2, 3, 4 engines-- what's actually safer?

From:         ckd@eff.org (Christopher Davis)
Organization: Electronic Frontier Foundation Tech Central
Date:         16 Mar 93 21:57:48 PST
References:   1 2 3
Followups:    1
Next article
View raw article
  or MIME structure

DAT> == Daryle A. Tilroe <data@beast.math.ualberta.ca>

 DAT> If my memory serves me a four engined aircraft is safer than two at
 DAT> HIGHER engine reliability.  I believe that four engines are safer at
 DAT> any p failure less than one third (which is VERY unreliable).  Thus
 DAT> for any reasonable failure rate four engines are better than two.
 DAT> BTW I am dreging this up from my recollection of a statistics course
 DAT> many years ago so if anyone else remembers this problem feel free to
 DAT> comfirm or dispute.

I know this topic has been hashed over in RISKS before (the question of
independent failures vs. common-mode failures, the question of how much
sooner an engine will fail if it's doing 33% more vs 50% more vs 100% more
work, and so on) but I'm interested in looking at actual data.

I can think of a few incidents where having three engines was not much
additional help (the EA L-1011 that was missing O-rings on *all three*
engines, for example) or actually a drawback (Sioux City; if #2 hadn't been
in the tail, it wouldn't have done the same damage).  And, of course, there
are the cases of four-engined aircraft losing all four due to volcanic ash
ingestion or the like.  Even some of the twinjet incidents (Gimli) would
have been no different with three or four engines, since the nature of the
failure would have affected them all.

In particular, I'm interested if anyone can think of an incident in which a
trijet survived (for some value of survived; a semi-controlled crash
landing, such as the Sioux City crash, would qualify) where a twinjet would
have (presumably) not done so.  (Arguably, the above-mentioned L-1011 might
qualify, since the early shutdown of one engine allowed them to fly just
that much longer...)

* Christopher Davis * <ckd@eff.org> * <ckd@kei.com> * [CKD1] * MIME * RIPEM *
226 Transfer complete. 17512509 bytes received in 5.2e+02 seconds (33 Kbytes/s)