Water Heater Hook Ups

Big-Al_2

I just bought a bigger water heater for our project/home, a 50 gallon natural gas unit.  I'll be installing it next week. At the same time, I also will replace a lot of my very old, congested galvanized pipe with copper.  The new water heater has built-in plastic-lined steel nipples for hook-up, rather than just female threaded openings.  I'd like to maximize its life, and want to avoid any galvanic problems.



With the plastic lined connections, am I covered?  Can I run the (pre-soldered) copper right to the water heater . . . or would I be better of with the dreaded dielectric unions . . . or should I put a short section of galvanized pipe between the water heater and the copper tubing . . . or should I transition from steel to brass, then to copper . . .  or maybe use the braided, stainless covered hook-up hoses?



I know this topic has been covered before, but I don't change out water heaters for a living, and this is the first one I've seen with the plastic lined nipples built-in . . . so the rules may have changed. 



By the way, my old water heater is 26 years old, and still doesn't leak.  Is this, in part, due to the fact that it was hooked up to a galvanized piping system and not copper? . . . or did they just make them better back in the 1980's . . . or did this unit just get lucky?



Opinions?

EricAune

Great water is most likely it

Al,



The fact that your existing water heater is that old and not leaking probably has more to do with the quality of your water than the piping connected to it.



I personally would not use DIE-electric unions for the new install.  I have been piping directly to the heater with a copper/brass male/female adapter for years without failure.  You could spend a little more for flexible stainless hoses but you will be left to rely on a rubber gasket for your high pressure seal.  I am not too keen on that.  If you do go that route, make sure you get at least 3/4" I.D. hoses so you do not sacrifice flow.



Good luck

Big-Al_2

Thanks

Thanks, all the hoses I've seen in the home centers seem mighty small in the ID, so I have been hesitant to use them.  The reason for re-piping is to get better flow, not to restrict it.   I used to serve as a water commissioner here, so I know a little about the water.  It's Lake Michigan water, generally soft, about 7 grains per gallon, and has a pH of around 7.5 to 8.  The utility also adds orthophosphate to the water.  It coats the pipes with a slimy kind of substance to prevent leaching of lead from old services.  Maybe that helps a little.  Plumbers around here do seem to connect steel directly to copper all the time . . . and get away with it.



Others?

Mark Eatherton

Use brass adapters...

They are neutral to the plastic lined steel nipples and the copper. They are an accepted method of dissimilar metal connections in potable water.



Unions are required by code, unless you are doing it in your own home, in which case I'd skip them.



ME

Big-Al_2

Ball Valves

OK, I have a couple of 3" brass ball valves with built-in unions.  I could come up from the water heater with 12" of 3/4" galvanized pipe, join it to the ball valves, and continue up with copper.  That should be good isolation, right? 



I just was in a neighbor's basement, and his plumber ran copper right into the female ports on his water heater.  I take it that some plumbers don't really take any precautions against galvanic action?  If galvanic corrosion did occur, would it be mostly localized around where the dissimilar metals meet . . . or would it affect the whole tank?

Mark Eatherton

Right...

Corrosion in dissimilar metal mixes generally attacks the lesser noble metal, in this case the steel tank.



The anode that is inside of the tank is supposed to protect the portions of the tank that did not get glass lining from the O2 in the water, but I don't think it will protect the tank from dissimilar connections like you saw. That would require the use of a cathode type of protection to insure the direction of flow of the electrons, to avoid dissolving the steel.



The average life expectancy of a residential tank style heater that does not receive regular anodic maintenance is 5 to 7 years. With regular anodic maintenance, the life expectancy is unknown.



ME

Mike Kusiak_2

Localized corrosion

Because of the high resistivity of tap water, corrosion between dissimilar metals mainly occurs close to the junction. As you get farther from the junction, less electron flow is possible due to the electrical resistance of the water and the greater distance between the dissimilar metals. This is precisely the reason the nipples supplied with the heater are lined with plastic. It serves as an insulator to isolate the inside metal surface of the nipple from the cathode formed by the copper or brass fitting connected to it. By the time you get to the heater end of the nipple, the high resistance of the length of water within the plastic lined section reduces the electrolytic current to a negligible value, thereby reducing the potential for corrosion.



If you connect 12" lengths of galvanized pipe to the nipples, corrosion will still occur in the galvanized pipe at the unprotected connection to the copper or brass, so you defeat the purpose of the plastic lined nipples. The sacrificial 12" pipes may protect the heater, but you may have to replace them periodically as they corrode at the dissimilar metal junction.

Big-Al_2

How Can This Be?

If a few inches of water in a plastic-lined nipple has enough resistance to negate galvanic action, how can the anode rod possibly protect the tank? It's more than a couple of inches from most of walls of the tank, yet it does help.

Mike Kusiak_2

Cross sectional area

The plastic nipple does not negate the galvanic action, but does reduce it to a value much less than if a plain nipple was used. If a regular nipple was used, the junction of dissimilar metals occurs right next to each other, so the current path is very short and a relatively high current can flow near the junction, even though the area involved may be small.



The electrical resistance of the current path is proportional to its length and cross sectional area. If you think of the 3/4" insulated nipple, the cross section is a fraction of a square inch. Very little current can flow through the small area of water for the relatively greater length of the insulation. Then think of how many square inches of surface the anode rod has in contact with the water in the tank. Even though the distance from the rod to the tank may be greater than the length of the nipple, the much greater area of current path effectively reduces the resistance to a much lower value.



Depending on the acidity and mineral content of the water, the galvanic current through the rod may actually too great and more than necessary to protect the tank, in which case the rod is rapidly consumed. To prevent this some manufacturers actually add a resistor in series with the rod's connection to the threaded plug to limit the amount of current generated. Rheem's R-tech rods are built this way and seem to have a much greater life, yet still adequately protect the tank.

Big-Al_2

Makes Sense!

Thanks, Mike! That does make a lot of sense.  I talked a little more with a plumber-friend today.  He tells me that the municipal water here is really gentle to water heaters, and that they seldom use any kind of galvanic isolation on gas water heaters anymore . . . and get unusually good life from them anyway.  It sounds like no matter what I do, the tank should last pretty well.



The new unit is a Richmond (Rheem) with their patented resistive anode rod.  If I'm still in the house in a few years, I'll make it a point to pull the rod and have a look at how its doing.  Maybe if I replace it once in a while, the tank will last as long as the old one!