Aviation Investigation Report Out Boeing 737-210C Controlled flight into terrain
 

Aviation Investigation Report A11H0002 (today)

Controlled flight into terrain
Bradley Air Services Limited (First Air)
Boeing 737-210C, C-GNWN
Resolute Bay, Nunavut
20 August 2011

Summary

On 20 August 2011, the Boeing 737-210C combi aircraft (registration C-GNWN, serial number 21067), operated by Bradley Air Services Limited under its business name First Air, was being flown as First Air charter flight 6560 from Yellowknife, Northwest Territories, to Resolute Bay, Nunavut. At 1642 Coordinated Universal Time (1142 Central Daylight Time), during the approach to Runway 35T, First Air flight 6560 struck a hill about 1 nautical mile east of the runway. The aircraft was destroyed by impact forces and an ensuing post-crash fire. Eight passengers and all 4 crew members sustained fatal injuries. The remaining 3 passengers sustained serious injuries and were rescued by Canadian military personnel, who were in Resolute Bay as part of a military exercise. The accident occurred during daylight hours. No emergency locator transmitter signal was emitted by the aircraft.

http://www.tsb.gc.ca/eng/rapports-re...2/a11h0002.pdf

Transportation Safety Board of Canada - Aviation Investigation Report A11H0002

Captain way off localizer with FO telling this to captain several times.

The Canadian Military and Nav Canada got of lightly on this one, the extreame intercept angle and total lack of any real IFR operating knowledge, along with totally non standard procedures were in fact the first link in this train of events. There have been other recent incidents due to the the military trying to usurp Nav Canada/ ATC authority at mixed use airports, in this case not helped by an old aircraft with numerous avionics updates, lets hope this does not happen again.

It doesn't take a brain surgeon of a captain to figure out that if the localizer is showing full deflection on one or both of the instruments during the ILS and the FO is repeatedly saying that you are off course toward high terrain that it is time to go-around no matter what airport you are at. That is regardless of the stuff that happened earlier in the flight. There is always an excuse, but there is responsibility that comes with being PIC. Can't handle the obvious then don't take up the position.

This is quite common in some companies, so let me explain my story on a future accident.

Last year I had to give line trainning to a captain fresh from sim on a new aircraft, 15.000 hours and had flown many different aircrafts. On the first days I noticed that something was going wrong with him. Another trainning captain took him over and had the same problems.

On my return with him I found out that he had no idea how to IFR, and I mean no idea!!! I don´t know how he passed sim at CAE. Further more asking other pilots they already new that, even OPS manager, but company didn´t care.

But that was not enough, he didn´t know how to read a METAR, avoid a thunderstorm or basic stuff like calculating fuel, etc.

We recommended to ground him and go back to basic flight school. What happened? company had him on trainning with a TRE for 10 months non-stop but still has no idea, fired us and let him fly...

The alarming thing for me is that despite all the emphasis on and training in CRM, and that this was in Canada, it still happened. Frightning!

After a quick skim through......
Copilot lacked assertiveness......he was aware of the danger , yet allowed himself to be browbeaten into accepting something he knew was wrong.

The Captain was complacent, lacking in humility and CRM.
IMO, As Commander, it was his duty to heed crew warnings and ask himself why Crew, GPS and Navaids all disagreed with the location he "thought" the Autopilot had them at.
I fail to see the relevance of the Military...they weren't flying the aircraft.

A very sad and sorry tale.."get there-itis?"

One of the finest air accident reports I have ever read in my over sixty years in the industry. A must read, especially for airline first officers whether brand new MPL's or highly experienced. It is a most comprehensive report and it is a dedicated pilot who will stick with it to the end. But it is gold standard.

A quote from the report.

Are SOPS considered as guidance for wise men and for the obedience of fools?

Alioth, like PEI372, I have spent a fair chunk of my life sitting around the edge, or sitting in, smoking holes in the ground trying not to puke and do my job, so forgive me if I find your assesment of this crash to be a little simplistic. I for one cannot find any reference to reading "needle impact traces" on the Captains HSI as to what his side of the office was in fact indicating, its maybe in there, but I havnt been able to find it as yet. The placing of a "lash up" military ATC l system using uncalerbrated equipment and personell was in fact the first link in this chain of events, this doesnt change the fact that the crew were in a state of obvious overload and confusion in the last seconds of flight, but it certainly played a part and deserves more than a passing mention in the report, if one adds in the highly modified state of the avionics on these older aircraft, with set ups which have many traps and pitfalls its suprising there are not more CFIT incidents. The post by PE1372 should be mandatory reading for all in Nav Canada and Transport Canada.

How many times you have been to a crash site means nothing and does not add anything to your argument.

There are many reasons why an aircraft can be off course. If your FO says to you multiple times things such as he is showing off course and full scale localizer deflection and that the GPS is also showing off course, reminds you about a hill that is on the off course track, suggests going around to figure things out and then you ignore him, continue and hit the hill, you are not going to look like a very professional pilot.


There are first links in the potential chain of events on every flight.

When asked whether a go-around was required with full-scale deflection of the localizer, all pilots interviewed stated that this condition required a go-around. However, all pilots interviewed agreed that an FO could not command a go-around, and could not see an FO taking control from the captain. They stated that the FO would likely make a suggestion or negotiate with the captain regarding the need for action.

Hope they think differently now

Alioth, I wish crash investigating was as simple as you seem to think it is, yes the aircraft was of track, yes the F/O stated this fact, but I can demostrate to you in the simulator how it is fairly easy to overlaod a pilot to the point that the computer between his/her ears just doesnt recieve or process what going into those ears, as for time spent at crash sites, Ive often thought that all pilots should do at least one "on site" investigation visit, it might make some pay more attention to to the little things which so often lead to such tragic events.

Incompetent first officers?
 

It is difficult to understand how any captain, no matter how bone headed could ignore such concern from a first officer. Have any of the captains here ever had a first officer express such discomfort with an approach, and continue it, being certain in your own mind that he was wrong? Not having flown commercially, I have no knowledge of the range of communication/mis-communication in professional pilots cockpits.

Peterwelch, the Captain was not a "bonehead", nor was he one to generate an atmosphere of intimidation in the flight deck, he was most likely in total overload and task saturation and/or task fixation, as I pointed out in a previous post we can replicate such conditions in the simulator with multipe distractions/mechanical failures. The final factor which doomed this flight was an EXPERIENCED non -asertive F/O, just the final factor in this sad affair, note I said "experienced" F/O, the USA has just increased the hours required to be in the right seat in a scheduled airline, Canada has just gone the other direction and introduced the "Multi Crew Lic",I can see problems with crewmembers ignoring an F/O who isnt allowed fly an aircraft solo and has less than a hundred hours of flight time, but that another discusion altogether.

I agree an excellent report, much to be preferred to the often anodyne reports that offer few lessons prevent future occurrences.

In some ways reminiscent of another excellent report into a CFIT incident to an American 757 at Cali, Columbia in 1995.

The report of the Columbian investigation is here:

Cali Accident Report

but one pertinent extract is:


The Cali investigators also commented on peoples reluctance to change a decision, once made, even in the face of new and conflicting information.


The Canadian report does not say how many times the Captain had operated
into Resolute Bay, but mentions that he had previously made several successful approaches in similarly difficult conditions and that he knew that another pilot had experienced a similar anomaly on that approach.

It seems that, 16 years on, we are still not recognising the dangers of Recognition Primed Decision Making. Nor are instrument approach charts graphically representing terrain near airports or flight paths as recommended in the Cali report

While company SOP go into detail on what the PNF should do when an approach become dangerously unstable and requiring a go-around, it does not give specific advice - purely the correct warm and fuzzy words to say like please captain we are too high I suggest we go-around. As numerous accident reports have testified captains have ignored the official calls and the aircraft has crashed.

That point is well covered in the Canadian Boeing 737-200 accident report.
For some reason it is placed in to too hard tray and the PNF is left to use his own judgement on how to physically take control from an aggressive captain intent on pressing on regardless. The result of two pilots fighting each other to keep or take control at low altitude is bound to finish in tragedy.
One airline in Australia years ago, took the initiative and advised first officers that if their SOP calls for a go-around were deliberately disregarded by the captain, they should call once more for a go-around and if disregarded, they should select the landing gear to up. It was assumed that no captain would then deliberately land wheels up just to make a point and he would be forced to go-around.

While it was conceded that this action could place the aircraft in a dangerous position if close to landing, the alternative of a crash into terrain was infinitely worse. It was stressed that selecting the gear lever to up to force a go-around was essentially a last ditch action to prevent a certain CFIT. If that SOP had been used in most of the crashes known to have followed seriously unstable approaches - including the Canadian 737 accident - then these tragedies would likely to have been avoided.

I completely understand the Canadian reasoning regarding protecting of the CVR, but I do feel this is one of those cases where a proper transcript of the words the crew used would have been immensely illuminating.

As one in the right seat, what one would do if a Captain refused calls to Go Around is high in the mind. I am aware of an incident in gusty conditions a few years back where the crew received a hard windshear warning on final approach and the Captain continued to land the aircraft with the FO shouting "GO AROUND". The FO considered any attempt to take control from the Captain was too dangerous.

CRM training thirty odd years ago emphasised the use of an airline specific emergency language by the FO and was incorporated in the Ops manual. Qantas was 'Captain you must listen to me'. If this was said then it was a mandatory incident report.

Different airlines and different cultures used different phrases. My favourite was Lufthansa -their response was to call the Captain by his christian name.

4greens,

he did call the cpt by his christian name !

This is stupid
 

We had defined terminology too: "first name; i dont feel comfortable with this approach, do you think we should go around"
-followed by-
"This is Stupid"
The idea of raising the gear in what's already an unstable approach is, er interesting, but that brings all a host of other problems too; especially if the PF is fixated or saturated. TOGA is perhaps a better idea..
I fly by the simple missive; if there is ANY doubt, then there is NO doubt..
~blue skies. Horrible accident.

Educating me
 

Clunkdriver, thanks for starting to educate me re: task saturation. How easy is it to create this in the sim, and are there any signs of it that identiy it well enough that in and of themselves could be a sign to go around. My apologies for use of the bonehead term. I know this poor guy didn't want to crash.

TOGA will certainly be an effective method of forcing a GA. But the strong pitch up caused by GA thrust would certainly catch the offending captain completely by surprise. Especially in IMC he may not have a clue what the co-pilot is up to and by the time he twigs to what's happening allow the aircraft to reach an alarmingly high body angles which could lead to a stall which is not what the co-pilot had in mind!! TOGA can be dangerous if unexpected...

But the PNF who pressed TOGA is expecting it and can help an overloaded PF by taking control or applying pressure or trim.

It's more a psychological issue of having FOs trusting Captains and believing that their vocal inputs will get them back in the groove. I've been there - I had two events where in hindsight the Captain was more incapacitated or confused than I had perceived and taking control would have been the better course of action rather than continued vocal intervention. I think it's something that really could be helped by training in the sim, where the Captain is told quietly by the TRI to start making mistakes and misjudgements but the FO isn't made aware of the scenario, a bit like when the typical RTO with incapacitated pilot is done, but much subtler. The idea is to show that mild incapacitation might allow the PF to give all the SOP responses but to make poor decisions and fail to react to other indications and that the FO has to make a decision to take control even though the other guy appears concious. Incapacitation training seems to be far to obvious at the moment, with the incapacitations always being total.

Very true. It is sometimes very ard for someone who has become disorientated or gone down a wrong mental avenue to recognise it, and so is difficult for them to agree with a PNF telling them a contradicting tale...

A trawl through any accident database will demonstrate that on the day experience counts for very little. There are just as many memorials to experienced crews as there are to inexperienced crews. The goal is to always be effective as crew, that way you may also live to be experienced.

According to the report the FO had just over 100 hours on the 737, so was recently line released. He was previously a turboprop captain and does not seem to have had any other jet experience. He gets paired up with a guy who should know what he is doing. The captain mishandles the approach and the FO who is obviously well behind the aircraft knows something is wrong but clearly cannot quite believe what is happening. I would not say that the FO was particularly experienced in this case at least on a relevant aircraft. In any case it is a sobering report. As they say better one go-around too many rather than one too few.

Hindsight bias
 

There is a lot of hindsight bias going on in this thread. :mad: It's easy to talk about what the FO could have done or might have done and the assume that everything works perfectly from there on out--maybe it does and maybe it doesn't. Maybe the aircraft gets into a situation where the FO is doing one thing and the Captain is doing another and it winds up in the dirt anyway.

What caught my eye is this:
So thirty seconds before the crash the Captain is in a mental place where he has got it all figured out. He's seen this before. And he is following this cognitive map in his head despite the fact that the FO officer is telling him the map is wrong.

So let me play the troublemaker a little bit.

I wonder how much the FO nagging added to the Captain's task saturation. Perhaps at some point in time the Captain simply started to tune him out. The Captain is may be thinking, "I've got it all figured out. Why the hell is he bugging me?" Maybe if the FO had just shut up the pressure would have been off and the Captian would have figured it out on his own.

There is more than one way to engage in hindsight.

I would have thought the last thing you want is having both pilots putting inputs into the controls at the same time and possibly in opposition. Recipe for confusion and disaster.


If the PNF is actuating the stab trim without the PF being aware of it, it is possible the PF may think he has a runaway (uncommanded) stab trim movement and act as per QRH Runaway Stabilizer which is Control column...Hold firmly. Stab Trim Cutout switches....Cutout.


Meanwhile the nose is rearing up under full thrust. The mind boggles at the scenario

A few decades ago, I was right seat in a Viscount simulator when PF managed to get inverted (with the help of a white on black WWII vintage AH) at 16,000'.

I spent the next minute or so shouting in his ear that:

1. We're upside down!
2. Roll it back up!

His auditory channel had very firmly shut down:uhoh:

I had a little talk with myself about how hard it is to take over -- and that if I want to walk away from a similar situation in a real a/c, I'll have to take over in time to recover.

There's major psychosocial barriers to the FO taking over.

I second the suggestion that the simulator is a good place for PF to set up a situation where PM has to intervene.

In some people's eyes, a go-around in the circumstances described in the 737 accident, is seen as an admission of mistake and a "it can't happen to me" mindset takes over momentarily. The wavering of the first officer is understandable as he contemplates the awful realisation the captain is wrong and yet the F/O is reluctant to shove open the power and haul back on the stick since he knows the consequences for his job if indeed the captain was right. I doubt all the SOP or education in the world can cover the situation that any first officer finds himself under similar circumstances.

It might even boil down to one's personal physical courage. I believe that to be the case here. An F/O is seen as making serious waves if he forcibly takes over the controls at any time. Ethnic culture also plays a large part in this sort of lack of decision making as we have frequently seen in accidents such as the Asiana B777 crash at SFO. Not that appeared to be a problem in the 737 accident.

As an avid reader of accident reports, instilled in me from my early years as a military pilot - and where these reports were always readily available in military crew rooms, I believe each operator's flight safety manager should ensure that both overseas and local (own country) accident and incident reports should be readily available to crews. It is all very well having to regurgitate endless published bumpf on CRM and TEM or whatever is the next fancy name. But studying the cold hard facts in an accident report should never be boring and will generally be remembered.


What should interest a professional pilot is why a pilot made or failed to make a critical decision and the deadly result. An accident report tells you that. It should not be left to individual pilots to research for themselves because few will, in my experience. They have too many other activities to concern them - unless they live for flying (and devour Pprune accident reports!)

"There but for the Grace of God go I" reading of accident reports is a powerful educational tool. Those of us who have followed this particular thread will have already learned a valuable lesson IMHO and hopefully apply it when they go flying next time. Forgive the somewhat inarticulate rambling on a tricky subject...

PEI_3721 it is blindingly obvious that the crew got beyond their limits. Various posters including the one before my original post made reference to the co-pilot being experienced and like you seem to suggest it is not a relevant factor. Fact is that the first officer had very little time on type. It is also a fact that they crashed. A lot of other things being said are assertion.

I have been flying the 737 for many years with increasingly inexperienced co-pilots due to the way our industry is changing. I make plenty of mistakes like most people so I have a keen interest that crew resource management works. Airline management also would like to think that experience is not a safety issue. I do not agree.

In this specific case the mental overload occurred at least partly because the crew jointly allowed the aircraft to get well above path and rushed the approach. There seems to have been considerable confusion in the nav set up and approach briefing. The sops for callouts and other relevant items were not complied with. Surely more support from the pilot non flying would have reduced the workload on the captain and avoided his becoming task saturated?

More "support" from the PNF can be interpreted two ways. From the PF point of view or the PNF point of view. Too much"supporting" is often seen by the PF as irritating, superfluous and worse still, distracting. It all depends on from whose viewpoint.


We have all seen the situation where the PNF thinks he is doing what he thinks is the right thing and offering ideas, and advice but which drives the other pilot nuts dealing with the stream of words of encouragement - some of which is out of pure nervousness and even arse covering for the sake of the CVR and Big Brother..


In reality, the PNF is only verbalising what his own mind is thinking - all of which is not genuine SOP support but a chatter. In turn, this becomes counter productive as the PF is forced to acknowledge each transmission or comment or risk the PNF thinking "the bastard is incapacitated or in the "Red Zone" and needs more help and so the babbling continues.


There is something to be said for the old days before CRM was invented as a cottage industry. Then the PNF closely monitored an approach and silence was golden allowing the PF to concentrate on flying. Only if something was out of the ordinary would the PNF assess the flight path progress or whatever concerned him, and then would make a succinct comment which nowadays would be called a "support" call.

TeeEmm, I have probably got the wrong end of the stick, but it sounds to me like you are stating the main CRM element comes down to how advocational/confident the PNF is. As my posts say, it's not down to character strength, confidence, fear of the Captain or company reaction that is the often the problem, and the standard accident response to that effect is part of the issue. It's more down to the PNF not realising just how mind-f***ed the PF is. Hindsight is a wonderful thing, and that is how we learn. PNFs must learn that unless they see corrective action promptly, they need to assume the overload or incapacitation of the PF is more extreme and they need to take physical control of the aircraft. It is not an unwillingness to speak up, as these transcripts clearly demonstrate, it's the lack of understanding by the PNF of how disorientated the PF is.

I have a very open, relaxed culture on my aircraft - I'm horrified at the company recommendation above that the Captain's first name be used in these circumstances - I expect all of my crew, even the newest cabin crew, to call me by name all the time, and my FOs are relaxed enough to banter and take the pee out of me as I do them during the non-critical times of the day (this has been the norm at all the companies where I have worked, with only a few Cpts being stern, pompous and moody). Even so, I have made mistakes where the FO should have commented but failed to do so because they thought I had some other cunning plan, despite deviating from the brief, and that's been with green and experienced FOs alike - it wasn't fear of my reaction but their over confidence in me that was at fault.

Because things normally run smoothly, I don't think many FOs realise how quickly things can go catastrophically wrong. That's why I think practicing more insidious incapacitations without warning or briefing in the sim may be useful, just like practicing missed approaches from below minimums with surprises like runway incursions and wind shear to eliminate the "we're through minimums, so we're in" mentality.

Lots of talk about the FO on here. The fact that he lacked assertiveness hardly mitigates the fact that the Captain, in spite of repeated calls from the FO, continued a unstable approach in IMC in hilly terrain with full deflection on the LLZ. Short of hitting TOGA and taking control (and if they subsequently never saw the ground losing his career), I don't see what else he reasonably could have done to save the aircraft from the PICs decisions

Disagreement
 

Why can't the fact that there is a disagreement between the Captain and the FO, IN AND OF ITSELF, be a requirement to go around when it occurs during an approach? That seems to me a simple solution. How often do such events occur?

Petercwelch's suggestion is a simple and brilliant one. Perhaps the PF should be trained to go around if he or she hears the words "Disagreement, go around" from the FO. No need to explain the cause of the disagreement, just have the word as a 911/999 emergency call?

This is the type of proposal that strikes me as "trying to solve the last accident". I am skeptical that it would do much good for two reasons. First, it might just transfer the indecision of a FO from one dilemma to another dilemma viz. from does he take control to how much disagreement is enough disagreement to call for a go around. Second, if the mere existence of any level of disagreement is enough to cause a go around then the number of go arounds will go through the roof. While a go around in an isolated location such as the one in this accident is no big deal a large increase in go arounds in congested airspace like NYC might cause far more problems that it solves.

One should avoid the temptation to solve the last accident. In this case the PNF was correct and the PF was incorrect. But how many times have the PNF been wrong and the plane landed without incident?