Failure to lower pantograph

[Turtle]

Nope. Still done manually.



And a meeting sans biscuits for someone!

And no coffee.

 

[norbitonflyer]

Surely that's fixed by moving the changeover point back a few train lengths, no?

In most cases the changeover point is where it is for a good reason - complexity of extending 3rd rail further back for example.

 

[Meerkat]

It seems set up to create consequences - the wires carry on to within 50m of a bridge for the pantograph to hit.
Are the Westway sliproads really too low for the wires? An extension of 60m would get the pantograph under the bridges and then it has 400m to go overheight and the ADD to drop it.

 

[zwk500]

Eurostars change from 25kV HSL to 25kV ET (and vice versa) on the move nesr Westenhanger. This involves lowering and raising the pan. Similar process in France and in Belgium.

Dropping the pan through a neutral section to change wire height/tension/current/voltage though is different to switching between third rail and OLE. For one thing, if you're going from HS1 to ET the wire rises. I presume that in both directions there is no break in the wire and there is a gradient in the contact wire that meets the standards (noting that wire height goes up and down for things like level crossings and bridges all the time), and so forgetting to lower the pan would be unlikely to result in catastrophic damage.

It seems set up to create consequences - the wires carry on to within 50m of a bridge for the pantograph to hit.
Are the Westway sliproads really too low for the wires? An extension of 60m would get the pantograph under the bridges and then it has 400m to go overheight and the ADD to drop it.

The third rail goes where it does for the former 3rd rail depot. AIUI the Westway sliproads as they are are currently either too low or would force a too steep wire gradient, although this problem is solvable - studies have looked at extending the wires to Shepherd's Bush station before, and other papers have mentioned Kensington as a future OLE changeover point.
When I last looked at extending the wires, it gave a tiny benefit to passenger trains and a reasonable benefit to freight but only if the freight could still change on the move.

 

[Meerkat]

. AIUI the Westway sliproads as they are are currently either too low or would force a too steep wire gradient, although this problem is solvable - studies have looked at extending the wires to Shepherd's Bush station before, and other papers have mentioned Kensington as a future OLE changeover point.
When I last looked at extending the wires, it gave a tiny benefit to passenger trains and a reasonable benefit to freight but only if the freight could still change on the move.

Not sure why wire gradient would be an issue - what does it need to be high for to the north of the Westway.
I admit I have no idea of costs, but even an isolated wire/bar under the Westway would stop the pantographs hitting the bridge - I assume pantographs aren't exactly cheap.

 

[mikeb42]

Unnecessarily complicated. Camera linked to a signal is all that's required.

Still unnecessarily complicated and fancy? The life-or-death safety critical nature of signalling seems to mean that tinkering with it in any way opens a hugely expensive and protracted can of worms, justifiably so perhaps.

Why not something as simple as a tripcock thing akin to those used on the tube to deal with SPADs then? The locations where this happens all seem to be low speed so it can't be the practicalities of high velocities. Doesn't even need to go up and down unless the lines are reversible etc, it could just sit there with no power supply needed. Clouts something correctly located under any train with a pantograph, and drops it. Or even just a fixed magnet type thing. Doesn't have to be 100% foolproof to drastically reduce the risk from low to ~0.

No doubt there must be a massive long list of reasons otherwise someone would have thought of it within 10 seconds the first time the question was posed.

Camera linked to an indicator that isn't part of the signalling system though - why not? Just needs to flash up a visual "You've got your pantograph up you ****" type warning. The argument that image recognition is imperfect seems like "perfection being the enemy of progress", especially given recent strides in AI type machine learning that could be applied to refine it over time. What's the worst thing that can happen? A few unnecessary emergency stops due to false positives? It seems unlikely that the false positive and negative probabilities couldn't be got low enough to make things better overall.

 

[zwk500]

Still unnecessarily complicated and fancy? The life-or-death safety critical nature of signalling seems to mean that tinkering with it in any way opens a hugely expensive and protracted can of worms, justifiably so perhaps.

Why not something as simple as a tripcock thing akin to those used on the tube to deal with SPADs then? The locations where this happens all seem to be low speed so it can't be the practicalities of high velocities. Doesn't even need to go up and down unless the lines are reversible etc, it could just sit there with no power supply needed. Clouts something correctly located under any train with a pantograph, and drops it. Or even just a fixed magnet type thing. Doesn't have to be 100% foolproof to drastically reduce the risk from low to ~0.

No doubt there must be a massive long list of reasons otherwise someone would have thought of it within 10 seconds the first time the question was posed.

The lines are reversible here and any tripcock device (fitted on the bottom of the train) would need to be linked to the pantograph systems on the top of the train with a pipe capable of dumping the pressure quickly enough. Also the tripcock would need to be located in such a place that it won't be struck inadvertently by any other trackside equipment on any other part of the network. And the stop block would need to be located not to knock the shoes off the unit at the same time. Costs of design are quickly starting to add up, then there's fitting and then maintenance.

Camera linked to an indicator that isn't part of the signalling system though - why not? Just needs to flash up a visual "You've got your pantograph up you ****" type warning. The argument that image recognition is imperfect seems like "perfection being the enemy of progress", especially given recent strides in AI type machine learning that could be applied to refine it over time. What's the worst thing that can happen? A few unnecessary emergency stops due to false positives? It seems unlikely that the false positive and negative probabilities couldn't be got low enough to make things better overall.

The argument about false recognition was about wrong-side failure in a device linked to the signalling system. A separate indicator designed to trigger a 'Check Pantograph' warning that doesn't impact the signalling system would not need to be so critically fail-safe but false negatives are a far bigger concern than false positives.
Also, this is a very busy line with trains feeding into the Brighton Mainline and WCML, with 4tph Overground, 1tph Croydon-Watford and numerous freight paths. False positives that force an emergency stop will cause delays that will quickly lead to the implementation of the far simpler solution of just stopping the trains to change over.

Longer term the better solution would be to push the wires down to Clapham Junction, where all trains are going to stop anyway. Although that'll cost a bomb to sort out the links under the SWR and Brighton Lines.

 

[DelW]

The third rail goes where it does for the former 3rd rail depot. AIUI the Westway sliproads as they are are currently either too low or would force a too steep wire gradient, although this problem is solvable - studies have looked at extending the wires to Shepherd's Bush station before, and other papers have mentioned Kensington as a future OLE changeover point.
When I last looked at extending the wires, it gave a tiny benefit to passenger trains and a reasonable benefit to freight but only if the freight could still change on the move.

Would the tests that were done a few years ago for the GWML at Steventon, and which AIUI resulted in an increase in permitted wire gradient for a given speed, be of any relevance or help? Or are they only applicable to high speed lines, which this certainly isn't ?

 

[43066]

Track side sensors that tell the train to drop the panto if it's still in the raised position?

This is sort of the balise solution, but it’s seemingly regarded as too expensive/impractical here.

It seems ridiculous that we are still reliant on trackside signs reminding the driver

As discussed above, and in relation to the last Thameslink core spectacular, once you start drilling into the detail of the seemingly obvious alternatives, they generally aren’t as practical as they first seem. They may also generate costs, or different operational risks, that outweigh the occasional pantograph strike.

 

[CarrotPie]

Or are they only applicable to high speed lines, which this certainly isn't ?

The wire gradient is based on line speed and the work at Steventon decreased the gradientpeed ratio AIUI.

 

[D7666]

Driver been relived and train's been inspected, pan isolated... awaiting on NR now

Relived ?????

That is a bit drastic isn't it ?????

 

[Skimpot flyer]

Relived ?????

That is a bit drastic isn't it ?????

Well it beats being reviled, I suppose

 

[zwk500]

Would the tests that were done a few years ago for the GWML at Steventon, and which AIUI resulted in an increase in permitted wire gradient for a given speed, be of any relevance or help? Or are they only applicable to high speed lines, which this certainly isn't ?

Given the problem here is almost certainly the physical clearance, rather than the close spacing of two obstructions, I doubt anything at Steventon makes a difference here. I'm not sure why I mentioned wire gradient tbh.
However whatever the problem is certainly isn't felt to be impossible or obstructively expensive to solve, as various proposals to extend the wires have been considered to detailed modelling stages.

The wire gradient is based on line speed and the work at Steventon decreased the gradientpeed ratio AIUI.

A special maintenance regime and a slight reduction in speed AIUI.

 

[Fincra5]

There's solutions to these, as others have said.. but ultimately it comes down to cost vs benefit.

The price of installing equipment and modifying train software vs the number of times this occurs. Representing value for money for a TOC/ROSCO.

Often No.

 

[SECR263]

Surely RAIB should take an interest. There is a pattern for some reason for these panto strikes due to crew error ( I think crews should be helped to avoid this by engineered means, it helps them in their jobs) and removes a hazard. which lead to massive disruption and in some cases passengers detaining which is life threatening. An order to Network rail is needed irrespective of cost to sort this out. Just accepting it as a fact of life is not satisfactory. West Coast has to fit CDL, Why not network rail required to fit a solution.

 

[ComUtoR]

Surely RAIB should take an interest.

Why ? Its not really their remit. Its a generic incident that has a known reason for occurring.

( I think crews should be helped to avoid this by engineered means, it helps them in their jobs)

As discussed many times. Engineering solutions do not work. A whole lot of cost for little to no gain.

and removes a hazard.

Some reduction yes, some mitigation yes, removes... nope.

 

[Meole]

Somewhat amusingly this problem arose on the quayside branch at Newcastle after the electrification of the tunnel back in 1904, third rail in the tunnel, overhead outside in the sidings where foot loading of freight prevented use of the ground electric supply.

 

[zwk500]

Surely RAIB should take an interest. There is a pattern for some reason for these panto strikes due to crew error ( I think crews should be helped to avoid this by engineered means, it helps them in their jobs) and removes a hazard. which lead to massive disruption and in some cases passengers detaining which is life threatening. An order to Network rail is needed irrespective of cost to sort this out. Just accepting it as a fact of life is not satisfactory. West Coast has to fit CDL, Why not network rail required to fit a solution.

If an order would be made to anybody, it would be to London Overground and the order would be to come to a stand to change traction, as GTR do at this location. No need for an infrastructure solution.

 

[norbitonflyer]

St Quintin Park & Wormwood Scrubs station, on North Pole Road, about midway alonmg the 2 mile stationless stretch between Willesden Junction and Shepherds Bush, was roughly where the changeover point is. If trains are going to stop there anyway, why not re-open the station? (Name it North Pole - think of the opportiunities for special excursions in December!) (The nearest existing station is White City, half a mile away)

 

[Bald Rick]

Surely RAIB should take an interest.

Its nothing to do with them.

An order to Network rail is needed irrespective of cost to sort this out.

An Order by which organisation? RAIB doesn’t make orders.

 

[ExRes]

It seems ridiculous that we are still reliant on trackside signs reminding the driver being the main way this is done...

Can't say as I agree at all, you might as well say that speed boards, whistle boards and neutral signs are all insufficient, it's what signing the route is about