Answers from "himself"
I posted my original question, and then several follow-up questions, on the Matronics AeroElectric forum, and received clear answers from Bob Nuckolls. I found the answers interesting, and educational. I thought it would be worthwhile posting the answers here:
Original Question:
>I am rebuilding the "Fat Wire" portion of my purchased RV-6A, moving
>the battery and contactors to the engine side of the firewall.
>Fig Z-11 does not show any protection on the 6AWG wire going from
>the Battery Contactor to supply the Main Power Distribution Bus. In
>my case, I am using a 12AWG wire, which passes through the firewall.
>Seems to me this wire needs protection of some sort, such as a
>fusible link, or an in-line fuse on the engine side of the firewall.
>
>Looking for suggestions/recommendations, thanks!
A 12AWG wire is probably too small to be
a bus feeder. When wiring with truly "fat"
wires (6AWG or larger) they're not at high
risk for burning due to shorts or overloads.
Take a look at the wiring diagrams for any
single engine TC aircraft and you'll find
that few if any will incorporate fuses or
current limiters in these pathways. This
philosophy is echoed in the FARS . . .
Sec. 23.1357 Circuit protective devices.
(a) Protective devices, such as fuses or circuit breakers, must be
installed in all electrical circuits other than--
(1) Main circuits of starter motors used during starting only; and
(2) Circuits in which no hazard is presented by their omission.
(b) A protective device for a circuit essential to flight safety may not be
used to protect any other circuit.
(c) Each resettable circuit protective device ("trip free" device in which
the tripping mechanism cannot be overridden by the operating control) must be
designed so that--
(1) A manual operation is required to restore service after tripping; and
(2) If an overload or circuit fault exists, the device will open the
circuit regardless of the position of the operating control.
(d) If the ability to reset a circuit breaker or replace a fuse is
essential to safety in flight, that circuit breaker or fuse must be so
located and identified that it can be readily reset or replaced in flight.
In particular, paragraph 12.1357(a)(2) applies here.
Faulted robust wires generally arc to ground and
burn their faults clear. Further, they're easily
installed with attention to mechanical details such
that faults to ground are as probable as loosing
one's propeller due to bolt failure.
The Z-figures are crafted with this philosophy
in mind supported by a confidence in nearly 100
years of field history. I'll suggest that none
of your fat wires should be smaller than 6AWG
and that protection beyond what is illustrated
in the Z-figures is no-value-added weight, cost
and complexity.
Bob . . .
Follow-up Question from another forum member:
>Bob,
>Trying to turn this into a learning/understanding moment, the above
>leaves me a bit puzzled as to exactly when the omission of a
>protective device would be considered to pose no hazard. Say one
>has an AWG-6 feeder connected to the switched side of the master
>relay and the conductor or a terminal faults to
>airframe. Alternator current would be limited by the alternator
>breaker, good; but battery current would only be limited by
>conductor, termination, device and internal battery resistances. At
>first estimate this would seem to be capable of generating a current
>well above the safe capacity of the AWG-6.
>
>This topic is of particular interest to me as we are currently
>working on electrical system modifications for a Bellanca BL-17 and
>found the local FSDO inspectors have conflicting thoughts
>with what you have stated.
>
>Appreciate any additional insight you may be able to provide.
The style of "protection" one might add to
fat wires in an array of battery/bus/cranking
feeders are like the ANL "current limiters"
found in the power distribution systems of
many larger aircraft.
http://aeroelectric.com/Pictures/Fuses/Fuses/ANL-ANN_Current_Limiter.jpg
Fusing characteristics for these puppies are
shown here . . .
http://aeroelectric.com/Mfgr_Data/Fuses_and_Current_Limiters/Bussman/ANL_Specs.pdf
Suppose you decided to put an ANL200 in the
battery feeder for your project. Note in
the fusing plots above the ANL200 will carry
300A without breathing hard.
Now, what kind of fault can you imagine
that ties your feeders down so firmly to
the airframe that you'd expect battery fault
currents of 1000+ amps to flow? You're
going to rub off some insulation and have
some low pressure, flying fault that
arcs a lot and intermittently draws hundreds
of amps . . . but is unlikely to open the
ANL limiter. In fact, you'll find that
the copper is barely damaged compared to
adjacent aluminum that simply burns clear
without even warming up your feed wire.
Have your skeptics research the wiring
diagrams for the host of S.E. aircraft
produced in the hundreds of thousands
for nearly 100 years and they'll note
a not so curious absence of fuses/breakers
in the fat wire feeders . . .
These wires seldom get faulted and when
they do, the event is so benign that
the current protection doesn't operate
before arcing burns the fault clear.
It's a lot of careful design and hard work
to bring two conductors together to be
an efficient conductor of hundreds of amps.
It just doesn't happen accidently.
Bob . . .
More follow-up questions from me:
>First, thank you for your response to my original question. If I am
>following your logic: A fat wire, if shorted to airframe
>ground, will burn away the nearby aluminum ground fault very
>quickly, and thereby eliminate the immediate problem.
>
>If that is so, then the next question would seem to be: Why protect
>the relatively short B-Lead that goes to-from the alternator, as it
>can be easily physically protected in a similar manner to the bus feed wire.
Because the fault risk there is not to the
wire itself but for the potential of diodes
shorting in the alternator. This failure
would probably burn wires in the alternator.
That's probably the least "useful" of the fat-wire
feeder fuses. Probability of diodes shorting
in modern alternators is exceedingly low.
That fuse or breaker can probably be eliminated
with little risk. Cars have never used
this fuse. Early cars burned some wires . . .
later ones added fusible links in the system
to limit the amount of wire that needed to
be replaced after a "burn".
>And a second question: Why use an AWG-6 wire to supply a main bus
>that only has a 20 amp max load?
Because to qualify as a fat-feeder with
little use for protection, the wire needs to
be more robust than the fault site conductors that
would put it at risk. We protect the little
guys because flying faults will often burn
them . . . a distribution fat-wire unworthy of
protection needs to be in the robust class.
I've never seen any rules of thumb but my
sense is that wires in the 10 to 6AWG range
are in a grey area and I'd probably fuse them.
A short 6AWG or larger bus feeder is probably
good to go without specific protection other than
extra attention to support and isolation from
potentially hazardous mechanical damage.