Holland Park incident - RAIB report

For reference the original thread is here.

In accordance with Forum Rule 7.3 I've added a link to this thread in the original, and it is here that any discussion should occur.

As someone who hasn't (and probably wouldn't) pick up on these errors, any clues as to what the errors are?

I vaguely remember that at that time this happened, the discussion about passenger use of butterfly cocks (or emergency egress devices), and we noted that whilst it would be good for situations like this, it could be rather disadvantageous for everyday operations, because as soon as you give people a big thing saying "pull to open the door, emergency use only) it would then start to get abused by people who have just had doors closed in their face, and various miscreants out for a 'laugh'. That said though, the EEDs on s stock are fairly obvious, but don't seem to have been operated maliciously enough to see it become 'abuse'

To be honest, in a real emergency, there are always alternative means of accesing/leaving the train, often the window if it is a 92, D, or S stock, it may be more problematic on the smaller windows of other stock (but not impossible)

I was surprised why was the Traction Current was not discharged promptly, as is usually the case in this type of situation...

I think it has to be assumed that passengers are children, will behave illogically, pull and press everything press-able, not read any notices, not speak english, be drunk and incapable of looking after themselves.

And then think, when they do all those things, how much damage can they do.

In actual fact, very little, apart from cause confusion.

The only thing that might improve things is to make the inter car barriers harder to climb over.

And some 'cockpit resource management' or equivalent for the driver to be able to interpret a sudden deluge of multiple sources of [sometimes conflicting] information. This incident highlights that drivers have a very responsible job , and should be paid and trained accordingly.

in most Metros around world have emergency exit door - example Newscatle and Mersyrail both have emergency exit handles above the doors and DLR does have those fitted at floor level - the problem is if opeartor is knocked out or killed then passengers has to try to open doors to get out of danager - remember centrail line derailment years ago - people found buttefly to open the doors and some tried to kick out window to escape from danger
that's reason emergency exit is there for safety to evacuate from danger of smoke or any situations
skytrain in canada is one of

A thought which occurred to me, which I thought was wisely brought out in the report (see paragraph 63) was that the evacuation message played in response to the activation of the fire alarms probably made things much worse for people on the train, as it would have lead them to believe that there was definitely a fire and that they were being instructed to get out, but they could not get out.

I myself noticed in footnote 5:

"When the train is being operated in coded manual mode (see footnote 4), the emergency brakes are used to control train speed to 25 mph (40 km/h)."

That is simply not true to the best of my knowledge (although was, I believe, the case on the old 67s wasn't it?), in the general case.

It is possible that the MSS at the time was 40 (the footnote reefers to paragraph 69 which states, in part, "At 18:33 hrs when the train was between Notting Hill Gate and Holland Park stations, the emergency brakes were again applied, this time for only a few seconds, to correct a train overspeed"), in which case that would be true in the specific case. It would be true that the emergency brakes were used to correct an overspeed where the train exceeded an MSS of 40. However, if that is what was intended, the sentence is misleading.

In a station like Holland Park, when there is a fire alarm, the procedure should be that the lifts get taken out of service as soon as possible, as is generally the case in tall buldings and the like. The risk of having people stuck in a lift when staff are dealing with a potential fire is too much.
For those unable physically to make the climb up the stairs, the means of evacuation would be via a train. In this case it would have needed to have been on the opposite platform to the stopped train of course. If needs be an empty train would be arranged.

Are LU obliged to follow the recommendations in the report?

The raib said it was a flashover, but there was also a dragging brake in one of the cars, the smell of which contributed to the panic

Sort of. They have to consider the report, those involved in health & safety at LU have to take all of the recommendations into account when discharging their duties. However, it is only compulsory that LU comply with H&S legislation generally and that give appropriate consideration to the recommendations. They do not have to implement them, but they do have to report any measures they do take to the RAIB and give reasons to the RAIB for not implementing any additional measures, if they do not deem it fit to follow a recommendation:

From footnote 11 (footnote to paragraph 134):

"Those identified in the recommendations, have a general and ongoing obligation to comply with health and safety legislation and need to take these recommendations into account in ensuring the safety of their employees and others.

Additionally, for the purposes of regulation 12(1) of the Railways (Accident Investigation and Reporting) Regulations 2005, these recommendations are addressed to the Office of Rail Regulation to enable it to carry out its duties under regulation 12(2) to:
(a) ensure that recommendations are duly considered and where appropriate acted upon; and
(b) report back to the RAIB details of any implementation measures, or the reasons why no implementation measures are being taken.
Copies of both the regulations and the accompanying guidance notes (paragraphs 200 to 203) can be found on the RAIB’s website www.raib.gov.uk."

Kind of. The T/Op and DTSM involved initially attributed the smell (and, indeed, earlier problems on the train) to a dragging brake, but there was only limited evidence of a dragging brake and it doesn't seem to have made much difference.

Paragraphs 67-70:

"At the time of the incident, the train operator and the DTSM attributed the smoke and the pungent smell to dragging brakes. This was because they had both experienced dragging brakes before and believed they recognised the smell. However, there is no direct evidence to substantiate a dragging brake as the source of the smoke and smell. The train’s data transmission system, which displays diagnostic information to the train operator, did not record a dragging brake fault on the journey either to or from Holland Park. The train was also found to be rolling freely when it was tested before taking it forward from Holland Park. The vibration problem, which the train operator had associated with a possible dragging brake, was later found to have been caused by wheels on car 1 that had worn out-of-round.

However, it is possible that train 003 had an undetected dragging brake fault. When vehicles in the train were later uncoupled, maintenance records reportedly showed that a brake cylinder on car 6 was found to contain two bar of air pressure when it should have been empty (a brake cylinder is pressurised when there is a demand for braking, and emptied to release the brakes). If air pressure in a brake cylinder is not fully released it could cause a brake block to be partially applied and to drag on a wheel. It is not possible to say whether this condition was present at the time of the incident because a computer log, which would have recorded this information, was erased when the train was split. The maintenance work to correct the cause of the fault was not recorded in enough detail to make it clear whether the fault would be undetected by the train’s data transmission system. However, maintenance records do not show that the wheels or brake blocks on car 6 required any remedial action for the sort of damage that would be caused by a dragging brake.

A smell from the brakes could also be explained by two heavy braking events in the minutes before the train arrived at Holland Park station. At 18:31 hrs, when the train was travelling between Queensway and Notting Hill Gate stations the emergency brakes were applied by the train’s automatic protection system because of an error detected in the speed monitoring system. The train was brought to a stand from 28.5 km/h in 14 seconds. At 18:33 hrs when the train was between Notting Hill Gate and Holland Park stations, the emergency brakes were again applied, this time for only a few seconds, to correct a train overspeed.

Although there may have been some smell from the brakes, the primary cause o the smoke and burning smell was almost certainly a faulty traction motor. When the train was examined at the depot it was found that one of the traction motors at the trailing end of car 6 had suffered ‘flashover’ damage. This type of damage occurs when a motor suffers from an insulation breakdown and there is a subsequent electrical short circuit, known as a flashover. It is often accompanied by arcing and smoke from the breakdown and burning of insulation. The train is fitted with safety systems to ensure that the safety of passengers is not at risk from a motor flashover."

I'm really not sure about this one. For one thing, an MoS on the platform may have been able to prevent the train from leaving. The train should not have left Holland Park in the first place, it was supposed to be met by staff at Holland Park and taken out because of the smell reported earlier. However, that message seems never to have been passed on, so in that regard the MoS would not have made any difference as they would not have known that the train needed to be held, any more than the T/Op knew. The T/Op would normally be told and should have been. But anyway the MoS may have noticed something amiss in those few seconds before the doors closed when people started to leave car 6 because they could see smoke.

More importantly, though, the MoS would have been in a position to (almost instantly) operate the butterfly cocks on any and all cars which required it effecting a quick and calm response and a controlled evacuation of the train, as is meant to happen. They would also have provided visual reassurance that something was being done. The absence of anyone seeming to take control and of any direction is what led people to panic. Some couldn't be sure anybody knew what was happening and they didn't know what to do. An MoS would have been able to take charge, calm people down and direct and effect an orderly evacuation.

I know you said 'may' - I was just disagreeing with you - I hope politely

I agree that it was pretty quick - although obviously it would not have felt nearly quick enough to the people on the train, but I maintain that the presence of a staff member seeming to be doing something and possibly able to give instructions would have made it very unlikely that anyone would believe that they would need to climb out of the cars.

To stop a train you can use the emergency stop plunger or wave your arms around - probably any members of staff can add other methods, but those are the two I know.

Well, no LUL trains can depart without a pilot light, in other words - unless you've disabled the round train circuit manually - you cannot depart until all doors are proved closed and locked, so once you operate that butterfly cock, that train's going nowhere in a hurry. There may be additional rules and procedures for going by the book, but if you take a look at the video on hobbayne's original thread you can see the MoS operates the butterfly cock without seeming to go through any motions or procedures. Edit: It takes a handful of seconds for one set of double doors to open, but hardly any time at all really.

tjw - I understand that "what to do when loads of handles are pulled down at once" is now included in ATOR but as my last one was not long after the incident I've still got that pleasure to look forward to. A helpful IOp who was riding in my cab on the way to the W&C showed me how it works.

The only things the TOp did wrong was not make a PA to the whole train, for some reason he thought that talkback could be heard all over the train rather than just the car with the handle down, and not carrying his handheld radio when he went back to investigate. I'll admit I've been guilty of that in the past but both he and I trained before the days of handhelds so it doesn't automatically spring to mind when dealing with an incident and I can imagine with multiple alarms going off it might easily be left behind.

The report mentions that the Super and the DSM had both complained about a lack of station staff, they should have had eight on duty, but it seems there was a staff shortage as not enough people volunteered to work RDs for Carnival. Most of the staff were busy at the station entrance and ticket hall on crowd control, trying to restrict the number of people coming in to avoid overcrowding, they could only spare one bod and they were on the other platform. If you don't have enough staff then there's not much you can do about it, you can't force them to work rest days.

In my opinion the big failures that seem to have been somewhat overshadowed were Wood Lane telephoning the DSM at Holland Park rather than calling the TOp on the radio and the DSM (who said they didn't remember the phone call) not sending anyone down or radioing the staff member on the platform.