Colgan Blames Faulty Plane, Pilots for Buffalo Crash

[Matt Molnar]

Pilots only partly to blame for Buffalo crash: report (http://www.cbc.ca/world/story/2009/12/14/buffalo-crash-report.html?ref=rss)
Last Updated: Monday, December 14, 2009 | 2:33 PM ET

A U.S. regional airline is critical of both aircraft manufacturer Bombardier and the cockpit crew in a report on the deadly plane crash near Buffalo, N.Y., last February.

Flight 3407 from Newark, N.J., went down in the community of Clarence Center shortly before it was due to arrive in Buffalo. All 49 people on board died, as well as a man in his house, which was destroyed when the Bombardier Dash 8 Q400 plane stalled and plunged to the ground on Feb. 12.

The operator of the twin-engine turboprop detailed what it believes happened in a report submitted last week to the U.S. National Transportation Safety Board.

Colgan Air said the cockpit warning system failed to adequately advise pilots when the plane's speed was set below the calculated stall warning speed. [Story (http://www.cbc.ca/world/story/2009/12/14/buffalo-crash-report.html?ref=rss)]

[NIKV69]

What a load of crap. How about hiring pilots that would be a little careful when flying into potential icing conditions and not whining about pay or whatever they were talking about when they should have been paying attention to the approach. I think all the facts are out and both pilots fell asleep as the switch. Big time. To blame the manufacture of the aircraft is downright silly.

[hiss srq]

Nick, your for the most part accurate. However, I think people are putting too much emphasis on things that are not really relivant. Your right that the sterile cockpit rule was violated. (No off task below 10K) Second off, they blew it on situational awareness. Big time. The airlines hiring practice really means nothing in terms of the findings though the media will certainly latch on to it because its a heartstring tugger for the public. The fact that they did not react to a stall correctly is a huge redflag in Colgan's training practices. The F/O retracting flaps when entering a stall also brings questions about. Dont touch the flaps in a stall. Flaps will do more to help than harm you gnerally speaking (Big wing means slower flying speed). The fact that they completely defied logic though by pulling instead of pushing in a stall to me is just un justifyable. There were a great deal of WTF moments in that crash but it does bring to question alot of questions within the practices at regionals. Generally speaking the regionals have fairly experinced crews contrary to beleif. Earlier in the decade alot of airlines were hiring these "300 hour wonders" but for the most part it was a small number within the ranks. Those that were hired with that kind of time at most regionals didnt have the seinority to protect themselves in the last round of furloughs though so they are most likely on the street anyway right now or CFI someplace. Pay is a huge issue in the airlines, not just the regionals and not just in the sharp side of the cockpit though. Right now there are alot of regionals with very experinced pilots contrary to beleif. Colgan has always been kind of the basterd airline though. Colgan is where those who wouldnt make it with say Comair, Republic, Skywest etc go. Fast upgrade times which means tons of PIC time in short order which makes them competitive for jet jobs down the road. Now if you look at Piedmont, there are alot of crew there with over a decade seinority there. That I think is in part due to the flowthrough carrot that was dangled in the early decade that they could go from the Dash to right seat on the 737 at mainline. That is a diffrent story for a diffrent time though What it all comes down to is that crew blew the donkey on situational awareness 100%. Flight into icing conditions defineately was the catalist for the crash but the primary reason it was a crash and not an incident was because situational awareness broke down. Plain and simple. On a personal note I still question wether there was tail icing as well.

[hiss srq]

And also, how convient of Colgan to try to blame the Dash on it. I don't buy it. The Q4 has it's issues by far but I do not agree that the airplane design in any way was a factor in this crash.

[NIKV69]

Any prop is bad in those conditions and quite frankly I think they shouldn't be flown in icing conditions. They can't break off ice or deice as good as jets and as we saw here even if no ice was on the tail they lose control long enough to kill everyone. Mostly due to the fact they didn't know what was going on.

[hiss srq]

Any prop is bad in those conditions and quite frankly I think they shouldn't be flown in icing conditions. They can't break off ice or deice as good as jets and as we saw here even if no ice was on the tail they lose control long enough to kill everyone. Mostly due to the fact they didn't know what was going on.
Agreed. Bleed air is defineately the way to go. Most turboprops use boots though which "pop" the ice off the leading edges.

[PhilDernerJr]

i was surprised by the flaps-up thing, too. They hit trouble and the first thing she did was raise the flaps on her own. This might have been because the flaps had just been extended, and maybe she thought whatever was happened was caused by that. *shrug*

What are the de-icing systems on the Q?

[hiss srq]

It is boots. I am not overly familliar with the Q4 as I am the 1 and 3 but it is boots.
Horiz. Stab Vert Stab wings and props. No heat blankets on the wings as some other airplanes have. Phil, part of me thinks maybe her reaction was that they had asymetrical flap expension which would cause a very odd yaw and roll motion but even then. It would escape me why they did what they did retracting the flaps and pulling against the stick shaker.

[USAF Pilot 07]

Phil, part of me thinks maybe her reaction was that they had asymetrical flap expension which would cause a very odd yaw and roll motion but even then. It would escape me why they did what they did retracting the flaps and pulling against the stick shaker.

In most cases the asymmetric flap checklist will direct you to return asymmetric flaps back to the last position where they were symmetrical. Could be if she had just moved them and they experienced this rolling motion right after she automatically thought the flaps were asymmetric and tried to put them back up.

If she's not flying, she's probably not going to be the first one to feel the rolling motion though and I would think the pilot flying would have said something prior to her moving them back up. Then again, any number of things could have been going through their minds; there's so many ways to speculate what could have happened. The fact that she moved them back up without verbalizing it or at least running it by the captain is pretty non-standard.

IIRC the biggest factor involved in the crash wasn't necessarily putting the flaps back up (which was a contributing factor but at ~3000' AGL shouldn't have proven fatal), it was the pilot's application of incorrect stall recovery procedures that caused them to roll over and get into an uncontrollable nose-down situation.

[NIKV69]

it was the pilot's application of incorrect stall recovery procedures that caused them to roll over and get into an uncontrollable nose-down situation.



Good point but I don't think it may have mattered. If the animation is how it exactly happened they got caught with their pants down, he wasn't flying the plane and even if he pushed the nose down abruptly at that altitude once the stick shaker went off I doubt they could have recovered. Not to mention if the ice build up was as bad as I think they couldn't have got the airspeed back up in time.

[hiss srq]

Ice does not have a corrilation on airspeed. It has a direct effect on lift. In a piston engine airplane if you left your Carb heat off ice would affect engine performance and thus airspeed. They screwed up on a fundimental procedure on top of the fact that they violated the on task below 10 k rule and bungled it with flaps and pulling. The critical points in the chain were when the F/O decided to retract the flaps and the captain pulled back. I don't know why they werent speed checking for flaps and how they didnt notice that the autopilt had trimmed them to the maximum pitch it could for the decaying airspeed. Where was their instremunt scan? Really....

[USAF Pilot 07]

it was the pilot's application of incorrect stall recovery procedures that caused them to roll over and get into an uncontrollable nose-down situation.



Good point but I don't think it may have mattered. If the animation is how it exactly happened they got caught with their pants down, he wasn't flying the plane and even if he pushed the nose down abruptly at that altitude once the stick shaker went off I doubt they could have recovered. Not to mention if the ice build up was as bad as I think they couldn't have got the airspeed back up in time.

If he would have applied correct stall recovery procedures I think we wouldn't be here talking about this right now unless the aircraft was so covered in ice that it couldn't maintain produce any type of lift.

The stick shaker is simply an indication of an approach to stall, not an indication of a stall. The mantra "max, relax, roll" is a familiar one to those who fly. Generally if you get the stick shaker unexpectedly the appropriate action is to firewall the throttles, relax the back pressure on the stick, roll wings level, get out of the stick shaker and then max perform the airplane (nibling in and out of the stick shaker) in a climb up to a safe altitude. It sounds like in this case instead of relaxing the back pressure, the pilot kept pulling back on the yoke in the stick shaker until the aircraft finally did stall and rolled off to one side and started to spin nose down. Just a monday morning armchair analysis though...

It sounds like there were a lot of factors involved in this crash (as is usually the case). Unfortunately all them combined produced the end result. :(

[hiss srq]

Situational awareness is really what it boils down too. They let the airspeed decay somehow without noticing the increasing pitch as the AP tried to maintain altitude given the loss of airspeed. I don't know alot about the Q4's systems and avionics outside of what is basic to all Dash8's but I cannot fathom the instrument scan that they were using. If I not mistaken there is a trend arrow on the Q4's ASI in the PFD screen which points to the airspeed trend. Those of you familliar with more modern glass cockpits like the Airbus and Embraer know what I am describing in the most basic terms. If anything how can you not notice that as well as that bright red tape rapidly converging with your indicated airspeed? To me that just makes no sense. Maybe the FAA needs to amp up the requirements in Part 121 etc for ice procedures in training. What to look for, how to handle it etc. In Turboprops you can generally manuver a bit better than you could in a jet as far as getting through it fast and still being on speed and sink etc when you cross the marker. In a jet it gets a bit more complex though as far as peramiters go on an approach. The requirement varies from airline to airline. At American its established, configured and spooled by 1,000 feet. At USAirways it was 500 feet. In a turboprop verses say a 757 or an Airbus you should be able to do it alot quicker. Refer above to advice a former ATR guy at Eagle gave me once.

[NIKV69]

If he would have applied correct stall recovery procedures I think we wouldn't be here talking about this right now unless the aircraft was so covered in ice that it couldn't maintain produce any type of lift.



I am still a bit skeptical. Maybe if he was flying by hand and felt everything but both pilots were caught off guard and confused to what to do for the few seconds after the stick shaker activated. If he was alert and did act properly he may of been able to save it but I don't think he was that good a pilot or he wouldn't have let his mind wander the whole approach.


It sounds like there were a lot of factors involved in this crash (as is usually the case). Unfortunately all them combined produced the end result

Factors come into play in everything. They are part of life but it usually one grave error that brings a plane down and as Ryan said losing awareness in icing conditions is basically why they are all dead. Not the aircraft as Colgan is very wrongly claiming.

[ch2tdriver]

While it wasn't a DHC-8, I've flown the DHC-6 which had the horizontal stab rather high up. IIRC we had a placard on the yoke that basically stated that if you even suspect tail icing don't lower flaps more than 10 deg. I was flying it in the mid 90's and had friends flying the ATR42 for Eagle when the one went in with suspected tail icing. After that they really harped on us with training for the icing proceedures, most likely as a result of the FAA pressure.

Reviewing this case I really believe that the crew was convinced that they had the tail stalled due to ice accumulation and were applying the training they received. Maybe not from Colgan but there previous training?

Here is what is mentioned in AC91-74:

Recognizing and Recovering from a Tail Stall

You are likely experiencing a tail stall if:
When flaps are extended to any setting, the pitch control forces become abnormal or erratic.

There is buffet in the control column (not the airframe).

Recovery from a tail stall is exactly opposite the traditionally taught wing stall recovery.

Remember, in a tail stall recovery air flow must be restored to the tail's lower airfoil surface, and in a wing stall recovery air flow must be restored to the wing's upper airfoil surface.

Here is how to recover from a tail stall:
Immediately raise flaps to the previous setting.
Pull aft on the yoke. Copilot assistance may be required.
Reduce power if altitude permits; otherwise maintain power.
Do not increase airspeed unless it is necessary to avoid a wing stall.


Pete

[USAF Pilot 07]

ch2tdriver:

Are crews trained extensively and regularly refreshed in the sim on not just wing but also tail stalls? Wouldn't a tail stall cause more of a nose-down attitude?

I just watched the NTSB "re-enactment" of the situation, I have some questions that maybe you can answer...

I'm assuming the Q doesn't have auto-throttles...

What are general speeds of the Q clean/dirty? What kind of stood out first was that at 0 flaps the aircraft was around 170 knots (getting down to 167), but at 5 flaps it seemed to stabilize around 185 knots (getting up to 188). Not that it's unsafe, maybe the airspeed just got away from the crew a little...

Also, haven't listened to the CVR for this in a while but would proper procedure be to increase bug speed in icing conditions? Did the crew talk about this at all? Wouldn't the Q's flight manual recommend the autopilot be off during icing encounters especially during approach and landing?

What struck me the most was the throttle position through all of this. The throttles are in idle the entire time until the stick shaker is encountered. When the capt calls for gear down airspeed is 174 knots. 20 seconds later, with the throttles still in idle airspeed is 130 knots with 10 degrees nose up. That seems like a pretty dramatic airspeed loss without at least being mentioned by anyone. I suspect it fell out of both pilots' crosscheck.
Those 20 seconds look exactly like the setup to practice a landing attitude stall recovery.

I don't know how the onset of a tail stall would look like, but I'm curious to see the similarities/difference. I want to say I may have practiced a tail stall once in the T-1 (BE400) sim, but I really can't remember. It definitely wasn't made a high priority issue. We used bleed air and electric heating elements for anti-ice in the T-1 though.

Ice and airplanes generally don't mix. If tail stalls are more prevalent in smaller, turboprop aircraft, it sounds like it could be a serious problem, especially with all these puddle-jumper planes flying into areas that may not deal with a lot of air traffic (lack of PIREPs) and where weather reports may not always be very reliable...

[hiss srq]

The Q does not have an autothrottle system built into the package. An increase in speed is always smart in icing and yes at most carriers handflying is the practice instructed upon in icing conditions. It give you a better control feel. Very good point!

[ch2tdriver]

USAF pilot 07

You are a lot more current than me as its been a while since I did any line flying. All my flying nowadays is as a CFI.
Like I said I only flew the DHC-6 so can't comment on the DHC-8 or Q for that matter. I remember that on the oral portion of a line check, the check airman basically covered icing proceedures, questions like "Would you climb or descend when encountered? Why would you do one or the other?" and discussion of when to deploy the boots as "Bridging" was the big topic back then. (Nowadays the accepted procedure is "blow the boots at the first sign of accumulation and don't worry its not an issue" thought)

Once the ATR-42 accident occured we suddenly received dedicated ground training as a group in icing proceedures and techniques. This is where I first really learned about tail icing and tail stalls and what are the indications. I remember the discussions where if the flaps are deployed and a buffett is encountered, to immediately retract them. A tail stall will cause the nose to pitch down. I saved all the material from those days.

While the Colgan case may not have been a tail stall, I think its possible that the crew may have been in the mindset that it was and began to act accordingly. It may have been emphasized in their training? IDK. Their actions regarding pull up, flap retraction and the power settings seem to indicate this. I guess what I'm trying to say is that the crew may have been properly trained, but misinterpreted events and applied the wrong training.

Pete

[USAF Pilot 07]

Thanks for the reply! Tail icing sounds like a bitch. Although it's pretty apparent they let their airspeed and throttles fall out of their crosscheck in this situation, it's very possible in the short amount of time they had to react to the situation they thought they were in a tail-stall (although it sounds like the Copilot was out to lunch and had no idea what was going on). We'll never know but, maybe the pilot had just had a refresher, just read about tailstalls or whatnot and it was fresh in his head. I'm a little surprised they never verbalized anything, although in that short of a situation I'm sure they had more than their hands full.