All evidential avenues being exhausted, we're resorting to arcane guesswork; our primary guess being specifically contradicted by flight data evidence.

We're all doomed!

The one with the parachute backpack accessory, please.

if there is evidence of low pressure damage on the engine fuel pumps, would that not indicate that the Rollers were trying to do what they should be, and throttling up?

Are there pumps to push the fuel from the tanks as well as to suck it into the engine, or does it rely on the drop in pressure along the fuel lines sucking more fuel from the tanks?

If there is only one set of pumps - at the engine end - then it would take time for the increased demand to be felt at the tanks, plus the same length of time for the increased fuel flow to reach the engine. If there are multiple sets of pumps, do any of them show signs of increased pressure due to the fuel getting "stuck" partway through the system, or does the 777 flight data recorder collect/retain pump/pipe pressure data and fuel flow at different points?

Are there any (licenced!) Boeing 777 engineers in ElReg lan who could tell us?

(Paris, cos there's no LiLo icon!)

<quote>

If there are multiple sets of pumps, do any of them show signs of increased pressure due to the fuel getting "stuck" partway through the system

<\quote>

Yes, the evidence was in the article, cavitation damage could be an indication of an over pressure situation. If you block/restrict the outlet of a pump then it can cavitate. Cavitation can cause damage to the impeller blades of the pump. (Depending on the type of pump used!)

Is reduced pressure "Euphemism" for lack of fuel thereof lack of thrust Blah Blah Blah

I would want to know if the cavitation damage could happen in 2 minutes or 2 hours. They took the engines off and stuck them on a test rig it seems - did they do the same for the control systems, I wonder.

Each engine is serviced by a separate pump; both show signs of damage from low inlet pressure. Therefore, further down the line, there was a restriction. The preliminary reports show that all valves, pumps, and plumbing are good. While the report does say that there isn't excessive water in the fuel, I'd suspect that there was just enough (condensation in the partially-empty tanks at some point?) for a little ball of ice to form and cause a temporary restriction(s) in the fuel flow. While the fuel temp wasn't low enough for the fuel to freeze, it was certainly low enough for any water to crystallize. If there was a momentary vacuum in the fuel line, the lower pressure would actually help it freeze.

Mine's the one with the parachute and the can of isopropyl alcohol in the pocket...

Vortex in the fuel tank?

Whirly pool spirraling down to the pump inlet?

any other guesses?

Perhaps we should form some sort of club?

I'm not an aviation engineer but...

There's a recirculating system to keep the fuel in motion and well mixed, which also helps reduce the effects of water ice in the fuel. There are various low-pressure pumps that feed the high-pressure pumps mentioned in the article, so the normal situation is that there's enough fuel being pushed from whichever tank is being used (normally on approach the engines would be fed independently from the adjacent wing tank, hence the trouble in explaining something that clobbered both engines). In the event of low-pressure pump failure, the suction of the high-pressure pump plus some bypass valves on the low-pressure pump should keep the system functioning well enough.

All the obvious answers have been examined so yes, they're onto guesswork. There have been several incidents of single-engine failures of this type (but not necessarily the same cause), but this is obviously a lot more serious. It may take them some time, but they've had fair warning and next time it happens they might not be so lucky. At the moment I'm sticking to non-777 flights.

So we are left with the only possible explanation - the almost simultaneous failure of separate fuel delivery systems on both engines caused by ice blockages. Or possibly an excess of monosodium glutamate in the fuel.

The real shame of this whole sorry affair is that the plane managed to avoid crashing into either Terminal 5 or Mr Gordon Brown. A disaster!

If the fuel freezes at -47°C and the minimum recorded fuel temperature was -34°C then the fuel did not freeze over Siberia. So frozen fuel is not the problem unless the air temperature at Heathrow was a little colder than our boys at the Met Office are telling us.....

Paris, because she *is* hotter than Siberia. (well a little anyway)

And I think the emphasis needs to be not on what happens when the system gets cold, but what happens when the system warms up again; like when you descend from 33,000ft to 720ft. Things can either melt or boil when you warm them up. The pressure increases as you descend, too. I presume that somebody has already looked in detail at the properties of various compounds (water, kerosene, water/kerosene mix, liquid nitrogen, liquid oxygen, fuel additives, cleaning fluids, CO2) with varying temperature and pressure? Phase transitions can also be exothermic or endothermic. If the compounds are particularly clean (condensed water vapour?), with nothing to nucleate on, you can also get super cooling or super heating, meaning that the phase transition occurs much later than you would expect, and happens very suddenly.

Not an Engineer.....

Why not just say:

We have no idea why the planes engines did what they did. However we are still trying to figure out what the fuck went wrong.

Simple and too the point and makes more sense that the 'report'

/mines the one with Simpleton written on it.

Cavitation? Paris? Joke? Anyone?

Oh go on, you know you want to...

Because they earn significantly more than you and take the matter seriously?

Interesting to note that during the final approach phase the aircraft attitude would probably be a little diffrerent to earlier in the flight (nose higher usually), which might have cause something to move in the tanks and ...

If you want some informed comment along with all the wild speculation about the cause, try pprune.org

I was lucky to be trained as a flight engineer many years ago.

We we taught to suspect everything if we had a fault in our aircraft.

We flew with it for up to 14 hrs on maritime reconnaisance out over the north Atlantic. It was MY responsibility to see we got where the navigator took us and get it back!!.

no mean task for a 20 yr old.

So, I still follow aviation news and this incident made me think, what set of conditions could cause fuel starvation after this uneventful flight?.

well, based on the facts as first reported by the AAIB, they flew through an area of extreme cold, this on its own isnt a problem with fuel flow to the engines as the fuel would be at a much higher temperature than the outside air, and even if the airframe in the wings was reduced to -70c the fuel would keep the pipe temperature up above the freezing point of any water in the fuel or the fuel itself.

BUT what happens when you throttle back to make the rapid decent from 40.000ft over the North sea? your fuel flow will drop very considerably in an airframe that still might be -70c.

Any water in the fuel will then have the chance to be cooled to below freezing inside the fuel pipes and thus restrict the fuel flow, as well as the possibility of the fuel freezing as well.

This doesnt matter when the engines are in descent mode, but the moment you want power to correct the landing glide path then opening the throttles to call for more fuel, wont happen, because the fuel flow may well have been restricted by ice or fuel slush like diesel fuel in winter..

By the time the AAIB got to work on the airframe it would have regained its normal temperature and the ice/fuel would have melted.

Especially as there was a fuel leak and any water residue may have spilled out.

Now they say the ambient air temp wasnt as low as -70c but much less.

So my initial thoughts now seem unfounded.

I commented to aviation frends one would need to replicate fuel flows in the temperature gradients they experienced to see if the fault could be duplicated.

Just an old engineer not a bold engineer.

Ted Frater

269 Sqn RAF.

Coastal command.

>>If there was a momentary vacuum in the fuel line, the lower pressure would actually help it freeze.

I thought low pressure helped things BOIL. If I remember my chemistry correctly, INCREASED pressure helps liquefy or freeze things, not decreased pressure...

Keep going guys and we're sure to beat the experts to the answer.

Has William Shatner been interviewed yet?

Ok this is pure speculation but jets have been flying over Canada & Siberia in winter for over 40 years in those kind of air temperatures with no obvious problem (OK 777's are fairly new). What gives?

I'll have the Full Monty please (parachute pack, thermal suit, furry boots and the DIY igloo kit)

that man wasn't supposed to fly?

And that all efforts to MAKE him do so entail some degree of risk, and that we WILL always have accidents even with well-designed systems and proceedures?

The 777 has been a remarkably safe and reliable airplane for over a decade. It has flown all over the world, in all kinds of temeratures. It represents some of the very best engineering work done in commercial aerospace. (No, I do not work for Boeing...) But I would happily fly one tommorrow...

It was a freak occurance. It happens, and it will always happen. Even birds who are meant to fly run into plate glass windows sometimes...

Do as the french do, and just shrug your shoulders and say 'Meh', or in English - 'iunno'

Nothing beats experience :)

Ted, that's by far the best analogy I've heard yet, I do hope it's sufficiently obvious to AAIB and RR and Boeing too.

I used to be an engineer, when I worked for the company that makes the control systems in this picture but that was a long time ago (probably about the same time the control system in this picture was designed... hmmm).

So who needs the fake so called time consuming liquid explosives with buckets of ice needed to create them , when simple software time bombs in the form of glitches and bugs can kill the modern fly by wire computer controlled aircraft !

Sad really , when you think about it , the best brains in the business and they are all still scratching their posteriors looking for a cause , that could be so simple and obvious that it may be staring them in the face or was it that mystery passenger in seat 57 turning on his Mac Book Air at that critical juncture in time ?

As one would say so many questions were answered , the one they wanted still eludes them all !

Some things can be very elusive indeed , when illogical thinking could solve the problem easily !

"I thought low pressure helped things BOIL. If I remember my chemistry correctly, INCREASED pressure helps liquefy or freeze things, not decreased pressure..."

Please do not confused "boiling" (generally recognised as an increase in heat in the liquid) and "vaporisation" (the turning of a liquid into gas).

Dropping pressure lowers the boiling temperature of a liquid - in other words, the liquid vaporises at a lower temperature, which is why mountain climbers have to use pressure cookers to cook anything properly.

Increasing pressure actually *increases* the temperature of the liquid - said increase in heat must be lost before the liquid will crystalise.

For an application of this principle, look no further than your fridge or air-conditioner: The liquid, upon leaving your house/fridge, enters a compression system. The compressed liquid (now hotter than it was) is fed through a series of radiators which allow the liquid to cool. This cooler liquid is now piped back into the house/fridge where it is then allowed to "decompress", dropping in temperature. A second set of radiators allow the liquid to "pick up heat" from its environment (house/fridge), at which point it is then piped back out to the compressor and the cycle begins again.

Lower pressure will indeed help things boil, if the pressure gets down to almost a vacuum. However it will also tend to remove heat from the substance in question (think of the cooling effect when gas is released from a spray can, or the instantly super cooled vapor when you open a beer), which (and I admit that I'm pulling this out of my ass) I think would tend to have a more immediate effect.

Two engines, two fuel pumps and two tanks of fuel.

Simultaneous fuel starvation?

It may not be impossible, but I doubt it very much. I'm thinking some subtle design fault. I'd like to know how long the pumps had suffered cavitation though. Perhaps the fuel lines are too skinny?

Because in China, making a bit extra justifies the means.

"Has William Shatner been interviewed yet?"

The engines cannae take it Jim.

No matter how much fuel you pump onboard, your plane will be starved a few hours later....

When this happens on a motorcycle -- particularly an old fashioned one -- then the culprit is the tiny vent hole in the tank filler cap that gets blocked. The result is the same -- you may potter on for ages at low throttle setting and everything's just fine, but then you open it up and it does a few seconds and then sputters back to low power.

The hallmark of a true engineer is to pull the entire fuel system and engine apart to analyze the malfunction. (Hint....been there, done that.....) The humiliation that results from someone sticking a paper clip into the vent hole to clean it is indescribable...

<quote>

Yes, the evidence was in the article, cavitation damage could be an indication of an over pressure situation. If you block/restrict the outlet of a pump then it can cavitate. Cavitation can cause damage to the impeller blades of the pump. (Depending on the type of pump used!)

</quote>

Correction is necessary

Cavitation damage results from the collapse of vapour bubbles in the impeller. Restriction in the outlet side causes low flow and high pressure. Restriction to the inlet reduces pressure causing the fluid to vaporise resulting in cavitation.

The important point about the discovery of cavitation is that it indicates the pump was trying to pump vapour not fluid and hence there was fuel starvation outside the engine core.

Here endeth the engineering lesson.

Grumpy Old Git

Given that the 777 is entirely managed by computerised systems it seems strange that nobody has yet suggested that the equivalent of an undeterminable BSOD in the engine management system might have been to blame; or is that prospect a little too terrifying to consider?

er, both engine management units correctly moved the fuel inlet valves to the position requested by the flight control systems indicating that the FCS and ECS were both functioning as expected, thus all the speculation about some sort of restriction to the fuel flow to the high pressure pumps being the cause.

Do try to keep up, 007.

I've pumped round liquids that can burst into flame when the pumps are doing the wrong thing (whisky pumping into a dead end can cause the pump to overheat and it all goes bang).

Their system didn't check for cavitation? Very poor design. You either need a system that checks for low pressure so you can do something about it or get pumps that don't get damaged when sucking on a blocked pipe.

The AAIB's further interim report into the prospective cause of Boeing 777 BA038 to crash land at Heathrow in January includes the comment: 'There is no evidence of any anomalous behaviour of any of the aircraft or engine systems that suggests electromagnetic interference.' But the elephant in the room for this incident, that makes this incident so different to probably every other time a 777 has ever come in to land at any airport in the world, is what? It is that the aircraft flew directly over the motorcade conveying the UK PM.

The roadway that the PM's motorcade was travelling on, the A30, is on the Eastern perimeter to the airport running along an acute angle to the glide path flight BA038 was cruising in on and passes close to the end of the runway the aircraft was heading for. These high security-risk road convoys are know to have specialist vehicles included within them that make multiple frequency transmissions designed to block the use of mobile phones in the vicinity along with other kinds of transmissions that could also be used to control & trigger devices such as a roadside bomb. In addition, over the past ten or more years, a great deal of research has been under-way to develop High Powered Microwave (HPM) and Electro Magnetic Pulse (EMP) systems designed sense/detect via radar various kinds of incoming missiles and attack the electronic systems of the detected threat in order to disrupt its control and navigation systems causing failure to reach it's target.

Aviation week - E-10 Radar Secretly Designed To Jam Missiles - May 29, 2005 - "You don't need to burn out the electronics to make a missile go off course," says the Pentagon radar specialist. A large pulse of energy can affect the sensor's infrared focal plane array or the processors that help guide the missile and identify its targets.

If such a system is now operational and deployed within Mr Brown's motorcade it is probable it could have detected the 777's approach and wrongly deducted it to be a threat since the aircrafts flight path was on a direct course to intersect with the PM's convoy - as is reported to have occurred; the plane flew directly overhead of Mr Brown's vehicle so close that the Daily Mail initially reported that the security team thought that they were under attack.

"The Boeing 777 was so close to the Prime Minister's car, on a perimeter road, that his security team at first feared a terror plot was unfolding, with Mr Brown as the target." Daily Mail 18th Jan 08

So on the balance of probabilities what is more likely the cause of the systems failures that cause both engines to independently fail? An unprecedented double occurrence of ice formation blocking the fuel feed (what was so special about this 777 flight to have caused that to occur) or the accidental deployment of a new technology defence system (and we know the unique event of this flight was it's heading towards & passing over the PM's motorcade) causing perhaps the electronic control system controlling the pumps feeding each engines fuel system to fail - both pump's systems falling venerable to the EMP's effect at slightly different moments.

Given the choice I would rather take a flight in a 777 than come in to land over the PM's motorcade.

The 777 has three, independently written (no communication between the teams) computer systems keeping it in the air. The redendant masters take advice from these and, if there is a disagreement or overly-long delay in presenting the data, discard the suspect input. Scream about the fact, and carry on flying the plane with one unit disabled.

The subsystems themselves (pumps and so on) are all instrumented so that their "I'm doing <foo> at <bar> RPM" or "I'm deflected at <baz> degrees" is independently verified by systems built specifically to detect control system errors.

Far from being terrifying, the equivalent of a BSOD on a modern airliner is expected and designed for.

Wave in a friendly manner to any Kondor you see going pass !! They're doing a job, same as you are !!

@Hunde Kotze - Yet another moronic China-basher who just refuses to see the facts even after they have been presented in the report !! Go and pray hard in your Creation church !! God might even grant you an extra brain cell if you try hard enough !!

somebody has not blamed microshaft yet..... it has to be bills fault doesn't it?

Ted Frater: "what happens when you throttle back to make the rapid decent from 40.000ft over the North sea"... nice to have your informed and intellectual approach. This is the same question I would be asking.

AC @ 04:30GMT: "No matter how much fuel you pump onboard, your plane will be starved a few hours later..." I understand that all flights aim to carry just enough fuel to complete their flight with a percentage (oh oh - statistics again) for error, delays, weather and possible stacking. So the fuel might well have been of perfect quality, just insufficient volume - no mention of that that I can find. True words are often spoken in Jest.

Marty: "Its a wonder" - check AC @ 09:16GMT.