Terrible accident at Santiago de Compostela, northern Spain - 24 July 2013

[whole diary]

[broadgage]

Media reports suggest that although relatively new, that this track was not equiped with ETRMS or any equivalent system.
In which case it would have no protection against accidental or intentional overspeed.

[ChrisB]

I think it was only the HSR track that was new - that train had just left the new track for the old....

[stuving]

As I see it, this was new track built on an existing alignment that was widened to make room for it. It only goes to/from the HS^▸ (High Speed (short for HSS^▸ (High Speed Services) High Speed Services)) line one way, but joins the other lines approaching the station. So it's moot if its a bit of HS line with a low speed limit, or a completely rebuilt old line as a link to the HS line. I prefer the former version. Either way it's new and its signalling must be too.

[JayMac]

The curve in question comes after many miles of straightish running over newly built dedicated high speed track. The new line was signalled with ETCS^▸ (European Train Control System) (European Train Control System) to a boundary point somewhere prior to the curve, where the signalling system switches to ASFA, Spain's near equivalent of the UK^▸ (United Kingdom)'s TPWS^▸ (Train Protection and Warning System).

I've watched an in-cab video of a train traversing the same stretch of line at the correct speed, but I don't know enough about Spanish signalling to work out where the demarcation point between modern and legacy signalling is.  The new line joins the old part way round the curve.

That video can been seen here:

http://www.youtube.com/watch?v=QvvvL3OTtXk (You'll need to switch to 480p or 720p as the default 360p doesn't seem to work)

Forward that video to around the 31 minute mark to see the approach to the curve.

[Chris from Nailsea]

Another update, from the BBC^» (British Broadcasting Corporation - home page):

Spain train crash: Driver Garzon declines to give evidence


Train driver Francisco Jose Garzon Amo is helped by a policeman after a train crashed near Santiago de Compostela, 24 July 2013.

The driver of the train that derailed killing 78 people has refused to answer initial questions, Spanish police say.

Francisco Jose Garzon Amo - hurt in Wednesday's crash - is under guard in hospital.

He is suspected of driving too fast round a bend. Reports say the train was travelling at more than double the speed limit at the time of the crash.

The case should now "proceed to a judicial process", the police added. Spain is in three days of mourning.

The police chief in the Galicia region, Jaime Iglesias, said earlier that the driver would be questioned "as a suspect for a crime linked to the cause of the accident".

No date has been fixed for his appearance before the judge, a spokeswoman for Galicia's High Court, which is leading the investigation, told the AFP news agency.

[Electric train]

The curve in question comes after many miles of straightish running over newly built dedicated high speed track. The new line was signalled with ETCS^▸ (European Train Control System) (European Train Control System) to a boundary point somewhere prior to the curve, where the signalling system switches to ASFA, Spain's near equivalent of the UK^▸ (United Kingdom)'s TPWS^▸ (Train Protection and Warning System).

I've watched an in-cab video of a train traversing the same stretch of line at the correct speed, but I don't know enough about Spanish signalling to work out where the demarcation point between modern and legacy signalling is.  The new line joins the old part way round the curve.

That video can been seen here:

http://www.youtube.com/watch?v=QvvvL3OTtXk (You'll need to switch to 480p or 720p as the default 360p doesn't seem to work)

Forward that video to around the 31 minute mark to see the approach to the curve.

I watched that vid through, there does not seem to be any fixed colour light signals until well past the accident site, I can only assume there is in cab signalling but it cannot be a speed controlling type of signalling whereby the system stops the train if the drive exceeds to max speed

[Red Squirrel]

...That video can been seen here:

http://www.youtube.com/watch?v=QvvvL3OTtXk (You'll need to switch to 480p or 720p as the default 360p doesn't seem to work)

Forward that video to around the 31 minute mark to see the approach to the curve.

I found watching that as horrifying as the video of the accident - the topography of the area where the train came off the rails couldn't have been much worse.

[stuving]

I watched that vid through, there does not seem to be any fixed colour light signals until well past the accident site, I can only assume there is in cab signalling but it cannot be a speed controlling type of signalling whereby the system stops the train if the drive exceeds to max speed

So, the emerging picture is that the new line uses ETCS^▸ (European Train Control System) (ERTMS^▸ (European Rail Traffic Management System.)) which can provide full supervision of train speed. But does it have to? It can operate as an in-cab signalling system using only inputs from train and track, so I presume the reverse link to the train to apply the brakes is a mode that can be omitted.

There are still some "why did/didn't they" questions. It seems the handover from ETCS to AFSA was 4 km back from the curve, just before the braking point at 0.05 g. I'm sure ETCS would signal this braking where the restriction is after the handover, but only if the braking starts before it. Why put the boundary there, when braking from straight high-speed line to a low speed curve has to come near the top of any risk assessment?

And some plain puzzles ... e.g. ERTMS say that all AVE lines were switched to level 2 (no lineside signals) in 2012*, whereas this one was opened in 2011 and, if it has no signals, clearly used level 2 from the start.

I also wondered why the old and very wiggly Spanish gauge line from Ourense, that joins the AVE line from the left at the curve, was of the same gauge. It turns out this AVE line was built to 1.668 m and will be converted later! These trains also go through a "drive-thru" gauge-changer near Valladolid, having left Madrid on a (1.435 m) AVE line, and before joining the existing line to Ourense. Spanish rail engineering is indeed unlike our own in many ways.

 (* Correction - I missed that they (UNIFE http://www.ertms.net/media/2434/ertms%20facts%20sheet%205%20-%20ertms%20deployment%20in%20spain.pdf) say the first was opened in 2011, which may have been this line.)

[Gordon the Blue Engine]

In the UK^▸ (United Kingdom) speed limits are strictly enforced with drivers at risk of dismissal for all but the most trivial breaches.
This is however fairly recent, I can remember trains routinely exceeding speed limits, was there not a "140 club" years ago of those HST^▸ (High Speed Train) drivers who had reached 140 MPH.

I'm not sure that's true Broadgage.  I was around when HST's came in on the Western region in1976, and I recall they all had speed limiters either from the outset of public service runs or very soon after which prevented them doing more than about 130 mph.  It maybe that early driver training runs etc pre-dated the speed limiters. 

I do remember doing 110 mph behind a Class 50 between Didcot and Reading: unlike Class 47's, which ran out of power above about 80-85 mph, Class 50's could really go!

[Electric train]

I watched that vid through, there does not seem to be any fixed colour light signals until well past the accident site, I can only assume there is in cab signalling but it cannot be a speed controlling type of signalling whereby the system stops the train if the drive exceeds to max speed

So, the emerging picture is that the new line uses ETCS^▸ (European Train Control System) (ERTMS^▸ (European Rail Traffic Management System.)) which can provide full supervision of train speed. But does it have to? It can operate as an in-cab signalling system using only inputs from train and track, so I presume the reverse link to the train to apply the brakes is a mode that can be omitted.

There are still some "why did/didn't they" questions. It seems the handover from ETCS to AFSA was 4 km back from the curve, just before the braking point at 0.5 g. I'm sure ETCS would signal this braking where the restriction is after the handover, but only if the braking starts before it. Why put the boundary there, when braking from straight high-speed line to a low speed curve has to come near the top of any risk assessment?

And some plain puzzles ... e.g. ERTMS say that all AVE lines were switched to level 2 (no lineside signals) in 2012*, whereas this one was opened in 2011 and, if it has no signals, clearly used level 2 from the start.

I also wondered why the old and very wiggly Spanish gauge line from Ourense, that joins the AVE line from the left at the curve, was of the same gauge. It turns out this AVE line was built to 1.668 m and will be converted later! These trains also go through a "drive-thru" gauge-changer near Valladolid, having left Madrid on a (1.435 m) AVE line, and before joining the existing line to Ourense. Spanish rail engineering is indeed unlike our own in many ways.

 (* Correction - I missed that they (UNIFE http://www.ertms.net/media/2434/ertms%20facts%20sheet%205%20-%20ertms%20deployment%20in%20spain.pdf) say the first was opened in 2011, which may have been this line.)

It seems to be a Level 2 system which I think is in cab signalling I am not sure how much the signalling system can override the drive (anymore than our AWS^▸ (Automatic Warning System) and TPWS^▸ (Train Protection and Warning System)) does now

Likny to ETRMS http://www.ertms.net/ertms/ertms-signalling-levels.aspx

[stuving]

It seems to be a Level 2 system which I think is in cab signalling I am not sure how much the signalling system can override the drive (anymore than our AWS^▸ (Automatic Warning System) and TPWS^▸ (Train Protection and Warning System)) does now

Likny to ETRMS http://www.ertms.net/ertms/ertms-signalling-levels.aspx

There's a longer version at http://www.ertms.net/media/2428/ertms%20facts%20sheet%203%20-%20ertms%20levels.pdf, where it says this:

ERTMS^▸ (European Rail Traffic Management System.) level 1 is designed as an add-on to or overlays a conventional line already equipped with lineside signals and train detectors. Communication between the tracks and the train are ensured by dedicated balises (known as ^Eurobalises^) located usually on the trackside adjacent to the lineside signals at required intervals, and connected to the train control centre. Receiving the movement authority through Eurobalises, the ETCS^▸ (European Train Control System) onboard equipment automatically calculates the maximum speed of the train and the next braking point if needed, taking into account the train braking characteristics and the track description data. This information is displayed to the driver through a dedicated screen in the cabin. The speed of the train is continuously supervised by the ETCS onboard equipment.
The main benefits brought by ERTMS Level 1 are interoperability (between projects and countries) and safety, since the train will automatically brake if exceeding the maximum speed allowed under the movement authority.

I presume that last statement goes for levels 2 and 3 as well. But, as I said, the link to the train that applies the brake is not required for signalling, and probably needs a specific interface box, so I could believe it is sometimes absent even with an ETCS level 1 or 2.

So you get a cab display with level 1, but it appears that it does not replace signals, it adds speed advice.

[stuving]

Oh, and should have said yes - ETCS^▸ (European Train Control System) level 2 does give in-cab signalling, and in ERTMS^▸ (European Rail Traffic Management System.) level 2 uses GSM-R^▸ (Global System for Mobile communications - Railway.) as its radio link. I note the words say "does not require lineside signals" - but do not spell out how things work for lines with multiple on-board systems. I see that several lines are described (elsewhere) as "level 1/2" - I wonder what that means (possibly more than one kind of mixture?).

[SandTEngineer]

Some further information here: http://elpais.com/elpais/2013/07/25/media/1374703338_483146.html and here: http://elpais.com/elpais/2013/07/26/media/1374872840_505323.html and on the specifics of the signalling system here: http://comofuncionanlostrenes.blogspot.com.es/2012/12/instalaciones-de-seguridad-asfa-digital.html?m=1  In the case of the latter link it is interesting to read the posts well down the document.

[eightf48544]

This seems to be a boundary conditon anomoly.

I thought the point of ERTMS^▸ (European Rail Traffic Management System.) level 2 and above is that it eliminates lineside signals giving the train a max speed for any point on the line, with the ability to slow/stop the train if the authorised speed is exceeded. Thus I would have expected that either the ERTMS would have given a slow speed coming off the new line or the old system to have a restricted aspect signal  before the bend which presumably the driver could overule. Similar to approach control at junctions as used in the UK^▸ (United Kingdom).

We have had problems with curves Morpeth comes to mind.

The lessons to be learnt is that boundary conditons must be carefully thought through to avoid such anomoies a lesson for the GWML^▸ (Great Western Main Line) resignallers.

There was apprently a SPAD^▸ (Signal Passed At Danger) on the first siganl out of Marylebone which is ATP^▸ (Automatic Train Protection) fitted. Unfortunately ATP requires to pass two signals before it knows where it is. 

[Chris from Nailsea]

A further update, from the BBC^» (British Broadcasting Corporation - home page):

Spain train driver Garzon questioned by judge


Francisco Jose Garzon Amo was taken to the courthouse from a police station

The driver of a train that crashed near the Spanish city of Santiago de Compostela, killing 79 people, has been questioned by a judge.

Francisco Jose Garzon Amo was detained on suspicion of reckless homicide after the accident. He is suspected of driving too fast on a bend.

Spanish media says Mr Garzon has been charged but allowed to leave the court. The judge has retained his passport, according to reports.

The investigation is still believed to be open and the driver will have to face trial at a later date. He will have to report to the judge once a week.

Reports say the train was travelling at more than double the speed limit at the time of the crash.

Mr Garzon, 52, was pictured being escorted away from the wreckage by police, blood pouring from a head injury. He left hospital on Saturday and was immediately taken to the central police station in Santiago. He had refused to make a statement or answer questions until now.

Spanish newspaper El Pais said Mr Garzon had admitted to "recklessness" in court. The judge reportedly said Mr Garzon was free to leave the court but was banned from driving trains.

On Sunday, an eyewitness to the disaster told the BBC that he had overheard the driver admitting minutes after the crash that he had been going too fast. Evaristo Iglesias, a resident of Santiago de Compostela, said he heard the driver saying he tried to slow down but "it was too late". Mr Iglesias said the driver, who was shocked and dazed, was repeatedly "saying he wanted to die" rather than see the damage at the scene.

Sunday's court hearing was closed but the judge was expected to decide whether to remand the driver as an official suspect, release him on bail, or free him without charge.

At least 130 people were injured in the accident. It emerged that one of them - identified as an American woman - died on Sunday. Dozens more remain in a critical condition.

All eight carriages of the train - packed with more than 200 passengers - careered off the tracks into a concrete wall as they sped around the curve on the express route between Madrid and the port city of Ferrol on the Galician coast. Leaking diesel burst into flames in some of the carriages.

The train's data recording "black box" is with the judge in charge of the investigation. Officials have so far not said how fast the train was going when it derailed.

Gonzalo Ferre, president of Spanish rail network administrator Adif, said the driver should have started slowing the train 4km (2.5 miles) before the spot where the accident happened.

The president of Spanish train operator Renfe, Julio Gomez Pomar, has said the train had no technical problems. He said the driver had 30 years' experience with the company and had been operating trains on the line for more than a year.

People from several nationalities were among the injured, including five US citizens and one Briton. Two Americans were among the dead.

The crash was one of the worst rail disasters in Spanish history.