Menu Sign In Contact FAQ
Welcome to our forums the scene of the crash (The second engine in a twin takes you to...)

Like Dave, I have had quite a lot of engine failures and shutdowns in twins. Unlike him, I have not had them in extraordinary temperatures and altitudes. I have had them in ordinary conditions. You know, the conditions we ordinarily fly in.

All of them have been pretty much exactly as in the video at the top of this thread. Shutdown, secure, set up for OEI cruise, engage autopilot and then a TDODAR of options available. Sometimes it has seemed important to land nearby (when it has not been obvious what the problem is, so could affect the other engine) other times the best option has been a transit to maintenance base (when the cylinder through the side of the engine cowling has made it fairly obvious what the issues is.)

There are a number of reasons why twin fatalities and statistics are up there with singles, and they essentially come down to risk compensation.

Simply put, twin pilots fly in circumstances when most single pilots would think twice (icing, long water crossings, mountains, night, weather minima etc) and that is usually what gets them. If twins were flown like singles they would be hugely safer, but the utility would be considerably lower.

That is no criticism. It is why I fly twins, because my dispatch rate is much, much higher than it would be in an SEP, because I am willing to take-off in R300, land in R550, fly into known icing, cross oceans and the polar ice cap, fly over the Alps, fly 1000nm at night and so on. None of these things would I do in an SEP (personally I wouldn’t be comfortable in an SET either).

So I accept that my risk in my MEP flights is much the same as in my SEP flights, which are mostly over farmland by day.

This is one of the things that makes the statistics very wobbly. But also a lack of critical thinking helps as well.

Incidentally, I am delighted that Dave is alive. He is alive because, even in those ridiculously difficult conditions, with an engine failure at the most critical moment, the second engine still didn’t take him to the site of the crash, but took him to a runway

EGKB Biggin Hill

Correlation is causation if there is only one variable (number of engines). Unless of course you believe the causal link is the other way (fatals cause engine failures).

You’re right about the fact that the engine failure rate is not double of course, but that makes it even worse for the safety benefit of the second engine.

Try etc.

P[ Fatal U EngineFailure | twin] > 2* P[ Fatal U EngineFailure | single]

Last Edited by denopa at 31 Dec 00:14

But this thread is not about fatalities, it is about crash scenes. That is the tabloid aspect of the misuse of statistics.

A very small number of engine failures in twins end in crashes. As Adam says, we cannot measure how small. But if we just take the contributors to this thread, Dave has had five engine failures in twins, I’ve had about ten. None of them has resulted in a crash. That’s nought out of fifteen.

Most engine failures in singles result in something that could reasonably be described as a crash. Again, we can’t be sure what proportion, but would 75% be a reasonable guess?

The pilots walk away from many of them, and are ambulanced away from many more.

So very few engine failures in twins result in a crash. Most engine failures in singles result in crashes, albeit non-fatal.

Fatalities are orthogonal to the argument. They are just thrown in to make a different point and muddy the water.

EGKB Biggin Hill

denopa wrote:

Correlation is causation if there is only one variable (number of engines).

But, as I said, there is a huge range of variables. Did you read about risk compensation?

May I recommend this?

EGKB Biggin Hill

Yes. I know about that. Thank you. My view is not coming from a lack of understanding of risk, statistics or conditional probabilities.


Well, you keep asserting

Given the fatality rate following an engine failure for MEPs is more than twice that of SEPs

Without giving any numbers, so where does it come from?

Your formula above simply restates your assertion in mathematical form.

Biggin Hill

That’s what I had been asked to do :-)

See this Mike Busch article. The interesting bit for this discussion is in the orange box page 3.


This orange box again only makes an assertion, no figures that even come close to proving the point. Some other figures, though, at least.

It may well be true, but so far – no evidence.

Last Edited by Cobalt at 31 Dec 09:08
Biggin Hill

Timothy wrote:

Dave has had five engine failures in twins, I’ve had about ten
Even without knowing your twin time, this seems like an awful lot.
Let’s assume (surely someone can chime in with figures backed up by statistics) that the non-fuel-related IFSR of an aircraft piston engine is 1 in 5000 hrs. This would mean that statistically you should have 25.000 hrs. twin time and Dave 12.500 hrs.
Is there anything that makes twins more prone to faliure of one engine than singles?

Friedrichshafen EDNY

I can’t talk for Dave, it sounds like his have been more varied than mine, but in my case a fair number have been a particular fuel icing issue in an Aztec. That was about five of them. Eventually I just took to mixing IPA with the fuel for every flight and the problem went away.

Then I had a cylinder detach, an induction pipe crack, two turbine disk failures, a bladder tank split and empty itself.

But I do seem to be unlucky with engines. My left engine is away having its third major rebuild in a year. None of the reasons for these rebuilds have been failures, they have either manifested themselves as major oil leaks, or been found on the ground, but it does seem unfair :-(

But my recognition of just how unreliable are piston engines does affect my choice to fly twins or, hopefully soon, a turboprop.

EGKB Biggin Hill
Sign in to add your message

Back to Top