'The Santiago Train Derailment Could Have Been Prevented with a Euro 6,000 beacon'

This week's Sunday Train reflected on the fatal train derailment in Santiago de Compostela, in Spain, the issue of mortality in both trains and our primary passenger transport system, self-chauffeured automobiles, and the "PTC" technology that can prevent this kind of accident, as well as the fatal rail collision in 2008 between an LA MetroLink commuter train and a Union Pacific freight train.

The Sunday Train reacted to an early account with: "the ultimate cause is stated directly after: the line is not within the version of Positive Train Control signal system used in Europe, the ERTMS."

More details have since come to light in the Spanish press, which I have seen as translated by European Tribune user Migeru. They reinforce the picture of a poorly designed transition between high speed rail and conventional rail corridors, a system needlessly exposed to the risk of operator error in braking down from 110mph to 50mph to avoid derailment in the curve.

Remember that the ERTMS safety system, which would not have allowed the train operator to go above the safe speed and would have automatically braked the train if the train operator had not applied braking in time, was not installed in the final 4miles of the 54mile stretch of the route to Santiago. As observed by Migeru in an update to his blog post, with the original posted shortly after the crash:

... presumably because the train is supposed to go below 200 on the entire 7-km stretch and ERTMS was seen as "overkill". But this left the line only with the ASFA security system, which can signal to the driver that he's going above the speed limit but requires driver input to slow down the train.

So it is a critical piece of information that Migeru shared in a comment to that blog post, that this section of corridor was not even equipped with the best available non Positive Train Control alert system:

El Diario: El accidente del Alvia podría haberse evitado con una baliza de 6.000 euros (27/07/2013)

Estas fuentes indicaron que no existe en toda la red española otro tramo en el que se obligue a los trenes a bajar de 200 a 80 kilómetros por hora de velocidad. Esta drástica reducción de la velocidad se ve además agravada por el hecho de que la operación está "desastrosamente mal señalizada" con dos señales de Asfa Analógico, un cartel indicador y un libro de abordo, lo que deja en las exclusivas manos del maquinista la realización de un frenado muy drástico sin ofrecer ningún soporte tecnológico alternativo para evitar las fatales consecuencias de un error humano.

El presidente de Adif, Gonzalo Ferre, en declaraciones a la Agencia Efe ha mantenido una postura completamente diferente, que coincide con la de su portavoz oficial. Señala que en los cuatro kilómetros fatídicos no se puede contar con una señalización distinta del Asfa Analógio, y que además, esta es más que adecuada.


Pese a la información oficial ofrecida por Adif de que las balizas de vía del Asfa Digital están todavía en desarrollo, todas las fuentes consultadas han corroborado que este sistema está plenamente operativo como sistema de apoyo en todas las líneas de alta velocidad, en el Corredor Mediterráneo y en muchos tramos de la propia línea entre Madrid y A Coruña, donde se produjo el siniestro.

The accident of the Alvia could have been prevented with a €6k beacon (27/07/2013)

[Engineering] sources pointed out that in the Spanish network there is no other point requiring trains to slow down from 200 to 80km/h. This drastic speed reduction is made worse by the fact that the operation is "disastrously badly signalled" by two analog ASFA signals, an indication sign, and the driver's handbook, which leaves in the hands of the driver the execution of a drastic braking without offering any alternate technological support to avoid the fatal consequences of a human error.

The CEO of ADIF, Gonzalo Ferre, speaking to Agencia Efe has kept a contrary position, consistent with that of his official spokesperson. He claims int he fatal 4km no signalling different from analog ASFA is available and, moreover, this is more than adequate.


Despite the information given by ADIF that digital ASFA beacons are in development, all sources consulted comfirmed that this system is fully operational as a support system in all high-speed lines, on the Mediterranean Corridor, and in many segments to the line between Madrid and A Coruña itself, where the accident took place.

The reason that the digital ASFA beacons are used as a backup throughout the high speed sections of the corridor is that the analog beacons cannot operate above 110mph.

It is not clear that ERTMS was in operation on this train on the day of the accident, as the Spanish system does not require operational ERTMS for 110mph operation. But it appears that an operational ERTMS and ERTMS installed on this section of the corridor would have prevented this accident. Installation of the more capable digital ASFA warning beacons would have provided the driver substantially more effective warning and would have increased the likelihood of preventing the derailment or reducing the severity of the accident if derailment had occurred.

Migeru sums up what at present appears to be the proximate and ultimate causes of this crash in this way:

In sum, it looks like there was a combination of four factors:

  1. the driver had a distraction. At 200km/h it takes 70-90 seconds to traverse the 4-5 km on which the train is supposed to slow down to 80km/h. The driver braked too late if at all.
  2. the Alvia train was not an AVE, it is halfway up from a conventional train, upgraded to run on high-speed track. It is possible that the on-board speed safety system was not certified to use ERMTS because its top speed is lower than regular AVE.
  3. the installed speed safety system on the last 7km of track was done on the cheap, going for fewer analog ASFA beacons as opposed to digital ASFA.
  4. the design of the speed safety system was not fail-safe: excluding ERTMS from the last 7km of track implies the assumption that the train will never enter that stretch above 200km/h.

Or, in short, a human erred, causing the crash. And humans will err. A well designed safety system takes this into account, but it appears that corners were cut in the sections of conventional rail corridor connecting to high speed rail corridor





Your rating: None Average: 3 (2 votes)


Eurotrib story updated

Migeru's picture

With links to sources and more information in the comments.

Your rating: None Average: 3 (3 votes)

Yes, I saw that later ...

BruceMcF's picture

... the point that the "distraction" for what seems to be the distracted train operator seems to have been a call from control, and the train operator may have been looking at a map, only underlines the need to extend the EU implementation of Positive Train Control through to the end of the 50mph curve, and to ensure that the hybrid 110mph "Rapid Passenger Rail" sets operate under the Positive Train Control where it is available.

Your rating: None Average: 3 (2 votes)