[rbt]

Practically, branches should be preferably 2 branches, and possibly 3. As the number of branches increase, the service frequency on the branch becomes too small. One thing that has not been mentioned regarding branches is that they should be less costly than the trunk line.

How would a branch cost less; except for the reduction is rolling stock required? Land prices might be lower but that's location dependent rather than branch dependent (branching Yonge down to Queens Quay would not be cheaper land wise).

* Station size is dictated mostly by fire code (2 full trains unloading simultaneously in an emergency situation with obstructed station exits).
* Tunnel size is dictated by train dimensions.
* Expropriations will be minimized at all costs (GO learned this on the Georgetown corridor).



Most of the operating costs are structural/equipment repair/maintenance/inspection which exists even if you mothball the line. This is why TTC runs 5 minute service at 1am; the cost basically rounds out compared to maintenance overhead.

Manned fare gates are still necessary or some type of fare-fraud prevention. Roaming inspectors in London don't seem to be enough for their gateless stations; so they've been installing manned gates.

I suppose it would be easier to clean.

 

[BurlOak]

How would a branch cost less; except for the reduction is rolling stock required? Land prices might be lower.

Station size is dictated by fire code (2 full trains unloading simultaneously in an emergency situation).
Tunnel size is dictated by train dimensions.
Expropriations will be minimized at all costs (GO learned this on the Georgetown corridor).

London has been learning that all stations need full fare gates or fare-fraud goes through the roof at those gateless access points and you need someone to guard the gates or they get jumped; they get jumped anyway sometimes.


If it's built in the same form (elevated vs. elevated; tunnel vs tunnel), the cost of a branch would be very close to that of the trunk.

Operations might be a bit lower (easier to clean, fewer trains, etc.) but structural maintenance is the vast majority of the expense for subway after it's 30+ years old.

My point was that the branch would not be the same form as the trunk. This can be done by switching from underground to elevated, hydro corridor or median (with LRT). Elevated to elevated is maybe the only exception to this that may work. Tunnel to tunnel is too expensive (to build 2 tunnels for the branches), median to median is not feasible since a median trunk would probably not work for high frequenies and hydro corridor to hydrol probably doesn't exist anywhere. The other one that works would be BRT trunk and mixed traffic buses in the branches - gweed 123 did mention this one.

 

[diminutive]

Manned fare gates are still necessary or some type of fare-fraud prevention. Roaming inspectors in London don't seem to be enough for their gateless stations; so they've been installing manned gates.

Source? There's been a surge of fare evasion in London, but almost all of that is on the bus network. The DLR, whose stations are mostly unmanned and ungated, has quite low levels of fare evasion and they've been stable according to that article. It hardly seems like there is some massive exploitation of the system.

Even in Vancouver, the decision to move to fare-gates has more to do with perceptions than anything. The installation of gates through the system is only supposed to save about ten million dollars a year, which is obviously less than the cost of installation and staffing.

 

[diminutive]

Practically, branches should be preferably 2 branches, and possibly 3. As the number of branches increase, the service frequency on the branch becomes too small. One thing that has not been mentioned regarding branches is that they should be less costly than the trunk line.

Agreed. Unless there is some way to reduce branch costs (you mention them in a subsequent post) branch lines could be bad investments; they would cost the same as the core piece of infrastructure but would have lower capacities.

Another criteria for successful branching could be that core route should be a destination for a substantial portion of rides originating along the branches.

Using the example of Jane->Eglinton, what would be the use? Intuitively, the origin-destination pairs along those routes don't match up. Only significant destination I could think of would be the transfers at Eglinton West and Eglinton. But then why not just bus East along Sheppard or Finch or Steeles?

Globally, it seems that most rail systems with branching are commuter or regional (RER, S-Bahn) basically following a trend of suburban branches->downtown core tunnel.

 

[gweed123]

Agreed. Unless there is some way to reduce branch costs (you mention them in a subsequent post) branch lines could be bad investments; they would cost the same as the core piece of infrastructure but would have lower capacities.

I think that the two biggest benefits of branching come in the form of more convenient trip patterns, and reduced pressure at transfer stations. Take interlining Sheppard with Spadina for example (which I brought up earlier). If Sheppard were to be extended westward, and terminate at Downsview, there would need to be a significant reconfiguration of the station in order to handle all of the transfer passengers (Downsview may not be the best example for this though, because it was designed as a future transfer station, but take a station like Sheppard which had to be jerry-rigged into becoming a transfer station).

If Sheppard is interlined with Spadina, not only does that extra infrastructure requirement in the form of a transfer station disappear, but so does the potential choke point it creates. I mean, if Bloor-Yonge was designed with a wye from WB to SB, and NB to EB, it's very likely that the mess there today wouldn't be nearly as bad, especially if B-D trains could round the loop and then merge with B-D again between St. George and Spadina.

And of course, either of these configurations would be much more user friendly, because let's face it, people hate transferring. I bet that if Sheppard is extended west, but not interlined, that you will still see the majority of Sheppard users transferring at Sheppard-Yonge. But if Sheppard is interlined with Spadina, I would bet that the majority of riders would stay on passed Sheppard-Yonge, and stay on that train to head downtown.

Another criteria for successful branching could be that core route should be a destination for a substantial portion of rides originating along the branches.

Using the example of Jane->Eglinton, what would be the use? Intuitively, the origin-destination pairs along those routes don't match up. Only significant destination I could think of would be the transfers at Eglinton West and Eglinton. But then why not just bus East along Sheppard or Finch or Steeles?

But that branch not only serves long distance patterns, it also serves local patterns too. A lot of people would ONLY use the Jane branch, and not stay on it until it reached Eglinton. Yes, it may not make sense for someone from Jane and Finch to travel all the way down to Eglinton just to turn in and transfer at Eglinton West Station. But if someone is coming from Wilson & Jane, or Lawrence & Jane? I'd bet that down & over via rapid transit would be faster than over with local transit, and then down with rapid transit.

Globally, it seems that most rail systems with branching are commuter or regional (RER, S-Bahn) basically following a trend of suburban branches->downtown core tunnel.

This is very true. For example, look at the S2 and S26 lines. Multiple suburban branches, but they all funnel through a single line through the central city (and in fact use a single central platform at Friedrichstaße). http://www.alice-dsl.net/o.debernardi/ode/ode/root_h/images/mapb/mapU-S bahn.gif

 

[rbt]

My point was that the branch would not be the same form as the trunk. This can be done by switching from underground to elevated,...

I see.

 

[rbt]

Source? There's been a surge of fare evasion in London, but almost all of that is on the bus network. The DLR, whose stations are mostly unmanned and ungated, has quite low levels of fare evasion and they've been stable according to that article. It hardly seems like there is some massive exploitation of the system.

I'm unable to find the list. There were a handful of suburban stations near the end of various lines which received fare-gates where previously they did not exist this was many years ago though (perhaps 2003?).

Sorry, no sources. I don't find searching old underground minutes very easy for something specific.

 

[diminutive]

But that branch not only serves long distance patterns, it also serves local patterns too. A lot of people would ONLY use the Jane branch, and not stay on it until it reached Eglinton. Yes, it may not make sense for someone from Jane and Finch to travel all the way down to Eglinton just to turn in and transfer at Eglinton West Station. But if someone is coming from Wilson & Jane, or Lawrence & Jane? I'd bet that down & over via rapid transit would be faster than over with local transit, and then down with rapid transit.

Local ridership along a given corridor though doesn't really impact whether to build it as a branch or separate line. If someone is going from Finch-> Lawrence along Jane, it doesn't matter that much what happens at Eglinton.

As for whether Jane itself would make a good branch, I don't have the ridership data to make an opinion. My point was more, notionally, just because two lines could be connected into one doesn't mean they should be; the branch-core set up should serve the bulk of the origin-destination pairs along the route.

Also, as to branching reducing the need for and cost of transfer stations, that would also depends partially on local situations. Building a curve and flyover to join two perpendicular lines together could itself be quite a task. 16-20 trains per hour going to Finch and Don Mills should work in terms of capacity.

The one line I'm surprised was never considered for branching was actually the Sheppard line, at least in its stubway form. Almost all of its trips would continue southbound at Sheppard, right?

 

[gweed123]

Local ridership along a given corridor though doesn't really impact whether to build it as a branch or separate line. If someone is going from Finch-> Lawrence along Jane, it doesn't matter that much what happens at Eglinton.

As for whether Jane itself would make a good branch, I don't have the ridership data to make an opinion. My point was more, notionally, just because two lines could be connected into one doesn't mean they should be; the branch-core set up should serve the bulk of the origin-destination pairs along the route.

That's a good point. I don't have the specific data to back it up, but it seems to me like a lot of the people on the Jane bus are taking it to Jane station, and then heading eastbound on the B-D subway. Interlining would remove that transfer, while still maintaining more or less the same trip pattern (assuming their destination is downtown and not a location along Bloor itself).

Also, as to branching reducing the need for and cost of transfer stations, that would also depends partially on local situations. Building a curve and flyover to join two perpendicular lines together could itself be quite a task.

That's true. It isn't just the capital cost though, it's the reduced pressure on the system by eliminating a transfer, as I said before. But yes, it's not applicable to every scenario, but there are quite a few where it would be beneficial.

Also, if the line is designed with that in mind from Day 1, they can build the necessary fly over/under when the line is initially being built, and then just have to "plug in" the branch line when it is built. If the branching is going to happen with Eglinton & Jane for example, they should be taking that into consideration in the tunnel design now.

The one line I'm surprised was never considered for branching was actually the Sheppard line, at least in its stubway form. Almost all of its trips would continue southbound at Sheppard, right?

The track infrastructure is certainly there. I suspect the reasons why it wasn't considered were:

1) The Yonge line is packed as it is, and they need all available trains originating from Finch. Slotting some in at Sheppard would reduce frequency coming into and out of Finch, which could be a problem.

2) Sheppard opened with 4 car trains and shorter platforms. There would have been an added capital as well as operating expense running longer trains and requiring longer platforms when neither of which were really needed.

By contrast, extending Sheppard westward may boost ridership and actually make 6 car trains justifiable. Also, the Spadina line doesn't have the same capacity concerns as the Yonge line does, especially on the future Vaughan extension, which is where the split would occur.

 

[M II A II R II K]

And branches don't have to be evenly used, there could be a higher percentage of trains running on one as opposed to the other.

 

[BurlOak]

The track infrastructure is certainly there. I suspect the reasons why it wasn't considered were:

1) The Yonge line is packed as it is, and they need all available trains originating from Finch. Slotting some in at Sheppard would reduce frequency coming into and out of Finch, which could be a problem.

2) Sheppard opened with 4 car trains and shorter platforms. There would have been an added capital as well as operating expense running longer trains and requiring longer platforms when neither of which were really needed.

By contrast, extending Sheppard westward may boost ridership and actually make 6 car trains justifiable. Also, the Spadina line doesn't have the same capacity concerns as the Yonge line does, especially on the future Vaughan extension, which is where the split would occur.

Can Sheppard run with the existing platforms and 6 (or 7) car trains? Only 4 cars would be line up at the station (with doors opening) and the others would just follow along until you branched onto the Spadina line. On the return north-east bound, passengers can still use the walk-through trains and go from a car they go onto (on the Spadina line) to a car the would actually have doors opening on the Sheppard portion.

 

[gweed123]

Can Sheppard run with the existing platforms and 6 (or 7) car trains? Only 4 cars would be line up at the station (with doors opening) and the others would just follow along until you branched onto the Spadina line. On the return north-east bound, passengers can still use the walk-through trains and go from a car they go onto (on the Spadina line) to a car the would actually have doors opening on the Sheppard portion.

The Sheppard platform are actually built for 6 car trains. What they have done however is left the ends of the platforms unfinished (and in some cases put up a temporary wall), so that the platform length is reduced to 4 cars. There would certainly be an expense to ripping down those walls and finishing the rest of the platforms, but the changes as I understand them would be purely cosmetic.

If an interlining option was chosen, I would imagine the cost of finishing the full platforms would just be added in.

 

[gweed123]

Two potential interlining options. The first is for Toronto, the second is for Ottawa.

 

[diminutive]

And branches don't have to be evenly used, there could be a higher percentage of trains running on one as opposed to the other.

That's true, but branch capacity is zero-sum. So the more heavily capacity is skewed to one branch, the less will be on another. The split doesn't necessarily have to be 50/50, but if you start talking about 70/30 or more it kind of begs the question of why are you paying for an asset you never intend to use more than 30% of? Maybe if specific technical circumstances permit very, very cheap construction, but that's rare.