I was pondering the concept of "monthly testing" of GFCI's the other
day, basically trying to determine if it had any merit. Here are my
thoughts...

What does a passing testing of a GFCI tell us? It tells us 2 things:

1 - That the testing circuitry worked at the time of the test; and
2 - Had there been a fault in the last month, there's a high
probability that it would have tripped.

What it doesn't tell us, in any certain terms, that the device will
work the *next* time there is a fault. A GFCI is an electromechanical
device and the possibility exists that the test we just performed was
the last time that particular unit was going to work.

Granted, if it fails the test, we know we should replace it, so we're
really checking for a failed device, not a working device. Perhaps we
should feel better (read: relieved) when the test fails, because we
have the opportunity to replace the failed device. Ah, but wait - once
we replace the device and test it, all we really know is that it passed
it's initial test - we still don't know that it will work when it is
needed.

You know that disclaimer the investment folks always use - "Past
performance is not a guarantee of future results"? It seems to me that
the same holds true for a GFCI.

I submit that we should not feel confident that a GFCI will protect us
just because it passed the test. In reality, all we can do is look back
and say whether or not it would have protected us since the last test.

Sleep well!