Branching
S-Bahns and similar systems have two defining features. One has been hashed to death on this blog: they reuse legacy rail lines, allowing urban rapid transit to extend arbitrarily deep into suburbia. The other, common also to many other transit technologies, is that they branch extensively, allowing them to run many services on the outer ends, where there’s no demand for rapid transit frequency, while interlining to produce high frequency in the center, where there is.
Since branching is a service planning decision independent of technology, any technology could branch. The branching-friendliest technology is subway-surface: the central subway segment has higher capacity measured in trains per hour than the outer surface segments, and this requires branching. For examples, consider the Boston Green Line, Muni Metro, the Frankfurt U-Bahn, and SEPTA’s Subway-Surface lines. However, even when the entire line is rapid transit, branching is useful to ensure higher service where there is higher demand, and infrastructure improvements will typically focus on boosting capacity in the center. For example, the RER A has moving-block signaling allowing 30 peak-direction trains per hour in the center, but fixed-block signaling on the branches, which do not need such capacity.
Even when rapid transit is built separate from both light rail and mainline rail, branching is useful for lines going into the suburbs or even outer-urban neighborhoods. This is practiced in both New York and London, both of which have extensive branching. Observe further that in both cities, the lines reaching farthest out – the A in the Queens-bound direction and the Metropolitan line in the west – are also the most highly branched.
It’s the opposite situation that is weird. When lines do not branch, there must be a strong outer anchor, or else trains need to run empty outside the center. The alternative is short-turns, and if there’s no space for this, the resulting service patterns can be awkward. Shanghai, which has little branching, runs Line 2 in two segments, a central segment with higher frequency and longer trains and an eastern one with lower frequency and shorter trains; trains do not run through. Beijing has a similar awkwardness with the split between Line 1 and the Batong Line, and Toronto has a split between the Bloor-Danforth line and the technologically incompatible Scarborough rapid transit. (The Sheppard line suffers from the same problem today, but it has the excuse that it was planned to continue west to the Spadina subway rather than stub-ending at Yonge.) Paris has little branching on the Métro as well, but the Metro only serves inner parts of the metro area, many lines have strong outer anchors (for example, La Défense on Line 1), and two others providing some of the farthest-out service branch. The RER branches much more heavily, as befits a suburban system. Tokyo has little branching on the subway proper, but the subway is for the most part inner-urban, and lines continue to the suburbs along commuter lines, which do branch.
In North America, this configuration has been common across a variety of new-build systems, especially ones that should have been S-Bahns. BART does this the most extensively, but the Washington Metro is also highly branched for its size, MARTA branches, the light rail systems branch once more than one line is built, and so on. BART in particular imitated the service planning aspect of commuter rail perfectly, and is an S-Bahn in all but the cost of extending the system further.
The problem with any branching is that it reduces frequency on the branches, potentially scaring away ridership. When a single rapid transit line splits in two it’s rarely a problem, and when city-center service splits into suburban services even more is easy to justify. I think the main issue in urban or inner-suburban cases is that with typical rapid transit frequencies (3-minute peak service or slightly better, a peak-to-base ratio of 2:1 or somewhat less) the trunk has about 5-minute off-peak service, and if it splits into two branches, this means 10-minute service on the branches. If the branching occurs early enough that dense neighborhoods with short-distance travel demand are on branches, it may be too little. In addition, if one branch has much more demand than the other, then it’s usually hard to match frequency on each branch to demand, since it requires trains to be unevenly spaced.
The issue is that branch frequency, 10-15 minutes, is in the transition zone between urban show-up-and-go frequency, where schedules do not matter, and suburban frequency, where they do. It’s perhaps less relevant in small cities with small enough transit systems that even 10-minute service is considered very good, but in large cities, people expect more, creating somewhat of an inner-urban metro envy effect.
That said, 10-minute suburban and outer-urban service can be done clockface, making the average wait much smaller. It is done on the RER A in the midday off-peak, with three 10-minute branches, and could be done with two 10-minute branches quite easily. Likewise, it could be done for 15-minute branches (the RER B already does this); the two A branches in New York have close to 15-minute frequency each, and if New York City Transit’s service planning considered it as a factor instead of focusing more on headway management it could ensure predictable schedules at Ozone Park and the Rockaways.
Pedestrian Observations 
If I understand branching correctly, the Chicago Transit Authority has a single branch line: the Purple Line. Sometimes, the Purple Line becomes the Purple Line Express and runs not only on its branch between the suburb of Wilmette, through Evanston to the northern boundary of Chicago, but along the Red and Brown Lines (during rush hours) to the Loop.
Chicago has much more branching than that, because the two-track Loop hosts 5-6 different services, counting the two Green Line branches separately.
I wouldn’t consider the shared services on the Loop something part of branching as there is likely extremely low demand to move from one part of the Loop to the other (which, with so many services makes for very high frequency). What is convenient about the offerings is that if you recall which direction around the Loop the services travel you can optimize your trip to take advantage of that over an “easier” transfer that may take a few minutes longer.
You might not consider it branching, but the planner who designs the service definitely has to. If you want to have clockface scheduling (and I don’t know the L well enough to know if this would even be useful), you need to treat the system as a giant branched network – there can’t be conflicts once a train enters the Loop.
Segregation of the Blue Line and Red Line in their own subways, separate from the Loop, simplifies the problem of service planning, but the flat junction of the Brown Line tangles the Red Line schedule up with the Loop schedule. I can see why the CTA has been itching to grade-separate that….
Great post Alon, as always, but it wasn’t the one I was looking for tonight! 😉
If a branch becomes too lopsided in relate to others, there is usually the alternative of converting that branch to a shuttle. If the branch ROW allows relatively predictable operations, then the branch shuttle schedule can be designed to match services in what is to become a “main line”.
Like the Parisian Metro routes 3bis and 7bis (née 7), the latter one having already seen one swap between mainline and shuttle.
The Metropolitan line in London is a good example. The outer reaches publish a timetable of the low frequencies. The inner sections just operate on a turn-up-and-go basis.
A way to mitigate frequency degradation is by splitting the train itself at branches – front car goes to the right, rear car goes to the left, or whatever. I’ve only seen this on the defunct NWP suburban rail line in Marin, but I’m sure other agencies still do it.
This would be most useful in places where train capacity isn’t as much of an issue as is frequency and in places where timing is impeccable.
Airports seem to be prone to Flügelung, as they “need” very frequent service but they are seldom able to fill long trains. Hamburg sends one third of its S1 to the airport every 10min, Munich’s S1 (20min interval) is split in Neufahrn for Freising and (AFAIK the longer part) the Airport.
Splitting is common in the UK and Swiss train networks.
“Very common”? There is zero in-service passenger train splitting/joining in Switzerland.
Really? There was gobs of train splitting/joining in Switzerland until quite recently. Are you sure you aren’t talking out your ass?
I am not sure there are zero splitting services, but those splitting services where heavily re-shuffled in 2006-7. Belgium and Netherlands also reduced by 2/3 or so their trains with split operations, substituting them from planned transfers.
Dearest Nathanael,
Yes, I’m sure.
Enjoy your gobs.
@Richard
http://de.wikipedia.org/wiki/Fl%C3%BCgelung#Schweiz
Nathanael will enjoy his gobs.
There may be one line where splitting/joining is done on a regular basis (although they may now have through train and shuttle, with the option to switch the destination of the through train. And there may be one other place where it happens with some trains. Adding or reducing additonal EMUs or “modules” at some stops occurs frequently, but that’s not splitting/joining.
Unless that just changed, my train most certainly split on my way to Neuchatel.
Your train to Neuchâtel comes from where?
That is a very bad idea. Splitting cars take time, and they automatically propagate delays.
It also requires some redundant workforce, at least one extra driver travelling as dead weight for the “joint” sector of the line.
The time taken can be reduced by technology. (Splitting seems to happen during an ordinary dwell time on the northern part of the Shinkansen, where it is common.) The extra driver can board at the splitting point to reduce dead weight time, though if a new line intends to split trains this is a strong reason to consider driverless operation. Schedule adherence is very important so that half-trains reach their merge point at the same time, but if frequency is low enough to justify splitting then schedule adherence is very important anyway.
Combining trains is a big source of delay for Amtrak’s Empire Builder, since the Portland and Seattle sections often don’t arrive on time. But then again, that’s Amtrak.
Splitting and combining is very, very tech-dependent. If you have fully automatic couplings, it can be done, well, fully automatically. Most trains don’t.
In addition to what Anon256 said, there’s train-splitting on the Shonan-Shinjuku Line: trains have 15 cars in the central and southern segments, but at the northern end 5 cars turn short while the other 10 continue to the end.
On busy stations, there are a serious of issues with platform occupancy and track interference when you start having many split services. That is much less of a problem when you just cut the “tail” of a train off at some point.
Nowadays, splitting/joining is done primarily with E/DMUs with automatic couplers, which gets the joined set operational again within a minute or so. The example mentioned elsewhere (München S-Bahn in Neufahrn is set up so that the first train arrives about 2 minutes before the second train, and the time allowed for joining is about 3 minutes. Also, the schedule is set up in the way that the trains meet a little bit outside Neufahrn, which means that one driver takes over the set left behind and brings it back (where it will become the leading unit. In the example, the driver’s operation diagram is: coming from München, driving through to the Airport; then Airport to Neufahrn, where he has the second unit, coupling to the firt unit, leave the train. Then taking over the next train to Freising (because that was the second unit of the train from München, driving to Freising; then driving back, now in the first unit in Neufahrn, driving through towards München… in other words, a simple and straightforward operation pattern.
A question of definition: would you call the operation with the München S-Bahn “branching”, where 8 lines are funnelled through the city tunnel, or the 7 overlapping lines between Zürich Hardbrücke and Stadelhofen?
I would call it branching, yeah. At least with rail, I’m going to stay neutral about branding questions; what I think is the important bit is the layout of providing high-frequency trunk service by interlining branches.
So basically the entire point of this post is “the way that most transit systems do things already is pretty good.” Dat’s some optimism right there.
I will note, however, that branching – when applied to bus routes – ABSOLUTELY REQUIRES A DIFFERENT NUMBER. Among people I know who have tried the bus but don’t ride it regularly, uncertainty about the final destination is a huge deterrent. Houston METRO took a huge step in this direction when they consolidated the 82 (which used to have four different endpoints) into two branches, then assigned each a different number. Likewise, if you look at a track map of DC, the line to Largo is a “branch,” but on the color map it’s just the blue line. No one is ever going to be like “wait, how the fuck did I end up in New Carrolton?”
FYI RE DC Metro: that used to be the case, but there’ve been some changes recently; there are now some Orange Line trains that run to Largo during rush hour. Therefore, there may in fact be a few people who will be like “wait, how the fuck did I end up in New Carrollton?”
More trains were needed on the Orange Line because of crowding on the Vienna “branch” of that line; however, because of capacity constraints in a tunnel under the Potomac river that the Orange Line shares with the Blue Line, Orange Line trains could only be added by removing some Blue Line slots in the tunnel. The additional trains weren’t needed on the New Carrollton branch of the Orange Line– and service reductions were definitely not warranted on the Largo branch of the Blue Line– so those additional Orange Line trains were re-routed to Largo.
Service reductions weren’t warranted on the Springfield-Franconia branch of the Blue Line either; however, because the Blue Line also shares a secondary trunk line with the Yellow Line south of the Potomac, some additional Yellow Line trains were re-routed to Franconia-Springfield. There is a good amount of extra capacity on the trunk line that the Yellow Line shares with the Green Line in its own Green/Yellow trunk downtown, but there isn’t a lot of demand for those additional Yellow line trains (and some turn-back and capacity problems further up the Green Line), so Yellow Line trains turn-back at different locations at different times of the day, creating additional confusion.
I think Alon is absolutely right about interlining and branches, but DC may not always be a good example for that. That said, Washington has long wanted to disentangle the trunk lines that have served us so well for the last 25 years; however, any new Blue or Yellow Line services in the city’s core will be insanely and obscenely expensive. Perhaps the Swiss or the Spaniards can do things better… but we can’t.
Cheaper options are considered from time to time (like new wyes below Rosslyn and above the Pentagon that could serve intra-Virginia travel), and there are definitely some electronic fixes that can improve capacity under the Potomac, but considering the District’s relationship with it’s Congressional Overlords, I don’t think any money will be available for anything– whether it be organization, electronics or concrete– unless we promise to divorce all of our gay-married couples, reanimate aborted children (provided it can be done without medical marijuana, stem cells, or gay parents), rename 16th Street “Ronald Reagan Boulevard”, and continue to pay taxes without representation in Congress; or secede. So never, basically.
One big advantage of joining is improving headways. Joining two short trains to make a long one means one less train in the crowded city core.
Another advantage is better use of an expensive piece of infrastructure. A cross river or cross bay tunnel is expensive to duplicate one bore to increase capacity. A recent proposal was to substitute for the cancelled ARC project by doubling the trains.
The LIRR and NJTransit regularly run 12 car trains. The longest platforms in Penn Station are 18 or 16 cars long depending on who you ask. Wanna run 24 car trains into Penn Station you have to rebuild Penn Station.
> you have to rebuild Penn Station.
That is a challenge. But building extra twelve-car platforms is most likely cheaper than an extra tunnel to Long Island. There is a “yard” west of 9th Avenue that could have platforms,
but would really need a west side subway (9th or 10th) which is a way off:
http://secondavenuesagas.com/2011/11/28/a-far-west-side-rebirth-but-without-a-key-subway-station/
Trains could split again after exiting the Tunnel, so the Plattforms at Penn don’t have to be so long. Or trains could stop twice on the same platform. Not ideal, but workable.
sounds like a great idea. How many switches does this double long train block while it’s loading of unloading passengers? Why stop twice, everyone could just walk forward through the train to the platform! why bother to go to a platform at all, just carve enough space out of the tunnels have everybody leap down to the moving sidewalk they install. Shouldn’t take more than 15-20 minutes for the moving sidewalk to make from 12th Ave over to 8th.
I don’t think your proposal is very realistic. I should think better before offering such ridiculous ideas.
But seriously, your whole argument is a complete red herring, because you point out that the longest trains that Penn can accommodate is 16 cars, which is much longer than the longest trains. Clearly, two 8 car trains could be coupled together to better utilize the tunnel capacity, using the existing infrastructure.
On few track in the center of the station. the rest of the platforms are 12 cars of shorter.
Some of the platforms could be lengthened without too much trouble. More could be lengthened with a little bit more trouble. Lengthening the southernmost platforms was part of ARC Alt G and Gateway
…piece of cake to move the stanchions holding up 7th Ave, 8th Ave and the subways running under them…. Or the stanchions holding up the station and Madison Square Garden. Something they could do over a holiday weekend….
Lengthening the platforms would require no changes between the outer edges of 7th and 8th Ave. You might have to move a few columns under the Farley Post Office.
Widening the platforms, on the other hand, which would likely do wonders for passenger circulation and dwell times, would require significant modification of the column structure.
The portals for the East River tunnels are just east of 7th Ave. The portals for the North River tunnels are at Tenth. Go ahead, move things around without tearing down Madison Square Garden and the Post Office.
You don’t suppose I’ve been using that map for reference as well?
You lengthen platforms to the West. The center 3 platforms need no lengthening. The LIRR platforms don’t need access to the tunnels and the southernmost platforms can be lengthened to the east too if need be (yes, you do need to underpin the subway but that’s nothing new). Yes, I already acknowledged that it requires moving some columns under the Post Office … not that much more complicated underpinning is anything new to transit construction in New York. Exactly how many columns you need to relocate depends on how many platforms you want to lengthen and by how much.
Again, much of this was part of ARC Alt. G and is now part of Gateway anyway.