• Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!

High Speed Derailment in Italy

Status
Not open for further replies.

Railperf

Established Member
Joined
30 Oct 2017
Messages
2,922
I consider that one to be on a lower-speed connecting track not an actual high speed line. Similarly for the one at Santiago de Compostela.

There have been accidents on the French high speed lines - a derailment on TGV Nord due to undetected WW1 excavations springs to mind. But no fatalities.
Yes Santiago de Compostela was the end of a high -speed line that was not fully commissioned as such. So speed limits were approximately 200km/h. The driver had been distracted as to his actual position and failed to slow the train in time for an 80 km/h bend. And yes the LGV Nord derailment caused some injuries but no fatalities.
The 125mph ICE derailment at Eschede may not have been on a high-speed line - but killed 101 people. That was due to a faulty wheel set initially. but the combination of the derailed train hitting points and switching them into another position caused part of the train to diverge onto a loop line , one of the carriage ends struck a bridge support - demolishing it and causing a road bridge to collapse down on the train.
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

Railperf

Established Member
Joined
30 Oct 2017
Messages
2,922
The track of the debris is unbelievable.
There could so easily have been severe casualties on the ground as well as on the train.
I can't see the line reopening quickly.

As you might expect, there is extreme disruption to High Speed services in Italy.
Services not truncated or cancelled are using the classic line Bologna-Piacenza-Milan, with local disruptions as a result.
It might even be quicker/more reliable to reroute via Verona or Padova.
https://www.trenitalia.com/it/informazioni/Infomobilita/notizie-infomobilita.html

Trenitalia also have a headline offering condolences to the deceased, with maximum support for their families.
Can anyone see the line opening for at least a couple of weeks minimum?
The investigation has to conclude first. The various authorities need to carry out a forensic examination of the area before any recovery work can be done and repairs. The recovery of the black box will confirm what the drivers actions were and the state of signalling etc.
There is significant damage to the track, overhead wiring and signalling. Maybe at the bare minimum they might be able to allow some bi-directional running at low speed through the affected area along the northbound track - but that will add safety risk to track workers. Probably unlikely to happen.

The Milan to Bologna high speed line has several connections to the 'classic' main line, so it looks like trains will use the 'classic' line from Milan Rogoredo to Piacenza - and then they can use the Piacenza Est connection back onto the high-speed line - a diversion of around 30 miles - but due to several low speed restrictions, will probably add another 20 to 30 minutes minimum to journeys.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
Yes Santiago de Compostela was the end of a high -speed line that was not fully commissioned as such. So speed limits were approximately 200km/h. The driver had been distracted as to his actual position and failed to slow the train in time for an 80 km/h bend. And yes the LGV Nord derailment caused some injuries but no fatalities.
The 125mph ICE derailment at Eschede may not have been on a high-speed line - but killed 101 people. That was due to a faulty wheel set initially. but the combination of the derailed train hitting points and switching them into another position caused part of the train to diverge onto a loop line , one of the carriage ends struck a bridge support - demolishing it and causing a road bridge to collapse down on the train.
Indeed so. My point is that Santiago de Compostela would not have happened if the route had been fitted with the train protection systems that are standard for high speed lines. And the bridge at Eschede was of a design that was vulnerable to train impact, which would not have been allowed on a high speed line. As far as we know this route has similar safety measures fitted but (as with the collision in China) they haven't prevented a serious accident.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
Indeed so. My point is that Santiago de Compostela would not have happened if the route had been fitted with the train protection systems that are standard for high speed lines. And the bridge at Eschede was of a design that was vulnerable to train impact, which would not have been allowed on a high speed line. As far as we know this route has similar safety measures fitted but (as with the collision in China) they haven't prevented a serious accident.
The scandal at Santiago de Compostela is that it would have been possible and comparatively easy to fit the standard 'ASFA' legacy train protection system in a special speed trap configuration, but authorities had been reluctant to use it in such applications (shades of BR AWS at speed restrictions until after Morpeth), and perhaps they also thought because it was only a temporary configuration, pending gauge conversion and provision of full ETCS throughout the route, it wasn't worth the very small expense. The train was equipped with ASFA but ETCS on board had never been commissioned on that class, even though ETCS was active on parts of the new line and worked with other train classes, but not for 4km on approach to the derailment site clearly, so there was a complete gap in protection for all traffic in the area. It is very telling that ADIF engineering staff very quickly designed and installed the appropriate ASFA trackside equipment WITHIN DAYS of the incident. In this respect, it was very similar to Amtrack's Frankford Jn and Metro North's Spuyten Duyvil derailment in curve incidents in the USA, where legacy speed protection equipment used widely elsewhere on the respective networks was not fitted in the specific locations, and then was installed quickly on the tracks immediately after the incidents, after the horse had bolted so to speak.
 

philabos

Member
Joined
19 May 2010
Messages
180
Location
Lancaster PA
The Frankford Jct derailment had over speed protection installed by the Pennsylvania Railroad in the 1950's. Amtrak deactivated the system for eastbound trains (including the train that derailed) , but it was still active for westbound trains. Service was restored quickly because they reactivated the existing system to cover until PTC was installed.
 

Railperf

Established Member
Joined
30 Oct 2017
Messages
2,922
Track workers named as suspects!

There is a turning point in the investigation into the Frecciarossa disaster. The Public Prosecutor of Lodi has entered the first five names in the register of suspects for railway disaster, manslaughter and multiple injuries. These are the workers of the team that worked in the night between Wednesday and Thursday on the turnout from which the derailment was triggered. The acceleration to the investigations came after a day of technical surveys of the Polfer and Scientific Research Unit operating on the remains of the ATR 1000. But now it is necessary to analyze in depth the exchange that triggered the disaster. Unrepeatable exams which must therefore involve the participation of lawyers and consultants of potentially suspect persons. Just to guarantee the defenses, therefore, the prosecutor of Lodi Domenico Chiaro has decided to notify the warranty information to verify any culpable liability in the accident. An act due in the case of the workers who worked on the exchange that night and on which the attention of the investigators was concentrated .

The sensors
Listened to as people informed of the facts by the investigators as early as Thursday afternoon, the workers said they had left the exchange in the correct position with respect to the running direction of the Frecciarossa. The "hydraulic diverter" was however found in an "inverted" position by the investigators,that is, open to a service track. A circumstance not reported by the monitoring system of the Bologna plant since the exchange had been "isolated" by the technicians during and after the maintenance works. And for this reason the safety sensors could not transmit the deviation signal which would immediately slow down the train's travel. Instead the convoy launched at 290 kilometers per hour came off the tracks and the leading car broke away from the rest of the train killing the train drivers Mario Dicuonzo and Giuseppe Cicciù.





carrello-treno-u31402250702662c0g--656x369corriere-web-nazionale_320x180_M.jpg

Video: Train derailed, the dynamics of the accident: the images a few meters from the cars







The phonogram
At 4.45, one of the five technicians investigated had sent a phonogram to Bologna to warn that the exchange had been "isolated" from the signaling system and above all to communicate that the turnout was placed "in a normal position" (therefore in line with the train track ). For this reason, the investigators want to reconstruct what actually happened and what works were carried out by the workers, not suspended by the Italian railway network but in the meantime destined for "another assignment". The RFI technicians worked on the line in the stretch between Livraga and Ospedaletto from midnight to 5. Thirty five minutes later, the passage of the train and the crash. During the works the exchange was "isolated" from the Bologna remote control center and operated by the console of the Livraga maneuver center,



The maneuvers
During their intervention (of «ordinary cyclical maintenance»), the technicians worked on mechanical and electrical parts of the turnout. Basically they opened the hydraulic and mechanical systems that are fundamental for operation. The exchange would have been handled several times both from the console and through the use of special levers that allow you to move the tracks manually without electricity. The repair, however, had not been sufficient and for this reason the technicians had decided - as per procedure - to keep the exchange "isolated" from the security system. Without realizing, however, that the turnout had remained "open" and that it would have mortally diverted the run of the Frecciarossa.
 

Chris M

Member
Joined
4 Feb 2012
Messages
1,057
Location
London E14
"Suspects" is probably a poor translation here. The term as used in a civil law jurisdiction like Italy doesn't have the same connotations as it does in a common law system like the UK, it means their actions are under investigation as a potential factor not that there is, necessarily, a reason to believe they have acted improperly or negligently.
 

SamYeager

Member
Joined
20 Mar 2014
Messages
337
"Suspects" is probably a poor translation here. The term as used in a civil law jurisdiction like Italy doesn't have the same connotations as it does in a common law system like the UK, it means their actions are under investigation as a potential factor not that there is, necessarily, a reason to believe they have acted improperly or negligently.
Perhaps "Persons of interest" is a more appropriate term in that case?
 

Railperf

Established Member
Joined
30 Oct 2017
Messages
2,922
Latest update in brief:

The five railway workers were questioned extensively by investigators to ascertain exactly what work was done to the points just hours before the fatal crash. The workers continue to claim that the points were left correctly set for the main line - and they cannot explain why the aftermath of the accidents shows the points set for the loop.

Investigators will today commence a thorough examination of the tracks and safety systems in order to verify exactly what happened. Hard discs from nearby cameras, and the trains black boxes have been retrieved.

The funerals of the two dead train drivers take place on Tuesday and Wednesday.
 

Chris M

Member
Joined
4 Feb 2012
Messages
1,057
Location
London E14
Perhaps "Persons of interest" is a more appropriate term in that case?
Maybe, but that sometimes also has connotations of it being suspected they have done something wrong. "Material witness" is best I can think of, as that is closer to being neutral about suspicion of guilt.
 

LNW-GW Joint

Veteran Member
Joined
22 Feb 2011
Messages
19,555
Location
Mold, Clwyd

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
Somehow the interlocking was fooled into thinking the straight route was set in order that the ETCS equivalent of an unrestricted proceed aspect was given to the train. Assuming the turnout was correctly 'called' (commanded) to the required position by the interlocking, either the detection limit switches themselves were incorrectly wired or maybe some inadequately controlled temporary test panel or strapping was left in circuit bypassing the contacts, like at Waterloo (luckily a very low-speed incident).
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
Somehow the interlocking was fooled into thinking the straight route was set in order that the ETCS equivalent of an unrestricted proceed aspect was given to the train. Assuming the turnout was correctly 'called' (commanded) to the required position by the interlocking, either the detection limit switches themselves were incorrectly wired or maybe some inadequately controlled temporary test panel or strapping was left in circuit bypassing the contacts, like at Waterloo (luckily a very low-speed incident).
Sounds very much as if it was something like this. As well as Waterloo we had something that narrowly averted being a disaster in 2009: https://www.gov.uk/raib-reports/incident-at-greenhill-upper-junction-near-falkirk
On 22 March 2009 as a passenger train traversed Greenhill Upper Junction, one of the trailing point ends, forming a set of switch diamonds, was in the incorrect position as the train trailed through them. It forced the switch blades to the correct position causing damage to the mechanism. The train was not derailed and continued normally to its destination. There were no injuries caused to either train crew or passengers, and there was no damage to the train.

The immediate cause of the incident was that the points were detected in the reverse position by the signalling system although the point ends were in the normal position. This allowed the protecting signal to clear to a proceed aspect for train 2N50 which subsequently trailed through the point ends.
Hitting a wrongly set switch diamond at over 60mph in the facing direction doesn't bear thinking about.
 

Railperf

Established Member
Joined
30 Oct 2017
Messages
2,922
Today, Ferrovie.info is claiming a wiring fault has been discovered in the turnout actuator - the part fitted hours before the crash. They are using the expression inverted wiring, which may suggest the actuator was working in reverse. I.E the actuator would would send the turnout in the opposite direction to where is should be.
The questions that still require answering are:

Did the workers simply rely on electronic actuation..and did no one physically check the points were in the correct position?
And why did no other safety sensors independently warn that the turnout was set to the loop and therefore causing the signalling system to impose a lower speed restriction through the ETCS in cab signalling?
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
Today, Ferrovie.info is claiming a wiring fault has been discovered in the turnout actuator - the part fitted hours before the crash. They are using the expression inverted wiring, which may suggest the actuator was working in reverse. I.E the actuator would would send the turnout in the opposite direction to where is should be.
The questions that still require answering are:

Did the workers simply rely on electronic actuation..and did no one physically check the points were in the correct position?
And why did no other safety sensors independently warn that the turnout was set to the loop and therefore causing the signalling system to impose a lower speed restriction through the ETCS in cab signalling?
With relatively short turnouts like the one concerned leading to a loop, often the only detection limit switch elements are actually contained within the actuating machine. Longer, very high-speed turnouts tend to have multiple detection elements distributed along the much longer moving switch rails to make sure the whole thing swings together properly, so that would normally catch such a 'non-correspondence' of commanded and detected position but clearly if the only limit switches are in the machine itself and the machine is wrongly wired internally or incorrectly connected externally, then there's nothing else to detect that error. This HAS happened in the UK, luckily never with such tragic consequences, and maintenance instructions are very clear today that the final activity for technicians maintaining or fixing points EVERY TIME is to motor the points back and forth and confirm verbally with the signaller that they are correctly indicating in the position expected at the box. As you imply, this is starting to look like some tragic human error in that final hand-back process.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
Switch machines often include detection, but as far as I know it's always connected to the blades by a different linkage from the one that does the actuation, as a common connection could fall off and lead to false detection. The actuation and detection circuits would have to be separate within the machine, as they do different things. So if one of those was mis-wired then the other one would have to be as well, otherwise the turnout would always be "out of correspondence".

The Greenhill Junction incident I referred to above was due to a problem at an early stage in the design, which was copied into the circuits themselves but also into the instructions used to test them. So when they were tested according to those instructions, they passed, but they were in fact working "back to front".
 

longhorn

Member
Joined
24 May 2018
Messages
40
Impressive how the rest of the train came to rest, relatively in line. Most derailments in America end up with the rail cars crashing with an accordion type effect.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
Switch machines often include detection, but as far as I know it's always connected to the blades by a different linkage from the one that does the actuation, as a common connection could fall off and lead to false detection. The actuation and detection circuits would have to be separate within the machine, as they do different things. So if one of those was mis-wired then the other one would have to be as well, otherwise the turnout would always be "out of correspondence".

The Greenhill Junction incident I referred to above was due to a problem at an early stage in the design, which was copied into the circuits themselves but also into the instructions used to test them. So when they were tested according to those instructions, they passed, but they were in fact working "back to front".

While separate rods for detection is a long-standing British requirement, dating back to mechanical practice, I don't think it has always applied to point machines for other continental European railway administrations. I agree both drive and detection would need to be mis-wired to be in correspondence. British point systems can be configured for LHSNC (left hand switch normally closed) or RHSNC (right hand...). If a complete new system was being installed in the Italian example it's possible the team chose or were given the wrong configuration to install. Techs may not have carried out adequate checks at the end of their work before handing the asset back into use & leaving site. In UK I'd expect them to ask the signaller to motor the points back & forth, verbally confirming observed positions match signal box indications. It's also possible a signaller or Automatic Route Setting system, seeing the turnout apparently working & available for use, set the route without realising the points had not yet been formally handed back. Final tests by techs would have revealed the problem, but signallers may have forgotten reminders and log entries and jumped the gun. A situation compounded by a shift change perhaps? Maybe the techs were having a difficult time with the equipment and were running out of time to complete and perhaps felt under some pressure. Was it new equipment they were not very familiar with? lots of questions no doubt being asked...
 
Status
Not open for further replies.

Top