signalling problems - what *really* are they?

signalling problems - what *really* are they?

Author
Discussion

phatmanace

Original Poster:

670 posts

209 months

Monday 19th June 2017
quotequote all
After todays problems at Cannon Street, I'm just curious (as someone who is an electrical engineer by training and IT by profession) - what , normally, really are 'signaling problems'?

- Hardware failures on the indicators that the drivers look at and/or sensors in the track?
- Software failures/bugs or machines running out of disk space or something arcane like that?

Really not mean to be a rant, I'm genuinely curious and hoping someone who works in the industry can share some insights.

paolow

3,209 posts

258 months

Monday 19th June 2017
quotequote all
phatmanace said:
After todays problems at Cannon Street, I'm just curious (as someone who is an electrical engineer by training and IT by profession) - what , normally, really are 'signaling problems'?

- Hardware failures on the indicators that the drivers look at and/or sensors in the track?
- Software failures/bugs or machines running out of disk space or something arcane like that?

Really not mean to be a rant, I'm genuinely curious and hoping someone who works in the industry can share some insights.
They can well be signalling problems - possibly the heat is exacerbating the situation.
However, suicides in front of trains are also referred to as 'signal failures' - again - perhaps the heat is a factor.

tog

4,536 posts

228 months

Monday 19th June 2017
quotequote all
Is theft of copper cabling still a problem?

Chrisgr31

13,478 posts

255 months

Tuesday 20th June 2017
quotequote all
It would be surprising if they were software related, although I am sure there have been times when the control offices have gone down. I have always assumed it is usually hardware related so a bit of kit that has failed, and that would appear more likely in extremes of temperature hence yesterday's issues.

alangla

4,795 posts

181 months

Tuesday 20th June 2017
quotequote all
As I understand it (I don't work on the railway!) they fall into a few basic categories:

Sensor failure as previously described - the lines have either track circuits or axle counters to detect trains - typically either the axle counter will get confused & think there's something there when there isn't or the track circuit will be shorted or the relay at the far end fails, again indicating something is there when it isn't.

Machinery failure - the motor or hydraulics that operates a set of points fails to make the points move, or the sensor attached to the points is unable to confirm if the points are properly positioned to run a train over them. Also failures of things like level crossings or their attached detection/CCTV etc.

Communications failure - often the equipment runs off what's basically a remote controlled signal box, sometimes the communication between the remote location & the manned signal box that supervises it is lost. I guess you could also get problems with communications between boxes, so trains move from one box to the next, but the next one has no idea what the train is.

Power failure - everything goes a bit dark. AIUI, signalling systems have separate power from the power at the stations, so the signals might work but the station lights could be out or vice versa, same with the overhead lines/3rd rail.

In areas with mechanical signalling, you can get things like snapped metal rods/wires that operate the old style semaphore signals/points as well, also the telegraph (yes, telegraph) between boxes can fail.


Can anyone with a bit more info confirm if I'm on the right lines or expand a bit?

bigandclever

13,789 posts

238 months

Tuesday 20th June 2017
quotequote all
A cynic might post this ...

Oct16 article said:
Analysis by the Labour Party found that between 2010 and 2015 Network Rail paid train operators £575 million for delays and cancellations caused by issues such as signalling problems and power failures, while the rail firms paid just £73 million in compensation to passengers over the same period.
... so it's in the interest of the train operator to give some spurious reason blaming Network Rail. My train this morning was "delayed because there is a delay" laugh

tight5

2,747 posts

159 months

Tuesday 20th June 2017
quotequote all
tog said:
Is theft of copper cabling still a problem?
Not in the areas with fibre optic cables !
wink


bigandclever said:
A cynic might post this ...

Oct16 article said:
Analysis by the Labour Party found that between 2010 and 2015 Network Rail paid train operators £575 million for delays and cancellations caused by issues such as signalling problems and power failures, while the rail firms paid just £73 million in compensation to passengers over the same period.
... so it's in the interest of the train operator to give some spurious reason blaming Network Rail. My train this morning was "delayed because there is a delay" laugh
You do know that not all train operators have passengers, don't you ?

phatmanace

Original Poster:

670 posts

209 months

Tuesday 20th June 2017
quotequote all
alangla said:
As I understand it (I don't work on the railway!) they fall into a few basic categories:

Sensor failure as previously described - the lines have either track circuits or axle counters to detect trains - typically either the axle counter will get confused & think there's something there when there isn't or the track circuit will be shorted or the relay at the far end fails, again indicating something is there when it isn't.

Machinery failure - the motor or hydraulics that operates a set of points fails to make the points move, or the sensor attached to the points is unable to confirm if the points are properly positioned to run a train over them. Also failures of things like level crossings or their attached detection/CCTV etc.

Communications failure - often the equipment runs off what's basically a remote controlled signal box, sometimes the communication between the remote location & the manned signal box that supervises it is lost. I guess you could also get problems with communications between boxes, so trains move from one box to the next, but the next one has no idea what the train is.

Power failure - everything goes a bit dark. AIUI, signalling systems have separate power from the power at the stations, so the signals might work but the station lights could be out or vice versa, same with the overhead lines/3rd rail.

In areas with mechanical signalling, you can get things like snapped metal rods/wires that operate the old style semaphore signals/points as well, also the telegraph (yes, telegraph) between boxes can fail.


Can anyone with a bit more info confirm if I'm on the right lines or expand a bit?
Great post. Thanks for that..

surveyor

17,825 posts

184 months

Thursday 22nd June 2017
quotequote all
Network rail are busy closing signal boxes and centres and having 'mega control rooms'. They will be great until the comms line goes down...

foxbody-87

2,675 posts

166 months

Thursday 22nd June 2017
quotequote all
alangla said:
Can anyone with a bit more info confirm if I'm on the right lines or expand a bit?
Pretty good explanation! As you have said the main ones tend to be:

Track circuit failures (can be the supply, the relay, leakage due to wet or contaminated ballast, or something as simple as a rogue metal can, lipped rail joint etc bridging two sections and causing them to show occupied on the signallers panel. Once this happens the signaller can't route anything through until the fault team clears the fault.

Points failure - could be something stuck in the points, lack of lubrication, electrical fault, point heater failure in winter, or on some old-fashioned pneumatic points it could be loss of air supply. The points will try to move and if they are obstructed they will move back to where they were and the signaller gets an alarm to notify the fault team.

Power failure - most signalling (except some old rural stuff) is fed from backed up supplies - more often than not the cause lies with linesode power cables. On old systems these are fed from one end only so if you lose the first cable you lose about 5 miles of signalling. They are aluminium with smaller copper tails (16 or 35sqmm copper) at each end which makes it easier to terminate, but that means there are 2 joints per cable and after 40 years in service some of them are starting to go bad.

A lot of this stuff is old technology, but really fascinating to see. Recommend a video on YouTube called "Points and Aspects". All that gear being installed by men with mustaches and flares is still ticking away today.

Network Rail have recently had their boffins working on some remote monitoring systems which help to predict the type of failures listed above. These systems are now in widespread use and there are plenty of stories about averted failures - of course none of those make the news!

foxbody-87

2,675 posts

166 months

Friday 23rd June 2017
quotequote all
alangla said:
Can anyone with a bit more info confirm if I'm on the right lines or expand a bit?
Pretty good explanation! As you have said the main ones tend to be:

Track circuit failures (can be the supply, the relay, leakage due to wet or contaminated ballast, or something as simple as a rogue metal can, lipped rail joint etc bridging two sections and causing them to show occupied on the signallers panel. Once this happens the signaller can't route anything through until the fault team clears the fault.

Points failure - could be something stuck in the points, lack of lubrication, electrical fault, point heater failure in winter, or on some old-fashioned pneumatic points it could be loss of air supply. The points will try to move and if they are obstructed they will move back to where they were and the signaller gets an alarm to notify the fault team.

Power failure - most signalling (except some old rural stuff) is fed from backed up supplies - more often than not the cause lies with linesode power cables. On old systems these are fed from one end only so if you lose the first cable you lose about 5 miles of signalling. They are aluminium with smaller copper tails (16 or 35sqmm copper) at each end which makes it easier to terminate, but that means there are 2 joints per cable and after 40 years in service some of them are starting to go bad.

A lot of this stuff is old technology, but really fascinating to see. Recommend a video on YouTube called "Points and Aspects". All that gear being installed by men with mustaches and flares is still ticking away today.

Network Rail have recently had their boffins working on some remote monitoring systems which help to predict the type of failures listed above. These systems are now in widespread use and there are plenty of stories about averted failures - of course none of those make the news!

valiant

10,227 posts

160 months

Friday 23rd June 2017
quotequote all
paolow said:
They can well be signalling problems - possibly the heat is exacerbating the situation.
However, suicides in front of trains are also referred to as 'signal failures' - again - perhaps the heat is a factor.

Never known a suicide to be classified as a signal failure, more a 'passenger incident' or 'person nit by train'. My lot call them a 'person on the track' as opposed to a 'tresspasser on the track' which is, err, a tresspasser on the track.

Also with my lot, a signal failure is a knackered signal or points but a 'signalling problem' tends to be a driver or signaller fkup e.g a Spad.

smile