Return to Archive -by date - by topic.
As the first anniversary of the Potters Bar derailment approaches it remains an unsolved whodunit
Whatever this column has said about the Health & Safety Executive in the past, its press team have always been helpful beyond the call of duty in the face of extreme provocation. But even they were unable to provide a copy of letter sent by the Investigation Board to those bereaved and injured in the Potters Bar derailment last December.
We know about the letter because of quotations published by one of the recipients in a newspaper article. It said:
‘While there remain some further avenues to pursue I can tell you , from the point of view of the technical investigation so far, that the (Investigation) Board was informed that there is a valid engineering explanation for the failure of Points 2182 A. and this has enabled several possible scenarios to be established independent of any reliance on sabotage'.
Well, we all knew that. As explained in detail in this column in August 2002, it was quite clear that the nuts on the right hand end of the second tie bar (Tiebar 2) had been screwed outward. This would move the right hand switch rail towards the stock rail, closing the flangeway when the points were set straight ahead.
In addition the nuts had been removed from the outside of the right hand end of Tiebar 2 and the left hand end of Tiebar 1. This meant that the only thing holding the two switch rails together was the lock stretcher.
Most likely, the narrowed flangeway meant that passing wheels hit the switch blade and the high frequency force caused the lock stretcher to fail from fatigue, allowing the right hand switch rail to move across to the stock rail, under the train.
So the Inquiry Board's comment is quite logical. We could all conceive a scenario where nuts could be removed or screwed out. An adjustment terminated by the end of a possession and left for next time, an incompetent workman turning nuts the wrong way. And so on.
But then the letter added: ‘The Board took advice from its legal adviser about issuing a definitive statement about sabotage…and came to the view that it had to respect that advice and not issue a statement on the basis of the evidence currently available'.
Now this seems eminently sensible advice to me. But you can understand the bereaved and injured seeing it as cautious lawyers stopping the Inquiry Board from ruling out sabotage.
Next consider the actions of the BR Police. As it is clear that the bolts could not have vibrated into the condition which caused the crash, someone must have done something. All of those responsible for working on the points deny having done anything that would have left 2182A in the pre-derailment condition. Impasse.
So, back in August the BT Police, BTP having taken around 850 statements, closed over 3000 lines of inquiry, collected 1600 exhibits and examined 489 documents, set up a confidential ‘hotline'. According to a spokesman, a number of those interviewed with company lawyers in attendance had indicated that they would like to speak confidentially and the hot line would provide the opportunity. As far as I know nothing significant emerged.
In safety terms, what matters is not who did it, but what was done, and how. Engineers from maintenance contractors Jarvis and Railtrack/Network Rail have only been allowed to see, but not touch, points 2182 A in the HSE laboratory. Fortunately, they were free to experiment with the set of replacement points intended for Potters Bar.
A series of tests were carried out to determine how the pre-derailment condition could be achieved. To recapitulate both tie bars had been adjusted to the point where there was not enough thread at one end for the main nut to be screwed on, let alone a second lock nut. This was at the left hand end of Tie bar 1 and the right hand end of Tiebar 2. The nuts at the other ends were tight on the bracket.
Each tie bar is clamped by a pair of nuts (main plus lock nut) on each side of a bracket. It is improbable that both pairs of locked nuts would have vibrated loose along the tie bar in the same direction, leaving insufficient thread for the outer pair of nuts.
On tie bar 2, where the ‘shiny thread' was observed, considerable force would have been required to turn the inner nuts out since this involves bending the switch rail. This was the obvious starting point in replicating the as-found condition.
Now the two Tiebars are connected by the ‘back-drive', the rod and bell cranks which transmit the drive of the point motor through Tie Bar 1 to Tie Bar 2. Winding out the nuts on Tie Bar 2 pushes the switch rail outward and this movement would be transmitted to Tie Bar 1 via the back drive. In other words, the back drive would be pulling the tips of the switch rails across towards the right hand stock rail.
When engineers removed the nuts on the outside of the bracket on the right hand end of tiebar 2 on the replacement points and started winding the inner nuts outwards, they found that well before the nuts had moved the 50mm indicated by the shiny thread, the back drive effect meant that the points would have lost detection when set to the diverging position.
To avoid this you had to isolate the back drive to Tiebar 1, releasing the ‘pull'. This explains what I assumed was an additional escapement – a gap in the drive system intended to provide the point motor with essential play – at the right hand end of Tiebar 1.
At the time, I noted that this ‘escapement' lacked the sleeve round the tiebar end to protect the thread and assumed it was the result of sloppy assembly. But once you appreciate that the back drive transmits motion of the switch rail at Tiebar 2 which affects detection it becomes clear that it might not have been an escapement at all, but a way of releasing pressure on Tiebar 1..
This detection issue could also explain the location of one end of the bar connecting the back-drive bell-cranks. This had been moved in towards the fulcrum, reducing the throw and thus the movement transmitted to Tiebar 1.
This is not theorising, You can see it happen in real time on a video Jarvis made of the replacement points being altered. It starts with the setting of the switch rails being checked with a gauge which showed that when the points were wound over manually, they detected correctly with a 1.5mm gauge between the tip of the switch rail and the stock rail. But with a 3.5mm gauge the points could not be wound over-centre to lock and detect.
As the nuts on Tiebar 2 are wound out and the flangeway pap it gets increasingly difficult to bend the switch rail. Eventually it gets too hard to turn the nuts.
So the man doing the work goes forward to Tiebar 1 and creates the false escapement. After 10 minutes of turning the nuts there is a visible bow in the right hand switch rail and the flangeway is virtually closed.
Remember that the outer nuts on the left hand end of Tiebar 1 were also off. To do this you have to undo all four nuts on the tiebar and move it across before doing up the nuts at the right hand end.
It took another 10 minutes to change the position of Tiebar 1 to as found. Overall it took just under 25minutes sec to get the points into the pre-derailment condition. Network Rail engineers have carried out a similar exercise. It took them nearer 30 minutes.
At one point during the video an experienced engineer said to me ‘I really, really can't get my mind round how it was all done with no loss of detection'. But to conclude the exercise the severely distorted points were pout on power and detected in both positions.
In both positions one of the switch rails was clearly bowed and should have been obvious to anyone walking the track – as was done the day before. It would probably have been noticeable to drivers too.
Interestingly, given the debate over police and HSE access to statements given to the new Rail Accident Investigation Board, the briefing I and several other technical journalists received was repeated at a meeting at the Institution of Civil Engineers convened by past President Mark Whitby last October. A selected group of engineers was briefed by Jarvis staff.
This was a private meeting, aimed at disseminating information on a matter of considerable interest to railway civil engineers. On January 13 Mr Whitby was visited at the Institution by two BTP officers, with evidence bags, who asked him for a statement plus details of the meeting an attendees.
He was not cautioned and was not obliged to give a statement. Afterwards Mr Whitby said that the Institution had a ‘right and a duty' to hold private discussions where members could speak freely ‘particularly where matters of public safety were concerned'.
He added ‘If such conversations were subject to police inquiry, then that would inhibit our right to hold such discussions. But I cannot imagine what information we have that is of the remotest interest to the police at this stage of their inquiry'.
Quite. I believe that the practical demonstrations are vital to the investigation. They have convinced me that Points 2182 were tampered with and by someone who knew what they were doing.
Equally I find it hard to conceive a scenario where the points could have been put into this condition by a maintenance gang. And the detection issue weakens the incompetence theory, as it does vandalism or terrorism.
Which leaves us with Mr Grudge, someone with experience of trackwork who wanted to discredit Jarvis and/or Railtrack, but didn't expect a catastrophic derailment. Well that's what my Occam's razor concludes