This is an interesting point. I've never thought about Scarb Junction needing grade-separation, nor heard about it being needed. I guess I don't really get why it would require grade-separation. If Stouff RER/ST is using the northernmost tracks, seems like LSE and VIA would be unaffected from a stop/proceed perspective. Granted I don't know how many tracks are envisioned at full buildout, nor how close these things can run.
If a separation is needed through there, which I can guess it is, it would likely be combined with a road/rail grade-separation @ Danforth Rd. This would be like a +1000m guideway or tunnel. Big bucks, many pitchforks, and seemingly quite complicated (since Stouff is somewhat narrow, the curvature, and all lines are active). Clearly politically unpalatable so maybe that's a reason Mlinx hasn't divulged much about it.
Here's the issue I see. The general spec we are hearing is, 4 tracks west of the Junction, 3 east on LSE, and 2 north to Unionville.
The service pattern at peak spec we are hearing - in each direction - is roughly
10 Unionville trains/hour - 6 ST/local, 4 express
8 local LSE trains/hour - 7.5 min headway
4 Oshawa-Bowmanville LSE express trains/hour - 15 min headway
1-2 VIA trains/hour - even with HFR there will still be local trains, and some will run at peak
A 'flat' junction will handle this volume, perhaps, but will require a lot of use of crossovers. Every time a crossover is used, that second track's route is blocked for 3-4 minutes minimum. A train on a conflicting route will encounter restrictive signals and begin to slow down well ahead of the junction, until the opposing movement has passed. Ten crossing movements at even three minutes' advance clearing = 30 minutes of restriction per hour.
One solution is to use a non-uniform direction of traffic for each track - eg put the eastbound moves to Unionville on the northmost track - but now think through where platforms would have to be located between Union and Scarborough.
A 'flyover' junction - as is common in the Northeast US and in other parts of the world - would be a lot more elegant and would remove much of the crossover conflicts. This would allow a standard slow/fast format to the four tracks with 'slow' (stopping) on the outside. The flyover would be fed from the two southmost tracks, with a short segment of a fifth track so that a local train could clear the "slow" (red) line and stop clear of the feed from the "fast" (green) track so that an overtaking 'express' train could pass through first, without having the local train block the 'slow' line until it has a clear path onto the Stouffville line.
(I'm sure
@steve will instruct us about the high-tech solutions used elsewhere, and frankly many old style signal box operations manage this, but my point is, why design the system with constraining features in the first place and then try to work around with electronics?)
This is all amateur design, but if I were building this in my basement, I would argue for the overpass. The fact that ML has paused with construction south of Kennedy makes me think that they haven't made a firm decision, either.
The biggest point is - a high performing project manager would realise that this question is material and drive to have it confirmed, one way or the other, early on so that work does not pause. With ML, decisions like this get avoided or debated well into critical path, putting the overall project schedule at risk. And, local residents, who never thought that they would see trains passing over their back yards on a flyover, will be told that this is not up for debate - the project is now behind schedule and there is no time for proper consultation without holding things up. #Davenport
- Paul
PS - yeah, my computer graphics skills wouldn't pass Kindergarten, sorry!