Menu Sign In Contact FAQ
Welcome to our forums

An unusual twin accident - fuel exhaustion?

Why he lost control during asymmetric flights is the real mystery here.

Not really. It is self evident – he got too slow.

If you fly an aircraft like that you owe it to yourself and your passengrrs to know your fuel system.

EGTK Oxford
With any significant pitch angle – and a 310 with just two POB can easily achieve that – there is a high risk of exposing the tank drain and feeding the engine with air instead of fuel,
How the hell can that get certified? Was it ever certified under Part 23?

The method of certification relates to the amount of useable fuel certified. My Twin Comanche has three pairs of tanks. The tip tanks and auxiliary tanks have (a) their full capacity declared as useable, and (b) a “level flight only” limitation in the AFM.

Part of the checklist for landing – and a critical part – is “main tanks selected”. The construction of the main tanks is the same as the above mentioned aux tanks, except that 3USG is declared unusable. The main tanks with 3USG in the bottom would mean no venting within certain credible pitch attitudes.

(As an aside, the twincom is also placarded with an alert to the risk of venting when there is the combination of low fuel in the main tanks and abrupt ground manoeuvring. You could end up with a big slug of air in the fuel lines that would make itself known at some point after applying full power – and probably on one engine only)

I only have a few hours on a 310, but recall the same item on the checklist for landing. I have not read the accident report, or know the construction of the 310 aux tanks, but it is possible that the failure in this case occurred due to pitch down on the approach. Perhaps someone might be able to comment?

The failure mode is – running down one’s auxiliary tanks in the cruise (good airmanship) – forgetting to select the main tanks for approach and landing (bad) – and then having to go around.

The twincom was certified under Part 3. I suspect the 310 as well – and have no idea whether the journey to Part 23 has any influence on the topic of this discussion.

…but don’t go around on one engine!”

As part of a LPC your circuit detail would include an EFATO with an asymmetric circuit with a low approach and go around, with a further asymmetric circuit to land.

The asymmetric circuit would include nominating an ACA (Asymmetric Committal Altitude), and you would be expected to countdown to this altitude and go-around asymmetric if not stable, runway not clear, not cleared to land – in the LPC you would announce a go around to demonstrate the ability to manage a go around asymmetric – while not descending below ACA. Once you descend below ACA you are committed to land. Some ATOs use 250’ AGL for ACA, I prefer 400’ AGL. In the USA MEP ACA is usually 500’ AGL.

Despite lots of briefing, a significant minority of students ‘lose it’ on their first asymmetric go around, which is a strong incentive for the MEP instructor to ensure good instruction on this element of the course. He is along for the ride, low level in a forty year piston twin with only some fibreglass for crash protection.

This accident is interesting as the pilot may not have been aware that he was asymmetric due to fuel issues (but pre landing checks should have covered the fuel status), and therefore did not apparently brief himself for an ACA. His decision to go around was below a typical ACA, so if he had briefed his brain should have been pre-set to continue and land on the grass.

As he only had 6 litres in the starboard main tank, this is somewhat moot, so even if he was current on an asymmetric go around he would have experienced a double engine failure on the go around. A further reason on this occasion to brief for a landing with no go around, in effect an emergency condition.

I believe Fedex take the ACA philosophy one step further – in addition to three gates on the approach (in effect red-blues and greens, approach stable calls for jets) – they brief all approaches as a go around as the default action, the landing decision being taken only when their approach checks/gates work out – which is good behavioral psychology.

The AAIB report might have given some helpful comment on the implications of this accident for renewed emphasis on systems knowledge, ACA and asymmetric circuits and explored whether the very recent LPC had covered these.

Statistics aside, I would suggest the current standards for an MEP rating remain quite minimal, and only adequate as a licence which provides a foundation for training to gain a Multi/IR where the old 50 hour course did raise the standard on drills.

Enstone (EGTN), Oxford (EGTK)

Robert, that is a very astute point. Even if he had gone around successfully, he would have ran dry very quickly.

I come back, a bit regardless of the certification process of MEP fuel systems, in that please put adequate fuel in your aircraft. I ALWAYS fill mine up, to the tabs, after every trip. Wherever I go, I fuel plan accordingly.

I still get guys at FBO, other pilots, speaking with me about duty drawback, go there to save 10p per litre, plan your trip to drop into Ireland to gain duty free fuel etc, etc, I say thank you for the advice, but fill her up please.

Those that fly for pleasure, and as a hobby, fly 200 miles for a coffee and a burger in a bun. It would be irony personified to die in the process, due to putting 10 liters a side in our planes, to save £2.00…

Fly safe. I want this thing to land l...
EGPF Glasgow

As I said earlier, I don’t get the desire to save a bit of money when you apparently burn more doing the landing and the climb back up. Or is there really an advantage? I don’t know the fuel burn of this aircraft.

In June I did a trip to Elba, and filled up to the top with Europe’s most expensive avgas while there. The flight was from Bergerac and back to Bergerac and the TB20 could have made those two legs without a refuel

and with about 2hrs’ reserve on the 2nd leg, but why cut off all your options in case of bad wx, etc? The saving is the fuel price difference multiplied by 3hrs of fuel burn – about €100. Or one could have taken on half of the empty space (reducing the cost of the “insurance” to €50).

Shoreham EGKA, United Kingdom

Fortunately I have only had one engine failure for real in a twin. Having set the aircraft up for the remaining engine, I recall, perhaps strangely, the one thing playing on my mind was what to do if the other engine quit. Part of the problem was understanding why the first engine had quit, and maybe, in most circumstances, you will never be quite certain.

I would entirely agree with the above post that the emphasis during recurrent training is entirely on deciding on a point of committing to the landing. Above that point you will go around and it would not concern me to do so unless there was some real reason to doubt the second engine would keep turning. I guess that means at least having arrived at a reasonable conclusion as to why it stopped in the first place. If fuel exhaustion is a likely explanation then I can understand why you might committ to a landing at “any cost”. Clearly a second engine failure on the go around is going to be incredibly dangerous and in the case of fuel exhaustion more than likely to occur as you bring the power and pitch up or shortly after. You will also have at the back of your mind how the aircraft is configured. If the carriage is down and the flaps are deployed you have already eaten into your performance and you will be thinking about whether you have a pump left to deal with the draggy bits, or the pump has gone with the failed engine.

So I think much depends on the twin, and the circumstances. As we know the “trouble” with so many lights twins is that there single engine performance is marginal at best so inevitably there is going to be a debate between the hazards of completing the landing even if there are some elements of the approach that are not perfect, contrasted with the risks of another circuit.

Rightly or wrongly, if I have options, that is another good reason for selecting the very longest runway available so everything is going your way on ensuring the approach is as stable and solid as possible.

As I said earlier, I don’t get the desire to save a bit of money when you apparently burn more doing the landing and the climb back up. Or is there really an advantage? I don’t know the fuel burn of this aircraft.

No, there is no advantage at all. The only reason for that would be issues with take-off-mass which forbid to fill the tanks for the whole trip. But with only two POB, they could have filled all tanks easily. Fuel burn is in the region of 150-180 lb/hr in the cruise and maybe 300 lb/hr for take-off and climb.

How the hell can that get certified? Was it ever certified under Part 23?

Most certified aeroplanes have limitations that, when not observed, lead to dangerous or deadly situations. The Cessna twins were and still are part 23 certified all over the world. They have plenty of safety built into them to prevent this kind of accident from happening. The AFM explains the fuel system in detail, the checklist mentions the required fuel selector settings, there are placards on the intrument panel and there are even two bright orange lights next to the fuel gauges that light up when the aux. tanks are selected. What more can you ask for?

but how does the tank return get handled in totalisers on such engines?

The fuel metering on the Cessna twins works very simply by measuring the fuel pressure at the injector nozzles. From this pressure, the mass flow is then calculated. This way, no attention has to be paid to the return flow. Using flow meters, both the forward and return flows would need to be measured. (In the 1980ies I installed an aftermarket “onboard computer” – the latest in gadgetry then… in my car that I had gotten cheap on an electronics fair. This actually came with two flow meters – only that my very simple carburetted engine did not have a return line at all).

EDDS - Stuttgart

Performance is marginal in some light twins…Seneca 1 is a shocker, for example…but is perfectly adequate in most light twins, certainly in the later Senecas, Baron, Aztec, PA 31 series and in the 300 and 400 series Cessnas.

This should have been a non-event, whether he opted for a go-around or land, but I agree that a go-around would be a bad idea, not because of performance, but because the other engine might have failed.

I have had quite a few engine failure in twins, and, of those, most have been fuel related. The trouble with a fuel related failure is that the engine surges, just as described in the report, meaning that it is impossible to identify from Dead Leg-Dead Engine. However, there is no doubt looking at the gauges, as one of the RPM needles will be all over the place, while the other is static.

On a wider front, with all the real engine failures I have had in twins, Dead Leg-Dead Engine has never helped once. I consider it merely as a technique for making life easier for test candidates and actually to be really dangerous if it’s the only weapon in the armoury when reality bites. It is far, far safer to make the judgement from the gauges.

The trouble is that there is no “engine fail” gauge on a piston engine, as there are circumstances which will allow any one of the gauges (ie MP, RPM, FF, EGT, CHT, TIT) to remain at pre failure levels, but there is no failure which will result in all the parameters remaining correct. I can assure you, after a lifetime of experience, that a quick glance will immediately reveal the failure, in one way or another.

So, I would beg people, once they have passed their MEP tests using the otherwise useless dead leg method, to only rely on gauges when the chips are down.

EGKB Biggin Hill

Would the EGTs not give it away instantly?

I happen to know that any fuel flow issue shows up immediately.

It’s one reason I moved the EDM700 right under my nose, rather than far right where Socata originally put them.

Shoreham EGKA, United Kingdom

Thank you for that advise Timothy, I must admit to being in he dead leg dead engine club, but will now change my ways. From your experience in real life, how much time do you have to identify and feather, also have you found it more or less to have total engine failure or a problem that causes you to shut the engine down.

Sign in to add your message

Back to Top