I'm going to take the time to respond inline here because of the dangerous nature of spreading incorrect or incomplete electrical information. This isn't about singling you out personally, but your response includes a number of partial truths and misinterpretations that can lead to unsafe outcomes.
No sir — that is not correct.
Drive two ground rods into open earth some distance apart, connect them with a voltmeter or ammeter, and you will measure a potential difference — often even current flow — between them.
When you drive separate ground rods and tie equipment to one and the service to another, you create a second, isolated reference point for “ground.” But earth is not a zero-resistance conductor.
That means voltage differences can exist between the two grounds. If there's a fault (e.g., hot wire touches a metal case), the fault current may try to return through earth rather than a dedicated, low-impedance equipment grounding path.
This creates two very real dangers:
- Shock hazard: Touching metal equipment bonded to that isolated ground could expose you to voltage, since the breaker won’t trip — there’s no proper return path for fault current.
- Equipment damage: Sensitive electronics may reference different ground potentials. This can cause circulating current and voltage spikes that destroy components.
This is not theoretical, it is easily measurable and dangerous. The misconception that all grounds share the same potential is exactly why the NEC mandates a single bonded grounding reference at the main service panel (NEC 250.24(A)(1)).
New code require two separate grounds to the panel where main breakers are.
That is absolutely not correct.
The NEC requires a single grounding electrode system, not “two separate grounds.”
--NEC 250.50: All available grounding electrodes must be bonded together to form one unified grounding electrode system.
--NEC 250.58: These electrodes must be connected to the main service panel using a single grounding electrode conductor (GEC).
There is no requirement for “two grounds”—only that if multiple qualifying electrodes exist, they must be used and bonded together (not separate). It also requires qualifying water systems to be bonded to that system, not as a second ground, but to render them at the same potential as the bonded service ground.
Note the recurring theme here... BONDED. That has a very narrow and specific meaning.
In simple terms: Code allows (and may require ) multiple rods, but those rods
must be connected with a single continuous conductor (no splices) and that conductor must be
bonded to the main panel. Any other electrodes (e.g., water pipes, additional rods) must be bonded to that
same point, and
nowhere else.
The entire purpose is to create a single common ground reference. Multiple, independent “grounds” violate code and create dangerous differences in potential.
If your plumbing enters the house underground it is grounded and can be used as such.
No Sir.
You are required to bond the ground to the water system at the point of entrance, but many water services are plastic derived and therefore can not act as a system ground. The NEC addresses this also in 250.
NEC 250.52(A)(1): A metal underground water pipe is a valid grounding electrode only if:
- It is metallic for at least 10 feet in contact with earth.
- It is electrically continuous into the building (i.e., no plastic sections or dielectric unions).
- It is bonded to the service equipment with an appropriate grounding electrode conductor (GEC).
Moreover, the water service is not allowed to be the sole means of ground. Also in 250
NEC 250.53(D)(2):
“A metal underground water pipe shall be supplemented by an additional electrode…”
So again -- If present, metal water systems
must be bonded, and if no other bonded electrode exists,
one must be added. Key point again --
bonded, not separate.
Could argue that a gfi that’s not on a grounded circuit is more dangerous because the current can travel through your tanks water back through neutral and the gfi won’t trip because it detects nothing.
There is absolutely no new danger introduced by adding a GFCI to an ungrounded circuit. An ungrounded outlet without a GFCI has zero protection. a fault won’t trip anything unless it draws enough current to overload the breaker.
A GFCI-protected circuit, however, will trip if any current leaks outside the hot-neutral loop, regardless of grounding.
To be 100% clear:
If you put hot in one end of a tank and neutral in the other, you’ve made a resistive load, no different than a heater or lightbulb. If the resistance is low enough, current flows. If the current stays below the breaker rating, the breaker won’t trip, and neither will the GFCI. That’s true with or without a ground.
The GFCI adds protection when current escapes the intended circuit, such as through you, stray metal, or water paths. That’s when it trips. Grounding is not required for this behavior.
However, introducing an isolated ground does create a new danger. The GFCI only monitors for imbalance between hot and neutral. It does not detect current that flows from neutral to an improperly grounded chassis or remote ground rod.
That current could pass through your body or cause a neutral to overload and overheat without tripping the GFCI or the breaker.
This is precisely why the NEC requires that all equipment grounding conductors be bonded to the grounding electrode system at the service panel.
Don’t want to argue. Research it if you’re a nonbeliever
This has absolutely nothing to do with belief. My comments are based on fundamental electrical principles and the NEC.