6K 
UrbanToronto is celebrating 20 YEARS throughout October with stories and images looking back over the last two decades. Today, Reece Martin looks forward to the next big decisions for Metrolinx and GO Transit that will shape our next 20 years.
* * *
Around the world, the common element seen in every great urban and indeed suburban rail system is electrification. As Toronto continues forward with its massive GO Expansion program, it’s clear that among the most important elements is the installation of overhead catenary along five different rail lines.
The scale of this project cannot be overstated: hundreds of kilometres of new infrastructure will be constructed, and that infrastructure will enable a sea change in the type of service that GO can provide. Plans go beyond the 15-minute service seen today on the Union Pearson Express (and all too briefly on the Lakeshore Line before the pandemic), and call for trains at major stations every couple of minutes for much of the day.
And yet, despite this complete revolution in the service and infrastructure that will outline the GO system, one element -- the GO trains themselves -- will not change very much at all.
Indeed, at present, plans appear to be to simply hook up electric locomotives to existing bilevel coaches, based on a design that is the better part of half a century old and that was not intended to provide the level of service passengers will hopefully see by the 2030s. In short, purchasing electric locomotives feels like taking the engine for a Ferrari and then mounting it in a jalopy.
To be clear, GO transit has an enormous fleet of bilevels with nearly 1,000 of the coaches, and it would obviously be wasteful to retire them before their time has come. But many of the bilevels are actually quite old — over ⅓ of them are 30 years old today — and by the time electrons are flowing across the GTHA rail corridors and e-locos are built, the fleet will be even older.
Further still, GO is clearly a network in transition. While spacing between stations has traditionally been wide, even before electrification works have started, numerous infill stations on the electrified sections of the network are moving forward. These tighter station spacings along with radically increased service will stretch the acceleration and deceleration capabilities of traditional locomotive-hauled sets and even electric locomotive-hauled trains. The reality is that GO ridership is likely to continue to grow for the foreseeable future, and with trains already struggling to quickly unload and board passengers at today's passenger numbers (largely thanks to the few and narrow doors on the bilevel cars, the step up into them, and their tight interior layout) it is not clear that the bilevel and locomotive model is a good one to form the backbone of service.
While it's clear that a large portion of the bilevel fleet can stick around and keep shuttling people around the GTHA for decades to come, it’s also clear that there is a gap to fill. While much attention has been paid to the huge growth in off-peak service that will happen under GO expansion, peak-hour travel demand will still grow substantially, which means a need to at a minimum maintain the current fleet size. This is simply inevitable in a region that continues to grow as it builds more housing around rail stations, expands connecting transit options, and does not build more highway lanes into the core.
Fortunately, Toronto is not the first city in the world to build a high-frequency electric suburban rail network. Such networks are the norm in cities from Europe, to Asia, Latin America, Africa, and Australia, and probably the single most common feature of all these systems — which are powered by varied means, run on tracks of different gauges, and play various different roles in their respective transportation systems — are their choice of rolling stock, which is almost always Electric Multiple Units, or EMUs.
Electric Multiple Units, like the subway trains already in use in Toronto, are the global standard technology for high frequency rail service, be it subways, suburban trains, or even high speed rail. Multiple Units are so called because — compared to locomotive-hauled trains where all the power comes only from wheels at the front or back — they feature multiple powered wheel sets spread across the train. This improves acceleration and deceleration, enables the climbing of more aggressive grades, and comes with a model of rail operations based on flexible consists of fixed-length units as opposed to the coaches and locomotives model of yesteryear.
To be clear, there are frequent services out there that do operate with locomotives, but they are increasingly uncommon, especially in urban contexts. Of all of the “S-Bahn” suburban train systems seen in the German-speaking world, the highest traffic networks all use multiple units, and holdouts like Zurich are moving in that direction. Indeed, the vast majority of passenger traffic in countries from Korea, to Sweden, to Brazil moves on multiple unit trains.
What's more is that the introduction of a new train model is completely appropriate at such a major inflection point in Toronto’s transit story, much of the reason mainline EMUs were developed in the first place was to serve the many suburban rail systems that Toronto seeks to emulate — even the original name of the project “GO RER” harkens back to Paris’ network — which does not use locomotives!
With the introduction of new trains, not only could Toronto bring something to the network that is well-suited to the dramatically different service the region will see, but it can also address the shortcomings of its current trains. New trains could have more doors, a design purpose-built for level-boarding, and seating that reflects the reality of busier trains and shorter trip times. Of course, to some extent bilevels could be retrofitted with some of these changes, but for the oldest cars in the fleet, the idea of spending substantial money on mitigating problems seems shortsighted when new trains could fix those same problems.
While a “wait-and-see” approach is often reasonable, given the electrification of GO is moving forward, the time for waiting has passed. The reality is that EMUs — or at least a plan for them — is not a “nice to have”, but a critical step in determining the shape of the future network and infrastructure that likely should have been taken years ago. Without a sense for what a fleet with a growing fraction of EMUs would look like, we miss a huge opportunity to optimize growth and our infrastructure for the trains of the future, rather than the trains of the past.
That's because the interaction between rolling stock and infrastructure is hugely important to the design of everything in a rail system. A network based on bilevels, even short sets of them pulled by electric locomotives, will necessarily look very different from a network that has some number of multiple units. Planning for multiple units today, and indeed starting with ordering some as we begin to retire the oldest bilevel cars, will set us up for future success.
For example, new trains will require new yards and maintenance facilities, and different operating procedures, but they will also open up opportunities. Fewer crews will be needed to staff train services that can complete trips more quickly with EMUs. Infrastructure designs can also be vastly different: curve radii, alignment types, and grades that would currently be unacceptable could open up new opportunities for things like tunnelled links to major destinations, lower-cost grade separations, and connections to locations like Pearson Airport with track designs that don’t place long coaches on excruciatingly slow curves. Of course, replacing the oldest bilevels in their 50s with new trains equipped with preventative maintenance technology and more robust designs, will also drastically reduce operating costs.
Now, the natural question to ask is… what about the cost? And the reality is that even a fairly large order of EMUs would only be a small fraction of the cost of GO Expansion and have the potential to pay for themselves in less expensive infrastructure (especially given the inflated prices we pay in Canada) and lower operating costs. The current proposed plan is for all locomotives as per Metrolinx, but there's obviously holes in that.
The UP Express trains for example have been notoriously unreliable and clearly will need to be supplemented in coming years as the service continues to grow in popularity and Pearson Airport’s passenger numbers increase dramatically. Clearly, given how slowly the (admittedly long) diesel multiple units on UP Express are forced to travel from the Kitchener Line to Pearson Airport, it would be unthinkable to run a locomotive over the connecting guideway — even if it increasingly sounds like that might be the plan — thus making the case for EMUs. And once EMUs are operating on the Kitchener line for UP, it would seemingly make sense to simply lower the platform at Pearson station, and through-run low-floor EMUs onto the Stouffville line east of Union.
Ultimately, while the cost of EMUs is real, the opportunity cost is obviously massive. EMUs would unlock the ability to reasonably add far more stations to existing GO corridors (each with the potential to unlock hundreds of millions if not billions in development), unlocking subway levels of service for far lower costs, while using express tracks and passing loops to maintain and even improve trip times even to the most distant locales. The idea that “we don’t need EMUs because our stations are widely spaced” has confused the cause and effect: our stations are widely spaced because of our low-performance trains.
Now, what could service look like for a version of GO that used EMUs?
Obviously many services would still be run using the large existing fleet of bilevels, which has always been the plan because significant portions of the network will remain unelectrified and suitable for that rolling stock well into the future, including both the Milton and Richmond Hill lines, Lakeshore West services to Aldershot and beyond, and Kitchener services beyond Bramalea. That leaves the valuable inner-suburban niche for EMUs, which could operate high-frequency stopping services from stations like Unionville and Burlington to and through Union Station. These services could often operate on their own 'local' tracks such as on the Lakeshore West, Stouffville, and Kitchener lines — much like how S-Bahn services in many German cities operate on dedicated tracks parallel to intercity and regional express services. Since EMUs would enable more flexible alignment planning and grades, the local services could even have short branches or diversions built off them, again, like is seen on German S-Bahn systems, such as a better alignment through Pearson airport or a connection to Mississauga City Centre.
To be perfectly clear, Toronto ordering EMUs does not mean retiring bilevel cars midlife or even cancelling plans for electric locomotives. Instead, embracing EMUs would allow the region to plan a network oriented towards the future, and embrace big service and planning opportunities sooner. EMUs could operate on the most “subway-like” services, opening up space for the bilevels to do what they do best.
With rail systems, matching rolling stock to infrastructure is crucial, and with Toronto’s plans for EMU levels of rail service, EMUs make obvious sense.
* * *
Reece Martin is the creator and host of RMTransit, a YouTube channel focused on transit, infrastructure, and development around the world, with extensive knowledge and professional experience as a transportation planner.
* * *
UrbanToronto will return Monday with another story celebrating 20 YEARS. In the meantime, check back often to our front page and Forum to keep an eye on all the current and emerging trends, and you can always leave your comments in the space below.
* * *
Thank you to the companies joining UrbanToronto to celebrate our 20 years in business.
