Nine dead, 50 seriously injured after head-on train crash in Bavaria, Germany

[Commander Wallace]

Two trains collided earlier this morning Southeast of Bavaria. Altogether, an estimated 100-150 people were injured, 50 of them seriously in addition to the 9 dead at the time of this writing.


Quote : The crash occurred shortly before 7 a.m. Tuesday local time near the spa town of Bad Aibling, about 60 kilometers (37 miles) southeast of the Bavarian capital of Munich.




It is estimated the 2 trains were traveling at 62 MPH or 100 KPH at the moment of impact. 2 of the 3 data recorders have been recovered and it's expected the 3rd will be recovered in short order. In an effort to determine the cause of the accident, Investigators are focusing their efforts on technical issues, human error and other causes.

Quote : The crash occurred on the Mangfall Valley Railway, a single-track regional rail line between the towns of Rosenheim and Holzkirchen, according to train company Meridian.
Read More

"This is a difficult hour for the railways in Germany," said German Transport Minister Alexander Dobrindt, describing the crash as "one of the biggest we have had for years."


http://www.cnn.com/2016/02/09/europe...ain-collision/


Hopefully, our German Subsim members and their familes are safe.


Condolences to those lost and best wishes for the recovery of those injured.

[JasonKR]

May each and every one of the 50 recover.

[I'mOnABoat]

And here I thought trains were the safer way to go!

[Schroeder]

Originally Posted by I'mOnABoat

And here I thought trains were the safer way to go!

They are compared to all other means of transportation.

[Oberon]

To say that something has gone terribly wrong would obviously be a drastic understatement, head on collisions are relatively rare in the railways of the western world, when there are crashes they're usually either into the rear of a preceding train or occasionally into another train which is changing lines in front (or of course, derailment due to poor track or train maintenance).
The fact that this was a single line as well only adds to the rarity of such a thing happening, generally speaking the safety systems of European trains are supposed to stop two trains being on the same track at the same time.
So either it was human negligence caused by overriding or ignoring safety systems (the PZB I believe it's called), or a failure of the safety systems that are supposed to stop this.

[August]

Originally Posted by JasonKR

May each and every one of the 50 recover.

Amen.

[Jimbuna]

Officials say they assume both train drivers had no visual contact before the crash as the site is on a bend - and therefore crashed into each other largely without braking.
The stretch of line had an automatic braking system designed to halt any train that passed a stop signal. It is not yet known why this failed to stop the crash but two of the three data recorders or "black boxes" on board the trains have been recovered so they should have a clearer indication of what went wrong soon.

[Oberon]

Originally Posted by Jimbuna

The stretch of line had an automatic braking system designed to halt any train that passed a stop signal. It is not yet known why this failed to stop the crash but two of the three data recorders or "black boxes" on board the trains have been recovered so they should have a clearer indication of what went wrong soon.

Yeah, that's the PZB system, it's a bit like the TPWS in the UK in that there's a magnet near the signals that operates on varying frequencies depending on the aspect of the signal, if the signal is red and the train approaches it too fast or tries to go past it it should automatically trip the brakes (the London Underground has a similar but more manual system in which a small lever comes out when a signal is red and if the train passes it, it hits the tripcock on the train and triggers the brakes). Likewise if the signal is at yellow or whatever the German equivilient is, there's automatically a speed restriction in place and if the train goes past or approaches the signal at a higher speed the brakes are applied.

That being said, PZB is not infallable, like our TPWS it can be overridden (a major disaster with a steam locomotive was narrowly averted the other year when a steam tour operator isolated the TPWS in the steam locomotive causing it to pass a signal at danger and come to a stop fouling a busy main line junction) but there are procedures in place that are meant to prevent this from happening.

There had been at least one major accident with the PZB90 in place - on 26 June 2000 an S-Bahn train left Hannover-Langenhagen station for a single-track section with an oncoming train. The PZB halted the train but the driver released the train ("Freitaste") without double-checking with the train director. The investigative report notes that there had been 22 similar recorded occurrences until that time when a driver related the PZB halt to a different cause than having overrun a main signal - the report concludes that the operations manual should be changed in that double-checking with train director should not only be required on a main signal overrun but explicitly on all PZB-related stops.

So how one of these trains (most likely the one which had just left Kolbormoor station which is, I believe, double tracked so would make a natural passing place) managed to drive through a red light and onto a single track section with another train coming towards it does lean heavily towards a signalling failure, as a driver leaving a station would have had more time to see the red signal and thus would have been less likely to pass it.
If so, it will be the first time that the PZB system has failed causing a fatal accident which was not directly linked to being overridden or altered by the driver.

Still, on the upside at least this was modern rolling stock, with better design and safety features than old stock, and with recording devices on board which will give investigators an idea of what happened. No consolation to the relatives of the dead or injured though, sadly.

[Skybird]

Media report that the investigating police now is very strongly focussed on human error made by a traffic regulator. All tracks seem to lead there.

[kraznyi_oktjabr]

Gentlemen, you seem to be making assumption which is not necesssarily true. That is that there was signal displaying danger aspect. In early 2000s in Savonia Railway in Finland between Kouvola and Mikkeli (don't remember exact spot) was incident where two passengers trains travelling in opposite directions had routes set and protected into same single tracked block with all main and distant signals displaying Proceed and Expect Proceed aspects.

Disaster was avoided only because one of train drivers remembered that there was supposed to be passing at loop and he should have had route set into loop track. Therefore distant signal should have had Expect 35 (proceed at maximum speed of 35 km/h, points ahead are set for diverging route) aspect instead of Expect Proceed.

This isn't very unlike scenario but still possible.

[Oberon]

Originally Posted by kraznyi_oktjabr

Gentlemen, you seem to be making assumption which is not necesssarily true. That is that there was signal displaying danger aspect. In early 2000s in Savonia Railway in Finland between Kouvola and Mikkeli (don't remember exact spot) was incident where two passengers trains travelling in opposite directions had routes set and protected into same single tracked block with all main and distant signals displaying Proceed and Expect Proceed aspects.

Disaster was avoided only because one of train drivers remembered that there was supposed to be passing at loop and he should have had route set into loop track. Therefore distant signal should have had Expect 35 (proceed at maximum speed of 35 km/h, points ahead are set for diverging route) aspect instead of Expect Proceed.

This isn't very unlike scenario but still possible.

Originally Posted by Skybird

Media report that the investigating police now is very strongly focussed on human error made by a traffic regulator. All tracks seem to lead there.

I presume by traffic regulator you mean what we in English would call a signalman or dispatcher? This would correlate with what krazyni put as well in that both trains got a clear aspect. One would have thought in this day and age that there would be something in the computing system that would have picked this up and flashed up a warning to the signalman that he was about to clear a train onto an occupied block.
Of course, it's possible that the signal itself was at fault and showed a clear aspect when it should have shown a danger, this is what caused the train crash at Clapham Junction back in 1989 because of rushed engineering work an old wire had not been isolated and as a result 35 people died and nearly 500 were injured.

Human error in the signalling department wouldn't surprise me though, there are plenty of crashes in history which can be traced to that, the worst train crash in the UK was primarily down to signalling error, and it goes to show that despite all the computer back-ups we put into a system, it's still only as strong as its weakest point, which is invariably the fleshy thing operating it.

[kraznyi_oktjabr]

Originally Posted by Oberon

I presume by traffic regulator you mean what we in English would call a signalman or dispatcher? This would correlate with what krazyni put as well in that both trains got a clear aspect. One would have thought in this day and age that there would be something in the computing system that would have picked this up and flashed up a warning to the signalman that he was about to clear a train onto an occupied block.
Of course, it's possible that the signal itself was at fault and showed a clear aspect when it should have shown a danger, this is what caused the train crash at Clapham Junction back in 1989 because of rushed engineering work an old wire had not been isolated and as a result 35 people died and nearly 500 were injured.

In Savonia case such conflicting route assignment should have been impossible as computer system which controlled signals and points in that railway at the time should have prevented it from happening. I don't remember what turned out to be the problem here.

Originally Posted by Oberon

Human error in the signalling department wouldn't surprise me though, there are plenty of crashes in history which can be traced to that, the worst train crash in the UK was primarily down to signalling error, and it goes to show that despite all the computer back-ups we put into a system, it's still only as strong as its weakest point, which is invariably the fleshy thing operating it.

Unfortunately that is very true. In 1996 Jokela rail accident was caused by train driver missing signal and as result coming at excessive speed into junction and derailed. In 1998 Jyväskylä rail accident train approaching Jyväskylä railway station came at excessive speed to points most likely because (tired) train driver was driving by routine (based on reconstruction) and missed signals indicating changed routing. In normal routing train goes through 80 km/h points closer to station, but in changed route there was 35 km/h points farther away from station.

In 1998 Suonenjoki rail accident InterCIty train departing station passed signal at danger, drove through points and crashed head on with freight train. In this case driving through points (forcing them off their original position) triggered alarm in dispatchers panel but by then trains had already crashed.

Common nominator in all those accidents was lack of any onboard proteection system like your TPWS or PZB. Implementation of such systems was progressing slowly and only after two consecutive fatal accidents (Jokela and Jyväskylä) was program accelerated and considered urgent. This despite several near disasters, rear endings speeding incidents and so on over years. Implementation - albeit in more limited form - began slightly before 1956 Kuurila rail disaster

 

Steam locomotive hauled sleeper express train and diesel multiple unit operated express train crashed because of signalman's error. Heavy steam locomotive essentially drove through first two carriages of relatively light weigth DMU. Additionally as locomotive's tender derailed it crushed several wooden passenger cars behind it.

and was completed in main lines in early 21st century. Many secondary and branch lines (some with passenger service) still do not have any such system.

Writing this somewhat off topic post reminded me of another failure point. I don't know what is situation with German safety systems but until last year (any maybe still in some locomotives) ours is as standard turned off and must be turned on by train driver. This is done manually every time train is moved first time from that control position (cab in MU and box cab locos, control stand in hood units). That changed only after 2014 Hyvinkää incident where driver of commuter train forgot to turn system on and as result there was no protection from his second error of coming at excessive speed into points. Fortunately that didn't end into derailment.

[Aktungbby]

A small silver cloud it seems:

Due to Bavaria's Fasching version of this week's carnival celebrations, the train was initially believed to be carrying fewer passengers than would normally be expected during commuter hours. Many children would also have usually been travelling to school, but are currently on holiday.

"Tuesday morning was caused by "human error," anonymous sources close to the investigation told German media on Tuesday. Who exactly was responsible for the collision that left 10 people dead and dozens injured remained unknown, authorities said. Before the confirmation of human error, earlier theories as to the cause of the crash had circled around the system that forces trains to brake when they are on the wrong track in order to circumvent such tragedies. According to rail operator Deutsche Bahn, this computer system was enabled at the time of the crash and had just recently been tested for functionality."

[Skybird]

It is confirmed that it was human error committed by the "Fahrdienstleiter" (signal man, dispatcher, traffic regulator, however you call him in English). No alcohol and no intention, drug test results still missing, but most likely nothing there, too. He tried to radio both train via emergency calls and sent out warnings, but to no avail.

The man is 39 years old, and serves in his job since 20 years.

Just a terrible human mistake, like throwing a glass of water off the table with the ellbow. I cannot condemn him. I do not want to be in his skin now.

Charges are about "fahrlässige Tötung" - negligent homicide.

[Jimbuna]

A potential maximum sentence of five years in prison if found guilty.