Skip to Main Navigation

The Buffalo News

Web Search
by YAHOO! SEARCH

How did complacency turn into catastrophe?

Published:February 22, 2010, 2:11 PM

Font Size:
  • E-mail
  • Share
  • Print

Updated: August 20, 2010, 11:02 PM

WASHINGTON — Federal investigators wrapped up three days of federal hearings Thursday into the crash of Continental Connection Flight 3407, without answering the question of how the final 30 seconds of the flight went from complacency to catastrophe, as one National Transportation Safety Board member put it.

How did a seemingly normal flight suddenly lose so much speed — from 182 knots to 130 knots — that it caused the air-craft’s stall-warning alarm to sound? And why did the pilots both take actions that experts say only worsened the problem and caused the plane to crash Feb. 12 in Clarence Center, killing 50 people?

Mark V. Rosenker, the acting NTSB chairman, pledged to the families and other loved ones who attended the hearings, or watched on the Internet or in simulcasts in Buffalo and Newark, N. J., where the flight originated, that his agency will find the answers.

“We are going to find out what happened here,” Rosenker said at the end of the hearings, “and I can assure you this will be a comprehensive report, with determination of probable cause, and recommendations will follow that, if implemented, will prevent this from happening again.”

Rosenker’s agency has investigated past fatal accidents, found the cause and made recommendations, only to see the Federal Aviation Administration delay or study them further.

He and fellow panel members grilled the final three witnesses, all FAA executives, about those past recommendations that safety board members say have been ignored.

Rosenker is confident that it won’t happen this time.

He said a Senate hearing into the cause of the crash will take place June 10, the new nominee to head the FAA will face confirmation hearings, and the FAA’s charter is coming up in Congress for renewal. New York’s congressional delegation has pledged to tie all of that into the Flight 3407 investigation.

“I’m optimistic,” Rosenker said. “I think we will get a series of changes because of this public hearing, and the attention we have received in the last three days.”

It was unwanted attention for Colgan Air, the regional carrier that operated Flight 3407; the Airline Pilots Association, the pilots union that just won a contract to represent Colgan pilots six weeks before the crash; and the regional airline industry.

Much of America learned for the first time that regional carriers now make up half of the nation’s airline flights.

Passengers aboard Flight 3407 bought their tickets from Continental Airlines, without knowing that Continental subcontracts the flights to Colgan Air, which is owned by Pinnacle Airlines.

The hearings brought out the low pay scales for regional airlines. Flight 3407’s first officer, Rebecca L. Shaw, Colgan executives testified at the hearings, was paid only $16,000 a year, and at one time had to take a second job at a coffee shop. Colgan later said its average pay for a first officer is $24,000.

How could Shaw, 24, who commuted from her home in Washington State, afford to rent an apartment in Newark on that salary, board member Kathyrn O’Leary Higgins asked.

She obviously couldn’t. To start off her shift, Shaw commuted from Seattle to Memphis, Tenn., on a FedEx red-eye flight, slept for four hours in a pilots lounge, and then caught a flight to Newark on the day of her flight to Buffalo, where it is believed she slept in the crew room.

Fatigue management, a key problem in an industry with 16-hour duty days, took up a large part of the hearings.

Flight 3407’s pilot, Capt. Marvin Renslow, 47, who came to flying as a second career, had to take a job stocking shelves in a Tampa, Fla., supermarket before he got hired at Colgan.

Renslow commuted from Tampa, and there is no record that he had an apartment or that he rented hotel rooms in Newark. It’s unknown where he slept on the night before the flight to Buffalo, but he had logged onto the computer in the Newark crew room at 3 a. m.

Colgan has stressed throughout the hearings that it has an excellent safety record and that its pilots are trained as well as any of the major air carriers.

“We want to know what happened as much as anyone else does and remain committed to do everything we can to prevent future accidents,” Colgan spokesman Joe F. Williams said at the hearing’s close.

“We are totally committed to safety and are determined to work closely with our industry colleagues and regulators to reinforce this primary industry objective.”

It was another rough day of testimony for Colgan and its Flight 3407 pilots Thursday as board member Deborah A. P. Hersman, who seemed to ask the most telling questions during the hearings, questioned a NASA scientist who studies pilot distractions.

“I think the crew went from complacency to catastrophe in 20 seconds,” Hersman said. “It seems like they were discombobulated.”

She said co-pilot Shaw put the airplane’s flaps up, which experts said was the wrong thing to do.

Renslow yoked the control back, causing the nose of the Bombardier Dash 8 Q400 twin-engine turboprop plane to pitch up, instead of putting the nose down to recover from the stall, she said.

“The room’s on fire at this point,” she said to Dr. Robert Key Dismukes of NASA’s Ames Research Center. Wouldn’t it make sense, she said, for aircraft manufacturers to install warnings that the aircraft is losing that much speed that quickly before the “stick shaker” alarm goes off?

“It would be sort of like a fire alarm,” she said. “The stick shaker is more like being in a room on fire.”

Dismukes, who spoke at length about what causes pilots to lose attentiveness, agreed with Hersman that for all the other warnings and chimes in a cockpit, this seems like one that makes sense.

Dismukes told the panel that he has spent 20 years studying such issues as pilot attention and holds pilot licenses from gliders to the 737, so he could appreciate their problems more and experience them firsthand.

The NASA exert also was surprised by the actions of Renslow and Shaw when the stick shaker went off, signaling that the plane was on the verge of a stall because insufficient air was flowing over its wings.

Dismukes said the only way for pilots to react to such emergencies was continual comprehensive training so that when it happens, they can automatically know what to do. Otherwise, he said, it’s like having to consult a book, not doing what needs to be done immediately.

He agreed with questioners who wondered whether stall training for pilots should be done the way it is now. Currently, pilots know they are going to encounter a stall in the simulator, and the element of surprise is gone.

Dismukes also agreed that having pilots encounter stalls when flying on autopilot would much better mimic the sudden terror that Renslow and Shaw experienced when the alarm went off as they approached Buffalo Niagara International Airport.

“Surprise and stress are problematic,” Dismukes said. “Even experienced pilots flail around.”

Dismukes said the pilot and co-pilot of Flight 3407 could have fallen victim to “automation complacency” — that is, putting so much faith in the autopilot that they were not paying enough attention.

“You’re a step removed from physical control of the aircraft,” Dismukes said.

Noting that the period just before landing is “one of the busiest periods in general” for flight crews, Dismukes noted that a lot was happening all at once as Flight 3407 descended toward Buffalo. The control tower was contacting the cockpit just as Renslow and Shaw were preparing for landing, adding one more distraction.

Meanwhile, the plane lost 50 knots of speed in 20 seconds — and the crew did not notice that the plane had decelerated to a dangerous level.

The plane’s stall-warning system activated when it fell to a speed of 130 knots. Sources with knowledge of the Q400 said that when flying a relatively full load in icy conditions, pilots generally would not want to slow the aircraft to much below 140 knots.

“I don’t see any evidence that he [Renslow] understood the situation he was in,” Dismukes said.

Dismukes noted that responding to a stall is much different in real life than it is for pilots training in a simulator.

Pilots have about 3 seconds to respond correctly to the stick shaker, which is far easier for pilots who are expecting a mock stall in training.

“If you’re expecting a stall, 3 seconds is plenty,” Dismukes said. “If you’re not expecting it, the response would be highly variable. Some [pilots] would respond correctly in 3 seconds. A lot would not.”

But safety board member Higgins noted that an earlier stall warning — an illuminated red barber pole — came on but went unnoticed.

“There was a warning,” she said. “It just was not recognized.”

mbeebe@buffnews.com and jzremski@buffnews.com

Comments

There are no comments on this story.