Alternative to dielectric union

[Bennett Price]

I've heard that a brass nipple can be used to join copper to galvanized
pipe instead of a dielectric union. Is this true? How long does the
nipple need to be? (.5" and .75" pipe are at issue.)

Thanks

[PLUMBGURU2]

Post by Bennett Price
I've heard that a brass nipple can be used to join copper to galvanized
pipe instead of a dielectric union. Is this true?

While it will NOT meet UPC code requirements, in my experience a brass nipple
does a pretty good job of electrically isolating the black tank of a water
heater from copper tubing.

I suppose it would work just as well for a galvanized to copper connections as
the material in galvanized or black you're trying to isolate the copper from is
iron.

Doug

[Jeff Wisnia]

I heartily disagree with you on your technical reasoning.

A brass nipple will not break the electrical continuity of the path through which
the galvanic current flows, and will do little at all to retard galvanic corrosion
of the dissimilar metals.

I think your experience, while I'm sure it's convinced you, was not a definitive
test, unless you set up two systems next to each other, identical in all respects
save for the addition of those brass nipples.

Those who are still using dielectric unions thinking they are "protecting" electric
hot water heaters should be aware that they are wasting their money, because the
insulating gap in the union is shorted out by the tank's electrical ground (per
code) through it's power wiring, and the copper piping (if done to code) which is
also grounded. Those two grounds produce an electrical short right across the
insulating gap of the dielectric unions, rendering them useless for their intended
purpose.

It is far wiser to spend the money and time checking the sacrificial anode rod
every couple of years and replacing it when it is badly corroded.

I learned this the hard way, because the last time I replaced my electric hot water
heater I went to the trouble of installing a pair of dielectric unions, standing on
2-1/2" galvanized nipples screwed into the tank's cold and hot ports. In just one
year one of the nipples started leaking right through its wall. I pulled things
apart and found both unions and nipples pretty well clogged with big rust crystals,
and badly corroded spots on the inside of the nipple walls.

I replumbed everything with copper and copper unions. I think it's better to let
the copper run righ to the tank ports, so that the galvanic currents have a better
shot at working on the sacrificial anode.

I sliced the leaking nipple open. You can see it at:

Loading Image...

The OP's appears to be at UCSF. It might be interesting for him to verify what I've
said above with a prof in the chemistry department who is a guro in
electrochemistry, I'm just a flatfooted electrical engineer, not a chemist.

Just my .02.

Jeff (Who sometimes misses the San Francisco he was born in back in 1936.)

Post by PLUMBGURU2

Post by Bennett Price
I've heard that a brass nipple can be used to join copper to galvanized
pipe instead of a dielectric union. Is this true?

While it will NOT meet UPC code requirements, in my experience a brass nipple
does a pretty good job of electrically isolating the black tank of a water
heater from copper tubing.
I suppose it would work just as well for a galvanized to copper connections as
the material in galvanized or black you're trying to isolate the copper from is
iron.
Doug

--

Jeff Wisnia (W1BSV + Brass Rat '57 EE)

"If you can smile when things are going wrong, you've thought of someone to blame
it on."

[PLUMBGURU2]

Post by Jeff Wisnia
I heartily disagree with you on your technical reasoning.

I didn't give any technical reasoning only what I've observed in over 30 years
of plumbing service.

Frankly your experience with your one electric water heater is pretty
insignificant compared to the number of water heaters I've changed out that had
brass nipples screwed into the black tank with no significant build-up or
damage to the copper tubing.

Labratory test are all well and good to give an indication of anticipated
results but quite often fields results vary dramatically from lab results.

Ask the other plumbers who visit here from time to time if a brass nipple does
a better job than a dielectric union to help control the effects of
electrolysis

Post by Jeff Wisnia
It is far wiser to spend the money and time checking the
sacrificial anode rod every couple of years and replacing
it when it is badly corroded.

I totally agree with the above statement, however the original question didn't
mention anything about water heaters. I only used water connections from
copper piping to brass nipple to the black tank in a storage water heater, as
an example of what I've observed.

Perhaps there's some other magic at work, I don't know, but before you dis my
statement observations and experience, go change a couple thousand water
heaters including gas fired and electric and get back to us.

Doug

[steve]

One of my supply houses is selling Rheem heaters instead of Bradford.
Rheem tanks have female tappings as does Home Depot and we all know
who makes the GE water heaters.
Anyway, I called the Cherry Hill, NJ plumbing inspector on this and he
said that he would be fine with brass nipples in the top of the tank.
Think of it this way... the drain valve is brass and the relief valve
is brass. I rest my case. As a matter of fact, if you can find CAST
male adapters, they would be fine too. Trying to find those
nowadays is tough. I still am installing Bradfords though. I am slow
to change. I have been installing bradfors since 78.
Also, the PEX piping... today I starting installing it finally since I
started the thread on it a week or so ago. Very easy to install. The
hardest part is getting the S/S rings engage in the ratchet tool. It
would be nice if they locked into the jaws somehow.
I am using RTI's fittings and tool and the stainless clamps. I
haven't pressure tested it yet but I'm sure it will be fine. It is
tough finding 20' lengths of 1/2" in blue though. The coils are a
little tough to work with. I bought some white in straight lengths
though at Weinstein Supply. (NJ)
That's it for now.

Post by PLUMBGURU2

Post by Jeff Wisnia
I heartily disagree with you on your technical reasoning.

I didn't give any technical reasoning only what I've observed in over 30 years
of plumbing service.
Frankly your experience with your one electric water heater is pretty
insignificant compared to the number of water heaters I've changed out that had
brass nipples screwed into the black tank with no significant build-up or
damage to the copper tubing.
Labratory test are all well and good to give an indication of anticipated
results but quite often fields results vary dramatically from lab results.
Ask the other plumbers who visit here from time to time if a brass nipple does
a better job than a dielectric union to help control the effects of
electrolysis

Post by Jeff Wisnia
It is far wiser to spend the money and time checking the
sacrificial anode rod every couple of years and replacing
it when it is badly corroded.

I totally agree with the above statement, however the original question didn't
mention anything about water heaters. I only used water connections from
copper piping to brass nipple to the black tank in a storage water heater, as
an example of what I've observed.
Perhaps there's some other magic at work, I don't know, but before you dis my
statement observations and experience, go change a couple thousand water
heaters including gas fired and electric and get back to us.
Doug

[CES]

Post by steve
One of my supply houses is selling Rheem heaters instead of Bradford.
Rheem tanks have female tappings as does Home Depot and we all know
who makes the GE water heaters.
Anyway, I called the Cherry Hill, NJ plumbing inspector on this and he
said that he would be fine with brass nipples in the top of the tank.
Think of it this way... the drain valve is brass and the relief valve
is brass. I rest my case. As a matter of fact, if you can find CAST
male adapters, they would be fine too. Trying to find those
nowadays is tough. I still am installing Bradfords though. I am slow
to change. I have been installing bradfors since 78.
Also, the PEX piping... today I starting installing it finally since I
started the thread on it a week or so ago. Very easy to install. The
hardest part is getting the S/S rings engage in the ratchet tool. It
would be nice if they locked into the jaws somehow.
I am using RTI's fittings and tool and the stainless clamps. I
haven't pressure tested it yet but I'm sure it will be fine. It is
tough finding 20' lengths of 1/2" in blue though. The coils are a
little tough to work with. I bought some white in straight lengths
though at Weinstein Supply. (NJ)
That's it for now.

Post by PLUMBGURU2

Post by Jeff Wisnia
I heartily disagree with you on your technical reasoning.

I didn't give any technical reasoning only what I've observed in over 30 years
of plumbing service.
Frankly your experience with your one electric water heater is pretty
insignificant compared to the number of water heaters I've changed out that had
brass nipples screwed into the black tank with no significant build-up or
damage to the copper tubing.
Labratory test are all well and good to give an indication of anticipated
results but quite often fields results vary dramatically from lab results.
Ask the other plumbers who visit here from time to time if a brass nipple does
a better job than a dielectric union to help control the effects of
electrolysis

Post by Jeff Wisnia
It is far wiser to spend the money and time checking the
sacrificial anode rod every couple of years and replacing
it when it is badly corroded.

I totally agree with the above statement, however the original question didn't
mention anything about water heaters. I only used water connections from
copper piping to brass nipple to the black tank in a storage water heater, as
an example of what I've observed.
Perhaps there's some other magic at work, I don't know, but before you dis my
statement observations and experience, go change a couple thousand water
heaters including gas fired and electric and get back to us.
Doug

Hope this goes to the right place; I've been having a little trouble posting
in news groups lately.



I'd like to add my experience to this. I've been a plumber for 30 years,
changed many a water heater in my day. Remember when water heaters lasted
20+ years? The old John Woods and Bradley WHs seemed to last the longest. I
still have a few customers with them, still going strong; never saw a
dielectric union on one of them. Do you think they would have lasted any
longer if they had? Isn't it a shame the crap they call WHs now hardly last
5, 6, 7 years, with or without dielectric unions.



Well enough of that. I, probably like most of us, started using dielectric
unions when they became code some time back. I really can't say I've seen
any WH's longevity increased by using them. I, like Jeff, found when
installing DEUs on steel nipples, an accelerated corrosion took place
attacking the steel nipples. In one case it was galvanized nipples on my own
water heater. It had been installed about 5 years when I disconnected it to
redo the flooring under it. The galvanized nipples where totally corroded on
the inside and just about closed to the point of restriction with rust. I
generally use dielectric nipples when using dielectric unions, but I was out
of them when I installed my water heater. I have since seen this condition
in other WHs.



Rheem Pro series WHs have dielectric nipples preinstalled. They also offer
an extended warranty kit which consists of an extra anode rod attached to a
Ÿ" dielectric nipple which installs in the hot side tapping. But get this;
Rheem gives you a brass, Ÿ" x 1" bushing for WHs with 1" tapings.



So, what's up with all of this? Is the key to always use dielectric nipples
with dielectric unions? Is it just as well to connect directly to the WH
whether it's copper, brass or steel? Does it all depend on water and
electrical conditions in a particular house and area?



As far as I'm concerned the jury is still out on DEUs. As I said earlier, I
haven't seen where they are making much difference and I haven't seen where
it makes much difference whether it's an electric, gas, or oil fired WH.



~C~

[Pipeline9000]

Post by steve
One of my supply houses is selling Rheem heaters instead of Bradford.
Rheem tanks have female tappings as does Home Depot and we all know
who makes the GE water heaters.
Anyway, I called the Cherry Hill, NJ plumbing inspector on this and he
said that he would be fine with brass nipples in the top of the tank.
Think of it this way... the drain valve is brass and the relief valve
is brass.

The drain valve and the T&P also dont have a direct connection to
ground.
Galvanic corrosion is most common in copper distribution systems
connected directly to the steel water heater tank and to the water
service piping under ground which provides the ground for the circuit.
Then the electrician comes along and compounds the problem by
connecting the electrical service ground to the incoming service
piping.....

[Jeff Wisnia]

Post by PLUMBGURU2

Post by Jeff Wisnia
I heartily disagree with you on your technical reasoning.

I didn't give any technical reasoning only what I've observed in over 30 years
of plumbing service.

I tried my best not to offend you, but the "technical reasoning" I choked over was
your statement that:

"in my experience a brass nipple does a pretty good job of electrically isolating
the black tank of a water heater from copper tubing."

If you can find any licensed electrician you know who'd agree with your statement
that brass, a metal which is a very good conductor of electricity, could possibly
electrically isolate two other metal items, then I'll offer a full apology to you.

Post by PLUMBGURU2
Frankly your experience with your one electric water heater is pretty
insignificant compared to the number of water heaters I've changed out that had
brass nipples screwed into the black tank with no significant build-up or
damage to the copper tubing.

Absolutely, but the tanks you've changed out were not in *my* basement, where they
were subjected to the water in our town.

And, FWIW, I'd gone through two heaters in this house in twenty years prior to
using those dielectric unions when installing the third heater. There was NO
build-up or damage to the copper tubing or fittings at the tank in all that time.

Go to the site below and you'll see that, since iron is higher in the galvanic
series than copper or brass, corrosion attacks the iron, not the copper. (Zinc is
even higher in the table than iron, that's why they use zinc anodes, which corrode
and "protect" the iron tank.) You'll also see that brass is right next to copper in
the series, so there shouldn't be much difference if it's used as you described.

http://www.mcnallyinstitute.com/CDweb/g-html/g002.htm

If you can honestly state that all of the tanks you've changed which had brass
nipples between them and copper piping had insignificant buildup or damage to the
copper tubing AND all the ones you've changed in the same town which didn't have
brass nipples had significant buildup or damage to the copper tubing, then I'll
apologize to you a second time.

Post by PLUMBGURU2
Labratory test are all well and good to give an indication of anticipated
results but quite often fields results vary dramatically from lab results.

Yes, but when they do, persons skilled in the science involved can discover *why*
there was a variance.

Post by PLUMBGURU2
Ask the other plumbers who visit here from time to time if a brass nipple does
a better job than a dielectric union to help control the effects of
electrolysis

I'll keep my eye on this thread and see how many do.

Post by PLUMBGURU2

Post by Jeff Wisnia
It is far wiser to spend the money and time checking the
sacrificial anode rod every couple of years and replacing
it when it is badly corroded.

I totally agree with the above statement, however the original question didn't
mention anything about water heaters. I only used water connections from
copper piping to brass nipple to the black tank in a storage water heater, as
an example of what I've observed.

Yes, but the OP has since posted that wants to use dielectric unions to separating
old iron plumbing in the walls of his home from new copper in the basement/crawl
space. I maintain that he'll have to bypass those couplings with an electrical
jumper to maintain the code required grounding of the pipes and fixtures in the
home. When he does that he'll lose any help those dielectric unions might have
given him Thge steel parts of those unions and the iron pipe adjacent to them will
suffer from corrosion.

Post by PLUMBGURU2
Perhaps there's some other magic at work, I don't know, but before you dis my
statement observations and experience, go change a couple thousand water
heaters including gas fired and electric and get back to us.

The "did" was, as I stated, disagreement with what you said, not disrespect of them
or you. I'm sorry if you misunderstood that, but a lifetime of experience has
shown me that there's there's probably little I can do to change your mind about
that now..

Jeff

Post by PLUMBGURU2
Doug

--

Jeff Wisnia (W1BSV + Brass Rat '57 EE)

"If you can smile when things are going wrong, you've thought of someone to blame
it on."

[Mark Monson]

Post by Jeff Wisnia

Post by PLUMBGURU2

Post by Jeff Wisnia
I heartily disagree with you on your technical reasoning.

I didn't give any technical reasoning only what I've observed in over 30 years
of plumbing service.

I tried my best not to offend you, but the "technical reasoning" I choked over was
"in my experience a brass nipple does a pretty good job of electrically isolating
the black tank of a water heater from copper tubing."
If you can find any licensed electrician you know who'd agree with your statement
that brass, a metal which is a very good conductor of electricity, could possibly
electrically isolate two other metal items, then I'll offer a full apology to you.

Post by PLUMBGURU2
Frankly your experience with your one electric water heater is pretty
insignificant compared to the number of water heaters I've changed out that had
brass nipples screwed into the black tank with no significant build-up or
damage to the copper tubing.

Absolutely, but the tanks you've changed out were not in *my* basement, where they
were subjected to the water in our town.
And, FWIW, I'd gone through two heaters in this house in twenty years prior to
using those dielectric unions when installing the third heater. There was NO
build-up or damage to the copper tubing or fittings at the tank in all that time.

I agree with Doug that when using galvanized nipples and die-electric unions
the nipples will quickly fill up with rust. I have come to the conclusion
that for steel tank water heaters on copper pipe the best way is come out
with male adapters and forget about D.E.U.'s.

The worst problems I've seen is when running a circulated domestic hot water
system on galvanized pipe with a COPPER water heater (A.O. Smith Burkay).
The hot water piping popped pin holes all over the system. ( 283 bed nursing
home) over a period of twenty years. I would definitely use die-electrics
with a Burkay piped to galvanized.

Post by Jeff Wisnia
Go to the site below and you'll see that, since iron is higher in the galvanic
series than copper or brass, corrosion attacks the iron, not the copper. (Zinc is
even higher in the table than iron, that's why they use zinc anodes, which corrode
and "protect" the iron tank.) You'll also see that brass is right next to copper in
the series, so there shouldn't be much difference if it's used as you described.
http://www.mcnallyinstitute.com/CDweb/g-html/g002.htm
If you can honestly state that all of the tanks you've changed which had brass
nipples between them and copper piping had insignificant buildup or damage to the
copper tubing AND all the ones you've changed in the same town which didn't have
brass nipples had significant buildup or damage to the copper tubing, then I'll
apologize to you a second time.

Post by PLUMBGURU2
Labratory test are all well and good to give an indication of anticipated
results but quite often fields results vary dramatically from lab results.

Yes, but when they do, persons skilled in the science involved can discover *why*
there was a variance.

Theory is great, but theory is tested in practice. Science doesn't discover
truth. The most science can do is come up with a hypothesis that fits for a
time. You won't convince plumbers that engineers and architects have all
the answers. We've seen too many so-called "engineered" designs that
weren't worth a damn.

MM

[j***@gmail.com]

Sounds like you’ve got some really corrosive water. I’ve a water heater I installed 27 years ago with dielectric nipples, and it’s still in good condition. I am on well water, but it is of good quality.

[Bennett Price]

Perhaps I should have added that the question came to mind since I'm
planning a changeout from galvanized to copper in the basement/crawl
space of my house but intend to leave the galvanized in the walls.

I'm having trouble getting homeowner's insurance (in California) because
of the old galvanized plumbing and am told that a 'substantial'
replumbing to copper will satisfy the insurance companies. Water heater
is in basement so it will go pure copper; house is 1 story high and the
galvanized vertical sections left will be about 4 to 5 feet long.

And as I asked initially, 'how long does the nipple need to be?'

Post by Bennett Price
I've heard that a brass nipple can be used to join copper to galvanized
pipe instead of a dielectric union. Is this true? How long does the
nipple need to be? (.5" and .75" pipe are at issue.)
Thanks

[kenny b]

Post by Bennett Price
Perhaps I should have added that the question came to mind since I'm
planning a changeout from galvanized to copper in the basement/crawl
space of my house but intend to leave the galvanized in the walls.
I'm having trouble getting homeowner's insurance (in California) because
of the old galvanized plumbing and am told that a 'substantial'
replumbing to copper will satisfy the insurance companies. Water heater
is in basement so it will go pure copper; house is 1 story high and the
galvanized vertical sections left will be about 4 to 5 feet long.
And as I asked initially, 'how long does the nipple need to be?'

Post by Bennett Price
I've heard that a brass nipple can be used to join copper to galvanized
pipe instead of a dielectric union. Is this true? How long does the
nipple need to be? (.5" and .75" pipe are at issue.)
Thanks

usually when doing a basement changover and leaving the galv. risers
I'd either use a diletric union, a brass ips valve or a min 3" brass
nipple. I haven't had any problems over the last 30 years. A simple
brass break between the copper and galv. slows down the reaction
enough that will out live the existing galv. pipe that is left.

[Dale Wilcox]

It can depend on the jurisdiction, In one city I work in they want 6" Brass
nipps, DEU not allowed. In another, ether is ok, yet another DEU only. So
you should check with your local inspecting agency.
Dale

Post by Bennett Price
Perhaps I should have added that the question came to mind since I'm
planning a changeout from galvanized to copper in the basement/crawl
space of my house but intend to leave the galvanized in the walls.
I'm having trouble getting homeowner's insurance (in California) because
of the old galvanized plumbing and am told that a 'substantial'
replumbing to copper will satisfy the insurance companies. Water heater
is in basement so it will go pure copper; house is 1 story high and the
galvanized vertical sections left will be about 4 to 5 feet long.
And as I asked initially, 'how long does the nipple need to be?'

Post by Bennett Price
I've heard that a brass nipple can be used to join copper to galvanized
pipe instead of a dielectric union. Is this true? How long does the
nipple need to be? (.5" and .75" pipe are at issue.)
Thanks

[Gary Slusser]

Post by Dale Wilcox
It can depend on the jurisdiction, In one city I work in they want 6" Brass
nipps, DEU not allowed. In another, ether is ok, yet another DEU only. So
you should check with your local inspecting agency.
Dale

Hey guys, the one common thing to all water heaters regardless of the
type of plumbing and anode rod materials, or the manufacturers, is the
water quality. I've only seen it mentioned what, twice in the whole
thread? All water is wet and most of it is clear but that's where the
similarities between one location and another cease. Unseen differences
in water quality ranges all over the board and you'll find that dictates
how well the plumbing/appliance materials it contacts do in regards to
their service lives.

Since the EPA changed the acceptable pH range of potable water in 1991
from 6.9-8.5 to 6.5-8.5, most 'city' water has been more acidic. A water
with pH 6.0 is 10 times more acidic than a water of 7.0 (neutral) pH.
That's not the only thing that will shorten heater and plumbing material
life. Things like bacteria, DO, CO2, chlorides, sulfates and TDS content
plus chlorine etc. all have a negative impact on material life. Also,
electrical grounding and grounds do too.

Take a look here for more info on water caused corrosion:
http://www.corrosion-doctors.org/NaturalWaters/Frames.htm

Gary
Quality Water Associates

[Pipeline9000]

Post by Bennett Price
Perhaps I should have added that the question came to mind since I'm
planning a changeout from galvanized to copper in the basement/crawl
space of my house but intend to leave the galvanized in the walls.
I'm having trouble getting homeowner's insurance (in California) because
of the old galvanized plumbing and am told that a 'substantial'
replumbing to copper will satisfy the insurance companies. Water heater
is in basement so it will go pure copper; house is 1 story high and the
galvanized vertical sections left will be about 4 to 5 feet long.
And as I asked initially, 'how long does the nipple need to be?'

Like the first gentleman so eloquently stated: Brass is an excellent
conductor. So wouldn't it stand to reason that a nonconductive
material would break the electrolytic circuit?? how 'bout that PEX!
The current will still travel through the water but the intensity of
the current is dependant upon the density of the disolved solids in
the water (mainly minerals), but its definately better than a metal
connector. Make sense?

[Bennett Price]

If I understand all of this correctly, the issue is not how well a
pipe conducts but how electrochemically reactive it is when placed in
contact with a pipe of different composition.

I found the following URL's helpful - though they don't directly address
the question I posed initially.

http://www.mtec.or.th/th/research/famd/corro%5Chowmetals.htm
http://www.corrprev.org.au/Galvanic.htm

Post by Pipeline9000

Post by Bennett Price
Perhaps I should have added that the question came to mind since I'm
planning a changeout from galvanized to copper in the basement/crawl
space of my house but intend to leave the galvanized in the walls.
I'm having trouble getting homeowner's insurance (in California) because
of the old galvanized plumbing and am told that a 'substantial'
replumbing to copper will satisfy the insurance companies. Water heater
is in basement so it will go pure copper; house is 1 story high and the
galvanized vertical sections left will be about 4 to 5 feet long.
And as I asked initially, 'how long does the nipple need to be?'

Like the first gentleman so eloquently stated: Brass is an excellent
conductor. So wouldn't it stand to reason that a nonconductive
material would break the electrolytic circuit?? how 'bout that PEX!
The current will still travel through the water but the intensity of
the current is dependant upon the density of the disolved solids in
the water (mainly minerals), but its definately better than a metal
connector. Make sense?

[Jeff Wisnia]

Post by Bennett Price
If I understand all of this correctly, the issue is not how well a
pipe conducts but how electrochemically reactive it is when placed in
contact with a pipe of different composition.

You've got it right, but what I was trying to say in my OP was that the "contact" doesn't have to be directly between the
two metals touching each other. The electrolytic corrosion takes place just as badly if they are connected together by some
other electrical conductor.

If you stick the bottom ends of a zinc and a copper rod in a glass jar of saltwater, you won't get much corrosion if
they're not touching each other, but connect the top ends with a piece of wire (even one a hundred feet long) and watch how
much faster the zinc corrodes.

Remember the old zinc/carbon flashlight batteries, and how they'd leak and ruin your flashlight if you discharged them too
much? That happened because the zinc shell got eaten through by the galvanic corrosion which took place when you connected
the carbon to the zinc through the flashlight bulb.

That's why I was barfing about dielectric unions helping only if there isn't a "sneak path" for electric current to flow
around the insulating gap in those unions.

Jeff
--
Jeff Wisnia (W1BSV + Brass Rat '57 EE)

"If you can keep smiling when things go wrong, you've thought of someone to place the blame on."

Post by Bennett Price
I found the following URL's helpful - though they don't directly address
the question I posed initially.
http://www.mtec.or.th/th/research/famd/corro%5Chowmetals.htm
http://www.corrprev.org.au/Galvanic.htm

Post by Pipeline9000

Post by Bennett Price
Perhaps I should have added that the question came to mind since I'm
planning a changeout from galvanized to copper in the basement/crawl
space of my house but intend to leave the galvanized in the walls.
I'm having trouble getting homeowner's insurance (in California) because
of the old galvanized plumbing and am told that a 'substantial'
replumbing to copper will satisfy the insurance companies. Water heater
is in basement so it will go pure copper; house is 1 story high and the
galvanized vertical sections left will be about 4 to 5 feet long.
And as I asked initially, 'how long does the nipple need to be?'

Like the first gentleman so eloquently stated: Brass is an excellent
conductor. So wouldn't it stand to reason that a nonconductive
material would break the electrolytic circuit?? how 'bout that PEX!
The current will still travel through the water but the intensity of
the current is dependant upon the density of the disolved solids in
the water (mainly minerals), but its definately better than a metal
connector. Make sense?

[s***@att.net]

Dielectric union

My water heater contains galvanized steel inlet water pipes.
The house water pipes are copper.

For a test, I did this:
- Turned off the gas input valve.
- Temporarily disconnected the gas line from the water heater.
Then I measured the open circuit voltage and short circuit current across the dielectric union.

Copper Iron
======||||====== <--- Water heater pipes
+ -
Measured across the dielectric union:
Open: 5.5 mV Using a voltmeter.
Shorted: 33. uA Using an ammeter.

The copper and iron pipes, with the mineral-filled water, create a chemical battery.
Connecting the copper and iron with an electric conductor allows current to flow.
Then ions migrate from the iron to the copper, slowly corroding the iron.
Eventually the iron pipe deteriorates enough that it leaks water.
That conductor makes the voltage between the copper and iron metals zero. A battery "short circuit".

If something anywhere electrically connects the gas and water lines,
then that invalidates the dielectric union. That "electrically" connects across it.
Here is the "test": Failure to sense any open-circuit voltage across the union indicates that condition.

To reduce the current flow to zero, I inserted an "electric insulator pipe" into the heater's input gas line.
Then the voltage between the copper and iron metals is no longer zero. No "short", so no current flows.
Then the dielectric union actually works. The water heater's inlet iron water pipe lasts much longer.

If all the house water pipes are plastic, then the dielectric union is not needed.
A plastic pipe / connector will serve the same function as a dielectric union.

Knowing the shorted current, you can calculate how much iron is lost per year. It's chemistry.

[S***@aol.com]

Post by s***@att.net
Dielectric union
My water heater contains galvanized steel inlet water pipes.
The house water pipes are copper.
- Turned off the gas input valve.
- Temporarily disconnected the gas line from the water heater.
Then I measured the open circuit voltage and short circuit current across the dielectric union.
Copper Iron
======||||====== <--- Water heater pipes
+ -
Open: 5.5 mV Using a voltmeter.
Shorted: 33. uA Using an ammeter.
The copper and iron pipes, with the mineral-filled water, create a chemical battery.
Connecting the copper and iron with an electric conductor allows current to flow.
Then ions migrate from the iron to the copper, slowly corroding the iron.
Eventually the iron pipe deteriorates enough that it leaks water.
That conductor makes the voltage between the copper and iron metals zero. A battery "short circuit".
If something anywhere electrically connects the gas and water lines,
then that invalidates the dielectric union. That "electrically" connects across it.
Here is the "test": Failure to sense any open-circuit voltage across the union indicates that condition.
To reduce the current flow to zero, I inserted an "electric insulator pipe" into the heater's input gas line.
Then the voltage between the copper and iron metals is no longer zero. No "short", so no current flows.
Then the dielectric union actually works. The water heater's inlet iron water pipe lasts much longer.
If all the house water pipes are plastic, then the dielectric union is not needed.
A plastic pipe / connector will serve the same function as a dielectric union.
Knowing the shorted current, you can calculate how much iron is lost per year. It's chemistry.

What a great post this takes away all the hand waving, and gives a real explanation why dielectrics don't work in the real world. Virtually all gas lines are grounded; therefore virtually all dielectrics are short circuited.

So the other way to break the galvanic cell (stop the corrosion) is to inhibit the conduction through the electrolyte (water), which is done in practice by having a several inch plastic lining inside the pipe known as a dielectric nipple. This effectively makes electrolyte (water) a much poorer conductor. Hence most water heater manufactures build these in or supply these.

Now the other unanswered question is why DOES brass seem to work when put between Copper and steel? Theoretically brass should corrode iron almost as fast as copper, but plumber after plumber says putting in brass nipples work, to the point that it has become code in some areas.

Perhaps this just moves the corrosion to just inside the thick large anode protected tank. Since there is no "nipple" failure it appears there is no acute problem. Under this explanation a dielectric nipple would still be preferable and would help the water heater itself last longer.

Does anybody have experience with putting brass between a copper pipe and galvanized pipe, then opening up some time later? In this scenario the galvanized pipe should still readily corrode.

A more out there idea is perhaps brass forms some natural surface oxidation or other surface layer that acts as an inhibition to the electrolyte similar to a plastic liner?

Any other ideas?

[t***@gmail.com]

Post by Bennett Price
I've heard that a brass nipple can be used to join copper to galvanized
pipe instead of a dielectric union. Is this true? How long does the
nipple need to be? (.5" and .75" pipe are at issue.)
Thanks

It should be at least 6 inch min.