Air France Flight 447: Who’s Fault Was It?
Clive Irving looks at who’s really to blame for flight 447’s disappearance in the South Atlantic seas.
The pilots did it. Put bluntly, that seems to be what the latest report by French air crash investigators on the loss of Air France flight 447 two years ago is saying. More precisely, the pilots had not been trained to deal with the sudden emergency they faced and lost control.
What the investigating body, France’s Bureau of Investigation and Analysis (BEA), are not saying, at least not overtly, is that the crash should never have happened–and, but for a technical failure, would not have.
Air crash investigation reports always deliver a torrent of technical jargon that has been through many drafts. The goal is to be sensitive toward those who can end up getting blamed, but then the language becomes impersonal. So let’s not forget the victims. And what they went through.
This is the stark reality of the end of Flight 447:
Eleven minutes after 2 a.m. on June 1, 2009, the 216 passengers bound for Paris from Rio de Janeiro were plunging toward the Atlantic Ocean at a rate of 10,000 feet a minute. But for the restraint of seat belts, they would have been tossed around the cabin like loose luggage.
They would have been fully aware of their plight. The airplane, an Airbus A330, had lost its ability to fly. During the fall its nose reared up far more steeply than in a takeoff, and it stayed there. It was in an aerodynamic stall. As it fell, the Airbus turned right, away from its heading to Europe and, three minutes after the descent began, it was heading back toward Brazil. With its nose still pointed sharply up, its wings dipped slightly left and at a forward speed of just over 120mph, the airplane hit the Atlantic.
The most violent force was the rate of vertical descent—so violent that the cargo hold below the cabin crumpled and galleys in the cabin were crushed from the bottom up upon impact. The vertical stabilizer sheered off and floated away into the night, becoming a lone beacon for those sent to search for the wreckage.
Fifty bodies were thrown into the water from gaps in the cabin—including that of the captain, 58-year-old Marc Dubois, and four flight attendants. The remaining passengers, five of the nine cabin crew and two of the three pilots went with the wreck deep to the ocean floor in one of the most remote and inaccessible parts of the south Atlantic.
There had never been a crash like this: an airplane representing state of the art flight deck automation, an airline with a long and esteemed record, flying a route that should have been routine disappeared without explanation and without trace. Then, spectacularly, this spring the airplane’s black boxes and cockpit voice recorder were recovered on a sandy plateau 13,000 feet beneath the ocean’s surface, after two previously aborted searches.
The details in today’s report confirm that this crash should not be considered on its own. It is another in a series of disasters involving what is technically called “loss of control” – a problem which has become the number one cause of air crashes. The trend is so alarming that it was the subject of heated debate at a Flight Safety Foundation conference in Turkey earlier this year.
At the heart of the issue is how pilots respond when faced with an imminent aerodynamic stall, which comes down to just one movement on the flight controls: the need to push down the nose of the airplane, not pull it up.
At the conference, Michael Coker, Boeing’s senior safety pilot, cited a series of crashes, beginning with that of the Colgan Air crash that killed 50 people at Buffalo in 2009 and including others in Venezuela, Amsterdam, and France. In each case the airplane had reached the brink of a stall, where the wings lose the ability to provide lift, and the pilots, rather than putting the nose down to regain speed—a basic tenet of Piloting 101 since the beginning of flight—had instead pulled up the nose and produced a fatal outcome.
The case of Air France Flight 447 now joins that list as the deadliest of all.
The French investigation shows that after the pilot in control had a second automatic warning that the Airbus was about to stall he pulled the nose up and even increased the angle. In the course of one minute and fifteen seconds the airplane soared up another 500 feet, from an altitude of 37,500 feet until it stalled so completely that it remained in that attitude until it hit the ocean.
Captain John Cox, CEO of Safety Operating Systems in Washington, an expert with experience of flying a similar Airbus, the A340, as well as an A330 in a simulator, told me that the climb that led to the stall “is one of the most perplexing questions in the accident.”
Before the black boxes were recovered, suspicion had focused on instruments called pitot tubes (the A330 has three of them) that measure the air speed. The pitot tubes on A330s had a record of freezing up and giving false readings to the automated flight management system, as the airplane was cruising and being flown by the autopilot.
This is indeed what happened to Flight 447, creating an emergency on the flight deck. Eight minutes earlier, Captain Dubois had left to take his rest period. The most junior pilot, 32-year-old Pierre-Cedric Bonin, with just 807 hours experience on A330s, was left in control, watched by a second pilot, 37-year-old David Robert, who with 4,479 hours was the most experienced. (Captain Dubois had 1,747 hours.)
With the autopilot disabled by the false air speed information, Bonin took over manually and remained in control. Today the BEA revealed that these pilots had received no training for this kind of emergency at cruise altitude when they had to fly the airplane manually. For a short time Bonin did put the nose down and the situation was recoverable. But when the second stall warning sounded, he reverted to the nose-up command.
The captain arrived too late to take over. Why had the second, more experienced pilot, not taken over?
John Cox explained to me that when a pilot says, as Bonin did, "I have thecontrols" that gives him authority over the co-pilot. "A transfer of controlis essential after a disconnection from the autopilot."
However, even though pilot proficiency is a huge question in this case, it is only fair to them to give a more complete view of what it must have been like on that flight deck.
The failure of the pitot tubes was no minor glitch. There had been 32 previous cases of flight deck emergencies caused by false speed readings. The European Aviation Safety Agency – Europe’s FAA – said in an internal report, disclosed in the magazine Aviation Week, that these failures represented “a large reduction in safety margins and a high work load” for the crews. In the thin air at cruise altitude, 36,000 feet, there is a narrow band of speed in which the airplane is stable, giving a crew very little time to recover if they get anywhere near a stalling speed.
There are at least three culpable parties here in addition to the pilots: the manufacturers of the flawed speed sensors, the French company Thales; Air France for negligent pilot training; and Airbus for not insisting that the faulty speed sensors be replaced when they were known to put the Airbus A330 in jeopardy.
Today’s BEA report was clearly designed to drop into the biggest news black hole of the year. It’s the afternoon of the last Friday in July. In Paris. The moment that millions depart for la grande vacance.
There is one grisly but intriguing detail yet to be explained. Why did Captain Dubois’s body float clear of the airplane, to be recovered a few days later, while the bodies of Robert and Bonin went down to the depths? Did Dubois not have enough time to strap himself into a seat? Was he fighting to remain on the flight deck during the hellish descent? These are questions that may never be answered.