Large Plumbing
 

The B777 uses two of just about the most powerful engines in the business, so each engine is presumably fed by some of the largest fuel lines in service.

Strange that this very fuel plumbing may have provided the first recorded example of an engine suffering serious fuel starvation due to a major blockage caused by gradual ice accretion, despite the contents of the fuel tanks being of good quality.

It is suggested that this may have been brought about because the engines were not operated at climb power again after the initial top-of-climb, but at slightly lower (cruise) power settings.

During the last several decades, how many other long-haul jet-transport engine-fuel systems − from those feeding Speys on G2s, through aircraft with engines of ascending fuel flows like (for example) JT3Ds on B707s, DC-8s and C-141s; to CF6s, JT9Ds and RB211s on various aircraft − have been operated successfully at similar altitudes and temperatures to BA038, for similar flight times, often without the application of climb power after initial TOC?

There’s nothing new or particularly unusual in crews avoiding the unnecessary use of climb power when an aeroplane is light.

What’s so special about the B777-Trent? Does anyone find the hypothesis convincing?

infrequentflyer789 (#1978)

- Agree about noise and switch-off of OJ pumps.

- I have been through posts but not seen centre tank scavenge inlets, OJ pump inlets or drain valves relative to each other or side on view as to how far off floor. Of few pictures have seen, they include:
Sumping - http://panchromat.org/.misc/B777-doc...stallation.jpg
Vent picture - http://panchromat.org/.misc/B777-doc...ent_system.jpg
Pump - http://panchromat.org/.misc/B777-doc...stallation.jpg
Centre tank - http://www.pprune.org/3929959-post284.html
Engine fuel supply – ImageShack - Hosting :: 777engineug1.jpg
RR Trent parameters - http://rgl.faa.gov/Regulatory_and_Gu...E/E00050EN.pdf
Plan of fuel supply: ImageShack - Hosting :: 777fuelpw3.jpg

- Agree re burn rate - slight difference in main tank flow rate just before and then after scavenge finishes.

- Agree re context of remark in AAIB report - my point was more that it seemed to suggest if gas then why not fluid could move through inactive OJ pump with suitable pressure differential? [I had an obscure "March Hare" theory a few posts back that air might be dissolved/entrained into main tank fuel as result of operation of centre tank fuel scavenge when CWT dry and as a result (i) subsequent release of undissolved air from air-saturated fuel into suction feed pipe and/or fuel manifold near OJ/ jettison check valve [or, before cavitation (as cause) ruled out, in HP pump] and (ii) so suction feed restricted by undissolved air and/or undissolved air near OJ/ jettison check valve end might allow localised low pressure fluctuations induced by cycling to result in OJ/ jettison check valve opening if surge of pressure in centre tank on fast descent, thereby introducing water slug (not air) to fuel manifold. In my dream world both the Mad Hatter and March Hare were coming to tea.]

- Agree that will have compared fuel data and that of previous flight.

Anyway it seems there is still quite a bit of Boeing testing to be done on the cascade of contributory factors leading to the location of the 95% restriction in the underlying fuel system, and the tentative, conflicting scenarios to date.

Perhaps the expense of putting a bit more flight test kit where not normally found in tanks and supply lines of a real plane (with correct geometry) and then flying it would be worth it (maybe could add water to centre tank and lower the temp of fuel in one of the main tank before uplifting it to cut to the chase, etc) ... since a double rollback is not just unusual but after all a surprisingly serious failure condition in an otherwise excellent aircraft.

The 777 FDR receives and stores selected parameters from various systems and sensors (to know which parameters must be equivalent to drawing Excalibur from the stone). A digital flight data acquisition function (DFDAF), part of the airplane information management system (AIMS) receives all the flight recorder data. The data from the systems is received through ARINC 429 and ARINC 629 data buses, and also analog signals and analog discretes. The DFDAF processes this data (converts it to a digital format) and sends it to the FDR through ARINC 717 data buses.


Most likely (if water sumping was correctly performed) there would not be that much ice adhering to the internal wing structure after a 20 hour downtime with temperatures between +6C and +11C before the flight to Beijing on Jan. 16. Add to that the 2 day maintenance downtime (inside a hangar?) before the flight to/from Shanghai on Jan. 14-15, where likely water draining and refueling would have taken place only hours before departure to Shanghai. Outside temperatures at Heathrow were between +4C and +11C on Jan. 12 thru 17.

It seems to me that whatever amount of water was present in the tanks during this period with ground times of 48 hours and 20 hours respectively, most of it, if present, would have been drained. In addition, refueling at Heathrow (according to the AAIB) occurred only after sumping for water, any condensation adhering to the interior structure (water, not ice with so many hours of temps. well above freezing) would have been absorbed by the fuel during refueling and would not have had time to settle at the bottom of the tanks before the flights took place.

Apart from the Trent being different from the other engines types, the plumbing in the airplane fuel system is no different then for the other engine types, except for the engine/fuel system plumbing interface.

No, i do not find the hypothesis convincing, especially when the weather conditions and actions taken to drain water while on the ground for considerable times at Heathrow are taken into account.



Green-dot

Chris Scott wrote:

"The B777 uses two of just about the most powerful engines in the business, so each engine is presumably fed by some of the largest fuel lines in service.

Strange that this very fuel plumbing may have provided the first recorded example of an engine suffering serious fuel starvation due to a major blockage caused by gradual ice accretion, despite the contents of the fuel tanks being of good quality."

Well, if you think about it, having a larger diameter pipe to allow higher flow at full power means that the effects of low power and fuel flow for long periods are exacerbated, the fuel is moving more slowly because a given volume travels more slowly through a larger diameter pipe. Hence, if this is down to ice accretion on the pipe walls, there is more time for an ice crystal passing accreted ice to stick to it.

Pumps in series
 

dxzh, in re your post #1969 of yesterday, on page 99:

Centrifugal pumps of conventional design can be operated in series; I have done design work with such arrangements. It does not matter if the upstream pump (in this case) loses its electrical power. There are no in-built internal obstructions to continued flow through a centrifugal pump whether or not its impeller is still free to turn, or siezed in its bearings.

It is conceivable that the housing of such a pump could be fitted with a stop valve held open by a solenoid energized by current from the pump power circuit. I haven't seen any indication that the 777 pumps of which I have details (Eaton Aerospace (UK) wing tank LP boost) have any such feature. This seems undesirable in an aircraft, just my opinion. I've seen systems (non-aircraft) that had such a feature, but only as a separate valve in the piping.

As to air entering the engine LP pump suction, I can only address the fuel path from the main wing tanks. Here the wing tank suction points are above the engine fuel feed in such a way that the engine LP pump would normally operate with a gravity-flooded suction if fed entirely from the wing on that side, with the LP boost pumps of that side off. I can't say I see this as even remotely likely in the air, and use it only by way of example (lacking as yet sufficient info regarding elevations in the center tank plumbing).

Anyway, with a fuel velocity of at most 7 feet per second (8 possibly), I cannot see it likely that significant air would be entrained should the suction of one of the wing LP boost pumps become unflooded itself due to fuel slosh. It isn't like the engine pump was above the fuel source. BTW, the velocity given is for a fuel rate of 35,000 lbs/hr. This is the conventional maximum velocity (for avoiding static charge on the fuel), and may also be seen from the near 2" discharge diameter of the wing LP boost pump casting.

The same principles would apply to the center tanks, but as noted I don't know the elevations of the various components clearly enough to comment. The dimensional detail of the wing LP boost pumps in on the Eaton Aerospace website; earlier I forgot to mention that it is not in the brochures-- instead look under "downloads" for the data sheets, including B777-300 for a comparison to the -200.

Following what I consider to be good example, I stand to be corrected in this matter if necessary. Well, it is a somewhat complex fuel system, although I think it is robust, simple in components and backup/bypass paths, and extremely reliable, at least until now.

OE

Quote: All of the testing and research of this accident has been conducted
on Boeing Model 777-200 and -300 series airplanes, equipped with Rolls-
Royce Model RB211-TRENT 800 series engines. Initial review of 777 other
airplane engine combinations has not revealed the same vulnerability to
the identified unsafe condition. End Quote.


Sounds like they are reasonably sure where the blockage occured and its somewhere specific to airframes plumbed for RR engines. Why wouldn't a similar release of ice be a hazard to other engines?

Edit: I was asked to post the source of the above quote. It came from this document posted earlier in this thread by "Precept"...

http://www.pprune.org/4391725-post1975.html

Justia Regulation Tracker Airworthiness Directives: Boeing Model 777-200 and -300 Series Airplanes Equipped with Rolls-Royce Model RB211-TRENT 800 Series Engines, - Federal Aviation Administration - 52909?52911 [E8?21138]

cwatters
please identify source of quote that you used in your post

FAA Airworthiness Directive
 

Regarding the above quotation, Google is your friend, and finds that exact string at:

FR Doc E8-21138

Assuming this site to be legitimate (I think it is), it comes from the discussion portion of the Supplementary Information section for the AD being promulgated by the FAA in response to the incident which is the subject of this thread.

This FAA AD (for comment) amazes me - to put it mildly !

It would seem that whilst the lead autority (the AAIB) cannot assess what has happened and only speculates - the FAA SEEMS to have decided what happened, and more importantly that it couldn't happen to other engines on 777s ? (The AAIB thinks proper testing of other engine/airframe combinations are required).

Well - IF the FAA know it can't happen to other engines, maybe they can tell us (and the AAIB) what happened with the Trent supplies.

PLEASE can someone explain, as this doesn't make any sense to me.

.

Part 'g' has been written (in beta form if you like ;)) for the AMM and is awaiting final approval. It has been tried and tested mainly to see if any improvements can be made (for example reducing the number of trips between the flt. deck and refuel panel) or whether it's a one or two person operation!
As regards the comment 'it can only happen to a RR Trent powered a/c' may stem from the fact that all the fuel system components are on the outside of the fan case. Whereas certainly for the GE90, all the components are around the core engine in a very warm environment. Other than the pylon area the airframe fuel system is identical. :8

just suppose
 

what if the sumping was not done? For example the engineer just rubber stamped the job. I've heard it goes on. Now the aircraft has the free water thats needed to create the ice that caused the blockage that brought the craft down prematurely.

Just a thought;-)

About ten pages back, there seemed to be an emerging consensus that the reason the RR engine was being singled out for changes in operating procedures was the presence/position/design of the FOHE. If blocking of the heat exchanger was the immediate cause of the loss of power, then the time between the demand for increased thrust and the start of uncommanded power reduction could give an indication of where in the system the ice came from. I could dig out the time from the AAIB report, but I have no information on the speed of fuel flows or the geometry of the pipework. Does anyone in possession of the relevant information care to do the sums?

Re: http://www.pprune.org/4392955-post1983.html

If refueling was done after sumping at LHR what about the possibility of contaminated fuel being loaded at LRH? Unlikely, but if I understand correctly they had a water alert at LHR? Say that water was correctly scavanged and sent to the engines causing a layer to build up in the plumbing (not enough to cause a problem in itself). The contaminated fuel would have been used up and replaced with good fuel in china. It was cold there so ice in plumbing (if not the tanks) might have remained frozen? Then on way back ice build up reduces until the last bit brakes off in chunks and causes the serious restriction. Result no water detected in the tanks.

However unlikely (it had crossed my mind as well), it could be a possibility but the AAIB most likely will have investigated such a scenario. Fuel in storage tanks, fuel hydrants and/or bowsers are also monitored for contamination and records are kept. If there would have been a contaminated batch of fuel originating from LHR, I'd suspect more aircraft would have been affected.

I know, it is quite speculative to say the AAIB will have investigated because none of their reports have confirmed that they did but knowing their thorough investigative track record, my guess is that they did.

Regarding the water alert, none of the AAIB reports sofar have confirmed the claim of a water alert for any of the flights referred to in these reports. I am sure that if it was an issue it would have surfaced in the most recent interim report.


Green-dot

I flew the 767 some years ago, I seem to remember that there weren't any water drains on that aircraft. Does the 777 have them?

"I flew the 767 some years ago, I seem to remember that there weren't any water drains on that aircraft. Does the 777 have them?"

They both do.

water drain
 

of course 767 has water drains! All aircraft do

I really don't think sky9 meant what he said!

As an ex 767 driver, he's obviously no fool!

Some editing required perhaps?

Any chance there is a correlation in progress between the investigation authorities with respect to the BA38 and QF72 events regarding possible uncommanded system behaviour?


Green-dot

Simple answer NO! BA038 was brought down due to ICE in the fuel system. Nothing more nothing less.

Simple answer NO! BA038 was brought down due to ICE in the fuel system. Nothing more nothing less.

Must be great to be so sure in your own mind - do you work for London Airways (BA)

Draughtsman, its not about being clear in ones mind, just read the AAIB report. Just they can't fathom were the water came from. Of course the answer is sure to be revealed in the near future.

I like your continual reference to London AIrways, maybe this should be a new thread. Is BA entitled to continue using British in their name when they have pulled out of all regional bases?

I guess you work for them out of GLA, sad day for you all. Lets hope your MP Sheridan and his motion in Parliament brings about a U turn.

Back on thread, ice brought down BA038, the Qantas flight cause is as yet unknown and totally unconnected.

I have read all the reports.

Ice accumulation is a hypothesis, nothing more.

No more than "backdoor coupling."



Green-dot

Winter is upon us, so maybe we will see a repetition of BA 038. But then, maybe not. I have flown BA038's profile before without mishap, so why did that accident happen just once?

I do not support the fuel icing scenario. I reaffirm my hypothesis that on BA038's flight the fuel uplifted in Beijing had an unusually high gas content. This excess gas became trapped in the suction feed pickup in the wing tanks during flight, was pressurised in the descent (above normal levels) and was later injected into the fuel manifold at low altitude, as the trapped gas exceeded LP pump output pressure. The ejected gas restricted fuel flow to both engines (almost) simultaneously.

On one of my infrequent visits to Heathrow last week, I saw what I presume is G-YMMM sitting behind screens not far from its original location on the threshold of 27L.

Can someone tell me why the fin has been removed? A source of spare parts or an obstruction issue?

I can think of 2 better reasons. Firstly it was a gigantic vertical reminder of the event to those not fortunate enough to be armchair theorists on PPRuNe, secondly in high winds the ac was getting blown about.

The fin was removed at the request of the A/C's owners (owned by the insurance co. and NOT BA.), for sale to anyone that wants to buy it! Plus as overstress points out the a/c was moving about as its no longer supported on jacks but lying on 'sleepers'.

Trent Differences
 

It is my understanding that one of the main differences between the Trent engine installation and others is the design of the FOHE and the position it is plumbed on the trent. Perhaps one of our engineering buddies can confirm what the difference is.

Yes. On the Trent, the Fuel passes through the FOHE first when it enters the engine. So there is a chance that any ice could block the passages in the heat exchanger. In the GE the fuel goes through the pumps first, so there is good chance that any ice will be broken up in the pump.

I'm still very unhappy with all the findings so far and have little confidence in the ice theory.

This feeling will remain until a fully rigged 777 can be sat in a cold store to exactly similate the flight and bring about the same engine problems a number of times just like any other proper scientific experiment - demonstrable by REPEATABLE experiment.

Until then I shall habour deep suspicions that some dark conflict lies buried in the software with sharp teeth waiting to bite some other poor souls when they least expect it.

You believe a software problem caused cavitation in the fuel pumps?
and what stopped the engines spooling up when the pilots advanced the throttles, isn't it a mechanical linkage in the 777?

The Fuel Metering Valves Opened Fully
 

I've read all the previous 2012 posts. Unless I missed something, it has been conclusively shown that the fuel metering valves moved to the full open position but there was no fuel to move through them. Since the valves opened, how could it be a software problem? The system responded to the movement of the throttles. The engines failed because of a scarcity of burnable materials to pass through that system. :ugh:

This accident will eventually fall into the same category as the 747
centre wing tank saga. An accident or two followed by speculation
and investigation over many years. There is obviously a design
fault that could recur. But economic expediency by airlines, boeing
and the engine manufacturers means we all take a calculated risk
until the real problem is fixed. My bet is a small mod will be made
to the fuel system, a proceedure change to fuel management etc.
Just as the 747 problem unfolded, they will not acknowledge a
known design fault but will piecemeal create a fix. Why did they
change the centre wing pumps on the 747, why have they built
a nitrogen enrichment system into the latest 747s centre tank.
Different technically, but identical legally. I guess its all part
of human progress, we learn as we evolve, build technology.

Quote from Swedish Steve:
On the Trent, the Fuel passes through the FOHE first when it enters the engine. So there is a chance that any ice could block the passages in the heat exchanger. In the GE the fuel goes through the pumps first, so there is good chance that any ice will be broken up in the pump.
[Unquote]
If memory serves, that's news on this thread. If relevant, would it be practicable to re-position the FOHE, Steve? If so, would you expect the mod to be in the pipeline already (pun not intended).

Quote from Phlap1:
Just as the 747 problem unfolded, they will not acknowledge a known design fault but will piecemeal create a fix.
[Unquote]
In the event of that being necessary, let's hope they don't hang about too long in an effort to avoid it being too obvious to potential litigants. If they did, they could end up wasting one of the luckiest breaks in aviation history.

Quote from Smilin Ed:
...it has been conclusively shown that the fuel metering valves moved to the full open position but there was no fuel to move through them. Since the valves opened, how could it be a software problem?
[Unquote]
Looking at hypothetical software problems, could the fact that the position sensors indicated and recorded the valves fully open not be part of it?

Quote from Smilin Ed:
...the fuel metering valves moved to the full open position but there was no fuel to move through them.
[Unquote]
Considering a given engine: if the valve was indeed fully open, there was certainly a flow restriction elsewhere, as you say. Was that due to a foreign-object blockage (ice, for example), a loss of pressure (and/or suction) to shift it through, or the anomalous closure of another valve?

Quote from Smilin Ed:
I've read all the previous 2012 posts. Unless I missed something...
[Unquote]
Well... There's a thought. ;)

Quote from Chris Scott
Wouldn't the software which moves the fuel metering valves be different from the flight recorder software which monitors the valve position? Surely they are independent. Is there someone here who can answer that?

I believe Swedish Steve may have stumbled on the answer.

If the fuel flow through the FOHE became more & more partially blocked throughout this long cold flt,then a continuous descent at low pwr would not have changed anything, until more fuel was required at engine spool up at 1000ft. Then the ice would be sucked further into the FOHE causing a blockage, it would have melted with time but time was not what they had.
This in fact could explain the cavitation damage of the pumps, it had been happening to a lesser scale on other flts but the temporary "hicups" had not been noticed.

This had been a long very cold flt, with maybe a marginal fuel spec.

It seems incredible to me that where there has been an obvious fault in an aircraft type whole fleets have been grounded until the problem is fixed or checked. Here we have a heavy, sophisticated aircraft that crashes for no expiclable reason and it just keeps on flying until, perhaps, it happens again! Can you imagine what must go through the mind of every 777 aircrew on finals? Surely the aviation authorities can do better than this or do we have the dreaded economic considerations outweighing the value of human lives?

Being inherently distrustful of computers I will always suspect 'wonky' software.

But just in case, why did we as Pilots accept losing total control of the fuel heat ?

From the 75 onward on Boeings (don't know about Airbus) not only do we have no control of such, but not even a way of monitoring its activation.

Bring back fuel heat switches !!

And flight engineers to use them as necessary!

I do NOT want to be too dogmatic as it has been so long since this (and the previous) thread started to remember all the details, let alone theories.

However, If I remember correctly, the engine thrust was increased and decreased several times during descent without problems. In addition, the actual problems occured at slightly different times but ended up at almost exactly the same amout of (lack of) thrust. It seems almost impossible that the conditions in each tank were such that the "icing" (OR WHATEVER) acted at different times but ended up with the almost exactly the same results.

People grasped at the cavitation damage because IF (AND IT IS A BIG IF) the cavitation meant that pumps would work at x% capacity and this would show why the thrust fell to the levels that actually concerned. HOWEVER, as no definitive statements accompanied the last AAIB report it is questionable as to whether this is possible, or not.

It seems odd that nothing has arisen in the more than 9 months since the accident (and especially since the Northern Hemisphere Winter is near enough here) to actually guide airlines in possible ways to minimise a re-occurance of such an incident.

(Sorry for the CAPITALS - but people do have a tendency to go off on tangents if you don't spell things out).

.