Train crash on the Spanish/Portugese border.

[roythebus]

Four people died and fifty were injured Friday in the northwest of Spain, in the derailment of a train from Vigo to Porto (Portugal), authorities said.

 

The driver of the small cross-border train, a Portuguese national, is among the dead. There are three years Galicia had already experienced the worst rail accident in Spain since 1944, which had 80 deaths.

At least "four people died," said Manuel Carrera, councilor of the town of O Porriño where the accident occurred. The regional government said that 47 people had been hospitalized.

The train of three cars derailed around 9:30 am as he was about to get in O Porriño station, south of Vigo, not far from the border with Portugal, according to the Spanish company Renfe. It belongs to the Portuguese railway company Comboios de Portugal, which manages the line with his Spanish colleague. According to the regional newspaper La Voz de Galicia, a controller of fifty years, living Vigo, also among the four dead.

The first wagon overturned

The train was carrying 65 people according to Renfe, 69 according Comboios de Portugal. "We were quietly sitting and suddenly the train started to do like that," testified a passenger of the second car, mimicking a rocking motion. "Suddenly I saw that did not stop and I fell," she added in an interview filmed by La Voz de Galicia.

Images broadcast by the Galician television showed the front of the process embedded in an electric pillar, completely reversed. The other two cars seemed almost intact. A 15-year-old Alex Ramilo, going cycling on a bridge over the road when the accident happened: "I heard a deafening noise. I turned my head and saw the train derailed. I was speechless, really shocked, "he told the boy in the Rami coffee in O Porriño.

[roythebus]

plusieurs-morts-dans-un-deraillement-de-train-reliant-lespagne-et-le-portugal  Newspaper article from a Luxembourg friend. The driver, a railway manager and 2 passengers killed. Our thoughts to their families.

[wombatofludham]

http://www.bbc.co.uk/news/world-europe-37315566

 

BBC Website coverage.

[Glover]

I have used this train, about two years ago; from Santiago de Compostela, change to the Portuguese train (still Spanish broad gauge) at Vigo and then on to Porto.

The photos I've seen of this accident seem to indicate that the train switched from the main line into sidings and, possibly, the lead unit struck the over bridge.

I suspect the key question is: are the points which allow access to the sidings controlled from a CTC type centre or are they manually operated at the actual location?

 

Glover

[roythebus]

It would appear, looking at the picture on the BBC article, that the points were set for the loop line and the train has derailed on the points, speculate on excessive speed? Points split under the train?

[Guest]

http://www.bbc.co.uk/news/world-europe-37315566

 

BBC Website coverage.

 

The first image on the BBC website (above) shows just how badly mangled the first carriage of the unit is. It's a Class 592 DMU, used by both RENFE and CP, vintage 1981-1984, so about the same vintage as say, a BR Class 455 EMU. The crash-worthiness of these DMUs will probably now be examined closely. As above, sympathies to all those affected.

 

This file is licensed under the Creative Commons Attribution-Share Alike 2.0 Generic license.

[Guest B Exam]

The first image on the BBC website (above) shows just how badly mangled the first carriage of the unit is. It's a Class 592 DMU, used by both RENFE and CP, vintage 1981-1984, so about the same vintage as say, a BR Class 455 EMU. The crash-worthiness of these DMUs will probably now be examined closely. As above, sympathies to all those affected.

 

Barcelos_L1180527_(24919684370).jpg

This file is licensed under the Creative Commons Attribution-Share Alike 2.0 Generic license.

The 592s as proven in yesterdays crash are not very strong in the body.

 

The units are still owned by RENFE, but on long term hire to Comboios de Portugal. The unit involved 592.056 is one of dedicated "Celta" (Arrow) branded units for this service.

 

158722 of parish has reported on another forum that the unit was travelling at 120km/h and took the turnout into a platform road which has a 30km/h restriction on it. Normally it would of travelled straight through on the centre line platform at normal speed, but for some reason it was diverted into one of the loop platforms. It looks to of derailed and struck the bridge parapet causing the catastrophic damage.

 

The points are maually operated. Possible engineering work needing to be diverted into one of the loops?

[phil-b259]

158722 of parish has reported on another forum that the unit was travelling at 120km/h and took the turnout into a platform road which has a 30km/h restriction on it. Normally it would of travelled straight through on the centre line platform at normal speed, but for some reason it was diverted into one of the loop platforms. It looks to of derailed and struck the bridge parapet causing the catastrophic damage.

 

If true - why does yet another European railway administration appear to have gaping great holes in their signalling system - I mean its not hard to fit some sort of approch control to routes that have shap turnouts.

[Guest B Exam]

I doubt the driver would of been on the phone as the train had just exited a tunnel (seen in the distance in the photos). Lack of reception etc.

 

As for the quote Spanish taking corners too fast, it was a Portuguese driver based at Porto driving a Portuguese train that provides the cross border service between Porto and Vigo. 158722 of this parish knew of the driver - friend of a friend. 158722 is a very good personal friend of mine so we need to cut some slack here please, and not jump to conclusions to the cause of the accident.

 

It has hit the all CP train crew very hard as its a very close knit community.

 

To the signalling system in use:

 

Its still a antiquated system on that line that relys on telephone calls from one station to the next. The driver of the train will stop and receive papers that he signs to give permission to travel to the next station or if the line is clear, to their next booked stop. The driver will have a carbon copy of this instructions given to them. On these papers will be details of engineering works, slacks and anything else to look out for.

 

At the previous stop before Porrino, the driver should of been notified that he will be diverted into line 3 and he would of signed the papers as message received and understood so he could take the necessary action on the approach to Porrino station.

 

Whether the driver received the necessary instructions, no idea, but it will no doubt be revealled in any inquiry to the accident.

 

The system in use has been used all over the world since the telephone was invented and on lightly used lines like this one, the most efficient. To spend millions on full CTC signalling for 3-4 trains in each direction a day, I think is a complete waste of money.

 

Edit:

 

This line does have a colour light signalling system. Portuguese press this morning suggesting a IT failure with the points set for line 3 loop, but the signal adpect saying line 1 main line. This could explain the excessive speed on the approach to Porrino?

[caradoc]

If the train was booked to stop at Porrino why was it (reportedly) travelling at 120 km/h on the approach to the station ?

If this was not the case, but the train was booked over a 30 km/h route instead of its normal 120 km/h route, Phil-b259 has made a perfectly valid point; Giving a Driver no indication of such a severe speed restriction than a piece of paper at a location which may be a long way away, in terms of both distance and time, is wide open to the tragic consequences of human error.

[Guest B Exam]

If the train was booked to stop at Porrino why was it (reportedly) travelling at 120 km/h on the approach to the station ?

If this was not the case, but the train was booked over a 30 km/h route instead of its normal 120 km/h route, Phil-b259 has made a perfectly valid point; Giving a Driver no indication of such a severe speed restriction than a piece of paper at a location which may be a long way away, in terms of both distance and time, is wide open to the tragic consequences of human error.

Train was not booked to stop at Porrino. The Vigo-Porto service in a very limited stop "premium" service.

 

As mentioned earlier in the thread, its been reported that there was a IT/signalling problem which suggested to the driver the line was clear for 120km/h through line 1 when in reality the points were set for line 3 with a 30km/h.

[phil-b259]

Train was not booked to stop at Porrino. The Vigo-Porto service in a very limited stop "premium" service.

 

As mentioned earlier in the thread, its been reported that there was a IT/signalling problem which suggested to the driver the line was clear for 120km/h through line 1 when in reality the points were set for line 3 with a 30km/h.

 

You do not have 'IT' Problems in railways signalling systems

 

Signalling should be Fail Safe. If the points are set for the loop it should be impossible (as in the physical wiring / relays / software) for the associated signal to indicate anything to the contrary.

 

As such, a failure that gives rise to a dangerous situation (known as a 'Wrong Side Failure' in the UK) is an extremely serious issue - and should never be refereed to as an 'IT problem'

[Guest B Exam]

You do not have 'IT' Problems in railways signalling systems

 

Signalling should be Fail Safe. If the points are set for the loop it should be impossible (as in the physical wiring / relays / software) for the associated signal to indicate anything to the contrary.

 

As such, a failure that gives rise to a dangerous situation (known as a 'Wrong Side Failure' in the UK) is an extremely serious issue - and should never be refereed to as an 'IT problem'

Only reporting what the Portuguese press are reporting this morning. Now I'm no expert on Spanish signalling so cannot coment on the ins and outs of their systems.

 

As such, I will shut me gob on the matter and let the you lot play amongst yourselves...

 

Good night!

[roythebus]

Thanks for the further information. As a former driver i always feel for the families of those killed in railway accidents, even those abroad. You mention a friend of a friend knew the driver, i happened to know one of the drivers killed in the Clapham disaster and the driver who had the SPAD at Wooton Basset, so know what their families (and the WD driver) went through.

 

I will reiterate that mobile phones and computer related items have been partly responsible for a couple of fatal crashes in recent years, the RENFE crash and the German head-on this year. Only the accident report will find what actually happened.

[phil-b259]

Only reporting what the Portuguese press are reporting this morning. Now I'm no expert on Spanish signalling so cannot coment on the ins and outs of their systems.

 

As such, I will shut me gob on the matter and let the you lot play amongst yourselves...

 

Good night!

 

I was not having a go at you, so if you are merely informing us what some idiot journalist put down then please don't take it personally.

 

However I stand by my assertion that attributing a train crash is due to 'IT Problems' is shows a shockingly cavalier attitude to what should be a fail safe system.

[The Stationmaster]

I was not having a go at you, so if you are merely informing us what some idiot journalist put down then please don't take it personally.

 

However I stand by my assertion that attributing a train crash is due to 'IT Problems' is shows a shockingly cavalier attitude to what should be a fail safe system.

 

Indeed so Phil but you are looking at it from a very UK signalling system viewpoint and in a some other - seemingly otherwise 'advanced' countries - things are not the same and lesser parts of their network are run on systems which a signal engineer friend of mine once described as 'not even as sophisticated as stick and string signalling'.  Where verbal messages or even printed or messages or those transmitted by modern machines and electronics are involved then things can go wrong.  

 

I can well remember coming across a junction station in southern Portugal in 2000 (just into the year) where all the points were set by hand and secured by padlocks, the 'signals' were worked by levers at the base of the framework on which they were situated and there was no interlocking of any sort whatsoever while a nearby level crossing was worked by a crossing keeper who seemed to have no communication with those who worked the signals.  However what we might deign to call a 'block system' was simply communication of messages using telephones and a pc while even the single line branch from the junction seemed to be worked by nothing more than verbal instruction with no sort of train staff in sight.

 

Equally on an SNCF single line I travelled on later in the same year sections of it lacked any sort of fixed signal and authority to pass a stop board and enter the section was a wave and a shout from the Station Master (although there was no doubt some sort of block system in use).

 

The plain fact is that systems of single line operation which were rendered illegal in Britain well over a century ago seems to have remained commonplace in some other western European countries with train crossings altred solely by a 'phone call on lines worked by what amounted to train order systems.

[roythebus]

A further twist to this in an unrelated article and incident, last paragraph states the train was doing 118km/h in a 30 km/h zone.AQLd_z7n6RKO9DYZt_q4Z7w

[Bruce Depot]

A bit in the El Pais

Excessive speed caused train crash in Galicia that killed four people

[Pacific231G]

Indeed so Phil but you are looking at it from a very UK signalling system viewpoint and in a some other - seemingly otherwise 'advanced' countries - things are not the same and lesser parts of their network are run on systems which a signal engineer friend of mine once described as 'not even as sophisticated as stick and string signalling'.  Where verbal messages or even printed or messages or those transmitted by modern machines and electronics are involved then things can go wrong.  

 

I can well remember coming across a junction station in southern Portugal in 2000 (just into the year) where all the points were set by hand and secured by padlocks, the 'signals' were worked by levers at the base of the framework on which they were situated and there was no interlocking of any sort whatsoever while a nearby level crossing was worked by a crossing keeper who seemed to have no communication with those who worked the signals.  However what we might deign to call a 'block system' was simply communication of messages using telephones and a pc while even the single line branch from the junction seemed to be worked by nothing more than verbal instruction with no sort of train staff in sight.

 

Equally on an SNCF single line I travelled on later in the same year sections of it lacked any sort of fixed signal and authority to pass a stop board and enter the section was a wave and a shout from the Station Master (although there was no doubt some sort of block system in use).

 

The plain fact is that systems of single line operation which were rendered illegal in Britain well over a century ago seems to have remained commonplace in some other western European countries with train crossings altred solely by a 'phone call on lines worked by what amounted to train order systems.

Staff and token working has always been rare on single track lines in Europe and indeed anywhere else outside the former British empire.

 

There does seem to have been a divergence a very long time ago with the British approach being to try to create inherently foolproof electro-mechanical systems while other administrations placed more reliance on strictly observed procedures and defined communication protocols. So,  for example in France, on lines with passenger trains, only the Chef du Gare in person was allowed to clear a train onto a single track line having established that it was safe to do so. That role could not be delegated to another member of staff, but on a lightly traficked line it may well not have involved mechanical signals as such. After a couple of single line collision disasters a few years ago the French augmented this tradtional telephone block with a system using specially adapted PCs known as CAPI. Cantonnement Assisté Par Informatique- computer assisted block. The dedicated computer in each station (don't worry it doesn't use Windows) communicates with those on either side and standard messages are passed between them with each machine recording everything that has happened and showing the status of the line. It's essentially a recording block instrument and its operation is burnt into it. It is not interlocked with any signals and it is still the responsibility of the Chef du Gare  to give each train permission to leave.

 

Portugese and Spanish practice was very heavily influenced by French so the system you describe of padlocked points- the release of whose keys may well have been interlocked- sounds typical of many small French stations and the locally operated signals do sound similar to the Lartigue block semaphire system once widely used in France.

 

Such systems are not as failsafe as those traditionally use in Britain (and are disappearing) but no system is completely foolproof. It has not been unknown for a British driver to be offered and accepted the wrong token for a section with disastrous results and there is an argument that, if a system appears to be foolproof, those operating it may become complacent about double checking it. 

Despite these different approaches overall rail safety seems to have been generally fairly comparable across Europe . At the moment Britain's railways appear to be among the safest but rail accidents that cause multiple passenger fatalities are thankfully so rare that a single disaster (such as Ladbroke Grove or Eschede) can skew the comparative figures for a country for years.

[Edwin_m]

It has not been unknown for a British driver to be offered and accepted the wrong token for a section with disastrous results and there is an argument that, if a system appears to be foolproof, those operating it may become complacent about double checking it. 

The last one of those was Abermule in 1921, where a contributory factor was that the tablet machine was located in the station building for supervision by the British equivalent of the Chef de Gare. 

 

So,  for example in France, on lines with passenger trains, only the Chef du Gare in person was allowed to clear a train onto a single track line having established that it was safe to do so. That role could not be delegated to another member of staff, but on a lightly traficked line it may well not have involved mechanical signals as such. After a couple of single line collision disasters a few years ago the French augmented this tradtional telephone block with a system using specially adapted PCs known as CAPI. Cantonnement Assisté Par Informatique- computer assisted block. The dedicated computer in each station (don't worry it doesn't use Windows) communicates with those on either side and standard messages are passed between them with each machine recording everything that has happened and showing the status of the line. It's essentially a recording block instrument and its operation is burnt into it. It is not interlocked with any signals and it is still the responsibility of the Chef du Gare  to give each train permission to leave.

I think any signal engineer brought up with the British way of doing things would have apoplexy at the thought of letting a PC anywhere near anything vital - even if it doesn't use Windows! 

 

Just goes to show, as you say, how much practices differ.  But with multiple-fatality head-on collisions in Italy and Germany this year, on lines worked respectively by telephone and by interlocking but with a manual override, I can't help thinking such systems ought to be on the way out. 

[The Stationmaster]

The last one of those was Abermule in 1921, where a contributory factor was that the tablet machine was located in the station building for supervision by the British equivalent of the Chef de Gare. 

 

 

 

Abermule was I believe the last incident where mistakes over tokens/tablets resulted in a collision - it very definitely was not the last time it happened and a train entered a section carrying the wrong token.  Incidentally the siting of single line instruments in station buildings instead of signalboxes was not unique to the Cambrian

 

Incidentally the Cambrian seem to have been rather unfortunate when it comes to getting two trains in a section heading in opposite directions on a single line having previously managed it at Parkhall although it that case something was very definitely amiss as both of the trains had the correct tablet for the section - i.e. two tablets had been somehow withdrawn and were given to trains in opposite directions.

[John_Hughes]

Abermule was I believe the last incident where mistakes over tokens/tablets resulted in a collision - it very definitely was not the last time it happened and a train entered a section carrying the wrong token.  Incidentally the siting of single line instruments in station buildings instead of signalboxes was not unique to the Cambrian

 

Incidentally the Cambrian seem to have been rather unfortunate when it comes to getting two trains in a section heading in opposite directions on a single line having previously managed it at Parkhall although it that case something was very definitely amiss as both of the trains had the correct tablet for the section - i.e. two tablets had been somehow withdrawn and were given to trains in opposite directions.

 

I know as a matter of fact of at least one instance when a passenger train departed from an intermediate single-line station with no token at all, a fact that the footplate crew only realised when they were well under way.

 

The missing token was duly biked over later in the day and the whole thing was duly swept under the carpet by all concerned.

 

Something similar may well have happened on other occasions as well; but those who know are keeping mum, and as far as further details are concerned so am I!

[Guest]

Historic mishaps and accidents involving tokens, whilst informative and interesting, are a long way removed from the OP, which is about a tragic accident in a completely different country. Sorry, but I just think things are getting a little...

As is usually the case with RMWeb topics about accidents, prior to the results of any enquiry, everything else really is just speculation.

[158722]

I've only just caught up with this thread now.

 

Most of what has been commented on so far is correct.

 

The data download has confirmed the train was travelling at 118kmh in a zone where 130kmh permitted for through trains on the normal line 1 at Porrino station. However, because of some works on or adjacent to this line (not confirmed what was being done, but I gather some sort of routine maintenance which required that line to inactive whilst staff were present), the train was required to be diverted via line 3, which is the secondary loop around the back of the island platform and which is reached via two 30 kmh limited turnouts. What is not yet confirmed is whether the driver was not advised correctly by the signalling system (i.e. incorrect aspect and/or route indicator) or whether the signalling was indeed correct and he simply did not slow in time.

 

B Exam's quote (where has he gone by the way?) about an IT problem is probably repeating what I have posted elsewhere, but which I subsequently corrected to a computer system error - perhaps not quite correctly translated on my part. This came from one press report here which suggested that there was indeed a signalling error and that whilst the route was set for 30kmh limited access to line 3, the approach signal was indicating the normal 130kmh limit line 1. This report has only appeared in one newspaper and not followed up.

 

As regards the systems in place here and in Spain, the Porrino area is covered by ASFA basic, which from my restricted knowledge appears to be similar to AWS - as in, an advisory system based on signal aspect with no speed control. In the case of Porrino, the driver should have been advised that he was coming up on an adverse signal, not a red stop, but a yellow and routing over to line 3. The investigation will no doubt be centering on whether or not he did receive that warning and I have not seen any confirmation or comment about this over the last few days.

 

ASFA Digital exists in many parts of Spain, which I gather is a newer, more advanced system which includes overspeed controls. This sounds similar to the Portuguese CONVEL system (CONtrolo de VELocidade, literally speed control) which is active on most modernised lines in the country. This works on the basis of trains picking up a signal from a track ballaise before every signal section or change of permitted speed with the trains then being limited to that speed until advised otherwise. Any overspeed of over 5kmh creates an audible alarm to the driver, then at either 8 or 10kmh over the limit, an automatic brake application to a dead stand. If the train fails to correctly pick up the signal, then the brakes are also applied to a stand and the train is required to run at 30kmh to the next ballaise to pick up the signal again. Anyone who has travelled on CP here and had their train come to a grinding halt in the middle of nowhere will know what I mean.

 

Pacific231G is spot on as regards comments about the controls in place on the non-modernised lines here with the telephone cantonamente system in place whereby the chefe de estação (station master) is responsable for despatching trains to the next station via a telephone call to his colleague down the line. He will also be responsable for the route setting at his own station with the locking and unlocking of points, as commented on. 

This manual system is in place still between;

Nine and Valença on the Minho line; Caide and Pocinho on the Douro; Abrantes to Elvas & Badajoz on the Leste; Louriçal to Mira Sintra on the Oeste; Casa Branca to Beja in the Alentejo; Tunes to Lagos and Faro to VRSA on the Algarve

 

This is obviously totally dependent on human error and this week (Sep 11th) saw the anniversary of Portugal's worst rail accident, the Sud Express disaster on the Beira Alta in 1985. The station master sent a local Regional service powered by CP 1439 out on to the single track section from Mangualde, not knowing that the Sud Express hauled by 1961 had already entered the section from the opposite direction.

https://en.wikipedia.org/wiki/Moimenta-Alcafache_train_crash

[Pacific231G]

Historic mishaps and accidents involving tokens, whilst informative and interesting, are a long way removed from the OP, which is about a tragic accident in a completely different country. Sorry, but I just think things are getting a little...

As is usually the case with RMWeb topics about accidents, prior to the results of any enquiry, everything else really is just speculation.

Except that it has once again brought up the very different approaches to operating and signalling single track lines that may be relevant to this and a number of other recent accidents.

The fact that different philosphies have been adopted in different parts of the world doesn't mean that other countries are operating in an inherently unsafe way. A "casual wave and order" from the stationmaster may seem decidedly iffy to anyone used to British signalling practice but is in reality the expression of a system that if carried out correctly  is completely safe. The computers used for the French CAPI system (which may be in use elsewhere) add a further level of security. They are not simply PCs running a block instrument programme, so no need for apoplexy, but are very heavily modified and dedicated to carrying out very specific functions and they don't do anything else.