RAIB report into 2013 Norwich collision

[fsmr]

Report into the Norwich station collision now published discussed originally here but now locked

http://www.railforums.co.uk/showthread.php?t=87042&highlight=Norwich


Report can be downloaded below
Some interesting observations focusing more and more on suitability of candidates in critical positions from the start. It would seem that reading between the lines, if you are of a deposition not suited to the role, then it should be picked up and presumably the driver moved off that role as it seems there is little in the way of training to remove the risk of microsleeps or periods of concentration loss save for the usual removal of distractions and ensuring adequate breaks and sleep

http://www.raib.gov.uk/publications/investigation_reports/reports_2014/report092014.cfm

The RAIB concluded that the accident occurred because during the last 20 seconds of the driver’s approach to the station, he either had a lapse in concentration or a microsleep.
The RAIB identified some factors which may explain the driver’s possible lapse in concentration (ie the noise made by the passengers immediately behind his cab and the various thoughts occupying his attention at the time of the approach). The RAIB also found that the driver had a previous operational history indicative that he was prone to lapses in concentration, and that this had not been identified by Greater Anglia’s competence management system.
Greater Anglia’s investigations of the previous incidents that the driver had been involved in had not raised any concerns about the driver’s ability to maintain concentration. This was because the driver manager who carried out the investigation had not been trained to consider that incidents, seemingly different in nature, could be linked by underlying behavioural issues. Opportunities to formally review the driver’s operational history were missed

 

[westcoaster]

Quiet some record that driver had, and lots of missed opportunities.

 

[Pumbaa]

Quite a record:

The RAIB reviewed the records of the driver’s career covering the period from when he joined British Rail in 1984 to the accident at Norwich on 21 July 2013. Since becoming a driver in 1989, the driver had been involved in more than
20 incidents. He had been held responsible by his employer for 14 of these incidents. These included signals passed at danger, speeding incidents, failures to call at a station, station overruns, stopping short at a station, an unscheduled stop at a station, AWS/TPWS interventions, a door incident and accepting a wrong route. The RAIB has been unable to source reliable data to compare
this driver with others in terms of operational performance. However, after the accident, Greater Anglia described the rate of one incident every two years (the driver had 14 incidents in 24 years) as not typical of even an average driver.

The RAIB analysed the OTDR records for the driver’s last two return journeys
on the day of the accident. This analysis showed that the driver exceeded the maximum permitted line speed on more than 10 separate occasions in the space of three hours. This erratic style of driving is indicative of either a driver prone
to lapses in concentration or of one unwilling to comply with rules. Given that Greater Anglia had no concerns regarding the driver’s willingness to comply with rules, it is likely indicative of the driver’s propensity to lapses in concentration and is particularly significant as the driver was on a development plan for a previous speeding incident at the time (paragraphs 78d and 88).

The report is also (subtly) critical of the lack of non-technical skills afforded to driver managers. Lots of missed chances.

 

[carriageline]

Is one incident every two years, REALLY that bad? Genuine question by the way, seeing as how many things he must of done that didn't end up in an operational incident?


Sent from my iPhone using Tapatalk

 

[jopsuk]

Well the TOC appear to be of the opinion that it is an abnormally high rate? And they should be able to back that up with stats.

 

[A-driver]

Is one incident every two years, REALLY that bad? Genuine question by the way, seeing as how many things he must of done that didn't end up in an operational incident?


Sent from my iPhone using Tapatalk

Depends on the incident. I havnt read the full report so don't know what the incidents were exactly but there are plenty of drivers who havnt had a single incident in their careers of 10+ years. Obviously near misses arnt always reported and I don't know what they are defining as 'speeding incidents' here...I'm sure every experienced driver in the world has on a few occasions looked at the speedo and seen that they are creeping up to 95 in a 90mph stretch of track etc and promptly brought the train back under control but that is not dangerous and very different to hitting a 20mph set of points at 45...

If they are seriousness incidents then I would say that it is abnormally high. Many TOCs get very concerned if you have 2-3 incidents classed as SOL such as SPAD, TPWS intervention (as opposed to activation unless that is the wrong way round, I can never remember!), stop short, wrong side release, overrun, speeding etc in a period of 2-5 years but are more forgiving of more minor incidents such as fails to call, stopping out of course, TPWS activations (again may be the wrong way round...), missed AWS etc.

I'm surprised that the report includes AWS, fails to call and stopping out of course actually as personally I don't count them as SOL and it looks like they are trying to prove how bad the drivers record was. I can see where they are coming from with lapses of concentration but without full details of the minor incidents it's hard to know if they all were concentration related-plenty of fails to calls happen to alert drivers simply misreading a diagram, missed AWS is often caused on some units by a sticky reset button or an open window drowning out the noise of the alarm etc but obviously these incidents are combined with more serious problems according to the report.

Obviously the RAIB just states the facts as it sees them and what we don't (and will never) know is the exact nature of the other incidents and how the driver and company dealt with them. It states the driver was on a plan so It seems the TOC were closely monitoring the driver.

It does highlight the seriousness of fatigue in these kinds of jobs. There really is nothing worse than getting half way through your job and finding yourself seriously fatigued to the extent that you actually think to yourself 'I don't know how I am going to get through the rest of the day'. Finding yourself struggling on the approach to stations, fanning the breaks, having to think really hard about the simplest elements of the job etc... It's why train driving is more than just sitting there pushing a few buttons and pulling a lever. It's not easy to break out of that kind of tiredness and force yourself to apply 100% concentration when in that kind of state. Unfortunately some people are completely unable to do so and they then end up having 'silly, incidents because of it and as I can understand it, this collision at Norwich was exactly that...a 'silly' incident caused by a lapse in concentration.

 

[edwin_m]

The driver also had a condition that required several prescriptions over the years of a medicine, not identified in the report but stated to be something that might cause drowsiness. TOC didn't appear to manage this too well - for the most recent prescription they sent him for assessment before he'd built up to the full dose - but we are not told whether they did anything or even knew about the previous prescriptions and whether there was any correlation with the driver's incident record.

 

[BestWestern]

The driver also had a condition that required several prescriptions over the years of a medicine, not identified in the report but stated to be something that might cause drowsiness. TOC didn't appear to manage this too well - for the most recent prescription they sent him for assessment before he'd built up to the full dose - but we are not told whether they did anything or even knew about the previous prescriptions and whether there was any correlation with the driver's incident record.

That is very poor management by the sounds of it, and really quite worrying.

 

[SPADTrap]

"Witness evidence indicates that the briefing note was displayed in the DTMs office at Norwich station too, but there was no signature sheet for DTMs to confirm that they had been briefed. The DTM who was on duty at the time of the accident on 21 July 2013 stated that he was unaware of this briefing note"

This is what needs to change.

 

[Lockwood]

TPWS intervention
TPWS activations

What's the difference between these?

 

[Llama]

TPWS interventions are seen as more serious because the system has intervened to bring the train to a stand, with the driver having taken no action prior to the TPWS brake demand.

Activations are when the driver has attempted corrective action prior to the brake demand, but the TPWS has still activated.

 

[A-driver]

TPWS interventions are seen as more serious because the system has intervened to bring the train to a stand, with the driver having taken no action prior to the TPWS brake demand.

Activations are when the driver has attempted corrective action prior to the brake demand, but the TPWS has still activated.

Did I get it the right war round then? Can never remember which is which!

But as you say, one is basically seen as the system stepping in as the driver appears not to have acted (ie not be braking for the speed board or red) and the other is take far less seriously in most cases like when a MU driver with sharp brakes slows towards a speed restriction but hits the approach grids a couple of mph above trigger speed. If tpws hadn't been fitted then the train would still have been well within the speed limit-many grids are set for freight trains with slower brakes.

 

[Llama]

There are different timings for passenger and freight, but only crudely represented by the fact the timers on the unit or loco differ - on all multiple units and locos with the brake distributor set to 'passenger' the system activates if the passage of arming and trigger loops is 0.974s or less, on locos with freight brake timings the activation time is 1.218 seconds.
I would suggest that the % difference in those timings is a bit restrictive for some units, ie modern disc braked units could probably easily end up being tripped by some TPWS installations if driven within their normal capabilities. For the poorest performing unit brakes (14x vehicles) the timing might be about right, so the 'lowest common denominator' seems to rule.

 

[A-driver]

There are different timings for passenger and freight, but only crudely represented by the fact the timers on the unit or loco differ - on all multiple units and locos with the brake distributor set to 'passenger' the system activates if the passage of arming and trigger loops is 0.974s or less, on locos with freight brake timings the activation time is 1.218 seconds.
I would suggest that the % difference in those timings is a bit restrictive for some units, ie modern disc braked units could probably easily end up being tripped by some TPWS installations if driven within their normal capabilities. For the poorest performing unit brakes (14x vehicles) the timing might be about right, so the 'lowest common denominator' seems to rule.

I didn't know they were set differently so thanks for that-very interesting.

A number of grids are certainly over cautious for many EMUs and modern driving - I know a few drivers who have been well within tolerances for a speed restriction but still been 1mph or so over.

Of course if NR would finally agree to publish loop trigger speeds then it would be more helpful...(although I can see why they won't equally).

 

[Llama]

There has been a policy over the last few months at my TOC leaning towards publishing 'set speeds' (or at least passenger set speeds!) wherever there has been an incident.

And definitely agree that some of the installations are too restrictive, I can think of one 'multi-trip' location where a set of OSS loops on a 60mph approach protecting a 30 PSR (OSS set at 48) would actually stop a train significantly before the PSR commencement even at line speed.

 

[carriageline]

I know on one route (have a feeling it's Wessex, but may be completely wrong) the TPWS trigger speeds have been handed out to/by the TOC in question because, as you say, they are not very giving at times!

 

[edwin_m]

"Witness evidence indicates that the briefing note was displayed in the DTMs office at Norwich station too, but there was no signature sheet for DTMs to confirm that they had been briefed. The DTM who was on duty at the time of the accident on 21 July 2013 stated that he was unaware of this briefing note"

This is what needs to change.

That might have prevented the accident in question (by making sure that there was no attempt to put six cars into the platform that wasn't quite long enough). However the length of the platform is actually a red herring here - the driver was aware of the trains already in the platform and knew where to stop but did not do so. The same thing could equally have happened in any other platform and if it didn't, with the driver's record of less serious incidents he would possibly have got into some sort of serious trouble sooner or later anyway.

 

[TorqueOfTheDevil]

Is one incident every two years, REALLY that bad? Genuine question by the way, seeing as how many things he must of done that didn't end up in an operational incident?

I can't give you a typical rate for train drivers, but presumably one might expect that somebody of his experience would be less accident-prone. And the fact that none of these episodes had a disastrous outcome doesn't make it okay!

The issue of performance being affected by a medical condition and/or medication is an interesting one. I work in aviation, and we are completely forbidden to fly if we have consumed any medication at all (including alternative remedies, dietary supplements etc), unless the medication has been prescribed by a doctor who explicitly cleared us to fly while taking it. Now some people with a fair degree of experience are confident that (for example) having a single painkiller for a mild headache won't affect their performance much, and will take one anyway rather than stay on the ground and go to the doc, but the management are quite clear: if we are involved in an incident/accident and it turns out that we have been self-medicating, they will wash their hands of us. I wonder if such stark guidelines will permeate the railway world in due course...

 

[TDK]

Is one incident every two years, REALLY that bad? Genuine question by the way, seeing as how many things he must of done that didn't end up in an operational incident?


Sent from my iPhone using Tapatalk

You need to remember that one incident every 2 years identified and proved is bad as how many other times has a driver had incidents that have not been reported or detected?

 

[SansPareil]

"Witness evidence indicates that the briefing note was displayed in the DTMs office at Norwich station too, but there was no signature sheet for DTMs to confirm that they had been briefed. The DTM who was on duty at the time of the accident on 21 July 2013 stated that he was unaware of this briefing note"

This is what needs to change.

Not the first time an unsigned note at Norwich station has had a role in an accident.

 

[A-driver]

I can't give you a typical rate for train drivers, but presumably one might expect that somebody of his experience would be less accident-prone. And the fact that none of these episodes had a disastrous outcome doesn't make it okay!



The issue of performance being affected by a medical condition and/or medication is an interesting one. I work in aviation, and we are completely forbidden to fly if we have consumed any medication at all (including alternative remedies, dietary supplements etc), unless the medication has been prescribed by a doctor who explicitly cleared us to fly while taking it. Now some people with a fair degree of experience are confident that (for example) having a single painkiller for a mild headache won't affect their performance much, and will take one anyway rather than stay on the ground and go to the doc, but the management are quite clear: if we are involved in an incident/accident and it turns out that we have been self-medicating, they will wash their hands of us. I wonder if such stark guidelines will permeate the railway world in due course...

Such stark guidelines are in force on the railway. I don't know specifics of this incident but it seems that either this driver wasn't declaring it or management were ignoring it.

If I take so much as a strong lemsip then I tell the foreman when booking on so it's on a recorded phone line. If they tell me I can't drive on it then I won't. If they say it's fine then I carry on. They tell bupa of any medication we are taking-self or doctor prescribed-and bupa tell them if we are ok to drive on that particular medication. Even if you have taken it before and know it's only a minor drug like a paracetamol we are still told to declare it and the decision as to weather we can drive goes down to management shifting and responsibility onto them.

 

[SansPareil]

Is there any mechanism, that won't reflect negatively on a driver, for informing management of a disturbed nights sleep and feeling tired. Say if neighbours had a late party, or roadworks, which prevented a driver being properly rested before a shift.

 

[Bald Rick]

Is one incident every two years, REALLY that bad? Genuine question by the way, seeing as how many things he must of done that didn't end up in an operational incident?

That would be approx one incident per 400 driving shifts. Given that there are in the region of 8,000 driving shifts a day (very approximately), if this were a normal rate that would equate to 20 similar incidents a day on the network. And there are nowhere near that many.

Qualifier - I've not read the report and thus don't know how serious this chaps previous was.

 

[A-driver]

Is there any mechanism, that won't reflect negatively on a driver, for informing management of a disturbed nights sleep and feeling tired. Say if neighbours had a late party, or roadworks, which prevented a driver being properly rested before a shift.

Not any more than there is for anyone else who works shifts (or works normal hours come to that...)

If you really don't feel that you can safely work then you would be best phoning in sick but if you made a habit of it then you may need to consider if it's the right job for you.

 

[DarloRich]

I can't give you a typical rate for train drivers, but presumably one might expect that somebody of his experience would be less accident-prone. And the fact that none of these episodes had a disastrous outcome doesn't make it okay!

The issue of performance being affected by a medical condition and/or medication is an interesting one. I work in aviation, and we are completely forbidden to fly if we have consumed any medication at all (including alternative remedies, dietary supplements etc), unless the medication has been prescribed by a doctor who explicitly cleared us to fly while taking it. Now some people with a fair degree of experience are confident that (for example) having a single painkiller for a mild headache won't affect their performance much, and will take one anyway rather than stay on the ground and go to the doc, but the management are quite clear: if we are involved in an incident/accident and it turns out that we have been self-medicating, they will wash their hands of us. I wonder if such stark guidelines will permeate the railway world in due course...

They all ready exist in an almost similar level of starkness!

 

[G0ORC]

127 On 19 August 2013, Greater Anglia re-issued its briefing note to all station staff (supervisors and DTM) on the length of each platform at Norwich station and the maximum number of vehicles that can be accommodated. Station supervisors and DTMs were asked to sign a form to confirm that they had read and understood the briefing note. The RAIB has been informed that since 19 August 2013 there have been further reported instances when six vehicles have been accommodated in platform 6. The RAIB observes that the various briefings notes have not been entirely adhered to and invites Great Anglia to clearly communicate the output of recommendation 5 of this report to all relevant employees

A very interesting report which I find particularly worrying - I can assure the non-railway employees on here that the number of incidents that this individual has been involved it VERY high. His manager, relatively untrained in root cause analysis, failed to spot the underlying cause to many of them. I'm sure Anglia quickly reviewed and amended their Driver Management System a long while before this report was released in order to deal with the issues raised by this incident.

Oh, and yes stopping short, station over-runs, missed calls and wrong station calls are considered safety of the line incidents and rightly so. It indicates failings in an area of the driver's competence that need to be addressed which, if left to go unchecked could lead to more serious incidents.

What concerns me even more is the paragraph I have highlighted above.

How, given the high profile of this incident, and instructions to the contrary, have station staff STILL requested trains to be allowed into platforms which have insufficient room to accommodate them? This indicates to me a "them and us culture" with a reluctance of the station management to accept that they played a role in this incident too.

Clearly there are major operating problems which affect the working of Norwich station. As the number of trains using it has increased, the demands on platform occupation has increased the risk of something like this happening. It is easy with hindsight to say that this could and should have been identified and mitigating measures introduced.

But apparently refusing to implement a simple and common sense instruction which does nothing more than tell the staff to accept no more vehicles into a platform than the platform can fully accommodate is something that I hope senior operations management at Anglia have already dealt with.

As far as I know SDO is not fitted to Class 156 units, so an the instruction to have the conductor standing by a door which was going to be off the platform, is very questionable as a safe method of working. Think of of all of the possible permutations which could lead to the conductor, having released the doors, not being able to stop someone (perhaps drunk and/or aggressive, as many of the passengers on this train appeared to be) using the door and falling out of a train from a door not on the platform.

 

[edwin_m]

The report says the last door was on the platform ramp. So if they really need six cars in this platform it could be rebuilt with a square end where the bottom of the ramp now is.

 

[kieron]

As far as I know SDO is not fitted to Class 156 units, so an the instruction to have the conductor standing by a door which was going to be off the platform, is very questionable as a safe method of working.

It's what they do at the bottom line platforms at Shotton when a 5 or 6 car 175 calls there. There's a big difference between doing something like this with management approval (with the risk assessment and training that implies) and doing it on an ad hoc basis when there aren't enough free platforms of a suitable length, though.

 

[Class 170101]

However the length of the platform is actually a red herring here - the driver was aware of the trains already in the platform and knew where to stop but did not do so.

The driver would have been advised by the signalling the platform was occupied. A driver cannot tell from thr signalling how many carriages are in the platform just that he is being signalled into an occupied platform.

 

[A-driver]

The driver would have been advised by the signalling the platform was occupied. A driver cannot tell from thr signalling how many carriages are in the platform just that he is being signalled into an occupied platform.

To be honest, I ignored the post which you quoted as regardless of signalling/ road knowledge or any other info the driver may have it's fairly obvious that 'where to stop' is before colliding with another train...

This seems to have been a concentration lapse rather than incompetence or ignorance of the situation.