At its meeting of February 11, 2015, Toronto Council debated a report from the Medical Officer of Health on Suicide Prevention. In response to this report, Council approved the following motion (which is a modified version of one of the MOH’s recommendations):

1. City Council request the Toronto Transit Commission to consider the following improvements to passenger safety and suicide prevention in future budget submissions as the automatic train control project is completed: a. in the design of stations for all future extensions or new lines include Platform Edge Doors or other means for restricting unauthorized access to the subway tracks by members of the public; b. retrofit existing stations with Platform Edge Doors or other means for restricting unauthorized access to the subway tracks by members of the public.

Please refer to the update at the end of this article for comment about the content of the debate which is now available online.

During the debate, various claims were made for the benefits of Platform Edge Doors (PEDs) on the advice of TTC staff, notably that it would not be possible to increase subway service from 28 trains/hour to 36/hour without the installation of PEDs.

28 trains/hour is equivalent to a headway of 128.6 seconds, somewhat shorter than the current scheduled level of 141 seconds, but within the capabilities of the existing signal system. 36 trains/hour is equivalent to a headway of 100 seconds which is well below the current infrastructure’s capacity.

This is the first time that the TTC has advanced PEDs not just as a “nice to have” option, but as a pre-requisite to improved subway service. The MOH cites a TTC report on the subject, but does not comment on its technical merit only regarding PEDs as a way to eliminate subway suicides, a noble goal.

The TTC received a presentation on this report in September 2010, but only a two-page covering report is online. (The TTC plans to post the longer version, but as I write this it is not yet online.)

According to this report:

In May 2010, SYSTRA Group (an affiliated company of Paris Metro) was retained to conduct a business case study for the installation of PEDS at TTC subway stations.

The SYSTRA report is not publicly available, but the presentation summary will be posted by the TTC soon. It is not yet on the TTC’s site as I write this article, but was provided to me by the TTC’s Brad Ross and is available here.

PEDs Business Case Presentation Sep 28, 2010

This presentation is misleading in that it combines benefits expected to flow from reduced headways through Automatic Train Control (ATC) and those specific to PEDs. A major benefit of the doors is to keep debris from falling onto the tracks where it creates a fire hazard. However, a separate review of TTC operations by an international consulting group noted that the TTC’s ability to operate its advertised service is compromised by several factors including equipment reliability and passenger illness (some of which is a result of overcrowding).

Among the statements included in the Business Case:

Current capacity is 26 trains/hour in peak

Need to move to 35 trains/hour to meet required forecast capacity in 2030

How do we increase capacity and reliability? Move more trains per hour; currently constrained by signal system headway Train capacity More efficient use of trains interior space Longer trains Reliability, ensuring that the line is operated at peak capacity



This is followed immediately by:

Industry Best Practices Toronto will not achieve the target level of reliability through automation alone Even with TR and ATC, Toronto needs to reduce total incidents by 75% to achieve the target reliability level of 1 peak failure per week Toronto needs to target key areas first and then evaluate every aspect of the subway in terms of reliability



The issue, then, is how that goal of 1 peak failure/week can be achieved and which of many possible approaches will contribute most. Anyone who follows the TTC’s E-Alerts for subway delays will know that common problems include equipment failures, passenger illness, investigations of possible fires, and unauthorized persons at track level, not to mention the unending series of signal failures.

The TTC does not publish breakdowns of the types of incidents nor of their severity. In other words, we don’t know as a starting point (at least for public debate) what type of incident should be our primary focus to reduce delays in number and severity. The TTC does not even publish reliability statistics for its various fleets of vehicles although past procurement reports both for the T1 and TR trainsets claimed reliability improvements (and hence reduction in provision for spares) would occur. The monthly CEO report tells us many things, but it does not show how the TTC might be achieving a most critical goal — reliability. Only aggregate numbers for service are shown and, as discussed here before, they are completely bogus because of the underlying methodology (half of the subway service could be missing and it would still score 100% provided that what remained was regularly spaced).

How can we know whether the TTC is working toward those “best practices” if it does not publish statistics to track progress?

Key areas for improvement are listed including:

New rolling stock

Modern signalling and train control

Management of trains in stations

Platform Edge Doors

By 2010 when this presentation was given, the entire T1 fleet had been in service for almost a decade, and delivery of the TR trainsets had started. It’s a bit of a stretch to continue pointing at old equipment as the source of rolling stock problems, and by today the retirement of all of the previous equipment generations should leave us with a marvel of fleet reliability.

Signalling and train control, as we know, will take a while.

Station management is important, but an essential part of this is to avoid overloading station capacity in the first place. As long as the subway runs at over 100% capacity (with even higher peaks from the effect of any delays), the movement of passengers between platforms and trains, as well as within the stations, will be a big problem for the TTC. PEDs will not eliminate the crowding, only eliminate the possibility of passengers being pushed onto the tracks.

The goal should be to avoid routine overcrowding. It is amusing that an illustration of the doors shows Rosedale Station with a handful of waiting passengers.

In the Business Case, the study assumes 1.5% annual increase in ridership, a six-year installation program, and a cost/station of $9.8m. That produces a total of about $676m for 69 stations, well below the current budgetary estimates. However, the “investment” according to the presentation would be only $511.6m, or about 52 stations’ worth. At no point does the analysis include the cost of major changes, notably $1b for Bloor-Yonge Station, as an essential part of capacity improvements, nor does the presentation account for the supposed “economic benefits” of $567.1m from the PED project.

The presentation concludes that

PEDs are necessary to achieve TTC performance objectives.

although nowhere does it explain the limitations the TTC would face in their absence.

In a chart (page 21), the presentation includes a chart showing the potential reliability improvement to take Toronto to the top of the range of European systems, or the low end of new Asian systems. However, it does not explain how much each of the four types of improvement (of which PEDs are only one) contributes to the overall goal.

In a further claim (page 22), the presentation claims that

Adopting these improvements creates the equivalent of another YUS branch.

This is a claim that at least one Councillor during debate made simply for PEDs, not for the several components of reliability and capacity improvements. In any event, another YUS branch would require a line with the capacity of at least 26 trains/hour. This is not achieved simply by bringing the YUS headway down to 100 seconds (36 trains/hour) even presuming this is operationally feasible. In that regard, the presentation misrepresents of the possible new capacity that could be provided. (Note that there would be additional trains on both the Yonge and University branches so that the change in trains/hour is effectively doubled.)

However, more trains/hour don’t solve all of the capacity problems.

Other TTC studies regarding shorter headways have made clear that they are impossible without reconfigurations of terminals and connection stations. Terminals must be changed to sustain shorter spacing between trains and minimize conflicts between inbound and outbound movements. Connection stations, notably Bloor-Yonge, require substantial additional capacity for movement on and between platforms.

Yonge Subway Headway Study (first of eight articles)

Yonge Subway Extension – Recommended Concept/Project Issues (see pages 32-33)

The TTC Capital Plan includes several “below the line” projects including Yonge-Bloor Capacity Improvements at a projected cost of $1.052-billion. This is on top of the estimated $550m for PEDs on the YUS plus $614m for PEDs on the BD line.

A pre-requisite for use of PEDs is Automatic Train Control to ensure proper train-to-station positioning. This project is now underway on the YUS, but will not complete until 2019. On the BD line, it will not finish until the mid-2020s by which time a new ATC-based fleet (or possibly the existing TR trains relocated from the Yonge line) would be available.

If Toronto is to study the viability of substantially increased capacity on the Yonge-University line, then many factors need to be included in the cost base for any “business case analysis” including:

How many additional trainsets will be required to operate the YUS at 36 trains/hour over much of its length? What will these cost? Where will they be stored and maintained? Note that the very frequent service must operate over much of a very long route (at least Finch to Downsview, possibly north to Richmond Hill in the future), and this magnifies the effect of shorter headways on fleet size.

What is the currently proposed design for expansion of capacity at Bloor-Yonge Station (the $1b unfunded project)?

Which other stations will require added capacity notably St. George but also other busy YUS stations with limited platform access (e.g. College, Dundas, King)?

How will the higher capacity of the YUS affect transfer traffic onto the BD subway during the PM peak period? What effect will this have on the required level of service, fleet size and train storage requirements?

What will be the additional operating cost of both the extra train service, of the station facilities and of the control systems to make all of this work?

In the context of the Yonge Relief Study now underway, these costs (and the construction issues related to various improvements) must be weighed against the perceived high cost of building a separate new subway line into the core area (aka the “Downtown Relief Line”). That project is often described as complex and unaffordable, but this is never done in the context of a comparison to the alternatives required to stuff more passengers into the YUS.

If the TTC is now claiming that it cannot achieve more frequent service without PEDs, let it say so, and be honest about the other costs Toronto will face just so that TTC staff don’t have to admit that their opposition to the DRL has been misplaced for decades. Even CEO Andy Byford speaks of the need for the DRL, and his organization has to present a full range of options rather than dragging out old, self-serving reports from the pre-Byford era.

A true “Business Case” looks at the big picture, not at one pet project however laudable it may be to the social cause of suicide prevention. The TTC must not wrap itself in an unchallengeable motherhood defence, especially when their proposal stresses service quality and capacity as its goals. There are many ways to get more commuters into downtown Toronto, and Platform Edge Doors are not an absolute pre-requisite.

Updated February 12, 2015 at 11:00 pm

The debate on this matter at Council is available on Rogers website.

Select the February 11 video and scroll forward to about the 15-minute mark.

Questions to and comments made by John O’Grady, TTC’s Chief Safety Officer:

[These have been reordered from the debate to keep related items together. My comments are in square brackets.]

The cost estimates did not include structural or electrical changes to stations needed to support PEDs and therefore the cost would be “substantially more”.

The cost estimate for the retrofit was cited as $551m. [Actually that only covers YUS, and with BD included it is over $1.1b. The discussion that followed was based on the incorrect lower number as if it were for the entire system.]

Why were doors not included in the TYSSE stations? The stations are designed to accept doors as a retrofit. There was a funding envelope for the line; this was an item that was removed to fit within in. [O’Grady made no mention of need for ATC as a pre-requisite for PEDs, nor of the fact that ATC was not part of the initial TYSSE design (it will be retrofit after the line opens).]

Could the TTC have trains approach stations at a slower speed? There is a Standard Operating Procedure to slow trains if there is a report of a passenger at risk, [but this is not a routine operation under normal circumstances].

Station entry would be quicker with ATC. Trains would get in quicker, get out quicker. [This is not strictly true because the constraint is the maximum G force from braking/acceleration that is safe for passengers. This only really applies to a situation where trains are queued near a station and might crawl in anyhow because they are too close to get back to speed before hitting the platform. The major benefit in faster loading comes from ATC which allows trains to be much closer together especially when moving at low speeds. I believe that there is a fundamental misunderstanding of the dynamics of how busy stations operate.]

PEDs make for more efficient loading. [This contradicts a statement made to me by Andy Byford.]

TTC will automate trains with or without PEDs, but doors will allow them to take better advantage of ATO – notably for driverless operation.

How many incidents occur? 34 in 2014 which was unusually high. This varies year to year, but lies in the high teens on average. There is no magnet station. [The subway fatalities represent less than 10% of all suicides in Toronto, although they do represent a major disturbance to subway service.]

With the direction from Council, the project would “have a lot of moral suasion”.

The point about “moral suasion” shows the danger of saying “we want this” without having a parallel discussion about spending and social policy priorities. A project with some sort of endorsement can acquire a heft it would not have otherwise.

Believe it or not, the Spadina Extension became an “add on” to the Ridership Growth Strategy when the engineering staff feared that the TTC would stop building rapid transit and focus entirely on surface routes, even though the whole intent of RGS was “short term improvements”. The Commission endorsed the TYSSE as its priority, but there was little concern by then-chair Howard Moscoe who claimed that there would never be money to build it. However, in the process the Commission said “this is our next priority” and this endorsement was repeatedly cited by TYSSE advocates. In for a penny, in for a pound.

Bad enough that this Council motion was informed, so to speak, by misleading staff statements, but the underlying report is itself flawed because it omits very large costs that would be integral to the capacity upgrades being discussed. The combined effect is that future debates will either continue in a fog of misinformation, or have to walk back a lot of mistaken impressions.