Croydon Tram Accident

[bécasse]

There is an easy way to overcome the fear culture, which I suspect the inquiry has used, poll the workforce anonymously on safety issues, on a regular basis, using an independent agency.

 

 

I worked for a railway company which decided to poll the workforce anonymously on a whole range of issues including attitude to safety. A popular and well-respected personnel manager actually commissioned the poll and the company he selected didn't make a bad job of it (although I did have considerable professional standards misgivings about quoting verbatims from identified small teams, which risked the respondent concerned being identified or, even worse, mis-identified). However, the overall results from the survey were at complete variance with the views of the senior management team - how dare they say that our staff think that we are a load of w*nk*rs when clearly we are near perfect - and not only weren't they taken on board, but the personnel manager concerned found himself selected for compulsory redundancy only a few weeks later.

 

What I would add, though, is that my view the issue of what caused the driver to "micro-sleep" was a total red herring. What caused the accident was the failure to foresee that some form of signalling was required in the area of the tunnels. The first stage of the Bruxelles pre-metro (tram in tunnel) had opened c1970 and, despite the fact that the Bruxelles tramway system already had long stretches of reserved track, it was considered essential to install an elaborate signalling system (displays include not only red/yellow/green lights in combinations but maximum permitted speeds) with system overrides to stop a speeding tram. The system was particularly pertinent because at the southern end of the pre-metro tunnel the tram lines turned sharply - a situation very similar to that at Sandilands (today's replacement Metro goes straight on).

 

I would suggest, in fact, that the curve didn't just present a serious hazard in in its own right (as the accident has clearly proved), but also presented (and indeed continues to present) a potential serious rear-end collision hazard. Croydon trams normally run (other than on the common Sandilands-central Croydon section) at well-spaced intervals, but if delays were to occur at the Sandilands junction (and the juxtaposition of a trailing junction and highway crossing make it one of the more likely locations for delays on the network) the maps and photographs in the report demonstrate clearly that the driver of the following tram would have very little prior vision of the stationary tram ahead (and indeed might initially think that it was a tram coming the other way). Even with a 20 mph speed limit and the hazard braking facility a rear end collision would be difficult to avoid, limited signalling would avoid the risk. I am rather surprised that RAIB report seems to make no mention of the possibility.

[Coryton]

 

And for rail and bus

Basically the DfT went 'you rail people can try to be safe, we are going to ignore the safety of bus passengers'

 

Interesting.

 

Rather like when a cyclist is in a collision with a truck and the investigation concludes that it wasn't the driver's fault because trucks have big blind spots.

 

Which is perhaps fine in itself, but there seems to be no mechanism to ask whether perhaps trucks need to be designed differently.

 

That's not to say that a lot of effort hasn't gone into safety of road vehicles, and not all aimed at the occupants, but it seems to work in a different way to rail.

[Mad McCann]

Aye. Bus design certainly has a fair few shortcomings. There's no doubt that,  for example, an Optare Solo or Enviro 300 are comfortable and generally good to drive, but in terms of crash-worthiness they offer little real protection for passengers or driver. There's not much in front of you except glass and plastic. I guess, in many situations, the driver of a large tractor unit has at least his height above the most likely impact zone and the engine itself for protection.

 

D4

[Coryton]

Aye. Bus design certainly has a fair few shortcomings. There's no doubt that,  for example, an Optare Solo or Enviro 300 are comfortable and generally good to drive, but in terms of crash-worthiness they offer little real protection for passengers or driver. There's not much in front of you except glass and plastic. I guess, in many situations, the driver of a large tractor unit has at least his height above the most likely impact zone and the engine itself for protection.

 

Which is all very well for the driver, but it doesn't help them see cyclists (and maybe pedestrians) in the way.

 

Bus collisions seem to be a lot more rare than those involving cars so to some extent it seems reasonable to worry less about bus design. But a huge amount of effort goes into crashworthiness of trains which very rarely do actually crash.

[jim.snowdon]

Alas I cant see that doing anything as it is only visual the use of an audible warning that needs to be acknowledged in the same way as AWS is far better!

 

The thing that strikes me is that there is no joined up thinking between tram and train with each having to learn by its own mistakes!

 

Mark Saunders

Trams and trains are not the same things - they may both run on rails, but there is a completely different operational environment. To suggest that the same sort of safety measures that are appropriate for 125mph trains where the driver cannot even see the length of his stopping distance are appropriate for a line of sight tramway is suggestive of a lack of understanding of the tramway environment.

 

Jim

[jim.snowdon]

Whether the driver was at fault or not does not seem to have been proved one way or the other but the fact that they have established there is a fear culture needs to be stamped on very quickly and acted upon.

The purpose of the RAIB is to determine the cause, not to ascribe blame. That is for the ORR, as the safety authority and, where appropriate, the Police and any guilt or otherwise is only proven by a Court of Law.

 

Jim

[jim.snowdon]

What I would add, though, is that my view the issue of what caused the driver to "micro-sleep" was a total red herring. What caused the accident was the failure to foresee that some form of signalling was required in the area of the tunnels. The first stage of the Bruxelles pre-metro (tram in tunnel) had opened c1970 and, despite the fact that the Bruxelles tramway system already had long stretches of reserved track, it was considered essential to install an elaborate signalling system (displays include not only red/yellow/green lights in combinations but maximum permitted speeds) with system overrides to stop a speeding tram. The system was particularly pertinent because at the southern end of the pre-metro tunnel the tram lines turned sharply - a situation very similar to that at Sandilands (today's replacement Metro goes straight on).

 

I would suggest, in fact, that the curve didn't just present a serious hazard in in its own right (as the accident has clearly proved), but also presented (and indeed continues to present) a potential serious rear-end collision hazard. Croydon trams normally run (other than on the common Sandilands-central Croydon section) at well-spaced intervals, but if delays were to occur at the Sandilands junction (and the juxtaposition of a trailing junction and highway crossing make it one of the more likely locations for delays on the network) the maps and photographs in the report demonstrate clearly that the driver of the following tram would have very little prior vision of the stationary tram ahead (and indeed might initially think that it was a tram coming the other way). Even with a 20 mph speed limit and the hazard braking facility a rear end collision would be difficult to avoid, limited signalling would avoid the risk. I am rather surprised that RAIB report seems to make no mention of the possibility.

The RAIB is there to determine the cause of the event, not to suggest ways in which any deficiencies may be resolved. The most that they can legally do is identify that there is an issue that requires resolution and make a formal recommendation to that end, addressed to the appropriate party. It is up to the investigated party, ie the railway or tramway authority, to come up with what it considers an appropriate solution; the arbiter of whether the solution is appropriate, or even necessary, is the ORR (by way of its function as HM Railway Inspectorate).

 

As regards the Brussels pre-Metro, although I have not experienced it, I would suggest that it was treated as exactly that - a section of signalled Metro, not an extension of line of sight tramway operation. In that respect, it would be no different to, say, the Market Street tunnel section on the San Fransisco MUNI system, or the Amsterdam tram routes that run onto their metro system. Essentially, appropriate control methods for the circumstances. Sandilands tunnels are straight, with clear visibility from one end to the other, and maintaining line of sight operation is not inappropriate. Railway signalling is not the answer, but better provision of visual aids to tram drivers in terms of establishing their position along the length of the dark, off-street sections of the system probably is.

 

Jim

[chriswright03]

Aye. Bus design certainly has a fair few shortcomings. There's no doubt that,  for example, an Optare Solo or Enviro 300 are comfortable and generally good to drive, but in terms of crash-worthiness they offer little real protection for passengers or driver. There's not much in front of you except glass and plastic. I guess, in many situations, the driver of a large tractor unit has at least his height above the most likely impact zone and the engine itself for protection.

 

D4

 

Actually I would say the A pillar blind spot on an Envro being big enough to hide a car at a junction is very poor. Certainly easy to miss someone on two wheels at a junction or roundabout.  I was surprised that it passed the Cab Design Committee from the Unions.

 

 

The purpose of the RAIB is to determine the cause, not to ascribe blame. That is for the ORR, as the safety authority and, where appropriate, the Police and any guilt or otherwise is only proven by a Court of Law.

 

Jim

I realise that Jim of course but the mere suggestion that the driver has had  a micro sleep is now out there in the public domain (being discussed today by Jeremy Vine) and what I was trying (poorly) to say was that there is no evidence really to show he did or he didn't but of course for a lot of people it is a case of it has been suggested therefore must be true.  Being an Ex Police Officer I know about evidence, burden of proof and the like but innocent until proved guilty seems to mean nothing to the press.

[Edwin_m]

As for the accident: I still believe a simple speed-trap system using computing power already available inside the vehicles would have prevented the entire nightmare to unfold. Total cost: a pair of magnets and 2 magnetic sensors for each vehicle, plus some wiring and a minor update to the vehicle software. I doubt it'll be any more the 30k with most of the cost taken up by installing the additional wiring and hiring the supplier to change the S/W side and roll-out of such to each vehicle. Cost of magnets including installation: not more than a grand. Magnets are installed a very specific distance apart approx 90-100yrds from the curve. The first magnet activates the sensor (under each cab, direction sensitive) and triggers a timer to run. If during the running of this timer a 2nd trigger signal arrives from the 2nd magnet, the system detects over-speed and thus triggers an unrecoverable emergency brake application the driver can only reset while stationary. Simple, cheap, very (cost-) effective and in use by the Swiss (amongst others) for decades. Including tramways!

That seems entirely sensible.  The TPWS system we use on UK main lines works on a similar timer principle but using inductive loops rather than magnets and probalby the space and time are more appropriate to a heavy rail environment (and it woudl also be a lot more expensive). 

 

The other option would be GPS, which could in principle measure both position and speed and contain a simple list of places to apply the brakes if the specified speed is exceeded.  I understand something similar is already used on some trams to dispense flange lubricant on the approach to certain tight bends.  There would have to be certainty that it would work in confined locations such as the tunnel exit in Croydon though. 

[Pacific231G]

In the case of the Herald of Free Enterprise I think the bigger issue goes further back than the crew or TT, why did the world ignore the known and understood vulnerabilities of Ro0Ro vessels for so long?

That's certainly also true and I suppose the answer is that undivided vehicle decks made them more profitable. But, even with my own very limited seafaring experience I could see a difference in culture between boring old Sealink and dynamic bucaneering (and corner cutting) Townsend Thoresen and it wasn't a Sealink ferry that sank. What's more to the point was meeting officers working for TT a year or two earlier who feared for the safetyof their ships though it was fire on the undermonitored vehicle deck rather than free surface effect they were most concerned about. Whether P&O, who'd only just taken over TT, would have introduced a safer culture if they'd had more time we'll never know.

In a horrible irony, I was on a TT ferry from Calais the week before the Herald disaster and they were showing a video banging on about how dangerous the Channel Tunnel was going to be.

Edited December 8, 2017 by Pacific231G

[jim.snowdon]

I realise that Jim of course but the mere suggestion that the driver has had  a micro sleep is now out there in the public domain (being discussed today by Jeremy Vine) and what I was trying (poorly) to say was that there is no evidence really to show he did or he didn't but of course for a lot of people it is a case of it has been suggested therefore must be true.  Being an Ex Police Officer I know about evidence, burden of proof and the like but innocent until proved guilty seems to mean nothing to the press.

Chris, we aren't to know that. More importantly, whilst the report refers to the possibility of a microsleep, it goes on to note that the driver, having recovered, failed to re-establish his perception of where he was, ie he became spatially disoriented, with no visual cues to help build any picture of where he really was.

 

That seems entirely sensible.  The TPWS system we use on UK main lines works on a similar timer principle but using inductive loops rather than magnets and probalby the space and time are more appropriate to a heavy rail environment (and it woudl also be a lot more expensive). 

 

The other option would be GPS, which could in principle measure both position and speed and contain a simple list of places to apply the brakes if the specified speed is exceeded.  I understand something similar is already used on some trams to dispense flange lubricant on the approach to certain tight bends.  There would have to be certainty that it would work in confined locations such as the tunnel exit in Croydon though. 

Edwin,

As ever, everyone seems to rush off looking to apply a combination of technology and/or railway thinking, which may be a disproportionate response. It is going to be for London Trams, working in the context of UK tramways in general, to decide what solution(s) they come up with, and for the ORR to consider whether or not that is sufficient. I don't doubt that that process is already in progress, as there will have been discussions between RAIB and the UK tramway industry well before the report was finally published, and the wrong thing to do is rush off coming up with knee-jerk solutions.

 

Jim

[The Stationmaster]

Membership of CIRAS, especially outside of the main line railway industry was never, as I understand it, compulsory, particularly as the body originated within the RSSB, which itself does not cover the non-regulated railway sector. I would very much doubt that RAIB would be in error over their statement, but you do need to allow for the 13 month time difference between the event and now. RAIB will report on the situation at the time; if London Trams and TOL have subsequently become members after the event, that will not be covered by the report.

 

Jim

 

The 'non-regulated railway sector' effectively means only those railways whose lines have a track gauge of less than 350mm (13.78") - thus all (I think) of the publicly known narrow gauge railways in the UK are 'regulated' but miniature railways, including those which carry passengers on a commercial basis, fall outside the scope of the Regulations.  RSSB does not therefore cover all of the regulated railway sector but specifically covers only the national network, i.e. lines owned and operated by NetworkRail and equally RSSB does not cover various privately owned standard gauge railways which are linked to the national network beyond their boundary with NR.

[Mad McCann]

Actually I would say the A pillar blind spot on an Envro being big enough to hide a car at a junction is very poor. Certainly easy to miss someone on two wheels at a junction or roundabout.  I was surprised that it passed the Cab Design Committee from the Unions.

 

 

 

 

 

 

There's a similar blind spot on the A post/lead door pillar on the Solo and you really do have to look carefully before turning out of a side road. It could get you into real trouble otherwise. It did occur to me that this was a significant weakness in the design.

 

D.

[JDW]

There's a similar blind spot on the A post/lead door pillar on the Solo and you really do have to look carefully before turning out of a side road. It could get you into real trouble otherwise. It did occur to me that this was a significant weakness in the design.

 

D.

 

Agreed, though I've driven worse too!  (And don't get me started on the half-height opening cab window on ours, where the horizontal divide is at many people's eye level...) But every vehicle will have blind spots, and even for different drivers the same vehicle can be better or worse depending on seat position

(as an aside to the topic - sorry - on some of our other types we recently started replacing the long nearside mirrors that stick forwards with ones that you view through the door glass.  Management and Engineering thought it was a great idea, to reduce the number of mirror strikes on shelters, etc.  Great.  Until some of us who are thinner and sit further forwards got in, and found we could no longer see it as the dividing bar for the cab door/assault screen was right in our line of sight.  It took a lot of arguing to get them to see it from our point of view, and agree that having to lean back/forwards to see it wasn't acceptable...)

 

Back to the original topic, and in the dark it seems that the easiest visual cue would be a set of highway-style  [///]  [//]  [/]  countdown boards which gives the driver an instant, cheap and failsafe way of judging the distance in the dark.  It doesn't require any fancy electronics and could have prevented the accident - indeed, its the kind of thing an experienced driver might have picked up on and suggested from the start.  Obviously, that wouldn't stop a runaway/overspeed/brake failure happening, but would deal with the root cause of this type of incident for a very small outlay.

[Mark Saunders]

Trams and trains are not the same things - they may both run on rails, but there is a completely different operational environment. To suggest that the same sort of safety measures that are appropriate for 125mph trains where the driver cannot even see the length of his stopping distance are appropriate for a line of sight tramway is suggestive of a lack of understanding of the tramway environment.

 

Jim

 

What's in a name, tram or train?

 

The effect of excess speed on a curve is exactly the same so a similar safety measure is appropriate!

 

Mark Saunders

[jim.snowdon]

It has the merits of both redundancy, ie missing one sign still leaves more than sufficient information to allow the driver to recover the situation, and providing more information over a longer period to keep the brain stimulated, which is pertinent to avoiding a micro sleep in the first place.

 

It is also a practical solution that should achieve a significant safety improvement without disproportionate cost, unlike more technically complex solutions that will yield an only marginally greater improvement for significantly greater cost. There has, ultimately, to be a balance between safety and cost, otherwise the whole exercise becomes self-defeating.

 

Jim

[Coryton]

What's in a name, tram or train?

 

The effect of excess speed on a curve is exactly the same so a similar safety measure is appropriate!

 

But the trams are driven in a very different way to trains (line of sight vs signals etc.). The effect of excess speed may be the same, but the appropriate measure to prevent it from happening may not be.

 

 

(as an aside to the topic - sorry - on some of our other types we recently started replacing the long nearside mirrors that stick forwards with ones that you view through the door glass.  Management and Engineering thought it was a great idea, to reduce the number of mirror strikes on shelters, etc.  Great.  Until some of us who are thinner and sit further forwards got in, and found we could no longer see it as the dividing bar for the cab door/assault screen was right in our line of sight.  It took a lot of arguing to get them to see it from our point of view, and agree that having to lean back/forwards to see it wasn't acceptable...)

 

Literally...

 

As for the lack of response to the RAIB pointing out what they saw as deficiencies in bus design...I wonder how they would react if the shoe were on the other foot and they received a similar letter, or what mechanism they could use to take it forwards if they wished. So far as I know they normally investigate rail accidents and write a report with recommendations. Can they make recommendations outside the context of an accident they're investigating?

[martin_wynne]

 

https://www.alibaba.com/product-detail/Freight-Car-Speed-Control-Retarder-for_60517226836.html

 

[Mark Saunders]

But the trams are driven in a very different way to trains (line of sight vs signals etc.). 

 

No still driven exactly the same! 

 

Permissive working is on line of sight!

 

Mark Saunders

[JDW]

True, that's a cheaper option, if the driver is alert. That might not always be the case (micro-sleep!) so the technical solution is the safer option.

 

It is indeed.  I was going to write an answer but jim.snowdon summed up exactly what I was thinking before I could answer myself:

 

It has the merits of both redundancy, ie missing one sign still leaves more than sufficient information to allow the driver to recover the situation, and providing more information over a longer period to keep the brain stimulated, which is pertinent to avoiding a micro sleep in the first place.

 

It is also a practical solution that should achieve a significant safety improvement without disproportionate cost, unlike more technically complex solutions that will yield an only marginally greater improvement for significantly greater cost. There has, ultimately, to be a balance between safety and cost, otherwise the whole exercise becomes self-defeating.

 

Jim

[jim.snowdon]

No still driven exactly the same! 

 

Permissive working is on line of sight!

 

Mark Saunders

Without signals, or TPWS, or AWS, either, at low speeds out of necessity. Permissive working constitutes, by railway standards, a tiny proportion of railway operations. Trains, at east in the UK, are driven under the absolute authority of fixed signals, with reliance on the driver to remember where he is and anticipate the fixed hazards of speed restrictions. Severe changes of speed should have AWS magnets as a warning device, but as we know from history, even the effectiveness of that single input depends upon the driver realising it for what it is and not just cancelling it as an automatic reaction.

 

Jim

 

Jim

[wombatofludham]

I think we have to make a note of caution. Finding technological "solutions" to the consequences of this accident - let's not forget, the only case of a modern light rail vehicle overturning on a sharp curve in the UK, albeit with tragic consequences - will increase the cost of building light rail in this country, which is already struggling to make a financial case for construction in some cases. In some cases it might be enough to make a new scheme unaffordable. If that were to happen, anyone who potentially would have used the new service would have to drive, walk, cycle or catch the bus, all of which, with the possible exception of bus travel, have far higher death and injury rates than a light rail scheme.

 

Most modern light rail schemes are classed as tramway throughout their length, which means they are driven as though they were road vehicles. Buses and coaches do not have overspeed protection when approaching sharp bends around the country, and the impact of a double decker bus coming off a motorway slip road at high speed - as happened in 2007 when a double deck coach overturned on the M4-M25 junction - could be just as catastrophic.

[Mark Saunders]

Which is why a system that interferes when the driver doesn't react is the safer option.

 

@Mark S.: have you ever driven a train, or tram for that matter? Take it from those that have (vast) experience with either or both, it's a world of difference.

 

As rule of thumb I do not advertise what I do for the day job, the closest I have came to tram driving is 142's but first principles apply to both! 

 

Mark Saunders

[Coryton]

Most modern light rail schemes are classed as tramway throughout their length, which means they are driven as though they were road vehicles. Buses and coaches do not have overspeed protection when approaching sharp bends around the country, and the impact of a double decker bus coming off a motorway slip road at high speed - as happened in 2007 when a double deck coach overturned on the M4-M25 junction - could be just as catastrophic.

 

Yes if that were on the railway there would no doubt be calls for some kind of automatic speed limiting system to be installed on all slip roads used by double deck coaches.

 

I think in that case the reaction was to stop using double deck coaches on that service...but there are plenty on other routes.

[Edwin_m]

I think we have to make a note of caution. Finding technological "solutions" to the consequences of this accident - let's not forget, the only case of a modern light rail vehicle overturning on a sharp curve in the UK, albeit with tragic consequences - will increase the cost of building light rail in this country, which is already struggling to make a financial case for construction in some cases. In some cases it might be enough to make a new scheme unaffordable. If that were to happen, anyone who potentially would have used the new service would have to drive, walk, cycle or catch the bus, all of which, with the possible exception of bus travel, have far higher death and injury rates than a light rail scheme.

 

Most modern light rail schemes are classed as tramway throughout their length, which means they are driven as though they were road vehicles. Buses and coaches do not have overspeed protection when approaching sharp bends around the country, and the impact of a double decker bus coming off a motorway slip road at high speed - as happened in 2007 when a double deck coach overturned on the M4-M25 junction - could be just as catastrophic.

I agree in general, which is why I suggested a GPS-based system or the Swiss magnet system rather than a railway-style ATP or even TPWS (I have some professional background on both TPWS and GPS in a rail context). 

 

However there is an imporant difference with road vehicles, which the report does hint at.  If a road vehicle runs into a curve too fast it will probably go off the road and remain upright, but a tram is constrained to follow the rails and when the laws of physics make that impossible it will overturn.  That's unless flange climbing happens first, which the report suggests would have been the case if it hadn't been raining and might well have been a less catastrophic outcome. 

 

I also thought it was a little harsh to suggest the tram industry wasn't aware of overturning risk before this accident.  The black and white chevrons where Metrolink does a tight turn off the former Oldham Loop at Werneth demonstrate that wasn't universally so.