Collision and derailment near Salisbury (Fisherton Tunnel) 31/10/21

[Driver0202]

Wow there’s a lot of pretend railway signalling engineers on this feed saying how things should work. Never have I seen a thread with so many myths posted.

 

[muz379]

A good question.

My observations are:
1 The 7 minute gap may not be accurate, this was from the Daily Mail who would want be on the sensationalist side and maybe there just wasn't enough time to protect with dets.

2 The driver assumed it was protected as he/she had contacted the signaller/ pressed the panic button to stop all trains in the area.

3 I think in these circumstances if dets were required then the driver would go forward and protect the opposite line so it would be up to the train manager/guard to go back and protect the rear and he/she would obviously be very shocked or may have been checking tickets at the time and their first thought would be for the passengers. Also what comms are there between driver and guard apart from the intercom to the rear cab?

4 There was a converging junction so which line would they protect first?

A horrendous situation for any train crew to be in and we can say all we like that is what they are trained for but it's a whole different scenario when one minute you're cruising along in a nice warm train then suddenly you're off the road in a tunnel, ballast flying everywhere and got injured passengers screaming etc.

One further thought, do they use track circuit operating clips still? Sometimes a better bet than dets, although maybe not in this case if the protecting signals had gone dark.

Absoloutely this , even if it was 7 minutes that really is not a terribly long time at all .

The first thought of any Driver or Guard in that situation is going to be checking on each other , you need to confer anyway before going out to carry out detonator protection if that is required . And chances are if the unit has split the cab to cab intercom isnt going to be working , so lets say you both start to walk through to the middle not knowing yet that the unit has split . You are going to be faced with the passengers of your train frightened and possibly injured . Not an easy situation to just tell people to sit tight whilst you speak to the driver/guard .

Its easy to sit in a classrrom and rattle off the rules around detanator protection and actions to be taken in the event of a train accident . But in a real world scenario like this you have multiple other things going on at the same time . The majority of traincrew are lucky enough to go their whole careers without having to deal with a sitaution like this , its not something that you can practice for .

 

[Sean Emmett]

Not sure where the 7 mins came from? RTT suggested gap between the two trains much less than 7 mins.

Honiton train seems to have stopped pretty quick, unlikely to have,been doing the full 50 mph?

 

[Bertiethebobby]

But 7 minutes isn't accurate.

Network Rail Wessex confirm on their twitter feed that the incident was 1845. Originally this was quoted as 1846.

That's just two and a quarter minutes after [the front] of the GWR train passed through the Tunnel Junction timing point.

Yes that's not very long at all, 135 seconds.

Assuming there had not been a derailment and 1F30 had proceeded normally and passed through the tunnel without incident, then with such a short time between trains 1L53 would already have received at least 1 if not 2 restrictive aspects (I believe they are R/Y/G LED's?) before reaching the junction? I suppose that the investigation will focus on what these signals were displaying, if anything at all.

 

[Signal Head]

Can’t be informative if it’s also largely incorrect.


For the record, SY31 is a 3 aspect signal. Interesting to read the above though as they are AWS protected on the routes that I sign

AWS was originally intended for, and fitted to, distant signals only. This included colour light Stop signals which also act as Distants by virtue of having a caution aspect. It was not applied to 2 aspect R/G colour light signals.

The standards were changed around 30 years ago to include provision at R/G colour light signals "in colour light signalled areas", so an isolated colour light Starting signal with semaphore Stop signal(s) in rear wouldn't be fitted, for example.

This addition provision was intended to remove the potential for confusion where drivers have been receiving a continuous 'run' of AWS indications and then encounter an unfitted R/G signal.

 

[Ken H]

AWS was originally intended for, and fitted to, distant signals only. This included colour light Stop signals which also act as Distants by virtue of having a caution aspect. It was not applied to 2 aspect R/G colour light signals.

The standards were changed around 30 years ago to include provision at R/G colour light signals "in colour light signalled areas", so an isolated colour light Starting signal with semaphore Stop signal(s) in rear wouldn't be fitted, for example.

This addition provision was intended to remove the potential for confusion where drivers have been receiving a continuous 'run' of AWS indications and then encounter an unfitted R/G signal.

Ah. I was out of date then. Sorry.

 

[Snow1964]

It is 20mph from the Romsey direction and 50mph from Andover.

It is not straight, the first line built (from Eastleigh, then Bishopstone) ended bit further South. When the line from Basingstoke was built, the original line was extended to a junction immediately east of the new Fisherton tunnel. The original Salisbury station then closed and was replaced by current location other end of tunnel.

Although speed limits are different, the lines from the tunnel are more of a Y layout than a straight main line and branch curving off.

 

[baza585]

JourneyCheck lists GWR are running between Romsey and Portsmouth Harbour calling additionally at Redbridge and Millbrook. SWR are stopping their Southampton to Waterloo trains at Swaythling and Basingstoke.

Practical solutions in a difficult situation. Creditable.

 

[Crossover]

Jack Boskett has some pictures from the scene (link should work without an account...ignore the Arabic!)

https://www.facebook.com/100058251992042/

EDIT: Link just to the page - the post itself is near the top currently

A couple of questions re LED signals
i've noticed that alongside the main aspect they have a little repeater, I've assumed that this was a battery back up in case of power failure. Secondly if the signal goes blank doesn't the TPWS still work? Surely that's the idea of it.

Layperson speaking, but I believe it depends on design. There is one LED lightweight signal which has the "repeater" but I have assumed it to be because the main aspect isn't so visible when the driver is close to it?

 

[D6130]

Jack Boskett has some pictures from the scene

There seem to be an awful lot of leaves in those photos....both on the overhanging trees and bushes and on the track.

 

[Dougal2345]

Worst thing about the Mail article is them now giving air to the "terrified" passengers whose first thought was somehow to get their phones out and start filming rather than help others.

Same type of person who in a plane evacuation would be filming or grabbing their bags from the overhead lockers.

We've talked with some relief about the fact that the physical injuries only seem to be minor, but I imagine being in a train crash where the carriage starts tilting over is going to be pretty terrifying, and I suspect a lot of people on that train will be having nightmares for a while, and some perhaps will never want to get on a train again...

 

[HowardGWR]

Practical solutions in a difficult situation. Creditable.

Yes, I am impressed with what they have put together so far.

 

[MarkyT]

Layperson speaking, but I believe it depends on design. There is one LED lightweight signal which has the "repeater" but I have assumed it to be because the main aspect isn't so visible when the driver is close to it?

In the later incandescent era, the long-range fresnel lenses used in signals had a special small section with revised optics to suit close-up viewing. This was known as the 'hot strip' and the lens oriented to place the strip closest to the driver's viewing position. My avatar clearly shows this in the bottom right quadrant of the lens. Some earlier signals achieved the same function using a separate auxiliary lens to the side, lit by the same lamp and known as a 'pig's ear'. Modern LED heads use various methods. Dorman products have a bright 'eye brow' along the edge of part of the lens for example.

 

[Del1977]

It did sound that something falling from the tunnel may have caused the derailment, but looking at the pictures, I would imagine they would also be checking the points to see if they hadn't somehow moved while the GWR set was going across the junction.

 

[bramling]

Not sure where the 7 mins came from? RTT suggested gap between the two trains much less than 7 mins.

Honiton train seems to have stopped pretty quick, unlikely to have,been doing the full 50 mph?

The 7 minutes seems suspect to me. Indeed, it seems *very* fortuitous the third train wasn't involved. This could easily have been worse - without wishing to big things up, this could very easily have resulted in fatalities.

We've talked with some relief about the fact that the physical injuries only seem to be minor, but I imagine being in a train crash where the carriage starts tilting over is going to be pretty terrifying, and I suspect a lot of people on that train will be having nightmares for a while, and some perhaps will never want to get on a train again...

Absolutely. I've been on a train which was involved in a very minor "rough shunt" in a depot, and that made enough of a bang and shudder, for what was essentially a nothing incident.

 

[Ken H]

We've talked with some relief about the fact that the physical injuries only seem to be minor, but I imagine being in a train crash where the carriage starts tilting over is going to be pretty terrifying, and I suspect a lot of people on that train will be having nightmares for a while, and some perhaps will never want to get on a train again...

PTSD...
(Post Traumatic stress Disorder)

Do we know the weather conditions at the time of the accident yet? Wind, rain?
Just looked at this site - its a network of personal weather stations. Its graphs of the days weather numbers. Shows the weather was dry and not particularly windy. https://www.wunderground.com/dashboard/pws/ISALIS25/graph/2021-10-31/2021-10-31/daily

 

[jumble]

I remember reading about train crashes of the past when coaches were quite flimsy and would break up in a crash. The fact this crash didnt cause serious injury or death is a tribute to the design of modern rolling stock.

One just has to take a look at the photos of the 1952 Harrow and Wealdstone crash and see how the coaches were smashed to matchwood

 

[Ken H]

One just has to take a look at the photos of the 1952 Harrow and Wealdstone crash and see how the coaches were smashed to matchwood

and compare to Colwich....

 

[spyinthesky]

PTSD...
(Post Traumatic stress Disorder)

Do we know the weather conditions at the time of the accident yet? Wind, rain?
Just looked at this site - its a network of personal weather stations. Its graphs of the days weather numbers. Shows the weather was dry and not particularly windy. https://www.wunderground.com/dashboard/pws/ISALIS25/graph/2021-10-31/2021-10-31/daily

I can confirm that the local conditions were calm and dry. Heavy downpour at about 4ish. Previous night and early morning were the opposite.

 

[seagull]

A few thoughts which come to mind and might perhaps be (or not be) factors:

1) Time of year = leaf fall (especially so after the wind and rain recently) causing build-up of dirt and detritus on rails, making track circuits more vulnerable to wrong-side failure. Also just possibly causing issues in stopping as quickly as optimal for the second train.
2) Derailed train = Class 158 which are (along with many Multiple Units) lightweight and poor at activating track circuits and hence have a powered Track Circuit Actuator. With one coach off, it leaves just one to do the actuating.
3) Certain areas of the rail network have dead spots on the GMS-R cab-signaller communication system - they are usually short in length and infrequent but they do occur, is this location perhaps one of them?
4) Emergency Protection of the line - imho something that is not always given enough briefing or training time these days, it (thankfully) being such a rare necessity, more so since GSM-R.

All speculation of course, a nasty incident and very glad that it wasn't a lot worse.

 

[Railman]

Looking at the pictures the junction points that are on view would suggest the last "Signalled move was the train entering the tunnel from the Left, and then stopped due to derailment. The following train would not have been given a signal with the junction still set for the first train, that was also still on the line, not to mention damage to the signalling system that would always go to a fail safe condition when faulty. So despite derailment the second train must have sufferd some operational or equipment failure (by who or what still to be investigated) to end up where it did.

 

[Skip 10]

I was going to ask a question about detonators, and found this link. If the 7 minute gap is true, shoudn't these have been used? (unless the GWR driver and guard were both too injured to do anything)

What is the purpose of railway detonator equipment? | RailUK Forums (railforums.co.uk)

One thing to remember here is that, if I've read rightly, 1/2 - 3/4 of the GWR train remained on the rails at the time? Therefore it would be assumed that the track circuit was still showing occupied and therefore protecting the train from the rear (the crew wouldn't know the signaling had failed at this point) and the priority would be to protect the train from on coming traffic (So heading towards Salisbury to stop anything coming along the other line) and to make contact with the signaller.

 

[paul1609]

It did sound that something falling from the tunnel may have caused the derailment, but looking at the pictures, I would imagine they would also be checking the points to see if they hadn't somehow moved while the GWR set was going across the junction.

There is at least one rumour doing the rounds that the derailment was only caused by the collision between the 2 trains on the junction.

 

[daikilo]

Looking at the pictures the junction points that are on view would suggest the last "Signalled move was the train entering the tunnel from the Left, and then stopped due to derailment. The following train would not have been given a signal with the junction still set for the first train, that was also still on the line, not to mention damage to the signalling system that would always go to a fail safe condition when faulty. So despite derailment the second train must have sufferd some operational or equipment failure (by who or what still to be investigated) to end up where it did.

Oh, so we can rule out a SPAD then (of either train).

 

[molecrochip]

If, the GWR train hitting an object was from the initial GWR driver reporting an issue, this is fully understandable. Surely, the driver's first reaction would be more towards hitting an object than another train hitting it based on frequency of these events happening. Hence the importance of the RAIB doing their job.

 

[43096]

Oh, so we can rule out a SPAD then (of either train).

No, we can’t rule out a SPAD.

 

[MarkyT]

Early 1980s era Salisbury ASC has a very early automatic route setting installation to assist the signallers in operating the NX panel and relay interlockings, using an experimental Southern Region system known as ERSE (electronic route setting equipment). My fear is that if sufficient of the first train had left the track in order for the track circuit to clear, then the automatic route setting could have stepped in to set the route for the following train immediately before anyone had any time to think, and because the track circuit was still showing clear, the signal would have cleared straight away as commanded.

Pictures of the signalling panel here:

Southern: Wessex (SW) / Salisbury | Signalling Photos

Southern: Wessex (SW) / Salisbury

photos.signalling.org

 

[anthony263]

Now that we've had the good news that no one thankfully was seriously hurt, I have some questions about service recovery. Obviously RAIB will take as long as they need at the site, then Network Rail will need to make extensive repairs. This will take some time.

Is there anywhere on SWR's network that can service 159s apart from Salisbury depot? There will be some units "trapped" at the London end, and I presume a few will be needed for a Basingstoke to Andover shuttle.

Does the infrastructure allow for the "Salisbury 6" services to be turned at Dean?

The stations served by GWR South of Salisbury have provision from other operators, so I guess there will be no GWR services here. Is there any provision to get any GWR units back to their home depot?

Saw something on Facebook last night of two class 159s at Exeter st davids. Could exeter depot maintain them?

 

[Ken H]

Would it be possible for someone to summarise what we actually know at this stage, please. It would be nice to separate actual known fact from speculation.

 

[Wilts Wanderer]

Even if signals were black, you’d expect the SWR service to have reacted to the AWS magnet warnings it will have received? Something still doesn’t sound right about this.