Croydon Tram Crash

[Tim R-T-C]

Unfortunately, reports made by passengers do tend to be treated with a degree of scepticism. This is because virtually every single one turns out to either be an error, or worse a deliberate lie. Far more times than not, when CCTV is viewed it paints a different picture to the one painted by the complainer.

That's not to say reports aren't or shouldn't be investigated, just that I'd treat *any* report made by someone who isn't railway staff as very suspect, until substantiated by some kind of evidence, be it CCTV or download.

Ironically, one of the issues above comes from someone who "worked on the railways", although they might have been a station cleaner for all we know.

I think the Tramlink would do well to find the data for the trips that have been mentioned in the above article and either release them to prove that the train was travelling correctly, or take action against the drivers if not.

Even if they are groundless allegations, given what has happened, looking into them will either reassure passengers that the line is safe and avoid a loss of travellers to cars (which are infinitely more dangerous) or show drivers that action will be taken if they over-speed on sections.

 

[bramling]

Ironically, one of the issues above comes from someone who "worked on the railways", although they might have been a station cleaner for all we know.

I think the Tramlink would do well to find the data for the trips that have been mentioned in the above article and either release them to prove that the train was travelling correctly, or take action against the drivers if not.

Even if they are groundless allegations, given what has happened, looking into them will either reassure passengers that the line is safe and avoid a loss of travellers to cars (which are infinitely more dangerous) or show drivers that action will be taken if they over-speed on sections.

Yes I agree, there really needs to be a definitive statement from someone associated with the investigation (be it RAIB, TFL or First) outlining as many of the facts as are known. It really isn't good for people to be picking up their information from the media.

Regarding the previous overspeed allegations, it will of course depend on how long the data in the data recorders is kept before being overwritten. Some can be as low as 72 hours, so unfortunately it's quite possible the data would have already been overwritten before the allegations came to light. Perhaps there might be some CCTV that's still around, though of course this isn't as good as a data recorder.

 

[BestWestern]

Unfortunately, reports made by passengers do tend to be treated with a degree of scepticism. This is because virtually every single one turns out to either be an error, or worse a deliberate lie. Far more times than not, when CCTV is viewed it paints a different picture to the one painted by the complainer.

That's not to say reports aren't or shouldn't be investigated, just that I'd treat *any* report made by someone who isn't railway staff as very suspect, until substantiated by some kind of evidence, be it CCTV or download.

This is of course very true, but in the vast majority of cases the evidence can be easily viewed. CCTV is present on most stations, and on a lot of rolling stock. If a report comes in, most of the time it should present little difficulty establishing whether or not it is accurate.

 

[matt_world2004]

This is of course very true, but in the vast majority of cases the evidence can be easily viewed. CCTV is present on most stations, and on a lot of rolling stock. If a report comes in, most of the time it should present little difficulty establishing whether or not it is accurate.

The problem with the fgw case is because they didn't believe the customer it was too late to examine the train.

 

[AlterEgo]

Given GPS is military-grade technology, though, it wouldn't be too prohibitively expensive to create a more industrial-grade solution based on it though, redundant antennas, redundant electronics, etc. - if it designed to fail safe, for example, if there was no GPS lock or an unconfirmed / weak GPS lock, or a system failure, the tram could be limited to the maximum street-running speed until it is taken out of service.

Certainly it would have prevented an incident like this.

The incident in question happened directly after a long tunnel section where there would have been no, or weak, GPS signal. Limiting the tram to maximum street running speed would still have been well in excess of 12mph. I'm just not convinced it's a solution.

 

[2HAP]

Thinking back to the Well Hall derailment, the speed the train derailed at was something like three times that of the speed restriction in force. It was revealed that the speed at which passengers felt comfortable at was much less than the speed at which the curve could have safely been taken.

Given that tram passengers had previously reported feeling unsafe on that curve, their claims must have some merit, even though estimates of speed may be wide of the mark in either direction.

 

[Islineclear3_1]

The RAIB's interim report is expected to be out this week, so watch this space

 

[Tim R-T-C]

The incident in question happened directly after a long tunnel section where there would have been no, or weak, GPS signal. Limiting the tram to maximum street running speed would still have been well in excess of 12mph. I'm just not convinced it's a solution.

Using GPS seems like overkill on this short route - local beacons could easily be used which would be far more effective. Sited near the curve, being picked up at the point where a unit should be starting to brake.

However, like all these sort of safety systems, like lane divergence warnings in a car, they run the risk of a driver becoming over-reliant on them and if they fail for any reason, catching a driver out.

 

[Crossover]

Thoughts with all those involved in this.

Does anyone know, given the tram was finally lifted by crane some distance from the incident scene, how it was righted? Was it done with the same crane but the stock then moved to a location with better road access to load it onto the trailers for onward road transport?

 

[Bletchleyite]

The incident in question happened directly after a long tunnel section where there would have been no, or weak, GPS signal. Limiting the tram to maximum street running speed would still have been well in excess of 12mph. I'm just not convinced it's a solution.

An easier solution might be to provide an approach-controlled stop signal before the curve, requiring the tram to fully stop before it clears. It would then both act as a reminder and ensure the tram would not have chance to reach a dangerous speed before the curve.

If a tram is basically a road vehicle, we probably need to think about road-style solutions, not rail ones. Approach controlled traffic lights are used to slow traffic on roads in places, e.g. on approach to high-risk pedestrian crossings. (They are a nuisance to cyclists as cycles often won't trigger them, but nobody is going to be cycling down the tram track).

 

[littlerock]

Can I return to the windows for a moment. According to various press reports, mainly the Croydon Advertiser, the windows fell out when the train toppled over. One passenger said they smashed.

"Fellow passengers on the early morning Croydon Tramlink service in South London told how one man was flung through a window."

"Rhys McCausland, 19, recounted the horror crash, saying his head was crushed against the ground as the tram windows smashed behind him.The teen said: "There was nothing unusual going on the tram. I had my headphones in sitting there in he middle. Going to Sandilands through the tunnel there was nothing unusual either. We go very fast through there on a regular basis. I didn't notice we were coming to a corner because I wasn't really paying attention. The next thing I knew the tram was rolling over on its side. There was no attempt to brake or anything like that. I rolled on my back. The glass has smashed behind me and my face rolled over the gravel. I was quite lucky not to be trapped like the others.He said the passengers were caught in the dark with rain pouring down on the wreckage.He said: "I could feel the blood dripping and I was in total shock about what happened."I have been getting on that tram nearly every day for four years. To realise it's on its side in pitch black you can't really believe it. Having no way out is the hardest."

Other passengers fell out - presumably no windows in place - and were trapped underneath the train. One survivor reported that he felt the train topple over and the next thing his hand was scraping along the gravel.

Another passenger, Rui De Sa, a builder on his way to work when disaster occurred at 6.10am, said: "One man beside me was screaming and had his arm trapped outside the tram as it skidded along on its side. He added: 'I was on my phone with my headphones on then we just heard screaming in the first carriage. That toppled over then we went over. It went black and we were all just thrown around. People were screaming and shouting for help. De Sa said once the train was on its side anything it hit was "like a blade'' through it.

Now I have read the trains had laminated glass windows, so did the windows just fall out or blow out?

Safety Standards Railways safety standards say laminated glass should be installed in all railway passenger vehicles. But the standards also seem to assume laminated windows will stay in place during a crash, which these did not. Does a light railway vehicle/tram count for these standards, or are they exempt? And if so why?

http://www.rssb.co.uk/Library/risk-...improvements-in-safety-in-train-accidents.pdf)

It says here that laminated windows which stay in place during a crash provide the best protection against injury as they stop people falling out. Clearly that was not the case here, as the windows either fell out or shattered. Should they? Otherwise what is the point in fitting them unless to to provide protection during a crash. Yet they were smashed/fell out on the side which hit the ground, when the train overturned. Was it the act of falling over which caused them to fall out/smash? And did they smash into small pieces or come out as individual panels.?

RSSB report in 2009

3.1.2 Passenger and crew containment

Analysis of train accidents that have occurred in the last 10 years showed that 20% of fatalities occurred through people being ejected through breakable windows (usually when a carriage has turned onto its side). If windows are strong or tough enough, they can prevent people from being ejected from trains during an accident. However, if breakable, windows can be used to escape from trains after accidents. The analysis showed that, if the breakable windows are replaced with essentially unbreakable windows (using laminated glass), then more people would be expected to survive accidents on those rare occasions when such events occur.

One consequence of having all laminated windows is that passengers would no longer be able to use windows for escape after an accident without the intervention of the emergency services. Further exploration of the accidents database, and consideration of an extensive review of emergencies on trains overseas (including fires), found no examples where a life has been lost, or would have been lost, because of an inability to get out through the window.

The research shows that either there is no imminent requirement to leave the train under the circumstances, or alternatively, there are better (and safer) egress routes through the bodyside doors and gangway ends. Indeed, a number of significant injuries have been sustained by people escaping from derailed trains through windows. Furthermore, most serious train fires arise from external sources; even after impact, laminated glass is generally retained in position and helps prevent fire from entering railway vehicles.

The conclusions that were reached in this work, presented to the ATOC Board and the Office of Rail Regulation (ORR), were: Risk analysis showed that it is important that the industry has a clear and consistent approach to escape as passengers seeking to break unbreakable windows would put themselves and others at risk; it was duly recommended that all hammers provided for breaking windows be removed, even on those vehicles with breakable windows.

Accordingly, the analysis supported an industry approach that passengers should be contained in a rail vehicle in the event of an accident, particularly when they overturn, and should not be encouraged to escape through windows at all.

All bodyside windows in passenger and train crew areas on new vehicles should be fitted with laminated glass and have a high degree of containment; consideration should also be given to the frame and mountings. Windows on existing vehicles should be considered for progressive replacement with laminated glass, but should always be replaced when broken, subject to cost-benefit analysis on refurbishment.

To facilitate the incremental fitment of laminated glass, train operating companies (TOCs) should remove hammers and alter signage such that the primary egress route, in the event of an evacuation being required, is recognised as being via the doors and gangways instead of breakable windows. In order to realise the full safety benefits of laminated windows, a consistent transition strategy should be developed and implemented across all TOCs.

All passenger trains built since 1993 have mostly laminated windows, but the research confirmed that the best approach was to dispense with all designated escape windows. The containment work in T424 (Requirements for train windows on passenger rail vehicles) has been developed into a series of measures for vehicle bodyside windows, including a comprehensive suite of test requirements. These are incorporated in GM/RT2100 Issue 4, which is currently undergoing industry consultation and is scheduled for publication in Spring 2010. The introduction of a common approach to escape is being coordinated by ATOC.

 

[507 001]

An easier solution might be to provide an approach-controlled stop signal before the curve, requiring the tram to fully stop before it clears. It would then both act as a reminder and ensure the tram would not have chance to reach a dangerous speed before the curve.

If a tram is basically a road vehicle, we probably need to think about road-style solutions, not rail ones. Approach controlled traffic lights are used to slow traffic on roads in places, e.g. on approach to high-risk pedestrian crossings. (They are a nuisance to cyclists as cycles often won't trigger them, but nobody is going to be cycling down the tram track).

All signals on tramways are approach controlled. Including the one on the approach to this curve/junction. The problem is if the tram goes through the signal there is nothing to physically stop it.

 

[Bletchleyite]

All signals on tramways are approach controlled. Including the one on the approach to this curve/junction. The problem is if the tram goes through the signal there is nothing to physically stop it.

In this instance, did it?

 

[littlerock]

Can I ask some more questions?

First, and this may not be popular but I will ask it anyway. Is it possible that, to enliven a dull job, some tram drivers were engaged in a competition to see how fast they could take the bend? Reports that the driver may have been texting at the time of the incident might support this, if he was sending a text/image showing his speed at that stage of the journey.

If the driver in fact blacked out before the incident, it must have been at the last minute, and in a 30 second interval between presses of the "dead mans handle" button, as no one reported any breaking or a klaxon sounding. Does this mean a different system is needed?.

Finally several accounts report that the train seemed to speed up as it approached the bend. Is that just how it seemed or is it likely to have happened? And if it did speed up, why? Anyone got a view?

 

[AlterEgo]

Can I ask some more questions?

First, and this may not be popular but I will ask it anyway. Is it possible that, to enliven a dull job, some tram drivers were engaged in a competition to see how fast they could take the bend? Reports that the driver may have been texting at the time of the incident might support this, if he was sending a text/image showing his speed at that stage of the journey.

If the driver in fact blacked out before the incident, it must have been at the last minute, and in a 30 second interval between presses of the "dead mans handle" button, as no one reported any breaking or a klaxon sounding. Does this mean a different system is needed?.

Finally several accounts report that the train seemed to speed up as it approached the bend. Is that just how it seemed or is it likely to have happened? And if it did speed up, why? Anyone got a view?

I don't think we have any information at this stage to even speculate on those questions - particularly the allegation that the tram drivers "had a competition".

Are you a journalist?

 

[MarkyT]

The incident in question happened directly after a long tunnel section where there would have been no, or weak, GPS signal. Limiting the tram to maximum street running speed would still have been well in excess of 12mph. I'm just not convinced it's a solution.

A transponder solution is more practical I think. In this case it could be a passive balise indicating a mandatory deceleration within given parameters.

Whatever the trackside system, on board there would be a lot of work in modifying the trams, interfacing with brake and traction control, incorporating new indications and controls into the drivers desk layout, rewriting rules and regulations. All difficult, time consuming and expensive.

In the interim, used with distant marker boards, such passive balises could be configured as a 'modern AWS' to activate a simple audio-visual cab warning. Like main line AWS, the warning would require acknowledgement within a time limit or it would trigger an emergency stop. The tram would need a balise reader and other changes on board, but not as complex as with the speed control option, and it would be a means to ensure drivers approaching such danger zones were awake and engaged to take control, exactly the justification for main line AWS sixty odd years ago and the GWR's ATC half a century earlier. Using standard balises could also be a useful stepping stone to achieving greater protection in the future, with upgrade potential once the trams were modified with some more sophisticated kind of ATP control or new ones purchased with that capability built-in. Then the trams might be sent messages conveying more sophisticated behaviour, such as to to slow down for a particular restriction within so many metres or seconds elapsed.

 

[Antman]

Can I ask some more questions?

First, and this may not be popular but I will ask it anyway. Is it possible that, to enliven a dull job, some tram drivers were engaged in a competition to see how fast they could take the bend? Reports that the driver may have been texting at the time of the incident might support this, if he was sending a text/image showing his speed at that stage of the journey.

If the driver in fact blacked out before the incident, it must have been at the last minute, and in a 30 second interval between presses of the "dead mans handle" button, as no one reported any breaking or a klaxon sounding. Does this mean a different system is needed?.

Finally several accounts report that the train seemed to speed up as it approached the bend. Is that just how it seemed or is it likely to have happened? And if it did speed up, why? Anyone got a view?

I really don't think anyone on here is in a position to answer such questions.

 

[Islineclear3_1]

Can I ask some more questions?

First, and this may not be popular but I will ask it anyway. Is it possible that, to enliven a dull job, some tram drivers were engaged in a competition to see how fast they could take the bend? Reports that the driver may have been texting at the time of the incident might support this, if he was sending a text/image showing his speed at that stage of the journey.

If the driver in fact blacked out before the incident, it must have been at the last minute, and in a 30 second interval between presses of the "dead mans handle" button, as no one reported any breaking or a klaxon sounding. Does this mean a different system is needed?.

Finally several accounts report that the train seemed to speed up as it approached the bend. Is that just how it seemed or is it likely to have happened? And if it did speed up, why? Anyone got a view?

Please see post above and don't speculate. No one on here knows what really happened and what the driver was doing. Wait for the RAIB report

 

[Robertj21a]

Can I ask some more questions?

First, and this may not be popular but I will ask it anyway. Is it possible that, to enliven a dull job, some tram drivers were engaged in a competition to see how fast they could take the bend? Reports that the driver may have been texting at the time of the incident might support this, if he was sending a text/image showing his speed at that stage of the journey.

If the driver in fact blacked out before the incident, it must have been at the last minute, and in a 30 second interval between presses of the "dead mans handle" button, as no one reported any breaking or a klaxon sounding. Does this mean a different system is needed?.

Finally several accounts report that the train seemed to speed up as it approached the bend. Is that just how it seemed or is it likely to have happened? And if it did speed up, why? Anyone got a view?

I think you, and all the others, might benefit from the initial report, due out shortly. No doubt that will help provide some more facts.

 

[Juniper Driver]

I have often marvled at the tightness of this curve and the skill of drivers easing off, taking it and rapidly getting up speed again. Really sad to hear things have gone wrong here and hope all are ok

Is it tighter than the 25mph outside Southampton? We used to come up there on the 442s when no TPWS was fitted and never had any problems.TPWS is fitted now which changes the braking points so to speak.Not that I have any problem with that.

 

[martin2345uk]

Is it tighter than the 25mph outside Southampton? We used to come up there on the 442s when no TPWS was fitted and never had any problems.TPWS is fitted now which changes the braking points so to speak.Not that I have any problem with that.

You can see on google earth, it's basically a 90 degree turn!

 

[littlerock]

Edit: wild speculation removed

My other two questions were quite separate.

1. Is the 30 second delay between presses of the dead man's handle, too long? Travelling at high speed it would not be enough to activate the brakes on a sharp bend before the train fell off the track.

2. I provided examples of passengers saying that the train speeded up after it left the tunnel and I was wondering if that was just how it felt, or if it really did and if so what could be the reason.

 

[Taunton]

I suspect the passengers notice the speeding up after the previous tram stop. Up to that point it is basically roadside tramway reservation, at that point it turns north onto the old fully separated railway alignment and is a long run between stops. Must be one of the fastest lengths of the system.

 

[littlerock]

Speaking for heavy rail, that is the general idea. But if there is still pressure or weight being applied after the driver blacks out you will have to wait untill the driver vigilance device kicks in which could be up to a minute.

So if, as some reports say, the driver blacks out and falls on the accelerator, just before a bend, the train will speed up before the driver vigilance device kicks in. With inevitable consequences. How does it work on conventional locomotives if the driver collapses at the controls? Would the same be true?
--- old post above --- --- new post below ---
Can I reiterate my point about the windows? Were they laminated? The recommended safety standards are for windows not to break in railway accidents as people being thrown out of the windows is the largest cause of fatalities.

Yet that is what happened at Croydon. Still Survivors reported the windows breaking. Some reported seeing a passenger thrown out, others that they felt onto the ground as the train toppled over and others were then trapped underneath.

Rui de Sa told the Mirror the man next to him had his head crushed under the tram as it slid on its side. Mr De Sa, 31, treated for back and shoulder strains and shock, said: “We just heard screaming in the first carriage. That toppled over then we went over.

“It went black and we were all thrown around like rag dolls. People were screaming and shouting for help.One man beside me was screaming with his arm trapped outside the tram as it skidded on its side.I looked at the man next to him, who had his whole head out of the tram and crushed *underneath it".

He said the tram was going “too fast round the bend”, adding: “I’ve been getting that tram for nine years and normally you feel it braking there. This time I couldn’t feel it braking at all. I don’t know whether the brakes weren’t working but it was going way too fast.

"But if those carriages had been built with unbreakable glass none of these people would have died.When it started sliding down the line on its side anything going through it was like going against a blade.”

 

[rebmcr]

Any answer to your questions, littlerock, as you have basically drilled down to the fundamental factors of the situation, would be beyond the scope of speculation and require actual facts from the investigation. Sit tight, it'll be out soon.

 

[507 001]

In this instance, did it?

In this instance it wouldn't have made any difference whether the signal was on stop or not. He would have just blown straight through it. It's a waste of time having approach control if you there isn't a system to intervene if the train fails to respond correctly.

Which kinda disproves the 'competition theory', because at the very least it would have resulted in a SPAS (Signal Passed At Stop) on the drivers record and potentially dismissal. Nobody is going to risk that, let alone the deaths of 7 people.

My other two questions were quite separate.

1. Is the 30 second delay between presses of the dead man's handle, too long? Travelling at high speed it would not be enough to activate the brakes on a sharp bend before the train fell off the track.

2. I provided examples of passengers saying that the train speeded up after it left the tunnel and I was wondering if that was just how it felt, or if it really did and if so what could be the reason.

AFAIK, and I'm going off my traction knowledge of the nigh on identical M5000 sets in Manchester, the CR4000 series don't have Vigilence. They have a simple dead mans handle on the TBC (the TBCs in M5000s and CR4000s are identical) which when released sounds a bell. If the driver doesn't put his/her hand back on to the TBC (and if motoring return it to the coast position) within 4 seconds, the tram will drop its track brakes.

There is no requirement, certainly on the newer M stock to repeatedly press a button.

 

[Antman]

In this instance it wouldn't have made any difference whether the signal was on stop or not. He would have just blown straight through it. It's a waste of time having approach control if you there isn't a system to intervene if the train fails to respond correctly.

Which kinda disproves the 'competition theory', because at the very least it would have resulted in a SPAS (Signal Passed At Stop) on the drivers record and potentially dismissal. Nobody is going to risk that, let alone the deaths of 7 people.



AFAIK, and I'm going off my traction knowledge of the nigh on identical M5000 sets in Manchester, the CR4000 series don't have Vigilence. They have a simple dead mans handle on the TBC (the TBCs in M5000s and CR4000s are identical) which when released sounds a bell. If the driver doesn't put his/her hand back on to the TBC (and if motoring return it to the coast position) within 4 seconds, the tram will drop its track brakes.

There is no requirement, certainly on the newer M stock to repeatedly press a button.

There are approach controlled signals at the Sandilands Junction.

 

[507 001]

There are approach controlled signals at the Sandilands Junction.

Yes I know, perhaps I should have said 'in this case it didn't make a difference' instead of 'wouldn't have'.

 

[jon0844]

Which kinda disproves the 'competition theory', because at the very least it would have resulted in a SPAS (Signal Passed At Stop) on the drivers record and potentially dismissal. Nobody is going to risk that, let alone the deaths of 7 people.

I think if there's ANY truth to this story, it's that some drivers may have over the years realised they can go faster than 12mph and perhaps some have tried a little bit faster each time.

However, given the alleged speed, I sincerely doubt someone was doing this regularly - just 1mph more each time, for example - and nobody was noticing besides one Facebook post?

It's not a very plausible story, as no driver would surely keep pushing until the train tipped over!

 

[AndrewE]

Going back (quite) a bit, talk about black-outs in the aftermath of any accident should be read in the light of the fact that one of the first consequences of concussion is complete loss of memory of the minutes before it happened.

A friend was told that she shouldn't drive her moped on the grounds that she couldn't remember what had happened just before she came off it [and hence must have had a black-out], when in fact there was no history of blackouts or any other suggestion that she might have had one. Not remembering what happened just before an accident isn't uncommon or incriminating in any way.
A