Lumo overspeed incident at Peterborough (17/04/2022)

[karlbbb]

Indeed, isn't the issue the junction signal being someway before the junction, and the 80x being very quick to accelerate?

The 80x does indeed seem to have impressive acceleration!

At 10:19:09 hrs, the JI for the Up Slow line to platform 1 illuminated, followed within a second by the main aspect of P468 signal changing from red to green (figure 5). The train was travelling at around 25 mph (40 km/h) and was approximately 790 metres from the signal when the aspect changed. Three seconds later and with the train approximately 760 metres from P468 signal, the driver applied full traction power to accelerate the train (figure 16). The train passed P468 signal travelling at around 64 mph (103 km/h) and was still accelerating.
At 10:20:08 hrs, the train was approximately 160 metres from 1243 points and had reached around 77 mph (123 km/h). The driver stated that they saw the platform banner repeater signal at the north end of platform 3 was ‘on’, indicating that the route for the Up Fast line through platform 3 was not set. They reacted by applying a full service brake (figure 17).

760m to get from 25mph to 64mph and soon after already at 77mph.

 

[Nottingham59]

Why aren't supplementary TPWS loops fitted on the approach to junctions, to be activated when the diverging route is set?

 

[railfan99]

I'm unfamiliar with the area but 80 vice 30 is a pretty nasty overspeed over points. They're not the first to go into a set of points too hard, and they sure as anything won't be the last; hopefully the folks on board who've had injuries have a speedy recovery and the driver isn't too shaken or injured.

Albeit not approaching a passenger platform (the standard gauge line from Albury to Melbourne parallels the two broad gauge tracks at Wallan, with the latter having local passenger trains), there was an eerily similar incident on 20 February 2020 in Victoria, Australia where the driver (aided by a trackwork assistant) took points into a crossing loop at up to 127kmh (almost 80mph) despite a temporary 15kmh speed restriction.

Unfortunately, neither survived, but thanks to the crashworthiness of the NSW TrainLink XPT between Sydney and Melbourne, no passengers were killed:

https://www.atsb.gov.au/publications/investigation_reports/2020/rair/ro-2020-002

‘The thing he loved to do’: Train driver killed in derailment identified

The driver who died when an XPT derailed north of Melbourne has been remembered as a very experienced worker with a passion for train driving.

7news.com.au

 

[Merle Haggard]

A few years ago I caught a Northern Manchester - Stoke train. Unusually, it was signalled into the up platform at Stoke, rather than the usual bay on the down side. The signal at the south end displayed feathers to the left indicating the siding on the up side. The driver left his cab and asked the Virgin/Avanti customer service person on the platform what the signal meant; the person, displaying a wider knowledge of railway operating, suggested it meant crossing over, but the driver initially seemed to gbelieve that it meant using the (trailing) crossover onto the down (to the right, obviously) - over which a Pendo was running at that moment.
So, anecdotally, feathers may not always be understood. Perhaps lessons could be learned.

See also light engine derailment (at speed), Bletchley, a few years ago.

 

[Snow1964]

Indeed, isn't the issue the junction signal being someway before the junction, and the 80x being very quick to accelerate?

Thats my understanding too, the approach control worked, train slowed approaching the red, but too much acceleration when it cleared.

It appears trains slowed by the approach signal can get upto about 70mph by the junction, so unless junction is relaid to increase speed, or moved nearer signal, probably needs an extra TPWS magnet that can indicate the lower speed when that low speed route is set.

The current system doesn't really limit speed even when there is not a junction, nothing to stop a train accelerating through a single yellow, as long as it can stop again before next signal.

 

[railfan99]

It's amazing what we each interpret from the report (and indeed many others)

76 in a 30 and there was no derailment and no damage caused. That, to me, highlights how robust and how good our tolerance's can/are. A testament to our safety culture for sure.

It also once again highlights how well designed and structurally sound much rollingstock is in many nations including yours and mine.

 

[185]

I wonder if a negligence investigation and prosecution by the CPS will follow, not only for the driver but for the operator too, to establish what level of training was provided to misinterpret an approach control / red protected diverging junction, when the signals prior had correctly brought the train's speed down.

Seems to be a lot of concentration on luggage - flying luggage has always been a problem, but a long distance train with inadequate luggage space for long distance journeys begs a question - did FirstGroup order trains with too many seats and not enough luggage space? Are the RAIB suggesting luggage on the top racks affected the upright stability?

 

[Neo9320]

The RAIB YouTube video of the incident if anyone is yet to see it.

 

[800001]

Indeed, isn't the issue the junction signal being someway before the junction, and the 80x being very quick to accelerate?

Except the 2nd incident involved a Grand Central 180.

 

[AlterEgo]

A few years ago I caught a Northern Manchester - Stoke train. Unusually, it was signalled into the up platform at Stoke, rather than the usual bay on the down side. The signal at the south end displayed feathers to the left indicating the siding on the up side. The driver left his cab and asked the Virgin/Avanti customer service person on the platform what the signal meant; the person, displaying a wider knowledge of railway operating, suggested it meant crossing over, but the driver initially seemed to gbelieve that it meant using the (trailing) crossover onto the down (to the right, obviously) - over which a Pendo was running at that moment.
So, anecdotally, feathers may not always be understood. Perhaps lessons could be learned.

See also light engine derailment (at speed), Bletchley, a few years ago.

A driver took a signalling instruction/interpretation from a member of platform staff? That’s surely one for the signaller to clarify.

 

[alistairlees]

The RAIB YouTube video of the incident if anyone is yet to see it.

The RAIB seems confused asto whether it is P468 or P486 signal

 

[43066]

A driver took a signalling instruction/interpretation from a member of platform staff? That’s surely one for the signaller to clarify.

Absolutely.

 

[class 9]

To me, the driver's lack of training on diverging routes has got to be the crucial role in the overspeed, which probably tells that Lumo haven't done their homework well, so this must put some uneasiness on passengers. This reminds me of the Ladbroke Grove collision, in which the lack of training provided by Thames Trains to the driver was one of the key roles in causing the accident. Perhaps Lumo needs to be scrutinised to a higher degree to ensure their training courses are of adequate standards?

My sentiments as well , the report mentions the amount of accompanied driving(aka handling hours) that Lumo require their trainee Drivers to do:175 hrs minimum, it goes on to mention other TOCs having a minimum of 225 hrs minimum and an unnamed TOC that also operates on the ECML a minimum of 300 hrs!!

 

[dk1]

My sentiments as well , the report mentions the amount of accompanied driving(aka handling hours) that Lumo require their trainee Drivers to do:175 hrs minimum, it goes on to mention other TOCs having a minimum of 225 hrs minimum and an unnamed TOC that also operates on the ECML a minimum of 300 hrs!!

At my TOC it's now 296 as opposed to 225hrs but when drivers pass out they now sign all routes rather than just the basic ones as this is built into the training programme.

 

[Dieseldriver]

A few years ago I caught a Northern Manchester - Stoke train. Unusually, it was signalled into the up platform at Stoke, rather than the usual bay on the down side. The signal at the south end displayed feathers to the left indicating the siding on the up side. The driver left his cab and asked the Virgin/Avanti customer service person on the platform what the signal meant; the person, displaying a wider knowledge of railway operating, suggested it meant crossing over, but the driver initially seemed to gbelieve that it meant using the (trailing) crossover onto the down (to the right, obviously) - over which a Pendo was running at that moment.
So, anecdotally, feathers may not always be understood. Perhaps lessons could be learned.

See also light engine derailment (at speed), Bletchley, a few years ago.

I disbelieve your interpretation of events

 

[66701GBRF]

The RAIB seems confused asto whether it is P468 or P486 signal

There is no confusion. The diagram and signal numbers in the video are correct.

== Doublepost prevention - post automatically merged: 10 Jul 2023 ==

A few years ago I caught a Northern Manchester - Stoke train. Unusually, it was signalled into the up platform at Stoke, rather than the usual bay on the down side. The signal at the south end displayed feathers to the left indicating the siding on the up side. The driver left his cab and asked the Virgin/Avanti customer service person on the platform what the signal meant; the person, displaying a wider knowledge of railway operating, suggested it meant crossing over, but the driver initially seemed to gbelieve that it meant using the (trailing) crossover onto the down (to the right, obviously) - over which a Pendo was running at that moment.
So, anecdotally, feathers may not always be understood. Perhaps lessons could be learned.

See also light engine derailment (at speed), Bletchley, a few years ago.

I find it near impossible to believe that a driver didn't/doesn't know what a route indicator is.

 

[12LDA28C]

A few years ago I caught a Northern Manchester - Stoke train. Unusually, it was signalled into the up platform at Stoke, rather than the usual bay on the down side. The signal at the south end displayed feathers to the left indicating the siding on the up side. The driver left his cab and asked the Virgin/Avanti customer service person on the platform what the signal meant; the person, displaying a wider knowledge of railway operating, suggested it meant crossing over, but the driver initially seemed to gbelieve that it meant using the (trailing) crossover onto the down (to the right, obviously) - over which a Pendo was running at that moment.
So, anecdotally, feathers may not always be understood. Perhaps lessons could be learned.

See also light engine derailment (at speed), Bletchley, a few years ago.

I find it impossible to believe that a driver wouldn't know the difference between a Junction Indicator indicating a route set to the left (1,2,3) and one to the right (4,5,6). It would indicate someone doesn't know their left from their right, or possibly their arse from their elbow.

 

[Cornish Guilt]

I do wonder if the driver got tunnel vision. He was aware that a service (1L06) was in front of him as the report states he had already been running on restrictive aspects and had reduced his speed to 95mph to ensure that the signals were stepping up to green aspects as he got to them. Then at Peterborough wrongly assumed that when P468 stepped up from red to green meant that 1L06 had got out of his way allowing him to return to line speed, rather than what was actually happening in being sent across into platform 1, and as such did not register the feather as a result of this mindset.

 

[karlbbb]

The RAIB YouTube video of the incident if anyone is yet to see it.

Holy moly! Quite the sideways lurch in the "dashcam" video there.

I do wonder if the driver got tunnel vision. He was aware that a service (1L06) was in front of him as the report states he had already been running on restrictive aspects and had reduced his speed to 95mph to ensure that the signals were stepping up to green aspects as he got to them. Then at Peterborough wrongly assumed that when P468 stepped up from red to green meant that 1L06 had got out of his way allowing him to return to line speed, rather than what was actually happening in being sent across into platform 1, and as such did not register the feather as a result of this mindset.

The RAIB investigation says as much - the strong expectation (thanks to almost "always" doing it) of having the up fast as normal meant, effectively, the junction indication was missed (or subconsciously ignored).

 

[Ashley Hill]

The sighting of the route indicator (feather) on P468 in the video isn’t very clear. This is not helped by being rather divorced from the green below it. Perhaps if the green was moved to the top position on the signal the feather would have been more obvious. To me it looks like the driver concentrated on the green without noticing the feather.

 

[AlterEgo]

The sighting of the route indicator (feather) on P468 in the video isn’t very clear. This is not helped by being rather divorced from the green below it. Perhaps if the green was moved to the top position on the signal the feather would have been more obvious. To me it looks like the driver concentrated on the green without noticing the feather.
View attachment 138864

I also think that the horizontal feather indicator being set against a horizontal structure like the bridge makes it less visible, especially if you aren't looking for it.

 

[MadMac]

The sighting of the route indicator (feather) on P468 in the video isn’t very clear. This is not helped by being rather divorced from the green below it. Perhaps if the green was moved to the top position on the signal the feather would have been more obvious. To me it looks like the driver concentrated on the green without noticing the feather.
View attachment 138864

That RI looks like something painted on the bridge.

 

[Nottingham59]

The RAIB YouTube video of the incident if anyone is yet to see it.

I would like to see the few seconds before that video starts. How many seconds was the green signal (and the feathers) visible to the driver before they went past it? And does the horizontal part of the OHLE gantry visible in the starting frame of that video obscure the horizontal line of lights that compose that particular feather?

EDIT: I see the driver was 700m away from signal P468 when it went green. How visible are the feather lights from that distance? Seems very similar to the overspeed incident at Bletchley when the driver saw the green and did not notice the (fainter) line indicator panel, and accelerated into a divergent route.

 

[karlbbb]

We shouldn't discount how much the image is compressed (loss of quality and dynamics) compared to what the human eye would have been seeing at the time, though I do admit that the signal being so close to the bridge structure almost makes the JIs much easier to dismiss, compared to seeing the signal against the sky which would give the silhouette of the JI surrounds much more prominence.

 

[Taunton]

My understanding is that Open Access operators, to get away from minimalist formations filling up the paths, have been set a minimum formation - which is done by setting 400 seats as the requirement. Presumably whoever set this believed that this would ensure at least a 6-car formation (average 67 seats per car).

However, someone from Lumo worked out that if they averaged 80 seats per car, they could, just, get away with a 5-car formation, by dumping all the frippery of baggage storage, etc, and minimum seat pitch. Cheaper to build and run. And nobody in the new world of just got to comply with the regulation said "that's not what we meant". Especially for a 400-mile mainstream haul, where such passengers do typically come with baggage. It's not a commuter run.

Now the genie is out of the bottle, but nobody can bring themselves to do a derogation for Lumo saying they can reduce the seats, but have to provide heavy luggage stacks so they are not up over people's heads. And Lumo are not allowed to do this unilaterally themselves, even if they wanted to, because they will get hit for dropping below the 400 seat limit. It would be nice if the RAIB report made such a point, and even better if they told those who frame the open access regulations not to be so silly, but it seems industry niceties do not allow them to draw conclusions like this. Regardless of the injuries sustained, now, and potentially in any future event.

 

[Railsigns]

Perhaps if the green was moved to the top position on the signal the feather would have been more obvious.

You can't put every aspect at the top. Apart from being non-standard, moving the green lens the top position would increase the distance between the feather and the yellow lenses.

 

[ExRes]

Personally I wouldn't take much notice of how visible, or not, the signal and route indicator appear on a video, I hit that signal countless times and it was clearly readable in any light or weather, the only way to foul it up was by having a complete lack of concentration and/or interest and failing to apply basic driving skills to the signals and associated warnings previously received

 

[Ashley Hill]

You can't put every aspect at the top of a standard multi-lens signal. Apart from being non-standard, moving the green lens the top position would increase the distance between the feather and the yellow lenses.

Maybe non-standard but perhaps G/Y in top position and R/Y (for double Y) in bottom position.

 

[Crossover]

Maybe non-standard but perhaps G/Y in top position and R/Y (for double Y) in bottom position.

But then there was also an accident which was at least part attributed to a non-standard signal (Southall?)

As for the service in question, it is probably also fortunate that the route wasn't set out of P3 at the time (where the driver would have seen the banner to be "off"), though I am unsure whether interlocking would have permitted this to happen

 

[12LDA28C]

That RI looks like something painted on the bridge.

It's a JI (for Junction Indicator) and that is a very poor photo, the human eye would see it much better / clearer than that.

== Doublepost prevention - post automatically merged: 10 Jul 2023 ==