SWH is on the money - but, um, how do I explain the mathematics of probability in relation to multi engine aircraft engine failures.......?
Here's a try......
The PT6A-114 installed in Cessna 208 Caravans and the PT6A-67 installed in the PC12 have exhibited engine failure rates less than 1 in 100,000. In effect, assuming all maintenance requirements are complied with, an engine failure should not occur in less than 100,000 hours of fleet time. It would be true to say that the greater the world wide fleet size, the greater the chance of engine faulure, solely because there are more examples of the type.
Mathematically, the C208 and PC12 should be more reliable than (for example) the Heron, with four Gypsy engines (or four Lycoming engines if modified), which do not meet the same failure testing criteria.
Similarly, with advances in engine technology and reliability, it would probably be reasonable to assume the chances of engine failure in an older technology Boeing 707, would be greater than the chances of engine failure in a late model Boeing 747.
Back to the mathematics of probability: If you were comparing a Boeing 767 (2 engines) with a Boeing 747 (4 engines), the chances of an engine failure on the 767 are probably
identical to the chances of an engine failure on the 747.
Because the 747 has double the number of engines, does not double the chance of engine failure.
As I said above, it's not how many engines you have - it's how many are left after one engine fails......
And that's probably why the 146 has five engines..........
I know I haven't explained this very well, but, hey, it's early Sunday morning..........
Woomera