I decided to build a simple statistical model and compare three cases to get
a feeling for the relative expected reliability of them:

• single screw / single engine
• twin screw / twin engine w/ common fuel supply
• twin screw / twin engine w/ independent fuel supplies

The failure scenarios I considered were:

• internal engine failure (e.g., oil pump failure, throw a piston rod and
  the like)
• internal engine cooling failures (e.g., an impeller fails)
• external engine cooling failures (e.g., seaweed over the raw water intake)
• fuel system failures (e.g., filter clogs up with junk)
• internal running gear failure (e.g., transmission dies)
• external running gear failure (e.g., hit something and bend a shaft)

Obviously, these aren't the only things that can go wrong, but I think
they're representative enough to get a feeling.

I then guestimated probabilities for each class of event. For instance, the
claim that in 300 hours of running per year, you might see a major internal
engine failure once every 10 years of running or so.  These guestimates are
just that: just useful for order-of-magnitude calculations.

The results aren't quite what I expected to see, although they showed the
expected relationships.  The expected down-times, given my model were:
single engine: 26 min/year, twin with shared fuel: 17 mins/year, and twins
with independent fuel: about 5 minutes/year.

The numbers don't matter, what I find fascinating are 1) the relationships,
and two, the model's sensitivity to the various failure scenarios.

The relationships say that you're less likely to be immobile if you have
twins than if you have a single engine...but if they share a fuel supply,
the twins are considerably less than twice as reliable as the single. The
introduction of independent fuel supplies increases their reliability by
about a factor of 3. But they have to be truly independent to get this
benefit: for example, you might have to buy fuel from different suppliers to
really achieve this degree of independence.

The model was most sensitive to three failures: fuel, external cooling
failure and external running gear failure.  Failures such as internal engine
failure and internal running gear failure have much less of an impact
(because they're frequently truly independent events from one another,
whereas the external failure causes are frequently not independent. The
independent failure cases gain the big benefits of redundant systems, the
co-dependent failures cases also benefit, but at a much lesser level).

not hard to replace a filter nor does it take long to clear seaweed, once
you've figured out that's the problem.

The moral of the story, from my perspective, is that singles are only
slightly more likely to experience outage than twins...but they are
slightly more likely to experience outages. The big reliability gains come
when you can isolate the twins from related events: independent fuel tanks,
anything you can do to remove external sources of cooling failure, and so
on.

Jeff East
Sand Dollar
53' Fleming

I've been reading with fascination about the issues discussed over the past
few weeks. One of the responses said that the real issue was what you wanted
to do from the standpoint of displacement versus semi-displacement hulls. I
was just wondering, is that really the determining factor. It seems like the
majority (perhaps vast majority) of displacement hulls have one engine, and
the majority of the semi-displacment hulls have two engines. Some of this
may just be related to the power factors needed and the market that they are
designed to meet, but that also determines the options that are available to
us as consumers (unless we design our own, which is a luxury few have the
time or talent to do right).

Have any designers of boats done the modeling. My fear is that more of them
are based on the economics of producing the boat rather than the reliability
factors. For the average boater, there may not be much of a difference until
you hit that one time were you wished you would have thought of it ahead of
time.

It would seem like overall we fall into two categories

1. The world cruiser to who one engine will work, but two would be the best
   of all worlds if he/she could afford it and if it was available in the boat
   that they had or wanted to have. Usually, this person would have a full
   displacement hull (due to economics of cross ocean cruising). But you have
   to give up speed.

2. The near world cruiser (Caribbean and close to North/South America). One
   engine will work, you can afford two from the fuel and range standpoint, and
   maybe you want to get somewhere a little faster. Usually, you'd have a
   semi-displacement hull. But you have to give up range.

No matter how you hope, the laws of physics, laws of economics, and marital
laws will always be in conflict with the issue of having a boat that is
perfect. In my field, we have a saying: "Cheap, Fast, Correct - you can have
two - so pick". I'll change that you "Range, Speed, Cost - you can have
two - so pick"

-----Original Message-----
From: owner-trawler-world-list@samurai.com
[mailto:owner-trawler-world-list@samurai.com] On Behalf Of Jeff East
Sent: Monday, April 26, 1999 7:24 PM
To: trawler-world-list
Subject: RE: Single vs Trwin

I decided to build a simple statistical model and compare three cases to get
a feeling for the relative expected reliability of them:

• single screw / single engine
• twin screw / twin engine w/ common fuel supply
• twin screw / twin engine w/ independent fuel supplies

The failure scenarios I considered were:

• internal engine failure (e.g., oil pump failure, throw a piston rod and
  the like)
• internal engine cooling failures (e.g., an impeller fails)
• external engine cooling failures (e.g., seaweed over the raw water intake)
• fuel system failures (e.g., filter clogs up with junk)
• internal running gear failure (e.g., transmission dies)
• external running gear failure (e.g., hit something and bend a shaft)

Obviously, these aren't the only things that can go wrong, but I think
they're representative enough to get a feeling.

I then guestimated probabilities for each class of event. For instance, the
claim that in 300 hours of running per year, you might see a major internal
engine failure once every 10 years of running or so.  These guestimates are
just that: just useful for order-of-magnitude calculations.

The results aren't quite what I expected to see, although they showed the
expected relationships.  The expected down-times, given my model were:
single engine: 26 min/year, twin with shared fuel: 17 mins/year, and twins
with independent fuel: about 5 minutes/year.

The numbers don't matter, what I find fascinating are 1) the relationships,
and two, the model's sensitivity to the various failure scenarios.

The relationships say that you're less likely to be immobile if you have
twins than if you have a single engine...but if they share a fuel supply,
the twins are considerably less than twice as reliable as the single. The
introduction of independent fuel supplies increases their reliability by
about a factor of 3. But they have to be truly independent to get this
benefit: for example, you might have to buy fuel from different suppliers to
really achieve this degree of independence.

The model was most sensitive to three failures: fuel, external cooling
failure and external running gear failure.  Failures such as internal engine
failure and internal running gear failure have much less of an impact
(because they're frequently truly independent events from one another,
whereas the external failure causes are frequently not independent. The
independent failure cases gain the big benefits of redundant systems, the
co-dependent failures cases also benefit, but at a much lesser level).

not hard to replace a filter nor does it take long to clear seaweed, once
you've figured out that's the problem.

The moral of the story, from my perspective, is that singles are only
slightly more likely to experience outage than twins...but they are
slightly more likely to experience outages. The big reliability gains come
when you can isolate the twins from related events: independent fuel tanks,
anything you can do to remove external sources of cooling failure, and so
on.

Jeff East
Sand Dollar
53' Fleming

Another counterintuitive statistic is that, at least as of a few years ago,
the accident rate due to engine failure was higher for twin-engine aircraft
than for singles. Reason: pilots have high expectations for the ability of
twin-engine aircraft to continue after an engine failure and therefore take
chances -  and, of course, twins have twice as many engine failures as
singles all else being equal.  Pilots of single-engine aircraft have no
such illusions and land the puppy right away when the fan stops turning. If
they can.
Mike T.

Boy this place is into misleading information.  The incident rate for light
twins is less than that for singles.  The fatality rate per mile flown is
higher - by a good bit.

So the statistics on light twins are in fact worse than singles.  The
statistics on heavy twins are far better than singles.  The difference is
that heavy twins can fly on one engine and light twins can't. These
statistics have  other information if you wish to reach for it.  The light
twin pilots are generally better trained and rated than their single
contemporaries - yet they die more often per mile flown.  Could it be that
sophisticated, often well educated pilots try to fly airplanes where they
should set down quickly?  Could a well educated and trained group of pilots
ignore the numbers and do dumb things that get them killed? Any correlation
to the existing discussion?

Jim

-----Original Message-----
From: owner-trawler-world-list@samurai.com
[mailto:owner-trawler-world-list@samurai.com]On Behalf Of Mike Taylor
Sent: Monday, April 26, 1999 5:00 PM
To: Jeff East; trawler-world-list
Subject: RE: Single vs Trwin

Another counterintuitive statistic is that, at least as of a few years ago,
the accident rate due to engine failure was higher for twin-engine aircraft
than for singles. Reason: pilots have high expectations for the ability of
twin-engine aircraft to continue after an engine failure and therefore take
chances -  and, of course, twins have twice as many engine failures as
singles all else being equal.  Pilots of single-engine aircraft have no
such illusions and land the puppy right away when the fan stops turning. If
they can.
Mike T.

My recollection is obviously flawed - my apologies. Your response is also
somewhat misleading. Light twins will fly on one engine, but not under all
circumstances. Stipulate discussion about Vmc, configuration, density
altitude, weights, critical engines, etc. etc. (I am instrument and
multi-rated so know at least somewhat of what I speak). The point is that
having two engines creates an expectation that you'll be OK in the event of
an engine failure. The single engine pilot is under no such misapprehension.
Similarly having two engines on a boat might well be less safe than a single
because, for example, maintenance is neglected.

Mike T.

I have been following this discussion with interest and as we all do,
applying it to my own situation. I have a 44' power catamaran being built.
This will have two engines, skegs and rudders under each hull, completely
separate water intake and cooling systems, and a two fuel systems which are
not even cross-linked.

I am left with the feeling that the only factor apt to cause both engines to
go out at roughly the same time is fuel. I can see two ways in which fuel
can kill both engines. First, it may be too old, have too much water and/or
algae, be stirred up by rough seas, etc. In this case, it is likely that
both tanks though separate will be affected and therefore could bring both
engines to a stop. The second way is if you take on bad fuel.

If we assume that each tank takes, say 200 gallons, to fill, how likely is
it that if you get troublesome fuel in the first tank filled that the second
200 gallons will also be bad. It seems to me that both water and sediment
drop out of solution fairly rapidly and by the time you get to filling the
second tank, the fuel may be back in spec.

Does anyone have experience with getting a load of bad fuel from a normal
U.S. distributor? Bahamian? Did you figure out what caused it? Did it
contaminate more than one tank?

Tom

Alumina 42' trawler in Key West