Stations in Scotland and many in England and Wales will fall silent for a minute on Wednesday at 09.43, one week after the derailment south west of Stonehaven in which two members of ScotRail’s staff and a passenger tragically lost their lives. The time has been chosen to mark the moment that the incident was reported. But this train was the 06.38 from Aberdeen to Glasgow Queen Street, which left punctually and was apparently still on time at Stonehaven (booked departure 06.54). The long period of time between then and when the alarm was raised is just one of the mysteries surrounding this incident, as Sim Harris reports.

RARELY has there been a railway accident in which so many vital details have remained vague for days after the crash.

The key facts about accidents, in the relatively recent past at least, have usually been clarified almost at once – trains involved, time of accident, immediate consequences, and so on.

Of course the complete picture would take railway inspectors months of intricate enquiries to assemble, but it was known very quickly after the double collision near Clapham Junction in December 1988 which trains were involved and when the accident had happened. Even the cause was suggested with amazing speed, when British Railways Board chairman Sir Bob Reid was questioned by reporters a few hours later.

The traditional answer in these circumstances had always been a ‘holding’ one: ‘This will be a matter for the inquiry to determine … no one will be helped by speculation at such an early stage …’ (etc.), but Sir Bob had already been briefed sufficiently to allow him to break with tradition and say frankly that it looked like ‘a problem with the signals’. He was, of course, right.

Winding forward almost 32 years, the first days after the Stonehaven derailment were notable for a lack of clarity about, well, almost everything.

The site of the crash was a remote one, and the best way of describing it seemed to be ‘near Stonehaven’. In fact, it was about 2.25km short of the next signal box and former station at Carmont. Stonehaven is more than 6km to the north-east.

The emergency services had been called at 09.43, but where was the train heading at the time?

Some reports said it had been travelling south-west from Stonehaven when it encountered flooding, and that it had been ‘switched to an alternative line’. This theory gained ground when Network Rail published a video on Twitter showing what appeared to be the Glasgow-bound up line almost out of sight in murky brown floodwater, while the parallel down road was dry.

But crossing to the down for the purpose of working ‘wrong line’ would have first meant reversing back to Stonehaven (again, working wrong line on the up line) where there is a crossover.

Assuming that this had been done, the train was now presumably using the down line when it came to grief at Carmont. The ‘wrong line’ school of thought was further supported by another report that both the driver and conductor had died in the leading cab.

This would have been in accordance with the rules, which require a ‘second man’ to accompany a driver who is working wrong line. But the speed of the train, which seems to have been quite high, to judge by the wreckage, would surely have been too much for wrong line working, which is usually done at caution (say, 20mph/32km/h at most).

We now know better. In spite of appearances to the contrary, there was no wrong line working and the train did not reverse back to Stonehaven on the up line.

Thanks to a briefing from the RAIB which was published on Friday afternoon, it seems that the train was travelling on the up line (and had already passed what would be the crash site as well as Carmont itself), when the Carmont signaller advised the driver by radio that a landslip had been reported further ahead, between there and the next station at Laurencekirk.

The train was stopped, and used another crossover at Carmont to join the down line with the intention of returning to Aberdeen. It was on the first leg of this return journey between Carmont and Stonehaven that disaster struck, when the train collided with another landslip and became derailed.

Even now there are some pieces missing.

Media reports speculated that it took quite a while (more than a couple of hours, seemingly) before the alarm could be raised, because the crash site is in a communications blackspot.

But there is something odd about this too. The signaller at Stonehaven would surely have received ‘train entering section’ from Carmont by 08.30 at the latest (even allowing generously for the time needed to set back to Carmont, talk to Control and carry out the reversing move on to the down line).

More than an hour would have gone by, and yet the train was still seemingly somewhere between Carmont and Stonehaven (as indeed it was). Surely the signallers at either end would have asked questions before 09.43?

To add to the confusion, some media reports have claimed that an off-duty member of staff managed to escape from the wreck and, in the best railway tradition, then set off to tramp the 2250m to Carmont signal box with the very proper intention of protecting the line and the train.

This might account for some of the time which elapsed between the accident and the calling of the emergency services, except that the RAIB says the third staff member on the train, who was indeed off-duty, was taken to hospital with the other surviving passengers. No mention of a walk down the line to Carmont.

Although the RAIB has clarified many vital details, it has unfortunately added a new layer of uncertainty by also saying that the derailment occurred ‘at around 09.40’.

No one could have called the emergency services just three minutes later, unless the crash site is not such a communications blackspot after all or, perhaps, there was a convenient signal telephone nearby. It is also possible that the wreckage fouled track circuits. Certainly, a track circuit flicking unexpectedly to ‘occupied’ has indicated trouble on previous occasions.

But if the derailment really took place at 09.40, where had the train been for the last couple of hours? And why was it travelling fairly fast in an area where landslips had been reported and parts of the railway were submerged in floodwater?

The RAIB has provided some clarification, which is welcome. But there must be more to come.

The August print edition of Railnews, RN282, was published on 30 July. The new edition and some previous issues can be obtained by calling 01438 281200 from UK numbers or +44 1438 281200 internationally, and selecting Option 2.