Talk:Eckwersheim derailment

Latest comment: 6 years ago by Mjroots in topic Final report
Good articleEckwersheim derailment has been listed as one of the Engineering and technology good articles under the good article criteria. If you can improve it further, please do so. If it no longer meets these criteria, you can reassess it.
Article milestones
DateProcessResult
January 5, 2016Good article nomineeListed
Did You Know
A fact from this article appeared on Wikipedia's Main Page in the "Did you know?" column on December 21, 2015.
The text of the entry was: Did you know ... that the Eckwersheim derailment in November 2015 is the only fatal accident of the TGV since it entered commercial service in 1981?
On this day...A fact from this article was featured on Wikipedia's Main Page in the "On this day..." column on November 14, 2023.

Nomenclature

edit

Should the more universal designation of "town" be used for Eckwersheim when in France they use "commune"? B137 (talk) 16:19, 14 November 2015 (UTC)Reply

commune seems fair to me UkPaolo/talk 16:36, 14 November 2015 (UTC)Reply

Image

edit
 
This map doesn't show enough detail, including Eckwersheim.

Inevitably an image will be added to the article. An identical locomotive or trainset model image would be second best to the actual site, but at least a map showing the location should be added. B137 (talk) 17:20, 14 November 2015 (UTC)Reply

Location

edit

Using Google Earth I seem to have nailed the location. 48.690233, 7.705263 it is under construction when the image was taken and not available in the greatest resolution. B137 (talk) 17:26, 14 November 2015 (UTC)Reply

Death toll

edit

Are we back to five on the death toll? Initially it was five dead seven injured, seven dead may have been a mix up with the injured number. B137 (talk) 17:33, 14 November 2015 (UTC)Reply

The majority of sources are saying 5&7. Mjroots (talk) 17:41, 14 November 2015 (UTC)Reply
The majority of our referenced sources seem to have updated to 7, despite showing 5 earlier. I've updated the article to reference sources for the death toll from the infobox (and updated the referenced headline titles, where this has subsequently been changed). UkPaolo/talk 17:57, 14 November 2015 (UTC)Reply
Looks like it's now reached 10... :-( UkPaolo/talk 17:59, 14 November 2015 (UTC)Reply

Title

edit

Would Eckwersheim derailment be a better title? Mjroots (talk) 17:41, 14 November 2015 (UTC)Reply

I think it's better because crash tends to imply it hit something or something hit it where this seems to be a pure derailment due to speed around a corner. The bridge seems to be intact, not itself the culprit. B137 (talk) 17:54, 14 November 2015 (UTC)Reply
I'd also support renaming to Eckwersheim derailment per B137's rationale UkPaolo/talk 17:58, 14 November 2015 (UTC)Reply
Page moved I hope I wasn't too hasty. B137 (talk) 18:16, 14 November 2015 (UTC)Reply
The train crashed into water in a canal. It didn't just gently float away. There are a good number of sources calling it a crash. BCdcc (talk) 18:53, 14 November 2015 (UTC)Reply
The crash followed the derailment, unlike the Marden rail crash, where the derailment followed the crash! Mjroots (talk) 22:54, 16 November 2015 (UTC)Reply

The locomotive and several cars derailed ...

edit

Even if this is a quote from a news site: A TGV has no locomotive. As a train set, it is driven by power cars. And not "the", but two of them. Shouldn't we try to avoid contradicting our articles with themselves? 2A02:2028:624:1E01:1030:C127:F475:1F10 (talk) 18:04, 14 November 2015 (UTC)Reply

You're right, but locomotive and power car are arguable interchangeable in non-technical discussion. "Several cars" was me sticking to the text despite the pictures showing fairly clearly that the whole shebang was off the rails. B137 (talk) 18:11, 14 November 2015 (UTC)Reply
How about re-wording to "The power car and several passenger carriages..." then, to align with the terminology of the TGV article? UkPaolo/talk 18:14, 14 November 2015 (UTC)Reply
Another point is that the curve almost certainly had something to do with it. But it may not be in the text of any source yet, though it is clear from pictures. A Google search suggests it hasn't been mentioned yet, so it would probably qualify as WP:SYNTH or even WP:OR. A picture may not be worth a thousand words after all. For something so obvious, it's tedious to have to wait for a lengthy official investigation for certainty. Hang me for my brazen assumptions if the curve ends up irrelevant. B137 (talk) 18:39, 14 November 2015 (UTC)Reply

Article structure

edit

The lede should be a brief summary of the article. There is no need for references in the lede as all facts there should be stated and referenced in the body of the article. Mjroots (talk) 18:40, 14 November 2015 (UTC)Reply

Fair enough, but as things stand all that's happening is a gradual migration of content from the "Accident" section, to the lead, causing duplication... UkPaolo/talk 18:46, 14 November 2015 (UTC)Reply
Migration of content into the lede is fine, but the refs are not needed there. Mjroots (talk) 19:15, 14 November 2015 (UTC)Reply

Accounting for all 49

edit

10 killed plus 32 injured plus five unaccounted for equals 47. There really aren't many options for the other two outside these three categories; unaccounted for is the most likely case...unless of course they were uninjured? Didn't give this a passing thought given the images. B137 (talk) 19:06, 14 November 2015 (UTC)Reply

Yep, surely the other option is that they are uninjured, and hence not fall into any of those categories. Or maybe some of the figures are wrong. Either way, these are the numbers being quoted in the latest articles from both DNA and the Guardian, so we should stick with it. UkPaolo/talk 19:10, 14 November 2015 (UTC)Reply
There does seem to be some disagreement between sources as to the make-up of the 49: the Guardian states explicitly "all those aboard the train were employees of national railway operator SNCF" and yet DNA states that it included "their guests & family" and that there "were children on the train". UkPaolo/talk 19:12, 14 November 2015 (UTC)Reply

In the news

edit

This is a Wikipedia:In the news/Candidates. I oppose its addition. It would cause too much confusion, not to mention the fact that it is only getting so much attention because it comes on the heals of a more worldwide crisis. This is certainly a big deal in France, and at most western Europe as it is part of an international railway system, but not universal enough for the highly selective news blip on the front page of an enormous website who's first purpose is not be a newspaper. B137 (talk) 19:45, 14 November 2015 (UTC)Reply

You know I may have to flip flop here. On a system second only to the flawless shinkansen on safety records, this is "...the first fatal accident since the TGV trains were introduced more than 30 years ago." I have seen this fact touted even recently on articles discussing the safety of high speed rail. B137 (talk) 00:07, 15 November 2015 (UTC)Reply

Speed rounding

edit

The article is about a crash that happened at a (presently) uncertain speed above an approximate 10% overspeed test of a speed limit with a 5% tolerance allowance. These values have imperial conversions for readability using {{convert}}. An edit[1] has now adjusted the {{convert}} rounding so that the values 99, 109, 199, 219 miles-per-hour appear instead of 100, 110, 200, 220 mph—the excess of 9s does not aid readability. Ideally the readability conversions values ending with the 0s would prevail. (The readability conversions ending with 0s are within 0.3%‒0.6%, even disregarding all uncertainities).Sladen (talk) 16:13, 16 November 2015 (UTC)Reply

Done. I didn't understand that from your original edit summary of "sensible rounding". AHeneen (talk) 13:40, 17 November 2015 (UTC)Reply

2369?

edit

"The train, TGV number 2369,". This came from The Daily Star. I added it but it was usurped by other numbers related to the train. Does anybody know what it means and where it came from (besides the Daily Star). I believe it may be specific to (one of) the power cars. B137 (talk) 19:29, 17 November 2015 (UTC)Reply

I don't know what exactly it references. High-speed trains are assembled as a complete set and, unlike freight trains, remain as a set. Only in rare circumstances are the cars replaced/changed individually. This was TGV Dasye set 744. The number 2369 may be for one of the cars, but as far as I can tell most sources use the trainset number 744, which is why it is used in the article. If 2369 is one of the cars in the train, then it probably isn't appropriate to list just one of the cars in the article (there are two power cars, one on each end). AHeneen (talk) 20:42, 17 November 2015 (UTC)Reply
"2369" may be an operating/timetable number. In any case, the Daily Star is not a reliable source. Mjroots (talk) 22:51, 17 November 2015 (UTC)Reply

Overturned vs. derailed

edit

I'm removing the use of "overturned" from the article, as any overturning was subsequent to the derailment due to excessive speed. I.e. derailment was primary, any overturning was secondary. As far as I can see, there's only one reference which actually uses "overturn" in reference to the start of the event, which is the early Railway Gazette article. The later Railway Gazette article doesn't use the word. Despite the flanges on railway wheels, they don't really provide a great deal of sideways resistance at speed, as the flange will climb on top of the rail under sustained extreme contact (something which I think is referenced in the French language official PDF report, but I can't be sure as my French isn't good enough for that). --Murph9000 (talk) 18:25, 20 November 2015 (UTC)Reply

The RG article was the reason for the use of the word "overturn" and I thought the 19 November DNA article ("Un freinage tardif...") also supported it as well. However, upon reading it again and looking up some of the words in a French-English dictionary (I have a working knowledge of French, but am far from fluent) it appears that the DNA article indicates that the train left the track to the left. Mentally, I had pictured the train tilting far to the left (not completely overturned) and striking the parapet. I had not seen the PDF report at the time I updated the article yesterday (19 Nov), only the summaries posted on the SNCF webpage about the accident and some news reports.
I don't have time right now to translate and add the details to the article, but essentially the scenario given by the investigators is found on page 8 (my comments in parentheses; this is not a good quality/exact translation of every important detail, just the main points):
  1. The rear bogie of the front power car derailed. Due to centrifugal forces and the violent transverse movements, the rear of the lead power car swung left and separated from the following car
  2. The rear of the power car then hit the concrete parapet which turned (I guess that means the concrete parapet was pushed outwards, note the way the rebar is twisted in the image in article), then hit the end of the upper flange of the left beam (again, see photo to understand how bridge was constructed) which destroyed the rear of the front power car.
  3. The main transformer was dislocated and projected down onto the east bank of the canal (the bank on the side the train exited the bridge). The oil it contained is spread on the left structure of the bridge. Due to shock and heat-related friction, the oil ignited.
  4. The rear bogie of the power car is embedded in the end of the upper flange of the left beam (visible in photo).
  5. According to calculations made by one of the internal departments in SNCF, given the same external conditions (mass of train, radius of curb, etc.), this phenomenon will occur at a speed between 220-240 km/h. According to the same source, it is the "extreme" bogies (rear bogies??) of cars 1 and 8 that are susceptible to derailing first in this scenario (the train consisted of 10 cars, incl. power cars at each end).
For discussion, the report groups the cars into three groups: 1)the lead power car, 2)the following six passenger cars, and 3)the last two passenger cars plus the rear power car. Continuing at page 20:
  1. After travelling for 15m on the bridge, the front bogie of the lead power car derailed to the left. At the end of the bridge, the lead power car overturned on its left side and slid down the embankment (still with forward momentum, see photo page 20 showing the latter two events) and came to a rest 150m past the end of the bridge. Much of the back of the lead power car disintegrated and parts of the power car were spread along the bridge and the left side of the embankment after the end of the bridge.
  2. The six cars after the lead power car separated from the power car shortly before the bridge (see above) and derailed to the left (outside of curve) due to centrifugal force. They had enough residual speed to overshoot the canal and land on the far (eastern) bank of the canal. They separated from each other and, due to shock, the bogies separated from the cars. These six cars came to a rest between 80-130m from the beginning of the bridge. (page 21)
  3. Passenger cars 7 & 8 (the 8th & 9th cars of the train) came to a rest beside each other on the far (east) bank of the canal. The rear power car, heavier, came to a rest in the canal.
On page 3 and in the conclusion (which are identical), it says something I don't think I fully understand (the parentheses appears in the original): "This excessive speed led to a [délestage] of the axles on the inside [path/rail] of the curve (the right side in the sense of its movement) due to centrifugal force. The [délestage] was sufficient enough to cause/provoke the derailment." When trying to determine what "délestage" means, most results point to a definition of load shedding (related to electricity) and it used in similar ways to indicate "unballasting" (removing ballast from something) or in verb form (délester) to leave something with someone. This sounds like the term "destabilized" used in the later RG article. AHeneen (talk) 03:23, 21 November 2015 (UTC)Reply
Wiktionary defines it as "load shedding". In the case of rail axles/wheels, it may mean "flange climb", a term I am familiar with from RAIB derailment reports. Mjroots (talk) 22:29, 21 November 2015 (UTC)Reply
I have asked for an explanation of the phrase on the talk page of the French version of this article. AHeneen (talk) 06:47, 22 November 2015 (UTC)Reply
French Wikipedia users have explained it clearer. This phrase doesn't indicate flange climb (although on page 10, there is a photo indicating that the flange [French: boudin] rode on the top of the outside rail and later outside the outside rail). The phrase in question above indicates that the wheels on the right side of the train (riding the inside rail of the curve) began to lift off the rail (they were no longer touching the rail...ie. the train began to lean to the left with the right side rising off the rail). This correlates with a statement on page 23 that there was no trace of soil on the right wheels of the power car, which supports the investigators' hypothesis that the right wheels were never in contact with the soil (which I interpret to mean anything other than the rail...ie. soil or ballast). In the first response, the user said they were surprised to see the term délestage used and that it was probably an indication that the weight had shifted to the left side (outside of the curve) of the train, hence the term délestage which I've mentioned could be translated as unballasting. AHeneen (talk) 16:12, 22 November 2015 (UTC)Reply

High-speed rail deaths

edit

This article mentions two accidents - Santiago de Compostela and Eschede as the only other high speed rail accidents causing deaths. Not so! Taking the High-speed rail article's definition of 200 kilometres per hour (120 mph) on existing tracks, there have been at least three in the UK, at Ladbroke Grove, Southall and Ufton Nervet which have resulted in fatalities. Mjroots (talk) 21:16, 20 November 2015 (UTC)Reply

Yes, I mostly agree with that. There's also the Grayrigg derailment. To me, "high speed rail" always started with 125mph max service speed. That said, at this point of the 21st century, I think we're now in the situation of having original generation high speed (maybe 200–250kph), and next generation high speed (>250kph). British Rail had up to 250kph at an advanced stage of development in the 1980s, as part of their original drive for "high speed" from the late 1960s onwards, and would very likely have introduced a home grown "InterCity 250" in the mid 1990s had history played out a little differently between the demise of the APT project and the hiatus caused by the breakup of BR itself. They actually built the 225kph trains, just failed to deliver the 225kph signalling to go with them. --Murph9000 (talk) 21:36, 20 November 2015 (UTC)Reply
I didn't think about the number of different definitions of high-speed rail there are when adding the Eschede incident to the see also section. I think it could be safely removed. However, the Santiago de Compostela incident was relatively similar and occurred in the same geographical region (Western Europe) and relatively close in time, so I think it should remain. AHeneen (talk) 03:28, 21 November 2015 (UTC)Reply
I think the current situation of only linking to Santiago de Compostela is the correct one. Both apparently due to similar causes. Mjroots (talk) 22:26, 21 November 2015 (UTC)Reply

Identity of the victims

edit

I don't think that this was a good faith edit –- I rather believe it was a "Sandy Hook conspiracy" kind of remark. Anyway, most victims *have* been named: [2]

  • Alain Cuccaroni, 59 years (the manager who oversaw the the whole project and the most senior SNCF employee who died that day)
  • Jérémy Landais, 28
  • Stéphane Briet, 42
  • Fabrice Jounot, 43
  • Daniel Heury, 58
  • Marc Lenglet, 48
  • Freddy Martial, 40
  • Christophe Miannay, 49

--Edelseider (talk) 12:54, 21 December 2015 (UTC)Reply

Noted, but as none of these people are notable enough to have a Wikipedia article, there is no need to name them in the article. Mjroots (talk) 15:02, 21 December 2015 (UTC)Reply
Cuccaroni may be notable enough - he was the man who helmed the planning and building of the line and a professor at a prestigious college - but my point was a completely different one: to put a mendacious claim to rest. --Edelseider (talk) 16:13, 21 December 2015 (UTC)Reply
If Cuccaroni gets an article, then he can be mentioned. Mjroots (talk) 17:48, 21 December 2015 (UTC)Reply

GA Review

edit
This review is transcluded from Talk:Eckwersheim derailment/GA1. The edit link for this section can be used to add comments to the review.

Reviewer: Adam Cuerden (talk · contribs) 05:00, 5 January 2016 (UTC)Reply


Well, let's begin. First of all, since most of the sources are in French, I'm going to WP:AGF on sourcing. I don't see why there's any reason to worry. All text (bar part of the lead's summary - as is normal per WP:LEAD) at least has a source, and I see no reason to doubt the factual accuracy and justification of the claims. Sources look strong. I've not checked for copyright violations; I trust you to assure that there's no close-translation of copyrighted text that would amount to a copyright problem.

Secondly, image licensing looks absolutely fine, and are a strong set of images; maybe not quite featured picture level, but that's hardly a requirement.

The text is mostly strong, but could perhaps explain some of the jargon, e.g. "at approximately KP 404.003" - I'd suggest adding a gloss to that phrase; I'd imagine something like "location code KP 404.003" but there's probably a better phrase.

Coverage seems broad enough; I could see naming the victims, but it doesn't seem necessary. I suppose there could be other views on (for example) whether there should have been automatic breaking in the leadup to curves, or whether the investigations are too harsh on the people involved - but that's speculation on things that could exist, not evidence that they do. for a recent subject, this is more than adequate.

Stability is probably the hardest criterion on this article. It's going to need to be kept up-to-date. If you can promise that you will do so, I'm happy to promote.

Current status: Waiting for a response on stability. Adam Cuerden (talk) 05:00, 5 January 2016 (UTC)Reply

@Adam Cuerden:I promise there's no close paraphrasing (except the quotation, of course!) and that the content matches the source. KP is explained the first time it appears (in the "Departure" section): "...it should have slowed from 352 to 176 km/h (220 to 110 mph) before reaching Kilometer Point (KP) 403.809,...". It's the metric equivalent to milepost/mile marker...a reference to the specific location along the line. In the "Background" section, there is the following: "Some automated safety systems were disabled to allow the test train to operate beyond normal operating conditions." I see that that could be expounded upon to mention that the automatic braking system was disabled for the testing.
When I expanded the article a couple days ago, I used a few search terms in Google with the results limited to the past month and there was nothing new about this incident. I think the main issues that will need updated are: 1) whether or not criminal charges are brought against the driver, 2) whether and how long the opening of the line for commercial service will be delayed, and 3) what procedural changes are made with respect to testing by French authorities. The first two are minor updates that I can check on in a few months (shouldn't be hard to learn with a quick web search). The third may take years to learn the impact, with minor updates from time to time, so I think that shouldn't impact the GA as the present article "addresses the main aspects of the topic"(GA criteria 3a). The footnote to that criteria says in part: "it allows shorter articles, articles that do not cover every major fact or detail, and overviews of large topics" (emphasis added). I think the fact that the impact of the incident on the testing process will take a while to realize is reasonable enough to consider criteria 3a met. The stability criteria is more about day-to-day changes, ie. edit wars or an ongoing event (for example, the article for an Olympics would not be appropriate to promote during the games). AHeneen (talk) 13:17, 5 January 2016 (UTC)Reply
Sounds good. Passed. Adam Cuerden (talk) 20:58, 5 January 2016 (UTC)Reply

Recent edit

edit

This IP edit is unreferenced, but may well be factually correct. Can we find a source for this? Mjroots (talk) 06:35, 8 February 2016 (UTC)Reply

I've edited the page to source the edit (and moved the detailed information out of the lead). The French WP uses the term guests (personnes invitées [invited persons]), but the source for that statement uses the term for someone who accompanies or escorts someone (accompagnants...companion [as in a "significant other"] is compagnon). According to the blog of the victims families, one death was a 25-year-old woman who had responded to an invitation to go on the test train with her "companion" [ie. boyfriend], a SYSTRA employee (subsidiary of SNCF that was responsible for testing of the line). AHeneen (talk) 05:49, 9 February 2016 (UTC)Reply

Final report

edit

The final report] was published last March. Needs someone with better French than me to extract the info and expand the article. Mjroots (talk) 20:02, 30 November 2017 (UTC)Reply