Green-dot

Swedish Steve,

As I have sumped fuel tanks on aircraft many times and with similar findings as you discribe. Most of the time there was no visual indication of water what so ever. The odd every-now-and then somewhere around that 1 cc you mentioned. Never experienced more than that.

Indeed, finding 5 liters of free water during a transit check (certainly when sumping is performed on a daily basis) is reason for an investigation.

The estimated 5 liters of water in the fuel loaded at Beijing, according to the report, was divided into 3 liters of dissolved water and 2 liters of undissolved water (entrianed or free, spread evenly over the main tanks and the center tank), plus perhaps a maximum of 0.14 liter which may have entered via the fuel tank vent system during the flight.

If evenly spread, that would imply a maximum of approx. 0.7 liter per main tank and center tank regarding entrained or free water in the fuel as loaded in Beijing. Main tanks probably a bit more than 0.7ltr and the ctr tank a bit less when compared to the fuel quantity loaded in each tank.

When the aircraft's fuel pumps are activated, most of the dissolved and entrianed water/ice would have been well stirred and consumed during the flight, with the aircraft sumped twice the previous few days, how much water would have had time to actually settle at the bottom of the tanks to form a layer of ice? And if so, if that estimated small amount broke loose or melted into slush, wouldn't it have blocked the water and/or fuel scavenge pumps before it could have collected somewhere in the boost pumps or engine feed manifolds?

After take-off at Beijing the fuel temperature remained -2 deg. C until reaching initial cruise altitude, as measured by the probe in the LH main tank. With the center tank above the airconditioning packs would it be possible that the local fuel temperature was slightly warmer at the bottom of the center tank during this flight phase, melting any ice at the bottom into water which would have been a very small amount, if any at all, as estimated in the report?

If so, wouldn't most of it have been scavenged by the ctr tank water scavenge pumps, subsequently well stirred by the OJ pumps and have been consumed with engines set at climb power before doing any harm?

Another question is, if the fuel temperature in the above scenario remained below 0 deg. C, but with an almost empty center tank (situated above the airconditioning packs), what would the temperature be at the bottom of the tank with the 800kg of fuel remaining for approx. 5.5 hours before those 800kg were scavenged? Would the temperature have reached above 0 deg. C, melting any ice which would have been scavenged as a mix of fuel/slush/water but well stirred after passing the boost pumps before re-freezing again in the engine feed manifold? (ctr tank OJ pumps are off at this stage, therefore water scavenge pumps in the ctr tank are off) Or would the ice have melted at lower altitude after the ctr tank was empty and then scavenged? (max. 0.3 liter LH and 0.3 liter RH in the ctr tank) Any free water at the bottom of the main tanks would probably have remained frozen until touchdown and would not have moved much, even if it broke free, due to the dihedral of the wing lower surface. The picture taken of the aircraft with frost on the wing lower surface seconds before touchdown may be evidence to that.

I find it hard to believe such small amounts of free water/ice in the ctr tank, which has a flat lower surface, would have been distributed so evenly between LH and RH engine feed systems with an aircraft in motion to contribute to the cause of identical problems within seconds to both systems.

The 3 liters of dissolved water in the fuel, however, might have but the question remains, wouldn't we have experienced such events more frequently?


Green-dot

sispanys ria

Come on, flaps reduction is fine when high and far from the runway as it requires a speed increase (nose down) loss of altitude. Anyway, the flaps reduction technique to increase distance has to be completed by flaps extension during the flare in order to benefit from final lift increase, ground effect and speed reduction to allow low speed/low vz impact.

Since the plane was under AP flying the glide, retracting the flaps definitely increased the AOA to compensate the loss of lift and accelerated the stall. To me, retracting the flaps at low speed/altitude while the AP is desperately trying to fly a path passing 50 ft over the threshold is a weird technique while the only thing that matters is to make the threshold. For sure, the glide is not the best path to the threshold, as retracting flaps doesn't help when you're plane is 14 degrees nose up with almost no power and the AP in charge.

snanceki

IMHO I can't help but think that we are still missing the point.

Accepting that sufficient evidence exists to point to fuel restriction I remain open minded regarding the cause.

For ice to have formed, to the extent required to restrict flow, a considerable amount of water would need to be present in the supply from the fuel tanks and in addition some mechanism would have to be present to make this situation common to both LH and RH systems.

Now the AAIB choose the word "accretion" I believe with considerable care. I suspect that used this word since it infers the presence of a "nucleation" site. It also indicates "growth".

So a mechanism has been proposed that ostensibly could be common LH to RH. However the conditions by which the growth takes place has not been indicated. i.e. Growth from passing ice crystals "aggregation" or "freezing" of dissolved water from solution.
Either way COULD explain why such a RELATIVELY small quantity of water caused the restriction since the restriction would build up over time.

IMHO I wonder whether the intuitive design practice to consideravly oversize fuel supply lines is in fact counter productive since this reduces fuel flow rates which ARE LIKELY to significantly accretion rate.

Add to this the fact that fuel economy is improving and thus reducing flow rates MAYBE we have reached the reached the edge of the envelope of our (lack of) understanding of permissible water content.

This might actually support why the AAIB actually only sited the Trent 800 and left the wider picture to other agencies.

Strikes me we are facing a design maturity / standards issue and that it is an issue of lack of understanding (technology maturity) rather than something specific being/having gone wrong.

Following this logic I would have have thought that some relatively simple experiments could be done to determine accretion rates at differing water concentrations, flow rates and temperatures, could be produced.

The difficult bit will be determining whether the subject aircraft meets these "new" requirements and then subsequently potentially all aircraft in operation.

TheShadow

snanceki said A small amount of water coming out of solution due to super-cold temperatures can form a sizeable ice-block if it all sinks to (or aggregates at) a common (left and right side) low point in the system. What could cause water to exit solution at a particular point (and later melt from its attachment point and migrate during the warming descent)? Perhaps it's a function of:
.
long-term exposure in the cruise to super-low temperature as compounded by.....
.
pressure drops at constrictive flow-points (such as in a venturi) and .....
.
a very cold area (not all areas outside the pressure hull stay at a homogenous temperature). Aerodynamic stagnation points can cause significant temperature variations, particularly where piping runs along or immediately adjacent to the aircraft's skin.
.

Reflect upon what water does in another fluid (i.e. air) when it's cold enough. It will form freezing rain (that hits and sticks and accumulates upon impact). It will also form snow and hail, depending upon relative humidity, the temperature strata (lapse-rates), condensation nuclei and the recirculation found in strong convective flows. Atmospheric water will form rough shapes upon wings and fuselage as clear or rime icing, particularly in areas where the airflow tends to be slowed or stagnates. The equivalent outcome within fuel tanks is dependent upon the thermal and fluid dynamics that are in internal play. If the fuel-feed obstruction was actually a coalescing mass of "not so small" ice particles (i.e. an icing "tumour" or growth), rather than a detaching mobile lump, we should not be amazed.

pax2908

I still find it annoying that "fuel temperature" is actually just one sensor, in a relatively big system. The explanation given in the report, why the measured temperature is representative, is quite short!

Someone mentioned that the FOHE is regulated. I don't remember this was mentioned in the report. Is it true? What other temperature probes are there to read out? Oil temperature maybe?

When did the "obstruction" start? I understand of course that by the time the FMV is fully open and the thrust is less than commanded, you can say the fuel flow is (already) restricted. But how about before this? When did the valve position start being inconsistent with the thrust demand?

snanceki

@ The Shadow.
I think you are reinforcing what I tried to say.
Yes, your analogy to a tumour is good.

I believe that this has been a "latent" problem for some time.
I believe that the key will be found to be low flow rates exacerbated by minimum use of thrust during descent, rather than unique temperature conditions.
Of course low sub zero (normal) fuel temps are a prerequisite.
I still believe that fuel distillation / composition may have a bearing on the issue by impacting the water solubility/temperature characteristics.

So I wonder where flow rates are at their lowest, away from potential energy sources and combined with profile changes that might act as a nucleation site.


I don't believe that the tumour broke away and blocked some downstream area. For this to have occurred at almost the same time in both independent systems is one step too far on the coincidence scale.

We'll have to wait and see.

Basil

DC8,
I intended my comments to be taken separately with the go-around as supporting comment.

sispanys ria,
You do not post your background in order for we experienced heavy jet pilots to accord appropriate weight to your comments. Please re-read what has been said about landing flap.
The accident took place eight months ago and here we are, in the cold light of day, following a good night's sleep, still discussing the actions of the flight crew who, following a long flight, had seconds to consider ways of ameliorating a very confusing situation. The aircraft landed within the airfield perimeter; no one was killed; rejoice!

sydneymike

I think that BA/Boeing/RR need to respond quickly. There seems to be a groundswell of (Albeit uninformed) public opinion developing that until the problems are clarified and resolved, better to avoid the 777. This was "explained" to me as follows: "If the chances of such a random event are remote, statistics dictate that it could happen again today. In the meantime, I prefer to fly 4 engined planes." Never mind that this is not PEK in winter.
The old adage that "Because they don't make em with 6" comes to mind?

You know, there is a kind of logic to this and the markrting gurus need to respond.

sispanys ria

As you are an experienced heavy jet pilot you should be able to read my words and understand their technical meaning without needing my CV.

I'm not discussing the crew actions. They are what they are but some lessons could be learned from these actions, should we have proper information. Unfortunately I could only find 2 sentences about these actions in this report following 9 months of investigations.

As you are an experienced heavy jet pilot please elaborate on how stalling an heavy jet 170 ft above the ground under AP and with retracting flaps is a skilled maneuver as to decide to let the AP try to follow a glide which is leading far beyond the runway threshold.

For the flaps, your heavy jet will certainly have less drag at 25 than 30, provided that you increase the speed (since your AP is following the glide slope). If you don't increase the speed to balance the loss of lift, you will need to increase the AOA and your nose up heavy jet will probably generate more induced drag than you could save with your 5 degrees flaps retraction. Anyway, in order to extend engine out landing distances, full flaps should be used for the final flare.

Do I need to be a NASA test pilot to expect you to consider my words ?

NigelOnDraft

Abolute B****ks

NoD

atakacs

My thoughts exactly !

Should be relatively simple to implement. Actually I would propose a BA038 EICAS message...

DC-ATE

sispanys ria -
Thank you for your supporting remarks regarding the flap issue. Not that I need it, but it's nice to know you're not alone!

Basil -
>>>....."discussing the actions of the flight crew who, following a long flight, had seconds to consider ways of ameliorating a very confusing situation. The aircraft landed within the airfield perimeter; no one was killed; rejoice!"

You are absolutely correct in that. That is NOT what I was attempting to prove/disprove, however. As -sispanys ria- wrote, we're trying to learn something here should another pilot/pilots ever be faced with a similar situation in the future. I stand by my remark that I think it was improper to raise the flaps at that point in time under the conditions that existed. BTW, I'm a retired DC-8 Captain. Not that that makes me any more qualified than many others here. But I DO know what it's like to have the flaps dumped on me! First thing that happens is.....you guessed it; airplane goes DOWN.

Now.....back to the report. I still question the preliminary findings about ice/fuel. Granted, I wasn't qualified on the 777, but I'd be interested to know about the fuel heat on board that machine. I was qualified on the 737 and we had to turn on the fuel heat manually; and did so often after observing the fuel temp gauge. Don't recall now the temp at which it was to be turned on, but it was something that was watched even though flight legs on that little bird were obviously much shorter than the flight in question.

The DC-8 had a fuel/oil heat exchanger that the crew had no control over and I've been in the air for 9 hours (KORD - PHNL) with no known problem.

Let's hope when the final report comes out we really know what happened so that it may never happen again.

Desk Jockey

Quote:

I think that BA/Boeing/RR need to respond quickly. There seems to be a groundswell of (Albeit uninformed) public opinion developing that until the problems are clarified and resolved, better to avoid the 777. This was "explained" to me as follows: "If the chances of such a random event are remote, statistics dictate that it could happen again today. In the meantime, I prefer to fly 4 engined planes." Never mind that this is not PEK in winter.

I don't think It's usual for Airlines to comment while an investigation is ongoing unless it's in answer to media questions. In fact BA did answer questions soon after the event.
As far as technical action is concerned airlines and engineers don't wait for investigations to report before taking steps to try to prevent another event whether the incident is their own or another airline.
In this case fuel was such an obvious common denominator steps would have been taken to focus on fuel and fuel systems servicing and maintenance and repair and their operation by the crew.
Safety information is shared in the aviation community, no one would want airlines to compete on the basis of safety records because the two are not compatible.
It's what the travelling public would expect and it's what they get.
As far as marketing gurus are concerned airlines don't overtly market safety. Safety statistics are around if people want to look.

barit1

Brainstorming for a minute - only half-facetiously:

Inasmuch as fuel is the common link in BA038, 2 or 4 or 6 or 8 engines would be equally vulnerable.

The obvious answer is a dual-fueled engine - a few minutes' supply of hydrogen (or methane) under pressure would at least allow you to pick a better crash site.

sydneymike

I think you're missing my point, which is, I conceed probably off-topic and anecdotal. The fact is, however, and the commercial side of the industry does need to deal with this, there is a growing PERCEPTION (Which is not the same as technically proven fact) amongst the travelling public that the 777 is not safe and that 4 engined planes are safer.

Rationalize it as you will mate.

md80fanatic

A good start would be to -reduce- the presence of additives in jet fuel, some of which are designed to absorb water (dessicant). What was once a smattering of tiny manageable crystals of ice is quick becoming larger chunks that cannot always pass cleanly through the system.

sooty655

barit1 said
It would have to be a fuel storable as a liquid to get the package down to a reasonable size - maybe propane - and it would still require a large, heavy, high pressure bottle. Why does that remind me of the Qantas B744 depressurisation??

Maybe a simpler alternative would be to provide a second fuel feed pipe from tank to engine. At present, the engine's ability to "suck" fuel only protects against a boost pump failure - a dedicated pipe would protect against ANY fault in the normal supply system.

Smilin_Ed

I think pax2908 has it right here. Following up on this would also prove, or disprove, snanceki's embracing of the accretion vs. break-off scenario. I agree with snanceki.

grebllaw123d

I think that I have read all the posts on this thread.

But I do not think that the following questions have been put forward - correct me if I am wrong!

Anybody knows the dimensions of the different fuel system pipes - especially the pipes coming from the wing tank booster pumps and terminates at the engine HP pump - the part of the fuel system where the AAIB suspects the blockage may have happened.

How many meters approx. and what is the diameter of the pipes?

I am just curious.

overthewing

I haven't seen this mentioned before, and didn't know that additives are water absorbers.

Having once overdosed a garden pot with water-retaining gel granules, I'm aware of how dramatically such materials can expand. (Looked like the patio had been attacked by flying jellyfish.) If the additives in fuel are in any way similar, is it plausible that they could form lumps large enough to cause problems?

If the fuel contained very little water, might this be a reason why?

Not a pro, ready to be put right...