Welcome To Zelo Street!

This is a blog of liberal stance and independent mind

Sunday, 4 August 2013

Spain Rail Crash – What Did I Tell You?

As soon as it looked like Francisco José Garzón was going to be scapegoated for the derailment near Santiago de Compostela, which left 79 dead and many more injured, I pointed out that there were, whatever the driver’s actions, questions over the safety systems in use, the stability of the 730 series Talgo/Bombardier hybrid trains, and the way the coaches protected their occupants in a crash.
Now, as the idea of using Garzón as a scapegoat appears to be receding, infrastructure operator Adif is belatedly revisiting not just the approach to the derailment point, but all similar locations across the Spanish network, with a view to installing overspeed protection, which as I pointed out at the outset, was so sadly lacking in the 4km before the A Grandeira curve.

So how is Garzón’s situation changing? Well, after guard Antonio Martín asserted that he was not responsible for the derailment, despite having apparently broken the rules and phoned the driver on his company mobile on a matter which certainly was not an emergency, the investigating judge decided he would not prosecute him. How a reckless homicide charge then sticks only to the driver is not clear.

What more significantly undermines the charge laid against Garzón is the decision of Adif to retrospectively enhance the installation of the ASFA safety system by installing additional “beacons” which will enforce progressively lower speeds as trains approach speed limited curves like the one where the derailment occurred. This is despite previously insisting that provision was adequate.

ASFA is rather like the UK’s TPWS system: it prevents signal overruns, but can also be used to slow trains before junctions, or, as in this case, on plain track before severe speed limits. And another factor that chips away at the assurance that all was previously well is that the speed limit to be enforced approaching the A Grandeira curve will be just 30km/h, not 80 as previously.

That last speaks to the questions that have been raised about the stability of the diesel generator cars in the 730 series trains. It was the lead generator car that tipped over and left the track first, dragging the lead power head and the coaches behind it into the concrete retaining wall. All that this blog questioned at the beginning, and that can be addressed in short order, has now been addressed.

This is how the railway remains the safest method of land based travel: not by pretending that problems do not exist, but by admitting that they do, and solving them. As Renfe’s CEO put it, “Every day we have 4,600 trains running that are completely safe”. They have now: let’s hope it stays that way, if only for the memory of the family whose stop at Pontedeume prompted the guard’s fatal phone call.

There were four of them, but, sad to relate, two never made it home.


Dave H said...

Tim - please watch the main crash video again and more closely.

There is a ripple that sets off in the train and lifts a Talgo car in the rake - roughly car 4-5 and it would appear that this is the action which splits out the central portion of the Talgo set, and the front portion of the train, with the loco and generator car still on the rails continues around the bend, until the flailing Talgo car catches the end of the retaining wall and the jerk pulls the loco over, with the rest of the front section then travelling outwards by centrifugal force.

The ripple moves back through the Talgo rake and another car lifts up, splitting out the centre section of the train, and without the weight of the loco and generator car this section heads tangentially out from the curve. Depending on which way the Talgo rake was aligned the leading car of the central section may well have been without a wheelset to guide the leading end and thus predisposed to head off from the track at a tangent.

With the weight of the generator car and rear locomotive keeping the train on the track, the rear section kept aligned with the rails, even if not all wheels were running on them. This then carried on round the curve and struck the carriages of the front section, which by then had come to a stand.

The ripple appears to start when the front of the train is hidden from the CCTV camera by an OHLE mast. It could be a small track irregularity, or soft spot, with the effect magnified by the forces acting at that raised speed.

Alternatively it could be the point at which the driver made a severe brake application, as appears to have been the case, and as you will probably be aware, the rate at which each individual wheel set is decelerated can vary and can lead to a braking 'wave' effect. In old loose coupled, and generally unfitted freight trains the forces building up with cumulative accelerations travelling along the train could rip the drawgear off wagons or crush them to matchwood as a ripple effect progressed. With a modern close coupled train, and fast acting, closely managed braking this effect is largely controlled under normal braking conditions, but may have created an undesirable event in this instance.

I also note the possible significance of the investigators call to measure very carefully the wheelsets of all vehicles in the train, as the possibility of a gauge change variation or uncontrolled movement of the variable gauge wheels needs to be eliminated as a direct causal factor, and as a factor in the development of the derailment.

Whilst there seems to be little doubt about the fact that the train entered the curve travelling at a speed in excess of the posted limit, both the conservative calculation of that speed limit, in terms of the train's ability to stay on the track, and the fact that for most situations where the train running gear and the track are in good order, the passengers would be thrown against the outer side of the carriages by centrifugal force, well before the train reached the limits for falling off the track.

In most derailment investigations the event initiation takes place a distance before the main area of mayhem. Unfortunately we have few pictures of this point, and the movement of the train as it passes the point where things start going wrong is masked by an OHLE mast on the main CCTV record available.

Tim Fenton said...

As I've already answered on Twitter, you should look again at the video.

It's not just my contention, but also that of Prof Sierra of the University of Granada (link in this post http://zelo.tv/1aYDzzQ).

RoSi said...

Very well written. Thank You. Please keep updating Santiago crash.

John Palmer said...

Looks like Garzon is going to ber made the only scapegoat after all: http://www.tumbit.com/news/articles/8172-driver-now-only-suspect-in-train-derailment.html