On Mon, 18 Feb 2008 15:31:12 -0800 (PST), wrote:
Doesn't it still need to be grounded to the tank?
yes it does,,,,,,,,,,,, no metal to metal contact no protection.....

In addition to the prior quoted article from Rheem which says to use Teflon
tape on the sacrificial anode, these guys on the plumbing forum ran an
experiment by wrapping 20 layers of Teflon tape around a threaded coupling
and then measuring the electrical resistance:
http://www.plbg.com/forum/read.php?1,285316

Here's what they say about it.
Donna

"I just put about 20 layers of teflon tape on each end of a nipple and
tightened one into female galv an one end into female copper. I then used a
multimeter to measure the resistance from one fitting to the other, through
the two teflon-tape joints. The resistance was less than a tenth of an ohm
(i.e. short circuit).

When I did the same across a dielectric union, I got about 2 mega-ohms
(i.e. there's just a tiny bit of conduction through the water, but the
metals aren't touching)

So the business about teflon tape being just as good as a dielectric union
is complete nonsense. The threads just cut through the teflon. It may work
in some cases, but it won't work in others and should not be recommended.

What's amazing to me is that there are all these "old plumbers' tales" out
there, so people are arguing about whether brass or teflon or stainless are
good or bad. Shouldn't this be scientifically determined? I realize that
corrosion happens over many years, but still, there must be ways of
measuring corrosion in the lab. It shouldn't be a matter of opinion or
first-person stories (anecdotal evidence)."