Posted byHere’s the press release from the NTSB:
****************************************************
NTSB PRESS RELEASE
*********************************************
National Transportation Safety Board
Washington, DC 20594
FOR IMMEDIATE RELEASE: February 2, 2010
SB-10-02
*********************************************
CAPTAIN’S INAPPROPRIATE ACTIONS LED TO
CRASH OF FLIGHT 3407 IN CLARENCE CENTER,
NEW YORK, NTSB SAYS
********************************************
The National Transportation Safety Board determined that the captain of Colgan Air flight 3407 inappropriately responded to the activation of the stick shaker, which led to an aerodynamic stall from which the airplane did not recover.
In a report adopted today in a public Board meeting in
Washington, additional flight crew failures were noted as causal to the accident.
On February 12, 2009, a Colgan Air, Inc., Bombardier DHC-8-400, N200WQ, operating as Continental Connection flight 3407, was on an instrument approach to Buffalo-Niagara International Airport, Buffalo, New York, when it crashed into a residence in Clarence Center, New York, about 5 nautical miles northeast of the airport.
The 2 pilots, 2 flight attendants, and 45 passengers aboard the airplane were killed, one person on the ground was killed, and the airplane was destroyed by impact forces and a post-crash fire. The flight was a 14 Code of Federal Regulations (CFR)Part 121 scheduled passenger flight from Newark, New Jersey. Night visual meteorological conditions prevailed at the time of the accident.
The report states that, when the stick shaker activated to warn the flight crew of an impending aerodynamic stall, the captain should have responded correctly to the situation by pushing forward on the control column.
However, the captain inappropriately pulled aft on the control column and placed the airplane into an accelerated aerodynamic stall.
Contributing to the cause of the accident were the
Crew members’ failure to recognize the position of the low-speed cue on their flight displays, which indicated that the stick shaker was about to activate, and their failure to adhere to sterile cockpit procedures.
Other contributing factors were the captain’s failure to effectively manage the flight and Colgan Air’s inadequate procedures for airspeed selection and management during approaches in icing conditions.
As a result of this accident investigation, the Safety Board issued recommendations to the Federal Aviation Administration (FAA) regarding strategies to prevent flight crew monitoring failures, pilot professionalism, fatigue, remedial training, pilot records, stall training, and airspeed selection procedures. Additional recommendations address FAA’s oversight and use of safety alerts for operators to transmit safety-critical information, flight operational quality assurance (FOQA) programs, use of personal portable electronic devices on the flight deck, and weather information provided to pilots.
At today’s meeting, the Board announced that two issues that had been encountered in the Colgan Air investigation would be studied at greater length in proceedings later this year.
The Board will hold a public forum this Spring exploring pilot and air traffic control high standards.
This accident was one in a series of incidents investigated by the Board in recent years – including a mid-air collision over the Hudson River that raised questions of air traffic control vigilance, and the Northwest Airlines incident last year where the airliner overflew its destination airport in Minneapolis because the pilots were distracted by non-flying activities – that have involved air transportation professionals deviating from expected levels of performance.
In addition, this Fall the Board will hold a public forum on code sharing, the practice of airlines marketing their services to the public while using other companies to actually perform the transportation. For example, this accident occurred on a Continental Connection flight, although the transportation was provided by Colgan Air.
A summary of the findings of the Board’s report are available on the NTSB’s website at:
http://www.ntsb.gov/Publictn/2010/AAR1001.htm
NTSB Media Contact: Keith Holloway
(202) 314-6100
*************************************************************************
Now – What do you think….
Does that sound like root causes?
I’d like to know why a trained pilot would pull back on a stick when the stick shaker activates.
And I don’t think the fact that they were talking in thew cockpit has anay thing to do with it.
Ideas????
Category: Accidents, Current Events, Investigations, Topic of the Week
14 Comments »
RSS feed for comments on this post.
As far as I can tell there is really no root cause evaluation. Sure, the pilot made the wrong move, but there is no looking into the background. Where was the co-pilot? There are so many questions that are unaswered and there seems to be no systematic examination.
Comment by Bear Slothower — February 23, 2010 @ 10:27 am
I agree that this investigation didn’t go far enough. They stopped at a symptom – the failure of the pilot to respond correctly, and never found out WHY the pilot responded incorrectly. Bad training? Mindset? Other diverse indications available? They can’t interview the pilot but they could interview other pilots, review training, etc.
Comment by Bill Mihalovits — February 23, 2010 @ 11:14 am
When one reads the entire report, it is made clear that icing did not occur, so why is it emphasized up front? Looks like the investigators had a few assumptions going in that influenced their evaluation, regardless of the factual evidence!
The report states that the correct response to an incipient stall is to push forward on the stick; fine if you are at a safe altitude, but problematic when you have an aircraft load of passengers near the ground. Lift can be rapidly generated by (novel thought) advancing throttles to increase thrust. This may often be a safer response.
Comment by Alan Weston — February 23, 2010 @ 11:15 am
The casual factor I believe is human error. However it’s a training issue, the Pilot and the crew didn’t perform based on their training, or lack of training. And to create a corrective action that captures training similiar to the event raises a flag that training was a precurser that lead up to this event.
Comment by Derwin — February 23, 2010 @ 11:27 am
According to the pilot I was talking to the pilot that day had enough altitude to lower the nose and advance the throttles to avoid the stall.
Comment by Mark Paradies — February 23, 2010 @ 11:54 am
First… It’s a waste of time for anyonoe who’s not a pilot to reach any conclusion.
Second.. Its a waste of time for anyone who’s never done any training flying in IMC (Insurment Metrological Conditions) to reach any conclusion.
Unless you’re a trained pilot, unless you understand the dynamics that occur inside the cockpit, you’re never going to be able to reach any meaningful conclusion.
I’m an insturment rated pilot trained to fly in the clouds. I know what it’s like. I understand crew resource management, cockpit management, weather, aerodynamics, and emergency proceedures.
I’d even argue that most folks don’t understand an aerodynamic stall. It’s when the plane gets so slow that the wings (or in some cases the horizontal stablizer) doesn’t have enough airflow to do their job. The end result is an abrupt nose down movement of the airplane as it tumbles to the ground.
============
Rule #1) There can only be 1 person flying the airplane.
Pilots have specific proceedures they use to transition control from one pilot to the other. When was the last time you tried driving down the interstate with you and your wife both trying to grab ahold of the steering wheel?
The co-pilot doesn’t matter in this equation.
Had the sequence of events been strung out over a longer period of time, the co-pilot might have caught the pilots mistakes. But if you’ve ever been in a stalled airplane, it happens NOW! And when you’re on an insturment approach, at night, in the clouds, at a couple of thousand feet – you’re done. You can’t recover. The pilot has to recognize the onset of the stall and take the appropriate actions.
The proper action is push forward on the stick (to pickup airspeed) and apply power to gain even more airspeed.
The cause of the crash is pretty simple. He didn’t respond correctly (and quick enough) to the warning signs of an approaching stall. His inappropriate response casued the crash.
The real question is to try and figure out why did he respond the way he did?
The answer to that question falls into the psychlogical catagory and we’ll never know what he was thinking at the time. So you’ll never get that answer.
What’s possible is that he got the impending stall warning indications (stick shaker, alarms, and wing buffetting) but didn’t feel like he could push the nose over and point it to the ground at such a low altitiude. That’s a tough position to be in because it goes compeltely against your instincts.
Most likely, that’s the case. But we’ll never know.
A similar example that has claimed many lives over the years is a failed engine shortly after take-off. Imagine leaving the runway, climbing and passing through 100 ft, and then your engine shuts off….. What’s your first instict???
Everything in your body says to pull back on the stick and extend the glide until you can find a place to land. But if you do, you stall the airplane and crash. The correct move is to push forward to lower the nose and protect your airspeed. Then do your best to land the airplane straight ahead – even if it’s in the trees, power lines, or lake.
Being low, slow, and stalling the airplane is a recipe for disaster. There’s just no time to recover. You’ve got to make the right decision in an instant; and it may go against all of your life-saving instincts. And in some cases (like when there isn’t much air under your butt) even the correct response will still drive the nose the airplane into the ground.
Comment by Phil Perry — February 23, 2010 @ 12:20 pm
The Pilot may have reacted incorrectly but why was the plane at the point of stalling in the first place?
Comment by Craig Stannard — February 23, 2010 @ 4:54 pm
One clue to a possible “cause” is revealed in Phil Perry’s response above. The pilot is in charge and “Unless you’re a trained pilot, unless you understand the dynamics that occur inside the cockpit, you’re never going to be able to reach any meaningful conclusion.” The culture reflected in this comment must be addressed. I’m not trying to start a war, just interested in overcoming barriers that can mask root causes and in turn contribute to further fatalities.
Root Cause: Work direction/supervision during work/crew teamwork ni
This is a serious allegation to make towards investigators. I facilitated an icing related stick-shaker incident with a Boeing 737 and the level of defensiveness was disturbing despite the presence of pilot’s on the team. The investigation was termed a “witch hunt”. Ultimately, we found causal factors and root causes that have resulted in improved operating conditions for these pilots.
Looking at the presentations, there are several findings that go unaddressed based on the corrective actions.
1. Action by the co-pilot setting the landing speed 13 k slower than stick-shaker. Doesn’t this point to a causal factor: The plane slowed to stick-shaker speed? So what resulted from the plane slowing below stick shaker speed? The pilot made the decision to pull back. The choice he made resulted in a crash and we’ll never know about the result had he chosen to push. Having defined this as a causal factor, what’s the root cause?
2. Another causal factor appears to be: Crew did not respond to the low-speed cue. So what resulted from the crew not responding to low-speed cue? The plane slowed to stick-shaker speed (impending stall speed)
3. The pilot’s abilities were questionable (see Capt.Roger Cox’s “Pilot training records/remedial training program” presentation) Causal Factor: “Colgan did not proactively address these issues.” So what resulted from Colgan not addressing the pilot’s performance issues? He was not prepared to adequately respond to the conditions. Possible root causes:
-Work direction/selection of worker/not qualified.
- Training/no training/decided not to train (“Colgan did not have program”
- Training/understanding ni/continuing training ni
I think a generic root cause was identified:
-”Few carriers had such programs before accident”
We all know this industry has a good safety record as compared to other modes of transportation but we must remain committed to zero.
Comment by Mike Rodriguez — February 23, 2010 @ 5:35 pm
Wow Mike – The NTSB needs you!
Mark
Comment by Mark Paradies — February 23, 2010 @ 6:16 pm
Another point to consider – The “embarrassment” factor. This is this pilot’s world, his own island, he is in control. Could the fact that he missed the speed cue and dropped to stall speed hit him like a ton of bricks and he failed to follow his training and even acknowledge that he had a problem? As a safety professional with air operations background, I knew many pilots who would tough out a bad approach instead of adding power, calling it out and going around again. The embarrassment of being singled out for poor performance would outweigh the possible crash or aircraft damage that could occur. Your right, we can’t speak to the pilots but we can find all of the factors that caused or allowed the event to take place.
Comment by Bob Whitfield — February 24, 2010 @ 8:06 am
Two things jump out from the list of findings:
The Q400 airspeed indicator LACKED LOW-SPEED AWARENESS FEATURES, such as an amber band above the low-speed cue or airspeed indications that changed to amber as speed decrease toward the low-speed cue, that would have facilitated the flight crew’s detection of the developing low-speed situation.
AN AURAL WARNING IN ADVANCE OF THE STICK SHAKER would have provided a redundant cue of the visual indication of the rising low-speed cue and might have elicited a timely response from the pilots before the onset of the stick shaker.
======
Having sufficient information in advance and the time to digest what it means is a major component of engineering hazard reduction. It is unclear at which point prior to the actual stall the stick should shake, but there needs to be time to digest the information. If you cannot eliminate the hazard, you must either cfontrol ir or warn the operator of the hazard. Administrative controls and protective equipment (PPE is not valid here) are last resorts better reserved for low risk hazards. It is unfortunate that such design features often fall by the wayside when cost becomes an issue.
Yet, the conclusions of the board ignore this entirely and blame the pilot. The Q400 may not be a Space Shuttle, but the risks in terms of lives is substantially greater.
Comment by Ron Montague — February 24, 2010 @ 1:03 pm
Good comments Mike – I was about to address the lack of training or pilot proficiency suggested by the findings but you nailed it – management didn’t address it when it came up.
Comment by Ron Montague — February 24, 2010 @ 1:09 pm
As a Certified Instrument Flight Instructor and somewhat schooled in aerospace and mechanical engineering and having 30 years of experience in nuclear power plant safety I have found out how little I know. However, I agree with the above responders that there was insufficient root cause analysis. That this sad fact occurs often with NTSB and is evidenced by the continuing increase in accidents.
As a Standards/Evaluation Officer for the CAP, I fear that the National Staff doesn’t fully understand what a root cause is either. The CAP management sincerely tries to increase safety. However, it is not common for people to believe that safety can not be improved by treating the proximate causes with solutions that have not worked in the past.
Getting back to the subject accident, understand that air speed is related to stall in stable flight, but angle of attack for a given airfoil and the turbulence in the air in which the aircraft is flying are the real key to stalls. (Turbulence can reduce the Reynolds Number by 20% which affects the stall angle of attack.)
In my limited experience with stick shakers, I recall that they react to airspeed and not to angle of attack. The wing and horizontal stabilizer (an upside down wing) simply don’t have their respective design air foils (hence different stall angle of attack) in even the smallest amount of icing. Granted there was not supposed to be any. And by the way, when the stick shaker starts its dance, your pulse rate doubles. Never mind how it was in the simulator.
If the tail stalls, usually the pilot would apply back pressure while if the wing stalls, the pilot would normally apply forward pressure. In either case one would first add full power. If properly trained for the given aircraft it would be only necessary to first try to recover (especially near terrain) by adding full power and gently ease the nose down and if the wing starts flying again, assume the best rate of climb attitude. This will usually stop the bleeding or come close. If it doesn’t, there may be ice causing tail stall, so some back pressure is advised with the power.
Most aircraft have the capability to indicate ground speed. A retired Air France pilot I know preaches that final approaches should be made at constant ground speed. This provides a little airspeed for grandma, and because when landing into the wind, helps when the airspeed decrease as near the ground. Near the ground can be pretty high when there is wind shear.
I don’t mean this to be a flying lesson, and for the given aircraft and the situation, it may not include the correct actions. My point is that the root cause may have been less than adequate training to avoid what happened. It may have been a design issue with the aircraft. It may have been poor cockpit management. It may have been crew impairment. It may have been the decision to continue the flight under the conditions. It could have been poor weather information. It could have been lack of a plan “B” for possible emergent situations. It may have been less than adequate flight testing. It could have been improper loading. If it had been me, it could have been sheer panic. Obviously we don’t have a root cause.
Comment by W Reuland — February 25, 2010 @ 4:17 am
Wow! These are great comments.
Comment by Mark Paradies — February 25, 2010 @ 10:52 am