http://www.taiwannews.com.tw/etn/new...php?id=2763968

For what it's worth.

Here's a link to the press release, which contains much more detail on the cockpit conversation, as well as links to some of the published reports.


https://www.asc.gov.tw/main_en/docDe...=318&docid=701

8 seconds before impact:
Not only did they shut down the wrong engine, but they also realised that they had made a mistake.

Google translate of CVR transcipt from ASC website
 

ATC time


Summary

1051: 39


GE 235 flights took off rolling line.

1051: 43


Monitor drivers (PM) call "No ATPCS armed" control pilot (PF) answered, "Well continue the takeoff."
(ATPCS: Automatic Take-off Power Control System)

1051: 51


PM calls "Oh there Ah ATPCS armed there."

1052: 01


The aircraft took off from the ground, after the completion of post-takeoff checklist and connect the autopilot began to climb at 1,000 feet by starting right.

1052: 38


Warning aircraft cockpit noise by 1,200 feet; Procedure "ENG2 FLAME OUT AT TAKE OFF" appears on the Engine and Warning Display (EWD).

1052: 41


Aircraft climbed through 1,300 feet, PF will autopilot lifted.

1052: 43


PF said: "I put the engine back to One", PM replied: "Wait a minute cross check" this time ENG1 throttle position record income from 75 degrees to 66 degrees.

1053: 00


PM responded: "Good engine flame out check" and went on to say: "check up-trim there, auto feather there."

1053: 06


PF said: "The number one back to" after ENG1 throttle is closed to 49 degrees. Meanwhile PM said: "Okay, now is to determine the 2nd engine flameout."

1053: 09


PF replied: "Good." ENG1 throttle position remains at 49 degrees. At this height the aircraft to reach 1,630 feet, and began to decrease speed 102 knots / hour.

1053: 13


Audible stall warning sounded and a stick shakers, PM said: "Good pushing back."

1053: 15


PF said: "shut"; PM replied: "Wait a minute ... throttle throttle." 1053: 13-1053: 15:00 ENG2 the accelerator pushed to the recovery of 86 degrees to 34.5 degrees ENG1 throttle (idle position).

1053: 19


PF said: "number one" and then said: "feather shut off." Meanwhile PM said: "The number feather"; after the stall warning and stick pusher action to start 1053: 27.

1053: 22


PM said: "okay"; PF then said: "uh number one."

1053: 24


ENG1 CL (Condition Lever) is close to the fuel shutoff position; 6 seconds after ENG1 propeller shown in the feathered position.

1053: 35


PM Matsuyama tower to declare an emergency engine flameout, followed in 1053: 46-1054: 04, the flight crew had two attempts connect autopilot, unsuccessfully, at the same time during the aircraft entered a stall condition.

1054: 05


PM said: "The two sides are not" two seconds after the PM said: "There is no engine flameout both sides gone."

1054: 09


Aircraft height of 545 feet, the speed of 105 nautical miles / hour, PF said: "The car again."

1054: 20


ENG1 CL leave fuel shutoff position.

1054: 25


Aircraft height of 401 feet, the speed of 106 nautical miles / hour, ENG1 of NH1 speed increased to 30%, while the PM said: "not open."

1054: 27


Aircraft height of 309 feet, the speed of 105 nautical miles / hour, PF said: "wow throttle closing wrong."

1054: 34


Aircraft height of 83 feet, the speed of 108 nautical miles / hour, ground proximity warning system "Pull-up" warning sounded.

1054: 35


Aircraft height of 55 feet, the speed of 106 nautical miles / hour, the slope increased from the left 10 degrees to 80 degrees to the left, then the aircraft position near the viaduct on the left bank of the Keelung River, and the aircraft hit the left wing tip of a viaduct by driving the vehicle and after another crashed into a guardrail and a viaduct viaduct's edge lights, fall within the Keelung River.

In the " old days " the mantra was " Dead leg ....Dead engine "

Was he using his legs or was the autopilot in, to reduce his work load?

The article stated he took control BEFORE stating left engine failed

nowadays, more and more pilots do not know to fly the plane when things gro wrong. No surprising that some people still talk about pilotless a/c. :ugh:
If pilots can not do their job properly (meaning, they are there to manage a plan and to fly if things go wrong) no wonder why accountants want pilots be paid less etc...I don't say it's good or not, but from an outside perspective, if I was a person who would not know anything about planes and pilots, yes, I would wonder in what is the role of pilots. And I don't want to say it's only to fly from A to B when everythings go fine. It is more than that.
I feel sad and angry that more and more pilots can not recover from a stall or from an engine failure or when autothrust if off during final approach. Something we learn from PPL 6th lesson or first week of MCC. Basics.

Today the mantra for engineers is "reduce the pilots workload"
...which unfortunately also means "reduce the information the pilot receives"
In IMC that might even be a good idea, as pilots may not immediately understand that they have to "kick the ball" since they are so used to have a screen showing the horizon, but not the yaw motion... Also the simulator can not give you the real sensation of the yaw due to an engine failure.

Sometimes the automatics save your day, sometimes they kill you or at least mislead you, so that you kill yourself. There is no easy solution to komplex problems. There are only skilled and well trained people solving problems. The more basic an aircraft behaves, the easier to understand the situation. But sometimes the harder to deal with it.
In the past pilots failed to do the right thing, today they typically fail already to understand the situation.

"In IMC that might even be a good idea, as pilots may not immediately understand that they have to "kick the ball" since they are so used to have a screen showing the horizon, but not the yaw motion... Also the simulator can not give you the real sensation of the yaw due to an engine failure."

As much as training has changed, the focus moving toward using the automation and away from flying the airplane, I find it incredible to think that pilots are not made quite familiar with engine out procedures.

Nor does it surprise me entirely that the pilots extinguished the wrong engine - but it SHOULD surprise me. The British Midland accident in 1989 is another example, I suppose there are others. But, it anything comes out of this, it should be an emphasis in training. There is just no excuse for letting this go on.

Oh, there's loads of excuses.
Crap pilots are cheaper.
They don't complain because their jobs are at risk anyway, so you can treat them like dirt.
The planes are safer.
Sufficient of the travelling public don't care about anything except the ticket price.
Governments don't want to rock the economic boat by regulating businesses properly.

From a news source...
 

The ASC's report also showed that Captain Liao Jian-zong had failed simulator training in May 2014, in part because he had insufficient knowledge of how to deal with an engine flame-out on take-off.

"Wow, pulled back the wrong side throttle," Liao, 41, was heard to say on voice recordings seconds before the crash.

There appeared to be confusion in the cockpit as the two captains tried to regain control of the plane after one engine lost power about three minutes into the doomed flight.

Liao reduced the throttle on the working engine but did not appear to realize his mistake until it was too late.

He tried to restart the engines several times before a junior first officer in the cockpit said: "Impact, impact, brace for impact."

Those were the chilling last words heard on the data recordings, according to the latest report of the ASC's investigation into the Feb. 4 crash.

Seconds later the almost new ATR 72-600, which had 58 people on board, crashed upside down into a shallow river in Taipei after it lurched between buildings, clipping an overpass and a taxi.

Fifteen people survived but all three pilots and 40 passengers and other crew died in the second crash involving a TransAsia ATR plane in a year.

A source with direct knowledge of the report told Reuters on Wednesday the working engine had been shut off.

FAILED SIMULATOR TRAINING

The ASC report, which neither assigns responsibility nor suggests recommendations, paints a more detailed picture than a preliminary report released days after the crash.

Liao, a former air force pilot, began flying commercial aircraft in 2009 and joined TransAsia the following year. He was promoted to captain in August 2014 and joined the ATR 72-600 fleet in November.

He had a total of 4,914 flight hours on ATR 72 planes.

However, the report showed that Liao failed the simulator check in May 2014 when he was being evaluated for promotion. Assessors found he had a tendency not to complete procedures and checks, and his "cockpit management and flight planning" were also found wanting.

However, he passed after a second simulator check on June 29 and 30 and was promoted to captain, although similar problems were detected during training from July 2-10 last year.

Instructors commented that he was "prone to be nervous and may make oral errors during the engine start procedure" and displayed a "lack of confidence", the report shows.

Issues cropped up again during training for the ATR 72-600 in November, when an instructor said Liao "may need extra training" when dealing with an engine failure after take-off.

After the crash, Taiwan's Civil Aeronautics Administration put TransAsia's ATR pilots through oral proficiency tests on how to handle an aircraft during engine failure.

All but one of the pilots passed the tests, although some needed more than one attempt. The lone failure was demoted in rank to vice captain from captain.

I think most multi-engine flight instructors will recognize that starting to shut down the wrong engine is one of the most common mistakes in multi-engine training.

Tscott that is exactly right. I trained multi conversion guys for 10 years and it was hands down the most dangerous kind of flying I've ever done because people would simply panic and use incorrect rudder or close the wrong throttle.
I would have thought with auto-feather on the ATR you would be better off doing nothing and focus on target speed and climbing.

There needs to be a fundamental change in the basic handling of EFATO/V1 cut.
1:fly the a/c.
2:400ftagl,announce the failure.
3:1000ft agl,crew cross confirm failure then STOP!
REVIEW!!
Action check list.
Burning holes or wet cockpits happen due to stupid errors mostly via rushed event managing. The key is to slow down and confirm with review.
This whole regulatory sim scenario for IR/PC renewals..base checks hasn't changed since 1975!!!!!!

This is not GA & it certainly isn't ab-initio twin training. This is airline flying & all pilots should have these issues well & truly sorted out before they come anywhere near the flight deck of an airliner.

The old adage that you get what you pay for may well hold true in this tragedy, but it shouldn't be accepted by anybody when people's lives are at stake. It is not a restaurant meal or a piece of clothing that we are talking about here, but it would seem that human life is cheap & getting cheaper all the time. Profit is all many care about.

soon I take my flight for vacation.

When I see this kind of accident, I just pray that automatism and plane work very well, and the pilots don't have to take controls (except for t/o and landing, oh well).
Not even a touch with one finger. :rolleyes::rolleyes:

As a pilot, I am just worried now about "my colleagues" (if they slept, if pay to fly or not, if a next lubitz or not etc).
Funny that since I work in this field, I feel less and less confortable as a passenger myself than before !

After the crash, Taiwan's Civil Aeronautics Administration put TransAsia's ATR pilots through oral proficiency tests on how to handle an aircraft during engine failure.

All but one of the pilots passed the tests, although some needed more than one attempt. The lone failure was demoted in rank to vice captain from captain.


This sounds a little off to me. Surely you don't get to be even a 'vice captain' if you can't handle a flame out situation?


Is there a desperate shortage of pilots in that part of the world? (Forgive me for asking)

Well, airfares for any typical flight are less than a quarter pounder with cheese and a milkshake. Or any medium-sized T-shirt with starbucks on it. So don't expect too much in return.

http://puu.sh/iLTL9/0f69c605c6.png

I don't understand this exchange. What is the flameout check and is that where the confusion between the two engines started? Or was it an incomplete check?

On a turboprop, it is not that simple. You need to make sure that the dead propeller is feathered, or else You will not get the required performance and might not even reach the 400ft AGL in Your suggestion. In fact, if You have not flown a turboprop in real life, I invite You to try a ATR or DH8 sim when You find the opportunity. Then You will see the difference: a failed jet simply stops pushing while a failed prop starts braking. Heavily. So a V1 cut will positively require a quick (but not hasty) analysis of the situation immediately after gear retraction. If the propeller is feathered and no fire is indicated, one can and indeed will wait until acceleration altitude to attack the problem. In case of a non-feathered prop and/or an engine fire, there will follow a by-heart shutdown of the engine and if needed the activation of the fire suppression systems (with the relevant selections confirmed and verified by the PF, but performed by PNF). After acceleration, it will be QRH time and all selections will be looked over once again.

This procedure comes with workload much higher than in a jet facing the same situation. It is not only strictly required due to performance reasons, but has been performed in real life many times without any bent metal resulting.

However, I strongly doubt that the proceedings in the TransAsia ATRs flight deck were even close to what any book says. The best SOP will not help anyone if it is not followed.

As one who pays to sit in the back, this does not fill me with confidence.

As a matter of fact, what is a "vice captain"?

I kind of doubt that this is an intermediate flight crew rank. Could it be that this is what Googoo Translate made of the Chinese word for a First Officer?

You will get even less performance by shutting down the live engine! Even with an unfeathered prop the aircraft climbs better than without any working engine.

Thus i completely agree that you should take your time to properly analyse the situation before rushing to action.

The turboprop myth of having to feather in a heroic blink should not be upheld religiously by proud prop drivers, it misleads others.

By the way, if you have not flown a jet: Even a jet engine has quite some braking action when severely damaged, especially on the big twins ....

NTS or Autofeather systems work far more reliably than pilots under stress. Even in an engine fire scenario the biggest risk is arguably the pilot rather than the fire itself. And in the real world what are the chances that performance is really critical?
Sure, pilots need to know how to handle the worst case but change the emphassis to leave the damn thing alone until you're good and ready to attend to it properly.
The prevailing training system is not based on sound risk vs. reward analysis but ego driven, outdated belt and braces dogma rooted in piston era instruction techniques and empire building.

About time it was taken outside and given a good kick up the arse.

TransAsia in the water - British Midland B737 into ground.
 

It is not only TransAsia pilots, or turboprop pilots generally, who mistakenly shut down the wrong engine. BM Flt92 accident at Kegworth also involved the crew closing down the wrong engine in a B737-400.

Oakape, I think the point of my pointing out that starting to shutdown the wrong engine is perhaps the most common mistake in twin-engine training, was not to suggest we all just have to accept it. The point I was hinting at was that it suggests this pilot was not proficient or had not mastered operation of his aircraft. Asian airlines are expanding at fast rates and it's not impossible for a fast-growing airline, in a fast-growing region, to not train as much as others or hold on to less than stellar pilots once hired. Reliance on automation and turbine reliability can cover for a lot of underlying pilot deficiency. We don't have info if this accident pilot ever had a genuine engine failure outside of a sim. Of course he had the license, and the training, and should have been proficient at the procedure. In every field, at all levels of proficiency, there are common mistakes. Decade after decade in field after field many of these items are persistent. It never helps to have fewer pilots than you want, or access to less-experienced pilots than you wish. Nobody suggests we should accept this.

The accident pilots may have been the best in the fleet/region/world and just had a bad day. I'm sure the accident pilots were much better pilots than I am. My background is in training and my initial thought upon seeing the video of the last moments of this crash were that it looks like a classic twin-engine flip after speed decay below Vmca. It turns out, as it often does, to have been crew screwup. The transcripts turn out to show it was a screwup AND not following procedures. Nearly all pilot screwups could be prevented by following procedures, which would catch screwups if followed. If the accident pilot had waited to pull the throttle/power lever back until the PNF confirmed which engine was failing and which throttle was to be pulled this accident would not have happened. Nobody is suggesting anything is acceptable. :ugh:

The Capt. had almost 5000 hrs on type.
This is no peanuts. I would expect such mistakes to be made by noobs and trainees, not Captains with this much time.
Bad training or selection process?

The LH seat pilot came from the military where I am guessing he flew single engine jets, then transitioned to civilian flying in 2009. 5000 hours on type is just doing the same routine hour 5000 times. In a modern turbine transport* most pilots will never see an engine shut down, so the only practice people get is in their sim rides...so after training you see a failure what? twice a year?

Maybe it was a bit of both: weak candidate/mediocre training. Its not like there is a vast pool of qualified candidates to choose from in that part of the world, and training programmes are only as good as the most knowledgeable and experienced pilot WHO IS ALSO CONNECTED POLITICALLY TO SENIOR DECISION MAKERS, as distinct from the ideal.

*contrast that to the big pistons which failed often...the world Constellation fleet for example had >1 shut down every day, and that was fewer than 500 aircraft in daily use.

Question for the ATR experts: Does the yaw damper agressively compensate for engine loss? What indications of yaw are displayed on the EFIS, and how prominent is such indication, if any? Is torque displayed for each engine? Or any other reliable power indications beyond just N1, N2 and temperatures?

Sitting in the cockpit hours is a lot different from 'flying' hours. I often wonder how many of the thousands of hours credited to individual pilots are ones actually flying an aircraft?

funfly,

Then it could me more interesting to note how many emergencies they have handled.

Rwy in Sight

I don't think it comes down to how many hours or experience someone has, to me someone flying 10 hours across the pond for 20 years is not really flying experience, its system management experience

End of the day these chaps unfortunately made an error from an event that probably took them by surprise, i don't think any experience in the world can stop you making a mistake, its all about training and procedures and it seems Transasia's is not up to the mark.

Sometimes in a TP EFATO there is not too much time to sit and look at things initially there is quite a bit that has to be checked, my Type will not achieve a climb at max weight with an Auto Feather failure so the prop has to be feathered fairly quickly sometimes you cannot get a confirmation from the other guy as he is maxed out controlling the yaw

There's a moment where you look down at the CL and just have a 3 second mental check yourself...easy to C*ck that up

Shutting down the wrong engine is not a unique event. I would have thought that in the event of an engine fie there could be a bloody big hand pointing to the lever saying "it's this one - shut me down"

No Funfly even if you mounted one as big as the one Kenny Everrett, may he rest in peace, used on his shows it would not work. If it has to be some kind of human appendage, much better to use big feet, the bigger the better.
Surely you must remember the drill for twins, " dead leg dead engine ". Is this no longer valid, has it been replaced with chimes, gongs, bells and whistles. If it has to be so, then what about an ambulance siren.

Things have changed sisnce then.

Firstly, we had a flight engineer on the DC7, whose primary job was looking after the engines. And flight engineers were a special breed, manufactured in a biological facility in Arizona. It was physically impossible to get them to break into a sweat, because it was not in their programming. If an engine was on fire, they would go back and check, rather doing something as stupidly rash as wrongly loosing 25% of our power, when we could only climb at 500ft/min on four.

Secondly, when something is an everyday event, you deal with it as such. Another day at the office. Another engine symbol in the logbook, to go alongside all the others. The real surprise was when you got to 10,000 ft with all four still running. That was a cause for much celebration - coffee with a booster and high fives from the loadmaster.

The only reason they built the Connie and -7 with four engines, is because there was no room on the wing for six. Modern engineers and accountants have forgotten that truism, and seem intent on hanging 400 tonnes on just one live engine.

A good idea? I think not.

Again, most avaition logical thinking has been thrown out of the window by the accountants. EFATOs don't appear on the bottom line, or the shareholder's report, so it is a trivial matter of little importance.

In the old days they knew such things were important, and so the 737 had the fire warnings and handles on the coaming, in the central field of view. But then they started adding MCP automation at the cheapest price possible, which meant no major redesigns. The result was the fire warnings and handles were moved to the aft center console, underneath the fuel cocks, where it is all but impossible for the PF to see what the hell the PNF is doing.

Was this a good idea? No, it was cheap. That's what the passengers want, and that is what the passengers have got.

I significant issue is still being overlooked. Why was a global restriction (quietly) issued in December 2104 on 'Engine Out Training, Testing and Checking associated with DC Gen fault on remaining engine' on 72-600 FSTDs?, subsequently (quietly) rescinded in April 2015!

Early in this thread there was a FDR sequence posted. If that one indeed was for this flight, it shows the event on engine #2 about 30 seconds after the Ground/Air switched to Air. It shows over 1000 feet at the time of the initial event. This event happened well after rotation and lift off. It has yet to be suggested as to the cause of #2 feathering and sitting there with an 80% NH. This was not an Engine Flameout at Take Off. Indeed it appears it was not a Flameout at all. It also does not show the spiked torque that accompanies a uncommanded PCU feathering, instead it looks like any ATPCS triggered event. But why?

...people died far more frequently and pilots opposed the introduction of cockpit voice recorders. Things have changed for the better since then.

Too true.

As well as flying hours, I keep a record of how many landings* I've done, which as you imply, does at least give an idea of how much actual manual handling I've had.

Having said that, don't discount the planning work one does while sitting in the cockpit: to avoid CB's and considering escape routes from high ground or oceans etc. in the event of emergencies.



*Obviously the number of landings should be the same as the number of take-offs, but owing to the occasional bounced landing, I probably have slightly more landings than take-offs !!:rolleyes:

Military Pilots only flying single engine jets.
 

Australopithecus, not all Military pilots fly only fast pointy nosed aircraft. As a former Military and Civil Flight Engineer I flew with many Military and ex-Military pilots in both turboprop and turbojet aircraft. Some had flown only fast jets prior to joining the airlines. None of those with whom I flew ever gave me cause to question their competence or their professionalism. We certainly saw many more than "twice per year" engine failures in the Simulator. Engine shutdowns were more common on the turboprop aircraft but not exclusively so. Well trained professional crews, regardless of whether they are ex-military or only ever civil, will generally have the capacity to handle any emergency which is able to be resolved. "Hasten slowly" has always been the norm in any outfit with which I have been associated.