AAIB BA38 B777 Initial Report Update 23 January 2008

[golfyankeesierra]

prolonged cold soak followed by a long low/idle thrust descent

prolonged cold soak?... yes
low/idle thrust descent into LHR?...... must be your lucky day!

And you have no need for additional fuel inbound LHR anymore.
BTW how much additional fuel do you BA(,BMI and other LHR-based) crews take normally for arrival LHR?

[M.Mouse]

The fuel temperature at the outlet of the fuel cooled oil cooler would be lowest at the moment the crew/AT increased the thrust at 600'. This is when the engines lost thrust.

I've spent far too much time thinking about this and my conclusion is that a prolonged cold soak followed by a long low/idle thrust descent cooled the whole system (engine oil and fuel) such that when thrust was requested the increase in fuel flow through the fcoc resulted in the fuel temperature and therefore viscosity being low enough to disrupt the proper flow through the fuel metering unit or other downstream fuel components such that the requested fuel did not reach the fuel injectors.

Your hypothesis is wrong on numerous accounts. If you understood BA SOPs, the normal approach pattern into LHR and the way the entire arcraft fuel system functions you would understand why.

BTW how much additional fuel do you BA(,BMI and other LHR-based) crews take normally for arrival LHR?

There is no standard answer to that question. Regular LHR pilots rely on experience to judge the possibility of delays taking into account forecast wind, visibility, time of day and day of week and load fuel accordingly over and above flight plan fuel which already includes diversion and contingency figures.

[FE Hoppy]

Quote:
The fuel temperature at the outlet of the fuel cooled oil cooler would be lowest at the moment the crew/AT increased the thrust at 600'. This is when the engines lost thrust.

I've spent far too much time thinking about this and my conclusion is that a prolonged cold soak followed by a long low/idle thrust descent cooled the whole system (engine oil and fuel) such that when thrust was requested the increase in fuel flow through the fcoc resulted in the fuel temperature and therefore viscosity being low enough to disrupt the proper flow through the fuel metering unit or other downstream fuel components such that the requested fuel did not reach the fuel injectors.

Your hypothesis is wrong on numerous accounts. If you understood BA SOPs, the normal approach pattern into LHR and the way the entire arcraft fuel system functions you would understand why.

Would you care to enlighten me? Particularly about the fuel system. I stand by the statement that the fuel temp at the outlet of the fcoc was lowest when the thrust was requested.

[411A]

The fuel temperature at the outlet of the fuel cooled oil cooler would be lowest at the moment the crew/AT increased the thrust at 600'. This is when the engines lost thrust.

Sounds entirely reasonable to me, if the Trent fuel/oil cooling arrangement is the same as on other RB.211 types....and yes, I have operated the RB.211 types for twenty eight years...I'm sure F/E Hoppy is near to this also, so this scenario is entirely plausable...so called BA SOP's or not.

And, just to show that I am not personally knocking the FD crew in question, yes, they did a superb job in keeping damage to an absolute minimum, and for this they should certainly be congratulated.
Well done.

[L337]

A long idle thrust descent from cruise is absolutely positively not going to happen at the time of day they arrived. Even arriving at 04:30 the other morning I still had the power up and down, vectors, step descents, the whole nine yards. So all day every day "idle" from cruise is not going to happen.

In terms of BA SOPS, these require us to be spooled up and stable by 1000'. And if not stable by 500' you should fly a go-around.

[davy123]

Couple of questions, more general than specific as I really don't think there is any point speculating until more information comes through from AAIB:

1. If there was an issue with insufficient fuel getting to the engine would this not be flagged up by the FADEC as a fuel demand versus Acutal crosscheck failure?
I thought the FADEC would demand a certain amount fo fuel based on the TRA and command the FMU actuator to adjust accordingly. It would then monitor the fuel flow rate and based on that make adjustments to keep things as close to steady state as possible.
If the actual flow rate was significantly below demand then the FADEC should have declared a warning to the cockpit EICAS/MCDU.

2. What diameter are the fuel supply pipes? Some aircraft have a system which oversupplies fuel to the engine and then returns the excess to the tank. I think that part of the purpose of this is to keep the fuel flowing. Does anyone know anything more about this design?

[grebllaw123d]

long low/idle descents...

FE Hoppy writes that
"my conclusion is that a prolonged cold soak followed by a long low/idle thrust descent cooled the whole system (engine oil and fuel) such that when thrust was requested the increase in fuel flow through the fcoc resulted in the fuel temperature and therefore viscosity being low enough to disrupt the proper flow through the fuel metering unit or other downstream fuel components such that the requested fuel did not reach the fuel injectors"

I very much agree with the contents in posts #309 (M.Mouse)and #312(L337) re this theory, which I do not find realistic either.

Another point to note:

For more that 5o years there have been performed millions of low idle descents from high altitude after long flights in low OAT with many different aircraft/engines - and as far as I know, no serious problems like the ones described in the theory brought forward by FE Hoppy have ever been reported. I may be wrong - if so, please correct me

During my 35 years of airline flying I have not noticed ONE instance of oil temperature not being in the green during descent - no matter how long time was spent at very low temperatures during cruise. Engines include P&W JT8(DC-9/MD-80), P&W 4060 (B-767-300ER) and CFM56(A340-300).
Of course this does not prove anything - but anyway..

My only encounter with water/ice in fuel goes back approx. 30 years, when I flew the DC-9. We got a fuel filter clogged caution on ONE engine after 3 hours of flying at high flight level - cleared acc. checklist.

In any case the BA incident is still puzzling

[CharlieChalk]

davy123,
"I thought the FADEC would demand a certain amount fo fuel based on the TRA and command the FMU actuator to adjust accordingly. It would then monitor the fuel flow rate and based on that make adjustments to keep things as close to steady state as possible."

The FADEC controls to its primary control parameter (EPR) which is derived from the throttle lever angle. The FADEC calculates the fuel flow demand needed to achieve the commanded EPR and then commands the fuel metering valve to deliver this fuel. The valve position will be altered to achieve the demanded EPR, if the engine does not respond then the control loop error will cause the metering valve to be driven more and more open. The FADEC does not monitor fuel flow directly and control to achieve this, as it is more important to control to achieve the demanded EPR (thrust).

[phil gollin]

Quote :

"..... The FADEC controls to its primary control parameter (EPR) which is derived from the throttle lever angle. The FADEC calculates the fuel flow demand needed to achieve the commanded EPR and then commands the fuel metering valve to deliver this fuel. The valve position will be altered to achieve the demanded EPR, if the engine does not respond then the control loop error will cause the metering valve to be driven more and more open. The FADEC does not monitor fuel flow directly and control to achieve this, as it is more important to control to achieve the demanded EPR (thrust)."

unquote

How does it measure EPR (thrust) - by revolutions or ?

[L337]

What are you telling us? Everyone is on SOP at 1000ft?

Of course I am not telling you that.

[Mad (Flt) Scientist]

How does it measure EPR (thrust) - by revolutions or ?

EPR, being Engine Pressure ratio, is measured by measuring various air pressures inside (and outside) the engine and calculating their ratio.

The general layout is similar to that in the GE engines, an Ad on which system was dissected here at some length a while ago. The RR system is analagous, but of course different in detail.

[barit1]

How does it measure EPR (thrust) - by revolutions or ?

I suspect (educated guess...) that the FADEC senses IEPR on a Roller.

If fan speed is the primary control parameter (CF6, CFM, ...), then N1 is the sensed value, and the FADEC seeks the target N1 by trimming FF up/down as required.

[Jetcal1]

EPR

EPR is measured as a ratio between PT2/PT7. I am new here and do not have much civil experience.

Earlier comments on oil coolers have got me to thinking, it is quite possible that the thermostat on the oil cooler could close preventing the heating of the fuel. The pilot would still see green oil temps. However, must fuels have anti-icing additives.

I have heard of PT2 probes icing over, that could cause a EPR issue. What was the temp on the day of the accident? Is there a ice detection sensor in the inlet?

[bsieker]

Originally Posted by Mad (Flt) Scientist

EPR, being Engine Pressure ratio, is measured by measuring various air pressures inside (and outside) the engine and calculating their ratio.

To get the EPR, you only really need to measure intake pressure and exhaust pressure, EPR is their quotient. The various intermediate pressures are also sensed to monitor correct operation of the engine and possibly adjust parameters such as variable stator vanes and/or bleed valves.

Nasa has a nice picture on engine pressures.

The general layout is similar to that in the GE engines, an Ad on which system was dissected here at some length a while ago. The RR system is analagous, but of course different in detail.

The major differences between the GE90 and the Trent 800 engines is that the GE is a 2-shaft design, and uses N1 (low pressure rotor rpm) as its primary control parameter, whereas the Trent is a 3-shaft design and uses EPR as its primary control parameter.


Bernd

[M.Mouse]

L337(post 312)
....What are you telling us? Everyone is on SOP at 1000ft? ............Were they?

If by that you are asking if the majority of BA aircraft conform to the BA stabilised approach criteria at 1000' then the answer is yes and BA have the data to confirm that. That is courtesy of the SESMA monitoring system which has been in place for years and is used as a valuable aid to monitor trends and address any potential issues before they become a problem. The system is anonymous in that BA cannot identify the pilots of a particular flight but the BALPA rep who sits on the monitoring committee can. He alone has that ability and may contact a pilot purely for further information to help in analysing an event that might need clarification. The pilots are never identified to the company SESMA is purely a tool used for safety purposes.

....What are They telling us ,now, was the problem noticed at 600ft or 400ft? Figures seem to have changed,

The problem occured at 600' and was noticed shortly thereafter. What has that got to do with being stable at an absolute gate of 500'?

[pavvyben]

Normally BA 777's land with 6-8tonnes, yes they vary a bit, but generally you can use that above as a guide.

[Fargoo]

I have heard of PT2 probes icing over, that could cause a EPR issue. What was the temp on the day of the accident? Is there a ice detection sensor in the inlet?

The probes are heated, no ice detection system is required.

[greekdalek]

Compressor Stall

As I've hinted before a compressor stall would also hinder engine acceleration from idle conditions. This could be caused by mis-scheduling of the bleeds.

[arizona]

Loosing momentum

It looks like this very thread, also loosing its momentum. Here comes some fresh (clean, no water or ice) fuel for the fire!
Philosophically (I say) the 777 is two single jets, glued together by an airframe.
I.e. it’s a very few things they have in common. When the centre tank is emptied and the cross-feed closed, you have two single jets, just happen to fly, in close formation.
In theory, they have their own wing, landing gear, fuel, electrical and controls.
So, if you got an engine failure, the next engine failure, statistically, must be in a different time and in a different airspace/airplane.
Hm, who was responsible for the crap? We tried to blame the pax, some green men, p3 lines and voodoo.
I don’t know. How about the loadmaster in Beijing? An air bubble without oxygen on short final??
This one is really tricky!

By the way, the crew was outstanding!

[bubbers44]

These two single engine jets glued together also got exactly the same Chinese fuel load, climbed into exactly the same temps over the polar route and descended exactly the same way with the Center tank fuel for quite some time supplying both of their engines. Also reports of other airliners having to descend because of fuel temp problems to stay in limits. We will eventually find out why this happened so lets wait for the final report.