Serious Accident in Bavaria

[Bertiethebobby]

And the TRB would read something like:
Instruction given to driver of..............to do whatever. Driver refused to carry out instruction. Control informed @ time"
--- old post above --- --- new post below ---


Don't worry, it even makes us siggies sweat too!
We're the ones who end up in Court!

I've become far more risk averse in my old age.
On the heritage railway I work on we sometimes have a situation where if one train stops a bit short in the platform and still occupies a track it stops you pulling off for the other one even though they're 50 yards apart. Now I wait for the first one to clear the track rather than authorise him past the signal at danger. Same with run rounds, I get them to set back behind the signal instead.
As you say, it's our neck on the line every time!

 

[sarahj]

To see a Zs signal in action https://www.youtube.com/watch?v=L27aq50M16c
just around the 9 min mark. ( I think )

SJ

 

[Bletchleyite]

I've become far more risk averse in my old age.
On the heritage railway I work on we sometimes have a situation where if one train stops a bit short in the platform and still occupies a track it stops you pulling off for the other one even though they're 50 yards apart. Now I wait for the first one to clear the track rather than authorise him past the signal at danger. Same with run rounds, I get them to set back behind the signal instead.
As you say, it's our neck on the line every time!

Quite right too. Bypassing first-line safety systems to excuse sloppiness (even if done according to the proper procedure) leads to accidents. The first choice should always be to remedy, in a proper manner, the actual problem, and only if that is not possible move onto alternatives.

 

[leb37]

BBC News is reporting that it was human error that is to blame for the crash

http://www.bbc.com/news/world-europe-35585302

 

[Zoidberg]

BBC News is reporting that it was human error that is to blame for the crash

http://www.bbc.com/news/world-europe-35585302

An extract from that webpage to help those unable to access links

Human error by a train controller was to blame for a train crash in Bavaria, Germany, last week that killed 11 people, the prosecutor has said.

More than 80 people were injured after two commuter trains collided on a single-track stretch of railway.

The controller who opened the track to the two trains and tried to warn the drivers caused the crash, according to the prosecutor.

 

[Bletchleyite]

Horrible echoes of Cowden...

 

[Taunton]

Horrible echoes of Cowden...

I don't believe so. At Cowden one train was signalled to stop, but ran past onto the single line (not noticing the points were set reverse). Now it was misty, the signal was found displaying a very reduced illumination, and the AWS on the train was suspect, but not like the German incident, where it appears (as has happened before in Germany) that it is all too easy to overlook things and clear trains from both ends at once.

 

[Bletchleyite]

I don't believe so. At Cowden one train was signalled to stop, but ran past onto the single line (not noticing the points were set reverse). Now it was misty, the signal was found displaying a very reduced illumination, and the AWS on the train was suspect, but not like the German incident, where it appears (as has happened before in Germany) that it is all too easy to overlook things and clear trains from both ends at once.

It wasn't so much on the basis of the similarity of the cause, which indeed was a different human error, but more on the "we tried to warn them but the crash was inevitable once the issue had occurred" thing, which just makes one shudder.

 

[zuriblue]

It wasn't so much on the basis of the similarity of the cause, which indeed was a different human error, but more on the "we tried to warn them but the crash was inevitable once the issue had occurred" thing, which just makes one shudder.

Indeed. I can't help thinking that the Zs1 signal which allows the driver to resume linespeed once over the pointwork, is a dangerously flawed concept. I would have thought that on a single line the Zs7, which means that the driver continues to drive at caution until he encounters a clear aspect, is much more appropriate.

Unfortunately, in the words of Pterry, "while it is easy to deal with stupid, bl**dy stupid is hard to erase"

 

[Bertiethebobby]

BBC News is reporting that it was human error that is to blame for the crash

http://www.bbc.com/news/world-europe-35585302

A chilling reminder of the burden of responsibility that we carry in our work.

 

[Bletchleyite]

Indeed. I can't help thinking that the Zs1 signal which allows the driver to resume linespeed once over the pointwork

I didn't realise it meant that. I therefore do not think Zs1 should exist at all on any kind of line. If instructed to SPAD, the rule should be "drive on sight until you pass a signal saying otherwise".

on a single line the Zs7, which means that the driver continues to drive at caution until he encounters a clear aspect, is much more appropriate.

On a single or bidirectional line I don't think there should be any pushbutton override, I'm afraid, as the situation is that there could be a train coming the other way that knows nothing of the "drive on sight" regulation. Extremely dangerous.

In such cases, either you fix the problem, or you have someone walk the line to check it is clear, in my view. If you can't absolutely prove there is not an "errant" train running against the one you're about to clear, no go.

Drive on sight is only safe if there isn't, and *absolutely* cannot be, a train coming towards you at linespeed, with no "drive on sight" caution and not knowing you are there.

 

[w0033944]

My understanding from the BBC report is that both trains were running at linespeed - 100 km/h - which seems inconsistent with one of them being under "drive on sight" regulations. Is this correct?

 

[Bletchleyite]

My understanding from the BBC report is that both trains were running at linespeed - 100 km/h - which seems inconsistent with one of them being under "drive on sight" regulations. Is this correct?

It seems from the above that Zs1 does not require drive-on-sight. A most dangerous concept.

 

[w0033944]

It seems from the above that Zs1 does not require drive-on-sight. A most dangerous concept.

Presumably, the west/northbound train was running under clear signals, while the east/southbound service was affected by the signalling failure, but was allowed through at full linespeed the red aspect of the main signal by this subsidiary aspect?hock:

 

[Bletchleyite]

Presumably, the west/northbound train was running under clear signals, while the east/southbound service was affected by the signalling failure, but was allowed through at full linespeed the red aspect of the main signal by this subsidiary aspect?hock:

This is the impression I have.

If true, if my journey home was going to be on a German single line I would be considering the bus.

 

[ComUtoR]

Presumably, the west/northbound train was running under clear signals, while the east/southbound service was affected by the signalling failure, but was allowed through at full linespeed the red aspect of the main signal by this subsidiary aspect?hock:

Potentially that may be true but it would also mean that the signal system has failed somewhere else too and that both the Signaller and Driver were both at fault. The reports are stating that this is human error by the Signaller.

There are still more questions to be asked and importantly we need to know for certain what signals were seen by the Drivers of both trains.

This being human error does not mean that the system is at fault or inherently unsafe.

 

[gordonthemoron]

This is the impression I have.

If true, if my journey home was going to be on a German single line I would be considering the bus.

If travelling outside a conurbation, it is unlikely that there would be a competing bus

 

[BRX]

This is the impression I have.

If true, if my journey home was going to be on a German single line I would be considering the bus.

I reckon I'd still go with the train on the basis that the risk of an accident being caused by human error is even higher on a bus.

 

[w0033944]

Potentially that may be true but it would also mean that the signal system has failed somewhere else too and that both the Signaller and Driver were both at fault. The reports are stating that this is human error by the Signaller.

There are still more questions to be asked and importantly we need to know for certain what signals were seen by the Drivers of both trains.

This being human error does not mean that the system is at fault or inherently unsafe.

My confusion is centred around the lack of certainty here as to what this emergency signalling system does in terms of how drivers are required to respond to it. Must they proceed within line of sight regulations, or are they allowed to do full linespeed?

 

[ComUtoR]

My confusion is centred around the lack of certainty here as to what this emergency signalling system does in terms of how drivrs are required to respond to it. Must they proceed within line of sight regulations, or are they allowed to do full linespeed?

If they are given the Zs1 signal then they proceed at no more than 40kph till they are past the part of the line to which the Zs1 relates to.

It is reasonable to assume that the Driver proceeded past a Zs1 signal at 40kph then accelerated towards the crossing and past the crossing signal at green. At the point of impact it could therefore be going line speed.

It is reasonable that the oncoming train was in fact being brought to a stand by INDUSI as it was also reported that it was going slower and at the point of impact. The fast moving train entered the section and the system tripped the INDUSI.

Why weren't both trains brought to a stand by INDUSI
Why did the signaller authorise the Driver one of the trains to enter an occupied section.
Why wasn't procedure followed for allowing a Zs1 signal to be shown (line must be checked ?)
Why didn't interlocking prevent the Zs1 from being shown.

At the moment there are too many unanswered questions. The police have arrested the Signaller based on the probability of it being human error on his part.
--- old post above --- --- new post below ---

within line of sight regulations

drive by sight

I am very confused to what is meant by these terms

 

[w0033944]

If they are given the Zs1 signal then they proceed at no more than 40kph till they are past the part of the line to which the Zs1 relates to.

It is reasonable to assume that the Driver proceeded past a Zs1 signal at 40kph then accelerated towards the crossing and past the crossing signal at green. At the point of impact it could therefore be going line speed.

It is reasonable that the oncoming train was in fact being brought to a stand by INDUSI as it was also reported that it was going slower and at the point of impact. The fast moving train entered the section and the system tripped the INDUSI.

Why weren't both trains brought to a stand by INDUSI
Why did the signaller authorise the Driver one of the trains to enter an occupied section.
Why wasn't procedure followed for allowing a Zs1 signal to be shown (line must be checked ?)
Why didn't interlocking prevent the Zs1 from being shown.

At the moment there are too many unanswered questions. The police have arrested the Signaller based on the probability of it being human error on his part.
--- old post above --- --- new post below ---



I am very confused to what is meant by these terms

The BBC reports that both trains were travelling at 100 km/h:

The trains crashed head-on while both were travelling at about 100km/h (60mph) east of Bad Aibling, a spa town about 60km south-east of Munich.

 

[30907]

-See EAD's post #110 a few days ago. I'm sure s/he will be back to help us understand.

On a quick skim of other reports, I don't think the BBC's claim about speed (which was made in earlier reports too) has been confirmed today.

However, it does seem that Zs1 allows travel at line speed (?), unlike Zs7. And the westbound train would have no reason to slow down until preparing for the stop at BA Kurpark (or a red intermediate signal before it).

 

[ComUtoR]

The BBC reports that both trains were travelling at 100 km/h:

However, it does seem that Zs1 allows travel at line speed, unlike Zs7.

My confusion is centred around the lack of certainty here as to what this emergency signalling system does in terms of how drivrs are required to respond to it. Must they proceed within line of sight regulations, or are they allowed to do full linespeed?

If they are given the Zs1 signal then they proceed at no more than 40kph till they are past the part of the line to which the Zs1 relates to.

Which in this case may be as close as the pointwork from the main entry signal to the single line. Hence my statement to say it is reasonable to assume the Driver was within regulations by passing at 40kph then accelerating.

It would then eliminate the Zs1 as the cause of the accident. From what I understand the Zs1 is for a localised problem specific to that signal. Even if, as Neil states, you were allowed to proceed at far as the next signal only; there is still the question of why that signal appears to have been showing green. As evidenced by the trains speed.

Immediate cause is that both trains occupied the same section of line. How that happened and why it was allowed to happen is still unknown.

 

[AlexNL]

Why weren't both trains brought to a stand by INDUSI

Indusi / PZB works with magnets located besides the signals, and a pick-up coil which is located to the right of the loco / first carriage. Once the train has passed the magnet, it can't be controlled until it passes the next magnet.

Why did the signaller authorise the Driver one of the trains to enter an occupied section.
Why wasn't procedure followed for allowing a Zs1 signal to be shown (line must be checked ?)

The only thing I can think of is human error: the signaller must've been under the impression that the opposite train was already in Bad Aibling, not underway from Kolbermoor.

Why didn't interlocking prevent the Zs1 from being shown.

I am not aware of how the interlocking works in Bad Aibling, but I would not be surprised if the Zs1 is not linked to the interlocking at all.

As Zs1 is meant to substitute the normal signal, it would not be possible to give a Zs1 when the normal signal displays a danger aspect because of a fault further down the line (broken train detection / axle counters).

Zs1 is equivalent to verbal communication by the signaller over GSM-R, authorising the driver to pass a signal at danger.

At the moment there are too many unanswered questions. The police have arrested the Signaller based on the probability of it being human error on his part.

From what I understood, the signaller has not been taken into custody. An investigation into unvoluntary manslaughter has been opened by the police, the man himself has been taken to a safe place as he is 'not doing well'.

 

[ComUtoR]

Indusi / PZB works with magnets located besides the signals, and a pick-up coil which is located to the right of the loco / first carriage. Once the train has passed the magnet, it can't be controlled until it passes the next magnet.

As this is beyond my knowledge I will keep that in mind. I was under the belief that the magnets would slow you to a stand if you did not acknowledge various in cab alerts before approaching the red. Is it then correct to assume that they do not have a system like TPWS and the INDUSI works the braking curve based on the signal passed and not approached ?

The only thing I can think of is human error: the signaller must've been under the impression that the opposite train was already in Bad Aibling, not underway from Kolbermoor.

Human error seems to be the most likely cause

I am not aware of how the interlocking works in Bad Aibling, but I would not be surprised if the Zs1 is not linked to the interlocking at all.

As Zs1 is meant to substitute the normal signal, it would not be possible to give a Zs1 when the normal signal displays a danger aspect because of a fault further down the line (broken train detection / axle counters).

This is where I am also in the dark. We still do not know what signals the Drivers saw or were given.

Zs1 is equivalent to verbal communication by the signaller over GSM-R, authorising the driver to pass a signal at danger.

As discussed earlier. This is much the same as PoSA's in the UK.

From what I understood, the signaller has not been taken into custody. An investigation into unvoluntary manslaughter has been opened by the police, the man himself has been taken to a safe place as he is 'not doing well'.

The Beeb are very bad at reporting and I still see much of their reporting as sensationalised to generate headlines


Alex,

Many thanks for further insights.

 

[amcluesent]

Echoes of Quintinshill with signallers ignoring the rules, then overlooked the presence of another train.

 

[Bertiethebobby]

My understanding from the BBC report is that both trains were running at linespeed - 100 km/h - which seems inconsistent with one of them being under "drive on sight" regulations. Is this correct?

I think we can discount the idea they were both running at line speed.
Although very bad, that crash didn't look like they were both doing 60mph to me. AFAIK only the 1st coach of each train derailed, if they'd been at line speed I would have expected most of the coaches to be off the road.

I saw somewhere a "leaked" report that one of the black boxes said 40kph.
--- old post above --- --- new post below ---

Echoes of Quintinshill with signallers ignoring the rules, then overlooked the presence of another train.

Another possible similarity is shift changeover time.
The accident occurred at 06.45 just when you would imagine the shift changing and 2 people in the box. This is well documented as a dangerous time in many occupations not just railways.
--- old post above --- --- new post below ---

Echoes of Quintinshill with signallers ignoring the rules, then overlooked the presence of another train.

In the harsh light of investigation it seems incredulous that a professional, well trained, qualified person in a safety critical role can simply "forget" something, especially something as big as a train. However, there have been many accidents caused by this very thing, the signalman convincing himself that the situation is not as it really is, thankfully getting fewer and fewer as technology is introduced to mitigate human failings.

Not just in the rail industry either, instances of pilots shutting down the wrong engine, flying into mountains because they thought the instruments were wrong, etc etc.

 

[AlexNL]

As this is beyond my knowledge I will keep that in mind. I was under the belief that the magnets would slow you to a stand if you did not acknowledge various in cab alerts before approaching the red. Is it then correct to assume that they do not have a system like TPWS and the INDUSI works the braking curve based on the signal passed and not approached ?

Yes and no.

When a train passes a Vorsignal (distant signal, placed 1250m before the main signal) which shows a yellow aspect, a 1000 Hz magnet initiates a braking curve on board of the train. The driver has to start slowing down, within 23 seconds he has to reduce his trains' speed to a maximum of 85 km/h. After 700 metres, the driver should be able to see the main signal - if it has cleared, he can end the braking curve, otherwise he'll have to continue braking.

Once the train passes over the second magnet (500 Hz, 250m before the main signal) the driver is not allowed to exceed 65 km/h and he should not have released from the curve. If these conditions are not met the emergency brakes are triggered, otherwise a new braking curve starts: within 150 metres the train has to decelerate further from 65 km/h to 45 km/h, expecting to slow to 0.

250 meters after the 50 Hz magnet is the main signal, with a 2000 Hz magnet besides it. If the signal displays a red aspect the emergency brakes will be triggered, unless when the driver uses the PZB Override button.

Here's the PZB braking curve in a graph, image courtesy of Wikipedia:

Please note that different limits apply to freight trains.

 

[ComUtoR]

That is how I understood it :/ I think we were talking cross purposes.

EAD mentioned that one of the units received the 500hz magnet for the approaching red. That made sense because it would be interlocked to the red to provide a mechanic should the signal revert to red in an emergency.

I'm trying to understand why the approaching train that is alleged to be going line speed appears to have either clear signals, was authorised to override or manually overrode the PZB. Where it has been stated the system was working correctly it had to have been human error.

I just cannot fathom why it was possible to go such a speed unless the signals were green.

We are all assuming the Driver had the Zs1 as that appears to be the only mechanism to allow someone into an occupied section. We are also assuming that it was the train travelling at speed that was at fault. It would be reasonable to assume the Signaller set the train from Kolbermoor in error as the point of impact means the train at speed passed 3 ? signals, none of which in my mind should have been showing green.

However it has been reported that both Drivers were shown greens. Which in itself, poses more questions as to how that is even possible and I would worry if it wasn't interlocked.

Thanks again Alex.

 

[MarkyT]

I am not aware of how the interlocking works in Bad Aibling, but I would not be surprised if the Zs1 is not linked to the interlocking at all. As Zs1 is meant to substitute the normal signal, it would not be possible to give a Zs1 when the normal signal displays a danger aspect because of a fault further down the line (broken train detection / axle counters).

As discussed earlier. This is much the same as PoSA's in the UK.

It is most surprising and worrying if the Zs1 signal circuits carry out no opposing route checking at all. The UK PoSA definitely DOES check opposing route locking which as I said before does not release until the train taking the conflicting route has cleared the section. in the UK if on a single line the route locking remains active in the opposing direction to the next intended movement because of a track circuit failure it is impossible to defeat the interlocking and display ANY kind of signal aspect so it becomes necessary to implement pilotman working. If the problem was casued by an axle counter miscount, a 'sweep train' going through in the same direction as the first movement might clear the section and release the locking. Apart from that any override to the interlocking is impossible in the UK with the exception of certain permissive moves, usually into platforms, in which case timeouts occur that would effectively prove that the first movement (if it was still there in the section) had already come to a stand.