China subway accident

[The_Ghan]

I'm only just hearing about it now.  I understand that there was a signal failure and the trains were being directed by telephone ... I'm not sure if there is a work method statement on that!

Here is a couple of links:

ABC News (Australia)

NDTV

Cheers

The_Ghan

[bikkuri bahn]

I understand that there was a signal failure and the trains were being directed by telephone

It beggars belief that a subway would be run manually w/o signaling and w/o any fail safe system.  If you don't have signals, then you shut down the operation until it's fixed.  But we're talking about China, so...

[Dani]

I understand that there was a signal failure.....

I would be aware from everything coming from China Railway Signal and Communication Corp.... even Ikea would be able to build more reliable signals...

[KenS]

I understand that there was a signal failure and the trains were being directed by telephone

It beggars belief that a subway would be run manually w/o signaling and w/o any fail safe system.  If you don't have signals, then you shut down the operation until it's fixed.  But we're talking about China, so...

Actually, railways in the U.S., including ones Amtrack runs on, often have rules that allow them to revert to radio and paper-based track authority systems in the event of signal failure.  The alternative is leaving a train out in the woods for hours until someone gets there to replace a blown bulb.  All a signals is, is a way to authorize a train to enter a block in such a way that two can't.  The dispatcher can do the same thing with a list on paper.

Signal systems have the advantage that they'll catch errors humans might miss (like someone going past a signal when they weren't authorized) and so they're safer than paper systems. But, particularly in a subway with all trains moving in one direction, I can see a manual system being used to run trains (likely at reduced speed) when the signal system failed, as an alternative to shutting the system down and stranding everyone.

Done right, it should be reasonably safe.  Obviously something more went wrong here.

[spacecadet]

I understand that there was a signal failure and the trains were being directed by telephone

It beggars belief that a subway would be run manually w/o signaling and w/o any fail safe system.  If you don't have signals, then you shut down the operation until it's fixed.  But we're talking about China, so...

Actually, railways in the U.S., including ones Amtrack runs on, often have rules that allow them to revert to radio and paper-based track authority systems in the event of signal failure. 

Not subways.  When signals fail on the NYC subway (just for example), the line shuts down.  We're talking about 2 minute headways here.  It's not the same on an Amtrak line, which often get one passenger train per day.

I don't know about the northeast corridor but I would be extremely surprised if it ever runs when the signals are out.  I have heard many times of trains on that line being stopped because of signal trouble.

[The_Ghan]

I'm hoping that MarkNewton can tell us what the policy is in Sydney.  I'm sure I can remember occasions where we travelled between stations at around 10km/h, possibly because of signal problems.

I would have imagined that work method statements for operating ANY subway with a signal failure would restrict speeds such that  a safe stoppage once the train ahead is sighted is ensured.  Subways are quite different to lines above ground.  I'll bet there are many occasions where visibility is restricted to less than 50m on tight bends in tunnels.

My opinion (for what it is worth, I don't work in the industry) is that irrespective of any communications failure over the telephone, radio, or whatever, that train should have been travelling slow enough to stop once another train is sighted.  It seems to me that lack of appropriate policy and / or training is to blame, rather than the actual signal failure itself.

Cheers

The_Ghan

[keitaro]

The signal between northsydney and milsons point fail a bunch when they do no trains go through.

When I lived in blue mountains uf the signal failed between valley heights the train would go through slowly where the teain would sit there at north sydney for like an hour.

However there is a big saftey difference between a train every 5mins to every 30 mins peak un the blue mountains.

That's my experience anyway

[bikkuri bahn]

IMO, in the event of a signal failure in a subway, at the most, you run the trains at snails pace to the nearest station to let passengers off, then shut the line down completely.  Amtrak's snail trains are a different matter, but then again, trains are run w/o ptc in America, so that modus opperandi is "normal".

Reports on Japanese TV today state the same line in July had a signal malfunction where a train ran the wrong way on the line.  Furthermore, quoting a southern Chinese newspaper, the signalling system currently is actually running on a demonstration or test mode, as the line is planned to go to full automatic (mujin ATO in Japanese parlance) operation next year.  Perhaps the demo mode was inadequate to run a full blown traffic pattern, leading to shutdowns?  Shanghai officials were rushing to have the line running to serve the World Expo last year...

[linkey]

I'm hoping that MarkNewton can tell us what the policy is in Sydney.  I'm sure I can remember occasions where we travelled between stations at around 10km/h, possibly because of signal problems.

I would have imagined that work method statements for operating ANY subway with a signal failure would restrict speeds such that  a safe stoppage once the train ahead is sighted is ensured.  Subways are quite different to lines above ground.  I'll bet there are many occasions where visibility is restricted to less than 50m on tight bends in tunnels.

My opinion (for what it is worth, I don't work in the industry) is that irrespective of any communications failure over the telephone, radio, or whatever, that train should have been travelling slow enough to stop once another train is sighted.  It seems to me that lack of appropriate policy and / or training is to blame, rather than the actual signal failure itself.

Cheers

The_Ghan

I know that is the similar policy for all Metro services for Melbourne if there is a signal failure where they travel at 10km/h until they get to a point where either the signal is in the correct setting or they can see where the train ahead in which they will stop at the nearest singal and wait til the signals can be resetted.

[The_Ghan]

"Resetted", linkey?

For an Aussie you have a fine Chinese accent, isn't it ??  lol

Just giving a bit of cheek.

Cheers

The_Ghan

[keitaro]

i'll cheek you :shakes fist:

haha

[Nick_Burman]

I've just received an email from a friend in Rio saying that a common friend of ours was in the train behind the one which crashed. Whew, lucky guy...


Cheers NB

[keitaro]

at least they didn't tear the subway down to bury the evidence this time ....

[marknewton]

I'm hoping that MarkNewton can tell us what the policy is in Sydney.

Depends on the circumstances. If it's only one or two signals that have failed, you stop at the signal, wait for the signaller's authority to pass the signal at stop, trip past the failed signal, run at reduced speed until you pass the next working signal, then resume normal speed. Reduced speed is generally defined as a speed at which you can "bring your train safely to stand short of any obstruction".

If there are many automatic/permissive signals that have failed, we can introduce what is known as C.A.N. block working - "Condition Affecting Network". It's a manual block system that allows trains to run at close to normal frequency and speed. If the failure involves controlled/absolute signals and interlockings, then we can go over to pilot staff working. That's a system that would take quite a while to explain, but it's a more sophisticated manual block sytem that involves paper and a staff to ensure the separation of trains.

Cheers,

Mark.

[marknewton]

The signal between northsydney and milsons point fail a bunch when they do no trains go through.

That sounds like a points failure, rather than signal failure.

Cheers,

Mark.

[Lawrence]

I've just received an email from a friend in Rio saying that a common friend of ours was in the train behind the one which crashed. Whew, lucky guy...


Cheers NB

Talking of luck, I don't know how many people remember the Kings Cross Fire in '87, but despite living in Scotland, somehow both my father and I transited through the effected area about 10-20 minutes before the disaster, neither of us even new the other was in London and luckily enough we were not caught up in the carnage that followed