• Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!

Another fatal rail crash, Bavaria, Germany (May 2018)

Status
Not open for further replies.

BRX

Established Member
Joined
20 Oct 2008
Messages
3,599
Is there something fundamentally wrong with Germany's signalling arrangements? This comes not long after the Bad Aibling disaster just a couple of years ago, and looks also to have been caused by signaller inattention.

German police on Tuesday arrested a rail traffic controller over the previous night's train collision that claimed two lives.
The unnamed 24-year-old man has been detained on suspicion that his negligence led to the accident, said the regional police service. Prosecutors are now examining whether to file a request to remand him in custody.

A commuter service late Monday rammed into a parked freight train just a few hundred metres from Aichach station, in southern Bavaria.

The Deutsche Bahn network operator said that a commuter service hit a freight train between Ingolstadt and Augsburg near Aichach station in the country's southern Bavaria region.

"There are some seriously injured and others only lightly," a Deutsche Bahn spokesman told the German DPA press agency.

German police said that a total of 14 people were hurt.

According to a police spokesman, The 37-year-old train driver and a 73-year-old female passenger were killed.

The company did not immediately provide additional details into the possible causes of the collision.

Shortly after the accident a spokesperson said the two trains had not derailed.

The passenger train which had almost 20 people on board had driven toward the still, empty freight train and jammed into it. How fast the passenger train was going could not be clarified late on Monday.

It was also unclear whether signals or switches were possibly set incorrectly.

The Ingolstadt-Augsburg train route has since been closed.

The closure around Aichach station will last at least until Tuesday afternoon. No long-distance trains run through this station, but rather, only regional and freight trains.

The collision brought back memories of a horrific train crash which happened in February 2016, when two trains collided in Bavaria. Nine passengers were killed and around 80 more were injured.
https://www.thelocal.de/20180508/2-dead-14-injured-as-trains-collide-in-bavaria
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

BRX

Established Member
Joined
20 Oct 2008
Messages
3,599
Passed me by at the time but it looks like there was another collision in 2017 -

http://www.dw.com/en/passenger-train-collides-with-freight-train-northwest-of-düsseldorf/a-41666107

thankfully no-one killed in that one. I've not been able to find yet whether a cause for that has been established yet.


edit - according to Wikipedia

The Eisenbahn-Unfalluntersuchungsstelle des Bundes (EUB) opened an investigation into the accident.[6] It was reported that the passenger train may have incorrectly been given permission to pass a signal indicating “stop”.[2] Having been given permission, the driver should have proceeded at a speed not exceeding 40 kilometres per hour (25 mph) and have been able to stop in the distance he could see ahead.[7]

https://en.wikipedia.org/wiki/Meerbusch_train_crash
 

AM9

Veteran Member
Joined
13 May 2014
Messages
14,191
Location
St Albans
Zs1 again, perchance? Really needs abolishing as it is clearly being completely misused.

From the Youtube video it looks like a class 648 Alsthom where collision structural damage was not much more than the front crumple zone and the driver's cab. Not much consolation for the two fatals (or any of the injured) but had it been at a higher speed or a derailment collision, the casualties could have been a lot higher. These low-floor designs don't seem to have much more crash resistance than a reasonable tramcar.
 

duesselmartin

Established Member
Joined
18 Jan 2014
Messages
1,902
Location
Duisburg, Germany
It does sound like a Zs1. Signalling ist often outdated and it seems that the rule book is ignored. That seems systemic.
The line in question has mechanical semaphore signalling.
 

Bletchleyite

Veteran Member
Joined
20 Oct 2014
Messages
97,539
Location
"Marston Vale mafia"
From the Youtube video it looks like a class 648 Alsthom where collision structural damage was not much more than the front crumple zone and the driver's cab. Not much consolation for the two fatals (or any of the injured) but had it been at a higher speed or a derailment collision, the casualties could have been a lot higher. These low-floor designs don't seem to have much more crash resistance than a reasonable tramcar.

This isn't that surprising as they are a vastly bigger "bog roll tube" than a UK design.
 

Bletchleyite

Veteran Member
Joined
20 Oct 2014
Messages
97,539
Location
"Marston Vale mafia"
It does sound like a Zs1. Signalling ist often outdated and it seems that the rule book is ignored. That seems systemic.
The line in question has mechanical semaphore signalling.

How many more will there be before they remove Zs1 completely as they need to?

I just can't see any sense in a system where the solution to a block being occupied is to think it isn't and send a train in anyway. If the system is so bad that you get enough failures to need to do that, it needs replacing, even if it's something as simple as a train staff.

Mind you, given that it was a rear-ending of a stationary train it would also seem to raise questions as to what the driver was doing rather than driving on sight as he is meant to when he receives a Zs1 in my understanding. That particular part of the incident could also occur in the UK.
 

carriageline

Established Member
Joined
11 Jan 2012
Messages
1,897
We have POSA (Proceed on sight authority) signals over here which work very similarly. I don’t think the answer should be to remove them over there, but questions need to be asked on why/when they are being used...
 

Bletchleyite

Veteran Member
Joined
20 Oct 2014
Messages
97,539
Location
"Marston Vale mafia"
We have POSA (Proceed on sight authority) signals over here which work very similarly. I don’t think the answer should be to remove them over there, but questions need to be asked on why/when they are being used...

There is one key difference between here and there (which isn't strictly relevant to this incident, but was to the previous one) - we don't (AIUI) use those onto single or bidirectional lines, they do.

The obvious difference in severity between this incident and the previous head-on is a good reason. Zs1 onto a single line is just asking for another Cowden.
 

carriageline

Established Member
Joined
11 Jan 2012
Messages
1,897
There is one key difference between here and there (which isn't strictly relevant to this incident, but was to the previous one) - we don't (AIUI) use those onto single or bidirectional lines, they do.

The obvious difference in severity between this incident and the previous head-on is a good reason. Zs1 onto a single line is just asking for another Cowden.

We can, and do use POSAs on BiDirectional lines. Heck, even Single lines COULD do.

But, ours are (nearly) fully interlocked, and will lock routes in the opposite, and require routes in the opposite direction to be cancelled. I imagine they do not have that over there!
 

Bletchleyite

Veteran Member
Joined
20 Oct 2014
Messages
97,539
Location
"Marston Vale mafia"
We can, and do use POSAs on BiDirectional lines. Heck, even Single lines COULD do.

But, ours are (nearly) fully interlocked, and will lock routes in the opposite, and require routes in the opposite direction to be cancelled. I imagine they do not have that over there!

Yes, that's quite different - the explicit purpose of Zs1 is to override a failed interlock, AIUI.
 

carriageline

Established Member
Joined
11 Jan 2012
Messages
1,897
Yes, that's quite different - the explicit purpose of Zs1 is to override a failed interlock, AIUI.

To be fair, we have a few more modern locations where we can also do that! But the entire process and procedure is very locked down and strict, and it’s in supervised locations where permission needs to be saught first.

It sounds like their procedure is very lacking, or is just not being followed
 

AM9

Veteran Member
Joined
13 May 2014
Messages
14,191
Location
St Albans
... Mind you, given that it was a rear-ending of a stationary train it would also seem to raise questions as to what the driver was doing rather than driving on sight as he is meant to when he receives a Zs1 in my understanding. That particular part of the incident could also occur in the UK.

Actually by the look of the scene in this video:


it looks like a head-on into a stationary freight as it collided into the loco.

Here is the first youtube video that I came across, (caution, it shows rather more than UK broadcast news does):


The centre cab diesel loco can be seen here.
 

Spoorslag '70

Member
Joined
29 Oct 2017
Messages
272
Location
Garching (b. München)
From what I understand, there are no track circuits down there - it seems to rely on the signaller's memory. A quite unpleasant situation, considering that a human has a limited memory. Is there a box in the UK with no detection?
 

AlexNL

Established Member
Joined
19 Dec 2014
Messages
1,683
From what I understand, Aichach has a mechanical box with semaphore signalling. A track occupation detection system is not present in that kind of boxes, everything is covered through operating procedures and log books.

Presumably, the signaller made an error by forgetting to check if the track was occupied when he sent a train there.
 

30907

Veteran Member
Joined
30 Sep 2012
Messages
17,867
Location
Airedale
As far as I can make out (from the discussion on Drehscheibe-online https://www.drehscheibe-online.de/foren/read.php?002,8553152,page=6) the collision took place within station limits. It is alleged that the signaller routed the passenger train onto the track occupied by the stationary freight, which had drawn right down to the exit signal and was therefore a considerable way from the signaller's position.
Zs1 hasn't been mentioned, but the absence of track circuits has. Remember that German signalling doesn't routinely indicate the route, and the incident happened around 21:15, half an hour after sunset.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
From what I understand, there are no track circuits down there - it seems to rely on the signaller's memory. A quite unpleasant situation, considering that a human has a limited memory. Is there a box in the UK with no detection?
The traditional "absolute block" system still used on some lines does not require detection. The signallers either end of the section keep track of whether a train has been accepted or is in the section using a three-position needle instrument, and the section can only be released when the signaller at the exit sees the train pass complete with tail lamp. As well as the block instrument the signaller is required to keep a register book which they can use to check back if they are uncertain about what is happening, as well as acting as a record in case of accidents.

The section around the signal box is known as "station limits" and here - as apparently in the accident in this thread - the signaller originally had to rely on observation and memory of where trains were located. However there were some bad accidents when a train was forgotten, usually because the crew had also failed to remind the siganall of their presence. So track circuits started to be fitted in the areas around the boxes, allied to various controls that aimed to enforce the correct sequence of operations, but retaining the principle of no detection in the block section itself. I would guess all of the surviving absolute block boxes now have these extra protections, and many otherwise traditional looking boxes no longer operate on the principle of absolute block.
 

Gostav

Member
Joined
14 May 2016
Messages
413

According to my observation, many European branch line in poor area use very simple system, the levers just near the switches non-interlocking and be changed by the train guard or ground staff.
 
Last edited:

BRX

Established Member
Joined
20 Oct 2008
Messages
3,599
According to my observation, many European branch line in poor area use very simple system, the levers just near the switches non-interlocking and be changed by the train guard or ground staff.
Sure...but Germany is one of the wealthiest and most technologically developed countries in Europe. It's surprising to me that safety systems on its railways appear to be significantly more lax than in the UK.
 

Bletchleyite

Veteran Member
Joined
20 Oct 2014
Messages
97,539
Location
"Marston Vale mafia"
Sure...but Germany is one of the wealthiest and most technologically developed countries in Europe. It's surprising to me that safety systems on its railways appear to be significantly more lax than in the UK.

I think there is a cultural issue - in Germany the sense of duty and of following rules would suggest that a procedure is adequate, whereas in the UK we have less confidence in people following procedures precisely because it isn't as much in our culture to do so, so we have provisions such as interlocks to ensure that a person can't make a mistake and kill people.

Mind you there is a simple, cheap and effective solution that would have prevented this one on a single line with a 2 track station - sprung points. Then the signaller only has to concern himself with departure rather than arrival as well.
 
Last edited:

duesselmartin

Established Member
Joined
18 Jan 2014
Messages
1,902
Location
Duisburg, Germany
According to my observation, many European branch line in poor area use very simple system, the levers just near the switches non-interlocking and be changed by the train guard or ground staff.

In this case the station is signal box operated by a signal man.
However it does show that the network suffers from underinvestment.
 

BR111

Member
Joined
28 May 2019
Messages
30
I'm late to the party, but someone linked this thread in one of mine to see what I thought of the situation.

How many more will there be before they remove Zs1 completely as they need to?

I just can't see any sense in a system where the solution to a block being occupied is to think it isn't and send a train in anyway. If the system is so bad that you get enough failures to need to do that, it needs replacing, even if it's something as simple as a train staff.

Mind you, given that it was a rear-ending of a stationary train it would also seem to raise questions as to what the driver was doing rather than driving on sight as he is meant to when he receives a Zs1 in my understanding. That particular part of the incident could also occur in the UK.
That's not how Zs 1 works. Zs 1 is used if there is a signal fault but the signaller is certain that there is nothing in there. If the signalbox still says that the track is occupied, then it is either done with a Zs 1 in addition to a Befehl ('command') to drive on sight until the next main Signal, or the Signal Zs 7 can be used (essentially a Zs 1 with the addition of the command to drive on sight). The problem with these signals isn't that they are inherently unsafe to use, it's that they aren't used properly. And that's what causes the accidents unfortunately. It might be worth noting that a good amount of drivers are always suspicious of a Zs 1 and ring the signaller up to confirm and ask why it was pressed. In my experience, a good deal of Signallers will tell you in advance that is is being pressed and why, although none of this is required. If I see Zs 1, that's the only permission I need to proceed. I could see a verbal confirmation being required in the future, but I don't think Zs 1 will be completely abolished.

Yes. The Signaller sits locally and can see the tracks, however. If there is something he cannot see (in exceptional circumstances due to bad weather or darkness - during the day everything must be visible), he is required to send the next train in on sight. Having no system to identify track occupation is exclusive to these kind of signalboxes.
 
Last edited:

BRX

Established Member
Joined
20 Oct 2008
Messages
3,599
Yes. The Signaller sits locally and can see the tracks, however. If there is something he cannot see (in exceptional circumstances due to bad weather or darkness - during the day everything must be visible), he is required to send the next train in on sight. Having no system to identify track occupation is exclusive to these kind of signalboxes.
This just applies to sections within stations - or the sections of tracks between stations too?
 

BR111

Member
Joined
28 May 2019
Messages
30
This just applies to sections within stations - or the sections of tracks between stations too?
Only sections within stations. For a train to leave a station first permission must be asked from the next signalbox. The following signaller would then have to look out for the train end markers/lights and only then could the line be declared empty again.
 

Bletchleyite

Veteran Member
Joined
20 Oct 2014
Messages
97,539
Location
"Marston Vale mafia"
I'm late to the party, but someone linked this thread in one of mine to see what I thought of the situation.

That's not how Zs 1 works. Zs 1 is used if there is a signal fault but the signaller is certain that there is nothing in there.

The trouble is (and the cause of this accident was) that it trusts one person's judgement that nothing is in there. In the UK we do not do that; if a driver was instructed to pass a signal at danger they would, I believe, do so driving on sight until the next correct signal. Perhaps the issue with Zs1 is that it doesn't require driving on sight by default, then?

(On single or bidirectional lines this would be different due to the risk of head-on collision - can Zs1 be used onto this kind of line? In the UK I believe this would require pilotman working?)
 

BRX

Established Member
Joined
20 Oct 2008
Messages
3,599
Only sections within stations. For a train to leave a station first permission must be asked from the next signalbox. The following signaller would then have to look out for the train end markers/lights and only then could the line be declared empty again.

But does it rely entirely on the signallers observation/memory - there is no token, physical or otherwise, involved?
 

BR111

Member
Joined
28 May 2019
Messages
30
The trouble is (and the cause of this accident was) that it trusts one person's judgement that nothing is in there. In the UK we do not do that; if a driver was instructed to pass a signal at danger they would, I believe, do so driving on sight until the next correct signal. Perhaps the issue with Zs1 is that it doesn't require driving on sight by default, then?

(On single or bidirectional lines this would be different due to the risk of head-on collision - can Zs1 be used onto this kind of line? In the UK I believe this would require pilotman working?)
I'm not sure I can answer this to be honest - I'm not signaller, but I know that Zs 1 is a lot more common than Zs 7 (which you almost only find on entry signals and a few block signals), but that if the reason for Zs 1 is due to some kind of track occupation error, then you will always additionally receive a command to drive on sight. An example of where Zs 1 without driving on sight would make sense is if the Signal cannot be put on green because a level crossing tied to the Signal has failed, but there is staff there to secure it - there is no need to drive on sight in this situation.

But does it rely entirely on the signallers observation/memory - there is no token, physical or otherwise, involved?
This depends on the system - we have something called Zugleitbetrieb which is used on very few routes, and it is more or less a token system. But one-tracked lines are by no means a reason to set up a token system. What exactly this requires, I'm afraid I also cannot answer. I drive on a few single-lined routes and on none of them we have a token system. I'm pretty sure the signallers have to confirm the complete arrival of the train every time but I'm not in the loop of the specifics of their job.
 

Bletchleyite

Veteran Member
Joined
20 Oct 2014
Messages
97,539
Location
"Marston Vale mafia"
I'm not sure I can answer this to be honest - I'm not signaller, but I know that Zs 1 is a lot more common than Zs 7 (which you almost only find on entry signals and a few block signals), but that if the reason for Zs 1 is due to some kind of track occupation error, then you will always additionally receive a command to drive on sight. An example of where Zs 1 without driving on sight would make sense is if the Signal cannot be put on green because a level crossing tied to the Signal has failed, but there is staff there to secure it - there is no need to drive on sight in this situation.

The problem is that that still requires the signaller to remember that that's the situation. To me it's a lot safer to have a strict rule of "you only ever pass a signal at danger on a unidirectional line, and you always drive on sight to the next correct signal" - there can't be any misunderstanding then.

The principle as used in the UK is that trusting one person's recollection can be quite dangerous.

This depends on the system - we have something called Zugleitbetrieb which is used on very few routes, and it is more or less a token system. But one-tracked lines are by no means a reason to set up a token system. What exactly this requires, I'm afraid I also cannot answer. I drive on a few single-lined routes and on none of them we have a token system. I'm pretty sure the signallers have to confirm the complete arrival of the train every time but I'm not in the loop of the specifics of their job.

That is quite an interesting difference, as in the UK you are on a single or bidirectional line either required to have a token/train staff system of some kind (can be physical or electronic), track circuiting/axle counters interlocked with the signalling (and if they fail, to assume a train is in section even if the signaller doesn't think there is) or ETCS. This has been the case for many, many years - since well before the modern technologies listed as alternatives.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
That is quite an interesting difference, as in the UK you are on a single or bidirectional line either required to have a token/train staff system of some kind (can be physical or electronic), track circuiting/axle counters interlocked with the signalling (and if they fail, to assume a train is in section even if the signaller doesn't think there is) or ETCS. This has been the case for many, many years - since well before the modern technologies listed as alternatives.
I think I've seen some somewhere (in the UK) where a dead-end section wasn't track circuited but the train had to occupy several track circuits in sequence when exiting the section before the next one could be signalled in, so as to protect against track circuit failures giving a false indication that the train was out.
 
Status
Not open for further replies.

Top