PDA

View Full Version : Air Transat thread (cont)


Ranger One
2nd Sep 2001, 07:14
We reached the 100 mark... we left off the old thread here:

Old thread, final page (http://www.pprune.org/cgibin/ultimatebb.cgi?ubb=get_topic&f=1&t=015188&p=7)

Here's a thought, provoked by Airboeings post which mentioned the enormous proportions fuel leaks can assume: this wasn't a catastrophic situation, the crew had time to analyse the problem and gather data. If you had been on that flight deck, do you think it would have occured to you to call the cabin and ascertain whether there was a *visible* leak, and whether your actions had ameliorated it?

This, and the Concorde thread, have been some of the best I've seen lately - some very thoughtful contributions. I'll be very interested to hear the final verdict on this one, it definitely falls into the 'can't happen... but it did' category.

R1

[ 02 September 2001: Message edited by: Ranger One ]

Flight Safety
2nd Sep 2001, 09:12
I find myself having to agree with the summary of events that Skybloke gave on page 6 of the previous thread. The events can be summarized in the phrase "they left the #2 engine running", which Skybloke used more than once in his summary. Clearly this was not good fuel management once they determined they had a fuel leak, and this explains the massive loss of fuel.

411A
2nd Sep 2001, 09:18
One thing is for SURE, we can all learn from this incident, and as a bonus, no injuries or bent equipment.

Apollo
2nd Sep 2001, 11:31
Ranger One....

I Think it would be pretty difficult to see a fuel leak at night coming from an engine.I'm sure it may also spark a bit of unrest with the pax with you poking your nose at the window.

With regards to the de-pressurization. I recall the pax stating that the masks dropped. Is there no help from the RAT for this system?

tunturi
2nd Sep 2001, 12:13
Regarding pressurisation problems; The outflow valve/s would have driven closed under stanby electical power (Rat/battery)and the cabin leakage rate would then determine how long before cabin altitude equalled aircraft altitude. Depending upon the condition of door seals etc. it could take some time. It (cabin alt) obviously reached the mask trigger point (approx 13000ft ??on Airbus) but the leakage rate is probably a lot less than most might think, maybe 700 ft/min for example.

This is not like an explosive decompression and can be equated to the volcanic ash scenarios in Indonesia (BA747) and Alaska (KLM 747) where all 4 engines flamed out and the crews were faced with conflicting requirements of best glide speed versus high airspeed for windmill relight of engines and cabin pressuisation.

RatherBeFlying
2nd Sep 2001, 18:04
AIRBOEING's statement that the probable positions of the fuel shutoff valves: null the LP by the wing spar (generally placed well away from the engine so that it can do some good in case of fire),
null the HP with the engine supports the pilots' action to restore/continue the crossfeed -- fine as long as the LP shutoff does not fail and how do you catch that in time?
So, we may have either: null two component failures -- pipe and LP valve (with invidious implications for maintenance)
pipe rupture above LP valve (less likely, but if so, a design shortcoming).
In either nasty case, a crossfeed valve is your last line of defense, as long as there's enough accessible fuel for the remaining engine(s) to get to an airport.

foxmoth
2nd Sep 2001, 18:39
Gulfer - I did say in my post that the 330 normaly does better than the 3nm/1,000', but obvoiusly the NORMAL slope would be degraded by windmilling engines and no residual thrust (though you would probably get this back by comming down at a more efficient speed)
CaptainX - sorry, but I would rather have enough fuel on ONE engine than insufficient on two, fine, start of that way, but then monitoring the situation should tell you it is NOT working and time to shut off the US side.
N.B. anyone tried dead stick in the sim?
not tried it yet, but I would probably aim for slightly high at 5 miles (say 1,600 - 1700') and clean, then drop everything and accept what I got on landing - probably fast, but safer that way than short and slow.

sky9
2nd Sep 2001, 21:13
Wouldn't the situation been a great deal easier if aircraft were equipped with a FF meter on the engine and another in the fuel line within the wing structure.

I am left with a feeling that the overcomplicated automatic fuel transferring on the A330 has contributed to any confusion that the crew may have had establishing whether they had an engine fuel leak or a wing fuel leak.

Nothing these days seems as simple as the checklist suggests, I suspect because the checklist was drawn up with both eyes on the product liability laws.

After every incident like this I question "how well would I do in that situation?" I have yet to find a situation where I know that I would have done any better and incidents like Soux City fill me with admiration for the guys involved.

No doubt the experts will tell us how to do it better; with the benefit of 20/20 hindsight.

Roadtrip
2nd Sep 2001, 21:21
1.AirTransat employes convicted drug smugglers as captains??

2. Did the captain try and isolate the fuel leak or not (crossfeed off)?? Yes or No. There appears to have been plenty of time to "fly the airplane, analyze the situation, and take appropriate action."

This whole episode smells, and it's not getting any better as more facts get known.

What kind of airline is AirTransat???

fly4fud
2nd Sep 2001, 21:51
all right, my turn now :rolleyes:

"great job", "congratulations to all", "well done", etc????? Come on guys, wake up and remove your shades :cool:

It looks to me that this once more a case where pilots have endangered a flight and then saved it :o Call'em heroes if you want, but don't forget one thing, pilots are (still) here to save the day and not to make the situation worse. The same was true of the AC story. Those deadstick landings were good (and extremely lucky ;) ), but guess who first placed the aircraft in this situation:

- Air Transat, rumor has it of a fuel leak on one side, all the fuel dumped overboard thru the xfeed valve, great... :(
- Air Canada, aircraft refuelled with liters, indications in pounds and pilots not knowing how much they have... :(
- Aero Lloyd, gear down flight, fuel consumption relying only on the FMS (who cares about the fuel gages anyway?)... :(

Understand that those lines are not written to attack colleagues, but more to the point, to point out how quickly a highly trained group of professionals can be mislead by the circumstances, stress or just fatigue :eek:

Capt PPRuNe
2nd Sep 2001, 21:52
I believe the A330 arrived overhead the airfield at about 15,000'.

As this is a technical discussion I am moving this thread to the Tech Log forum.

Ranger One
3rd Sep 2001, 01:54
Apollo:

Doh. I'm short on sack time and I'd completely overlooked the fact that this incident took place at night.

zzzzzzzzzzz

R1

broadreach
3rd Sep 2001, 02:00
Thanks Danny for the reminder re altitude they arrived over Lages at. I believe it was mentioned only once during one of the press conference newspaper stories.

Rockhound
3rd Sep 2001, 04:27
Capt PPrune,
In your message above, were you referring to Air Transat TS 236 overhead Lajes (as Broadreach seems to think)? According to Paul Koring in the Globe and Mail of Aug 31, at 0634 hrs local, 8 mins after the left engine failed at FL 345, the aircraft was at FL 130 and 13 miles from the field. The landing run ended at 0646 hrs.
Apologies if we're not talking about the same thing.
Rockhound

Scud Runner
3rd Sep 2001, 06:10
Obviously, there are still details to be learned before we pass final judgement on this incident.

Some of you are suggesting that there needs to better technology to prevent mishandling of a fuel leak on what is already a high technology aircraft. Others think the A330 fuel management system is overly complicated, an opinion which is completely without factual basis. The A330 and its' relatives have a very simple little feature which can be quite helpful in this regard. It is called EFOB (estimated fuel on board) at destination. It seems to me that keeping an eye on this figure, combined with consistent fuel checks over waypoints, should yield clues to fuel leaks rather simply. Why complicate things further?

It's a poor chef that blaims his utensils!

Scud Runner

Traffic
3rd Sep 2001, 13:03
Capt PPR

Is someone seriously suggesting our friends arrived overhead Lajes at FL150??

I have seen two numbers for the FL at which the second engine stopped FL 320 and now FL 345.

Either way how do you travel 100nm deadstick from either of thsoe levels and arrive overhead at FL 150??

The original propsition that they departed FL 320 and got there by a foreskin sounds more believable unless there is something the 330 can do that no other a/c can manage.

Rockhound
3rd Sep 2001, 15:42
Today's Globe and Mail carries an exclusive by Graeme Smith, which reports that, five days before the dead-stick landing in the Azores, a senior Air Transat mechanic advised his supervisor that the aircraft was not yet ready for line flight because not all the work associated with an engine change had been completed. He was overruled by his (non-union) supervisor and the aircraft was released for duty. The mechanic recorded his conversation with his boss on tape. An AT spokesperson confirmed that the supervisor was suspended with pay last week.
Rockhound

Capt PPRuNe
3rd Sep 2001, 18:48
Just stating what I overheard some AirTransat pilots saying about the incident. I cannot confirm that that is the actual case and we will have to wait for the official report to come out.

It would seem a bit more logical if there was enough height remaining after the second engine flamed out to arrive overhead and then descend from there for a deadstick landing rather than plan for a one off approach from some distance away.

This weeks Flight International has a report that says DeJager (the F/O) said that he made a fuel check at 0457 (z?) and at that time there was sufficient remaining. 28 minutes later, one hour before the fuel starvation he first reported a fuel problem and requested a diversion to Lajes. The emergency was declared 23 minutes later at 0548 and according to the report he told ATC that the right engine had failed 25 minutes later at 0613. Around 13 minutes later at about 0626 at FL320, 100nm from Lajes the left engine stopped. The crew prepared the a/c for ditching but made a flapless heavy landing 20 minutes later at 0646.

It would need to be verified what distance they were from Lajes when the second engine flamed out to determine whether they made the deadstick approach from some distance away or from overhead the field. Either way it will have been a high adrenaline exercise and having had to deadstick a measly C182 from 4,000' overhead an airport I do not envy those guys.

[ 03 September 2001: Message edited by: Capt PPRuNe ]

Techman
3rd Sep 2001, 21:52
This incident is starting to smell of more than just fuel.

Rockhound
4th Sep 2001, 03:17
From everything I've read and heard in the Canadian media, the aircraft made a straight-in approach to Lajes Field from the northeast. A reputable source is Paul Koring of the Globe and Mail newspaper, who is considerably more knowledgeable on aviation-related matters than the average Canadian journalist; Koring wrote extensively on SR 111. In his article of Aug 31, he gives the following sequence of events (all times local = Zulu):
0458hrs: Fuel check by crew shows no abnormalities. A/C is NNW of the Azores at Fl 390.
0536: Crew notices fuel imbalance.
0541: Concerned about fuel quantity, pilot changes course for the Azores, 540 km away to the SW.
0548: Pilot decides there is a fuel leak and declares emergency.
0613: Still at FL 390, right engine fails, 217 km from Lajes.
0625: At FL 345, left engine fails, 137 km from Lajes.
0634: A/c has reached FL 130 and is 13 km from Lajes.
0646: Landing at Lajes.
Note that my earlier post contains an error: at FL 130 in the glide, the a/c was 13 km, not miles, from the field.
Rockhound

Traffic
4th Sep 2001, 09:52
Maybe we should wait for the FDR numbers to come out. I don't want to be anally retentive but the data put forward by Koring doesn't seem logical.

A STRAIGHT-IN approach with an initial step-down from FL 345 to 130 in 9 minutes covering a distance over the water of 137-13 km = 124km at a TAS of about 450kts.The rate of descent sounds achievable but 450kts?

Then the final 13km travelled in 12 minutes, meaning he did get within spitting distance with plenty of altitude that then needed to be bled off, belying the straight-in theory.

Sounds more like an excocet missile with a parachute than a scarebus.

Scud Runner
4th Sep 2001, 14:45
What will be most interesting from the CVR and DFDR will not be the glide performance, but rather, how this Captain got into this mess in the first place. The underlying rumours, if true, are quite disturbing. Since they are still only rumours, posting them would be premature.

Wait for it!

RatherBeFlying
4th Sep 2001, 23:51
Paul Koring has written several avaition pieces for the Toronto Globe and Mail's over quite a long time. While he may be better than the average run of popular press writers on aviation, in my opinion he still has a considerable way to go.

One of his earlier pieces on this incident implies severe criticism of the pilots for leaving a crossfeed valve open. Assuming that the checklists call for the crossfeed to be open once the engine is secured (LP and HP shut) and the difficulty of seeing even a heavy leak at night, my opinion is that Koring has taken a position while important facts are still missing. But it was an eloquent article that would appeal to the general public.

Skybloke
5th Sep 2001, 15:52
Rather Be Flying,
I`m not sure if you appreciate that there are TWO possible drills for the fuel leak case.
1. If the leak can be confirmed as coming from the engine then it calls for a shut down. Once that has occured the X-Feed can be opened because it is upstream of the LP valve and should not therefore pump fuel under pressure through the leak.
2. If the leak cannot be identified (probably likely in their situation) then the drill says keep X-feed closed and descend to gravity feed ceiling which in their case would be FL200. Once reaching this level ALL pumps are turned off in an attempt to minimise the fuel leak.
It would appear what they did was keep both engines running at altitude, which is rather like doing a bit of both drills.I figure if they had done EITHER drill completely they would have probably ended up doing a SE approach and landing.
Just my fourpenny worth.

PPRuNe Towers
5th Sep 2001, 18:31
Hope Keiron and the folks at ATI don't mind but they have put something together that does clear up some queries. Obviously we all await the eventual official reports but this is the best precis I've seen of what is known publicly at the moment.



Flameout stopped A330's FDR & CVR for last 20min of flight
Chris Kjelgaard, Washington DC (04Sep01, 22:44 GMT, 633 words)

Portuguese and Canadian safety officials investigating why an Air Transat
Airbus A330-200 had to make an unpowered landing in the Azores on 24 August
will be unable to derive any information from the aircraft's flight
recorders for the last 20min of the flight.

The Transportation Safety Board (TSB) of Canada confirms the flameout of
both of the A330's engines stopped the supply of electrical power to its
alternating current (AC) power buses. As in many other commercial aircraft
types, the AC buses in the A330-200 provide the electricity supply to power
both the digital flight data recorder (DFDR) and the cockpit voice recorder
(CVR).

The A330's AC power buses are supplied with electrical current by the
generators attached to each engine and by the aircraft's auxiliary power
unit (APU). In the Air Transat incident, however, the engines flamed out
because of fuel starvation following a serious fuel leak and the draining of
all fuel also meant the pilots of the aircraft were unable to use the APU to
provide electrical power.

The initial findings of the investigation, led by the Portuguese civil
aviation agency INAC, show the A330's right engine - the engine in whose
fuel supply the leak occurred - flamed out at 06:13 GMT on 24 August, but
that the left engine of the aircraft remained lit for another 13min, until
06:26 GMT.

At this point the A330 was still at an altitude of 34,000ft and was 85nm
distant from Lajes airport on the island of Terceira, at which it landed at
06:46 GMT after gliding without engine power or main electrical power for 20
min. At 06:39 GMT, 7min before landing, the aircraft was still at an
altitude of 13,000ft and was 8nm from the threshold of Lajes airport's
runway 33.

Neither the DFDF nor the CVR was operational during the last 20min of the
flight as a result of the loss of electrical power. This is not preventing
the officials investigating the incident from using the readouts from the
recorders in trying to piece together why both engines flamed out, as the
data from the period before the fuel ran out is likely to be of more
immediate importance in finding out what went wrong than what happened
afterwards.

Nevertheless, one official says that as well as understanding the Air
Transat crew's actions during the last 20min of the flight, the
post-flameout data could have been of considerable use to the investigating
agencies in modeling the flight characteristics and systems behavior of the A330 for future simulations of the aircraft in conditions of unpowered
flight.

The TSB has been concerned about interruptions in power supply to flight
data recorders of commercial aircraft ever since it began its investigation
into the crash of a Swissair MD-11 off the coast of Newfoundland in
September 1998.
In its first safety action arising from the Swissair Flight 111
investigation, the TSB recommended on 9 March 1999 that from 1 January 2005
every aircraft with a CVR offering a recording capacity of at least 2h
should have an independent power supply to power the CVR and the cockpit
area microphone for a period of 10min.
This power supply would allow the microphone and CVR to keep working
whenever the normal power source to each was interrupted.
§
The TSB also recommended that in aircraft required to have two flight
recorders, each recorder should be powered by a separate generator bus, to
minimize the possibility of an interruption in the power supply from one bus
affecting the recording capabilities of both the recorders.
In its Swissair 111 recommendations the TSB did not specifically address the
need for the DFDR in each aircraft to be linked to an independent power
supply, but the agency might review its position as a result of the findings
of the Air Transat incident.
Source: Air Transport Intelligence news

[ 05 September 2001: Message edited by: PPRuNe Towers ]

RatherBeFlying
5th Sep 2001, 21:00
Skybloke has commented that there are two drills for a fuel leak on the A330 depending on where the leak is identified. I would like to know what those drills are and am interested in knowing what guidance and/or instrumentation the crew has in identifying where the leak is so that they can know which drill to follow. The information in the last post that the second engine flamed out at FL340 indicates that the pumps were likely on as the maximum gravity feed altitude is FL200.

Paul Koring in today's Toronto Globe and Mail indicates that in the case of a fuel leak, the "trim" (stabiliser?) tank would first drain into the less full tank and prevent the fuel imbalance from developing until that tank is emptied -- in this case gone overboard.

In the same article, the new engine was a earlier model that AT maintenance had not installed before and found itself short of parts to conform to one service bulletion and used a procedure in another SB.

Old Aero Guy
6th Sep 2001, 04:46
With regard to powering the FDR when all engines are inoperative, remember you need more power than that required to run only the FDR. For the FDR data to be useful, you must also power all the sensors used to supply the FDR with data to record.

Having both engines inoperative puts a premium on available electrical power. Battery and/or RAT generated electrical power is devoted to essential flight systems needed to maintain control of the aircraft in a potential IFR environment.

A fully functional FDR would require a significantly larger set of batteries or larger/multiple RAT's. The airframe builders have been unable to see this as a good trade in terms of the increased equipment, weight and maintenance costs. After all, although the Air Transat dead stick landing was a significant piece of airmanship, the data necessary to understand the incident was recorded before the engines quit.

Flight Safety
6th Sep 2001, 23:48
I agree with Old Aero Guy, the high priority use of available electrical power after all engines are shut down, must be given to the critical systems and instruments that help the pilot get the aircraft down safely. All other priorities for use of the limited standby electrical power are secondary.

After thinking about this for a while, perhaps the current methods of supplying power to the FDR and CVR are correct just as they are. If the CVR unit in this aircraft has a 30 minute tape (correct me if I'm wrong about this), then the CVR stopped at just the right moment when the left engine flamed out. As others have pointed out, the events leading up to the fuel starvation of both engines, are far more important to the investigators that what happened afterwards. Because the CVR stopped when the left engine flamed out, the maximum amount of CVR data relevant to the events leading up to the fuel starvation, were preserved. BTW, I think that if anyone wants to know how an A330 performs in a glide, flight tests should be scheduled, which would provide a far better test data set wouldn't they?

To me, the most important question after what caused the fuel leak (which is already known), was what events lead to the premature fuel starvation of the left engine. This question is very critical regarding this incident, as a future event of this kind could send an aircraft into the water. The investigation may or may not implicate the pilots, but a thorough understanding of those events is crucial.

As the events unfolded, and with the FDR and CVR particularly stopping when they did, the maximum amount of data was preserved for the most critical part of this investigation. Perhaps nothing needs to be changed in the power sources of the FDR/CVR.