Boeing 777-300ER fuel indication

[Homo Ludens]

Been flying the bird for about 1000 hours now, and noticed a strange phenomenon.
During a, say, 14-hour flight, the fuel indication on the EICAS and the fuel, calculated by the FMC, would show a difference, as much as 2 tonnes sometimes. After the end of the flight, however, the two figures would be equal. Allways.
Any takes, please?
Your kind contribution is greatly apreciated!
Cheers!

[Old Smokey]

I'm getting a kind of feeling of deja vu.... Did we fly together yesterday in said aircraft, and discuss this very topic and it's "fix". If so, see you for a coffee downstairs, if not..... I can shed some light on the matter.:ok:

Regards,

Old Smokey

[mach 84]

happens on the 200 too, the biggest difference i have observed was 1100kgs,

[doishquattroserche]

i know to get the fuel disagee message on the triple requires a difference between totaliser and calculated of 4080kgs which must be a round number in lbs but does sound rather a lot . see i can remember the twaddle,but where's my car in the car park?!!

[Techman5]

Very interested in this story. It sounds like a regular occurence. Does it always happen at a specific point, e.g. when the centre tank empties and you start burning the wings, or when the fuel is at a specific level?

Does it occur during manouevres, or just in level straight flight?

[Homo Ludens]

It's happening gradually. The difference increases, then decreases, sometimes it goes the other way - the calculated becomes greater then the indicated. In general, if there is some kind of pattern, I can not get it.
Most of the observations were done on trans-Atlantic (E-W and W-E direction) routes.
Do you think it could have something to do the sun heating the wings unequally?
And, BTW, I've never set foot in Singapore, unfortunately. :)
It looks like the matter has been discussed in a lot of 777 cockpits...

Good flying to all!

[Homo Ludens]

Smokie,
Kindly share the "fix", please.
Thanks!

[lion-g]

My guess will be due to the lower fuel tempreture after a long flight. As the 777 fuel sensor measured the quantity based on velocity of sound, with in turn is directly propotional to the density of the fuel. Therefore, with a colder fuel, this will results in a denser fuel thus, with the formula, density X volume = mass, the totaliser tends to "over read".

Just purely my guess, standing by to get corrected.

Cheers,
lion-g

[TURIN]

Noticed an associated fuel indication issue on the -300ER sometime ago.

On a quick turnround the crew will enter the arrival fuel in the tech log very soon after 'chocks on'. About 15 mins later when the eng/tech gets to the flt deck the indicated fuel in tanks suddenly changes by as much as 800KG.

There was a company Tech News knocking about concerning the VOS in fuel but there was supposed to be a software fix for it.

Caused a few fuel discrepancies until we got used to the idea and always, always check the onboard fuel before allowing the refueller to start pumping.
On a quick turn round everyones in a rush!:ugh:

[Old Smokey]

OK Homo Ludens, it's a bit long winded, but here goes. Bear in mind that what I say here came about after experimentation, and VERBAL advice from Boeing, I have no reference to quote.

It is an FMC problem, specifically, VNAV, it has nothing to do with Flowmeter performance or fuel density, and applies to RR Trent versions as well as GE 115 versions.

Like yourself, I had often noticed the difference en-route between Indicated and Calculated fuel, but unlike you, I often landed with about 1 Tonne less than the FMC prediction, even after flying the approach and arrival exactly as programmed in the FMC (no additional radar vectors etc.). The Calculated prediction was in error on the "wrong" side, leading one to believe that there would be more fuel on arrival than actually was.

The first "fix" was accidental. Most pilots calculate Vref by adding the predicted fuel on arrival to the ZFW, and entering this figure on the APPROACH REF page to get the Vref. After leaving the APPROACH REF page, the manual weight entry reverted to Actual Weight (Important to remember).

In earlier times, the Boeing technique for calculating Vref was messy and seldom used (most operators used the technique described above). Hints of it are still there in Vol 2 if you want to go searching. The technique was to note the FMC Fuel prediction for arrival, go to the PERF INIT page, and Manually enter the FMC Fuel prediction into the Fuel line. (This immediately resulted in an "INSUFFICIENT FUEL" message, but we knew why it was there). Then proceed to the APPROACH REF page to get the Vref.

VERY IMPORTANT : Now return to the PERF INIT page and DELETE the Manual fuel entry, it DOES NOT automatically erase when you leave the page. This is done by pressing DELETE and Line Selecting the Fuel Quantity line.

After doing this return to the PROGRESS page, and Huh!, where did 800 to 1000 Kg of fuel go to?, the Arrival Fuel Prediction often dropped by this amount. (Yesterday the Arrival fuel dropped from 9.1T to 8.2T after this procedure, necessitating a second calculation of Vref). Go now to PROGRESS Page 2, and et voila, the Calculated and Actual Fuel matched, or nearly so.

And what fuel did we land with after STAR/Approach as planned ..... 8.2T:ok:

My Boeing informant tells me that it's a VNAV glitch that will be addressed ..... someday. By going to the PERF INIT page, and inserting a Manual figure, and then deleting it, it forces the FMC to do a "Soft Reboot", removing the cumulative VNAV errors. (Being an INIT page, you're forcing a "Soft" Re-Initialisation and internal re-work of the figures).

So there you have it. The DANGER in the procedure is that most crews are accustomed to temporary Manual entries erasing after moving to a new page. It DOES NOT for the PERF INIT page, and some crews have been perplexed by subsequent INSUFFICIENT FUEL warnings. For this reason our company (and most others) avoid the original Boeing technique, but a side effect of this Vref calculation, although messy, is correction of a known Calculated Fuel error caused by a VNAV glitch.:D

BE CAREFUL!!!!!!

Regards,

Old Smokey

[lion-g]

So, can i conclude that the totalizer figures is considered more accurate than the calculated figures on the FMC Progress Page 2 ? Provided there is no fuel leak or other external factors affecting quantity of fuel ?

Thanks in advance,
Lion-g

[Old Smokey]

What you say lion-g is absolutely true. One of the caveats that I did not mention is that the procedure should not be used if there was a suspected fuel leak, as the Non Normal Procedure for this requires a cross check between the two readings, as you are undoubtedly aware. There could even be a case argued for actually doing it in the event of a suspected fuel leak, but the known error is still being investigated by the makers.

The cynics will say that following the procedure merely re-initialises the FMC to Actual Fuel on Board from the Totaliser, but, in flight, this is not so.

Until the software glitch is repaired, my attitude is to use the prediction (the higher figure) for Vref calculation, and the prediction MINUS the error for Flight Planning purposes.

Regards,

Old Smokey

[Homo Ludens]

Old-man Smokie,
thanks for the info.
I've seen this technique been used.. My concern is, if the totalizer is at fault, by doing this we loose the good indication.
It's interesting to know if a similar problem existed on 767 or 744.
cheers!

[Techman5]

Very interesting about the VNav.

My only addition is that all fuel systems, be it the ultrasonic FQIS, the capacitance fuel guages or even sticking the tanks, actually generate a depth of fuel at a given point. When everything is working and the aircraft is level, it is straightforward to work out a volume. Obviously, if probes are misreading, or if you are manoeuvring, it becomes more complex.

Obviously, you need to multiply by a density (usually around 0.8Kg/lit) to get a mass. If you have a 50T load, clearly, it only takes a slight change in density (which will obviously change as the temperature changes) to create a significant difference in fuel remaining.

The one advantage that the ultrasonic system has proved to have, is that it is much better at getting a density than other systems.

When you measure fuel flow, that will measure the volume pumped to the engines. I don't know what is used to get density in this system.

[chksix]

Isn't the computer compensating for the temp to get the correct density? :8

[sequ]

It happens the same on the 767.

SEQU

[lion-g]

Thanks a lot Old Smokey, hope to fly with you one day :ok:

[GE90115BL2]

Yes, I too have seen some strange Fuel TOT readings on the ER, around 1000 kg's difference in CALC and TOT in cruise.

The Left and or Right gauges can fluctuate quite a bit after shutdown.

And we do seem to shutdown with up to 1000 kg's less than FMC predicted FOB. ( even after flying the exact app we programmed in )

Thanks the info Old Smokey I'll ask out tech reps what gives.

[woodyspooney]

All these posts are pretty interesting. I believe Boeing had a bulletin dealing with B777 fuel totalizer discrepancies.

All is well and dandy when you do not have other problems like fuel leak. Several years ago a colleague in a major Pacific Rim carrier had this fuel totalizer discrepancy together with a small fuel leak. Sometime after departure from Seoul he had noticed that the fuel totalizer was reading about 1000-2000 lbs more than calculated and then fluctuated to slightly less than calculated. The fluctuation of the totalizer from several hundreds of lbs more and then to several hundreds less than calculated continued for sometime; this was consistent with the contents of the Boeing bulletin. A couple of hours after that, he began noticing a persistent reading less than calculated only to be followed by an increase to more than calculated as they entered turbulence. Once he climbed above the turbulent layer in smooth air, the totalizer versus FMC calculated reading gradually decreased at a small rate of 200 to 300lbs an hour. He began to think he had a fuel leak problem! He tried visual inspection but since it was dark and the fuel leak rate was pretty small, he could not visually ascertain it was fuel leak. He accomplished the fuel leak checklist and indeed, it showed a leak rate fluctuating between 200-800lbs per hour........however he could not ascertain the source of the leak.

Being an ETOPS flight and not wanting to shutdown any engine unnecessarily, he opted to make a precautionary landing. However the airline dispatch and maintenance DID NOT concur with his decision. According to the company maintenance and dispatch, the totalizer discrepancy was within limits as made out in the Boeing bulletin and real time maitenance datalink from the aircraft revealed no problems! They demanded that he continued the flight to the destination.

Nevertheless, he countermanded company directive to continue and diverted to Anchorage where upon landing it was indeed found that he had a small engine fuel leak. Of course he was exonerated but the company officials did gave him a hard time for countermanding their directive twice; once for deciding to divert and secondly, diverting to an enroute alternate despite their later stated preference he return to Seoul in case he really insisted on diverting.

All this grief came because of assumptions made because of the B777 fuel totalizer discrepancies which we see everyday so much so that any fuel discrepancy betweeen calculated and totalizer are just glossed over and ignored. Be careful here!

[Paishinel]

Do airlines' maintenance really access real time maintenance reports via datalink all the time? Can they access the reports as needed with datalink queries to the MAT onboard?

The B777 fuel totaliser problem has always been there and Old Smokey's " solution " is good for reconciling the totaliser and FMC calculated readings to the end of the flight. However, inflight fluctuations can be a headache if you have other related problems which need decisions based on remaining fuel.

[Molokai]

Me thinks it's a myth like all urban legends. Been to a maintenance base in the sandpit once and we have guys yowling out bets on English soccer ; they ain't got no time for any real or pseudo time datalinks!

[Mat Sabo]

Me thinks it's a myth like all urban legends. Been to a maintenance base in the sandpit once and we have guys yowling out bets on English soccer ; they ain't got no time for any real or pseudo time datalinks!


Very correct...the grease monkeys at the maintrol in the sandpit are soccer mad and the EPL games take precedence over any real or bluff time maintenance datalinks!:=

[Geragau]

Well, Old Smokey's trick to " soft reboot " the VNAV function of the FMC does seem to work in getting the calculated figure to reconcile with the totalizer. However, in many many flights where the flight was exactly conducted as per computer flight plan ( as far as TOWs, routes and flight levels are concerned ), the calculated figures are almost equal to the computed flight plan figures at the various check waypoints. On the other hand the totalizer figures fluctuated from plus to minus several hundreds of kgs! By " soft rebooting " the VNAV function, the calculated figure almost immediately equals the totalizer and as time goes on, the two values start to drift apart again.

I still think that the totalizer suffers from inaccuracies in the FQIS; B777 drivers should heed woody spooney's warning about the totalizer misleading the unlucky ones who get slow fuel leaks.

[FullWings]

I've noticed this happening on about one out of three flights over the last decade. It has improved since the early days with software mods but you still expect some discrepancies - it's even got a name now: "mid-Atlantic dip"!

The good thing is that the accuracy returns as the amount reduces, so the danger of running dry unexpectedly is small.

I remember being over the middle of Khazakstan about 10 years ago, looking at the scenery, when the "fuel imbalance" checklist pops up... OK, 2T difference, let's see what's in the tanks... Oh, the total seems to be going down by at least several tonnes a minute - that's not good - might have to shut one down. Got the 3rd pilot to go back and see if there was any indication of a fuel leak: no. Were just contemplating chopping the motor on the offending side when we noticed that the "leak" had stopped and we seemed to getting a bit of in-flight refuelling.

As the indicated imbalance was in the order of 15T, we took the autopilot out and let go: steady as a rock. Hmmm. Over the next 20mins or so the fuel displays returned to normal. We also checked the expanded engineering page which shows the fuel at each of the measuring stations - this was very uneven and depicted somewhat of a 'slope' on something liquid, which was rather unlikely. We were told later it was a software issue with failed sensors messing up the readings, rather than being excluded from the calculation...

[Jon Lei]

It's interesting to know if a similar problem existed on 767 or 744.

It almost always happens on the 744s. Sometimes, the figures differ by almost 2000 kgs after a 10 hour cruise.

[haughtney1]

Happened when I was on the 757/767..and we used Smokeys fix

It damn well happens on the Falcon 900EASy I fly now....:ugh:

[Old Smokey]

Be Careful .....

In my original post, I did describe the "soft reboot" fix (which works), but don't forget that we have three possible reasons for strange B777 fuel indications, i.e.

(1) A VNAV problem being addressed by Boeing, the "soft reboot",

(2) Suspect FQI for the centre tank in some instances,

(3) The "Phantom" Wing Fuel assymetry problem, still being addressed by Boeing (which yours truly experienced last night).

Geragau did point out that whilst the "soft reboot" does much to realign the computed fuel with the Totaliser, they tended to drift apart again thereafter (much like INS after a DME update). What he says is TRUE, BUT, although the in-flight discrepancies might begin again, the more important to me is the PREDICTED Fuel at arrival, and, it is my observation that even though in-flight differences may slowly recommence, the Predicted Arrival Fuel is MUCH more in line with reality.:ok:

Three FQI oddities for the aircraft, I say again Be Careful .....

Regards,

Old Smokey

[MMBenar]

Woodyspooney mentioned real time maintenance data to maintenance control...........is this an every practice? Do ACARS datalink download full real time aircraft performance data to maintenance control all the time or only when there is an anomaly as claimed in the recent Air France 330 AF447 case over the pond?

[halas]

Another way to check the rate of change from the Totalizer to Calculated is as follows:
First take note of Calculated value and Totalizer value.
On the PERF INIT page, type the Totalizer value onto the scratch pad and line select it over Calculated figure then execute. 'MAN' will replace 'CALC'
Then select DELETE and line select over the same figure and execute. 'CALC' will return.

Now they are synched. Then each hour you can monitor the rate of change.
You will be surprised how much this the rate does vary over a long flight.

halas