Lou Scheffer wrote:

w_tom wrote in message ...
Good idea if only using theoretical knowledge. [...]
Why is the refrigerator,
specifically demanded by code, not on a GFCI? Because GFCIs
are good in some places and not desirable in others - as has
been proven by experience. A blown GFCI on a refrigerator can
create food poisoning - something learned by field
experiments.

This would be a great example for considering all the effects of a
'safety' improvement. Has there ever been a paper, or a summary of
the field trials, published on this result? Something that said, for
example, that each year, in the USA, X people get electrocuted by
refrigerators. This could be virtually eliminated by requring GFCIs,
but then Y people would die of food poisoning, and YX.

Or is it considered 'obvious' that this is the case? For example,
refrigerators are usually grounded, so maybe X is near 0, and it would
certainly be an inconvenience to have a GCFI breaker trip and spoil
your food.

Basically, does anyone know of any formal study or review on this
topic? I could not find any on-line, but such a study would probably
predate the internet.

Thanks,
Lou Scheffer

The National Electrical Code requires the metal case of
refrigerators and freezers to be grounded, rendering the
issue you raise meaningless.