Thameslink Core pantograph strike

[Bald Rick]

In theory yes, in practice I think they're very hesitant to do it and as above says it absolutely kills the service on the branches.

The bi di is used during incidents principally to release trapped trains and get them out of the core. If the incident is prolonged (like on Weds) then it is used for a very, very limited service, but most of the service is suspended. This keeps delays down and makes it easier to recover.

 

[swt_passenger]

The system for pan control for Class 700s is designed the way it is because Network Rail made it an absolute requirement that voltage change in the Thameslink core had to be done whilst stationary - so there is no way that an ECS train should ever run through without stopping. The train is, from what I remember, perfectly capable of voltage change on the move, and it uses ETCS balises to determine position so could initiate the pan drop just as the Class 800 does. It's just that the "on the move" case wasn't allowed, so wasn't designed for. I think NR preferred the odd pantograph being wiped off to the perceived risk to the infrastructure of changeovers on the move.

However, changeover on the move was explicitly required in the DfT‘s original train/infrastructure interface specification, I remember that being quoted in a discussion here over 10 years ago. So does anyone know why Network Rail would subsequently reject that?

“The Unit shall be capable of performing an AC to DC or DC to AC changeover without stopping. TIIS_1216”
Train Infrastructure Interface Specification, Network Rail Thameslink Programme, Nov 2008.
(No online version found.)

Isn’t the infrastructure now rigid conductor bar throughout as well, in which case what’s the major risk?

 

[jdp30]

Isn’t the infrastructure now rigid conductor bar throughout as well, in which case what’s the major risk?

Only a bar around St Pancras, it's wires again by Farringdon.

 

[Peter Mugridge]

A pantograph incident doesn't mean that one has been left up and had an unscheduled meeting at Puddle Dock.

True, but he did say it had been left up, so whether or not it actually struck the building I'd say it still counts?

 

[ItCouldBeJae]

5K15 didn't stop at Farringdon or City Thameslink which prevented the train from automatically dropping the pantograph when it receives the signal to do so. When the power switch is in auto, the train is required to stop at either 2 stations to allow the auto changeover to complete.

 

[AM9]

In theory yes, in practice I think they're very hesitant to do it and as above says it absolutely kills the service on the branches.

Isn't the bi-di in the core more often used to release the queue behind a breakdown?

 

[jdp30]

Isn't the bi-di in the core more often used to release the queue behind a breakdown?

It's very rare for it to be used for anything other than that I think - I suppose by the time it is considered the issue is usually close to fixed.

Either way these completely avoidable incidents absolutely destroy the service and given some branches are only 2tph it causes major problems for passengers. GTR often don't do an amazing job of diverting in to KGX and it can take a long time to get it set up when they do.

 

[CaptainBen]

Possibly naïve question: why does this happen on Thameslink and not on the Northern City Line? In 10 years of commuting regularly on the latter, I've only ever once seen a problem. That was a (old and clapped out) 313 which the driver couldn't get the pan down at Drayton Park, so had to go out of service and reverse. I've never heard of any trains going south with the pan up and hitting the tunnel.

 

[swt_passenger]

Possibly naïve question: why does this happen on Thameslink and not on the Northern City Line? In 10 years of commuting regularly on the latter, I've only ever once seen a problem. That was a (old and clapped out) 313 which the driver couldn't get the pan down at Drayton Park, so had to go out of service and reverse. I've never heard of any trains going south with the pan up and hitting the tunnel.

AIUI they can’t hit the tunnel portal because an earthed overhead wire continues further south of Drayton Park.

Presumably it ends somewhere though?

 

[boiledbeans2]

From reading the posts on this thread, it seems that as long as the train stops at the station - the pan will lower automatically.

So possibly the simplest solution would be to force every train to stop? i.e. the signal at the end of the station would remain red until the train stops.

 

[Nicholas Lewis]

It's very rare for it to be used for anything other than that I think - I suppose by the time it is considered the issue is usually close to fixed.

Either way these completely avoidable incidents absolutely destroy the service and given some branches are only 2tph it causes major problems for passengers. GTR often don't do an amazing job of diverting in to KGX and it can take a long time to get it set up when they do.

At least KGX is an alternative on the Brighton line the entire TLK was cancelled for hours with some stations having no services and told to use a local bus service rather than put stop orders on Southern services.

The bi di is used during incidents principally to release trapped trains and get them out of the core. If the incident is prolonged (like on Weds) then it is used for a very, very limited service, but most of the service is suspended. This keeps delays down and makes it easier to recover.

Thing is the service didn't recover even six hours after the core had reopened the service wasn't fully restored (not even hourly) on several service groups. This happens everytime there is chaos in the core the service gets written off for the day particularly the 9J/9R's leaving Redhill line with a 4 car 2 tph service from the North and a GWR 3 car once per hour from the South.

 

[Skie]

With my "I know nothing about the finer points of OHLE, but have a dangerous understanding of circuits" hat on there seem to be 2 simple(ish) ways to provide a warning that wouldn't cost millions or add additional risk:

1. Add a distance of earthed/dead wire that has an induced current in it of a certain frequency/pulse width modulated that a bit of kit on the train can detect. If it detects it the train knows it's in an area where it shouldn't still have the pan up and it can either auto-drop it or scream at the driver.
2. Add a distance of wire that is electrically hooked up to the rails (through a box of tricks). If the circuit completes via the traction return, lights/signs/whatever should light up trackside where a driver can see and take action before it's too late.

Probably incredibly naive, but they feel like they could be methods that wouldn't be too challenging for actual experts in the field to figure out how to make work. No need to integrate with the signalling, nor use flaky pigeon activated optical/laser sensors that would produce endless false alarms.

 

[AM9]

With my "I know nothing about the finer points of OHLE, but have a dangerous understanding of circuits" hat on there seem to be 2 simple(ish) ways to provide a warning that wouldn't cost millions or add additional risk:

1. Add a distance of earthed/dead wire that has an induced current in it of a certain frequency/pulse width modulated that a bit of kit on the train can detect. If it detects it the train knows it's in an area where it shouldn't still have the pan up and it can either auto-drop it or scream at the driver.
2. Add a distance of wire that is electrically hooked up to the rails (through a box of tricks). If the circuit completes via the traction return, lights/signs/whatever should light up trackside where a driver can see and take action before it's too late.

Probably incredibly naive, but they feel like they could be methods that wouldn't be too challenging for actual experts in the field to figure out how to make work. No need to integrate with the signalling, nor use flaky pigeon activated optical/laser sensors that would produce endless false alarms.

The transformers on class 700s have a very narrow frequency band of operation, (as became obvious in the August 2019 power cuts that closed the whole TL network. So any low power ac within that band would be drowned by the 5MW load of those traqnsformers.
Making ad hoc changes to the signalling, especially to ETCS, would involve very high costs with extended qualification testing before being set loose on a live system.

 

[jdp30]

At least KGX is an alternative

Is London Bridge never used in a similar way? No capacity presumably?

 

[class717]

I do take the point though if something happens "in the core", it screws over train services all over from Brighton, Cambridge, Luton and Peterborough. Are the lines in the core bi-directional? So that you could work round a stranded train and run a very limited service on the other line? Or is this not allowed for safety?

For a while this was what was happening

 

[RGM654]

The transformers on class 700s have a very narrow frequency band of operation, (as became obvious in the August 2019 power cuts that closed the whole TL network. So any low power ac within that band would be drowned by the 5MW load of those traqnsformers.
Making ad hoc changes to the signalling, especially to ETCS, would involve very high costs with extended qualification testing before being set loose on a live system.

The special section of overhead wire would carry only the special frequency. The transformers would present some sort of finite impedance at that frequency. Exactly what impedance or what the transformers did with the current shouldn't matter.

 

[jamieP]

Changes don't need to be made. Just better training for drivers. They drive in and out the core everyday so should know better about how the changeover works.

 

[GN Boy]

Possibly naïve question: why does this happen on Thameslink and not on the Northern City Line? In 10 years of commuting regularly on the latter, I've only ever once seen a problem. That was a (old and clapped out) 313 which the driver couldn't get the pan down at Drayton Park, so had to go out of service and reverse. I've never heard of any trains going south with the pan up and hitting the tunnel.

The starting signals at Drayton Park are always set to danger on approach, so all trains are forced to stop. Although, these days, the signals appear to come off quicker, so if you was to coast slowly enough through the station, in theory you could pass without stopping!

 

[Nicholas Lewis]

Changes don't need to be made. Just better training for drivers. They drive in and out the core everyday so should know better about how the changeover works.

The drivers are trained sufficiently its a simple human error situation that needs protecting against. The risk of not changing over is low particuliarly as 99.9% trains stop at Farringdon but has huge impact as we've now seen on several occasions. Solutions are available but i would have thought signalling controls on TL1049 to be held at Red, maybe for class 5's only, is a non physical change albeit i appreciate anything involving signalling software could be much more involved that it appears. Perhaps the simplest thing is a plethora of signs from Farringdon to City Thameslink "Are You on DC"

 

[OscarH]

Changes don't need to be made. Just better training for drivers. They drive in and out the core everyday so should know better about how the changeover works.

Should we stop spending money maintaining AWS and TPWS then, as drivers should know better

 

[jamieP]

Should we stop spending money maintaining AWS and TPWS then, as drivers should know better

Obviously not but sometimes it is the simple things that make a big difference. I work on the railway but not as a driver. If I make a mistake its on me. If the mistake is then made by others the fix is often extra training/ briefings. When a driver makes a mistake we always seem to go and find a more costly/ expensive fix when sometimes it is just not needed.

 

[ComUtoR]

When a driver makes a mistake we always seem to go and find a more costly/ expensive fix when sometimes it is just not needed.

There is a very big difference when discussing this on a forum and what will actually happen in real life.

The incident will be investigated and all aspects will be looked at. For the Driver, I guarantee this will involve some kind of re-brief/training. However, there could be a process error or even a unit fault that occurred.

Ultimately, there will be something that says "how can this be prevented". "Driver Training" will not prevent this happening again. It was most likely human error. Additional training, signage, processes, will all mitigate against it. Multi million pound technical solutions could prevent it but fall far short of reasonable or practical

 

[Samzino]

Changes don't need to be made. Just better training for drivers. They drive in and out the core everyday so should know better about how the changeover works.

Human error still exists. Not to insult you but its this thinking that can be dangerous. In a ideal world you have a fail safe upon fail safe. Human fails then system prevents or reduces the impact of that error, if even that fails then something else covers its back.

It's why companies stress on the Non technical skills. These you can't easily train into someone yet can have a big impact if severely lacking when fatigue kicks in.

 

[Skie]

Making ad hoc changes to the signalling, especially to ETCS, would involve very high costs with extended qualification testing before being set loose on a live system.

Which is why the suggestion is to not have anything that interfaces with the signalling. Obviously it's cheaper to suffer the occasional disruption than it is to make something interlocked, otherwise they'd have done it by now.

Something that completes a circuit by having the pan remain in contact with the 'naughty wire' and illuminates a light/sounds a buzzer is all that's needed and would be, in railway terms, cheap.

 

[sh24]

AIUI they can’t hit the tunnel portal because an earthed overhead wire continues further south of Drayton Park.

Presumably it ends somewhere though?

It ends at the southern headwall of the southbound tunnel at Highbury & Islington.

 

[swt_passenger]

It ends at the southern headwall of the southbound tunnel at Highbury & Islington.

Thanks for the extra detail, does it rise high enough in the station tunnel to force an ADD activation?

 

[philthetube]

Changes don't need to be made. Just better training for drivers. They drive in and out the core everyday so should know better about how the changeover works.

Obviously not but sometimes it is the simple things that make a big difference. I work on the railway but not as a driver. If I make a mistake its on me. If the mistake is then made by others the fix is often extra training/ briefings. When a driver makes a mistake we always seem to go and find a more costly/ expensive fix when sometimes it is just not needed.

So after your mistake following the training does no one else make that mistake? If the driver is at fault it will be on the driver, this doesn't mean that you shouldn't take steps to prevent it happening again.

If this was a mistake it will have cost a good few quid in delay repay as well as inconveniencing many passengers

It seems to me that you could go dead simple to solve this, a piece of string, across the track at pantograph height tied to two spring loaded switches, if either switch is operated then a flashing light alerts the driver, no need to mess with signals. it would only fail safe and a member of staff could then be on the platform to check pantographs before the train departs.

 

[sh24]

Thanks for the extra detail, does it rise high enough in the station tunnel to force an ADD activation?

That goes well beyond my scope of knowledge! It certainly look high, but I have no idea the details of forcing an ADD activation.

 

[boiledbeans2]

It ends at the southern headwall of the southbound tunnel at Highbury & Islington.

Interesting, so that means there is space to fit overhead wires in the tunnels of the Northern City Line?

 

[Mr. SW]

I love all these suggestions, all of which add complication. All you need is an illuminated sign saying "Check Pan" set at an appropriate position. And maybe a shore-mounted arrangement so the driver has to leave their controls to acknowledge the notice by pressing a button. Nothing else needed. No lasers, no strings, no insulated run-off, no modulated signal superimposed on the traction current, no balises, no triggers mounted on gantries, no track-mounted magnets, no software. Just a light source, a sign, a switch and a power source.