Egyptair Flight MS804 Disappears From Radar — UPDATED

UPDATED 5/21/16: Egyptian authorities have released photos of MS804 debris recovered from the ocean. Here’s a cropped version of one of them:


The size, the shape of the edges, the amount of exposed honeycomb and even the presence of fasteners is quite reminiscent of MH370 debris found in the western Indian Ocean, especially “No Step.” Of course, marine fouling is absent.

UPDATE 5/20/16: CNN has posted a screengrab showing ACARS error messages just before MS804 crashed:

Right-click and open in a new tab to expand


As you can see in the diagram below, there is a lavatory directly behind the captain’s seat. If there is thick smoke in there, it could penetrate down into the avionics bay below:


Gerry Soejatman points out: “It appears that aircraft may have had an in-flight fire and if so, the aircraft maneuvers could be due to Smoke Removal Emergency Procedures, which involves descending the aircraft to 10,000 feet and also opening the cockpit window.”

A reader interprets the ACARS messages:

00:26 ANT-ICE R WINDOW; a fault is in either the right sliding window or fixed window (not the windshield).
00:26 R SLIDING WINDOW SENSOR; the right window heat control unit detects a problem with the sliding window sensor circuit.
00:26 SMOKE LAVATORY SMOKE; smoke detected in the lavatory.
00:27 AVIONICS SMOKE; smoke is detected in the Avionics bay.
00:28 R FIXED WINDOW SENSOR; the right window heat control unit detects a problem with the fixed window sensor circuit.
00:29 AUTO FLT FCU 2 FAULT; autopilot flight control unit (Mode control panel) channel 2 is faulted. Channel 1 still OK so no big deal.
00:29 F/CTL SEC 3 FAULT; the number 3 spoiler elevator computer is faulted. Number 1 and 2 still OK.

If a bomb has gone off near the forward toilet the blast may have damaged the right window heating somehow. There would be a short delay until the toilet smoke detector goes off.


ORIGINAL POST, 5/19/2016:

At timing of writing, 6.15am Eastern Daylight Time, Egyptian and Greek military boats and planes are still hunting for a missing airliner, Egyptair Flight 804, which disappeared over the Mediterranean Sea at approximately 2.30am local time. The plane was three and a half hours into a scheduled flight from Paris, France to Cairo, Egypt.

Airbus has put out a statement which reads, in part:

The aircraft involved, registered under SU-GCC was MSN (Manufacturer Serial Number) 2088 delivered to Egyptair from the production line in November 2003. The aircraft had accumulated approximately 48,000 flight hours. It was powered by IAE engines. At this time no further factual information is available.

Here’s a screenshot of the ADS-B data reported by FlightAware. Note that this data is considered highly unreliable–but at the moment it’s all I’ve got:

MH804 FlightAware

Open in a separate tab to see full resolution. Note that the disappearance is sudden — the ADS-B doesn’t show any descent profile. This would be consistent with a catastrophic loss of electrical power (as perhaps due do a bomb or missile strike) or to someone deliberately turning off the electronic forms of communication, as was done in MH370. Below is a plot of the plane’s last known location.


Worth pointing out that the weather at the time was fair, suggesting that the incident was not weather-related, like AirAsia 8501 or Air France 447. I find it interesting that the disappearance took place right after the plane crossed from one Flight Information Region (FIR) to another–that is to say, when transferring from one air traffic control zone to another. MH370 disappeared under similar circumstances (also in fair weather in the middle of the night.) I would be very interested to see the ATC transcripts–in particular, to know if the plane signed off with the Greek controllers and failed to contact Egyptian ones.

The Mediterranean is a heavily-traveled body of water, both by sea and by air. It is heavily monitored. One can only presume that at the time it vanished from secondary radar screens it was being tracked by primary (military) radar as well. What’s more, based on historical precedent, when planes get into trouble at altitude like this, they tend to come down very close to their last known position. At the exact moment I write this, no debris has been found, but given the good weather conditions and the very small area to search, we should expect wreckage to start turning up very soon.

The Independent notes, “In March, an EgyptAir plane flying from Alexandria to Cairo was hijacked and forced to land in Cyprus by a man wearing what authorities said was a fake suicide belt. He was arrested after giving himself up.” In air crash circles, the name Egyptair is synonymous with EgyptAir 900, which crashed off Nantucket when one of the junior pilots deliberately steered into the ocean. At this point, both terror and suicide remain possible causes in the current incident.

UPDATE 7am EDT: The Guardian has just published this timeline, put out by Greece’s civil aviation department:

02:24: EgyptAir flight 804 from Paris to Cairo enters Greek airspace, air traffic controller permissions it for the remainder of its course.

02:48: The flight is transferred to the next air traffic control sector and is cleared for exit from Greek airspace. “The pilot was in good spirits and thanked the controller in Greek.”

03:27: Athens air traffic control tries to contact the aircraft to convey information on the switch of communications and control from Athens to Cairo air traffic. In spite of repeated calls, the aircraft does not respond, whereupon the air traffic controller calls the distress frequency, without a response from the aircraft.

03:29: It is above the exit point (from Greek airspace).

03:29:40: The aircraft signal is lost, approximately 7 nautical miles south/southeast of the KUMBI point, within Cairo FIR.
Immediately the assistance of radars of the Hellenic Air Force is requested to detect the target, without result.

03:45: The processes of search and rescue are initiated, simultaneously informing the Flight Information Region of Cairo.

It seems, then, that unlike MH370, the flight crew here did not sign off with ATC before leaving their airspace.

358 thoughts on “Egyptair Flight MS804 Disappears From Radar — UPDATED”

  1. Momentarily saw the screen grab of an ACMS computer with the log for Egyptair 804 on pprune before it was deleted by the mods. Avherald has pretty much the same info. Firm evidence of a fire onboard.

    “On May 20th 2016 The Aviation Herald received information from three independent channels, that ACARS (Aircraft Communications Addressing and Reporting System) messages with following content were received from the aircraft:

    00:26Z 3044 ANTI ICE R WINDOW
    00:27Z 2600 AVIONICS SMOKE
    00:28Z 561100 R FIXED WINDOW SENSOR
    00:29Z 2200 AUTO FLT FCU 2 FAULT
    00:29Z 2700 F/CTL SEC 3 FAULT
    no further ACARS messages were received”

  2. Air cond smoke (I think) will normally trigger various warnings eg lav, avionics etc so it doesn’t tell us where the smoke was coming from.

    If it just said avionics it would be more straightforward.

    Nonetheless it could be a significant clue. Protocol would be to descend to 10,000 and land asap as well as messing with switches.

  3. Initial reaction -looks like a fire in the avionics bay leading directly to a catastrophic loss of the flight control system, fbw etc.

  4. According to Jeff’s post above, at 00:48Z “02:48: The flight is transferred to the next air traffic control sector and is cleared for exit from Greek airspace. “The pilot was in good spirits and thanked the controller in Greek.””

    That doesn’t fit, does it?

  5. @JS, Of course in the Valujet case the flight crew communicate with ATC about their problem, but it’s easy to imagine a flight crew being too overwhelmed/surprised to do that.

    Not sure how confident we can feel about the validity of the ACARS data, but a fire does seem to me to explain the non-responsiveness to ATC for nearly three minutes in a way that bomb doesn’t.

  6. @Victor

    That raises the question of why the radar data is not available as it was in the case of QZ8501? Exner nailed the location perfectly from the QZ8501 radar data. I would like to get his take on it.


    I will not be held to the accuracy of the tables I present, Just found them and think they are interesting to read through as we speculate.

    Last cabin fire that resulted in fatalities was 1998.

    Last Incapacitation of the crew due to depressurization of a 2005 flight near Greece.

    Interesting tables, veracity I cannot say
    I withhold guessing what happened to MS804.

    Site also has some of the CVR and other investigation data.
    e.g. Smolensk CVR –

    Fare thee well

  8. @Jeff: “a fire does seem to me to explain the non-responsiveness to ATC for nearly three minutes in a way that bomb doesn’t.”

    It could have been a fire-bomb set off in the lavatory. You’re allowed as many 1/10th of liter bottles of liquid that will fit in a quart sized bag. A group of 4 suicide bombers could perhaps smuggle well over 2 liters of gasoline. Dump the gas on a carry on packed with polyester clothes, light it off with a single, strike-anywhere match, you’ve got a hell of fire on your hands.

  9. If those warnings appeared on the ECAM between 00:26 and 00:29, with associated master warning, master caution attention getters and chimes, how could the pilot be unaware of those failure conditions at 00:48?

  10. @DennisW, It took us a while to get that radar data, as I recall. Info trickled out of Indonesia in dribs and drabs, it was very frustrating! I expect this story may well provide more of the same.

    @All, CNN and the WSJ are now reporting the ACARS data and sourcing it to “sources.” I’m not sure whether they know something we don’t, or would just rather risk being wrong than risk missing a chance to announce “breaking news.”

  11. @jeffwise – agreed, except there’s one difference in this case.

    The Valujet pilot immediately sought an emergency landing in Miami.

    In this case, there’s nowhere to go. So in the aviate-navigate-communicate sequence, you would presumably communicate sooner if an emergency landing was possible. I am not a pilot, but I could see some slight bias towards communicating the closer the plane is to a runway.

  12. @Warren, Well, let’s call a molotov cocktail a “fire” and something that blows up a “bomb.” Yes, either could be triggered as an act of terror. A fire though could also have an accidental cause.

    @Gysbreght: The cheerful exchange in Greek was, I believe, at 11:48Z.

    It’s starting to form a picture, isn’t it? Fire breaks out. Pilots try to deal with it, get distracted, miss ATC call. Then things get quickly out of control…

  13. @Gysbreght

    And the final ACARS message was level 3, if I’m understanding right, which is the highest level, requiring immediate action. The crew may have had a problem interpreting the warnings, especially if the particular failure hadn’t been practiced in the simulator.

    This is the potential problem when systems are computerized – identifying the problem, and determining what to do about it, particularly if in this case the ARINC Databus itself could have affected, and they could have been getting conflicting messages.

    In such situations, the “Communicate” is going to be very low on the list of priorities. AF447 was an example. The crew didn’t trust their instruments and they couldn’t get a handle on the situation.

  14. Regarding the sudden turn in midair, CNN also noted this:

    “What Greek officials described as swerving was likely pieces of the aircraft being picked up on radar as they fell from the sky, U.S. officials told CNN on Thursday.”

    Interesting cos MH370 also swerved before it disappeared altogether. And I always thought it did so a. to avoid something or b. deflected off course by shrapnel from an explosion. Just wondering

  15. @TBill – the empty center fuel tank on 800 was heated by hot July weather sitting on the tarmac. The center tank exploded at 13,000 feet or so.

    After several hours at 37,000 feet, you would think the tanks on 804, empty or not, would be sufficiently cool to prevent a similar spontaneous explosion.

    Also note that 804 wasn’t even built until 7 years after TWA 800 exploded. One would hope that the type of wiring defects and fuel practices suspected in TWA 800 have been corrected.

    Of course, in 1996, lithium (rechargeable) batteries weren’t in widespread use. So, maybe we fixed one cause and created others.

  16. The ACARS messages:
    00:26Z 3044 ANTI ICE R WINDOW
    00:27Z 2600 AVIONICS SMOKE
    00:28Z 561100 R FIXED WINDOW SENSOR
    00:29Z 2200 AUTO FLT FCU 2 FAULT
    00:29Z 2700 F/CTL SEC 3 FAULT

    I know what the terms ‘smoke’. ‘lavatory’, ‘avionics’, ‘window’, ‘fixed’
    sliding’ mean, in general. But in the context of an ACARS message, exactly, not. Further, the bottom two lines look ominous, but what, word-for-word, do the mean?

  17. @Millhouse, I’ve added some interpretation to the updated post. If anyone would like to add further detail that would be very welcome.

  18. A partial answer to my question comes from ITV News:

    ‘FCU’ refers to a unit in the cockpit which the pilot uses to input instructions into the flight computer, while ‘SEC 3’ is the computer which controls the spoilers and elevator computers.

    Complete story at:'FCU‘ refers to a unit in the cockpit which the pilot uses to input instructions into the flight computer, while ‘SEC 3’ is the computer which controls the spoilers and elevator computers.

  19. The below article provides a good summary of the threat presented by high yield explosives, eg PETN.

    80 grams of PETN planted in the lavatory or the forward cargo hold in the course of one of the day’s previous flights would do it.

    Favoring some sort of electrical fire given other elements of the mix (beginning with the low prability of a catastrophic failure in an Airbus airframe at cruising altitude) is perhaps not indicated in this instance.

    The fact that more people do

  20. Does this narrow down the possible placement ( if that is the case) of a combustible substance or explosive device solely in first class then? Sorry if i am being captain obvious here. Enjoy this blog so much and everyone’s input herein.

  21. A cockpit oxygen-fire maybe like EgyptAir 667?

    The sequence reading this ACARS looks to me like a fire starting in the right side of the cockpit, first detected by the right window heat sensors, then the smoke going aft in the lavatory and further down in the avionics bay.

    Knowing how fast that cockpit-fire spread on EgyptAir 667 and burned a hole in the fuselage I can imagine the crew had no time to communicate.

  22. Woes started on the ground ??

    QAR MEDIA LOW — backup FDR out of memory?
    QAR (Quick Access Recorder): Copy of the FDR and thus records the same
    data. It allows a quick and easy recovery of the raw data recorded in FDR.
    Only raw data is used for the event analysis.

    The MS804 ACARS mssgs say “ADR 1”, the following is for an otherwise identical “ADR 2” situation. Not entirely sure what it all means, but please note the ultiate text “NO DATA FROM ADR” and the requirement to replace the ADIRU. This implies MS804 suffered severe failure of its ADIRU-1 device ??

    Internal Failure of the ADR 2 Identified by the ECAM 2
    1. Possible Causes
    – ADIRU-2 (1FP2)
    2. Job Set-up Information
    A. Referenced Information
    ——————————————————————————-REFERENCE DESIGNATION ——————————————————————————-
    AMM 31-50-00-710-001 Ground Scanning of the Central Warning System
    AMM 34-12-34-000-001 Removal of the ADIRU (1FP1, 1FP2, 1FP3)
    AMM 34-12-34-400-001 Installation of the ADIRU (1FP1, 1FP2, 1FP3)
    3. Fault Confirmation
    A. Test Do the operational test of the central warning systems (system 2) (Ref. AMM TASK 31-50-00-710-001).
    4. Fault Isolation
    A. If the test gives the maintenance message FWC 2 : NO DATA FROM ADC 2:
    – replace the ADIRU-2 (1FP2) (Ref. AMM TASK 34-12-34-000-001) and (Ref. AMM TASK 34-12-34-400-001).
    B. Do the test given in Para. 3.

    FCU = Flight Control Unit = ATs & APs

    F/CTL SEC 3 = “Spoiler & Elevator Computer” 3 of a/c’s Flight-Control surfaces:
    “Consequently, two types of computers may be distinguished:
    2 ELAC (elevator and aileron computers) and 3 SEC (spoiler and elevator computers) on A320/A321”

    “Q2. If a/c is dispatched with SEC1 inoperative {under MEL}, then which of the flight control surfaces are not available for you in flight?

    A2. Spoilers 3 and 4 would not be available.”

    “Q57. How many SEC computers are installed and what are their functions?

    A57. There are 3 SEC computers on board the A320. The SEC’s provide output control to the Spoilers and will be back-up for control of the Elevators and the THS.”

    SPOILERS 3-4 are the third and fourth, moving outwards, of five spoilers on each A320 wing, not including two additional at the tips of both wings

  23. Re: MH370 and the recent “SDU logon” thread…

    Isn’t that a great idea ? Compared to the cost of two years of seabed searching, what would be the cost of a couple solid days of simulated MH370 failure scenarios ?? Especially if Boeing could incorporate such scenarios into its Pilot Training Program… what would be more meaningful & relevant to aspiring pilots than arguably real-life situations, from major headline events ?

    Some quick thoughts:

    ELEC AC BUS L main powers
    1) Window Heat (L)
    2) Pri. hydrlc Engine Driven Pump (EDG) (L)

    So, logically, ELEC AC BUS R powers:
    1) Window Heat (R)
    2) Pri. hydrlc Engine Driven Pump (EDG) (R)

    The backup pumps L/R are AC motor pumps (ACMPs), obviously requiring AC power, and implying that they run off of the L/R AC XFR buses…

    Similarly, the primary CENTER pumps are also ACMPs, likely drawing power from the L/R XFR buses…

    So, if you imagine some scenario, w/ MH370 flying w/o either MAIN AC bus, then…

    no cockpit FD window heat anti-icing
    no primary L/R hydraulic EDPs

    Next thought is the XFR buses still receiving power from the Backup Generators on the engines… Probably that kicks on the L/R DEMAND secondary hyd. pumps and keeps powering the CTR main hyd. pumps… probably enough pressure for cruise-style flying ??

    How long until the FD windows ice over, obscuring pilot sight ? About 1:03 ?? Pilots flub the FMT towards Banda Aceh b/c they can’t see straight ahead, through caked-over windshields, which haven’t defrosted yet? Maybe some psychology, they were holding on to the thought of landing at Banda Aceh, enduring cold and airless conditions for an hour, and after missing the turn, they lost their will to survive ??

    AIMS cabinets still receiving uninterrupted DC power… how to wipe out the Flight ID ? Manually erase it from memory ?? That implies hijacking… Shutdown the computers ? That requires pulling CBs and could be construed as an electrical emergency response, to some sort of catastrophic, widespread danger.

    What happens without AIMS ? Most likely that implies something in the vicinity of vestigial backup mechanical flight mode, on two spoilers and manual trim of the stabilizer. B-777 mechanical flight mode is sufficient only for flat / straight / level flight maintenance, and seems inconsistent with the second IGARI turn back towards Khota Baru & Penang, but perhaps consistent with the weaving & porpoising path after the U-turn. However, if the co-pilot actually did respond in some way to radio at 1:30am, then they still had some sort of residual electric power during the first 10 minutes or so… maybe, like OZ stated, they had reserve battery power for 10 minutes, and responding to the radio call used up the remaining battery charge ??

    There’s actually alot of information out there, but it’s very hard to grasp and navigate. Pilot Training flight simulations look like a great option, no extra cost to anybody, meaningful relevant real-life-like situations to train on, maybe the next graduating class could solve the MH370 mystery ??

  24. The cockpit fire on EgyptAir667 breached the hull within 30sec. I read.
    Photos of the right cockpit windows, side and interior give an indication of the damage done within a few minutes.
    If this happened at altitude things went very fast with no change of saving the plane.

  25. @Jeff Wise

    Maybe something to correct in your timeline update?
    It still mentions 3:39:40 as the aircraft signal was lost.
    Yesterday this was corrected by the Greek to 3:29:40

  26. @Jeff,

    ‘As you can see in the diagram below, there is a lavatory directly behind the first officer’s seat’

    Is that correct? It doesn’t look quite right to me.

  27. A sudden, severe fire in the cockpit is looking increasingly likely, with smoke percolating through to detectors in the lavatory and avionics bay. But opening a sliding window at 37,000ft would be highly problematic imo.
    The normal procedure would be to make an emergency descent and possibly depressurize and or open a window then. Possibly the onset was too sudden to permit normal procedures.

  28. There have been an awful lot of smoke/fire indications reported on the A320 over the last several years, most resulting in a precautionary landing or divert or similar.

    I’m fairly certain this is usual with any type and I doubt it indicates a problem.

    Hull losses (A320 family) due to fire are not so frequent. Last due to fire, I believe, was in 1993.


    Sketch of Scenario to explain MH370…

    1) As the Inmarsat JON article has always shown, consistent with BTO data actual & calculated, the a/c slowed significantly, from 500-510kts along the radar track, to 420-440kts along N571 from MEKAR to NILAM towards IGOGU

    2) Actual BTO data imply the FMT at 2:35-36am, according to the intersection of the best-fit trendlines for pre- & post-FMT BTOs

    1+2) At 2:35-36am, travelling at 420-440kts along N571 towards IGOGU, the a/c reached a point, along that air-route, perfectly aligned, towards the SSE, towards the main runway at Banda Aceh

    3) The aircraft completed FMT onto a heading of ~170 towards Banda Aceh
    3A) En route, the pilots received but somehow didn’t hear the Sat-Phone call at 2:39-40am
    3B) then radioed distress and “cabin disintegrating” to unspecified sensitive ears about 2:43am

    4) Knowing that the a/c was doomed, and unable to save the lives of any aboard, pilots Shah & Hamid used their few remaining moments to divert the aircraft 20-30 degrees to starboard, onto a SSE heading, AWAY from LAND out into the MIDDLE of NOWHERE, intentionally so as to spare the lives of all those on the ground, when the plane invariably crashed… SIO crash site was, in fact, Shah & Hamid’s desired crash location, effecting their last desired flight outcome, of injuring no other persons on the ground, when the a/c crashed

  30. …and that was a fire which started due to inopportune landing circumstances/crash. Not an electrical fault per se.

  31. Ge Rijn: That fire was on an Egyptair 777. The faults were identified and advisories were sent to all users. But there is some doubt as to whether they were done on 9M-MRO which flew on MH 370.
    I don’t think there is any similar incident on A320s in recent years, so it is unlikely to be a factor in MS 804.
    For more on this 2011 incident see:

    which contains a link to the full report.

  32. WRT smoke procedure on the A320 the QRH was revised so that it’s no longer advised to reset the generators to try and isolate the source of smoke. You could potentially lose both buses if this is done while batts are charging and not be able to get them back on.

    The ECAM may still instruct you to reset the generators but the revision directs not to do this even if so.

    I wonder if this may have caused any confusion in the heat of the moment.

  33. @All

    WRT MH370: Regrettably, this does tend to support what we discussed earlier on, ie. that a catastrophic event such as a fire, from whatever cause, almost invariably results in the loss of the a/c within a short space of time.

  34. @ABN397
    Yes it’s a different plane offcourse.

    To me it is only an example to give another possibility for a very rapid and intense fire which started near the right cockpit windows.
    Oxygen bottles, hoses and masks are generaly in the same places I assume.
    Imo it could fit the ACARS sequence well instead of an explosion or a slower evolving fire.

    We’ll see when the black boxes are recovered.
    Anyway information from them might be also of importance in understanding more about the vanishing of MH370 imo.

  35. How sad to look at.
    This must have been a very forcefull high speed impact.
    Looking at those twisted torne apart pieces the plane must have been shredded.
    How sad for all those families and friends who will also watch this pictures.

  36. @Erik Nielsen. To me the most relevant/interesting thing arising from the MS804 tragedy is the lack of comms in spite of several minutes elapsed time between event initiation and crash. To all of the nay-sayers who regard “non-communication in spite of multiple redundancy on VHF/HF radios and associated power circuits”: what say you?? I have never been able to understand why the majority of posters on JW, inclusive pilots and avionics boffins, subscribe to the “deliberate” [malevolent] action thesis rather than serious cascading problem affecting multiple systems. In which instance the urgent and intense aviate/navigate challenge, plus serial (deliberate, per protocol) isolation/disabling of circuits that could render radios unusable. Is this not a perfectly plausible explanation for initial turn (to head for alternate) and lack of comms in the minutes after a serious problem? Why are we all so quick to reach the conclusion of deliberate and nefarious diversion by persons unknown? If you choose (as I do) not to regard the MY radar “record” as unequivocal, why persist in discounting the “accident and attempted diversion” scenario?

Comments are closed.