Shinkansen, technical specs, timetabling and margins

[tnbb]

I've been trying to scour the internet for this info and can't find anything, I have an interest in timetabling and rail operations and have always been in awe of the Tokaido Shinkansen with its 16 tph and different stopping patterns. What really gets me though is how Hikari and Kodama services are released into the mainline traffic with relatively tiny gaps.

There seem to be instances where Hikari and Kodama services are released into a 6 minute window between two Nozomi trains. Since it takes even the N700 (with it's very quick acceleration for a high speed train) a couple of minutes to get up to line speed it would seem that these slower services would have to depart a station moments after (approx 1 min) a fast service passes though in order to maintain the 3 min headways,

Ultimately what question boils down to this:
Are there provisions for Shinkansen trains, that are accelerating from a stop, to do so with less than what appears to be a standard 3 min headway?
Is 3 min the minimum headway or is this just a timetabling choice and the real minimum headway is shorter?
What are the junction margins for the shinkansen, they appear to be well below the 2/2.5 min that is standard for the UK network?
Does anyone know of any publicly accessible technical documents in English that could shed more light on this, I can find timetables but not much else

 

[Wandering Pom]

You might find the Japanese Railway Society is a good place to ask: http://www.japaneserailwaysociety.com/
Their website has a series of articles on different aspects of Japanese railways. They don't seem to have a forum as such, but there's a Facebook group, and they have e-mail addresses for contacts.

 

[edwin_m]

HS2 were also planning to drop a stopping train into a gap of 6min or not much more, at Birmingham Interchange. There was a short paper from Prof McNaughton somewhere on the net describing how it could be done, which I imagine is also what is done in Japan.

The turnout speed was (and probably still is) 230km/h and the platform loops were long enough that trains joined and left the main line at that speed. Shinkansens have more motored axles than HS2 units, so may be able to accelerate in a shorter distance. The technical headway on the route was actually less than 3min, so a train joining or leaving on that speed didn't delay the train behind (a train at lower speed would need a shorter headway anyway). Timings would probably have to be to the second, and there was also some kind of advisory system to optimise speed and acceleration, so for example a joining train didn't have to slow down on the approach to the junction due to being too close to the one in front, and end up joining at a lower speed. I block sections around the junction were also shorter than elsewhere for a further headway improvement.

 

[etr221]

The East Japan Railway Culture Foundation (https://www.ejrcf.or.jp/english/index.html) may have something useful, in particualar the Japan Railway & Transport Review - while no longer published, I think it's all available on line at that hey

 

[Route115?]

On the Tokaido Shinkansen trains are timetabled at headways of 3 minutes. Trains can pass at most stations and the slower Kodama services are overtaken at many stops. It will then run two paths (six minutes) behind - or three paths behind if overtaken by two trains which happens. I attach pdfs for the 2010 TT - you can download the current timetable at https://global.jr-central.co.jp/en/info/timetable/ Extra trains are run on bank holidays.

I have also attached an excel spreadsheet I knocked up for 18 March 2020 (just before covid) westbound. Maybe someone has a timetable graph they could share.

It works but requires split second accuracy. Only in Japan.

 

[KK2109]

The signaling system Tokaido Shinkansen currently use is called ATC-NS, which uses track circuit to detect the location of trains with an on-broad computer processing trackside data to determine the braking curve. It is not quite moving blocks, but with short track circuit sections, it can have a very similar result. To achieve minimum headways and improve passenger comfort, ATC-NS included the concept of one-step braking curve. In the graph, the bold line shows the target speed and the thin line shows the actual speed of the train.


Operationally, despite Tokaido Shinkansen seems to have various stopping patterns, it is a clockface timetable with four stopping patterns:
1. Nozomi (Fast): Tokyo, Shinagawa, Shin-Yokohama, Nagoya, Kyoto, Shin-Osaka
2. Hikari 6xx (Semi-Fast): Tokyo, Shinagawa, Shin-Yokohama, Odawara or Toyohashi, then all stations from Nagoya
3. Hikari 5xx (Semi-Fast): Tokyo, Shinagawa, Shin-Yokohama, Atami/Mishima/Shizuoka/Hamamatsu (2or3 out of 4), Nagoya, Kyoto, Shin-Osaka
4. Kodama (Stopping): All stations en-route

This timetable pattern is designed around the infrastructure of the line, with all station except Atami have passing loops, the stopping Kodama trains can be overtake by fast Nozomi trains behind it at every station. This also shows on the timing of Hikari Semi-Fast services, despite they stops at different stations every hour, the overall timing between Shin-Yokohama and Nagoya for each pattern is the same. The fast Nozomi trains can overtake them at different stations every hour.

The minimum headway of the line is about 2mins 15seconds, however the trains are timed at a minimum of 3mins as to protect the reliability of the line with some buffer times.

The junction margins on the Tokaido Shinkansen are about 2mins, this is the headway in which a following train can pass the points at full speed after the previous train went into the loop. However, there are nearly no conflicting movements on the line, except at Tokyo and Shin-Osaka. At Shin-Osaka, most trains run through towards Sanyo Shinkansen and terminating trains can turnaround in the sidings west of the station, so it is less of a problem. On the other hand, at Tokyo, all 6 platforms are terminus platforms. It is timetabled to take a maximum of 19 trains per hour, between 18:00-19:00 when all extra trains are running. (12 Nozomi+2 Hikari+3 Kodama+2 ECS moves)

This is part of the Tokaido Shinkansen 2022 timetable's train graph, it shows that Nozomi trains are often flight in pairs, dash lines are extras that do not run every day.
The whole graph can be found on this Website(In Japanese).


One thing worth noticing is that in recent years JR Central has spent significant effort on improving the performance of its trains, from upgrading the whole N700 fleet to N700A to improve acceleration and include cruise control, to introducing even more efficient N700S series. Having a uniform high-performance fleet really helps them to run this highspeed line to the maximum frequency on minimum headways.

Unfortunately, most technical documents are in Japanese, maybe in the future more will be released as both JR Central and JR East are looking to export their technology overseas.

Hope this answers most of your questions.

 

[edwin_m]

The signaling system Tokaido Shinkansen currently use is called ATC-NS, which uses track circuit to detect the location of trains with an on-broad computer processing trackside data to determine the braking curve. It is not quite moving blocks, but with short track circuit sections, it can have a very similar result. To achieve minimum headways and improve passenger comfort, ATC-NS included the concept of one-step braking curve. In the graph, the bold line shows the target speed and the thin line shows the actual speed of the train.
View attachment 161810

Operationally, despite Tokaido Shinkansen seems to have various stopping patterns, it is a clockface timetable with four stopping patterns:
1. Nozomi (Fast): Tokyo, Shinagawa, Shin-Yokohama, Nagoya, Kyoto, Shin-Osaka
2. Hikari 6xx (Semi-Fast): Tokyo, Shinagawa, Shin-Yokohama, Odawara or Toyohashi, then all stations from Nagoya
3. Hikari 5xx (Semi-Fast): Tokyo, Shinagawa, Shin-Yokohama, Atami/Mishima/Shizuoka/Hamamatsu (2or3 out of 4), Nagoya, Kyoto, Shin-Osaka
4. Kodama (Stopping): All stations en-route

This timetable pattern is designed around the infrastructure of the line, with all station except Atami have passing loops, the stopping Kodama trains can be overtake by fast Nozomi trains behind it at every station. This also shows on the timing of Hikari Semi-Fast services, despite they stops at different stations every hour, the overall timing between Shin-Yokohama and Nagoya for each pattern is the same. The fast Nozomi trains can overtake them at different stations every hour.

The minimum headway of the line is about 2mins 15seconds, however the trains are timed at a minimum of 3mins as to protect the reliability of the line with some buffer times.

The junction margins on the Tokaido Shinkansen are about 2mins, this is the headway in which a following train can pass the points at full speed after the previous train went into the loop. However, there are nearly no conflicting movements on the line, except at Tokyo and Shin-Osaka. At Shin-Osaka, most trains run through towards Sanyo Shinkansen and terminating trains can turnaround in the sidings west of the station, so it is less of a problem. On the other hand, at Tokyo, all 6 platforms are terminus platforms. It is timetabled to take a maximum of 19 trains per hour, between 18:00-19:00 when all extra trains are running. (12 Nozomi+2 Hikari+3 Kodama+2 ECS moves)

This is part of the Tokaido Shinkansen 2022 timetable's train graph, it shows that Nozomi trains are often flight in pairs, dash lines are extras that do not run every day.
The whole graph can be found on this Website(In Japanese).
View attachment 161814

One thing worth noticing is that in recent years JR Central has spent significant effort on improving the performance of its trains, from upgrading the whole N700 fleet to N700A to improve acceleration and include cruise control, to introducing even more efficient N700S series. Having a uniform high-performance fleet really helps them to run this highspeed line to the maximum frequency on minimum headways.

Unfortunately, most technical documents are in Japanese, maybe in the future more will be released as both JR Central and JR East are looking to export their technology overseas.

Hope this answers most of your questions.

That's really interesting. Would you happen to know the turnout speed for the station loops?

 

[KK2109]

That's really interesting. Would you happen to know the turnout speed for the station loops?

IIRC, the turnout speed is 70km/h for station loops.

 

[TheWierdOne]

IIRC, the turnout speed is 70km/h for station loops.

That seems surprisingly slow for a high speed rail line. Presumably the loops are fairly short and the acceleration of the trains makes up for it?

 

[tbwbear]

This might not answer your questions but it is a lot of fun to watch -

It is taken at Kakegawa after a period of extreme disruption (3 hour delays) with the trains passing each other and stopping all out of pattern. Lots of shots of trains entering the loop and being passed before they have even stopped.

The red text of the title is pointing to the Kodama and says "this hasn't stopped yet, and the yellow text is pointing to the Nozomi and says "this is already passing"

Watch at 5:50 - quite interesting ! The commentator is asking himself "is this really happening?"

Towards the end (42:00) the footage is from a different day and (by using the shinkansen speed check app) shows the speeds dropping significantly as the Nozomi that the cameraman is travelling on follows the Kodama in front.

 

[HSTEd]

That seems surprisingly slow for a high speed rail line. Presumably the loops are fairly short and the acceleration of the trains makes up for it?

My understanding is that Japanese operators prefer extreme reliability to high turnout speed - slower speed turnouts are easier to maintain and more reliable, so there we are.

Indeed they tend to go with very simple layouts, as can be seen at the Tokyo terminus of the Tohoku Shinkansen (youtube video). Four platforms and only six point ends, including an unswitched diamond - operating at ~15tph in the peaks.

If we take Shiroishi-Zaō Station on the Tohoku Shinkasen (as) an arbitrarily chosen example, the loop length on Google Earth (turnout to turnout) is only ~800m for a ~400m platform.

 

[eldomtom2]

Watch at 5:50 - quite interesting !

Would such a movement (train still moving in block as second train enters the block immediately after) be permitted in the UK?

 

[MisterT]

Watch at 5:50 - quite interesting ! The commentator is asking himself "is this really happening?"

I don't speak Japanese, but I don't see anything strange or special happening around that time? Is this the right timestamp?

Around 5:25, the video shows a train pulling into the station and stopping, then there are some cuts in the footage, and we see the train pulling away from the station, while another train is coming to a stop on the middle track to wait for the first train to have left the station.
I don't know how much footage was cut and I don't know what the minimum dwell time of the Shinkansen is exactly (I didn't pay attention to that on my last visit to Japan), but I assume it's around the 1 minute mark if there aren't many passengers, which would mean that, including stopping and pulling away again, the time between those trains would have been around 90 seconds, maybe 2 minutes, which is not that strange at all with a modern, short block length safety system in this situation at slower speeds - even with our old ATB safety system here in the Netherlands we are able to do this at certain places where the track lay-out and signal placement allows this.

Would such a movement (train still moving in block as second train enters the block immediately after) be permitted in the UK?

I don't see any train entering the same block there? Just a train on the middle track coming to a stop before the signal, and waiting until the train on the outer platform track has left the block before it?

I'm pretty confused about these comments, am I missing something here?

 

[Gerald Fiennes]

My own observation (but I have never timed it) is that the 'secret' is the fantascially high acceleration/deacceleration of the trains at the smaller stations (most are now the N700S). Essentially the same as tube stock (well, they are very light weight...) It felt like the trains rejoining the mainline were doing so at about 2/3rds of line speed and then regained full line speed about 30 seconds later.

As well as this, it is important to recall in Japan the trains leave mega-punctually and the passengers know that once the 'hustle alarm' goes they have just a few seconds to board, and do so. Everyone knows what they doing and just does it. Thus the dwell times are tiny by UK standards (had the pleasure of seeing c. 4 platform staff wave bats and whistles and things for c. 90 seconds at York station before the doors closed during the despatch of a Northbound LNER the other day, quite a different operating philosophy!)

Many, but certainly not all, platforms also have platform gates that close as well. The high approach speed of the Shinkansen at intermediate platforms is something that the UK would find somewhat uncomfortable, I think!

 

[eldomtom2]

I don't see any train entering the same block there? Just a train on the middle track coming to a stop before the signal, and waiting until the train on the outer platform track has left the block before it?

I didn't say a train was entering the same block. I was wondering if the UK would allow to a train to be still moving while a second train is let out immediately in front of it.

 

[MisterT]

Apologies, I have misread your post.
But why wouldn't that be allowed? I would actually be suprised if it wasn't. I don't think trains in the UK wait for junctions with an extra seperation block either? Honestly, that would be pretty weird I'd say.
Both trains moving at the same time in blocks next to each other is an everyday occurrence. The same happens when, for example, two trains are moving at block distance from each other and the latter one is waiting for their signal to clear.
If I'm behind another train, I ususally 'crawl' towards the signal, waiting for it to clear. That is the same thing that can be seen in the video, with the only difference in the video being that the train in front is coming onto the main line, but the result for the train behind it is the same.

Even two moving trains within the same block is not that weird, although I can't speak for the UK. Many countries (including the Netherlands) allow that, even though it wouldn't be a common occurrence. I think I did that twice in the past ten years or so, but it happens.

 

[eldomtom2]

Apologies, I have misread your post.
But why wouldn't that be allowed? I would actually be suprised if it wasn't. I don't think trains in the UK wait for junctions with an extra seperation block either? Honestly, that would be pretty weird I'd say.

I'm not an expert on signalling, but I thought there were rules regarding overlap.

 

[tbwbear]

What is happening at 5.25 is very unusual. In normal operation the Kodama train would stay in the loop until the Nozomi train has passed and it would then be released after it. What is happening here of course is the Kodama is being released onto the main line and the Nozomi made to wait on the fast tracks.

This almost never happens and I haven't seen it on film before. I once experienced, again during time of extreme disruption, being on a Kodama which was being held in the platform road at Mikawa Anjo whilst a Nozomi was also stopped on the through tracks alongside, but generally the faster trains are given priority.

The digital ATC system is working perfectly normally though, in stopping the Nozomi ahead of the junction. What makes it seem strange might be the lack of signals - if you imagine a red stop signal at the place where the Nozomi eventually stops, it seems normal to me.

What the video demonstrates is the flexibility in the ATC system, the trains will alter their speed based on the speed of the train in front. You can see this in the way that the Nozomi trains pass through Kakegawa at different speeds.

In normal operation, there has to be an element of the Nozomi trains having to slow down below line speed to allow for the slower trains in front to accelerate or get out of the way. You can see this in the timetable where Nozomi #1 - leaves Tokyo at 6:00 and encountering little traffic ahead of it has a time of 2hr 22 to Shin Osaka, a full 5 minutes faster than subsequent trains. That would suggest 5 minutes built in for a bit of slowing down here and there.

This faster transit time for the first train enables the company to claim a faster headline time for the journey of course, but it only appeared as the line got busier, until the mid 1980s every single Hikari train was timed for 3hr 10.

if you look on the graph for the difference in the speed profile between Nozomi #1 and Nozomi #5 (departing at 6:33, making the same stops, and taking 5 minutes more) it might be interesting. The system is absolutely incredible but certainly there is a sacrifice of speed for extra capacity here.

 

[KK2109]

This almost never happens and I haven't seen it on film before. I once experienced, again during time of extreme disruption, being on a Kodama which was being held in the platform road at Mikawa Anjo whilst a Nozomi was also stopped on the through tracks alongside, but generally the faster trains are given priority.

JR Central's regulation policy priorities the right sequence of trains as much as possible during disruptions. They believe this can help the services recover faster after the disruptions.

 

[Mcv378]

You all might also be interested in the NHK Japan series Japan Railway Journal available on the NHK app and YouTube.

It's not just about the Shinkansen, but many different railway subjects across Japan. It's very interesting.

 

[jfollows]

You all might also be interested in the NHK Japan series Japan Railway Journal available on the NHK app and YouTube.

It's not just about the Shinkansen, but many different railway subjects across Japan. It's very interesting.

“Video unavailable” in both cases for me.

 

[Gag Halfrunt]

If you click on "Watch on YouTube" a new window will open.

Anyway, here is the Japan Railway Journal page on the NHK website. IIRC new episodes always appear there first before being uploaded to YouTube.

Japan Railway Journal | NHK WORLD-JAPAN

Go behind the scenes of Japan's amazing railways. Interesting facts, new technologies, expert tips, plus where to go, how to get there and more!

www3.nhk.or.jp

 

[Mcv378]

“Video unavailable” in both cases for me.

“Video unavailable” in both cases for me.

The YouTube links aren't working on here for some reason.

But search for "Japan Railway Journal Shinkansen" on YouTube, there's quite a few episodes on there.

 

[Gostav]

Shinkansen

Shinkansen route maps in Japan, real-time train animations

www.railaround.com

Here is a good website shows the operation of the trains on the line and is available in English.

 

[Three-Nine]

Regards shinkansen boarding procedures, based on my visit to Japan a couple months ago theres a bit of evidence that the large influx of tourists unfamiliar with them is causing a bit of an issue in maintaining the boarding times, at least at the busier stations like Kyoto. I witnessed a couple of occasions where Japan Railway staff were trying to hurry non-Japanese passengers along and heard an announcement that one shinkansen had been delayed by about five minutes "due to a large number of passengers" (may not have been the exact wording).

This was at the tail-end of the very popular cherry blossom season so may not be entirely representative of how things operate normally though.

 

[Beebman]

There's some major disruption today on the Tokaido Shinkansen due to a collision overnight involving maintenance vehicles. NHK World has this news article in English: https://www3.nhk.or.jp/nhkworld/en/news/20240722_23/

Tokaido Shinkansen bullet train services remain suspended between Nagoya and Hamamatsu stations following an accident involving maintenance vehicles.

The operator, Central Japan Railway Company, says the recovery work will continue until Monday evening and it cannot predict when trains will be able to use the affected section.

Central Japan Railway says a maintenance vehicle was rear-ended between Toyohashi and Mikawa-Anjo stations at around 3:30 a.m. on Monday. Both vehicles were derailed.

The driver of the vehicle that rear-ended the other one suffered a cut to his head and was taken to a hospital. A maintenance worker has a minor leg injury.

The operator is urging travelers not to use the Tokaido Shinkansen line until services are fully restored, to avoid congestion at stations.

Here's a video with aerial pictures of the collision from the channel of the Asahi Shimbun newspaper (no commentary):

Video description (using Google Translate):

At around 3:37 a.m. on the 22nd, two maintenance vehicles collided during nighttime work on the Tokaido Shinkansen line in Gamagori, Aichi Prefecture, and both derailed. The recovery work, including removing the maintenance vehicles that could no longer move, took a long time, and the Tokaido Shinkansen line was suspended from the first train between Hamamatsu and Nagoya. There is no prospect of resuming operations.

Currently YouTube has a large number of video reports posted in the latest few hours from Japanese TV channels showing the effects of the disruption to passengers at stations along the line.

 

[tnbb]

My own observation (but I have never timed it) is that the 'secret' is the fantascially high acceleration/deacceleration of the trains at the smaller stations (most are now the N700S). Essentially the same as tube stock (well, they are very light weight...) It felt like the trains rejoining the mainline were doing so at about 2/3rds of line speed and then regained full line speed about 30 seconds later.

What are the crash standards for these trains like? I know Gareth Dennis has pointed out that we could never have a Shinkansen style HSR in the UK unless we basically ripped up or heavily altered the UK's rules when it comes to rail risk and safety in the UK. I've felt the N700A pull away like a daemon for a high speed train and the N700s is meant to be even better (around 0.72mps^2 I believe?). I remember when I went on them they felt extremally smooth and light but I'm not sure if that was because they are or because I know they are and just though I felt it. I know JR's approach to Shinkansen safety is one of absolutely no conflicts what so ever and they've gone to extreme lengths across infrastructure, signalling, and training, to achive this. Does this allow them to make trains that are far lighter and perform better (all be it worse in a crash) than the European ones?

As well as this, it is important to recall in Japan the trains leave mega-punctually and the passengers know that once the 'hustle alarm' goes they have just a few seconds to board, and do so. Everyone knows what they doing and just does it. Thus the dwell times are tiny by UK standards (had the pleasure of seeing c. 4 platform staff wave bats and whistles and things for c. 90 seconds at York station before the doors closed during the despatch of a Northbound LNER the other day, quite a different operating philosophy!)

The number of times I've seen Brits desperately hitting the door button in some vain hopeful attempt to get on, desperately thinking that somehow it'll be different this time while 3 different station/train crew scream at them to stand back amazes me, mind you, maybe if our trains were as punctual and frequent as some of Japan's people wouldn't get so worked up? Although considering some people still do it on the tube I think it's just a mindset we might have

 

[HSTEd]

I think disciplined boarding and such is more a function of familiarity with the conventions of the system than due to any innate cultural factors.

If a railway is as succesful as the Tokaido Shinkansen, most people who will ever use it will use it fairly regularly (at least ignoring tourists), and will thus learn the relevant etiquette.
You see this, for example, on the London underground and elsewhere.

 

[Beebman]

Shinkansen and other stations often have lines on platforms marking queue positions which line up with the doors. At some non-Shinkansen stations with different types of trains there can be some quite complicated markings, I've attached a photo I took last year at the Keikyu Railway Shinagawa Station in Tokyo:

 

[MatthewHutton]

What are the crash standards for these trains like? I know Gareth Dennis has pointed out that we could never have a Shinkansen style HSR in the UK unless we basically ripped up or heavily altered the UK's rules when it comes to rail risk and safety in the UK. I've felt the N700A pull away like a daemon for a high speed train and the N700s is meant to be even better (around 0.72mps^2 I believe?). I remember when I went on them they felt extremally smooth and light but I'm not sure if that was because they are or because I know they are and just though I felt it. I know JR's approach to Shinkansen safety is one of absolutely no conflicts what so ever and they've gone to extreme lengths across infrastructure, signalling, and training, to achive this. Does this allow them to make trains that are far lighter and perform better (all be it worse in a crash) than the European ones?


The number of times I've seen Brits desperately hitting the door button in some vain hopeful attempt to get on, desperately thinking that somehow it'll be different this time while 3 different station/train crew scream at them to stand back amazes me, mind you, maybe if our trains were as punctual and frequent as some of Japan's people wouldn't get so worked up? Although considering some people still do it on the tube I think it's just a mindset we might have

To answer the first question, there have been no in service non suicide passenger fatalities ever on the Shinkansen. I believe one person has got stuck in the doors at some point and died.

With regards to pressing the door button, firstly there are platform edge doors in Japan so it’s more difficult but secondly the Japanese train would just depart.