TGV Derails

[Northroader]

I notice that during the course of the crash, all the cars became detached. Most of the recent crashes involving speed over here, the couplers did at least stay together, which did help to minimise the results. Presumably the speed was that much higher and caused all the couplings to fail?

[Titan]

TGVs are articulated, so there are not couplings as such, except perhaps between the power car and first vehicle.

[Fat Controller]

TGVs are articulated, so there are not couplings as such, except perhaps between the power car and first vehicle.

There are UIC screw couplings between the power-cars and the ends of the rake, I believe; otherwise all the vehicles are articulated; this how they have managed previously to have derailments at very high speeds without serious injuries to passengers, let alone fatalities.

[Fat Controller]

Some further news here: the train was travelling at 176 kph, which would be the 10% overspeed for the 160 kph band on TVM. This was the permitted speed for the section of line in queation.

http://www.capital.fr/a-la-une/actualites/le-conducteur-du-tgv-accidente-nie-tout-exces-de-vitesse-1085930

Other reports from official sources indicate the power car hit the bridge before the train derailed.

[Nearholmer]

FC

 

I read it to be saying that the train derailed as it entered the right-hand curve, with the power car off the road to the left, and that the power car then struck the bridge abutment. If that is what it is saying, it looks consistent with what can be seen in the aerial photo.

 

K

[Mike Storey]

Some further news here: the train was travelling at 176 kph, which would be the 10% overspeed for the 160 kph band on TVM. This was the permitted speed for the section of line in queation.

http://www.capital.fr/a-la-une/actualites/le-conducteur-du-tgv-accidente-nie-tout-exces-de-vitesse-1085930

Other reports from official sources indicate the power car hit the bridge before the train derailed.

 

AFP confirms the same report, almost word for word, but neither say what you have said - they say the investigating magistrate reported that "He (the Driver) has indicated that he respected the speed limit imposed by the route plan" and that meant that the train was moving at 176kph, but that they await confirmation of this from the black box. No reason for the Driver to fib of course, as he would be found out anyway, but the word "respected" (other translations are possible) suggests that the normal TVM speed control was turned off and the train speed was solely at the discretion of the driver.

 

Further down, the reports state that there appears to be no credibility to suggestions of sabotage or terrorism. But they do not completely rule it out.

[jjb1970]

I had to switch off the news over the weekend. After the ghastly events in Paris and this dreadful train crash I just found it terribly depressing. We can only express our deepest sympathy to those who have lost loved ones and the injured.

With regards the technicalities, I have every confidence that a thorough investigation will be completed which will inform us of what happened in due course. Hopefully some good will come out of this by identifying improvements to prevent future incidents.

[roythebus]

there was an interesting aspect of this published in the Huffington Post on Facebook earlier today, probably in the British Rail Old School group; a well-written piece by someone who knows about these things. Regrettably i couldn't manage to make the link copy across to here.

[petethemole]

This one? http://www.huffingtonpost.co.uk/karl-davis/48-hours-in-france-rules-_b_8572734.html

 

I couldn't find anything else on HP in a quick search.

[The Stationmaster]

The interesting thing about the Huffington Post article - which is very sound - is the reference to the presence of children.  While SNCF very senior management have said they shouldn't have been there what we do not know as yet is if someone had actually authorised the presence of families/family members on the train.  It has certainly been the case in the past that on SNCF trial runs (perhaps not quite the same as a 'test train'?) persons other than those involved in the actual test/trials work have travelled on such trains and it has not just been SNCF employees but also family members travelling with them for a day out.

 

Whether or not that was 'custom & practice' or official policy I do not know, neither do I know if (assuming it was unofficial) it has subsequently been banned by SNCF senior management.  It has definitely been the case in Britain that family members have travelled on test or trial trains, and were specifically invited to do so as part of the tests being carried out, but this has to my knowledge only occurred on fully commissioned trains over fully commissioned routes which of course was not the case with the recent SNCF derailment.

[Oldddudders]

The Eurostar service introduction in 1994 was preceded by a period of 'Discovery' service, when all sorts of people were invited to try it out gratis. But this was in line with Mike's note that the service was running under full operating conditions.

 

In the real world, dads are proud of what they do. Bringing the family along for a free ride sounds innocent enough.

[Nearholmer]

OD

 

I thought the same, and that the Huffington poster was extending a point slightly too far.

 

Late on in the testing phase, just before things move on to "operational trials", things settle down into a sort of 'routine data gathering' pattern; it was a Saturday; you're proud of the work;and, maybe son or daughter is beginning to show an interest in these things, doing well in maths and physics at school ........

 

K

[Edwin_m]

It must depend on whether the run in question is "bedding down" the service under all the rules applicable to passenger operation but without fare-paying passengers on board.  Staff and their relatives will cause some of the issues that the railway has to get experience in dealing with when transporting real passengers, but are less likely to complain (especially if they haven't paid) if the staff aren't completely slick when handling them.  Obviously this must not extend to safety issues!

 

It is more open to question in this case where it seems to have been established that the train was intended to exceed normal permitted speeds and (perhaps more importantly) the train protection that would normally prevent such overspeeds was suspended.  I think we are still waiting independent confirmation of what the actual speed was, and we still don't know whether this was some event that could only arise during testing or something that could equally well have happened after the line opened to service. 

[The Stationmaster]

The Eurostar service introduction in 1994 was preceded by a period of 'Discovery' service, when all sorts of people were invited to try it out gratis. But this was in line with Mike's note that the service was running under full operating conditions.

 

In the real world, dads are proud of what they do. Bringing the family along for a free ride sounds innocent enough.

SNCF were running trains with some 'passengers' (i.e. not staff members) in them before that Ian, Eurostar UK ran some with other than directly involved operational staff aboard as well but I don't think any of theirs actually involved shopping trips to Paris.

 

Not exactly railway related safety issues (in the normally understood sense) were tested using families of Eurostar staff and a few other folk in a situation where a particular location was not open for Eurostar public operation but in that case the specific need was for a test involving the equivalent of a train loaded with ordinary passengers in order to test and prove safety arrangements for public operation.  This of course is still not directly comparable with the recent incident on SNCF.

[dullsteamer]

In the real world, dads are proud of what they do. Bringing the family along for a free ride sounds innocent enough.

In my experience its not an uncommon thing to do. Back when I was on the tools we used to do late-night trial trips with power cars that had just completed component change-out. We'd couple the CCO car and the spare power car back to back and then set off for a run over to Glenfield. Most nights the whole depot would come along for the ride and it would be standing room only. Many times the drivers would bring their kids along as well.

 

Cheers,

 

Mark.

[Fat Controller]

SNCF were running trains with some 'passengers' (i.e. not staff members) in them before that Ian, Eurostar UK ran some with other than directly involved operational staff aboard as well but I don't think any of theirs actually involved shopping trips to Paris.

 

Not exactly railway related safety issues (in the normally understood sense) were tested using families of Eurostar staff and a few other folk in a situation where a particular location was not open for Eurostar public operation but in that case the specific need was for a test involving the equivalent of a train loaded with ordinary passengers in order to test and prove safety arrangements for public operation.  This of course is still not directly comparable with the recent incident on SNCF.

'

I took part in several of these; in the Channel Tunnel before E* services were allowed to restart after the 1st fire (as Lynne had written the evacuation procedures, this was a three-line Whip), in North Downs tunnel and on the Medway Tunnel for CTRL Section 1 and in the London Tunnels for CTRL Section 2. On some of them, in order that conditions should be as realistic as possible, participants were invited to bring along small children and ageing relatives. During the Channel Tunnel one, I walked a couple of kms with some's offspring on my shoulders.

There was at least one jolly to Paris after one section of CTRL was completed.

Eurotunnel had a long period of Marche a Blanche before formally opening; at various points, staff and families, local residents and dignitaries were invited. It used to run until 22:00 local time, which led to a couple of 'Channel Dashes' to get the last train.

[Fat Controller]

Latest news that I've found:-

 

http://www.lefigaro.fr/societes/2015/11/19/20005-20151119ARTFIG00209-un-freinage-trop-tardif-a-l-origine-du-deraillement-du-tgv-pres-de-strasbourg.php

The report suggests that the driver started braking a kilometre or so later than he should have done, and was travelling at 265 kph, rather than the 176 kph he was supposed to be doing. The precision of the figures suggest they've come from the ATESS disc, which records speed, signal aspects and brake applications. As to why he was travelling so much above the planned test speed, that'll have to wait until the BEA-TT interim report.

As I have previously stated, in order to do the 10% overspeed tests, the TVM would have had been isolated, so the driver would have been driving 'on sight'.

[JDW]

Latest news that I've found:-http://www.lefigaro.fr/societes/2015/11/19/20005-20151119ARTFIG00209-un-freinage-trop-tardif-a-l-origine-du-deraillement-du-tgv-pres-de-strasbourg.phpThe report suggests that the driver started braking a kilometre or so later than he should have done, and was travelling at 265 kph, rather than the 176 kph he was supposed to be doing. The precision of the figures suggest they've come from the ATESS disc, which records speed, signal aspects and brake applications. As to why he was travelling so much above the planned test speed, that'll have to wait until the BEA-TT interim report.As I have previously stated, in order to do the 10% overspeed tests, the TVM would have had been isolated, so the driver would have been driving 'on sight'.

And at 256kmh - or even 'just' 176 - a kilometre will not take long at all, less than 20 seconds...

[Fat Controller]

Since the last post, someone forwarded this posting from the UIC to me:-

France: Test train accident at Eckwersheim – initial findings of the immediate internal SNCF investigation and statement from the management team

[http://uic.org/com/local/cache-gd2/b14ccc1937152535a8e1c1b5ab842aaf.jpg]<http://uic.org/com/IMG/png/carte-2.png>

 

In the minutes following the test train accident which took place on Saturday 14 November at Eckwersheim on the new Est Européenne line, the managing director of the Railway Safety System launched, at the request of the SNCF management team, an immediate investigation via the SNCF Audits and Safety directorate general. Investigators from the Audits and Safety directorate general went to the accident site in order to gather and understand the initial findings of the accident. The internal report on the immediate investigation was submitted to the Chairmen on Thursday 19 November.

 

Background to the accident

 

The accident took place outside the operational national rail network, as running on this line section is managed by those in charge of testing. The tests are intended to analyse and gain experience of the properties of the new line, and include running at speeds up to 10% greater than commercial speeds. To do this, the some automated speed control systems on board the test train are disabled as part of the test protocol and are instead controlled by a team of experts. This is not the case for commercial services running on the operational national rail network.

This test train run was performed as part of dynamic testing to study the properties of the line before it was approved. For information, its scheduled commercial opening was the 3 April 2016.

 

All of the components of the line’s sub-systems had already been verified. Since 28 September almost 200 technical test runs had been carried out on this future new line (cf “Temps Réel”, issue No. 77). This was to be the final run of the dynamic testing phase. The test train was fitted with measuring equipment and composed of two power cars and eight coaches. 53 people were on board: SNCF Group engineers and technicians, and their companions. There were 7 people in the driving cab.

 

The facts at the time of the accident

 

The test train left the Meuse TGV station at 14:18 and in accordance with the test protocol, gradually increased speed on the line to reach 352 km/h, 10% above the maximum commercial speed of 320 km/h. It then began a deceleration phase once it reached Vendenheim, as this is where the connection to the conventional line is located, allowing access to Strasbourg.

 

Immediate cause of the accident

 

After examining the black boxes (Atess recording tapes), the investigators established that the speed of the test train as it entered the section was greater than the speed assigned for the test run. The assigned operating speed was 176 km/h and the speed recorded was 265 km/h. The accident took place at a speed of 243 km/h.

This increased speed was due to a late braking sequence of the test train.

As a result of the centrifugal force exerted, the lead power car derailed and came to rest on the embankment. All the other investigations of the internal investigators have confirmed that no abnormalities were found with regard to the infrastructure of the line or the rolling stock. In addition, there were no abnormalities in terms of the traffic management of the test train.

 

Continuation of the interval investigations

 

It is too early to know the causes of this late braking sequence. Investigations will continue, focusing on procedures and human error.

 

(Source: SNCF)

[Mike Storey]

It will be interesting to learn why the driver, reportedly, stated to the prosecutor the day after the accident, that he believed he was travelling at the authorised speed (176 kph). The report does not say whether this driver had undertaken the run under the same circumstances before.

[Nearholmer]

Hmmm ....... All poses more questions than it answers, methinks.

 

1) was the driver actually in charge of train speed, or was the "team of experts", through the automated systems?

 

2) who were the seven people in the cab?

 

3) why were they there?

 

4) what affect might the presence of that number of people have had on the driver's concentration?

 

5) come to that, was the assigned driver actually in the cab, and was he/she at the controls?

 

6) ......... Lots more questions.

 

Still loads we don't know!

 

Kevin

[The Stationmaster]

There should be plenty of room for 7 people in a TGV cab (I've been in one with 5 of us, including the Driver, and we were hardly fighting for space) - far more important is what were the 7 doing and your point about their impact on the Driver's concentration.

 

Your point about the 'team of experts' is very pertinent as it is possible to 'drive' some TGV variants from a laptop computer allowing various indications, fault reports, and controls in the ATESS to be over-ridden.

[Nearholmer]

Before saying too much more about that topic, it is worth considering that not every clever bod with a laptop is well-intentioned, and that anyone in the world can read RMWeb.

 

K

[Fat Controller]

The 'automated systems' had been isolated for test purposes, as I have previously said, and the report confirms. They would have to be, as even a slight overspeed sets in course a full emergency brake application, automatic lowering of pantographs, all sorts of bells and whistles in the cab and control centre, and a flurry of paperwork. The problem then is that the driver has to be very conscious of where he or she is- it's to be hoped that distraction was not a factor.

Even with SNCF levels of staffing ('Sur Neuf, Cinq Faites rien' is how I've heard SNCF personnel refer to the organisation) seven does seem excessive. I'd guess at driver, Traction Inspector, Chef des Essais and perhaps someone logging the info from the power-car gauges as a base, but who were the others?

[Nearholmer]

"The 'automated systems' had been isolated for test purposes, as I have previously said, and the report confirms."

 

Except, that isn't precisely what the report says. What it does say on this point can be read to have several different meanings, and I take it to be deliberately ambiguous, pending full discovery/publication of the facts.

 

K