[mdrejhon]

The Great Platform Height Debate (converting train stations to high platform)

Over the next ten to fifteen years, we will have lots of trains with various platform height requirements, running over the Metrolinx rail:
- GO Trains (low only, with center car high platform)
- VIA (low and high)
- GO RER (unknown)
- SmartTrack (if using different trainsets than GO RER)
- High Speed Rail (usually high only)

The biggest impact may be the GO RER / SmartTrack initiatives, some which MAY be high platform (in short: they're still deciding), and the potential need for station sharing with GO stations. Metrolinx has already proposed diversifying their trains during the upcoming GO RER boom. This will be huge challenge in station-sharing, in the coming years. Inevitably, we may need certain stations modified, or transitioned (or even all of them).

The question is no longer "if", but "when". In fact, it already happened. We already had to build a UPX station at Union that is high platform. We had to duplicate some stations for UPX. But this is nothing compared to the future explosion of train diversity. Europe had to go through this headache, as well as California with their upcoming HSR.

Various methods that are used or proposed:
- Different heights of platforms on either side of trains (see example photo below)
- Separate stations for high vs low platform trains (UPX method)
- Transitioning the trains and/or platforms (including long term slow transitions)
- Different levels of sinking of trackage within the station (tracks for high platform trains are sunk lower in the station, to keep platform at the same level as for the low platform trains in other tracks)

One example of the European solution is pictured below.
High speed train at left (high platform train in sunken track), and commuter train at right (low platform train on raised track).

Photo by Mundus Gregorius from flickr
Source: http://sf.streetsblog.org/2014/10/07/caltrain-and-high-speed-rail-to-pursue-shared-level-platforms/

But not many stations have enough land space to allow both platform heights in the same locations. So tough platform choices has to be made. California is going through this debate, with lots of interesting diagrams that are relevant to discussion. Because California are using the same bilevel trains as we are!

Source:
http://caltrain-hsr.blogspot.ca/2014/07/the-path-to-level-boarding.html

Source:
http://caltrain-hsr.blogspot.ca/2014/07/the-path-to-level-boarding.html

But that can be insanely expensive, so some systems have proposed trainsets with doors at both heights, for supporting both kinds of stations, high and low. It also permits long (50-year) transitions in station platform heights, by living with trainsets that support both. Example:

Source: http://caltrain-hsr.blogspot.ca/2014/10/level-boarding-plan-b.html

Knowledge of precedent is useful. Lots of bilevel EMUs in Europe were low level, but some of them have high level doors in theirs, so bilevel EMUs exists for both low and high platforms. So it is theoretically possible

Moscow's airport train. Peris bilevel RER is high platform too.
Source: http://caltrain-hsr.blogspot.ca/2014/11/the-train-that-shouldnt-exist.html

It is worth mentioning our GOtrains are modifiable to an extent, if we needed such a transition over time but for 25" platforms and not the UPX height of four feet.

Source: http://caltrain-hsr.blogspot.ca/2014/09/compatibility-done-backwards.html

In less than 15 years (or even 15 months!), something (HSR, RER or SmartTrack) is going to force expanded discussion about this. It possibly is not a question of "if" but "when" the big platform height debate goes public within Toronto.

 

[rbt]

While I'm strongly in favour of high platforms as dual-level cars are significantly more efficient with the doors on the mid-level; this may be very expensive.

All elevators and escalators would require significant adjustment. It's possible to build ramps down to these areas but then you create a collection area for snow/rain exactly where you don't want it.

Discussions around adjusting Sheppard platform levels don't have weather to contend with; GO does.

 

[mdrejhon]

This be very true. But we've got a few monkey wrenches being thrown at us:
- UPX (high platforms)
- SmartTrack (proposals seem to suggest high-level platforms)
- HSR (traditionally high platforms)

Most GO stations do not have escalators, although some do. Elevators are definitely going to be a challenge, however. Shaft extension would probably be more feasible than trying to create ramps towards elevators. It may even be a station-by-station case, where some stations are split-level (low on one side, high on the other side) with ramps/stairs between the heights.

Another way to think of this is:
In fifteen or twenty years, will it be even more problematic to make sure all trains are all always permanently low-platform compatible?

We may be forced to choose less desirable HSR trainsets and MMUs, or expensive separate stations, because of our massive investment in low platforms. Is there room for hybridization -- one line becoming converted to high platforms? We might have cemented our investment in low platforms for a long time, but we might end up revisiting the question. And, of course, SmartTrack is currently a big wildcard, but if SmartTrack is high platform as it is suggested to be -- and it plans to share/interchange with several existing GO stations -- we've got a powderkeg of a debate waiting in the wings.

 

[gweed123]

Indications seem to be that GO RER will be using shorter trains than the current 12-car bi-levels. If that's the case, it could potentially be possible to, as an interim solution, retrofit half of a GO platform to serve high floor vehicles, and keep the other half at the current height. Of course, this would necessitate reducing the length of the non-electrified (current diesel fleet) GO trains.

Ex: west end of the platform is for GO RER trains, east end of the platform is for GO Regional (diesel) trains.

Once the entire fleet is electrified over the course of a decade or two, the other half of the station can be retrofitted for higher platforms as well. That should correspond in a lot of cases to a natural upgrade cycle anyway.

 

[mdrejhon]

Some ideas:

-- They could also lengthen the platforms to allow berthing of either, but they are already almost a third of a kilometer long (300 meters!). You stand in the wrong location, it takes a few minutes to walk to the opposite end of the platform.

-- For the "SmartTrack" GO RER route, one idea is they may even discontinue local stops for Kitchener/Stoufville trains (going express to Union in the "SmartTrack" GO RER segment), allowing full conversions of the local GO stations to high platform. The "SmartTrack" GO RER would replace all local diesel GOTrain service.

I would consider 150 meter to be the bare minimum length practical for GO RER in these corridors -- regardless of high or low platform. If they go with 150-meter SmartTrack trains (~7.5-min peak, 15-min offpeak), proper selection of the trainset (e.g. electric Paris RER double decker EMUs) can still push more people down the corridor than the 300-meter GO trains which don't run as frequently. Single deckers may even work at this length; simple sheer frequency (150m single-decker every 5-minutes during peak period) can push more people down the corridor than an hourly 300meter bilevel GOtrain. Georgetown corridor should be able to sustain that during peak period, though the Stoufville corridor will require major upgrades to do so. Sheer service frequency will be critical to the success of SmartTrack / GO RER.

 

[44 North]

I would consider 150 meter to be the bare minimum length practical for GO RER in these corridors -- regardless of high or low platform.

Is this an overall ballparked consensus, to have 150m platforms? And would anyone happen to know the rules concerning new stations and safety/accessibility? For elevated stations, will there have to be two elevators on either side of the platform? What about a station attendant?

 

[mdrejhon]

Is this an overall ballparked consensus, to have 150m platforms?

Not necessarily, but this is simple capacity math.

If we're going to replace huge infrequent 300 meter bilevel GOtrains with smaller frequent 150 meter single-level trains, we may be pushing only ~1200 people max per train, rather than ~4000 people max per train. Train frequency would need to go up 3-4x to compensate (15min instead of hourly). Shorter trains may be extremely difficult to replace GOtrain capacity, since, SmartTrack consuming GOtrain corridor passenger-movement capacity. UPX-length trains will not be a practical passenger-capacity replacement for GOtrain traffic.

There's some give there, e.g. 100 meter bilevel electrics, etc. Some countries now have "bilevel train subways" (I found at least two, one European, and one Australian). Mainly commuter routes with underground stations in the urban segment. All of them are high-floor and accessible. SmartTrack could concievably use these, and should, if we choose very short trainsets (e.g. 100-150 meter) with a level boarding system. Replacing most local-section GO service with SmartTrack is an excellent idea, and possibly turn Kitchener/Stoufville diesels into express gotrains that only stops at select stations and beyond the SmartTrack endpoints.

But, 150 meter-ish, not a hard-exact number, is right in the minimum ballpark of being able to maintain current existing "GOtrain peak passenger throughput" (compensating lack of per-train capacity with extra frequency, can only go so far).

And would anyone happen to know the rules concerning new stations and safety/accessibility? For elevated stations, will there have to be two elevators on either side of the platform? What about a station attendant?

They will have to have this conversation.
For accessibility compliance, you'd need an arrangement such as:
(1) hire extra staff/attendants at every train or station to lower a bridge to the existing raised center platform; or
(2) electronically lower a ramp from the train whenever requested.
(3) brand new stations with level platforms to doors (new platform height of whole train length).

We could also end up using low-level electric GOtrains, like this electric bilevel with retractable steps [wikipedia], but only certain platform heights will be accessible to wheelchairs for this train.

Otherwise, SmartTrack is going to be inaccessible for wheelchairs.

This is yet another reason for the Great Platform Debate for another reason: Accessibility!

 

[reaperexpress]

If we're going to replace huge infrequent 300 meter bilevel GOtrains with smaller frequent 150 meter single-level trains, we may be pushing only ~1200 people max per train, rather than ~4000 people max per train. Train frequency would need to go up 3-4x to compensate (15min instead of hourly). Shorter trains may be extremely difficult to replace GOtrain capacity, since, SmartTrack consuming GOtrain corridor passenger-movement capacity. UPX-length trains will not be a practical passenger-capacity replacement for GOtrain traffic.

There's some give there, e.g. 100 meter bilevel electrics, etc. Some countries now have "bilevel train subways" (I found at least two, one European, and one Australian). Mainly commuter routes with underground stations in the urban segment. All of them are high-floor and accessible. SmartTrack could concievably use these, and should, if we choose very short trainsets (e.g. 100-150 meter) with a level boarding system. Replacing most local-section GO service with SmartTrack is an excellent idea, and possibly turn Kitchener/Stoufville diesels into express gotrains that only stops at select stations and beyond the SmartTrack endpoints.

I had to do these quick calculations to make sense of this discussion:

300 metres is 984 feet, or eleven 85-foot (standard) train cars. I'm guessing we're actually talking about 311 metres, which is 12 standard cars.
150 metres (or rather, 155 metres) would be a six-car train.

A 12-car Bombardier Bilevel train has 1800 seats, so a practical capacity of around 2400 people. At crush load it could theoretically carry 5000 people, but the narrow aisles would preclude movement and people would not be able to get out at their stop. A six-car bi-level train would carry around 1200 people.

A 6-car Toronto Rocket subway consist (75' cars) can practically carry around 1000 people. My guess is that a single-deck suburban train would carry about the same amount - the extra 10' of space per car being used up by a higher percentage of people seated rather than standing.

So if we figure a track can support up to 10 scheduled trains per hour, we're talking 12 000 people per hour per direction per track with a 155 metre platform, or 24 000 people per hour per direction per track with a 310 metre platform.

For reference, the current scheduled peak-direction capacity of GO lines ranges from 4800 to 16 800 pph (all tracks combined, and assuming entirely 12-car consists).

So my quick back-of-the-envelope verdict would be that 6-car platforms just barely cut it. We'd be talking around 24 000 pphpd for a quad-tracked line (i.e. Lakeshore and Kitchener), which is not that much above where we are now. On the other hand, 12-car platforms seems quite excessive. I think the sweet spot is around 8 to 10 cars.

 

[UD2]

Bombardier's OMNEO double decker EMU trains used in France has a capacity of 1380 at 135 meters. It uses low platform doors and are fully accessible at the low platform level. Design speed 140 - 200 km/h

Therefore at 310M platform length and 10 trains per hour we are looking over 27K pphpd.

And we can keep our hard working people of Thunderbay working for some years to come.

Sounds like our solution.

 

[reaperexpress]

To actually weigh in on the Great Platform Debate:

I am definitely in the high-platform camp for the following 4 reasons:

1. Dwell Time (A.K.A. Speed and Capacity)
GO's dwell time is frustratingly long, because of the accessibility ramp procedure. Level boarding would eliminate the need for an accessibility ramp, and also increase the flow rate through the other doors.

But most importantly, we could add more doors by having single-deck trains. I acknowledge that the bilevel Paris RER example above has 3 doors per car, but that arrangement is really space inefficient due to the number of stairs required. At that point there's not much capacity advantage over a single-deck train anyway.

VIA's current loading procedure is horrendously inefficient because they use high-floor vehicles at low platforms. Staff have to open each door individually and put out a step stool, and often they don't even open all the doors. Then they have to help people into the train (elderly, disabled, luggage, etc), which inevitably takes a while.

VIA is not going low-floor anytime soon, so the only solution is high platforms.

2. Accessibility
Look at how small the platform gap is at Pearson Terminal 1 Station!

Image by Metrolinx

Instead of an Accessibility Car we could have an entirely accessible train! Level boarding benefits more than just wheelchairs, but other people don't necessarily go to the accessibility car. I have experienced both myself and others struggling to accommodate strollers and/or luggage in non-accessibility cars.

3. Flexibility
The accessibility car "locks" trains into a specific location on the platform because it is always a 5(?) cars behind the locomotive. So when we run smaller trainsets such as the 6-car sets on the Barrie Line, the train ends up stopping all the way at one end of the platform, which is not necessarily the end where the exits are.

4. Sense of Quality
I'm not sure exactly why, but high-platform stations seem so much more proper. I've heard this from non transit-interested people too.

Implementation
Not all lines necessarily need to be high-platform (Richmond Hill comes to mind), but I think it's worth moving in that direction. With UP Express introducing high-level platforms to Union, Bloor, Weston and Pearson stations, the dual-height scenario has already begun.

The Kitchener line seems like a natural candidate for high-platform conversion given that so many of the stations will be extensively rebuilt or replaced in the relatively near future, and some stations already have high-level platforms. The line is also shared with VIA, which would really benefit from high-level platforms.

A first step could be to build mini high platforms (like at Weston and Bloor) at Etobicoke North, Bramalea, Brampton and Mount Pleasant, and operate the existing 3-car DMUs on the local service to Brampton when the UPX gets its EMUs.

To accommodate the progressive rollout of full-length high level platforms, we should switch from buying Bombardier Bilevels to a model compatible with both high- and low-level platforms, such as the Bombardier Multilevel:

Image by Bombardier

The end doors can serve either high-level or low level platforms thanks to a retractable trap, and the middle doors serve only high-platforms.

The new AMT Mascouche Line in Montreal operates MultiLevels and has both high-level and low-level platforms. In addition to Gare Centrale, the new Mascouche, Terrebonne and Repentigny stations have high-level platforms.

Image by Nicolas Houde on Flickr

 

[mdrejhon]

So my quick back-of-the-envelope verdict would be that 6-car platforms just barely cut it. We'd be talking around 24 000 pphpd for a quad-tracked line (i.e. Lakeshore and Kitchener), which is not that much above where we are now. On the other hand, 12-car platforms seems quite excessive. I think the sweet spot is around 8 to 10 cars.

We will still be running the diesel 12 car trains during peak for the surge traffic, that will give relief to the 6-car to 8-cae situation.

Some 8-car trains are the same length as 6-car trains using longer cars; GO uses relatively long cars. Either way, those train lengths hover approximately 150-meterish, so our napkin calculations seem to be on the money for a minimum SmartTrack length capable of taking over GO service.

Good additions to this thread! And thanks for a photo of a real dual-platform-height train!
I was not aware Montreal now operates dual-platform-height bilevel trains.

 

[mdrejhon]

Expanding on reaperexpress revealation that the new 2014 Montreal commuter train & train route is dual-platform-height.
So I dug up images for this thread.

Studying dual-platform-height-capable commuter train cars, this is quite interesting we already have a precedent as reaperexpress mentioned. Montreal's new AMT Mascouche Line opened in 2014. It is line with stations of multiple platform heights, and trains capable of servicing both platform heights. So, we have here, right here in Canada, a real precedent of a Canadian line running dual-platform-capable traincars, built by a Canadian train company, Bombardier. It gives a hint on what our future GO RER may become, whether it be MMU/EMU or locomotive-driven.

This may not be the GO RER or SmartTrack trainset, but it illustrates one of the many possible solutions for The Great Platform Height Debate. The flexibility of being able to service both low-platform and high-platform stations, while having enough capacity to replace GOtrains, is a very interesting option that Metrolinx is probably seriously considering.

Dual-height doors
Credit: Montreal Gazette

Wheelchair boarding, attendant assisted:
Credit: AMT

It appears it would not be too difficult to make it attendantless with a few future modifications, especially if Metrolinx owns the line, standardizes the clearances & use a smaller platform gap.

Low platform boarding; electricified
Credit: Montreal Gazette

High platform boarding; electricified
Credit: Montreal Gazette

One very major con with these cars, is that the boarding doors are noticeably narrower than GO's current Bombardier bilevels. That can slow down boarding significantly. On current GO trains, the doors are wide enough that people can disembark simultaneously while people embark. Or often, embark double-file. On these specific train cars, that can't happen due to the narrower width of the doors. Though, stepless boarding may compensate significantly.

 

[smallspy]

The Multilevel cars are a terrible idea. Besides the issues with the doorways, there are no locations anywhere on the system with restrictive overhead clearances, so why would you force yourself to be limited in size? They are considerably heavier than the BiLevels, and that means that trains won't be able to accelerate as quickly as they do today.

There seems to be no particular reason why you couldn't modify the BiLevel design to have high-level doorways if you wanted to go that way.

Dan
Toronto, Ont.

 

[mdrejhon]

The Multilevel cars are a terrible idea. Besides the issues with the doorways, there are no locations anywhere on the system with restrictive overhead clearances, so why would you force yourself to be limited in size?

Yes, that's a good argument against the trains. It's simply a demonstration that several options exist -- and might conceivably be pondered by Metrolinx as they figure out what to do.

There seems to be no particular reason why you couldn't modify the BiLevel design to have high-level doorways if you wanted to go that way.

It can be incredibly disruptive and expensive to modify all the stations to high level platforms, all at once. How do you handle a GO network expansion with high level platforms? Especially since all the GO stations tend to be 300-meter-long low platforms.

The way the current trains are set up, is that existing GOtrains aren't compatible with UPX height platforms (~48" platforms) even after modification. You actually would have to step downwards to enter the train.

Do we go with current 8" platforms, interim ~24" platforms or full height (HSR / UPX) 48" platforms?

It's not as simple as you think.

 

[Catenary]

Having read a bit about the similar discussion incto this currently going on in the San Francisco Bay area, I'm a fan of 760mm. CalTrain operates Bombarider BiLevels as well as other bi-level cars with high lower levels, with three steps to the lower level. There are a number of issues that must be reconciled there:

- North American high platforms are ~48 inches ATOR (Above Top of Rail)
- North American low platforms are ~8 inches ATOR
- Most High Speed Trains have platform heights of 48-51 inches ATOR
- The EU has standardized on platform heights of 550mm (21.5 inches) and 760mm (30 inches)

CAHSR is looking at 48 inch platforms, while CalTrain has already invested in 8 inch platforms, which are not level and have the same issues as GO. There is a growing support for standardizing both systems to the 760mm height, and train manufacturers are beginning to introduce high speed trains that have level 760mm boarding. A Bombardier BiLevel car has a lower floor height of about 25 inches, so it is incompatible with 760mm platforms, but other bilevel railcars with higher deck heights would fit nicely.

I'm a personal fan of the 30 inch platforms, but they do pose a challenge for systems like GO that rely on Bombardier BiLevels.