My 1970 house with slab foundation originally had a connection to the copper water line as the only ground. I had a pair of ground rods 6' 6" apart added and would like to disconnect the ground to the copper water supply pipe. Is this allowable and advisable?

The reason I would want to do this is to save the water pipe from damage. We had 20 pinhole leaks in a vertical vertical hot out from the hot water tank. This tank was sweated copper tubing all the way to the tank (advice from city plumbing inspector 25 years ago) instead of flexible supply lines with insulating connections.

There is at least the possibility that the pinhole leaks were due to electrolysis from dissimilar metals (steel tank-copper water lines). Twelve years ago I had the tank replaced with a tankless heater connected with insulated ("dielectric" connections). I want to eliminate the possibility that the ground rods might produce a similar effect in the copper tubing.

Also, I have heard of lightning strikes damaging copper water lines if the panel is grounded to the copper pipes.

If it is OK to disconnect the ground wire to the copper water supply line, should I do this in the panel and at the pipe? Is it OK to leave the old ground wire in place in the panel but disconnected from the ground bar?

  • 1
    Installing a ground rod is difficult and unpredictable, and ground piping is a better grounding system in 99% cases. Had your new ground rod been measured and declared as within parameters?
    – Agent_L
    Nov 14, 2016 at 10:30
  • There is a limestone layer under my house at a depth of maybe 4.5 ft. The electrician put in two ground rods about 4 ft apart. He drove the rods as far as he could with a baby sledge and then cut off about 18 inches on each one. The rods have not been measured and declared to meet standards. Nov 14, 2016 at 10:42
  • Well, measuring of a ground rod involves hammering in temporary measurement rods so it's understandable why he skipped it but on your place I would put more trust into those pipes than a trimmed rods in a limestone.
    – Agent_L
    Nov 14, 2016 at 10:46
  • Since I have two rods could I test the system by disconnecting the wire to the panel from both rods and measuring the resistance between them with a VOM? Nov 14, 2016 at 11:05
  • 1
    AFAIR "ground meter" is a dedicated device. A regular multimeter is not enough. I've last used one at school 20 years ago, but I recall one had to take several measurements in pre-defined intervals to estimate ground resistance falloff. Plus it was connected to the measured ground and 2 ground probes, I forgot why 1 is not enough. My point is that the process is overly complicated because there is much more than a matter of simple resistance.
    – Agent_L
    Nov 14, 2016 at 11:58

2 Answers 2


No -- that bond wire could save your life!

It is very important that all non-current-carrying metal, including the cold and hot water pipes, be bonded to the electrical system ground. This prevents highly unpleasant surprises should a wire in the dishwasher come loose and contact a water pipe! Furthermore, this requirement is enshrined in 250.104(A) of the Code:

(A) Metal Water Piping. The metal water piping system shall be bonded as required in (A)(1), (A)(2), or (A)(3) of this section. The bonding jumper(s) shall be installed in accordance with 250.64(A), (B), and (E). The points of attachment of the bonding jumper(s) shall be accessible.

(1) General. Metal water piping system(s) installed in or attached to a building or structure shall be bonded to the service equipment enclosure, the grounded conductor at the service, the grounding electrode conductor where of sufficient size, or to the one or more grounding electrodes used. The bonding jumper(s) shall be sized in accordance with Table 250.66 except as permitted in 250.104(A)(2) and (A)(3).

Also, the pinhole leakage you saw in your hot water pipe is a water chemistry issue (excessive chlorine residuals can attack copper pipes), not an electrolytic corrosion issue, and has nothing to do with the bonding of the water system or any dissimilar metals -- in fact, in an iron-copper couple, it's the iron that corrodes first.

  • OK ThreePhaseEel, I reconnected the copper wire attached to a cold water pipe to the ground-neutral bar. Nov 13, 2016 at 21:33
  • 1
    I'd like to emphasise that it's the copper pipe that's grounded itself, not your network that's grounded through it.
    – SQB
    Nov 14, 2016 at 10:55
  • ThreePhaseEel, I take your point about the relative positions of iron and copper in the electrochemical series of metals, but there are some complicating factors. The tank is glass lined steel at a higher temperature than the vertical downward run of copper out-flow tubing. The vertical run of tubing was the thinnest allowed (red marked Class M). Nov 14, 2016 at 10:55
  • 1
    Or the corrosive factor could just be your water. Dec 5, 2016 at 20:41

You used "Dielectric Unions" on your tankless connections, and you have a copper plumbing system? Those Dielectric unions won't last long, just sweat the copper directly to the isolator valves, that are threaded on the tankless water connects. Even with a grounding rod, you will want your copper pipe grounded to it, put the grounding rod in the vicinity of the copper pipe, not on the other side of the house, I have had an electrical engineer tell me that two different ground points on the opposite sides of a structure create an antenna of your wire and copper plumbing system. Excessive chlorine itself won't cause leaks in a copper system, as far as I am aware, but in a chlorimanated water system, you can get excessive ammonia as a by product of nitrification, and that can corrode copper. But that condition would have to be going on for some time.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.