MS804: smoke in avionics bay, then flight control computers begin to fail

Egyptair operations centre plane screenshot


This is a screenshot of ACARS messages automatically datalinked from Egyptair flight MS804 to the airline’s operations headquarters shortly before the aircraft appeared – on radar – to go out of control.

News of the ACARS datalinked message first appeared in the Aviation Herald, a respected Salzburg, Austria-based journal that publishes on aviation safety issues. The French accident investigator BEA, which is helping Egypt with the inquiry, has confirmed the existence of the information. (ACARS = Aircraft Communications Addressing and Reporting System).

The printout/screenshot shows that all was more-or-less well in the aircraft until smoke was recorded in the forward lavatory just behind the flight deck.

A minute later smoke was recorded in the avionics bay, which is below the flightdeck floor, therefore close to the forward lavatory. The avionics bay contains all the electronic sensors and computers that provide the pilots with the information they need, and connections to the computers that direct the flight controls.

Two minutes after the smoke warning for the avionics bay the FCU 2 (second flight control unit) recorded a fault, and the SEC 3 (the N0 3 spoiler/elevator computer) also recorded a fault.

According to the Aviation Herald’s source, the ACARS feed then stopped.

Warnings about all these would have appeared on the pilots’ instrument panel.

There were other symptoms in the ACARS messages that popped up before the smoke was sensed by the system. These related to cockpit window de-icing and related window sensors, and were nothing to do with the aircraft controls but may have been yet another symptom of early fire damage to electrical systems.

The question now is whether the fire that caused the smoke was the result of an electrical fault – for example a short-circuit caused by damaged wiring – or whether some form of explosive or incendiary device was used – for example by a terrorist – to generate a fire or other damage.

The fire appears to have propagated fast. Flight control computers were failing within two minutes of the avionics bay smoke warning. If more of them failed subsequently the pilots would have limited means for controlling the aircraft, and with fire present, crew stress would have been extremely high.

That might explain the fact that there was no distress call. The pilots were trying to understand what the cause of the fire warnings was, where the source of the fire was located, whether they could do anything to stop it, and coping with a gradual degradation of their ability to control the aircraft.

These facts, providing they are not some kind of macabre coincidence, may have provided the basic reason why the aircraft went out of control and crashed.

But it is still not clear whether this situation was the result of terrorist action or an aircraft fault. Certainly no terrorist group has claimed responsibility.

And that answer is unlikely to be forthcoming soon. Even the recovery of the cockpit voice recorder and flight data recorded may not provide absolute proof of terrorist action – or lack of it – although the data will probably provide compelling circumstantial evidence.



17 thoughts on “MS804: smoke in avionics bay, then flight control computers begin to fail

  1. You’re spot on. Just the recorders may not contain the conclusive evedence for the cause(s), just insight into the symptoms. Hopefully the wreckage will be recovered soon before possible traces of explosives will be washed away. Sabotage or technical, either way, it will take a lot of research to find out what caused this accident and it is imperative to do so. It will probably also encourage a rethink on more comprehensive recording and means of retention of data


  2. Hello David, Thank you for yet another good article. We always enjoy your stories on here and in the Flight international and also the interviews you give on the news and Aviation programmes. I however want to point out that the Aviation Herald is run By Simon Hradecky from Salzburg in Austria and not from the US as stated in your article. I hope you don’t object to me pointing this out.


  3. Hello David .. thank you for this article.. but about “That might explain the fact that there was no distress call” I think that any strange thing about the plane or any error the crew’s first reaction is a distress call .
    I rule out any hypothesis about technical error since if there was any technical error I think there was plenty of time to make a distress call , and the crew is highly qualified to deal with any stress ?!
    this my humble opinion


    • In an ideal world I would agree with you.

      But just for context, in this case a distress call would not have saved the aircraft, and would have made little difference to the search.

      The classic set of priorities for pilots is aviate, navigate, communicate. All three need to be done, but the priorities are in that order.


  4. Terrifying information you are presenting here. A fire spreading from the forward lav to the avionic bay within two minutes would require quite some energy.


    • The smoke alarm triggered first in the lavatory, then in the avionics bay, but the smoke may have registered first in the lavatory, but the heat source was not necessarily there.


  5. Why does everybody who has some sort of unrelated experience in aviation, be it Microsoft Flight Simulator, amateur level or professional level, always feel the irresistible need to voice their opinion or their own little speculations about air accidents. Let the professional investigators take care of that, and stick to objective media instead of spreading this social media mumbo jumbo. How would you feel if you were a victim’s relative and you’d get every average Joe feeding you different “theories”? Whatever happened to integrity in aviation??!


    • Just another perspective; the accident aircraft was an A320. thousands in operation. Operated, maintained dispatched loaded by thousands every day. These people believe what they’re doing ensures safe operation. The engineer that released the aircraft for its last flight believed it was safe. But if he was not on the flight himself, rest assured this person had a good number of sleepless nights, not knowing what happened and having to face investigators and prosecutors.
      People just have a genuine interest and ignoring the event for one or more years is not good enough for them. The internet is a good place to exchange thoughts, not speculate. David has been very competent and authoritive for many years on operational safety and this site is well moderated. I don’t think it’s mumbo jumbo


  6. Bart I agree with your point about the release of information, and the inevitability of discussion about the released information. The importance of respecting the public’s intelligence by telling them facts when they become known has been a theme of mine for many years, and as you can see I have added another blog entry on the subject.


  7. Hi David,
    I think we know, don’t we, that heat rises, and that the avionics bay is lower down…


  8. Hi David
    I have been reading your words since 1979 when my dad got me my first copy of Flight International (I think you were Air Transport Editor back then) and turn to this blog for your opinions now. I read at the weekend a comment attributed to an aviation professional which has puzzled me about this accident If it were a fire, then the time between the aircraft systems reporting it, the last ACARS communication and loss of the aircraft was very short (too short?), and if it were a bomb then that time scale was too long. Have you any thoughts on this? David


    • The information we have is too sparse. It doesn’t contain any proof that this fire was the result of sabotage, nor of an electrical fault. I have heard it proposed that this was the result of an amateur device that failed to explode and just burned instead. Maybe it was, but we have absolutely no evidence to back this theory.

      And all the other theories that are doing the rounds may be interesting too, and they might fit the facts, but they are still guesses.

      Even when they find the recorders we may not be able to prove the cause, but it will help rule some theories out, which is an improvement on where we are now.


  9. If you want “Mumbo Jumbo”, look no further than this lame-named broadcast blog called, APG on the Egyptair 408 hull loss. Egyptair 408 is not only lost at sea, it seems. APG features Jeff Nielsen, a Delta(Don’t Ever Leave The Airport) MD-80 Captain and Nick Whatsisname, a Virgin A-330/340 Captain. If only Jeff didn’t ?!…To ease the monotonous drudgery, he co-opted a couple of others too for anesthesia , but not on this occasion.
    Jeff makes CNN-CABLE NOISE NUMPTY’S anal-ist, Mary-“Flying Blind”, deaf and dumb- Schiavo, sound coherent, among other “expert” panellists of repetitious talking heads, without a single central -let alone peripheral- pertinent qualification to anchor what she pontificates upon with such certainty. Still, if Jeff replaced her it would ease CNN’s inanity somewhat, as he is at least ATP type-rated. And their imposition of time constraints would make him bearable in a shortcast sense. Less is more and hope springs eternal.
    He can’t come to Europe, as there is an ICC arrest warrant out for him, for crimes against broadcasting, of which I am a lead prosecution advocate. I’ve appealed to him to reduce it before, alas to no avail. In fact, Jeff ultra long-haul broadcasting or long narrowcasting propagates without attenuation in inverse proportions to appeals for curtailment.
    Capt. Nick may be a pending ICC indicted co-defendant who may need to seek political refuge in the USA. Thus far, being the thin end of a wedge, he is merely aiding and abetting Jeff’s inanity.
    Jeff’s previous broadcasting alias was, “Catholic Pilot”, no, I’m not joking. When I said to him, that being involuntarily co-opted into the Irish branch of the Church of Rome after birth, I considered it in hindsight -Catholic retrospective foresight-an asset of life, as it gave me so much to recover from, it went over his head at FL990, without contrails. Jeff shows no signs of recovery, thus far.
    For illustration, I even invoked the Learmount broadcast narrative of between 5 and 10 minutes. Alas Jeff, thinks his Mad Dog-80 is 180 Minutes ETOPS/LROPS qualified. I’m sure he’s a nice guy socially, but hopefully his CRM is more effective than his shambolic broadcasting. And he makes up for short-haul flying with long haul broadcasting. Oh for a tailwind for any haul.
    I’ve never been able to listen to more than 15 minutes in the interest of preserving my First Class Medical. I’ve noticed that he uses his hat on the pillow as a background prop. If this hat had a female topless cabin crew member or hotel hooker in suspenders under it gyrating or pole dancing, it would be worth watching, but alas Jeff has not been so innovative.
    EICAS/ECAM ALERT: Viewing and listening to more than 10-15 minutes of Delta drivel will lead to a comatose and neural vegetative state meltdown.


  10. David, I’m surprised that you assumed the Aviation Herald was of the United States. If it was, they would not be using UTC, assuming it had something to do with the Ms. Universe of the Zulu tribe. All times would be of the ambiguous AM/PM format, as it is in all US civil airports, amazingly. How many missed flights occasioned, I wonder?. After all, the FAA has only in recent years recovered from using the ridiculous Fahrenheit scale in METARS’. Alas they are still infected with using LBS instead of KGS for weights. B-744 MTOW =396 TONS or 780000LBS. In my simple mind, the latter has been hyper-inflated to Lira meaninglessness to convey weight. Japan, Korea and Taiwan are also guilty due to US post-war aviation legacy provision.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s