On Wed, 02 Dec 2015 20:04:37 -0700, Don Y
wrote:


I.e., why would it "leak" then NOT leak a second later
(when power applied from the breaker).

And, when *hot*, not leak for 8 hours of continuous use;
not leak after being disconnected for 1 second; not leak
after being disconnected for 10 seconds; then "leak"
when disconnected for 2 minutes??

Let's say you have a leakage of 4 mA and the GFCI is designed to trip
at a leakage of 3 to 5 mA. At 4 mA, it may or may not trip. For
whatever reason, when first turned on, its trip point is 3 mA (within
spec) but it then rises to nearer 5 mA (still within spec). Wouldn't
that explain what you are seeing? I made up those numbers, but you
get the idea. Look for a small leakage of current from hot to ground
(or a low resistance connection from neutral to ground) that is on the
hairy edge of tripping the GFCI. As others have said, water or bugs
in one of the boxes on the line is the most likely. A neutral to
ground short in one of the boxes is also possible. Good luck. These
things are not easy to find.

Pat