eglinton. Isn’t Eglinton already undergoing a renovation with regards to the Eglinton-Crosstown LRT? Is there anywhere in the system already with this setup?

    Steve: Adding a third platform is extremely difficult because the station is hemmed in by buildings. TTC does have a scheme to expand Bloor-Yonge, but what it would require is (a) a very large amount of money and (b) massive disruption while the subway attempts to continue operation.

    Another line is what is required, not another platform, because the latter does not address the basic issue of overall network capacity into and out of the core area.

    If you want three platforms, you build the line that way from the outset.


  23. When the Yonge subway was built, they ordered about 100 cars for it, expecting that to suffice for ages. They also ran 4-car trains, expanding to 6 in rush hour. (IIRC).
    The table in JFB’s book suggests that in 1960 rush hours there might be 140 PCCs on the Bloor/Danforth routes.
    So that is what the capacity was designed for at the Bloor station.

    Steve: The 1964 TTC schedules are linked in an article I wrote two years ago for the 50th anniversary of the BD line. The AM peak service at Bloor-Yonge consisted of a train of PCCs on the Bloor (Jane-Luttrell) service every 2’30” (24 trains, 48 cars/hour) plus trains on the Danforth (Bedford-Coxwell) service every 3’20” (18 trains, 36 cars/hour) for a total of 42 trains, 84 PCC cars per hour. At a service design capacity of about 70, that would be 5,880/hour, but the cars on Bloor-Danforth ran packed and so their average load would be higher. Demand was cited in the 7,000 to 8,000 per hour range.

    The service came to Yonge from both directions, although the eastbound Danforth trippers would be lightly loaded as they originated at Bedford Loop. Leaving these out, this meant that 132 cars/hour arrived at Bloor Station.


  24. A lot of the demand on Line 1 is local traffic, particularly downtown traffic. A surface route on Yonge Street could provide relief by off-loading this local traffic.

    An example of this in action can be seen on Vredenberg in Utrecht. Vredenberg is the main East-West street in Utrecht. Private cars were removed from this street in 1996. In a recent traffic count, 195 buses and almost 4,000 cyclists were counted in an hour.

    Yonge Street has been car-free downtown as recently as the 1970’s.

    Using 150 passenger articulated buses on a 2′ headway gives a capacity of 4,500 passengers per hour. Another 4,000 cyclists would give a total surface capacity of Yonge Street at 8,500 people per hour.

    This would be fairly cheap and easy to implement. The public transit component would require buses, garage space for those buses, and transit signal priority at intersections on an otherwise car-free Yonge Street. The cycling component requires nothing more than paint to delineate bus vs. bicycle street space.

    For local traffic, surface routes have a key advantage over subways. Depending upon the subway station, it can take up to five minutes to walk from the street to the subway platform. And the same again from the platform at the destination station to the street. Giving a built-in 10 minute advantage to the surface route.

    Steve: I beg to differ with your premise about “local” traffic downtown. We know that the subway is jammed well north of Bloor, and so adding surface capacity from there south will not address the problem further up the line. That’s not to say that taking over a major street (and it does not have to be Yonge itself) as a cycling street is not worthwhile, just that these are two separate issues. Like the problem with surface routes being necessary to feed rapid transit, a cycling corridor downtown needs feeder streets with good cycling facilities, and there must be capacity/infrastructure downtown to disperse cyclists to their destinations.


  25. Steve wrote:

    “We know that the subway is jammed well north of Bloor, and so adding surface capacity from there south will not address the problem further up the line.”

    Very true. I certainly did not mean to imply that this was a “one magic bullet” proposal that solves all problems. If that impression was given, then my comment was poorly worded.

    I was writing only about local demand downtown. There is, of course, a large amount of such local demand. We can see this in the crowding of local services such as on King.

    In my opinion, effective relief of the Yonge line requires three major solutions.

    1. A downtown relief line (Don Mills subway) extending to Eglinton.
    2. 15 minute GO service on the Richmond Hill line, with the existing Oriole GO station moved 250 metres north up the track and co-located with the Leslie subway station.
    3. Local downtown traffic offloaded onto surface routes on a car-free Yonge Street.

    Although item #1 will deliver the greatest amount of relief, constructing an entirely new subway line is very costly and takes a very long time. However, it must be done. When the Eglinton Crosstown opens, it will create a minor crisis. There is no more existing capacity on the Yonge line from Eglinton south during the AM peak. All the new riders on the Crosstown will not be able to get on the subway at Yonge. They need a DRL to get downtown.

    For item #2, 15 minute GO service on the Richmond Hill line is currently planned by Metrolinx. However, relocating the GO station to co-locate it with the Leslie subway station on the Sheppard line is not part of that plan. This is foolish, because a large amount of traffic on the Sheppard line is currently headed downtown via the Yonge subway. Diverting this traffic onto the GO line would provide some much-needed relief to the Yonge subway.

    Item #3 is largely a matter of political will to reallocate street space. However, that political will is currently being generated by current downtown residential construction. Only considering what is approved and/or under construction right now, 10 years from now the downtown population will be sharply higher. It is a matter of simple mathematics, specifically geometry, to conclude that the only way to get those people moving at all is to severely reduce the number of private automobiles in downtown Toronto. And that higher downtown population will have the electoral clout to make it happen.


  26. Thanks for all of this Steve, and comments. The neutral recitation and compilations of facts make your tart political remarks even more cutting. It sure feels like most everyone elected at all levels should be ousted, though in some ways, better the devils you know…

    But there are some short term improvements possible that won’t be silver bullets, but silver buckshot, though it requires political will and some thought. The eg. of Curitiba where they had political will to clear the private cars off of some roads meant very effective and cheap-to-do bus transit. On some streets, like Yonge and Danforth, I think we’d be able to do road repaints to produce centre-lane reversible busways, though of course, how do the buses return. If we thought of providing bike safety at the same time ie. ensured safer, continuous. and kept-up road surfaces, we could have some bikeway relief. This is especially true on Bloor/Danforth and it has been the best place for east-west in a 1992 study, but the pathetic nature of the City is such that even the remnant bit of Stage 1 of that 1992 study that made it into the 2001 Bike Plan, that little bit of Bloor St. E. between Sherbourne and Church, it’s STILL unsafe/undone, and for a whopping $25,000. Ms. Wong Tam is part of the problem perhaps, and I’m tired of ward-by-ward bike safety when corridor thinking is required, as per Places to Grow, not that the provincial Liberals are at all keen on actually doing what that says should be done.

    While repainting the roads for busways is ‘roadical’ and disruptive, the sheer volumes of folks who are PAYING for inferior service greatly outweighs the numbers of people in often-1-each cars, and the public transit users must have a clear claim to the public road as the transit is what is supposed to be favoured by Official Plan and sustain-the-bull right??

    As I can go on about a lot of options that aren’t part of our thinking apparently, I’ll limit to a need to look at surface options first ahead of diggings etc., and maybe we should go for a pretty good fix, if long, vs. ‘best’ of subway Relief, which I’m almost against at this point as the entire region needs big change, not just a Fix, which does Ford-all for the Yonge line. So what about using the Don Valley, and Parkway and adjacent rail corridor for priority surface routes of some robust nature (busway, LRT). And bypass Bloor/Danforth up to Thorncliffe, and from Thorncliffe have two further routes that get up to Don Mills Road and then north as one leg, and another leg crossing the Don with a new construct to start using the very long, direct, wide and diagonal Gatineau Hydro corridor to reach out through Scarborough to the Zoo.

    This would still require a couple of years worth of heavy construction to keep the construction interests happy-ish, and concrete consumed, but I think we could get a two-for-one SURFACE relief line that could feed in to the core somewhere.

    With the flood control, about 70% of the Don flooding is from storm drains etc., apparently, in a recent book about the History of the Don. (Buried somewhere, Cottrell??), And when Mr. Tory was chair of CivicAction we got a D or an F on the water control issue, in a c. 2013 report, but when it was possible to start having a drainage fee for stormwater, oh nope, too much to put in to place, so he’s responsible for continuing the subsidy for cars and screwing up a cheaper and better/faster option. Total cost for opting for Suspect Subway Extension is c. $200M right? I hope he barely wins Scarborough and loses everywhere else – and Mr. Colle is also unworthy, and more than a few others.

    I’m avoiding thoughts on doing express subways comment btw, as I go on. Thanks again Steve.


  27. For using surcharges on certain lines, there must be tap in and out at every mode. For example, someone must tap out of the 199 bus and then tap in again at Finch Station. It keeps everyone honest and provides a lot of OD data for future transit planners. One can count how many people ride the 199 bus per hour, but the passengers are using it because the 42 bus is not running frequent enough, the TTC will need to reconsider. A $2 surcharge until the relief line is paid off is not a bad strategy. It also equalizes the fares with GO. GO fares within Toronto are around $5.60. This way passengers will flow to the best form of transportation for them as oppose to choosing the cheapest one.

    Tap in and out can also allow for rewards. Since King Station is so crowded, Presto can be programmed to give a small reward for people using the Melinda Street exit instead of the main one. It could be a $0.10 credit to Presto or something like 1 Aeroplan Mile. It also has the side benefit of getting people to walk more. So, it is healthy too.

    The Richmond Hill Line is not really easy to buy for Metrolinx. CP and CN are holding up transit in this province. An extreme solution would be to leverage the asset of Hydro One and OLG and do a leveraged buyout of these two companies. Even with the corridor in hand, there would still be a lot of upgrades. Flood prevention is one. The other are twinning it and banking (upgrading) the curves. Tilting trains cannot go around any curves. In order for tilting to work, the curves must be banked. Money can solve problems, but at what cost.

    Yes, the Richmond Hill Line should be upgraded to RER standards. However, it takes time. Look at how long it takes to upgrade the Stouffville Line. It is taking more than 2 years just to twin the tracks between Marilyn Ave to Kennedy Rd. This is on level ground with no major obstructions. Metrolinx’s record is not great at building things fast either. Even if the Richmond Hill Line is upgraded, it will take at least 7 years.

    This is why a solution needs to be done now. If someone actually gets injured waiting at Bloor Yonge, will the politicians be held responsible for not doing anything? A monorail is not the most elegant solution as there is some visual pollution. However, relief in 3 years vs relief in 10 years (for a metro) must weighted against whether overcrowding is a present and credible danger. Bombardier’s monorail can move 30000+ people per hour. At the stations, buildings nearby can be redeveloped, so that the trains enter the buildings as oppose to having a station that cover the entire intersection. Monorails have the smallest cross section, so the visual impact is kept to a minimum.

    Steve: Your naivety about Bombardier’s technology continues to amaze me. The assumption that the problem of stations over streets can be wished away by having them dive into nearby (and redeveloped) buildings ignores the fact that in many locations the buildings are there to stay. There are geometric issue too that I won’t go into.

    As for saying “we cannot do X because Metrolinx will take too long”, well that would apply to any solution we would dream up. Of course we could always farm this out to the private sector, maybe a company like Bombardier with such a sterling record in such matters.

    The length of time taken for some projects is a direct result of the spend rate at Queen’s Park and the desire not to do everything at once. One reason a P3 looks attractive is that a private builder wants to start making money on an asset as quickly as possible rather than dragging out construction to limit spending in the short term. The tradeoff, of course, is that the public partner is then on the hook for decades of operating and lease cost in which are buried the borrowing cost plus markup.

    To be quite clear, I have never argued that the Richmond Hill GO corridor can be operated at a comparable frequency to, say, the wholly-owned Barrie line, but this does not mean RH GO can’t play a part, especially before the Relief Line “long” comes on stream. Yes it floods now and then. Does this mean we should not ever operate service there? What I hear too often is excuses, not commitments to do what can be done give the long history of putting off any meaningful work on what is now a critical, dangerous problem in the network.


  28. It seems the ALRV rebuild program is done. From another forum

    “4205 … was gutted to prepare for the refurbishment, but someone higher up wrapped up the ALRV life extension program. 4213, 4241 were slated to go for refurbishment, but now that’s done and gone.”

    While the Richmond Hill line is limited by Doncaster diamond in how many total trains can be run, I think with a reasonable application of funds it might be possible to operate a few peak trains to/from Old Cummer by building a bypass track so the 905 trains can zip past the station and some additional double track so the 416 train can get up to OC – and if the hydro wires were rerouted from beside Bayview (ambitious!) it might even be possible to plunk down a 20 inch height platform beside the west track with ramps and elevators to the Queen and Dundas bridges/streetcars. The question would be how to get enough people in North York to board the train.

    This would be expensive to do but all about reducing peak pressure on Yonge-Bloor rather than conventional cost benefit.

    Steve: I am not surprised that the ALRV program was shut down considering that these are the least reliable cars in the fleet, way worse than the CLRVs. The refurbishment should be a deep embarrassment to whoever designed it.

    As for the RH corridor, I really don’t see the point of trying to intercept passengers at Queen and Dundas as this is well south of the Bloor-Yonge crunch point. A few riders might be tempted to use the streetcars to/from the core, but they can just as easily to to Union.

    The biggest issue on that line is the service frequency and the penalty that the transfer time to GO presents to anyone contemplating shifting their travel to it.


  29. Steve, I am sure that we can agree the relief line is needed and hopefully soon. We may disagree on the technology, but that is okay. We may pay more with a P3 solution where a company like Bombardier will design, build and operate a transit line. However, it may be worth the extra cost if we can get it sooner and the business case is sound.

    The federal goverment does not have a good record of funding transit. Think about it this way, VIA Rail serves 9 provinces in the dominion. Yet, the Liberal MPs cannot even agree on $2 billion to buy new coaches for the crown corporation. How many MPs we have in Toronto? If a technology exist where the relief line can be built cheaply and quickly, we ought to look at it. It might be $4 billion to build the relief line right now using metro technlogy. It could be $6 or even $8 billion due to inflation by the time all the funding partners can be found.

    If we can buld the relief line now, it gives Toronto an insurance that should GO Transit cannot bring additional capacity on board. I want them to bring more capacity on board, but we need to hope for the best and plan for the worst. On the non GO owned corridors, negotiations can go bad. CP and CN have their own interest, but they are playing hard ball.

    I am not making excuses for Metrolinx here because I do not want to play their hand right now. I am just writing what I see in this political envrionment. Right now, Metrolinx is bogged down in negotiations in non GO corridors with CN and CP. If the 407 bypass can be constructed now, the Kitchener Line (beyond Mount Pleasant) and Milton Line can go farther and more often. Every community in southern Ontario is pleading for more GO lines. Think of all the areas in Toronto demanding their own metro lines except on a different scale. Unless there is funding for everything in one go, projects might be priortized. For example, Kitchener GO service may go ahead before work begin for a GO train service to Cambridge. HSR is suppose to be running by 2025. Will the provincial government have an appetite to spend a few billion upgrade the Richmond Hill line concurrently with other lines? I am not going to speculate.

    The regulatory envrionment for Metrolinx is also very challenging despite good intentions. Even though Metrolinx, CP and CN are federally regulated railways (under Federal Railways Act), it does appear that Metrolinx’s interest are looked after. Here is an example, since the Barrie Line became mid day 2 ways service, service has been short turned at Maple GO Station more than once. The reason is to accomodate freight trains. This is a Metrolinx owned corridor, but CN still have freight rights that somehow trump GO’s dispatch.

    CN and CP already vetoed overhead wires in their corridors. They may even veto using CBTC signalling. The federal government needs either re-write the Federal Railways Act or make a law in the same manner as the Air Canada Act. If CBTC with PTC (Positive Train Control) is required for all railways in urban areas, then Metrolinx’s work will be easier. The EU already requires that all airport serving commercial flights needs instrumented landing system. This allows a plane to land using computers in bad weather. Everyone just have it and there is no argument. If negotiations cannot be made, Metrolinx should be allowed to build tracks right next to the private company’s right of way.


  30. When I started working by the 1970’s, the “work day” had changed away from the old “9 to 5” spread. We could work “8 to 4”, “8:30 to 4:30”, even “7:55 to 3:55”, as long as we were in the office during a “core period”.

    That helped some of the crowding on the TTC (and the expressways)… until 2018.


  31. Benny Cheung wrote: “If the 407 bypass can be constructed now, ….”

    This is dead now that the province has killed off this highway bypass. The RR line was to go in that same corridor.


  32. The CN-owned portion of the Bala Subdivision required to provide a higher-frequency, bi-directional GO service is very small. There would need to be a fly-under at the current Doncaster diamond crossing of the CN York Subdivision and a short section of GO-only trackage north to the existing station. It’s not that it can’t be done, it’s that the politicians don’t want to do it. Somewhat like how the city politicians didn’t want to let Skoda bring its 70% low-floor demonstrator car to Toronto in 2008. But that’s another story.


  33. Steve: I am not surprised that the ALRV program was shut down considering that these are the least reliable cars in the fleet, way worse than the CLRVs. The refurbishment should be a deep embarrassment to whoever designed it.

    I *could not* believe that the rebuild/refurb was $25M on COSMETIC. Un-f*****g real. “We don’t have the money for better service…” (the line they’ve offered for my entire life), and I believe them, and then they do something like this. Truly outrageous.


  34. Malcolm N says “So even if magically found a route, to give meaningful relief would mean moving thousands of riders per hour, like 3 or 4 thousand or something like 60-100 buses per hour.”

    In light of the changes to King St. I would revisit the Don Valley Corridor study, fully implement it picking the most appropriate route from Don Mills to King St. Most of the needed structures are already in place depending on the route selected.

    I personally prefer adding shoulder lanes to the DVP south of Don Mills to accommodate high occupancy vehicles.

    Depending on how aggressive they want to be a mixed effort between GO and the TTC can reasonably provide relief, while at the same time providing better service to the broader community.

    Steve: There are a few pinch points that could be troublesome. One is at the Prince Edward Viaduct, and another is on the ramps to/from downtown street, and on those streets. We’re talking about a lot of buses, and a lot of new pedestrian activity on, say, Richmond and Adelaide.


  35. “Steve: There are a few pinch points that could be troublesome. One is at the Prince Edward Viaduct, and another is on the ramps to/from downtown street, and on those streets. We’re talking about a lot of buses, and a lot of new pedestrian activity on, say, Richmond and Adelaide.”

    Yes there are issues, but the cost estimates to correct them are quite reasonable. Frankly of all the options to relieve Yonge this is likely the cheapest and most effective, and easiest to implement. As for the downtown streets that will have to accommodate the added traffic that issue is really a matter of choice. The city already has bike lanes that handle a fraction of the volume of a priority bus route, and the plans on King St have created a lot of wasted street capacity in the effort to prevent through traffic. The city does have the space to implement a downtown bus route found in the don valley corridor study. It merely needs the will to optimize the performance of its assets in the area.

    Steve: The absence of a rebuttal here is not an indication I agree with you, merely that I tire of this sort of debate.


  36. Regarding the GO bus terminal at Scarborough Town Centre, do most GO riders use it to connect to the SRT? Can you comment on the breakdown of the trip types and connections that are being made?

    Thank you

    Steve: I have never seen a breakdown of this traffic, but from my time working at Scarborough City Hall, I would say most GO riders transfer to the TTC either to the RT or to a bus route. Where they would be headed from there, I don’t know.


  37. Radical idea. A service rearrangement as follows. Reinstall moving walkways at Spadina for a few million dollars and replace escalators with speed ramps for accessibility. Terminate all Spadina trains at Lowther Stn during AM rush periods for the 90 minute crunch.

    Reactivate BDU wye, even if it means forcing traffic through the wye whereby Bloor trains WOULD NOT STOP at Yonge Stn. during that time.

    Passengers from the WEST will no longer be changing at St. George and will be forced down University and up Yonge (distributing traffic to use the stairwell capacity on Northbound platforms at KING and QUEEN). Passengers from the EAST — trains will not STOP at Yonge Stn. but will instead proceed to Bay Lower and STOP there. Bay and Yonge are within walking distance.

    Passengers from the WEST may, on if their next train happened to be an E-W train, transfer at Bay to a southbound University train.

    Public safety is paramount. Bloor-Yonge has been dangerous since the 1980s. I remember being scared to the point that I would actually transfer at Spadina/Lowther and go through the tunnel.

    This will undoubtedly inconvenience many passengers on SPADINA until a real solution can be found, but passengers need to stay on trains. There was talk in 1968 and again in 1995 of reactivating the wye during the busy hours. It’s the passenger volumes from the Bloor trains that are the problem here.

    The load at Lowther southbound would unfortunately have to totally disembark, go through the tunnel, and either board a Downtown train at Spadina Eastbound, or use the Spadina LRT, or use an EW train and transfer at Bay.

    All of that transfer traffic has to be spread out until the DRL is built. Does anybody from the TTC read this blog?

    Steve: You have a future as a totally out of touch transit planner. There should be lots of work for you in Toronto.


Comments are closed.