MH 370: Fire as a Black Swan
Curiosity may have killed the cat, but at the very least, it can create some vivid nightmares and I had one last night. Like everyone else, I’ve been following the sputtering investigation of Malaysian Flight MH 370, especially the fire-in-flight theories. So I spent the day reading fire-in-flight accident reports and woke up in a cold sweat last night, imagining myself choking on some smoky, non-descript flight deck. Awful.
But not nearly so awful as the real thing as described in five accidents I researched. Watching the cable channels and following news reports, it seems like the MH 370 theories are going through a bizarre cycling loop as factual leads become ever harder to come by. Now, the talking heads have latched onto the presence of lithium-ion batteries in the cargo hold as a likely fire source and two plus two naturally equals five. If this turns out to be true—and we may never know—MH 370 will be a bit of a black swan, a unique case in which fire disabled the crew and partially disabled the airplane, which then continued to fly for seven hours.
In the real world of transport aircraft fires, it hasn’t worked that way. Just the reverse. In the five in-flight fires I looked at, none flew longer than 30 minutes after smoke or fire was detected and two flew barely 20 minutes and two were fully involved in five minutes or less.
Remember ValuJet 592 (PDF)? It crashed in the Florida Everglades in May 1996 after a shipment of improperly packed oxygen generators ignited. Time from report of first fire detection to the crash: Three minutes and 32 seconds. Swissair 111, an MD-11, crashed off the coast of Nova Scotia in September 1998 after overhead wiring in the entertainment system ignited. Time from report of first fire detection to the crash: just under 21 minutes. In February 2006, a UPS DC-8 freighter (PDF) was fortuitously on approach to Philadelphia International when the engineer confirmed a cabin smoke indication. Five minutes later, the flight landed and the crew evacuated in dense, billowing smoke. The aircraft burned to destruction on the ground. The ignition source wasn’t determined, but lithium-ion batteries, which the airplane was carrying, were mentioned.
More recently, in September 2010, UPS 6 (PDF), a 747 freighter, crashed in Dubai following a cargo hold fire believed to have probably been caused by a shipment of lithium ion batteries. Time from report of first fire detection to the crash: 27 minutes.
Then there was FedEx 1466 (PDF) in September 1996—September seems to be a bad month for fires. It flew for 18 minutes from the time of first smoke detection until the end. But it didn’t crash. It made an emergency landing at Newburgh, New York and was destroyed by fire on the ground. The crew and two passengers escaped with minor injuries. In some ways, it’s the most interesting of the four because it vividly illustrates how quickly fires propagate and how survival can turn on timely decision making and discipline.
The airplane was a DC-10 enroute from Memphis to Boston when, at FL330, the flight engineer reported a cabin smoke indication. Although the NTSB would later ding the Captain for poor crew resource management, it didn’t give him much credit for rapid decision making that probably saved the crew and passengers’ lives. As more smoke alarms appeared, just 3:30 into the incident he acted: “We’ve definitely got smoke guys. We need to get down, right now, let’s go.”
On approach to Newburgh, he coached the First Officer to ignore the 10,000-foot 250-knot speed restriction and get the airplane on the deck. He did, just in time. As the crew was exiting, they were propped in windows venting dense smoke like chimneys. Another minute or two might have made the situation unsurvivable. As it was, the NSTB criticized the Captain for interrupting or failing to supervise the Flight Engineer who neglected to close a cabin vent and to depressurize the airplane after landing, momentarily delaying egress.
As with other hull-loss fires, the source of ignition in FedEx 1466 was never determined, but lithium-ion wasn’t suspect as it was in the two UPS incidents. And, adhering to guilt by association, in MH 370, too, at least by some willing to weave a theory without confirming data. Malaysia Airlines confirmed that the flight was carrying more than 400 pounds of lithium-ion batteries. To be fair, it’s not unreasonable to think these could have been a fire source. There are various ways to explain how this could have selectively disabled certain systems—the comms and ACARs—while leaving the fly-by-wire control system functional. And the 777 does have sophisticated fire detection and suppression systems in its cargo compartments. While this turn of events might not be a high probability, it’s not zero probability either.
But I keep coming back to those other accidents in which fire rapidly propagated and destroyed aircraft, even as two of which were being attacked by airport fire crews. If speculation is a mold, lack of information is its Petri dish, so I remain skeptical of all the theories thus far. But with regard to the fire postulation, I’ll concede the logical fallacy of believing something can’t happen because it never has—the black swan.
One thing is certain, however. As the accidents above describe, a fire in any aircraft rarely leaves other than one option: get it on the deck right now. It’s not arbitrary that the very first line in the 777 Smoke, Fire or Fumes checklist, even ahead of the mask and goggles, is to consider a diversion. It’s clearly meant to be a negative option. In other words, divert now, unless you have a really good reason not to.