Central line - shortage of serviceable trains

[Nym]

You can't run short sets away from the loop right now - there are no stop boards for short sets.

... and put a different 2 car unit in...

Kinda the whole point of them being 2 car units. And in spite of several attempts over the years to save a hell of a lot of money on autocoupler maintenance, no-one has though to reform them into four car units with some double enders.

 

[bramling]

So now you'd like the COO Operational engineers to change the DISI and depot dispatch instructions, where something might fail out on the railway.

Not to go too far into this rant, but someone very irritating about the 'big railway' attitude is that stopped is safe, this is not the case on LUL in deep level tunnels, and if you have failed motors on a unit, on 1992TS I beleive you're isolating a whole bogie or potentially even a whole unit, which is a significant hit to the performance and will make the other motors work harder, and therefore, fail sooner, increasing the likelyhood of an in service failure.

Combine this with the lack of LUL controllers nowerdays to take positive actions (likely due to the joys of oversight by comittee after the sight saying, "well if you'd have done this", when compared by those who are not operational, but like to think they'd know better, hence resulting in a decision paralysis due to fear of issues later).

And now you have increased the risk of stalled trains, this is not an acceptable hazard to the operational railway, so no sensible engineer, and I'm sure there are still some left in TfL after the multiple re-orgs that pushed out so many good people...
... it's not a good idea, and anyone who signs it off does not have the interest of the railway or the safety case in mind.
A change of the maintenance requirement needs to be analysed under the TfL Pathway and CSM-RA safety assessment and against LU Standards S1037 and S1180 for the change, it's unlikely that any change would be able to be done by the very short staffed engineering teams, even if they were willing. It would be much better trying to focus efforts on being a better company to work for, and a better company to work with as part of the supply chain.

It does occasionally happen on LU that trains will run around with motors out, though I’m not sure trains would be offered for service knowingly in that state.

Certainly I can think of a train that ran about on the Northern Line for most of a day with a traction fault on one car. Needless to say, despite service control being supposedly fully aware, it didn’t stop one or more calls from service control enquiring as to why the train was losing time, and the odd report from station staff regarding burning smells due to the friction brakes having to do extra work.

 

[Mawkie]

If there are units parked up with a motor failure, what would be the impact of running it in service? How many motors could be isolated on a full set before it caused a performance issue?

I don't know much about the Central Line, but the Defective in Service Instructions require a 92 Stock train with a defective traction package to be worked in service to the depot.

One traction package not
working on entire train.

09 S Remain

72 73 92 95 96 In service to depot

(On 92 stock only, a
traction package is defined
as the traction equipment
on a two car unit.)

 

[trebor79]

So now you'd like the COO Operational engineers to change the DISI and depot dispatch instructions, where something might fail out on the railway.

Not to go too far into this rant, but someone very irritating about the 'big railway' attitude is that stopped is safe, this is not the case on LUL in deep level tunnels, and if you have failed motors on a unit, on 1992TS I beleive you're isolating a whole bogie or potentially even a whole unit, which is a significant hit to the performance and will make the other motors work harder, and therefore, fail sooner, increasing the likelyhood of an in service failure.

Combine this with the lack of LUL controllers nowerdays to take positive actions (likely due to the joys of oversight by comittee after the sight saying, "well if you'd have done this", when compared by those who are not operational, but like to think they'd know better, hence resulting in a decision paralysis due to fear of issues later).

And now you have increased the risk of stalled trains, this is not an acceptable hazard to the operational railway, so no sensible engineer, and I'm sure there are still some left in TfL after the multiple re-orgs that pushed out so many good people...
... it's not a good idea, and anyone who signs it off does not have the interest of the railway or the safety case in mind.
A change of the maintenance requirement needs to be analysed under the TfL Pathway and CSM-RA safety assessment and against LU Standards S1037 and S1180 for the change, it's unlikely that any change would be able to be done by the very short staffed engineering teams, even if they were willing. It would be much better trying to focus efforts on being a better company to work for, and a better company to work with as part of the supply chain.

That all makes a lot of sense. But perhaps in extremis its appropriate to at least consider how things might be improved with some short term measures whilst the situation is sorted out?
And how about buying a bigger stock of spare motors/motor parts?
Could the acceleration and deceleration profiles of the trains be tweaked to put less stress on the motors, maybe just in the short term? If the service has big gaps anyway then a couple of extra minutes running time isn't going to cause issues.

I get it's difficult, but I detect a sense of resignation that nothing can be done about it. That's rarely the case. There's any number of small things which on their own don't solve it or make much difference but together can improve things for everyone.

 

[Nym]

Things can of course be done, more stock can be ordered etc etc etc.

But... Can it be done in TfL? The answer 10 years ago was no, and with the more apparent financial difficulties and loss of a lot of very good engineers, the answer will be even more no today.

 

[mrmartin]

Good thread here on twitter from a former rolling stock engineer at LU: https://twitter.com/ajf43/status/1750931929732468829?t=3KCW83qNRRb7vBaJIWD_wQ&s=19

Basically central line stock are some of the last with DC motors. Northern and Jubilee have AC motors so aren't at risk of the same failure.

Apparently it's not really known why so many are all failing at the same time.

Don't really think this has much to do with "managed decline" per se. It could have happened whatever the funding situation was (unless the trains were completely replaced, but 30 years is a poor show for this stock). Just defective design kicking in...

 

[Route115?]

Would it be worth introducing an emergency timetable? It seems that only about 10% of trains are cancelled (more at weekends) but this leads to long gaps because of the way that the timetable is structured.

 

[D7666]

I can't say specifically for the DC motors in 92TS but it is not good for many DC motors to be dead but rotating for long term, which is what happens with cut out motors and what would happen if trains were deliberately placed in service with isolated motors. One ill effect is commutator glazing which is NBG for them when you next put power on. You are likely to add more motor breakages which is a bit counter productive.

There is also the issue of what the actual motor defect is - there are many defects that won't allow the motor be drawn dead anyway. Again no idea what the 92TS faults are but loose armature banding or even loose windings, or loose commutators, or bearing troubles you can't just tow them around.

Another issue is not that of motoring but of braking; the all out dynamic braking load needed to run the timetable - the braking effort won't be there if 2car an 8car are isolated.

Working a defective train in service to depot is one thing, sending it out, all day, for weeks maybe, and expecting it to perform is unlikely to work for very long.

Apparently it's not really known why so many are all failing at the same time.

Don't really think this has much to do with "managed decline" per se. It could have happened whatever the funding situation was (unless the trains were completely replaced, but 30 years is a poor show for this stock). Just defective design kicking in...

I don't work with rolling stock but do in other asset areas; this sort of "systematic failure" more often than not has little connection with funding gaps as can almost never be predicted. OK one could argue the general case of general under investment, but even then more general money does not head off specific issues. Some while back, well mid last decade, one signals control system hardware went into systematic failure, entirely unpredictable, and one particular processor board of one type, with a low total count in use BUT essential for the whole, more or less every board failed within a 12 mo period. The only reason we got through it was a high spares holding (higher than accountants would like), the ability to get the early failures flown out repaired and returned before the later failures occurred, and procure of new build boards - of which both repairs and new build relied on USA specialists; such outfits even now still support x86 286 386 486 etc devices even if the average "expert" will tell you no. (The biggest issue was actually getting customs clearance through Heathrow; it was more difficult doing that than finding someone with Intel OEM rights.)

 

[ThameslinkUser]

Would it be worth introducing an emergency timetable?

One is already in use. TTN88/23 according to this post on District Dave

Due to long term train cancellations, the following temporary service arrangements will apply and supersede Working Timetable 70, until further notice.
6 trains have been withdrawn from the weekday timetable of WTT70. This is comprised of 3 trains which operate throughout the traffic day and 3 trains which operate in the AM peak only.
2 out of the 6 trains are cancelled from after the PM peak to close of traffic.
Existing trains have been rescheduled to minimise the impact of wide intervals that have opened up from the cancelled train paths. However some wider than usual intervals will remain in certain parts of the timetable, due to pathing constraints. Intervals are to be regulated and services adjusted where necessary.
In consequence of the re-allocation of Signalling Trip numbers and Routeing information, the whole Timetable is published in this notice.

Trains in service: WTT70 M-F peak: 77; TTN88/23 M-F peak: 71

This TTN was brought into use on June 5th 2023, mind

 

[Zeremony]

One is already in use. TTN88/23 according to this post on District Dave

See this article by Diamond Geezer
Three diagrams were cancelled, six had their operating hours reduced and six less trains were required for the M-F peak

So there are 85 Central line trains but five are being refurbished at Acton.
71 trains are scheduled to be in service during the M-F peaks (-6 in June 2023 and -1 in Jan 2020 with the introduction of the Woodford-Hainault shuttle)
However currently only about 50 trains are operating day to day with a 18 tph service at peak times

Using intertube (very useful for checking where the next train is) I measured the amount of trains in a 2 hour period between 0700-0800 and 1630-0830 at Oxford Circus on various dates:

FRIDAY 26
MORNING PEAK: there were 72 trains passing in both directions in the 2 hours which works out to 18 tph
EVENING PEAK: there were 74 trains passing which works out to 18½ tph
THURSDAY 25
MORNING PEAK: 72 trains total in the two hours = 17½ tph
EVENING PEAK: 64 trains total / 2 hours = 16 tph
WEDNESDAY 24
MORNING PEAK: 68 trains = 17 tph
EVENING PEAK: 73 trains = 18¼ tph
TUESDAY 23
MORNING PEAK: 67 trains = 16¾ tph
EVENING PEAK: 75 trains = 18¾ tph
MONDAY 22
MORNING PEAK: 73 trains = 18¼ tph
EVENING PEAK: 75 trains = 18¾ tph

SUNDAY 21
EVENING: 70 trains = 17½ tph
SATURDAY 20
EVENING: 73 trains = 18¼ tph

FRIDAY 19
MORNING PEAK: 65 trains = 16¼
EVENING PEAK: 77 trains = 19¼ tph
THURSDAY 18
MORNING PEAK: 67 trains = 16¾ tph
EVENING PEAK: 71 trains = 17¾ tph
WEDNESDAY 17
MORNING PEAK: 71 trains = 17¾ tph
EVENING PEAK: 65 trains = 16¼ tph
TUESDAY 16
MORNING PEAK: 71 trains = 17¾ tph
EVENING PEAK: 76 trains = 19 tph
MONDAY 15
MORNING PEAK: 69 trains = 17¼ tph
EVENING PEAK: 76 trains = 19 tph

SUNDAY 14
EVENING: 67 trains = 16¾ tph
SATURDAY 13
EVENING: 67 trains = 16¾ tph

Using the Working Timetable I worked out:
- 78 trains are required for the full peak time M-F service (from August 2015 with Woodford via Hainault services)
- 73 trains are required for 26½ tph - every 2¼ mins (Sat afternoons)
- 71 trains are required for 26¼ tph (M-F peak since June 2023)
- 65 trains are required for 24 tph - every 2½ mins (Sun evenings and Saturday morning)
- 41 trains are required for around 14-16 tph (2300 - close)

- Around 44 trains are required for 16 tph
- Around 46 trains are required for 17 tph
- Around 49 trains are required for 18 tph
- Around 52 trains are required for 19 tph

Conclusion​

- There are around 50 out of 80 trains in service at peak times
- This suggests there are approximately 30 defective trains with motor issues!
- There are almost always more units available for the evening peak compared to the morning
- Some Woodford via Hainault trains operate with Woodford - Hainault services every 30-40 minutes
- Gaps of around 10 minutes to both branches (Epping/Hainault and Ruislip/Ealing) are common leading to overcrowding
- If a diagram does not run it leaves a 8-10 gap to that branch because the trains still run to the times on the WTT which are scheduled every 4-5 minutes
- I think there should be new working timetable with frequencies of every 7 minutes to each branch will help even out the service more
- However, big TT changes like this take a lot of time so probably will never happen
- A temporary TT will probably just be cancelling diagrams and rescheduling trains slightly to reduce the interval a little like the June TT amendment

Based on the data from the last 2 weeks:
- Average wait: 3½ minutes
- Average peak tph: 17.9
- Average tph morning peak: 17.6
- Average tph evening peak: 18.2
- Average tph evening weekends: 17.5

 

[StewLane]

I assume it is the DC motors that are failing so with 5 trains at Acton being converted to AC motors I would expect that there should be 5 trains worth of DC motors available for use as spares at any one time? Or am I missing something.
Confused..

 

[Goldfish62]

I assume it is the DC motors that are failing so with 5 trains at Acton being converted to AC motors I would expect that there should be 5 trains worth of DC motors available for use as spares at any one time? Or am I missing something.
Confused..

A former LU senior rolling stock engineer gave a bit of background to all this on X. Sorry, I wouldn't be able it again now such is X these days.

Basically the DC traction motors, which by their nature are high maintenence anyway, have had an inherent design fault from new. It's never been fully got to the bottom of, but has been managed over the years.

It now seems it's got to the stage where it can no longer be effectively managed.

 

[Nym]

Or that TfL is unwilling to pay to manage it any more. (Seems more likely)

DC motors are maintenance heavy, but they're all a very known value if you're not trying to squeeze money out of Ruislip's budget.

 

[StewLane]

A former LU senior rolling stock engineer gave a bit of background to all this on X. Sorry, I wouldn't be able it again now such is X these days.

Basically the DC traction motors, which by their nature are high maintenence anyway, have had an inherent design fault from new. It's never been fully got to the bottom of, but has been managed over the years.

It now seems it's got to the stage where it can no longer be effectively managed.

So they are not able to use any serviceable DC motors that come off trains being converted to AC motors or are there just not enough being replaced?

 

[Goldfish62]

So they are not able to use any serviceable DC motors that come off trains being converted to AC motors or are there just not enough being replaced?

Perhaps that's what they're doing and stock availability would be even worse if they weren't.

 

[StewLane]

Perhaps that's what they're doing and stock availability would be even worse if they weren't.

Thanks everyone for the replies.
suspect that DC motors may not be being overhauled as the plan is to replace them but they are failing faster than they are being replaced

 

[notverydeep]

There was an article on this subject in the Evening Standard on Friday: https://www.standard.co.uk/news/tra...delays-trains-tfl-b1134996.html#comments-area

The line has been operating at times with barely half of the 78 trains required to run a peak service, the Standard has learned.

This is because the ageing motors on the trains, which date to 1992, are suffering catastrophic failures and have to be taken out of service to be replaced.

Sources said the number of trains available over the last fortnight had ranged from 44 to 66, and typically little more than 50.

 

[Goldfish62]

There was an article on this subject in the Evening Standard on Friday: https://www.standard.co.uk/news/tra...delays-trains-tfl-b1134996.html#comments-area

Wow, never realised it was that bad!

I've just looked at live departures at Bank. Eastbound is OK, but westbound there's an 8 min gap followed by a 14 min gap.

 

[Dstock7080]

22 trains cancelled from 71 scheduled yesterday PM.

 

[Goldfish62]

22 trains cancelled from 71 scheduled yesterday PM.

The staff canteens must be full of drivers!

 

[RingArm]

This is a very sad time for the Central, but I suppose if all the trains were introduced around the same time, they will all age and ultimately conk out at about the same time. On the other hand it's considered good practice to reduce maintenance if a refit (or withdrawal) is anticipated. Pity they can't borrow a few trains from the other tube lines.

 

[Nym]

This is a very sad time for the Central, but I suppose if all the trains were introduced around the same time, they will all age and ultimately conk out at about the same time. On the other hand it's considered good practice to reduce maintenance if a refit (or withdrawal) is anticipated. Pity they can't borrow a few trains from the other tube lines.

Well that's the joys of preparatory signalling systems rather than settling on something standard.

 

[rebmcr]

Well that's the joys of preparatory signalling systems rather than settling on something standard.

I think you mean 'proprietary'.

 

[Thirteen]

Let's be honest, there isn't spare trains available even if the Central Line was compatible with other lines.

I think TfL needs to pitch to Labour for a long term funding deal so that we can get new rolling stock.

 

[D7666]

I think you mean 'proprietary'.

That is what I read too.

On the subject of proprietary, any new system, and definitely any ATO, is always proprietary these days.

TBTC (NL and JL) and CBTC (SSL) is Thales proprietary, VL DTG Siemens/Westinghouse, and CLATP also Siemens/Westinghouse of an earlier generation; Eliz line CBTC is Siemens CBTC; Thameslink core ATO is yet another different Siemens product, add so on. ETCS ERTMS etc are standard in the sense that principles and the outputs align, but how these are implemented by the different suppliers is also proprietry. That is why NR have a whole test centre at Hitchin dedicated to homologising the different suppliers kit; Hitchin ENIF was not just about a 313, Thameslink, and the Moorgate branch which popular myth will have you believe.

While the 1990s LU Central Line kit is proprietary, it was not unique in the world (although it is in GB), its concepts at the signalling ATP level are not that different to DTG (the communication is very different). CL (and DTG) use coded track circuits, and any supplier can provide those. The CL interlockings, depending which of the 19 sites you look at, are part all relay or part relay + part Westcode or all Westcode; the differences lie in the project installation time line no other reason. The relays are Q types, known on main line as BR930, standard items, and are more or less plug and play kit; the code generator relays on the all relay interlockings are standard items set to generate unique codes; Westrace is modular anyway, despite signalling engineers surrounding a Westrace in mystery, they are but high integrity PLCs; and Westrace is a world wide product. Then there are the signalling PLC supervising the interlockings, industry standard PLCs from a well know maker. The scanning kit is unique within LU but a world wide product also used elsewhere and follow generic concepts. The ATS ATO kit is a little more bespoke in the sense it was implemented in Intel Multibus hardware, but rationalise that into modern terms, Multibus*** is but a 'blade server', and, being Intel based, is not difficult, indeed, is being done now, to migrate to modern PC platforms; the Sun Sparc control room workstations again rationalise as desktop PC: ok they run Solaris o/s but not rocket science to migrate; ditto the central timetable server is another Solaris box.

Agree it is all obsolete, none directly one for one replaceable, but none of it is a blocking point, it is far more modular than first sight, and if push came to shove, different makers kit could fit the bill.

I am less certain of the train borne kit, that is outside my domain, but although high integrity, they can't be highly complex simpy because of the generation they were built in. My ATO colleagues might debate that with me, as we have in the past. Even then, there are Q relays on board, that is the clicking you can hear if you sit right next to the relevant bulkheads. Other stuff like antennae pale into trivia.


*** Re. Multibus, it was a widely utilised industry control system, most if not all the USA FAA air traffic control systems used and a lot still uses exactly the same modules; the US military still use a lot; GB Navy used the same kit in diesel subs; world wide a number of petrochemical plants use it. Sure these are not railway control but are all high integrity control systems built out of standard modules. If not familiar with Multibus, all they are is 19 in rack mount PC processor boards with certain on board peripherals; imagine your deskptop PC mother board, not in its own box, but plugged into a backplane; data comms is via industry standard 2 MB IEC backplane bus. Reason for mentioning all this is to demonstrate its modularity at hardware level. The o/s is the same; only the application software is different, but application software is always always always bespoke and customised site by site.

 

[RingArm]

Even if all the tube stock could be used on any other line, the trains themselves are all different, so the staff training required would make it impossible.

 

[357]

Even if all the tube stock could be used on any other line, the trains themselves are all different, so the staff training required would make it impossible.

No it wouldn't. Look at how on the District line Earls Court used to be the only depot to sign C stock and High Street Kensington - Edgware Road, and look on the mainline how different depot's sign varying routes and traction types.

Epping to Ongar used to use a different type of stock if you want a Central line example.

If the will is there it could be done. You'd need to find a place where it could be done where disruption would not risk having a train in a platform with no driver who signs it.

 

[RingArm]

No it wouldn't. Look at how on the District line Earls Court used to be the only depot to sign C stock and High Street Kensington - Edgware Road, and look on the mainline how different depot's sign varying routes and traction types.

Epping to Ongar used to use a different type of stock if you want a Central line example.

If the will is there it could be done. You'd need to find a place where it could be done where disruption would not risk having a train in a platform with no driver who signs it.

Had it been the usual way of working, I would agree; but that isn't how the underground works. To borrow a few trains from around the system, even if they were compatible, to overcome a serious but temporary problem, would require a massive staff training exercise as things are operated today. Not just drivers, platform staff, maintenance staff and even cleaners have to be trained.

 

[Dstock7080]

Look at how on the District line Earls Court used to be the only depot to sign C stock and High Street Kensington - Edgware Road,

Acton Town operated C Stock until the end and Upminster worked them until they were refurbished 1991/92

 

[D7666]

Had it been the usual way of working, I would agree; but that isn't how the underground works. To borrow a few trains from around the system, even if they were compatible, to overcome a serious but temporary problem, would require a massive staff training exercise as things are operated today. Not just drivers, platform staff, maintenance staff and even cleaners have to be trained.

The general case quasi blocking point is not will or technical or operational inability to train any on anything else, but there is no money to do it and no other train types that could be moved from any one line to any other line to meet the bill. The whole idea is a non starter.

Re. the Central line using other stock, seeing as someone else already cited an example of other stock on the line, that was in non line wide CLATP ATO days. Whatever you put on the CL *TODAY* needs to have CLATP in /some form/ installed. Money. I'd be pretty sure if someone had money to throw at this problem it would be far better spent on overhauling some motors than on fairyland suggestions of running non existent trains from other lines.

Other train incompatibilty is not the blocking issue - non availability of other trains is.