The Invention of Bad Railroad Timetables
The rail advocate Shaul Picker has uploaded a fascinating potpourri of studies regarding commuter rail operations. Among them, two deserve highlight, because they cover the invention of bad timetable practices in New York, and, unfortunately, not only think those practices are good, but also view their goodness as self-evident. They are both by Donald Eisele, who was working for the New York Central and implemented this system on the lines that are now Metro-North, first introducing the concept to the literature in 1968, and then in 1978 asserting, on flimsy evidence, that it worked. Having implemented it in 1964 based on a similar implementation a few years earlier in the Bay Area, Eisele must be viewed as one of the people most responsible for the poor quality of American mainline service, and his idea of zone theory or zonal operations must be discarded in favor of the S-Bahn takt.
Eisele’s starting point is that commuter rail service should be exclusively about connecting the suburbs with city center. He contrasts his approach with urban transit, which is about service from everywhere to everywhere; trips short of Manhattan were 20% of single-trip ticket revenue for New York Central suburban operations and 5% of multi-ride pass revenue, and the railroad wanted to eliminate this traffic and focus on suburb-to-city commuters. From this inauspicious starting point, he implemented a timetable in which suburban stations are grouped into zones of a few contiguous stations each, typically 2-4 stations. At rush hour, a train only stops within one zone, and then expresses to city center, which in the original case means Grand Central.
The idea behind zone theory is that, since all that matters is a rapid connection to city center, trains should make as few stops as possible. Instead of trying to run frequently, it’s sufficient to run every 20 or 30 minutes, and then once a train fills with seats it should run express. This is accompanied by a view that longer-haul commuters are more important because they pay higher fares, and therefore their trips should not be slowed by the addition of stops closer in.
It’s important to note that what zone theory replaced was not an S-Bahn-style schedule in which all trains make all stops, and if there’s more demand in the inner area than the outer area then some trains should short-turn at a major station in the middle. American railroads had accumulated a cruft of timetables; Eisele goes over how haphazard the traditional schedules were, with short but irregular rush-hour intervals as some trains skipped some stations, never in any systematic way.
The first paper goes over various implementation details. For example, ideally a major station should be the innermost station within its zone, to guarantee passengers there a nonstop trip to city center. Moreover, considerable attention goes to fare collection: fares are realigned away from a purely distance-based system to one in which all stations in a zone have the same fare to city center, simplifying the conductors’ job. The followup paper speaks of the success of this realignment in reducing fare collection mistakes.
The failure of zone theory
We can see today that zone theory is a complete failure. Trains do not meaningfully serve anyone except 9-to-5 suburban commuters to the city, a class that is steadily shrinking due to job sprawl and a change in middle-class working hours. Ridership is horrendous: all three New York commuter railroads combined have less ridership than the Munich S-Bahn, a single-trunk, seven-branch system in a metropolitan area of 3 million. Metro-North would brag about having an 80% market share among rush hour commuters from its suburban shed to Manhattan, but that only amounts to about 90 million annual riders. In contrast, the modal split of rail at major suburban job centers, even ones that are adjacent to the train station like White Plains and Stamford, is single-digit percent – and Metro-North is the least bad of New York’s three railroads in this category.
Even on the original idea of providing fast service from the suburbs to city center, zone theory is a failure. The timetables are not robust to small disturbances, and once the line gets busy enough, the schedules have to be padded considerably. I do not have precise present-day speed zones for Metro-North, but I do have them for the LIRR courtesy of Patrick O’Hara, and LIRR Main Line service is padded 30% over the technical travel time of present-day equipment on present-day tracks. A textbook I have recently read about scheduling best practices cites a range of different padding factors, all single-digit percent; Switzerland uses 7%, on a complex, interlined network where reliability matters above all other concerns. With 30% padding, the LIRR’s nonstop trains between Ronkonkoma and Penn Station, a distance of 80 km, take about as long as local trains would with 7% padding.
Eisele is right in the papers when he complains about the institutional inertia leading to haphazard schedules. But his solution was destructive, especially in contrast with contemporary advances in scheduling in Europe, which implemented the all-day clockface schedule, starting with Spoorslag ’70 and then the Munich S-Bahn takt in 1972.
Zone theory and reliability
The first paper claims as self-evident that zonal timetables are reliable. The argument offered is that if there is a short delay, it only affects trains within that zone, and thus only affects the stations within the zone and does not propagate further. There is no attempt at modeling this, just claims based on common sense – and transport is a field where intuition often fails and scientific analysis is required.
The problem is that zone theory does not actually make trains in different zones independent of one another. The second paper has a sample timetable on PDF-p. 4 for the evening rush hour, and this can also be reversed for the morning. In the morning, trains from outer zones arrive in city center just after trains from inner zones; in the afternoon, trains serving outer zone stations depart city center first, always with a gap of just a few minutes between successive trains. In the morning, a delay in a suburban zone means that the trains in the zones behind it are delayed as well, because otherwise they would clash and arrive city center at literally the same minute, which is impossible.
This isn’t purely an artifact of short headways between running trains. Subway systems routinely have to deal with this issue. The key is that on a subway system, trains do not have much of their own identity; if a train is delayed, the next train can perfectly substitute for it, and cascading delays just mean that trains run slightly slower and (because the equipment pool is fixed) are slightly more crowded. The principle that individual suburban stations should only be served every 20-30 minutes means no such substitution is possible. S-Bahn trains are not as interchangeable as subway trains, which is why they cannot run as frequently, but they still manage to run every 2-3 minutes with 7% padding, even if they can’t reach the limit values of a train very roughly 1.5 minutes achieved by some big city subways.
Eisele did not think this through and therefore made an assertion based on intuition that failed: reliability did not improve, and with long-term deterioration of speed and lack of reduction in operating expenses, the express timetables at this point are slower than an all-stops S-Bahn would be.
How many zones do you need before it’s considered zonal expresses? Because I think it’s quite common in Europe to have all-stop inner regional trains that terminate at a major suburban station, and express trains that become all-stop trains after that. Zürich has this with S6 and S7 (one overlapping transfer stop at Meilen). The half hour frequencies are supplemented by S20 in rush hour, which creates a 15 minute express pattern, but has a slightly different stopping pattern. The 7 skipped stops save about 10 minutes.
They also do this with S8, S24 and S2, S25 and IR/RE trains. Those have a few overlapping stops to transfer. These trains have 30 minute frequencies as well. There are some more examples, and they want to roll it out across the system with 15 minute frequencies: https://www.zvv.ch/zvv/en/about-us/projects/s-bahn-2g.html
The Netherlands has some “Intercity” trains that become local trains after some point where the all-stop train turns short, like with Enkhuizen and Hoorn trains to Amsterdam. It’s a 30 minute frequency as well.
With these examples, you could have a 15 minute frequency in the inner part, at a cost of 10 minutes extra travel time from outer stations, and 10 minutes extra operating costs (which can be anything depending on current turning times)
Is this bad practice, or is it acceptable if you have highest demand for city centre, run a regular schedule with limited padding, and have overlapping transfer stations with timed same/cross-platform transfers?
The crucial difference is that in the European zone-express system, the outermost stop of the local and the innermost stop of the zone express are at the same station, thus the passengers who don’t travel to the city can connect between these trains. Ideally, this connection would take a very short time, but in practice that’s often not feasible. (Often the local is too fast, and throwing so many infill stations at it that it would take long enough would be stupid in some fashion. This is not a problem, an extra trainpath for long-distance service can go through this “hole” at the zone border station.) One way to look at the zone express system is as a relative of the timed overtake, except with very unbalanced travel demand.
Elaboration: the “default” alternative is to have a single local going all the way (let’s assume a 30min takt). This, being slower than long-distance/intercity trains, either clogs up the system (slowing down the ICs), requires a timed overtake (adding several minutes to the short-distance passengers’ travel time), or an extra (pair of) track(s) for 1-3 interstations to keep the local moving while getting overtaken (the worst case is if the ovetake falls around a station that are also served by some long-distance trains). Furthermore, the occupancy of the local is very lopsided. Assuming for the sake of the argument that it’s an EMU in multiple traction, at the cost of just a few additional drivers, the vehicle fleet serving this local could just as well be divided to serve the combination of short-turn local plus the zone express (with a few trains being left over) without any part exceeding capacity. Since from the outer segment, in fact more passengers want to go to city center than to another stop on the inner segment, forcing a cross-, or even same-platform transfer on the latter to give the former a faster ride is overall worth it.
S5/S14 is similar, just with the difference that S14 serves one stop beyond the transfer station on another line.
Also note that the SBB style padding is not everything just before the central station, but at stations about 15 minutes or so apart. This ensures that the trains arrive more or less on time in the innermost core, and cause minimum disturbance there. The Swiss network is that much interlocked, that a delay can ripple through the whole country, and often hard to be caught up.
The Swiss system being so on time once even led to serious discussion after a power outage that overlapping load spikes caused by an above average number of trains all starting or climbing at the same time might have caused it. It was found out that it didn’t, but the fact that experts thought it plausible for a while shows just how “Swiss clockwork” like the system works…
If I remember correctly, the effect was in context with the bigger number of locomotives with three-phase drives, which have a different reactive current characteristic than conventional locomotives.
Isn’t it similar to Commuter Express trains in Japan, which have been working great? Although, certainly, there are still abundant amount of regular trains and trains with different stop patterns on Japanese commuter railway network, and frequencies at each stops are still vastly higher
Those express/limited express trains work great because they are complemented by local trains and semi-expresses. Those local services act as feeders for the expresses. Everything is EMUs with taut scheduling for timed overtakes and in Tokyo lots of dedicated 4 track sections. They are usually off-peak around 3-5 times per hour for just the expresses (e.g. Tobu Tojo, Joban). Its harder at the very edges of the megacities (Tobu in Gunma or Meitetsu in Gifu) but those locations are more 60km from city centre while LIRR or Metra services have low infill, low frequency often DMU “commuter rail” within 10km of city centre. Sure there are commuters who want rapid service straight to city centre that’s some service, but its too limited to make best use of the infrastructure.
Thing is Japan did have a milder version of this pre-privatization because JNR neglected urban services, on a lot of the mainlines feeding into the major cities there was a substantial increases in frequency and substantial station infill (Sanyo line especially which has added 20 stations since privatization). But they knew what good urban rail looked like because their competitors Kintetsu/Hankyu and Tobu already did it. Who will Metra and LIRR learn from?
“Everything is EMUs with taut scheduling for timed overtakes”
Maybe LIRR’s problem is that it doesn’t do that. Not the zone system.
By building a long third track it’s baked the current operating model into the concrete. But then again this is a railroad that thought not reverse peak service was acceptable.
Hamburg S-Bahn mixes third rail and overhead wire electrification. It’s not ideal, but it is possible.
Herbert, Nilo is not referring to a third-rail method of power, but a railroad widening of a dual track to triple track.
The triple track will be used for Manhattan bound expresses in the morning peak, and LI bound expresses in the evening.
That is what solidifies the operating model, though with conveniently placed switches you maybe could re-purpose the track as a series of overtake tracks for expresses in either direction.
All 3 tracks will use 3rd rail for power, as all other electric tracks operated by the LIRR.
(There are no DMUs in either New York or Chicago; Chicago uses diesel locomotives outside Metra Electric, the LIRR and Metro-North use EMUs except for outer diesel tails with craptastic dual-mode locomotives, NJ Transit uses electric locomotives or sometimes EMUs and recently also craptastic dual-mode locomotives.)
Whoops got confused, I, Nilo and Max Wyss were talking EMU/DMU vs Locomotives yesterday.
@Nilo The third track is not the main problem. Lots of Japanese urban rail systems manage with just two tracks and relief platforms. Though an easy comparison isn’t easy given so many do subway through-running too. But say Nishitetsu and Keihan with isolated networks have terminal stations whose users daily average is about 1/3 of LIRR’s “weekday ridership”.
It really does baffle me that they didn’t electrify all of the LIRR’s systems in the interwar years, literally the wealthiest region on the planet then and maybe still now. Southern at least electrified the coastal towns south of London in that period even if it was third rail.
@Eric2 The ideal of zoning system contributes to not having this. Japanese operators look to squeeze as much as they can from existing rights of way. That’s why they developed out of centre shopping/entertainment/tourism assets. Or at least the competent ones…
The LIRR spent most of its time until the ’60s under struggling private ownership.
Once the state took over the MTA, electrification & grade separation took place in earnest, with electrification extended to Ronkonkoma, Huntington & Babylon, with the Babylon line also becoming fully grade separated on elevated track and lengthening to 310m platforms.
It is only in the late ’80s that this stopped and attention was switched to dual-modes. Today on the Port Jefferson Branch which was supposed to be next in line, there is room along the entire line for a double track, the platforms are all 310m long, and there are even electrical substations along the line. They just need to actually install a third rail.
Historically electrification was often driven by a scarcity of coal (or later in East Germany, a scarcity of oil) and not so much the operational benefits. In the Scandinavian countries and the alpine region cheap hydropower also played a role.
To my knowledge the US at that time was the world’s largest coal producer
No, NYC electrification was predicated on two things:
* a massive speedup in journey times for commuters (NY suburban voters are the swing vote in statewide elections)
* direct services to Penn Station, which does not permit diesel services. (East Side Access will be even more restrictive, since it was pretty much designed around the M1 railcar shape.)
Earlier in what was then Germany, electrification was driven by economics—number of trains, steep inclines or whatever. Leipzig – Dessau – Magdeburg – Köthen – Halle – Leipzig and Görlitz – Riesengebirge (Karkonosze) – Breslau (Wrocław) certainly weren’t electrified for a lack of coal. (1938 maps, electrified lines have little crosswise ‘teeth’)
They can put a dual mode into service in a few years without much trouble. Passengers actually get service instead of waiting decades for some idealized perfection.
Why would electrification take decades?
Because it will take that long for the boomers who are obstructing it (like Amtrak management, commuter rail management, and the governors – not NIMBYs or other extra-governmental scourges) to die.
Wouldn’t it be sufficient for boomers who obstruct stuff in their professional capacity to retire from said professions?
China seems to have a problem like that too?
The opposite I would think. China doesn’t really do much regional rail like Japan, partly because they don’t have a large legacy network and partly because they are so subway centric. They just extend subway lines out into the subways without express services.
Sorry “extend subway lines out into the suburbs without express services”.
But eventually a subway line gets too long. That’s one of the arguments in Berlin against further extending U7 which for a time was the longest tunnel transporting humans in the world.
Other arguments against extension are that it would be the second urban rail connection to the airport and Berlin needs other stuff more than that and that crossing the state line creates all sorts of headaches
Yet I keep hearing news keep prompting commuters train, although China Railway aren’t motivated due to low revenue, and the result of those attempts always seems lacking.
Subway centric is a side effect of lack of usable commuter railway, I would say?
And there are also like the old Guangzhou-Shenzhen line which is a not bad alignment for developing into commuter railway?
In China, all conventional railways belong to China Railway Company (aka the State) and there are a lot of regulations, rituals, and bureaucratic inertia surrounding it. Meanwhile, cities have much more control about the planning, building, and operation of the rapid transit.
By the way, most if not all suburban metro lines in China are elevated.
@Frederick that’s kinda the impression I got. Not that that sort of terf war/terf neglect doesn’t happen in Japan or elsewhere. JR East and JR West put a lot of effort into their conventional urban lines because much of the revenue and infrastracture was there before privatisation. JR Central dumped a bunch of perfectly useable urban lines that could of made an orbital system (Aonami line/Aichi loop line) because they thought it was too hard for the effort.
I was reading an article in Japanese recently comparing Shanghai and Tokyo metro and why the latter has double the passenger per km. Emphasizing through-running, integration with conventional rail (which means expresses of course). It also pointed to how much effort was put into 4 tracking the major radial lines and increasing train length. I didn’t know this till writing this but a lot of Beijing subway lines are 6 cars! Tokyo metro is usually 10 and even old deep-tunnel London Underground lines manage 8. They are working on that. But its quite a weird oversight. Not that Japan doesn’t do that (looks at Nagoya subway).
Its interesting that China didn’t learn from Hong Kong.
@borners your comment about Nagoya’s subways struck a chord with me, as whenever I visit the city, not only the shorter train lengths but also the headways always struck me as not befitting a city/metro of its size. 10 min. weekend headways on the Tsurumai or Sakuradori Lines, which is a bigger gap than the weekend services on Sapporo’s lines- a smaller metro area with fewer commercial/retail subcenters. Which may reflect the auto dominance of the region- perhaps people are mainly using public transport only on weekdays for commuting, while leisure travel demands the family car.
Oh boy don’t get me started on Nagoya’s heavy rail mistakes. And you’re right about Tsurumai and Sakuradori being somewhat underutilised. Higashiyama is the opposite which one of the reasons they built the stupid Linimo (I’ve ridden it its stupid) as an “extension”. Whoever runs or ran Aichi and Nagoya’s mass transit policy doesn’t seem to know about the transfer penalty. The lack of Kaneyama-Komaki-Seto through-running tunnel(s) via Sakae is the worst, but not the stupidest, that’s the Sakuradori missing a connection by 500m with the Meitetsu Nagoya main line. The Nagoya eastern suburbs have good connections to Sakae/Nagoya station but getting to other locations in Aichi require 2 transfers and going in the opposite direction via Nagoya station which is pointedly the only truly big hub in Aichi. And then there’s the failed Aichi loop where Aichi owns three seperate railways that are connected to JR Central’s with no connecting service because….reasons?
People say Aichi’s road oriented but I think that misses that they made bad mistakes at pretty much every subway line. I mean Kyoto’s less rail oriented and 100% of their lines have through-operations.
Sapporo is less offensive like maybe they should have through-run the Hakodate mainline? But Sapporo has relatively few heavy rail lines to start with, so a newer radial network was appropriate.
@borners A number of Chinese metro lines were built very early and they have to keep the construction standard of everything low in order to save cost so that they can actually pay for them. And as those lines were built befire economy started developing rapidly, demand was still limited and they deemed it possible to save cost and invrease service level by running short and frequent trains. Experience from intercity trains in China in 1990s also seemed to reflect the success of “frequent short trains” in term of attracting passengers better than “long, infrequent trains”.
True enough especially for the non-megacity level 6-car sets aren’t necessarily wrong. But for the megacities? I guess they didn’t bother learn from anybody else? Doing a brief Moscow, some of the earlier lines were short, but even the oldest is now 8 cars.
@ borners and @Andrew in Ezo
For history of transportation planning in Nagoya and their questionable moves, descriptions in “History of Urban Planning in Nagoya” series might answer some of your question. PDF copies are available for free from Nagoya Urban Planning Institute website:
Regarding Aichi Loop Line, Japanese Wikipedia entry for the line has somewhat detailed description of why the third-sector transition was the best solution in 1980s when JNR and the local governments were deciding on future of the line, including the third-sector transition.
@borners: Last time I checked, Line 1 and Line 2 of Shanghai Metro have 8-car trains. Newer lines, though, have 6 cars.
Seoul is similar; the oldest numbered lines have 10 cars, newer numbered lines and GJ line have 8 cars, and newer named lines and Line 9 run 6-car trains.
Is it because when the older lines opened, the demand was unable to fill the trains, so they opted for a shorter train? Or maybe, newer lines are more peripheral, which warrants using shorter trains?
@Frederick And Beijing is upgrading their busier lines too. And its not like Japan doesn’t make similar calls that turn out to be wrong. Tsukuba express is 6 cars and it is definitely not enough. The reason in that case was the upfront cost, but with Tokyo and Osaka, my guess is that the legacy urban rail through-running reduced tunneled sections as a share of route-length enabling and encouraging longer trains, if you upgrade the legacy lines to 10 cars then you are sending 10 car trains to city centre legacy terminals and through-running subway lines. And the Koreans seem to have learned the lesson well for Seoul.
No. The express trains in Japan make major stops. For examples:
1. Six-to-eight-track section of the Tokaido Main Line: express trains (Tokaido, Yokosuka) make major stops like Yokohama, either Kawasaki or Shin-Kawasaki, Shinagawa, and Shimbashi. Moreover, local trains (Keihin-Tohoku, Yamanote) run extremely frequently.
2. Four-track section of the Chuo Main Line: Chuo Rapid Line trains make major stops like Nakano, Shinjuku, and Yotsuya. Moreover, local trains run extremely frequently.
3. Two-track section of the Chuo Rapid Line beyond Mitaka: at rush hour all trains make all stops, but otherwise some trains have timed overtakes and make major stops like Tachikawa and Kokubunji. Moreover, vanilla rapid trains that do make all rapid stops, including all stops from Mitaka out, run extremely frequently.
This also cascades to the Shinkansen, where all trains make major stops, unlike on the TGV, where Paris-Marseille trains skip Lyon.
In contrast, the New York (and I think also Chicago) system first of all has no local trains to speak of; the New Haven Line’s basic local pattern of trains making all stops to Stamford only runs 4 times between 7 and 10 in the morning inbound, arriving Grand Central 7:04, 7:21, 8:03, 9:11, i.e. with longer gaps at rush hour than off-peak. Then, express trains do not serve major stations – New Rochelle is never served as a major station and therefore has 25-minute gaps at rush hour, 8 out of 20 trains entering Grand Central 8-9 am skip Harlem, 4 out of 12 from past Stamford skip Stamford.
The Chicago Metra does have all-stop (or nearly all-stop) locals on most lines running infrequently all-day. The rush hour trains generally use the zone system, but with the ability to either wait one train or make a closely scheduled transfer to access the entire line.
Metra is also piloting about-hourly bidirectional local-ish trains on many lines starting in 3 days.
The MBTA is doing something like this too, but the Fairmount Line is getting a train every 45 minutes because a line that takes 30 minutes end to end clearly is fine with 45 minutes.
That’s yet another example of American railroad operations being stuck somewhere in the steam age…
Not all Paris – Marseille trains skip Lyon. If I remember correctly, the ones serving Lyon Part Dieu also serve Valence TGV. For the whole route, they take about 20 minutes longer. I can’t remember seeing any Paris Gare de Lyon – Lyon Saint-Exupéry – … – Marseille trains.
I mean, like the old Commuter Rapid service on Tokaido Main Line, which would skip Yokohama and Kawasaki
The old Commuter Rapid on Tokaido Line was there to separate the shorter-distance passengers (ones going to Ofuna and other stations west of it) and longer-distance passengers so that they can reduce overcrowding on rapid trains, which are meant to be for longer-distance passengers. It looks like American “Zone Express” concept, but like Alon mentioned, it’s quite different from American “Zone Express” because JNR/JR East offered Local service concurrently with timed transfer from/to Commuter Rapid trains at Ofuna while being overtaken. If it were operated like American way, people going to Totsuka would have no way traveling from Tokyo unless taking Yokosuka Line instead.
Some Japanese commuter railways offer Commuter-Limited Express/Express/Semi-Express/Suburban-Express/Section-Express/etc. trains skipping larger stations intentionally to even out number of passengers on board and to be picked up on all trains they are operating during the morning peak period because creating popular trains and unpopular trains in peak hours trigger delays (like Tokyu Den-en-Toshi Line having delay issues before Tokyu converted all Express trains in the morning peak hour to Semi-Express trains because Express trains were more popular than Local trains). The most prominent examples are the morning peak period service on Hanshin Main Line (Section Limited Express skips Nishinomiya, a regular Limited Express stop) and the way Seibu operated Ikebukuro Line during morning peak period (all patterns with “Commuter” prefix stop at Local-only stops like Oizumi Gakuen and Hoya while skipping regular Express stops like Hibarigaoka [Commuter Express] or Shakujii Koen [Commuter Semi-Express], the stop served by all types but Limited Express; now-defunct Commuter Rapid stopped only at Oizumi Gakuen while skipping all regular Express stops when it was first introduced even though Rapid is considered as slower type than Express on Seibu lines).
They call it, Chidori stopping pattern?
Speak of which, can the comparative failure of Hong Kong section of High Speed Rail be analyzed under the commuter railway aspect? As, despite the short section of railroad is now an integral part of Chinese High Speed Rail network, it only have about 17 trains a day serving destinations beyond 1-hour range from Hong Kong, while remaining ~80% high speed trains only serve Shenzhen and Dongguan and Guangzhou (pre-pandemic schedule), and is thus in line with the original proposed role of express regional connection within the pearl river delta area. However, many of these short haul high speed trains have strange schedule, and even when they do serve multiple stops, most of them won’t stop at Shenzhen’s most convenient station Futian together with Guangzhou’s main station Guangzhou South, causing demand to spread apart and result in low load factor. It seems like they have focused too much on advertising the benefit of the High Speed Rail as “Arrive Guangzhou in 48 minutes and arrive Shenzhen in 14 minutes”, which is only achievable on non-stop trains, and that make it hard for them to add extra stops at Futian on trains going to Guangzhou which would cost extra time, and then for some reason these short haul trains are run by additional trainsets instead of extending other trains between Guangzhou and Shenzhen down to Hong Kong, probably because the Hong Kong segment is operated by a separate company instead of CR despite it only have 1 single station.
And then, despite the focus of advertise the benefit of getting to Guangzhou in 48 minutes, to use the train it require passengers arriving the station at least 45 minutes before departure time, in order to pick up physical train ticket after making electronic reservation and to clear custom and border check, in addition to the design of both the Hong Kong train station as well as the station at Guangzhou South being unnecessary grandiose which cause extra time to navigate, and also the location of Guangzhou’s South station being away from urban center, location of the train station in Hong Kong being south of the center of population, in addition to security measures being implemented against possibility of terror attack, also seems to cause the rail line traffic underperform (before the pandemic hit).
Or could this be more simply explained by the international nature of such line?
The 2019 protests killed much of the demand between HK and mainland China.
And there was protests over having border security zone in Hong Kong under mainland laws.
For the one year of operation before that, the result wasn’t particularly impressive either.
And cross border traffic through other transportation means are also significantly higher
Several years ago I looked into a recently opened peripheral HSR line in China. At the time it only ran about 5 times a day and I was scratching my head why this expensive infrastructure was so underused. Recently I looked again, it’s up to 30 times per day. So maybe they just prefer to start slow.
1. They are too expensive. The fare is more than double than the other ways of transportation.
2. They are not that fast, once you consider door-to-door times. (Guangzhou South is 20km off the CBD.)
3. As you mentioned, the procedures for riding the train are complicated, so people go back to use the old methods.
Anyways, why do you have to ask an off-topic question here? Doesn’t Hong Kong have vibrant railfan communities in, say, Hong Kong Discuss or hkitalk?
No more off topic than talking about trains in Britain on a thread about Germany.
We need to learn from each other and that includes learning from Asia which includes China as well as South Korea, Taiwan and Japan.
The Chinese have twice as much high speed rail as the rest of the world combined and have probably built more metro lines than the rest of the world combined – it’s just a shame they’ve been so untransparent and have gone so far backwards on human rights.
International comparisons are of course fine but he is talking about HSR on a thread about commuter rail. I’m pretty sure that Alon has blogged about HSR in the past and, at the very least, he could take his question there.
I’m talking about how the line inherited commuter shuttle concept from 1990s and most trains running on the line are of such nature.
@Matthew Hutton Don’t feel too much PRC envy man. Remember Spain built more HSR per capita with more innovative technology and probably lower cost. And Beijing subway lines have lower frequency and shorter trains than London’s older deep lines. Democratic Asia>the greatest kleptocratic autocracy humankind has ever known.
People are allowed to ask questions about trains here. It’s good because it helps everyone learn.
The sad thing about the mess that is Metro-North peak service is that Metro-North’s off-peak schedules (pre-COVID at least) were the closest thing the NYC metropolitan area had to a takt. 30-minute locals (all lines) and 30 minute expresses (hourly on Hudson) with timed transfers at the end of the local zone. All they would have to do for solid peak service was to take their off-peak schedule and up the frequency.
The fact that the 5:15 to Scarsdale is 10 minutes quicker if it skips stops in the Bronx and Eastchester doesn’t save that much time if missing the 5:15 means the next train isn’t until 5:45 and your last call of the day ran 5 minutes over.
Yeah. Thing is, Metro-North management isn’t as stuck in the mud as LIRR management. I can actually imagine them improving the timetables. Part of the timetable design is based on the location of the shops & yards; they keep the terminals of any short-turns near the yards.
We could probably design viable local / limited schedules just with this information and the track / platform information.
In order from Grand Central outward from NYC………
GCT East Yard
BN Yard (MOW)
Mount Vernon West Yard (MOW, also siding for ACE/Endico Foods, served by CSX B-750)
North White Plains Yard
New Haven Line:
East Bridgeport Yard
New Haven Yard
I take it that this scheme also failed to improve the financial burden upon the New York Central Railway?
Isn’t part of NYC financial problem partly related to historic drain of lucrative inner city transit fares toward construction of vehicular traffic bridges, such as Verranzzano-Narrows Bridge? So, increased automobile access killed the golden goose.
NYC did spend an inordinate sum of money on bridges, tunnels, and highways from the 1940s through the 1960s, but it wasn’t funded via transit fares. Robert Moses’ Triborough Bridge and Tunnel Authority was a self-sustaining beast bonding against toll revenues to continue to expand, not to dissimilar from the modern-day tollway authorities in Texas. One of the reasons the state created the MTA was to dismantle Moses’ fiefdom by redirecting toll revenues towards propping up commuter rail, subways, and buses.
What NYC should do now, and I know motorists would hate it, is raise the bridge tolls and parking fees. I used to drive commercial trucks to deliver into NYC. Probably raising truck restrictions and tolls is not such good idea. Commercial traffic in middle of the night is good for refurbishing city retail business. But, single passenger (driver only) automobiles should definitely be charged premium rates to bring their space wasting contraption into the city.
NYC is planning just that by implementing a congestion pricing scheme in 2023!
Today, I took bus €1.40 to outskirts of town to Home Depot style store to buy bricks and concrete, that cost €15-. However, taxi changed policy regarding return €12 trip I would make because cargo is “dirty”. So, I had to return cargo and refund my credit card because big box store would charge €39- to deliver this cargo two days later.
At my local Obi you can rent a cargo bike for a 50€ deposit…
It looks to me that Eisele’s methodology was just what such types had in that era (Beeching in the UK comes to my mind). It is a complete narrow-mindedness. Eisele seemed to look only at the revenue side, but not at the cost (we all know that peak operation is the most expensive operation by far). I guess people with social science would also call Eisele’s action as extremely racist.
Zones (for fares) per se are not bad. Any Tarifverbund works with zones. BUT, those zones include all operations within, and contribute very much to the success of the system.
Also interesting that, instead of optimising the payment system, Eisele insisted on conductors. In that era, urban operators elsewhere started POP, first only for passes, later on also for individual tickets. FWIW, in Zürich, the trailer of a two-car tram train was dedicated to POP, and the throne for the conductor ripped out and replaced with seats. Then, the same happened to the motor car.
But then, also think about that only a few years later, München started their Verkehrsverbund.
Unless a rail line has strategic benefit like the far north line in Scotland which goes to a nuclear power station I think it is seriously debatable as to whether any railway that can’t support an hourly service should be closed.
No one would choose to go by train on such a line when they can drive.
How do you define “support an hourly service”?
I’m not sure. But keep open branch lines with all their stations getting 5000 passengers a year seems pretty pointless – why not have a connecting working bus service?
5.000 passengers a year is send-a taxi level of demand.
If you bustitute that, you get much lower ridership and people won’t do long distance trips by train from those stations but instead drive all the way…
5.000 a YEAR there aren’t enough people to need a bus. To keep the arithmetic real simple 5,200 people a year is one 50 passenger bus, once a week. 50 passengers departing and 50 passengers arriving. 49 roundtrips a week would average out to 7 a day. You need a taxicab.
Ok let’s take the Atlantic coast line in Cornwall – https://en.m.wikipedia.org/wiki/Atlantic_Coast_Line,_Cornwall. The busiest station, Newquay has 100k passengers a year, the others have a max of 5.5k.
A train from Newquay to Bodmin parkway takes just over an hour, runs every 2 hours and the line is 20 miles long. Personally I’d shut it, maybe do some sort of taxi thing for the intermediate stations and run an express bus from Bodmin Parkway to Newquay which can be driven in 30 minutes so would be a little slower than that – but hopefully which could meet every train to Bodmin Parkway. The express bus would be clearly better and you’d get to shut 20 miles of railway with its associated costs.
There were I’m sure lots of lines like that shut by Beeching that would see a handful of trains a day and be pretty much pointless – as well as a few such as the Alexandria palace line and I believe the Newcastle Sunderland line that probably shouldn’t have been closed.
While zone theory is a failure, what Eisele advocated is at least a partial remedy to the vilest disease ever known to rail scheduling — the skip-stop pattern. Refer to “Example B” in page 9 of Eisele’s 1968 article for an illustration of the skip-stop pattern. You can also read the article on Wikipedia: https://de.wikipedia.org/wiki/Skip-stop
Anybody with some knowledge about passenger railway will agree that the skip-stop pattern is utter rubbish. That’s why no country on Earth uses such scheduling “technique”, except for the exceptionally bad US of A.
Sadly, Eisele’s zone theory does not succeed in exorcising the skip-stop demon from Yankeeland. Just look at the current weekday schedule of Caltrain: https://www.caltrain.com/schedules/weekdaytimetable.html
What’s so bad about skip-stop? It seems like a good option for lines with lots of close stops and 3-5 minute headways (more frequent than that and the two skip-stop variants have signalling conflicts, less frequent and the headways on skipped stops are too infrequent).
It assumes that both your origin and destination happen to be either both As or both Bs, else you get a linear transfer. This means that a lot of trips are likely not being done (or are massively inconvenienced) due to said linear transfers. This problem can be fixed by a high percentage of AB stops, but at then you’re defeating the point.
Skip-stop relies on the idea that most trips contain a select few stops (variable suburb to fixed downtown), meaning that demand is very polarised. This idea is generally considered outdated since it does not work for a growing number of casual trips, and areas with less polarised demand.
Not to mention that skip-stop branding is often very cumbersome and is quite hard to figure out.
Early 20th century subways frequently have stops very close together (like 500m). When there is no parallel express line, this makes long trips very slow. If 1/3 of stops are As, 1/3 Bs, 1/3 ABs, then skip-stop decreases the overall stop penalty by 33%, a pretty significant gain. Not only downtown but also major neighborhood stops will be ABs. Since stops are so close, it is often possible to make an A-A or B-B ride that is only slightly longer than the A-B ride which requires a transfer. And if you do need to transfer, it adds 3-5 minutes which is not a big deal for most trips – for a long trip you’ll make that back from the skip stops.
In more recent builds where stops are 1km or more apart, I agree skip-stop makes much less sense (no matter what the headways or population distribution).
I mean you’ve outlined a very limited use case but I’m not sure where in the real world has:
– short stop spacing
– very long lines
– no parallel express routes
– heavily uneven land-use distribution (such that only a few stops have the sort of demand to be ABs)
The average stop spacing on all German subways is less than 1000m and with good reason. Every transit trip starts as a walking trip and if you have to walk a kilometer to your subway stop, the time saved is consumed walking…
To say nothing of accessibility concerns and an aging society with more and more mobility impairments
It would be completely insane for the early 20c subway system where I live or where I lived until 2.5 years ago to do such a thing, because these are both tight meshes with many interchange points and shorter average trip distance than you’d intuit.
@fjod: You see, “heavily uneven land-use distribution” is why the skip-stop was invented in the US. And in the real world there is the NYC Subway which fits the use case, but they are already moving away from the practice.
@Eric2: Not everybody has their destination to be right next to the station. If the destination is 1 km perpendicular to station A, and if you now have to walk 1 km down the line from a station B to station A, that will be 2 km of walking. Also, 3-5 minutes increase of trip time is large; for a 30-minute trip, it is a 10% increase of trip time.
Anyways, I guess we can all agree that, for the case of suburban commuter rail, the skip-stop pattern is totally, utterly, completely stupid.
@fjod: this case is so rare in NYC that it only exists in one location anymore on the subway, the Jamaica Line (J/Z), which on the eastern skip-stop portion is overwhelmingly residential. And it only exists because significant engineering works would be required either to consolidate stations, or the more traditional local approach of adding a third peak-direction express track, and the MTA can barely keep the lights on.
And even in that very limited use case the skip stop service runs for one hour before reverting to an all-local pattern.
I’m not even sure the J/Z train fits the bill on the land-use factor. It runs through uniform mid-low density mixed-use (tending residential) development from Jamaica to Broadway Junction, and then uniform mid-high density mixed-use development from there to the East River. There does not seem to me to be a reason why Crescent Street or Alabama Avenue needs to be an AB stop but others do not. Nor does there seem to be a reason why suburb-to-suburb journeys are particularly unlikely, especially those starting/ending at the denser western section.
If you have stops 500 meters apart, for many trips the time spent getting up/down to the level of a subway or el becomes non-negligible so that for some cases trams can actually offer faster door to door trip times – especially if you have to rely on elevators…
@fjod: some examples that meet these criteria include NYC’s J/Z, Philadelphia’s Market Frankford Line, and several Chicago North Side lines. Not surprisingly these have all had skip stop at some point.
@Herbert Germany has relatively small cities (in both population and physical size), which means that overall line length is short and travel times are more reasonable even with closely spaced stops.
When it opened in its current length in the eighties Berlin’s U7 was the longest tunnel in the world
It is said that, if all stops on a train/metro line are distributed into type A/B/C equally, and you have 3 stopping patterns, aka AB/AC/BC, then all trains can skip one third of all stops on the line, and all passengers can reach their destination with 0 transfer albeit a bit longer waiting time, and trains can also be soaced more closely together since each stations always have 1/3 of all trains skipping it and can thus increase capacity, yet no rail line in real world that I know of actually implemented such system completely.
@Eric2 – Market Frankford Line I think fits only the ‘no parallel express routes’ criterion. It’s 25 minutes all-stops from the outermost stop to 13th St, so quite short by transit line standards. Average stop spacing north from 2nd St is 900m so not short. Land-use distribution is fully unfitting of the criterion, with Tioga to 2nd Street being uniform density and the outer few stops being a little less dense (yes people are disproportionately travelling to/from downtown but that’s the same across almost every radial line worldwide).
Fine if you want to get people to a city centre, not much use to link local centres together.
Note that both A and B patterns would stop at higher ridership stops (aka AB stops). So local centers would in fact be served well.
What if you – for example – wish to take a trip that is local to the local center? So to take a Berlin example, an U7 trip internal to Spandau or, you take a Nuremberg example, an U1 trip internal to Fürth?
Skip-stop pattern might more sense for busy buses lines with bus stop packets?
The major bus stops (ie connection mass transit line) gets AB stop.
There is trade-off between having frequent stops which reduces walking distance and stop consolidation which reduce stop penalty.
Skip-stop pattern buses can have both at expense of frequency but this trade off might worth it because very frequent buses also have frequent bus bunching.
Ideally you won’t have all that many buses serving your inner city subway nodes as they’re already well served by rail, including for most connecting trips. Certainly there are many places where, with one or two exceptions, outlying subway stops tend to have more bus lines terminating there as they distribute people to where a subway isn’t (yet) built and might never be… Or maybe that is just Nuremberg making the wrong mode choice, idk
In the UK it’s overlapping services on the same line. In the London commuter system, most lines work with the outer most routes are local services until they get to major town, then often become semi express , stopping at major towns until closer to London where they are mostly express through the suburbs. Then another service will start a local stopping service from that outer commuter town making local stops through smaller towns before becoming semi express through closer in suburbs stopping at major inner suburb interchanges. Finally a local stopping service will run through the London built up area.
Nearly all minor stations will have a half hourly all day service, Busier stations will have overlapping services that stop at the same station so frequency will increase to 4, 6, 8 10 etc trains an hour as more services serve the station. Some routes have 4 trains an hour running the same stops.
Other lines have long distance trains overlaid over the top.
As a long time user of the Windsor lines out of London Waterloo, changing trains at Staines, Feltham, Twickenham, Richmond, Putney or Clapham to swap between routes to access different stations became 2nd nature.
Yeah the overlapping services are quite tedious and over complicated.
Not in the age of smartphones. An app can easily show the optimal trip for you.
* a working smartphone (one may not own one, one’s may have died, left it at home, etc.)
* reliable cell service where you are
* reliable information on the app
and even then, it really means nothing if your train is cancelled, or you show up late and now need to arrange additional transport but your specific service pattern is not available for a good amount of time. A turn-up-and-go simple schedule is still a much better customer experience.
To say nothing of the elderly who are disproportionately more likely to ride transit and disproportionately less likely to own a smartphone or have a bunch of apps on it if they do own one
Do you have survey data on this? In the US, they only track work trips, but the use of public transport peaks for young adults and is somewhat lower for near-retirement-age people.
The standard practice timetable (off-peak) for the major private railways in Japan provides reasonable diversity of services without undue complication and relatively easy user-friendliness for the semi-regular/regular rider- this “pattern diagram” is a regular *repeating* interval of services, ranging from a simple local-express pairing (e.g. the Seibu Shinjuku line @Kami-Shakujii) repeated 6 x/hour to more diverse service patterns such as a rapid express – local – express – rapid express – local – semi express – express “impulse” over 20 min., repeating 3x/hour (Odakyu Odawara Main Line @Shimokitazawa). If you miss the desired train, you just wait 10 min. to catch the next train with the same stopping pattern (except the semi-express above, which repeats every 20 min.)
That public transit in the “mode of last resort” places is basically only used by pupils, retirees, unemployed people and those at the margins of society is a widespread common assumption in Germany. Whether it is actually true, idk, but anecdotally there are indeed lots of old people riding buses…
https://www.nordbayern.de/1.11203393 the city of Erlangen and the adjacent Landkreis are planning a bus redesign. Isn’t that just about in your wheelhouse? Maybe you can advise and/or critique them…
About the excessive padding on the schedule, not all railroads around the world can operate with consistent amount of dwell time and schedule pad all day long. If passenger volume becomes high, dwell time needs to be longer because it takes more time to let people go though the doors at each station stop. Also, once train count per hour per direction goes beyond the certain level, trains start speed restrictions due to less favored signal aspect/speed code. In some cases, operators schedule express trains with fewer or no mid-line overtakes during peak hours to draw more people to slower or local trains, and this increase the trip time or additional pad assuming pad is amount of time added in the schedule from the best-effort, physically-possible fastest run time.
The case similar to what you mentioned for LIRR happens some commuter railways in Tokyo. For example, inbound Express and Semi Express trains on Keio Line arriving at Shinjuku in the peak hour is 10 to 15 minutes slower than these operating in early morning in Chofu to Shinjuku segment because with train count (28 trains per hour per direction on a double-track railway) and passenger traffic volume (60,000+ passengers traveling through the segment from Shimo-Takaido and Meidai-mae, the busiest one, in the busiest one hour; it takes time to load and unload passengers on jam-packed train), trains go stop-and-go like cars in the freeway traffic jam even though they upgraded signal system to “single-step ATC” system with very short block length and use both platform faces at 4-track mid-line stations . Having only 3 stub-end tracks at Shinjuku Station (end of line) adds more time as trains line up to wait for turn as it approaches to Shinjuku Station (the station cannot be expanded unless almost all of existing Odakyu/JR East/subways/station plaza developments in the Shinjuku West Gate area are demolished).
With 10 to 15 minutes of additional time, these peak-hour Semi Express trains are actually slower than some Local trains in early morning in Chofu-Shinjuku segment, just like your LIRR example, even though Keio’s Semi Express makes 7 fewer stops than Local.
Here is a cab-view video of one of the peak-hour Semi Express trains going stop-and-go:
I’m not sure how bad peak hour crowding is at LIRR, I guess train count during peak hours makes trains slower (= more padding) during the peak period since Systra said in Caltrain’s signal headway report that “their published peak-period scheduled running times (especially between Jamaica and Manhattan) reflect the expected level of signal delays”: