Zurich, the Northern Line, Toronto and some thoughts on a new Regional Rail Project.
An interesting observation about transfer capacity, and a potential new way to approach network design.
So, the cat is out of the bag: I am working on an explainer video on the Zurich S-Bahn. Now, despite thoughts that anything Swiss must be a world away from transit in North America, the Zurich S-Bahn in particular actually has a surprising amount in common, at least at a fundamental level with regional rail systems here - most notably GO transit. With a large network of often very long line which carry double decker trains with low floor boarding and even some sets pulled by electric locomotives.
Of course, there are some serious differences — forone, Zurich’s trains and operations are much more refined, with one-person operation and gap filler-based level-boarding. But fundamentally, the way service is deployed is also radically different.
A good way of explaining how the service is different can actually be with the streetcar network in Toronto, which is quite different in operation from European systems — which in my opinion tend to have more compelling service offerings. In Toronto (and probably more widely in English-speaking countries as well, although Australia is better at this than Canada) there usually seems to be a belief that the best way of providing service is to run one type of trip on every line at as high frequency as possible and thus minimizing overlap and codependence as well as interlining. This is very different from how a tram system tends to operate in Europe and in particular the German-speaking countries: In such locales, it’s much more common to see a network operated with more direct point to point trips at lower frequencies. This might initially seem worse as you can’t always get your trip started as quickly (or be spontaneous to the same degree), but it does mean less wasted capacity and shorter trip times, while also massively reducing the need to transfer, which often is where a lot of friction is generated.
Zurich’s-S Bahn network operates in much the same way. If you looked at a map you might assume there are many quad-track-or-greater corridors due to the number of services operating in parallel, but because of the way German systems tend to denote branches more precisely as discrete services most individual lines on a map only equate to at most two trains per hour per direction. This very low amount of service for each individual route makes a lot of sense and is what enables so many different services to operate in Zurich despite a much lower number of actual corridors (there are still a lot!) relative to the actual number of services that run on them.
The interesting thing to consider here is congestion at individual stations. Zurich’s are not that massive despite the very large trains and high amount of services — even the main station is not actually that large considering the crazy connectivity of it both within the region and Europe more broadly. I think this may well be a benefit of the way services operate in Zurich and actually reduces the need for shoulder stations which I have advocated for in Toronto.
I think this notion can actually be compared to the Northern Line on the London Underground. You see, the Northern Line — or as I prefer, lines — are two north-south metro lines that come together at two points but are only actually interacting at one — Camden Town, which enables through services from either northern leg to either southern leg. Now, this interlining limits frequency on both lines, but removing it proves tricky — the ability to take a train from either northern branch to either southern branch means the restrictive size of Camden Town, while far from ideal, is not a complete disaster — which TFL thinks it would be if service was isolated.
A recent discussion in Toronto actually brought this up as well: While I don’t think running Line 2 trains from Kennedy into downtown via the University subway is a good idea, this was how the system was designed to operate (poorly I might add), and so the idea of using it as a way of limiting congestion at Bloor Yonge is enticing if unlikely to work well.
This is actually a pretty interesting consideration that, I’ll be honest, I haven’t thought enough about in the context of regional rail, but which is clearly helpful. If you can get a direct train, you don’t need to transfer and so while your minor stations might be more busy, your centre city stations don’t need to be enormous, which is quite good for cost effectiveness.
You might be wondering how this plays into regional rail more broadly, and this will likely be a topic I dive into in greater detail in the future, but I like the idea of designing some rail plans for cities including Toronto that do not rely on the construction of shoulder stations — not necessarily because I think they shouldn’t exist, but because I am under no illusion that we will see as many built as I would like. Such rail plans would be complex in terms of operations and infrastructure, but I think they present a compelling challenge to designing effective networks for common trips and so I’d like to design some!