NT Air King Air Accident - Pilot Discussion Thread

[MacStork]

My sincere condolences to the families of the crew that passed away as a result of this accident.
I have read all of the posts regarding this very sad and tragic accident. There are some very profound speculations as to what happened, others are simply bizarre, to say the least. However, everyone is entitled to his or her opinion, and I respect that.
Here are my thoughts.
The aircraft departed NE bound for Kelowna late in the afternoon on a nice weather day. The sun was behind them as they climbed to altitude. The crew noticed an oil leak from the left engine (with the sun behind them, this oil leak would be quite noticeable.)
The oil system on a PT-6-27 or 28 engine is pressurized and the oil cap must be properly secured to prevent any oil loss. The oil cap may be snapped down and appear to be locked, but if the O ring in the oil dip stick cap has not been compressed, the oil will slowly leak out. The one important check that you must complete after closing the oil cap, is the eagle claw check, where you run your fingers over the edge of the oil filler cap to ensure that the O ring has been compressed and a proper seal has been made. There are several documented cases where this exact scenario has actually happened and the oil has leaked out and eventually caused an engine oil pressure issue that has resulted in an In Flight Engine Shut Down. It can take up to an hour for the oil to leak out and eventually cause a low oil pressure indication (if the cap has been snapped to the down and locked position without compressing the O ring).
This is what I believe happened. There was a slow oil leak, but ALL of the engine instrument indications (particularly the oil pressure guage) would have appeared normal. The very experienced crew , after discovering the oil leak, would have immediately checked the engine guages. There is no emergency check list for any kind of oil leak, only for an abnormal oil pressure indication. If the oil pressure was in the normal operating range (the green arc or between 80 -100psi) then no emergency action would be called for. If the oil pressure was in the yellow or caution range (80-40 psi), then the emergency checklist requires that the Ng be reduced to 75%. Below 40 psi requires that the engine be shut down.
The crew reacted to this oil leak situation in a predictable and controlled manner. They decided to return to Vancouver to have the oil leak rectified. They had the choice of several airports to divert to if there was an emergency; Pitt Meadows, Abbotsford and Langley. But, at this point in time, there was no emergency.
They were asked twice if they wanted the Crash Rescue Vehicles alerted. Both times they refused this offer, most likely because there was no indication of any oil pressure problem. The second time was when they were on relatively short final to RW 26L at YVR. If there was an oil pressure problem where they had reduced power to 75% Ng or had shut down and feathered the affected engine, then I believe they would have requested the CRV’s to be on stand by for them. It was a judgment call, the crew did not feel it was necessary for this action at this point in time. In hind sight, perhaps it would have been a better call to have the CRV’s standing by.
At about 900 meters from the runway threshold, for unknown reasons, the aircraft made a hard left turn, probably stalled and then hit a lamp standard and came to rest on the Russ Baker Way, outside the perimeter fencing.
The question that I would ask is, what caused the aircraft to make that hard left turn on short final?
Let me digress and relate an incident that happened to me in the fall of 1971. I was the pilot of a Grumman G21A Goose landing at Seal Cove in Prince Rupert. I was at about a 1000 feet asl, downwind left for a Southeast bound landing at Seal Cove. I was checking the landing area for logs and debris, my eyes and head were looking out the side window at the water landing area. I looked ahead and an eagle was filling my windscreen. I reacted immediately and rolled left and dove the aircraft to avoid hitting the large bird. It was a survival type of reaction …. violent and instantaneous. I did not want that bird to hit my windshield. I had the altitude to dive to avoid the collision and the room to roll left. Somehow I missed the bird. There was no collision and no damage.
At this time of the year at YVR there are many migrating and resident birds in the general area.
The weather at the ETA was about 280 at 5 knots, and sunny. Weather or wind turbulence was not a factor. The aircraft was on short final with no apparent problems, other than the oil leak.
At 900 meters from the runway at an altitude of about 400 feet asl, the aircraft made a hard turn to the left. The aircraft was in a stable landing profile about to land on 26 L at YVR.
On short final, even a catastrophic engine failure would not cause that aircraft to turn hard left. All that would be required to control the aircraft would be some rudder input to maintain the direction and a bit more power on the good engine to maintain the approach speed. All the crew had to do, if there was a major failure, was to land the airplane and then deal with the problem on the runway.
I believe that it may have been a bird that filled the windscreen of the flying pilot. The reaction may have been similar to mine, a violent roll to the left to avoid the impending collision coupled with a hard pull up. At this point there would have been insufficient altitude to push forward and dive the aircraft. It would have been a survival action to avoid a collision. If this did happen, and considering the configuration and the weight of the aircraft, pulling up and rolling hard left would have caused an immediate stall. At that low altitude, recovery would not be possible and the aircraft then pan-caked onto the road way in a full stall.
I think there were two separate issues; the oil leak and the possibility of hitting a significantly sized bird. Fate brought all the forces into position to create the perfect storm that led to this tragedy.
IF the bird and the crew did manage to avoid the collision, there would be no indication of this near miss. Perhaps, not even a feather touched the aircraft in the last two seconds. IF they had landed at YXX or Langley …… IF….. We are pilots. We make hundreds of potential life altering decisions every day. Most of them are good decisions. Occasionally, FATE will intervene …. And, we have no control over that.
The TSB will eventually determine the cause of this accident. What I have described is my speculation as to what may have caused this terrible crash.
Just my thoughts.

[GyvAir]

Heard on the radio this morning that the investigation report is to be released today. Preliminary word is that the chain of events leading to the accident started with an insecure oil cap.

[av8ts]

http://www.cbc.ca/news/canada/british-c ... eport.html

[pelmet]

My experience with PT-6 oil caps were that they were an incident waiting to happen. Initially, I had nothing to do with them but then I started operating out of a different base at my company and part of the walkaround procedure was to visually check the oil level on the walkaround. It seemed to me that the oil cap could easily be put back in place and due to its design, could appear to be secure when it was not. This happened to me when I was being shown how to use it. So I was kind of paranoid after that and always triple checked that it was secure but it just seemed to be difficult in my experience.

Of course, soon after my experience with this, someone else did not secure the oil cap properly on one aircraft resulting in a precautionary shutdown.

Triple check your oil cap security on the PT-6 if you are the one securing it. Unlike some turbine engines, the PT-6 oil will leave the engine if the cap is not secure.

[Sidebar]

TSB investigation report: http://www.tsb.gc.ca/eng/rapports-repor ... 1P0149.asp

[pelmet]

Just to be clear, it has been a long time since I saw a PT-6 cap as shown in a picture in the report. But I remember that when I put the latch down my very first attempt to put the cap on, it appeared to be secure but the person showing me the walkaround indicated that it was not securely held in even though the latch was down. The cap has a fairly thin flange on it and there is a similar diameter flange on the filler neck that it attaches to. With the latch down, my experience was that if you want to pull on the cap to see if it was securely on, because the two flanges of equal size were so close to each other, I would put my fingernails around the edge of the cap flange to be able to pull on the cap to ensure it was secure.

Hope this is understandable.

[Rudy]

This is my worst nightmare. I once left home and drove back to the airport in the middle of the night because I couldn't specifically remember double checking the oil cap was secure.

[The Mole]

Ive left an oil cap ajar during my initial training. The backend of the heli was covered in oil, probably lost about half the engine oil. There was no indication or the oil loss in flight. Was an O-F*** didnt die moment after landing.

[xsbank]

I think the oil cap is a problem - however, what killed these guys was an inability to fly on one engine. All to do with the inability to fly the aircraft at the correct speed and to compensate for assymetric thrust. It was a nice day, a steady descent, a long runway and lots of power on the remaining.

If you do not have access to a sim (or even if you do!) practise your drills, your memory items, your calls, and thoroughly brief your approach and especially the missed approach (I'll bet that many of you have not recited an engine out/fire drill since your last recurrent). Unless absolutely necessary, do not do anything out of the ordinary, fly your normal approaches and follow your SOPs; do your checklists - an emergency is when you have to get things right and unless you are already on fire, DO NOT RUSH! Do things at your normal speed. Make sure your copilot is not overloaded, you are both ready and then get 'er done.

If you do not think you can handle an engine fail/fire, ask for some more training! Only you know...

[CpnCrunch]

I think the oil cap is a problem - however, what killed these guys was an inability to fly on one engine. All to do with the inability to fly the aircraft at the correct speed and to compensate for assymetric thrust. It was a nice day, a steady descent, a long runway and lots of power on the remaining.

I'm not sure that is correct, although the accident report is a bit vague in terms of what actually happened. It says the props were turning at 1900rpm during the descent and neither was feathered, so they weren't "flying on one engine" (at least, not for most of the descent). When the upset happens, it just says "When power was applied, likely only to the right engine", which I assume to mean that the pilot pushed both thrust levers forward but only one engine actually increased thrust.

It's a little unclear, but my reading between the lines is that the pilot assumed both engines were working normally (as oil pressure was normal). However at the last moment when increasing power, one of the engines didn't develop any thrust and that caused the upset (being too close to the ground to recover).

It does say "If there were an indication of declining oil pressure, it would go from normal pressure to no pressure quite quickly", so perhaps the low oil pressure checklist which says to "reduce power and continue flight" when oil pressure is 40-80 psi isn't the best thing to do.

[CFR]

I think the oil cap is a problem - however, what killed these guys was an inability to fly on one engine. All to do with the inability to fly the aircraft at the correct speed and to compensate for assymetric thrust. It was a nice day, a steady descent, a long runway and lots of power on the remaining.

I'm not sure that is correct, although the accident report is a bit vague in terms of what actually happened. It says the props were turning at 1900rpm during the descent and neither was feathered, so they weren't "flying on one engine" (at least, not for most of the descent). When the upset happens, it just says "When power was applied, likely only to the right engine", which I assume to mean that the pilot pushed both thrust levers forward but only one engine actually increased thrust.

It's a little unclear, but my reading between the lines is that the pilot assumed both engines were working normally (as oil pressure was normal). However at the last moment when increasing power, one of the engines didn't develop any thrust and that caused the upset (being too close to the ground to recover).

It does say "If there were an indication of declining oil pressure, it would go from normal pressure to no pressure quite quickly", so perhaps the low oil pressure checklist which says to "reduce power and continue flight" when oil pressure is 40-80 psi isn't the best thing to do.

You're right there's some ambiguity in how it was written, however I read it that only the right engine had the throttle pushed forward. There was no conclusion of any problems with the left engine in the report. With respect to the cause, they fell below Vref seemingly because they were unfamiliar with the need to add 10 kts in a Low Oil scenario. I wonder what the circumstances were at the time he applied power, was it a quick application of full power, or a gradual one to feel how the aircraft yawed.

In any event as with most accidents there was chain with many links. If they had checked out the oil leak ...

[boeingboy]

I talked to a few of the employees shortly after TSB held a meeting with them advising them of what they found. The pilots think that the left yaw was caused by a firewalling or very rapid advancement of the throttles. Apparently that KA 100's with these mods done to it is known to roll over to the left if the power is advanced too rapidly.

Why the captain allowed the aircraft to slow so much is not known.

[crazy_aviator]

Reminds me of the failed landing light accident , where a perfectly serviceable large plane ran out of fuel whilst the 3 man crew was fiddling with a gear lamp ,,,NOTHING to see here folks PILOT FOOLISH ERROR and the "fraternity" ( like the cops) gleefully trying to find a "reason" why such superior beings were let down by the brute beast of a plane ,,,,,,,,

[pelmet]

I talked to a few of the employees shortly after TSB held a meeting with them advising them of what they found. The pilots think that the left yaw was caused by a firewalling or very rapid advancement of the throttles. Apparently that KA 100's with these mods done to it is known to roll over to the left if the power is advanced too rapidly.

Seems like something that could happen to many different aircraft but I have never heard of mods on the aircraft making it more likely to do so. Usually the mods are by Raisbeck. We had dual aft body strakes, wing fences and "quiet" four bladed props along with a belly pod. The aircraft was quite unstable in roll and if you did not pay attention in cruise, you could find yourself well off proper heading. I saw one guy 90° off course while chatting in IMC. No autopilot of course. Good thing the airspace was uncontrolled.

[Diadem]

I watched the TSB news conference when the report was released, and the lead investigator said something along the lines of "the Raisbeck mod, with the propellor disced and the power at idle, produces so much drag that it's like having -300 ft/lbs of torque". It was actually worse to have the engine running than to shut it down and feather the prop. He also phrased the portion about advancing the power on one side a bit differently than the report; I can't remember exactly what he said, but he made it sound like the pilot had kept the left engine at idle to keep the oil pressure down and only advanced the right power lever. The combination of high torque on the right and negative torque on the left resulted in a much higher minimum control speed than published and caused a VMC roll.
Edit: Here's a source from the report regarding the 300 ft/lbs of drag: "However, the propeller manufacturer calculated the drag produced by the aircraft’s 4-bladed propeller, turning at about 1900 rpm, to be about 300 pounds." And here's one implying that the left engine was kept at idle throughout the descent without the prop feathered: "oth propellers maintained about 1900 rpm. An engine igniter was on throughout the descent [indicating a power setting below 425 ft/lbs], and continued through to the upset."

[C-FABH]

Reminds me of the failed landing light accident , where a perfectly serviceable large plane ran out of fuel whilst the 3 man crew was fiddling with a gear lamp ,,,NOTHING to see here folks PILOT FOOLISH ERROR and the "fraternity" ( like the cops) gleefully trying to find a "reason" why such superior beings were let down by the brute beast of a plane ,,,,,,,,

Good point, I couldn't help but chuckle at some of the speculation in this thread. Am glad the TSB employs a tad more logic in their investigation than the members of Avcanada.

[MUSKEG]

I marvel at the idiousy of several posts here. It was not a big bird nor the supposed instability of the Be-100 with the raisebeck mod that caused this crash. Loss of situational awareness much like the Asianic 777 accident was the direct cause. Letting the speed fall behind the curve often has a deadly outcome especially close to the ground. The effect of 1 engine at idle and prop fine and the other at some level of power should be demonstrated at altitude during training. Its an eye opening exercise. Especially on this airceaft. Just one thing more. A 90 degree heading drift is NOT the fault of any aircraft Mod. Just poor airmanship. Nothing more, nothing less. Stop the stupid excuses.

[Diadem]

The aircraft was configured in a way that didn't meet the definition of VMC, which meant the speed at which they lost control was unpublished and unpredictable. Based on what they knew and how they'd been trained, the pilots had no reason to assume they would lose control for another ten knots. Even though they were five knots below VREF they expected to have another ten knots of buffer, and began correcting once they realized they were slow. There was no reason for them to expect an application of power would cause a loss of control.

[pelmet]

I watched the TSB news conference when the report was released, and the lead investigator said something along the lines of "the Raisbeck mod, with the propellor disced and the power at idle, produces so much drag that it's like having -300 ft/lbs of torque". It was actually worse to have the engine running than to shut it down and feather the prop. He also phrased the portion about advancing the power on one side a bit differently than the report; I can't remember exactly what he said, but he made it sound like the pilot had kept the left engine at idle to keep the oil pressure down and only advanced the right power lever. The combination of high torque on the right and negative torque on the left resulted in a much higher minimum control speed than published and caused a VMC roll.
Edit: Here's a source from the report regarding the 300 ft/lbs of drag: "However, the propeller manufacturer calculated the drag produced by the aircraft’s 4-bladed propeller, turning at about 1900 rpm, to be about 300 pounds." And here's one implying that the left engine was kept at idle throughout the descent without the prop feathered: "oth propellers maintained about 1900 rpm. An engine igniter was on throughout the descent [indicating a power setting below 425 ft/lbs], and continued through to the upset."

I have not had time to read the report. However, I can tell you that leaving an engine on a turboprop unfeathered at idle instead of shutting it down is or at least potentially is bad news(in a performance limited situation such as maintaining high altitude or landing configuration) and this should not be done without careful consideration. It leads to a high drag situation that we all learned about in our multi-engine training about a windmilling propeller and all the drag associated with it. Even though the engine is at idle, there is still drag and it makes sense that if they had the four-blade Raisbeck "Quiet Turbofan" propellers installed that there would be even more drag than with the original three-bladed props.

Perhaps there is an assumption on occasion that if the engine is still running that there is not a significant amount of drag but there is. This information should be known by all pilots flying these sorts of aircraft. It is obvious if you think about it but perhaps we don't think about it and we don't normally train this specific scenario so in the heat of action, the consequences of such an action may not be fully considered. Likely, the engine was left running as it was considered to be useful and safer, but in reality based on how it was operated, it became a liability and a causal factor in the accident.

So what can we as pilots do to prevent such a scenario or other similar scenarios from happening. There are so many possible "Gotcha's" out there and we just won't be given each one of them as an example during our training.

My suggestion...read as many accident reports as possible. I have read hundreds of them and continue to do so. There are so many available on the internet nowadays from various safety boards.

Therefore....I have encountered the scenario that these King Air guys encountered. Where? In an accident report about a Saab 340 in the Netherlands in 1990 that made me realize that it is risky to just keep one engine like this at idle and better to just feather it (assuming is not needed for other reasons such as terrain clearance) and then continue on. No instructor on aircraft ever detailed the consequences of leaving an engine running unfeathered at low power on approach. We always just shut it down right away in training. But frequently, the real world is quite different than the standard training procedure that we seem to encounter over and over again.

It reminds me of that Air France situation over the Atlantic where they raised the nose so high after losing their airspeed. No one ever specifically emphasized to me in jet training how bad it is to be at such a high pitch attitude(more than 10°)like they went to at cruise altitude. Perhaps the instructors will say that it is obvious but there are thousand upon thousands of jet pilots out there and there are so many other things to study that this fact might fall by the wayside for a few when things fall apart. But, I had read about this being bad on technical forums(and AvCanada can be a great source of info) and of course it is obvious when you think about it, but it is best to consider it for the first time in your armchair. Therefore, you lose your airspeed, maybe just hold your pitch attitude or lower slightly to make sure you don't stall and then deal with the issue at hand.

Bottom line, you are on your own when it comes to really improving your knowledge about flight operations and need to read as much as you can as the training world won't and can't give you each and every possible scenario.

Flight KL433 took off from Amsterdam runway 24 at 14:19 for a 1h18min flight to Cardiff. At 14:30 the flight passed FL165, climbing to FL200 when the Master Warning was triggered by the no. 2 engine oil pressure light. The Master Warning was reset and the captain retarded the no. 2 engine power lever slowly to flight idle, causing the engine torque to decrease from 78% to 10%. Retarding the power lever is not part of the Emergency Checklist (ECL) procedure. Possibly this was done to prevent damage to the right-hand engine. The ECL Engine Oil Pressure Low procedure was then started. The ECL states that the engine must be shut down if the oil pressure warning light is on and oil pressure below 30psi. If the warning light is on and the engine oil pressure is above 30psi, normal operation should be continued. As the no. 2 engine pressure was above 50psi, both flight crew members agreed normal operation should be continued. The captain didn't want to continue to Cardiff and elected to return to Amsterdam. At 14:33 Amsterdam Radar was contacted with a PAN-call and a request to maintain FL160 and return to Amsterdam. One minute later a clearance was received to descend to FL070. KL433 was offered a straight in approach for either runway 06 or runway 01R. The captain decided to use runway 06 and was cleared to descend to 2000 feet. Landing clearance was given at 14:42 and one minute later the ILS localizer and glide slope were intercepted. The captain, not realising the consequences of flying with one engine in flight idle and was not able to anticipate correctly on the airspeed variations which resulted in an approach not stabilized in power, airspeed and pitch during the final approach. The speed had decreased to 115 knots and the crew reacted by applying an aggressive increase in torque (from 40% to 65%) with limited corrections for asymmetry. This placed the aircraft to the right of the extended centreline at an altitude of 90 feet. At 14:45:53 the captain called for a go-around and no. 1 engine torque was set at 98% with the no. 2 engine remaining at flight idle. Flaps were set at 7deg and the gear was retracted. The aircraft rolled right and pitched up to a maximum of 12deg. The stall warning system activated as the airspeed decreased and continued until impact. The Saab struck the ground with an approx. 80deg right bank and broke up.

PROBABLE CAUSES: "Inadequate use of the flight controls during an asymmetric go around resulting in loss of control.
Contributing factors: Insufficient understanding of the flight crew of the SAAB 340B engine oil system; lack of awareness of the consequences of an aircraft configuration with one engine in flight idle; poor crew resource management."

http://aviation-safety.net/database/rec ... 19940404-1

A couple of other thoughts to add.....

If you have an engine operating at idle and you need it, then use it. Perhaps mention this during the landing briefing if you decide to leave the engine running at idle. "If we encounter any control issues, I will add power on the bad engine as it is still operating, it just has a leak, that's all." It may be obvious but if it was briefed, it is more likely to be considered when suddenly there is a problem. Control is more important than wear and tear on the problem engine.

Also, if you decide to leave the bad engine at idle on approach, is there any reason to attempt to be at Vref on approach instead of carrying some extra speed when landing on a very long runway like they had in YVR.

[TA/RA]

The aircraft was configured in a way that didn't meet the definition of VMC, which meant the speed at which they lost control was unpublished and unpredictable. Based on what they knew and how they'd been trained, the pilots had no reason to assume they would lose control for another ten knots. Even though they were five knots below VREF they expected to have another ten knots of buffer, and began correcting once they realized they were slow. There was no reason for them to expect an application of power would cause a loss of control.

+1

[pelmet]

O.K., I read the report.

Not checking the puddle of oil under the engine after taxiing it to another location is a questionable action. It has been a long time since I flew a PT-6 operated aircraft but it seems to me that while some oil drips are normal on the bottom cowl(along with many other turbine engines) a puddle is something to be looked into. Unfortunately, the report does not say how big this puddle was.

The crew didn't check the oil cap on the walkaround but we never did either as it was not easy to remove the cowl or access it cap however it was done. I don't know to this day how to check the oil on a King Air(maybe I should). And I seem to remember that you had to be quite cautious about making sure that those cowls were latched properly and that this was another thing that could be easily missed(an unsecured cowl). Apparently someone in our company did so with a company exec on board. Not sure how much the cowl or cowl(s) opened on that particular flight but I bet the bigwig wasn't impressed. Seem to remember lines painted on the cowl and the fasteners to show proper alignment but somehow you could still be fooled into thinking they were secure when they were not....or so I was told.

Maybe some experience King Air guys can confirm. I was only on it for a short period of time.

[Doc]

This is my worst nightmare. I once left home and drove back to the airport in the middle of the night because I couldn't specifically remember double checking the oil cap was secure.

Shit.....I've done that.

[MUSKEG]

No you do not need to remove the cowls to check the oil. A convenient quick latch door is for that purpose. I find it odd that at least in pelmet's case a thorough walk around pointing out the location of basic service items wasn't included. Wasn't his fault. Who's doing the training and what's the company culture like that doesn't demand that? They do that for you when you first get your car from a dealership.

[Siddley Hawker]

Not checking the puddle of oil under the engine after taxiing it to another location is a questionable action.

Are there any no-go drains on the King Air? There were 5 drains on the underside of the cowling of the F-27. If there was anything dripping out of any of them, it was no-go till it was investigated/corrected.

[pelmet]

No you do not need to remove the cowls to check the oil. A convenient quick latch door is for that purpose. I find it odd that at least in pelmet's case a thorough walk around pointing out the location of basic service items wasn't included. Wasn't his fault. Who's doing the training and what's the company culture like that doesn't demand that? They do that for you when you first get your car from a dealership.

I have flown 9 different types of turbine powered aircraft in my career so far with 6 different types of engines. In all 9 cases, oil cap security was not checked on the walkaround except on one aircraft type when I was temporarily deployed to a different base where they were doing this. The only time I ever heard about an oil cap improperly secured was at this base where pilots were to check the oil level in which case, it was the co-pilot who created the problem(or maybe the company policy). That being said, doing just a check for the security of the cap instead of actually checking oil level may make sense but it doesn't seem to be a standard in the turbine world in my experience.

In reality, it is difficult to reach the oil cap on many if not most airline aircraft.