Jubilee Line Evening Peak

[Deleted]

How well does the Jubilee Line timetable work during the evening rush hours?

I ask because after looking into it closely, I’m amazed it can work! Running 24 TPH from North Greenwich to Willesden Green is fair enough – but then having trains being reversed at Willesden Green, Wembley Park and Canons Park (to run empty straight into Stanmore depot), with the associated detrainment stands, plus the constraint of having only two reversing platforms at Stanmore – it’s an extremely tight working. I wanted to know how reliably it works in practice – as well as if there were any thoughts about how things could be improved considering what we have got at present – no new platforms, sidings or higher density signalling for a few years yet and the contractual requirement to run 24 TPH tidal flow.

(Of course if there’s another thread that’s already discussed, then please link me there! I did have a search, but didn’t find anything).

The next timetable change is due in May 2007, to coincide with the extended hours change. But this will be used as a chance to rewrite most of the Jubilee Line timetable using a mixture of 15 and 18 TPHs during the off peaks and weekends. Some of that is driven by the expected Wembley stadium traffic (surely it will be open by then!)

[Chris M]

The third platform at Stanmore can't be too far off completion

[jimini]

It works swimmingly well. Don't ask me to quote on the technical side of things, but from a commuter perspective it rarely lets us Wharfers down of an evening. A train every two minutes is the norm.

[towerman]

To the casual observer the service appears to be running well,but most trains are running 6 to 10 minutes late in the evening peak,if there is a delay in excess of 15 minutes for any reason that's the service sh****d till COT.

[Deleted]

As a long suffering Jubilee Line commuter (Baker St-Finchley Rd, and previously Green Park-Willesden Green, and even prior to that Canary Wharf-Green Park), I think things are working much better now with the 24tph than they have when 22tph was being run. Trains are usually pretty regular, but then again that isn't surprising if the line is operating at capacity, as the bottlenecks on the JLE will even out the headways.

However, I've noticed that later in the evening peaks (approx 6pm onwards) trains do tend to queue to get through Finchley Road NB, with slowing or stopping for home signals sometimes reaching as far back as St John's Wood. This is probably thanks to the long dwell times at Finchley Road, due to the cross platform interchange with the Met. The NB starter at Finchley Rd usualy holds the train there for an extended time as well, I'm not too sure if this due to headway control, or just a long signal block ahead (or both). This doesn't help the journey time, and must cause late running?

When I used to commute to Willesden Green, the tipping out times were variable. On a good day the train can be tipped out in approx 65secs. However, I sometimes observed tip outs of over 2 mins, which is not acceptable. Given that the roll out run in time behind a train going into a siding at Willesden Green is approx 1min40secs-2mins, then the headway for a tip out usually varied between 3-4mins. For a line running 24tph, this would require the headways for the next three trains (as 1/4 trains reverse at Willesden Green) to be approx 2mins to reach that 24tph target and prevent late running. This is obviously pushing the line to it's limit!

So whilst the Jubilee Line is waiting for it's new signalling, I think to keep things moving then the following could be done:-
1) Improve tipping out efficiency at Willesden Green (e,g always having 2 platform staff available to assist tipping out).
or
2) Send Willesden Green reversers to Stanmore's third platform (if signalling beyond Wembley Park can support 3trains per 10mins) .
and
3) Stop regulating trains at Finchley Road (if they are already).

Certainly I don't think reducing train frequency is the answer, as the Jubilee Line is at crush loading between Baker St and Finchley Rd during the busiest parts of the peaks (particularly the am peak).

[Deleted]

if you are taking the jub from Baker street to Finchley Road, why not take the Met and have an infinitely more comfortable ride?

Back on topic, how full are trains with such a small headway and how much more frequent will the trains be after re-signalling? also ATO will make a difference when it eventually comes.

[Deleted]

rob said:

if you are taking the jub from Baker street to Finchley Road, why not take the Met and have an infinitely more comfortable ride?

Because I change onto the Jubilee from the Bakerloo. Travelling upstairs to the Met, and waiting for less frequent train, still with no chance of getting a seat would be fruitless.

Back on topic, how full are trains with such a small headway and how much more frequent will the trains be after re-signalling? also ATO will make a difference when it eventually comes.

24tph/2mins 30secs isn't that small a headway for a metro system! The evening peak can be busy on trains that have left Canary Wharf just after 5pm. The morning am is often at crush capacity during the busiest period (often many passengers cannot board between Finchley Rd and Baker Street (inclusive)).

I have heard figures of either 31tph or 33tph for the Jubilee post moving block signalling. I think the former is more realistic given the permitted 40sec dwell at Finchley Road. The JLE extension was originally planned to run 36tph.

[amershamsi]

stephenk said:

I have heard figures of either 31tph or 33tph for the Jubilee post moving block signalling.

the question is - does Stanmore need a large number of those trains? OK, some will terminate early, however that's rather overkill for the outer edge of London, surely? Perhaps throw a branch to Harrow (changing the Met slightly) in order to use some of that spare capacity? also 31tph seems a strange number - a 2 minute frequency would surely make more sense.

[Deleted]

amershamsi said:

stephenk said:

I have heard figures of either 31tph or 33tph for the Jubilee post moving block signalling.

the question is - does Stanmore need a large number of those trains? OK, some will terminate early, however that's rather overkill for the outer edge of London, surely? Perhaps throw a branch to Harrow (changing the Met slightly) in order to use some of that spare capacity? also 31tph seems a strange number - a 2 minute frequency would surely make more sense.

31tph is a train every 1min55secs, 33tph a train every 1min50secs. There is no point in installing moving block signalling, and then not running the line to full potential, so as to run trains at headways that equate to nice round figures like 2mins.

In my opinion, if LU want to run headways below 2mins (as is planned post resignalling on many lines), then services need to be timetabled at 1,5,6, or 10sec accuracy (as is done on every other metro system running 31tph+), not 30secs as at present on LU. When you get down to very tight headways, and the service is regulated by computers, trains should to be timetabled at even intervals. This cannot be done with 30sec accuracy timetabling, or even 15secs (unless you are running 34tph with the latter).

Anyway, I think we are heading a bit off topic.....

[Deleted]

barney said:

You'll never get under two minute headways anywhere unless you also reduce station dwell times also. The pax that hold the doors and generally all try to get in the same door at once will see to that. Saw a geezer the other day dash for a wobbly and tried to have a fight with the platform edge doors. He lost. The language that resulted would have made a sergeant major blush! If they want close headway services then the hoi polloi MUST be made to alight and board as fast as safely possible. Otherwise is Irish offensives.

I would disagree with your statement there. I've observed headways of 1min25secs on the Central during rush hours, and 1min15secs - 1min 35secs on the Northern (around the Claphams). However, these are minimum headways, and obviously timetabled headways need some degree of operating margin on top of those figures.

There are three elements in station headway - the roll out run in time (aka dynamic headway or close in time), the dwell time, and the operating margin. So even if the dwell times stay the same after resignalling, the improved roll out roll in times may allow for timetabled headways below 2mins. The wider doors of the 09TS and S-stock should also help reduce the dwell time, as they have appeared to have done on the 92TS.

[Harsig]

soultrain said:

How well does the Jubilee Line timetable work during the evening rush hours?

I ask because after looking into it closely, I’m amazed it can work! Running 24 TPH from North Greenwich to Willesden Green is fair enough – but then having trains being reversed at Willesden Green, Wembley Park and Canons Park (to run empty straight into Stanmore depot), with the associated detrainment stands, plus the constraint of having only two reversing platforms at Stanmore – it’s an extremely tight working. I wanted to know how reliably it works in practice – as well as if there were any thoughts about how things could be improved considering what we have got at present – no new platforms, sidings or higher density signalling for a few years yet and the contractual requirement to run 24 TPH tidal flow.

Going back to the original question, as someone responsible for signalling of the Jubilee Line between Stanmore and Chring Cross, my impression is that the service works reasonably well during the peak, in fact far better than the later evening service where the entire service goes to Stanmore; the least little hiccup at Stanmore and there is a queue of trains back to Queensbury or beyond.
Our other big bugbear relates to trains entering and leaving service on the shoulders of the peaks. In particular at the end of the evening peak it is quite common for delays to occur at Wembley Park while trains wait to get to depot. There have been cases where there was a stabling train in the southbound platform, having arrived empty from Stanmore sidings, one in the siding waiting to go to depot behind the first train and a third train waiting to go into the siding and then to depot, thus blocking the Jubilee entirely. This sort of situation tends to crop up reasonably reguarly when there is some late running on the northbound Jubilee i.e. the train in the siding should have gone to depot before the one from Stanmore arrived at Wembley. In fact the Jubilee Controllers, of late have taken to terminating some of the Canon's Park trains at Neasden and sending them straight to depot. (These are the trains which otherwise then run empty from Stanmore Sidings to Wembley and Neasden Depot).

From the point of view of improving the timetable the best solution to us in the SCC at Baker St would be to re-arrange matters so that the Canon's Park trains genuinely do stable in Stanmore sidings, with some of the late night stablers there running empty to Neasden depot after the last Stratford train. Similarly in the morning we feel there should be some earlier trains from Neasden depot to Stanmore to allow some of the trains stabled in Stanmore sidings to start much later as the trains which currently run empty from Neasden Depot to Stanmore sidings.

stephenk said:

The NB starter at Finchley Rd usualy holds the train there for an extended time as well, I'm not too sure if this due to headway control, or just a long signal block ahead (or both).

Headway control is almost never applied here. The delay is due to the fact that the signal cannot clear until the train in front has departed West Hampstead platform. Because of the rising gradient approaching Finchley Road trains can approach quite close to a train in Finchley Road platform. Thus when that train leaves the following train arrives in the platform at Finchley Road before the departing train has even arrived at West Hampstead and thus the second train must wait while the first train arrives at West Hampstead, completes its station duties and departs. In the meantime a third train has closed up behind the second train thus repeating the cycle. Of course the delay is even greater if thetrain at West Hampstead terminates there.

[citysig]

awkward42 said:

The third platform at Stanmore can't be too far off completion

Just to add, the third platform will never be controlled by us at Baker Street. The Stanmore area including the new platform will be controlled - following re-signalling and ATO equipping - by Neasden SCC.

Timescale is obviously dependant on politics, money and speed of the project. Depending on who you talk to, it's either 18 months or 4 years.

Work has commenced at most sites, but then I could have made the same statement a year ago.

[Deleted]

harsig said:

stephenk said:

The NB starter at Finchley Rd usualy holds the train there for an extended time as well, I'm not too sure if this due to headway control, or just a long signal block ahead (or both).

Headway control is almost never applied here. The delay is due to the fact that the signal cannot clear until the train in front has departed West Hampstead platform. Because of the rising gradient approaching Finchley Road trains can approach quite close to a train in Finchley Road platform. Thus when that train leaves the following train arrives in the platform at Finchley Road before the departing train has even arrived at West Hampstead and thus the second train must wait while the first train arrives at West Hampstead, completes its station duties and departs. In the meantime a third train has closed up behind the second train thus repeating the cycle. Of course the delay is even greater if thetrain at West Hampstead terminates there.

Thanks for explaining that.

A situation I have observed quite a few times is when a NB Jubilee at Finchley Rd is waiting for the starter to clear, and during that time a Met train arrives. Thus even when the signal has cleared, the Jubilee driver is then having to wait for the interchanging passengers before departing, thus extending the dwell time further. I'm not too sure what, if anything can be done to improve that situation.

[agoodcuppa]

barney said:

A standing train earns no money and more work could be scheduled from crews if running times were reduced.

Much of what you say is true, however if the schedule is as tight as you propose there is then no flexibility to get it running right time again when something untoward occurs.

[Deleted]

There has to be some 'leeway' built in to the running times, for an unexpected event, like a fault on the train. If everything runs to the minimum, then, if there is a problem, it will be made worse by the huge number of trains trying to go through the section.... I mean, West Ruislip to Epping is what... 86 minutes. Years ago this took some 94 minutes, but that was with the old 62ts.... Effectively, you could lop a few more minutes off, but only if the track was upgraded in certain places. This is probably another reason why it is taking longer to do a circle, or any run for that matter.... Staff also have thier side in the time taken to do trips...

[Tom]

barney said:

There have been cases where there was a stabling train in the southbound platform, having arrived empty from Stanmore sidings, one in the siding waiting to go to depot behind the first train and a third train waiting to go into the siding and then to depot, thus blocking the Jubilee entirely.

That is not the fault of the train crews or any operating staff. That is entirely down to timetables office not doing their job properly!

Hmmm, I remember this style of posting from someone else who used to rant about office staff...

Unfortunately the Timetables office don't have a massive (or even a small) crystal ball with which they can predict service disruptions. As Harsig's post explained, this scenario is quite common when late running occurs on the NB service, which could occur for any number of reasons outside the control of the timetables office, including infrastructure faults (InfraCos), staff arriving late (train crew) etc.

While the timetable should be robust it doesn't have to be rocket-proof and claiming the timetables office are at fault looks to me like someone who either doesn't understand how the railway works, or is trying to pass the blame on to someone else.

[Deleted]

agoodcuppa said:

barney said:

A standing train earns no money and more work could be scheduled from crews if running times were reduced.

Much of what you say is true, however if the schedule is as tight as you propose there is then no flexibility to get it running right time again when something untoward occurs.

Decisions in the design and operations of metro systems always have advantages and disadvantages, this is why I find it so fascinating! In this case the tighter the operation, the better the capacity if the line is running well, but also the increased chance of delays caused by problems.

Interestingly in Japan, operations HAD to be precision perfect, because lack of available resources post-war meant that that expensive trains sitting around in terminals were unacceptable. Also very high property prices, and lack of space, meant that most busy commuter line terminals only have space for 2-4 platforms. For example, the outer suburban Chuo Rapid line runs 30tph with 200m trains into a 2 platform terminus - thus trains have to run bang on time! Unfortunately it was this running on time at all costs ethos, that cost the lives of over 100 people in the derailment that occured last year when an inexperienced driver was trying to recover from late running.

[citysig]

barney said:

That is not the fault of the train crews or any operating staff. That is entirely down to timetables office not doing their job properly! Relying too much on computers and not common sense. It seems to be case nowadays of trying to mix the incompatable. 10 minute head ways here, 8 minute headways there, 12 minutes somewhere else. If the headways on each branch and central sections were multiples/divisibles of an hour things would improve.

This 2 hour yardstick was only introduced at the behest of signallers who are only controllers. It does NOT make for good railway operating. A controller/signaller should be able to cope with any timetable set before them. I think that it may not be totally their fault though. Again too much reliance on computers. Who ever decided that circle trains need 56 minutes to go round want's his head testing. 50 minutes is more than ample to do it in. That is the trouble with todays schedules. TOO MUCH time is given for trips. What used to be 80min flat time from end-to-end of the District now seems to be 86 minutes. A standing train earns no money and more work could be scheduled from crews if running times were reduced.

If you don't mind me asking, what are you on - apart from some kind of huge dig. I too, like Tom, am suffering a severe bout of de-ja-vu.

The scenario my colleague gave as an example is one which occurs not because of bad timetabling, but on a multitude of minor problems combining as one larger problem at Wembley. If everything is bang on time then it rarely occurs.

Headways may seem complicated to the outsider, but if you properly study services and have the big pitcure to look at - including passenger numbers, acurate line capacity graphs etc. as well, you will notice an awfully large chunk of common sense.

Signallers and/or controllers do not make decisions on timetabling. We are occasionally called upon to attend meetings to table our own suggestions based upon our experiences - as are train operators, station staff and many other grades.

For your information, I would be able to cope with any timetable set before me. The question of course is whether the infrastructure available to run it on can cope.

Circle Line trains can easily complete a circuit in 48 minutes. The timetable used to have them on 50 minute trips. However, in light of the fact that every other trip they turned up 5-10 late on a good day, someone used their common sense (and not a computer) and added 6 minutes into the timetable. This gave an 8-minute service, which is the same as the H&C. The two services therefore mesh perfectly.

The additional minutes are added at various points along the route, and for example one slot of recovery may be given at Aldgate. Allowing a Circle train an extra minute or two there means they then mesh perfectly with the District service.

Much of the trips that used to run would be pretty un-realistic on todays world. The lack of speed restrictions, passengers and, to be frank, regulations back then made it possible.

I have a feeling you already knew the answer to your questions. As I said, have we been here before?

[Deleted]

citysig said:

barney said:

Circle Line trains can easily complete a circuit in 48 minutes. The timetable used to have them on 50 minute trips. However, in light of the fact that every other trip they turned up 5-10 late on a good day, someone used their common sense (and not a computer) and added 6 minutes into the timetable. This gave an 8-minute service, which is the same as the H&C. The two services therefore mesh perfectly.

The additional minutes are added at various points along the route, and for example one slot of recovery may be given at Aldgate. Allowing a Circle train an extra minute or two there means they then mesh perfectly with the District service.

Personally (from a passenger and enthusiasts point of view) I have found the SSLs to run much more reliably since the headways were increased slightly (from 8 to 8.5min cycles), and extra running time was incorporated in the timetable a few years back. The same goes for the Piccadilly which became much more reliable after the headways were reduced to 24tph. Before the changes the line was considerably prone to bunching. It seems to be that from these observations that if reversing or junction capacity is causing the problems, then allowing slightly more operating margin can help considerably.

[Deleted]

stephenk said:

rob said:

if you are taking the jub from Baker street to Finchley Road, why not take the Met and have an infinitely more comfortable ride?

Because I change onto the Jubilee from the Bakerloo. Travelling upstairs to the Met, and waiting for less frequent train, still with no chance of getting a seat would be fruitless.

Not always fruitless. I often travel from Green Park to Canons Park, and many a time am too lazy to go up the escalator and stairs to change to the faster Met at Baker Street, waiting instead to do so at Finchley Road. However, on a number of occasions, I have done so and as my Met train drew out of Finchley Road, I see the original Jubilee that I was on pulling in. Had I stayed on the Jubilee until Finchley Road, I would have missed my Uxbridge all stations Met...

However, I concede that you need to run pretty fast at Baker Street, and be lucky, to make it worthwhile.

amershamsi said:

stephenk said:

I have heard figures of either 31tph or 33tph for the Jubilee post moving block signalling.

the question is - does Stanmore need a large number of those trains? OK, some will terminate early, however that's rather overkill for the outer edge of London, surely? Perhaps throw a branch to Harrow (changing the Met slightly) in order to use some of that spare capacity? also 31tph seems a strange number - a 2 minute frequency would surely make more sense.

I think that could be a really good idea. Even better could possibly be a connection to the Bakerloo to Harrow and Wealdstone and on to Watford Junction. The Bakerloo from Kenton to Baker Street takes about 31 minutes, whereas the Jubilee from the same point (stopping at Preston Road and all stops between Wembley Park and Baker Street) would take only 24 minutes. With a 2-minute wait for the Met at Wembley Park, it would be possible to get from the following destinations to Baker Street (and from there to the West End) in much better times (and much improved frequencies):

• From Harrow and Wealdstone in 20 minutes
  
• From Hatch End in 25 minutes
  
• From Watford High Street in 33 minutes
  

Of course, the Croxley link would improve times to/from Watford High Street anyway, but wouldn't the above be better value for money (at least from the punter's POV) than extending the Bakerloo to Watford Junction?

[Deleted]

alantan said:

I think that could be a really good idea. Even better could possibly be a connection to the Bakerloo to Harrow and Wealdstone and on to Watford Junction. The Bakerloo from Kenton to Baker Street takes about 31 minutes, whereas the Jubilee from the same point (stopping at Preston Road and all stops between Wembley Park and Baker Street) would take only 24 minutes. With a 2-minute wait for the Met at Wembley Park, it would be possible to get from the following destinations to Baker Street (and from there to the West End) in much better times (and much improved frequencies):

• From Harrow and Wealdstone in 20 minutes
  
• From Hatch End in 25 minutes
  
• From Watford High Street in 33 minutes
  

Of course, the Croxley link would improve times to/from Watford High Street anyway, but wouldn't the above be better value for money (at least from the punter's POV) than extending the Bakerloo to Watford Junction?

Sounds too complicated, for little benefit to me. As I've mentioned before on this forum, the most reliable metro lines, are almost always simple end to end lines with no branches.

[mandgc]

Delays at Finchley Road and Wembley Park.

Regarding the buildup of trains at Finchley Road could West Hampstead reversers continue on to Willesden Green ? (Agreed this would entail extra running time) This would allow trains to clear Finchley Road when station duties are completed.

At Wembley Park, is there a connection to enable Empty trains from Stanmore toaccess the Depot on the Southbound side 'on the level' ?

[Harsig]

mandgc said:

Delays at Finchley Road and Wembley Park.

Regarding the buildup of trains at Finchley Road could West Hampstead reversers continue on to Willesden Green ? (Agreed this would entail extra running time) This would allow trains to clear Finchley Road when station duties are completed.

They do. There are no trains scheduled to reverse at West Hampstead. It is used only for turning late running trains, usuallly the Willesden Green Reversers.

At Wembley Park, is there a connection to enable Empty trains from Stanmore toaccess the Depot on the Southbound side 'on the level' ?

There is but the problem is that at the time the Jubilee trains are stabling, many Met trains are doing so also and even if there weren't, I've had situations where that route was not available because it was blocked by a southbound Met train with a clear signal that was not moving because it was awaiting a crew relief.

[Deleted]

barney said:

You'd think timetables would make signallers job easier by arranging the after peak stablers to be at Stanmore or SMD and leave the Neasden stables until later when it is less busy at Wembley park.

Similarly the Met after-peak stables could be Rickmansworth, Uxbridge ect. Would give the cleaners/fitters at outlying places more time to the job as there are generally only one at each point

Don't forget that some trains have to be serviced or have checks done overnight. To allow more time for these to happen, it makes sense for the trains that are booked for servicing and checks to go into the depot after the evening peak. The trains that only need work doing that takes a shorter amount of time, such as cleaning, should go into the stabling sidings at the end of the working day.

[mandgc]

Re - Delays at W. Hampstead and W,Park.

So - what are we to do ??

[citysig]

mandgc said:

Re - Delays at W. Hampstead and W,Park.

So - what are we to do ??

Do you want a manager's job? That's what they do. State the facts then leave it in our laps ;D

As pointed out by my colleague, West Hampstead has no booked reversers, and so would only be used during service disruption - when delays are inevitable anyway.

As for moving trains through that particular faster, that will have to wait for the new signalling to be introduced.

As for Wembley Park, that too will see a recovery over the next few years - namely when the Jubilee is "kicked-out" of Neasden depot. Delays due to stabling trains will then no longer be a problem ;D

[Deleted]

Yeah, but where is Tube Lines going to find the space to put another ten sidings at Stanmore?

[Deleted]

Stanmore currently has ten sidings and stables ten trains. Neasden stables another 10-15 trains, and Stratford Market stables the rest. Trains are not regularly outstabled AFAIK.

If the Jubilee Line is to be booted from Neasden, they will want to move those 10-15 trains somewhere else in northwestern London, and NOT to east London. Otherwise they will end up writing a very painful timetable with lots of useless ECS movements to get trains in position for the morning peak.

Ergo, another ten sidings at Stanmore go in and a minimum of 20 trains become available from the west, with the balance coming from the east.

[DWS]

barney said:

Why does it have to be Stanmore? Theres plenty of waste ground in the West Ham area to build sidings. The old chemical factory next the river Lea is still dereclict I think. Has road access and is right next to the Jubbly.. I am sure some space could be found out east for extra stabling points. The empty Charing Cross and over runs could also be utilised.

The site at West Ham is due to have a Mosque built on it.

[trc666]

Are there any sidings on the Met that could be used?