For fun, look up the MTBF of a human 
maxbc wrote:
I think you need to take MTBF*1.5 in your calculation (because e^-1 does not equal 0.5, it’s more like e^-1.5).
No, there is no need to take MTBF*1.5 because M in that abbreviation stands for “mean”, not “median”. 0.074 is the (1 – CDF) of the first failure, in other words, it’s the probability of the engine first failing not earlier than at 13000h mark.
My mistake, I was sure it was median ! (i.e. half engines would fail before xxx)
Taking the (correct) mean instead, numbers lean even more towards 10k and 30k.
“Basic engineering / operating common sense” tells us the number must be in the tens of k hrs. If it was much lower, as some say, then much GA would be dead. SEP flying would simply be totally unviable.
In other sectors e.g. 2-stroke UL engines (previous threads) it appears widely agreed that the number is in the 100-300hrs range, which seems staggering, but with a Vs around 20-30kt the outcome is rarely bad for the pilot and the failure will be mostly unreported.
Peter wrote:
“Basic engineering / operating common sense” tells us the number must be in the tens of k hrs. If it was much lower, as some say, then much GA would be dead. SEP flying would simply be totally unviable.In other sectors e.g. 2-stroke UL engines (previous threads) it appears widely agreed that the number is in the 100-300hrs range, which seems staggering, but with a Vs around 20-30kt the outcome is rarely bad for the pilot and the failure will be mostly unreported.
Agree, on the basis that:
1) In certified GA (and uncertified using lyco/conti) the vast majority of private pilots never experience an engine failure or anything close to it during their entire flying career.
2) Microlight pilots were traditionally trained on the basis of “expect to have to dead-stick it in every time”.
Yes but private pilots have a very limited number of hours and come nowhere close to these numbers. 3k hours is maybe the top 2-5% of PPLs, possibly less. For all we know the average career could be 200hrs. On such small numbers it’s hard to judge by the « average pilot ». You could easily have 10 pilots without issues, and a 3000hr MTBF.
Again small sample biased, my instructor has maybe 5k hours in GA (18k total) and experience two engine failures (at least one catastrophic).
Right, but if an engine fails after 1000hrs, and on average pilots pack up after 10 hrs, it would mean (crude stats but you get the idea) the first 99 would be fine and the 100th would get hit. So the individual risk is low (which is why I was happy to train in the crap I did my PPL in; my time in them was maybe 1% of the total 
) but the operator would still end up with a wrecked fleet.
And the instructors would be in hospital regularly.
Yes the operator primarily cares, but even for the pilot since we’re talking about potential lethal risks (especially over terrain and water), we can’t ignore minority risks above 1%. Another way to compare it would be that it multiplies your accidental death rate by 50-100 compared to not flying. So knowing if it’s 1,000 or 30,000 hrs is interesting. My current guesstimate based on the above is 15,000 to 30,000 hours MTBF. That would make turbines roughly 10-20x more reliable.
If a flying club in flordia does 1k hours per year I think reasonable to expect 1 engine failure every 5-10 years.
Most likely with an experienced instructor on board.
50/50 that it’s nearby the airfield
So you won’t end up with a hospital full of instructors I don’t think.
I reckon the serious injury/mortality rate of an engine failure with an instructor on board is close to 10pct so will never make the news or the insurance company;)
I agree.
Also we have the old chestnut of whether the crankshaft snapping ranks the same as total loss of power due to some accessory failure. I think most would include a failure of the ignition system as an “engine failure”, but what about carb icing which is not addressed quick enough?
So in this sense “the system” is a part of the discussion, and fuel injected engines are a lot more reliable than carburetted engines – especially as the conditions required for fuel servo icing (-15C, forgetting to use alternate heat) will never be seen by some 99% of private pilots.
So, while maintenance of the engine block barely affects reliability, getting magneto(s) looked at periodically will do so.
And to confuse things more, the choice of the engine shop for the engine overhaul will also have an effect
Re TP engines, one hears of too many failures to support the fantastic reliability figures sometimes mentioned, but much of that could be mismanagement…
