Do metallic DHW pipes need to be isolated

Solid_Fuel_Man

Question is for those (much) smarter than I! For proper sacrificial anode protection, would copper lines need to be disconnected electrically from a water heater? I'd guess that's what those pesky nipples with the plastic sleeve in them are doing, separating the (conductive) water from the piping system.

Could I do this with a short run or PEX on the H and C port then back to copper, I know code doesnt allow PEX to be directly connected to the water heater....but I'm trying to get more life out of my pin holes SS indirect. Got a magnesium rod today for it.

GroundUp

I'd run a flex connector, personally. They're typically isolated on both ends with gasketed unions. If you still have the dielectric nipples in place it shouldn't be necessary though, that's what they're for.

Larry Weingarten

Hi, I like to use plastic lined nipples with flex connectors that have true dielectrics in the ends. This gives good separation of different metals. Do make sure to bond and ground the pipes back to the ground in your main panel, if they are metal, so no current wants to flow near the tank.

Also, the copper pipes affect only what's close to them, so they tend to make the steel rust that's very nearby. Stainless is a different beast than steel with different reactivities from the same metal.

And finally, you may get a LOT of gas from the magnesium anode if the water is very conductive and an air vent at a high point in the hot line will help with that.

Yours, Larry

Solid_Fuel_Man

I peered into the tank when I pulled the plug out of the too, the coils are still shiny with very little mineral deposits even after 9 years. The tank has male S.S. threads coming from it. Will brass adapters act as a cathode, all the flex lines I know of are nickel plated brass with S.S. braided poly pipe.

Larry Weingarten

Hello, It might be that all you need is something like this: Notice the white sleeve between copper and brass? That along with the rubber washer inside give a true dielectric.

Yours, Larry

ChrisJ

Uponor says you can run their pex directly to electric water heaters, but not gas. I assume because of drafthoods, so you'd think powervents would be ok too but they specifically say electric.

Also, local codes may prohibit it regardless.

Bradford White heaters come with dielectric nipples from what I've seen. The plastic inserts are heat traps. I'm installing a domestic HW return on my heater soon and I'll just be using a brass nipple into the tank.

You'll get a lot of opinions on all of this and who knows what's right. I've yet to see issues when using brass to connect steel and copper though. I have had issues with brass and aluminum though.

hot_rod

There is always some potential from electrolysis between dissimilar metals. The three driving factors are conductivity of the water, O2 and effect area. Measuring the TDS occasionally will give you a good indicator of the conductivity. Metals are often tested in sea water as it is very conductive.

Relative area effect is interesting. A large iron plate with copper rivets would show small amount of electrolysis. A large copper plate with iron rivets would corrode the rivets quickly in a sea water test.

I believe the same happens in plumbing or hydronics. If everything in the hydronic system is non ferrous, boiler, piping, pex, etc and only the expansion tank is steel, it seems the tanks tend to fail quickly.

On a system with a cast boiler, the copper to iron area relationship changes and you rarely have an electrolysis issue with copper adapters directly into the boiler. The conductivity of the fluid increases the potential.

I think that has to do with why dielectric unions tend to degrade and plug, rapidly in some areas, not so much in other. The electrolysis is related to the relative area, and potable water with high TDS provides a strong electrolyte.
In the area where I had a plumbing business city water was sourced from several mine tunnels, and that tends to be aggressive water, high TDS and also somewhat acidic, so dielectric union leaks and plugging issues were very common. Copper flex tubes solved the problem.

I would use any insulator between the two metals, I like the stainless or copper flexes, or pex on an indirect transition if no flue temperature is involved. Some of those stainless flex are a rubber or EPDM formula hose, the stainless braid around it for burst strength.

Solid_Fuel_Man

I guess the question is more of if I do not isolate copper directly connected to the SS water heater, what effect will that have on anode performance? Will it short out, and sacrifice the copper? Or is a magnesium rod ok in a 316L SS tank with copper directly connected?

Larry Weingarten

Hi, I like to think of the way an anode works like this; the anode is lit up and anywhere the light shines is protected. Of course it's less protected the further from the light it is. So, not much light will be shining up into the copper connections, meaning they will not consume much of the anode's potential. I'd still use the built-in dielectrics on the copper flex lines so you don't have direct metal contact between stainless and copper. Also, I'd bond and ground the piping. So, the anode should be OK with the right copper flex connectors.

Yours, Larry