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.
Steve Munro 
Steve.. did you see Councillor Doucette’s remarks on CP24? She claims that trains can come into stations faster increasing capacity with platform doors. I shudder to think what she defines as faster given the distance to stop safely without injuring passengers.
I will say this some stations need PEDs… take Dundas for example its bordering on bursting on any given day.
Steve: I was watching Council when Cllr Doucette made that statement. I believe that she has been “badly advised”, as we say politely at City Hall, by the TTC. Trains enter stations at full speed today except when they are queued up behind a delay. At most, this remark might refer to entering a very crowded Bloor-Yonge if a train approached from far enough away to make acceleration to speed worthwhile. However, the new signal system will allow trains to creep up behind each other, and so if there is a queue, they will be accelerating from very close to the station. This sounds as if someone at the TTC has invented a rationale without thinking through what they are saying.
LikeLike
Steve other than the possibility of running into station at higher speed, how would platform doors actually assist headway. I understand allowing more crowding in the station safely, however this only applies to a handful of stations.
Steve: Please see my previous response re entering at speed. It’s not going to happen under most conditions, or at least not on a scale that will affect headways. In theory the doors allow better crowd control, but even Andy Byford has said that this does not always work and he is of two minds on the subject.
As for more crowding in the station, Bloor-Yonge and other stations are already packed full. “More” crowding is inconceivable.
LikeLike
Yes what is the relative cost of making 3-5 stations of 3 platform design to allow fast enough alighting and boarding in areas where real estate is extremely limited and platform doors in a large enough area to permit the trains to come into station at a high enough speed and additional turn and yard space to actually support 90 second or shorter headway, compare to simply building a new line? A new line which will also support an LRT that will allow the shortening a goodly number of bus routes, and keep them from having to operate through the most congested areas of current operation. The project just at Bloor is what $1 billion? Would we not need to spend similar money at St George and Eglinton among other stations in order to allow the trains to move that quickly?
Would not the combined costs of doors, station reconstruction, land acquisition, the extension to say Steeles (extra turn) and additional yard construction required to support the headway needed, be on the order of a $4+ billion project? How much more would the the cost of building the DRL cross core to Eglinton and the section of LRT from Sheppard to Eglinton on Don Mills be?. How would the overall service delivered and cost of operation compare afterwards? It seems to me that we would need to target a headway of about 80 seconds with longer trains in order to stay ahead of the curve, and provide adequate service at Bloor and south on Yonge. Also as Steve has noted would mean running a huge number of trains a large distance in order to support this.
Steve: You are not supposed to notice the inconsistencies in the TTC’s position on spending money. Pet projects good. DRL very bad. Remember, by the way, that all of this work to build up capacity on YUS is in aid of the Richmond Hill extension. If we are going to spend that much money, I would rather it be on lines that serve Toronto.
LikeLike
I always find it ridiculous when city council debates something to do with the TTC and they don’t have someone other then the councillor appointed to the board there.
Steve: I believe TTC staff were there providing background info (I was watching online and so couldn’t see all the action).
LikeLike
My basic issue with this Steve is the capacity is also required in order to support any significant increase in demand withing the 416, or from people crossing from beyond in the 905. Regardless of what is done in Scarborough, a substantial improvement in service (LRT or Subway) will increase the modal mix towards transit of core bound riders. Any changes in zoning in the area anywhere along the BDL or the Crosstown LRT will result in additional load to Yonge as well (not to mention growth along the corridor that directly feeds Yonge now). The idea of being fixated only on getting Yonge subway to Barrie (sorry but close enough) makes for a ridiculously expensive system.
If the RHC and north riders are headed to the core, Yonge is a goofy place for them to be. Far better to find someway to offer frequent service in the Richmond Hill ROW (however this is done) that is fairly express (say current GO plus maybe a stop at Eglinton) to speed them on their way there, without burdening them with dozens of stops, or the subway with being overloaded and built to handle loads that make no sense.
I know neither you nor Robert like the idea of converting Richmond Hill GO to a rapid transit application, however to me, finding a way of doing this cost effectively would be cheaper, more appropriate, and less painful, than trying to find a way of increasing Yonge capacity by 25% every decade, starting at 40k.
Based on a capacity of 40k, RHC extension projections showed Yonge overloaded south of Sheppard. Extend that a decade, with 2% growth each year. Need to get close to about 94 seconds with current trains to support 40k, so after that would need add a car and get to 90 within the decade. So about 10 years after we build it, regardless of what we do on Yonge we are either at, through or brutally close to the wall.
Do a DRL to Eglinton, frequent GO (whatever tech) in Richmond Hill, and all of a sudden you have the ability to at least hope to keep up from a technical perspective. Richmond Hill as a train every 8 minutes is something like 14k. If you could abandon the line LRT could make that 18+K. EMU if you find a way of getting other tracks to RHC – in a dedicated ROW, and well … Start with GO and keep your powder dry (especially the powder on Yonge).
LikeLiked by 1 person
Platform doors are nice to have. There is nothing stopping the TTC from painting lines on the platform once ATO is running. This way, passengers can form proper lines at each door. This way, people will not pushing each other to get on the train. The doors do not allow for more crowding. Passengers will still be only able to stand behind the yellow strip. The yellow strip area will house the doors, motors and barriers. There will be no net increase in platform capacity.
Platform doors will make the TTC a better experience. With the doors, stations can install climatization systems. This way, people will not be boiling hot or freezing cold while waiting for trains. In addition, the doors themselves can be used as an advertising platforms. If projectors are installed behind the doors, it can be a moving video wall. Even if static posters are put there, it is better than lining station walls with ads. Station wall ads has cutout for station names. With the doors, it will be one continuous advertising space.
There must be someone who can supply the ad revenue to make this project less expensive for the TTC. There is nothing wrong with seeking proposals, but using exclusively public money for this project is a bit wasteful.
Steve: Considering the pricetag, that’s a lot of advertising. The entire contract for advertising on the TTC only brings in about $26m/year.
LikeLike
Last July, Global News ran an item discussing the frequency and causes of subway delays, including a variety of interactive tools.
The top 10 reasons for delays in 2013:
“Injured / ill customer” doesn’t differentiate between a customer that has fallen ill aboard a train and pushed the emergency alarm, and a so-called “personal injury at track level”.
Steve: It is noteworthy that most of these problems would not be affected by the presence of platform doors. Indeed, given that “door problems” are the second most popular on this list, we could be adding yet another major failure point to the system. As the external review of TTC operations stated, the TTC has to get its many sources of delay under control if it has any hope of achieving the level of reliability needed to run extremely frequent service every day of the year, rain, snow, or shine.
LikeLike
From what I could tell watching on Rogers community 10 it was only a counselor on hand answering questions. I would have expected them to have someone at the podium beside the chair’s desk to answer questions directly.
Steve: Earlier in the debate there is an opportunity to question staff. I didn’t see that part and do not know if it happened “online” or if there were private conversations between Councillors and staff. I do know that at least one member was misinformed about the issue directly by TTC staff.
LikeLike
I cannot imagine how you could close Bloor-Yonge for an extended stretch without an existing DRL. I don’t see how St. George could hold all those extra peak commuters, even if some went over to Spadina to transfer.
Just curious if there’s any concern over the future capacity of St. George and/or the University stations? It seems like an awful lot of the transit expansion over the next decade or so is adding pressure there too: the TYSSE, the Crosstown, the Finch LRT, even little things like the Mississauga BRT and new bus terminal at Kipling, and increased development along Bloor West all seem likely to add to the load on the Union to St. George stretch, to varying degrees. I know the belief has always been that the University-Spadina side has capacity to spare, and it will obviously benefit from increased headways from ATC, but that capacity can’t be unlimited. Is anyone at the TTC thinking about or concerned about this, or is a similar problem as Bloor-Yonge way off still?
Steve: St. George got a passing mention, but much of the focus is on Bloor-Yonge. The TTC expects several thousand riders to shift to the University line from Yonge when the TYSSE opens. This will reduce the space available to riders at St. George southbound (also the “short turn” trains will originate further north, at Downsview, not at St. Clair West, and won’t be as empty), although it should free up space on the Yonge line for a time.
LikeLike
Just reading that makes me cringe at the thought of the years of disruption on Front St that accompanied the construction of the new platform at Union. I know any kind of station renovation is going to create some construction chaos, but just imagine how paralyzing it would be at Yonge and Bloor, or Yonge and Dundas.
LikeLike
I would think that the primary benefit isn’t about capacity or suicide prevention: it’s reliability.
To me the benefit would be reducing amount of incidents where a section of the subway shuts down completely, disrupting the transit network significantly, whether the cause is a suicide attempt, accidental fall, or garbage.
That one benefit alone seems worth it to me, since I think such incidents will increase in frequency the future as the system ridership grows.
Does it have to happen at every station all at once? Wouldn’t installing them at the busiest few stations first provide the biggest benefit?
Steve: Suicides happen at many locations on the system and not necessarily at the busy times or locations. As noted in an earlier comment, most of the reasons for delays have nothing to do with the problems PEDs are supposed to address.
What is particularly frustrating here is that suicide, an extremely serious topic, is being used as cover for a very expensive scheme when there are other things Toronto (and Queen’s Park to the extent that they fund the TTC) could be doing both to improve subway service and to address the problems of mental health care. I sympathize with the MOH’s goals, but the TTC is making claims for service improvements using PEDs that are not credible, especially in the absence of substantial spending on signals, trains and station expansion. To claim that PEDs have a positive “net benefit” when most of the overall costs are not included in the equation is dishonest, and that has to be said.
I would be more than happy to see a full study of all of the changes needed to beef up the subway system for higher capacity, but to pretend (as was claimed at Council) that we can get the equivalent of another Yonge Subway into the core area for about half a billion dollars (while avoiding construction of the DRL) is complete nonsense. The TTC has until quite recently done everything in its power to downplay the importance of the DRL, and this is only one of the more blatant examples of the bad advice they are giving the city on that file.
LikeLiked by 1 person
Re the comments by Clr. Doucette, the TTC staff did very much give her idea that it would improve through put.
Other things of note was there was 34 incidents in 2014 but most years it is in the high teens with about a third being fatal. There is no particular “magnet” station. The Vaughan extension stations are being build with provisions for doors. All other stations will have to be examined to determine whether the platforms can hold the doors as well as electrical hookup.
Glenn
Steve: Thanks for the link. I have incorporated this and comments about the Q&A into the main article.
LikeLike
What I find really unfortunate about this whole thing is that, after the Prince Edward Viaduct experience, I do not believe that the installation of PEDs will actually prevent suicides; they will only divert suicides to other locations, as the Leaside Bridge has demonstrated. Restoration of funding for mental health could do a lot more for suicide prevention, not just in Toronto but across the GTA if not all of Ontario (as health is provincial jurisdiction), than PEDs in the TTC subway system ever would. The real issue – making help available to people who need support – is not addressed by PEDs.
LikeLiked by 1 person
On the SRT, which had ATO, people spontaneously form proper lines at each door without any markers showing where to stand.
LikeLike
I worked in the construction industry all of my career, half of which was with a major contractor, and which included three subway contracts.
Construction projects are very dangerous sites. The industry and the Ministry of Labour spend mega dollars trying to deal with construction site safety.
I am indoctrinated when it comes to safety issues and I am a pain in the ass around the house.
When I walk onto a TTC subway platform I am appalled. Such dangerous conditions on a construction site would lead to ‘Stop Work Orders’ and fines.
On some subway lines, not all, but one in Kyoto, Japan there are barriers.
Steve: I do not want this to sound like a trite response, but if one were to look at the operation of a typical intersection, even one with traffic signals, through the eyes of a safety inspector, the whole thing would be shut down. The basic point with rapid transit, as with many other things, is that the mode of operation is long established and the danger level is sufficiently low as not to raise a concern. It would be a valid comment to say “build everything new to a tighter code”, but retrofits are quite another matter.
As I said earlier, if the goal is to improve safety at very busy stations, then doors are only needed at those locations. If the goal is suicide prevention, then doors are needed pretty much everywhere because there is no such thing as a “busy” station on that account.
This discussion is also taking place in the context of a city that refuses to fund accessibility improvements at a small fraction of the cost of platform doors. One must ask just what we should be spending our money on, and what we really hope to achieve.
LikeLike
Indeed, there is a double standard with respect to regular road traffic. Regular traffic is allowed to be incredibly dangerous; public transit, including all forms of buses and especially school buses, is required to be safe.
Simple example: near where I live, there is a one-way three-lane road (Erb St. in Waterloo). It has normal-width sidewalks on both sides. So one can be walking along the sidewalks, maybe with kids, while traffic goes by over the 60km/h limit literally less than a meter away. If a kid were to trip and fall the wrong way at the wrong time, instant dead kid (or adult, for that matter); same if a vehicle mounts the curb at the wrong time for any reason. This is considered normal, and as a result I minimize my use of that stretch of sidewalk, exercise paranoid caution while doing so, and explain to the kids before we do it that we need to do something dangerous.
Contrast that with the situation that exists on the Waterloo Spur where a hiking trail is to be installed later this year. This line sees two trains a day, one in each direction, going at a jogging speed at best; once our LRT opens they will only operate between 23:00 and 05:00. Safety supposedly requires that a 6m space be reserved for the trains; outside of this there will be a 1m drainage area, then the 3m wide path.
Or consider the LRT itself. People have expressed worry about what happens if it hits people. There have been suggestions that it should be fenced off from the park and other areas through which it will travel. Yet it will stay on its track, come once every several minutes, and be driven by professionals. If we take the road situation as acceptable, there is really no reason at all to insist on any more separation than exists between the Erb St. traffic and sidewalks.
Of course, we should cure the double standard mostly by adjusting the road network, not by going back to allowing everything to be dangerous, but still it seems like there is a fair bit of room for transit to accept some manageable risk of injury.
LikeLike
Same thing happens on the GO train which usually stop at almost the same location all the time.
LikeLike
@Isaac Morland – reminds me of working at DuPont when I was a kid. The company was extremely safety conscious, and ended up doing a big push to fix safety for workers at home and their most dangerous workplace – head office. Point being people tend to be contemptuous of risks they are used to, and as such filing cabinet drawers can be more dangerous than heavy equipment.
LikeLike
Steve quick question with regards to the TTC presentation. They talk about 28k peak 2031 ridership based on current south of Bloor, does this not mean they (based on actual current) greatly underestimated ridership growth?
If we did jump to a 105-110 second headway, and added enough buses to address the peak overcrowding issues would not at least 4k additional peak riders suddenly appear? If we restore service levels on the surface network, and provide capacity,would not additional capacity through 35k be used in a handful of years, 40K by the 2031 time frame?
I may be misguided, but I was under the impression that part of the provincial idea was to try increase density inside the existing built up area. Should not part of the plan to be increasing density within the 416, in areas that are or can (LRT) be reasonably be served by rapid transit? Would this not mean a substantial continued growth of within 416 ridership?
Steve: The history of TTC ridership and capacity estimates is closely linked to the Richmond Hill subway for which, at one time, there was a “fifth column” working inside the TTC. During the late 80s, the subway was bursting with riders, and that’s where some of the early plans to add platforms at Bloor-Yonge originated. There had also been talk of a loop subway to eliminate the terminal delays associated with turnbacks and to allow a shorter headway on the line.
Then came the recession of the early 1990s. Riding dropped by 20% and suddenly there was lots of room on the subway for growth, and pressure to look at a DRL fell off. The Richmond Hill subway advocates regarded the DRL as a mortal threat because it would soak up a lot of money and possibly pre-empty extension of the Yonge line further north. The TTC made some quite outrageous claims about the potential capacity of the line with various technological changes even to the point of getting to 90 second headways. Combined with larger trains and the diversion of some riding to the Spadina branch, this would give a huge increase (over 60%) in Yonge line capacity thereby eliminating “the need” for a DRL and guaranteeing that a Richmond Hill extension could be accommodated.
As things worked out, a lot of the then-surplus capacity was eaten up by growth and as we know today the demand is capacity constrained. It does not matter how many more buses pull into Finch, there is no place for riders to get onto the trains. Moreover, some degree of common sense (but not enough) filtered into the TTC and they no longer make as large a claim for increased capacity as they once did. A simple example is that the extra 10% or so given by the TRs used to be included as “future capacity” but it’s now in operation and has been consumed by the backlog of demand. Talk of a 7th car has been put off until at least the next order of trains in the 2020s (new trains would go to YUS with the TRs switching to BD).
The TTC no longer talks about 90 second headways, and the lower bound they consider reasonable now sits at about 105. Even that has challenges for terminal operations that will require much better crewing practices that are today the source of random delays. On a 141 second headway, small delays can be absorbed, but if the line gets down to the absolute minimum, there will be no wiggle room. Frankly I don’t think the TTC can pull that off given their inability to manage service.
What this all comes back to is that it is much more practical to build a new line into downtown to provide parallel capacity and divert traffic away from the main interchanges. There will still likely be a problem again in two or three decades, but I don’t think any of us really know what the travel demand patterns will look like. Just to bring the 905 up to the modal split we see in Toronto will require a huge increase in transit service and usage. The province’s $29 billion is just playing catchup with what we have ignored for years, and shows little sign of addressing demands that are not focused on downtown Toronto.
LikeLike
To play devil’s advocate, the big issue the with PED is not the deaths or potential injuries, but the inconvenience to everyone else on the system. To be fair, either these people are intent and going to try anyway at a different location or with a different means or it is a passing matter and better platform security is sufficient. Specifically, if they are intent on death by subway, there are several open air sections, such as Belmont St. between Bloor and Rosedale.
If it’s about health and safety, even if every track-level event were a fatality, this represents 0.2% of all deaths in Toronto per year. To put it in perspective, 34 deaths is 2.5% of the annual mortality rate amongst those under 20, 6.9% of the deaths caused by flu and pneumonia, or 17.4% of infant mortalities. $23.3M per year for 50 years can save a lot more lives in other areas. Specifically, Toronto has a number of neighbourhoods with higher than average premature mortality rates: Moss Park, New Toronto, Birchcliffe-Cliffside, Woodbine Corridor, East End-Danforth, and Church-Yonge Corridor.
If it’s about convenience/reliability, there are a lot of other sources that can be improved first cheaper and with a bigger impact.
If it’s about cleanliness, garbage and litter will still get on the tracks from other points in the system. $23.3M per year can buy a lot of cleaning.
If it’s about station comfort, this is $0.40 per trip without any costs for climate control, which would be another order of magnitude for all the ventilation and energy use. On top of that, unless your trip is only from station to station, you’ll be prepared for the elements, whether hot or cold. If it’s just to warm the stations up, $23.3M buys a lot of space heaters.
If it’s about headways/speed, then it’s a waste of money because it won’t change how fast a train can decelerate.
If it’s about driverless trains, then $23.3M per year pays for 100 drivers earning $111.88 per hour.
Did I miss any of the supposed reasons? All in all, it’s a good idea with a strong sentimental component, but with a price tag well out of proportion with the effect.
LikeLiked by 1 person
It was always my understanding that ATC would facilitate a maximum deceleration curve for a particular station (which takes into account station grades and passenger comfort), that is repeated every time. Since a driver cannot offer such consistency, his/her curve will differ every time, which he/she must allow for by starting to decelerate earlier and thus take more time.
I am persuaded that PED’s principal benefit has to do with reliability, and should be selectively installed to address evident reliability concerns. Suicide is a health issue and should be approached with a health budget. I must admit though that the climate control benefits offered by full height doors are attractive.
In case your readers are not aware, PEDs are in use with the airport people mover and soon on the UPX, end point stations.
Steve: Actually, the braking points for trains are clearly marked on the tunnel walls and with signs in the open cut areas, and operators quickly learn that they can actually “push” this because the signs are for a more conservative style of operation than regular crews practice. If anything, ATO may slow down the trains by operating them in the “official” way as opposed to the way operators know will get them fastest from point A to B.
If you recall from the Russell Hill crash, the junior operator had learned how to drive through a long section within timing signals in a way that the actual change from red to amber was rarely visible, but happened just as the train reached each signal. Therefore, the fact he had run a red that didn’t clear, but where the train stop device was not working, wasn’t evident until he came upon the next train stopped around a blind curve. Anyone who rides the subway knows that this is how it is actually operated. To claim that somehow ATO will speed things up really is not credible and sounds more like marketing than fact.
LikeLiked by 1 person
Which is a big issue for me. We can build a system that is not just core focused. When we start to use limited funds to extend subway where LRT is a more appropriate means, you end up with much more compromise than required. Most “world class cities” have found a way to get the public to adopt LRT as the best means for most new lines where demand will be below 15k. Toronto region as a whole really need 100s of km of rapid transit lines, however, only a very small portion of it can be subway. If we actually ever want to create a region wide functional transit system, it needs to be capacity appropriate. Should be looking to build BRT and going to LRT where and when demand justifies, and subway only where it is really required.
LikeLike
People on the Scarborough RT mostly line up at the door position. There are so many people standing where the doors will never line up. They do this to beat the line. If the lines are present on the platform, people who cheat the line will stand out. Platform doors should encourage better manners. It will be an expensive way of doing it.
Climatization of stations is a benefit. One of the benefits of automobiles are that it is climate controlled from beginning to end. Imagine arriving to a meeting in a heavy coat or sweaty. Most downtown stations are connected to the PATH. So having climatization systems will allow people to wear suits in the summer and no jackets in the winter.
Steve: Be careful what you expect to see. The TTC was not looking at floor to ceiling enclosures, and many of the samples people have been mentioning on social media would leave stations like Rosedale just as drafty as they are today.
LikeLike
I don’t know that much about ATO but I assume the programming would include ‘regular’ operations and ‘catch up’ operations much like on the Vancouver Skytrain with the ‘regular’ operation using conservative driving parameters and the ‘catch up’ operations using the system maximums for speed…So I would not expect regular service to be much faster….but it would be more reliable after delays…and I would expect less wear and tear on the trains.
Obviously ATO will result in the ability to decrease headways. Scheduled headways in Vancouver are 108 seconds in the peaks without significant issues. In theory the Expo line could be run at 90 seconds (but would likely become unreliable at this level due to random delays). If the Expo and Millenium lines were separated (no switch at Columbia) it would technically be able to have 75 second headways so 90 seconds would probably still be reliable. I am not familiar enough with the tracks and switches in Toronto, are there tracks and switches that would still limit headways to greater than 105 seconds with ATO? The Vancouver headways are driven by the switch to the Millenium line and Waterfront station, both of which could be fixed (granted with lots of money…although I would expect the Columbia switch issue would probably be addressed if Vancouver ever builds the Broadway extension to the Millenium line).
Steve: The constraint on headway comes mainly from the track layout at terminals and the length of the trains. There are limits on acceleration that any train can use without throwing passengers onto the floor, as well as maximum spees through curves and switches. It is physically impossible to dispatch a 90 second headway from the TTC’s terminals today because of the cycle time of trains moving through the crossovers. The TTC raised this itself years ago in a study of subway capacity. I have timed trains at terminals to see how long each move takes and it would be tight once you get below 120 seconds, never mind down to 90. ATO can give you ideal operations, but it cannot get past the laws of physics.
LikeLike
Oops, the previous post does not mean to come across as saying ATO and screen doors are a better transit investment than a DRL. I do not know enough about either to suggest that, but ATO should be able to significantly improve capacity and reliability. More I think than is being implied by TTC. Is there a reason for this? Are they assuming they still need an operator and he/she needs to switch ends?
Steve: The operator is not an issue. There are ways of handling crews at terminals to avoid changeover delay. The main benefit from ATO comes from allowing trains to run close to each other. At a very busy station with a long dwell time, the “next” train can be waiting just outside of the station rather than two signal blocks away. TTC actually did operate this way once upon a time until they changed operating rules to make single red signals a “stop and stay” rather than a “stop and proceed” indication. I quite clearly remember trains leaving Bloor Station with the next train following into the station half a train length behind.
LikeLike
Having said that Steve, assuming that you actually ran the equivalent of a loop at the end of line (unlimited turn capacity). Would not achieving 90 second headway or less still be very difficult at Bloor on Yonge? This simply due to dwell time for alighting and boarding? To achieve that kind of headway in real life conditions with the volumes of riders at this location, would you not require 3 platforms so passengers could board and alight simultaneously?
Steve: Probably. The TTC had a scheme for rebuilding the station with a centre platform, but it was extremely complex and would require closing (!) the station for an extended period during construction. It is no longer an active proposition for obvious reasons.
It becomes tiring hearing about 90 second headways because these usually refer to operations where trains do not actually stop, but share trackage between junctions that are very tightly scheduled. Station dwell times, especially with their variability, make a very short headway quite difficult to sustain, and it’s the sustained headway that gives the high hourly capacity.
LikeLike
Everyone seems to agree that at a cost of $10M per station seems ridiculously expensive. We should pick a lessor used station (Maybe Castle Frank or Bessarian) and see if we can do it for half a million. Open it up to an RFP where proponents are free to propose solutions. If something goes wrong, it won’t be at a very busy station and it can be removed. If it works, it can be installed to other stations.
Steve: “Everyone”? You seem to have access to a larger sample of informed engineering opinion that I have seen talking about this issue recently.
LikeLike
…but just imagine how paralyzing [major construction] would be at Yonge and Bloor, or Yonge and Dundas.
Major construction actually happened several years ago with expansion of the Bloor platform. Just how paralyzing was that construction period?
Steve: The construction took advantage of other work in the area that allowed the subway structure to be exposed. The northern part of the station is physically inside of the Hudson’s Bay building, and very close to other buildings further north. Also, that section of Bloor was originally wider because of the old streetcar-subway transfer platforms in the middle of the street. This gave extra staging room for the construction diversion.
At Yonge and Dundas, the station is directly under Yonge Street. A second entrance was proposed at Gould Street as part of the Ryerson construction now in progress, but nobody wanted to pay for it.
LikeLike
There are a number of factors that affect minimum headway; these are:
1. Train length: shorter trains can clear a block faster and run closer together. TTC subways are 135m and Skytrain is between 68 to 76m depending on train type.
2. Station dwell time: TTC cars hold a lot more passengers than Skytrain per unit length because they are about 205 [mm] wider.
3. Block length of signal system: ATO can let trains get closer together because it keeps trains separated by a combination of speed and distance. If the trains are going slower they can close up. Fixed block signals keep them 2 blocks apart regardless of speed, except at terminal stations where there are special signals to allow them to close up. According to Transportation Research Board’s TCRP report 100 “Transportation Capacity and Quality of Service Manual – 2nd edition” the use of ATO will only get a capacity increase of between 2 – 4%.
4. Maximum station dwell time and the time for the passengers to clear the station: this is determined by the busiest station on the line, Bloor Yonge for the TTC.
5. Terminal time: This can be decreased by using a loop or by using multiple turn backs, i.e. short turn service.
Actual passenger capacity per hour is the product of the number of passengers per train times the number of trains per hour. Using manufacturer’s data for Skytrain a Mark III train can hold 580 passenger while a TTC train can hold 1,000. So if Skytrain can run at a 90 second headway this would give 40 trains per hour times 580 passenger per train for a capacity of 23,200 pphpd. The TTC’s service at 120 second headway gives 30 trains per hour for a capacity of 30,000 pphpd. I have doubts about the accuracy of the value of 580 passenger for a Mark III train.
Steve: Also, the service design capacity for a TR train is 1,080 which is below the crush load of 200+/car. Service is never planned to be at crush for a sustained period (such as the full AM peak period) because at those levels dwell times go through the roof.
Designing the system becomes an optimization problem. Increasing train length increases train capacity but reduces the number of trains that can be run per hour. The relationship is not linear in that increasing train length does not necessarily cut the number of trains that can be run in half. The TTC cannot run headways much under 120 seconds because Bloor-Yonge would not clear between trains.
LikeLike
Transport for London sees Platform edge doors as essential in it’s Deep Tube project where it plans to go up to 36 trains an hour. It is also seen as prerequisite to going driverless.
Preventing people under trains is seen as a nice add on, not a primary reason to look at PED’s.
The first line to be upgraded is the Piccadilly line.
Lots of interesting articles and discussion on London Reconnections.
http://www.londonreconnections.com/2014/getting-radical-piccadilly-line-upgrade-2/
http://www.londonreconnections.com/2014/new-tube-london-driverless-train-driver/
http://www.londonreconnections.com/2014/driverless-trains-piccadilly/
Steve: Yes there’s lots of interesting stuff there, but an important point is that the PEDs are not a pre-requisite for 36 trains/hour, simply another aspect in the redesign of the lines. The problem we face in Toronto is that there are many possible improvements in operations, design and maintenance, and we must avoid the artificial creation of a link (particularly one that has never been articulated before) that would radically increase the challenges in going to closer headways. At the same time, forgetting to include station capacity as a key design feature suggests that PEDs are being oversold when they are only one part of a more complex system.
LikeLike
In the 2014 loading standards document, there’s a footnote specifying that “standards for [the TRs] are subject to confirmation after in-service experience.” Now that the Y-U-S is pretty much fully changed over to TRs, has the TTC ever confirmed whether they’ve resulted in the oft-touted 10% increase in capacity? (Right off the bat, the loading standard only went up by 8%.)
Steve: The TTC seems to be sticking with that number. One problem is that when they talk about increased capacity on YUS, that is one component of a cumulative increase, but it has already been eaten up by latent demand and cannot be counted as a “future” improvement. The next bump will come from diverting some ridership to Spadina-University when the TYSSE opens in a few years.
LikeLike
Steve, I have to admit, I have trouble with this being on the list of projects. There is substantial doubt in my mind whether is the city was handed $30 billion to do extensive repairs and build new transit lines, whether it should make the list then, so why should it be listed as a below the line project, that seems like it is a priority to the city.
Balance of transit city, DRL east, re-signaling Bloor-Danforth, new trains on BDL, Waterfront West LRT, East Bay front LRT, even DRL West side, Kipling and Markham road BRTs, extensions to the streetcar network, Finch East LRT … would all likely come before this in terms of actually making a substantive improvement to the quality of transit, and improving access. Having this on a list, actually makes it seem like there is nothing more important. Would the TTC not be better off just pulling this off a list entirely, or at least make it a project for only a handful of stations where the crowding is likely to get dangerous?
Steve: The question is just what PEDs are supposed to achieve. If it is safety and crowd control, then yes only the busy stations need them. If it is suicide prevention, then far more (all?) require them because there is no “magnet” station for suicides. Similarly, if the goal is fire prevention (from blowing newspapers and other debris), then there are far more candidate stations, and some of them have debris blown in from outside (e.g. Victoria Park) where PEDs won’t do any good. If someone can answer the question about goals, then we can talk about budget.
Meanwhile, it would be nice to get on with accessibility, a project with about 1/10 of the cost.
LikeLike
Amen Steve, to the extent that they are about deterring suicide, we are really talking about avoiding delay on the subway line. People will simply choose instead to jump off the bridge over the DVP into traffic instead, it does not inconvenience as many people but well it will do.
I do not really buy the argument about garbage, as there are many points not in stations exposed to this sort of this as well that are wide open. Does not debris mess up the electrical system where it is in the open? So in the end, I cannot see it really being effective in achieving much beyond safety at the most crowded stations. Accessibility is really important, especially given the large number of people who want to travel off peak, and who are now getting to a point in their lives where their mobility is starting to be reduced. We do not want all of these who do not want to drive (especially if it is because they are nervous) being forced to, not partaking in society, or having to call Wheel Trans.
We will lose much as many still have much to give socially, and not doing so will mean their lives will be filled with other issues fairly quickly which will also cost society dearly.
LikeLike
I was aware of most of your points, but had not considered point 1 (train length, makes sense though). About point 3, isn’t the system being upgraded to a moving block system with the upgrades in Toronto?
Steve: Yes, there will be moving blocks with the new signal system, but this is of real value only in congested portions of the line by allowing trains to move physically closer together and, thereby, to reduce the cycle time at a busy line station such as Bloor southbound in the AM peak. However, the congestion at terminals remains because this is due to physical track layout: length of trains and crossovers, speed controls through curves, speed controls in stations. (For an example of the latter, if there is a train in the storage track beyond, say, Kennedy, the approach speed of an incoming train is forced to be much slower than under “normal” conditions, and this compromises the headway that can be operated from the terminal.)
What I don’t know is the track/switch layout in Toronto and whether these are the bottlenecks to improving headways or if other factors like platform crowding are the bottlenecks. So if the terminal station track layout limits headways can it be fixed easily? Couldn’t you also improve the terminal configuration just by increasing the length of the tail track and temporarily storing a train on it ready to go (instead of a loop)? If so how much will it cost?
Steve: You require more than a tail track, but storage areas beyond the station so that the turnarounds are, in effect, done after the station, and the “terminal” operates as a line station. An alternative, workable when the demand is not very high at the outer terminal, is to have a short turn point where half of the service turns back. On an extended YUS, this would be at Finch (or maybe Steeles) for a Richmond Hill service, and at Downsview for the Vaughan service. Building a short extension simply to provide this would be expensive because one would have to mobilize for a very piece of three-track tunnel rather than building it as part of a larger project.
If the problem is station passenger flow increasing platform capacity will be expensive, they are starting to do that on the Vancouver stations now as part of a plan to eventually run 5 car Mark III trains on the Expo line (86m length I think), for the Commercial station they will add another platform with doors opening on the disembarking side first. I doubt the problem would be with the number of passengers per train as long as the station could handle the people. I would assume the nice shiny new TR trains have a better door configuration that the old Mark Is in Vancouver and that there would be better access to doors per passenger on the TRs so the problem will be getting people on/off the platforms.
Steve: Yes, the problem in Toronto is platform and station circulation capacity, not train loading/unloading.
LikeLike
Of course the next question is how much closer would this actually bring the trains even if we did cut the boarding and alighting time in half? I do not see us getting much below 105 based on Bloor on Yonge now, however, even with 3 platforms at Bloor and St George, and say Sheppard and Eglinton would we be able to get past say 95 or 90 seconds? If you assume you can save 15 seconds or more to the point add say 6 trains per hour, is that really worth what it would cost to avoid building what 8 km of subway? Just as important you are talking about from say 37k to 43k, even at 43k does that address your long terms problems (assuming that is you can actually achieve 90 second sustained headway by making all your busiest stations 3 platform. After this going to 7 cars really only means the trains are 10% longer – yes you added an entire car, but each one is shorter so you might achieve 47k.
I cannot help but wonder when the TTC talks about capacity projects if there is not an engineering drive there to simply squeeze the line for capacity, to see what can be had, not because it makes sense. The idea that is should flow to Yonge stops making sense for bus routes, once the area is already sufficiently congested. The development around a lot of Yonge already means it would make a lot more sense to try and find an alternate destination for these routes. If we are going to spend billions to redevelop Yonge, it would likely be no more expensive (especially after disruption is factored in), and much more effective to build parallel capacity, instead of doubling down where we are. The big bang on a line with new signals would come from extending one stop and increasing turn capacity would it not (25-27 trains to 32-34 trains) ? After that does not every tweak for capacity not get increasingly expensive and uncertain? Does it not just end up being more expensive to build up capacity on Yonge? Some part of me loves the idea of 3 platform stations, and all the associated flow dynamics, but then I think of what would be required to get there, and the impact on surface transit over time and well…
LikeLike
That is a better way of phrasing what I was thinking. I was thinking a configuration like this would be much cheaper than station redesign … if it was the problem rather than platform passenger flows … which I guess it is not …
LikeLike
Most of the lines with ATO that I have ridden on usually slow almost to a stop a couple of seconds before they stop and then for the last couple of seconds travel slowly to the correct stopping point. This allows for problems such as increased mass, dirty rail or worn brakes that can throw off the deceleration curve. PEDs with human operator have a much longer slow time to ensure that the train doors are opposite the PEDs.
LikeLike