Mayday (2013) s10e02 Episode Script

VOICEOVER: January 2008.
The Boeing 777 is the gold standard for commercial aviation.
For more than 10 years, the plane has logged 2 million flights without a single major accident.
Clear to land But less than 1,000 feet above London's Heathrow Airport, all that changed.
Hey, I can't get power on the engines.
What do you mean? It's not giving me power.
Mayday! Mayday! Speedbird! Speedbird! Whatever brought down one of the world's most technologically advanced air planes can't be found at the crash site.
And they had the pilots, they had the passengers, they had the aircraft.
They had everything there except the thing that caused it.
That had disappeared, as if by magic.
British Airways Flight 38.
A 10.
5-hour journey from Beijing to London.
Captain Peter Burkill has made this trip many times before.
I was operating that route a lot.
It was one of my favourite flights, because it was daylight, you could see the views the whole way along the route.
Burkill and First Officer John Coward both have thousands of hours flying the 777 under their belts.
It's a joy to fly.
She handles really well, she's got modern equipment, the computers are easy to use.
She's got the range, the ability to do short haul as efficiently as long haul.
For these two, as for many pilots, a day on the job consists of sitting in the cockpit for long hours as computers fly the plane.
Powered by two massive Rolls-Royce engines, the Boeing 777 is one of the safest planes in the industry.
Nick Harris is flying back to London after a business trip to China.
The good thing was it seemed to be that we were coming in early into Heathrow.
I was looking forward to going home to see the family.
Flight 38's destination is Heathrow Airport, one of the busiest international airports in the world.
Nestled in the south-west corner of London, it's bordered by the A30 motorway.
People who live in the nearby neighbourhood of Hounslow are well accustomed to the sound of jets taking off and landing.
The area's tightly packed houses come into distant view for British Airways Flight 38.
The autopilot is in control as the crew lines up with runway 27 left.
As it nears the ground, the plane is buffeted by some high winds.
Just a little turbulence.
It's the first wrinkle in an otherwise smooth journey.
It's not me.
That's the auto throttle is doing that, I think.
We were picking up wind gusts of about 20 to 30 knots and we were fully aware that the auto throttles would be able to cope with that so they'd be moving up and down quite a lot.
You may want to keep the autopilot on a little longer than usual until the wind calms down.
Will do.
In turbulent weather, the autopilot can make faster adjustments to keep a plane level than the pilots.
1,200 feet above ground You'll have to turn that off for now, sir.
- Sure.
- Thank you.
Two minutes before landing, Coward takes over flying the plane.
You have control? Just as the two pilots had planned.
I have control.
I had a few seconds to spare before clearance to landing and I was just tidying up my map charts.
CONTROL TOWER: Speedbird 38, prepare to land 27 left.
Cleared to land 27 left, Speedbird 38.
I remember looking out to see if our gate was available.
COMPUTERISED VOICE: 500 feet.
500 feet.
Stable? Well, sort of.
Suddenly, there's a problem.
Pete, I can't get power on the engines.
It's not giving me power.
What's going on? What do you mean? I was looking at the engine instruments and they didn't make sense to me because we were asking for full power but the engine instruments were not giving us any power at all.
What's going on? Flight 38 is crippled in the worst way imaginable.
It looks like we have double engine failure.
With neither engine providing enough power, the jet won't make it to the airport.
With 152 people on board, Flight 38 is falling toward London with only seconds until it hits the ground.
This can't be happening.
This is one of the most modern jets in the world.
In the cabin, there's no indication that anything's gone wrong.
On approach to land, everybody seemed relaxed in the cabin.
We were just looking forward to landing.
COMPUTERISED VOICE: Airspeed low.
Airspeed low.
Airspeed low! By now, I was looking at our impact point.
I could see a set of buildings around the Hatton Cross area and a petrol station.
I just knew if we were to nose, then it was certain, 100% fatalities.
That point, I felt the weight of my four bars on my shoulder.
I am the captain.
If I don't do anything, then everybody will die.
This close to the ground, Burkill doesn't have many options.
He could take back control of the plane from Coward, but doesn't think that's the right move.
First decision, really, was to not take control.
I remember looking over at John.
He was still flying the plane well.
Good.
He's doing a good job.
He's doing what I need him to do.
Burkill knows that even if they manage to clear at Hounslow, the jet could still smash into the busy A30 motorway or the antennas at the edge of the airport.
I thought about raising the gear.
That's the biggest drag on a commercial jet, but I needed the gear to crash on.
We were going to impact and that was going to take the brunt of the impact, so I had to leave the gear down.
With a crash imminent, Burkill has only one move left.
But it comes with enormous risk.
I knew I had about 15 seconds to make a huge decision.
I needed to get past those buildings.
It was obvious to me that I needed to raise the flaps.
Retracting the flaps will reduce drag, but also lift.
The plane will fly further but drop faster.
Burkill must determine which is more important.
I wanted to discuss it with the crew, but this was all in a matter of seconds.
I remember holding that lever for a fraction of a second.
He hopes this move will help his plane avoid catastrophe.
The effect was immediate.
Now, Burkill makes the announcement every pilot dreads.
Mayday, mayday! Speedbird, speedbird! It was hard, very hard.
I remembered a couple of impacts and the noise.
Good God.
The landing was hard.
Just a huge bang.
The noise, bits falling off the ceiling.
Then I also became a passenger because we were now in an uncontrolled aeroplane.
We were sliding along the ground and I didn't know what we were going to hit next.
And then I thought about my wife and kids and I said goodbye to them.
The lights went out.
I could hear the wreckage breaking up.
And then we stopped.
When we actually stopped .
.
I was surprised that I was still there.
And then I looked round the flight deck and I was more surprised to see that all three of us were unscarred.
WOMAN SHOUTS: Everybody, stay still with your seatbelts fastened! And I quickly became the captain again.
The crew switches off the fuel APU fire switch.
.
.
hydraulic and electrical systems.
Control switches, cut.
Check list complete.
I then had to look after my survivors.
I thought 20% were dead because this aircraft was broken up.
This is an emergency.
Evacuate, evacuate.
The air stewardess then took control completely.
As she opened the door, I smelt the fuel.
We should hurry.
I realised that we need to get off this aircraft.
The big danger now is fire.
Having survived the landing, passengers could be killed if flames sweep through the jet.
Jump! Jump! I left all my luggage and I just got myself down the slide.
Jump! Let's get out of here.
I went through the galley and checked the right-hand aisle and I was surprised to see nobody there.
Despite Burkill's fears, there are no casualties.
Outside in the cool January air, the scale of the disaster is clear.
The right wheel had actually been ripped off during the impact.
The engines were just sad, they were ripped apart.
The cowlings were ripped off and the engines were half buried.
It wasn't an aircraft any more.
I now started to realise that we'd been involved in a major accident.
I saw the damage to the undercarriage and to the wing.
The plane narrowly missed the A30 and the antennas at the foot of the runway.
It could have been total devastation.
If the impact had been on anything other than soft mud and grass, the likelihood of a fire was immense.
Given the amount of damage, it's incredible that only 47 people were injured.
I felt quite lucky that I'm still alive.
It was quite a surreal feeling.
From China, British prime minister Gordon Brown praises Burkill and the crew.
I think it's right to pay tribute to the calmness and professionalism of the British Airways staff, the captain, Peter Burkill, for what he achieved in engineering a landing that protected lives.
Investigators have some early hunches about what caused the engines to fail.
They also have plenty of clues.
But a key piece of evidence is already missing.
Heathrow's emergency planning swung into action.
The crash of British Airways Flight 38 dominates the news.
It's the most serious accident at Heathrow in 30 years.
Heathrow Airport is a very busy airport.
It is the main airport in the UK, so anything that happens in Heathrow Airport immediately hits the media.
More significantly, the 777 is one of the world's most modern and reliable jets.
The accident unsettles the airline industry.
The 777 is as safe as an aeroplane can be.
It had such a superb safety record.
What on earth happened? You don't like not knowing whether it might happen to the rest of your fleet.
Investigators quickly arrive on the scene.
Phil Sleight is the lead engineer for Britain's AAIB, the Air Accidents Investigation Branch.
When you first approach an accident site like this, the first thought you've got is, "Where do I start?" The plane landed about 1,000 feet short of the runway.
The landing gear's pushed up through the wings, the nose landing gear collapsed.
Just hours after the crash, investigators interviewed the crew.
Captain Peter Burkill has a first-hand account of what happened.
I think I wanted to talk about it because I knew of the innocence of the whole crew.
This isn't our fault.
We didn't do anything wrong.
But the engines weren't giving us any more power.
It looks like we have double engine failure.
We tried to give them more power manually, but there was nothing.
Burkill explains that something had caused both of the jet's Rolls-Royce engines to roll back at the same time, robbing the aircraft of power.
When you hear something of a double engine failure, you start looking at what's common.
What could cause both engines to fail at the same time? So our first thoughts were that it may have run out of fuel.
A fuel shortage is rare, but it has happened before.
In 1983, a Boeing 767 ran out of fuel flying over central Canada.
A mistake in converting between metric and imperial measurements left the plane with much less fuel than the crew thought.
At 26,000 feet, they lost power to both engines.
The captain was able to glide the plane to a safe landing at an abandoned airbase.
In the case of Flight 38, this theory is immediately questioned.
SLEIGHT: Jet fuel has a certain odour.
Anyone who has worked with an aircraft will know the distinctive smell of jet A1 fuel.
There was certainly a lot of fuel leaking from the aircraft from the ruptures to the bottom of the engines when we arrived on site.
A check of the plane's dipsticks confirms two of the tanks did, in fact, have fuel.
So we knew that there was plenty of fuel on board this aircraft to complete the flight.
There remains several other possible explanations for such a loss of power and investigators must tackle them all.
There is great confidence that the mystery can be solved, as any and every piece of evidence investigators could ever want is readily available.
They have access to the pilot .
.
the crew .
.
and most importantly, the entire plane.
We were quite fortunate to have the amount of data that we had on this aircraft because the aircraft remained intact.
We were able to interrogate lots of computers and also get lots of data from the recorders.
Additionally, we had data external to the aircraft, such as radar data, and also the radio transmissions as well.
Mark Ford retrieves the plane's Flight Data Recorders, which promise to give investigators important clues.
He has another resource - the QAR, or Quick Access Recorder.
While the black boxes are in the rear of the plane, the QAR is a data recorder in the front used mainly for diagnostics.
It does have an advantage over the flight recorder, though, in that it has a greater memory capacity and can record additional parameters over and above those available in the flight recorder.
You can recover the data from a Quick Access Recorder within a matter of minutes, in some cases.
Speed, altitude, control settings, cockpit conversations.
In all, the various recorders have preserved 1,400 different pieces of data, which should help the investigators close the case.
But when they study the QAR, what they find is chilling.
The Quick Access Recorder data stopped about 45 seconds prior to the accident itself.
Initial thoughts were that there was an electrical problem possibly that had affected not only the QAR, but possibly the engines as well.
The 777 is one of the latest advances to computer-assisted airplanes.
Pilots don't directly control the plane.
Instead, their inputs are sent to a computer.
Those signals are then relayed to the engines, flaps and other systems on the aircraft.
Pete, I can't get power on the engines.
A massive failure of the plane's information systems could explain why the engines stopped working.
Perhaps the computers that control them froze.
Farnborough is home to the AAIB.
The Flight Data Recorder has been sent here for analysis.
An electrical problem would be recorded on this device.
But interpreting the information will take some time - time investigators don't have.
There are hundreds of 777s landing every day.
They need to find answers before whatever brought down Flight 38 strikes again.
While they wait, authorities turn their attention to the fuel recovered from the ruined plane.
A bad batch of fuel could have deprived the aircraft of power just when it needed it most.
Fuel can be contaminated in many different ways.
It could be particle contamination, it could be biological contamination and they can get into filters to restrict the flow.
We took several samples from various points within the fuel system, so not only from the fuel tank itself, but also from remnants within fuel lines and also within various fuel components.
We managed to trace the fuel back through the documentation to a shipment of jet A1 which came from South Korea, and that was shipped across in a tanker to China and was then transported by pipeline to Beijing Airport.
We compared it with over 1,200 other batches of fuel in the UK.
And it came out to be very good.
Investigators next consider the possibility of a blockage in the fuel tanks.
They discover some bizarre evidence.
Several small pieces of tape are recovered, along with something else.
When we were in the fuel tanks, we did find a few small articles.
One of them was a red scraper.
The items were likely left over from when the plane was built seven years ago.
But while they're peculiar, there's no indication that any of them played a role in bringing down the plane.
Obviously, it's not ideal to find anything within a fuel tank.
We obviously looked to see whether this would have had an effect on the fuel flow to the engines.
The items were so small, they would not have caused a restriction.
As the search for answers continues, a puzzling find surfaces.
Less than three years before the Heathrow crash, another 777 suffered a serious midair problem.
After taking off in Perth, Australia, a Malaysian passenger jet was climbing through 38,000 feet when suddenly, the plane's autopilot pitched the nose up and climbed steeply.
(BEEPS) Calamity was only averted when the pilot took manual control of the jet.
They did manage to get it safely back to base.
The investigators established that this was a computer problem.
British aviation authorities consider that the electronic brain of one of the world's most reliable jets might be faulty.
The Malaysia Airlines aircraft was really quite an unnerving one.
The investigators sort of pulled out the files on it and started looking over it to try and get some clues.
Did we have a problem with the electronic control system, causing the engines to roll back? If flaws are found with Flight 38's computers, it means 777s around the world could all be disasters in the making.
This thought hangs heavily over the industry.
When the readout of the plane's flight data recorder is conducted in Farnborough, investigators learn that the crash could have been much worse.
There was very little they could actually have done.
It looks like we have double engine failure.
There was only 30 seconds between when they were fully aware of what was going on Airspeed low! .
.
and the accident was inevitable.
Records show that, soon after the engines failed, the plane started falling faster than 1,800 feet a minute - a very steep descent.
But lack of power was only one factor working against the crew.
They had extended their flaps on descent.
It helps control the plane at lower speeds by increasing lift, but the extended flaps create drag, which slows the plane down.
It takes more power to keep the plane flying with a wider wing.
Airspeed low.
I needed to raise the flaps.
I knew that raising it one notch to flap 25, would be the right thing to do.
To reduce the drag, that's what I had to do.
What he did here .
.
gave them a few extra feet.
Had the flaps been left at flap 30, the aircraft would have struck the ground just before a set of ILS antennas.
The plane then would have slammed into the antennas, sustaining even more damage and increasing the chances of serious injuries.
For their efforts, Burkill and his team are considered heroes.
Flying is about teamwork and we had an outstanding team.
There is no doubt without the flight crew's actions, Flight 38's landing would have been a catastrophe.
Recordings As they continue examining data from the flight data recorders .
.
continue right to the moment of impact.
.
.
investigators are no closer to solving the case.
Analysis of the flight recorder didn't identify any particular fault with the aircraft electrical system that would have resulted in both engines rolling back.
They are at least able to determine why no data was found on the quick access recorder for the last 45 seconds of the flight.
It's not an electrical fault, as they first feared.
The 45-second delay in the QAR recording was a result of the fact that the QAR doesn't record in real time.
It buffers the data and then will record approximately 45 seconds later.
With all the information from the plane's various data recorders now collected, investigators have a precise picture of what happened.
The jet was on the proper course as it made its approach to land when things went horribly wrong.
At 720 feet, the right engine rolled back.
Seven seconds later, the left engine rolled back.
The autothrottle attempted to command greater thrust.
It's not me.
That's the autothrottle.
But the engines didn't respond.
I can't get power on the engines.
While electrical and computer problems are ruled out as causes of the crash, the flight data recorders do point investigators to the likely source of the problem.
The two most significant parameters from the quick access recorder were the fuel-metering valves.
Both those valves indicated that they had opened to their fully open position.
The plane's computers were calling for as much fuel from the tanks as possible.
The valves were fully opened.
But not nearly enough fuel was reaching the engines.
A plane with plenty of fuel and a working computer system still crash-landed and investigators can't determine why.
They focus on the weather on the day of the accident.
BURKILL: It was a very cold air mass over Siberia, so we were aware of the cold conditions, but the flight was expected to be smooth and also quick.
The plane's path took them high over Russia.
The outside temperature went as low as -74 degrees Centigrade.
It's a potentially dangerous temperature if not monitored closely.
In cold air masses, you have to be very aware of the fuel temperatures.
SLEIGHT: When fuel starts to freeze, it produces wax.
As the fuel temperature drops, the wax becomes greater within the fuel itself until it comes to a point where it can no longer flow.
BURKILL: We have a fuel temperature gauge on the flight deck, which is monitored at all times, and I personally look at that every time I do a fuel check, which, minimum, is once an hour.
Basically, you don't want to get it below -34 degrees Centigrade.
While they were cold, temperatures never dipped into the danger zone where the fuel could wax and clog the system.
And the fuel temperature increased significantly as the jet approached London.
We were coming through 20,000 feet and I remember seeing -20 degrees Centigrade, so it had already started to warm up.
Another potential cause of the crash is ruled out.
Investigators are now running out of possible explanations.
And the expectation from the media, and also from aviation experts, was that we would have the answer within days.
REPORTER: But there was no comment today about the ongoing investigation.
It very quickly became apparent that that would not be the case.
They had the pilots, they had the passengers, they had the aircraft, they had everything there .
.
except the thing that caused it.
That had disappeared, as if by magic.
Despite mountains of data and physical evidence at their disposal, investigators are still unable to say what caused the crash of a British Airways 777 at Heathrow Airport.
They turn to a failure of the fuel delivery system as a culprit.
To prove it was responsible, they need to pinpoint anything out of the ordinary about Flight 38's journey on 17 January, 2008.
It's a daily flight operating from Beijing to Heathrow.
We wanted to work out what was unique about this flight.
Why this flight in particular? If you've got a very puzzling incident, like the 777 accident at Heathrow, you're going to mine all the data you can.
Justput them over there for now.
So we approached many operators to obtain some data of previous flights so that we could then do a comparison.
Flight data is collected from companies around the world.
And we had something like 144,000 flights in all.
But it will take months of work to analyse this enormous volume of information.
To study Flight 38's fuel system more closely .
.
huge sections of it are brought to the AAIB hangar and reassembled.
McDERMID: It allowed us to basically just sit there and look at it and to brainstorm and come up with ideas as to what might have caused it.
McDermid and his team consider all possibilities, from design flaws to a malfunction specific to this aircraft.
The examination comes up empty.
We carried out an exhaustive and very thorough inspection and tests of the fuel system and we found nothing wrong.
Frustrated, investigators are at a dead end.
SLEIGHT: We kept going through the cycle.
We kept going back over what we'd done before.
Had we missed something? And every time we went back over, we kept coming back tono.
There was no trouble with the plane's computers.
It had plenty of fuel.
And no problems with the fuel itself.
But somehow, when it was most vital Mayday! .
.
the engines didn't get the fuel they needed.
The AAIB detectives have run out of suspects.
There was no clear explanation.
Nothing that you'd seen before It's not getting the power.
.
.
in any aeroplane, let alone the 777, could have explained this.
Now they turn away from the clues they have to those they don't.
Phil started comin' out with a mantra which was a quote from Sherlock Holmes.
Whatever is left, however improbable .
.
must be the cause.
Has to be.
In this case, we always kept coming back to what's there - what could've been there that's not there now.
And we come back to ice.
Suspicion that the cause of the accident was ice grew stronger and stronger.
But investigators are faced with a nearly impossible task.
How can they prove that ice had brought down the plane when the evidence they need would have melted by the time they arrived? Three months after the crash, Brian McDermid flies to Seattle to work with engineers at Boeing, the company that built the 777.
If ice brought down Flight 38, they need to prove how and why.
We took the pipes from the right side of the fuel system and those pipes and those couplings were used on the test rig at Boeing.
Ice in the fuel lines has long been a concern for jets of all kinds.
But decades ago, engineers found an ingenious way to deal with the ice forming in the fuel of passenger jets.
The heart of the system is the fuel oil heat exchanger, or FOHE.
Cold fuel runs through thin tubes, which are surrounded by the hot oil used to lubricate the engines.
The purpose of a fuel heater is to heat the fuel up to prevent icing of the delicate fuel control system.
The FOHE is designed to prevent exactly the problem investigators suspect crippled Flight 38.
But it's also a potential bottleneck in the system.
McDERMID: During the fuel testing, we were looking to establish where we could get a restriction that would restrict the fuel flow to 6,000 pounds per hour .
.
and the only place that we could get such a restriction occurring was on the face of the fuel oil heat exchanger.
Investigators have to prove three things - first, that substantial amounts of ice can accumulate inside the fuel pipes.
Second, that the ice can suddenly be released.
And, finally, prove that it can block a device specifically designed to melt it.
McDERMID: We used different methods of controlling the environment around the pipes that ranged from dry ice to cold fuel and to hot air to simulate the environment around the pipes that the aircraft experienced during the flight.
After each fuel-cooling test, investigators look inside the pipes for ice.
But they can never get much ice to form, let alone block the fuel lines.
The fuel temperature was -30 degrees Centigrade.
Very little ice would stick to the inside of the fuel pipes.
Investigators know the temperature on the accident flight didn't get much colder, so how had the ice formed? We still hadn't answered the question of where does the ice come from in the first place? And, indeed, was it ice? Or was there something else that we've missed? Surprisingly, when the fuel temperatures are warmer, the breakthrough arrives.
If you have water in the fuel, then that water will freeze and form ice crystals and then when the temperature gets to about -20 degrees Centigrade, those ice crystals will start to stick together and also stick to the inside of the pipes.
Below -20, any ice crystals in the fuel are too cold to attach to the pipes.
But in the sticky range, between -20 and -8 Centigrade, slushy ice forms and sticks to the sides of the fuel pipes.
So, when we demonstrated that ice could actually build up and grow onto the walls of the fuel pipes, there was a certain amount of surprise as to how much could actually grow.
Now we're gettin' somewhere.
Well, the question for us at this stage was - how could that ice come off the pipes? Investigators think they have the smoking gun.
But they still can't prove how a small bit of ice could bring down a 200 million plane.
They continue analysing thousands of comparative flights to see what made Flight 38 unique.
From 35,000 Rolls-Royce-powered flights, it was less than 1% that had the same features as the accident flight.
After seven months, they still can't crack the case.
On November 26, 2008, more than 10 months after the crash, the importance of finding an answer is underscored when the elusive culprit strikes again.
Another 777 runs into trouble 39,000 feet above the United States .
.
when one of its engines simply stops working.
Fortunately, in that case, the engine recovered and the aircraft landed safely in Atlanta.
Even though it didn't result in an accident, it reinforces concerns that there's a potentially dangerous flaw on every 777 around the world.
The incident captures the attention of the British Airways crash investigators.
LEARMOUNT: The Delta Shanghai flight really was investigated with huge interest.
The engines were Rolls-Royce engines of the same type, so they immediately went looking for evidence of the same problem.
And, in fact, they do find important similarities between the Delta airliner and British Airways Flight 38.
The aircrafts had operated a long sector from Shanghai, China to Atlanta.
The fuel temperatures were within the sticky range that we had defined during the research.
We were able to see that the reaction of the engine was very similar to that of the 777 at Heathrow.
Later examination of the Delta flight shows no evidence of any electronic, mechanical or fuel system problems .
.
strongly suggesting that the culprit was, indeed, ice.
When we heard about the Delta flight, we were more convinced than ever that we were on the right track.
McDermid and the investigators at Boeing in Seattle continue their testing.
It's been a year since the accident and the problem hasn't been found or fixed.
They've run hundreds of simulations and spent millions of dollars.
They still can't get the ice in the fuel lines to clog the fuel oil heat exchanger and cause the kind of blockage that brought down Flight 38.
We came to a stage where we had not actually come up with what had caused this accident.
This time, the media and also the general public, there was a bit of unrest as to why we had not yet come up with an answer.
Industry pressure to resolve this case is intensifying.
Investigators still need to know what made Flight 38 different from thousands of others that had flown under similar conditions.
To find out, they retrace every moment of Flight 38's journey.
The voyage from Beijing involved a gradual climb, a steady cruise and a gradual descent.
The fuel temperatures fell and rose accordingly, causing ice to accumulate in the fuel lines.
That ice posed no danger unless it was released.
Now, investigators study the fuel flow and find that it was kept steady for most of the flight.
The autopilot maintained constant speed and low power for hours, never demanding an abrupt increase in engine power until just before they reached the runway.
The approach into Heathrow was quite a turbulent approach .
.
and the engines were demanding various levels of thrust power.
That's not me.
That's the auto throttles doing that.
And there were four changes in fuel flow, one of which was a very high fuel flow of around about 12-13,000 pounds per hour demanded.
A closer review of the American Delta flight reveals that its engine rolled back under the same circumstances - a sudden demand for power after a long period of consistent speed.
(ALARM BLARES) Investigators are now set to duplicate these precise conditions.
They hope that by reproducing the moments just before landing, which was the only part of the flight where the crew suddenly required more power, they may finally get the answer they're looking for.
The problem that we were looking at is very, very difficult to replicate, and during what was one of the last tests, we actually managed to get all the elements together.
After simulating running engines at constant speed, investigators increase the power.
We allowed the ice to accumulate for three hours, and then the flow rate was increased.
What happens next breaks the case wide open.
In Boeing's Seattle lab, investigators' sudden demand for more engine power causes the fuel pressure downstream of the fuel oil heat exchanger to drop dramatically.
And when investigators examine this vital component MCDERMID: We then saw that ice had formed across the face of it.
.
.
soft ice that had formed deep inside the fuel lines broke free when the pressure was abruptly increased.
It restricted the flow of fuel to the engines.
Investigators have finally found the Achilles heel hidden deep inside the 777.
The tubes that bring the fuel through the FOHE jut out just above the container of hot oil.
It's just a few millimetres, but was enough to prevent the ice from coming in contact with the hot surface below and melting.
Now, the ice doesn't totally restrict the fuel flowing through it.
It has some porosity.
So the fuel will continue to flow through but the flow rate is much slower than it should be.
In all of the tests, only one gave investigators the result they were looking for - the one that exactly matched the journey of Flight 38.
While the plane flew over Russia, water in the fuel turned to ice.
At -20, it began to build up along the inside of the pipes.
The steady speed of the aircraft ensured this accumulation was never interrupted.
Then, as the plane approached Heathrow, turbulence resulted in the first demand for power since much earlier in the flight and things cascaded into a serious problem.
You may want to keep the autopilot on a little longer than usual.
The gushing fuel washed the ice through the fuel system until it built up against the face of the FOHE - with disastrous results.
Pete, I can't get power on the engines.
It's not giving me power.
Facing a threat no-one knew existed Mayday! Speedbird! .
.
the pilots didn't have a chance to solve the problem.
(ALARM BLARES) But why had the American Delta crew been able to clear the blockage while the British Airways flight ended in near-catastrophe? It's discovered that after just a few seconds of reducing engine power to idle .
.
the blockage in the FOHE clears.
But this manoeuvre was not available to the British Airways pilots because they were so close to the ground.
Soon after the Seattle tests, Rolls-Royce redesigned the fuel oil heat exchanger.
So what Rolls-Royce did was to actually remove those protruding tubes so you now have a flush face.
And they found that if they flattened the surface, even if ice crystals did form in the fuel, they'd go straight down the holes in the block.
Very, very easy fix.
Other aircraft manufacturers are ordered to ensure that their systems aren't vulnerable to the same problem.
Peter Burkill, John Coward and the rest of the crew of Flight 38 received the British Airways Safety Medal for their performance during the accident.
It's the company's highest honour.
The crew had been presented with an unprecedented failure and they did the best they could in the time that they had available.
The crew did as good a job as they could, and since they really only had about 30 seconds to think about what they were going to do about this, they couldn't have done any better than they did.
The ice that brought down one of the world's most sophisticated airplanes was gone by the time investigators showed up.
The weird thing about this investigation was that the culprit had fled the scene.
The inquiry into what happened consumed thousands of hours of manpower and cost millions of dollars.
These people really pulled out the stops to find out, and the reason, we have to know.
With unrelenting diligence .
.
and a little inspiration from Sherlock Holmes, the mystery was finally solved.
SLEIGHT: Whatever is left, however improbable .
.
must be the cause.