Spanish Rail Crash

[Glorious NSE]

With the Spanish crash, surely the second driver should have warned the man in the chair he was going too fast?

 

From comments elsewhere, there were 2 drivers on board, only one was allowed to be in the cab, the second driver on the train was back in the coaches...

[Colin_McLeod]

What was the point in that?  My recollection of the HSTs is that both drivers are in the cab.

[phil-b259]

On Potters Bar, I totally agree with your last sentence Mike. Well Put.

 

On Ladbroke Grove, I never got a proper explanation of why the last set of facing points the Thames train went through were set towards the Up Main rather than the Down Relief. If set for the Dwon Relief, there would just have been a st of trailing points run through on that morning. The worst case scenario would have been a side swipe with a train going in the same direction, not a head-on crash with a train doing 100 MPH in the opposite direction as did happen. My personal opinion is that the signalling fraternity bamboozled the Inquiry with, to my mind, spurious excuses about flank protection (which I was later informed was not provided anyway) and the bloke who did much of the design work having passed away in the meantime so the design logic was untraceable. The other excuse used was used that the work was supposed to incorporate ATP (or whatever it was called at the time) but that was never installed.

 

The explanation was very simple.

 

The signalling at that point was fairly new having  been renewed as part of the Heathrow Express upgrades. What occurred was that when the signalling designer was writing the software for the SSI interlocking he neglected to add the requirement that, unless required for a route, the default position of points (and the position they would "self normaise" to) would be towards the down relief rather than the up. There was absolutely nothing wrong with this in design terms and still isn't, signal S109 had in excess of the 200ard overlap before the pointwork and therefore flank protection is not required (There are plenty of locations on the railway network where this is the case today especially with the trend for single lead crossovers). Moreover signalling engineers have a right to expect that drivers will obey the instructions given by the signals such that they will be able to stop within the overlap provided, it is after all a fairly fundamental part of driving. While ATP would certainly have been advantageous it certainly wasn't the case that they layout was deficient without it according to modern signalling practice.

[phil-b259]

What was the point in that?  My recollection of the HSTs is that both drivers are in the cab.

 

Question is, were they both on duty at the time? Just because there may be two drivers on a train doesn't mean both were scheduled to be in the cab at that point.

[Glorious NSE]

What was the point in that?  My recollection of the HSTs is that both drivers are in the cab.

 

HSTs are single-manned these days...

[Edwin_m]

The explanation was very simple.

 

The signalling at that point was fairly new having  been renewed as part of the Heathrow Express upgrades. What occurred was that when the signalling designer was writing the software for the SSI interlocking he neglected to add the requirement that, unless required for a route, the default position of points (and the position they would "self normaise" to) would be towards the down relief rather than the up. There was absolutely nothing wrong with this in design terms and still isn't, signal S109 had in excess of the 200ard overlap before the pointwork and therefore flank protection is not required (There are plenty of locations on the railway network where this is the case today especially with the trend for single lead crossovers). Moreover signalling engineers have a right to expect that drivers will obey the instructions given by the signals such that they will be able to stop within the overlap provided, it is after all a fairly fundamental part of driving. While ATP would certainly have been advantageous it certainly wasn't the case that they layout was deficient without it according to modern signalling practice.

I seem to recall that although the setting of the points beyond SN109 as they were brought the train onto a high speed line in the wrong direction, there was a lot more distance available to realise the mistake and stop than if the point had been set the other way where the convergence with the adjacent line was almost immediate.  So it would be a trade-off against a greater likelihood of a less severe (but still bad) accident. 

 

The decision on the default position of points should have been made when producing the control tables rather than at the SSI data stage - or are are suggesting that it was in the tables but omitted from the data? 

[beast66606]

I seem to recall that although the setting of the points beyond SN109 as they were brought the train onto a high speed line in the wrong direction, there was a lot more distance available to realise the mistake and stop than if the point had been set the other way where the convergence with the adjacent line was almost immediate.  So it would be a trade-off against a greater likelihood of a less severe (but still bad) accident. 

 

The decision on the default position of points should have been made when producing the control tables rather than at the SSI data stage - or are are suggesting that it was in the tables but omitted from the data? 

 

That's my understanding too - damned if you do and also damned if you do the other.

 

Perhaps a phraseology confusion.

[The Stationmaster]

I seem to recall that although the setting of the points beyond SN109 as they were brought the train onto a high speed line in the wrong direction, there was a lot more distance available to realise the mistake and stop than if the point had been set the other way where the convergence with the adjacent line was almost immediate.  So it would be a trade-off against a greater likelihood of a less severe (but still bad) accident. 

 

The decision on the default position of points should have been made when producing the control tables rather than at the SSI data stage - or are are suggesting that it was in the tables but omitted from the data? 

I suspect the default position of the points was related to other possible moves as they worked as one end of a crossover.  the ideal answer would have been wide-to-gauge trapping but because it was a fast set of crossovers the switches were very long and I suspect incorporating w-t-g would have been a step too far.  

 

But on the other side of the coin SN109 had a safe overrun margin of around half a mile, longer than any Clearing Point or Overlap ever thought of in British signalling practice including that on single lines.  Added to that the ability for a Driver (who knew the road properly) to know where he was was probably as good as it gets - from SN109 you could only go left (first turnout) to the Down Main or right (second turnout) towards the Down Relief, there was no route straight ahead although there was obviously track there.  So half a mile to realise you've got it wrong, easy landmarks to pick out where you would go if you'd got it right, and a nice brick wall on the left to count the number of tracks from if you weren't sure of your position - quite honestly even the addition of numbers at least 8 feet tall hasn't made most of the running lines in that vicinity mush more easily identified than they already were.

 

But once again we're back - as in Spain to relying on the man in the driving seat understanding and realising all those things and alas to me it's plainly obvious that the Thames Train Driver hadn't got a clue where he was (unlike an FGW Driver who'd previously passed SN109 at danger and then realised from the landmarks that he had done so, and stopped).  So in an ideal world the machine needs to recognise the fallibility of the man - and at Ladbroke Grove at that time it didn't - the world (but perhaps not in Spain) has improved since then, I think.

[phil-b259]

I seem to recall that although the setting of the points beyond SN109 as they were brought the train onto a high speed line in the wrong direction, there was a lot more distance available to realise the mistake and stop than if the point had been set the other way where the convergence with the adjacent line was almost immediate.  So it would be a trade-off against a greater likelihood of a less severe (but still bad) accident. 

 

The decision on the default position of points should have been made when producing the control tables rather than at the SSI data stage - or are are suggesting that it was in the tables but omitted from the data? 

 

IIRC the control tables were checked with the SSI data and both were found to be correct to each other but as I pointed out earlier there was no need to consider the lie of the points - the layout and the signalling control tables meet all the requirements in force then and they haven't changed much to this day.

 

As to the location of the points, the convergence with the down line would still have been for 200 yards past signal SN109 otherwise the points concerned would have had to be written into the control tables. Also if what you say about the distance till the convergence on the up is correct and the points were within the 200 yard overlap of SN109 then (ignoring things like restricted overlap routes) the train would still have to have been routed towards the up line when the signal was at danger due to the flank protection requirements that apply with regards to overlaps.

[martin_wynne]

So half a mile to realise you've got it wrong

 

Is there any place in modern signalling for the detonator-placer?

[Dunsignalling]

I was always under the impression that the cabin crew's first responsibility was passenger safety. Using them as "trolley dollies" gives them a secondary role whilst the aircraft is airborne which enhances the passenger experience. IMHO the budget airlines would have done away with them as an unnecessary expense long ago if they could.

Never mind cabin staff, some of them would like to do away with the co-pilot!

 

Pay-as-you-Enter Airliners anyone?

 

John

[Oldddudders]

Is there any place in modern signalling for the detonator-placer?

Almost certainly - but where overlaps are very sub-standard, so the driver must be alerted the instant he runs by. Borough Market Junction and environs have traditionally had them, maybe still do, as the overlaps are tiny, but of course linespeed is very low, too. SN109 appears from the evidence above to have had a very long and therefore presumed safe overlap - in common with many thousands of other signals.

 

I'm sure ASLEF would say that "boil-in-the-bag" drivers cannot be expected to be competent.

[The Stationmaster]

Is there any place in modern signalling for the detonator-placer?

The problem with many modern driving cabs is hearing dets - that was a problem with diesels back in the late 1960s and it's probably a case of even more so nowadays (in fact you are probably as likely to feel them as hear them judging by what I've come across riding in diesel loco cabs or as a passenger on a 'modern' dmu.  the second problem is renewing them - multishot placer still have to be reloaded and the tend to place only a single det (if they can still be obtained?).

 

The alternatives of the likes of AWS, TPWS or full ATP don't need much attention at site once they're in and tested,  the latest kit doesn't involve contact with the train so there's no wear although presumably components will still age and eventually require replacement.  And they create a positive action in many cases - if the Driver takes no notice of the message the machine takes over and applies the brakes and/or cuts off the power.  Not at all like the old story of the Fireman asking the Driver 'what was that bang mate?' and receiving a 'dunno' in reply - only to be followed by a somewhat bigger bang.

['CHARD]

Question is, were they both on duty at the time? Just because there may be two drivers on a train doesn't mean both were scheduled to be in the cab at that point.

 

I read that the non-driving 'driver' was in coach 7, which suggests to me that he was travelling on the cushions, and quite possibly off-duty. 

 

Assuming he was off duty, he has every right to be dozing off or whatever, certainly not trying to calculate actual vs line-speed at any given moment.

 

The 'two drivers' detail seems to me to be another bit of sensationalist - and selective, cut'n'paste - reportage.

[Ozexpatriate]

For the media no news is bad news. There is also today's "Blame" culture which seems to have spread across the world - if there's an accident, it must be somebody's fault. So the news media goes looking for someone to blame and can declare someone guilty even before any proper investigation is under way.

People are still rehashing the TWA 800 disaster and perpetuating conspiracy theories to find *someone* to blame.

[TheSignalEngineer]

I suspect the default position of the points was related to other possible moves as they worked as one end of a crossover.  

Agreed. Lie of points beyond the overlap in layouts like LG is always a matter of debate. If 8059 crossover had been single ended with the B end lying the other way and a legitimate move taking place along the Down Relief there could have been a side-on collision instead but the wreckage from that would have ended up on both the Up Main and Up Relief, so as Beast said you are shafted whatever you decide.

[Ozexpatriate]

I read once that passengers in any vehicle are actually safer if they are facing backwards. If that became the norm, I wouldn't travel anywhere! Can't stand going backwards in a train and usually feel really sick within a couple of minutes.

It is common to have backward facing seats in military aircraft - particularly those that land on aircraft carriers - that way the seat absorbs the landing shock - rather than the seatbelt. (I believe the seats of the Grumman C2 Greyhound face the unloading ramp in the rear.)

[Ozexpatriate]

If you think that was "relatively gently", I dread to think what you consider a "hard" landing!

Look at the video.

After hitting the wall (hard) and ripping the tail section off it slid/bounced on the runway, cartwheeled over and bits broke off

That is incorrect the Asiana 777 did not cartwheel, the undercarriage was ripped from the fuselage and the tail hit the wall during the commencement of a go around.

Eyewitnesses described the airplane as "cartwheeling". They were wrong. They used the wrong word.

 

The right undercarriage hit the rip-rap wall, followed by the empennage (tail plane and horizontal stabilizers) which came off. Most of the time the airframe slid diagonally down the runway on its belly

 

After sliding off the runway and hitting the dirt, the airplane spun (approximately around its nose) with the damaged tail end in the air. The last motion of the airframe is the rear crashing down hard.

 

The engines came off (as designed). The wings remained intact. Most of the damage (before the fire) to the fuselage was subfloor crushing damage underneath a virtual plane from the rear wing root to the bottom half of the aft pressure bulkhead.

[Fat Controller]

It is common to have backward facing seats in military aircraft - particularly those that land on aircraft carriers - that way the seat absorbs the landing shock - rather than the seatbelt. (I believe the seats here face the unloading ramp in the rear.)

My dad's brother-in-laws (from his first marriage) were two of the three survivors from the Llandow air crash ( http://en.wikipedia.org/wiki/Llandow_air_disaster), as they occupied the two rear-facing seats at the back of the plane (the other survivor was in the toilet). Had my father accepted the invitation to go to the match, they'd have sat elsewhere, and I wouldn't be typing this..

[Oldddudders]

My dad's brother-in-laws (from his first marriage) were two of the three survivors from the Llandow air crash ( http://en.wikipedia.org/wiki/Llandow_air_disaster), as they occupied the two rear-facing seats at the back of the plane (the other survivor was in the toilet). Had my father accepted the invitation to go to the match, they'd have sat elsewhere, and I wouldn't be typing this..

 "the worst disaster in the history of aviation". I'm sorry to say we've come a long way since then.

[Edwin_m]

Thanks to all for the further useful information on Ladbroke Grove. 

 

The TPWS train stop does roughly the job of a detonator placer - applies the brakes if a danger signal is passed, assuming the signal has TPWS.  If TPWS had been fitted at Ladbroke Grove at the time it would have stopped the Thames unit well before any possible collision. 

[RJS1977]

Thanks to all for the further useful information on Ladbroke Grove. 

 

The TPWS train stop does roughly the job of a detonator placer - applies the brakes if a danger signal is passed, assuming the signal has TPWS.  If TPWS had been fitted at Ladbroke Grove at the time it would have stopped the Thames unit well before any possible collision. 

 

Though it was my understanding that AWS (which was installed) also stops a train if it passes a danger signal (though I know TPWS also has speed limiters built in). As the driver evidently over-rode his AWS, there's no reason to think he wouldn't have done the same with TPWS (or can't TPWS be over-ridden?).

[Glorious NSE]

Though it was my understanding that AWS (which was installed) also stops a train if it passes a danger signal (though I know TPWS also has speed limiters built in). As the driver evidently over-rode his AWS, there's no reason to think he wouldn't have done the same with TPWS (or can't TPWS be over-ridden?).

 

AWS gives an audible warning, which the driver has to acknowledge, it doesn't take control unless the driver does not acknowledge - so in this case the driver presumably acknowledged the noise (as he had for the previous two caution signals) and carried on...

 

TPWS is a little more intrusive and will actively intervene when it thinks you're likely to pass a signal at red, I don't *think* you can override at TPWS intervention at speed?

[Oldddudders]

AWS gives an audible warning, which the driver has to acknowledge, it doesn't take control unless the driver does not acknowledge - so in this case the driver presumably acknowledged the noise (as he had for the previous two caution signals) and carried on...

 

 

And therein lies the problem - in busy suburban areas the driver can be hitting the "acknowledge" button every signal, because he's running on single or double yellows all the time. It must become a reflex - and then  you acknowledge one too many....

[stadman]

It is common to have backward facing seats in military aircraft - particularly those that land on aircraft carriers - that way the seat absorbs the landing shock - rather than the seatbelt. (I believe the seats of the Grumman C2 Greyhound face the unloading ramp in the rear.)

Been on  couple 737-200's which had rear facing seats, only couple of rows behind internal bulkheads.