5 Lessons from the Buffalo Crash

More than any other crash in recent memory, the tragedy of Continental Connection Flight 3407, which went down on approach to Buffalo killing 49, made me think about how I fly my small airplane. How might the traps that snared the relatively inexperienced crew of that Colgan Air flight be waiting for GA drivers like you and me?

If you haven't read this elsewhere, it's come to the attention of the world that crew of the Q400 (Dash 8) that crashed was sleep challenged, relatively inexperienced and poorly paid. (This last part doesn't sound like a safety issue, but when you look at crew rest and travel factors, you can see that it is.) 

The NTSB transcripts of the cockpit voice recorder are absolutely chilling. If you haven't read through them yet, here is a link to the transcript (Buffalo Crash CVR Transcript) . Here, also, is a link to an animation of the last two minutes of the ill-fated flight. (Buffalo Crash Animation)  Be forewarned: its' not light reading or viewing. And while the NTSB hasn't issued a final report and its finding of probable cause, the transcript raised numerous questions that the final report will surely address.

How big a causal role did these five factors play in the crash of 3407? I have my ideas about it. But when it comes to examining our own flying, there's no question that there are risk factors exposed by the crash and the subsequent, ongoing, investigation that should give us pause and prompt us to take a minute to reflect on our flying.

5 Lessons

5. Training Did the crew of 3407 have sufficient training in their airplane? A Colgan Air official probably answered that indirectly when he was asked whether the crew should have been able to recover from the stick shaker incident: "Yes." was his one word response. But they didn't, and it has been reported widely that the pilot never had stick shaker training in an actual airplane or in the sim. What blind spots do you have for your airplane? Can you do the emergency gear extension procedure in your sleep? Do you know where the extinguisher is and can you get to and activate it without having to puzzle through the process? Do you know how far your airplane glides and how to get the best distance out of a glide in an engine out scenario? Can you fly on the backup gages in the soup? Bottom line: if you're like most pilots, you could probably know your airplane better and be more proficient in it.

4. Fatigue    I don't think I've heard anyone argue that the amount or quality of sleep that the Colgan Air pilots had was ideal, never mind adequate. I'd argue that it was insufficient. People who are sleep deprived can do all kinds of things--witness me making coffee in the morning. And flying an airplane is one of those things they can do when they're sleepy. I have a close friend who used to fly corporate in the 70s when rest rules were non-existent or unenforced, and the stories he's told me about schedules--post midnight arrivals with one set of bosses followed by pre-dawn departures with another set--make my hair stand on end. How would he deal with it? He and his co-pilot would swap naps. On more than one occasion they'd both be awakened by a radio call. And I know that I've flown when I've been too tired to be at my best, and I'm not proud of that fact. I did fine, but how would I have handled a complicated emergency? Probably not was well as I could have, and who knows, if that would have been good enough?

3. Preparation    Read the transcripts to see just how nonchalant the crew of 3407 appears. They chatted about their commutes, their experience levels, their career plans and their schedules. When it came time to do the approach briefing, they rushed through it as though it were an afterthought. Would something as simple as better advanced planning have prevented the accident? You could make a good argument that it would have.Dick Collins once told me that if on a long flight you think there's nothing to do in the cockpit, nothing to monitor, no information to gather, nothing to prepare for, you're just not paying close enough attention. Good advice.

2. Sterile Cockpit When I was based on the East Coast, I used to fly with friend and collegue Tom Benenson a lot. And I loved our chats while flying. But one day on departure from White Plains (KHPN), a couple hundred feet off the ground, we found ourselves talking about the imporance of a sterile cockpit. The irony was immediately apparent to us, and a good lesson. The crew of 3407 did not maintain a sterile cockpit below 10,000 feet, as the regs require. Would it have made a difference if they had? It's sure a possiblity.  The bottom line is, when you're close to the ground, you need to be paying attention.The sterile cockpit policy helps ensure that always happens.

1. Auto Pilot Stalls What was the cause of the crash of Flight 3407? It sure look as though it was the onset of an autopilot stall and a failed recovery from it. Many pilots, I'm certain, don't even know what an autopilot stall is, even those pilots who regularly fly with and use an autopilot. I had to figure it out myself. The first airplane with an autopilot I ever flew was a great little Grumman Tiger. The flying club owner checked me out in it himself. He showed me how to avoid PIO landing accidents in it, he put me through a series of accelerated stalls, and he put me under the hood, too. But he never said word one about how to use the autopilot, never mind about autopilot stalls.

I learned it myself one day when I was climbing out of Bridgeport, CT, (KBDR). I set the vertical speed at a good rate--I forget what it was--one that the airplane could hold down low. Problem was, I was going up to a higher altitude, and as I climbed, I got distracted, with a revised clearance and a new heading. By the time I got that copied down and deciphered, I got a funny feeling. I glanced up at the airspeed indicator, and I was slow, right around best angle of climb airspeed. Had I kept writing for another minute or so, I would have stalled. I saw what was happening, disconnected the autopilot, lowered the nose, and thought, "Hmm, that's a real trap. Wonder why nobody every taught me that?"

The problem is, unless you've got an advanced autopilot, one that knows the airspeed, it will keep flying the airplane for you just as you asked it to do. On approach, a similar thing can happen. You set an altitude, reduce power, and the autopilot will keep flying the airplane for you at the altitude you told it to maintain. If the power's low enough, the airplane will slow down. If there's a stick shaker, that will activate. If there's a pusher, it will push.

And if you're tired, under-trained, not focused and not ready, you're going to be in a tough spot.

By Robert Goyer

Comments

[Charles Severs]

We have not been told but it is fairly apparent that they let the speed bleed off with an ice load to stall. The stall was a fairly high stall speed and above normal stall but a LOT of speed bled off while they were commenting about the ice build up. When the stick pusher pushed the pilot thought it was a runaway autopilot or something other than stall and FOUGHT IT
right into the ground.
Does anyone get a pilots license without a fear of ice and stall training even in a Cessna 150??

[Marrese]

Agree with above. Cause: Failure to maintain flying speed --STALL SPIN.

In reading the Cogan air transcript, I am not sure: Were the deicing boots ever deployed? How can the so called "24 hr ice check" be performed enroute? If the anti-ice was on, why was there ice on the windshield.

[Timothy Burling]

On AP stalls, the very first thing my CFII told me on our first climbout during my first instrument lesson was, "Watch the autopilot - carefully. It wants to kill you!". I never forget that. I now prefer to handfly the whole climbout so I can manage the airspeed to safely climb as fast as possible. My AP (KAP140) doesn't have selectable airspeed. I know some newer ones do (GFC700) which sound safer but not immune...

[Bill Unternaehrer]

How does this relate to the FAA / NASA video on tail ice where the proper technique is to pull back in a tail stall due to ice?

[Harold Snyder]

Could he have thought it was a tail stall due to ice and pulled back??

Too many I's in the report unkle sky!

[AJL]

The 24 hour ice check is just to check and make sure the anti-ice systems are working. They can be turned back off if they aren't required, which changes with company policy.

They might have had the boots in slow mode and not fast,which changes the speed at which the boots cycle. Too slow of a cycle and too much ice too fast,well...

Ice can accumulate on a windshield, even with anti ice on, if the windshield heat isn't on high enough or the accumulation rate is too high. I think there could have been a combination of the two in this case.

The author mentions "They rushed through the approach brief". When you fly professionally, especially for a regional airline, you've shot the same ILS tons of times. I know the ILS approach at my local airport without using the chart and I'm sure many of you do too.

Does that make it right? No. Think about your non-flying job. Do you have thing you do automatically, without looking at a book,chart,etc, even though you should?

Flying for an airline,especially a regional, is much different then flying single pilot IFR or for pleasure or business. You fly ALL THE TIME. 7 hours a day sometimes, for possibly 3 days in a row (or more). The more you do it, the more automatic it becomes. The more complacent you become. Combine that with inadequate rest and training, you have a Colgan 3407. Every FAR or procedure is written in someones blood, this case will be no different. Unfortunately.

[Jason]

I had to quit flying for my airline. We never got any sleep. I never got to go home. My day off whas in whatever hotel in whatever city I was in with no car to go get food. I was in the plane sun up to sun down with no food. They gave me pretzels and cokes. They only paid me $600 per month for that and treated me less than lugguage. The TSA would not let me get through again after a short weather break to get a muffin in the terminal one day. I had enough so I quit my life long dream of being an airline pilot.

[Tony Hansberger]

These are the kind of thing the FAA and the NTSB should be more concerned about instead of badges on the flight line. The terriost in this case was the responsible agencies that manage us..

[Jacie]

I read the transcripts; chilling is correct. I am a low time private pilot who learned to fly because and air carrier just about killed us in a 777 on take off (before we did a turn). We lost the horizon and only saw a growing reef out our side of the aircraft. We were told the airspeed indicator was the problem and we were flying on the back up airspeed indicator. The pilots took us over the Pacific Ocean with out ever turning back to the airport we just took off from. This was on a commercial carrier! We reported this to the FAA and they never bothered to call us back.
As long as the airlines have the confidence to lie to the passengers the truth will never be know in some cases about what really happens on a commercial flight. The use of recording the critical phases of flights are removed from the paying passenger because we have to have all electronic equipment off during landings and take offs....... the consumer who pays for the service is the looser in this game.
Professional pilots can do so much to help the paying passenger understand what is happening by simply telling the truth. If more pilots would simply act human instead of covering their ass then we could get down to the real nuts and bolts of why airplanes fall from the sky. We need to include integrity as part of our checklists. What is the solution? I do not know, but I know when I launch my ship in the air I am as honest as I can be to myself about my abilities. Pony up if there are issues.

[Joel Godston]

Major--MAJOR--MAJOR MESSAGE! FLY the airplane.....anything else can Kill...and it did in this really sad illustration.

[Joel Godston]

major--Major--MAJOR MESSAGE! Know the POH procedures and FLY the Airplane....Anything else can Kill; and it it did in this sad illustration

[Ron]

The most chilling part of the simulation/transcript is where the flaps (which appeared to be wrongly set at ten instead of fifteen degrees)go to zero and the FO states that she put the flaps to zero. The Captain is to busy to do anything but grunt, trying to hand fly the airplane out of the impending stall.
The second most chilling part is watching the airspeed fall below Vref with less than approach flaps on.

[T.J. Kai]

This all stems back to a wonderful thing called "Deregulation". In addition to gaining a passenger compartment filled with "Bus People" in various stages of dress and undress, cleanliness and the lack thereof, we also found ourselves flying lots of legs for fewer and fewer dollars in aircraft that may or may not be maintained to the standard we would like to see.

As a result, we have nearly lost our once vaunted air transport industry altogether. And guess what? It is not going to get any better.

[Jim Cosson]

First of all, my sympathies go out to the families of the victims. In a sense, all pilots were victims in this incident in some small way. The reputation of aviation was tarnished and it will take time to recover. The media now covers all aviation incidents, no matter how small, and it makes us all look bad. Media reports are generally “factual” but the conclusions often miss the mark. The impression the public now has is that an airplane could fall out of the sky at any time and kill you. We, as pilots, know that those fears are irrational but that is a hard sell to the rest of the population. Add to that the impression that GA is for the criminally rich corporate CEO and we're in a real crisis.

We need to be doing better, private pilots and those flying for hire. We should demand good training for ourselves and expect no less than good decisions from every pilot, every day. Make a “no go” decision with confidence, if there is a question about safety. Know your airplane and train for the worst case scenario. Understand that technology does not always make us safer (see the fatal Cirrus incident at Front Royal, VA last year, N141SR) and yes, the autopilot is trying to kill you. Technology is great but it doesn't remove our responsibility to fly the airplane.

There were many factors that led to the crash of 3407. When viewed in isolation, no single factor should have been fatal. Combined… well, we see the result. I viewed the animation of the last two minutes of 3407 and my heart broke. I saw what was happening with perfect hindsight and I found myself yelling at the screen, willing the airspeed to increase to no avail. I’ve learned a lot from this story and I continually challenge my own skills and preparations for every flight. I hope that this report and article encourages all pilots to look carefully at themselves and their skills. I hope that the legacy of flight 3407 is that it made us all safer in the sky.

[Eric]

Interesting comments...I think that icing played a minimal role in this accident. Most simply it was a trim stall. With no increase in power (for whatever reason), a whole lot of drag increase, the AP just trimed up and up. It is my understanding that it was full up trim on an airplane that only uses 1-2 degrees of trim. When the shaker kicked the AP off and the pilot increase the power to max the airplane pitched up causing a full stall and over it went.
Fly the airplane and let nothing distract you from that, especially when you are in such a critical phase of flight.

[Tom Thunnell]

The scenario is simple - the evidence is prima facie, unqualified pilots !

Tom Thunnell

[Scott]

Jacie, as a professional airline pilot I have to say that it is nearly impossible to know whats happening in an airplane when you're sitting in the back. Things that look bad in the back are usually no big deal from the office view.

Case in point, my father is a former F14 driver with about 27,000 hours in the 737, I'm just trying to illustrate that he's a pretty competent pilot with just a little bit of experience. He called me up the other day to tell me about some passenger that barged into the cockpit during de-planing to tell him that there was an airplane "Less Than a Hundred Feet" from them on takeoff. Then the pax proceeded to tell him that he couldn't BS him because he "Was a Private Pilot." My dad tried to explain to him that no such event ever occured. If said event occured there would have been numerous TCAS warnings followed by a very aggressive RA maneuver. Of course the dude wouldn't hear it, after all he's a private pilot.

Honestly the entire situation would be laughable if it wasn't so infuriating. I'm guessing your situation in the 777 was pretty much the same thing. Losing altitude and overbanking during takeoff in my lowly Saab 340 would result in multiple GPWS (Ground Proximity Warning System) callouts, let alone the warnings in the 777. Believe me Jacie, the second those of us in the pointy end of the aluminum tube messes up everyone knows. From ATC to the dispatcher watching our flight to the Maintenance dude that pulls the FDR to review our profiles.

As someone that spent a few years in GA before transitioning to the airlines I can tell you that the two are completely different animals. Tooling around in your 172 isn't going to give you even the slightest idea of what's happening during a flight in a 777. Give the pilots the benefit of the doubt, to get to a 777 they've probably been flying for more than 20 years. I'm guessing they know what's going on.

[John]

Not to make judgement because enough of that is going around, hopefully we can all learn from mistakes, ours and our peers. To help with a couple questions here.

1) A tail stall is not going to be noticed in a 400 due to the horizontal is hydraulic. If ice was an issue you would never get the tail stall issues, it only appears in aircraft with a mechanical connection, ie; cables or rods

2) The icing system was on and this aircraft will carry more ice that you could possibly imagine. Even with the windshield heat on a small corner does not heat well and it gets ice on it. It is not a problem and does not reduce visibility. I do not think we will ever know if ice had much to do with this.

3) Anyone doing training in the simulator for this airplane has done stalls, quite a few. In training and on his Type ride. They are all run to the stick shaker when we are waiting for it and a recovery is immediate. Unless it is not done correctly you would never get a pusher. But think about it when was the last time you practiced a stall when you were not ready?

We all respond to emergencies differently. The question here is with some due diligence could this be prevented? Absolutely

[JR]

The pilots here were more than qualified, and had plenty of experience. The problem, as mentioned in several other comments, is complacency. At the regional level we fly 6 legs a day, 4 or 5 days a week. Once a year do we practice steep turns and stalls, and it's in the sim. So when something like this happens, we are not prepared for it. Were the pilots tired? Probably. Pilots with far less rest are asked to drop ordinance on precise targets in the middle of the night every day, with less flight time than these pilots. It's not training, it's not rest, it's complacency and not thinking. In the words of one of my sim instructors, "you need to stop and think". There's not much time, but there is time. We have the worlds fastest supercomputer in our skull, use it!

Those that say the pilots were inexperienced or unqualified are not educated on the matter (and usually not pilots!). And, unfortunately, publications such as Flying and those from AOPA are speculating, not educating.

[JOJO]

Should never use the Autopilot when in icing conditions. Hand fly it...it will tell you what its trying to do

[Kenneth R. Rusnak]

It is time that new pilots had to study aviation history as part of their training. Almost this same senario occurred with an American Eagle Flight in Indiana a number of years ago and the aircraft was a high wing ATR 72 turbo prop. It was from this crash that the FAA warned pilots in heavy icing not to use the autopilot! I am sure that the Colgan company regulations also prohibited autopilot use in icing conditions.
So, why was the autopilot on?

[GE]

I would say hand-flying in icing conditions does indicate a training issue and/or an experience issue. As pilots, there are certain conditions (atmospheric or otherwise) which should "trigger" a response, paricularly as PIC, to prevent a snowball effect. If not, to me, that is a qualification problem.

[B-737576777DC10]

Speed tape airspeed indications are not as obvious as round-dial indications. You have to get used to them. FO: "I put the flaps up"!!! What a comedy of errors. Stick Shaker is probably the most unambiguous warning: tied into flap configuration and angle of attack vanes on every transport plane I have flown. A damn shame deep stall and stick pusher training wasn't required by the FAA. The DO sign off on all this. Old argument: do we teach spins or not? Do we just do stall recognition and recovery or do stalls? I read Sterile Cockpit was being violated. The transcript presented, if complete, didn't reveal they were doing extraneous BS.

[B-737576777DC10]

Full long transcript does reveal too much well into the arrival.

[David Dibbell]

From the media reports soon after the crash, until recently when I reviewed and compared the flight data recorder report and the cockpit voice recorder transcript from the NTSB hearings, I was speculating that the captain misinterpreted the stick shaker as a tail stall induced by icing. How else to explain the "aft force on the control column" which was being reported? But now my speculation seems unfounded. Consider that immediately following the stick shaker activation the power levers were advanced to a high setting and remained there. Engine torque came up to a high value within seconds. So one of the captain's immediate reactions (to add power) was correct as a response to the stick shaker warning of an impending conventional stall. But then he simply failed to fly the airplane. The high pitch attitude indication should have produced a learned response to push to capture the correct pitch attitude. The mystery remains as to why aft force was applied to the control column. I can understand fatigue and/or simple inattention to explain the airplane's decaying airspeed with too little power, but it doesn't explain the incorrect pitch input after the stick shaker activated. Is that not a more basic airmanship issue than an aircraft-specific training issue?