Overspeeding incident at Grantham South Junction (RAIB investigation) -- 25/2/25

[Jimini]

BBC report

Staff onboard a train were left hurt after it travelled through a junction near a market town at more than twice the speed limit, investigators found.
An LNER train service travelled at about 55mph (87km/h) through Grantham South Junction, in Lincolnshire, where a speed restriction of 25mph (40km/h) was in place, the Rail Accident Investigations Branch (RAIB) said.
While staff onboard reported receiving minor injuries during the incident on 25 February, no passengers were injured, investigators had noted.
Train operator LNER said it had supported RAIB's findings and the safety of its staff and passengers was its "top priority".
The train, which was travelling from Hull to London Kings Cross, had stopped at Grantham station before passing through the junction at about 08:25 GMT.
After leaving the station, a signal warned the driver the train was switching to a slower railway line.
In its findings, the RAIB said there was a "strong likelihood" factors in this case were similar to those identified in two separate train speeding incidents at Spital Junction, in Peterborough, on 17 April 2022 and on 4 May 2023.
In both of those incidents, passengers had been left injured.
Following the preliminary examination, the RAIB said it had written to the Office of Rail and Road to draw attention to the incident and copied the letter to Network Rail, LNER and the Rail Safety and Standards Board.

'Ever busier railway'​

The report stated: "This incident again illustrates the issues associated with relying completely on train drivers reacting appropriately to a junction or route indicator to control the risks presented by trains taking diverging low-speed turnouts on high-speed through routes.
"This risk may be increased by the introduction of higher performing train fleets, and possibly by routing patterns on an ever busier railway."
An LNER spokesperson said: "The safety of our customers and colleagues is our top priority.
"We have supported RAIB in their preliminary examination, note the recommendations made and will continue to work with partners in the industry."

RAIB incident report

At around 08:25 on 25 February 2025, train reporting number 1A12, the 0700 Hull to London Kings Cross service, operated by LNER, was involved in an overspeeding incident at Grantham South Junction.

The train departed after a planned stop at Grantham station with the signal indicating that the train was to diverge from the main line onto a parallel slow line at Grantham South Junction. This divergence has a permanent speed restriction of 25 mph (40 km/h). However, train 1A12 was travelling at around 55 mph (87 km/h) when it traversed the junction. Staff on board reported receiving minor injuries due to the overspeed, although there were no reported passenger injuries. The permissible speed in this area for a train which remains on the main line is 115 mph (184 km/h).

RAIB was notified of the incident soon after it occurred. We have since gathered evidence from the railway industry and carried out a preliminary examination into the circumstances surrounding the incident.

Our preliminary examination found that there is a strong likelihood that factors were present in this incident that were similar to those identified during RAIB’s investigations into a train overspeeding at Spital Junction, Peterborough, 17 April 2022, (report 06/2023) and a similar event at the same location on 4 May 2023 (report 10/2024). Some of the recommendations from these investigations remain open and we have concluded it is unlikely that further investigation will lead to new recommendations for the improvement of railway safety. Consequently, RAIB will not investigate further or produce an investigation report.

However, this incident again illustrates the issues associated with relying completely on train drivers reacting appropriately to a junction or route indicator to control the risks presented by trains taking diverging low-speed turnouts on high-speed through routes. This risk may be increased by the introduction of higher performing train fleets, and possibly by routing patterns on an ever busier railway.

RAIB has written to the Office of Rail and Road, to draw its attention to this incident when considering industry responses to the recommendations made in the Spital Junction reports and the wider questions within the industry around protection against overspeeding.

We have copied the letter to Network Rail, LNER and the Rail Safety and Standards Board so that they are aware of the contents.

 

[FrontSideBus]

Oops.

 

[75A]

A case of always expect the unexpected.

 

[Tram203]

A case of always expect the unexpected.

Was it the unexpected?

I believe 1A12 is routed on to the up slow on departure from Grantham to allow 1E01 - The Flying Scotsman 0540 ex Edinburgh - to overtake prior to Stoke Tunnel.

 

[Unobrow]

Was it the unexpected?

I believe 1A12 is routed on to the up slow on departure from Grantham to allow 1E01 - The Flying Scotsman 0540 ex Edinburgh - to overtake prior to Stoke Tunnel.

The routing was expected, the speed wasn’t……….

 

[High Dyke]

For context. The signal for the turnout is situated at the South end of Platform one at Grantham. The points are just under half a mile south of there.

That's some going from a standing start. I don't know what the acceleration speed is for an Azuma.

 

[zwk500]

Another incident that ETCS would have intervened in to prevent/mitigate...

 

[Taunton]

An LNER spokesperson said: "The safety of our customers and colleagues is our top priority."

The old adage. When a company puts out a statement that "Safety is our first priority", it generally follows an incident which shows that it isn't.

 

[JordR]

The old adage. When a company puts out a statement that "Safety is our first priority", it generally follows an incident which shows that it isn't.

What else have LNER prioritised? There's no advantage to them in overspeeding through a junction.

 

[Haywain]

Reading the quotes in the opening post suggests that RAIB are strongly suggesting that it was driver error without wanting to actually say so.

 

[Watershed]

Reading the quotes in the opening post suggests that RAIB are strongly suggesting that it was driver error without wanting to actually say so.

Certainly if there was any fault with the infrastructure or train, you'd expect that to have been mentioned.

It's clear that technological safeguards are needed to prevent this kind of 'near miss' from escalating into an accident causing serious injury or worse. Adding ever more policies - which will inevitably slow services down as drivers are encouraged to be more cautious - will not solve the underlying problem here, which is that drivers are human and do make mistakes occasionally.

 

[mike57]

Adding ever more policies - which will inevitably slow services down as drivers are encouraged to be more cautious - will not solve the underlying problem here, which is that drivers are human and do make mistakes occasionally.

In most other industries on anything safety critical you have a 'plant control system' and a 'safety system' The plant control system runs the plant effectivley most of the time, and may make decisions based on human or other machine input. The safety system sits there monitoring critical plant parameters, separately and simply, and takes no information from the plant control system. If any of the critical operating parameters exceed certain thresholds then the safety system shuts the process down safely and independently. In rail terms the 'plant control system' is the driver, but the safety system does not have appeared to developed to the same exent on railways.

If I tried to get the HSE or ONR to accept a human driven control system for something safety critical with no (or limited) back up safety systems they would just say 'No way' or worse. Bearing in mind simple is best there are surely things that could be done, it feels like most of the technology is already there. I am not going to dive off into the possible engineering solutions, but signalling system knows the route set, and approach speed at a given point(s) is easily determined. A simple equation - If actual speed is 10% greater (or some other number) than expected speed based on current route and signalling state then apply brakes and a 3 minute timeout.

 

[43066]

Certainly if there was any fault with the infrastructure or train, you'd expect that to have been mentioned.

It's clear that technological safeguards are needed to prevent this kind of 'near miss' from escalating into an accident causing serious injury or worse. Adding ever more policies - which will inevitably slow services down as drivers are encouraged to be more cautious - will not solve the underlying problem here, which is that drivers are human and do make mistakes occasionally.

It’s equally clear that “technological safeguards” don’t stack up cost/benefit wise given how safe the railway is. ETCS will gradually eliminate these incidents as it rolls out.

 

[Crossover]

In most other industries on anything safety critical you have a 'plant control system' and a 'safety system' The plant control system runs the plant effectivley most of the time, and may make decisions based on human or other machine input. The safety system sits there monitoring critical plant parameters, separately and simply, and takes no information from the plant control system. If any of the critical operating parameters exceed certain thresholds then the safety system shuts the process down safely and independently. In rail terms the 'plant control system' is the driver, but the safety system does not have appeared to developed to the same exent on railways.

If I tried to get the HSE or ONR to accept a human driven control system for something safety critical with no (or limited) back up safety systems they would just say 'No way' or worse. Bearing in mind simple is best there are surely things that could be done, it feels like most of the technology is already there. I am not going to dive off into the possible engineering solutions, but signalling system knows the route set, and approach speed at a given point(s) is easily determined. A simple equation - If actual speed is 10% greater (or some other number) than expected speed based on current route and signalling state then apply brakes and a 3 minute timeout.

Your description sounds like ERTMS/ETCS levels of supervision, which is being (very slowly) worked towards

 

[Matt2725]

For context. The signal for the turnout is situated at the South end of Platform one at Grantham. The points are just under half a mile south of there.

That's some going from a standing start. I don't know what the acceleration speed is for an Azuma.

They can get a shift on from a stand I've found.

 

[RichSwitch]

More importantly: tea and biscuits, tea no biscuits, or no tea no biscuits?

 

[Egg Centric]

It’s equally clear that “technological safeguards” don’t stack up cost/benefit wise given how safe the railway is. ETCS will gradually eliminate these incidents as it rolls out.

1000 times this. Quick wins with little or no cost if they exist should obviously be seized upon but (UK) railways are too expensive for many people and already incredibly safe. We need to recognise that when it comes to rail, increasing safety is generally not where we should be allocating scarce resources.

 

[30907]

I am puzzled at the (booked) use of the Grantham loop rather than the Slow from Stoke Jn where there is a higher speed turnout - the timetable would allow it.
I see a potential for confusion when trains might be regulated at either location.

 

[mike57]

which is being (very slowly) worked towards

My view would be that it would be better to get on with it, rather than pile yet more pressure on to the person at the front.

To go back to my original reply, where correct human intervention is a requirement we take 1% as the human error percentage. I realise railways do far better than that, through training and the correct selection of people, but there will always be human errors, which is why we have safety systems. Again relating this to railways, the safety system is not automatic train control, it just sits there looking at a few simple parameters, in railway terms speed, location and signal state and will initiate a shutdown, i.e. remove power and apply brakes. On the systems I am involved in, in other industries, there is therefore no credible situation where the safety system can be the initiator of a safety breach. It cannot physicaly appliy power or release brakes in a railway analogy even if the safety system fails in some catastrophic and novel way. The only failure of the safety system is the probability of failure on demand, i.e. not take the required action to shutdown, and this is calculated based on system design and components used.

Another interesting thought is that for many years there was 'double manning', with steam the fireman, which carried over when steam was phased out. When did it cease? Did the second person in the cab help "You do know the route is set to take us on to the slows". It would be interesting to categorise incidents and accidents by cause and see if there was a change in percentages as double manning was phased out. I am not suggesting we return to that by the way.

 

[edwin_m]

These incidents haven't resulted in serious injuries or damage, demonstrating how much tolerance there is in turnout speeds before derailment becomes likely. But there have been several of them since the introduction of much higher performance trains on the ECML, plus one at Bletchley where a Class 90 was unusually running light so also had a very high acceleration. While ETCS will address this eventually, there may be a case to extend TPWS protection to sites where this could occur (approach controlled signal with a relatively long distance to a low speed turnout). This would be more complicated than a standard TPWS because it needs to know which route was set when the train passed the signal, either directly from the interlocking (route sticks or sub-routes) or by remembering the aspect of the signal before it reverted to danger as the train passed it. Also defining the positioning and setting of the overspeed sensor would need some new principles, because TPWS settings don't take account of the possibility of the train accelerating over the loops. However, a greater tolerance of overspeed should be acceptable compared with the standard speed restriction installation.

 

[AdamWW]

If I tried to get the HSE or ONR to accept a human driven control system for something safety critical with no (or limited) back up safety systems they would just say 'No way' or worse.

If you want to look at an area that would be shut down instantly if we applied the same safety standards as just about anything else in the UK try road transport.

I can walk right in front of a machine capable of suddenly moving towards me and causing serious injury if the operator makes a mistake yet it doesn't have to be locked and tagged out first.

 

[43066]

When did it cease? Did the second person in the cab help "You do know the route is set to take us on to the slows".

It ended in the early 90s AIUI (and at that stage was only for trains above 100mph).

I don’t think it was ever proved that having a second person made mistakes less likely and indeed the opposite approach is taken today. People other than the driver are only allowed into the cab under certain circumstances, and newly qualified drivers, and drivers on “action plans” after incidents, generally won’t be allowed anyone with them other than their manager when assessing them.

 

[High Dyke]

I am puzzled at the (booked) use of the Grantham loop rather than the Slow from Stoke Jn where there is a higher speed turnout - the timetable would allow it.
I see a potential for confusion when trains might be regulated at either location.

Probably the shorter section to Great Ponton, as previously mentioned. To be put onto the Up Slow at Stoke would mean running all the way on there until Tallington.

 

[800001]

I am puzzled at the (booked) use of the Grantham loop rather than the Slow from Stoke Jn where there is a higher speed turnout - the timetable would allow it.
I see a potential for confusion when trains might be regulated at either location.

The timetable only works by putting 1A12 in loop at the point, as 1E01 Flying Scotsman is right up behind it, to run it slows from Stoke Junction means running slow line all the way to Talkington which then causes clashes with other services between there and Kings Cross.

It is booked in that loop each and every weekday and has done for several years. There is a feather on the signal displaying it will be diverging into the loop.
There is no confusion if the driver pays attention to there route.

1A12 is also only driven (99%) of the time by the Doncaster drivers, so a small group who should all be aware that service gets looped!

 

[mike57]

try road transport

I agree, without going to far off topic... "So you plan on using a low flash point fuel, with a 1.4% LEL in air to propel your vehicle, and there will be several gallons of this fuel in a standard steel tank.The person operating this vehicle may have no understanding of the principles of operation. Come back with more acceptable proposals"

 

[Tetragon213]

A tad concerning to see similar things happen 3 times in 3 years. Seems like we've been lucky 3 times to not have either had a Nuneaton 2.0, or a passenger getting critically injured from having heavy luggage falling on top of them.

I'm sure ETCS will reduce the chances of this happening again, but in the meantime, it feels that something should probably be done while we're waiting on that to be rolled out. Maybe a TPWS grid to monitor speeds through the junction, if a diverging route is set?

 

[High Dyke]

A tad concerning to see similar things happen 3 times in 3 years. Seems like we've been lucky 3 times to not have either had a Nuneaton 2.0, or a passenger getting critically injured from having heavy luggage falling on top of them.

I'm sure ETCS will reduce the chances of this happening again, but in the meantime, it feels that something should probably be done while we're waiting on that to be rolled out. Maybe a TPWS grid to monitor speeds through the junction, if a diverging route is set?

A feasible idea to fit an additional overspeed grid, only activated when the route to the Up Slow is set/signalled. I fear that there may be some 'excuse' from Network Rail why they wouldn't be able to do this.

Historically, the train could've been routed onto the Up Slow just after the signal, but that connection was removed many years ago.

 

[Silverlinky]

A tad concerning to see similar things happen 3 times in 3 years. Seems like we've been lucky 3 times to not have either had a Nuneaton 2.0, or a passenger getting critically injured from having heavy luggage falling on top of them.

I'm sure ETCS will reduce the chances of this happening again, but in the meantime, it feels that something should probably be done while we're waiting on that to be rolled out. Maybe a TPWS grid to monitor speeds through the junction, if a diverging route is set?

There will be other similar incidents which RAIB may or may not be aware of.

 

[Tetragon213]

There will be other similar incidents which RAIB may or may not be aware of.

That makes it all the more eye-raising that no physical measures have been taken to either stop or at least the mitigate the consequences of overspeeds at these junctions under such circumstances.

A feasible idea to fit an additional overspeed grid, only activated when the route to the Up Slow is set/signalled. I fear that there may be some 'excuse' from Network Rail why they wouldn't be able to do this.

Historically, the train could've been routed onto the Up Slow just after the signal, but that connection was removed many years ago.

I'm equally fairly certain a weak excuse will be given. If Spital I and Spital II weren't enough to convince the powers that be that a more active form of prevention is required, what will convince them? TPWS would at least be reasonably quick to implement while we await ETCS.

 

[Silverlinky]

That makes it all the more eye-raising that no physical measures have been taken to either stop or at least the mitigate the consequences of overspeeds at these junctions under such circumstances.

I'm equally fairly certain a weak excuse will be given. If Spital I and Spital II weren't enough to convince the powers that be that a more active form of prevention is required, what will convince them? TPWS would at least be reasonably quick to implement while we await ETCS.

TPWS would not really be effective in these cases. How far back from the junction do the arming and trigger loops go? At that point the train would be travelling at a lower speed anyway. Using the Grantham incident as an example, it could be that the train might only be travelling at 40mph over the loops, does it need to be tripped then? The loops cant tell whether the train is already braking or in this case still accelerating.

I know of an incident at West London Junction where a train went over a crossing at 40mph (crossing speed 15mph). The distance between signal and junction was not that far but the acceleration of trains these days makes incidents like this more likely and perhaps more dangerous.

What will convince them? Perhaps only a derailment and loss of life?