Why we think a fire on MH370 is unlikely; 777 has lots of smoke/fire detection to give ample warning; hypoxia theory

One of the early theories about the disappearance of Malaysian Airlines Flight MH370 has been the prospect of a fire in the electronics bay, or elsewhere in the airplane, that led pilots to begin pulling circuit breakers to isolate the fire. This theory continues that the transponders and radio communications, including ACARS, systematically failed as a result of the fire.

This scenario, while it can’t be totally ruled out, is unlikely, in our view.

The best example of this scenario is Swissair Flight 111, a McDonnell Douglas MD-11, that crashed off Nova Scotia in 1998. Shortly after takeoff from New York for Geneva, Switzerland, the pilots smelled smoke and smoke eventually began to emerge in the cockpit. However, the pilots radioed Air Traffic Control of the situation and began to divert to Halifax. It is true they began to pull circuit breakers, per procedure, to isolate the problem. However, even as it got worse, the pilots radioed updates and the decision to dump fuel prior to an emergency landing. Thus, this fire scenario hypothesized for MH370 breaks down. Tragically, Swissair 111 crashed as the fire went out of control. This is an important point that will be addressed below.

Associated with the theory that a fire somehow led to hypoxia is another of which we’re skeptical. This would require a hull breach and the failure warning equipment to alert the crew.

Another theory that surfaced last week is the presence of lithium-ion batteries in the cargo manifest. At least one or two plane crashes occurred because lithium-ion batteries in cargo caught fire and went out of control. These are the UPS Boeing 747F crash in Dubai and the Asiana Airlines 747 crash in the Pacific. However, in both cases, the pilots radioed that a fire existed. As we know, there was no communication from MH370.

Another reason a fire is unlikely on MH370 is this: according to an Airbus study in 2012, it takes just eight minutes for a fire to go out of control and the airplane has to be on the ground within 15 minutes, with the implication being that the airplane would be lost beyond this time. We know from the satellite pings that MH370 was airborne for up to seven hours after contact was lost. If any fire occurred strong enough to cause the loss of all radios and transponders, it certainly would have resulted in the loss of the airplane well before seven hours.

Furthermore, any pilot knows that with a fire, you land at the nearest available airport. Additionally, flipping the transponder to 7700 is a mere flick of the wrist and alerting potential air traffic in the vicinity (so, among other things, you don’t collide with it) is also important.

The theory that a fire overcame the crew and passengers and the airplane flew on for hours on auto pilot or in a computerized stable state also seems unlikely. The fire would essentially had to have snuffed itself out to avoid destroying the airplane.

Finally, there is the airplane itself. We have consulted two 777 pilots, both of whom believe that a fire or electrical failure of such magnitude as to take out all five radios, two transponders and the ACARS system simultaneously is extremely unlikely.

“There is a big effort [by Boeing] to do separation of the electronics,” one of the pilots tells us. “It’s unlikely a fire would take them down.”

As for pilots being overcome by smoke, one of the 777 pilots we talked with points out that first, the pilots have a full-face smoke mask to protect them from smoke and second, “the airplane [design] does a god job of keeping smoke out of the cockpit.”

Finally, all cargo compartments and the electronics bay have smoke and fire detection systems that would provide ample warning to the pilots of something amiss–which in turn would allow the pilots time to radio Air Traffic Control.

For these reasons, we do not believe there was a fire on board.

Debunking the 45,000 ft hypoxia theory

An early area of speculation evolved around the report that MH370 had gone as high as 45,000 ft and descended to as low as some 25,000 ft. Theorists suggested this high-altitude report was to depressurize the cabin and kill all the passengers, cabin crew (and potentially the legitimate pilots), enabling the Bad Guys (whoever they might be) to steal the airplane without having more than 200 people to contend with.

Both of our 777 pilots we consulted pointed out the obvious: you don’t have to go to 45,000 ft, which is nearly 2,000 ft above the certificated service ceiling of the 777, to suffocate the people on board. This can be done at 25,000 ft.

Furthermore, 45,000 ft is probably achievable only with a “light” airplane (i.e., one where most fuel has been burned off). MH370 was less than an hour into a six hour flight (plus reserves) and probably wasn’t light enough (though our pilots did say actual calculations with the known payload and fuel on board would have to be undertaken to be sure).

Finally, 45,000 ft would be flirting with the “coffin corner,” the flight envelope in which engines and aerodynamics would be hard-pressed to remain stable. Any slip and the airplane could suffer a high altitude upset.

Such an upset could explain why the airplane descended from this reported 45,000 ft to 25,000 ft until the aircraft was recovered–but this suggests a pilot with a skill level.

Perhaps a more plausible theory–if the altitude reports are accurate–is a fight in the cockpit, according to our pilots, that caused the airplane to go out of control.

23 Comments on “Why we think a fire on MH370 is unlikely; 777 has lots of smoke/fire detection to give ample warning; hypoxia theory

  1. What about some freak collision with a (long range) drone ?
    ( I still expect this to solve into an accident )

  2. Days ago, I too began walking back from the fire on board theory.
    I also no longer believe this airplane rapidly climbed to FL-450, then a rapid decent to FL-230. The autopilot does not do those types of aggressive maneuvers, so someone would have had to be on the yoke. Also, I have heard nothing from RR that engine thrust was increased for the climb.
    Finally, depressurizing the airplane would make it extremely cold inside, with temps bottom out well below 0 F or -17 C. This drop in temp would be very rapid, and descending to FL-230 alone would not warm up the airplane. The airplane would have to be repressurized, only then will it begin to get warmer. 15-20 minutes at these temps would be deadly to anyone not properly dressed.

    • “Also, I have heard nothing from RR that engine thrust was increased for the climb.”

      I know those are powerful engines, but can they really take a 777 up to 45,000 ft at cruise settings? I am assuming it was at cruise when this occurred?

    • I have no experience as a pilot that is why I ask your input on the following assumption
      If pilots are, one way or the other , disabled and as a consequence pull the stick inavertendly, does the plane can go up to 45000 feet and then as it stall fall down quite abruptly to 25000 and continue on a “self stabilized” course ???

      • At those altitudes there is no margin. Special training is required to fly at that height and it remains a delicate operation under the best circumstances. The airplane would not remain stable under the scenario you describe and would quickly fall out of the sky.

        It is called coffin corner for a reason. At that altitude the speed and the angle of attack have to be maintained very accurately within an extremely narrow margin. All this has to do with the basic fact that the higher you fly the thinner the air is. It is indeed equivalent to skating on thin ice.

  3. There remains a lack of information, and I believe this is by design as it maintains the most flexibility for authorities to settle into their own best-fit explanation to share months down the road. Not ‘best fit’ as in factual, but ‘best fit’ as in how to minimize financial liability and maximize a public sense that all is groovy in commercial aviation.

    Regarding the analysis, timing is everything. In the event of even a relatively minor system problem, any competent flight crew leaving the safety of land for an over-water, non-radar crossing is going to be quick to (a) communicate and (b) navigate directly to a landing. So, if the timing shows nav/com elements being shut down PRIOR to the timing of a potential fire, yeah, that points toward mal intent on the flight deck. As for the FL450 climb, the data has always seemed sketchy on this, given that the transponder was shut down early. What is the definitive source for FL450 as a ‘fact’? Also, perhaps those with actual 777 experience can explain: does the autopilot allow for a max altitude input, wherein the aircraft climbs until the system says no more climb is possible? Or, what exactly would a pilot have to do to get this aircraft to go that high?

  4. By determining the elapsed time for each of the hourly handshakes between ACRS and MD 370, a series of concentric circles can be constructed with the ACRS satellite at the center. The radius of, each of those circles can be established according to the handshake’s elapsed time just as was done for the two arcs shown above that are actually two segments of the original circle drawn according to the handshake elapsed time when MD 370’s position was last known.

    The several concentric circles could be useful in several ways. If the radius of each successive circle were decreasing, this could indicate a flight path toward the ACRS satellite, therefor trending toward the west. If the radius of each successive circle were increasing, this could indicate a course trending away from the ACRS satellite, so, toward the east. Better yet, one of two possible courses can be established using the concentric circles.

    Based on an assumed ground speed for the 777, a vector (scaled in length according to the map scale) can be drawn from that last known location on the initial circle and rotated until the tip of that vector touches the second circle. Two vectors would need to be drawn from the origin on the first circle to an intersection with the second circle, one vector rotated clockwise the other counterclockwise,

    One of these vectors could provide a fair approximation of the course angle flown when departing from that last known position, the counterclockwise vector illustrating a possible course into the Bay of Bengal, the clockwise vector illustrating a possible course toward the South Indian Ocean. The angles of the two courses could be reset from circle to circle with diminishing accuracy as the hourly time progressed from one handshake to the next. But this concept could greatly reduce the size of the areas to be searched.

    And if two or more concentric circles were to fall atop each other, this would indicate that the 777 was stationary during those particular handshakes, so it had either landed somewhere or crashed. The area to be searched would be greatly reduced in this case.

    This concept is difficult to put into words and could be better understood with a graphical presentation of the concentric circles drawn on a map and with the possible course vectors intersecting the successive circles

    • The Southern route has been concluded using data analysis by the AAB. Based on the confidence of the Aussies a few days ago they probably had a preliminary finding at least since before then.

    • Hourly pings were very usefull for the search … your explanation of the math behind it is cristal clear thanks

  5. IMHO EXCELLENT analysis as to probable cause. Regarding the gazillion theories put out by ( insert pundits, press, ex spurts, etc here )- ALL require a linked series of low probability items- which while NOT impossible, stretch the plausible and probability games beyond belief.

    An overly simplified explanation of ‘ odds’ and ” probability ” follow. The old saw abut flipping a coin n times and each time it comes up heads. So the question becomes ” what are the ODDS that the next ( n + 1 ) flips comes up heads ? ((.9 percent give the wrong answer . The ODDS are the same for each flip eg 1/2.
    However the PROBIBILITY of the next flip being the same ( heads ) is for example 4 flips and what is the probibility of the 5th flip being heads.. it is 1/2 * 1/2*1/2*1/2*1/2 or 1 in 2^4 or 1 in 32 . . . or 1/64 depending on how one counts the first flip.

    Now in the case at hand … the ODDS of say a comm failure and a fire and a depressurization and a ghost flight for x hours are NOT simply 1/2 for each event Even if we assume a 1 in 100,000 chance for each event ( coupling 2 or 3 or 4 for each event gets to be a probability of 1 in MILLIONS OR BILLIONS.

    Not impossible – but VERY VERY . . . . . .VERY improbable !!!

  6. All true. HOWEVER, something incredibly unlikely had to have happened, or else the plane would have landed on schedule. Or, at the very least, it would have been found along the initial flight path.

    We are only left with the possibility of an incredibly unlikely event or series of events.

  7. Yep.
    ValuJet 592, another SwissAir 111-like incident except that it was cargo that set of the fire. But the pilots did radio in ATC

  8. I do not agree with your opinion rexcluding a fire on board. The SWR 111 fire was a slow progressing fire. Now if you have a fire similar to the 777 fom Egypt Air, this can and will knock out all crew eforts to even try to comunicate so the radios are not disabled but merely not used dur to a very intense fire. If because of this fire a rupture in the fuselage occurs then we have a decompression in altitude leading to the extinthion of the fire for lack of oxigen. Passenger masks will not drop as per procedure for in flight fires and crew only has the last cpability of changing course.

    • And the plane makes several turns, changes altitudes, slows down and speeds up?

      The five radios are routed in separate paths to avoid losing all at once.

      There are two transponders and one ACARS and these are silenced….

      The pilots of SR111 had time to radio and radioed several times.

    • grooooan- any fire enough to disable physically separated cables and power supplies within a few seconds/minutes is still going to allow plane on autopilot or even under manual control for another 5 or 6 hours ?

      Puuuuhhhhlese — get real !!

  9. Certainly the theory of a disturbance or struggle on the flight deck would explain the erratic flight path over the Straits of Malacca and also perhaps the extreme changes in altitude.

    The fact that this was preceded by a routine call exiting Malaysian airspace and that at that very precise moment between exiting Malaysian airspace and entering and checking in with Vietnam ATC, everything failed all at once is just ridiculous. This was a pre-planned moment when the unusual behavior of the aircraft would be least likely to be detected. For a period the Vietnam ATC authorities would not have been unduly worried, but after perhaps 5 minutes the failure to check in would cause concern. The assumption would be radio problems and hence the request for another aircraft en route to Tokyo at the time to try to raise MH370 on the emergency frequency. The pilot did establish contact although rather muffled. This would have once again given the Vietnam ATC authority some comfort. However by the time that nothing further had been heard and the alarm was raised the aircraft was already established on a new reciprocal route and effectively invisible.

    If however the aircraft settled on to a southerly course after various turns over the Straits of Malacca I am mystified why the air defence radars in Singapore did not notice. Certainly Singapore has the most advanced military hardware in Asia (and better than most European countries) and to my personal knowledge they have radar coverage that extends to 250 nm. They also take the Island’s security very seriously and would have mapped anything that differed from or did not have a logged flight plan.

    To avoid detection the aircraft would have had to have flown well to the west of Sumatra (or at low level) and then adjusted course again to the south east (both requiring pilot intervention). Why do we not hear anything from Singapore? They are extremely professional and anything untoward would have be noticed and tracked. However Singapore is saying nothing, unless of course the aircraft headed north, which may explain why Singapore has nothing to say and perhaps why nothing can be found in the southern Indian Ocean.

    One final question to our readers and that concerns the auto pilot. In this case my own knowledge is lacking and if anyone knows the real answer I would like to hear it. My question is this – if an aircraft is on autopilot and both engines flame out, for whatever reason (lack of fuel), what is the default action of the auto pilot, without any human intervention? Does it continue to try to maintain altitude and end up putting the aircraft into a stall or does airspeed become the determining factor, or being unable to maintain altitude does it disengage? And then how would the aircraft react – once again without human intervention? Gentle descent, or being out of control may it drop a wing and spiral into the ground / sea?

    • The autopilot works with engine power and is very good at it when power is available. But with both engines out the autopilot becomes useless. In that case only an experience pilot can glide the aircraft and avoid an abrupt stall high in the air.

      There are many precedents. The most recent was an Airbus A320 on the Hudson River; but prior to that there was an A330 from Air Transat in the Açores and a Boeing 767 in Canada. All three aircraft landed safely with both engines out.

Leave a Reply

Your email address will not be published. Required fields are marked *