One Of The WORST Reports I’ve EVER Read! | Thai International Airways 261

It's a stormy wet [music] night in the

city of Surat Thani in southern

Thailand.

And above the city's main airport, an

Airbus A310 is hanging in the sky like a

cross just about to [music] fall down.

Now, there's no technical reason for

this to be happening. So,

what has led up to this?

On December 11th, 1998, a Thai Airways

crew arrived at Don Mueang International

Airport in Bangkok to operate flight

261, which was a short domestic hop over

towards Surat Thani.

The aircraft they had been scheduled to

fly was a wide-body Airbus A310, which

is likely the smallest wide-body ever

made, but it was still more than large

enough for this 1-hour domestic flight

with only 132 scheduled passengers.

It's not clear why exactly this A310 was

used when one of Thai Airways' Boeing

737-400s would have been better suited,

but maybe it was just what was available

on that night.

In any case, with that low passenger

load, the A310 would only be little over

half full, but it still carried a full

complement of 12 cabin crew, which

together with the two pilots in the

front brought the total number of people

on board up to 146.

Now, the ages of the flight crew weren't

specified in the final report, but we do

know that the captain had been flying

for Thai Airways for many years at this

point and had 10,167

total flying hours.

And that included over 3,000 on the

Airbus A310 and its larger cousin, the

A300, which have a common type rating.

The first officer was less experienced,

but wasn't new either. He had 2,839

total hours, so which almost all had

been flown on this particular type.

Both pilots had had a good rest before

the flight with more than 20 hours of

duty.

Although it is worth noting here that

the captain had logged more than 108

hours during the preceding 30 days,

which would have been close to or maybe

even over modern duty time limits. So,

he had been working a [music] lot.

Now, the final report doesn't mention

any duty time violations here. So, most

likely Thailand's duty time requirements

back in 1998 was

not as stringent. But, obviously, we

have strict limits now because working

more hours could cause chronic fatigue,

and that is worth keeping in mind.

Anyhow, before walking out to the

aircraft, the pilots reviewed their

dispatch paperwork, which included the

weather forecast, notices to airmen or

NOTAMs, as well as the aircraft

technical log.

And from there, they could see that the

aircraft was in good working order with

no defects that would become relevant to

this story.

But, the same could not be said about

their destination airport, Surat Thani.

It had been undergoing renovations for

the last [music] couple of years with

the pace increasing in late 1998. And

Thailand's Department of Aviation had

therefore issued several NOTAMs

highlighting several systems that were

inoperative.

First of all,

>> [music]

>> the instrument landing system, the ILS,

which provides precision approach

guidance, had been out of service since

January of 1997. And in September of

'98, the approach light system had also

been [music] turned off.

Then, in November, the precision

approach path indicator [music] or PAPI

lights, which tell us whether we are

approaching too high or too low, also

stopped working.

And whilst the PAPI lights on the

left-hand side of the runway was soon

restored,

the right side wasn't. [music]

Then, the non-directional beacon that

was located at the airport, which was

used as a backup navigation aid, was

[music] being moved, so it didn't work,

either. And finally, the airport had

shut down one of its two redundant

electrical circuits powering the runway

edge and threshold lighting. So, as a

result of that, only every other [music]

light was illuminated.

A specific NOTAM about that reported

that the spacing between the runway edge

lights were therefore increased to 120

instead of 60 m [music]

and 6 m instead of 3 m for the threshold

lights.

Now, all of this was clearly some pretty

serious degradations of Surat Thani's

[music] capabilities when it came to

night or low visibility operations. And

in fact, the airport no longer met the

minimum equipment requirements for

>> [music]

>> low vis operations under the rules of

the International Civil Aviation

Organization, ICAO.

You see, ICAO sets out very clear rules

for what type of equipment can be

missing or downgraded depending on what

type of approach [music] is going to be

flown. So, if several items are missing,

the [music] runway might only be allowed

to be used in very good visibility or

with a high cloud base.

Also, for runways that are used in

[music] low visibility, ICAO imposes a

maximum spacing of 60 m between runway

edge lights in order to make the runway

as distinctive as [music] possible,

which obviously was not being met here.

So,

with all of this, you might be wondering

whether it was even allowed to dispatch

a flight towards this airport at night,

and the answer is actually

yeah.

Like I mentioned before, [music] we

pilots are supposed to increase the

minimum visibility required for an

approach depending on [music] which

lighting systems are out of service or

working, but even if the forecasted

visibility [music] doesn't meet those

new requirements, we can still dispatch

towards that airport as long as we have

enough fuel to reach at least two

alternate airports, which have better

weather and equipment enough to satisfy

all of the rules.

As a result, these pilots [music] most

likely read those NOTAMs and the weather

forecast and concluded that they would

definitely need to take these [music]

precautions, but if worse came to worse,

they could just divert.

The forecast for Surat Thani on that day

showed that in general the conditions

were pretty good with a visibility of 7

km and intermittent rain showers

scattered to broken clouds at 1,800 ft

[music] and an overcast layer at 11,000

ft.

But,

it did also warn that intermittent

storms could pass over the airport

temporarily reducing the visibility down

to 1,500 [music]

m in rain with scattered clouds at 1,200

ft and broken at 2,000, as well as with

variable winds at 15 kn gusting to 25.

Now, if flight 261 were to encounter

those conditions,

well, then landing at the airport with

its degraded lighting and without an ILS

could be

difficult to even impossible.

But, since those conditions were

expected to be just intermittent, the

pilots probably expected that they could

just

>> [music]

>> wait for the visibility to improve.

In any case, due to this, they now

loaded the aircraft up with a full 15.7

tons of fuel, which was enough to fly to

their destination, make several

approaches, then hold for a bit, and

then finally fly all the way back to

Bangkok again if needed.

And that extra fuel brought the total

takeoff weight up to about 108 tons,

which was still far below the A310's

maximum takeoff weight of 142 tons,

meaning that they would be fairly light.

And whilst that might seem like a good

thing, in this particular story, it

would come with some truly nasty

consequences.

And talking of which, did you know that

scammers can now use AI to clone voices?

Well, what they do is first, they buy

your phone number from data brokers, and

then they make a robo call to you to

sample your voice, which is then cloned

using AI.

After that, they can use that voice

clone to call up your family and friends

pretending to be you in an emergency

asking for money. And since that will

sound exactly like you,

it often works.

In my family, we have actually adopted a

code word for this exact reason, but

another thing that I highly recommend is

to use today's sponsor, Incogni.

As soon as you sign up with them, they

will start scanning all data brokers out

there for your data. And when they find

it, they will send out an instruction to

delete your data from their services.

Incogni will then handle all paperwork

needed, follow up on objections, and so

on until all of your data has been

removed. [music] And they will keep

searching for your data out there for as

long as you stay with them.

Now, you could theoretically do all of

this yourself, but it would take up a

lot of your time because there are just

so many data [music] brokers out there.

So, I would highly suggest to use

Incogni instead, which is what I did.

Now, after you've signed up, you can

then track their progress in real time

through their dashboard, which is

awesome. Plus, with their family and

friends plan, you can add up to four

loved ones to your plan, which is a

really nice touch.

So, to get started and get an awesome

60% discount, click on this QR code or

go to incogni.com/pilot.

Just remember to use the code pilot at

checkout.

Thank you, Incogni. Now, let's get back

to the video.

After having completed all of the

pre-flight checks, flight 261 pushed

back, taxied out to the runway, and

finally took off at time 17:54,

climbing into the darkening evening with

the captain at the controls as pilot

flying and the first officer handling

the radios.

The climb to their cruise altitude of

31,000 ft was then completed uneventful

and they proceeded south for about half

an hour without any issues.

Then at 18:26, when they were about

halfway through the flight, the first

officer called up Surat Thani approach

and was told to expect a VOR DME

approach to runway 22.

Now the pilots accepted this, set up and

briefed for that approach, and the

controller then began to give them radar

vectors and descent clearances down

towards that approach.

But before we get to that, we need to

talk a little bit about the specific

procedures that were involved in this

approach that they were soon about to

fly.

Or actually procedures, because in a

very strange twist, there were actually

two of them.

A VOR DME approach is a non-precision

approach, which means that while it does

provide pilots with lateral guidance to

align with the runway, it doesn't

provide any vertical guidance to help

them descend on the correct slope.

In order to fly this type of approach,

the pilots must first tune in to a VOR

radio beacon, or very high frequency

omnidirectional range, located at or

near the airport. And this VOR will then

emit modulated signals that allow the

aircraft's VOR receivers to determine

the exact bearing from the VOR to the

plane.

Similar to the spokes of a bicycle

wheel, divided into 360° of so-called

radials.

During the VOR approach, the pilots will

then fly inbound towards the VOR on a

specific radial, which will be lining up

close to the runway heading, whilst also

descending to predetermined altitudes to

certain distances from the runway.

And those distances are monitored by

tuning in to a distance measuring

equipment, or DME, which is usually

co-located with VOR, hence the name VOR

DME.

Now in the case of runway 22 at Surat

Thani, the VOR approach was a bit

unusual because the approach course was

actually offset from the runway

direction by quite a large amount.

That could have been for many reasons,

but in this case, the VOR wasn't located

on the extended centerline, but instead

on the left side of the airport seen

from the runway 22 direction. And that

meant that flying straight towards it on

the runway heading of 225 degrees would

cause the plane to fly on the left side

of the actual final approach track

throughout the entire approach. So,

instead, this VOR DME approach was

designed with an approach course that

was offset in a way that meant that the

plane would start off too far to the

right of the centerline before crossing

the extended runway centerline close to

the threshold. So, that if the aircraft

would be visual at that point, the

pilots could just make a slight right

turn and line up with the runway

centerline just before landing.

Now, the Thai Department of Aviation had

produced a VOR DME approach chart with

an inbound course of 204 degrees, which

would mean a very large 21-degree

offset. And if that course was followed,

it would cause the aircraft to cross the

extended runway centerline at a distance

of 1,800 m from the threshold,

which was pretty okay.

And that chart was available in the

cockpit of flight 261,

but it wasn't the only chart they had.

You see, Thai Airways had also produced

its own chart for the runway 22 VOR

approach, and it contained a number of

key differences.

For example, the two charts had

different missed approach points where

the pilots must go around if they

couldn't see the runway. They also

showed different altitudes that the

pilots should climb to in case of a

go-around, and the Department of

Aviation's chart strangely didn't

specify a minimum visibility, whilst the

Thai Airways chart imposed a minimum of

2,000 m.

But finally, the Thai Airways chart also

specified an approach course of 215°

instead of the 204° shown on the DOA

chart. That meant that it was much

closer to the runway heading of 225° and

therefore would require less of a turn

to align with the runway.

But that also meant that the approach

course would cross the runway extended

center line further away from the

threshold at a distance of 4,800 m to be

exact.

Which in practical terms meant that a

higher visibility was needed to fly the

Thai Airways variant of the approach

since the point where the pilots would

have to make the

>> [music]

>> right turn to visually align with the

runway was further away.

But like I mentioned before, the minimum

visibility printed on the Thai Airways

chart was only 2,000 m. So, if the

pilots caught sight of the runway at

that distance, they would have already

passed the center line to the left

making it difficult to maneuver in for

landing.

In reality, the practical minimum

visibility for this approach was much

higher than stated, but that wasn't

mentioned anywhere.

Maybe you're starting to see the

problem?

Back on board flight 261, the

controllers vectors had soon brought

them onto the approach course and by the

time the cockpit voice recorder started

at time [music] 18:37,

they were already established towards

the VOR.

The two pilots were using the Thai

Airways approach chart with the inbound

course of 215°, which is the one where

the visual turn towards the runway had

to be made about 4,800 m away. But when

the pilots contacted the tower, they

were told that the latest weather report

gave winds calm, visibility 1,500 m in

light rain, and a cloud base at 1,800

ft. Which was obviously much less than

the minimum visibility of 2,000 m

depicted on the approach chart, even

before taking into account the effects

of all of those inoperative lights at

the airport.

Now, in that situation, in most

countries, we have something known as an

approach ban, which states that it is

not legal to continue an approach beyond

the final approach fix or equivalent

position unless the reported visibility

would improve above the minimums.

And we have that because if the

visibility is below the minimum, the

chances of a successful landing are just

very, very low. So, there's really no

reason to put yourself in a situation

where you're close to the ground in

instrument conditions, just [music]

praying for the runway to miraculously

appear in front of you.

This approach ban actually gives crews

the tools needed to just say no at an

early stage, to climb back out, and then

hold until the visibility either

improves or a diversion is initiated.

But in this case, the pilots decided to

attempt the approach anyway. And the

final report doesn't really explain why,

nor points to this as a serious

violation.

It is possible that there was confusion

about the minimum visibility because the

DOA chart didn't specify one.

Maybe

>> [music]

>> it was just normal at Thai Airways to

ignore their own company minimums, or

they thought that the conditions might

improve before they reached the runway

since the forecast said that those

periods of low visibility should be

temporary.

We

>> [music]

>> just don't really know.

In any case, after passing the

intermediate approach fix, 11 mi out

from the VOR, the tower controller

reminded the pilots that the right side

papi lights were out of service, and

that the spacing of the runway edge and

threshold lights was increased. [music]

The pilots acknowledged this and then

continued the landing checklist,

lowering the landing gear, extending the

flaps to full, and notifying the cabin

crew to take their positions for

landing.

In front of them, their landing lights

now illuminated only clouds [music] and

rain, while the windscreen wipers

swished forth and back trying to keep

the windows clear.

At time 18:41, the first officer

reported that they had reached the final

approach fix, located 6 nautical miles

from the VOR, which, like I mentioned,

is the furthest that you can normally

legally go if the visibility is below

the minimum.

But instead of abandoning the approach

here, the pilots just continued their

descent down to the minimum descent

altitude or MDA of 430 ft, as if this

was the most normal thing in the world.

Now today, this would be a quite serious

violation, which would definitely make

you fail a line check if you were doing

one, but

it's possible that the rules were

different in Thailand back then, or that

these kind of things have been

normalized at Thai Airways for some

reason.

The controller soon reported that the

aircraft was not in sight, but cleared

them to land anyway, along with giving

them a new reported wind of 310° at 5

knots.

The first officer read back that landing

clearance, and the tower then cautioned

them that the runway was wet.

After this, the first officer quickly

verified the landing checklist had been

completed, and a few moments later, the

tower controller asked again whether

they had the runway in sight.

Around this time, the dim lights of the

runway actually did appear through the

mist. So, the first officer now

confirmed, "Runway in sight."

And that was then followed by the

captain disconnecting the autopilot and

autothrottle, and beginning to [music]

steer towards the runway heading.

But now, the first officer noticed that

they weren't properly positioned to

land, so he called out, "Pull up! 2,000

set, sir." Referring to the go-around

altitude of 2,000 ft.

It's likely because of the combination

of the offset approach and the low

visibility, they had simply spotted the

runway too late. So, making that visual

right turn to align properly with it was

now not possible, which

to be honest, was quite predictable.

But, in spite of this, the captain now

asked, "Is that the right one?" And the

first officer repeated, "Right, right,

right hand side."

"Where?" said the captain, to which the

first officer repeated himself again,

and at that point, the captain just

uttered a curse and said, "Okay, okay,

that's right. [music] Go around."

He initiated the go-around procedure by

slowly increasing thrust towards the

[music] takeoff go-around or toga

setting, allowing the pitch to gently

increase towards the go-around attitude.

>> [music]

>> Now, the procedure for a go-around on

the Airbus A310 was to increase thrust

to toga, hold the pitch angle at the

maximum of 18° [music] pitch up, and

then retract the landing gear, and then

flaps before climbing to the go-around

altitude, which in this case was 2,000

ft.

But, one counterintuitive aspect of this

procedure was that, in order to maintain

a pitch of 18°, pilots might actually

have to push forward on their controls

instead of pulling back. And that was

because aircraft with engines mounted

[music] below the center of mass, like

the A310 had, will experience [music] a

pitch up moment when the engine thrust

is increased, and that's especially

noticeable when the aircraft is light,

like it was in this case.

Now, if you imagine a 2D aircraft pinned

to a wall with a pin through the center

of mass, if you then grab the plane

below that pin and push it forward, it

will then rotate about the pin, bringing

the nose up, right?

Well, that is essentially what happens,

and these aircraft will do the same

thing, but when the autopilot is

engaged, it will counteract this pitch

up moment by applying nose down elevator

as needed to prevent the pitch from

exceeding 18°. [music]

But, if the autopilot is off,

well, then the pilot will need to adjust

that pitch manually. And in some cases,

that might require some quite quick and

assertive control inputs to prevent the

pitch from getting too high.

In this case though, the captain was

fairly conservative with his thrust

increase, taking a full 23 seconds to

reach the toga setting. [music]

Then

even though that was the case, he still

didn't use enough pitch down inputs

here. So, the pitch initially increased

to 24° before he was able to get it back

down to 18.

Now, just for context, we consider an

unintentional pitch up of 25° or more to

be an in-flight upset.

So, this wasn't [music] great.

Nevertheless, the captain kept the plane

under control, and the pilots executed

the rest of the missed approach

procedure without any real [music]

issues, climbing back to 2,000 ft and

then requesting a return back to the

final approach fix, which the tower

controller quickly granted them.

After having completed the initial turn,

the captain then clarified to his first

officer that he would now do a teardrop

course reversal and then attempt the

approach again. And he then added, "Too

close. We were off to the left, weren't

we?"

On top of that, he also then mentioned

that there had been a high rate of climb

during the go-around, to which the first

officer correctly replied,

"I it might be because we are very

light."

But there was no further discussion of

this or any kind [music] of internal

debrief on how to avoid it the next

time, nor any discussion about the poor

weather at this point. So, the pilots

just executed that teardrop maneuver and

went straight back into a second

approach.

Now, at this point, it would have made a

lot of sense to instead enter into

holding and wait for the visibility to

improve. After all, they had plenty of

fuel with them, but neither pilot seemed

to even wanting to suggest that.

Instead, the first officer just

obliquely asked,

"Do you want me to ask if it's raining

at the airfield or not?" To which the

captain told him to ask for the

visibility instead, which makes sense.

But the first officer just went ahead

and asked whether it was raining anyway.

To which he was told by the controller

that, "Yeah, there's some light rain

over the field."

The investigators noted that the first

officer might not have actually

correctly heard the captain's

instruction to ask for the visibility

instead because the captain seemed to be

very [music] soft-spoken. But if that

was the case, the captain could have

easily just repeated [music]

his requests.

But that did not happen here.

Which possibly points to a bit of lack

of

crew resource management, which is

something that this crew would

definitely come to need.

Once again, the pilots now lined up for

the VOR approach after having made a

quick PA to the passengers about not

having been able to land and that they

would try again.

They then assumed the approach course of

215 degrees, completed the landing

checklist again, and began descending.

At time 18:52, the controller again

cleared flight 261 to land. And just 1

minute later, the captain could be heard

complaining about the offset approach

course, noting that it was difficult to

land like this.

And the first officer agreed.

In the background, the cockpit voice

recorder captured the sound of the

windscreen wipers running at high

setting. And as we could hear that the

tower controller also came in and asked

for their intentions.

The first officer responded that the

runway still was not in sight at that

point. And 5 seconds later, the captain

asked if the runway was off to the

right. To which the first officer said

he wasn't sure. He could only see

intermittent red lights.

The captain now added that they were too

far left of the runway. And asked

whether they had passed it already,

suggesting that they should go around.

So the first officer again confirmed

that the go-around altitude of 2,000 ft

was set and the captain again used the

go-around switches on the thrust levers.

But, this time this put the autopilot

and autothrottle into go-around mode for

an automatic go-around since he had

never disconnected the automatics on

this approach.

The autothrottle now quickly spooled the

engines up to TOGA thrust in about 8

seconds. Well, the autopilot applied

nose-down elevator [music] to keep the

pitch from rising above 18° in a smooth

and controlled way.

The aircraft then climbed to 2,000 ft

again and executed the missed [music]

approach checklist. And after that, the

captain said that this time he would

bring them all the way out to the

intermediate approach fix at 11 mi from

the VOR before coming back in.

Maybe he wanted to do that in an effort

to slow things down a bit and allow for

the weather to improve, but it was never

really discussed.

He also continued to complain about the

fact that the approach was not in line

with the runway. And the first officer

explained that he had seen some lights

earlier, but with so many runway lights

out of service, he wasn't sure whether

it was the runway or a fishing boat that

he had seen and he didn't want to

mislead the captain.

>> [music]

>> And to be honest, if there is a doubt

between those two things, that was

probably a good idea.

Now, at this point, the pilots also had

to make a choice about how they were

going to proceed and if they should even

try for an approach a third time, which

in my view should have been a pretty

easy choice.

At my old airline, we weren't even

allowed to attempt a third approach

unless the visibility had doubled from

the previous attempts and obviously was

above the minimums.

So, realistically speaking, the best

course of action here would have been to

either hold and wait for the visibility

to improve substantially or just divert

back to Bangkok.

>> [music]

>> But, instead of having this crucial

discussion, the captain just told the

first officer to request another

approach, which he promptly did.

Now, most likely the captain was feeling

some pressure here to get his passengers

to their destination, [music] whether

that pressure was self-induced or

otherwise. So, he now became narrowly

focused on accomplishing that goal.

And with that mindset, he must have

concluded that there was no harm in just

trying again.

Maybe this time it would work.

But the truth here is that if you have

failed twice, chances are that you'll

fail a third time as well. And

if you do that, well, then you will be

even more tired than you were on the

first two.

>> [music]

>> And this is where the pilot monitoring

must step in and become the voice of

reason.

But unfortunately, that didn't happen.

Instead, the captain now made another PA

to the passengers informing them that he

would try the approach one more time and

then return to Bangkok if it failed.

[music]

But

there was still no discussion around

that decision with his first officer.

As they positioned for the third

approach, pushing through the clouds and

rain, the captain was clearly still

focused on those earlier failed

approaches.

He complained that they couldn't turn in

in time because they had gained visuals

too close to the runway. And he joked

nervously that if they didn't land on

time, well, then this plane might not be

available for its next scheduled flight

over to Chiang Mai.

Then, as they ran through the landing

checklist for the third time, he

repeatedly said,

"It's too close." And he then asked his

first officer to inform Thai Airways

operations of their plans to go back to

Bangkok if this approach was

unsuccessful.

Just a few moments later, at time 19:05,

the captain's worst fears were confirmed

as the tower controller informed them

that the visibility had now reduced down

to only 1,000 m.

>> [music]

>> On the cockpit voice recorder, you could

hear the captain mumble to himself,

"1,000 m, unable."

And that clearly showed that he knew

that continuing this approach beyond

this point was likely a just a fool's

errand.

But

perhaps out of some misplaced hope that

somehow this approach would be

different, he still didn't abandon it.

Now,

remember what I said about how minimums

exist to keep [music] us pilots from

pushing ahead into bad conditions hoping

for a miracle?

Well, this is an excellent example of

why that is.

The first officer now asked the

controller again whether it was raining

at the field, and she replied that there

was indeed still some light rain, and

that the runway was wet, and she then

cleared them to land.

And to this, the captain responded that

since they were now cleared to land,

they might as well try the approach

anyway, showing that maybe this

clearance was what pushed this decision

over the edge, even though obviously the

controller had no say in that matter.

After finishing the landing checklist,

the captain confirmed that they were on

course, and he then asked the first

officer if he could see the runway this

time,

but he didn't get any reply back.

At 100 ft above the MDA, the captain

disconnected the autopilot and attempted

[music] to turn inbound to where he

thought that the runway should be, but

just seconds later, the runway lights

appeared through the rain, predictably

still way too far off to the right. So,

the captain now exclaimed, "Oh, we can't

make it. We cannot land." And he then

called for the third go-around for the

night.

The first officer called for him to

press the go-around switch or go lever,

which he did, but there was a crucial

difference this time.

You see,

he hadn't disconnected the auto

throttle.

Remember, during the first go-around,

both the autopilot and the auto throttle

had been disconnected, [music] and the

captain then increased thrust manually,

really, really slowly over 23 seconds

and was therefore able to keep the pitch

under control,

albeit barely.

And then, during the second go-around,

both the automatic systems had been

engaged. So, while the autothrottle had

advanced thrust rapidly in just 8

seconds, the autopilot had quickly

compensated for the sudden pitch up to

keep the pitch

at or under 18°.

Now, those two previous go-arounds had

likely set the captain's unconscious

expectations of what would happen now

during the start of this third

go-around.

But, of course, this time, when the

autothrottle increased thrust [music] to

TOGA in 8 seconds, the autopilot wasn't

there to help. So, it was now entirely

up to him to keep the pitch angle under

control.

In such a light aircraft, by this point

down to only 102 tons and with slightly

aft center of gravity, a rather

aggressive pitch down input would be

needed. [music]

And this was especially true also since

in the go-around mode, the Airbus A310's

autothrottle appears to always use

maximum thrust, which in this case was

102% N1.

Now, that is way more than this

particular aircraft needed for this

go-around, and I am surprised that it

was set up that way.

On the Boeing 737 that I fly, pressing

the TOGA switches once will give you a

reduced go-around thrust of around 90%

N1, which will lead to a smoother,

controlled go-around with around 2,000

ft per minute rate of climb. And if you

want max thrust for whatever reason,

well, then you have to then press the

switch again.

But, as far as I can tell, the Airbus

A310 doesn't seem to have that option.

Now, if you are an A310 pilot or used to

be,

let me know if that's the case in the

comments below.

In any case, the result of the engines

now spooling up this way was a very fast

pitch up momentum, which had to be

countered [music] aggressively. But, it

seems like the captain's mind was at

this point still focused on the reasons

behind the failed approaches, because

even as he made the standard go-around

callouts, he also added things like

"Too close. Cannot land."

All meanwhile, the nose of his aircraft

was now rising higher and higher,

shooting through 18° in just a few

seconds.

So,

he needed to react immediately to this,

but in his distraction and complacency,

that didn't happen.

Now,

this sequence of events is actually

eerily similar to another accident

involving the very similar Airbus A300,

which had happened earlier in that same

year.

In February of 1998, China Airlines

flight 676 was approaching Taipei,

Taiwan, when the captain decided to

execute a go-around. But, moments before

that, he had applied some force onto the

control column, enough to actually

exceed the autopilot's automatic

disconnect function, which he had done

noticed, but likely hadn't fully

processed when he pressed the go-around

switches.

As a result of that, the auto throttles

pulled up the engines to go-around power

rapidly, and the nose started to climb,

but the captain initially failed to

intervene because he was expecting the

autopilot to hold the pitch.

And as a result, the pitch angle quickly

exceeded 40° nose up. The aircraft

stalled, and flight 676 plummeted into a

residential street before

the captain had any chance to recover.

The sad outcome was the loss of 202

lives.

Now, both that China Airlines captain

and this captain likely understood the

physics involved here. They knew that

their low slung engines would cause a

pitch-up moment when the thrust was

increased. So, the reason both of them

failed to action it wasn't because they

didn't understand the principle,

but instead likely because their mental

model of how the aircraft was going to

react to their inputs was based on

inaccurate starting conditions.

In the case of the China Airlines

captain, he hadn't realized that the

autopilot was off and therefore hadn't

mentally prepared himself to take manual

control of the airplane's pitch. And the

captain of this aircraft knew that the

autopilot was off because he had

intentionally disconnected it just a few

seconds earlier,

but he hadn't taken the time to

anticipate how the reaction would be

from the aircraft

>> [music]

>> when the autothrottle was still on

and how that would be different from the

previous two go-arounds.

On top of that, he was likely still

under stress from these three failed

approaches and was probably feeling

quite frustrated, which might have

reduced his focus on the basic piloting

skills that was now in front of him.

And also, by the way,

on the 737, we don't have automatic

go-around capability for most of our

approaches, only autolands. So, we are

quite familiar with the handling of

manual go-arounds.

But on an aircraft like the Airbus A310,

which can perform automatic go-arounds,

I can kind of see how the demands of a

sudden manual go-around could catch a

pilot off guard.

But the result of all of these factors

was that the captain now pushed forward

way too gently to stop the pitch from

increasing.

In fact, he used less than 50% of the

control authority that was available to

him and consequently the pitch increased

past 18° to 25, then 30, then nearly

40°.

At that point, the captain did push

forward for about 5 to 6 seconds, which

briefly brought the nose back down to

about 30°,

but

then he let go again. The nose rotated

back up past 40° and then to almost 50,

which is incredible.

Now, obviously, this caused the aircraft

speed to just quickly wash off, and its

angle of attack soon approached what was

needed for an aerodynamic stall.

From the ground, if anyone would have

been able to see this, it would have

looked like the aircraft was basically

hanging almost motionless in the sky

like a big cross.

In fact, only 19 seconds after the

captain had pressed the go around

switches, the stall warning activated,

shaking the pilot's [music] control

columns and triggering a loud cricket

alarm in the cockpit.

Now, the correct response to that

warning would have been to immediately

push the nose down to reduce the angle

of attack and increase the airspeed,

which is something that we pilots

practice so frequently that it should

become second nature to us.

But, time and time again, history has

shown that sometimes pilots react in the

complete opposite way [music] by pulling

back.

Now, we don't know exactly why this

happens, but there are a couple of

hypotheses that we can put out there.

One possibility is that when a stall

warning catches a pilot by surprise, a

deeper, more basal instinct takes over,

which is the instinct to reach out, grab

the control column, and pull it closer

to one's body.

This is possibly a reaction to acute

stress, much like when a situation

catches you by surprise while you're

driving, and suddenly you're white

knuckling the steering wheel and sitting

up extra straight in your seat.

But, of course, if you do this in an

airplane under these kind of

circumstances, you'll end up pulling

back on the yoke and raising the nose

even further, which in a stall situation

could kill you.

Now, the whole purpose of stall training

is to make sure that our instinctive

reaction to the sound of the stick

shaker is to push the controls forward

without even having any time to think

about it.

But, like I said, occasionally for some

individuals, maybe that training doesn't

quite stick. [music]

Alternatively, another reason for the

captain's behavior could have been a

sort of reverse somatogravic illusion.

Now, if you're a regular viewer of this

channel, you might remember that the

somatogravic illusion occurs when an

acceleration force pushes the pilot back

[music] in their seat, causing them to

falsely believe that the airplane is in

a pitch up.

Well, in this case, the plane really was

pitching up very steeply. [music] So,

gravity was pressing the pilots back

into their seats while the engines were

screaming [music]

at max thrust.

So,

it is possible that the captain

therefore became disoriented and

interpreted these cues to mean that the

airplane was accelerating rapidly and

would overspeed [music]

unless he pulled up to slow it down.

Either way, what we know is that as the

first officer exclaimed in surprise at

the stall warning, the captain now

grabbed his control column and for a

couple of brief panicked seconds, pulled

it towards his body, turning a dangerous

situation into

something truly catastrophic.

The pitch quickly increased to nearly

60° as the airspeed plummeted to [music]

below 30 kn,

which obviously caused the aircraft to

go into a fully developed aerodynamics

stall.

And at a height of just 1,880 ft, this

situation [music] was now completely

unrecoverable. The only way to regain

energy would have been to descend, but

there wasn't enough altitude left to get

the speed [music] back up before they

hit the ground.

As the airplane entered the stall, the

loss of airflow over the ailerons also

led to a loss of roll control and the

airplane therefore swayed wildly from

side to side as it fell, banking first

120° right, then back 50° left, 60°

right, and then left again.

The first officer now shouted and called

for full thrust,

>> [music]

>> but the engines were already producing

that.

And still, the captain clung to the

control column, holding it in the nearly

full nose up position.

But of course, with the aircraft in a

fully developed aerodynamic stall, its

effect was now limited. So, the [music]

nose soon dropped anyway, falling below

the horizon as the plane plunged towards

the ground below.

There were just enough time for the

ground proximity warning system to call

out

>> Don't sink.

>> Minimum.

>> 100.

>> And then, it was over.

At time 19:08 and 25 seconds, Flight 261

crashed into a flooded plantation with a

nearly neutral pitch attitude and a high

rate of descent. [music]

The impact severely damaged the

aircraft, but due to its very low

forward speed and the swampy terrain at

the impact site, only 51 of the 146

occupants perished in the impact [music]

itself.

Unfortunately, though, a fire soon

erupted from the left wing and then

spread quickly through the fuselage,

forcing the survivors to flee as best as

they could from the badly mangled cabin.

And tragically, many were [music] unable

to do so.

Medical examinations later showed that

50 people died primarily due to the

effects of the fire, bringing the total

[music] death toll to 101, including

both of the pilots, while only 45 were

able to escape.

Now, the final report of this accident

was among the very worst that I've ever

read.

>> [music]

>> It was missing loads of key information.

The analysis was only surface level, and

it had some factual errors that were

thankfully called out in the comments

left by the [music] French BEA, who

participated in the investigation on

behalf of Airbus.

As a result, my team and I had to

combine the raw flight data and the

cockpit transcript with our knowledge of

aircraft performance and human factors

in order to explain why the pilots of

this flight made certain decisions, as

well as why they ultimately lost control

of their airplane. And I want to

especially highlight two things that I

think that we can learn from this.

[music]

Firstly, the captain of this flight

didn't effectively control his aircraft

during either of the manual go-arounds

because he wasn't anticipating its

behavior and his control inputs were

just too [music] timid.

These are both signs of a pilot who

maybe didn't get enough manual flight

time and therefore wasn't comfortable

using the control authority that he had.

So,

if you are a pilot out there, try to get

some hand flying in when it's safe and

suitable to do so.

The second lesson has to do with risk

taking.

This crew decided to attempt a third

approach and basically all of these

approaches on the conditions that they

knew would likely preclude a successful

landing.

And the captain still [music]

disconnected the autopilot and tried to

line up with a runway that he couldn't

clearly see, putting himself and his

crew and aircraft in a situation where

he had to hand fly a go-around at night

in marginal conditions close to [music]

the ground.

And this is a situation that we'd prefer

not to put ourselves in if we can help

it, which is why we're supposed to

comply with stipulated weather minima in

the first place.

Now, I want to really shout out my

script writer, Kira, for this script,

where she has done a lot of great work.

I am so proud of my entire team for all

of the hard work that they do and that

is made possible both by the support of

our sponsors, which I hope you will

check out,

but especially by our fantastic private

sponsors in the Patreon crew.

Now, if you want to support us as well

and get to see [music] my upcoming

videos both here on Mentor Pilot and on

Black Box and Mentor Now before everyone

else, well then scan this QR code or go

to patreon.com/join/mentorpilot because

Cuz I would really love to see you

there.

Here are also some great comments that

you guys posted in our last video and I

love both responding to them and reading

them. They're great. [music]

My name is Petter Hornfeldt and you're

watching Mental Pilot. Have an

absolutely fantastic day and I'll see

you [music] next time. Bye-bye.