Original subtitles

(narrator): The Pilots of Air Transport International

Flight 7-82 prepare to depart from Kansas City

with just three of their four working engines.

- A three-engine ferry, it's a difficult manoeuvre.

- A hundred knots... - Okay.

(narrator): Their takeoff goes tragically wrong.

- We're off the runway.

- Go max power!

- You got it!

(crashing)

- This accident, of course, was very tragic

and they did not make it.

(narrator): Examination of the wreckage reveals very little.

- From stem to stern, wingtip to wingtip, it was airworthy.

(narrator): But when investigators

visit the simulator used to train the pilots...

- This is not how their aircraft would have behaved.

(narrator): ...they begin to understand

the origins of this accident.

- The simulators are extremely accurate.

We were shocked that this one, uh, was not.

(intro)

- Okay, first things first,

let's review our three-engine ferry procedure.

- Sure thing. - Yup.

(narrator): Air Transport International Flight 7-82

has been cancelled.

One of its four engines won't start.

- Maintain directional control using rudder

and nose wheel steering.

(narrator): The replacement crew is taking the cargo plane

for repairs to a maintenance facility in Massachusetts.

- The only crews that are permitted to do

a three-engine ferry are the very experienced crews,

because it is very dangerous.

It does not happen very often.

(narrator): Captain Walter Miga has worked his way up

from flying the DC-6 and DC-7

and is now captain on the airline's DC-8 aircraft.

- The captain had about four thousand hours in the DC-8.

He'd been flying the DC-8 for years.

- Max wind for takeoff crosswinds 10 knots,

tailwinds 5 knots. - Okay,

let's check the weather on that.

(narrator): First Officer Mark Ulmer is new to the company

and the plane, but he's been flying for 14 years.

- He only had about two hundred hours in the DC-8.

However, uh, he was commended for being enthusiastic,

for trying to learn, and being an important part of the crew.

(narrator): There's a third crew member onboard:

Flight Engineer Kerry Hardy.

- The flight engineer had more than 4,000 hours of flight time,

and in large airplanes.

He was a very experienced pilot.

(narrator): It's a two-hour flight from Kansas City

to Chicopee, Massachusetts where the engine is to be repaired.

- Here you go.

- Thank you.

(narrator): The crew has concerns about their schedule.

- There was a curfew on the airport in Massachusetts.

The airline had told the crew

that they needed to land by 11 p.m.

because that is when that airport would close.

- What we're gonna need to do is get as much direct as we can.

That will allow us to fly a little bit better

than the 2-88 knots indicated.

- Well, as pilots, we're always trying to hurry.

The crew had decided that they wanted to fly

a more direct route at a faster speed.

- So, we'll be using 1-9, right?

(narrator): First Officer Ulmer wants the longest runway

for their takeoff.

- I'm gonna request right

because we'll get an extra 13,000 feet.

- Okay.

(narrator): In order to take off, the DC-8 needs power

from at least three of its four engines.

- You always want the longest runway that you can get

with a takeoff like this so that you have more room

to get the airplane up to the proper speeds.

- Kansas City, this is Air Transport 7-82 ready to taxi.

We're gonna be a three-engine departure.

(narrator): Flight 7-82 checks with Air Traffic Control

about their runway assignment.

- Air Transport, 7-82, south on bravo taxi runway one left.

- Okay, south on bravo taxi one left, Air Transport 7-82.

What's the winds?

- Winds 240 at 4.

- Roger. - That's a tailwind, right?

- Yeah.

- You don't want a tailwind with a three-engine takeoff

because it's going to take more runway.

- 240 and running into what?

- Four.

- Four knots.

- The tailwind was acceptable.

The maximum tailwind allowed was five knots,

and they had four knots.

(narrator): Flight 7-82 taxis to the longer runway.

- Takeoff data?

- Okay, this is a max power takeoff.

(narrator): The pilots review the required speeds

for the three-engine takeoff.

- Okay. I got VMCG of 1-0-7, 1-23 for VR,

1-40 for V2, and 2-10 for the cleanup.

- Set.

(narrator): This takeoff requires raising both inboard engines,

Engine 2 and Engine 3, to max power.

Only when the plane reaches a certain speed

can the third working engine, Engine 4,

be increased to max power.

- Controlling the airplane on the runway

while you are adding the power from the third engine is vital.

It must be done slowly and you must be able to keep control

of the airplane at all times.

- And of course, we'll all be watching real close

for loss of directional control.

- The two engines on the right side

are gonna cause the aircraft to veer to the left.

It's only gonna have one engine operating on the left side,

so directional control is the critical issue

on a three-engine takeoff like this.

- Number 1 engine is inoperative.

So what we're gonna do is set max power on number 2

and number 3. - Right.

(narrator): As the crew nears the end of the taxi,

they finalize the takeoff plan.

- Okay, then I'll ease in number 4.

- And I'll call increments of point one.

- Yeah.

(narrator): Flight 7-82 is now ready for takeoff.

(whirring)

- Okay, coming up on 2 and 3.

(narrator): The crew lets the inboard engines spool up

before the plane starts moving.

- Set max power.

- Max power 2 and 3.

(narrator): At 8:20 p.m.,

Air Transport International Flight 7-82 begins its takeoff.

(tense music)

- One point three.... point four...

(narrator): The flight engineer closely monitors engine power.

- Airspeed's alive. - ...one point six.

- Eighty knots.

(grunts) - Ninety knots.

(narrator): But 20 seconds into the takeoff roll,

the plane veers left of the centreline.

- One point eight.

- A hundred knots.

(narrator): The captain uses his rudder

to redirect the plane to the centre of the runway,

but it doesn't work.

- Abort.

(tense music)

Spoilers.

- Air Transport 7-82 aborting takeoff.

(narrator): Flight 7-82 aborts its three-engine takeoff.

- Air Transport 7-82, roger.

When able, turn right off the runway.

Do you need any assistance?

- Negative assistance? - No. Negative.

- Negative assistance, Air Transport 7-82.

(narrator): The crew considers what went wrong.

- We were at 1.6 and then power went all the way up

to 190 as you ran it up, so it went up real fast.

- And number 4 jerked up

while I was trying to ease it in.

- You brought it up too fast, or it jerked up, or what?

- The power on it just came on too fast, that's what happened.

(narrator): Hardy proposes a solution.

- If you want to try it again,

I can try adding the power if you like.

- Okay, let's do it that way.

- It was decided that the flight engineer

would help the captain and he would add the power

on the third engine for the second takeoff.

- Tell them we wanna taxi back and have another try at it.

- Kansas City ground, Air Transport 7-82

we'd like to taxi back to depart one left again.

- Air Transport 7-82, roger, taxi one left.

- Okay.

(narrator): While returning to the runway,

the crew revisits the takeoff procedure.

- Okay, when do I have to have max power

in on the outboard engine?

- One hundred and seven.

(tense music)

(narrator): Flight 7-82 is again ready for takeoff.

- Kansas City control, Air transport 7-82,

ready to go at the end of one left, a three-engine takeoff.

- Air Transport 7-82 heavy, control, one left,

turn right 030, cleared for takeoff.

- I've lined up just to the right of the centreline here.

- That's a good idea.

- The captain did line up on the right side of the runway

because they veered off to the left on the first takeoff

and he wanted to give himself more runway width available

in order to do the second takeoff.

- Okay, make sure that 2 and 3 are set...

- Max power?

- Yeah.

(narrator): At 8:26 in the evening,

Flight 7-82 begins its second takeoff attempt.

- One one... one two...

(whirring)

one three...

One four...

One five...

One six...

- Airspeed's alive.

(narrator): Ten seconds into the takeoff roll,

the plane again starts veering to the left.

- One seven.

(narrator): The captain tries to direct it back

to the centreline with the rudder.

- God bless it.

(tense music)

- Keep it goin'. - Keep it goin'?

- Yeah.

- Eighty knots, ninety knots... - Okay.

(narrator): The captain attempts to get airborne...

- We're off the runway.

(narrator): ...but he doesn't have the speed.

- Go max power!

(narrator): He tries to salvage the takeoff.

Flight 7-82 is barely in the air.

- Get the nose down!

- Go max power. - You got it!

- Max power!

(crash, explosion)

- Alert 3. Runway 0-1, north end, DC-8 on takeoff.

(narrator): Air Transport International Flight 7-82

has crashed beside the runway at Kansas City Airport.

Fire and rescue units are nearby on a night-training exercise.

They rush to the crash site.

- We heard a real loud boom like something hit the dock real hard.

- It didn't take long for fire rescue crews to arrive on the scene, I'm assuming?

- No, no. Just in a matter of about 60 seconds.

(sirens)

(narrator): All three crew members are dead.

The National Transportation Safety Board, or NTSB,

is notified, and Bob Benzon is named Lead Investigator.

- The aircraft had broken apart, although it three large pieces.

The fuselage, behind the wings seemed to be reasonably intact.

The wing assembly itself seemed to be in one piece.

(narrator): The question Benzon must answer:

Why was the crew unable to execute this takeoff,

not once, but twice?

- We were aware that it was an attempted three-engine takeoff.

And because that is an unusual manoeuvre,

we wondered if that could have been part of the reason

the aircraft crashed.

- Thanks for taking the time to meet.

(narrator): While the plane wreckage is collected,

investigators speak to the controller,

to learn more about the crash.

- When did you realize there was an issue with Flight 7-8-2?

- I guess the first thing that caught my attention

was that they aborted their first takeoff.

- Why?

- I was told the plane veered left on the runway.

- And then?

- They lined up for their next takeoff

and the same thing happened. Only worse.

- The plane veered left,

and this time they rotated awkwardly, nose high.

- There was a fire coming out of one of the engines.

And then there was a loud bang.

Then the plane stalled, rolled left, and crashed.

(foreboding music)

(crash, explosion)

- The fire, what side?

- Left.

- That had to be engine number 2.

(narrator): Plane engines are numbered from left to right.

For Flight 7-82, the number one engine was inoperative.

If engine two failed,

the crew would be left with only the right engines.

- We knew because this was a three-engine takeoff attempt,

and if we'd lost any more power on the left side of the aircraft

it may have caused an immediate crash.

- What have you got?

(narrator): Investigators examine engine number 2

after its recovery from the crash site.

- Ah, no penetration on the nose cowl.

Fuel is present throughout.

Compressor bleed valve was closed.

(narrator): They discover that it was operative

and it didn't fail on takeoff.

- That fire in the engine was probably a compressor surge.

- Hmm.

- A compressor surge is a phenomenon

that occurs when the tilt of the engine is high enough

that it disturbs the airflow going into the intake

and causes a big bang, flames coming out both ends perhaps.

But technically the... the engine isn't damaged at all.

- No unusual wear.

(narrator): If all three remaining engines were working,

could one of the brakes have locked,

causing the plane to veer left?

- The treads look almost new.

No flat spots.

(narrator): The tires show no sign of abnormal braking.

- Okay, let's have a look at the brakes.

- We thought that because they had aborted the earlier takeoff,

maybe they, uh, they used the brakes very heavily

and caused them to, uh, fail, frankly, or weaken,

and then fail completely during the second takeoff.

- Brake stack assembly looks fine.

- No discolouration or pitting on the pads.

- Discs are clean.

No signs of overheating.

- We looked at brakes very carefully.

Uh, we found nothing really physically wrong.

There was no overheating or melding together.

So in essence the brake system did not become an issue.

(narrator): The NTSB is still unable to determine

why the crew lost control.

- Let's check out the rudder.

- The rudder plays a massive role in a three-engine takeoff.

It is the key to keeping the airplane going straight

and being safe.

(narrator): Three-engine takeoffs

require the two inboard engines

to reach V-M-C-G, or Minimum Control Speed on the Ground.

Only then can the rudder counteract the directional force

of the third working engine as it's powered up.

(whirring)

- The faster you are going, the more rudder control you have,

and you need that rudder control to be able to keep

the airplane on the runway.

- How's the rudder control cable look?

(narrator): Did the rudder malfunction,

causing the crew to lose control of the plane?

- The cable is severed in several different places, see.

- There's no corrosion or fraying.

Probably severed in the crash.

- Agreed.

Let's test the rudder.

- We looked very carefully at the cabling

used in the rudder system.

And the only thing we did note were simple breaks in the cables

because of overload, instant overload,

which occurred during the crash.

- Control cable ready?

- Good to go.

(narrator): If the rudder cables didn't fail,

then could the rudder itself have malfunctioned.

- Okay, deflect left.

(narrator): Investigators test its movement.

- Turning to the left.

(narrator): But the rudder moves freely.

- We're good.

- On occasion in previous accidents rudders have locked up

or become blocked by some... some broken part.

And we found that, uh, this did not occur during this event.

(narrator): The mystery of Flight 7-82's crash deepens.

- In the end, uh, we looked at the entire wreckage

very, very carefully, uh, from stem to stern,

wingtip to wingtip,

and determined that it was airworthy.

We had to look into other areas to determine what happened.

- Okay, we found skid marks on the runway

for the first takeoff.

Starting here all the way to here.

Then they aborted.

(narrator): Investigators now turn to evidence left behind

by Flight 7-82 to determine why it veered off the runway.

- In the second takeoff, the skid starts even earlier: here.

And the marks last...

...until here.

- The skid marks were there indicating that,

although the aircraft was rolling forward,

it was skidding to the left side of the runway.

(narrator): The skid marks are a telltale sign.

- It looks like they had too much power

to the right side of the plane.

And it happened early on. Look.

On both takeoffs, they skidded to the left after only 500 feet.

- Correct.

So maybe there was an issue with the power

in that third working engine.

- Let's find out.

- We discovered that the engine operation

would probably be some kind of an issue.

And the only way we could determine that

was to go to the flight data recorder.

(narrator): Did engine number 4

somehow cause the plane to veer too far to the left?

- There's the airspeed data there.

(narrator): Investigators examine the flight data recorder

to learn how the crew managed their engines.

- It looks like we had a glitch here.

- To learn more about how the crew manipulated the throttles,

we needed parametres from the FDR:

airspeed and the engine pressure ratio, or the EPR.

We had airspeed, but unfortunately,

this FDR did not record correctly the EPR.

- We need a workaround to get that engine data.

- What about a sound spectrum analysis?

- Sure, let's give it a shot.

- Okay. Coming up on 2 and 3.

(powering up)

- There's a microphone in the cockpit.

It records every sound.

And it can record and hear the engines

as they are increased or decreased.

And they're very accurate.

(narrator): The team conducts a digital analysis

of the engine sounds.

- Done. Great work.

- Let's compare the third working engine, Engine 4,

with the plane's overall airspeed data.

- In the end, we used both the FDR and the CVR

to determine the airspeed throughout the entire flight,

and the power setting, for all three of the operating engines.

- It looks like...

...the third working engine, engine number 4,

almost reaches max power in 10 seconds

when the airspeed is only about 65 knots.

- That seems awfully slow.

(narrator): Was the power to engine 4

increased too soon in the takeoff roll?

- Let's check the three-engine takeoff chart.

(narrator): Investigators examine the chart

used by the crew to determine the speed they needed to reach

before applying max power to engine number 4.

- What was their weight?

- Ah, their weight was 220,000 pounds.

Temperature was zero degrees.

- They shouldn't have put the third working engine

to max power until they reached 116 knots.

That's the target speed.

- At 65 knots, it was way too much power, too soon.

- One three, one four, one five, one six.

- Airspeed's alive.

- One seven.

(narrator): Investigators conclude engine 4

was powered up too fast, pushing the plane to the left.

- They never reached the speed they needed

for the rudder to counteract the power of both engines

on the right side of the plane.

- That's why they veered off the runway.

(narrator): Why did the pilots increase power

to the right outboard engine so quickly?

- We knew that the issue here was reaching VMCG

at exactly the right time, with the correct airspeed,

and everything. But the crew didn't do that

and that one became the big question.

Why? Why didn't they do that?

- Whenever you're ready.

(narrator): Benzon's team listens to the cockpit voice recorder, or CVR,

to better understand how the pilots of Flight 7-82

performed the three-engine procedure.

(Captain): Okay, this is a max power takeoff.

Speeds: VR 1-23, 1-40 and 2-10.

VMCG of 1-0-7.

- Okay, I got VMCG of 1-0-7,

VR of 1-23, 1-40 for V2, and 2-10 on the cleanup.

- Both pilots think the VMCG is 107 knots.

It should be 116.

- Yup, VMCG is 116 knots.

- So, how did the flight engineer get that wrong?

- What a minute.

Check this out.

107 is the speed for a temperature of 30 degrees...

Celsius.

(narrator): Investigators discover the flight engineer

used the speed for a temperature of 30 degrees Celsius,

but on the night of the crash it was 30 degrees Fahrenheit,

approximately zero degrees Celsius.

- That C for Celsius is so small.

He must have used 30 Celsius

thinking the chart is in Fahrenheit.

(somber music)

- They selected the wrong line to get their V speeds.

- Here you go.

- Thanks.

- They were nine knots off on what VMCG should have been.

(narrator): It's happened before.

Four years earlier, a flight engineer

on another Air Transport DC-8

miscalculated the plane's weight,

providing the captain with the wrong takeoff speeds.

- I was one of the first investigators

from the NTSB to arrive on scene.

When I went into the cockpit,

it didn't take any smarts at all to see the takeoff data

was 100,000 pounds too low.

- And as a result, the company very seriously

began to emphasize the need to check those calculations

and make sure they're accurate.

And so it's disturbing that in this accident,

that did not happen.

- Was that the only thing the pilots missed?

(tapes whirring)

- A question to consider, Captain,

once we get near VMCG, if we've used all our rudder control,

you might wanna consider abort, because when we get higher,

we're gonna be in worse trouble, correct?

- That's correct, absolutely.

(narrator): The CVR of Flight 7-82

reveals that the captain and first officer

were uncertain about the impact of VMCG speed on rudder control.

- The first officer made the comment that you have...

rudder control is less effective once you go beyond VMCG,

when actually the opposite is true.

- Uh, no, actually.

(narrator): The flight engineer realizes their mistake.

- Above VMCG, your rudder has more authority.

It's helping you more.

- I understand.

(narrator): But the captain fails to acknowledge

the correction.

- The flight engineer had the right answer

that the faster the airplane goes,

the more effective the rudder becomes,

and you need that effectiveness

to keep the airplane on the runway.

- Why didn't the crew stop

and discuss it until they reached a consensus?

- That's the advantage of having three people in the cockpit.

- The right answer generally exists somewhere in the cockpit.

In this instance the flight engineer had the right answer.

- Maybe their lack of understanding

about the procedure caused them to abort their first takeoff.

- It's possible.

- Here it is.

Check that out.

- The power went all the way up to 190 as you ran it up,

so it went up real fast.

- And number 4 jerked up while I was trying to ease it in.

- You brought it up too fast?

(narrator): Investigators learn that the captain struggled

with engine 4's thrust during the first takeoff roll.

- The power on it just came on too fast, it's what happened.

- If you wanna try it again, I can try adding the power,

if you like.

(narrator): The flight engineer proposes an unorthodox change

to the procedure.

- Okay, let's do it that way.

- The flight engineer took over control of engine 4?

- That's crazy.

- They couldn't have understood the difficulty of the procedure

if they split the responsibility.

- The core problem is the captain could not tell

how much rudder he was going to need

because he wasn't the one adding the thrust to the airplane.

It's like trying to have two people drive a stick shift car,

one doing the steering, one doing the shifting.

It just doesn't work.

You need to have one person in control at one time.

- The flight engineer and the captain thought

they could coordinate the takeoff together.

Let's figure out how they did.

- We worked with the Douglas Aircraft Company

and they produced a, uh, a graph of what an ideal

three-engine takeoff should look like with airspeed

and engine speed.

We compared that with what actually happened.

- So this is what the crew was supposed to fly.

- It seems like the power and speed rise steadily,

in unison until they reach VMCG.

- Very, very smooth and synchronized.

But in the accident flight, the flight engineer

brings up power to the engine number 4 really quickly.

(narrator): The plane begins to veer left, off the runway.

- Then, he brings down the power abruptly, here.

Probably in reaction to the plane being out of control.

- The most probable thing we think happened

was that the flight engineer saw the aircraft going off

to the left of the runway and kinda panicked

and tried to pull the engines back a little bit.

- God bless it.

(whirring)

- Keep it goin'.

- Keep it goin'? - Yeah.

(narrator): Instead of aborting,

the captain attempts to get airborne

because they're about to veer off the runway.

- He was countermanded by the captain, who said,

"No, no, keep going." So the engines went back up.

- We're off the runway. - Go max power!

- Then the captain insists

that he goes to max power a few seconds later.

- They don't have enough speed for takeoff.

- The captain never decided to abort

and it appeared to be that he thought

he could get the airplane in the air.

We will never know.

- The captain and the flight engineer

could never coordinate their actions.

- One two...

One three.

(narrator): Why did the crew fail to execute

a takeoff procedure that they were specifically sent

to Kansas City to perform?

(tense music)

- You got it!

- Find anything?

(narrator): Investigators review the work history

of Flight 7-82's crew to determine their experience

with three-engine takeoffs.

- Our flight engineer and our first officer were new hires.

They were still on probation.

It's almost certain that neither

had performed a three-engine takeoff.

- And the captain?

- He'd done it three times, but only as first officer.

He'd never actually flown as a captain.

- They were all green.

(narrator): The closest the captain came

to flying this type of takeoff was in a simulator.

He only trained on it twice.

Investigators travel to Colourado to assess the simulator

the airline used to train its pilots

to perform three-engine takeoffs.

- Okay...

This is a three-engine takeoff,

and we are watching for directional control.

(narrator): Did the pilots have sufficient training

to perform this difficult procedure?

- I'll call out engine power until you get to max.

- Part of the examination was to see if the simulator

accurately portrayed what a three-engine takeoff should entail.

- Okay, let me give this a try.

Engines 2 and 3 are spooled up.

Engine 4 is set.

Here we go.

Slowly bringing up 4.

(powering up)

- One five...

One six...

One seven...

- I am barely using my rudder.

There is no force pushing me off the runway.

- One eight.

- We have VMCG.

Max power to engine 4 and...

rotate.

(narrator): In the simulator, the three-engine takeoff

is no different from a regular takeoff.

- Okay, cut.

- The simulator was not calibrated properly

in that it did not require the normal amount of rudder

that would be required with a three-engine takeoff.

- This is not how their aircraft would have behaved.

- Not a chance.

- Simulators are for the most part extremely accurate,

and in this particular case, we were...

frankly, shocked that this one was not.

It also happened to be the one the crew trained on.

And this would really give them, I think,

a false sense of security when it came to the difficulty

of making a three-engine takeoff.

- I'm calling from the NTSB.

(narrator): How did Air Transport International,

or ATI, compared to other cargo operators

on three-engine takeoffs?

- I'd like to talk to your chief pilot.

- And we checked with the other cargo operators

who were operating the DC-8. There were nine of them.

- So only your most qualified pilots do it, huh?

- The vast majority of them used their most experienced crews

and limited the operation to daytime in good weather.

Now, this led us to believe that the industry, as a whole,

understood the danger and the difficulty

of a three-engine takeoff better than ATI did.

- Why did the airline choose such an inexperienced crew

for such a difficult flight?

- I think it's time we asked them that.

- Our team went to Little Rock,

which was the headquarters of the ATI operation,

and we met with all of the key people there.

- Thanks for meeting with me.

So, I've been told

that you didn't use your most experienced crew.

Why not?

- I needed my experienced crew for a revenue flight.

The Flight 7-8-2 crew was only legal for a non-revenue flight.

- How so?

- They'd only had enough rest for a non-revenue flight.

- You don't say.

(narrator): They discover that the crew was selected

because of a critical regulation

governing non-revenue flights and rest times.

- Normal flights where the company makes revenue

are covered by flight and duty time regulations.

For safety reasons, they have to provide a crew

that has been properly rested. But there is a loophole

that if the flight is not revenue,

then there are no flight and duty time regulations.

In theory, a crew could fly non-revenue flights indefinitely.

- Abort.

(narrator): Investigators now wonder how rested the crew was

when Flight 7-82 prepared for takeoff.

- So, take us through this.

(narrator): Investigators closely examine

the crew of Flight 7-82's recent flight assignments.

- Okay, the morning before the accident,

our crew lands in Germany after a seven-hour flight...

...from Delaware.

- And how many time zones did they cross?

- Six.

Later that same day, they return,

landing first in Gander, Newfoundland.

Then, another flight to Delaware.

(scoffs)

- So, they cross another six time zones heading back?

- We were aware that their circadian rhythm had to be

pretty much out of sync.

It's simply not a good thing.

- Fatigue is not easy to recognize,

but it degrades all aspects of thinking,

decision-making and performance.

- They got into their hotel in Delaware at 2:40 a.m.

on the day of the crash.

- How much uninterrupted rest did the captain get?

- Less than 5 hours.

- So, they didn't have enough rest to fly anything

except this flight, and they were sent,

despite lacking the training and experience they needed.

(narrator): The NTSB concludes its investigation,

outlining the fateful steps that caused the crash of Flight 7-82.

It begins with the poorly rested crew.

- One of the consistent findings of fatigue research

is that people tend to underestimate

how tired they are.

Much of the time you can conduct a safe flight doing that.

But if something unexpected happens,

you're at a real disadvantage.

In this case, the takeoff procedure

was much more complicated

than the crew had been trained to expect.

- Here you go. - Thanks.

(narrator): The flight engineer makes an error in a calculation

that the other pilots don't catch.

- Once we get near VMCG,

if we've used all our rudder control

you might wanna consider abort,

because when we get higher,

we're gonna be in worse trouble, correct?

- That's correct, absolutely.

(narrator): The pilots are confused

about the takeoff procedure.

- Uh, no, actually.

- Above VMCG your rudder has more authority.

It's helping you more.

- This was a very delicate procedure

and they seemed to think otherwise.

- If you wanna try it again, I can try adding the power,

if you like. - Okay, let's do it that way.

(narrator): A decision to improvise a difficult takeoff

leads to the plane veering out of control.

- It was a very bad decision to improvise on this procedure

because only the person operating the rudder

can tell how effective that is when advancing the throttle.

- Okay.

(narrator): The captain decides not to abort a second time.

The plane lacks enough speed to safely get airborne.

(shaking)

- We're off the runway.

- Go max power!

(whirring, crash)

- It's hard to say what was going through his mind,

but I think he believed he could get the aircraft airborne.

(narrator): In their report,

the NTSB makes several key recommendations.

One is to revise the training procedures

for three-engine takeoffs.

- The crew not only was not trained well,

it was inexperienced in the particular manoeuvre

they were talking about, and they were tired.

(narrator): Closing the loopholes

that allow flight crews to perform non-revenue flights

without adequate rest is another recommendation.

- The reason this is relevant is because the company

elected to have this tired crew fly the three-engine ferry

rather than the rested crew that was already present in Kansas City.

It was legal.

But it was not a good idea.

(narrator): The NTSB also recommends

that three-engine ferry flights

be assigned to specially designated, experienced crews.

Training instructors advised pilots

of the simulator's shortcomings for three-engine takeoffs.

- The company exceeded the Board's requests.

They really took it very seriously

and did change their whole procedure to prevent this.

(narrator): The airline industry as a whole does not follow suit.

Non-revenue flights remain exempt from flight

and duty time regulations.

- There should be no difference with safety

whether there are people on the airplane,

freight on the airplane, or the airplane is empty.

One rest rule requirement should apply to everybody,

every time, everywhere.

Subtitling: difuze