Emergency stop - 08:20 Newcastle to London (Lumo) 17.4.2022

[whole diary]

[jamestheredengine]

3. I note this comment from Clarence Yard on railforums (he's in GWR^▸ (Great Western Railway) fleet management, but Lumo is a "TOC^▸ (Train Operating Company)-lite" so relies on expertise from the rest of First Group):

For those of you that haven’t experienced an emergency brake application from a relatively high speed on an 800 series unit, it’s completely different to that experienced on an HST^▸ (High Speed Train) or modern EMU^▸ (Electric Multiple Unit). If you are standing up or not sitting securely in your seat, you will be at risk. Luggage will move too, depending on size and how it is stowed.

I guess that implies that, with good adhesion, the combination of regenerative (traction motor) and friction braking can manage significantly more than 1.2 m/s/s - even from high speeds, where older high-speed trains struggle. Mind you, this was not really from high speed.

I've been on one that did an emergency stop from reasonably high speed (75mph, I think). Wednesday February 5th, 2020. This was on 1B21, the 1548 Paddington (1742 Cardiff) to Swansea. The guard had checked our tickets after we'd left Cardiff and gone into the kitchen, and we'd got our free wine (glass of red, nice) and thankfully the host had also just made it back into the kitchen with the trolley. We get to the last signal before Pontyclun and I was very glad I had a napkin under my wine to slow its journey across the table enough for me to catch it (could have been messy!). We came to a stop with First Class short of the platform at Pontyclun – very impressive. No sooner had we come to a stop than the guard emerges from the kitchen muttering "this isn't good at all", followed by a louder "is everyone okay?" What it turned out had happened was that there was a thankfully totally incompetent "distressed person" on the line in front of TfW's stationary stopping service at Pencoed (there was a steel train between them and us (yes, brilliant work from the signallers there...) – hence the quite so awkward location for an emergency stop – the natural worry there is that that's the signal where you'd find out that something had gone awry with the barriers at Llantrisant West Level Crossing, but the front of the train wouldn't have reached that). We had about a 40-minute delay as the police persuaded the distressed person to leave the railway. Everything was all right in the end (and we all got seconds of wine). But the rate at which that stopped was truly staggering.

[stuving]

RAIB^▸ (Rail Accident Investigation Branch) have - a bit later than you'd expect - announced an enquiry into this incident:

At around 10:20 hrs on 17 April 2022, the 08:20 hrs Newcastle to King’s Cross service, operated by Lumo, passed over three sets of points at the north end of Peterborough station at a speed of 75 mph (121 km/h). This was above the maximum permitted speed limit for these points of 25 mph (40 km/h). Passing over the points at this speed meant that the train suddenly lurched sideways.

The sudden movement of the train resulted in some passengers being thrown from their seats and in luggage falling from the overhead storage in the passenger compartments. This resulted in a number of minor injuries being sustained by passengers. The train subsequently came to a stand beyond the far end of Peterborough station. The train did not derail during the incident, and no damage was caused to the infrastructure of the railway, or to the vehicles involved.

The train involved in the incident was not due to call at Peterborough and had originally been approaching the station on a fast line, before being routed onto a slower line via the points where the overspeed occurred. The route onto this slower line was being displayed on the signal situated on the approach to the points.

Our investigation will seek to identify the sequence of events which led to this incident. It will include consideration of:

    the way in which the train was driven
    any factors which may have influenced the actions of the driver of the train
    the condition of the signalling system at the time of the incident
    the nature of the reported injuries
    any underlying management factors.

[stuving]

RAIB^▸ (Rail Accident Investigation Branch) have - a bit later than you'd expect - announced an enquiry into this incident:

RAIB have now issued some "urgent" safety advice, following that Lumo incident and a strikingly similar one for a GC^» (Great Central Railway - link to heritage line) train earlier this month. It all comes down one of those complicated signalling issues to do with flashing yellow aspects for turnouts, but the potential - and actual - results are far from subtle.

Urgent Safety Advice 02/2023: Overspeeding through Spital Junction
Published 25 May 2023

1. Safety issue

Suitable arrangements may not be in place to mitigate the risk of trains travelling southbound through Spital Junction at excessive speeds when signalled from the Up Fast line onto the Up Slow lines at Peterborough station.

P468 signal, which controls this junction, is located 700 metres on the approach to the point of divergence. The signal is fitted with a position light junction indicator informing drivers of their signalled route. Under certain circumstances, the signal clears from red as a train approaches when a diverging route is set. The maximum permitted speed through the diverging junction is initially 30 mph (48 km/h), before reducing further to 25 mph (40 km/h).

Drivers who rarely experience being routed towards the slow lines when approaching Peterborough station from the north, and whose trains are not scheduled to stop at the station, may develop an expectation that their train will remain on the Up Fast line and miss some of the information provided at P468 signal when their train is being signalled onto the diverging route.

In these circumstances the distance from which a proceed aspect on P468 signal can be seen by approaching trains, and the distance from the signal to the junction, is sufficient to result in some trains being able to accelerate to speeds which could lead to derailment by overturning when passing through the junction.

2. Safety advice

Duty holders should take immediate steps, either operationally, or by technical means, to mitigate this risk.

3. Issued to:

Network Rail and transport undertakings who operate trains on the East Coast Main Line through Peterborough station.

4. Background

On 17 April 2022, the driver of a Lumo service from Newcastle to London King’s Cross did not observe and react to the junction indicator on P468 signal which showed that the train was being signalled towards the slow lines at Spital Junction. Believing that his train was going to stay on the Up Fast line, he accelerated the train towards the junction and passed over it at 76 mph (122 km/h) instead of the 30 mph (48 km/h) maximum permitted speed.

On 4 May 2023, the driver of a Grand Central service from Sunderland to London King’s Cross did not observe and react to the junction indicator on P468 signal which showed that the train was being signalled towards the slow lines at Spital Junction. Believing that his train was to stay on the Up Fast line, he accelerated his train towards the junction and passed over it at 65 mph (105 km/h).

(If you look at the timeline of RAIB's outputs you'll see why I put quotes round "urgent".)

[stuving]

RAIB^▸ (Rail Accident Investigation Branch) have today announced an investigation into the second overspeed incident (of a GC^» (Great Central Railway - link to heritage line) train on 4th May 2023). The statement refers to the earlier incident, and the urgent safety notice, but does not say anything about the investigations being combined, or even that the second one will rely heavily on the first.

Station CCTV^▸ (Closed Circuit Tele Vision) image of the train slowing down at Peterborough platform 1 following the overspeeding incident (courtesy of LNER^» (London North Eastern Railway - about)).

At around 13:00 hrs on 4 May 2023, the 09:54 hrs Sunderland to King’s Cross service, operated by Grand Central, passed over three sets of points at Spital Junction on the northern approach to Peterborough station at a speed of around 65 mph (104 km/h). This was above the maximum permitted speed over the junction which is initially 30 mph (48 km/h) reducing to 25 mph (40 km/h).

The excessive speed over the points led to the train lurching sideways. This resulted in a number of minor injuries being sustained by passengers, with some being thrown from their seats.

The train subsequently came to a stand at Peterborough station under emergency braking. The train did not derail during the incident, and subsequent inspections found that no damage was caused to the infrastructure of the railway, or to the vehicles.

The train involved in the incident was not due to call at Peterborough. It had been approaching the station on a fast line, before being routed onto a slower line via the points where the overspeed occurred. The route onto this slower line was being displayed by the junction signal situated on the approach to the points.

[stuving]

The RAIB^▸ (Rail Accident Investigation Branch) report into the Lumo overspeeding incident has just been published:

Report 06/2023: Train overspeeding at Spital Junction

Train overspeeding at Spital Junction, Peterborough, 17 April 2022.

Summary

At around 10:20 hrs on 17 April 2022, the 08:20 hrs Lumo service from Newcastle to London King’s Cross, passed over three sets of points at Spital Junction at the northern approach to Peterborough station at excessive speed. The maximum permitted speed over the junction is initially 30 mph (48 km/h) reducing to 25 mph (40 km/h). The data recorder from the train indicated that the points had been traversed at a speed of 76 mph (122 km/h).

The speed of the train over the junction resulted in sudden sideways movements of the coaches. This led to some passengers being thrown from their seats and luggage falling from the overhead storage, with some passengers receiving minor injuries. Although the train did not derail, and no damage was caused, post-incident analysis has indicated that the train was close to a speed that would have led to it overturning, and it was likely that some of the wheels of the vehicles lifted off the rails.

RAIB’s investigation found that the overspeeding was caused by the driver of train 1Y80 not reacting appropriately to the signal indication they had received on approach to the junction. This signal indication was a warning that the train was to take a diverging route ahead which had a lower speed limit than the straight-ahead route which they were expecting to take. The driver’s awareness of the signal conditions that could be presented on approach to this junction and their training were not sufficient to overcome this expectation.

RAIB found that Lumo had not assessed and controlled the risk associated with trains being unexpectedly routed on a slower, diverging route at this location and that it had not adequately trained the driver to prepare for this eventuality. Network Rail had also neither assessed nor effectively controlled the risk of overspeeding at locations where there is a long distance between the protecting signal and the junction itself. The investigation also found that half of the passenger injuries were as a result of falling luggage that had been stowed in the overhead luggage racks.
Recommendations

RAIB has made four recommendations. The first recommendation is for Lumo to review its processes to ensure that it effectively controls the risk of overspeeding at diverging junctions. The second recommendation asks Network Rail to identify junctions where there is a greater potential for overspeeding to occur and to work with operators to share information on the associated risks. The third recommendation asks Network Rail and train operators to consider and implement risk control measures at those junctions identified in the second recommendation. The fourth recommendation is intended to ensure that Lumo minimises the risks from falling luggage on its services.

RAIB has also identified two learning points. These relate to the need for drivers to maintain alertness when approaching junction signals and that train operator emergency plans should specifically include processes to deal with the aftermath of overspeeding incidents.
May 2023 overspeeding incident

At around 13:00 hrs on 4 May 2023, another overspeeding incident occurred at the same location involving a Grand Central service. As a result, RAIB issued urgent safety advice to the industry and announced its intention to investigate this second incident. Further details can be found at paragraph 195 of this report.

Published 10 July 2023

[ChrisB]

Simulator training insufficiently detailed for that stretch of track?

[bradshaw]

What surprised me was that the signal could clear from red directly to green with that diverging route having such a low speed restriction. I feel that there might be a case to hold the advanced signal on platform one at red until P468 had been passed.

[stuving]

Simulator training insufficiently detailed for that stretch of track?

Not exactly. Lumo didn't use a simulator for that purpose (gaining experience of all the things the signals might do at a given place), and indeed have never had one of their own.

But I'm afraid this one is so complicated that reading the whole report is a necessary condition of making a meaningful comment on it. Sorry about that!

[ChrisB]

In that case, 3 stills from the CCTV^▸ (Closed Circuit Tele Vision)?....

[Electric train]

At first I thought surely TPWS^▸ (Train Protection and Warning System) would have prevented this, it is interesting to read in the report of an exception NR^» (Network Rail - home page) has in place.

Crudely in my simplistic view there were 2 factors came into play human error, the driver training / route knowledge, and the lack of and automated system to reduce human error, TPWS.

[stuving]

Cases like this make me think of how signalling is always designed to fail safe wherever possible, but the signalling does nothing in the real world until the train driver acts on it. And that driver can't be made to fail safe; here we are back to procedural safety rather than engineered or inherent safety.

ETCS^▸ (European Train Control System)/ERTMS^▸ (European Rail Traffic Management System.) will be a big step up in this area, so it's a pity that (as this report reminds us) something that it was hoped would be on all high-speed main lines by 2010 is still .... very much not. And its first big implementation, here on the ECML^▸ (East Coast Main Line), they now say will be in 2029 at the earliest. High-speed, eh?

[IndustryInsider]

We do of course have the ageing ATP^▸ (Automatic Train Protection) on the busiest and fastest parts of the GWML^▸ (Great Western Main Line), even if it’s a bit ‘windy’.  Though the similar system on Chiltern has now been switched off

[paul7575]

NR^» (Network Rail - home page) have temporarily modified the signal sequence since though, (report paragraph 198), now if the route is set through one of the various junctions the driver will only see a yellow, so shouldn’t accelerate to full speed.

As it was before when the approach control to a red was used the signal then went straight to green.  (As mentioned by Bradshaw in his post #21 yesterday.)  So it looks like a relatively straightforward change was possible.

Paul

[Electric train]

RIAB youtube video detailing the sequence of events  https://www.youtube.com/watch?v=w3D4m9--2h0

[stuving]

The RAIB^▸ (Rail Accident Investigation Branch) report into the second incident, involving a Grand Central class 180, has been released.

Report 10/2024: Overspeed at Spital Junction, Peterborough, 4 May 2023.

Summary

At around 13:00 hrs on 4 May 2023, the 09:54 hrs Sunderland to London King’s Cross Grand Central service passed over three sets of points forming part of Spital Junction at excessive speed. The maximum permitted speed over the junction, which is to the north of Peterborough station, is initially 30 mph (48 km/h) reducing to 25 mph (40 km/h). The data recorder from the train indicated that the points had been traversed at a speed of 66 mph (106 km/h).

The speed of the train over the junction resulted in sudden sideways movements of the coaches. This led to some passengers being thrown from their seats, with some receiving minor injuries.

RAIB’s investigation found that the overspeeding was caused by the driver of the train not reacting appropriately to the signal indication they had received on the approach to the junction. This signal was indicating that the train was to take a diverging route ahead which had a lower speed limit than the straight-ahead route. The driver’s expectation was that the train was being routed straight ahead and their application of driving awareness skills was not sufficient to overcome that expectation.

UK^▸ (United Kingdom) railway signalling principles mean that the control of speed at diverging junctions such as this is dependent on drivers reacting to signal information given at considerable distances. This, and exemptions granted in the past from fitting engineered protective measures beyond the signal, places the reliance on drivers correctly observing and responding to all the information given by the signal. This was a factor in this incident.

Testing and analysis by RAIB also found that the junction indicator element of the signal may not have been as conspicuous as the main aspect of the signal at the point the driver observed and reacted to the signal. This is a possible factor.

Three underlying factors were identified by the investigation. Grand Central had not provided the driver with the necessary non-technical skills or additional strategies to manage the risk present at this signal. This was a possible underlying factor. Network Rail and East Coast Main Line train operators had not effectively controlled the risk of overspeeding at this junction both at the time the signal’s operation was changed in 2013 and following a previous overspeeding incident at the same location in April 2022. Thirdly, Network Rail does not control the risk of overspeeding at locations where there is a long distance between the approach released protecting signal and the junction itself, once a proceed aspect has been given to drivers.

RAIB observed that Grand Central had not identified the risks associated with the signal in its route risk assessment and was not managing the development plans for the driver in accordance with its own processes. RAIB also observed that Network Rail’s reliability centred maintenance regime does not include a means to effectively manage degradation of junction indicator modules fitted with light emitting diodes (LEDs).

A similar incident occurred at the same location (RAIB report 06/2023), 13 months before this incident. Following this more recent incident, RAIB issued urgent safety advice in May 2023 to Network Rail and operators of trains on the East Coast Main Line through Peterborough station. This advice alerted them that suitable arrangements may not be in place to mitigate the risk of trains travelling southbound through Spital Junction at excessive speeds when signalled from the Up Fast line on to the Up Slow lines at Peterborough station. Duty holders were advised that they should take immediate steps, either operationally, or by technical means, to mitigate this risk.
Recommendations

RAIB has made four new recommendations as a result of this investigation. The first recommendation is for Grand Central to review and amend its training and competence management processes to provide its drivers with the necessary non-technical skills or additional strategies to manage the risk encountered at signals which may show different aspects to those usually encountered.

The second recommendation asks Network Rail and train operators to review the processes by which they derive, share and implement safety learning from accidents and incidents that involve shared risks across organisations.

The third recommendation is for RSSB^▸ (Rail Safety and Standards Board) to review the standards specifying the relative brightness of main aspects and junction indicators on signals to understand the effects on conspicuity of the complete signal up to the maximum distance at which a signal is required to be readable, to minimise the risk of drivers not correctly reading signals.

The fourth recommendation, arising from an observation, is for Network Rail to manage the risk of a driver not seeing a route indication because of the gradual reduction in light output of LED signals, which occurs over time.

RAIB has also identified two learning points during the investigation. The first relates to train operators ensuring that their route risk assessments include the risks to their services from signals which may show different aspects to those usually encountered. The second reminds transport undertakings of the importance of managing the competence of safety critical staff effectively and in accordance with their own processes.

Also mentioned within this report are the learning points from RAIB’s investigation into a previous incident at this location on 17 April 2022. These relate to the need for train operators to ensure that drivers maintain alertness when approaching junction signals and that train operator emergency plans should specifically include processes to deal with the aftermath of overspeeding incidents.

The incident itself was almost identical to the first one, specifically in that the sequence of signal aspects seen by the driver was the same (not always the case here). However, the report and its conclusions are very different. That's probably because the main explanation given before, inadequate training and preparation of the driver, doesn't fit this case. In addition, with two such very similar occurrences, the signalling was looked at much more critically.

One point in the report, but not leading to a recommendation, is arguably the main issue here. The driver saw the relevant signal clear to green, but did not see the junction indicator above it. That JI was the only unambiguous indication to the driver that a diverging route had been set.