|
|
|
GO Transit: Union Station Shed Replacement & Track Upgrades (Zeidler)
mdrejhon
Senior Member
I remember the brutalist architecture of Terminal 1. I'll have to agree with you on that one (if it weren't for flight capacity reasons) that some of my best memories (my first trip to Canada's Wonderland as a kid) started with my memories of that terminal. That said, that's quite a low bar to compare Union's shed against.
I wouldn't have minded keeping the ironwork if they housed it all under full-length glass including the concrete ducts. It would not be quite St. Pancras Station, London -- but they have the luxury of doing that due to 100% electrification already.
Then we can keep heritage AND gain a Crystal Palace on top of it all -- maybe even 19th century greenhouse style (making the shed look good) outside the modern glass centre. Alas, if only we were electrified 20 years ago...
Maybe in 50 years from now when they need to do *another* renovation of the shed.
I now miss the views of the illuminated CN tower during my evening commute during the depressing dark hours of these shortening days towards winter. I shouldn't have to go to Platform 24/25 to enjoy CN tower views. Green gardens, I support them practically everywhere else, but.... CN tower! Tourism! World city! Glass was a no-brainer. Take the train, exit the train and see beautiful Toronto views above. We had that for a while when it was all open during construction. Now it's a grand missed opportunity, IMHO.
Last edited:
mdrejhon
Senior Member
Allandale25
Senior Member
AlvinofDiaspar
Moderator
Thanks for posting that - good that someone at the top finally dare to mention out loud the potential and problems at Union although...
It will have to change - platform in/egress is fundamentally tied to circulation elements at the concourse/teamway level. Those narrow staircases are part of the problem and it is unclear how the current works will handle modifications to that system.
AoD
Allandale25
Senior Member
It's part of the challenge with adjusting the platforms at Union the location of the support columns under the tracks? Would they remove a few tracks to make wider platforms? Would they also have to address the support columns for the shed which could be complicated because of the heritage protections?
AlvinofDiaspar
Moderator
I think their plan is not reorganizing the locations of the tracks per se (because the columns are right underneath - moving them will require structural modification) but eliminate some tracks and fill the space up as platforms. Fundamentally there is no reason why you can't carve new openings there for staircases/escalators and close up the existing ones but that's not what I would consider as minor modification.
Best way to conceputalize this is to look at the cross section:
(Via Archdaily http://www.archdaily.com/342301/tor...-roof-proposal-zeidler-partnership-architects)
AoD
Markster
Senior Member
I think the entire question about Union is worth quoting. I particularly note the statement about the signalling system, since it's clear that the Union station rail corridor capacity is constrained by that, and not physical space. Other rail stations do much more with less.
What has to be done to improve Union?
I'm very keen on level boarding. It allows lower dwell times, it allows for better throughput of fleets. If we are going to move towards a higher-frequency service, operations through that corridor needs to be segmented. The key in a busy corridor is to minimize [train] crossings. Even if it's half a minute, it's a half a minute lost. Definitely [we're looking to buy] a signalling system solution that is more automated. It allows you to safely manage a huge capacity of trains through a narrow corridor with great effect.
Part of that whole solution will include a rebuild of the platforms. We must build up the platforms, but the platforms aren't right. They're too narrow. In order to get increased capacity through that corridor, we going to have to sacrifice some of [the current platforms]. We have to come up with a combination of bay and through platforms, which we'll have less of because we need wider platforms where people come up. We have to create wider width around those entry points, because it's simply unsafe. I've asked the team to look at a Union station platform layout that's different. As we're going to redo the platforms, [we'll] lift the platforms to the right level.
What has to be done to improve Union?
I'm very keen on level boarding. It allows lower dwell times, it allows for better throughput of fleets. If we are going to move towards a higher-frequency service, operations through that corridor needs to be segmented. The key in a busy corridor is to minimize [train] crossings. Even if it's half a minute, it's a half a minute lost. Definitely [we're looking to buy] a signalling system solution that is more automated. It allows you to safely manage a huge capacity of trains through a narrow corridor with great effect.
Part of that whole solution will include a rebuild of the platforms. We must build up the platforms, but the platforms aren't right. They're too narrow. In order to get increased capacity through that corridor, we going to have to sacrifice some of [the current platforms]. We have to come up with a combination of bay and through platforms, which we'll have less of because we need wider platforms where people come up. We have to create wider width around those entry points, because it's simply unsafe. I've asked the team to look at a Union station platform layout that's different. As we're going to redo the platforms, [we'll] lift the platforms to the right level.
smallspy
Senior Member
And doing that will drastically reduce the number of trains that can be put through the station.
Thus, it doesn't seem like the wisest idea until such a point as they've figured out how to build more tracks into the facility.
Dan
Toronto, Ont.
mdrejhon
Senior Member
I believe you if it was done today, but:
Aren't Metrolinx suggesting this under ~2031 assumptions?
-- Full deployment of a PTC/CBTC type system to European-quality standards (moving-block or smaller-blocks). Train chasing each other at much tighter distances.
-- Full deployment of USRC digital signalling infrastructure replacing 1930s infrastructure
-- USRC approach speed upgrades from both east and west
-- Short dwells similar to other GO stations (1 minute dwell) instead of Union (3 minute dwell)
-- Full electrification through Union with faster-accelerating trains
-- New GO Transit Control Centre (GTCC) at Oakville
Even peak period trains (when they arrive late to overcrowded platforms) manage to have all their seats filled within 60 seconds. I would project that the raised platform + much wider platform (3x wider) + fewer people waiting in a lineup at bottom of stairs, would allow full pack standee loading in ~60 seconds with level boarding.
Wouldn't this more than compensate? Pushing trains through the blocks at popular RER platforms, with Paris/Japan precision, at closer to subway frequencies (~3-5min for specific upgraded platforms, closer to 3 for EMUs and closer to 5 for elec-loco driven BiLevels)
Assuming track-elimination did not happen until all the above happens?
It's easy to be skeptical that the pre-requisites will happen, but the $13.5bn budget allocates several hundreds million dollars to a PTC/CBTC compatible system, according to the Metrolinx documents.
I did some napkin math: (with a BIG caveat):
Mathematically, at some point, elimination of a track actually increases people-throughput.
It certainly requires a big checklist of pre-requisites before the bottleneck of narrow platforms exceeds the bottleneck of carefully removed tracks (possibly only 2 removed tracks only where platforms are narrowest). There's actually a mathematical crossover point -- assuming Metrolinx checks off enough pre-requisites. Trusting Metrolinx to do all the pre-requisites to cause that mathematical crossover (pun!) to occur, is understandable -- it should not be done until it actually increases throughput.
The upgrades being done, reportedly nearly doubles the train-per-hour during peak -- the earlier Metrolinx document suggested apporximately 90 trains in one hour is doable without the underground tunnel. Ouch.
So mathematically, there comes a point where PTC/CBTC/faster USRC approaches are less useful if narrow platforms becomes a major bottleneck in pushing trains more frequently through Union. So we should not close a track until we're very sure that it actually increases people throughput (via refilling a bigger platform faster with wider staircases on wider platforms + and boarding trains faster via level boarding).
Once the pre-requisites are checked off -- I think only 1 or 2 track needs to be closed to maximize train throughput -- and only where there are consecutive narrow platforms with lots of obstructions -- where narrow platforms is causing short-dwells to become unable to fill the train quickly. Some doorsets are 3x slower loading than the doorsets on the wide platforms -- I watched this and it's amazing how platform obstructions slow down boarding. Person boards, long delay, person boards, long delay, person boards, especially when there's also disembarkation fighting in the opposite direction to (it takes only 1 disembarker to slow about 20 people from boarding along the diagonal yellow zebra striped "NO STANDING" zones). I predict, as a commuter, myself, that boarding a train to standees will double in speed, if the "NO STANDING" obstructions are slightly adjusted to double-file room when the platform is tripled in width. This could enable 2000-2500 to fully board in just a mere 60 seconds flat, based on my calculations from the fastet-loading coaches (I've seen >200 people board one coach in less than 60 seconds, so don't tell me it cannot be done).
Try being a commuter one day at peak in the obstruction zone, and compare the commuter embark/disembark flow to the most unobstructed zones on the widest platforms.
You'll quickly agree with me that 2000 people can board a GOTrain in less than 60 seconds, if there was no platform-bottlenecks + level boarding. That's simply not possible today, but you can *see* individual coaches (1/12th of a train) successfully fill to standees in ~45 seconds flat; so mathematically it IS possible. But not with those NO STANDING zones, especially with disembarkers fighting the flow.
If you laugh at 60 seconds, then even 90 seconds would be an improvement. GO transit dwells are so luxuriously long at Union, I can board a random train to go to the bathroom to do Number 1 & 2, and get off the train on time before the doors close. If you went electric EMUs with fast acceleration, and short headways on the EMU-ified platforms, the removed-track platform can easily handle over 25,000 people per hour for that one train track, possibly closer to 30,000 per hour. Today, that's simply impossible on the narrow platforms.
Assuming one track removal doubles boarding speed, mathematically:
Math Assumptions Made:
--> Define "X" as one track's train-filling speed (in people per minute)
--> Define "0" as a no-longer-extant track
--> Removed track now having a 3x wider platform plus all "NO STANDING" zones removed via relocation of stairways into middle of former tracks. Resulting in approximately halving train-filling time for any trains that abuts one edge of a wider platform
Napkin Math:
3 tracks adjacent to narrow platforms
X+X+X = 3X
2 tracks adjacent to a wide platform
2X+0+2X = 4X
So, a removed track, assuming doubled train-fill speed, theoretically results in 33% more people throughput due to complete elimination of a track.
You don't have to widen the platform on both of the train, to double a coach-filling speed, so removing track can be strategic:
narrow-TRACK-wide-TRACK-narrow-TRACK-wide-TRACK-narrow
Turning 6 tracks into 4 tracks would still be:
six tracks abutting narrow platforms:
X+X+X+X+X+X = 6X people filling speed for six tracks
four tracks abutting two ultra-wide platforms (rest narrow)
2X+2X+2X+2X = 8X people filling speed for four tracks
It finally becomes easier to encourage a stronger spanish solution via digital signage announcing only one platform instead of two. Wide platform for the bottlenecked flow direction, narrower side for the less bottlenecked. Open doors first for disembarking commuters to the appropriate platform first. Now you've signalled a reliable spanish solution of an embarkation platform and a disembarkation platform.
-- Embarking passengers are veered via digital signs announcing only 1 platform not 2. It would now be efficient, fast, and safe to do so, because of ultra-wide platform (in a fully revitalized Union).
-- Disembarking passengers are veered via a train door opening first.
The platforms would swap during the opposite-direction peak.
The 3x-wider platform now becomes big enough to hold 2000 people simultaneously. If you stagger opposite side trains (north edge versus south edge), 5-min headways by 2.5-min offsets, you can fill 2000 people onto the platform every 2.5 minutes via a steady one-way flow of people onto the platforms. Even the existing stairways (shifted to center of removed track) are fast enough throughput to handle that if there's no counter-direction people.
Trains leave Union quicker after fuller fillings, 1 minute dwell instead of 3 minute dwell, 3-to-5 minute headways for tracks specifically abutting ultrawide platforms that has ZERO "NO STANDING" zones.
As a commuter, I intentionally stand near the NO STANDING zones, and board there because coaches are often less full -- Occasionally, I've even had a single quad seater all to myself on a 5:30pm peak train (train was late, people scrambled to board as fast as possible, train left as soon as most boarded, the NO STANDING bottlenecked guaranteed emptiness of specific coaches!) -- just because I am familiar with how constrained commuter flows are on narrow platforms, helping me select coaches that cannot be mathematically filled in 60 seconds like at other parts of Union. Late trains at Union (those that also need to clear the track quickly for a subsequent train) often leave far emptier than 2000 people even at peak, interrupting late boarders trying to step onto them while there are still empty seats on the coaches abutting lots of NO STANDING zones.
I think that is how you increase train throughput with 1 fewer track. But no more than 2 tracks removed -- then removed tracks then becomes the bottleneck again.
But do we trust them to do it correctly; shifting the staircases closer to the center of a closed track -- to eliminate the NO STANDING zones? That's a big ask. And mandatory to speed up trains (via improving sheer people throughput more than compensating for loss of track), but it's a lot of pre-requisites to fully double train-fill speed.
Last edited:
cplchanb
Senior Member
man those corrugated steel roofs on those photos from the 12th make the building look so cheap......
TOareaFan
Superstar
they don't look that bad in person......that guy is just a crappy photographer using a useless iPhone7
mdrejhon
Senior Member
I didn't know Drum118 uses a useless iPhone?they don't look that bad in person......that guy is just a crappy photographer using a useless iPhone7
I also do not think that he is a crappy photographer, that's a bit mean to him.
(I don't have an iPhone)
BTW, my photo was intentionally adjusted to expose the CN tower, rather than the roof structure. As your eyes adjust to bright light, dark things become dark. It's also what the human iris does, and when you stare in the bright spots, the roof darkens as your iris shrinks.
For some humans with disabled/impaired vision, much more dramatically -- like walking indoors to a dim room from a bright sunny place. They take much longer to adjust to dark rooms than you; we must not make fun or diss such.
It is THAT dark as in my photo, for some humans, compared to more vision-privileged people. Not all of us sees bright-dim contrasts in quite the same way. If you're walking for 15 minutes on a sunny Bay street, and suddenly walk up the Bay Teamway, the platform roof are darker than in my photo especially for a slower-than-average-vision-adjusting person. Such vision sensitivity variances are even more common than color blindness (8% are colorblind), not every human sees (including iris speed) in quite the same way! So saying such generalities as you are, is quite overpriveleged.
Anyway:
I hope Union/Metrolinx installs some nice lighting soon to compensate for the dark. The Platform 26/27 continuous uninterrupted warm-white LED light strip is relatively nice -- I hope they duplicate that to all platforms; it's much brighter on 26/27 at night than the rest of the interior platforms.
Last edited:
Leo_Chan
Senior Member
Since when were iPhones useless? And how does that impact the pictures taken? Almost all cell phone cameras are subpar to a “real” camera.
TOareaFan
Superstar
You see I was completely joking (and since I was joking at myself and my post I really didn't think anyone would challenge it) but since you ask.....I have had/used several cell phones with better cameras than my iphone7