Passenger train collision in Rafz, Switzerland

sprinterguy · 20 Feb 2015

A Zurich to Stuttgart inter-city train has collided with a commuter train at Rafz in Switzerland (19 miles north of Zurich) this morning:

http://www.msn.com/en-gb/news/uknews/dozens-hurt-in-swiss-train-collision/ar-BBhMevx?ocid=U218DHP
'Dozens Hurt' In Swiss Train Collision
4 hrs ago

Two trains have collided in Switzerland injuring up to 50 people, according to reports.

The collision happened at the train station in the Swiss town of Rafz, around 30 km (19 miles) north of Zurich.

The crash was between a commuter train and a a high-speed train on its way from Zurich to Stuttgart.

Eyewitnesses said the high-speed train hit the other from behind.

"There was an accident this morning, it's serious, there are injured," a police spokeswoman said.

"Ambulances from all regions have been mobilised," she added.

One rescue worker said as many as 49 people had been injured.

A passenger on the local train told the 20 Minutes newspaper that his train had been pulling out of the station when the driver braked suddenly.

"An express train from Zurich came up from behind and hit the side of our train - the intercity train derailed," said the man, who did not give his name.

The 18-year-old went on to say that passengers had quickly been evacuated from the trains, both of which were "quite damaged".

Rail services have been suspended.

The Swiss rail service is generally admired for its safety and efficiency - the Swiss are the biggest user of trains in Europe.

More pictures and video on the BBC report here:
http://www.bbc.co.uk/news/world-europe-31547979

Looking at the track layout at the station, and based on this account, it would appear that this was a side-swipe collision where the fast inter-city service has collided with the commuter train as it was departing, where a through line and the platform line merge into one.

RailUK Forums

Wild Swan · 20 Feb 2015

According to Swiss media, the express train involved was a Zürich-Schaffhausen Interregio service, I reckon he 06.05 from Zürich (apparently running 15 late), not a Zürich-Stuttgart high speed (!) service - there is no Stuttgart service at the time the collision occurred.

Simon11 · 20 Feb 2015

Looking at both of these sources, I would have thought that British rail service is actually safer than in Switzerland? There have been quite a few crashes/serious derailment over the last few years in Switzerland. By this, just looking over the last few years and based on accidents which are the fault of the operators rather than suicide.

"The Swiss rail service is generally admired for its safety and efficiency - the Swiss are the biggest user of trains in Europe."

"Switzerland's rail system is considered among the best and safest in the world."

Would the Swiss also be the biggest user of trains in Europe????

gordonthemoron · 20 Feb 2015

Journalist friend in Zurich says both trains were being driven by trainees

A-driver · 20 Feb 2015

gordonthemoron said:
Journalist friend in Zurich says both trains were being driven by trainees

Sounds completely irrelevant to me...I would guess trainees are supervised same as in the UK and it's just a coincidence, but one which is a journalists dream come true?!

Bletchleyite · 20 Feb 2015

Simon11 said:
Looking at both of these sources, I would have thought that British rail service is actually safer than in Switzerland? There have been quite a few crashes/serious derailment over the last few years in Switzerland. By this, just looking over the last few years and based on accidents which are the fault of the operators rather than suicide.

They also have a downright dangerous dispatch procedure on LHCS services.

1. Guard blows whistle (if multiple guards, they wave a signal to the main one)

2. Guard goes to platform panel and signals driver the RA

3. Guard returns to train and closes the doors, often without looking properly to see if they are obstructed

4. Driver departs (he has a small mirror to see back, but he can't see the whole train in many or even most cases).

Can anyone see the flaws in that?

I have on multiple occasions seen trains depart with doors open, FWIW. There is an automatic system that closes them periodically (closure pressure is applied about every 20 seconds I think, can be a pain when alighting from an old coach with a UIC slamdoor if that happens just before the train stops) after the train exceeds 5km/h, though.

Neil

richa2002 · 20 Feb 2015

Neil Williams said:
They also have a downright dangerous dispatch procedure on LHCS services.

1. Guard blows whistle (if multiple guards, they wave a signal to the main one)

2. Guard goes to platform panel and signals driver the RA

3. Guard returns to train and closes the doors, often without looking properly to see if they are obstructed

4. Driver departs (he has a small mirror to see back, but he can't see the whole train in many or even most cases).

Can anyone see the flaws in that?

I have on multiple occasions seen trains depart with doors open, FWIW. There is an automatic system that closes them periodically (closure pressure is applied about every 20 seconds I think, can be a pain when alighting from an old coach with a UIC slamdoor if that happens just before the train stops) after the train exceeds 5km/h, though.

Neil

It's called running a proper railway with minimal faffing. The Swiss obviously don't have much trouble with it but is a bit of an eye opener for the MDTR-bred railwaymen of this country.

Back on topic though, looks like the weather could have been foggy for this incident judging by the photos?

Bletchleyite · 20 Feb 2015

richa2002 said:
It's called running a proper railway with minimal faffing.

I don't think that changing it to:

1. Guard blows whistle

2. Guard checks doors are unobstructed and closes all but local door

3. Guard uses platform device to give RA having had a glance both ways to check the doors have closed

4. Driver departs

5. Guard closes local door

which is DB's procedure (add a 0 for DB of guard using platform device to give the "An Gleis N bitte einsteigen" autoannouncement), would be slow enough for it to be a problem.

Neil

D1009 · 20 Feb 2015

Some photos on this site.

http://www.zol.ch/blaulicht/

30907 · 20 Feb 2015

...which says one driver was seriously injured, 5 pax minor injuries.

ilkestonian · 20 Feb 2015

Neil Williams said:
They also have a downright dangerous dispatch procedure on LHCS services.

1. Guard blows whistle (if multiple guards, they wave a signal to the main one)

2. Guard goes to platform panel and signals driver the RA

3. Guard returns to train and closes the doors, often without looking properly to see if they are obstructed

4. Driver departs (he has a small mirror to see back, but he can't see the whole train in many or even most cases).

Can anyone see the flaws in that?

Neil

Only "guard checks signal aspect to see if he can give RA to driver" seems to be absent from their procedures.

ChiefPlanner · 20 Feb 2015

ilkestonian said:
Only "guard checks signal aspect to see if he can give RA to driver" seems to be absent from their procedures.

As brought in , on BR , in or around 1980 - to prevent what were called "ding / ding and away SPADS" (and in the case of some Scottish area issues - head on collisions at Paisley etc)

Apart from AWS and TPWS - here in the UK we have the "DRA" driver reminder appliance to be put on when standing at a red signal .....

Groningen · 20 Feb 2015

The only highspeed line in Switserland is the Zuerich to Bern line; further non can be found.

D1009 · 20 Feb 2015

Hearing the news item on BBC radio 2 this morning I remember being surprised a non fatal accident in another country was even reported, but I suppose it's a slow news day. It seems to have been confirmed that the driver of the IR train was seriously injured, but not life threatening,

http://www.blick.ch/news/schweiz/zu...unfall-zuege-von-rafz-geborgen-id3504217.html

I thought Swiss railways had a form of ATP system which should prevent this kind of incident?

Groningen · 21 Feb 2015

First there is the S-bahn train that goes from Schaffhausen to Rafz and back. From what i read in the NZZ did the IR train from Zuerich to Schaffhausen a track with no platform and did the S-bahn train at that same moment also leave the station. On Google Maps satelite you can see a train standing at the platfrom. And as it is single track around 300 to 400 meters outside of the station they had to come together. As maybe said that IR train had a delay of 15 minutes. International trains are diverted via Winterthur.

Carlisle · 21 Feb 2015

ilkestonian said:
Only "guard checks signal aspect to see if he can give RA to driver" seems to be absent from their procedures.

Wouldn't the RA device be interlocked so as to only work when the signal is showing a proceed aspect

A-driver · 21 Feb 2015

Carlisle said:
Wouldn't the RA device be interlocked so as to only work when the signal is showing a proceed aspect

Only if associated with a signal. Not if by RA people mean rig the bells/wave a green flag etc.

ilkestonian · 21 Feb 2015

A-driver said:
Only if associated with a signal. Not if by RA people mean rig the bells/wave a green flag etc.

Sorry, sloppy shorthand on my part. Yes, by RA I was referring to the guard doing whatever is required, eg whistle, flag, bells etc. as A-driver correctly surmised.

D1009 · 21 Feb 2015

Groningen said:
First there is the S-bahn train that goes from Schaffhausen to Rafz and back. From what i read in the NZZ did the IR train from Zuerich to Schaffhausen a track with no platform and did the S-bahn train at that same moment also leave the station. On Google Maps satelite you can see a train standing at the platfrom. And as it is single track around 300 to 400 meters outside of the station they had to come together. As maybe said that IR train had a delay of 15 minutes. International trains are diverted via Winterthur.

http://www.zol.ch/blaulicht/standar...tzten-trotz-modernster-Technik/story/14454673

From what I can see from this site, it appears that the S bahn formed of a double decker unit moved away from the platform as the IR train was passing through on the outer line with no platform. The 2 lines converge before the bridge over Ruedlingerstrasse so the derailed IR train continued over the bridge and came to rest with the coaches leaning against the bridge parapet. It looks to me that thanks to the solid construction of the bridge, there could have been a far worse outcome.

MarkyT · 27 Feb 2015

D1009 said:
I thought Swiss railways had a form of ATP system which should prevent this kind of incident?

Yes they have an intermittent trainstop and overspeed protection system called Signum which is supplemented in many places by ZUB (including this line, recently modernised) which overlays a precise location specific braking curve from the distant at caution to a stop. The problem is sometimes 'ding ding and away' incidents can defeat a trainstop system and my first guess is that is what may have occurred here.

Look at this aerial image:

https://maps.google.com/maps?ll=47.604494,8.545338&spn=0.001557,0.004128&t=h&z=19

A 4 car double deck S-bahn train of the type involved is clearly stopped on the inner platform track probably at the normal stopping position next to the canopy for such a short train. From the front of the train heading north east the signal protecting the junction is approximately 130 metres away, and the junction fouling point is another 70 or 80 metres beyond that. The questions the investigators may be asking are: If the S-bahn train accelerated away towards the red signal from the normal stopping position, what speed could it attain immediately before passing the signal at danger? And if the emergency braking was engaged immediately at that speed by the protection equipment at the signal, is there sufficient overlap for the train to brake safely to a stand before reaching the fouling point. I think the answers may be 'surprisingly fast' and 'no'. There was a sudden braking of the S-bahn train just before the collision according to some passenger reports in the press, which could have been the protection system cutting in, albeit too late. Assuming all the equipment was working correctly, the fast train on the outer track would have been obeying Signum and ZUB which would not have allowed the train to SPAD at the high speed it clearly collided and derailed at, which suggests the high speed through route was set correctly for the fast train.

http://www.blick.ch/news/schweiz/os...-zh-ich-haette-sterben-koennen-id3500987.html

Edit:

The signals at Rafz are not 'in line', being staggered by about 30 metres and attached to simple straight posts to the left of the line they apply to. The signal for the through line is actually closer to the S-bahn stopping position on the platform line than the correct signal, and is situated between the two tracks to the right of the platform track. If this signal was showing proceed, maybe the S-Bahn driver misinterpreted it as applying to his own line. I don't know if Swiss signals are ever to the right of a running line. There may be other 'route knowledge' complications around the area as it is very close to Germany and some drivers may be accustomed to different standards and practices on cross border workings. Swiss railways normally drive on the left with left side signals, Germany drives on the right with right side signals.
--- old post above --- --- new post below ---
A new press release from SBB has been published:

http://www.sbb.ch/sbb-konzern/medien/medienmitteilungen.newsdetail.2015-2-2702_1.html

A Google translation:

Media release, 27 February 2015

Rafz - the passing of signals led to grazing collision. SBB takes precaution.
SBB has initiated extensive investigations on 20 February after the accident in Rafz. After the first knowledge grazing collision is due to the crossing of the STOP field signal by the S-Bahn. The exact circumstances of the accident is still under investigation. Safety is a central corporate goal top priority. As a precautionary measure, the SBB performs the
reduction of speed by a turns to the first signal.

When glancing collision five passengers and two slightly SBB employees were heavy and medium seriously injured. The travelers were treated as outpatients and the hospital could leave the same day. The engineer of the Inter Regio after surgery except death. The engineer candidate who was also in the cab of the Inter Regio, the hospital can leave in the next few days expected.
According to initial findings of the SBB train 18014 was if the signal at 06.40 clock after the turn in Rafz set off on platform 4 towards Schaffhausen , At a speed of 59, the emergency brake is triggered at the signal. About 100 meters after the signal came the train to a halt and stood there easily into the light profile of the adjacent track. Seconds later approached on the adjacent track from Zurich Inter Regio 2858 with Tempo 110. Due to a delay in Interregio went through the Rafz Station on track 5 instead of track 3. During the side collision of Interregio derailed, both trains were severely damaged. The exact circumstances of the accident is clarified by the prosecutor of the canton of Zurich and the independent accident investigation authority of the federal government.

Extensive repairs to roads and vehicles.
Since the night of the 22th of February, the Rafz train station can be ridden again. SBB replaced a total of 60 meters of track, a switch and a catenary. For work on catenary and Soft additional barriers were necessary. The accident vehicles were transferred to SBB industry works to Olten and Yverdon. The damage to the affected train type DTZ and the IR composition, consisting of five Euro City cars and the locomotive Re460 is, according to initial estimates, several million francs.

Station Rafz has modern security system.
As after any accident SBB has initiated its own extensive research. The Rafz Station has modern security systems. The signals in Rafz are set up according to the rules. The train control system complies with Swiss standard. According to initial findings, the safety devices are working properly. The Abfahrverhinderung is aimed at passing trains, which make up the majority in Rafz. Because the S-Bahn was retracted from Schaffhausen and had turned, she has not slowed down by the train control system. Additional security components, but which have not foolproof, are provided only when a weekly average investment of at least one train per day. This is not the case in Rafz direction Schaffhausen.

SBB introduces a precaution.
As a precautionary measure, the SBB leads from next Monday and until further speed reduction for phrases. Trains a maximum of 40 km / h up to the first signal. This increases the likelihood that in situations such as in Rafz the train can be stopped before the danger point. The SBB will now examine whether this precaution has no other adverse effects. After the accident in Granges-Marnand SBB has developed its own warning app, which additionally warns drivers if the exit signal is not on the descent to ride. The system is currently being tested intensively in the background. The launch is scheduled for successful tests in spring 2015. The current limitations of security systems in phrases to be solved with the development of the train control system in the future: The modern ETCS Level 2 system ensures that erroneous runs are technically impossible after the turnover. The network-wide implementation of ETCS Level 2 is planned from 2025 onwards. At present, it is checked whether this can be accelerated. SBB has a high safety standard, which was confirmed by the accident of Granges-Marnand third party appraisals. The accident Rafz as each event carefully analyzed in order to further increase the security. However, a residual risk can never be ruled out.

30907 · 27 Feb 2015

The end of the first translated paragraph says that as an interim measure trains that have reversed will proceed at reduced speed as far as the first signal (ie the station exit signal). 40kph is mentioned later.

MarkyT · 28 Feb 2015

30907 said:
The end of the first translated paragraph says that as an interim measure trains that have reversed will proceed at reduced speed as far as the first signal (ie the station exit signal). 40kph is mentioned later.

Yes a 40kph limit following reversal until the first signal is passed is what I understand. This had been suggested earlier this week by the president of the Swiss Train Drivers Association, Hubert Giger. See here:

http://www.zurich4you.ch/sbb_comes_under_criticism_of_president_of_train_drivers_union.html

The comment in the article above about a right hand signal puzzled me at first, as all the station signals at Rafz are clearly on the left, as can be seen on the very recent Google aerial images. There is, however, a right-hand signal just beyond the junction on the single line. It appears to have a diamond plate on top so may be a distant for the next break section block signal further along the single line.

See: http://www.styria-mobile.at/home/Austro-SwissRailwaySignalling/asr/ensigchsysn.html

Finding a couple of images for the Shaffhausen end of the station I drew this up to illustrate: http://www.townend.me/files/rafz.pdf

The problem only occurs after a reversal. For a station stop followed by a continuation in the same direction, there is a braking curve applied by ZUB when passing the previous cautionary aspect. After a station stop that constraint still applies, even if the signal has cleared subsequently, unless there is a 'fill-in' loop from the platform stopping point to the signal that can update the on-board system, for which the expense can be difficult to justify. With no loop the train is subject to a low 'release speed' that allows a train to sidle up to the next signal at a safe speed. If it passes the signal the release speed is determined so a trip when passing the signal at red will always bring the train to a halt before the conflict point. At Rafz following a reversal and without the ZUB 'fill-in' or 'update' loop there was no mechanism to prevent departure or control speed subsequently on approach to the signal.

eastwestdivide · 1 Mar 2015

retranslating the google version into something closer to English:

After the first knowledge grazing collision is due to the crossing of the STOP field signal by the S-Bahn

from the original:

Nach ersten Erkenntnissen ist die Streifkollision auf das Überfahren des Halt zeigenden Signals durch die S-Bahn zurückzuführen

to:

According to preliminary findings, the sideswipe collision can be attributed to the S-Bahn passing a signal that was showing stop

Passenger train collision in Rafz, Switzerland

Established Member

RailUK Forums

Member

Established Member

Established Member

Established Member

Veteran Member

Established Member

Veteran Member

Established Member

Veteran Member

Member

Established Member

Established Member

Established Member

Established Member

Established Member

Established Member

Member

Established Member

Established Member

Veteran Member

Established Member

Established Member