There are new procedures in place for flight crew during flight and engineering prior to flight, to ensure fuel moves!

Smilin_Ed posted:-
I have also read all posts on this and other disasters - though I must enter pleading at being most confused at times by disjointed thread creep - I think I've stayed on track and the one overiding factor common to all is that despite many "experts" on aircraft systems and engineering there seems to be a plethora of "conclusively proved" items that turn out to be absolutely wrong and "impossibilities" that turn out to be anything but.

I am reminded of the expert poster with years of experience in oxygen bottles saying that one had never let go in an aircraft and people suggesting they should be sectioned - yet a large part of one had come through the floor to embed itself in the doorframe - visible to all who looked. We also had explosive experts saying that the viewable photos were not of an explosion and other guys confidently pronouncing metal fatigue - all wrong.

In this particular case it certainly seems that the fuel pumps were fighting against something but ice to me is just a convenient answer when nothing else seems to fit. Just how easy is it for some valves to be closed against the pumps? Does anyone know ? Of course, if the actual temperature of the fuel was known we wouldn't be arguing.

FlyGooseFly

In threads like this you have to read much deeper than individual posts. While I'll admit that the obvious hunches do vary and are discarded along the way as available evidence surfaces, only as ignorance would one profess a conclusively proven point before the final report of facts and analysis is presented. In almost al cases, portions of the postulations of causes have arguments of pro and cons in their consideration, and it is those arguments that become the basis for the support of readership of these threads.

I suggest that you do not assign words like "conclusively proven", "absolutely wrong" or "impossibility" to postulations within these type of discussion threads.

Photographic evidence posted earlier shows the fuel was cold enough to cause frost on the lower wing surface.

Rightbase,
That's an everyday occurrence following high alt coldsoak as is also frost on upper surface.

Frost on the lower wing surface.

In 30 years or more it is only on very short sectors (i.e. less than 3 hours) that I haven't experienced that. With TAT's on the nippy side of comfortable what can anybody expect?


Regards
Exeng

Perhaps an engineer could confirm this....
There is a difference between the Trent and the GE90 in terms of fuel piping layout. If I interpret correctly the various photos I have googled the GE engine has all the fuel pipery and pumps around the core. On the Trent the fuel starts outside the fan casing and then goes through pipes which pass through the fan airflow on their way to the core.
True or false?

The Trent certainly has some fuel system components mounted on the outside of the fan case. I'm not certain but afaik the GE90 has all the system mounted on the core.

However, both engines have to get the fuel through the fan airflow. The only difference is which components are outside and which are inside.

Where Is The Fuel Heater?
 

I had envisioned the fuel heater in the wing. From reading this, I gather that it is in the engine nacelle. It looks to me like they needed one in the wing before the pipes enter the engine pylon.

There are hydraulic oil-to-fuel heat exchangers.
- 1 heat exchanger in the left main tank for the left hydraulic system,
- 2 heat exchangers in the right main tank for the center and right hydraulic systems.

Through the heat exchangers fuel cools the hydraulic oil but due to the large volume of fuel, heating of the fuel is negligeble.

Regards,
Green-dot

Same applies to the Fuel Oil Heat Exchanger (also known as a fuel-cooled oil cooler). It is controlled only on oil temperature. Any heating of the fuel is co-incidental and uncontrolled.

something else....
 

sorry, gentlemen and ladies,

there's abit about fuel management...centre tank runs to x -thousand
lbs./kgs and then stays off until main tanks reduce to y-thousand, then
centre tank pumps run again. could this have been shortly before the approach commenced ? possibly bringing very cold fuel into the situation?

been following thread for some time....

well done all crew, especially the f/o!

Not quite, to be more precise the centre tank override jettison (OJ) pumps are switched off manually at approx. 800kg of fuel remaining in the centre tank. Float-operated shutoff valves prevent fuel scavenge when the main tanks are full. When the main tanks reduce to y-thousand the float-operated shutoff valves open and the fuel scavenge jet pumps will start scavenging (operated by motive flow from the boost pumps) the remaining fuel from the centre tank. The centre tank OJ pumps remain switched off.

Centre tank fuel is relatively warmer than main tank fuel because most of the centre tank is situated above the airconditioning system which radiates heat, it is between forward and aft cargo compartments, and the passenger cabin is above it. Hence, main tank fuel is relatively colder than centre tank fuel.


Green-dot

Was any impact damage to the impeller blades found in the wing tank(s) pump(s) ?

To answer one of Sooty655's previous postings:

You are correct in the understanding that Rolls-Royce RB211 and it's Trent derivatives have the Fuel Control System mounted on the outside of the Fan case where General Electric have generally favoured mounting much of it's systems around the core.

There are pros and cons for both design philosophies. Whilst at first glance it seems a good plan to keep all the fuel system nice and warm by keeping it close to the core, the transient extremes of temperature encountered during flight in such circumstances impose design issues regarding valve tolerances/clearances. There can also be problems with fuel laquering of metering and control valves that operate in fuel luricated systems caused by higher fuel temperatures.

To answer your question though, the RB211-Trent 895 (and all it's RB211 predecessors way back to the original -22B variant of the late 60s/70s) have the following mounted on the fan case:

Low Pressure Fuel Pump
Low Pressure Fuel filter
Fuel Oil Heat Exchanger
High Pressure Fuel Pump
Fuel Metering Unit (and it's constituent control system)
Fuel Manifold Drains Tank and it's ejector pump.


As an aside, the Main Engine Fuel pump designed and fitted to the R-R Trent 895 may well have demonstrated evidence of bearing face cavitation upon strip, however on such pump designs there is no escaping cavitation and it is more a case of designing the pump to minimise what cavitation there is and to encourage the cavitation to take place in the least problematic areas of the unit - more akin to 'managing' the cavitation than anything else . It is the degree and agressiveness of the cavitation scarring that tell the story of the unit's unhappy existence up to it's untimely death and from what information I have seen, the unit was in good health up until a short time before the incident.

The manufacturer of the pump are very well respected in the Industry and were selected by Rolls-Royce for designing a product with a good pedigree and having an extremely good record for dealing with cavitation issues. The pump itself has a delivery capacity that far exceeds the requirements of the 95,000 lb thrust variant fitted to the R-R powered BA 777 as it was targeted for a much higher thrust Trent variant that didn't make it into production for reasons other than design viability or reliability. Having been closely involved in the project's design, I can say with a high degree of confidence that even in the pump's post cavitation degraded state I can not imagine a situation where the pump was incapable of delivering the required fuel other than through starvation/blockage.

I realise that I have probably re-iterated some points previously made and for that I apologise.

GFYA

Thanks, GFYA,

How clear and illuminating. 2035 (+) posts on this thread, and many on its predecessor, since 17th January. How forbearing of you not to comment until now.

Guess you must have been occupied elsewhere... Welcome back.

Any comment on the theoretical relative susceptibility (to icing blockage) of large cross-section plumbing - versus smaller?

GFYA

I may have been reading more into this than intended but should I interpret that the unit did suffer some ill health shortly before the accident?

Guys,
Be reasonable, and read GFYA's post properly.
He is saying cavitation damage may be pointing a finger, but is not unique to this case, if I read it properly.

lomapaseo, I agree he should explain what he means with "I have seen, the unit was in good health up until a short time before the incident."

Chris Scott, We're not all frequent PPRuNe readers.....
It's true nobody really answered your post about "large cross-section plumbing" either....which I thought made sense.

CJ

Fascinating, helpful, and VERY THOUGHT PROVOKING.
 

Thank-you GFYA, very informative.

I would highlight SOME of your post, for others who have posted since - although it should be read as a whole :

".............. the Main Engine Fuel pump ........... may well have demonstrated evidence of bearing face cavitation upon strip, however on such pump designs there is no escaping cavitation .......... and from what information I have seen, the unit was in good health up until a short time before the incident.

............ The pump itself has a delivery capacity that far exceeds the requirements of the 95,000 lb thrust variant fitted ............. I can say with a high degree of confidence that even in the pump's post cavitation degraded state I can not imagine a situation where the pump was incapable of delivering the required fuel other than through starvation/blockage."


Very Thought provoking.

.

Hey.My first post here so please go easy.
Just wondering,any particular (minimum) amount fuel quantity in the centre tanks before you manually switch them off? Or is it procedure to just put the pumps off when the "fuel low centre" message pops up?
Thx in advance.

7 posts before you by Green Dot:

Thx.
I just hadnt understood the term "OJ".
Now i think i understand........
I assume these are the same as the centre tank fuel boost pumps.

Flaperon777,
Yes, basically they are centre tank boost pumps but they have a dual function and therefore are officially named Override / Jettison pumps because they:
a. deliver a higher boost pressure than the main tank boost pumps to ensure the centre tank empties before the main tanks (they override the main tank boost pumps which normally also operate while fuel is fed to the engines from the centre tank).
b. they also function as jettison pumps in case a situation develops whereby fuel needs to be dumped.


Green-dot

Can anyone provide a photograph of the fuel cooled oil cooler pipe matrix? because I am having difficulty visualizing the following scenario:

Aircraft takes off with 5 litres of water in tanks (AAIB) IF none passes through engines in normal manner & IF it evenly distributes itself evenly between engines, it means that a maximum of 2.5 litres of frozen water managed to block both oil coolers within seconds of each other (with a bit of uneven distribution, less than 2.5 litres would have caused blockage).

If as has been stated in a previous post, the primary purpose is to cool oil & any fuel heating is a secondary uncontrolled effect, one would have thought that a some sort of fuel bypass woud have been designed in to counter any possibility of ice blockage.

I have no experience of modern airliners but I have 35 years experience on the engineering side of older military aircraft. I am desperately trying to visualize how such a small quantity of water could cause such problems, in fact to tell the truth, I don't believe it.

Regards Flight_Idle

- Chris Scott.

As you can see, I am not a frequent poster on Pprune, more of a watcher. The quiet one on the night out if you will, the one that listens to what others have to say and only infrequently comments with a few 'golden nuggets'. Part of that is a result of how busy this year has been for me and the rest is how frequently I see things get taken out of context on internet forums. That is no reflection on you guys who have remarked since my posting, merely my general perception of such facilities.

WRT your remark regarding icing and pipe diameter, there will of course be a far greater susceptibilty for blockage of narrow Internal diameter fuel pipework. I don't have any facts or figures to present to you I'm afraid, but fuel system icing is taken very seriously in gas turbine design and many components are tested on rigs using fuel that does not have the FSII (Fuel System Icing Inhibitor) additive included.

Having come into this discussion somewhat late, I hope you'll forgive me for not reading all the previous (2035 was it?) postings so I will apologise once more if I touch on issues that have already been addressed.

The RB211-Trent LP fuel filter has a dual purpose, not only will it filter out rubbish/swarf etc that may remain in the fuel tanks after the manufacture of the aircraft or post maintenance, it also filters out ice crystals. Any blockage of the LP fuel filter beyond a critical pressure differential will give rise to the LP fuel filter bypass valve opening and an associated caution and chime in the flightdeck. I am not aware of any such bypass taking place and would have thought that the AAIB report would have declared a filter bypass in their reports.

It is also indicated that there was at least a certain amount of fuel flowing through the pump and metering system because:

a) The engines did not flame out and;

b) The Fuel Metering Valve position is controlled using the fuel pressure delivered by it's associated high pressure fuel pump. By that you may interpret that if there is no fuel flow or pressure that Metering Valve cannot move. In this accident the Fuel Metering Valve operated exactly as per the requirements of the FADEC closed loop software.

In crude terms, the FADEC has a spreadsheet, if you will, of typical Power Lever Angle vs Fuel Metering Valve position vs Engine power output. In this case the Engine power output was significantly less than that expected for the Power Lever Angle. In response to this the FADEC scheduled the Fuel Metering Valve to open further in response to this shortfall. The engines did not respond accordingly.


- Lompaseo.

My comment regarding evidence of ill health before the accident is a result of remarks made in the AAIB report that an unexpected degree of cavitation was found post unit strip at the pump manufacturer's facility. I was not suggesting that there were indications that the pump was about to fail. Far from it, it was a very good pump design. My interpretation of the AAIB's remarks wrt the cavitation found in the subject pumps is that the cavitation scarring is indicative of the pump working in pressure and flow conditions that it was not dsigned for. In this situation, a very low pump outlet pressure.


- Christiaan J:

Thank you for you remarks, may be I should speak up more often?

Absolutely right Sir, my comments about pump cavitation were not suggesting that the degree of cavitation was the cause of the failure, more the symptom.

Regards,

GFYA.

GFYA
Could the cavitation damage be due to the blockage in the fuel lines restricting the amount of fuel to less than demanded, effectively causing the fuel pumps to stall and further reduce fuel flow to the engine? The engines did initially accelerate then fall back.

Sky9,

Yes, blockages do indeed aggravate cavitation in such pump designs, personally I wouldn't have used the term 'stall' though. I must re-iterate the point that cavitation in such gear pump designs is a given. The important factor is the severity of the cavitation scarring. This is influenced by the specifics of the pump design and the environment in which it is operating. Given the wealth of experience the pump manufacturer had, the design was well understood and was a development of a previous model with a very significant back catalogue of service history. It is the environment (ie the local fuel pressure in the pump) that I feel influenced the degree of cavitation scarring.

Regards,

GFYA

The fuel cooled oil cooler (FCOC)

http://i337.photobucket.com/albums/n...OC-cutaway.jpg

is mounted near the top on the fan case right side (black in photo)

http://i337.photobucket.com/albums/n...og/800-79b.jpg

and connected to the gear driven pumps at bottom of engine via a network of tubes:

http://i337.photobucket.com/albums/n...og/800-66b.jpg

Thanks for those photos Machaca, I assume that the oil passes through the capillary pipes & the fuel passes over them?.

I wonder if the oil is pumped through on a constant flow basis, or is there a mechanism for the oil to be pumped through intermittently to keep engine oil temperature within specified limits?

If the latter stationary oil in the FCOC would reach fuel temperature, would oil viscosity be OK at this temperature?, if this was the case no fuel heating would occur.

Thanks again for the excellent photos.

Afaik it's the other way round. The fuel makes a single uncontrolled pass through the tubes. The oil passes around the outside of the tubes, guided by the baffles.

Oil temperature is controlled by a thermostatic valve, which allows the oil to bypass the cooler unless cooling is required.

Didn't the AAIB initial enquiry state that the pump cavitation was consistent with an aircraft of its age and was unrelated? And why would the fuel freeze during the approach, 30 minutes or more after starting descent and the fuel beginning to warm up again? Surely fuel temperatures are at their lowest at the beginning of the descent, not the bottom of it?

Pretty simple stuff.
Any mass of ice forming in the high-level cruise and adhering to a surface would, in a gradually warming environment of a descent, firstly become dislodged,move with the flow and then lodge somewhere downstream in the fuel-flow and cause an obstruction to that flow.

That simple deduction is a direct pointer to the true nature of BA038's problem.

Shadow
 

I don't really see how you could make that 'simple deduction'. How can you come to that conclusion?

Whilst I think I know something about the characteristics of cold fuel and ice formation I have seen nothing from the AAIB releases that specifically points you to your deduction.

It is very difficult to prove anything about this incident because it would seem the evidence melted!


Regards
Exeng

Whippersnapper,
The AAIB's theory is that ice had (probably) adhered to the walls of the fuel feed pipes in the wings and was released during the descent (because of the increase in temperature) causing a restriction in fuel flow at the FOHE. They have been unable to simulate the operational environmental conditions to cause a rollback, except by shoveling lots of ice into their test rig just before the FOHE. There is doubt that the amount of water believed to be in the tanks of G-YMMM could produce the amount of ice required to restrict the fuel flow in this manner. This is a simple precis of part of the AAIB's findings so far, and is an inadequate substitute for their report. Their investigations continue.

Pettifogger,

I think calling the AAIB suggestion a "theory" is a little bit too optimistic - they do not think there was enough water to form sufficient ice, and can't model a reasonable event to cause damage.

I think it is merely something they are investigating.


.

Phil Golin

good explanation.

It would be nice if we on Pprune could agree on the use of words like "theory", "postulation" and "speculation"

If a consensus can agree I'm willing to adopt them

my current use is:

"theory" starts with a premise at the beginning while explaining links in the chain and ends with a conclusion to the exclusion of all other paths (linkages)

"postulation" starts with a premise anywhere in the chain of events and only to explain the linkage to another link in the chain

"speculation" starts with a premise and jumps directly to a conclusion while excluding confirmation of the links in the chain

My current creed is:

I am inclined to start with "postulations" and only after I see enough links fitting that postulation do I assign a "theory" to it.

I take "cons" against a theory as serious as I do the "pros"

I apply Occams razor to determining both the "Pros" and the "cons"

Or perhaps "hypothesis", an unproved theory, tentatively accepted to explain certain facts or to provide a basis for further investigation.


Green-dot

fine by me, is this a substitue for postulation or an "as-in"

Incidently I'm fine with folks who openly hypothesize and/or postulate and/or even theorize later after many facts are known but not with speculation as it wastes time and gives the appearance to the casual reader that one can conclude

Green-dot, lomapaseo,
Funny, really.
I've always used "working hypothesis" as my 'worldview system'.

As a mere human, with limited knowledge, that's all one can do.

To me, "God" is a convenient hypothesis thought up by primitive people thousands of years ago to explain phenomena unexplainable to them at that time (be it thunder, or chance).

We no longer need "God" for that.

But we all need a "working hypothesis" about the world around us. Sadly, too many people revert to "belief", instead.....

CJ

What is a typical fuel flow at flight idle in landing configuration with engine anti-ice selected on ?
Is it correct to guess around 2600 pounds per hour per engine ?

According to the last report, FF maintained for the last 40 sec was twice that amount for each engine, which must not be far from a typical FF in order to maintain the G/S and cross the threshold at 50ft ?
I know someone said earlier the approach was quite bumpy with thrust variations, but in the same time, at this stage, BA38 was pretty light.

I would consider an airplane touches down in such a low energy situation 1000ft before threshold in case of dual engines flame out ... but I find it too much when each FF was between 5 and 6000 pounds per hour ?

Just a thought ...