Accident at Parkhall 1918

In his book ‘Obstruction Danger’ (1989) Adrian Vaughan describes the accident which occurred on a single line at Parkhall, near Oswestry, in 1918. Two goods trains met head-on at a closing speed of 35 mph. Both crews had a valid tablet for the section.

The Inspecting Officer reported that it was the most difficult case he had ever had to deal with. He could find no explanation for the accident but suggested the signalmen were not working in accordance with the regulations.

Does any member of the forum know anything more about this event – or have a theory of how it occurred?

Jimbob52 said:

In his book ‘Obstruction Danger’ (1989) Adrian Vaughan describes the accident which occurred on a single line at Parkhall, near Oswestry, in 1918. Two goods trains met head-on at a closing speed of 35 mph. Both crews had a valid tablet for the section.

The Inspecting Officer reported that it was the most difficult case he had ever had to deal with. He could find no explanation for the accident but suggested the signalmen were not working in accordance with the regulations.

Does any member of the forum know anything more about this event – or have a theory of how it occurred?

Click to expand...

There is an extract from the accident investigation report here: http://www.railwaysarchive.co.uk/docsummary.php?docID=1482

Not sure if that helps?

A look at the full accident report http://www.railwaysarchive.co.uk/documents/BoT_Parkhall1918.pdf shows that there was significant disagreement between the two signalmen over their actions. Either there was an temporary electrical fault - rather unlikely - which allowed tablets to be withdrawn from both ends of the line, or perhaps one signalman had unofficial access to his machine's interior and was therefore able to extract a tablet when he shouldn't have! (These single line tablet machines had locked side panels for which the S&T technicians were supposed to hold the only keys.) The report does draw attention to a specialist's examination suggesting a polarising magnet was rather weak which might have led to a wrong actuation of a tablet release, but the Inspector was still unable to reconcile this with the differences in accounts between the two signalmen.
It may be that snow had brought two wires into contact which allowed irregular operation, or which prevented proper operation and the signalmen had resorted to improper working and then lied about it. Who knows?

There's a common belief among those who are not familiar with token working that there is only one token for each section, and thus some disbelief when you tell them that there were commonly 20 to each section, and then describe why, and the interlocking arrangements. Given the standard of remote electrical controls of more than a century ago, it's surprising this didn't happen more often for a variety of reasons. By no means all were interlocked with the starting signal, right to the end of some lines. I wonder in this case what the signalbox registers showed. Vaughan was an ex-signalman and commonly has a good opinion on matters, so what were his conclusions.

In Somerset the GWR tokens were plain gunmetal, but the LSWR lines had substantial brass tablets, and it was not uncommon on a quiet Sunday morning to get them all out and give them a good Brasso. I've written about this previously here and some were appalled at this. The "keys" to these token machines were pretty simplistic and the signalbox fireirons turned the lock just as well. Have a read of this thread from a while back.

https://www.railforums.co.uk/threads/token-operated-signalling-problems.109546/#post-2009952

Taunton said:

……. I wonder in this case what the signalbox registers showed. Vaughan was an ex-signalman and commonly has a good opinion on matters, so what were his conclusions...….

Click to expand...

The signal box Train Registers were incomplete from the official report and the two signalmen disagreed over what each had done. Adrian Vaughan could only say that the secret of what happened that morning went to the grave with the two signalmen. (Their employment history after the accident is not recorded in the official report or, it seems, anywhere else.)

In his book O S Nock who worked for Westinghouse said that they had to make their machines tamper proof. He gave an example from Africa of a Stationmaster upset that another train had got right of way, broke open the tablet machine and gave the tablet to the driver leading to a head on collision.

Apart from Parkhall and Abermule, as far as I know there were no other collisions on lines in the British Isles worked by Tyers electric token machines. Which is a pretty impressive safety record.

Pure speculation, but my inclination is that the accident arose from both signalmen implementing an unauthorised method of working to overcome a temporary failure of the Ellesmere tablet instruments (possibly due to snow, exhasperated by the poor design features highlighted later).

After the failure at Ellesmere, Evans and Edwards perhaps agree that they’ll continue running traffic via telephone, as Evans perhaps has a means of extracting tablets from the machine independently. It’s wartime and they’ll both be reticent to prevent goods traffic moving freely

Evans does so for the no.9 goods. Let’s say he fabricates entries for this into the register, documentating an exchange between him and Edwards that never took place via the tablet instruments, but instead by phone. Edwards leaves his register blank in order to fill in the full details later based on Evan’s entries and once all movements of the train are completed. As long as an inspector doesn’t turn up before this can take place, they’ll be covered.

In the intervening minutes, Edwards loses track of the agreement and reverts to regular working. Perhaps he’s distracted by other events or just tired at 1am; the temporary issues with the Ellesmere instruments have rectified themselves. He asks Evans for a tablet release. Surprised by this, given seemily insufficient time has elapsed for the train to arrive, Evans has to telephone Edwards to confirm if the casually signalled goods has indeed arrived - obvious misunderstanding here as Edwards loses track of which train is being referred to, forgetting one is already in the section.

Once the accident occurs, both signalmen cover their actions by denying all knowledge, given they’re both compromised.

Given that the only fully recorded and agreed token was the one issued by Edwards, I think it points to Evans having had a greater role in whatever incorrect working led to the issuing of a token to the first goods at Ellesmere. Edwards had no reason not to enter corresponding entries if that train was signalled via regular means; it all points to irregularities on that train/tablet issue in particular.

That’s my take anyway. We’ll never know for sure, but it sure is a fascinating incident!

AW has offered a plausible explanation for the accident but I have my doubts. I am not convinced that the motivation for the irregular way of working was to keep the traffic moving. There would have been strict rules in place on how to operate the section if the electric interlocking had failed because of snow and no incentive not to follow them.

I suggest a more likely explanation is that the signalmen had worked out a way to circumvent the interlocking so that one of them could sleep on duty.

After the crash, Edwards ran to the site and asked the driver if he had the right tablet, strange behaviour if he had in fact handed the tablet over in the correct manner.

The accident happened at 2.00 a.m. when the section was seven and a half miles long because an intermediate box had been switched out. A slow goods train would take about 20 minutes to get through the block, say half an hour from when the ‘is line clear?’ was requested. By devising an unauthorised method of handing over the tablet the signalman could sleep for this time plus the interval until the next train was due. But this relies on the co-operation of the footplate crews to collect the tablet in some unauthorised way and there is no suggestion in the report that this was happening.

With hindsight it is a pity the official report did not give more information on the interval between trains or question the drivers on whether they were aware the section was being worked in an irregular manner at night.

As AW has said, this is a fascinating accident. I hope that a member of the forum with detailed knowledge of single line working will offer further thoughts on what may have taken place. In the meantime, I agree with AW’s conclusion: we’ll never know for sure.

Jimbob52 said:

There would have been strict rules in place on how to operate the section if the electric interlocking had failed because of snow and no incentive not to follow them.

Click to expand...

I think there would be every incentive given some of the rules in place in case of failure. David L Smith's books of the G&SW give some detail of this. One was that someone had to walk the length of the section, maybe with a tablet, to ensure it was clear. Here was the middle of the night, in a snowstorm, with older men (all the younger ones being off at the war), Ha Ha indeed.

MP33 said:

In his book O S Nock who worked for Westinghouse said that they had to make their machines tamper proof.

Click to expand...

I've written before about how in Somerset the signalmen could readily get all the tablets out, maybe for such a trivial event as just to polish up the brass tablets on a Sunday morning. Apparently the stove poker fitted the Tyer's tablet machine lock quite straightforwardly. Never an accident.

I've just read the accident report, and noticed that the machine in use at the time of the collision wasn't actually a Tyers but some sort of copy by another manufacturer - does that have any bearing does anyone think?

Fleetwood Boy said:

I've just read the accident report, and noticed that the machine in use at the time of the collision wasn't actually a Tyers but some sort of copy by another manufacturer - does that have any bearing does anyone think?

Click to expand...

It wasn't just some sort of casual copy, but there were a number of comparable systems by other manufacturers.

The GWR did their own (of course they would). In Victorian times they developed their Electric Train Staff, which looked like a policeman's truncheon, and later devised their key Token, patented, which did so well that they licensed production of it back to Tyer's company to manufacture and sell it to railways other than the GWR, in parallel with Tyer's own circular brass tablets.

Am I correct that the GWR key Token system was the first to interlock the token machine with the starting signal, a further safeguard?

There's an article about the different types here

http://www.trainweb.org/railwest/gen/signal/sl-intro.html

Everyone wants to say "Abermule" in such discussions, suffice to say that I believe the Cambrian was still using Tyer's, not interlocked. The accident happened very shortly before the GWR absorbed the Cambrian, and they very quickly replaced the equipment with their own. The hulks of the two severely damaged locomotives were at the back of Oswestry works at the grouping, and were taken into the GWR accounts at their full book value, to the benefit of the Cambrian shareholders, but were never repaired and were broken up for scrap shortly afterwards.

matchmaker said:

Apart from Parkhall and Abermule, as far as I know there were no other collisions on lines in the British Isles worked by Tyers electric token machines. Which is a pretty impressive safety record.

Click to expand...

There was a fatal serious head-on on the Longmoor Military Railway in 1956. Although not the national system, the whole purpose of the line was a training for mainstream railway operation, and I would expect they would have full single line controls. It was discussed afterwards in Parliament although the system in use is not detailed. See Hansard here

https://api.parliament.uk/historic-...956/oct/30/longmoor-military-railway-accident