Tube Trains Cabling

[V41LEY]

Travelling into London at the weekend on Chiltern Railways you reach the sections where the tube lines and mainline track are adjacent. Despite doing the journeys thousands of times, it suddenly struck me how much cabling and wiring is needed to run the tube. Racks and racks of wire fixed in brackets, all different diameter and colour. It was as if every function had a wire which ran the whole length of track. Enlighten me !

[mrmistoffelees]

Cabling for:
Power for different rail sections (you can't put power for section A and B down the same cable, realistically), negative feeders.
Power for services (trackside equipment, lighting, etc)
Monitoring - CCTV, track monitoring (sensors for *everything* temperature for instance, air quality, smoke (is there a fire?))
Signalling
Track circuit counters
Points
Trackside telecoms

Some lines have special requirements - things like Jubilee and Northern which allow for accurate location of the trains.

[andy97]

All just part of the reason why trains (not just Tube trains) are so expensive to run and why Crossrail is so late/ expensive.

[outnumbered]

This was asked on Railforums the other day.

https://www.railforums.co.uk/threads/london-underg...

[Talksteer]

mrmistoffelees said:

Cabling for:
Power for different rail sections (you can't put power for section A and B down the same cable, realistically), negative feeders.
Power for services (trackside equipment, lighting, etc)
Monitoring - CCTV, track monitoring (sensors for *everything* temperature for instance, air quality, smoke (is there a fire?))
Signalling
Track circuit counters
Points
Trackside telecoms

Some lines have special requirements - things like Jubilee and Northern which allow for accurate location of the trains.

This why in the long term rail won't win and wheeled autonomous vehicles will (ultimate expression of boring company solution).

Chargers are cheaper than electrification, intermittent electrification and batteries is even cheaper.

Who needs signalling if the vehicles carry all the intelligence on board.

If you need capacity a few dozen minibus sized vehicles can run in formation. Who needs stations if you don't have to stop every vehicle at every station, also as the vehicles can break off make tight turns, reverse and climb steep gradients stations can be smaller and closer to the surface.

[Condi]

Talksteer said:

This why in the long term rail won't win and wheeled autonomous vehicles will (ultimate expression of boring company solution).

Chargers are cheaper than electrification, intermittent electrification and batteries is even cheaper.

Who needs signalling if the vehicles carry all the intelligence on board.

If you need capacity a few dozen minibus sized vehicles can run in formation. Who needs stations if you don't have to stop every vehicle at every station, also as the vehicles can break off make tight turns, reverse and climb steep gradients stations can be smaller and closer to the surface.

Conversely, why bother with chargers if you can simply run the vehicles on rails? Connect all the trains together or via a central system wirelessly and you need no signalling, no wiring for sensors, nothing other than a power feed. Much more efficient (in terms of people moved) than putting them on the road. There is nothing faster or more efficient for moving large numbers of people than the underground.

[Talksteer]

Condi said:

Conversely, why bother with chargers if you can simply run the vehicles on rails? Connect all the trains together or via a central system wirelessly and you need no signalling, no wiring for sensors, nothing other than a power feed. Much more efficient (in terms of people moved) than putting them on the road. There is nothing faster or more efficient for moving large numbers of people than the underground.

It costs ~£1 million to fit a train "for but not with" ERTMS the most advanced rail signalling system which is taking 20+ years to roll out and uses 2G mobile phone signals.

If you don't maintain rails properly the train derails, if points are incorrect set trains crash. Everything with rail is high stakes and everything must interact with each other, the vehicles must be a certain size or you will either collide with the station or require passengers to jump over a yawning gap. This is why progress is slow and projects expensive.

Autonomous vehicles on the other hand have progressed massively in a short time period and can be made to work with off the shelf technology produced at massive volume and hence very low prices. They are inherently much more tolerant of failures compare a pothole to a broken rail.

The Boring Company went from answering the RFI for the Vegas Loop to building and deploying it in less time than Cross Rail overran.

Yes the first system is clunky, regulations require drivers but imagine what it will look like when the vehicles are custom designed for the purpose.

Underground's are efficient because one driver can move a train with hundreds of passengers down a single small tunnel with headways of a few minutes. Cars couldn't do that because they need a driver per vehicle, much larger tunnels and to deal with the combustion products. Make the cars autonomous and electric and you get rid of the advantages of the train. (Ok you don't get the friction reduction of rails, but that is trivial and totally offset by the cost maintaining the rail)

[Condi]

You're comparing existing technology and costs on one hand with a future time when road vehicles are autonomous. A vehicle on tracks in a relatively predictable and controlled environment is always going to cheaper and more efficient than one which has to cope with all the difficulties and encounters a car has to deal with - pedestrians, emergency vehicles, objects in the road etc. The big difference so far is the investment into autonomous cars is in the 10s of billions, and the investment into autonomous trains much less so.

[Talksteer]

Condi said:

You're comparing existing technology and costs on one hand with a future time when road vehicles are autonomous. A vehicle on tracks in a relatively predictable and controlled environment is always going to cheaper and more efficient than one which has to cope with all the difficulties and encounters a car has to deal with - pedestrians, emergency vehicles, objects in the road etc. The big difference so far is the investment into autonomous cars is in the 10s of billions, and the investment into autonomous trains much less so.

The later point is very much the point, automotive is a much bigger market, but it is also much more dynamic and with much lesser permissions and interdependencies.

As I was saying in my previous post the nominal simplicity of rail is an illusion because it requires much higher levels of engineering integrity. You need complex central control and signalling. Whereas rubber wheeled autonomous vehicles in tunnels can do everything rail can with much less complex central planning and with much greater fault tolerance.

[Condi]

Talksteer said:

Whereas rubber wheeled autonomous vehicles in tunnels can do everything rail can with much less complex central planning and with much greater fault tolerance.

But wheeled autonomous vehicles don't run in tunnels, they run on roads, they run through woods where trees might fall down, where deer may run out; they run through towns where boys play football, where utility companies put roadworks in the carriageway; they run on motorways and duel carriageways where situations can change very quickly. An autonomous wheeled vehicle on the road has to contend with other wheeled vehicles driven by humans.

The 2 are not comparable. Yes, obviously putting cars in tunnels would be cheaper and easier than rail, but that's not how our transport system works, and I would argue that drilling an extensive network of tunnels all over the UK, or indeed putting cars on the underground, would be more expensive than developing driverless trains!