 |
Re: Multi-fatality train wreck in NW Spain
Author: OldPoleBurner
Date: 07-26-2013 - 12:39
According to several railroad trade journals now, the train was operating under ETCS level 1 - limited supervision mode; which is functionally equivalent to the Congress's current PTC mandate in this country . Limited supervision mode merely means that supervision is limited to enforcing various safety requirements; such as signal aspects, following distance, maximum speed limits, curve restrictions, brake applications, etc.
Braking supervision in PTC and ETCS enforces the totality of the braking profile, from the 1st command to reduce speed to a complete stop, or to the new commanded speed. Any violation of the braking profile (measured by accelerometers) would result in service brake re-application, or if already applied and ineffective - a full open loop emergency application. It also would cut off the throttle - all automatically!
Given the above safety system in place, it is hard to believe that the train driver could have managed to achieve nearly double the authorized speed, without intervention long before, of the ETCS (PTC), to bring the train back under control. Either the ETCS failed in some way, or the onboard connections between propulsion and braking systems and the ETCS were faulty. However, in any proper design, a fault here should have automatically tripped the propulsion system offline. But it didn't!
It is possible the safety equipment was tampered with, as in the 1987 Chase Maryland Amtrak collision, but nothing so far indicates this. Reports that prior to the accident, the panicked train driver used a cellphone instead of regular communication channels to report his excessive speed; suggests it probably wasn't him; and that some kind of railroad radio system outage was also occurring.
While ETCS level 1 generally does not use radio ranging to fix a trains position, or the radio to transmit safety related data; The ETCS-1 specification permits it. If that was the case here, the apparent radio outage could have also interfered with the ETCS.
Aside from the sad (and probably unnecessary) enormous loss of human life; this accident has huge implications here at home! The mechanical implications posed by the very odd way this derailment occurred are huge. Rather than overturning the lead unit, a second or third unit climbed the rails (flange over gauge corner) without first turning over. The turnover occurred later in the process.
Moreover, the total disintegration of the passenger compartment of several cars, should not have happened by only the glancing blows seen on the video. This has also been a problem with Germany's ICE. Its crashes have been more deadly than this one! Weak car bodies have also been a problem over the years with BART, and with WMATA cars as well! The odd way of derailing and the weak car-bodies all suggests to me the possibility that there are also inherent weakness in the "Talgo" design itself, that may have contributed once the other failures occurred. But that is outside my engineering discipline, so I will leave it to the mechanical engineers to figure out.
But after a 40 year career in railway signal engineering (and many accident investigations); I am still almost speechless that such failures continue to occur, especially in Europe. This is the forth I know of in the last ten years, where ETCS was implicated. This has huge implications here; since in the USA, we are now installing similar systems, which ARE using radio for safety data and train location via GPS (radio ranging).
If these poor souls are NOT to have died in vain; we must not continue to insist on using technologies with known flaws and hazardous failure modes. Especially when more reliable and inherently safer technologies are readily available. We must LEARN that it will no good whatsoever, to replace flawed human responses and errors; with flawed machinery, along with its also very deadly errors. Murphy's Law says that if it CAN happen - It WILL happen.
If we Do Not learn this, Such deadly accidents WILL happen here too!; and will continue in Europe unless they learn it too!
OPB
Braking supervision in PTC and ETCS enforces the totality of the braking profile, from the 1st command to reduce speed to a complete stop, or to the new commanded speed. Any violation of the braking profile (measured by accelerometers) would result in service brake re-application, or if already applied and ineffective - a full open loop emergency application. It also would cut off the throttle - all automatically!
Given the above safety system in place, it is hard to believe that the train driver could have managed to achieve nearly double the authorized speed, without intervention long before, of the ETCS (PTC), to bring the train back under control. Either the ETCS failed in some way, or the onboard connections between propulsion and braking systems and the ETCS were faulty. However, in any proper design, a fault here should have automatically tripped the propulsion system offline. But it didn't!
It is possible the safety equipment was tampered with, as in the 1987 Chase Maryland Amtrak collision, but nothing so far indicates this. Reports that prior to the accident, the panicked train driver used a cellphone instead of regular communication channels to report his excessive speed; suggests it probably wasn't him; and that some kind of railroad radio system outage was also occurring.
While ETCS level 1 generally does not use radio ranging to fix a trains position, or the radio to transmit safety related data; The ETCS-1 specification permits it. If that was the case here, the apparent radio outage could have also interfered with the ETCS.
Aside from the sad (and probably unnecessary) enormous loss of human life; this accident has huge implications here at home! The mechanical implications posed by the very odd way this derailment occurred are huge. Rather than overturning the lead unit, a second or third unit climbed the rails (flange over gauge corner) without first turning over. The turnover occurred later in the process.
Moreover, the total disintegration of the passenger compartment of several cars, should not have happened by only the glancing blows seen on the video. This has also been a problem with Germany's ICE. Its crashes have been more deadly than this one! Weak car bodies have also been a problem over the years with BART, and with WMATA cars as well! The odd way of derailing and the weak car-bodies all suggests to me the possibility that there are also inherent weakness in the "Talgo" design itself, that may have contributed once the other failures occurred. But that is outside my engineering discipline, so I will leave it to the mechanical engineers to figure out.
But after a 40 year career in railway signal engineering (and many accident investigations); I am still almost speechless that such failures continue to occur, especially in Europe. This is the forth I know of in the last ten years, where ETCS was implicated. This has huge implications here; since in the USA, we are now installing similar systems, which ARE using radio for safety data and train location via GPS (radio ranging).
If these poor souls are NOT to have died in vain; we must not continue to insist on using technologies with known flaws and hazardous failure modes. Especially when more reliable and inherently safer technologies are readily available. We must LEARN that it will no good whatsoever, to replace flawed human responses and errors; with flawed machinery, along with its also very deadly errors. Murphy's Law says that if it CAN happen - It WILL happen.
If we Do Not learn this, Such deadly accidents WILL happen here too!; and will continue in Europe unless they learn it too!
OPB
