Circles to Go

[mrfs42]

Jun 30, 2009 22:21:02 GMT @plymothian said:

The Circle should become the Hammersmith & Circle

I think we'd named it on the Forum 'Circlesmith' - which I quite like on several levels:

1. It never actually makes a circle. ;D ;D
2. It is reminiscent of 'whitesmith' [1] and 'wintersmith'.

[1] I'm sure there's a comment about tin gods and Whitechapel reversers in there somewhere.

[Deleted]

Ideally for this service pattern Edgware Road would be expanded. The works would be very expensive and disruptive, so it's unlikely to ever happen.

Ideally Hammersmith-bound trains would have an new route underneath the Circle/District towards Paddington (Earl's Court-style); and a pair of reversing sidings between (and connected to at the east end) the running lines towards Baker Street (Queen's Park-style) would make for the perfect layout.

[citysig]

Jun 30, 2009 16:20:28 GMT @upfast said:

MetControl - Say a train is stuck in platform 2 or 3 that isn't going to move will you still be allowed to bring the next train into platform 1 and carry out a reform or stock and crew to get things going? Or would this be prohibited under the new service?

In terms of utilising the rolling stock and available platforms, nothing will change from as it is now. So the scenario above will indeed still be possible.

[citysig]

Jun 30, 2009 23:46:44 GMT @a60k said:

Ideally for this service pattern Edgware Road would be expanded. The works would be very expensive and disruptive, so it's unlikely to ever happen.

Hitting the nail on the head of course, that maybe, as the works will never happen, then maybe the new service shouldn't either ;D

[Phil]

ADMIN

Almost half the recent post have been along the 'what if....' lines. Since we all know there is no money for doing anything sensible to expand, please can we confine ourselves to how the thing is working in practice, including of course suggestions for minor tweaks which could be introduced for the start of the real thing in December. If that proves impossible I'm afraid some other unnamed admin/mod ( ) will rightly move the whole thread to the "ideas" category - or just go though deleting all the posts with fantasy elements.

Thanks

[Deleted]

Jun 30, 2009 22:21:02 GMT @plymothian said:

The Circle should become the Hammersmith & Circle

then the Hammersmith & City could either stay the same, or become the Hammersmith & Barking, or revert back to the Metropolitan (or even Hammersmith & Metropolitan) or something completely different. Erm...
The Regent line??

I'm not remotely keen on the unofficial designations this forum creates for sections of track be it "Wimbleware" for the District Line Edgware Road - Wimbledon service or "Circlesmith" for the experimental Circle Line service. They may serve for ease of description on this forum, though I don't use them. But they are not used either oficially or unofficially within London Underground and I worry that casual observers to the forum may think this is how LU internally refer to them or they some how get into print as a mater of historic record !

The Hammersmith and Circle has some greater validity, in so far as internally the generic descriptor for the H&C and Circle Line is "Circle and Hammersmith" or "C&H" because of the interworked duties, stock and staff. Officially the new pattern of service remains "Circle" and "Hammersmith & City" for LU internal communications only the new service is unoficially refered to as the "T-Cup" service so, that term has more justification in useage, though I doubt it would ever be used oficially ... because it sounds daft !

Moving on, I gather during the pilot run of this service Edgware Road had to be closed to incoming passengers at least once due to overcrowding, in a first and obviously not peak hours, when several terminators arrived during a gap in the through service ....

edit for spelling

[Deleted]

Jul 1, 2009 1:22:37 GMT @citysig said:

Hitting the nail on the head of course, that maybe, as the works will never happen, then maybe the new service shouldn't either ;D

If it doesn't prove possible to improve the implementation of the new service, or tweak the timetable in the future trials, that would be my opinion as well. I won't condemn the new service on the basis of a first weekend of trials, but if it can't work any better there doesn't seem much point in changing.

[Phil]

With these problems being highlighted over a weekend, is anyone in TfL going to be courageous enough to go for a full week trial before December? The weekend trials are fine for stock and crew, signalling and controlling, but hopeless for passenger loading/movement at critical points in the peaks (thinking particularly of Paddington H&C).

[Deleted]

courage or not, there simply isn't the time to implement a full weeks trial before December, nor is there the train op resource

[Phil]

Jul 1, 2009 9:04:23 GMT @shedcompnodosh said:

courage or not, there simply isn't the time to implement a full weeks trial before December, nor is there the train op resource

That doesn't surprise me, somehow.

[Chris M]

How long would it take to organise a one week trial?
By my calculations (which could be wrong) there are 21 full weeks between now and the 7th of December (the first Monday of that month). Obviously it will take some planning so not all are available and you will need time to determine the impact of the trail and any issues arrising. Picking a figure out of the air, if it took 10 weeks planning, that gives a week's trial the week of the 7th or 14th of September, and it also gives 9-10 weeks available to digest the results and implement any changes necessary for the start of the real thing, possibly longer if the start date is later in the month (I don't recall seeing anything more specific than "December").
From a layman's point of view that would seem perfectly possible - but I'm happy to be corrected if someone can explain what I'm missing.

I'm assuming that the number of operational staff (in all roles) and the number of trains in service is approximately the same for both current Circle and new T-cup services. I'm basing this on the assumption that this would be the cheapest scenario - but we all know the danger of assumptions!

[Deleted]

And I'm not sure that a 'full week trial' (with going back to the current system after) is at all sensible.

While I appreciate the apparent desirability of such, assuming the weekend trials are deemed successful, then it' s a case of going for it - and hoping it all works come the Monday (or Tuesday or Wednesday) morning peak - at least to a minimally acceptable standard. And if it doesn't, then you have to think again (optimally, drag the contingency plan out of the drawer - but (excuse my cynicism) I doubt one has been thought about). But planning a week's trial and then reverting prior to implementation is asking for three weeks chaos instead of (hopefully) a couple of days worth...

[Deleted]

Jul 1, 2009 9:59:05 GMT Chris M said:

How long would it take to organise a one week trial?
By my calculations (which could be wrong) there are 21 full weeks between now and the 7th of December (the first Monday of that month). Obviously it will take some planning so not all are available and you will need time to determine the impact of the trail and any issues arrising. Picking a figure out of the air, if it took 10 weeks planning, that gives a week's trial the week of the 7th or 14th of September, and it also gives 9-10 weeks available to digest the results and implement any changes necessary for the start of the real thing, possibly longer if the start date is later in the month (I don't recall seeing anything more specific than "December").
From a layman's point of view that would seem perfectly possible - but I'm happy to be corrected if someone can explain what I'm missing.

I'm assuming that the number of operational staff (in all roles) and the number of trains in service is approximately the same for both current Circle and new T-cup services. I'm basing this on the assumption that this would be the cheapest scenario - but we all know the danger of assumptions!

Firstly, it would take far longer than 10 weeks (very much an air based figure) to put together a timetable (MF for the entire sub surface !!!), duty schedules for train ops for each of these timetables. I'd guestimate this to be approx 8 to 10 weeks, per timetable, and the same again for the duty schedule.

Then you have to consider is there the resource available to produce these timetables (approx 1 years person resource!!!)

When a timetable and duty schedule are completed, signal control data, machine rolls etc are then created. This would normally occur at T - 6 weeks. That's more than half your allotted time, just creating behind the scenes elements of the process, to give you a better idea on timescales involved.

And your assumption on ops staff is incorrect. There will be more train ops based on the H&C/SSR come December.

[max]

I can feel some irony coming on.

When you think about where the trains go, and what stations they serve, the new 'Circle Line' is actually more 'Hammersmith & City' than the Hammersmith & City Line.

Can't see this ever happening until the maps, signs, and announcements are updated to match the service. Even with perfect working, the confusion will cause very bad press, and confused passengers slow down services (e.g. slower to get on etc.).

[citysig]

Basically, the weekend trials are a sort of "toe in the water" type thing, and are being used to generate feedback on the sort of items raised on here. In particular the customer information thing.

The rest of the timetable has already been written, and on paper (as has been said on here many times) it works. Very few railways take the opportunity to "test" a timetable at all. Many simply plunge in on the day of introduction and sort it out from there. The weekend tests are only really a "representation" of the real thing - as I have already said some of the recovery time is not incorporated, and there are no real peaks.

There could be, in theory, time to implement a week-long test - if this had been planned from the start. However, if you switch to the new timetable for a week, there is a real danger that people - especially customers - will become used to it and then when it is withdrawn everybody will have to revert back to old habits. Then a few more weeks down the line, it's back to the new one. I think we're messing people about enough without that sort of chopping and changing.

Best to keep any tests "hidden" on the odd weekend. Unrealistic as they probably are, we are never going to know all the ins and outs of the new timetable until probably the end of January 2010 - when the odd bit of disruption or failure has highlighted both good and bad points.

[mrfs42]

Jul 1, 2009 10:41:00 GMT @shedcompnodosh said:

Firstly, it would take far longer than 10 weeks (very much an air based figure) to put together a timetable (MF for the entire sub surface !!!), duty schedules for train ops for each of these timetables. I'd guestimate this to be approx 8 to 10 weeks, per timetable, and the same again for the duty schedule.

I think you'll find that including production and proofing, it takes a fair bit shorter than that to make a respectable and robust WTT.

I reckon, in all frank seriousness, I could make a decent stab at producing a weekday WTT for all the SSR lines (Met., District, Circle, H&C) in just under 10 days. That assumes starting with a blank series of worksheets and finishing with a machine-readable draft copy. What would stymie me is the production of the duty schedule, as I don't have (quite rightly) a list of the staffing complements.

Jul 1, 2009 10:41:00 GMT @shedcompnodosh said:

When a timetable and duty schedule are completed, signal control data, machine rolls etc are then created. This would normally occur at T - 6 weeks. That's more than half your allotted time, just creating behind the scenes elements of the process, to give you a better idea on timescales involved.

It only takes 15 minutes to punch a programme machine roll. As an example I've got two adjacent PM rolls - same location, same timetable, but for two of the machines at the site. The TTNs were dated on the 24th April, the rolls were punched on the 12th May (one at 09:30:34 and the next one at 09:46:54) and the altered working came into force on 24th May. Obviously this is too small a sample size to draw any accurate conclusions (and it assumes that there is only one machine to punch the rolls) [1], but it would seem to indicate that there is a month lead time on the timetable, and a fortnight lead time on the signalling data production. Other PM rolls I've got would indicate that there is less than a month between punching and use.

I'm not saying that you are wrong in what you've said, but I think your timescales are slightly longer than what actually happens. In whatever case, there certainly needs to be a lead time of about 6-7 weeks from a blank start - and that assumes the Timetable Office are just allowed to get on with it, rather than sending the suggestions through a seemingly interminable procrss of review meetings.

[1] in reading the collection of rolls there is very little to tell that there could be more than one punching machine, there is only an occasional variance of less than ¼" between the end of the roll and the three locating holes for the carrier.

[mikebuzz]

Jun 30, 2009 23:07:13 GMT mrfs42 said:

1. It never actually makes a circle. ;D ;D
2. It is reminiscent of 'whitesmith' [1] and 'wintersmith'.

[1] I'm sure there's a comment about tin gods and Whitechapel reversers in there somewhere.

Or Rhodesia?

How about terminating Wimbleware's at High Street Ken, keeping 12 Tph H&C but terminating 4 at Edgware Road and 2 at Baker Street (therefore 6 carrying on as usual to Barking) and reverting to service as usual on the Circle?

Would this work and what are the drawbacks?

[mrfs42]

2 main line shunts at Baker St. in the hour could stuff the job up the wall quite easily; possibly better to have the BkS terminators going on to Moorgate.

[mikebuzz]

Better for Baker St. to Moorgate which would have Circle, Met and 8 tph H&C. Would the extra 2 tph along this section be a problem?

[mrfs42]

I don't think so, but I'm ready to be persuaded otherwise.

Provided you could squeeze in a bit of stand time at Farringdon OR and delay the IR departures from Aldgate I'd suggest that it is doable. Possibly too easy to start delays building up though. The mental picture of where everything would be seems feasible, although it would be nice to tweak things around Chalton St. in this circumstance.

[Colin]

Folks, a reminder - just 16 posts ago, on the previous page - was put up by Phil on behalf of the admins stating that drifting into the realms of fantasy should be done in the appropriate area below and not in this thread. Despite that I'm seeing fantasy posting!!

Once again:

Jul 1, 2009 6:59:46 GMT Phil said:

ADMIN

Almost half the recent post have been along the 'what if....' lines. Since we all know there is no money for doing anything sensible to expand, please can we confine ourselves to how the thing is working in practice, including of course suggestions for minor tweaks which could be introduced for the start of the real thing in December. If that proves impossible I'm afraid some other unnamed admin/mod ( ) will rightly move the whole thread to the "ideas" category - or just go though deleting all the posts with fantasy elements.

Thanks

Failure to heed this request will see the thread moved - you've been warned twice now and there won't be a third!

[Chris M]

I don't want to be seen as second-guessing the admins or anything like that, but could you say which posts since Phil's message you consider to be fantasy postings please. I don't see any that I would personally class that way*, and as it is the admin's opinions that count and not mine I would like to understand what I should be avoiding as I don't want to see this thread moved or messages deleted.

*I see posts that are asking whether X would be possible and would Y be better that what is currently proposed, where neither X nor Y require any fantasy developments, structural alterations, or other ways to spend huge amounts of money, and appear to me to be fully implementable realistically if they would be of benefit and the powers that be decided they wanted to implement them.
If you would prefer this thread now to be solely and strictly about what actually happened and what it is known will be happening in future, with all other discussion - including realistic questions, proposals and suggests based on these - moved to a new thread please say that to avoid misunderstanding (as this is not how I interpreted Phil's request).

[Deleted]

Jul 1, 2009 16:12:36 GMT mikebuzz said:

Better for Baker St. to Moorgate which would have Circle, Met and 8 tph H&C. Would the extra 2 tph along this section be a problem?

I suspect that the powers that be, have decided to reduce the service at this time and end up with a slightly more robust service. The way the service is planned obviously will end up with 2½ mins headways over all the busy shared sections, Praed Street to Edgware Road, Baker Street - Aldgate and Tower Hill - Gloucester Road. An improvement on the current tight 2 mins headways on most of the circle which is increasingly more difficult to maintain. The various overlap changes and removal of signals at different locations has made this so and the lack of any recovery allowance makes it even less likely to recover from minor problems let alone a major one. I feel this may be going in the right direction, that is until the SSR resignalling, if that sees the light of day!

[metman]

Out of interest, what is the chance of converting the space where the two car sidings are at Edgware Road into another platform? Would it be costly and is there room? If I remember, there is a covered section there with offices above. The reason I ask is that I think an extra platform would be required at Eg Rd for the T Cup to work properly. Hopefully this does not incur the wrath of the admins!

Alternatively, if the Wimbleware service is made into 6tph, could one of the platforms at Edgware Road be made available to side track Circle Line trains and provide a bit of lay over of the existing Circle Line service? Which bay roads can you get into from the east end of the station?

[Phil]

Jul 1, 2009 18:08:07 GMT Chris M said:

I don't want to be seen as second-guessing the admins or anything like that, but could you say which posts since Phil's message you consider to be fantasy postings please.

If you want specifics, ChrisM I'll name them - normally the admins do not do this since everyone usually understands. But since you seem not to here we go.

1.Terminating at HSK. This has already been discussed in depth, and the reasons why this is impossible are above in this thread.

2.Terminating at Moorgate. Several LU staff including controllers have already posted as to the restrictions and bottlenecks caused, but again it has been brought up.

3. Reversing some at Baker Street. Apart from this also having been settled, anyone with real knowledge of the situation there will realize how ridiculous this is in operational terms except when the line ahead is blocked for some reason.

4. Conversion of storage roads etc. The whole point of the exercise is that it is cost-neutral and suggestions of spending large amounts of money (on something which still may yet not work......) is indeed pure fantasy.

I hope, Chris, that this explains exactly the sort of thing that is going to get this thread locked by A.N.Other.

[Chris M]

Thank you Phil, I do now understand where you are coming from.

[citysig]

Jul 1, 2009 21:22:52 GMT metman said:

Out of interest, what is the chance of converting the space where the two car sidings are at Edgware Road into another platform? Would it be costly and is there room? If I remember, there is a covered section there with offices above. The reason I ask is that I think an extra platform would be required at Eg Rd for the T Cup to work properly. Hopefully this does not incur the wrath of the admins!

Alternatively, if the Wimbleware service is made into 6tph, could one of the platforms at Edgware Road be made available to side track Circle Line trains and provide a bit of lay over of the existing Circle Line service? Which bay roads can you get into from the east end of the station?

The first part is in the fantasy world the admins have highlighted. If the timetable works - eventually - all well and good. But there will be no structural works to accomodate the workings of the timetable.

On paper, as I have said many times, this timetable works. The timetable has been (of course) written based around the railway we have today. If it runs as on paper, there is no need for flying junctions, additional platforms or sidings. They would, at times, be nice to have even on todays timetable. But they are simply not being thought of by management.

The key question of this thread is really whether this timetable is going to work on paper - which it will, whether this timetable will have noticeable drawbacks during disruption - which it will, and what things can hopefully be improved before the timetable's introduction.

Let's not forget, there may be people influential to the introduction of the timetable reading what is being written here, and endless talk of new sidings won't give them the valuable feedback they could glean from places such as this.

For the second part, if the new timetable was still to be a full all-round Circle, then yes, in theory the reduction in District services would enable the additional layover to be given at Edgware Road. Both middle roads at Edgware Road can be accessed from the eastbound approach.

[Phil]

Good - now Metcontrol has pulled us firmly back to reality (if anyone's in the hot seat he is....) please can we keep it there. Someone is already starting a thread on this topic in the 'ideas' board.

[Deleted]

As a 'former' Schedule Compiler, who is still closer than most to the LUL Timetable Office 'machine' I feel that I must comment on the recent postings by 'mrfs42'.

With all due respect, you have no real idea of the huge amount of work that has been involved in the December 2009 recast of the SSR Lines, to facilitate the Extended Circle Line project.

Your estimate of 'just under 10 days work' is a total and utter fantasy.

From December, in addition to the basic Extended Circle new timetable for the H&C/Circle Lines, train run-times across the entire District and Metropolitan Lines have been increased significantly to better reflect actual performance.
Metropolitan and District Line services have been revised between Baker Street - Aldgate (Met) and Earl's Court - Tower Hill (Dist) to inter-work with the revised H&C/C 'EC', in addition, with this rare recast of the entire SSR, new confliction values at a number of critical sites, including amongst others, Wembley Park, Baker Street and Aldgate have been introduced, again to better reflect actual performance and make the timetable more robust.

As an aside, the recent & future trials for the 'EC' rely on existing run-times/confliction values, so will not offer a 'true' reflection of what can be expected from December 14th, which should be much more robust than what is currently scheduled.

Back to the 'under 10 days to compile a TT' post....

It is my understanding that around six schedule compilers were working full time on the December changes for up to approximately 12 weeks each.

Below is a rough summary of what is involved in a SSR WTT recast for each line (this is not everything - just the headlines).
Also bear in mind that this timetable change affects all days of the week - M/F, Sats and Suns.

Review new run times/confliction values.

Develop Peak and Off-Peak/Weekend 'standards' - Often the morning peak standard is different from the evening peak due to different run times - Ensuring sufficient Trains & T/ops are available.

Construct MF morning peak, from start of traffic/First Trains - building services from nothing, introducing peak run-times. Check first trains against available night turn T/Op duties.

Transition from morning peak to off-peak, removing peak run-times. Stabling required number of trains as appropriate in correct locations.

Midday Off-Peak.

Transition from off peak to evening peak, starting correct number of trains from appropriate locations to mirror morning peak stabling strategy. Introducing Peak run-times.

Transition from evening Peak to evening off peak, removing peak run times, stabling trains in correct locations.

Evening Off-Peak

Transition from evening off-peak to last trains/close of traffic, checking last trains against available night turn T/Op duties. Ensuring trains stabled in correct locations and stock is balanced.

Review Run times - between off-peak and peak and vice versa.

Add Stock moves/ Run As Required test or turning moves/ 'Rusty Rail' moves.

Check all intervals and headways.

Review platform workings/conflicts.

Inter-work with other Lines as required.

Review intervals and headways again - adjusting other lines services as required.

Re do platform workings/conflict checks and all headways/intervals again once inter-worked.

Polish/ Review.

Double check stock balance.

De-ice.

Note any turned trains.

Review stock rotation - ensuring all trains reach a major depot at required interval - especially important for 'C' stock.

Send to Line Business Unit for their comments/feedback.

Incorporate LBU comments.

Pass over to Duty Schedules so they can begin to construct the T/Op Duty Schedule & Rotas.

Once Duty Schedule is complete and cleared timetable has to be formally notified to Infraco via Contracts at not less than three months before operation.

Prepare Front Pages/depot Workings/Rolling stock Workings etc.
Getting ready for press.

Six Weeks out, begin Programme Machine Roll/signals output. Yes it may only take 15 minutes per roll, but there are over 75 rolls to be produced for the Dec 09 WTT change alone.

Finished WTT delivered from printers to whole business a minimum of 4 weeks prior to first operation.


So as you can see there is a wee bit more to it all than you may at first be aware of!

[londonstuff]

What do the program machine roll look like?