Air India crash fuel switches turned off

[limetrees32]

I've not found a thread on this , although it's taken me so long to search out the knowledgeable posters
on the Washington crash that there probably is one now.
But @notimagain what do you think ?

As I recall it in the air if you manually put the switch back to "run" the associated engine will begin an automated relight cycle - that's probably what the Times comment was about.

It's not an instant process and the engines on the Air India aircraft were in the process of doing that when it hit the buildings.

Apart from scary turbulence, I've only been in one serious incident and it was on landing. Aer Lingus flight from Heathrow to Dublin 1990.

The pilots didn't turn off the cockpit intercom so we heard it all, curses and sheer panic. Obviously everyone was distressed and bizarrely three nuns behind me started saying the rosary. Thankfully landed without incident and it was reassuring and frightening at the same time to see fire engines and ambulances accompanying us down the runway.

Thank you for your posts @notimagain , I’ve found them fascinating.

Engines which are separated from each other, but do they have a common electrical supply source? The Air Transat case, though entirely different had a single pump which pumped fuel from one tank to the next. And the computer (this one programmed) there could not make a simple deduction that the rate of fuel loss was so excessive that no more fuel should be allowed to be pumped into the faulty tank.

If this AI case can be pinned on human error, liability shifts from a company which can be fought in American courts to an Indian company where it may be difficult to fight this.

I tend to read Indian local legislation on and off, the language is misleading at the best of times, not by design, just lack of English experience. Thus, to clutch onto each word and dissect the report is futile in the Indian context. If people here have experience of reading FAA crash reports, they can opine if the language is indeed "tighter" than what they are reading in the Indian report.

For people here who may want to look at electrical inteference, they can look up the BA Belly Landing at Heathrow case where a Boeing 777 was written off with no deaths. At that time, the first "speculation" surfaced that the then PM, I think Gordon Brown's car was passing on the M25 when the plane passed it and that electrical jamming from his convoy may have caused an electrical issue. It was found to be icing in fuel.

@SuratNuJaman

Engines which are separated from each other, but do they have a common electrical supply source?

I'm sure the 787 wiring diagrams are out there if you want to investigate..I dumped the specifics of my types years ago.

What I can say is that generally speaking each engine has it's own main generator which feeds the main multiple, usually split, AC and DC busbars for supply of the whole aircraft and then often tucked away each engine is running one or more secondary or back up generators.

I can't swear to it but I think each FADEC has it's own dedicated generator or taps into one of the secondary gens. Overall all the systems are highly compartmented.

The Air Transat case, though entirely different had a single pump which pumped fuel from one tank to the next. And the computer (this one programmed) there could not make a simple deduction that the rate of fuel loss was so excessive that no more fuel should be allowed to be pumped into the faulty tank.

Err, hang on, no...you are not characterising that at all correctly.

That accident is quite famous in the multi engined flying world because it became a classic "how not to manage a fuel imbalance"..and you can't pin that on computers or AI, it was a pure Human Factors stuff up.

The crew were presented with all the data they needed to handle the imbalance correctly and the system gave them clear warnings, but for some reason they made a decision and took a course of action (without reference to checklists) that led to them continuing to pump fuel into a leaking engine (and so over the side).

The two cutoff commands that are being referred to were registered on the Flight Data Recorder (FDR) and were apparently separated by approx 1 second

Oh, that's interesting. So, given that there is the safety system set on the switches to avoid accidental cutoff, is it possible/has it been tested that a human could indeed cutoff both switches, with their safety system, within the space of one second?

I'll just point out the timing is subject to some inaccuracy (maybe plus/minus a second).
due to how often the data recorder looks at (samples) the switch position..

With that in mind if you wanted you could easily get both from run to cutoff in around a second..

It's simply a case of lift and pull down one switch, move the hand laterally 6 inches or so,.lift and pull down the second switch.

I would say you wouldn't expect anyone flying for a living to act in that way or that quickly, certainly on the aircraft, when the aircraft was in motion..there's supposed to a strict cross checking process applied before operating switches controling critical systems.

notimagain - sorry to keep asking you questions (and apologies if someone already said this and I missed it...),

Given this all happened extremely quickly, a matter of seconds, how would the pilot asking the question about cutoff, know that cutoff had occurred? Would it likely have been a visual check and he saw that the switches were in the wrong position? Or would it have been that he felt the aircraft didn't feel right and knew it was a fuel issue? And if neither of those things would likely have happened, then maybe it is because he saw it happen I guess.

Chances are performance going out of the window (e.g. feeling thrust reduction, speed starting to decay) might be the first clue...then possibly seeing systems failure messages on one of the screens...

We can only speculate that at some point somebody did a fast scan around all the various panels and spotted the switch positions.

Pull switch out, flick down; pull second switch out, flick down certainly seems like a sequence of actions that would only take a few seconds. As I understand it, the pull out part isn’t meant to be particularly resistant, it’s there to stop an accidental knock moving the switch, not a deliberate action happening quickly (and rightly so, given fuel flowing to a flaming engine needs to be stopped ASAP in the envisaged use)

@notimagain in the Air Transat case you mentioned, was that somewhat similar in the human error area to the Air France crash - where, IIRC, the pilots “couldn’t believe” what the sensors were saying until it was too late to correct the stall, because the information was unusual?

One of the news channels had an ex-pilot on who was saying it could be possible that the vibrations from a bumpy runway could have flicked the switch.
Is this something still possible, has it been discussed here?

I skim read the thread but couldn't see it discussed.

unlikely with safety switches like that.
they are double action switches (somewhat similar to the reverse gear of a car, where you have to push or pull in addition to moving the gear stick).

& you be looking at it happening twice?

No. Not a chance.

@HereForTheFreeLunch unlikely given the lift and flick action, but also - because both switches were affected and because if the runway had jiggled them off, then I’m not sure the plane could have got off the ground at all, with no fuel to the engines

It's been shown that at the time of the crash the fuel switches were “on”. They had both been turned off briefly but in a matter of seconds switched back to the Run position.

If anybody wished to sabotage an aircraft by cutting off the fuel supply then why would they almost immediately turn it back on again?

Could one of the pilots on here explain whether there is ever any legitimate technical reason for performing such a rapid on/off/on procedure?

@MrsGusset because as a pilot, you would know that cutting the fuel at such a low height would not give sufficient time to restart the engines and save the plane. Additionally, there are two pilots - presumably the switches are within reach of both, so one may have turned off and the other back on.

I think the AF crew to some extent weren't well served by the aircraft or some of their training and I think they very quickly got into test pilot and then WTF territory.

AirTransat was a bit different, crews (not just at that airline) had got a bit relaxed about handling minor fuel imbalances. The odd few hundred kilos of imbalance was not that unusual, was usually written off as being down to minor differences in engine consumption and was usually handled without dragging a checklist out...

The crew that night weren't under time pressure but didn't recognise that despite the symptoms and warnings that they were dealing an actual leak..might have been denial but??

Whatever the reason it certainly led to fuel inbalance coming up a lot in sim checks

It's looking like a suicide/murder.

Dual Engine Failure checklist on the 777 ,.maybe the same on the 787 is the only one I can think of that is close to that..but it's a rapid selection of cutoff then straight back to run, no > 10.sec..pause with a discussion in the middle.

But of course you don't start doing that if either or both engines are already producing thrust, you'd try to cross check with your partner before doing it, and you haven't a hope in heck of it working in time if you have only just lifted off.

Doubt it, TBH if a switch was going to somehow bounce out of the gate you'd expect it to happen during taxy out (taxyways can be rough.)

Chances of both switches bouncing out of their gates and then dropping down, one second or so apart, after you've actually got airborne ...not a lot.

Presumably because the pp has not corrected having been referred to as such before.

It seems to me that all the pilots with real life experience are saying there's no way this could have been accidental.

And all the armchair pilots are trying to come up with ways it could be accidental.

So it was most likely a deliberate act by one of the pilots?

How awful. I feel for everyone involved, especially now the families of the 2 people in the cockpit. Someone's son or brother or father did this dreadful thing.