If everything is plastic, you shouldn't be required to bond. The risk of current in the water itself is IMHO negligible in a house.
Having a dielectric union and not using the underground pipe as an electrode is a good idea. To the extent that we should have grounding electrodes, we shouldn't be confusing them with plumbing.
I don't see how GFCI or AFCI tech changes the need (or lack of need) for grounding. You want to provide a well controlled path for low current, high voltage upsets ('static', HV leakage, etc.)
To ground or not to ground was seriously debated at the start of electrification. Soares Book on Grounding and Bonding had some history on this subject.
So this seems to be similar to the Stray Voltage that Mike holt was talking about in his Stray Voltage video where he references he was on the Stray voltage foundation.
"Stray Voltage" should be changed to "Stray Current"
"It has been shown previously that it is the current that is dangerous to the human animal and to the animal kingdom. Voltage is only the pressure that pushes the current. Voltage does not burn the body. It is the current that burns the body, sets the heart into fibrillation, halting the pumping action of the heart, resulting in death of both".
No need to post a picture of what I have installed hundreds of times. The neutral serves to support the drop cable. So what? Do you propose a separate insulated neutral and a separate messenger?
No need to post a picture of what I have installed hundreds of times. The neutral serves to support the drop cable. So what? Do you propose a separate insulated neutral and a separate messenger?
If both the secondary and primary neutral are earthed (grounded), then they will be at (approximately) the same potential. What is the advantage of having them only connected through earth, rather than solidly bonded together?
Bonding imposes primary neutral-to-earth voltage on every metal device via the EGCs. This is a well known problem for farms, e.g. animals drinking heated water. Utilities sometimes install a neutral isolator which keeps primary and secondary neutral seperate as long as the voltage difference is low.
I’ve been gone a while, and I’m surprised at the prevalence of plastic water mains these days—that’s news to me.
We moved into a 1933 brick house with original galvanized pipe running 30’ through a slab, connecting two zones. The GI pipe is visibly corroding, especially on the meter side at the copper connection. Notably, a black gas pipe runs literally right next to it (feeding furnaces and gas water heaters in the second zone), yet shows basically zero corrosion.
A neighbor says a plastic water main went in about 15 years ago with a single grounding rod. But I see no deliberate bonding anywhere—just plenty of dodgy UF feeders, Romex through damp brick, old cloth wiring, etc.
Until I can replace that galvanized pipe (which may take a while), we’re installing a new water filter and water heater (my wife hates the water taste). I plan to clamp at accessible points—gas pipe, copper pipe, grounding conductor—and isolate conditions with the main breaker on and off. I’ll use temporary jumpers to bright metal for testing. With the main off, proper bonding should theoretically route stray currents up the mast neutral to the transformer center tap. With the main on, I’d isolate and abandon any leaky circuit.
The serious corrosion at the GI-to-Cu junction strongly suggests galvanic reaction and hidden grounding paths—pipes touching appliances, conduit, older wiring, etc. But I’m still SMH how the black gas pipe right next to it remains unaffected. I haven’t clamped the gas pipe itself yet; with 10 feet in earth, it looks like a fantastic grounding electrode conductor. I’m expecting current flow there because it’s connected to gas appliances. I don’t see any deliberate bonding to the GEC, and I can’t imagine it’s completely isolated.
Given our isolated position (transformer at least 200’ downhill, uphill neighbors 500’ away—though plenty downhill), I think this stray-current issue is most likely internal, not from neighbors or utility. My plan:
Verify a solid “primary neutral” connection first.
Short-term (until GI replacement):
Install dielectric unions at Cu-to-GI joints (though finding bright GI pipe is tricky).
Deliberately bond directly to any accessible bright GI pipe sections to the GEC, or as close as I can on the copper side.
Bonding copper alone without bonding GI could worsen corrosion at the joint.
And starting right away, bond bond and bond everywhere to the GEC.
I wonder still if there’s some hidden scenario I’ve overlooked.
Could deliberate bonding now, even assuming a solid utility neutral, somehow attract more strays and worsen corrosion?
When I was a kid, the water main was the real GEC, and everything else was just for the equipotential.
Anyhoo—I might dig out my 1987 copy of Soares’ grounding book or my 1999 superbeast—but hey, I’m retired now, haha.
Concise Situation:
Galvanized (GI) water pipe embedded in slab severely corroded at copper junction, indicating galvanic corrosion and stray current flow.
Black gas pipe running parallel shows zero corrosion, suggesting electrical isolation (coatings, fittings, sealants).
Copper piping indirectly bonded via appliances, creating galvanic potential difference with GI pipe.
Corrosion at GI-Cu joint increases impedance, separating potentials.
GI pipe unintentionally conducting stray currents due to damp slab conditions.
Planned Tests (Clamp Meter - Amps, AC/DC):
• Neutral bus → GI pipe jumper: Significant AC current likely, otherwise driving severe corrosion. Briefly test for DC current to rule out electrolysis.
• Neutral bus → Copper pipe jumper: Moderate AC current possible, otherwise accelerating galvanic corrosion. Minimal/no DC expected.
• Neutral bus → Gas pipe jumper: Minimal/no AC/DC current expected; measurable current indicates unintended bonding.
• GEC to grounding rod: Minimal/no AC/DC current expected; measurable current indicates unwanted stray currents.
• Primary neutral (mast/panel): Normal measurable AC current expected, confirming proper return path; minimal/no DC current.
Planned Tests (LoZ Voltmeter - Volts, AC/DC):
• GI pipe → Copper pipe: Significant galvanic voltage expected (AC/DC), otherwise directly causing corrosion.
• GI pipe → Neutral bus: Measurable AC/DC voltage likely, indicating stray current path through GI.
• Copper pipe → Neutral bus: Low/minimal AC/DC voltage expected; measurable voltage indicates bonding/corrosion issue.
• Gas pipe → Neutral bus: Minimal/no AC/DC voltage expected; measurable voltage indicates unexpected connection.
Should voltage and LoZ tests be performed on both sides of slab?
• Yes, measure voltage and clamp currents at both slab ends (meter side and far side) to confirm extent of stray currents.
How to Prove Bonding Fix Helps GI Pipe:
• Temporarily bond GI pipe directly to neutral bus or GEC.
• Re-test clamp current and LoZ voltage (GI pipe → Neutral bus); reduction in current and voltage after bonding confirms improvement.
• If significant DC voltage/current measured, standard AC bonding will not help; identify and eliminate DC source separately.
Safety Cautions (Shock Hazard Potential):
• Neutral bus connections (energized conditions pose shock hazard).
• Temporary jumpers to GI, copper, gas pipes (potential shock hazard from stray voltage).
• Gas pipe measurements (confirm zero voltage before touching pipe).
• GEC and grounding rod connections (possible shock hazard if stray current present).
Different Types of Corrosion: stray current corrosion -causes and prevention. All Different Forms of Corrosion are explained by NACE certified Corrosion Specialist . WebCorr provides corrosion consultancy services, corrosion expert witness and corrosion short courses for in-house training...
www.corrosionclinic.com
In my opinion general AI is not good with hard science/engineering topics.
Different Types of Corrosion: stray current corrosion -causes and prevention. All Different Forms of Corrosion are explained by NACE certified Corrosion Specialist . WebCorr provides corrosion consultancy services, corrosion expert witness and corrosion short courses for in-house training...
www.corrosionclinic.com
In my opinion general AI is not good with hard science/engineering topics.
AI searches only returns information based on what they find, including inaccurate claims. They possibly can at least mention the possibility of inaccuracy if they also find related content pointing out the inaccurate content.
