[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Corona breakout voltages?



Original poster: "Jim Lux by way of Terry Fritz <twftesla-at-qwest-dot-net>" <jimlux-at-earthlink-dot-net>

> >But, were those "bundled" lines, which increases the effective radius?
>
> Yes,  Steel rope core with thick aluminum wires spiraled on the outside.

So called ACSR (aluminum conductor steel reinforced...) making use of skin
effect too...

But, bundled conductors are the usual thing you see on HV lines.. a pair or
quad of wires spaced a foot or so apart.


> Yes!!!!  that's what I was thinking...  The area around the conductor just
> charges up an no more corona...  I little thougher at 38kHz I would think
> but maybe not...

Yep.. there's even empirical formulae for how much power you lose to the
corona at a given voltage, given diameter transmission line, etc...

> >>
> >
> >HV is strange...
>
> This I have noticed :o)))))
>
> >
> >Mostly, there's a big difference between uniform and nonuniform fields,
in
> >terms of observed effects.  Running small diameter conductors at high
> >voltages, you've got a nonuniform field.
>
> If you have any further info or references on such effects, we ARE
> interested ;-)))  We are studying this on the TSSP list and we seem to
have
> opened a barrel of monkeys here ;-))
>
Oh ho... non-uniform fields are something that there's some literature on,
but not a whole lot.  Lots of empiricism, not much theory. Most of the
research is tied to working with EHV lines (say, 500 kV and up), mostly
because of the observed phenomenon that as you get above that, the clearance
distance needed gets longer a whole lot faster than you'd think from just
the field calculation.  After all, lightning occurs in an average field of
just a few thousand  V/meter, and that's the mother of all flashovers.

Part of it is just the fact that scaling up the radius of curvature gets
impractical at really high voltages (haven't seen many 1 Meter diameter
wires recently...), and the other is some basic phenomena with leader jump
lengths, and so forth. The latter is probably related to fundamental things
like mean free path, recombination times, and so forth.  There's probably
some basic reason why the steps in a spark are all pretty much the same
length, regardless of the overall spark length.

Some folks have looked at numerical models, but, for a variety of reasons,
they do 1D, 1.5D (assuming axisymmetric) or 2D models, not 3D.  The
underlying physics at a micro scale is pretty well known (at least compared
to other areas..), but there's a healthy random component, and numerical
simulation is going to take very fine grids, very fast time steps, and lots
of runs.  And, because the grid and time steps are so fine, compared to the
overall phenomenon, the numerical conditioning problems become significant,
so small roundoff errors cause radical changes in simulated phenomena.