Electrical Fire on Board Malaysia Airlines 370?

My friends are emailing me with the latest theories about Malaysia Airlines 370. There are many articles and blog postings (example) that posit that an electrical fire on board the airplane caused the crew to try to divert to a nearby airport. The original posting from Chris Goodfellow suggests the following:

  • an electrical fire caused smoke in the cockpit
  • the pilots pulled lots of breakers
  • the breakers that they pulled disabled everything that sends signals out of the airplane, e.g., the transponders and the ACARS
  • the breakers that they pulled had no effect on the air data computers, attitude reference systems, or autopilots, thus enabling the plane to continue to fly on autopilot for 6 more hours
  • the pilots tried to divert to a nearby airport

This does not match up very well with another bit of information we have about Flight 370, i.e., that the airplane was diverted (via the FMS (like the GPS in your car)) to an IFR intersection, which is an arbitrary point defined by a five-character code. If the pilots wanted to go to an airport they would presumably have typed in the four-letter airport ID instead of a five-letter IFR intersection in the middle of nowhere. (e.g., one could go to BOSOX with an airplane GPS and land on top of an exurban dentist’s McMansion and SUV collection or one could go to KBOS and find an assortment of two-mile-long runways; which would you prefer?).

The other problem is that autopilots just love to disconnect (I wrote about this in my first conjecture on Air France 447). So if one were to pull breakers at random one would be much more likely to cause autopilot disconnection (and a crash much sooner than 6 hours later unless the plane was being hand-flown) than to cause transponder and ACARS disconnection.

Finally you have to remember that, unlike in a crummy four-seat plane, all of that fancy stuff in front of the pilots in an airliner is not the real stuff that runs the airplane. It is mostly switches, knobs, and displays that connect through wires to the actual stuff, which is typically in “electronics bays” underneath the passenger seats (photos; don’t spill your Diet Coke if you want to get to Denver!). So even if a fire burned up the cockpit the transponders would continue to operate because the thing on the dashboard that says “transponder” is in fact just a control panel for a transponder located elsewhere. (I answered the question of Why is it possible to turn off the transponder? in a comment on an earlier posting.)

So I’m still as confused as anyone about what happened to this beautiful B777 and the passengers but based on the other information that we’ve received (many hours of pings, turn to an IFR waypoint,  etc.) I am pretty sure that there was not an electrical fire on board that yet left the autopilot and associated systems untouched.


  1. Lon Sobel

    March 19, 2014 @ 11:24 am


    I didn’t see any mention of this elsewhere:

    This does not match up very well with another bit of information we have about Flight 370, i.e., that the airplane was diverted (via the FMS (like the GPS in your car)) to an IFR intersection

    I agree with you that pilots would put an airport ID into the FMS, rather than an intersection. But how do we know the FMS was programmed to fly to an intersection?

    I ask because I am sincerely curious — not to criticize your analysis.

  2. Tiago

    March 19, 2014 @ 1:39 pm


    If it was a fire, isn’t it too much of a coincidence that it just happened in the border between Malaysian and Vietnamese airspace? So far most of the evidence points towards a deliberate move from the pilot(s). Most puzzling is the ascent to 43,000 ft and then the drop of 20,000 ft in one minute.

  3. philg

    March 19, 2014 @ 6:12 pm


    Lon: http://www.nytimes.com/2014/03/18/world/asia/malaysia-airlines-flight.html talks about the FMS being reprogrammed but doesn’t give the actual waypoint, just says that it is a 5-letter code.

    http://www.smh.com.au/world/malaysia-airlines-flight-mh370-may-have-been-deliberately-diverted-to-andaman-islands-20140314-34sp1.html mentions some of the waypoints by name. None of them are airports.

  4. anon

    March 20, 2014 @ 2:47 pm


    forgive me if I misunderstood something, but doesn’t the original post by Chris Goodfellow imply that the military radar data is misinterpreted? If I understand this right, the notion is that plane likely crashed before reaching Pulau Langkawi; the flight path, as if proceeding through waypoints, is deduced from the military radar data. No one is saying just how robust is the interpretation of that data – and it was revealed at the time when malaysian govt was scrambling to come up with anything. Likewise with Inmarsat ping data – no one has revealed just how reliably that ping data has been interpreted.

  5. Gary Drescher

    March 22, 2014 @ 11:35 am


    The article’s “five-letter” remark is by way of background information, explaining the “normal procedure” for entering a waypoint. The article does say that the flight’s FMS had a “waypoint” programmed, but I wouldn’t make much of that passing remark. It would not be surprising if the authors–or someone who conveyed the information to them–was unaware that not all FMS-programmed locations are waypoints.

  6. Lon Sobel

    March 22, 2014 @ 11:51 am



    Thanks for the links to articles about the waypoints. I hadn’t read about this aspect of the investigation before. The waypoints add to the mystery, because I agree that in an emergency, pilots would input airport IDs, not waypoints. So the waypoints suggest there was no emergency but someone did want to go to an alternate destination, and that someone had to know enough about the plane’s FMS to be able to make the change. If lives were not at stake, this would be an absolutely wonderful puzzle.


Log in