Paint breakdown at high voltage

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Christopher \"CajunCoiler\" Mayeux by way of Terry Fritz <teslalist-at-qwest-dot-net>" <cajuncoiler-at-cox-dot-net>


Anyone else having problems with paint breaking down under high
voltage?  I recently put a coat of paint on the outside of my RQ-style
spark gap, and it goes nuts with arcing when a 15/30 NST is applied to
test it... problem is, the arcing is on the OUTSIDE, around the heads
of the bolts that hold the electrodes in.  I also notice some lines
forming in between the bolt heads that closely resemble the trails
left by a mole...although on a MUCH smaller scale.

Next, I decided to make a simulation, to see what was going on under
a controlled condition.   I took a left-over slice of PVC, and drove
some screws and copper couplings in, making a simple replica of the
original gap.  Then I attached the 15/30 NST to the two end electrodes,
and connected the NST to my bench variac.  I cranked up the whole thing
at the 0 setting of the variac, and slowly turned it up.  The hiss of
corona was present, telling me all was well, all the way up to full
voltage.

Next, I covered the open ends of the pvc, and sprayed on 2 coats of
the paint in question, letting each coat dry solidly.  Then I took my
Fluke77 DVM, and tested to see if perhaps the paint was conductive at
this point... it wasn't.  So I ran it through the voltage test as
before, and this time, when the input was between 90 and 95 volts
(probably about 11,250v and 11,875v out of the NST), it once again
broke down, and started arcing and making mole trails.

Anyone else encountered this?  Or is it just me again?
If it is common, what workarounds are available?  BTW, I checked, and
the paint is NOT a metal flake or metal particle type.