GC incident at Peterborough (04/05)

[Signal Head]

Interesting to note pages 45/46

They go into a lot of detail about the alignment of the signal aspect and the associated junction indicator.
The Junction indicator is pointing far too high and as everyone will know with LEDs, they have a much smaller viewing angle than filament lamps - just look at a motorway VMS sign next time you pass under it, they almost disappear when you get close.

I have noted this myself on occasion at some signals where the JI is much harder to see than the aspect itself. More needs to be done to ensure proper alignment of signals!

It does make interesting reading. The vertical 'misalignment' of the JI is, I think, shocking, when you look at where most of the light is going, I'm not surprised it didn't stand out sufficiently to get the driver's attention.

The comparison between filament and LED JIs is also interesting. The comparison was between a 'modern' filament JI using SL18 (or SL35) lamps. An SL35 is the type used in main aspects and is provided with dual filaments, one acting as an auxiliary for when the main fails. The SL18 is the same rating but without the aux filament.

Older JIs used a different lamp, from memory SL33, which were a 110v lamp with a similar rating but, being 110v (SL18/35 are 12v) the filament is longer and not straight, a bit more like those in domestic lamps. This means much less of the filament is at the focal point of the lens and the beam is therefore less concentrated. I remember the first time I saw an SL35 type how much brighter it was than the old one it had replaced, and this was reinforced at another location where old and new were positioned side by side, the older 1960s vintage JI was very much less visible at distance than the newer one, which actually tended to 'drown out' the main aspect below it when first sighted, especially when displaying a Green, exactly the opposite of what appears to have happened here.

 

[bengley]

It does make interesting reading. The vertical 'misalignment' of the JI is, I think, shocking, when you look at where most of the light is going, I'm not surprised it didn't stand out sufficiently to get the driver's attention.

The comparison between filament and LED JIs is also interesting. The comparison was between a 'modern' filament JI using SL18 (or SL35) lamps. An SL35 is the type used in main aspects and is provided with dual filaments, one acting as an auxiliary for when the main fails. The SL18 is the same rating but without the aux filament.

Older JIs used a different lamp, from memory SL33, which were a 110v lamp with a similar rating but, being 110v (SL18/35 are 12v) the filament is longer and not straight, a bit more like those in domestic lamps. This means much less of the filament is at the focal point of the lens and the beam is therefore less concentrated. I remember the first time I saw an SL35 type how much brighter it was than the old one it had replaced, and this was reinforced at another location where old and new were positioned side by side, the older 1960s vintage JI was very much less visible at distance than the newer one, which actually tended to 'drown out' the main aspect below it when first sighted, especially when displaying a Green, exactly the opposite of what appears to have happened here.

Indeed, I really do think the RAIB have downplayed quite how much this could have contributed to the incident, instead concentrating more on the fact the driver only looked at the signal once despite it being displayed to him for over a minute...
Perhaps he only looked at it once because he was certain what had been displayed (despite his certainty being misplaced)

 

[Meerkat]

There was a warning board with AWS magnet, according to the RAIB Report.

The signal is effectively displaying a restrictive aspect (a route you need to slow down for) so it would be consistent for it to get an AWS alarm on approach, as a yellow would.

 

[357]

The signal is effectively displaying a restrictive aspect (a route you need to slow down for) so it would be consistent for it to get an AWS alarm on approach, as a yellow would.

No it wouldn't.

 

[norbitonflyer]

The horizontal "position 2" feathers don't seem as distinctive as the diagonal position 1 or 3 ones, which may also have contributed in this case. I wonder if they would be better if they could flash in sequence (like some modern car indicators do)

 

[Meerkat]

No it wouldn't.

Are you going to explain why not?

 

[Sheridan]

Are you going to explain why not?

I obviously can’t speak for the other poster but it would seem to be a very large change in basic signalling principles to have a warning associated with a green signal. If what you’re suggesting was felt desirable (and I do understand where you’re coming from) then the signal showing an actual restrictive aspect would probably be an easier way of doing it. But regardless of that there is a separate warning for the reduction in speed anyway. Others have suggested some form of reminder of this nearer to the junction - I would tend to agree that of any solution, that may be the best one.

 

[357]

I obviously can’t speak for the other poster but it would seem to be a very large change in basic signalling principles to have a warning associated with a green signal. If what you’re suggesting was felt desirable (and I do understand where you’re coming from) then the signal showing an actual restrictive aspect would probably be an easier way of doing it. But regardless of that there is a separate warning for the reduction in speed anyway. Others have suggested some form of reminder of this nearer to the junction - I would tend to agree that of any solution, that may be the best one.

You're bang on my train of thought.

If I get an AWS warning on a green signal then I should be reporting it to the signaller.

Route indicators are NOT a warning of a speed reduction. In fact, the diverging route could have no speed reduction at all or have an increase in speed immediately after the points.

 

[Meerkat]

You're bang on my train of thought.

If I get an AWS warning on a green signal then I should be reporting it to the signaller.

Route indicators are NOT a warning of a speed reduction. In fact, the diverging route could have no speed reduction at all or have an increase in speed immediately after the points.

Thank you. I’m not in the industry so have a different angle of thought.
It wouldn’t be needed if there wasn’t a dramatic speed reduction.
As there is one the JI is effectively a warning of a speed difference (Isnt that it’s main purpose - why else does it really matter?).
The AWS warning on a green would be explained by the JI, and bring attention to it if it hadn’t already been noticed.
I accept the current obvious method is to show yellow rather than green, but personally I don’t like the idea of a yellow meaning something other than ‘get ready to stop’ as it risks familiarity that the next signal ‘is always green’…..until it isn’t.

 

[zwk500]

As there is one the JI is effectively a warning of a speed difference (Isnt that it’s main purpose - why else does it really matter?).

The primary purpose of the JI is to indicate the *Route*. Speed is one factor relevant to this, but also route clearance, driver route knowledge, electrification, further routing for the service, platform acceptability, etc...

 

[43066]

The AWS warning on a green would be explained by the JI, and bring attention to it if it hadn’t already been noticed.

No, it wouldn’t. It would be a right side failure, and reportable to the signaller as such.

I accept the current obvious method is to show yellow rather than green, but personally I don’t like the idea of a yellow meaning something other than ‘get ready to stop’ as it risks familiarity that the next signal ‘is always green’…..until it isn’t.

Yellow (single or double, two, three or four aspect, flashing or steady, whereabouts is it, do you have a feather, what does your route knowledge tell you about this precise location etc.) doesn’t always mean “get ready to stop”. On my patch there are single yellows I’d happily blow past at 100mph+, and others I’d crawl by at no more than 20mph.

With respect, you clearly don’t have the foggiest idea how UK railway signaling actually works, so should you really be opining on what signal aspects should mean, or when AWS warnings should or shouldn’t be sounding?

 

[357]

As there is one the JI is effectively a warning of a speed difference (Isnt that it’s main purpose - why else does it really matter?

No. It's to indicate where you're going.

The passengers tend to want to go to the advertised destination and not on a mystery tour of the UK rail network.

 

[norbitonflyer]

I obviously can’t speak for the other poster but it would seem to be a very large change in basic signalling principles to have a warning associated with a green signal. If what you’re suggesting was felt desirable (and I do understand where you’re coming from) then the signal showing an actual restrictive aspect would probably be an easier way of doing it. But regardless of that there is a separate warning for the reduction in speed anyway. Others have suggested some form of reminder of this nearer to the junction - I would tend to agree that of any solution, that may be the best one.

I thought AWS at speed restrictions had been standard practice since Morpeth (and at temporary ones since Nuneaton) so I'm surprised it doesn't seem to be fitted at junctions such as this.

 

[357]

I thought AWS at speed restrictions had been standard practice since Morpeth (and at temporary ones since Nuneaton) so I'm surprised it doesn't seem to be fitted at junctions such as this.

It is. There is a warning triangle with AWS that is only active when the signal is red or set for a diverging route, however to comply with braking distances it is a long way back before the prior signal.

What that poster suggested was a warning on the green signal if it's showing a feather.

 

[43066]

I thought AWS at speed restrictions had been standard practice since Morpeth (and at temporary ones since Nuneaton) so I'm surprised it doesn't seem to be fitted at junctions such as this.

Only large speed reductions, and then not everywhere. Not all Morpeth boards have AWS magnets either.

 

[Tomnick]

I accept the current obvious method is to show yellow rather than green, but personally I don’t like the idea of a yellow meaning something other than ‘get ready to stop’ as it risks familiarity that the next signal ‘is always green’…..until it isn’t.

The concept of an AWS warning is a non-starter – routinely acknowledging the warning there on a green undermines the primary function of AWS. If you wanted that, it'd be better to have it associated with an additional warning board on the immediate approach to the signal, suppressed with the straight route set.

The bit that I've quoted is spot on, though. There's a few places on the routes that I sign in in the North West where there's a bit of a compromise on that arrangement – the junction signal is held at single yellow until you've passed over the magnet (and received the warning) before being released to step up to a less restrictive aspect. There's a similar arrangement at Stoke Junction, further north on the ECML and part of the same signalling scheme, where the junction signal is held at YY (in this case, it doesn't release, with full braking distance still available from the Y to the R). If anything, I'd suggest that would be the "least worst" arrangement here.

 

[John Webb]

The horizontal "position 2" feathers don't seem as distinctive as the diagonal position 1 or 3 ones, which may also have contributed in this case. I wonder if they would be better if they could flash in sequence (like some modern car indicators do)

The RAIB report clearly states the 'Position 2' feather was badly aligned - had it been better aligned it seems it would have been rather more obvious and therefore wouldn't need anything else done to it?

 

[66701GBRF]

RAIB video explanation on the incident.

 

[TreacleMiller]

I thought AWS at speed restrictions had been standard practice since Morpeth (and at temporary ones since Nuneaton) so I'm surprised it doesn't seem to be fitted at junctions such as this.

It is fitted here. You get a warning for the 30 points which you technically hit first. You then get a further two magnets depending on the aspects show.

Aws isn't standard on Morpeth boards though. There's numerous places where the boards are and are not aws or tpws protected.

 

[theageofthetra]

It is fitted here. You get a warning for the 30 points which you technically hit first. You then get a further two magnets depending on the aspects show.

Aws isn't standard on Morpeth boards though. There's numerous places where the boards are and are not aws or tpws protected.

One ridiculous example where AWS is not fitted is on the 60 down to 20 PSR.on the up approaching Elmers End on the Hayes Kent branch. That curve is one of the sharpest on the UK rail network

 

[PhilipRdNGauge]

If this area had been signalled with ETCS, would the movement authority at the first signal have allowed for the speed restriction implied by the route set further down the line & forced the train to slow down, so it was safe by the time it reached the points.

 

[zwk500]

If this area had been signalled with ETCS, would the movement authority at the first signal have allowed for the speed restriction implied by the route set further down the line & forced the train to slow down, so it was safe by the time it reached the points.

Is this signal within the current proposal for ETCS L2 Overlay at Peterborough?

 

[357]

One ridiculous example where AWS is not fitted is on the 60 down to 20 PSR.on the up approaching Elmers End on the Hayes Kent branch. That curve is one of the sharpest on the UK rail network

Gas Factory Curve and Limehouse have similar speed changes with no warning too.

 

[Taunton]

Gas Factory Curve and Limehouse have similar speed changes with no warning too.

The difference there is the drivers go round these every trip, rather like I turn into my cul-de-sac at home every time I take the car out, without needing a warning. The situation at Peterborough was an abnormal turnout, not the scheduled route, which I believe the driver had never needed to take before. Giving a green on the main line, with a "hope they notice the route indicator", seems something to have slipped through any meaningful Risk Assessment - even after the previous Lumo event.

Back in semaphore days you would never have splitting distants pulled off for such a sharp turnout - which is what a green on the main line with a route indicator is the equivalent of.

 

[43066]

The bit that I've quoted is spot on, though. There's a few places on the routes that I sign in in the North West where there's a bit of a compromise on that arrangement – the junction signal is held at single yellow until you've passed over the magnet (and received the warning) before being released to step up to a less restrictive aspect. There's a similar arrangement at Stoke Junction, further north on the ECML and part of the same signalling scheme, where the junction signal is held at YY (in this case, it doesn't release, with full braking distance still available from the Y to the R). If anything, I'd suggest that would be the "least worst" arrangement here

That is a perfectly normal signalling arrangement, though. As professionals we should be able to cope with simultaneously knowing that approach release exists, and also not assuming it will happen. There are rules course instructors who deny approach release exists, of course, which is embarrassing (for them).

The difference there is the drivers go round these every trip, rather like I turn into my cul-de-sac at home every time I take the car out, without needing a warning. The situation at Peterborough was an abnormal turnout, not the scheduled route, which I believe the driver had never needed to take before. Giving a green on the main line, with a "hope they notice the route indicator", seems something to have slipped through any meaningful Risk Assessment - even after the previous Lumo event.

Back in semaphore days you would never have splitting distants pulled off for such a sharp turnout - which is what a green on the main line with a route indicator is the equivalent of.

But ultimately if you sign the route, you sign the signalling turnoffs, even the unusual ones. Unless there’s a defect with the feather in question it’s down to a failure of route knowledge/lack of situational awareness.

 

[66701GBRF]

The difference there is the drivers go round these every trip, rather like I turn into my cul-de-sac at home every time I take the car out, without needing a warning.

Given you drive on line of sight it's not really the same thing is it. What about engineering trains, tampers, RHTTs etc?

But ultimately if you sign the route, you sign the signalling turnoffs, even the unusual ones. Unless there’s a defect with the feather in question it’s down to a failure of route knowledge/lack of situational awareness.

That doesn't mean there shouldn't be safeguards or mitigation against human error.

 

[43066]

That doesn't mean there shouldn't be safeguards or mitigation against human error.

Absolutely, and the safeguard in this case is the approach release. I have a location where, with flashing yellows, the line speed goes from 110mph to a (tight) 30mph turnoff. Nothing other than own knowledge prevents me from ploughing into it at 110 which would likely result in a major derailment and fatalities.

Ultimately getting this stuff right is what we’re paid for. The existing system is incredibly safe and will ultimately be superseded by ETCS and enforced braking curves.

 

[Signal Head]

You're bang on my train of thought.

If I get an AWS warning on a green signal then I should be reporting it to the signaller.

Route indicators are NOT a warning of a speed reduction. In fact, the diverging route could have no speed reduction at all or have an increase in speed immediately after the points.

Quite. There is a deep-seated reluctance to give drivers an AWS indication that varies from that displayed by the signal.

In the past a 'two stage' release has been applied, where the signal clears from R>Y quite early, to avoid unnecessary braking, but then steps up to a G once the train has passed the AWS magnet.

Although it wasn't covered by standards, I've put a similar arrangement in more recently where it was desirable to relax an overly restrictive Approach Release from Red, but there were concerns that an unrestricted clearance may lead to premature acceleration owing to the complex sighting of the JI. This was accepted condional upon the change to Green occuring before the AWS so that the driver didn't receive a 'horn' but then pass the signal on Green.

 

[Ghostbus]

Anything where route knowledge alone is the difference between life and death, is clearly unsafe. It seems nobody picked up on the urgency of the pertinent issue from the Luno incident - new trains accelerate quickly, changing the risks of certain routes and signalling arrangements.

 

[godfreycomplex]

Is this signal within the current proposal for ETCS L2 Overlay at Peterborough?

It is indeed