As the wreckage is taken away, the dead identified, and the injured nursed back hopefully to full health, the investigation has begun as to how a Madrid to Ferrol train entered an 80km/h curve at well over twice that speed, derailing in the process. Focus will inevitable turn to the train’s safety systems, and why they did not step in and slow the train down.
For starters, let’s bin the idea that you can’t safely operate a high speed line that has an 80km/h curve at the end of it: the very first Spanish AVE has an even more severe 70km/h reverse curve as it passes through the town of Puertollano, and has been in use with no mishap for around 20 years. But it does have its train control system switched on at all times.
The reason I mention this is because El Pais has reported news which, if true, would explain how the accident happened, and moreover, suggests that such an accident has been waiting to happen ever since the high speed link from Ourense to Santiago de Compostela was opened in 2011. And the safety system installed at the point of derailment would have been unable to intervene and slow the train.
This is what El Pais reported: “Thales ... confirmed to this newspaper that ... ERTMS Level 1 was installed to Kilometer 80 of the line ... and tested and delivered in November 2011 ... ERTMS works perfectly on ... Madrid-Valladolid, but these trains do not use it ... in ... Ourense-Santiago, even in the 80km where already installed and tested, confirmed sources” [my emphasis].
It’s entirely possible that something has been lost or gained in translation from Castellano, but consider this: as shown in the diagram above, ETCS hands over to the older Spanish ASFA system very close to the 80km/h curve, and knowing that ASFA could not intervene in time to deal with a serious overspeed incident, any competent ERTMS/ETCS installation would have provided for slowing trains.
And, if ETCS was working, why the need for two drivers, as appears to have been the case here? All that was needed was for whoever was at the controls to miss his braking point for the curve, and with a number of tunnels immediately before the curve, the potential for disorientation is clearly present. Miscount the tunnels, and there is the recipe for the disaster that followed.
On top of that is the downhill gradient before the curve, meaning a longer braking distance would be needed, and the increased weight of the 730 series train (with its diesel generator cars in place of two more coaches) over the more conventional 130 series, making the braking distance longer still. All of this fits with the assertion in the El Pais article that ERTMS was not in use.
That’s the grim conclusion to which my Occam’s Razor is pointing right now.