I'm reading up on some plumbing work I'm considering doing myself. I
have galvanized pipes throughout my house. I want to put in copper for
the new supply lines I want to install. I understand a Dielectric
union will allow me to "join" the copper with the galvanized iron and
protect from corrosion.
Continuing my reading I read someplace that since most older homes are
grounded to the plumbing the addition of a dielectic union will undo
The alternative the person recommended was to use a "brass nipple at
least 6" long" and use "brass couplings" for the copper and galvanized
This is totally new to me. So here come my newbie questions:
I looked up brass nipples and I understand the concept there. How do
the coupling pieces come into play? Can someone give me a URL to see
what they look like?
Put a clamp on the old work, and a clamp on the new work, and run a
heavy gauge wire between the two clamps, to jump the union,
This is identical to what is done for water meters to insure ground
This would work if you put in a short plastic piece, or something like a
water filter, which is plastic.
What you suggest is what HAS to be done, because code says that all
metallic plumbing in the house has to be grounded.
The perfectly reasonable explanation for that is because is some piece
of electrical equipment associated with the plumbing, like maybe an
electric water heater, clothes washer or a dishwasher has an improperly
connected or broken ground lead and then develops an insulation fault,
every faucet in the house could be electrified with 120VAC. And worse,
the faulty appliance may well "run OK" without giving any warning of
what's waiting to happen. (Think about a kid with bare feet standing on
damp soil reaching to turn on an outside sillcock.)
A less likely but not impossible reason the piping may become
electrified is when someone drives a long screw or nail into a wall
which pierces the insulation of a live wire, touches a hot conductor,
and then kisses a pipe.
But, galvanic corrosion is sorely misunderstood by most of the trades,
and using an insulated coupling bypassed by that necessary bonding wire
won't do squat to prevent corrosion of the steel pipe. The electrons
have a fine return path through that bonding wire and the galvanic
corrosion of the steel part will take place just as it would have if the
two dissimilar metals were touching directly. Well, maybe a little less
quickly if the insulating coupling consists of many feet of plastic pipe.
An insulated coupling will prevent galvanic corrosion only if the two
disimilar metals are NOT electrically connected together by another
The "big boys" use "impressed current" protection of buried tanks and
pipes, but these require an electrical supply to create a bucking
current. Such systems are used less and less these days as buried street
piping switches over to noncorroding plastic.
Just my .02,
My name is Jeff Wisnia and I approved this message....
Jeff, I agree with everything you said. But the OP's question got me
thinking. As I understand it, these dielectric unions are sold for
precisely the purpose that the OP is talking about, which is for use
to prevent galvanic corrosion where dissimilar metals meet. Yet, code
would seem to indicate that all the metal plumbing needs to be bonded
together for electrical safety. That seems reasonable, since if you
have a section on plumbing which is isolated from ground, then if any
spot became energized through a short the entire section of ungrounded
pipe would be live. But, as you pointed out, bonding defeats the
whole purpose of the isolation.
So, that begs the question, how are you supposed to be able to use one
of these dielectric unions and pass code? I even saw Richard
Trethewey on Ask This Old House install one on a boiler to solve a
homeowner's corrosion problem. He put it in with no bonding, which
solves the corrosion, but what about code?
BTW, if you ever see that episode, it has a nice blooper. The grand
master plumber has two pipe wrenches on the old union and is trying to
unscrew it, but he's pulling the wrong way. The camera cuts away, a
second later it's back and now the wrenches are on the right way.
Brass Rat 78
ONLY takes place when two dissimilar metals actually touch each other in
an electrolyte. That's not right. Think about the simple old fashioned
carbon-zinc flashlight battery. The carbon and zinc certainly aren't
touching each other. If the battery is left alone will last a long time
without "leaking". But, when you let the battery deliver current through
a load, by "connecting" the two materials through a flashlight bulb
filament or other load, the zinc shell gets "corroded" away pretty fast
and will develop leaks right through it.
The reasons I pontificate so much on this subject is because I could
understand the principles involved because of an electrical engineering
education, and also because I had a first hand experience two years ago
which caused me to question and later decry the "conventional wisdom"
about using dielectric couplings at the inlets and outlets of home water
Our electric water heater rusted out so I bought and schlepped home a
replacement. The old heater had copper piping running right to it, but
I'd heard about dielectric couplings over and over again and they were
displaying them right next to the water heaters at Home Cheepo so I
bought a couple and installed them when I put in the new water heater,
thinking that they would somehow help the new heater tank last longer. I
stood them off the top of the heater on 3 inch galvanized steel nipples.
Less than six month's later I noticed some rust on the heater's outlet
boss and on closer inspection I saw a bead of water oozing from where
the steel half of the dielectric union screwed onto the galvanized
nipple. I figured I'd just messed up on teflon taping the pipe threads
or didn't tighten them enough, but when I took things apart I was
shocked, just shocked, to find the nipple nearly clogged with rust and a
pinhole leak corroded through it, breaking through at the root of the
pipe thread just outside the portion of the threads which were screwed
into the steel half of the dielectric coupling.
I started thinking about what had happened and soon realized that the
water heater's tank was grounded via the safety ground lead on the 240
volt circuit feeding it (per code) and the copper piping in our home was
also grounded (again per code). Those two grounds provided a low
resistance shunt circuit across the insulation of the dielectric
couplings, so the galvanic current flowed through that path and the
corrosion took place just as though there was no insulation in those
The reason why the corrosion is worst right at the junction of the
copper and steel, even if there's an insulator there, but bypassed by
another conductor, is because the electrolyte (i.e. less than totally
pure water.) is not a perfect conductor, so the galvanic current density
into and out of the surfaces of the dissimilar metals isn't the same all
over, it's greatest close to the junction point and gets lower the
further away from the junction it travels because it has has to flow
through the resistance of the electrolyte to get there. (If you're an
electrical engineer, think of a resistive ladder network.)
BTW, the other nipple on my water heater was similarly clogged with rust
and also well corroded.
I pulled out both those dielectric unions and replaced them with copper
unions standing on "copper nipples".
I was gratified to see that the Rheem Water Heater Company recently
published a technical paper advising NOT to use dielectric couplings
when installing their products. They say to run copper all the way to
the tank shell and the sacrificial anode inside the tank will protect
the iron. That paper is at:
I sawed the leaking pipe nipple I described above in half and snapped a
photo of it to email to a buddy. I scraped out the big soft rust clumps
prior to taking the picture, so you can clearly see the craters which
corroded into the inside wall in less than six months time. See:
Jeff (Who won't drink the unfiltered tap water in our home <G>)
My name is Jeff Wisnia and I approved this message....
I'm just a DIY plumber/electrician, who has seen this first hand. I
know not to mix metals, and use copper plated stuff on copper pipes, and
zinc on galvanized.
I don't know the point, purpose, or intent of the original posters
di-electric unions, only how to jump around a break in the plumbing.
On 19 Oct 2004 14:07:47 -0700, email@example.com (coolneo) wrote:
OK, here's my inexpert take on it.
The dielectric union inserts an electrical insulator in the water
line. So if the dielectric union is between the grounding lug and the
city water supply, it would indeed disable the ground, as there would
not be an electrical connection to the earth, so to speak.
But in my house the grounding lug is attached to the cold water pipe
inches from where it first enters the house. I imagine that this is
typical. If yours is like that too I can't see how a dielectric union
further into the house plumbing would affect the grounding.
I have heard of larger buildings where people have "grounded" any
number of stupid things randomly to any pipe they could find, but I
hope that doesn't happen much in private homes.
I can even think of a few ways of testing for that, but I think you
should get more expert advice.
Actually, the grounding of switch boxes by virtue of the nearest
Cold water pipe was very common practice in much of the US
from the turn of the century till roughly 1965. After that,
Romex w/gnd was in widespread use.
There were exceptions, of course. Big cities mandated the use
of metallic raceway, other places used AC (BX) cable as a grounding
means. But millions upon millions of houses relied (still rely)
on the ground connection at the nearest Cold water pipe for
grounding of boxes for: bath switches, bath recepts, kitchen
counter recepts and switches, etc.
That's why plumbers have to be extra careful when disturbing
old runs of galv iron pipe.
much of the US from th
line it has to be included with the grounding circuit. Beyond that
we're discussing bonding, not grounding. The point of bonding the
metallic water lines is to provide a path to clear a ground fault and
trip the circuit breaker or fuse controlling the circuit. The jumper
suggested earlier will work for that. If the ground fault does not
have a path to clear, it can leave a section of the piping charged,
with possibly tragic results.
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.