GO Transit's big problem: Too much west and not enough east.
Balance is key in all places — transit is no different!
GO Transit is a system I talk a lot about — for obvious reasons. It’s a very good commuter railway that should be a very good regional railway, and it’s definitely getting there. Electrification, better stations, grade separations — GO is getting transformed piece by piece from the track up into the type of system most European cities take for granted and most North American cities can only dream of.
But there is a big problem (ok. there are a few — future post subjects!), and that is the unbalanced nature of the railway.
If you’ve followed my work for a longer period of time, you’re likely familiar with the term “through-running”. But, if not, here’s an overview.
Through running is the idea of lines not ending at a major node in a transport system, but passing through it. The rationale here is simple — turning vehicles around, especially big trains, is almost always harder than keeping them going in one direction, and doubly so with space constraints that are inevitable at transport nodes where land is highly developed and expensive. This is because to turn a train around you must either use a loop (super space consuming and almost never used for heavy rail / metro), or use crossovers (which frequently create conflicts and require personnel moving around). It’s much easier to have a frequent transit line — particularly with big and long trains — pass through major hubs by continuing in the same direction, and deal with the turning of trains in a low density area, possibly on a branch where train movements are reduced.
The benefits of through running are exactly the reason you see S-Bahn/RER tunnels through cities. By connecting railways on either side of a city with a (relatively speaking) short tunnel, you can eliminate two turning operations and increase convenience and capacity.
Fortunately, Toronto and its downtown hub of Union Station are almost unbelievably well set up for through running. The wide above ground Union Station rail corridor has room for 8-10 tracks at a minimum along its length, which are more cross city tracks than are seen in most significant S-Bahn systems, and we didn’t have to dig a single inch of tunnel for the privilege. Unfortunately, all is not well — because while we have the downtown capacity to push trains through Union Station, there is a serious balance issue. The problem here is not at all immediate, but it is very significant and will require careful planning and thought to address.
Right now, approaching Union Station from the west are four lines (Lakeshore West, Milton, Kitchener, Barrie), while from the east there are three (Richmond Hill, Stouffville, Lakeshore East). Here, you can already start to see the lack of balance — more lines come in from the west than from the east, and thus in the perfect scenario only about 75% of trains from the west could through run!
Now, you might hope that the fact that the Milton Line has been a real laggard in terms of transit use (being owned mostly by Canadian Pacific Railway) would mean that we essentially have three western lines, but unfortunately much of the Richmond Hill line is owned by Canadian National, and Metrolinx does not seem particularly interested in enhancing the part it does own, apparently because of flood risk (this deserves a future post!), but we are doing a lot to mitigate this!
So we are left with three lines to the west and two to the east, right? Well, not exactly…
As it turns out, Toronto’s Western rail lines are a LOT beefier than its eastern ones — the Lakeshore West Line is quad-tracked as far as Mimico, and triple-tracked or better for much of the rest of the way out into the suburbs (I’m only really talking about passenger priority tracks here), with full quad-tracking also on the table. The Kitchener Line is also being quad-tracked all the way to Woodbine, and could likely be quad-tracked all the way to Brampton. Two of the tracks do indeed feed Pearson Airport at the moment, but especially if larger trains ever head to the airport, or (unlikely as it is) a diversion through the airport grounds happens, very little capacity is actually going to be eaten up by the airport itself. We’ve also got the Barrie Line to the west of course, which is being double-tracked right now with projects like the Davenport Guideway. It’s also worth pointing out that even though the Milton Line doesn’t have confirmed frequent all-day service yet, the giant Strachan Avenue grade separation trench and the rail corridor up past Bloor Station has room for 8 and 6 tracks respectively, so even more trains could be coming.
This compares to just four tracks heading east from downtown Toronto, which, sure, is the same as the number through running east out of Manhattan right now, but certainly does not match up with the 10-to-12 that come in from Toronto’s west.
This creates a few major issues:
Where do extra trains coming into Union from the west turn around?
Where else could we send the extra trains coming into Union?
Why are we not creating additional stations to the east of Union on the Union Station Rail Corridor (where there is quite a bit of space!) to turn trains and relieve demand on Union, while also allowing passengers to transfer to trains to the west on the east of Union itself?
Right now, we have a few answers… sort of. Metrolinx is reactivating some tracks in the southern Don Valley (which could someday be the starting point of an enhanced Richmond Hill Line) so that some trains can be turned up that way. This is okay, but doesn’t have any of the connectivity or transportation benefits of having actual stations east of Union. Some trains will also be turned at Union itself, and of course we won’t exactly be hitting the theoretical limits of the track capacity we have, but it also probably doesn’t make sense to run twice the frequency on the Stouffville and Lakeshore East lines to compensate, and that would create its own issues.
The natural question is… how do we solve this problem? Here are 3 options ranked from Least to Most likely to see the light of day.
Option 1
One solution that has been tried at least once (or twice) is a City Loop, which is Melbourne’s name for its downtown suburban rail loop that essentially functions to turn trains from the east around so they can head back east, since Melbourne has the reverse problem to Toronto — far more eastern than western lines. A City Loop is an option for Toronto, but it would clearly be extremely challenging to build, and would require a lot of space under the densest part of Canada’s largest city that would likely be better used for rapid transit (a City Loop would not be very useful for getting around Downtown). It would also likely require major construction near Union, which would be immensely tricky. I also think it’s worth asking whether this option was even good for Melbourne, as the loop has led to confusing service and the entrenchment of the unbalanced network.
Option 2
Another solution that is less outlandish as it does not require a giant underground loop would be new eastern GO lines, though these would not be a reasonable thing to build based on our current model for GO (bilevel trains). Such lines would likely be mostly tunnelled given the developed nature of Old Toronto east of the Don. While ideally, we would have three eastern corridors to handle the minimum of three extra track pairs we have west of Toronto, doing something reasonable with the Richmond Hill line should definitely come before new tunnels, so I’ll propose two.
One would be a tunnel that starts just west of East Harbour (creating a new underground connection with the various lines there), before running east towards Main Street and turning north to terminate at Bloor — with a large concourse or the platforms themselves being usable as an underground connection between Main Street Subway and Danforth GO stations. This project wouldn’t preclude further northern extensions, and would be a smart way of freeing up lots of capacity on the Ontario Line east of East Harbour, while also giving you a purpose-built facility to turn trains.
Another option would be a much shorter tunnel into the Port Lands as well as potentially southern and eastern Leslieville. This idea reminds me of Tokyo’s Rinkai Line and would likely need to come along with some major intensification in the aforementioned areas, but there is already pressure to do just that!
Option 3
The most infrastructure-light option (that doesn’t have me ripping my hair out) is probably to add a few stations on the Union Station Rail Corridor. I’ve always been a fan of putting a major station at Spadina that could intercept all or most of the western rail lines, and while Metrolinx is not building that, a station is going to be built at Spadina for the Barrie Line. One scary potential outcome could be seeing the Barrie Line cut back to terminate at this station, since Spadina isn’t that far from Union — people could walk it, or take the Spadina Streetcar, and offices seem to be moving west anyways. The better solution is to have Barrie trains continue through Union of course, as with other services that would otherwise terminate — such as locals from the Lakeshore West and Kitchener Lines, and Airport Express Trains. A new station at Sherbourne (which would reach all the way to Jarvis if it had 300-metre platforms) could be one place to turn trains. Some have also suggested a station at Cherry Street, but personally I think given a major station is being built at East Harbour, some additional platforms there (which may have to be built directly west of the station with a connecting corridor across the Don) would make much more sense. This might even be the boost that takes East Harbour from a satellite business district to something that can compete with King and Bay — since almost all GO services could travel here.
Now, spitballing these ideas is a lot of fun, but there is a serious conversation to be had here. Smart regions recognize problems like the lack of balance in a regional rail network from a mile away and plan space for future transport right-of-ways as well as the development/land use to support them in order to create an integrated proactive plan.
Toronto is not there… yet.
Regarding option 1, a big challenge with circular routes is that the trains can't make up as much time if they are delayed and if the tracks are shared with other services, can delay other services and has a knock on effect. London's Circle Line comes to mind and the branch to Hammersmith hasn't really helped solve this problem but if done right, a circular line can work but as you said will cost a lot of money for Toronto. Great article BTW!
I like Option 2. It may also be the thing to finally force an indoor pedestrian connection between TTC/GO there as well. Future thoughts if you include a Midtown and Bolton line? I suppose Midtown helps create room at Union as Milton (and maybe other lines) would have trains run that way instead, and avoid the core's bottleneck. Is Bolton not as much of an issue since it wouldn't be a unique trackset into Union?