IE 11 is not supported. For an optimal experience visit our site on another browser.

Pilots question speed sensors' troubled history

Airbus knew since at least 2002 about problems with the type of speed sensor that malfunctioned on an Air France passenger plane that crashed in June, The Associated Press reports.
Brazil Plane
Brazilian navy sailors recover debris from the missing Air France jet on June 8.AP
/ Source: The Associated Press

Airbus knew since at least 2002 about problems with the type of speed sensor that malfunctioned on an Air France passenger plane that went down in June, The Associated Press has learned. But air safety authorities did not order their replacement until after the crash, which killed all 228 people aboard.

The tubes, about the size of an adult hand and fitted to the underbelly of a plane, are vulnerable to blockage from water and icing. Experts have suggested that Flight 447's sensors, made by French company Thales SA, may have iced over and sent false speed information to the computers as the plane ran into a thunderstorm at about 35,000 feet.

The exact role the sensors — known as Pitots — played in the crash may never be known without the flight recorders, which have not been recovered and which have stopped emitting signals. Investigators insist sensor malfunction was not the cause of the crash, but many pilots think false speed readings may have triggered a chain of events that doomed the plane.

Fernando Alonso, head of Flight Operations at Airbus, maintains the doomed Airbus A330 plane was "totally airworthy."

"There is no question for me the safety, the reliability of the airplane nor of the maintenance and operation procedures used by our operators," he said.

Cockpit confusion
The plane was flying from Rio de Janeiro back to the French capital when it went down in a remote area of the Atlantic, 930 miles off Brazil's mainland and far from radar coverage. Automatic messages transmitted by the plane show its computer systems no longer knew its speed, and the automatic pilot and thrust functions were turned off.

Several European airline pilots, including former Air France captain Gerard Feldzer, believe a reading of the messages suggests Air France pilots were suddenly forced to take manual control in near impossible conditions: a cockpit ringing with warning bells and flashing lights, some of them contradictory, with few clues to speed, altitude and nighttime weather conditions.

"It's very difficult when you are already experiencing turbulence in the middle of the night, to know what to do," said Feldzer, adding that the plane's automated warning system could have been issuing incorrect instructions. "It's very difficult to resist what you are being ordered to do because they are false orders."

Air France is now starting a training program for pilots on how to manage a Pitot malfunction at high altitudes of the type experienced on Flight 447. Previously, Air France had only offered simulator training for Pitot malfunction on take-off and landing.

Pilots are angry about what they see as an attempt to pin the crash on pilot error. Eric Tahon, an Air France pilot, defended the role of the Flight 447 crew.

"We are trained to deal with multiple failures of the plane," he said. "We are convinced that without the breakdown of the Pitots, Air France 447 that day would have set down at (Paris') Roissy (airport)."

Feldzer, however, said that while the dangers now appear to have been underestimated, he believes Airbus and Air France would not have risked their reputations had they thought Pitot faults were critical.

Memo warned or problems
A series of industry documents verified by investigators show that regular warnings on Airbus Pitots popped up as far back as 1994, although for a different model that was later banned in 2001 by French aviation officials.

An Airbus memo from July 2002 warns of blocked drainage holes on the Thales AA Pitot — the type fitted onto the doomed Air France jet — and says "this issue can affect all Airbus aircraft fitted with Thales Pitot probes."

Airbus recommended replacing Thales Pitots with a newer model in 2007 but did not make the change mandatory. And Air France decided to replace Pitots on its longer range Airbus fleet only when they broke down.

Air France says it started having problems with speed-measuring equipment on long-range Air France A330 and A340 jets in May 2008, which Airbus blamed on ice crystals blocking Pitot tubes. But functioning sensors were not replaced at that stage.

Paul-Louis Arslanian, head of BEA, the French agency investigating the crash, said the body knew of around seven Pitot incidents on long range planes of the A330-A340 family before the crash, and that Airbus knew of around 20. Europe's air safety authorities say they had been monitoring Thales Pitots on A330-340 long-range planes since 2008, when it was aware of nine incidents of malfunctions.

'Perfect storm'
Michael Barr, who teaches aviation safety at the University of Southern California, said it appears Airbus recognized the Pitot problem needed to be fixed but did not make it an urgent priority.

"What they hoped for was that the perfect storm wouldn't come up before they got it fixed," Barr said. "They were in the process of doing that when this one hit that perfect storm over the ocean."

The cost of replacing all Pitots on a worldwide fleet of planes is hardly major: For Air France, a full speed-sensor overhaul on its A330-A340 fleet would cost around $222,000, according to AP calculations based on pilot estimates. Both Thales and Goodrich declined to disclose the cost of their sensors.

But if prosecutors rule that the crash could have been avoided, the financial penalties and loss of reputation for Air France and especially Airbus, whose aircraft fill the skies every day, would be devastating.

Ban issued
There is no hard evidence that faulty Pitots caused the Air France crash, and in past reported incidents of Pitot malfunction, pilots have been able to regain control of the plane. Furthermore, what looks to be a likely cause in the beginning of a crash investigation sometimes shifts after investigators get more information from the flight recorders.

However, more than two months after the crash, the European Aviation Safety Agency has reassessed the dangers of faulty Pitots, ordering a continent-wide ban on the types of Thales sensors that were fitted onto Flight 447 on all long range planes.

The directive also extends to a newer Thales model, of which there can no longer be more than one on each plane. EASA said at least two of the three probes fitted to each Airbus plane should be made by North Carolina-based Goodrich. This month the U.S. Federal Aviation Administration issued a similar directive for U.S. airlines.

The crash, and what EASA spokesman Daniel Hoeltgen calls "an avalanche" of reported Pitot problems among European carriers since the accident, has caused the industry to reconsider its view of the sensors.

"Prior to 2008, operators and manufacturers did not deem Pitot issues to be safety-critical," he said.

"Now, with all the reporting and testing that's going on, we're saying that we believe as a precautionary measure, that it would be best to have either 3 Goodrich probes or 2 Goodrich probes and 1 Thales."

Newer models
In its airworthiness directive, EASA notes that faulty speed readings can lead to the disconnection of the autopilot and auto-thrust functions.

"Depending on the prevailing aeroplane altitude and weather environment, this condition could result in increased difficulty for the crew to control the aeroplane," the agency said.

The directive does not apply to the smaller single-aisle and shorter range A320s, because the rule is designed to address the type of Pitot malfunction that occurs only at the altitudes and temperatures flown by the longer-range planes.

The newer Thales model had addressed the different problem of water getting into the tubes but is not as good as the Goodrich probe in icy conditions, said EASA's Hoeltgen.

EASA has also received reports of incidents with Goodrich probes, but of a "lower magnitude" than with Thales Pitots. Hoeltgen said problems occur "no matter which make and series."

Airbus rival Boeing Co. admits to Pitot "incidents" on some of its planes, but says "unreliable airspeed due to Pitot icing or torrential rain is a temporary condition" and procedures exist "to ensure continued safe flight," according to spokeswoman Sandy Angers. She declined to say what make of Pitot Boeing uses on its planes, but said there have been no accidents "attributable to high-altitude icing of heated Pitot probes."

Assessing blame
Experts caution that without more evidence it's impossible to pin blame on Pitots for the crash of Flight 447.

"Pitot tubes have been highlighted in regards to AF 447 but there are a whole bunch of other things that may have gone wrong," said Chris Yates, Janes' aviation security editor.

Still, with only the debris and a series of automatic messages transmitted from Flight 447 — which highlight the Pitot problem — some pilots are asking why top officials at Air France, Airbus and the BEA appear reluctant to talk about the role of the speed sensors.

Lead investigator Alain Bouillard says Pitots are "an element but not the cause." EADS CEO Louis Gallois insists that Flight 447 was brought down by a "convergence of different causes." Air France CEO Pierre-Henri Gourgeon says he's "not convinced that the sensors are the cause of the accident."

Leon Cremieux of the Sud Aerien union at Air France said airlines receive a lot of service bulletins and rely on planemakers and aviation authorities to flag urgent maintenance issues.

"There was a dialogue between Airbus and Air France" over the Pitots, he said. "Clearly it wasn't judged important enough to be a question of security."

Barr, the USC aviation safety professor, said serious efforts to correct safety problems in the airline industry tend to increase in direct proportion to the number of people killed, while too little action is often taken in response to incidents in which no one is killed.

It even has a name: blood priority.

"The more blood that is spilled, the more corrective action is taken. The less blood that is spilled, the less corrective action that's taken," Barr said.