Colgan Air 3407

There has been some good blog coverage of the Colgan 3407 accident lately.  For a perspective on the Catch-22 finances and schedules that the regional airlines (and their pilots) find themselves in, read Aviation Mentor‘s terrific post.  For an in-depth examination of the FDR (flight data recorder) and CVR (cockpit voice recorder) data, see Blogging at FL250.  Probably the best overall description of the events leading up to the fatal crash, far better than anything you will find in bowdlerised media accounts, whether print, web or television.

It’s hard to know what to make of this incident, as the accounts you get in the media inevitably focus on the sensational (CVR, shrieking final moments, etc) as opposed to the instructive (FDR).  What has become apparent out of the data is that the aircraft didn’t suffer a tailplane stall so much as an ordinary (wing) stall, something pilots all over the world are taught to recover from by pushing the nose down and increasing the throttle.  Yet the captain of Colgan 3407 did the opposite, and it’s not entirely clear why.  Media accounts paint him as inexperienced at best, and at worst they hint at several failed checkrides and come nigh to declaring him incompetent.

My own sense is that one doesn’t get past a PPL, instrument rating, night rating, commercial rating, multi-engine rating, ATP, type rating, several FO checkrides and captain checkrides by being an incompetent boob.  It strains credulity.  Which is not to say that pilots can’t occasionally be stupid, but they are not habitually stupid.  So how to explain pulling up on the yoke when standard stall recovery is to push the yoke forward?  A panic reaction?  Or did the captain mistakenly believe they were experiencing a tailplane stall (whose recovery procedure is to pull up), only to belatedly realise that it was a wing stall—but by that time, all manoeuvring altitude had run out.

The “how” is readily apparent from the FDR, but it’s hard to answer the “why” without having the aircrew able to answer for themselves.

UPDATE:  Sam at Blogging at FL250 cranks out another home run with his latest post.  In it, he unlocks the mystery of why the captain might have acted contrary to standard to stall recovery procedures.  The short version is that the Practical Test Standards for the ATP rating do not require realistic simulation of an approach or takeoff stall, where the aircrew is distracted.  This allows airlines to condition their pilots to “ride the shaker”, maintaining attitude and altitude during the stall, waiting for power to wrestle the airplane back up to a safe airspeed.  The practical outcome is that the aircraft remains in a stall longer than necessary, and when you’ve got relatively little altitude to play with—on takeoff or approach—that may have dire consequences.

Like one of his commenters says, if he’s not in the training department at his local airline, he ought to be.

You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
2 Responses
  1. M says:

    No One explains why Flight 3407 stalled in the first place, Did the pilots let the speed decrease incorrectly?

  2. Chris Taylor says:

    The short answer, I suppose, is yes.
    An aircraft’s speed is supposed to decrease as it gets closer to its destination, naturaly. For the Q400 that would be from a high of about 0.58 mach (360 kts) cruise speed, down to 250 kts by 10,000 feet, and even slower (150-130kts) on the approach, where you are down to about 2500 feet.
    When you put down flaps, it changes the shape of the wing to generate more lift at slower speed, but it also slows the aircraft down by producing some additional drag at higher speeds. Putting the landing gear down also incurs a drag penalty and slows the aircraft down further.
    In order to maintain airspeed and keep from stalling, a pilot will typically advance the throttles slightly as more degrees of flap are extended, and especially when the gear is extended, too.
    A Q400 can decelerate at about 3 knots per second with the gear and flaps down, so it doesn’t take very long for it to decelerate from a safe maneuvering speed (say, 160 kts) to something approaching stall speed (in this particular case, under 130 kts). That would be all of ten seconds.
    Because the Q400 doesn’t have an auto-throttle, unlike its larger jet-engined cousins, the pilots have to remember to advance the throttles as they extend the flaps and gear. And it appears that in this case, they either forgot, or didn’t add enough throttle to overcome the the drag imposed by the gear and the flaps.