Regarding squeezing UPX trains along with SmartTrack trains (or, if you prefer, call it GO RER with extra infill stations)...
People keep saying Georgetown Corridor, even after expansion, cannot handle the increased services that SmartTrack will also bring. This is a fallacy, as there is a lot of optimizations that have not yet been milked out of it. With $13.5 billion, can buy a huge number of corridor capacity improvements including electricification, faster-accelerating EMUs, USRC, running through trains through Union (connecting Kitchener-Stoufville for efficiency), positive train control, improved dispatching, resignalling of corridors to permit tighter headways, etc). Many countries have milked their corridor capacity far more than Metrolinx has with their tracks. Even the Lakeshore corridor capacity is actually not efficiently used, compared to Japanese and European standards. With a number of modifications to the SmartTrack to make the urban parts of Kitchener-Stoufville corridors compatible with the GO RER initiative and 100% corridor-compatible with all Georgetown corridor plans (e.g. optimizing the services around an express track / allstop track architecture, and slotting fast-accelerating EMU trains between UPX EMU trains, and short headways, as I've explained in previous posts), it is really a very good plan, albiet with some questionable attributes (E.g. the incredibly expensive Eglinton spur, and funding mechanism). It is fairly harebrained if they go with LRT trains fruitlessly attempt to slot frequent-stopping 12-car bilevels between UPX trains. But with the RIGHT kind of train, the many-stop GO RER service branded as "SmartTrack", can be slotted into a 4-track corridor with other services such as UPX. Kitchener express 12-car bilevels (that goes express after Bramalea, or only stops at Weston/Bloor as a limited-stop express), and discontinue 12-car service completely in the 416 stations, to free up corridor room, exclusively servicing the 22 stations (of SmartTrack suggestion) using the GO RER electric train only. Sheer frequency (15min offpeak, possibly ~5min peak) even of a 150-meter single level EMU, will move more people than an hourly 12-car bilevel. And you can use the 12-car bilevels as the limited-stop express. After all, RER = Regional Express Rail. On some commuter-rapid-transit routes (France's RER service is a hybrid subway-commuter train that has subway frequencies during peak period), France operates a bunch of allday limited-stop-expresses interspersed with all-stop trains. During express operation and limited-stop operation, use the 12-car bilevels. During the allstop operations (with infill stations), use the high-performance EMUs that can keep up being between trains despite having to stop more, or kick both UPX and 12-car semiexpresses onto the express track, assuming the performance envelopes can fit the headways safely, especially with enhanced signalling (optimized for reliably tighter headways) and positive train control, to make it easier for the train drivers to safely keep the trains slotted in their schedule without interfering with the train ahead/behind. If the performance envelopes of the 12-car bilevels don't allow them to stop at Weston/Bloor between UPX trains, then just run them fully express to Bramalea.
Assuming infill stations are done from Bramalea to Unionville (i.e. that element of SmartTrack plan is kept), and assuming they plan to keep UPX operating (since it'll be profitable for Metrolinx) then EMUs pretty much becomes essential for allstop service in the Georgetown corridor, or it becomes a hugely expensive waste of corridor capacity (outweighing the cost of EMU trains), unless you want to slow down the UPX trains. Only EMU trains can be reliably slotted at subway-headway densities; no other types of train technology can pull razor-thin headway (when signalling and trackage is optimized as such). Among all train technologies, EMU have a super-huge performance envelope that allows them to either gently or strongly accelerate, all the way to gently or strongly braking, perfect for expert slotting between a diverse range of other train performance envelopes with the tightest possible headways. By eliminating the need to add more trackage, this more than pays for the premium of EMUs.
Possible trains
- Diesel locomotive pulling 12-car bilevels (goes to unelectricified Kitchener)
- Electric locomotives pulling 12-car bilevels (goes to Bramalea, possibly for peak period overflow)
- GO RER EMUs running all-stop service on high density of infill stations (ala SmartTrack)
- UPX EMUs pr DMUs
- Merging SmartTrack and UPX (not financially or politically likely in near future)
Possible service schemes (not all may be done)
- Express to Bramalea (no stops until after Pearson)
- Express to Bloor, Weston, Bramalea (skips infills)
- Stop at Bloor, Weston, Pearson (UPX train)
- All stops at all infills (ala the "SmartTrack" part of GO RER)
Possible corridor-slowdown problems that can happen
- UPX catching up behind a slow 12-car bilevel
- UPX catching up an allstop SmartTrack EMU
- Express 12-car bilevel fully accelerated, but having to slow down because UPX train stopped ahead
- Express 12-car bilevel having to slow down for an allstop GO RER (SmartTrack) train.
- EMU catching up to a 12-car bilevel ahead, because of EMU faster acceleration
- Electric-locomotive driven 12-car bilevel catching up to diesel-locomotive driven 12-car bilevel due to acceleration
- etc, etc, etc.
Various possible solutions:
- The use of common express-allstop track architecture (4 tracks, one express, one allstop, for each direction)
- Share express track between UPX and 12-car express to Bramalea. Dedicate one allstop track to GO RER (SmartTrack)
- Share express track between UPX and 12-car limited stop (Weston, Bloor). Dedicate one allstop track to GO RER (SmartTrack)
- Share express track between UPX and GO RER (SmartTrack). Dedicate one track for 12-car bilevels.
This is very dependant on train performance envelopes:
- Slow accelerating: 12-car with diesel locomotive
- Medium accelerating: 12-car with electric locomotive
- Fast accelerating: DMU (diesel multiple unit, like current UPX train)
- Fastest accelerating: High-performance EMU
The top speed of the UPX and 12-car bilevels are similar, somtop speed is currently a nonissue, but delays caused by stop length (affwcted by stopping distance, dwell, and acceleration performance), which is why EMUs are so criticial for efficient use of corridor capacity if you make stops, otherwise, one needs to avoid doing the stops and just focussing on staying slotted between UPX trains, as part of the word "Express" in "Regional Express Rail", relegating 12 car bilevels to only stations beyond Pearson (aka Kitchener thru Bramalea).
There are a lot of other examples, the above is just a sprinkling of the jigsaw puzzle that Metrolinx needs to fit together, and decide what is possible and what is not.
There is sometimes overlap depending on what is being done, a 3-car diesel locomotive will likely outperform a 12-car electric locomotive. Likewise, an electric locomotive pulling 1 car, may outperform many models of EMU. But for simplicity, the above is usually true.
What they need to do is do schedule modelling of what compatible train performances fit in the service plan. Let's assume the whole Kitchener-USRC-Stoufville corridor has been later upgraded to the latest in state-of-art signalling (e.g. capable of full speed 100kph+ trains chasing each other full speed 100kph+ trains with 3-minute headways, at least at certain moments). Pinch points like USRC, the Bloor station, the crossovers, will have to be modelled for, but if a train can safely crossover between tracks at very high speed while there's a train only a few minutes ahead on both tracks and a train only a few minutes behind on both tracks, that requires a very damn good signalling system like the type of stuff used on some European or Japanese trackage, to make sure it's able to be safely done, and the quality of the control and signalling is so good, that trains behind trains immediately knows to decelerate on time, if there's any problem with trains ahead, during very tight headways. This is frequently done on subways but much harder to do on commuter rail. But with EMUs and complete grade separation, and a very good signalling system, it's certainly altogether possible to do far better than today. Often, a lot of train signalling is extremely granular -- if a train is on a section of track, a whole lot, even a few kilometers, can be off-limits to other trains because of block-based signalling. That can unnecessarily create larger headways than is necessary to be safe. But with modern signalling, positive train control, continuously moving train blocks (or at least extremely fine granularity of blocks), efficient automated dispatch that don't keep trains unnecessarily waiting -- there's such systems in use in some more modern commuter rail systems than Metrolinx's GO system which currently still use older signalling and older operation (no positive train control or automatic train control at all). But let's assume that electrification includes various corridor efficiency upgrades such as these, so with this in mind, let's continue...
EXAMPLE:
(A) -- In schedule modelling they may find the 12-car diesel expresses perform well enough to slot between UPX trains provided they go express to Bramalea. I already know that the 12-car diesels go 140kph; I have used a GPS speedomoeter on the Lakeshore West express and I have seen them reach 140kph whenever they are running late. This is fast enough to keep behind a UPX train. So they can at least start accelerating a 12-car bilevel right AFTER a UPX leaves its station. The UPX accelerates ahead like a jackrabbit, while the diesel slowly accelerates, but eventually the diesel exceeds' UPX /average/ speed, so stays slotted exactly between the UPX trains. They then find out that one 12-car express bilevel then becomes doable with good continuous moving-block signalling (upgraded signalling on entire corridor). Therefore, as a result, they upgrade the signalling to make it possible to slot a 12-car express bilevel between UPX trains. Okay, now you can fit an express-to-Bramalea 12-car GOtrain between UPX trains, so you can run that type of express service to all stops beyond Bramalea, without interfering with UPX. Good.
(B) -- Now, we try to do one better. Now, they try to model if that train has enough time to stop at Bloor and/or Weston (turn the full express into a semiexpress). The problem is that 12-car diesels accelerate much more slowly than either a DMU or EMU, so eventually the next UPX will eventually catch up behind the 12-car. Now, you continue to model to see if there's enough time to do a brief stop and reacceleration. You may find that with upgraded non-granular moving-block signalling that permits a minimum of 3-minute headway moments of full-speed trains, you have a 9-minute sliding window between UPX trains (no closer than 3-minutes behind UPX train ahead, and no further than 3-minutes behind a UPX tran behind). Due to dramatic difference of performance envelope, permit 12-car disels to begin accelerating, say, 1.5 minutes after UPX has left its berth. The 1.5 number varies, but by the time the diesel is entering the USRC and then into the corridor going the same speed as the UPX ahead, the UPX is 3-minute headway ahead because of the UPX's faster acceleration. At this stage, you're now going the same speed of the UPX train 3 minutes ahead. You can slow down the 12-car bilevel GOtrain by up to 9 minutes worth before the subssequent UPX train (15 min frequency) is only 3 minutes behind you (and the 12-car is threatening to block the UPX train's journey). You barely have enough time to stop and then re-accelerate once or twice, managing to squeeze a 12-car semiexpress to stop at both Bloor and Weston (two stops). By the time the 12-car semiexpress is past Weston heading to Bramalea, the subsequent UPX train would be only about a few minutes behind. Hopefully 6-minutes behind, so that it rarely delays a UPX train. Ok, even better, now you can run a limited-stop 12-car GOtrain that stop at the pre-existing GO stations. Perhaps they may find it's only possible with electric-locomotive-driven 12-car bilevel (which then has to short-turn at Bramalea, the end of electrification of this phase of GO RER). Or may be even doable with the 12-car diesels (So it can go all the way to Kitchener). They find out what mathematically works, with sufficient enough safety margin for slower-than-usual trains and reliability problems (and they may find electric locomotives are more reliable).
(C) -- Computer generated modelling of random failures of diesels will calculate where to install crossover tracks so a maximum number of UPX trains can be able to bypass a stalled diesel. They may find it works. But they may find that they can't optimize corridor operations efficiency enough to work around the reliability differences of their various trainsets, and then they may have to avoid using diesel locomotives on the UPX track. But they would have to computer-model and test for this, to find out if it is doable.
(D) -- Instead....alternatively.....They may, also, consequently find that a very good performing EMU (that they were shopping for) actually has enough time for a high-performance decelerate-dwell-accelerate, to be able to service all SmartTrack stations between UPX trains. Let's assume UPX trains are now EMU by then, and the GO RER SmartTrack trains are well-matched EMU. They would then have very similiar acceleration performance. Assuming corridor now has signalling good enough for full-speed 3-minute headways, you have a 9-minute sliding window (excluding 3 mins for UPX ahead/3 mins for UPX behind) to do all the non-Bloor non-Weston SmartTrack stops (the extra stops above-and-beyond the UPX trains). In modelling, they simulate start the GO RER SmartTrack allstop train from Union 3 minutes after the UPX train leaves its berth. If one decelerate-dwell-reaccelerate very quickly (20-to-30-second subway-style dwell at non-major stations), they may find there's enough performance to stop at some or all the "SmartTrack" stations, while still having enough safety margin to not delay a subsequent UPX train. So they find they may actually are able to buy a specific model of EMU that performs well enough to do allstop service between UPX trains. They might find they can do it, but only if they remove one or two of the planned infills, for example (and modify the SmartTrack plan to fit). They will be limited to the UPX frequency, e.g. 1 UPX train and 1 SmartTrack train, interspersed between each other.
(E) -- They may find they prefer to dedicate one track (per direction) completely to the GO RER SmartTrack trains, so that it can eventually increase to subway-style frequencies (20 trains per hour), so that they can run high-level-platform 150-meter trains moving more people than 300-meter 12-car bilevels (that are headway-limited because they aren't as fast-accelerating as EMUs), considering that only seven pre-existing go RER Platforms need to be modified to go high-level, in order to make it possible for Metrolinx to purchase a specific EMU that they might be looking to buy, in order to make it possible to squeeze a specific service plan, as an example (...see, that's why I think Kitchener GO RER route is a de-facto EMU, if they want extra infill stations, and if they don't want to interfere with UPX....)
Obviously, they have to simulate and jiggle things around, to see what works, what fits in the corridor, but if commited to keep the services they plan, they will definitely come up with a service plan that fits all the planned services. There is already four tracks on the Georgetown corridor (and extra tracks can redirect Milton/Barrie service, keeping four tracks for UPX/SmartTrack/Kitchener trains). Yes, the difficult Bloor area can actually be rejigged to fit six tracks (which will help frequent Barrie/Milton trains a lot). My prediction is the four-track express-allstop architecture, with express track assigned to UPX/expresses/HSR and the allstop track dedicated to allstop GO RER with infills (ala SmartTrack).
Please bear with me, sometimes my train terminology is incorrect; the ones that railroad engineers use is sometimes different than what I am trying to describe, but the Europeans and Japanese have some amazing train-service-modelling computer programs, and Metrolinx will need to push the corridors to highly efficient signalling that permits tight full-speed-train headways, for maximum flexibility in designing service plans (like above examples).
The people who say Metrolinx will not be forced to choose EMU, do not realize the challenge they are going to hit hard, if they want to keep the UPX train AND also add lots GO RER infill stations (Even if not as many as SmartTrack suggests).
The people who say Metrolinx can't squeeze all the services into the corridor, do not realize what is safely truly possible after a lot of efficiency improvements on a fully-grade-separated, extremely-well-signalled, extremely-well-operated electrified corridor. The georgetown corridor has great potential to reach this impressive European/Japanese status after a few more billion, and manage to successfully simultaneously squeeze GO RER (equals SmartTrack), express Kitchener gotrains (which may eventually be high speed Kitchener trains in the future), and UPX trains, onto just four tracks (two per direction), if tweaked accordingly.
There are going to be some challenges for sure, but nothing that several billion dollars can't eventually solve.
Things become a lot easier if you cancel UPX and/or SmartTrack-like infill stations, but, assuming, of course, they keep all these elements, they are going to have to run the corridor to highly modern commuter train standards closer to many well-operated European countries (e.g. France). It does not even have to be legendary Japanese train-slotting super high precision, but that, too, can further help if the corridor eventually runs high speed trains in conjunction with all the other services.
